ATMEGA16U4PRE_技术文档

Features

• High Performance, Low Power AVR^®8-Bit Microcontroller

• Advanced RISC Architecture

– 135 Powerful Instructions – Most Single Clock Cycle Execution

– 32 x 8 General Purpose Working Registers

– Fully Static Operation

– Up to 16 MIPS Throughput at 16 MHz

– On-Chip 2-cycle Multiplier

• Non-volatile Program and Data Memories

8-bit

Microcontroller

with

16/32K Bytes of

ISP Flash

and USB

– 16/32K Bytes of In-System Self-Programmable Flash (ATmega16U4/ATmega32U4)

– 1.25/2.5K Bytes Internal SRAM (ATmega16U4/ATmega32U4)

– 512Bytes/1K Bytes Internal EEPROM (ATmega16U4/ATmega32U4)

– Write/Erase Cycles: 10,000 Flash/100,000 EEPROM

– Data retention: 20 years at 85°C/ 100 years at 25°C⁽¹⁾

– Optional Boot Code Section with Independent Lock Bits

In-System Programming by On-chip Boot Program

True Read-While-Write Operation

All supplied parts are preprogramed with a default USB bootloader

– Programming Lock for Software Security

• JTAG (IEEE std. 1149.1 compliant) Interface

Controller

– Boundary-scan Capabilities According to the JTAG Standard

– Extensive On-chip Debug Support

– Programming of Flash, EEPROM, Fuses, and Lock Bits through the JTAG Interface

• USB 2.0 Full-speed/Low Speed Device Module with Interrupt on Transfer Completion

– Complies fully with Universal Serial Bus Specification Rev 2.0

– Supports data transfer rates up to 12 Mbit/s and 1.5 Mbit/s

– Endpoint 0 for Control Transfers: up to 64-bytes

– 6 Programmable Endpoints with IN or Out Directions and with Bulk, Interrupt or

Isochronous Transfers

ATmega16U4

ATmega32U4

– Configurable Endpoints size up to 256 bytes in double bank mode

– Fully independent 832 bytes USB DPRAM for endpoint memory allocation

– Suspend/Resume Interrupts

Preliminary

– CPU Reset possible on USB Bus Reset detection

– 48 MHz from PLL for Full-speed Bus Operation

– USB Bus Connection/Disconnection on Microcontroller Request

– Crystal-less operation for Low Speed mode

• Peripheral Features

– On-chip PLL for USB and High Speed Timer: 32 up to 96 MHz operation

– One 8-bit Timer/Counter with Separate Prescaler and Compare Mode

– Two 16-bit Timer/Counter with Separate Prescaler, Compare- and Capture Mode

– One 10-bit High-Speed Timer/Counter with PLL (64 MHz) and Compare Mode

– Four 8-bit PWM Channels

– Four PWM Channels with Programmable Resolution from 2 to 16 Bits

– Six PWM Channels for High Speed Operation, with Programmable Resolution from

2 to 11 Bits

– Output Compare Modulator

– 12-channels, 10-bit ADC (features Differential Channels with Programmable Gain)

– Programmable Serial USART with Hardware Flow Control

– Master/Slave SPI Serial Interface

7766E–AVR–04/10

ATmega16U4/ATmega32U4

– Byte Oriented 2-wire Serial Interface

– Programmable Watchdog Timer with Separate On-chip Oscillator

– On-chip Analog Comparator

– Interrupt and Wake-up on Pin Change

– On-chip Temperature Sensor

• Special Microcontroller Features

– Power-on Reset and Programmable Brown-out Detection

– Internal 8 MHz Calibrated Oscillator

– Internal clock prescaler & On-the-fly Clock Switching (Int RC / Ext Osc)

– External and Internal Interrupt Sources

– Six Sleep Modes: Idle, ADC Noise Reduction, Power-save, Power-down, Standby, and Extended Standby

• I/O and Packages

– All I/O combine CMOS outputs and LVTTL inputs

– 26 Programmable I/O Lines

– 44-lead TQFP Package, 10x10mm

– 44-lead QFN Package, 7x7mm

• Operating Voltages

– 2.7 - 5.5V

• Operating temperature

– Industrial (-40°C to +85°C)

• Maximum Frequency

– 8 MHz at 2.7V - Industrial range

– 16 MHz at 4.5V - Industrial range

Note:

1. See “Data Retention” on page 8 for details.

2

7766E–AVR–04/10

ATmega16U4/ATmega32U4

1. Pin Configurations

Figure 1-1. Pinout ATmega16U4/ATmega32U4

PE2 (HWB)

33

32

31

30

29

28

27

26

25

24

23

1

2

3

4

5

6

7

8

9

(INT.6/AIN0) PE6

UVcc

PC7 (ICP3/CLK0/OC4A)

PC6 (OC3A/OC4A)

PB6 (PCINT6/OC1B/OC4B/ADC13)

PB5 (PCINT5/OC1A/OC4B/ADC12)

PB4 (PCINT4/ADC11)

PD7 (T0/OC4D/ADC10)

PD6 (T1/OC4D/ADC9)

PD4 (ICP1/ADC8)

INDEX CORNER

D-

D+

UGnd

UCap

VBus

ATmega32U4

ATmega16U4

44-pin QFN/TQFP

(SS/PCINT0) PB0

(PCINT1/SCLK) PB1

AVCC

(PDI/PCINT2/MOSI) PB2 10

(PDO/PCINT3/MISO) PB3 11

GND

2. Overview

The ATmega16U4/ATmega32U4 is a low-power CMOS 8-bit microcontroller based on the AVR

enhanced RISC architecture. By executing powerful instructions in a single clock cycle, the

ATmega16U4/ATmega32U4 achieves throughputs approaching 1 MIPS per MHz allowing the

system designer to optimize power consumption versus processing speed.

3

7766E–AVR–04/10

ATmega16U4/ATmega32U4

2.1

Block Diagram

Figure 2-1. Block Diagram

PF7 - PF4

PF0

PC7

PC6

PF1

VCC

GND

PORTF DRIVERS

PORTC DRIVERS

DATA REGISTER

DATA DIR.

REG. PORTF

DATA REGISTER

DATA DIR.

REG. PORTC

PORTF

PORTC

8-BIT DA TA BUS

POR - BOD

RESET

INTERNAL

CALIB. OSC

OSCILLATOR

WATCHDOG

TIMER

PROGRAM

COUNTER

STACK

POINTER

JTAG TAP

TIMING AND

CONTROL

PROGRAM

FLASH

MCU CONTROL

REGISTER

SRAM

ON-CHIP DEBUG

TIMERS/

COUNTERS

BOUNDARY-

SCAN

INSTRUCTION

REGISTER

GENERAL

PURPOSE

REGISTERS

INTERRUPT

UNIT

UVcc

UCap

X

Y

Z

PROGRAMMING

LOGIC

INSTRUCTION

DECODER

ON-CHIP

USB PAD 3V

REGULATOR

EEPROM

TEMPERATURE

SENSOR

CONTROL

LINES

ALU

1uF

PLL

AVCC

HIGH SPEED

TIMER/PWM

ADC

AGND

AREF

STATUS

REGISTER

VBUS

DP

USB 2.0

DM

ANALOG

TWO-WIRE SERIAL

INTERFACE

USART1

SPI

COMPARATOR

DATA REGISTER

DATA DIR.

REG. PORTE

DATA REGISTER

DATA DIR.

REG. PORTB

DATA REGISTER

DATA DIR.

REG. PORTD

PORTE

PORTB

PORTD

PORTB DRIVERS

PORTD DRIVERS

PORTE DRIVERS

PE2

PE6

PB7 - PB0

PD7 - PD0

The AVR core combines a rich instruction set with 32 general purpose working registers. All the

32 registers are directly connected to the Arithmetic Logic Unit (ALU), allowing two independent

registers to be accessed in one single instruction executed in one clock cycle. The resulting

architecture is more code efficient while achieving throughputs up to ten times faster than con-

ventional CISC microcontrollers.

The ATmega16U4/ATmega32U4 provides the following features: 16/32K bytes of In-System

Programmable Flash with Read-While-Write capabilities, 512Bytes/1K bytes EEPROM,

1.25/2.5K bytes SRAM, 26 general purpose I/O lines (CMOS outputs and LVTTL inputs), 32

general purpose working registers, four flexible Timer/Counters with compare modes and PWM,

one more high-speed Timer/Counter with compare modes and PLL adjustable source, one

USART (including CTS/RTS flow control signals), a byte oriented 2-wire Serial Interface, a 12-

4

7766E–AVR–04/10

ATmega16U4/ATmega32U4

channels 10-bit ADC with optional differential input stage with programmable gain, an on-chip

calibrated temperature sensor, a programmable Watchdog Timer with Internal Oscillator, an SPI

serial port, IEEE std. 1149.1 compliant JTAG test interface, also used for accessing the On-chip

Debug system and programming and six software selectable power saving modes. The Idle

mode stops the CPU while allowing the SRAM, Timer/Counters, SPI port, and interrupt system

to continue functioning. The Power-down mode saves the register contents but freezes the

Oscillator, disabling all other chip functions until the next interrupt or Hardware Reset. The ADC

Noise Reduction mode stops the CPU and all I/O modules except ADC, to minimize switching

noise during ADC conversions. In Standby mode, the Crystal/Resonator Oscillator is running

while the rest of the device is sleeping. This allows very fast start-up combined with low power

consumption.

The device is manufactured using ATMEL’s high-density nonvolatile memory technology. The

On-chip ISP Flash allows the program memory to be reprogrammed in-system through an SPI

serial interface, by a conventional nonvolatile memory programmer, or by an On-chip Boot pro-

gram running on the AVR core. The boot program can use any interface to download the

application program in the application Flash memory. Software in the Boot Flash section will

continue to run while the Application Flash section is updated, providing true Read-While-Write

operation. By combining an 8-bit RISC CPU with In-System Self-Programmable Flash on a

monolithic chip, the ATMEL ATmega16U4/ATmega32U4 is a powerful microcontroller that pro-

vides a highly flexible and cost effective solution to many embedded control applications.

The ATmega16U4/ATmega32U4 AVR is supported with a full suite of program and system

development tools including: C compilers, macro assemblers, program debugger/simulators, in-

circuit emulators, and evaluation kits.

2.2

Pin Descriptions

2.2.1

VCC

Digital supply voltage.

2.2.2

2.2.3

GND

Ground.

Port B (PB7..PB0)

Port B is an 8-bit bi-directional I/O port with internal pull-up resistors (selected for each bit). The

Port B output buffers have symmetrical drive characteristics with both high sink and source

capability. As inputs, Port B pins that are externally pulled low will source current if the pull-up

resistors are activated. The Port B pins are tri-stated when a reset condition becomes active,

even if the clock is not running.

Port B has better driving capabilities than the other ports.

Port B also serves the functions of various special features of the ATmega16U4/ATmega32U4

as listed on page 72.

2.2.4

Port C (PC7,PC6)

Port C is an 8-bit bi-directional I/O port with internal pull-up resistors (selected for each bit). The

Port C output buffers have symmetrical drive characteristics with both high sink and source

capability. As inputs, Port C pins that are externally pulled low will source current if the pull-up

resistors are activated. The Port C pins are tri-stated when a reset condition becomes active,

even if the clock is not running.

5

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Only bits 6 and 7 are present on the product pinout.

Port C also serves the functions of special features of the ATmega16U4/ATmega32U4 as listed

on page 75.

2.2.5

Port D (PD7..PD0)

Port D is an 8-bit bi-directional I/O port with internal pull-up resistors (selected for each bit). The

Port D output buffers have symmetrical drive characteristics with both high sink and source

capability. As inputs, Port D pins that are externally pulled low will source current if the pull-up

resistors are activated. The Port D pins are tri-stated when a reset condition becomes active,

even if the clock is not running.

Port D also serves the functions of various special features of the ATmega16U4/ATmega32U4

as listed on page 77.

2.2.6

Port E (PE6,PE2)

Port E is an 8-bit bi-directional I/O port with internal pull-up resistors (selected for each bit). The

Port E output buffers have symmetrical drive characteristics with both high sink and source

capability. As inputs, Port E pins that are externally pulled low will source current if the pull-up

resistors are activated. The Port E pins are tri-stated when a reset condition becomes active,

even if the clock is not running.

Only bits 2 and 6 are present on the product pinout.

Port E also serves the functions of various special features of the ATmega16U4/ATmega32U4

as listed on page 80.

2.2.7

Port F (PF7..PF4, PF1,PF0)

Port F serves as analog inputs to the A/D Converter.

Port F also serves as an 8-bit bi-directional I/O port, if the A/D Converter channels are not used.

Port pins can provide internal pull-up resistors (selected for each bit). The Port F output buffers

have symmetrical drive characteristics with both high sink and source capability. As inputs, Port

F pins that are externally pulled low will source current if the pull-up resistors are activated. The

Port F pins are tri-stated when a reset condition becomes active, even if the clock is not running.

Bits 2 and 3 are not present on the product pinout.

Port F also serves the functions of the JTAG interface. If the JTAG interface is enabled, the pull-

up resistors on pins PF7(TDI), PF5(TMS), and PF4(TCK) will be activated even if a reset occurs.

2.2.8

2.2.9

2.2.10

D-

USB Full speed / Low Speed Negative Data Upstream Port. Should be connected to the USB D-

connector pin with a serial 22 Ohms resistor.

D+

USB Full speed / Low Speed Positive Data Upstream Port. Should be connected to the USB D+

connector pin with a serial 22 Ohms resistor.

UGND

USB Pads Ground.

6

7766E–AVR–04/10

ATmega16U4/ATmega32U4

2.2.11

2.2.12

UVCC

UCAP

USB Pads Internal Regulator Input supply voltage.

USB Pads Internal Regulator Output supply voltage. Should be connected to an external capac-

itor (1µF).

2.2.13

2.2.14

VBUS

USB VBUS monitor input.

RESET

Reset input. A low level on this pin for longer than the minimum pulse length will generate a

reset, even if the clock is not running. The minimum pulse length is given in Table 8-1 on page

50. Shorter pulses are not guaranteed to generate a reset.

2.2.15

2.2.16

2.2.17

XTAL1

XTAL2

AVCC

Input to the inverting Oscillator amplifier and input to the internal clock operating circuit.

Output from the inverting Oscillator amplifier.

AVCC is the supply voltage pin (input) for all the A/D Converter channels. If the ADC is not used,

it should be externally connected to V_CC. If the ADC is used, it should be connected to V_CC

through a low-pass filter.

2.2.18

AREF

This is the analog reference pin (input) for the A/D Converter.

7

7766E–AVR–04/10

ATmega16U4/ATmega32U4

3. About

3.1

Disclaimer

Typical values contained in this datasheet are based on simulations and characterization of

other AVR microcontrollers manufactured on the same process technology. Min and Max values

will be available after the device is characterized.

3.2

3.3

Resources

A comprehensive set of development tools, application notes and datasheets are available for

download on http://www.atmel.com/avr.

Code Examples

This documentation contains simple code examples that briefly show how to use various parts of

the device. Be aware that not all C compiler vendors include bit definitions in the header files

and interrupt handling in C is compiler dependent. Please confirm with the C compiler documen-

tation for more details.

These code examples assume that the part specific header file is included before compilation.

For I/O registers located in extended I/O map, "IN", "OUT", "SBIS", "SBIC", "CBI", and "SBI"

instructions must be replaced with instructions that allow access to extended I/O. Typically

"LDS" and "STS" combined with "SBRS", "SBRC", "SBR", and "CBR".

3.4

Data Retention

Reliability Qualification results show that the projected data retention failure rate is much less

than 1 PPM over 20 years at 85°C or 100 years at 25°C.

8

7766E–AVR–04/10

ATmega16U4/ATmega32U4

4. AVR CPU Core

4.1

Introduction

This section discusses the AVR core architecture in general. The main function of the CPU core

is to ensure correct program execution. The CPU must therefore be able to access memories,

perform calculations, control peripherals, and handle interrupts.

4.2

Architectural Overview

Figure 4-1. Block Diagram of the AVR Architecture

Data Bus 8-bit

Program

Counter

Status

and Control

Flash

Program

Memory

Interrupt

Unit

32 x 8

General

Purpose

Registrers

Instruction

Register

SPI

Unit

Instruction

Decoder

Watchdog

Timer

ALU

Analog

Comparator

Control Lines

I/O Module1

I/O Module 2

I/O Module n

Data

SRAM

EEPROM

I/O Lines

In order to maximize performance and parallelism, the AVR uses a Harvard architecture – with

separate memories and buses for program and data. Instructions in the program memory are

executed with a single level pipelining. While one instruction is being executed, the next instruc-

tion is pre-fetched from the program memory. This concept enables instructions to be executed

in every clock cycle. The program memory is In-System Reprogrammable Flash memory.

9

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The fast-access Register File contains 32 x 8-bit general purpose working registers with a single

clock cycle access time. This allows single-cycle Arithmetic Logic Unit (ALU) operation. In a typ-

ical ALU operation, two operands are output from the Register File, the operation is executed,

and the result is stored back in the Register File – in one clock cycle.

Six of the 32 registers can be used as three 16-bit indirect address register pointers for Data

Space addressing – enabling efficient address calculations. One of the these address pointers

can also be used as an address pointer for look up tables in Flash program memory. These

added function registers are the 16-bit X-, Y-, and Z-register, described later in this section.

The ALU supports arithmetic and logic operations between registers or between a constant and

a register. Single register operations can also be executed in the ALU. After an arithmetic opera-

tion, the Status Register is updated to reflect information about the result of the operation.

Program flow is provided by conditional and unconditional jump and call instructions, able to

directly address the whole address space. Most AVR instructions have a single 16-bit word for-

mat. Every program memory address contains a 16- or 32-bit instruction.

Program Flash memory space is divided in two sections, the Boot Program section and the

Application Program section. Both sections have dedicated Lock bits for write and read/write

protection. The SPM instruction that writes into the Application Flash memory section must

reside in the Boot Program section.

During interrupts and subroutine calls, the return address Program Counter (PC) is stored on the

Stack. The Stack is effectively allocated in the general data SRAM, and consequently the Stack

size is only limited by the total SRAM size and the usage of the SRAM. All user programs must

initialize the SP in the Reset routine (before subroutines or interrupts are executed). The Stack

Pointer (SP) is read/write accessible in the I/O space. The data SRAM can easily be accessed

through the five different addressing modes supported in the AVR architecture.

The memory spaces in the AVR architecture are all linear and regular memory maps.

A flexible interrupt module has its control registers in the I/O space with an additional Global

Interrupt Enable bit in the Status Register. All interrupts have a separate Interrupt Vector in the

Interrupt Vector table. The interrupts have priority in accordance with their Interrupt Vector posi-

tion. The lower the Interrupt Vector address, the higher the priority.

The I/O memory space contains 64 addresses for CPU peripheral functions as Control Regis-

ters, SPI, and other I/O functions. The I/O Memory can be accessed directly, or as the Data

Space locations following those of the Register File, 0x20 - 0x5F. In addition, the

ATmega16U4/ATmega32U4 has Extended I/O space from 0x60 - 0x0FF in SRAM where only

the ST/STS/STD and LD/LDS/LDD instructions can be used.

4.3

ALU – Arithmetic Logic Unit

The high-performance AVR ALU operates in direct connection with all the 32 general purpose

working registers. Within a single clock cycle, arithmetic operations between general purpose

registers or between a register and an immediate are executed. The ALU operations are divided

into three main categories – arithmetic, logical, and bit-functions. Some implementations of the

architecture also provide a powerful multiplier supporting both signed/unsigned multiplication

and fractional format. See the “Instruction Set” section for a detailed description.

10

7766E–AVR–04/10

ATmega16U4/ATmega32U4

4.4

Status Register

The Status Register contains information about the result of the most recently executed arithme-

tic instruction. This information can be used for altering program flow in order to perform

conditional operations. Note that the Status Register is updated after all ALU operations, as

specified in the Instruction Set Reference. This will in many cases remove the need for using the

dedicated compare instructions, resulting in faster and more compact code.

The Status Register is not automatically stored when entering an interrupt routine and restored

when returning from an interrupt. This must be handled by software.

The AVR Status Register – SREG – is defined as:

Bit

7

6

5

4

3

2

1

0

I

T

H

S

V

N

Z

C

SREG

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

• Bit 7 – I: Global Interrupt Enable

The Global Interrupt Enable bit must be set for the interrupts to be enabled. The individual inter-

rupt enable control is then performed in separate control registers. If the Global Interrupt Enable

Register is cleared, none of the interrupts are enabled independent of the individual interrupt

enable settings. The I-bit is cleared by hardware after an interrupt has occurred, and is set by

the RETI instruction to enable subsequent interrupts. The I-bit can also be set and cleared by

the application with the SEI and CLI instructions, as described in the instruction set reference.

• Bit 6 – T: Bit Copy Storage

The Bit Copy instructions BLD (Bit LoaD) and BST (Bit STore) use the T-bit as source or desti-

nation for the operated bit. A bit from a register in the Register File can be copied into T by the

BST instruction, and a bit in T can be copied into a bit in a register in the Register File by the

BLD instruction.

• Bit 5 – H: Half Carry Flag

The Half Carry Flag H indicates a Half Carry in some arithmetic operations. Half Carry Is useful

in BCD arithmetic. See the “Instruction Set Description” for detailed information.

• Bit 4 – S: Sign Bit, S = N ⊕ V

The S-bit is always an exclusive or between the Negative Flag N and the Two’s Complement

Overflow Flag V. See the “Instruction Set Description” for detailed information.

• Bit 3 – V: Two’s Complement Overflow Flag

The Two’s Complement Overflow Flag V supports two’s arithmetic complements. See the

“Instruction Set Description” for detailed information.

• Bit 2 – N: Negative Flag

The Negative Flag N indicates a negative result in an arithmetic or logic operation. See the

“Instruction Set Description” for detailed information.

• Bit 1 – Z: Zero Flag

The Zero Flag Z indicates a zero result in an arithmetic or logic operation. See the “Instruction

Set Description” for detailed information.

11

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 0 – C: Carry Flag

The Carry Flag C indicates a carry in an arithmetic or logic operation. See the “Instruction Set

Description” for detailed information.

4.5

General Purpose Register File

The Register File is optimized for the AVR Enhanced RISC instruction set. In order to achieve

the required performance and flexibility, the following input/output schemes are supported by the

Register File:

• One 8-bit output operand and one 8-bit result input

• Two 8-bit output operands and one 8-bit result input

• Two 8-bit output operands and one 16-bit result input

• One 16-bit output operand and one 16-bit result input

Figure 4-2 shows the structure of the 32 general purpose working registers in the CPU.

Figure 4-2. AVR CPU General Purpose Working Registers

7

0

Addr.

R0

0x00

0x01

0x02

R1

R2

…

R13

R14

R15

R16

R17

…

0x0D

0x0E

0x0F

0x10

0x11

General

Purpose

Working

Registers

R26

R27

R28

R29

R30

R31

0x1A

0x1B

0x1C

0x1D

0x1E

0x1F

X-register Low Byte

X-register High Byte

Y-register Low Byte

Y-register High Byte

Z-register Low Byte

Z-register High Byte

Most of the instructions operating on the Register File have direct access to all registers, and

most of them are single cycle instructions.

As shown in Figure 4-2, each register is also assigned a data memory address, mapping them

directly into the first 32 locations of the user Data Space. Although not being physically imple-

mented as SRAM locations, this memory organization provides great flexibility in access of the

registers, as the X-, Y- and Z-pointer registers can be set to index any register in the file.

4.5.1

The X-register, Y-register, and Z-register

The registers R26..R31 have some added functions to their general purpose usage. These reg-

isters are 16-bit address pointers for indirect addressing of the data space. The three indirect

address registers X, Y, and Z are defined as described in Figure 4-3.

12

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 4-3. The X-, Y-, and Z-registers

15

XH

XL

0

X-register

7

0

7

R27 (0x1B)

R26 (0x1A)

15

YH

YL

ZL

0

Y-register

Z-register

7

R29 (0x1D)

R28 (0x1C)

15

ZH

0

7

0

R31 (0x1F)

R30 (0x1E)

In the different addressing modes these address registers have functions as fixed displacement,

automatic increment, and automatic decrement (see the instruction set reference for details).

4.6

Stack Pointer

The Stack is mainly used for storing temporary data, for storing local variables and for storing

return addresses after interrupts and subroutine calls. The Stack Pointer Register always points

to the top of the Stack. Note that the Stack is implemented as growing from higher memory loca-

tions to lower memory locations. This implies that a Stack PUSH command decreases the Stack

Pointer.

The Stack Pointer points to the data SRAM Stack area where the Subroutine and Interrupt

Stacks are located. This Stack space in the data SRAM must be defined by the program before

any subroutine calls are executed or interrupts are enabled. The Stack Pointer must be set to

point above 0x0100. The initial value of the stack pointer is the last address of the internal

SRAM. The Stack Pointer is decremented by one when data is pushed onto the Stack with the

PUSH instruction, and it is decremented by three when the return address is pushed onto the

Stack with subroutine call or interrupt. The Stack Pointer is incremented by one when data is

popped from the Stack with the POP instruction, and it is incremented by three when data is

popped from the Stack with return from subroutine RET or return from interrupt RETI.

The AVR Stack Pointer is implemented as two 8-bit registers in the I/O space. The number of

bits actually used is implementation dependent. Note that the data space in some implementa-

tions of the AVR architecture is so small that only SPL is needed. In this case, the SPH Register

will not be present.

Bit

15

14

13

12

11

10

9

8

SP15

SP7

7

SP14

SP6

6

SP13

SP5

5

SP12

SP4

4

SP11

SP3

3

SP10

SP2

2

SP9

SP1

1

SP8

SP0

0

SPH

SPL

Read/Write

Initial Value

R/W

0

R/W

0

R/W

1

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

1

13

7766E–AVR–04/10

ATmega16U4/ATmega32U4

4.6.1

Extended Z-pointer Register for ELPM/SPM - RAMPZ

Bit

7

6

5

4

3

2

1

0

RAMPZ7

RAMPZ6

RAMPZ5

RAMPZ4

RAMPZ3

RAMPZ2

RAMPZ1

RAMPZ0

RAMPZ

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

For ELPM/SPM instructions, the Z-pointer is a concatenation of RAMPZ, ZH, and ZL, as shown

in Figure 4-4. Note that LPM is not affected by the RAMPZ setting.

Figure 4-4. The Z-pointer used by ELPM and SPM

Bit (Individually)

Bit (Z-pointer)

7

0

7

0

8

7

0

RAMPZ

ZH

15

ZL

7

23

16

The actual number of bits is implementation dependent. Unused bits in an implementation will

always read as zero. For compatibility with future devices, be sure to write these bits to zero.

4.7

Instruction Execution Timing

This section describes the general access timing concepts for instruction execution. The AVR

CPU is driven by the CPU clock clk_CPU, directly generated from the selected clock source for the

chip. No internal clock division is used.

Figure 4-5 shows the parallel instruction fetches and instruction executions enabled by the Har-

vard architecture and the fast-access Register File concept. This is the basic pipelining concept

to obtain up to 1 MIPS per MHz with the corresponding unique results for functions per cost,

functions per clocks, and functions per power-unit.

Figure 4-5. The Parallel Instruction Fetches and Instruction Executions

T1

T2

T3

T4

clk_CPU

1st Instruction Fetch

1st Instruction Execute

2nd Instruction Fetch

2nd Instruction Execute

3rd Instruction Fetch

3rd Instruction Execute

4th Instruction Fetch

Figure 4-6 shows the internal timing concept for the Register File. In a single clock cycle an ALU

operation using two register operands is executed, and the result is stored back to the destina-

tion register.

14

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 4-6. Single Cycle ALU Operation

T1

T2

T3

T4

clk_CPU

Total Execution Time

Register Operands Fetch

ALU Operation Execute

Result Write Back

4.8

Reset and Interrupt Handling

The AVR provides several different interrupt sources. These interrupts and the separate Reset

Vector each have a separate program vector in the program memory space. All interrupts are

assigned individual enable bits which must be written logic one together with the Global Interrupt

Enable bit in the Status Register in order to enable the interrupt. Depending on the Program

Counter value, interrupts may be automatically disabled when Boot Lock bits BLB02 or BLB12

are programmed. This feature improves software security. See the section “Memory Program-

ming” on page 346 for details.

The lowest addresses in the program memory space are by default defined as the Reset and

Interrupt Vectors. The complete list of vectors is shown in “Interrupts” on page 61. The list also

determines the priority levels of the different interrupts. The lower the address the higher is the

priority level. RESET has the highest priority, and next is INT0 – the External Interrupt Request

0. The Interrupt Vectors can be moved to the start of the Boot Flash section by setting the IVSEL

bit in the MCU Control Register (MCUCR). Refer to “Interrupts” on page 61 for more information.

The Reset Vector can also be moved to the start of the Boot Flash section by programming the

BOOTRST Fuse, see “Memory Programming” on page 346.

When an interrupt occurs, the Global Interrupt Enable I-bit is cleared and all interrupts are dis-

abled. The user software can write logic one to the I-bit to enable nested interrupts. All enabled

interrupts can then interrupt the current interrupt routine. The I-bit is automatically set when a

Return from Interrupt instruction – RETI – is executed.

There are basically two types of interrupts. The first type is triggered by an event that sets the

Interrupt Flag. For these interrupts, the Program Counter is vectored to the actual Interrupt Vec-

tor in order to execute the interrupt handling routine, and hardware clears the corresponding

Interrupt Flag. Interrupt Flags can also be cleared by writing a logic one to the flag bit position(s)

to be cleared. If an interrupt condition occurs while the corresponding interrupt enable bit is

cleared, the Interrupt Flag will be set and remembered until the interrupt is enabled, or the flag is

cleared by software. Similarly, if one or more interrupt conditions occur while the Global Interrupt

Enable bit is cleared, the corresponding Interrupt Flag(s) will be set and remembered until the

Global Interrupt Enable bit is set, and will then be executed by order of priority.

The second type of interrupts will trigger as long as the interrupt condition is present. These

interrupts do not necessarily have Interrupt Flags. If the interrupt condition disappears before the

interrupt is enabled, the interrupt will not be triggered.

When the AVR exits from an interrupt, it will always return to the main program and execute one

more instruction before any pending interrupt is served.

15

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Note that the Status Register is not automatically stored when entering an interrupt routine, nor

restored when returning from an interrupt routine. This must be handled by software.

When using the CLI instruction to disable interrupts, the interrupts will be immediately disabled.

No interrupt will be executed after the CLI instruction, even if it occurs simultaneously with the

CLI instruction. The following example shows how this can be used to avoid interrupts during the

timed EEPROM write sequence.

Assembly Code Example

in r16, SREG

; store SREG value

cli ; disable interrupts during timed sequence

sbiEECR, EEMPE ; start EEPROM write

sbiEECR, EEPE

outSREG, r16

; restore SREG value (I-bit)

C Code Example

char cSREG;

cSREG = SREG;/* store SREG value */

/* disable interrupts during timed sequence */

__disable_interrupt();

EECR |= (1<<EEMPE); /* start EEPROM write */

EECR |= (1<<EEPE);

SREG = cSREG; /* restore SREG value (I-bit) */

When using the SEI instruction to enable interrupts, the instruction following SEI will be exe-

cuted before any pending interrupts, as shown in this example.

16

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Assembly Code Example

sei ; set Global Interrupt Enable

sleep; enter sleep, waiting for interrupt

; note: will enter sleep before any pending

; interrupt(s)

C Code Example

__enable_interrupt(); /* set Global Interrupt Enable */

__sleep(); /* enter sleep, waiting for interrupt */

/* note: will enter sleep before any pending interrupt(s) */

4.8.1

Interrupt Response Time

The interrupt execution response for all the enabled AVR interrupts is five clock cycles minimum.

After five clock cycles the program vector address for the actual interrupt handling routine is exe-

cuted. During these five clock cycle period, the Program Counter is pushed onto the Stack. The

vector is normally a jump to the interrupt routine, and this jump takes three clock cycles. If an

interrupt occurs during execution of a multi-cycle instruction, this instruction is completed before

the interrupt is served. If an interrupt occurs when the MCU is in sleep mode, the interrupt exe-

cution response time is increased by five clock cycles. This increase comes in addition to the

start-up time from the selected sleep mode.

A return from an interrupt handling routine takes five clock cycles. During these five clock cycles,

the Program Counter (three bytes) is popped back from the Stack, the Stack Pointer is incre-

mented by three, and the I-bit in SREG is set.

17

7766E–AVR–04/10

ATmega16U4/ATmega32U4

5. AVR ATmega16U4/ATmega32U4 Memories

This section describes the different memories in the ATmega16U4/ATmega32U4. The AVR

architecture has two main memory spaces, the Data Memory and the Program Memory space.

In addition, the ATmega16U4/ATmega32U4 features an EEPROM Memory for data storage. All

three memory spaces are linear and regular.

Table 5-1.

Memory

Memory Mapping.

Mnemonic

ATmega32U4

ATmega16U4

Size

Flash size

32K bytes

16K bytes

-

Start Address

Flash

0x0000

0x7FFF⁽¹⁾

0x3FFF⁽²⁾

0x3FFF⁽¹⁾

0x1FFF⁽²⁾

End Address

Flash end

Size

-

32 bytes

0x0000

0x001F

64 bytes

0x0020

0x005F

160 bytes

0x0060

0x00FF

2,5K bytes

0x100

32 bytes

0x0000

32

Registers

Start Address

End Address

Size

-

0x001F

64 bytes

0x0020

-

I/O

Start Address

End Address

Size

-

Registers

-

0x005F

160 bytes

0x0060

-

Ext I/O

Start Address

End Address

Size

-

Registers

-

0x00FF

1.25K bytes

0x100

ISRAM size

ISRAM start

ISRAM end

Internal

SRAM

Start Address

End Address

0x0AFF

0x05FF

External

Memory

Not Present.

Size

E2 size

E2 end

1K bytes

0x03FF

512 bytes

0x01FF

EEPROM

End Address

Notes: 1. Byte address.

2. Word (16-bit) address.

5.1

In-System Reprogrammable Flash Program Memory

The ATmega16U4/ATmega32U4 contains 16/32K bytes On-chip In-System Reprogrammable

Flash memory for program storage. Since all AVR instructions are 16 or 32 bits wide, the Flash

is organized as 16K x 16. For software security, the Flash Program memory space is divided into

two sections, Boot Program section and Application Program section.

The Flash memory has an endurance of at least 100,000 write/erase cycles. The

ATmega16U4/ATmega32U4 Program Counter (PC) is 16 bits wide, thus addressing the 32K

program memory locations. The operation of Boot Program section and associated Boot Lock

bits for software protection are described in detail in “Memory Programming” on page 346.

18

7766E–AVR–04/10

ATmega16U4/ATmega32U4

“Memory Programming” on page 346 contains a detailed description on Flash data serial down-

loading using the SPI pins or the JTAG interface.

Constant tables can be allocated within the entire program memory address space (see the LPM

– Load Program Memory instruction description and ELPM - Extended Load Program Memory

instruction description).

Timing diagrams for instruction fetch and execution are presented in “Instruction Execution Tim-

ing” on page 14.

Figure 5-1. Program Memory Map

Program Memory

0x00000

Application Flash Section

Boot Flash Section

0x7FFF (32KBytes)

5.2

SRAM Data Memory

Figure 5-2 shows how the ATmega16U4/ATmega32U4 SRAM Memory is organized.

The ATmega16U4/ATmega32U4 is a complex microcontroller with more peripheral units than

can be supported within the 64 location reserved in the Opcode for the IN and OUT instructions.

For the Extended I/O space from $060 - $0FF in SRAM, only the ST/STS/STD and LD/LDS/LDD

instructions can be used.

The first 2,816 Data Memory locations address both the Register File, the I/O Memory,

Extended I/O Memory, and the internal data SRAM. The first 32 locations address the Register

file, the next 64 location the standard I/O Memory, then 160 locations of Extended I/O memory

and the next 2,560 locations address the internal data SRAM.

The five different addressing modes for the data memory cover: Direct, Indirect with Displace-

ment, Indirect, Indirect with Pre-decrement, and Indirect with Post-increment. In the Register file,

registers R26 to R31 feature the indirect addressing pointer registers.

The direct addressing reaches the entire data space.

19

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The Indirect with Displacement mode reaches 63 address locations from the base address given

by the Y- or Z-register.

When using register indirect addressing modes with automatic pre-decrement and post-incre-

ment, the address registers X, Y, and Z are decremented or incremented.

The 32 general purpose working registers, 64 I/O registers, and the 1.25/2.5Kbytes of internal

data SRAM in the ATmega16U4/ATmega32U4 are all accessible through all these addressing

modes. The Register File is described in “General Purpose Register File” on page 12.

Figure 5-2. Data Memory Map

Data Memory

$0000

$0020

$0060

-

$001

$005

$00FF

F

32

64 I/O

160

R

eg

xt I/O

i

st

e

r

s

e

R

i

st

r

s

eg.

E

ISRAM start : $0100

In

t

e

r

na

l

S

R

AM

ISRAM end : $05FF / $0AFF

$

FFFF

5.2.1

Data Memory Access Times

This section describes the general access timing concepts for internal memory access. The

internal data SRAM access is performed in two clk_CPUcycles as described in Figure 5-3.

Figure 5-3. On-chip Data SRAM Access Cycles

T1

T2

T3

clk_CPU

Address valid

Compute Address

Address

Data

WR

Data

RD

Memory Access Instruction

Next Instruction

20

7766E–AVR–04/10

ATmega16U4/ATmega32U4

5.3

EEPROM Data Memory

The ATmega16U4/ATmega32U4 contains 512Bytes/1K bytes of data EEPROM memory. It is

organized as a separate data space, in which single bytes can be read and written. The

EEPROM has an endurance of at least 100,000 write/erase cycles. The access between the

EEPROM and the CPU is described in the following, specifying the EEPROM Address Regis-

ters, the EEPROM Data Register, and the EEPROM Control Register.

For a detailed description of SPI, JTAG and Parallel data downloading to the EEPROM, see

page 360, page 365, and page 349 respectively.

5.3.1

EEPROM Read/Write Access

The EEPROM Access Registers are accessible in the I/O space.

The write access time for the EEPROM is given in Table 5-3. A self-timing function, however,

lets the user software detect when the next byte can be written. If the user code contains instruc-

tions that write the EEPROM, some precautions must be taken. In heavily filtered power

supplies, V_CCis likely to rise or fall slowly on power-up/down. This causes the device for some

period of time to run at a voltage lower than specified as minimum for the clock frequency used.

See “Preventing EEPROM Corruption” on page 25. for details on how to avoid problems in these

situations.

In order to prevent unintentional EEPROM writes, a specific write procedure must be followed.

Refer to the description of the EEPROM Control Register for details on this.

When the EEPROM is read, the CPU is halted for four clock cycles before the next instruction is

executed. When the EEPROM is written, the CPU is halted for two clock cycles before the next

instruction is executed.

5.3.2

The EEPROM Address Register – EEARH and EEARL

Bit

15

14

13

12

11

10

9

8

–

EEAR11

EEAR10

EEAR9

EEAR1

1

EEAR8

EEAR0

0

EEARH

EEAR7

EEAR6

EEAR5

EEAR4

EEAR3

EEAR2

EEARL

7

6

5

4

3

2

Read/Write

Initial Value

R

R/W

X

R/W

X

R/W

X

R/W

X

R/W

0

R/W

0

R/W

0

R/W

0

X

• Bits 15..12 – Res: Reserved Bits

These bits are reserved bits in the ATmega16U4/ATmega32U4 and will always read as zero.

• Bits 11..0 – EEAR8..0: EEPROM Address

The EEPROM Address Registers – EEARH and EEARL specify the EEPROM address in the

512Bytes/1K bytes EEPROM space. The EEPROM data bytes are addressed linearly between

0 and E2_END. The initial value of EEAR is undefined. A proper value must be written before

the EEPROM may be accessed.

5.3.3

The EEPROM Data Register – EEDR

Bit

7

6

5

4

3

2

1

0

MSB

R/W

0

LSB

R/W

0

EEDR

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

21

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bits 7..0 – EEDR7.0: EEPROM Data

For the EEPROM write operation, the EEDR Register contains the data to be written to the

EEPROM in the address given by the EEAR Register. For the EEPROM read operation, the

EEDR contains the data read out from the EEPROM at the address given by EEAR.

5.3.4

The EEPROM Control Register – EECR

Bit

7

6

5

4

3

2

1

0

–

EEPM1

R/W

X

EEPM0

R/W

X

EERIE

R/W

0

EEMPE

R/W

0

EEPE

R/W

X

EERE

R/W

0

EECR

Read/Write

Initial Value

R

0

R

0

• Bits 7..6 – Res: Reserved Bits

These bits are reserved bits in the ATmega16U4/ATmega32U4 and will always read as zero.

• Bits 5, 4 – EEPM1 and EEPM0: EEPROM Programming Mode Bits

The EEPROM Programming mode bit setting defines which programming action that will be trig-

gered when writing EEPE. It is possible to program data in one atomic operation (erase the old

value and program the new value) or to split the Erase and Write operations in two different

operations. The Programming times for the different modes are shown in Table 5-2. While EEPE

is set, any write to EEPMn will be ignored. During reset, the EEPMn bits will be reset to 0b00

unless the EEPROM is busy programming.

Table 5-2.

EEPROM Mode Bits

Programming

EEPM1 EEPM0

Time

3.4 ms

1.8 ms

–

Operation

0

1

0

1

0

1

Erase and Write in one operation (Atomic Operation)

Erase Only

Write Only

Reserved for future use

• Bit 3 – EERIE: EEPROM Ready Interrupt Enable

Writing EERIE to one enables the EEPROM Ready Interrupt if the I bit in SREG is set. Writing

EERIE to zero disables the interrupt. The EEPROM Ready interrupt generates a constant inter-

rupt when EEPE is cleared.

• Bit 2 – EEMPE: EEPROM Master Programming Enable

The EEMPE bit determines whether setting EEPE to one causes the EEPROM to be written.

When EEMPE is set, setting EEPE within four clock cycles will write data to the EEPROM at the

selected address If EEMPE is zero, setting EEPE will have no effect. When EEMPE has been

written to one by software, hardware clears the bit to zero after four clock cycles. See the

description of the EEPE bit for an EEPROM write procedure.

• Bit 1 – EEPE: EEPROM Programming Enable

The EEPROM Write Enable Signal EEPE is the write strobe to the EEPROM. When address

and data are correctly set up, the EEPE bit must be written to one to write the value into the

EEPROM. The EEMPE bit must be written to one before a logical one is written to EEPE, other-

22

7766E–AVR–04/10

ATmega16U4/ATmega32U4

wise no EEPROM write takes place. The following procedure should be followed when writing

the EEPROM (the order of steps 3 and 4 is not essential):

1. Wait until EEPE becomes zero.

2. Wait until SELFPRGEN in SPMCSR becomes zero.

3. Write new EEPROM address to EEAR (optional).

4. Write new EEPROM data to EEDR (optional).

5. Write a logical one to the EEMPE bit while writing a zero to EEPE in EECR.

6. Within four clock cycles after setting EEMPE, write a logical one to EEPE.

The EEPROM can not be programmed during a CPU write to the Flash memory. The software

must check that the Flash programming is completed before initiating a new EEPROM write.

Step 2 is only relevant if the software contains a Boot Loader allowing the CPU to program the

Flash. If the Flash is never being updated by the CPU, step 2 can be omitted. See “Memory Pro-

gramming” on page 346 for details about Boot programming.

Caution: An interrupt between step 5 and step 6 will make the write cycle fail, since the

EEPROM Master Write Enable will time-out. If an interrupt routine accessing the EEPROM is

interrupting another EEPROM access, the EEAR or EEDR Register will be modified, causing the

interrupted EEPROM access to fail. It is recommended to have the Global Interrupt Flag cleared

during all the steps to avoid these problems.

When the write access time has elapsed, the EEPE bit is cleared by hardware. The user soft-

ware can poll this bit and wait for a zero before writing the next byte. When EEPE has been set,

the CPU is halted for two cycles before the next instruction is executed.

• Bit 0 – EERE: EEPROM Read Enable

The EEPROM Read Enable Signal EERE is the read strobe to the EEPROM. When the correct

address is set up in the EEAR Register, the EERE bit must be written to a logic one to trigger the

EEPROM read. The EEPROM read access takes one instruction, and the requested data is

available immediately. When the EEPROM is read, the CPU is halted for four cycles before the

next instruction is executed.

The user should poll the EEPE bit before starting the read operation. If a write operation is in

progress, it is neither possible to read the EEPROM, nor to change the EEAR Register.

The calibrated Oscillator is used to time the EEPROM accesses. Table 5-3 lists the typical pro-

gramming time for EEPROM access from the CPU.

Table 5-3.

Symbol

EEPROM Programming Time

Number of Calibrated RC Oscillator Cycles

Typ Programming Time

EEPROM write

(from CPU)

26,368

3.3 ms

The following code examples show one assembly and one C function for writing to the

EEPROM. The examples assume that interrupts are controlled (e.g. by disabling interrupts glob-

ally) so that no interrupts will occur during execution of these functions. The examples also

assume that no Flash Boot Loader is present in the software. If such code is present, the

EEPROM write function must also wait for any ongoing SPM command to finish.

23

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Assembly Code Example⁽¹⁾

EEPROM_write:

; Wait for completion of previous write

sbic EECR,EEPE

rjmp EEPROM_write

; Set up address (r18:r17) in address register

out EEARH, r18

out EEARL, r17

; Write data (r16) to Data Register

out EEDR,r16

; Write logical one to EEMPE

sbi EECR,EEMPE

; Start eeprom write by setting EEPE

sbi EECR,EEPE

ret

C Code Example⁽¹⁾

void EEPROM_write(unsigned int uiAddress, unsigned char ucData)

{

/* Wait for completion of previous write */

while(EECR & (1<<EEPE))

;

/* Set up address and Data Registers */

EEAR = uiAddress;

EEDR = ucData;

/* Write logical one to EEMPE */

EECR |= (1<<EEMPE);

/* Start eeprom write by setting EEPE */

EECR |= (1<<EEPE);

}

Note:

1. See “Code Examples” on page 8.

The next code examples show assembly and C functions for reading the EEPROM. The exam-

ples assume that interrupts are controlled so that no interrupts will occur during execution of

these functions.

24

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Assembly Code Example⁽¹⁾

EEPROM_read:

; Wait for completion of previous write

sbic EECR,EEPE

rjmp EEPROM_read

; Set up address (r18:r17) in address register

out EEARH, r18

out EEARL, r17

; Start eeprom read by writing EERE

sbi EECR,EERE

; Read data from Data Register

in r16,EEDR

ret

C Code Example⁽¹⁾

unsigned char EEPROM_read(unsigned int uiAddress)

{

/* Wait for completion of previous write */

while(EECR & (1<<EEPE))

;

/* Set up address register */

EEAR = uiAddress;

/* Start eeprom read by writing EERE */

EECR |= (1<<EERE);

/* Return data from Data Register */

return EEDR;

}

Note:

1. See “Code Examples” on page 8.

5.3.5

Preventing EEPROM Corruption

During periods of low V_CC,the EEPROM data can be corrupted because the supply voltage is

too low for the CPU and the EEPROM to operate properly. These issues are the same as for

board level systems using EEPROM, and the same design solutions should be applied.

An EEPROM data corruption can be caused by two situations when the voltage is too low. First,

a regular write sequence to the EEPROM requires a minimum voltage to operate correctly. Sec-

ondly, the CPU itself can execute instructions incorrectly, if the supply voltage is too low.

EEPROM data corruption can easily be avoided by following this design recommendation:

Keep the AVR RESET active (low) during periods of insufficient power supply voltage. This can

be done by enabling the internal Brown-out Detector (BOD). If the detection level of the internal

BOD does not match the needed detection level, an external low V_CCreset Protection circuit can

be used. If a reset occurs while a write operation is in progress, the write operation will be com-

pleted provided that the power supply voltage is sufficient.

25

7766E–AVR–04/10

ATmega16U4/ATmega32U4

5.4

I/O Memory

The I/O space definition of the ATmega16U4/ATmega32U4 is shown in “Register Summary” on

page 408.

All ATmega16U4/ATmega32U4 I/Os and peripherals are placed in the I/O space. All I/O loca-

tions may be accessed by the LD/LDS/LDD and ST/STS/STD instructions, transferring data

between the 32 general purpose working registers and the I/O space. I/O Registers within the

address range 0x00 - 0x1F are directly bit-accessible using the SBI and CBI instructions. In

these registers, the value of single bits can be checked by using the SBIS and SBIC instructions.

Refer to the instruction set section for more details. When using the I/O specific commands IN

and OUT, the I/O addresses 0x00 - 0x3F must be used. When addressing I/O Registers as data

space using LD and ST instructions, 0x20 must be added to these addresses. The

ATmega16U4/ATmega32U4 is a complex microcontroller with more peripheral units than can be

supported within the 64 location reserved in Opcode for the IN and OUT instructions. For the

Extended I/O space from 0x60 - 0xFF in SRAM, only the ST/STS/STD and LD/LDS/LDD instruc-

tions can be used.

For compatibility with future devices, reserved bits should be written to zero if accessed.

Reserved I/O memory addresses should never be written.

Some of the Status Flags are cleared by writing a logical one to them. Note that, unlike most

other AVRs, the CBI and SBI instructions will only operate on the specified bit, and can therefore

be used on registers containing such Status Flags. The CBI and SBI instructions work with reg-

isters 0x00 to 0x1F only.

The I/O and peripherals control registers are explained in later sections.

5.4.1

General Purpose I/O Registers

The ATmega16U4/ATmega32U4 contains three General Purpose I/O Registers. These registers

can be used for storing any information, and they are particularly useful for storing global vari-

ables and Status Flags. General Purpose I/O Registers within the address range 0x00 - 0x1F

are directly bit-accessible using the SBI, CBI, SBIS, and SBIC instructions.

5.4.2

5.4.3

5.4.4

General Purpose I/O Register 2 – GPIOR2

Bit

7

6

5

4

3

2

1

0

MSB

R/W

0

LSB

R/W

0

GPIOR2

GPIOR1

GPIOR0

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

General Purpose I/O Register 1 – GPIOR1

Bit

7

6

5

4

3

2

1

0

MSB

R/W

0

LSB

R/W

0

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

General Purpose I/O Register 0 – GPIOR0

Bit

7

6

5

4

3

2

1

0

MSB

R/W

0

LSB

R/W

0

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

26

7766E–AVR–04/10

ATmega16U4/ATmega32U4

6. System Clock and Clock Options

6.1

Clock Systems and their Distribution

Figure 6-1 presents the principal clock systems in the AVR and their distribution. All of the clocks

need not be active at a given time. In order to reduce power consumption, the clocks to modules

not being used can be halted by using different sleep modes, as described in “Power Manage-

ment and Sleep Modes” on page 43. The clock systems are detailed below.

Figure 6-1. Clock Distribution

General I/O

Modules

High Speed

Timer

Flash and

EEPROM

USB

CPU Core

ADC

RAM

clk_ADC

clk

clk_CPU

AVR Clock

Control Unit

I/O

(1)

(2)

PLL Postcaler

clk_FLASH

clk^PLL

PLL

Reset Logic

Watchdog Timer

Source clock

System Clock

Prescaler

Watchdog

clock

PLL Input

Multiplexer

clk^PllPresc

PLL Clock

Prescaler

Clock

Multiplexer

Clock Switch

Watchdog

Oscillator

Calibrated RC

Oscillator

Crystal

Oscillator

External Clock

6.1.1

6.1.2

6.1.3

CPU Clock – clk_CPU

The CPU clock is routed to parts of the system concerned with operation of the AVR core.

Examples of such modules are the General Purpose Register File, the Status Register and the

data memory holding the Stack Pointer. Halting the CPU clock inhibits the core from performing

general operations and calculations.

I/O Clock – clk_I/O

The I/O clock is used by the majority of the I/O modules, like Timer/Counters, SPI, and USART.

The I/O clock is also used by the External Interrupt module, but note that some external inter-

rupts are detected by asynchronous logic, allowing such interrupts to be detected even if the I/O

clock is halted. Also, TWI address recognition is handled in all sleep modes.

Flash Clock – clk_FLASH

The Flash clock controls operation of the Flash interface. The Flash clock is usually active simul-

taneously with the CPU clock.

27

7766E–AVR–04/10

ATmega16U4/ATmega32U4

6.1.4

6.1.5

ADC Clock – clk_ADC

The ADC is provided with a dedicated clock domain. This allows halting the CPU and I/O clocks

in order to reduce noise generated by digital circuitry. This gives more accurate ADC conversion

results.

PLL Prescaler Clock – clk_PllPresc

The PLL requires a 8 MHz input. A prescaler allows user to use either a 8MHz or a 16MHz

source (from a crystal or an external source), using a divider (by 2) if necessary. The output of

the prescaler goes into the PLL Input multiplexer, that allows the user to select either the pres-

caler output of the System Clock Multiplexer, or the Internal 8MHz Calibrated Oscillator.

6.1.6

6.1.7

PLL Output Clock – clk_Pll

When enabled, the PLL outputs one frequency among numerous choices between 32MHz and

96 MHz. The output frequency is determined by the PLL clock register. The frequency is inde-

pendent of the power supply voltage. The PLL Output is connected to a postcaler that allows

user to generate two different frequencies (clk_USBand clk_TMR) from the common PLL signal,

each on them resulting of a selected division ratio (/1, /1.5, /2).

High-Speed Timer Clock– clk_TMR

When enabled, the PLL outputs one frequency among numerous choices between 32MHz and

96 MHz, that goes into the PLL Postcaler. The High Speed Timer frequency input is generated

from the PLL Postcaler, that proposes /1, /1.5 and /2 ratios. That can be determined from the

PLL clock register. The High Speed Timer maximum frequency input depends on the power sup-

ply voltage and reaches its maximum of 64 MHz at 5V.

6.1.8

USB Clock – clk_USB

The USB hardware module needs for a 48 MHz clock. This clock is generated from the on-chip

PLL. The output of the PLL passes through the PLL Postcaler where the frequency can be either

divided by 2 or directly connected to the clk_USBsignal.

6.2

Clock Sources

The device has the following clock source options, selectable by Flash Fuse bits as shown

below. The clock from the selected source is input to the AVR clock generator, and routed to the

appropriate modules.

Table 6-1.

Device Clocking Options Select⁽¹⁾

Device Clocking Option

CKSEL[3:0]

(or EXCKSEL[3:0])

Low Power Crystal Oscillator

Reserved

1111 - 1000

0111 - 0110

0101 - 0100

0011

Low Frequency Crystal Oscillator

Reserved

Calibrated Internal RC Oscillator

External Clock

0010

0000

Reserved

0001

Note:

1. For all fuses “1” means unprogrammed while “0” means programmed.

28

7766E–AVR–04/10

ATmega16U4/ATmega32U4

6.2.1

6.2.2

6.2.3

Default Clock Source ATmega16U4 and ATmega32U4

The device is shipped with Low Power Crystal Oscillator (8.0MHz-16MHz) enabled and with the

fuse CKDIV8 programmed, resulting in 1.0MHz system clock with an 8 MHz crystal. See Table

28-5 on page 348 for an overview of the default Clock Selection Fuse setting.

Default Clock Source ATmega16U4RC and ATmega32U4RC

The device is shipped with Calibrated Internal RC oscillator (8.0MHz) enabled and with the fuse

CKDIV8 programmed, resulting in 1.0MHz system clock. See Table 28-5 on page 348 for an

overview of the default Clock Selection Fuse setting.

Clock Startup Sequence

Any clock source needs a sufficient V_CCto start oscillating and a minimum number of oscillating

cycles before it can be considered stable.

To ensure sufficient V_CC, the device issues an internal reset with a time-out delay (t_TOUT) after

the device reset is released by all other reset sources. “On-chip Debug System” on page 48

describes the start conditions for the internal reset. The delay (t_TOUT) is timed from the Watchdog

Oscillator and the number of cycles in the delay is set by the SUTx and CKSELx fuse bits. The

selectable delays are shown in Table 6-2. The frequency of the Watchdog Oscillator is voltage

dependent as shown in Table 6-2.

Table 6-2.

Number of Watchdog Oscillator Cycles

Typ Time-out (V_CC= 5.0V)

Typ Time-out (V_CC= 3.0V)

Number of Cycles

0 ms

4.1 ms

65 ms

0 ms

4.3 ms

69 ms

0

512

8K (8,192)

Main purpose of the delay is to keep the AVR in reset until it is supplied with minimum Vcc. The

delay will not monitor the actual voltage and it will be required to select a delay longer than the

Vcc rise time. If this is not possible, an internal or external Brown-Out Detection circuit should be

used. A BOD circuit will ensure sufficient Vcc before it releases the reset, and the time-out delay

can be disabled. Disabling the time-out delay without utilizing a Brown-Out Detection circuit is

not recommended.

The oscillator is required to oscillate for a minimum number of cycles before the clock is consid-

ered stable. An internal ripple counter monitors the oscillator output clock, and keeps the internal

reset active for a given number of clock cycles. The reset is then released and the device will

start to execute. The recommended oscillator start-up time is dependent on the clock type, and

varies from 6 cycles for an externally applied clock to 32K cycles for a low frequency crystal.

The start-up sequence for the clock includes both the time-out delay and the start-up time when

the device starts up from reset. When starting up from Power-save or Power-down mode, Vcc is

assumed to be at a sufficient level and only the start-up time is included.

6.3

Low Power Crystal Oscillator

Pins XTAL1 and XTAL2 are input and output, respectively, of an inverting amplifier which can be

configured for use as an On-chip Oscillator, as shown in Figure 6-2. Either a quartz crystal or a

ceramic resonator may be used.

This Crystal Oscillator is a low power oscillator, with reduced voltage swing on the XTAL2 out-

put. It gives the lowest power consumption, but is not capable of driving other clock inputs.

29

7766E–AVR–04/10

ATmega16U4/ATmega32U4

C1 and C2 should always be equal for both crystals and resonators. The optimal value of the

capacitors depends on the crystal or resonator in use, the amount of stray capacitance, and the

electromagnetic noise of the environment. Some initial guidelines for choosing capacitors for

use with crystals are given in Table 6-3. For ceramic resonators, the capacitor values given by

the manufacturer should be used.

Figure 6-2. Crystal Oscillator Connections

C2

XTAL2

C1

XTAL1

GND

The Low Power Oscillator can operate in three different modes, each optimized for a specific fre-

quency range. The operating mode is selected by the fuses CKSEL[3..1] as shown in Table 6-3.

Table 6-3.

Low Power Crystal Oscillator Operating Modes⁽³⁾

Recommended Range for Capacitors

Frequency Range⁽¹⁾(MHz)

CKSEL3..1

100⁽²⁾

101

C1 and C2 (pF)

0.4 - 0.9

0.9 - 3.0

3.0 - 8.0

8.0 - 16.0

–

12 - 22

110

111

Notes: 1. The frequency ranges are preliminary values. Actual values are TBD.

2. This option should not be used with crystals, only with ceramic resonators.

3. If 8 MHz frequency exceeds the specification of the device (depends on V_CC), the CKDIV8

Fuse can be programmed in order to divide the internal frequency by 8. It must be ensured

that the resulting divided clock meets the frequency specification of the device.

The CKSEL0 Fuse together with the SUT1..0 Fuses select the start-up times as shown in Table

6-4.

Table 6-4.

Start-up Times for the Low Power Crystal Oscillator Clock Selection

Start-up Time from

Power-down and

Power-save

Additional Delay

from Reset

Oscillator Source /

Power Conditions

(V_CC= 5.0V)

CKSEL0 SUT1..0

Ceramic resonator, fast

rising power

258 CK

1K CK

14CK + 4.1 ms⁽¹⁾

14CK + 65 ms⁽¹⁾

14CK⁽²⁾

0

00

01

10

11

Ceramic resonator,

slowly rising power

Ceramic resonator,

BOD enabled

Ceramic resonator, fast

rising power

14CK + 4.1 ms⁽²⁾

30

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 6-4.

Start-up Times for the Low Power Crystal Oscillator Clock Selection (Continued)

Start-up Time from

Power-down and

Power-save

Additional Delay

from Reset

Oscillator Source /

Power Conditions

(V_CC= 5.0V)

CKSEL0 SUT1..0

Ceramic resonator,

slowly rising power

1K CK

16K CK

14CK + 65 ms⁽²⁾

1

00

01

10

11

Crystal Oscillator, BOD

enabled

14CK

Crystal Oscillator, fast

rising power

14CK + 4.1 ms

14CK + 65 ms

Crystal Oscillator,

slowly rising power

Notes: 1. These options should only be used when not operating close to the maximum frequency of the

device, and only if frequency stability at start-up is not important for the application. These

options are not suitable for crystals.

2. These options are intended for use with ceramic resonators and will ensure frequency stability

at start-up. They can also be used with crystals when not operating close to the maximum fre-

quency of the device, and if frequency stability at start-up is not important for the application.

Table 6-5.

Start-up times for the internal calibrated RC Oscillator clock selection

Start-up Time from Power-

down and Power-save

Additional Delay from

Reset (V_CC= 5.0V)

Power Conditions

BOD enabled

SUT1..0

00

6 CK

14CK

Fast rising power

Slowly rising power

14CK + 4.1 ms

14CK + 65 ms⁽¹⁾

01

6 CK

10

Reserved

11

Note:

1. The device is shipped with this option selected.

6.4

Low Frequency Crystal Oscillator

The device can utilize a 32.768 kHz watch crystal as clock source by a dedicated Low Fre-

quency Crystal Oscillator. The crystal should be connected as shown in Figure 6-2. When this

Oscillator is selected, start-up times are determined by the SUT Fuses and CKSEL0 as shown in

Table 6-6.

Start-up Times for the Low Frequency Crystal Oscillator Clock Selection

Start-up Time from

Power-down and

Power-save

Additional Delay

from Reset

Power Conditions

BOD enabled

(V_CC= 5.0V)

CKSEL0 SUT1..0

1K CK

14CK⁽¹⁾

0

1

00

01

10

11

00

01

10

11

Fast rising power

Slowly rising power

14CK + 4.1 ms⁽¹⁾

14CK + 65 ms⁽¹⁾

1K CK

Reserved

32K CK

Reserved

BOD enabled

14CK

Fast rising power

Slowly rising power

14CK + 4.1 ms

14CK + 65 ms

31

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Note:

1. These options should only be used if frequency stability at start-up is not important for the

application.

6.5

Calibrated Internal RC Oscillator

The calibrated internal RC Oscillator by default provides a 8.0 MHz clock. This frequency is

nominal value at 3V and 25°C. The device is shipped with the CKDIV8 Fuse programmed. See

“System Clock Prescaler” on page 37 for more details. This clock may be selected as the system

clock by programming the CKSEL Fuses as shown in Table 6-7. If selected, it will operate with

no external components. During reset, hardware loads the calibration byte into the OSCCAL

Register and thereby automatically calibrates the RC Oscillator. At 3V and 25°C, this calibration

gives a frequency of 8 MHz 1%. The oscillator can be calibrated to any frequency in the range

7.3 - 8.1 MHz within 1% accuracy, by changing the OSCCAL register. When this Oscillator is

used as the chip clock, the Watchdog Oscillator will still be used for the Watchdog Timer and for

the Reset Time-out. For more information on the pre-programmed calibration value, see the sec-

tion “Calibration Byte” on page 349

Table 6-7.

Internal Calibrated RC Oscillator Operating Modes⁽¹⁾⁽³⁾

Frequency Range⁽²⁾(MHz)

CKSEL[3:0]

0010

7.3 - 8.1

Notes: 1. The device is shipped with this option selected.

2. The frequency ranges are preliminary values. Actual values are TBD.

3. If 8 MHz frequency exceeds the specification of the device (depends on V_CC), the CKDIV8

Fuse can be programmed in order to divide the internal frequency by 8.

When this Oscillator is selected, start-up times are determined by the SUT Fuses as shown in

Table 6-5 on page 31.

Table 6-8.

Start-up times for the internal calibrated RC Oscillator clock selection

Start-up Time from Power-

down and Power-save

Additional Delay from

Reset (V_CC= 5.0V)

Power Conditions

BOD enabled

SUT1..0

00

6 CK

14CK

Fast rising power

Slowly rising power

14CK + 4.1 ms

14CK + 65 ms

01

6 CK

10

Reserved

11

6.5.1

Oscillator Calibration Register – OSCCAL

Bit

7

6

5

4

3

2

1

0

CAL7

R/W

CAL6

R/W

CAL5

R/W

CAL4

R/W

CAL3

R/W

CAL2

R/W

CAL1

R/W

CAL0

R/W

OSCCAL

Read/Write

Initial Value

Device Specific Calibration Value

• Bits 7..0 – CAL7..0: Oscillator Calibration Value

The Oscillator Calibration Register is used to trim the Calibrated Internal RC Oscillator to

remove process variations from the oscillator frequency. The factory-calibrated value is automat-

ically written to this register during chip reset, giving an oscillator frequency of 8.0 MHz at 25°C.

The application software can write this register to change the oscillator frequency. The calibra-

tion range is +/- 40% and linear (calibration step ~0.4%). With typical process at 25°C the code

should be 127 for 8 MHz. Input value of 0x00 gives the lowest frequency, and 0xFF the highest.

32

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The temperature sensitivity is quite linear but as said previously depends on the process. To

determine its slope, the frequency must be measured at two temperatures. The temperature

sensor of the ATmega16U4/ATmega32U4 allows such an operation, that is detailed on Section

24.6.1 ”Sensor Calibration” on page 300. It is then possible to calibrate the oscillator frequency

in function of the temperature measured.

Note that this oscillator is used to time EEPROM and Flash write accesses, and these write

times will be affected accordingly. If the EEPROM or Flash are written, do not calibrate to more

than 8.8 MHz. Otherwise, the EEPROM or Flash write may fail.

6.5.2

Oscillator Control Register – RCCTRL

Bit

7

6

5

-

4

-

3

-

2

-

1

-

0

-

RCFREQ

RCCTRL

Read/Write

Initial Value

R

0

R

0

R

0

R

0

R

0

R

0

R

0

R/W

0

Bits 7..1 – Reserved

Do not set these bits. Bits should be read as ‘0’.

Bit 0– RCFREQ: RC Oscillator Frequency Select

When this bit is cleared (default value), the RC Oscillator output frequency is set to 8 MHz.

When the bit is set, the RC output frequency is 1 MHz. Note that the OSCCAL value has the

same effect on both 8 MHz and 1 MHz output modes (~0.4% / step).

6.6

External Clock

The device can utilize a external clock source as shown in Figure 6-3. To run the device on an

external clock, the CKSEL Fuses must be programmed as shown in Table 6-1.

Figure 6-3. External Clock Drive Configuration

NC

XTAL2

EXTERNAL

CLOCK

XTAL1

GND

SIGNAL

When this clock source is selected, start-up times are determined by the SUT Fuses as shown in

Table 6-9.

33

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 6-9.

Start-up Times for the External Clock Selection

Start-up Time from Power-

down and Power-save

Additional Delay from

Reset (V_CC= 5.0V)

Power Conditions

SUT1..0

00

BOD enabled

6 CK

14CK

Fast rising power

Slowly rising power

14CK + 4.1 ms

14CK + 65 ms

01

6 CK

10

Reserved

11

When applying an external clock, it is required to avoid sudden changes in the applied clock fre-

quency to ensure stable operation of the MCU. A variation in frequency of more than 2% from

one clock cycle to the next can lead to unpredictable behavior. If changes of more than 2% is

required, ensure that the MCU is kept in Reset during the changes.

Note that the System Clock Prescaler can be used to implement run-time changes of the internal

clock frequency while still ensuring stable operation. Refer to “System Clock Prescaler” on page

37 for details.

6.7

Clock Switch

The ATmega16U4/ATmega32U4 product includes a Clock Switch controller, that allows user to

switch from one clock source to another one by software, in order to control application power

and execution time with more accuracy.

6.7.1

Example of use

The modification may be needed when the device enters in USB Suspend mode. It then

switches from External Clock to Calibrated RC Oscillator in order to reduce consumption and

wake-up delay. In such a configuration, the External Clock is disabled. The firmware can then

use the watchdog timer to be woken-up from power-down in order to check if there is an event

on the application. If an event occurs on the application or if the USB controller signals a non-

idle state on the USB line (Resume for example), the firmware switches the Clock Multiplexer

from the Calibrated RC Oscillator to the External Clock. in order to restart USB operation.

This feature can only be used to switch between Calibrated 8 MHz RC Oscillator, External Clock

and Low Power Crystal Oscillator. The Low Frequency Crystal Oscillator must not be used with

this feature.

34

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 6-4. Example of clock switching with wake-up from USB Host

resume

Resume from Host

1

USB

(Suspend)

non-Idle

Idle

Ext

non-Idle

Ext

1

CPU Clock

RC

External

Oscillator

RC oscillator

3ms

watchdogwake-up

frompower-down

Figure 6-5. Example of clock switching with wake-up from Device

upstream-resume

2

Upstream Resume from device

USB

(Suspend)

non-Idle

Idle

Ext

non-Idle

Ext

2

CPU Clock

RC

External

Oscillator

RC oscillator

3ms

watchdogwake-up

frompower-down

6.8

Clock switch Algorithm

6.8.1

Switch from external clock to RC clock

if (Usb_suspend_detected())

// if (UDINT.SUSPI == 1)

{

Usb_ack_suspend();

Usb_freeze_clock();

Disable_pll();

// UDINT.SUSPI = 0;

// USBCON.FRZCLK = 1;

// PLLCSR.PLLE = 0;

// CLKSEL0.RCE = 1;

// while (CLKSTA.RCON != 1);

// CLKSEL0.CLKS = 0;

// CLKSEL0.EXTE = 0;

Enable_RC_clock();

while (!RC_clock_ready());

Select_RC_clock();

Disable_external_clock();

}

35

7766E–AVR–04/10

ATmega16U4/ATmega32U4

6.8.2

Switch from RC clock to external clock

if (Usb_wake_up_detected())

{

// if (UDINT.WAKEUPI == 1)

Usb_ack_wake_up();

// UDINT.WAKEUPI = 0;

// CKSEL0.EXTE = 1;

Enable_external_clock();

while (!External_clock_ready()); // while (CLKSTA.EXTON != 1);

Select_external_clock();

Enable_pll();

// CLKSEL0.CLKS = 1;

// PLLCSR.PLLE = 1;

Disable_RC_clock();

while (!Pll_ready());

Usb_unfreeze_clock();

// CLKSEL0.RCE = 0;

// while (PLLCSR.PLOCK != 1);

// USBCON.FRZCLK = 0;

}

6.8.3

CLKSEL0 – Clock Selection Register 0

Bit

7

6

5

4

3

2

1

-

0

RCSUT1

RCSUT0

EXSUT1

R/W

0

EXSUT0

R/W

0

RCE

R/W

EXTE

R/W

CLKS

R/W

CLKSEL0

Read/Write

Initial Value

R/W

0

R/W

0

R

See Bit Description

• Bit 7-6 – RCSUT[1:0]: SUT for RC oscillator

These 2 bits are the SUT value for the RC Oscillator. If the RC oscillator is selected by fuse bits,

the SUT fuse are copied into these bits. A firmware change will not have any effect because this

additional start-up time is only used after a reset and not after a clock switch.

• Bit 5-4 – EXSUT[1:0]: SUT for External Clock/ Low Power Crystal Oscillator

These 2 bits are the SUT value for the External Clock / Low Power Crystal Oscillator. If the

External Clock / Low Power Crystal Oscillator is selected by fuse bits, the SUT fuses are copied

into these bits. The firmware can modify these bits by writing a new value. This value will be

used at the next start of the External Clock / Low Power Crystal Oscillator.

• Bit 3 – RCE: Enable RC Oscillator

The RCE bit must be written to logic one to enable the RC Oscillator. The RCE bit must be writ-

ten to logic zero to disable the RC Oscillator.

• Bit 2 – EXTE: Enable External Clock / Low Power Crystal Oscillator

The OSCE bit must be written to logic one to enable External Clock / Low Power Crystal Oscilla-

tor. The OSCE bit must be written to logic zero to disable the External Clock / Low Power Crystal

Oscillator.

• Bit 0 – CLKS: Clock Selector

The CLKS bit must be written to logic one to select the External Clock / Low Power Crystal Oscil-

lator as CPU clock. The CLKS bit must be written to logic zero to select the RC Oscillator as

CPU clock. After a reset, the CLKS bit is set by hardware if the External Clock / Low Power Crys-

tal Oscillator is selected by the fuse bits configuration.

The firmware has to check if the clock is correctly started before selected it.

6.8.4

CLKSEL1 – Clock Selection Register 1

Bit

7

6

5

4

3

2

1

0

RCCKS

EL3

RCCKS

EL2

RCCKS

EL1

RCCKS

EL0

EXCKS

EL3

EXCKS

EL2

EXCKS

EL1

EXCKS

EL0

CLKSEL1

Read/Write

Initial Value

R/W

0

R/W

0

R/W

1

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

36

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 7-4 – RCCKSEL[3:0]: CKSEL for RC oscillator

Clock configuration for the RC Oscillator. After a reset, this part of the register is loaded with the

0010b value that corresponds to the RC oscillator. Modifying this value by firmware before

switching to RC oscillator is prohibited because the RC clock will not start.

• Bit 3-0 – EXCKSEL[3:0]: CKSEL for External Clock / Low Power Crystal Oscillator

Clock configuration for the External Clock / Low Power Crystal Oscillator. After a reset, if the

External Clock / Low Power Crystal Oscillator is selected by fuse bits, this part of the register is

loaded with the fuse configuration. Firmware can modify it to change the start-up time after the

clock switch.

See “Device Clocking Options Select(1)” on page 28 for EXCKSEL[3:0] configuration. Only Low

Power Crystal Oscillator, Calibrated Internal RC Oscillator, and External Clock modes are

allowed.

6.8.5

CLKSTA – Clock Status Register

Bit

7

6

-

5

-

4

-

3

-

2

-

1

0

-

RCON

EXTON

CLKSTA

Read/Write

Initial Value

R

0

R

0

R

0

R

0

R

0

R

• Bit 7-2 - Reserved bits

These bits are reserved and will always read as zero.

• Bit 1 – RCON: RC Oscillator On

This bit is set by hardware to one if the RC Oscillator is running.

This bit is set by hardware to zero if the RC Oscillator is stopped.

• Bit 0 – EXTON: External Clock / Low Power Crystal Oscillator On

This bit is set by hardware to one if the External Clock / Low Power Crystal Oscillator is running.

This bit is set by hardware to zero if the External Clock / Low Power Crystal Oscillator is

stopped.

6.9

Clock Output Buffer

The device can output the system clock on the CLKO pin. To enable the output, the CKOUT

Fuse has to be programmed. This mode is suitable when the chip clock is used to drive other cir-

cuits on the system. The clock also will be output during reset, and the normal operation of I/O

pin will be overridden when the fuse is programmed. Any clock source, including the internal RC

Oscillator, can be selected when the clock is output on CLKO. If the System Clock Prescaler is

used, it is the divided system clock that is output.

6.9.1

System Clock Prescaler

The AVR USB has a system clock prescaler, and the system clock can be divided by setting the

“CLKPR – Clock Prescaler Register” on page 38. This feature can be used to decrease the sys-

tem clock frequency and the power consumption when the requirement for processing power is

low. This can be used with all clock source options, and it will affect the clock frequency of the

CPU and all synchronous peripherals. clk_I/O, clk_ADC, clk_CPU, and clk_FLASHare divided by a factor

as shown in Table 6-10.

When switching between prescaler settings, the System Clock Prescaler ensures that no

glitches occurs in the clock system. It also ensures that no intermediate frequency is higher than

neither the clock frequency corresponding to the previous setting, nor the clock frequency corre-

sponding to the new setting.

37

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The ripple counter that implements the prescaler runs at the frequency of the undivided clock,

which may be faster than the CPU's clock frequency. Hence, it is not possible to determine the

state of the prescaler - even if it were readable, and the exact time it takes to switch from one

clock division to the other cannot be exactly predicted. From the time the CLKPS values are writ-

ten, it takes between T1 + T2 and T1 + 2 * T2 before the new clock frequency is active. In this

interval, 2 active clock edges are produced. Here, T1 is the previous clock period, and T2 is the

period corresponding to the new prescaler setting.

To avoid unintentional changes of clock frequency, a special write procedure must be followed

to change the CLKPS bits:

1. Write the Clock Prescaler Change Enable (CLKPCE) bit to one and all other bits in

CLKPR to zero.

2. Within four cycles, write the desired value to CLKPS while writing a zero to CLKPCE.

Interrupts must be disabled when changing prescaler setting to make sure the write procedure is

not interrupted.

6.9.2

CLKPR – Clock Prescaler Register

Bit

7

6

–

R

0

5

–

R

0

4

–

R

0

3

2

1

0

CLKPCE

CLKPS3

R/W

CLKPS2

R/W

CLKPS1

R/W

CLKPS0

R/W

CLKPR

Read/Write

Initial Value

R/W

0

See Bit Description

• Bit 7 – CLKPCE: Clock Prescaler Change Enable

The CLKPCE bit must be written to logic one to enable change of the CLKPS bits. The CLKPCE

bit is only updated when the other bits in CLKPR are simultaneously written to zero. CLKPCE is

cleared by hardware four cycles after it is written or when CLKPS bits are written. Rewriting the

CLKPCE bit within this time-out period does neither extend the time-out period, nor clear the

CLKPCE bit.

• Bits 3..0 – CLKPS[3..0]: Clock Prescaler Select Bits 3 - 0

These bits define the division factor between the selected clock source and the internal system

clock. These bits can be written run-time to vary the clock frequency to suit the application

requirements. As the divider divides the master clock input to the MCU, the speed of all synchro-

nous peripherals is reduced when a division factor is used. The division factors are given in

Table 6-10.

The CKDIV8 Fuse determines the initial value of the CLKPS bits. If CKDIV8 is unprogrammed,

the CLKPS bits will be reset to “0000”. If CKDIV8 is programmed, CLKPS bits are reset to

“0011”, giving a division factor of 8 at start up. This feature should be used if the selected clock

source has a higher frequency than the maximum frequency of the device at the present operat-

ing conditions. Note that any value can be written to the CLKPS bits regardless of the CKDIV8

Fuse setting. The Application software must ensure that a sufficient division factor is chosen if

38

7766E–AVR–04/10

ATmega16U4/ATmega32U4

the selected clock source has a higher frequency than the maximum frequency of the device at

the present operating conditions. The device is shipped with the CKDIV8 Fuse programmed.

Table 6-10. Clock Prescaler Select

CLKPS3

CLKPS2

CLKPS1

CLKPS0

Clock Division Factor

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

2

4

8

16

32

64

128

256

Reserved

6.10 PLL

The PLL is used to generate internal high frequency (up to 96MHz) clock for USB interface

and/or High Speed Timer module, the PLL input is supplied from an external low-frequency

clock (the crystal oscillator or external clock input pin from XTAL1).

6.10.1

Internal PLL

The internal PLL in ATmega16U4/ATmega32U4 generates a clock frequency between 32MHz

and 96 MHz from nominally 8MHz input.

The source of the 8MHz PLL input clock is the output of the internal PLL clock prescaler that

generates the 8MHz from the clock source multiplexer output (See Section 6.10.2 for PLL inter-

face). The PLL prescaler allows a direct connection (8MHz oscillator) or a divide-by-2 stage for a

16MHz clock input.

The PLL output signal enters the PLL Postcaler stage before being distributed to the USB and

High Speed Timer modules. Each of these modules can choose an independent division ratio.

39

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 6-6. PLL Clocking System

PLOCK

PINDIV

PINMUX

PLLE

PLLTM1:0

CKSEL3:0

/1.5

10

01

11

Lock

Detector

clk

TMR

XTAL1

XTAL2

XTAL

OSCILLATOR

PLL clock

Prescaler

0

1

PLL

/2

RC OSCILLATOR

8 MHz

clk

8MHz

1

0

To System

Clock Prescaler

clk

USB

PDIV3..0

PLLUSB

6.10.2

PLL Control and Status Register – PLLCSR

Bit

7

6

5

4

3

2

1

0

$29 ($29)

Read/Write

Initial Value

PINDIV

R/W

0

PLLE

PLOCK

PLLCSR

R

0

R

0

R

0

R

0

R

0

R/W

0

R

0

• Bit 7..5 – Res: Reserved Bits

These bits are reserved bits in the ATmega16U4/ATmega32U4 and always read as zero.

• Bit 4 – PINDIV PLL Input Prescaler (1:1, 1:2)

These bits allow to configure the PLL input prescaler to generate the 8MHz input clock for the

PLL from either a 8 or 16 MHz input.

When using a 8 MHz clock source, this bit must be set to 0 before enabling PLL (1:1).

When using a 16 MHz clock source, this bit must be set to 1 before enabling PLL (1:2).

• Bit 3..2 – Res: Reserved Bits

These bits are reserved bits in the ATmega16U4/ATmega32U4 and always read as zero.

• Bit 1 – PLLE: PLL Enable

When the PLLE is set, the PLL is started. Note that the Calibrated 8 MHz Internal RC oscillator is

automatically enabled when the PLLE bit is set and with PINMUX (see PLLFRQ register) is set.

The PLL must be disabled before entering Power down mode in order to stop Internal RC Oscil-

lator and avoid extra-consumption.

• Bit 0 – PLOCK: PLL Lock Detector

When the PLOCK bit is set, the PLL is locked to the reference clock. After the PLL is enabled, it

takes about several ms for the PLL to lock. To clear PLOCK, clear PLLE.

6.10.3

PLL Frequency Control Register – PLLFRQ

Bit

$32

7

6

5

4

3

2

1

0

PINMUX

PLLUSB

PLLTM1

R/W

0

PLLTM0

R/W

0

PDIV3

PDIV2

PDIV1

PDIV0

PLLFRQ

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

1

R/W

0

R/W

0

40

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 7– PINMUX: PLL Input Multiplexer

This bit selects the clock input of the PLL:

– PINMUX = 0: the PLL input is connected to the PLL Prescaler, that has the Primary

System Clock as source

– PINMUX = 1: the PLL input is directly connected to the Internal Calibrated 8MHz RC

Oscillator. This mode allows to work in USB Low Speed mode with no crystal or

using a crystal with a value different of 8/16MHz.

• Bit 6– PLLUSB: PLL Postcaler for USB Peripheral

This bit select the division factor between the PLL output frequency and the USB module input

frequency:

– PLLUSB = 0: no division, direct connection (if PLL Output = 48 MHz)

– PLLUSB = 1: PLL Output frequency is divided by 2 and sent to USB module (if PLL

Output = 96MHz)

• Bit 5..4 – PLLTM1:0: PLL Postcaler for High Speed Timer

These bits codes for the division factor between the PLL Output Frequency and the High Speed

Timer input frequency.

Note that the division factor 1.5 will introduce some jitter in the clock, but keeping the error null

since the average duty cycle is 50%. See Figure 6-7 for more details.

PLLTM1

PLLTM0

PLL Postcaler Factor for High-Speed Timer

0

1

0

1

0

1

0 (Disconnected)

1

1.5

2

Figure 6-7. PLL Postcaler operation with division factor = 1.5

Fi

2

Fi x ---

3

• Bit 3..0 – PDIV3:0 PLL Lock Frequency

These bits configure the PLL internal VCO clock reference according to the required output fre-

quency value.

PDIV3

PDIV2

PDIV1

PDIV0

PLL Output Frequency

Not allowed

40 MHz

0

1

0

1

0

1

0

1

0

48 MHz

41

7766E–AVR–04/10

ATmega16U4/ATmega32U4

PDIV3

PDIV2

PDIV1

PDIV0

PLL Output Frequency

56 MHz

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

Not allowed

72 MHz

80 MHz

88 MHz

96 MHz

Not allowed

The optimal PLL configuration at 5V is: PLL output frequency = 96 MHz, divided by 1.5 to gener-

ate the 64 MHz High Speed Timer clock, and divided by 2 to generate the 48 MHz USB clock.

42

7766E–AVR–04/10

ATmega16U4/ATmega32U4

7. Power Management and Sleep Modes

Sleep modes enable the application to shut down unused modules in the MCU, thereby saving

power. The AVR provides various sleep modes allowing the user to tailor the power consump-

tion to the application’s requirements.

To enter any of the five sleep modes, the SE bit in SMCR must be written to logic one and a

SLEEP instruction must be executed. The SM2, SM1, and SM0 bits in the SMCR Register select

which sleep mode (Idle, ADC Noise Reduction, Power-down, Power-save, or Standby) will be

activated by the SLEEP instruction. See Table 7-1 for a summary. If an enabled interrupt occurs

while the MCU is in a sleep mode, the MCU wakes up. The MCU is then halted for four cycles in

addition to the start-up time, executes the interrupt routine, and resumes execution from the

instruction following SLEEP. The contents of the Register File and SRAM are unaltered when

the device wakes up from sleep. If a reset occurs during sleep mode, the MCU wakes up and

executes from the Reset Vector.

Figure 6-1 on page 27 presents the different clock systems in the ATmega16U4/ATmega32U4,

and their distribution. The figure is helpful in selecting an appropriate sleep mode.

7.0.1

Sleep Mode Control Register – SMCR

The Sleep Mode Control Register contains control bits for power management.

Bit

7

–

R

0

6

–

R

0

5

–

R

0

4

–

R

0

3

2

1

0

SM2

R/W

0

SM1

R/W

0

SM0

R/W

0

SE

R/W

0

SMCR

Read/Write

Initial Value

• Bits 3, 2, 1 – SM2..0: Sleep Mode Select Bits 2, 1, and 0

These bits select between the six available sleep modes as shown in Table 7-1.

Table 7-1.

Sleep Mode Select

SM2

SM1

SM0

Sleep Mode

Idle

0

1

0

1

0

1

0

1

0

1

0

1

0

1

ADC Noise Reduction

Power-down

Power-save

Reserved

Standby⁽¹⁾

Extended Standby⁽¹⁾

Note: 1. Standby modes are only recommended for use with external crystals or resonators.

• Bit 1 – SE: Sleep Enable

The SE bit must be written to logic one to make the MCU enter the sleep mode when the SLEEP

instruction is executed. To avoid the MCU entering the sleep mode unless it is the programmer’s

purpose, it is recommended to write the Sleep Enable (SE) bit to one just before the execution of

the SLEEP instruction and to clear it immediately after waking up.

43

7766E–AVR–04/10

ATmega16U4/ATmega32U4

7.1

Idle Mode

When the SM2..0 bits are written to 000, the SLEEP instruction makes the MCU enter Idle

mode, stopping the CPU but allowing the USB, SPI, USART, Analog Comparator, ADC, 2-wire

Serial Interface, Timer/Counters, Watchdog, and the interrupt system to continue operating. This

sleep mode basically halts clk_CPUand clk_FLASH, while allowing the other clocks to run.

Idle mode enables the MCU to wake up from external triggered interrupts as well as internal

ones like the Timer Overflow and USART Transmit Complete interrupts. If wake-up from the

Analog Comparator interrupt is not required, the Analog Comparator can be powered down by

setting the ACD bit in the Analog Comparator Control and Status Register – ACSR. This will

reduce power consumption in Idle mode. If the ADC is enabled, a conversion starts automati-

cally when this mode is entered.

7.2

ADC Noise Reduction Mode

When the SM2..0 bits are written to 001, the SLEEP instruction makes the MCU enter ADC

Noise Reduction mode, stopping the CPU but allowing the ADC, the external interrupts, 2-wire

Serial Interface address match and the Watchdog to continue operating (if enabled). This sleep

mode basically halts clkI/O, clkCPU, and clkFLASH, while allowing the other clocks to run

(including clkUSB).

This improves the noise environment for the ADC, enabling higher resolution measurements. If

the ADC is enabled, a conversion starts automatically when this mode is entered. Apart form the

ADC Conversion Complete interrupt, only an External Reset, a Watchdog System Reset, a

Watchdog interrupt, a Brown-out Reset, a 2-wire serial interface interrupt, an SPM/EEPROM

ready interrupt, an external level interrupt on INT6, an external interrupt on INT3:0 or a pin

change interrupt can wake up the MCU from ADC Noise Reduction mode.

7.3

Power-down Mode

When the SM2..0 bits are written to 010, the SLEEP instruction makes the MCU enter Power-

down mode. In this mode, the external Oscillator is stopped, while the external interrupts, the 2-

wire Serial Interface, and the Watchdog continue operating (if enabled). Only an External Reset,

a Watchdog Reset, a Brown-out Reset, 2-wire Serial Interface address match, an external level

interrupt on INT6, an external interrupt on INT3:0, a pin change interrupt or an asynchronous

USB interrupt sources (VBUSTI, WAKEUPI), can wake up the MCU. This sleep mode basically

halts all generated clocks, allowing operation of asynchronous modules only.

Note that if a level triggered interrupt is used for wake-up from Power-down mode, the changed

level must be held for some time to wake up the MCU. Refer to “External Interrupts” on page 85

for details.

When waking up from Power-down mode, there is a delay from the wake-up condition occurs

until the wake-up becomes effective. This allows the clock to restart and become stable after

having been stopped. The wake-up period is defined by the same CKSEL Fuses that define the

Reset Time-out period, as described in “Clock Sources” on page 28.

7.4

Power-save Mode

When the SM2..0 bits are written to 011, the SLEEP instruction makes the MCU enter Power-

save mode. For compatibility reasons with AT90USB64/128 this mode is still present but since

Timer 2 Asynchronous operation is not present here, this mode is identical to Power-down.

44

7766E–AVR–04/10

ATmega16U4/ATmega32U4

7.5

Standby Mode

When the SM2..0 bits are 110 and an external crystal/resonator clock option is selected, the

SLEEP instruction makes the MCU enter Standby mode. This mode is identical to Power-down

with the exception that the Oscillator is kept running. From Standby mode, the device wakes up

in six clock cycles.

7.6

Extended Standby Mode

When the SM2..0 bits are 111 and an external crystal/resonator clock option is selected, the

SLEEP instruction makes the MCU enter Extended Standby mode. For compatibility reasons

with AT90USB64/128 this mode is still present but since Timer 2 Asynchronous operation is not

present here, this mode is identical to Standby-mode.

Table 7-2.

Active Clock Domains and Wake-up Sources in the Different Sleep Modes.

Active Clock Domains

Oscillators

Wake-up Sources

Sleep Mode

Idle

X

ADCNRM

Power-down

Power-save

Standby⁽¹⁾

X⁽²⁾

X

Extended

Standby

X⁽²⁾

X

Notes: 1. Only recommended with external crystal or resonator selected as clock source.

2. For INT6, only level interrupt.

3. Asynchronous USB interrupts are VBUSTI and WAKEUPI.

7.7

Power Reduction Register

The Power Reduction Register, PRR, provides a method to stop the clock to individual peripher-

als to reduce power consumption. The current state of the peripheral is frozen and the I/O

registers can not be read or written. Resources used by the peripheral when stopping the clock

will remain occupied, hence the peripheral should in most cases be disabled before stopping the

clock. Waking up a module, which is done by clearing the bit in PRR, puts the module in the

same state as before shutdown.

Module shutdown can be used in Idle mode and Active mode to significantly reduce the overall

power consumption. In all other sleep modes, the clock is already stopped.

45

7766E–AVR–04/10

ATmega16U4/ATmega32U4

7.7.1

Power Reduction Register 0 - PRR0

Bit

7

6

5

4

–

R

0

3

2

1

-

0

PRTWI

R/W

0

PRTIM2

R/W

0

PRTIM0

R/W

0

PRTIM1

R/W

0

PRSPI

R/W

0

PRADC

R/W

0

PRR0

Read/Write

Initial Value

R

0

• Bit 7 - PRTWI: Power Reduction TWI

Writing a logic one to this bit shuts down the TWI by stopping the clock to the module. When

waking up the TWI again, the TWI should be re initialized to ensure proper operation.

• Bit 6 - Res: Reserved bit

This bits is reserved and will always read as zero.

• Bit 5 - PRTIM0: Power Reduction Timer/Counter0

Writing a logic one to this bit shuts down the Timer/Counter0 module. When the Timer/Counter0

is enabled, operation will continue like before the shutdown.

• Bit 4 - Res: Reserved bit

This bit is reserved and will always read as zero.

• Bit 3 - PRTIM1: Power Reduction Timer/Counter1

Writing a logic one to this bit shuts down the Timer/Counter1 module. When the Timer/Counter1

is enabled, operation will continue like before the shutdown.

• Bit 2 - PRSPI: Power Reduction Serial Peripheral Interface

Writing a logic one to this bit shuts down the Serial Peripheral Interface by stopping the clock to

the module. When waking up the SPI again, the SPI should be re initialized to ensure proper

operation.

• Bit 1 - Res: Reserved bit

These bits are reserved and will always read as zero.

• Bit 0 - PRADC: Power Reduction ADC

Writing a logic one to this bit shuts down the ADC. The ADC must be disabled before shut down.

The analog comparator cannot use the ADC input MUX when the ADC is shut down.

7.7.2

Power Reduction Register 1 - PRR1

Bit

7

6

5

–

R

0

4

3

2

–

R

0

1

–

R

0

PRUSB

–

PRTIM4

PRTIM3

R/W

0

PRUSART1

PRR1

Read/Write

Initial Value

R/W

0

R

0

R

0

R/W

0

• Bit 7 - PRUSB: Power Reduction USB

Writing a logic one to this bit shuts down the USB by stopping the clock to the module. When

waking up the USB again, the USB should be re initialized to ensure proper operation.

• Bit 6..5 - Res: Reserved bits

These bits are reserved and will always read as zero.

46

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 4- PRTIM4: Power Reduction Timer/Counter4

Writing a logic one to this bit shuts down the Timer/Counter4 module. When the Timer/Counter4

is enabled, operation will continue like before the shutdown.

• Bit 3 - PRTIM3: Power Reduction Timer/Counter3

Writing a logic one to this bit shuts down the Timer/Counter3 module. When the Timer/Counter3

is enabled, operation will continue like before the shutdown.

• Bit 2..1 - Res: Reserved bits

These bits are reserved and will always read as zero.

• Bit 0 - PRUSART1: Power Reduction USART1

Writing a logic one to this bit shuts down the USART1 by stopping the clock to the module.

When waking up the USART1 again, the USART1 should be re initialized to ensure proper

operation.

7.8

Minimizing Power Consumption

There are several issues to consider when trying to minimize the power consumption in an AVR

controlled system. In general, sleep modes should be used as much as possible, and the sleep

mode should be selected so that as few as possible of the device’s functions are operating. All

functions not needed should be disabled. In particular, the following modules may need special

consideration when trying to achieve the lowest possible power consumption.

7.8.1

7.8.2

Analog to Digital Converter

If enabled, the ADC will be enabled in all sleep modes. To save power, the ADC should be dis-

abled before entering any sleep mode. When the ADC is turned off and on again, the next

conversion will be an extended conversion. Refer to “Analog to Digital Converter - ADC” on page

292 for details on ADC operation.

Analog Comparator

When entering Idle mode, the Analog Comparator should be disabled if not used. When entering

ADC Noise Reduction mode, the Analog Comparator should be disabled. In other sleep modes,

the Analog Comparator is automatically disabled. However, if the Analog Comparator is set up

to use the Internal Voltage Reference as input, the Analog Comparator should be disabled in all

sleep modes. Otherwise, the Internal Voltage Reference will be enabled, independent of sleep

mode. Refer to “Analog Comparator” on page 289 for details on how to configure the Analog

Comparator.

7.8.3

Brown-out Detector

If the Brown-out Detector is not needed by the application, this module should be turned off. If

the Brown-out Detector is enabled by the BODLEVEL Fuses, it will be enabled in all sleep

modes, and hence, always consume power. In the deeper sleep modes, this will contribute sig-

nificantly to the total current consumption. Refer to “Brown-out Detection” on page 52 for details

on how to configure the Brown-out Detector.

7.8.4

Internal Voltage Reference

The Internal Voltage Reference will be enabled when needed by the Brown-out Detection, the

Analog Comparator or the ADC. If these modules are disabled as described in the sections

above, the internal voltage reference will be disabled and it will not be consuming power. When

47

7766E–AVR–04/10

ATmega16U4/ATmega32U4

turned on again, the user must allow the reference to start up before the output is used. If the

reference is kept on in sleep mode, the output can be used immediately. Refer to “Internal Volt-

age Reference” on page 54 for details on the start-up time.

7.8.5

7.8.6

Watchdog Timer

If the Watchdog Timer is not needed in the application, the module should be turned off. If the

Watchdog Timer is enabled, it will be enabled in all sleep modes, and hence, always consume

power. In the deeper sleep modes, this will contribute significantly to the total current consump-

tion. Refer to “Interrupts” on page 61 for details on how to configure the Watchdog Timer.

Port Pins

When entering a sleep mode, all port pins should be configured to use minimum power. The

most important is then to ensure that no pins drive resistive loads. In sleep modes where both

the I/O clock (clk_I/O) and the ADC clock (clk_ADC) are stopped, the input buffers of the device will

be disabled. This ensures that no power is consumed by the input logic when not needed. In

some cases, the input logic is needed for detecting wake-up conditions, and it will then be

enabled. Refer to the section “Digital Input Enable and Sleep Modes” on page 69 for details on

which pins are enabled. If the input buffer is enabled and the input signal is left floating or have

an analog signal level close to V_CC/2, the input buffer will use excessive power.

For analog input pins, the digital input buffer should be disabled at all times. An analog signal

level close to V_CC/2 on an input pin can cause significant current even in active mode. Digital

input buffers can be disabled by writing to the Digital Input Disable Registers (DIDR1 and

DIDR0). Refer to “Digital Input Disable Register 1 – DIDR1” on page 291 and “Digital Input Dis-

able Register 1 – DIDR1” on page 291 for details.

7.8.7

On-chip Debug System

If the On-chip debug system is enabled by the OCDEN Fuse and the chip enters sleep mode,

the main clock source is enabled, and hence, always consumes power. In the deeper sleep

modes, this will contribute significantly to the total current consumption.

There are three alternative ways to disable the OCD system:

• Disable the OCDEN Fuse.

• Disable the JTAGEN Fuse.

• Write one to the JTD bit in MCUCR.

48

7766E–AVR–04/10

ATmega16U4/ATmega32U4

8. System Control and Reset

8.0.1

Resetting the AVR

During reset, all I/O Registers are set to their initial values, and the program starts execution

from the Reset Vector. The instruction placed at the Reset Vector must be a JMP – Absolute

Jump – instruction to the reset handling routine. If the program never enables an interrupt

source, the Interrupt Vectors are not used, and regular program code can be placed at these

locations. This is also the case if the Reset Vector is in the Application section while the Interrupt

Vectors are in the Boot section or vice versa. The circuit diagram in Figure 8-1 shows the reset

logic. Table 8-1 defines the electrical parameters of the reset circuitry.

The I/O ports of the AVR are immediately reset to their initial state when a reset source goes

active. This does not require any clock source to be running.

After all reset sources have gone inactive, a delay counter is invoked, stretching the internal

reset. This allows the power to reach a stable level before normal operation starts. The time-out

period of the delay counter is defined by the user through the SUT and CKSEL Fuses. The dif-

ferent selections for the delay period are presented in “Clock Sources” on page 28.

8.0.2

Reset Sources

The ATmega16U4/ATmega32U4 has five sources of reset:

• Power-on Reset. The MCU is reset when the supply voltage is below the Power-on Reset

threshold (V_POT).

• External Reset. The MCU is reset when a low level is present on the RESET pin for longer

than the minimum pulse length.

• Watchdog Reset. The MCU is reset when the Watchdog Timer period expires and the

Watchdog is enabled.

• Brown-out Reset. The MCU is reset when the supply voltage V_CCis below the Brown-out

Reset threshold (V_BOT) and the Brown-out Detector is enabled.

• JTAG AVR Reset. The MCU is reset as long as there is a logic one in the Reset Register, one

of the scan chains of the JTAG system. Refer to the section “IEEE 1149.1 (JTAG) Boundary-

scan” on page 320 for details.

• USB End of Reset. The MCU is reset (excluding the USB controller that remains enabled and

attached) on the detection of a USB End of Reset condition on the bus, if this feature is

enabled by the user.

49

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 8-1. Reset Logic

DATA BUS

MCU Status

Register (MCUSR)

Power-on Reset

Circuit

Brown-out

Reset Circuit

BODLEVEL [2..0]

Pull-up Resistor

SPIKE

FILTER

JTAG Reset

Register

USB Reset

Detection

Watchdog

Oscillator

Delay Counters

Clock

CK

Generator

TIMEOUT

CKSEL[3:0]

SUT[1:0]

Table 8-1.

Reset Characteristics

Symbol Parameter

Condition

Min

Typ

1.4

1.3

Max Units

Power-on Reset Threshold Voltage (rising)

Power-on Reset Threshold Voltage (falling)⁽¹⁾

2.3

V

V_POT

V_CCStart Voltage to ensure internal Power-on

Reset signal

V_POR

-0.1

0.3

+0.1

V

VCC Rise Rate to ensure internal Power_on

Reset signal

V_CCRR

V/ms

0.2

Vcc

0.85

Vcc

V_RST

t_RST

RESET Pin Threshold Voltage

V

Minimum pulse width on RESET Pin

5V, 25°C

400

ns

Notes: 1. The Power-on Reset will not work unless the supply voltage has been below V_POT(falling)

8.0.3

Power-on Reset

A Power-on Reset (POR) pulse is generated by an On-chip detection circuit. The detection level

is defined in Table 8-1. The POR is activated whenever V_CCis below the detection level. The

POR circuit can be used to trigger the start-up Reset, as well as to detect a failure in supply

voltage.

A Power-on Reset (POR) circuit ensures that the device is reset from Power-on. Reaching the

Power-on Reset threshold voltage invokes the delay counter, which determines how long the

50

7766E–AVR–04/10

ATmega16U4/ATmega32U4

device is kept in RESET after V_CCrise. The RESET signal is activated again, without any delay,

when V_CCdecreases below the detection level.

Figure 8-2. MCU Start-up, RESET Tied to V_CC

V_POR

V_POT

V_CC

V_RST

RESET

t_TOUT

TIME-OUT

INTERNAL

RESET

Figure 8-3. MCU Start-up, RESET Extended Externally

V_POR

V_POT

V_CC

V_RST

RESET

t_TOUT

TIME-OUT

INTERNAL

RESET

8.0.4

External Reset

An External Reset is generated by a low level on the RESET pin. Reset pulses longer than the

minimum pulse width (see Table 8-1) will generate a reset, even if the clock is not running.

Shorter pulses are not guaranteed to generate a reset. When the applied signal reaches the

Reset Threshold Voltage – V_RST– on its positive edge, the delay counter starts the MCU after

the Time-out period – t_TOUT– has expired.

Figure 8-4. External Reset During Operation

CC

51

7766E–AVR–04/10

ATmega16U4/ATmega32U4

8.0.5

Brown-out Detection

ATmega16U4/ATmega32U4 has an On-chip Brown-out Detection (BOD) circuit for monitoring

the V_CClevel during operation by comparing it to a fixed trigger level. The trigger level for the

BOD can be selected by the BODLEVEL Fuses. The trigger level has a hysteresis to ensure

spike free Brown-out Detection. The hysteresis on the detection level should be interpreted as

V

_BOT+= V_BOT+ V_HYST/2 and V_BOT-= V_BOT- V_HYST/2.

Table 8-2.

BODLEVEL Fuse Coding

BODLEVEL 2..0 Fuses

Min V_BOT

Typ V_BOT

Max V_BOT

Units

111

110

101

100

011

010

001

000

BOD Disabled

1.8

2.0

2.2

2.4

3.2

3.3

4.0

2.0

2.2

2.4

2.6

3.4

3.5

4.3

2.2

2.4

2.6

2.8

3.6

3.7

4.5

V

Table 8-3.

Symbol

V_HYST

Brown-out Characteristics

Parameter

Min

Typ

50

Max

Units

mV

Brown-out Detector Hysteresis

Min Pulse Width on Brown-out Reset

t_BOD

ns

When the BOD is enabled, and V_CCdecreases to a value below the trigger level (V_BOT-in Figure

8-5), the Brown-out Reset is immediately activated. When V_CCincreases above the trigger level

(V_BOT+in Figure 8-5), the delay counter starts the MCU after the Time-out period t_TOUThas

expired.

The BOD circuit will only detect a drop in V_CCif the voltage stays below the trigger level for lon-

ger than t_BODgiven in Table 8-1.

Figure 8-5. Brown-out Reset During Operation

V_BOT+

V_CC

V_BOT-

RESET

t

TOUT

TIME-OUT

INTERNAL

RESET

52

7766E–AVR–04/10

ATmega16U4/ATmega32U4

8.0.6

Watchdog Reset

When the Watchdog times out, it will generate a short reset pulse of one CK cycle duration. On

the falling edge of this pulse, the delay timer starts counting the Time-out period t_TOUT. Refer to

page 55 for details on operation of the Watchdog Timer.

Figure 8-6. Watchdog Reset During Operation

CC

CK

8.0.7

USB Reset

When the USB controller is enabled and configured with the USB Reset CPU feature enabled

and if a valid USB Reset signalling is detected on the bus, the CPU core is reset but the USB

controller remains enabled and attached. This feature may be used to enhance device reliability.

Figure 8-7. USB Reset During Operation

CC

End of Reset

t_USBRSTMIN

DP

USB Traffic

DM

8.0.8

MCU Status Register – MCUSR

The MCU Status Register provides information on which reset source caused an MCU reset.

Bit

7

–

R

0

6

–

R

0

5

4

3

2

1

0

USBRF

JTRF

R/W

WDRF

R/W

BORF

R/W

EXTRF

R/W

PORF

R/W

MCUSR

Read/Write

Initial Value

R

0

See Bit Description

• Bit 7..6 - Reserved

These bits are reserved and should be read as 0. Do not set these bits.

53

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 5– USBRF: USB Reset Flag

This bit is set if a reset is being caused by a logic one in the JTAG Reset Register selected by

the JTAG instruction AVR_RESET. This bit is reset by a Power-on Reset, or by writing a logic

zero to the flag.

• Bit 4 – JTRF: JTAG Reset Flag

This bit is set if a reset is being caused by a logic one in the JTAG Reset Register selected by

the JTAG instruction AVR_RESET. This bit is reset by a Power-on Reset, or by writing a logic

zero to the flag.

• Bit 3 – WDRF: Watchdog Reset Flag

This bit is set if a Watchdog Reset occurs. The bit is reset by a Power-on Reset, or by writing a

logic zero to the flag.

• Bit 2 – BORF: Brown-out Reset Flag

This bit is set if a Brown-out Reset occurs. The bit is reset by a Power-on Reset, or by writing a

logic zero to the flag.

• Bit 1 – EXTRF: External Reset Flag

This bit is set if an External Reset occurs. The bit is reset by a Power-on Reset, or by writing a

logic zero to the flag.

• Bit 0 – PORF: Power-on Reset Flag

This bit is set if a Power-on Reset occurs. The bit is reset only by writing a logic zero to the flag.

To make use of the Reset Flags to identify a reset condition, the user should read and then

Reset the MCUSR as early as possible in the program. If the register is cleared before another

reset occurs, the source of the reset can be found by examining the Reset Flags.

8.1

Internal Voltage Reference

ATmega16U4/ATmega32U4 features an internal bandgap reference. This reference is used for

Brown-out Detection, and it can be used as an input to the Analog Comparator or the ADC.

8.1.1

Voltage Reference Enable Signals and Start-up Time

The voltage reference has a start-up time that may influence the way it should be used. The

start-up time is given in Table 8-4. To save power, the reference is not always turned on. The

reference is on during the following situations:

1. When the BOD is enabled (by programming the BODLEVEL [2..0] Fuse).

2. When the bandgap reference is connected to the Analog Comparator (by setting the

ACBG bit in ACSR).

3. When the ADC is enabled.

Thus, when the BOD is not enabled, after setting the ACBG bit or enabling the ADC, the user

must always allow the reference to start up before the output from the Analog Comparator or

54

7766E–AVR–04/10

ATmega16U4/ATmega32U4

ADC is used. To reduce power consumption in Power-down mode, the user can avoid the three

conditions above to ensure that the reference is turned off before entering Power-down mode.

Table 8-4.

Symbol Parameter

V_BGBandgap reference voltage

t_BG

Internal Voltage Reference Characteristics⁽¹⁾

Condition Min

Typ

1.1

40

Max Units

TBD

70

V

Bandgap reference start-up time

µs

Bandgap reference current

consumption

I_BG

TBD

10

TBD

µA

Note:

1. Values are guidelines only. Actual values are TBD.

8.2

Watchdog Timer

ATmega16U4/ATmega32U4 has an Enhanced Watchdog Timer (WDT). The main features are:

• Clocked from separate On-chip Oscillator

• 3 Operating modes

– Interrupt

– System Reset

– Interrupt and System Reset

• Selectable Time-out period from 16ms to 8s

• Possible Hardware fuse Watchdog always on (WDTON) for fail-safe mode

Figure 8-8. Watchdog Timer

128kHz

OSCILLATOR

WDP0

WDP1

WATCHDOG

WDP2

RESET

WDP3

WDE

MCU RESET

WDIF

INTERRUPT

WDIE

The Watchdog Timer (WDT) is a timer counting cycles of a separate on-chip 128 kHz oscillator.

The WDT gives an interrupt or a system reset when the counter reaches a given time-out value.

In normal operation mode, it is required that the system uses the WDR - Watchdog Timer Reset

- instruction to restart the counter before the time-out value is reached. If the system doesn't

restart the counter, an interrupt or system reset will be issued.

In Interrupt mode, the WDT gives an interrupt when the timer expires. This interrupt can be used

to wake the device from sleep-modes, and also as a general system timer. One example is to

limit the maximum time allowed for certain operations, giving an interrupt when the operation

has run longer than expected. In System Reset mode, the WDT gives a reset when the timer

55

7766E–AVR–04/10

ATmega16U4/ATmega32U4

expires. This is typically used to prevent system hang-up in case of runaway code. The third

mode, Interrupt and System Reset mode, combines the other two modes by first giving an inter-

rupt and then switch to System Reset mode. This mode will for instance allow a safe shutdown

by saving critical parameters before a system reset.

The Watchdog always on (WDTON) fuse, if programmed, will force the Watchdog Timer to Sys-

tem Reset mode. With the fuse programmed the System Reset mode bit (WDE) and Interrupt

mode bit (WDIE) are locked to 1 and 0 respectively. To further ensure program security, altera-

tions to the Watchdog set-up must follow timed sequences. The sequence for clearing WDE and

changing time-out configuration is as follows:

1. In the same operation, write a logic one to the Watchdog change enable bit (WDCE)

and WDE. A logic one must be written to WDE regardless of the previous value of the

WDE bit.

2. Within the next four clock cycles, write the WDE and Watchdog prescaler bits (WDP) as

desired, but with the WDCE bit cleared. This must be done in one operation.

The following code example shows one assembly and one C function for turning off the Watch-

dog Timer. The example assumes that interrupts are controlled (e.g. by disabling interrupts

globally) so that no interrupts will occur during the execution of these functions.

56

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Assembly Code Example⁽¹⁾

WDT_off:

; Turn off global interrupt

cli

; Reset Watchdog Timer

wdr

; Clear WDRF in MCUSR

in

andi r16, (0xff & (0<<WDRF))

out MCUSR, r16

r16, MCUSR

; Write logical one to WDCE and WDE

; Keep old prescaler setting to prevent unintentional time-out

in

r16, WDTCSR

ori

out

r16, (1<<WDCE) | (1<<WDE)

WDTCSR, r16

; Turn off WDT

ldi

out

r16, (0<<WDE)

WDTCSR, r16

; Turn on global interrupt

sei

ret

C Code Example⁽¹⁾

void WDT_off(void)

{

__disable_interrupt();

__watchdog_reset();

/* Clear WDRF in MCUSR */

MCUSR &= ~(1<<WDRF);

/* Write logical one to WDCE and WDE */

/* Keep old prescaler setting to prevent unintentional time-out

*/

WDTCSR |= (1<<WDCE) | (1<<WDE);

/* Turn off WDT */

WDTCSR = 0x00;

__enable_interrupt();

}

Note:

1. The example code assumes that the part specific header file is included.

Note: If the Watchdog is accidentally enabled, for example by a runaway pointer or brown-out

condition, the device will be reset and the Watchdog Timer will stay enabled. If the code is not

set up to handle the Watchdog, this might lead to an eternal loop of time-out resets. To avoid this

situation, the application software should always clear the Watchdog System Reset Flag

(WDRF) and the WDE control bit in the initialization routine, even if the Watchdog is not in use.

57

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The following code example shows one assembly and one C function for changing the time-out

value of the Watchdog Timer.

Assembly Code Example⁽¹⁾

WDT_Prescaler_Change:

; Turn off global interrupt

cli

; Reset Watchdog Timer

wdr

; Start timed sequence

in

r16, WDTCSR

ori

out

r16, (1<<WDCE) | (1<<WDE)

WDTCSR, r16

; -- Got four cycles to set the new values from here -

; Set new prescaler(time-out) value = 64K cycles (~0.5 s)

ldi

out

r16, (1<<WDE) | (1<<WDP2) | (1<<WDP0)

WDTCSR, r16

; -- Finished setting new values, used 2 cycles -

; Turn on global interrupt

sei

ret

C Code Example⁽¹⁾

void WDT_Prescaler_Change(void)

{

__disable_interrupt();

__watchdog_reset();

/* Start timed sequence */

WDTCSR |= (1<<WDCE) | (1<<WDE);

/* Set new prescaler(time-out) value = 64K cycles (~0.5 s) */

WDTCSR = (1<<WDE) | (1<<WDP2) | (1<<WDP0);

__enable_interrupt();

}

Note:

1. The example code assumes that the part specific header file is included.

Note: The Watchdog Timer should be reset before any change of the WDP bits, since a change

in the WDP bits can result in a time-out when switching to a shorter time-out period.

8.2.1

Watchdog Timer Control Register - WDTCSR

Bit

7

6

5

4

3

2

1

0

WDIF

WDIE

WDP3

R/W

0

WDCE

R/W

0

WDE

R/W

X

WDP2

R/W

0

WDP1

R/W

0

WDP0

R/W

0

WDTCSR

Read/Write

Initial Value

R/W

0

R/W

0

• Bit 7 - WDIF: Watchdog Interrupt Flag

This bit is set when a time-out occurs in the Watchdog Timer and the Watchdog Timer is config-

ured for interrupt. WDIF is cleared by hardware when executing the corresponding interrupt

58

7766E–AVR–04/10

ATmega16U4/ATmega32U4

handling vector. Alternatively, WDIF is cleared by writing a logic one to the flag. When the I-bit in

SREG and WDIE are set, the Watchdog Time-out Interrupt is executed.

• Bit 6 - WDIE: Watchdog Interrupt Enable

When this bit is written to one and the I-bit in the Status Register is set, the Watchdog Interrupt is

enabled. If WDE is cleared in combination with this setting, the Watchdog Timer is in Interrupt

Mode, and the corresponding interrupt is executed if time-out in the Watchdog Timer occurs.

If WDE is set, the Watchdog Timer is in Interrupt and System Reset Mode. The first time-out in

the Watchdog Timer will set WDIF. Executing the corresponding interrupt vector will clear WDIE

and WDIF automatically by hardware (the Watchdog goes to System Reset Mode). This is use-

ful for keeping the Watchdog Timer security while using the interrupt. To stay in Interrupt and

System Reset Mode, WDIE must be set after each interrupt. This should however not be done

within the interrupt service routine itself, as this might compromise the safety-function of the

Watchdog System Reset mode. If the interrupt is not executed before the next time-out, a Sys-

tem Reset will be applied.

Table 8-5.

Watchdog Timer Configuration

WDTON

WDE

WDIE

Mode

Action on Time-out

None

0

1

0

1

0

Stopped

Interrupt Mode

System Reset Mode

Interrupt

Reset

Interrupt and System

Reset Mode

Interrupt, then go to

System Reset Mode

0

1

x

1

x

System Reset Mode

Reset

• Bit 4 - WDCE: Watchdog Change Enable

This bit is used in timed sequences for changing WDE and prescaler bits. To clear the WDE bit,

and/or change the prescaler bits, WDCE must be set.

Once written to one, hardware will clear WDCE after four clock cycles.

• Bit 3 - WDE: Watchdog System Reset Enable

WDE is overridden by WDRF in MCUSR. This means that WDE is always set when WDRF is

set. To clear WDE, WDRF must be cleared first. This feature ensures multiple resets during con-

ditions causing failure, and a safe start-up after the failure.

• Bit 5, 2..0 - WDP3..0: Watchdog Timer Prescaler 3, 2, 1 and 0

The WDP3..0 bits determine the Watchdog Timer prescaling when the Watchdog Timer is run-

ning. The different prescaling values and their corresponding time-out periods are shown in

Table 8-6 on page 60.

59

7766E–AVR–04/10

ATmega16U4/ATmega32U4

.

Table 8-6.

Watchdog Timer Prescale Select

Number of WDT Oscillator

Cycles

Typical Time-out at

V_CC= 5.0V

WDP3 WDP2 WDP1 WDP0

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

2K (2048) cycles

4K (4096) cycles

16 ms

32 ms

64 ms

0.125 s

0.25 s

0.5 s

8K (8192) cycles

16K (16384) cycles

32K (32768) cycles

64K (65536) cycles

128K (131072) cycles

256K (262144) cycles

512K (524288) cycles

1024K (1048576) cycles

1.0 s

2.0 s

4.0 s

8.0 s

Reserved

60

7766E–AVR–04/10

ATmega16U4/ATmega32U4

9. Interrupts

This section describes the specifics of the interrupt handling as performed in

ATmega16U4/ATmega32U4. For a general explanation of the AVR interrupt handling, refer to

“Reset and Interrupt Handling” on page 15.

9.1

Interrupt Vectors in ATmega16U4/ATmega32U4

Table 9-1.

Reset and Interrupt Vectors

Program

Vector

No.

Address⁽²⁾Source

Interrupt Definition

External Pin, Power-on Reset, Brown-out Reset,

Watchdog Reset, and JTAG AVR Reset

1

$0000⁽¹⁾

RESET

2

$0002

$0004

$0006

$0008

$000A

$000C

$000E

$0010

$0012

$0014

$0016

$0018

$001A

$001C

$001E

$0020

$0022

$0024

$0026

$0028

$002A

$002C

$002E

$0030

$0032

$0034

$0036

$0038

INT0

External Interrupt Request 0

External Interrupt Request 1

External Interrupt Request 2

External Interrupt Request 3

Reserved

3

INT1

4

INT2

5

INT3

6

Reserved

INT6

7

Reserved

8

External Interrupt Request 6

Reserved

9

Reserved

PCINT0

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

Pin Change Interrupt Request 0

USB General Interrupt request

USB Endpoint Interrupt request

Watchdog Time-out Interrupt

Reserved

USB General

USB Endpoint

WDT

Reserved

TIMER1 CAPT

TIMER1 COMPA

Reserved

Timer/Counter1 Capture Event

Timer/Counter1 Compare Match A

TIMER1 COMPB Timer/Counter1 Compare Match B

TIMER1 COMPC Timer/Counter1 Compare Match C

TIMER1 OVF

Timer/Counter1 Overflow

TIMER0 COMPA

Timer/Counter0 Compare Match A

TIMER0 COMPB Timer/Counter0 Compare match B

TIMER0 OVF

SPI (STC)

Timer/Counter0 Overflow

SPI Serial Transfer Complete

USART1 Rx Complete

USART1 RX

USART1 UDRE

USART1TX

USART1 Data Register Empty

USART1 Tx Complete

ANALOG COMP

Analog Comparator

61

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 9-1.

Reset and Interrupt Vectors (Continued)

Program

Vector

No.

Address⁽²⁾Source

Interrupt Definition

30

31

32

33

34

35

36

37

38

39

40

41

42

43

$003A

$003C

$003E

$0040

$0042

$0044

$0046

$0048

$004A

$004C

$004E

$0050

$0052

$0054

ADC

ADC Conversion Complete

EEPROM Ready

EE READY

TIMER3 CAPT

TIMER3 COMPA

Timer/Counter3 Capture Event

Timer/Counter3 Compare Match A

TIMER3 COMPB Timer/Counter3 Compare Match B

TIMER3 COMPC Timer/Counter3 Compare Match C

TIMER3 OVF

TWI

Timer/Counter3 Overflow

2-wire Serial Interface

SPM READY

TIMER4 COMPA

Store Program Memory Ready

Timer/Counter4 Compare Match A

TIMER4 COMPB Timer/Counter4 Compare Match B

TIMER4 COMPD Timer/Counter4 Compare Match D

TIMER4 OVF

TIMER4 FPF

Timer/Counter4 Overflow

Timer/Counter4 Fault Protection Interrupt

Notes: 1. When the BOOTRST Fuse is programmed, the device will jump to the Boot Loader address at

reset, see “Memory Programming” on page 346.

2. When the IVSEL bit in MCUCR is set, Interrupt Vectors will be moved to the start of the Boot

Flash Section. The address of each Interrupt Vector will then be the address in this table

added to the start address of the Boot Flash Section.

Table 9-2 shows reset and Interrupt Vectors placement for the various combinations of

BOOTRST and IVSEL settings. If the program never enables an interrupt source, the Interrupt

Vectors are not used, and regular program code can be placed at these locations. This is also

the case if the Reset Vector is in the Application section while the Interrupt Vectors are in the

Boot section or vice versa.

Table 9-2.

Reset and Interrupt Vectors Placement⁽¹⁾

BOOTRST

IVSEL

Reset Address

0x0000

Interrupt Vectors Start Address

0x0002

1

0

1

0

1

0x0000

Boot Reset Address + 0x0002

0x0002

Boot Reset Address

Boot Reset Address + 0x0002

Note:

1. The Boot Reset Address is shown in Table 27-8 on page 344. For the BOOTRST Fuse “1”

means unprogrammed while “0” means programmed.

9.1.1

Moving Interrupts Between Application and Boot Space

The General Interrupt Control Register controls the placement of the Interrupt Vector table.

62

7766E–AVR–04/10

ATmega16U4/ATmega32U4

9.1.2

MCU Control Register – MCUCR

Bit

7

6

–

R

0

5

–

R

0

4

3

–

R

0

2

–

R

0

1

0

JTD

R/W

0

PUD

R/W

0

IVSEL

R/W

0

IVCE

R/W

0

MCUCR

Read/Write

Initial Value

• Bit 1 – IVSEL: Interrupt Vector Select

When the IVSEL bit is cleared (zero), the Interrupt Vectors are placed at the start of the Flash

memory. When this bit is set (one), the Interrupt Vectors are moved to the beginning of the Boot

Loader section of the Flash. The actual address of the start of the Boot Flash Section is deter-

mined by the BOOTSZ Fuses. Refer to the section “Memory Programming” on page 346 for

details. To avoid unintentional changes of Interrupt Vector tables, a special write procedure must

be followed to change the IVSEL bit:

a. Write the Interrupt Vector Change Enable (IVCE) bit to one.

b. Within four cycles, write the desired value to IVSEL while writing a zero to IVCE.

Interrupts will automatically be disabled while this sequence is executed. Interrupts are disabled

in the cycle IVCE is set, and they remain disabled until after the instruction following the write to

IVSEL. If IVSEL is not written, interrupts remain disabled for four cycles. The I-bit in the Status

Register is unaffected by the automatic disabling.

Note:

If Interrupt Vectors are placed in the Boot Loader section and Boot Lock bit BLB02 is programmed,

interrupts are disabled while executing from the Application section. If Interrupt Vectors are placed

in the Application section and Boot Lock bit BLB12 is programed, interrupts are disabled while

executing from the Boot Loader section. Refer to the section “Memory Programming” on page 346

for details on Boot Lock bits.

• Bit 0 – IVCE: Interrupt Vector Change Enable

The IVCE bit must be written to logic one to enable change of the IVSEL bit. IVCE is cleared by

hardware four cycles after it is written or when IVSEL is written. Setting the IVCE bit will disable

interrupts, as explained in the IVSEL description above. See Code Example below.

63

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Assembly Code Example

Move_interrupts:

; Enable change of Interrupt Vectors

ldi r16, (1<<IVCE)

out MCUCR, r16

; Move interrupts to Boot Flash section

ldi r16, (1<<IVSEL)

out MCUCR, r16

ret

C Code Example

void Move_interrupts(void)

{

/* Enable change of Interrupt Vectors */

MCUCR = (1<<IVCE);

/* Move interrupts to Boot Flash section */

MCUCR = (1<<IVSEL);

}

64

7766E–AVR–04/10

ATmega16U4/ATmega32U4

10. I/O-Ports

10.1 Introduction

All AVR ports have true Read-Modify-Write functionality when used as general digital I/O ports.

This means that the direction of one port pin can be changed without unintentionally changing

the direction of any other pin with the SBI and CBI instructions. The same applies when chang-

ing drive value (if configured as output) or enabling/disabling of pull-up resistors (if configured as

input). Each output buffer has symmetrical drive characteristics with both high sink and source

capability. The pin driver is strong enough to drive LED displays directly. All port pins have indi-

vidually selectable pull-up resistors with a supply-voltage invariant resistance. All I/O pins have

protection diodes to both V_CCand Ground as indicated in Figure 10-1. Refer to “Electrical Char-

acteristics” on page 378 for a complete list of parameters.

Figure 10-1. I/O Pin Equivalent Schematic

All registers and bit references in this section are written in general form. A lower case “x” repre-

sents the numbering letter for the port, and a lower case “n” represents the bit number. However,

when using the register or bit defines in a program, the precise form must be used. For example,

PORTB3 for bit no. 3 in Port B, here documented generally as PORTxn. The physical I/O Regis-

ters and bit locations are listed in “Register Description for I/O-Ports” on page 82.

Three I/O memory address locations are allocated for each port, one each for the Data Register

– PORTx, Data Direction Register – DDRx, and the Port Input Pins – PINx. The Port Input Pins

I/O location is read only, while the Data Register and the Data Direction Register are read/write.

However, writing a logic one to a bit in the PINx Register, will result in a toggle in the correspond-

ing bit in the Data Register. In addition, the Pull-up Disable – PUD bit in MCUCR disables the

pull-up function for all pins in all ports when set.

Using the I/O port as General Digital I/O is described in “Ports as General Digital I/O” on page

66. Most port pins are multiplexed with alternate functions for the peripheral features on the

device. How each alternate function interferes with the port pin is described in “Alternate Port

Functions” on page 70. Refer to the individual module sections for a full description of the alter-

nate functions.

65

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Note that enabling the alternate function of some of the port pins does not affect the use of the

other pins in the port as general digital I/O.

10.2 Ports as General Digital I/O

The ports are bi-directional I/O ports with optional internal pull-ups. Figure 10-2 shows a func-

tional description of one I/O-port pin, here generically called Pxn.

Figure 10-2. General Digital I/O⁽¹⁾

PUD

Q

D

DDxn

Q _CLR

WDx

RDx

RESET

1

0

Q

D

Pxn

PORTxn

Q _CLR

RESET

WPx

WRx

SLEEP

RRx

SYNCHRONIZER

RPx

D

Q

D

L

Q

PINxn

Q

clk _I/O

WDx:

RDx:

WRx:

RRx:

RPx:

WPx:

WRITE DDRx

READ DDRx

WRITE PORTx

PUD:

SLEEP:

clk_I/O

PULLUP DISABLE

SLEEP CONTROL

I/O CLOCK

:

READ PORTx REGISTER

READ PORTx PIN

WRITE PINx REGISTER

Note:

1. WRx, WPx, WDx, RRx, RPx, and RDx are common to all pins within the same port. clk_I/O

,

SLEEP, and PUD are common to all ports.

10.2.1

Configuring the Pin

Each port pin consists of three register bits: DDxn, PORTxn, and PINxn. As shown in “Register

Description for I/O-Ports” on page 82, the DDxn bits are accessed at the DDRx I/O address, the

PORTxn bits at the PORTx I/O address, and the PINxn bits at the PINx I/O address.

The DDxn bit in the DDRx Register selects the direction of this pin. If DDxn is written logic one,

Pxn is configured as an output pin. If DDxn is written logic zero, Pxn is configured as an input

pin.

If PORTxn is written logic one when the pin is configured as an input pin, the pull-up resistor is

activated. To switch the pull-up resistor off, PORTxn has to be written logic zero or the pin has to

be configured as an output pin. The port pins are tri-stated when reset condition becomes active,

even if no clocks are running.

66

7766E–AVR–04/10

ATmega16U4/ATmega32U4

If PORTxn is written logic one when the pin is configured as an output pin, the port pin is driven

high (one). If PORTxn is written logic zero when the pin is configured as an output pin, the port

pin is driven low (zero).

10.2.2

10.2.3

Toggling the Pin

Writing a logic one to PINxn toggles the value of PORTxn, independent on the value of DDRxn.

Note that the SBI instruction can be used to toggle one single bit in a port.

Switching Between Input and Output

When switching between tri-state ({DDxn, PORTxn} = 0b00) and output high ({DDxn, PORTxn}

= 0b11), an intermediate state with either pull-up enabled {DDxn, PORTxn} = 0b01) or output

low ({DDxn, PORTxn} = 0b10) occurs. Normally, the pull-up enabled state is fully acceptable, as

a high-impedance environment will not notice the difference between a strong high driver and a

pull-up. If this is not the case, the PUD bit in the MCUCR Register can be set to disable all pull-

ups in all ports.

Switching between input with pull-up and output low generates the same problem. The user

must use either the tri-state ({DDxn, PORTxn} = 0b00) or the output high state ({DDxn, PORTxn}

= 0b11) as an intermediate step.

Table 10-1 summarizes the control signals for the pin value.

Table 10-1. Port Pin Configurations

PUD

DDxn PORTxn (in MCUCR)

I/O

Pull-up Comment

0

1

X

0

Input

No

Tri-state (Hi-Z)

Pxn will source current if ext. pulled

low.

Input

Yes

0

1

0

1

X

Input

Output

No

Tri-state (Hi-Z)

Output Low (Sink)

Output High (Source)

10.2.4

Reading the Pin Value

Independent of the setting of Data Direction bit DDxn, the port pin can be read through the

PINxn Register bit. As shown in Figure 10-2, the PINxn Register bit and the preceding latch con-

stitute a synchronizer. This is needed to avoid metastability if the physical pin changes value

near the edge of the internal clock, but it also introduces a delay. Figure 10-3 shows a timing dia-

gram of the synchronization when reading an externally applied pin value. The maximum and

minimum propagation delays are denoted t_pd,maxand t_pd,minrespectively.

67

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 10-3. Synchronization when Reading an Externally Applied Pin value

SYSTEM CLK

INSTRUCTIONS

SYNC LATCH

PINxn

XXX

in r17, PINx

0x00

t_{pd, max}

0xFF

r17

t_{pd, min}

Consider the clock period starting shortly after the first falling edge of the system clock. The latch

is closed when the clock is low, and goes transparent when the clock is high, as indicated by the

shaded region of the “SYNC LATCH” signal. The signal value is latched when the system clock

goes low. It is clocked into the PINxn Register at the succeeding positive clock edge. As indi-

cated by the two arrows tpd max and tpd min, a single signal transition on the pin will be delayed

between ½ and 1½ system clock period depending upon the time of assertion.

When reading back a software assigned pin value, a nop instruction must be inserted as indi-

cated in Figure 10-4. The out instruction sets the “SYNC LATCH” signal at the positive edge of

the clock. In this case, the delay tpd through the synchronizer is 1 system clock period.

Figure 10-4. Synchronization when Reading a Software Assigned Pin Value

SYSTEM CLK

0xFF

r16

out PORTx, r16

nop

in r17, PINx

INSTRUCTIONS

SYNC LATCH

PINxn

0x00

t_pd

0xFF

r17

The following code example shows how to set port B pins 0 and 1 high, 2 and 3 low, and define

the port pins from 4 to 7 as input with pull-ups assigned to port pins 6 and 7. The resulting pin

values are read back again, but as previously discussed, a nop instruction is included to be able

to read back the value recently assigned to some of the pins.

68

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Assembly Code Example⁽¹⁾

...

; Define pull-ups and set outputs high

; Define directions for port pins

ldi r16,(1<<PB7)|(1<<PB6)|(1<<PB1)|(1<<PB0)

ldi r17,(1<<DDB3)|(1<<DDB2)|(1<<DDB1)|(1<<DDB0)

out PORTB,r16

out DDRB,r17

; Insert nop for synchronization

nop

; Read port pins

in

r16,PINB

...

C Code Example

unsigned char i;

...

/* Define pull-ups and set outputs high */

/* Define directions for port pins */

PORTB = (1<<PB7)|(1<<PB6)|(1<<PB1)|(1<<PB0);

DDRB = (1<<DDB3)|(1<<DDB2)|(1<<DDB1)|(1<<DDB0);

/* Insert nop for synchronization*/

__no_operation();

/* Read port pins */

i = PINB;

...

Note:

1. For the assembly program, two temporary registers are used to minimize the time from pull-

ups are set on pins 0, 1, 6, and 7, until the direction bits are correctly set, defining bit 2 and 3

as low and redefining bits 0 and 1 as strong high drivers.

10.2.5

Digital Input Enable and Sleep Modes

As shown in Figure 10-2, the digital input signal can be clamped to ground at the input of the

Schmidt-trigger. The signal denoted SLEEP in the figure, is set by the MCU Sleep Controller in

Power-down mode, Power-save mode, and Standby mode to avoid high power consumption if

some input signals are left floating, or have an analog signal level close to V_CC/2.

SLEEP is overridden for port pins enabled as external interrupt pins. If the external interrupt

request is not enabled, SLEEP is active also for these pins. SLEEP is also overridden by various

other alternate functions as described in “Alternate Port Functions” on page 70.

If a logic high level (“one”) is present on an asynchronous external interrupt pin configured as

“Interrupt on Rising Edge, Falling Edge, or Any Logic Change on Pin” while the external interrupt

is not enabled, the corresponding External Interrupt Flag will be set when resuming from the

above mentioned Sleep mode, as the clamping in these sleep mode produces the requested

logic change.

69

7766E–AVR–04/10

ATmega16U4/ATmega32U4

10.2.6

Unconnected Pins

If some pins are unused, it is recommended to ensure that these pins have a defined level. Even

though most of the digital inputs are disabled in the deep sleep modes as described above, float-

ing inputs should be avoided to reduce current consumption in all other modes where the digital

inputs are enabled (Reset, Active mode and Idle mode).

The simplest method to ensure a defined level of an unused pin, is to enable the internal pull-up.

In this case, the pull-up will be disabled during reset. If low power consumption during reset is

important, it is recommended to use an external pull-up or pull-down. Connecting unused pins

directly to V_CCor GND is not recommended, since this may cause excessive currents if the pin is

accidentally configured as an output.

10.3 Alternate Port Functions

Most port pins have alternate functions in addition to being general digital I/Os. Figure 10-5

shows how the port pin control signals from the simplified Figure 10-2 can be overridden by

alternate functions. The overriding signals may not be present in all port pins, but the figure

serves as a generic description applicable to all port pins in the AVR microcontroller family.

Figure 10-5. Alternate Port Functions⁽¹⁾

PUOExn

PUOVxn

1

PUD

0

DDOExn

DDOVxn

1

Q

D

0

DDxn

Q _CLR

WDx

RDx

PVOExn

PVOVxn

RESET

1

0

1

0

Pxn

Q

D

PORTxn

PTOExn

Q _CLR

DIEOExn

DIEOVxn

SLEEP

WPx

RESET

WRx

1

0

RRx

SYNCHRONIZER

RPx

SET

D

Q

D

L

Q

PINxn

_CLRQ

CLR

clk _I/O

DIxn

AIOxn

PUOExn: Pxn PULL-UP OVERRIDE ENABLE

PUOVxn: Pxn PULL-UP OVERRIDE VALUE

DDOExn: Pxn DATA DIRECTION OVERRIDE ENABLE

DDOVxn: Pxn DATA DIRECTION OVERRIDE VALUE

PVOExn: Pxn PORT VALUE OVERRIDE ENABLE

PVOVxn: Pxn PORT VALUE OVERRIDE VALUE

PUD:

WDx:

RDx:

RRx:

WRx:

RPx:

WPx:

PULLUP DISABLE

WRITE DDRx

READ DDRx

READ PORTx REGISTER

WRITE PORTx

READ PORTx PIN

WRITE PINx

DIEOExn: Pxn DIGITAL INPUT-ENABLE OVERRIDE ENABLE

DIEOVxn: Pxn DIGITAL INPUT-ENABLE OVERRIDE VALUE

clk_I/O

:

I/O CLOCK

SLEEP:

SLEEP CONTROL

DIxn:

DIGITAL INPUT PIN n ON PORTx

AIOxn:

ANALOG INPUT/OUTPUT PIN n ON PORTx

PTOExn: Pxn, PORT TOGGLE OVERRIDE ENABLE

70

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Note:

1. WRx, WPx, WDx, RRx, RPx, and RDx are common to all pins within the same port. clk_I/O

,

SLEEP, and PUD are common to all ports. All other signals are unique for each pin.

Table 10-2 summarizes the function of the overriding signals. The pin and port indexes from Fig-

ure 10-5 are not shown in the succeeding tables. The overriding signals are generated internally

in the modules having the alternate function.

Table 10-2. Generic Description of Overriding Signals for Alternate Functions

Signal Name Full Name

Description

If this signal is set, the pull-up enable is controlled by the

PUOV signal. If this signal is cleared, the pull-up is

enabled when {DDxn, PORTxn, PUD} = 0b010.

Pull-up Override

Enable

PUOE

PUOV

DDOE

DDOV

If PUOE is set, the pull-up is enabled/disabled when

PUOV is set/cleared, regardless of the setting of the

DDxn, PORTxn, and PUD Register bits.

Pull-up Override

Value

If this signal is set, the Output Driver Enable is controlled

by the DDOV signal. If this signal is cleared, the Output

driver is enabled by the DDxn Register bit.

Data Direction

Override Enable

If DDOE is set, the Output Driver is enabled/disabled

when DDOV is set/cleared, regardless of the setting of

the DDxn Register bit.

Data Direction

Override Value

If this signal is set and the Output Driver is enabled, the

port value is controlled by the PVOV signal. If PVOE is

cleared, and the Output Driver is enabled, the port Value

is controlled by the PORTxn Register bit.

Port Value

Override Enable

PVOE

Port Value

Override Value

If PVOE is set, the port value is set to PVOV, regardless

of the setting of the PORTxn Register bit.

PVOV

PTOE

Port Toggle

Override Enable

If PTOE is set, the PORTxn Register bit is inverted.

If this bit is set, the Digital Input Enable is controlled by

the DIEOV signal. If this signal is cleared, the Digital Input

Enable is determined by MCU state (Normal mode, sleep

mode).

Digital Input

Enable Override

Enable

DIEOE

DIEOV

Digital Input

Enable Override

Value

If DIEOE is set, the Digital Input is enabled/disabled when

DIEOV is set/cleared, regardless of the MCU state

(Normal mode, sleep mode).

This is the Digital Input to alternate functions. In the

figure, the signal is connected to the output of the schmitt

trigger but before the synchronizer. Unless the Digital

Input is used as a clock source, the module with the

alternate function will use its own synchronizer.

DI

Digital Input

This is the Analog Input/output to/from alternate

functions. The signal is connected directly to the pad, and

can be used bi-directionally.

Analog

Input/Output

AIO

The following subsections shortly describe the alternate functions for each port, and relate the

overriding signals to the alternate function. Refer to the alternate function description for further

details.

71

7766E–AVR–04/10

ATmega16U4/ATmega32U4

10.3.1

MCU Control Register – MCUCR

Bit

7

6

–

R

0

5

–

R

0

4

3

–

R

0

2

–

R

0

1

0

JTD

R/W

0

PUD

R/W

0

IVSEL

R/W

0

IVCE

R/W

0

MCUCR

Read/Write

Initial Value

• Bit 4 – PUD: Pull-up Disable

When this bit is written to one, the pull-ups in the I/O ports are disabled even if the DDxn and

PORTxn Registers are configured to enable the pull-ups ({DDxn, PORTxn} = 0b01). See “Con-

figuring the Pin” on page 66 for more details about this feature.

10.3.2

Alternate Functions of Port B

The Port B pins with alternate functions are shown in Table 10-3.

Table 10-3. Port B Pins Alternate Functions

Port Pin Alternate Functions

OC0A/OC1C/PCINT7/RTS (Output Compare and PWM Output A for

PB7

PB6

PB5

Timer/Counter0, Output Compare and PWM Output C for Timer/Counter1 or Pin

Change Interrupt 7 or UART flow control RTS signal)

OC1B/PCINT6/OC.4B/ADC13 (Output Compare and PWM Output B for

Timer/Counter1 or Pin Change Interrupt 6 or Timer 4 Output Compare B / PWM

output or Analog to Digital Converter channel 13)

OC1A/PCINT5/OC.4B/ADC12 (Output Compare and PWM Output A for

Timer/Counter1 or Pin Change Interrupt 5 or Timer 4 Complementary Output

Compare B / PWM output or Analog to Digital Converter channel 12)

PB4

PB3

PCINT4/ADC11 (Pin Change Interrupt 4 or Analog to Digital Converter channel 11)

PDO/MISO/PCINT3 (Programming Data Output or SPI Bus Master Input/Slave

Output or Pin Change Interrupt 3)

PDI/MOSI/PCINT2 (Programming Data Input or SPI Bus Master Output/Slave Input

or Pin Change Interrupt 2)

PB2

PB1

PB0

SCK/PCINT1 (SPI Bus Serial Clock or Pin Change Interrupt 1)

SS/PCINT0 (SPI Slave Select input or Pin Change Interrupt 0)

The alternate pin configuration is as follows:

• OC0A/OC1C/PCINT7/RTS, Bit 7

OC0A, Output Compare Match A output: The PB7 pin can serve as an external output for the

Timer/Counter0 Output Compare. The pin has to be configured as an output (DDB7 set “one”) to

serve this function. The OC0A pin is also the output pin for the PWM mode timer function.

OC1C, Output Compare Match C output: The PB7 pin can serve as an external output for the

Timer/Counter1 Output Compare C. The pin has to be configured as an output (DDB7 set “one”)

to serve this function. The OC1C pin is also the output pin for the PWM mode timer function.

PCINT7, Pin Change Interrupt source 7: The PB7 pin can serve as an external interrupt source.

RTS: RTS flow control signal used by enhanced UART.

72

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• OC1B/PCINT6/OC.4B/ADC12, Bit 6

OC1B, Output Compare Match B output: The PB6 pin can serve as an external output for the

Timer/Counter1 Output Compare B. The pin has to be configured as an output (DDB6 set “one”)

to serve this function. The OC1B pin is also the output pin for the PWM mode timer function.

PCINT6, Pin Change Interrupt source 6: The PB7 pin can serve as an external interrupt source.

OC.4B: Timer 4 Output Compare B. This pin can be used to generate a high-speed PWM signal

from Timer 4 module. The pin has to be configured as an output (DDB6 set “one”) to serve this

function.

ADC13: Analog to Digital Converter, channel 13.

• OC1A/PCINT5/OC.4B/ADC12, Bit 5

OC1A, Output Compare Match A output: The PB5 pin can serve as an external output for the

Timer/Counter1 Output Compare A. The pin has to be configured as an output (DDB5 set (one))

to serve this function. The OC1A pin is also the output pin for the PWM mode timer function.

PCINT5, Pin Change Interrupt source 5: The PB7 pin can serve as an external interrupt source.

OC.4B: Timer 4 Output Compare B. This pin can be used to generate a high-speed PWM signal

from Timer 4 module, complementary to OC.4B (PB5) signal. The pin has to be configured as an

output (DDB5 set (one)) to serve this function.

ADC12: Analog to Digital Converter, channel 12.

• PCINT4/ADC11, Bit 4

PCINT4, Pin Change Interrupt source 4: The PB7 pin can serve as an external interrupt source.

ADC11, Analog to Digital Converter channel 11.

• PDO/MISO/PCINT3 – Port B, Bit 3

PDO, SPI Serial Programming Data Output. During Serial Program Downloading, this pin is

used as data output line for the ATmega16U4/ATmega32U4.

MISO: Master Data input, Slave Data output pin for SPI channel. When the SPI is enabled as a

master, this pin is configured as an input regardless of the setting of DDB3. When the SPI is

enabled as a slave, the data direction of this pin is controlled by DDB3. When the pin is forced to

be an input, the pull-up can still be controlled by the PORTB3 bit.

PCINT3, Pin Change Interrupt source 3: The PB7 pin can serve as an external interrupt source.

• PDI/MOSI/PCINT2 – Port B, Bit 2

PDI, SPI Serial Programming Data Input. During Serial Program Downloading, this pin is used

as data input line for the ATmega16U4/ATmega32U4.

MOSI: SPI Master Data output, Slave Data input for SPI channel. When the SPI is enabled as a

slave, this pin is configured as an input regardless of the setting of DDB2. When the SPI is

enabled as a master, the data direction of this pin is controlled by DDB2. When the pin is forced

to be an input, the pull-up can still be controlled by the PORTB2 bit.

PCINT2, Pin Change Interrupt source 2: The PB7 pin can serve as an external interrupt source.

73

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• SCK/PCINT1 – Port B, Bit 1

SCK: Master Clock output, Slave Clock input pin for SPI channel. When the SPI is enabled as a

slave, this pin is configured as an input regardless of the setting of DDB1. When the SPI0 is

enabled as a master, the data direction of this pin is controlled by DDB1. When the pin is forced

to be an input, the pull-up can still be controlled by the PORTB1 bit.

PCINT1, Pin Change Interrupt source 1: The PB7 pin can serve as an external interrupt source.

• SS/PCINT0 – Port B, Bit 0

SS: Slave Port Select input. When the SPI is enabled as a slave, this pin is configured as an

input regardless of the setting of DDB0. As a slave, the SPI is activated when this pin is driven

low. When the SPI is enabled as a master, the data direction of this pin is controlled by DDB0.

When the pin is forced to be an input, the pull-up can still be controlled by the PORTB0 bit.

Table 10-4 and Table 10-5 relate the alternate functions of Port B to the overriding signals

shown in Figure 10-5 on page 70. SPI MSTR INPUT and SPI SLAVE OUTPUT constitute the

MISO signal, while MOSI is divided into SPI MSTR OUTPUT and SPI SLAVE INPUT.

PCINT0, Pin Change Interrupt source 0: The PB7 pin can serve as an external interrupt source..

Table 10-4. Overriding Signals for Alternate Functions in PB7.PB4

Signal

Name

PB7/PCINT7/OC0A/ PB6/PCINT6/OC1 PB5/PCINT5/OC1 PB4/PCINT4/A

OC1C/RTS

B/OC.4B/ADC13

A/OC.4B/ADC12

DC11

PUOE

PUOV

DDOE

DDOV

0

OC0/OC1C

ENABLE

PVOE

OC1B ENABLE

OC1A ENABLE

0

PVOV

DIEOE

DIEOV

DI

OC0/OC1C

OC1B

OC1A

0

PCINT7 • PCIE0

PCINT6 • PCIE0

PCINT5 • PCIE0

PCINT4 • PCIE0

1

PCINT7 INPUT

–

PCINT6 INPUT

–

PCINT5 INPUT

–

PCINT4 INPUT

–

AIO

74

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 10-5. Overriding Signals for Alternate Functions in PB3.PB0

PB3/PD0/PCINT3/

MISO

PB2/PDI/PCINT2/

MOSI

PB1/PCINT1/

SCK

PB0/PCINT0/

SS

Signal

Name

PUOE

PUOV

DDOE

DDOV

PVOE

PVOV

SPE • MSTR

PORTB3 • PUD

SPE • MSTR

0

SPE • MSTR

PORTB2 • PUD

SPE • MSTR

0

SPE • MSTR

PORTB1 • PUD PORTB0 • PUD

SPE • MSTR

0

SPE • MSTR

0

SPE • MSTR

SPI SLAVE OUTPUT SPI MSTR OUTPUT SCK OUTPUT

PCINT1 •

PCINT0 •

PCIE0

DIEOE

DIEOV

DI

PCINT3 • PCIE0

PCINT2 • PCIE0

PCIE0

1

SPI MSTR INPUT

PCINT3 INPUT

SPI SLAVE INPUT

PCINT2 INPUT

SCK INPUT

SPI SS

PCINT1 INPUT

PCINT0 INPUT

AIO

–

10.3.3

Alternate Functions of Port C

The Port C alternate function is as follows:

Table 10-6. Port C Pins Alternate Functions

Port Pin

Alternate Function

ICP3/CLKO/OC4A(Input Capture Timer 3 or CLK0 (Divided

System Clock) or Output Compare and direct PWM output A

for Timer 4)

PC7

OC.3A/OC4A (Output Compare and PWM output A for

Timer/Counter3 or Output Compare and complementary

PWM output A for Timer 4)

PC6

PC5

PC4

PC3

PC2

PC1

PC0

Not present on pin-out.

• ICP3/CLKO/OC.4A – Port C, Bit 7

ICP3: If Timer 3 is correctly configured, this pin can serve as Input Capture feature.

CLKO: When the corresponding fuse is enabled, this pin outputs the internal microcontroller

working frequency. If the clock prescaler is used, this will affect this output frequency.

OC.4A: Timer 4 Output Compare A. This pin can be used to generate a high-speed PWM signal

from Timer 4 module. The pin has to be configured as an output (DDC7 set “one”) to serve this

function.

75

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• OC.3A/OC.4A – Port C, Bit 6

OC.3A: Timer 3 Output Compare A. This pin can be used to generate a PWM signal from Timer

3 module.

OC.4A: Timer 4 Output Compare A. This pin can be used to generate a high-speed PWM signal

from Timer 4 module, complementary to OC.4A (PC7) signal. The pin has to be configured as an

output (DDC6 set “one”) to serve this function.

Table 10-7 relate the alternate functions of Port C to the overriding signals shown in Figure 10-5

on page 70.

Table 10-7. Overriding Signals for Alternate Functions in PC7.PC6

Signal

Name

PC7/ICP3/CLKO/OC.4

A

PC6/OC.3A/OC.4A

SRE •

PUOE

SRE • (XMM<1)

(XMM<2)|OC3A

enable

PUOV

DDOE

DDOV

PVOE

0

SRE • (XMM<1)

1

SRE • (XMM<2)

1

SRE • (XMM<1)

SRE • (XMM<2)

if (SRE.XMM<2)

then A14

PVOV

A15

else OC3A

DIEOE

DIEOV

DI

0

–

0

ICP3 input

–

AIO

76

7766E–AVR–04/10

ATmega16U4/ATmega32U4

10.3.4

Alternate Functions of Port D

The Port D pins with alternate functions are shown in Table 10-8.

Table 10-8. Port D Pins Alternate Functions

Port Pin

Alternate Function

T0/OC.4D/ADC10 (Timer/Counter0 Clock Input or Timer 4 Output Compare D /

PWM output or Analog to Digital Converter channel 10)

PD7

T1/OC.4D/ADC9 (Timer/Counter1 Clock Input or Timer 4 Output

Complementary Compare D / PWM output or Analog to Digital Converter

channel 9)

PD6

XCK1/CTS (USART1 External Clock Input/Output or UART flow control CTS

signal)

PD5

PD4

ICP1/ADC8 (Timer/Counter1 Input Capture Trigger or Analog to Digital

Converter channel 8)

PD3

PD2

PD1

INT3/TXD1 (External Interrupt3 Input or USART1 Transmit Pin)

INT2/RXD1 (External Interrupt2 Input or USART1 Receive Pin)

INT1/SDA (External Interrupt1 Input or TWI Serial DAta)

INT0/SCL/OC0B (External Interrupt0 Input or TWI Serial CLock or Output

Compare for Timer/Counter0)

PD0

The alternate pin configuration is as follows:

• T0/OC.4D/ADC10 – Port D, Bit 7

T0, Timer/Counter0 counter source.

OC.4D: Timer 4 Output Compare D. This pin can be used to generate a high-speed PWM signal

from Timer 4 module. The pin has to be configured as an output (DDD7 set “one”) to serve this

function.

ADC10: Analog to Digital Converter, Channel 10.

• T1/OC.4D/ADC9 – Port D, Bit 6

T1, Timer/Counter1 counter source.

OC.4D: Timer 4 Output Compare D. This pin can be used to generate a high-speed PWM signal

from Timer 4 module, complementary to OC.4D (PD7) signal. The pin has to be configured as

an output (DDD6 set “one”) to serve this function.

ADC9: Analog to Digital Converter, Channel 9.

• XCK1/CTS – Port D, Bit 5

XCK1, USART1 External clock. The Data Direction Register (DDD5) controls whether the clock

is output (DDD5 set) or input (DDD5 cleared). The XCK1 pin is active only when the USART1

operates in Synchronous mode.

CTS: Clear-To-Send flow control signal used by enhanced UART module.

• ICP1/ADC8 – Port D, Bit 4

ICP1 – Input Capture Pin 1: The PD4 pin can act as an input capture pin for Timer/Counter1.

ADC8: Analog to Digital Converter, Channel 8.

77

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• INT3/TXD1 – Port D, Bit 3

INT3, External Interrupt source 3: The PD3 pin can serve as an external interrupt source to the

MCU.

TXD1, Transmit Data (Data output pin for the USART1). When the USART1 Transmitter is

enabled, this pin is configured as an output regardless of the value of DDD3.

• INT2/RXD1 – Port D, Bit 2

INT2, External Interrupt source 2. The PD2 pin can serve as an External Interrupt source to the

MCU.

RXD1, Receive Data (Data input pin for the USART1). When the USART1 receiver is enabled

this pin is configured as an input regardless of the value of DDD2. When the USART forces this

pin to be an input, the pull-up can still be controlled by the PORTD2 bit.

• INT1/SDA – Port D, Bit 1

INT1, External Interrupt source 1. The PD1 pin can serve as an external interrupt source to the

MCU.

SDA, 2-wire Serial Interface Data: When the TWEN bit in TWCR is set (one) to enable the 2-wire

Serial Interface, pin PD1 is disconnected from the port and becomes the Serial Data I/O pin for

the 2-wire Serial Interface. In this mode, there is a spike filter on the pin to suppress spikes

shorter than 50 ns on the input signal, and the pin is driven by an open drain driver with slew-

rate limitation.

• INT0/SCL/OC0B – Port D, Bit 0

INT0, External Interrupt source 0. The PD0 pin can serve as an external interrupt source to the

MCU.

SCL, 2-wire Serial Interface Clock: When the TWEN bit in TWCR is set (one) to enable the 2-

wire Serial Interface, pin PD0 is disconnected from the port and becomes the Serial Clock I/O

pin for the 2-wire Serial Interface. In this mode, there is a spike filter on the pin to suppress

spikes shorter than 50 ns on the input signal, and the pin is driven by an open drain driver with

slew-rate limitation.

OC.0B: Timer 0 Output Compare B. This pin can be used to generate a PWM signal from the

Timer 0 module.

Table 10-9 and Table 10-10 relates the alternate functions of Port D to the overriding signals

shown in Figure 10-5 on page 70.

78

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 10-9. Overriding Signals for Alternate Functions PD7..PD4

PD7/T0/ PD6/T1/

OC4D/ADC10 OC4D/ADC9

PD4/ICP1/

ADC8

Signal Name

PUOE

PD5/XCK1/CTS

0

PUOV

XCK1 OUTPUT

ENABLE

DDOE

DDOV

PVOE

0

1

XCK1 OUTPUT

ENABLE

PVOV

DIEOE

DIEOV

DI

0

XCK1 OUTPUT

0

T0 INPUT

–

T1 INPUT

–

XCK1 INPUT

–

ICP1 INPUT

–

AIO

Table 10-10. Overriding Signals for Alternate Functions in PD3.PD0⁽¹⁾

PD0/INT0/SCL/

OC0B

Signal Name PD3/INT3/TXD1 PD2/INT2/RXD1

PD1/INT1/SDA

PUOE

PUOV

DDOE

DDOV

TXEN1

RXEN1

PORTD2 • PUD

RXEN1

0

TWEN

0

PORTD1 • PUD PORTD0 • PUD

TXEN1

1

TWEN

SDA_OUT

SCL_OUT

TWEN | OC0B

ENABLE

PVOE

TXEN1

0

TWEN ENABLE

PVOV

DIEOE

DIEOV

DI

TXD1

0

OC0B

INT3 ENABLE

INT2 ENABLE

1

INT1 ENABLE

1

INT0 ENABLE

1

INT3 INPUT

–

INT2 INPUT/RXD1 INT1 INPUT

– SDA INPUT

INT0 INPUT

SCL INPUT

AIO

Note:

1. When enabled, the 2-wire Serial Interface enables Slew-Rate controls on the output pins PD0

and PD1. This is not shown in this table. In addition, spike filters are connected between the

AIO outputs shown in the port figure and the digital logic of the TWI module.

79

7766E–AVR–04/10

ATmega16U4/ATmega32U4

10.3.5

Alternate Functions of Port E

The Port E pins with alternate functions are shown in Table 10-11.

Table 10-11. Port E Pins Alternate Functions

Port Pin Alternate Function

PE7

PE6

PE5

PE4

PE3

PE2

PE1

PE0

Not present on pin-out.

INT6/AIN0 (External Interrupt 6 Input or Analog Comparator Positive Input)

Not present on pin-out.

HWB (Hardware bootloader activation)

Not present on pin-out.

• INT6/AIN0 – Port E, Bit 6

INT6, External Interrupt source 6: The PE6 pin can serve as an external interrupt source.

AIN0 – Analog Comparator Negative input. This pin is directly connected to the negative input of

the Analog Comparator.

• HWB – Port E, Bit 2

HWB allows to execute the bootloader section after reset when tied to ground during external

reset pulse. The HWB mode of this pin is active only when the HWBE fuse is enable. During nor-

mal operation (excluded Reset), this pin acts as a general purpose I/O.

Table 10-12. Overriding Signals for Alternate Functions PE6, PE2

Signal

Name

PUOE

PUOV

DDOE

DDOV

PVOE

PVOV

DIEOE

DIEOV

DI

PE6/INT6/AIN0

PE2/HWB

0

1

0

INT6 ENABLE

1

0

INT6 INPUT

AIN0 INPUT

HWB

-

AIO

10.3.6

Alternate Functions of Port F

The Port F has an alternate function as analog input for the ADC as shown in Table 10-13. If

some Port F pins are configured as outputs, it is essential that these do not switch when a con-

version is in progress. This might corrupt the result of the conversion. If the JTAG interface is

80

7766E–AVR–04/10

ATmega16U4/ATmega32U4

enabled, the pull-up resistors on pins PF7(TDI), PF5(TMS), and PF4(TCK) will be activated even

if a Reset occurs.

Table 10-13. Port F Pins Alternate Functions

Port Pin

PF7

Alternate Function

ADC7/TDI (ADC input channel 7 or JTAG Test Data Input)

ADC6/TDO (ADC input channel 6 or JTAG Test Data Output)

ADC5/TMS (ADC input channel 5 or JTAG Test Mode Select)

ADC4/TCK (ADC input channel 4 or JTAG Test ClocK)

PF6

PF5

PF4

PF3

Not present on pin-out.

PF2

PF1

ADC1 (ADC input channel 1)

ADC0 (ADC input channel 0)

PF0

• TDI, ADC7 – Port F, Bit 7

ADC7, Analog to Digital Converter, Channel 7.

TDI, JTAG Test Data In: Serial input data to be shifted in to the Instruction Register or Data Reg-

ister (scan chains). When the JTAG interface is enabled, this pin can not be used as an I/O pin.

• TDO, ADC6 – Port F, Bit 6

ADC6, Analog to Digital Converter, Channel 6.

TDO, JTAG Test Data Out: Serial output data from Instruction Register or Data Register. When

the JTAG interface is enabled, this pin can not be used as an I/O pin.

The TDO pin is tri-stated unless TAP states that shift out data are entered.

• TMS, ADC5 – Port F, Bit 5

ADC5, Analog to Digital Converter, Channel 5.

TMS, JTAG Test Mode Select: This pin is used for navigating through the TAP-controller state

machine. When the JTAG interface is enabled, this pin can not be used as an I/O pin.

• TCK, ADC4 – Port F, Bit 4

ADC4, Analog to Digital Converter, Channel 4.

TCK, JTAG Test Clock: JTAG operation is synchronous to TCK. When the JTAG interface is

enabled, this pin can not be used as an I/O pin.

• ADC3 – ADC0 – Port F, Bit 1..0

Analog to Digital Converter, Channel 1.0

81

7766E–AVR–04/10

ATmega16U4/ATmega32U4

.

Table 10-14. Overriding Signals for Alternate Functions in PF7..PF4

Signal

Name

PUOE

PUOV

DDOE

PF7/ADC7/TDI

JTAGEN

1

PF6/ADC6/TDO

JTAGEN

0

PF5/ADC5/TMS

JTAGEN

1

PF4/ADC4/TCK

JTAGEN

1

JTAGEN

SHIFT_IR +

SHIFT_DR

DDOV

0

PVOE

PVOV

DIEOE

DIEOV

DI

0

JTAGEN

0

TDO

0

JTAGEN

0

–

0

–

0

–

0

–

TMS/ADC5

INPUT

TCK/ADC4

INPUT

AIO

TDI/ADC7 INPUT ADC6 INPUT

Table 10-15. Overriding Signals for Alternate Functions in PF1..PF0

Signal Name

PUOE

PUOV

DDOE

DDOV

PVOE

PVOV

DIEOE

DIEOV

DI

PF1/ADC1

PF0/ADC0

0

–

AIO

ADC1 INPUT

ADC0 INPUT

10.4 Register Description for I/O-Ports

10.4.1

Port B Data Register – PORTB

Bit

7

6

5

4

3

2

1

0

PORTB

7

PORTB

6

PORTB

5

PORTB

4

PORTB

3

PORTB

2

PORTB

1

PORTB

0

PORTB

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

82

7766E–AVR–04/10

ATmega16U4/ATmega32U4

10.4.2

10.4.3

10.4.4

Port B Data Direction Register – DDRB

Bit

7

6

5

4

3

2

1

0

DDB7

R/W

0

DDB6

R/W

0

DDB5

R/W

0

DDB4

R/W

0

DDB3

R/W

0

DDB2

R/W

0

DDB1

R/W

0

DDB0

R/W

0

DDRB

Read/Write

Initial Value

Port B Input Pins Address – PINB

Bit

7

6

5

4

3

2

1

0

PINB7

R/W

N/A

PINB6

R/W

N/A

PINB5

R/W

N/A

PINB4

R/W

N/A

PINB3

R/W

N/A

PINB2

R/W

N/A

PINB1

R/W

N/A

PINB0

R/W

N/A

PINB

Read/Write

Initial Value

Port C Data Register – PORTC

Bit

7

6

5

4

3

2

1

0

PORTC

7

PORTC

6

-

PORTC

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

10.4.5

10.4.6

10.4.7

Port C Data Direction Register – DDRC

Bit

7

6

5

4

3

2

1

0

DDC7

R/W

0

DDC6

R/W

0

-

DDRC

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

Port C Input Pins Address – PINC

Bit

7

6

5

4

3

2

1

0

PINC7

R/W

N/A

PINC6

R/W

N/A

-

PINC

Read/Write

Initial Value

R/W

N/A

R/W

N/A

R/W

N/A

R/W

N/A

R/W

N/A

R/W

N/A

Port D Data Register – PORTD

Bit

7

6

5

4

3

2

1

0

PORTD

7

PORTD

6

PORTD

5

PORTD

4

PORTD

3

PORTD

2

PORTD

1

PORTD

0

PORTD

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

10.4.8

10.4.9

Port D Data Direction Register – DDRD

Bit

7

6

5

4

3

2

1

0

DDD7

R/W

0

DDD6

R/W

0

DDD5

R/W

0

DDD4

R/W

0

DDD3

R/W

0

DDD2

R/W

0

DDD1

R/W

0

DDD0

R/W

0

DDRD

Read/Write

Initial Value

Port D Input Pins Address – PIND

Bit

7

6

5

4

3

2

1

0

PIND7

R/W

N/A

PIND6

R/W

N/A

PIND5

R/W

N/A

PIND4

R/W

N/A

PIND3

R/W

N/A

PIND2

R/W

N/A

PIND1

R/W

N/A

PIND0

R/W

N/A

PIND

Read/Write

Initial Value

83

7766E–AVR–04/10

ATmega16U4/ATmega32U4

10.4.10 Port E Data Register – PORTE

Bit

7

6

5

4

3

2

1

0

-

PORTE

6

-

PORTE

2

-

PORTE

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

10.4.11 Port E Data Direction Register – DDRE

Bit

7

6

5

4

3

2

1

0

-

DDE6

R/W

0

-

DDE2

R/W

0

-

DDRE

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

10.4.12 Port E Input Pins Address – PINE

Bit

7

6

5

4

3

2

1

0

-

PINE6

R/W

N/A

-

PINE2

R/W

N/A

-

PINE

Read/Write

Initial Value

R/W

N/A

R/W

N/A

R/W

N/A

R/W

N/A

R/W

N/A

R/W

N/A

10.4.13 Port F Data Register – PORTF

Bit

7

6

5

4

3

2

1

0

PORTF

7

PORTF

6

PORTF

5

PORTF

4

-

PORTF

1

PORTF

0

PORTF

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

10.4.14 Port F Data Direction Register – DDRF

Bit

7

6

5

4

3

2

1

0

DDF7

R/W

0

DDF6

R/W

0

DDF5

R/W

0

DDF4

R/W

0

-

DDF1

R/W

0

DDF0

R/W

0

DDRF

Read/Write

Initial Value

R/W

0

R/W

0

10.4.15 Port F Input Pins Address – PINF

Bit

7

6

5

4

3

2

1

0

PINF7

R/W

N/A

PINF6

R/W

N/A

PINF5

R/W

N/A

PINF4

R/W

N/A

-

PINF1

R/W

N/A

PINF0

R/W

N/A

PINF

Read/Write

Initial Value

R/W

N/A

R/W

N/A

84

7766E–AVR–04/10

ATmega16U4/ATmega32U4

11. External Interrupts

The External Interrupts are triggered by the INT6, INT3:0 pin or any of the PCINT7..0 pins.

Observe that, if enabled, the interrupts will trigger even if the INT[6;3:0] or PCINT7..0 pins are

configured as outputs. This feature provides a way of generating a software interrupt.

The Pin change interrupt PCI0 will trigger if any enabled PCINT7:0 pin toggles. PCMSK0 Regis-

ter control which pins contribute to the pin change interrupts. Pin change interrupts on PCINT7

..0 are detected asynchronously. This implies that these interrupts can be used for waking the

part also from sleep modes other than Idle mode.

The External Interrupts can be triggered by a falling or rising edge or a low level. This is set up

as indicated in the specification for the External Interrupt Control Registers – EICRA (INT3:0)

and EICRB (INT6). When the external interrupt is enabled and is configured as level triggered,

the interrupt will trigger as long as the pin is held low. Note that recognition of falling or rising

edge interrupts on INT6 requires the presence of an I/O clock, described in “System Clock and

Clock Options” on page 27. Low level interrupts and the edge interrupt on INT3:0 are detected

asynchronously. This implies that these interrupts can be used for waking the part also from

sleep modes other than Idle mode. The I/O clock is halted in all sleep modes except Idle mode.

Note that if a level triggered interrupt is used for wake-up from Power-down, the required level

must be held long enough for the MCU to complete the wake-up to trigger the level interrupt. If

the level disappears before the end of the Start-up Time, the MCU will still wake up, but no inter-

rupt will be generated. The start-up time is defined by the SUT and CKSEL Fuses as described

in “System Clock and Clock Options” on page 27.

11.0.1

External Interrupt Control Register A – EICRA

The External Interrupt Control Register A contains control bits for interrupt sense control.

Bit

7

6

5

4

3

2

1

0

ISC31

R/W

0

ISC30

R/W

0

ISC21

R/W

0

ISC20

R/W

0

ISC11

R/W

0

ISC10

R/W

0

ISC01

R/W

0

ISC00

R/W

0

EICRA

Read/Write

Initial Value

• Bits 7..0 – ISC31, ISC30 – ISC00, ISC00: External Interrupt 3 - 0 Sense Control Bits

The External Interrupts 3 - 0 are activated by the external pins INT3:0 if the SREG I-flag and the

corresponding interrupt mask in the EIMSK is set. The level and edges on the external pins that

activate the interrupts are defined in Table 11-1. Edges on INT3..INT0 are registered asynchro-

nously. Pulses on INT3:0 pins wider than the minimum pulse width given in Table 11-2 will

generate an interrupt. Shorter pulses are not guaranteed to generate an interrupt. If low level

interrupt is selected, the low level must be held until the completion of the currently executing

instruction to generate an interrupt. If enabled, a level triggered interrupt will generate an inter-

rupt request as long as the pin is held low. When changing the ISCn bit, an interrupt can occur.

Therefore, it is recommended to first disable INTn by clearing its Interrupt Enable bit in the

EIMSK Register. Then, the ISCn bit can be changed. Finally, the INTn interrupt flag should be

cleared by writing a logical one to its Interrupt Flag bit (INTFn) in the EIFR Register before the

interrupt is re-enabled.

85

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 11-1.

Interrupt Sense Control⁽¹⁾

ISCn0 Description

ISCn1

0

1

0

1

0

1

The low level of INTn generates an interrupt request.

Any edge of INTn generates asynchronously an interrupt request.

The falling edge of INTn generates asynchronously an interrupt request.

The rising edge of INTn generates asynchronously an interrupt request.

Note:

1. n = 3, 2, 1or 0.

When changing the ISCn1/ISCn0 bits, the interrupt must be disabled by clearing its Interrupt

Enable bit in the EIMSK Register. Otherwise an interrupt can occur when the bits are changed.

Table 11-2. Asynchronous External Interrupt Characteristics

Symbol Parameter Condition Min

Minimum pulse width for

Typ

Max

Units

t_INT

50

ns

asynchronous external interrupt

11.0.2

External Interrupt Control Register B – EICRB

Bit

7

6

5

4

3

2

1

0

-

ISC61

R/W

0

ISC60

R/W

0

-

EICRB

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

• Bit 7..6 – Res: Reserved Bits

These bits are reserved bits in the ATmega16U4/ATmega32U4 and always read as zero.

• Bits 5, 4 – ISC61, ISC60: External Interrupt 6 Sense Control Bits

The External Interrupt 6 is activated by the external pin INT6 if the SREG I-flag and the corre-

sponding interrupt mask in the EIMSK is set. The level and edges on the external pin that

activate the interrupt are defined in Table 11-3. The value on the INT6 pin are sampled before

detecting edges. If edge or toggle interrupt is selected, pulses that last longer than one clock

period will generate an interrupt. Shorter pulses are not guaranteed to generate an interrupt.

Observe that CPU clock frequency can be lower than the XTAL frequency if the XTAL divider is

enabled. If low level interrupt is selected, the low level must be held until the completion of the

currently executing instruction to generate an interrupt. If enabled, a level triggered interrupt will

generate an interrupt request as long as the pin is held low.

Table 11-3.

ISC61 ISC60 Description

Interrupt Sense Control⁽¹⁾

0

1

The low level of INT6 generates an interrupt request.

Any logical change on INT6 generates an interrupt request

The falling edge between two samples of INT6 generates an interrupt

request.

1

0

1

The rising edge between two samples of INT6 generates an interrupt

request.

1

Note:

1. When changing the ISC61/ISC60 bits, the interrupt must be disabled by clearing its Interrupt

Enable bit in the EIMSK Register. Otherwise an interrupt can occur when the bits are changed.

86

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 3..0 – Res: Reserved Bits

These bits are reserved bits in the ATmega16U4/ATmega32U4 and always read as zero.

11.0.3

External Interrupt Mask Register – EIMSK

Bit

7

6

5

4

3

2

1

0

-

INT6

-

INT3

R/W

0

INT2

R/W

0

INT1

R/W

0

IINT0

R/W

0

EIMSK

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

• Bits 7..0 – INT6, INT3 – INT0: External Interrupt Request 6, 3 - 0 Enable

When an INT[6;3:0] bit is written to one and the I-bit in the Status Register (SREG) is set (one),

the corresponding external pin interrupt is enabled. The Interrupt Sense Control bits in the Exter-

nal Interrupt Control Registers – EICRA and EICRB – defines whether the external interrupt is

activated on rising or falling edge or level sensed. Activity on any of these pins will trigger an

interrupt request even if the pin is enabled as an output. This provides a way of generating a

software interrupt.

11.0.4

External Interrupt Flag Register – EIFR

Bit

7

6

5

4

3

2

1

0

-

INTF6

-

INTF3

R/W

0

INTF2

R/W

0

INTF1

R/W

0

IINTF0

R/W

0

EIFR

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

• Bits 7..0 – INTF6, INTF3 - INTF0: External Interrupt Flags 6, 3 - 0

When an edge or logic change on the INT[6;3:0] pin triggers an interrupt request, INTF7:0

becomes set (one). If the I-bit in SREG and the corresponding interrupt enable bit, INT[6;3:0] in

EIMSK, are set (one), the MCU will jump to the interrupt vector. The flag is cleared when the

interrupt routine is executed. Alternatively, the flag can be cleared by writing a logical one to it.

These flags are always cleared when INT[6;3:0] are configured as level interrupt. Note that when

entering sleep mode with the INT3:0 interrupts disabled, the input buffers on these pins will be

disabled. This may cause a logic change in internal signals which will set the INTF3:0 flags. See

“Digital Input Enable and Sleep Modes” on page 69 for more information.

11.0.5

Pin Change Interrupt Control Register - PCICR

Bit

7

6

5

4

–

R

0

3

–

R

0

2

–

R

0

1

–

R

0

–

R

0

PCIE0

R/W

0

PCICR

Read/Write

Initial Value

R

0

R

0

• Bit 0 – PCIE0: Pin Change Interrupt Enable 0

When the PCIE0 bit is set (one) and the I-bit in the Status Register (SREG) is set (one), pin

change interrupt 0 is enabled. Any change on any enabled PCINT7..0 pin will cause an interrupt.

The corresponding interrupt of Pin Change Interrupt Request is executed from the PCI0 Interrupt

Vector. PCINT7..0 pins are enabled individually by the PCMSK0 Register.

11.0.6

Pin Change Interrupt Flag Register – PCIFR

Bit

7

6

5

4

–

R

0

3

–

R

0

2

–

R

0

1

–

R

0

–

R

0

PCIF0

R/W

0

PCIFR

Read/Write

Initial Value

R

0

R

0

87

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 0 – PCIF0: Pin Change Interrupt Flag 0

When a logic change on any PCINT7..0 pin triggers an interrupt request, PCIF0 becomes set

(one). If the I-bit in SREG and the PCIE0 bit in EIMSK are set (one), the MCU will jump to the

corresponding Interrupt Vector. The flag is cleared when the interrupt routine is executed. Alter-

natively, the flag can be cleared by writing a logical one to it.

11.0.7

Pin Change Mask Register 0 – PCMSK0

Bit

7

6

5

4

3

2

1

0

PCINT7

PCINT6

PCINT5

R/W

0

PCINT4

R/W

0

PCINT3

R/W

0

PCINT2

R/W

0

PCINT1

R/W

0

PCINT0

R/W

0

PCMSK0

Read/Write

Initial Value

R/W

0

R/W

0

• Bit 7..0 – PCINT7..0: Pin Change Enable Mask 7..0

Each PCINT7..0 bit selects whether pin change interrupt is enabled on the corresponding I/O

pin. If PCINT7..0 is set and the PCIE0 bit in PCICR is set, pin change interrupt is enabled on the

corresponding I/O pin. If PCINT7..0 is cleared, pin change interrupt on the corresponding I/O pin

is disabled.

88

7766E–AVR–04/10

ATmega16U4/ATmega32U4

12. Timer/Counter0, Timer/Counter1, and Timer/Counter3 Prescalers

Timer/Counter0, 1, and 3 share the same prescaler module, but the Timer/Counters can have

different prescaler settings. The description below applies to all Timer/Counters. Tn is used as a

general name, n = 0, 1 or 3.

12.1 Internal Clock Source

The Timer/Counter can be clocked directly by the system clock (by setting the CSn2:0 = 1). This

provides the fastest operation, with a maximum Timer/Counter clock frequency equal to system

clock frequency (f_{CLK_I/O}). Alternatively, one of four taps from the prescaler can be used as a

clock source. The prescaled clock has a frequency of either f_{CLK_I/O}/8, f_{CLK_I/O}/64, f_{CLK_I/O}/256, or

f_{CLK_I/O}/1024.

12.2 Prescaler Reset

The prescaler is free running, i.e., operates independently of the Clock Select logic of the

Timer/Counter, and it is shared by the Timer/Counter Tn. Since the prescaler is not affected by

the Timer/Counter’s clock select, the state of the prescaler will have implications for situations

where a prescaled clock is used. One example of prescaling artifacts occurs when the timer is

enabled and clocked by the prescaler (6 > CSn2:0 > 1). The number of system clock cycles from

when the timer is enabled to the first count occurs can be from 1 to N+1 system clock cycles,

where N equals the prescaler divisor (8, 64, 256, or 1024).

It is possible to use the prescaler reset for synchronizing the Timer/Counter to program execu-

tion. However, care must be taken if the other Timer/Counter that shares the same prescaler

also uses prescaling. A prescaler reset will affect the prescaler period for all Timer/Counters it is

connected to.

12.3 External Clock Source

An external clock source applied to the Tn pin can be used as Timer/Counter clock (clk_Tn). The

Tn pin is sampled once every system clock cycle by the pin synchronization logic. The synchro-

nized (sampled) signal is then passed through the edge detector. Figure 12-1 shows a functional

equivalent block diagram of the Tn synchronization and edge detector logic. The registers are

clocked at the positive edge of the internal system clock (clk_I/O). The latch is transparent in the

high period of the internal system clock.

The edge detector generates one clk_Tnpulse for each positive (CSn2:0 = 7) or negative (CSn2:0

= 6) edge it detects.

Figure 12-1. Tn/T0 Pin Sampling

Tn_sync

(To Clock

Tn

D

Q

D

Q

D

Q

Select Logic)

LE

clk_I/O

Synchronization

Edge Detector

The synchronization and edge detector logic introduces a delay of 2.5 to 3.5 system clock cycles

from an edge has been applied to the Tn pin to the counter is updated.

Enabling and disabling of the clock input must be done when Tn has been stable for at least one

system clock cycle, otherwise it is a risk that a false Timer/Counter clock pulse is generated.

89

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Each half period of the external clock applied must be longer than one system clock cycle to

ensure correct sampling. The external clock must be guaranteed to have less than half the sys-

tem clock frequency (f_ExtClk< f_{clk_I/O}/2) given a 50/50% duty cycle. Since the edge detector uses

sampling, the maximum frequency of an external clock it can detect is half the sampling fre-

quency (Nyquist sampling theorem). However, due to variation of the system clock frequency

and duty cycle caused by Oscillator source (crystal, resonator, and capacitors) tolerances, it is

recommended that maximum frequency of an external clock source is less than f_{clk_I/O}/2.5.

An external clock source can not be prescaled.

Figure 12-2. Prescaler for synchronous Timer/Counters

clk_I/O

Clear

PSR10

Tn

Synchronization

Tn

Synchronization

CSn0

CSn1

CSn2

CSn0

CSn1

CSn2

TIMER/COUNTERn CLOCK SOURCE

clk_Tn

TIMER/COUNTERn CLOCK SOURCE

clk_Tn

Note:

T3 input is not available on the ATmega16U4/ATmega32U4 products. “Tn” only refers to

either T0 or T1 inputs.

12.4 General Timer/Counter Control Register – GTCCR

Bit

7

6

–

R

0

5

–

R

0

4

–

R

0

3

–

R

0

2

–

R

0

1

0

TSM

R/W

0

PSRASY

PSRSYNC

GTCCR

Read/Write

Initial Value

R/W

0

R/W

0

• Bit 7 – TSM: Timer/Counter Synchronization Mode

Writing the TSM bit to one activates the Timer/Counter Synchronization mode. In this mode, the

value that is written to the PSRASY and PSRSYNC bits is kept, hence keeping the correspond-

ing prescaler reset signals asserted. This ensures that the corresponding Timer/Counters are

halted and can be configured to the same value without the risk of one of them advancing during

configuration. When the TSM bit is written to zero, the PSRASY and PSRSYNC bits are cleared

by hardware, and the Timer/Counters start counting simultaneously.

• Bit 0 – PSRSYNC: Prescaler Reset for Synchronous Timer/Counters

When this bit is one, Timer/Counter0 and Timer/Counter1 and Timer/Counter3 prescaler will be

Reset. This bit is normally cleared immediately by hardware, except if the TSM bit is set. Note

that Timer/Counter0, Timer/Counter1 and Timer/Counter3 share the same prescaler and a reset

of this prescaler will affect all timers.

90

7766E–AVR–04/10

ATmega16U4/ATmega32U4

13. 8-bit Timer/Counter0 with PWM

Timer/Counter0 is a general purpose 8-bit Timer/Counter module, with two independent Output

Compare Units, and with PWM support. It allows accurate program execution timing (event man-

agement) and wave generation. The main features are:

• Two Independent Output Compare Units

• Double Buffered Output Compare Registers

• Clear Timer on Compare Match (Auto Reload)

• Glitch Free, Phase Correct Pulse Width Modulator (PWM)

• Variable PWM Period

• Frequency Generator

• Three Independent Interrupt Sources (TOV0, OCF0A, and OCF0B)

13.1 Overview

A simplified block diagram of the 8-bit Timer/Counter is shown in Figure 13-1. For the actual

placement of I/O pins, refer to “Pinout ATmega16U4/ATmega32U4” on page 3. CPU accessible

I/O Registers, including I/O bits and I/O pins, are shown in bold. The device-specific I/O Register

and bit locations are listed in the “8-bit Timer/Counter Register Description” on page 102.

Figure 13-1. 8-bit Timer/Counter Block Diagram

Count

TOVn

(Int.Req.)

Clear

Control Logic

Clock Select

Direction

clk_Tn

Edge

Detector

Tn

TOP

BOTTOM

( From Prescaler )

Timer/Counter

TCNTn

=

0

OCnA

(Int.Req.)

Waveform

Generation

OCnA

OCnB

=

OCRnA

Fixed

TOP

Value

OCnB

(Int.Req.)

Waveform

Generation

=

OCRnB

TCCRnA

TCCRnB

13.1.1

Registers

The Timer/Counter (TCNT0) and Output Compare Registers (OCR0A and OCR0B) are 8-bit

registers. Interrupt request (abbreviated to Int.Req. in the figure) signals are all visible in the

Timer Interrupt Flag Register (TIFR0). All interrupts are individually masked with the Timer Inter-

rupt Mask Register (TIMSK0). TIFR0 and TIMSK0 are not shown in the figure.

The Timer/Counter can be clocked internally, via the prescaler, or by an external clock source on

the T0 pin. The Clock Select logic block controls which clock source and edge the Timer/Counter

uses to increment (or decrement) its value. The Timer/Counter is inactive when no clock source

is selected. The output from the Clock Select logic is referred to as the timer clock (clk_T0).

91

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The double buffered Output Compare Registers (OCR0A and OCR0B) are compared with the

Timer/Counter value at all times. The result of the compare can be used by the Waveform Gen-

erator to generate a PWM or variable frequency output on the Output Compare pins (OC0A and

OC0B). See “Output Compare Unit” on page 93. for details. The Compare Match event will also

set the Compare Flag (OCF0A or OCF0B) which can be used to generate an Output Compare

interrupt request.

13.1.2

Definitions

Many register and bit references in this section are written in general form. A lower case “n”

replaces the Timer/Counter number, in this case 0. A lower case “x” replaces the Output Com-

pare Unit, in this case Compare Unit A or Compare Unit B. However, when using the register or

bit defines in a program, the precise form must be used, i.e., TCNT0 for accessing

Timer/Counter0 counter value and so on.

The definitions in the table below are also used extensively throughout the document.

Table 13-1.

BOTTOM The counter reaches the BOTTOM when it becomes 0x00.

MAX

TOP

The counter reaches its MAXimum when it becomes 0xFF (decimal 255).

The counter reaches the TOP when it becomes equal to the highest

value in the count sequence. The TOP value can be assigned to be the

fixed value 0xFF (MAX) or the value stored in the OCR0A Register. The

assignment is dependent on the mode of operation.

13.2 Timer/Counter Clock Sources

The Timer/Counter can be clocked by an internal or an external clock source. The clock source

is selected by the Clock Select logic which is controlled by the Clock Select (CS02:0) bits

located in the Timer/Counter Control Register (TCCR0B). For details on clock sources and pres-

caler, see “Timer/Counter0, Timer/Counter1, and Timer/Counter3 Prescalers” on page 89.

13.3 Counter Unit

The main part of the 8-bit Timer/Counter is the programmable bi-directional counter unit. Figure

13-2 shows a block diagram of the counter and its surroundings.

Figure 13-2. Counter Unit Block Diagram

TOVn

(Int.Req.)

DATA BUS

Clock Select

count

clear

Edge

Detector

Tn

clk_Tn

TCNTn

Control Logic

direction

( From Prescaler )

bottom

top

Signal description (internal signals):

count

Increment or decrement TCNT0 by 1.

direction

Select between increment and decrement.

92

7766E–AVR–04/10

ATmega16U4/ATmega32U4

clear

clk_Tn

top

Clear TCNT0 (set all bits to zero).

Timer/Counter clock, referred to as clk_T0in the following.

Signalize that TCNT0 has reached maximum value.

Signalize that TCNT0 has reached minimum value (zero).

bottom

Depending of the mode of operation used, the counter is cleared, incremented, or decremented

at each timer clock (clk_T0). clk_T0can be generated from an external or internal clock source,

selected by the Clock Select bits (CS02:0). When no clock source is selected (CS02:0 = 0) the

timer is stopped. However, the TCNT0 value can be accessed by the CPU, regardless of

whether clk_T0is present or not. A CPU write overrides (has priority over) all counter clear or

count operations.

The counting sequence is determined by the setting of the WGM01 and WGM00 bits located in

the Timer/Counter Control Register (TCCR0A) and the WGM02 bit located in the Timer/Counter

Control Register B (TCCR0B). There are close connections between how the counter behaves

(counts) and how waveforms are generated on the Output Compare outputs OC0A and OC0B.

For more details about advanced counting sequences and waveform generation, see “Modes of

Operation” on page 96.

The Timer/Counter Overflow Flag (TOV0) is set according to the mode of operation selected by

the WGM02:0 bits. TOV0 can be used for generating a CPU interrupt.

13.4 Output Compare Unit

The 8-bit comparator continuously compares TCNT0 with the Output Compare Registers

(OCR0A and OCR0B). Whenever TCNT0 equals OCR0A or OCR0B, the comparator signals a

match. A match will set the Output Compare Flag (OCF0A or OCF0B) at the next timer clock

cycle. If the corresponding interrupt is enabled, the Output Compare Flag generates an Output

Compare interrupt. The Output Compare Flag is automatically cleared when the interrupt is exe-

cuted. Alternatively, the flag can be cleared by software by writing a logical one to its I/O bit

location. The Waveform Generator uses the match signal to generate an output according to

operating mode set by the WGM02:0 bits and Compare Output mode (COM0x1:0) bits. The max

and bottom signals are used by the Waveform Generator for handling the special cases of the

extreme values in some modes of operation (“Modes of Operation” on page 96).

Figure 13-3 shows a block diagram of the Output Compare unit.

93

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 13-3. Output Compare Unit, Block Diagram

DATA BUS

OCRnx

TCNTn

= (8-bit Comparator )

OCFnx (Int.Req.)

top

bottom

FOCn

Waveform Generator

OCnx

WGMn1:0

COMnX1:0

The OCR0x Registers are double buffered when using any of the Pulse Width Modulation

(PWM) modes. For the normal and Clear Timer on Compare (CTC) modes of operation, the dou-

ble buffering is disabled. The double buffering synchronizes the update of the OCR0x Compare

Registers to either top or bottom of the counting sequence. The synchronization prevents the

occurrence of odd-length, non-symmetrical PWM pulses, thereby making the output glitch-free.

The OCR0x Register access may seem complex, but this is not case. When the double buffering

is enabled, the CPU has access to the OCR0x Buffer Register, and if double buffering is dis-

abled the CPU will access the OCR0x directly.

13.4.1

Force Output Compare

In non-PWM waveform generation modes, the match output of the comparator can be forced by

writing a one to the Force Output Compare (FOC0x) bit. Forcing Compare Match will not set the

OCF0x Flag or reload/clear the timer, but the OC0x pin will be updated as if a real Compare

Match had occurred (the COM0x1:0 bits settings define whether the OC0x pin is set, cleared or

toggled).

13.4.2

13.4.3

Compare Match Blocking by TCNT0 Write

All CPU write operations to the TCNT0 Register will block any Compare Match that occur in the

next timer clock cycle, even when the timer is stopped. This feature allows OCR0x to be initial-

ized to the same value as TCNT0 without triggering an interrupt when the Timer/Counter clock is

enabled.

Using the Output Compare Unit

Since writing TCNT0 in any mode of operation will block all Compare Matches for one timer

clock cycle, there are risks involved when changing TCNT0 when using the Output Compare

Unit, independently of whether the Timer/Counter is running or not. If the value written to TCNT0

equals the OCR0x value, the Compare Match will be missed, resulting in incorrect waveform

94

7766E–AVR–04/10

ATmega16U4/ATmega32U4

generation. Similarly, do not write the TCNT0 value equal to BOTTOM when the counter is

down-counting.

The setup of the OC0x should be performed before setting the Data Direction Register for the

port pin to output. The easiest way of setting the OC0x value is to use the Force Output Com-

pare (FOC0x) strobe bits in Normal mode. The OC0x Registers keep their values even when

changing between Waveform Generation modes.

Be aware that the COM0x1:0 bits are not double buffered together with the compare value.

Changing the COM0x1:0 bits will take effect immediately.

13.5 Compare Match Output Unit

The Compare Output mode (COM0x1:0) bits have two functions. The Waveform Generator uses

the COM0x1:0 bits for defining the Output Compare (OC0x) state at the next Compare Match.

Also, the COM0x1:0 bits control the OC0x pin output source. Figure 13-4 shows a simplified

schematic of the logic affected by the COM0x1:0 bit setting. The I/O Registers, I/O bits, and I/O

pins in the figure are shown in bold. Only the parts of the general I/O Port Control Registers

(DDR and PORT) that are affected by the COM0x1:0 bits are shown. When referring to the

OC0x state, the reference is for the internal OC0x Register, not the OC0x pin. If a system reset

occur, the OC0x Register is reset to “0”.

Figure 13-4. Compare Match Output Unit, Schematic

COMnx1

Waveform

Generator

COMnx0

FOCn

D

Q

1

0

OCnx

D

Q

PORT

D

Q

DDR

clk_I/O

The general I/O port function is overridden by the Output Compare (OC0x) from the Waveform

Generator if either of the COM0x1:0 bits are set. However, the OC0x pin direction (input or out-

put) is still controlled by the Data Direction Register (DDR) for the port pin. The Data Direction

Register bit for the OC0x pin (DDR_OC0x) must be set as output before the OC0x value is visi-

ble on the pin. The port override function is independent of the Waveform Generation mode.

The design of the Output Compare pin logic allows initialization of the OC0x state before the out-

put is enabled. Note that some COM0x1:0 bit settings are reserved for certain modes of

operation. See “8-bit Timer/Counter Register Description” on page 102.

95

7766E–AVR–04/10

ATmega16U4/ATmega32U4

13.5.1

Compare Output Mode and Waveform Generation

The Waveform Generator uses the COM0x1:0 bits differently in Normal, CTC, and PWM modes.

For all modes, setting the COM0x1:0 = 0 tells the Waveform Generator that no action on the

OC0x Register is to be performed on the next Compare Match. For compare output actions in

the non-PWM modes refer to Table 13-2 on page 102. For fast PWM mode, refer to Table 13-3

on page 102, and for phase correct PWM refer to Table 13-4 on page 103.

A change of the COM0x1:0 bits state will have effect at the first Compare Match after the bits are

written. For non-PWM modes, the action can be forced to have immediate effect by using the

FOC0x strobe bits.

13.6 Modes of Operation

The mode of operation, i.e., the behavior of the Timer/Counter and the Output Compare pins, is

defined by the combination of the Waveform Generation mode (WGM02:0) and Compare Output

mode (COM0x1:0) bits. The Compare Output mode bits do not affect the counting sequence,

while the Waveform Generation mode bits do. The COM0x1:0 bits control whether the PWM out-

put generated should be inverted or not (inverted or non-inverted PWM). For non-PWM modes

the COM0x1:0 bits control whether the output should be set, cleared, or toggled at a Compare

Match (See “Compare Match Output Unit” on page 95.).

For detailed timing information see “Timer/Counter Timing Diagrams” on page 100.

13.6.1

Normal Mode

The simplest mode of operation is the Normal mode (WGM02:0 = 0). In this mode the counting

direction is always up (incrementing), and no counter clear is performed. The counter simply

overruns when it passes its maximum 8-bit value (TOP = 0xFF) and then restarts from the bot-

tom (0x00). In normal operation the Timer/Counter Overflow Flag (TOV0) will be set in the same

timer clock cycle as the TCNT0 becomes zero. The TOV0 Flag in this case behaves like a ninth

bit, except that it is only set, not cleared. However, combined with the timer overflow interrupt

that automatically clears the TOV0 Flag, the timer resolution can be increased by software.

There are no special cases to consider in the Normal mode, a new counter value can be written

anytime.

The Output Compare Unit can be used to generate interrupts at some given time. Using the Out-

put Compare to generate waveforms in Normal mode is not recommended, since this will

occupy too much of the CPU time.

13.6.2

Clear Timer on Compare Match (CTC) Mode

In Clear Timer on Compare or CTC mode (WGM02:0 = 2), the OCR0A Register is used to

manipulate the counter resolution. In CTC mode the counter is cleared to zero when the counter

value (TCNT0) matches the OCR0A. The OCR0A defines the top value for the counter, hence

also its resolution. This mode allows greater control of the Compare Match output frequency. It

also simplifies the operation of counting external events.

The timing diagram for the CTC mode is shown in Figure 13-5. The counter value (TCNT0)

increases until a Compare Match occurs between TCNT0 and OCR0A, and then counter

(TCNT0) is cleared.

96

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 13-5. CTC Mode, Timing Diagram

OCnx Interrupt Flag Set

TCNTn

OCn

(Toggle)

(COMnx1:0 = 1)

1

2

3

4

Period

An interrupt can be generated each time the counter value reaches the TOP value by using the

OCF0A Flag. If the interrupt is enabled, the interrupt handler routine can be used for updating

the TOP value. However, changing TOP to a value close to BOTTOM when the counter is run-

ning with none or a low prescaler value must be done with care since the CTC mode does not

have the double buffering feature. If the new value written to OCR0A is lower than the current

value of TCNT0, the counter will miss the Compare Match. The counter will then have to count to

its maximum value (0xFF) and wrap around starting at 0x00 before the Compare Match can

occur.

For generating a waveform output in CTC mode, the OC0A output can be set to toggle its logical

level on each Compare Match by setting the Compare Output mode bits to toggle mode

(COM0A1:0 = 1). The OC0A value will not be visible on the port pin unless the data direction for

the pin is set to output. The waveform generated will have a maximum frequency of f_OC0

_{clk_I/O}/2 when OCR0A is set to zero (0x00). The waveform frequency is defined by the following

equation:

=

f

clk_I/O

f

= -------------------------------------------------

OCnx

2 ⋅ N ⋅ (1 + OCRnx)

The N variable represents the prescaler factor (1, 8, 64, 256, or 1024).

As for the Normal mode of operation, the TOV0 Flag is set in the same timer clock cycle that the

counter counts from MAX to 0x00.

13.6.3

Fast PWM Mode

The fast Pulse Width Modulation or fast PWM mode (WGM02:0 = 3 or 7) provides a high fre-

quency PWM waveform generation option. The fast PWM differs from the other PWM option by

its single-slope operation. The counter counts from BOTTOM to TOP then restarts from BOT-

TOM. TOP is defined as 0xFF when WGM2:0 = 3, and OCR0A when WGM2:0 = 7. In non-

inverting Compare Output mode, the Output Compare (OC0x) is cleared on the Compare Match

between TCNT0 and OCR0x, and set at BOTTOM. In inverting Compare Output mode, the out-

put is set on Compare Match and cleared at BOTTOM. Due to the single-slope operation, the

operating frequency of the fast PWM mode can be twice as high as the phase correct PWM

mode that use dual-slope operation. This high frequency makes the fast PWM mode well suited

for power regulation, rectification, and DAC applications. High frequency allows physically small

sized external components (coils, capacitors), and therefore reduces total system cost.

In fast PWM mode, the counter is incremented until the counter value matches the TOP value.

The counter is then cleared at the following timer clock cycle. The timing diagram for the fast

97

7766E–AVR–04/10

ATmega16U4/ATmega32U4

PWM mode is shown in Figure 13-6. The TCNT0 value is in the timing diagram shown as a his-

togram for illustrating the single-slope operation. The diagram includes non-inverted and

inverted PWM outputs. The small horizontal line marks on the TCNT0 slopes represent Com-

pare Matches between OCR0x and TCNT0.

Figure 13-6. Fast PWM Mode, Timing Diagram

OCRnx Interrupt Flag Set

OCRnx Update and

TOVn Interrupt Flag Set

TCNTn

(COMnx1:0 = 2)

(COMnx1:0 = 3)

OCnx

1

2

3

4

5

6

7

Period

The Timer/Counter Overflow Flag (TOV0) is set each time the counter reaches TOP. If the inter-

rupt is enabled, the interrupt handler routine can be used for updating the compare value.

In fast PWM mode, the compare unit allows generation of PWM waveforms on the OC0x pins.

Setting the COM0x1:0 bits to two will produce a non-inverted PWM and an inverted PWM output

can be generated by setting the COM0x1:0 to three: Setting the COM0A1:0 bits to one allows

the OC0A pin to toggle on Compare Matches if the WGM02 bit is set. This option is not available

for the OC0B pin (See Table 13-3 on page 102). The actual OC0x value will only be visible on

the port pin if the data direction for the port pin is set as output. The PWM waveform is gener-

ated by setting (or clearing) the OC0x Register at the Compare Match between OCR0x and

TCNT0, and clearing (or setting) the OC0x Register at the timer clock cycle the counter is

cleared (changes from TOP to BOTTOM).

The PWM frequency for the output can be calculated by the following equation:

f

clk_I/O

f

= -----------------

OCnxPWM

N ⋅ 256

The N variable represents the prescaler factor (1, 8, 64, 256, or 1024).

The extreme values for the OCR0A Register represents special cases when generating a PWM

waveform output in the fast PWM mode. If the OCR0A is set equal to BOTTOM, the output will

be a narrow spike for each MAX+1 timer clock cycle. Setting the OCR0A equal to MAX will result

in a constantly high or low output (depending on the polarity of the output set by the COM0A1:0

bits.)

A frequency (with 50% duty cycle) waveform output in fast PWM mode can be achieved by set-

ting OC0x to toggle its logical level on each Compare Match (COM0x1:0 = 1). The waveform

generated will have a maximum frequency of f_OC0= f_{clk_I/O}/2 when OCR0A is set to zero. This

98

7766E–AVR–04/10

ATmega16U4/ATmega32U4

feature is similar to the OC0A toggle in CTC mode, except the double buffer feature of the Out-

put Compare unit is enabled in the fast PWM mode.

13.6.4

Phase Correct PWM Mode

The phase correct PWM mode (WGM02:0 = 1 or 5) provides a high resolution phase correct

PWM waveform generation option. The phase correct PWM mode is based on a dual-slope

operation. The counter counts repeatedly from BOTTOM to TOP and then from TOP to BOT-

TOM. TOP is defined as 0xFF when WGM2:0 = 1, and OCR0A when WGM2:0 = 5. In non-

inverting Compare Output mode, the Output Compare (OC0x) is cleared on the Compare Match

between TCNT0 and OCR0x while up counting, and set on the Compare Match while down-

counting. In inverting Output Compare mode, the operation is inverted. The dual-slope operation

has lower maximum operation frequency than single slope operation. However, due to the sym-

metric feature of the dual-slope PWM modes, these modes are preferred for motor control

applications.

In phase correct PWM mode the counter is incremented until the counter value matches TOP.

When the counter reaches TOP, it changes the count direction. The TCNT0 value will be equal

to TOP for one timer clock cycle. The timing diagram for the phase correct PWM mode is shown

on Figure 13-7. The TCNT0 value is in the timing diagram shown as a histogram for illustrating

the dual-slope operation. The diagram includes non-inverted and inverted PWM outputs. The

small horizontal line marks on the TCNT0 slopes represent Compare Matches between OCR0x

and TCNT0.

Figure 13-7. Phase Correct PWM Mode, Timing Diagram

OCnx Interrupt Flag Set

OCRnx Update

TOVn Interrupt Flag Set

TCNTn

(COMnx1:0 = 2)

OCnx

(COMnx1:0 = 3)

OCnx

1

2

3

Period

The Timer/Counter Overflow Flag (TOV0) is set each time the counter reaches BOTTOM. The

Interrupt Flag can be used to generate an interrupt each time the counter reaches the BOTTOM

value.

In phase correct PWM mode, the compare unit allows generation of PWM waveforms on the

OC0x pins. Setting the COM0x1:0 bits to two will produce a non-inverted PWM. An inverted

PWM output can be generated by setting the COM0x1:0 to three: Setting the COM0A0 bits to

99

7766E–AVR–04/10

ATmega16U4/ATmega32U4

one allows the OC0A pin to toggle on Compare Matches if the WGM02 bit is set. This option is

not available for the OC0B pin (See Table 13-4 on page 103). The actual OC0x value will only

be visible on the port pin if the data direction for the port pin is set as output. The PWM wave-

form is generated by clearing (or setting) the OC0x Register at the Compare Match between

OCR0x and TCNT0 when the counter increments, and setting (or clearing) the OC0x Register at

Compare Match between OCR0x and TCNT0 when the counter decrements. The PWM fre-

quency for the output when using phase correct PWM can be calculated by the following

equation:

f

clk_I/O

f

= -----------------

OCnxPCPWM

N ⋅ 510

The N variable represents the prescaler factor (1, 8, 64, 256, or 1024).

The extreme values for the OCR0A Register represent special cases when generating a PWM

waveform output in the phase correct PWM mode. If the OCR0A is set equal to BOTTOM, the

output will be continuously low and if set equal to MAX the output will be continuously high for

non-inverted PWM mode. For inverted PWM the output will have the opposite logic values.

At the very start of period 2 in Figure 13-7 OCnx has a transition from high to low even though

there is no Compare Match. The point of this transition is to guarantee symmetry around BOT-

TOM. There are two cases that give a transition without Compare Match.

• OCR0A changes its value from MAX, like in Figure 13-7. When the OCR0A value is MAX the

OCn pin value is the same as the result of a down-counting Compare Match. To ensure

symmetry around BOTTOM the OCn value at MAX must correspond to the result of an up-

counting Compare Match.

• The timer starts counting from a value higher than the one in OCR0A, and for that reason

misses the Compare Match and hence the OCn change that would have happened on the

way up.

13.7 Timer/Counter Timing Diagrams

The Timer/Counter is a synchronous design and the timer clock (clk_T0) is therefore shown as a

clock enable signal in the following figures. The figures include information on when Interrupt

Flags are set. Figure 13-8 contains timing data for basic Timer/Counter operation. The figure

shows the count sequence close to the MAX value in all modes other than phase correct PWM

mode.

Figure 13-8. Timer/Counter Timing Diagram, no Prescaling

clk_I/O

clk_Tn

(clk_I/O/1)

TCNTn

TOVn

MAX - 1

MAX

BOTTOM

BOTTOM + 1

Figure 13-9 shows the same timing data, but with the prescaler enabled.

100

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 13-9. Timer/Counter Timing Diagram, with Prescaler (f_{clk_I/O}/8)

clk_I/O

clk_Tn

(clk_I/O/8)

TCNTn

TOVn

MAX - 1

MAX

BOTTOM

BOTTOM + 1

Figure 13-10 shows the setting of OCF0B in all modes and OCF0A in all modes except CTC

mode and PWM mode, where OCR0A is TOP.

Figure 13-10. Timer/Counter Timing Diagram, Setting of OCF0x, with Prescaler (f_{clk_I/O}/8)

clk_I/O

clk_Tn

(clk_I/O/8)

TCNTn

OCRnx

OCFnx

OCRnx - 1

OCRnx

OCRnx + 1

OCRnx + 2

OCRnx Value

Figure 13-11 shows the setting of OCF0A and the clearing of TCNT0 in CTC mode and fast

PWM mode where OCR0A is TOP.

Figure 13-11. Timer/Counter Timing Diagram, Clear Timer on Compare Match mode, with Pres-

caler (f_{clk_I/O}/8)

clk_I/O

clk_Tn

(clk_I/O/8)

TCNTn

(CTC)

TOP - 1

TOP

BOTTOM

BOTTOM + 1

OCRnx

TOP

OCFnx

101

7766E–AVR–04/10

ATmega16U4/ATmega32U4

13.8 8-bit Timer/Counter Register Description

13.8.1

Timer/Counter Control Register A – TCCR0A

Bit

7

6

5

4

3

–

2

–

1

0

COM0A

1

COM0A

0

COM0B

1

COM0B

0

WGM0

1

WGM0

0

TCCR0A

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R

0

R

0

R/W

0

R/W

0

• Bits 7:6 – COM01A:0: Compare Match Output A Mode

These bits control the Output Compare pin (OC0A) behavior. If one or both of the COM0A1:0

bits are set, the OC0A output overrides the normal port functionality of the I/O pin it is connected

to. However, note that the Data Direction Register (DDR) bit corresponding to the OC0A pin

must be set in order to enable the output driver.

When OC0A is connected to the pin, the function of the COM0A1:0 bits depends on the

WGM02:0 bit setting. Table 13-2 shows the COM0A1:0 bit functionality when the WGM02:0 bits

are set to a normal or CTC mode (non-PWM).

Table 13-2. Compare Output Mode, non-PWM Mode

COM0A1

COM0A0

Description

0

1

0

1

0

1

Normal port operation, OC0A disconnected.

Toggle OC0A on Compare Match

Clear OC0A on Compare Match

Set OC0A on Compare Match

Table 13-3 shows the COM0A1:0 bit functionality when the WGM01:0 bits are set to fast PWM

mode.

Table 13-3. Compare Output Mode, Fast PWM Mode⁽¹⁾

COM0A1

COM0A0

Description

0

Normal port operation, OC0A disconnected.

WGM02 = 0: Normal Port Operation, OC0A Disconnected.

WGM02 = 1: Toggle OC0A on Compare Match.

0

1

0

1

Clear OC0A on Compare Match, set OC0A at TOP

Set OC0A on Compare Match, clear OC0A at TOP

Note:

1. A special case occurs when OCR0A equals TOP and COM0A1 is set. In this case, the Com-

pare Match is ignored, but the set or clear is done at TOP. See “Fast PWM Mode” on page 97

for more details.

102

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 13-4 shows the COM0A1:0 bit functionality when the WGM02:0 bits are set to phase cor-

rect PWM mode.

Table 13-4. Compare Output Mode, Phase Correct PWM Mode⁽¹⁾

COM0A1

COM0A0

Description

0

Normal port operation, OC0A disconnected.

WGM02 = 0: Normal Port Operation, OC0A Disconnected.

WGM02 = 1: Toggle OC0A on Compare Match.

0

1

0

1

Clear OC0A on Compare Match when up-counting. Set OC0A on

Compare Match when down-counting.

Set OC0A on Compare Match when up-counting. Clear OC0A on

Compare Match when down-counting.

Note:

1. A special case occurs when OCR0A equals TOP and COM0A1 is set. In this case, the Com-

pare Match is ignored, but the set or clear is done at TOP. See “Phase Correct PWM Mode” on

page 99 for more details.

• Bits 5:4 – COM0B1:0: Compare Match Output B Mode

These bits control the Output Compare pin (OC0B) behavior. If one or both of the COM0B1:0

bits are set, the OC0B output overrides the normal port functionality of the I/O pin it is connected

to. However, note that the Data Direction Register (DDR) bit corresponding to the OC0B pin

must be set in order to enable the output driver.

When OC0B is connected to the pin, the function of the COM0B1:0 bits depends on the

WGM02:0 bit setting. Table 13-2 shows the COM0A1:0 bit functionality when the WGM02:0 bits

are set to a normal or CTC mode (non-PWM).

Table 13-5. Compare Output Mode, non-PWM Mode

COM01

COM00

Description

0

1

0

1

0

1

Normal port operation, OC0B disconnected.

Toggle OC0B on Compare Match

Clear OC0B on Compare Match

Set OC0B on Compare Match

Table 13-3 shows the COM0B1:0 bit functionality when the WGM02:0 bits are set to fast PWM

mode.

Table 13-6. Compare Output Mode, Fast PWM Mode⁽¹⁾

COM01

COM00

Description

0

1

0

1

0

1

Normal port operation, OC0B disconnected.

Reserved

Clear OC0B on Compare Match, set OC0B at TOP

Set OC0B on Compare Match, clear OC0B at TOP

Note:

1. A special case occurs when OCR0B equals TOP and COM0B1 is set. In this case, the Com-

pare Match is ignored, but the set or clear is done at TOP. See “Fast PWM Mode” on page 97

for more details.

103

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 13-4 shows the COM0B1:0 bit functionality when the WGM02:0 bits are set to phase cor-

rect PWM mode.

Table 13-7. Compare Output Mode, Phase Correct PWM Mode⁽¹⁾

COM0A1

COM0A0

Description

0

1

Normal port operation, OC0B disconnected.

Reserved

Clear OC0B on Compare Match when up-counting. Set OC0B on

Compare Match when down-counting.

1

0

1

Set OC0B on Compare Match when up-counting. Clear OC0B on

Compare Match when down-counting.

Note:

1. A special case occurs when OCR0B equals TOP and COM0B1 is set. In this case, the Com-

pare Match is ignored, but the set or clear is done at TOP. See “Phase Correct PWM Mode” on

page 99 for more details.

• Bits 3, 2 – Res: Reserved Bits

These bits are reserved bits in the ATmega16U4/ATmega32U4 and will always read as zero.

• Bits 1:0 – WGM01:0: Waveform Generation Mode

Combined with the WGM02 bit found in the TCCR0B Register, these bits control the counting

sequence of the counter, the source for maximum (TOP) counter value, and what type of wave-

form generation to be used, see Table 13-8. Modes of operation supported by the Timer/Counter

unit are: Normal mode (counter), Clear Timer on Compare Match (CTC) mode, and two types of

Pulse Width Modulation (PWM) modes (see “Modes of Operation” on page 96).

Table 13-8. Waveform Generation Mode Bit Description

Timer/Counter

Mode of

Update of

OCRx at

TOV Flag

Mode WGM2 WGM1 WGM0 Operation

TOP

Set on⁽¹⁾⁽²⁾

0

1

0

1

Normal

0xFF

Immediate

TOP

MAX

PWM, Phase

Correct

0xFF

BOTTOM

2

3

4

0

1

0

1

0

CTC

OCRA Immediate

MAX

–

Fast PWM

Reserved

0xFF

–

TOP

–

PWM, Phase

Correct

5

1

0

1

OCRA

TOP

BOTTOM

6

7

1

0

1

Reserved

Fast PWM

–

OCRA

TOP

Notes: 1. MAX

= 0xFF

2. BOTTOM = 0x00

104

7766E–AVR–04/10

ATmega16U4/ATmega32U4

13.8.2

Timer/Counter Control Register B – TCCR0B

Bit

7

6

5

–

R

0

4

–

R

0

3

2

1

0

FOC0A

FOC0B

WGM02

R/W

0

CS02

R/W

0

CS01

R/W

0

CS00

R/W

0

TCCR0B

Read/Write

Initial Value

W

0

W

0

• Bit 7 – FOC0A: Force Output Compare A

The FOC0A bit is only active when the WGM bits specify a non-PWM mode.

However, for ensuring compatibility with future devices, this bit must be set to zero when

TCCR0B is written when operating in PWM mode. When writing a logical one to the FOC0A bit,

an immediate Compare Match is forced on the Waveform Generation unit. The OC0A output is

changed according to its COM0A1:0 bits setting. Note that the FOC0A bit is implemented as a

strobe. Therefore it is the value present in the COM0A1:0 bits that determines the effect of the

forced compare.

A FOC0A strobe will not generate any interrupt, nor will it clear the timer in CTC mode using

OCR0A as TOP.

The FOC0A bit is always read as zero.

• Bit 6 – FOC0B: Force Output Compare B

The FOC0B bit is only active when the WGM bits specify a non-PWM mode.

However, for ensuring compatibility with future devices, this bit must be set to zero when

TCCR0B is written when operating in PWM mode. When writing a logical one to the FOC0B bit,

an immediate Compare Match is forced on the Waveform Generation unit. The OC0B output is

changed according to its COM0B1:0 bits setting. Note that the FOC0B bit is implemented as a

strobe. Therefore it is the value present in the COM0B1:0 bits that determines the effect of the

forced compare.

A FOC0B strobe will not generate any interrupt, nor will it clear the timer in CTC mode using

OCR0B as TOP.

The FOC0B bit is always read as zero.

• Bits 5:4 – Res: Reserved Bits

These bits are reserved bits and will always read as zero.

• Bit 3 – WGM02: Waveform Generation Mode

See the description in the “Timer/Counter Control Register A – TCCR0A” on page 102.

• Bits 2:0 – CS02:0: Clock Select

The three Clock Select bits select the clock source to be used by the Timer/Counter.

Table 13-9. Clock Select Bit Description

CS02

CS01

CS00 Description

0

1

0

1

0

1

No clock source (Timer/Counter stopped)

clk_I/O/(No prescaling)

clk_I/O/8 (From prescaler)

clk_I/O/64 (From prescaler)

105

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 13-9. Clock Select Bit Description (Continued)

CS02

CS01

CS00 Description

1

0

1

0

1

0

1

clk_I/O/256 (From prescaler)

clk_I/O/1024 (From prescaler)

External clock source on T0 pin. Clock on falling edge.

External clock source on T0 pin. Clock on rising edge.

If external pin modes are used for the Timer/Counter0, transitions on the T0 pin will clock the

counter even if the pin is configured as an output. This feature allows software control of the

counting.

13.8.3

Timer/Counter Register – TCNT0

Bit

7

6

5

4

3

2

1

0

TCNT0[7:0]

TCNT0

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

The Timer/Counter Register gives direct access, both for read and write operations, to the

Timer/Counter unit 8-bit counter. Writing to the TCNT0 Register blocks (removes) the Compare

Match on the following timer clock. Modifying the counter (TCNT0) while the counter is running,

introduces a risk of missing a Compare Match between TCNT0 and the OCR0x Registers.

13.8.4

13.8.5

13.8.6

Output Compare Register A – OCR0A

Bit

7

6

5

4

3

2

1

0

OCR0A[7:0]

OCR0A

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

The Output Compare Register A contains an 8-bit value that is continuously compared with the

counter value (TCNT0). A match can be used to generate an Output Compare interrupt, or to

generate a waveform output on the OC0A pin.

Output Compare Register B – OCR0B

Bit

7

6

5

4

3

2

1

0

OCR0B[7:0]

OCR0B

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

The Output Compare Register B contains an 8-bit value that is continuously compared with the

counter value (TCNT0). A match can be used to generate an Output Compare interrupt, or to

generate a waveform output on the OC0B pin.

Timer/Counter Interrupt Mask Register – TIMSK0

Bit

7

6

5

4

–

R

0

3

–

R

0

2

1

0

–

OCIE0B

R/W

0

OCIE0A

R/W

0

TOIE0

R/W

0

TIMSK0

Read/Write

Initial Value

R

0

R

0

R

0

• Bits 7..3, 0 – Res: Reserved Bits

These bits are reserved bits and will always read as zero.

106

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 2 – OCIE0B: Timer/Counter Output Compare Match B Interrupt Enable

When the OCIE0B bit is written to one, and the I-bit in the Status Register is set, the

Timer/Counter Compare Match B interrupt is enabled. The corresponding interrupt is executed if

a Compare Match in Timer/Counter occurs, i.e., when the OCF0B bit is set in the Timer/Counter

Interrupt Flag Register – TIFR0.

• Bit 1 – OCIE0A: Timer/Counter0 Output Compare Match A Interrupt Enable

When the OCIE0A bit is written to one, and the I-bit in the Status Register is set, the

Timer/Counter0 Compare Match A interrupt is enabled. The corresponding interrupt is executed

if a Compare Match in Timer/Counter0 occurs, i.e., when the OCF0A bit is set in the

Timer/Counter 0 Interrupt Flag Register – TIFR0.

• Bit 0 – TOIE0: Timer/Counter0 Overflow Interrupt Enable

When the TOIE0 bit is written to one, and the I-bit in the Status Register is set, the

Timer/Counter0 Overflow interrupt is enabled. The corresponding interrupt is executed if an

overflow in Timer/Counter0 occurs, i.e., when the TOV0 bit is set in the Timer/Counter 0 Inter-

rupt Flag Register – TIFR0.

13.8.7

Timer/Counter 0 Interrupt Flag Register – TIFR0

Bit

7

6

5

4

–

R

0

3

–

R

0

2

1

0

–

OCF0B

R/W

0

OCF0A

R/W

0

TOV0

R/W

0

TIFR0

Read/Write

Initial Value

R

0

R

0

R

0

• Bits 7..3, 0 – Res: Reserved Bits

These bits are reserved bits in the ATmega16U4/ATmega32U4 and will always read as zero.

• Bit 2 – OCF0B: Timer/Counter 0 Output Compare B Match Flag

The OCF0B bit is set when a Compare Match occurs between the Timer/Counter and the data in

OCR0B – Output Compare Register0 B. OCF0B is cleared by hardware when executing the cor-

responding interrupt handling vector. Alternatively, OCF0B is cleared by writing a logic one to

the flag. When the I-bit in SREG, OCIE0B (Timer/Counter Compare B Match Interrupt Enable),

and OCF0B are set, the Timer/Counter Compare Match Interrupt is executed.

• Bit 1 – OCF0A: Timer/Counter 0 Output Compare A Match Flag

The OCF0A bit is set when a Compare Match occurs between the Timer/Counter0 and the data

in OCR0A – Output Compare Register0. OCF0A is cleared by hardware when executing the cor-

responding interrupt handling vector. Alternatively, OCF0A is cleared by writing a logic one to

the flag. When the I-bit in SREG, OCIE0A (Timer/Counter0 Compare Match Interrupt Enable),

and OCF0A are set, the Timer/Counter0 Compare Match Interrupt is executed.

• Bit 0 – TOV0: Timer/Counter0 Overflow Flag

The bit TOV0 is set when an overflow occurs in Timer/Counter0. TOV0 is cleared by hardware

when executing the corresponding interrupt handling vector. Alternatively, TOV0 is cleared by

writing a logic one to the flag. When the SREG I-bit, TOIE0 (Timer/Counter0 Overflow Interrupt

Enable), and TOV0 are set, the Timer/Counter0 Overflow interrupt is executed.

The setting of this flag is dependent of the WGM02:0 bit setting. Refer to Table 13-8, “Waveform

Generation Mode Bit Description” on page 104.

107

7766E–AVR–04/10

ATmega16U4/ATmega32U4

14. 16-bit Timers/Counters (Timer/Counter1 and Timer/Counter3)

The 16-bit Timer/Counter unit allows accurate program execution timing (event management),

wave generation, and signal timing measurement. The main features are:

• True 16-bit Design (i.e., Allows 16-bit PWM)

• Three independent Output Compare Units

• Double Buffered Output Compare Registers

• One Input Capture Unit

• Input Capture Noise Canceler

• Clear Timer on Compare Match (Auto Reload)

• Glitch-free, Phase Correct Pulse Width Modulator (PWM)

• Variable PWM Period

• Frequency Generator

• External Event Counter

• Ten independent interrupt sources (TOV1, OCF1A, OCF1B, OCF1C, ICF1, TOV3, OCF3A, OCF3B,

OCF3C and ICF3)

14.1 Overview

Most register and bit references in this section are written in general form. A lower case “n”

replaces the Timer/Counter number, and a) lower case “x” replaces the Output Compare unit

channel. However, when using the register or bit defines in a program, the precise form must be

used, i.e., TCNT1 for accessing Timer/Counter1 counter value and so on.

A simplified block diagram of the 16-bit Timer/Counter is shown in Figure 14-1. For the actual

placement of I/O pins, see “Pinout ATmega16U4/ATmega32U4” on page 3. CPU accessible I/O

Registers, including I/O bits and I/O pins, are shown in bold. The device-specific I/O Register

and bit locations are listed in the “16-bit Timers/Counters (Timer/Counter1 and Timer/Counter3)”

on page 108.

The Power Reduction Timer/Counter1 bit, PRTIM1, in “Power Reduction Register 0 - PRR0” on

page 46 must be written to zero to enable Timer/Counter1 module.

The Power Reduction Timer/Counter3 bit, PRTIM3, in “Power Reduction Register 1 - PRR1” on

page 46 must be written to zero to enable Timer/Counter3 module.

108

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 14-1. 16-bit Timer/Counter Block Diagram⁽¹⁾

Count

TOVn

(Int.Req.)

Clear

Control Logic

Clock Select

Direction

TCLK

(2)

Tn

Edge

Detector

TOP BOTTOM

( From Prescaler )

Timer/Counter

TCNTn

=

= 0

OCFnA

(Int.Req.)

Waveform

Generation

OCnA

OCnB

OCnC

=

OCRnA

OCFnB

(Int.Req.)

Fixed

TOP

Values

Waveform

Generation

=

OCRnB

OCFnC

(Int.Req.)

Waveform

Generation

=

OCRnC

( From Analog

Comparator Ouput )

ICFn (Int.Req.)

Edge

Detector

Noise

Canceler

ICRn

ICPn

TCCRnA

TCCRnB

TCCRnC

Note:

1. Refer to “Pinout ATmega16U4/ATmega32U4” on page 3, Table 10-3 on page 72, and Table

10-6 on page 75 for Timer/Counter1 and 3 and 3 pin placement and description.

2. Tn only refers to T1 since T3 input is not available on the ATmega16U4/ATmega32U4

product.

14.1.1

Registers

The Timer/Counter (TCNTn), Output Compare Registers (OCRnA/B/C), and Input Capture Reg-

ister (ICRn) are all 16-bit registers. Special procedures must be followed when accessing the 16-

bit registers. These procedures are described in the section “Accessing 16-bit Registers” on

page 110. The Timer/Counter Control Registers (TCCRnA/B/C) are 8-bit registers and have no

CPU access restrictions. Interrupt requests (shorten as Int.Req.) signals are all visible in the

Timer Interrupt Flag Register (TIFRn). All interrupts are individually masked with the Timer Inter-

rupt Mask Register (TIMSKn). TIFRn and TIMSKn are not shown in the figure since these

registers are shared by other timer units.

The Timer/Counter can be clocked internally, via the prescaler, or by an external clock source on

the Tn pin. The Clock Select logic block controls which clock source and edge the Timer/Counter

uses to increment (or decrement) its value. The Timer/Counter is inactive when no clock source

is selected. The output from the clock select logic is referred to as the timer clock (clk ).

n

T

109

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The double buffered Output Compare Registers (OCRnA/B/C) are compared with the

Timer/Counter value at all time. The result of the compare can be used by the Waveform Gener-

ator to generate a PWM or variable frequency output on the Output Compare pin (OCnA/B/C).

See “Output Compare Units” on page 117.. The compare match event will also set the Compare

Match Flag (OCFnA/B/C) which can be used to generate an Output Compare interrupt request.

The Input Capture Register can capture the Timer/Counter value at a given external (edge trig-

gered) event on either the Input Capture pin (ICPn) or on the Analog Comparator pins (See

“Analog Comparator” on page 289.) The Input Capture unit includes a digital filtering unit (Noise

Canceler) for reducing the chance of capturing noise spikes.

The TOP value, or maximum Timer/Counter value, can in some modes of operation be defined

by either the OCRnA Register, the ICRn Register, or by a set of fixed values. When using

OCRnA as TOP value in a PWM mode, the OCRnA Register can not be used for generating a

PWM output. However, the TOP value will in this case be double buffered allowing the TOP

value to be changed in run time. If a fixed TOP value is required, the ICRn Register can be used

as an alternative, freeing the OCRnA to be used as PWM output.

14.1.2

Definitions

The following definitions are used extensively throughout the document:

Table 14-1.

BOTTOM

The counter reaches the BOTTOM when it becomes 0x0000.

MAX

The counter reaches its MAXimum when it becomes 0xFFFF (decimal 65535).

The counter reaches the TOP when it becomes equal to the highest value in the

count sequence. The TOP value can be assigned to be one of the fixed values:

0x00FF, 0x01FF, or 0x03FF, or to the value stored in the OCRnA or ICRn

Register. The assignment is dependent of the mode of operation.

TOP

14.2 Accessing 16-bit Registers

The TCNTn, OCRnA/B/C, and ICRn are 16-bit registers that can be accessed by the AVR CPU

via the 8-bit data bus. The 16-bit register must be byte accessed using two read or write opera-

tions. Each 16-bit timer has a single 8-bit register for temporary storing of the high byte of the 16-

bit access. The same Temporary Register is shared between all 16-bit registers within each 16-

bit timer. Accessing the low byte triggers the 16-bit read or write operation. When the low byte of

a 16-bit register is written by the CPU, the high byte stored in the Temporary Register, and the

low byte written are both copied into the 16-bit register in the same clock cycle. When the low

byte of a 16-bit register is read by the CPU, the high byte of the 16-bit register is copied into the

Temporary Register in the same clock cycle as the low byte is read.

Not all 16-bit accesses uses the Temporary Register for the high byte. Reading the OCRnA/B/C

16-bit registers does not involve using the Temporary Register.

To do a 16-bit write, the high byte must be written before the low byte. For a 16-bit read, the low

byte must be read before the high byte.

The following code examples show how to access the 16-bit timer registers assuming that no

interrupts updates the temporary register. The same principle can be used directly for accessing

the OCRnA/B/C and ICRn Registers. Note that when using “C”, the compiler handles the 16-bit

access.

110

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Assembly Code Examples⁽¹⁾

...

; Set TCNTn to 0x01FF

ldir17,0x01

ldir16,0xFF

outTCNTnH,r17

outTCNTnL,r16

; Read TCNTn into r17:r16

in r16,TCNTnL

in r17,TCNTnH

...

C Code Examples⁽¹⁾

unsigned int i;

...

/* Set TCNTn to 0x01FF */

TCNTn = 0x1FF;

/* Read TCNTn into i */

i = TCNTn;

...

Note:

1. See “Code Examples” on page 8.

The assembly code example returns the TCNTn value in the r17:r16 register pair.

It is important to notice that accessing 16-bit registers are atomic operations. If an interrupt

occurs between the two instructions accessing the 16-bit register, and the interrupt code

updates the temporary register by accessing the same or any other of the 16-bit Timer Regis-

ters, then the result of the access outside the interrupt will be corrupted. Therefore, when both

the main code and the interrupt code update the temporary register, the main code must disable

the interrupts during the 16-bit access.

The following code examples show how to do an atomic read of the TCNTn Register contents.

Reading any of the OCRnA/B/C or ICRn Registers can be done by using the same principle.

111

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Assembly Code Example⁽¹⁾

TIM16_ReadTCNTn:

; Save global interrupt flag

in r18,SREG

; Disable interrupts

cli

; Read TCNTn into r17:r16

in r16,TCNTnL

in r17,TCNTnH

; Restore global interrupt flag

outSREG,r18

ret

C Code Example⁽¹⁾

unsigned int TIM16_ReadTCNTn( void )

{

unsigned char sreg;

unsigned int i;

/* Save global interrupt flag */

sreg = SREG;

/* Disable interrupts */

__disable_interrupt();

/* Read TCNTn into i */

i = TCNTn;

/* Restore global interrupt flag */

SREG = sreg;

return i;

}

Note:

1. See “Code Examples” on page 8.

The assembly code example returns the TCNTn value in the r17:r16 register pair.

112

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The following code examples show how to do an atomic write of the TCNTn Register contents.

Writing any of the OCRnA/B/C or ICRn Registers can be done by using the same principle.

Assembly Code Example⁽¹⁾

TIM16_WriteTCNTn:

; Save global interrupt flag

in r18,SREG

; Disable interrupts

cli

; Set TCNTn to r17:r16

outTCNTnH,r17

outTCNTnL,r16

; Restore global interrupt flag

outSREG,r18

ret

C Code Example⁽¹⁾

void TIM16_WriteTCNTn( unsigned int i )

{

unsigned char sreg;

unsigned int i;

/* Save global interrupt flag */

sreg = SREG;

/* Disable interrupts */

__disable_interrupt();

/* Set TCNTn to i */

TCNTn = i;

/* Restore global interrupt flag */

SREG = sreg;

}

Note:

1. See “Code Examples” on page 8.

The assembly code example requires that the r17:r16 register pair contains the value to be writ-

ten to TCNTn.

14.2.1

Reusing the Temporary High Byte Register

If writing to more than one 16-bit register where the high byte is the same for all registers written,

then the high byte only needs to be written once. However, note that the same rule of atomic

operation described previously also applies in this case.

14.3 Timer/Counter Clock Sources

The Timer/Counter can be clocked by an internal or an external clock source. The clock source

is selected by the Clock Select logic which is controlled by the Clock Select (CSn2:0) bits

located in the Timer/Counter control Register B (TCCRnB). For details on clock sources and

prescaler, see “Timer/Counter0, Timer/Counter1, and Timer/Counter3 Prescalers” on page 89.

113

7766E–AVR–04/10

ATmega16U4/ATmega32U4

14.4 Counter Unit

The main part of the 16-bit Timer/Counter is the programmable 16-bit bi-directional counter unit.

Figure 14-2 shows a block diagram of the counter and its surroundings.

Figure 14-2. Counter Unit Block Diagram

DATA BUS (8-bit)

TOVn

(Int.Req.)

TEMP (8-bit)

Clock Select

Count

Clear

Edge

Detector

Tn

TCNTnH (8-bit)

TCNTnL (8-bit)

clk_Tn

Control Logic

Direction

TCNTn (16-bit Counter)

( From Prescaler )

TOP

BOTTOM

Signal description (internal signals):

Count

Increment or decrement TCNTn by 1.

Direction

Clear

Select between increment and decrement.

Clear TCNTn (set all bits to zero).

clk_T

Timer/Counter clock.

n

TOP

Signalize that TCNTn has reached maximum value.

BOTTOM

Signalize that TCNTn has reached minimum value (zero).

The 16-bit counter is mapped into two 8-bit I/O memory locations: Counter High (TCNTnH) con-

taining the upper eight bits of the counter, and Counter Low (TCNTnL) containing the lower eight

bits. The TCNTnH Register can only be indirectly accessed by the CPU. When the CPU does an

access to the TCNTnH I/O location, the CPU accesses the high byte temporary register (TEMP).

The temporary register is updated with the TCNTnH value when the TCNTnL is read, and

TCNTnH is updated with the temporary register value when TCNTnL is written. This allows the

CPU to read or write the entire 16-bit counter value within one clock cycle via the 8-bit data bus.

It is important to notice that there are special cases of writing to the TCNTn Register when the

counter is counting that will give unpredictable results. The special cases are described in the

sections where they are of importance.

Depending on the mode of operation used, the counter is cleared, incremented, or decremented

at each timer clock (clk ). The clk _ncan be generated from an external or internal clock source,

n

T

selected by the Clock Select bits (CSn2:0). When no clock source is selected (CSn2:0 = 0) the

timer is stopped. However, the TCNTn value can be accessed by the CPU, independent of

whether clk_Tis present or not. A CPU write overrides (has priority over) all counter clear or

n

count operations.

The counting sequence is determined by the setting of the Waveform Generation mode bits

(WGMn3:0) located in the Timer/Counter Control Registers A and B (TCCRnA and TCCRnB).

There are close connections between how the counter behaves (counts) and how waveforms

are generated on the Output Compare outputs OCnx. For more details about advanced counting

sequences and waveform generation, see “Modes of Operation” on page 120.

114

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The Timer/Counter Overflow Flag (TOVn) is set according to the mode of operation selected by

the WGMn3:0 bits. TOVn can be used for generating a CPU interrupt.

14.5 Input Capture Unit

The Timer/Counter incorporates an input capture unit that can capture external events and give

them a time-stamp indicating time of occurrence. The external signal indicating an event, or mul-

tiple events, can be applied via the ICPn pin or alternatively, for the Timer/Counter1 only, via the

Analog Comparator unit. The time-stamps can then be used to calculate frequency, duty-cycle,

and other features of the signal applied. Alternatively the time-stamps can be used for creating a

log of the events.

The Input Capture unit is illustrated by the block diagram shown in Figure 14-3. The elements of

the block diagram that are not directly a part of the input capture unit are gray shaded. The small

“n” in register and bit names indicates the Timer/Counter number.

Figure 14-3. Input Capture Unit Block Diagram

DATA BUS (8-bit)

TEMP (8-bit)

ICRnH (8-bit)

ICRnL (8-bit)

TCNTnH (8-bit)

TCNTnL (8-bit)

ICRn (16-bit Register)

TCNTn (16-bit Counter)

WRITE

ACO*

ACIC*

ICNC

ICES

Analog

Comparator

Noise

Canceler

Edge

Detector

ICFn (Int.Req.)

ICPn

Note:

The Analog Comparator Output (ACO) can only trigger the Timer/Counter1 ICP – not

Timer/Counter3, 4 or 5.

When a change of the logic level (an event) occurs on the Input Capture Pin (ICPn), alternatively

on the analog Comparator output (ACO), and this change confirms to the setting of the edge

detector, a capture will be triggered. When a capture is triggered, the 16-bit value of the counter

(TCNTn) is written to the Input Capture Register (ICRn). The Input Capture Flag (ICFn) is set at

the same system clock as the TCNTn value is copied into ICRn Register. If enabled (TICIEn =

1), the input capture flag generates an input capture interrupt. The ICFn flag is automatically

cleared when the interrupt is executed. Alternatively the ICFn flag can be cleared by software by

writing a logical one to its I/O bit location.

115

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Reading the 16-bit value in the Input Capture Register (ICRn) is done by first reading the low

byte (ICRnL) and then the high byte (ICRnH). When the low byte is read the high byte is copied

into the high byte Temporary Register (TEMP). When the CPU reads the ICRnH I/O location it

will access the TEMP Register.

The ICRn Register can only be written when using a Waveform Generation mode that utilizes

the ICRn Register for defining the counter’s TOP value. In these cases the Waveform Genera-

tion mode (WGMn3:0) bits must be set before the TOP value can be written to the ICRn

Register. When writing the ICRn Register the high byte must be written to the ICRnH I/O location

before the low byte is written to ICRnL.

For more information on how to access the 16-bit registers refer to “Accessing 16-bit Registers”

on page 110.

14.5.1

Input Capture Trigger Source

The main trigger source for the input capture unit is the Input Capture Pin (ICPn).

Timer/Counter1 can alternatively use the analog comparator output as trigger source for the

input capture unit. The Analog Comparator is selected as trigger source by setting the analog

Comparator Input Capture (ACIC) bit in the Analog Comparator Control and Status Register

(ACSR). Be aware that changing trigger source can trigger a capture. The input capture flag

must therefore be cleared after the change.

Both the Input Capture Pin (ICPn) and the Analog Comparator output (ACO) inputs are sampled

using the same technique as for the Tn pin (Figure 12-1 on page 89). The edge detector is also

identical. However, when the noise canceler is enabled, additional logic is inserted before the

edge detector, which increases the delay by four system clock cycles. Note that the input of the

noise canceler and edge detector is always enabled unless the Timer/Counter is set in a Wave-

form Generation mode that uses ICRn to define TOP.

An input capture can be triggered by software by controlling the port of the ICPn pin.

14.5.2

Noise Canceler

The noise canceler improves noise immunity by using a simple digital filtering scheme. The

noise canceler input is monitored over four samples, and all four must be equal for changing the

output that in turn is used by the edge detector.

The noise canceler is enabled by setting the Input Capture Noise Canceler (ICNCn) bit in

Timer/Counter Control Register B (TCCRnB). When enabled the noise canceler introduces addi-

tional four system clock cycles of delay from a change applied to the input, to the update of the

ICRn Register. The noise canceler uses the system clock and is therefore not affected by the

prescaler.

14.5.3

Using the Input Capture Unit

The main challenge when using the Input Capture unit is to assign enough processor capacity

for handling the incoming events. The time between two events is critical. If the processor has

not read the captured value in the ICRn Register before the next event occurs, the ICRn will be

overwritten with a new value. In this case the result of the capture will be incorrect.

When using the Input Capture interrupt, the ICRn Register should be read as early in the inter-

rupt handler routine as possible. Even though the Input Capture interrupt has relatively high

priority, the maximum interrupt response time is dependent on the maximum number of clock

cycles it takes to handle any of the other interrupt requests.

116

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Using the Input Capture unit in any mode of operation when the TOP value (resolution) is

actively changed during operation, is not recommended.

Measurement of an external signal’s duty cycle requires that the trigger edge is changed after

each capture. Changing the edge sensing must be done as early as possible after the ICRn

Register has been read. After a change of the edge, the Input Capture Flag (ICFn) must be

cleared by software (writing a logical one to the I/O bit location). For measuring frequency only,

the clearing of the ICFn Flag is not required (if an interrupt handler is used).

14.6 Output Compare Units

The 16-bit comparator continuously compares TCNTn with the Output Compare Register

(OCRnx). If TCNT equals OCRnx the comparator signals a match. A match will set the Output

Compare Flag (OCFnx) at the next timer clock cycle. If enabled (OCIEnx = 1), the Output Com-

pare Flag generates an Output Compare interrupt. The OCFnx Flag is automatically cleared

when the interrupt is executed. Alternatively the OCFnx Flag can be cleared by software by writ-

ing a logical one to its I/O bit location. The Waveform Generator uses the match signal to

generate an output according to operating mode set by the Waveform Generation mode

(WGMn3:0) bits and Compare Output mode (COMnx1:0) bits. The TOP and BOTTOM signals

are used by the Waveform Generator for handling the special cases of the extreme values in

some modes of operation (See “Modes of Operation” on page 120.)

A special feature of Output Compare unit A allows it to define the Timer/Counter TOP value (i.e.,

counter resolution). In addition to the counter resolution, the TOP value defines the period time

for waveforms generated by the Waveform Generator.

Figure 14-4 shows a block diagram of the Output Compare unit. The small “n” in the register and

bit names indicates the device number (n = n for Timer/Counter n), and the “x” indicates Output

Compare unit (A/B/C). The elements of the block diagram that are not directly a part of the Out-

put Compare unit are gray shaded.

Figure 14-4. Output Compare Unit, Block Diagram

DATA BUS (8-bit)

TEMP (8-bit)

OCRnxH Buf. (8-bit)

OCRnxL Buf. (8-bit)

TCNTnH (8-bit)

TCNTnL (8-bit)

OCRnx Buffer (16-bit Register)

TCNTn (16-bit Counter)

OCRnxH (8-bit)

OCRnxL (8-bit)

OCRnx (16-bit Register)

=

(16-bit Comparator )

OCFnx (Int.Req.)

TOP

OCnx

Waveform Generator

BOTTOM

WGMn3:0

COMnx1:0

117

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The OCRnx Register is double buffered when using any of the twelve Pulse Width Modulation

(PWM) modes. For the Normal and Clear Timer on Compare (CTC) modes of operation, the

double buffering is disabled. The double buffering synchronizes the update of the OCRnx Com-

pare Register to either TOP or BOTTOM of the counting sequence. The synchronization

prevents the occurrence of odd-length, non-symmetrical PWM pulses, thereby making the out-

put glitch-free.

The OCRnx Register access may seem complex, but this is not case. When the double buffering

is enabled, the CPU has access to the OCRnx Buffer Register, and if double buffering is dis-

abled the CPU will access the OCRnx directly. The content of the OCR1x (Buffer or Compare)

Register is only changed by a write operation (the Timer/Counter does not update this register

automatically as the TCNT1 and ICR1 Register). Therefore OCR1x is not read via the high byte

temporary register (TEMP). However, it is a good practice to read the low byte first as when

accessing other 16-bit registers. Writing the OCRnx Registers must be done via the TEMP Reg-

ister since the compare of all 16 bits is done continuously. The high byte (OCRnxH) has to be

written first. When the high byte I/O location is written by the CPU, the TEMP Register will be

updated by the value written. Then when the low byte (OCRnxL) is written to the lower eight bits,

the high byte will be copied into the upper 8-bits of either the OCRnx buffer or OCRnx Compare

Register in the same system clock cycle.

For more information of how to access the 16-bit registers refer to “Accessing 16-bit Registers”

on page 110.

14.6.1

Force Output Compare

In non-PWM Waveform Generation modes, the match output of the comparator can be forced by

writing a one to the Force Output Compare (FOCnx) bit. Forcing compare match will not set the

OCFnx Flag or reload/clear the timer, but the OCnx pin will be updated as if a real compare

match had occurred (the COMn1:0 bits settings define whether the OCnx pin is set, cleared or

toggled).

14.6.2

14.6.3

Compare Match Blocking by TCNTn Write

All CPU writes to the TCNTn Register will block any compare match that occurs in the next timer

clock cycle, even when the timer is stopped. This feature allows OCRnx to be initialized to the

same value as TCNTn without triggering an interrupt when the Timer/Counter clock is enabled.

Using the Output Compare Unit

Since writing TCNTn in any mode of operation will block all compare matches for one timer clock

cycle, there are risks involved when changing TCNTn when using any of the Output Compare

channels, independent of whether the Timer/Counter is running or not. If the value written to

TCNTn equals the OCRnx value, the compare match will be missed, resulting in incorrect wave-

form generation. Do not write the TCNTn equal to TOP in PWM modes with variable TOP

values. The compare match for the TOP will be ignored and the counter will continue to 0xFFFF.

Similarly, do not write the TCNTn value equal to BOTTOM when the counter is downcounting.

The setup of the OCnx should be performed before setting the Data Direction Register for the

port pin to output. The easiest way of setting the OCnx value is to use the Force Output Com-

pare (FOCnx) strobe bits in Normal mode. The OCnx Register keeps its value even when

changing between Waveform Generation modes.

Be aware that the COMnx1:0 bits are not double buffered together with the compare value.

Changing the COMnx1:0 bits will take effect immediately.

118

7766E–AVR–04/10

ATmega16U4/ATmega32U4

14.7 Compare Match Output Unit

The Compare Output mode (COMnx1:0) bits have two functions. The Waveform Generator uses

the COMnx1:0 bits for defining the Output Compare (OCnx) state at the next compare match.

Secondly the COMnx1:0 bits control the OCnx pin output source. Figure 14-5 shows a simplified

schematic of the logic affected by the COMnx1:0 bit setting. The I/O Registers, I/O bits, and I/O

pins in the figure are shown in bold. Only the parts of the general I/O Port Control Registers

(DDR and PORT) that are affected by the COMnx1:0 bits are shown. When referring to the

OCnx state, the reference is for the internal OCnx Register, not the OCnx pin. If a system reset

occur, the OCnx Register is reset to “0”.

Figure 14-5. Compare Match Output Unit, Schematic

COMnx1

Waveform

Generator

COMnx0

FOCnx

D

Q

1

0

OCnx

D

Q

PORT

D

Q

DDR

clk_I/O

The general I/O port function is overridden by the Output Compare (OCnx) from the Waveform

Generator if either of the COMnx1:0 bits are set. However, the OCnx pin direction (input or out-

put) is still controlled by the Data Direction Register (DDR) for the port pin. The Data Direction

Register bit for the OCnx pin (DDR_OCnx) must be set as output before the OCnx value is visi-

ble on the pin. The port override function is generally independent of the Waveform Generation

mode, but there are some exceptions. Refer to Table 14-2, Table 14-3 and Table 14-4 for

details.

The design of the Output Compare pin logic allows initialization of the OCnx state before the out-

put is enabled. Note that some COMnx1:0 bit settings are reserved for certain modes of

operation. See “16-bit Timers/Counters (Timer/Counter1 and Timer/Counter3)” on page 108.

The COMnx1:0 bits have no effect on the Input Capture unit.

14.7.1

Compare Output Mode and Waveform Generation

The Waveform Generator uses the COMnx1:0 bits differently in normal, CTC, and PWM modes.

For all modes, setting the COMnx1:0 = 0 tells the Waveform Generator that no action on the

OCnx Register is to be performed on the next compare match. For compare output actions in the

119

7766E–AVR–04/10

ATmega16U4/ATmega32U4

non-PWM modes refer to Table 14-2 on page 130. For fast PWM mode refer to Table 14-3 on

page 130, and for phase correct and phase and frequency correct PWM refer to Table 14-4 on

page 131.

A change of the COMnx1:0 bits state will have effect at the first compare match after the bits are

written. For non-PWM modes, the action can be forced to have immediate effect by using the

FOCnx strobe bits.

14.8 Modes of Operation

The mode of operation, i.e., the behavior of the Timer/Counter and the Output Compare pins, is

defined by the combination of the Waveform Generation mode (WGMn3:0) and Compare Output

mode (COMnx1:0) bits. The Compare Output mode bits do not affect the counting sequence,

while the Waveform Generation mode bits do. The COMnx1:0 bits control whether the PWM out-

put generated should be inverted or not (inverted or non-inverted PWM). For non-PWM modes

the COMnx1:0 bits control whether the output should be set, cleared or toggle at a compare

match (See “Compare Match Output Unit” on page 119.)

For detailed timing information refer to “Timer/Counter Timing Diagrams” on page 127.

14.8.1

Normal Mode

The simplest mode of operation is the Normal mode (WGMn3:0 = 0). In this mode the counting

direction is always up (incrementing), and no counter clear is performed. The counter simply

overruns when it passes its maximum 16-bit value (MAX = 0xFFFF) and then restarts from the

BOTTOM (0x0000). In normal operation the Timer/Counter Overflow Flag (TOVn) will be set in

the same timer clock cycle as the TCNTn becomes zero. The TOVn Flag in this case behaves

like a 17th bit, except that it is only set, not cleared. However, combined with the timer overflow

interrupt that automatically clears the TOVn Flag, the timer resolution can be increased by soft-

ware. There are no special cases to consider in the Normal mode, a new counter value can be

written anytime.

The Input Capture unit is easy to use in Normal mode. However, observe that the maximum

interval between the external events must not exceed the resolution of the counter. If the interval

between events are too long, the timer overflow interrupt or the prescaler must be used to

extend the resolution for the capture unit.

The Output Compare units can be used to generate interrupts at some given time. Using the

Output Compare to generate waveforms in Normal mode is not recommended, since this will

occupy too much of the CPU time.

14.8.2

Clear Timer on Compare Match (CTC) Mode

In Clear Timer on Compare or CTC mode (WGMn3:0 = 4 or 12), the OCRnA or ICRn Register

are used to manipulate the counter resolution. In CTC mode the counter is cleared to zero when

the counter value (TCNTn) matches either the OCRnA (WGMn3:0 = 4) or the ICRn (WGMn3:0 =

12). The OCRnA or ICRn define the top value for the counter, hence also its resolution. This

mode allows greater control of the compare match output frequency. It also simplifies the opera-

tion of counting external events.

The timing diagram for the CTC mode is shown in Figure 14-6. The counter value (TCNTn)

increases until a compare match occurs with either OCRnA or ICRn, and then counter (TCNTn)

is cleared.

120

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 14-6. CTC Mode, Timing Diagram

OCnA Interrupt Flag Set

or ICFn Interrupt Flag Set

(Interrupt on TOP)

TCNTn

OCnA

(Toggle)

(COMnA1:0 = 1)

1

2

3

4

Period

An interrupt can be generated at each time the counter value reaches the TOP value by either

using the OCFnA or ICFn Flag according to the register used to define the TOP value. If the

interrupt is enabled, the interrupt handler routine can be used for updating the TOP value. How-

ever, changing the TOP to a value close to BOTTOM when the counter is running with none or a

low prescaler value must be done with care since the CTC mode does not have the double buff-

ering feature. If the new value written to OCRnA or ICRn is lower than the current value of

TCNTn, the counter will miss the compare match. The counter will then have to count to its max-

imum value (0xFFFF) and wrap around starting at 0x0000 before the compare match can occur.

In many cases this feature is not desirable. An alternative will then be to use the fast PWM mode

using OCRnA for defining TOP (WGMn3:0 = 15) since the OCRnA then will be double buffered.

For generating a waveform output in CTC mode, the OCnA output can be set to toggle its logical

level on each compare match by setting the Compare Output mode bits to toggle mode

(COMnA1:0 = 1). The OCnA value will not be visible on the port pin unless the data direction for

the pin is set to output (DDR_OCnA = 1). The waveform generated will have a maximum fre-

quency of f_{OC A}= f_{clk_I/O}/2 when OCRnA is set to zero (0x0000). The waveform frequency is

n

defined by the following equation:

f

clk_I/O

f

= --------------------------------------------------

OCnA

2 ⋅ N ⋅ (1 + OCRnA)

The N variable represents the prescaler factor (1, 8, 64, 256, or 1024).

As for the Normal mode of operation, the TOVn Flag is set in the same timer clock cycle that the

counter counts from MAX to 0x0000.

14.8.3

Fast PWM Mode

The fast Pulse Width Modulation or fast PWM mode (WGMn3:0 = 5, 6, 7, 14, or 15) provides a

high frequency PWM waveform generation option. The fast PWM differs from the other PWM

options by its single-slope operation. The counter counts from BOTTOM to TOP then restarts

from BOTTOM. In non-inverting Compare Output mode, the Output Compare (OCnx) is set on

the compare match between TCNTn and OCRnx, and cleared at TOP. In inverting Compare

Output mode output is cleared on compare match and set at TOP. Due to the single-slope oper-

ation, the operating frequency of the fast PWM mode can be twice as high as the phase correct

and phase and frequency correct PWM modes that use dual-slope operation. This high fre-

quency makes the fast PWM mode well suited for power regulation, rectification, and DAC

applications. High frequency allows physically small sized external components (coils, capaci-

tors), hence reduces total system cost.

121

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The PWM resolution for fast PWM can be fixed to 8-, 9-, or 10-bit, or defined by either ICRn or

OCRnA. The minimum resolution allowed is 2-bit (ICRn or OCRnA set to 0x0003), and the max-

imum resolution is 16-bit (ICRn or OCRnA set to MAX). The PWM resolution in bits can be

calculated by using the following equation:

log(TOP + 1)

R

= ----------------------------------

FPWM

log(2)

In fast PWM mode the counter is incremented until the counter value matches either one of the

fixed values 0x00FF, 0x01FF, or 0x03FF (WGMn3:0 = 5, 6, or 7), the value in ICRn (WGMn3:0 =

14), or the value in OCRnA (WGMn3:0 = 15). The counter is then cleared at the following timer

clock cycle. The timing diagram for the fast PWM mode is shown in Figure 14-7. The figure

shows fast PWM mode when OCRnA or ICRn is used to define TOP. The TCNTn value is in the

timing diagram shown as a histogram for illustrating the single-slope operation. The diagram

includes non-inverted and inverted PWM outputs. The small horizontal line marks on the TCNTn

slopes represent compare matches between OCRnx and TCNTn. The OCnx Interrupt Flag will

be set when a compare match occurs.

Figure 14-7. Fast PWM Mode, Timing Diagram

OCRnx / TOP Update

and TOVn Interrupt Flag

Set and OCnA Interrupt

Flag Set or ICFn

Interrupt Flag Set

(Interrupt on TOP)

TCNTn

(COMnx1:0 = 2)

OCnx

(COMnx1:0 = 3)

OCnx

1

2

3

4

5

6

7

8

Period

The Timer/Counter Overflow Flag (TOVn) is set each time the counter reaches TOP. In addition

the OCnA or ICFn Flag is set at the same timer clock cycle as TOVn is set when either OCRnA

or ICRn is used for defining the TOP value. If one of the interrupts are enabled, the interrupt han-

dler routine can be used for updating the TOP and compare values.

When changing the TOP value the program must ensure that the new TOP value is higher or

equal to the value of all of the Compare Registers. If the TOP value is lower than any of the

Compare Registers, a compare match will never occur between the TCNTn and the OCRnx.

Note that when using fixed TOP values the unused bits are masked to zero when any of the

OCRnx Registers are written.

The procedure for updating ICRn differs from updating OCRnA when used for defining the TOP

value. The ICRn Register is not double buffered. This means that if ICRn is changed to a low

value when the counter is running with none or a low prescaler value, there is a risk that the new

ICRn value written is lower than the current value of TCNTn. The result will then be that the

counter will miss the compare match at the TOP value. The counter will then have to count to the

MAX value (0xFFFF) and wrap around starting at 0x0000 before the compare match can occur.

The OCRnA Register however, is double buffered. This feature allows the OCRnA I/O location

122

7766E–AVR–04/10

ATmega16U4/ATmega32U4

to be written anytime. When the OCRnA I/O location is written the value written will be put into

the OCRnA Buffer Register. The OCRnA Compare Register will then be updated with the value

in the Buffer Register at the next timer clock cycle the TCNTn matches TOP. The update is done

at the same timer clock cycle as the TCNTn is cleared and the TOVn Flag is set.

Using the ICRn Register for defining TOP works well when using fixed TOP values. By using

ICRn, the OCRnA Register is free to be used for generating a PWM output on OCnA. However,

if the base PWM frequency is actively changed (by changing the TOP value), using the OCRnA

as TOP is clearly a better choice due to its double buffer feature.

In fast PWM mode, the compare units allow generation of PWM waveforms on the OCnx pins.

Setting the COMnx1:0 bits to two will produce a non-inverted PWM and an inverted PWM output

can be generated by setting the COMnx1:0 to three (see Table on page 130). The actual OCnx

value will only be visible on the port pin if the data direction for the port pin is set as output

(DDR_OCnx). The PWM waveform is generated by setting (or clearing) the OCnx Register at

the compare match between OCRnx and TCNTn, and clearing (or setting) the OCnx Register at

the timer clock cycle the counter is cleared (changes from TOP to BOTTOM).

The PWM frequency for the output can be calculated by the following equation:

f

clk_I/O

f

= ----------------------------------

OCnxPWM

N ⋅ (1 + TOP)

The N variable represents the prescaler divider (1, 8, 64, 256, or 1024).

The extreme values for the OCRnx Register represents special cases when generating a PWM

waveform output in the fast PWM mode. If the OCRnx is set equal to BOTTOM (0x0000) the out-

put will be a narrow spike for each TOP+1 timer clock cycle. Setting the OCRnx equal to TOP

will result in a constant high or low output (depending on the polarity of the output set by the

COMnx1:0 bits.)

A frequency (with 50% duty cycle) waveform output in fast PWM mode can be achieved by set-

ting OCnA to toggle its logical level on each compare match (COMnA1:0 = 1). This applies only

if OCR1A is used to define the TOP value (WGM13:0 = 15). The waveform generated will have

a maximum frequency of f_{OC A}= f_{clk_I/O}/2 when OCRnA is set to zero (0x0000). This feature is

n

similar to the OCnA toggle in CTC mode, except the double buffer feature of the Output Com-

pare unit is enabled in the fast PWM mode.

14.8.4

Phase Correct PWM Mode

The phase correct Pulse Width Modulation or phase correct PWM mode (WGMn3:0 = 1, 2, 3,

10, or 11) provides a high resolution phase correct PWM waveform generation option. The

phase correct PWM mode is, like the phase and frequency correct PWM mode, based on a dual-

slope operation. The counter counts repeatedly from BOTTOM (0x0000) to TOP and then from

TOP to BOTTOM. In non-inverting Compare Output mode, the Output Compare (OCnx) is

cleared on the compare match between TCNTn and OCRnx while upcounting, and set on the

compare match while downcounting. In inverting Output Compare mode, the operation is

inverted. The dual-slope operation has lower maximum operation frequency than single slope

operation. However, due to the symmetric feature of the dual-slope PWM modes, these modes

are preferred for motor control applications.

The PWM resolution for the phase correct PWM mode can be fixed to 8-, 9-, or 10-bit, or defined

by either ICRn or OCRnA. The minimum resolution allowed is 2-bit (ICRn or OCRnA set to

123

7766E–AVR–04/10

ATmega16U4/ATmega32U4

0x0003), and the maximum resolution is 16-bit (ICRn or OCRnA set to MAX). The PWM resolu-

tion in bits can be calculated by using the following equation:

log(TOP + 1)

R

= ----------------------------------

PCPWM

log(2)

In phase correct PWM mode the counter is incremented until the counter value matches either

one of the fixed values 0x00FF, 0x01FF, or 0x03FF (WGMn3:0 = 1, 2, or 3), the value in ICRn

(WGMn3:0 = 10), or the value in OCRnA (WGMn3:0 = 11). The counter has then reached the

TOP and changes the count direction. The TCNTn value will be equal to TOP for one timer clock

cycle. The timing diagram for the phase correct PWM mode is shown on Figure 14-8. The figure

shows phase correct PWM mode when OCRnA or ICRn is used to define TOP. The TCNTn

value is in the timing diagram shown as a histogram for illustrating the dual-slope operation. The

diagram includes non-inverted and inverted PWM outputs. The small horizontal line marks on

the TCNTn slopes represent compare matches between OCRnx and TCNTn. The OCnx Inter-

rupt Flag will be set when a compare match occurs.

Figure 14-8. Phase Correct PWM Mode, Timing Diagram

OCRnx/TOP Update and

OCnA Interrupt Flag Set

or ICFn Interrupt Flag Set

(Interrupt on TOP)

TOVn Interrupt Flag Set

(Interrupt on Bottom)

TCNTn

(COMnx1:0 = 2)

OCnx

(COMnx1:0 = 3)

OCnx

1

2

3

4

Period

The Timer/Counter Overflow Flag (TOVn) is set each time the counter reaches BOTTOM. When

either OCRnA or ICRn is used for defining the TOP value, the OCnA or ICFn Flag is set accord-

ingly at the same timer clock cycle as the OCRnx Registers are updated with the double buffer

value (at TOP). The Interrupt Flags can be used to generate an interrupt each time the counter

reaches the TOP or BOTTOM value.

When changing the TOP value the program must ensure that the new TOP value is higher or

equal to the value of all of the Compare Registers. If the TOP value is lower than any of the

Compare Registers, a compare match will never occur between the TCNTn and the OCRnx.

Note that when using fixed TOP values, the unused bits are masked to zero when any of the

OCRnx Registers are written. As the third period shown in Figure 14-8 illustrates, changing the

TOP actively while the Timer/Counter is running in the phase correct mode can result in an

unsymmetrical output. The reason for this can be found in the time of update of the OCRnx Reg-

124

7766E–AVR–04/10

ATmega16U4/ATmega32U4

ister. Since the OCRnx update occurs at TOP, the PWM period starts and ends at TOP. This

implies that the length of the falling slope is determined by the previous TOP value, while the

length of the rising slope is determined by the new TOP value. When these two values differ the

two slopes of the period will differ in length. The difference in length gives the unsymmetrical

result on the output.

It is recommended to use the phase and frequency correct mode instead of the phase correct

mode when changing the TOP value while the Timer/Counter is running. When using a static

TOP value there are practically no differences between the two modes of operation.

In phase correct PWM mode, the compare units allow generation of PWM waveforms on the

OCnx pins. Setting the COMnx1:0 bits to two will produce a non-inverted PWM and an inverted

PWM output can be generated by setting the COMnx1:0 to three (See Table 14-4 on page 131).

The actual OCnx value will only be visible on the port pin if the data direction for the port pin is

set as output (DDR_OCnx). The PWM waveform is generated by setting (or clearing) the OCnx

Register at the compare match between OCRnx and TCNTn when the counter increments, and

clearing (or setting) the OCnx Register at compare match between OCRnx and TCNTn when

the counter decrements. The PWM frequency for the output when using phase correct PWM can

be calculated by the following equation:

f

clk_I/O

f

= ---------------------------

OCnxPCPWM

2 ⋅ N ⋅ TOP

The N variable represents the prescaler divider (1, 8, 64, 256, or 1024).

The extreme values for the OCRnx Register represent special cases when generating a PWM

waveform output in the phase correct PWM mode. If the OCRnx is set equal to BOTTOM the

output will be continuously low and if set equal to TOP the output will be continuously high for

non-inverted PWM mode. For inverted PWM the output will have the opposite logic values. If

OCR1A is used to define the TOP value (WGM13:0 = 11) and COM1A1:0 = 1, the OC1A output

will toggle with a 50% duty cycle.

14.8.5

Phase and Frequency Correct PWM Mode

The phase and frequency correct Pulse Width Modulation, or phase and frequency correct PWM

mode (WGMn3:0 = 8 or 9) provides a high resolution phase and frequency correct PWM wave-

form generation option. The phase and frequency correct PWM mode is, like the phase correct

PWM mode, based on a dual-slope operation. The counter counts repeatedly from BOTTOM

(0x0000) to TOP and then from TOP to BOTTOM. In non-inverting Compare Output mode, the

Output Compare (OCnx) is cleared on the compare match between TCNTn and OCRnx while

upcounting, and set on the compare match while downcounting. In inverting Compare Output

mode, the operation is inverted. The dual-slope operation gives a lower maximum operation fre-

quency compared to the single-slope operation. However, due to the symmetric feature of the

dual-slope PWM modes, these modes are preferred for motor control applications.

The main difference between the phase correct, and the phase and frequency correct PWM

mode is the time the OCRnx Register is updated by the OCRnx Buffer Register, (see Figure 14-

8 and Figure 14-9).

The PWM resolution for the phase and frequency correct PWM mode can be defined by either

ICRn or OCRnA. The minimum resolution allowed is 2-bit (ICRn or OCRnA set to 0x0003), and

125

7766E–AVR–04/10

ATmega16U4/ATmega32U4

the maximum resolution is 16-bit (ICRn or OCRnA set to MAX). The PWM resolution in bits can

be calculated using the following equation:

log(TOP + 1)

R

= ----------------------------------

PFCPWM

log(2)

In phase and frequency correct PWM mode the counter is incremented until the counter value

matches either the value in ICRn (WGMn3:0 = 8), or the value in OCRnA (WGMn3:0 = 9). The

counter has then reached the TOP and changes the count direction. The TCNTn value will be

equal to TOP for one timer clock cycle. The timing diagram for the phase correct and frequency

correct PWM mode is shown on Figure 14-9. The figure shows phase and frequency correct

PWM mode when OCRnA or ICRn is used to define TOP. The TCNTn value is in the timing dia-

gram shown as a histogram for illustrating the dual-slope operation. The diagram includes non-

inverted and inverted PWM outputs. The small horizontal line marks on the TCNTn slopes repre-

sent compare matches between OCRnx and TCNTn. The OCnx Interrupt Flag will be set when a

compare match occurs.

Figure 14-9. Phase and Frequency Correct PWM Mode, Timing Diagram

OCnA Interrupt Flag Set

or ICFn Interrupt Flag Set

(Interrupt on TOP)

OCRnx/TOP Updateand

TOVn Interrupt Flag Set

(Interrupt on Bottom)

TCNTn

(COMnx1:0 = 2)

OCnx

(COMnx1:0 = 3)

OCnx

1

2

3

4

Period

The Timer/Counter Overflow Flag (TOVn) is set at the same timer clock cycle as the OCRnx

Registers are updated with the double buffer value (at BOTTOM). When either OCRnA or ICRn

is used for defining the TOP value, the OCnA or ICFn Flag set when TCNTn has reached TOP.

The Interrupt Flags can then be used to generate an interrupt each time the counter reaches the

TOP or BOTTOM value.

When changing the TOP value the program must ensure that the new TOP value is higher or

equal to the value of all of the Compare Registers. If the TOP value is lower than any of the

Compare Registers, a compare match will never occur between the TCNTn and the OCRnx.

As Figure 14-9 shows the output generated is, in contrast to the phase correct mode, symmetri-

cal in all periods. Since the OCRnx Registers are updated at BOTTOM, the length of the rising

and the falling slopes will always be equal. This gives symmetrical output pulses and is therefore

frequency correct.

126

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Using the ICRn Register for defining TOP works well when using fixed TOP values. By using

ICRn, the OCRnA Register is free to be used for generating a PWM output on OCnA. However,

if the base PWM frequency is actively changed by changing the TOP value, using the OCRnA as

TOP is clearly a better choice due to its double buffer feature.

In phase and frequency correct PWM mode, the compare units allow generation of PWM wave-

forms on the OCnx pins. Setting the COMnx1:0 bits to two will produce a non-inverted PWM and

an inverted PWM output can be generated by setting the COMnx1:0 to three (See Table 14-4 on

page 131). The actual OCnx value will only be visible on the port pin if the data direction for the

port pin is set as output (DDR_OCnx). The PWM waveform is generated by setting (or clearing)

the OCnx Register at the compare match between OCRnx and TCNTn when the counter incre-

ments, and clearing (or setting) the OCnx Register at compare match between OCRnx and

TCNTn when the counter decrements. The PWM frequency for the output when using phase

and frequency correct PWM can be calculated by the following equation:

f

clk_I/O

f

= ---------------------------

OCnxPFCPWM

2 ⋅ N ⋅ TOP

The N variable represents the prescaler divider (1, 8, 64, 256, or 1024).

The extreme values for the OCRnx Register represents special cases when generating a PWM

waveform output in the phase correct PWM mode. If the OCRnx is set equal to BOTTOM the

output will be continuously low and if set equal to TOP the output will be set to high for non-

inverted PWM mode. For inverted PWM the output will have the opposite logic values. If OCR1A

is used to define the TOP value (WGM13:0 = 9) and COM1A1:0 = 1, the OC1A output will toggle

with a 50% duty cycle.

14.9 Timer/Counter Timing Diagrams

The Timer/Counter is a synchronous design and the timer clock (clk_Tn) is therefore shown as a

clock enable signal in the following figures. The figures include information on when Interrupt

Flags are set, and when the OCRnx Register is updated with the OCRnx buffer value (only for

modes utilizing double buffering). Figure 14-10 shows a timing diagram for the setting of OCFnx.

Figure 14-10. Timer/Counter Timing Diagram, Setting of OCFnx, no Prescaling

clk_I/O

clk_Tn

(clk_I/O/1)

TCNTn

OCRnx

OCFnx

OCRnx - 1

OCRnx

OCRnx + 1

OCRnx + 2

OCRnx Value

Figure 14-11 shows the same timing data, but with the prescaler enabled.

127

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 14-11. Timer/Counter Timing Diagram, Setting of OCFnx, with Prescaler (f_{clk_I/O}/8)

clk_I/O

clk_Tn

(clk_I/O/8)

TCNTn

OCRnx

OCFnx

OCRnx - 1

OCRnx

OCRnx + 1

OCRnx + 2

OCRnx Value

Figure 14-12 shows the count sequence close to TOP in various modes. When using phase and

frequency correct PWM mode the OCRnx Register is updated at BOTTOM. The timing diagrams

will be the same, but TOP should be replaced by BOTTOM, TOP-1 by BOTTOM+1 and so on.

The same renaming applies for modes that set the TOVn Flag at BOTTOM.

Figure 14-12. Timer/Counter Timing Diagram, no Prescaling

clk_I/O

clk_Tn

(clk_I/O/1)

TCNTn

TOP - 1

TOP

BOTTOM

TOP - 1

BOTTOM + 1

TOP - 2

(CTC and FPWM)

TCNTn

(PC and PFC PWM)

TOVn (FPWM)

and ICFn (if used

as TOP)

OCRnx

(Update at TOP)

New OCRnx Value

Old OCRnx Value

Figure 14-13 shows the same timing data, but with the prescaler enabled.

128

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 14-13. Timer/Counter Timing Diagram, with Prescaler (f_{clk_I/O}/8)

clk

I/O

clk

(clk^T/ⁿ8)

I/O

TCNTn

TOP - 1

TOP

BOTTOM

TOP - 1

BOTTOM + 1

TOP - 2

(CTC and FPWM)

TCNTn

(PC and PFC PWM)

TOVn(FPWM)

and ICFn(if used

as TOP)

OCRnx

(Update at TOP)

Old OCRnx Value

New OCRnx Value

14.10 16-bit Timer/Counter Register Description

14.10.1 Timer/Counter1 Control Register A – TCCR1A

Bit

7

6

5

4

3

2

1

0

COM1A1

COM1A0

COM1B1

COM1B0

COM1C1

COM1C0

WGM11

R/W

0

WGM10

R/W

0

TCCR1A

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

14.10.2 Timer/Counter3 Control Register A – TCCR3A

Bit

7

6

5

4

3

2

1

0

COM3A1

COM3A0

COM3B1

COM3B0

COM3C1

COM3C0

WGM31

R/W

0

WGM30

R/W

0

TCCR3A

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

• Bit 7:6 – COMnA1:0: Compare Output Mode for Channel A

• Bit 5:4 – COMnB1:0: Compare Output Mode for Channel B

• Bit 3:2 – COMnC1:0: Compare Output Mode for Channel C

The COMnA1:0, COMnB1:0, and COMnC1:0 control the output compare pins (OCnA, OCnB,

and OCnC respectively) behavior. If one or both of the COMnA1:0 bits are written to one, the

OCnA output overrides the normal port functionality of the I/O pin it is connected to. If one or

both of the COMnB1:0 bits are written to one, the OCnB output overrides the normal port func-

tionality of the I/O pin it is connected to. If one or both of the COMnC1:0 bits are written to one,

the OCnC output overrides the normal port functionality of the I/O pin it is connected to. How-

ever, note that the Data Direction Register (DDR) bit corresponding to the OCnA, OCnB or

OCnC pin must be set in order to enable the output driver.

129

7766E–AVR–04/10

ATmega16U4/ATmega32U4

When the OCnA, OCnB or OCnC is connected to the pin, the function of the COMnx1:0 bits is

dependent of the WGMn3:0 bits setting. Table 14-2 shows the COMnx1:0 bit functionality when

the WGMn3:0 bits are set to a normal or a CTC mode (non-PWM).

Table 14-2. Compare Output Mode, non-PWM

COMnA1/COMnB1/

COMnC1

COMnA0/COMnB0/

COMnC0

Description

Normal port operation, OCnA/OCnB/OCnC

disconnected.

0

1

0

1

0

1

Toggle OCnA/OCnB/OCnC on compare

match.

Clear OCnA/OCnB/OCnC on compare

match (set output to low level).

Set OCnA/OCnB/OCnC on compare match

(set output to high level).

Table 14-3 shows the COMnx1:0 bit functionality when the WGMn3:0 bits are set to the fast

PWM mode.

Table 14-3. Compare Output Mode, Fast PWM

COMnA1/COMnB1/

COMnC0

COMnA0/COMnB0/

COMnC0

Description

Normal port operation, OCnA/OCnB/OCnC

disconnected.

0

WGM13:0 = 14 or 15: Toggle OC1A on

Compare Match, OC1B and OC1C

0

1

disconnected (normal port operation). For all

other WGM1 settings, normal port operation,

OC1A/OC1B/OC1C disconnected.

Clear OCnA/OCnB/OCnC on compare

match, set OCnA/OCnB/OCnC at TOP

1

0

1

Set OCnA/OCnB/OCnC on compare match,

clear OCnA/OCnB/OCnC at TOP

Note:

A special case occurs when OCRnA/OCRnB/OCRnC equals TOP and

COMnA1/COMnB1/COMnC1 is set. In this case the compare match is ignored, but the set or clear

is done at TOP. See “Fast PWM Mode” on page 97. for more details.

Table 14-4 shows the COMnx1:0 bit functionality when the WGMn3:0 bits are set to the phase

correct and frequency correct PWM mode.

130

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 14-4. Compare Output Mode, Phase Correct and Phase and Frequency Correct PWM

COMnA1/COMnB/

COMnC1

COMnA0/COMnB0/

COMnC0

Description

Normal port operation, OCnA/OCnB/OCnC

disconnected.

0

WGM13:0 = 8, 9 10 or 11: Toggle OC1A on

Compare Match, OC1B and OC1C

0

1

disconnected (normal port operation). For all

other WGM1 settings, normal port operation,

OC1A/OC1B/OC1C disconnected.

Clear OCnA/OCnB/OCnC on compare

match when up-counting. Set

OCnA/OCnB/OCnC on compare match

when downcounting.

1

0

1

Set OCnA/OCnB/OCnC on compare match

when up-counting. Clear

OCnA/OCnB/OCnC on compare match

when downcounting.

Note:

A special case occurs when OCRnA/OCRnB/OCRnC equals TOP and

COMnA1/COMnB1//COMnC1 is set. See “Phase Correct PWM Mode” on page 99. for more

details.

• Bit 1:0 – WGMn1:0: Waveform Generation Mode

Combined with the WGMn3:2 bits found in the TCCRnB Register, these bits control the counting

sequence of the counter, the source for maximum (TOP) counter value, and what type of wave-

form generation to be used, see Table 14-5. Modes of operation supported by the Timer/Counter

unit are: Normal mode (counter), Clear Timer on Compare match (CTC) mode, and three types

of Pulse Width Modulation (PWM) modes. (See “Modes of Operation” on page 96.).

Table 14-5. Waveform Generation Mode Bit Description ⁽¹⁾

WGMn2

(CTCn)

WGMn1

WGMn0

Update of

OCRnx at

TOVn Flag

Set on

Mode WGMn3

(PWMn1) (PWMn0) Timer/Counter Mode of Operation

TOP

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

Normal

0xFFFF

0x00FF

0x01FF

0x03FF

OCRnA

0x00FF

0x01FF

0x03FF

Immediate

TOP

MAX

PWM, Phase Correct, 8-bit

PWM, Phase Correct, 9-bit

PWM, Phase Correct, 10-bit

CTC

BOTTOM

MAX

2

TOP

3

TOP

4

Immediate

TOP

5

Fast PWM, 8-bit

TOP

6

Fast PWM, 9-bit

TOP

7

Fast PWM, 10-bit

TOP

8

PWM, Phase and Frequency Correct ICRn

PWM, Phase and Frequency Correct OCRnA

BOTTOM

TOP

BOTTOM

MAX

9

10

11

12

PWM, Phase Correct

CTC

ICRn

OCRnA

ICRn

TOP

Immediate

131

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 14-5. Waveform Generation Mode Bit Description (Continued)⁽¹⁾

WGMn2

(CTCn)

WGMn1

WGMn0

Update of

OCRnx at

TOVn Flag

Set on

Mode WGMn3

(PWMn1) (PWMn0) Timer/Counter Mode of Operation

TOP

–

13

14

1

0

1

0

1

(Reserved)

Fast PWM

–

ICRn

OCRnA

TOP

15

Note:

1. The CTCn and PWMn1:0 bit definition names are obsolete. Use the WGMn2:0 definitions. However, the functionality and

location of these bits are compatible with previous versions of the timer.

14.10.3 Timer/Counter1 Control Register B – TCCR1B

Bit

7

6

5

–

R

0

4

3

2

1

0

ICNC1

R/W

0

ICES1

R/W

0

WGM13

R/W

0

WGM12

R/W

0

CS12

R/W

0

CS11

R/W

0

CS10

R/W

0

TCCR1B

Read/Write

Initial Value

14.10.4 Timer/Counter3 Control Register B – TCCR3B

Bit

7

6

5

–

R

0

4

3

2

1

0

ICNC3

R/W

0

ICES3

R/W

0

WGM33

R/W

0

WGM32

R/W

0

CS32

R/W

0

CS31

R/W

0

CS30

R/W

0

TCCR3B

Read/Write

Initial Value

• Bit 7 – ICNCn: Input Capture Noise Canceler

Setting this bit (to one) activates the Input Capture Noise Canceler. When the Noise Canceler is

activated, the input from the Input Capture Pin (ICPn) is filtered. The filter function requires four

successive equal valued samples of the ICPn pin for changing its output. The input capture is

therefore delayed by four Oscillator cycles when the noise canceler is enabled.

• Bit 6 – ICESn: Input Capture Edge Select

This bit selects which edge on the Input Capture Pin (ICPn) that is used to trigger a capture

event. When the ICESn bit is written to zero, a falling (negative) edge is used as trigger, and

when the ICESn bit is written to one, a rising (positive) edge will trigger the capture.

When a capture is triggered according to the ICESn setting, the counter value is copied into the

Input Capture Register (ICRn). The event will also set the Input Capture Flag (ICFn), and this

can be used to cause an Input Capture Interrupt, if this interrupt is enabled.

When the ICRn is used as TOP value (see description of the WGMn3:0 bits located in the

TCCRnA and the TCCRnB Register), the ICPn is disconnected and consequently the input cap-

ture function is disabled.

• Bit 5 – Reserved Bit

This bit is reserved for future use. For ensuring compatibility with future devices, this bit must be

written to zero when TCCRnB is written.

• Bit 4:3 – WGMn3:2: Waveform Generation Mode

See TCCRnA Register description.

• Bit 2:0 – CSn2:0: Clock Select

The three clock select bits select the clock source to be used by the Timer/Counter, see Figure

13-8 and Figure 13-9.

132

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 14-6. Clock Select Bit Description

CSn2

CSn1

CSn0

Description

0

1

0

1

0

1

0

1

0

1

0

1

0

1

No clock source. (Timer/Counter stopped)

clk_I/O/1 (No prescaling

clk_I/O/8 (From prescaler)

clk_I/O/64 (From prescaler)

clk_I/O/256 (From prescaler)

clk_I/O/1024 (From prescaler)

External clock source on Tn pin. Clock on falling edge

External clock source on Tn pin. Clock on rising edge

If external pin modes are used for the Timer/Countern, transitions on the Tn pin will clock the

counter even if the pin is configured as an output. This feature allows software control of the

counting.

14.10.5 Timer/Counter1 Control Register C – TCCR1C

Bit

7

6

5

4

–

R

0

3

–

R

0

2

–

R

0

1

–

R

0

–

R

0

FOC1A

FOC1B

FOC1C

TCCR1C

Read/Write

Initial Value

W

0

W

0

W

0

14.10.6 Timer/Counter3 Control Register C – TCCR3C

Bit

7

6

–

R

0

5

–

R

0

4

–

R

0

3

–

R

0

2

–

R

0

1

–

R

0

–

R

0

FOC3A

TCCR3C

Read/Write

Initial Value

W

0

• Bit 7 – FOCnA: Force Output Compare for Channel A

The FOCnA/FOCnB/FOCnC bits are only active when the WGMn3:0 bits specifies a non-PWM

mode. When writing a logical one to the FOCnA/FOCnB/FOCnC bit, an immediate compare

match is forced on the waveform generation unit. The OCnA/OCnB/OCnC output is changed

according to its COMnx1:0 bits setting. Note that the FOCnA/FOCnB/FOCnC bits are imple-

mented as strobes. Therefore it is the value present in the COMnx1:0 bits that determine the

effect of the forced compare.

A FOCnA/FOCnB/FOCnC strobe will not generate any interrupt nor will it clear the timer in Clear

Timer on Compare Match (CTC) mode using OCRnA as TOP.

The FOCnA/FOCnB/FOCnB bits are always read as zero.

• Bit 4:0 – Reserved Bits

These bits are reserved for future use. For ensuring compatibility with future devices, these bits

must be written to zero when TCCRnC is written.

133

7766E–AVR–04/10

ATmega16U4/ATmega32U4

14.10.7 Timer/Counter1 – TCNT1H and TCNT1L

Bit

7

6

5

4

3

2

1

0

TCNT1[15:8]

TCNT1[7:0]

TCNT1H

TCNT1L

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

14.10.8 Timer/Counter3 – TCNT3H and TCNT3L

Bit

7

6

5

4

3

2

1

0

TCNT3[15:8]

TCNT3[7:0]

TCNT3H

TCNT3L

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

The two Timer/Counter I/O locations (TCNTnH and TCNTnL, combined TCNTn) give direct

access, both for read and for write operations, to the Timer/Counter unit 16-bit counter. To

ensure that both the high and low bytes are read and written simultaneously when the CPU

accesses these registers, the access is performed using an 8-bit temporary High Byte Register

(TEMP). This temporary register is shared by all the other 16-bit registers. See “Accessing 16-bit

Registers” on page 110.

Modifying the counter (TCNTn) while the counter is running introduces a risk of missing a com-

pare match between TCNTn and one of the OCRnx Registers.

Writing to the TCNTn Register blocks (removes) the compare match on the following timer clock

for all compare units.

14.10.9 Output Compare Register 1 A – OCR1AH and OCR1AL

Bit

7

6

5

4

3

2

1

0

OCR1A[15:8]

OCR1A[7:0]

OCR1AH

OCR1AL

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

14.10.10 Output Compare Register 1 B – OCR1BH and OCR1BL

Bit

7

6

5

4

3

2

1

0

OCR1B[15:8]

OCR1B[7:0]

OCR1BH

OCR1BL

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

14.10.11 Output Compare Register 1 C – OCR1CH and OCR1CL

Bit

7

6

5

4

3

2

1

0

OCR1C[15:8]

OCR1C[7:0]

OCR1CH

OCR1CL

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

134

7766E–AVR–04/10

ATmega16U4/ATmega32U4

14.10.12 Output Compare Register 3 A – OCR3AH and OCR3AL

Bit

7

6

5

4

3

2

1

0

OCR3A[15:8]

OCR3A[7:0]

OCR3AH

OCR3AL

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

14.10.13 Output Compare Register 3 B – OCR3BH and OCR3BL

Bit

7

6

5

4

3

2

1

0

OCR3B[15:8]

OCR3B[7:0]

OCR3BH

OCR3BL

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

14.10.14 Output Compare Register 3 C – OCR3CH and OCR3CL

Bit

7

6

5

4

3

2

1

0

OCR3C[15:8]

OCR3C[7:0]

OCR3CH

OCR3CL

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

The Output Compare Registers contain a 16-bit value that is continuously compared with the

counter value (TCNTn). A match can be used to generate an Output Compare interrupt, or to

generate a waveform output on the OCnx pin.

The Output Compare Registers are 16-bit in size. To ensure that both the high and low bytes are

written simultaneously when the CPU writes to these registers, the access is performed using an

8-bit temporary High Byte Register (TEMP). This temporary register is shared by all the other

16-bit registers. See “Accessing 16-bit Registers” on page 110.

14.10.15 Input Capture Register 1 – ICR1H and ICR1L

Bit

7

6

5

4

3

ICR1[15:8]

ICR1[7:0]

R/W

2

1

0

ICR1H

ICR1L

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

14.10.16 Input Capture Register 3 – ICR3H and ICR3L

Bit

7

6

5

4

3

ICR3[15:8]

ICR3[7:0]

R/W

2

1

0

ICR3H

ICR3L

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

The Input Capture is updated with the counter (TCNTn) value each time an event occurs on the

ICPn pin (or optionally on the Analog Comparator output for Timer/Counter1). The Input Capture

can be used for defining the counter TOP value.

The Input Capture Register is 16-bit in size. To ensure that both the high and low bytes are read

simultaneously when the CPU accesses these registers, the access is performed using an 8-bit

temporary High Byte Register (TEMP). This temporary register is shared by all the other 16-bit

registers. See “Accessing 16-bit Registers” on page 110.

135

7766E–AVR–04/10

ATmega16U4/ATmega32U4

14.10.17 Timer/Counter1 Interrupt Mask Register – TIMSK1

Bit

7

–

R

0

6

–

R

0

5

4

–

R

0

3

2

1

0

ICIE1

R/W

0

OCIE1C

R/W

0

OCIE1B

R/W

0

OCIE1A

R/W

0

TOIE1

R/W

0

TIMSK1

Read/Write

Initial Value

14.10.18 Timer/Counter3 Interrupt Mask Register – TIMSK3

Bit

7

–

R

0

6

–

R

0

5

4

–

R

0

3

2

1

0

ICIE3

R/W

0

OCIE3C

R/W

0

OCIE3B

R/W

0

OCIE3A

R/W

0

TOIE3

R/W

0

TIMSK3

Read/Write

Initial Value

• Bit 5 – ICIEn: Timer/Countern, Input Capture Interrupt Enable

When this bit is written to one, and the I-flag in the Status Register is set (interrupts globally

enabled), the Timer/Countern Input Capture interrupt is enabled. The corresponding Interrupt

Vector (See “Interrupts” on page 61.) is executed when the ICFn Flag, located in TIFRn, is set.

• Bit 3 – OCIEnC: Timer/Countern, Output Compare C Match Interrupt Enable

When this bit is written to one, and the I-flag in the Status Register is set (interrupts globally

enabled), the Timer/Countern Output Compare C Match interrupt is enabled. The corresponding

Interrupt Vector (See “Interrupts” on page 61.) is executed when the OCFnC Flag, located in

TIFRn, is set.

• Bit 2 – OCIEnB: Timer/Countern, Output Compare B Match Interrupt Enable

When this bit is written to one, and the I-flag in the Status Register is set (interrupts globally

enabled), the Timer/Countern Output Compare B Match interrupt is enabled. The corresponding

Interrupt Vector (See “Interrupts” on page 61.) is executed when the OCFnB Flag, located in

TIFRn, is set.

• Bit 1 – OCIEnA: Timer/Countern, Output Compare A Match Interrupt Enable

When this bit is written to one, and the I-flag in the Status Register is set (interrupts globally

enabled), the Timer/Countern Output Compare A Match interrupt is enabled. The corresponding

Interrupt Vector (See “Interrupts” on page 61.) is executed when the OCFnA Flag, located in

TIFRn, is set.

• Bit 0 – TOIEn: Timer/Countern, Overflow Interrupt Enable

When this bit is written to one, and the I-flag in the Status Register is set (interrupts globally

enabled), the Timer/Countern Overflow interrupt is enabled. The corresponding Interrupt Vector

(See “Interrupts” on page 61.) is executed when the TOVn Flag, located in TIFRn, is set.

14.10.19 Timer/Counter1 Interrupt Flag Register – TIFR1

Bit

7

6

5

4

–

R

0

3

2

1

0

–

ICF1

OCF1C

R/W

0

OCF1B

R/W

0

OCF1A

R/W

0

TOV1

R/W

0

TIFR1

Read/Write

Initial Value

R

0

R

0

R/W

0

136

7766E–AVR–04/10

ATmega16U4/ATmega32U4

14.10.20 Timer/Counter3 Interrupt Flag Register – TIFR3

Bit

7

–

R

0

6

–

R

0

5

4

–

R

0

3

2

1

0

ICF3

R/W

0

OCF3C

R/W

0

OCF3B

R/W

0

OCF3A

R/W

0

TOV3

R/W

0

TIFR3

Read/Write

Initial Value

• Bit 5 – ICFn: Timer/Countern, Input Capture Flag

This flag is set when a capture event occurs on the ICPn pin. When the Input Capture Register

(ICRn) is set by the WGMn3:0 to be used as the TOP value, the ICFn Flag is set when the coun-

ter reaches the TOP value.

ICFn is automatically cleared when the Input Capture Interrupt Vector is executed. Alternatively,

ICFn can be cleared by writing a logic one to its bit location.

• Bit 3– OCFnC: Timer/Countern, Output Compare C Match Flag

This flag is set in the timer clock cycle after the counter (TCNTn) value matches the Output

Compare Register C (OCRnC).

Note that a Forced Output Compare (FOCnC) strobe will not set the OCFnC Flag.

OCFnC is automatically cleared when the Output Compare Match C Interrupt Vector is exe-

cuted. Alternatively, OCFnC can be cleared by writing a logic one to its bit location.

• Bit 2 – OCFnB: Timer/Counter1, Output Compare B Match Flag

This flag is set in the timer clock cycle after the counter (TCNTn) value matches the Output

Compare Register B (OCRnB).

Note that a Forced Output Compare (FOCnB) strobe will not set the OCFnB Flag.

OCFnB is automatically cleared when the Output Compare Match B Interrupt Vector is exe-

cuted. Alternatively, OCFnB can be cleared by writing a logic one to its bit location.

• Bit 1 – OCF1A: Timer/Counter1, Output Compare A Match Flag

This flag is set in the timer clock cycle after the counter (TCNTn value matches the Output Com-

pare Register A (OCRnA).

Note that a Forced Output Compare (FOCnA) strobe will not set the OCFnA Flag.

OCFnA is automatically cleared when the Output Compare Match A Interrupt Vector is exe-

cuted. Alternatively, OCFnA can be cleared by writing a logic one to its bit location.

• Bit 0 – TOVn: Timer/Countern, Overflow Flag

The setting of this flag is dependent of the WGMn3:0 bits setting. In Normal and CTC modes,

the TOVn Flag is set when the timer overflows. Refer to Table 14-5 on page 131 for the TOVn

Flag behavior when using another WGMn3:0 bit setting.

TOVn is automatically cleared when the Timer/Countern Overflow Interrupt Vector is executed.

Alternatively, TOVn can be cleared by writing a logic one to its bit location.

137

7766E–AVR–04/10

ATmega16U4/ATmega32U4

15. 10-bit High Speed Timer/Counter4

15.1 Features

• Up to 10-Bit Accuracy

• Three Independent Output Compare Units

• Clear Timer on Compare Match (Auto Reload)

• Glitch Free, Phase and Frequency Correct Pulse Width Modulator (PWM)

• Enhanced PWM mode: one optional additional accuracy bit without effect on output frequency

• Variable PWM Period

• Independent Dead Time Generators for each PWM channels

• Synchronous update of PWM registers

• Five Independent Interrupt Sources (TOV4, OCF4A, OCF4B, OCF4D, FPF4)

• High Speed Asynchronous and Synchronous Clocking Modes

• Separate Prescaler Unit

15.2 Overview

Timer/Counter4 is a general purpose high speed Timer/Counter module, with three independent

Output Compare Units, and with enhanced PWM support.

The Timer/Counter4 features a high resolution and a high accuracy usage with the lower pres-

caling opportunities. It can also support three accurate and high speed Pulse Width Modulators

using clock speeds up to 64 MHz. In PWM mode Timer/Counter4 and the output compare regis-

ters serve as triple stand-alone PWMs with non-overlapping, non-inverted and inverted outputs.

The enhanced PWM mode allows to get one more accuracy bit while keeping the frequency

identical to normal mode (a PWM 8 bits accuracy in enhanced mode outputs the same fre-

quency that a PWM 7 bits accuracy in normal mode). Similarly, the high prescaling opportunities

make this unit useful for lower speed functions or exact timing functions with infrequent actions.

A lock feature allows user to update the PWM registers and

A simplified block diagram of the Timer/Counter4 is shown in Figure 15-1. For actual placement

of the I/O pins, refer to “Pinout ATmega16U4/ATmega32U4” on page 3. The device-specific I/O

register and bit locations are listed in the “Register Description” on page 162.

138

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 15-1. Timer/Counter4 Block Diagram

TOV4

OCF4A

OCF4B

OCF4D

OC4A

OC4B

OC4D

FAULT_PROTECTION

DEAD TIME GENERATOR

OCW4A

OCW4B

OCW4D

T/C INT. MASK

REGISTER (TIMSK4)

T/C INT. FLAG

REGISTER (TIFR4)

T/C CONTROL

REGISTER C (TCCR4D)

T/C CONTROL

REGISTER A (TCCR4A)

T/C CONTROL

REGISTER B (TCCR4B)

T/C CONTROL

REGISTER C (TCCR4C)

CLK

COUNT

CLEAR

TIMER/COUNTER4

(TCNT4)

TIMER/COUNTER4 CONTROL LOGIC

DIRECTION

T/C CONTROL

REGISTER D (TCCR4E)

10-BIT COMPARATOR

10-BIT OUTPUT

COMPARE REGISTER B

COMPARE REGISTER A

COMPARE REGISTER C

COMPARE REGISTER D

8-BIT OUTPUT COMPARE

REGISTER A (OCR4A)

8-BIT OUTPUT COMPARE

REGISTER B (OCR4B)

8-BIT OUTPUT COMPARE

REGISTER C (OCR4C)

8-BIT OUTPUT COMPARE

REGISTER D (OCR4D)

2-BIT HIGH BYTE

REGISTER (TC4H)

8-BIT DATABUS

15.2.1

15.2.2

Speed

The maximum speed of the Timer/Counter4 is 64 MHz. However, if a supply voltage below 4

volts is used, it is recommended to decrease the input frequency, because the Timer/Counter4

is not running fast enough on low voltage levels.

Accuracy

The Timer/Counter4 is a 10-bit Timer/Counter module that can alternatively be used as an 8-bit

Timer/Counter. The Timer/Counter4 registers are basically 8-bit registers, but on top of that

there is a 2-bit High Byte Register (TC4H) that can be used as a common temporary buffer to

access the two MSBs of the 10-bit Timer/Counter4 registers by the AVR CPU via the 8-bit data

bus, if the 10-bit accuracy is used. Whereas, if the two MSBs of the 10-bit registers are written to

zero the Timer/Counter4 is working as an 8-bit Timer/Counter. When reading the low byte of any

8-bit register the two MSBs are written to the TC4H register, and when writing the low byte of

any 8-bit register the two MSBs are written from the TC4H register. Special procedures must be

followed when accessing the 10-bit Timer/Counter4 values via the 8-bit data bus. These proce-

dures are described in the section “Accessing 10-Bit Registers” on page 159.

The Enhanced PWM mode allows to add a resolution bit to each Compare register A/B/D, while

the output frequency remains identical to a Normal PWM mode. That means that the TC4H reg-

ister contains one more bit that will be the MSB in a 11-bits enhanced PWM operation. See the

section “Enhanced Compare/PWM mode” on page 148 for details about this feature and how to

use it.

139

7766E–AVR–04/10

ATmega16U4/ATmega32U4

15.2.3

Registers

The Timer/Counter (TCNT4) and Output Compare Registers (OCR4A, OCR4B, OCR4C and

OCR4D) are 8-bit registers that are used as a data source to be compared with the TCNT4 con-

tents. The OCR4A, OCR4B and OCR4D registers determine the action on the OC4A, OC4B and

OC4D pins and they can also generate the compare match interrupts. The OCR4C holds the

Timer/Counter TOP value, i.e. the clear on compare match value. The Timer/Counter4 High

Byte Register (TC4H) is a 2-bit register that is used as a common temporary buffer to access the

MSB bits of the Timer/Counter4 registers, if the 10-bit accuracy is used.

Interrupt request (overflow TOV4, compare matches OCF4A, OCF4B, OCF4D and fault protec-

tion FPF4) signals are visible in the Timer Interrupt Flag Register (TIFR4) and Timer/Counter4

Control Register D (TCCR4D). The interrupts are individually masked with the Timer Interrupt

Mask Register (TIMSK4) and the FPIE4 bit in the Timer/Counter4 Control Register D (TCCR4D).

Control signals are found in the Timer/Counter Control Registers TCCR4A, TCCR4B, TCCR4C,

TCCR4D and TCCR4E.

15.2.4

Synchronization

In asynchronous clocking mode the Timer/Counter4 and the prescaler allow running the CPU

from any clock source while the prescaler is operating on the fast peripheral clock (PCK) having

frequency up to 64 MHz. This is possible because there is a synchronization boundary between

the CPU clock domain and the fast peripheral clock domain. Figure 15-2 shows Timer/Counter 4

synchronization register block diagram and describes synchronization delays in between regis-

ters. Note that all clock gating details are not shown in the figure.

The Timer/Counter4 register values go through the internal synchronization registers, which

cause the input synchronization delay, before affecting the counter operation. The registers

TCCR4A, TCCR4B, TCCR4C, TCCR4D, OCR4A, OCR4B, OCR4C and OCR4D can be read

back right after writing the register. The read back values are delayed for the Timer/Counter4

(TCNT4) register, Timer/Counter4 High Byte Register (TC4H) and flags (OCF4A, OCF4B,

OCF4D and TOV4), because of the input and output synchronization.

The system clock frequency must be lower than half of the PCK frequency, because the syn-

chronization mechanism of the asynchronous Timer/Counter4 needs at least two edges of the

PCK when the system clock is high. If the frequency of the system clock is too high, it is a risk

that data or control values are lost.

140

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 15-2. Timer/Counter4 Synchronization Register Block Diagram.

8-BIT DATABUS

IO-registers

OCR4A

Input synchronization

registers

Timer/Counter4

Output synchronization

registers

OCR4A_SI

OCR4B_SI

OCR4C_SI

OCR4D_SI

TCCR4A_SI

TCCR4B_SI

TCNT4

TC4H

OCR4B

OCR4C

OCR4D

TCCR4A

TCCR4B

TCNT4_SO

TC4H_SO

OCF4

OCF4A_SO

OCF4B_SO

TCCR4C

TCCR4D

TCNT4

TCCR4C_SI

TCCR4D_SI

TCNT4_SI

TCNT4

OCF4B

TC4H

TC4H_SI

OCF4A

OCF4B

OCF4D

TOV4

OCF4A_SI

OCF4B_SI

OCF4D_SI

TOV4_SI

OCF4D

OCF4D_SO

TOV4_SO

TOV4

PLLTM1:0

!= '00'

CK

S

A

S

A

PCK

(clk

)

TMR

SYNC

MODE

1/2 CK Delay

~1/2 CK Delay

1 CK Delay

1 PCK Delay

1/2 CK Delay

~1 CK Delay

ASYNC

MODE

1 PCK Delay

15.2.5

Definitions

Many register and bit references in this section are written in general form. A lower case “n”

replaces the Timer/Counter number, in this case 0. A lower case “x” replaces the Output Com-

pare Unit, in this case Compare Unit A, B, C or D. However, when using the register or bit

defines in a program, the precise form must be used, i.e., TCNT4 for accessing Timer/Counter4

counter value and so on. The definitions in Table 15-1 are used extensively throughout the

document.

Table 15-1. Definitions

BOTTOM

The counter reaches the BOTTOM when it becomes 0.

MAX

The counter reaches its MAXimum value when it becomes 0x3FF (decimal 1023).

The counter reaches the TOP value (stored in the OCR1C) when it becomes equal to the

highest value in the count sequence. The TOP has a value 0x0FF as default after reset.

TOP

141

7766E–AVR–04/10

ATmega16U4/ATmega32U4

15.3 Counter Unit

The main part of the Timer/Counter4 is the programmable bi-directional counter unit. Figure 15-

3 shows a block diagram of the counter and its surroundings.

Figure 15-3. Counter Unit Block Diagram

DATA BUS

TOV4

clk_T4

Timer/Counter4 Count Enable

count

clear

( From Prescaler )

PLLTM1:0

PCK

TCNT4

Control Logic

direction

CK

bottom

top

Signal description (internal signals):

count

direction

clear

TCNT4 increment or decrement enable.

Select between increment and decrement.

Clear TCNT4 (set all bits to zero).

clk_Tn

Timer/Counter clock, referred to as clk_T4in the following.

Signalize that TCNT4 has reached maximum value.

Signalize that TCNT4 has reached minimum value (zero).

top

bottom

Depending of the mode of operation used, the counter is cleared, incremented, or decremented

at each timer clock (clk_T4). The timer clock is generated from an synchronous system clock or an

asynchronous PLL clock using the Clock Select bits (CS4<3:0>) and the PLL Postscaler for High

Speed Timer bits (PLLTM1:0). When no clock source is selected (CS4<3:0> = 0) the timer is

stopped. However, the TCNT4 value can be accessed by the CPU, regardless of whether clk_T1

is present or not. A CPU write overrides (has priority over) all counter clear or count operations.

The counting sequence of the Timer/Counter4 is determined by the setting of the WGM10 and

PWM4x bits located in the Timer/Counter4 Control Registers (TCCR4A, TCCR4C and

TCCR4D). For more details about advanced counting sequences and waveform generation, see

“Modes of Operation” on page 149. The Timer/Counter Overflow Flag (TOV4) is set according to

the mode of operation selected by the PWM4x and WGM40 bits. The Overflow Flag can be used

for generating a CPU interrupt.

15.3.1

Counter Initialization for Asynchronous Mode

To change Timer/Counter4 to the asynchronous mode follow the procedure below:

1. Enable PLL.

2. Wait 100µs for PLL to stabilize .

3. Poll the PLOCK bit until it is set.

4. Configure the PLLTM1:0 bits in the PLLFRQ register to enable the asynchronous mode

(different from 0:0 value).

142

7766E–AVR–04/10

ATmega16U4/ATmega32U4

15.4 Output Compare Unit

The comparator continuously compares TCNT4 with the Output Compare Registers (OCR4A,

OCR4B, OCR4C and OCR4D). Whenever TCNT4 equals to the Output Compare Register, the

comparator signals a match. A match will set the Output Compare Flag (OCF4A, OCF4B or

OCF4D) at the next timer clock cycle. If the corresponding interrupt is enabled, the Output Com-

pare Flag generates an Output Compare interrupt. The Output Compare Flag is automatically

cleared when the interrupt is executed. Alternatively, the flag can be cleared by software by writ-

ing a logical one to its I/O bit location. The Waveform Generator uses the match signal to

generate an output according to operating mode set by the PWM4x, WGM40 and Compare Out-

put mode (COM4x1:0) bits. The top and bottom signals are used by the Waveform Generator for

handling the special cases of the extreme values in some modes of operation (See “Modes of

Operation” on page 149.). Figure 15-4 shows a block diagram of the Output Compare unit.

Figure 15-4. Output Compare Unit, Block Diagram

8-BIT DATA BUS

TCnH

TCNTn

OCRnx

10-BIT OCRnx

10-BIT TCNTn

=

(10-bit Comparator )

OCFnx (Int.Req.)

TOP

BOTTOM

FOCn

PWMnx

Waveform Generator

WGMn0

COMnX1:0

OCWnx

The OCR4x Registers are double buffered when using any of the Pulse Width Modulation

(PWM) modes. For the normal mode of operation, the double buffering is disabled. The double

buffering synchronizes the update of the OCR4x Compare Registers to either top or bottom of

the counting sequence. The synchronization prevents the occurrence of odd-length, non-sym-

metrical PWM pulses, thereby making the output glitch-free. See Figure 15-5 for an example.

During the time between the write and the update operation, a read from OCR4A, OCR4B,

OCR4C or OCR4D will read the contents of the temporary location. This means that the most

recently written value always will read out of OCR4A, OCR4B, OCR4C or OCR4D.

143

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 15-5. Effects of Unsynchronized OCR Latching

Compare Value changes

Counter Value

Compare Value

Output Compare

Waveform OCWnx

Synchronized WFnx Latch

Unsynchronized WFnx Latch

Compare Value changes

Counter Value

Compare Value

Output Compare

Wafeform OCWnx

Glitch

15.4.1

Force Output Compare

In non-PWM waveform generation modes, the match output of the comparator can be forced by

writing a one to the Force Output Compare (FOC4x) bit. Forcing Compare Match will not set the

OCF4x Flag or reload/clear the timer, but the Waveform Output (OCW4x) will be updated as if a

real Compare Match had occurred (the COM4x1:0 bits settings define whether the Waveform

Output (OCW4x) is set, cleared or toggled).

15.4.2

15.4.3

Compare Match Blocking by TCNT4 Write

All CPU write operations to the TCNT4 Register will block any Compare Match that occur in the

next timer clock cycle, even when the timer is stopped. This feature allows OCR4x to be initial-

ized to the same value as TCNT4 without triggering an interrupt when the Timer/Counter clock is

enabled.

Using the Output Compare Unit

Since writing TCNT4 in any mode of operation will block all Compare Matches for one timer

clock cycle, there are risks involved when changing TCNT4 when using the Output Compare

Unit, independently of whether the Timer/Counter is running or not. If the value written to TCNT4

equals the OCR4x value, the Compare Match will be missed, resulting in incorrect waveform

generation. Similarly, do not write the TCNT4 value equal to BOTTOM when the counter is

down-counting.

The setup of the Waveform Output (OCW4x) should be performed before setting the Data Direc-

tion Register for the port pin to output. The easiest way of setting the OCW4x value is to use the

Force Output Compare (FOC4x) strobe bits in Normal mode. The OC4x keeps its value even

when changing between Waveform Generation modes.

Be aware that the COM4x1:0 bits are not double buffered together with the compare value.

Changing the COM4x1:0 bits will take effect immediately.

144

7766E–AVR–04/10

ATmega16U4/ATmega32U4

15.5 Dead Time Generator

The Dead Time Generator is provided for the Timer/Counter4 PWM output pairs to allow driving

external power control switches safely. The Dead Time Generator is a separate block that can

be used to insert dead times (non-overlapping times) for the Timer/Counter4 complementary

output pairs OC4x and OC4x when the PWM mode is enabled and the COM4x1:0 bits are set to

“01”. The sharing of tasks is as follows: the Waveform Generator generates the Waveform Out-

put (OCW4x) and the Dead Time Generator generates the non-overlapping PWM output pair

from the Waveform Output. Three Dead Time Generators are provided, one for each PWM out-

put. The non-overlap time is adjustable and the PWM output and it’s complementary output are

adjusted separately, and independently for both PWM outputs.

Figure 15-6. Output Compare Unit, Block Diagram

OCnx

top

OCWnx

bottom

FOCn

Waveform Generator

Dead Time Generator

OCnx

CK OR PCK

CLOCK

PWMnx WGMn0 COMnx

DTPSn

DTnH

DTnL

The Dead Time Generation is based on the 4-bit down counters that count the dead time, as

shown in Figure 15-7. There is a dedicated prescaler in front of the Dead Time Generator that

can divide the Timer/Counter4 clock (PCK or CK) by 1, 2, 4 or 8. This provides for large range of

dead times that can be generated. The prescaler is controlled by two control bits DTPS41..40.

The block has also a rising and falling edge detector that is used to start the dead time counting

period. Depending on the edge, one of the transitions on the rising edges, OC4x or OC4x is

delayed until the counter has counted to zero. The comparator is used to compare the counter

with zero and stop the dead time insertion when zero has been reached. The counter is loaded

with a 4-bit DT4H or DT4L value from DT4 I/O register, depending on the edge of the Waveform

Output (OCW4x) when the dead time insertion is started. The Output Compare Output are

delayed by one timer clock cycle at minimum from the Waveform Output when the Dead Time is

adjusted to zero. The outputs OC4x and OC4x are inverted, if the PWM Inversion Mode bit

PWM4X is set. This will also cause both outputs to be high during the dead time.

Figure 15-7. Dead Time Generator

PWMnX

COMPARATOR

OCnx

CK OR PCK

CLOCK

DEAD TIME

CLOCK CONTROL

4-BIT COUNTER

PRE-SCALER

OCnx

PWMnX

TCCRnB REGISTER

DTn I/O REGISTER

OCWnx

DATA BUS (8-bit)

145

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The length of the counting period is user adjustable by selecting the dead time prescaler setting

by using the DTPS41:40 control bits, and selecting then the dead time value in I/O register DT4.

The DT4 register consists of two 4-bit fields, DT4H and DT4L that control the dead time periods

of the PWM output and its' complementary output separately in terms of the number of pres-

caled dead time generator clock cycles. Thus the rising edge of OC4x and OC4x can have

different dead time periods as the t_{non-overlap / rising edge}is adjusted by the 4-bit DT4H value and the

t_{non-overlap / falling edge}is adjusted by the 4-bit DT4L value.

Figure 15-8. The Complementary Output Pair, COM4x1:0 = 1

OCWnx

OCnx

(COMnx = 1)

t _{non-overlap / rising edge}t _{non-overlap / falling edge}

15.6 Compare Match Output Unit

The Compare Output Mode (COM4x1:0) bits have two functions. The Waveform Generator uses

the COM4x1:0 bits for defining the inverted or non-inverted Waveform Output (OCW4x) at the

next Compare Match. Also, the COM4x1:0 bits control the OC4x and OC4x pin output source.

Figure 15-9 shows a simplified schematic of the logic affected by the COM4x1:0 bit setting. The

I/O Registers, I/O bits, and I/O pins in the figure are shown in bold. Only the parts of the general

I/O Port Control Registers (DDR and PORT) that are affected by the COM4x1:0 bits are shown.

In Normal Mode (non-PWM) the Dead Time Generator is disabled and it is working like a syn-

chronizer: the Output Compare (OC4x) is delayed from the Waveform Output (OCW4x) by one

timer clock cycle. Whereas in Fast PWM Mode and in Phase and Frequency Correct PWM

Mode when the COM4x1:0 bits are set to “01” both the non-inverted and the inverted Output

Compare output are generated, and an user programmable Dead Time delay is inserted for

these complementary output pairs (OC4x and OC4x). The functionality in PWM modes is similar

to Normal mode when any other COM4x1:0 bit setup is used. When referring to the OC4x state,

the reference is for the Output Compare output (OC4x) from the Dead Time Generator, not the

OC4x pin. If a system reset occur, the OC4x is reset to “0”.

The general I/O port function is overridden by the Output Compare (OC4x / OC4x) from the

Dead Time Generator if either of the COM4x1:0 bits are set. However, the OC4x pin direction

(input or output) is still controlled by the Data Direction Register (DDR) for the port pin. The Data

Direction Register bit for the OC4x and OC4x pins (DDR_OC4x and DDR_OC4x) must be set as

output before the OC4x and OC4x values are visible on the pin. The port override function is

independent of the Output Compare mode.

The design of the Output Compare Pin Configuration logic allows initialization of the OC4x state

before the output is enabled. Note that some COM4x1:0 bit settings are reserved for certain

modes of operation. For Output Compare Pin Configurations refer to Table 15-2 on page 151,

Table 15-3 on page 152, Table 15-4 on page 154, and Table 15-5 on page 155.

146

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 15-9. Compare Match Output Unit, Schematic

WGM41

clk

I/O

OC4OE1:0

COM4A1:0

Output Compare

Pin Configuration

D

Q

PORTC6

0

1

OC4A

1

D

Q

0

DDRC6

OC4A

OCW4A

D

Q

Dead Time

Generator A

clk

Tn

PORTC7

1

0

OC4A

D

Q

DDRC7

WGM41

OC4OE3:2

COM4B1:0

Output Compare

Pin Configuration

Q

D

PORTB5

2

1

0

1

0

OC4B

Q

D

DDRB5

OCW4B

OC4B

1

0

Q

D

Q

Dead Time

Generator B

clk

Tn

1

0

PORTB6

OC4B

Q

D

DDRB6

WGM41

OC4OE5:4

COM4D1:0

Output Compare

Pin Configuration

D

Q

PORTD6

2

1

0

1

0

OC4D

D

Q

DDRD6

OCW4D

OC4

OC4D

1

0

D

Q

Dead Time

Generator D

clk

Tn

1

0

PORTD7

OC4D

D

Q

DDRD7

15.6.1

Compare Output Mode and Waveform Generation

The Waveform Generator uses the COM4x1:0 bits differently in Normal mode and PWM modes.

For all modes, setting the COM4x1:0 = 0 tells the Waveform Generator that no action on the

OCW4x Output is to be performed on the next Compare Match. For compare output actions in

the non-PWM modes refer to Table 15-6 on page 162. For fast PWM mode, refer to Table 15-7

on page 162, and for the Phase and Frequency Correct PWM refer to Table 15-8 on page 163.

A change of the COM4x1:0 bits state will have effect at the first Compare Match after the bits are

written. For non-PWM modes, the action can be forced to have immediate effect by using the

FOC4x strobe bits.

147

7766E–AVR–04/10

ATmega16U4/ATmega32U4

15.6.2

Enhanced Compare/PWM mode

When the bit ENHC4 of TCCR4E register is set, the Enhanced Compare/PWM mode is enabled.

This mode allows user to add an accuracy bit to Output Compare Register OCR4A, OCR4B and

OCR4D. Like explained previously, a compare condition appears when one of the three Output

Compare Registers (OCR4A/B/D) matches the value of TCNT4 (10-bits resolution). In basic

PWM Mode, the corresponding enabled output toggles on the Compare Match. The Enhanced

Compare/PWM mode introduces a bit that determines on which internal clock edge the Com-

pare Match condition is actually signalled. That means that the corresponding outputs will toggle

on the standard clock edge (like in Normal mode) if the LSB of OCR4A/B/D is ‘0’, or on the oppo-

site (next) edge if the LSB is ‘1’.

User will notice that between Normal and Enhanced PWM modes, the output frequency will be

identical, while the PWM resolution will be better in second case.

Writing to the Output Compare registers OCR4A/B/D or reading them will be identical in both

modes. In Enhanced mode, user must just consider that the TC4H register can be up to 3-bits

wide (and have the same behavior than during 2-bits operation). That will concern OCR4A,

OCR4B and OCR4D registers accesses only. Indeed, the OCR4C register must not include the

additional accuracy bit, and remains in the resolution that determines the output signal period.

Figure 15-10. How register access works in Enhanced mode

(TC4H)

(OCR4A/B/D)

7

9

6

5

4

3

2

1

0

10

8

9

User Interface Side

Timer Logic Side

True

5

4

3

2

8

OCR4A/B/D

Output Compare Module A/B/D

Waveform Generation

TCNT4<9:0>

Enhanced

Mode

OCR4C<9:0>

Configuration

bits

ENHC4

Pin Toggle

That figure shows that the true OCR4A/B/D value corresponds to the value loaded by the user

shifted on the right in order to transfer the least significant bit directly to the Waveform genera-

tion module.

The maximum available resolution is 11-bits, but any other resolution can be specified. For

example, a 8-bits resolution will allow to obtain the same frequency than a Normal PWM mode

with 7-bits resolution.

Example:

– PLL Postcaler output = 64 MHz, No Prescaler on Timer/Counter4.

– Setting OCR4C = 0x7F determines a full 7-bits theoretical resolution, and so a

500kHz output frequency.

148

7766E–AVR–04/10

ATmega16U4/ATmega32U4

– Setting OCR4A = 0x85 (= b’10000101’) signifies that the true value of “Compare A”

register is 0x42 (b’01000010’) and that the Enhanced bit is set. That means that the

duty cycle obtained (51.95%) will be the intermediate value between duty cycles that

can be obtained by 0x42 and 0x43 Compare values (51.56%, 52.34%).

15.7 Synchronous update

To avoid unasynchronous and incoherent values in a cycle, if a synchronous update of one of

several values is necessary, all values can be updated at the same time at the end of the PWM

cycle by the Timer controller. The new set of values is calculated by software and the effective

update can be initiated by software.

Figure 15-11. Lock feature and Synchronous update

TLOCK4=1

TLOCK4=0

Regulation Loop

Calculation

Request for an

Update

Writing to Timer

Registers Set j

Cycle with

Set j

Cycle with Cycle with

Set i Set i

Cycle with

Set i

Cycle with

Set i

In normal operation, each write to a Compare register is effective at the end of the current cycle.

But some cases require that two or more Compare registers are updated synchronously, and

that may not be always possible, mostly at high speed PWM frequencies. That may result in

some PWM periods with incoherent values.

When using the Lock feature (TLOCK4=1), the values written to the Compare registers are not

effective and temporarily buffered. When releasing the TLOCK4 bit, the update is initiated and

the new whole set of values will be loaded at the end of the current PWM cycle.

See Section 15.12.5 ”TCCR4E – Timer/Counter4 Control Register E” on page 169.

15.8 Modes of Operation

The mode of operation, i.e., the behavior of the Timer/Counter and the Output Compare pins, is

defined by the combination of the Waveform Generation mode (bits PWM4x and WGM40) and

Compare Output mode (COM4x1:0) bits. The Compare Output mode bits do not affect the

counting sequence, while the Waveform Generation mode bits do. The COM4x1:0 bits control

whether the PWM output generated should be inverted, non-inverted or complementary. For

non-PWM modes the COM4x1:0 bits control whether the output should be set, cleared, or tog-

gled at a Compare Match.

15.8.1

Normal Mode

The simplest mode of operation is the Normal mode (PWM4x = 0), the counter counts from

BOTTOM to TOP (defined as OCR4C) then restarts from BOTTOM. The OCR4C defines the

TOP value for the counter, hence also its resolution, and allows control of the Compare Match

output frequency. In toggle Compare Output Mode the Waveform Output (OCW4x) is toggled at

Compare Match between TCNT4 and OCR4x. In non-inverting Compare Output Mode the

149

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Waveform Output is cleared on the Compare Match. In inverting Compare Output Mode the

Waveform Output is set on Compare Match.

The timing diagram for the Normal mode is shown in Figure 15-12. The counter value (TCNT4)

that is shown as a histogram in the timing diagram is incremented until the counter value

matches the TOP value. The counter is then cleared at the following clock cycle The diagram

includes the Waveform Output (OCW4x) in toggle Compare Mode. The small horizontal line

marks on the TCNT4 slopes represent Compare Matches between OCR4x and TCNT4.

Figure 15-12. Normal Mode, Timing Diagram

TOVn Interrupt Flag Set

OCnx Interrupt Flag Set

TCNTn

OCWnx

(COMnx=1)

1

2

3

4

Period

The Timer/Counter Overflow Flag (TOV4) is set in the same clock cycle as the TCNT4 becomes

zero. The TOV4 Flag in this case behaves like a 11th bit, except that it is only set, not cleared.

However, combined with the timer overflow interrupt, that automatically clears the TOV4 Flag,

the timer resolution can be increased by software. There are no special cases to consider in the

Normal mode, a new counter value can be written anytime.

The Output Compare Unit can be used to generate interrupts at some given time. Using the Out-

put Compare to generate waveforms in Normal mode is not recommended, since this will

occupy too much of the CPU time. For generating a waveform, the OCW4x output can be set to

toggle its logical level on each Compare Match by setting the Compare Output mode bits to tog-

gle mode (COM4x1:0 = 1). The OC4x value will not be visible on the port pin unless the data

direction for the pin is set to output. The waveform generated will have a maximum frequency of

f

_OC4x= f_clkT4/4 when OCR4C is set to zero. The waveform frequency is defined by the following

equation:

f

clkT4

f

= ------------------------------------------

OC4x

2 ⋅ (1 + OCR4C)

Resolution shows how many bit is required to express the value in the OCR4C register. It is cal-

culated by following equation:

Resolution_PWM= log₂(OCR4C + 1).

150

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The Output Compare Pin configurations in Normal Mode are described in Table 15-2.

Table 15-2.

Output Compare Pin Configurations in Normal Mode

COM4x1

COM4x0

OC4x Pin

Disconnected

OC4x

0

1

0

1

0

1

Disconnected

OC4x

15.8.2

Fast PWM Mode

The fast Pulse Width Modulation or fast PWM mode (PWM4x = 1 and WGM40 = 0) provides a

high frequency PWM waveform generation option. The fast PWM differs from the other PWM

option by its single-slope operation. The counter counts from BOTTOM to TOP (defined as

OCR4C) then restarts from BOTTOM. In non-inverting Compare Output mode the Waveform

Output (OCW4x) is cleared on the Compare Match between TCNT4 and OCR4x and set at

BOTTOM. In inverting Compare Output mode, the Waveform Output is set on Compare Match

and cleared at BOTTOM. In complementary Compare Output mode the Waveform Output is

cleared on the Compare Match and set at BOTTOM.

Due to the single-slope operation, the operating frequency of the fast PWM mode can be twice

as high as the Phase and Frequency Correct PWM mode that use dual-slope operation. This

high frequency makes the fast PWM mode well suited for power regulation, rectification, and

DAC applications. High frequency allows physically small sized external components (coils,

capacitors), and therefore reduces total system cost.

The timing diagram for the fast PWM mode is shown in Figure 15-13. The counter is incre-

mented until the counter value matches the TOP value. The counter is then cleared at the

following timer clock cycle. The TCNT4 value is in the timing diagram shown as a histogram for

illustrating the single-slope operation. The diagram includes the Waveform Output in non-

inverted and inverted Compare Output modes. The small horizontal line marks on the TCNT4

slopes represent Compare Matches between OCR4x and TCNT4.

Figure 15-13. Fast PWM Mode, Timing Diagram

OCRnx Interrupt Flag Set

OCRnx Update and

TOVn Interrupt Flag Set

TCNTn

OCWnx

(COMnx1:0 = 2)

OCWnx

(COMnx1:0 = 3)

1

2

3

4

5

6

7

Period

151

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The Timer/Counter Overflow Flag (TOV4) is set each time the counter reaches TOP. If the inter-

rupt is enabled, the interrupt handler routine can be used for updating the compare value.4In

fast PWM mode, the compare unit allows generation of PWM waveforms on the OC4x pins. Set-

ting the COM4x1:0 bits to two will produce a non-inverted PWM and setting the COM4x1:0 to

three will produce an inverted PWM output. Setting the COM4x1:0 bits to one will enable com-

plementary Compare Output mode and produce both the non-inverted (OC4x) and inverted

output (OC4x). The actual value will only be visible on the port pin if the data direction for the

port pin is set as output. The PWM waveform is generated by setting (or clearing) the Waveform

Output (OCW4x) at the Compare Match between OCR4x and TCNT4, and clearing (or setting)

the Waveform Output at the timer clock cycle the counter is cleared (changes from TOP to

BOTTOM).

The PWM frequency for the output can be calculated by the following equation:

f

clkT4

f

= ------------

OCnxPWM

N

The N variable represents the number of steps in single-slope operation. The value of N equals

either to the TOP value.

The extreme values for the OCR4C Register represents special cases when generating a PWM

waveform output in the fast PWM mode. If the OCR4C is set equal to BOTTOM, the output will

be a narrow spike for each MAX+1 timer clock cycle. Setting the OCR4C equal to MAX will result

in a constantly high or low output (depending on the polarity of the output set by the COM4x1:0

bits.)

A frequency (with 50% duty cycle) waveform output in fast PWM mode can be achieved by set-

ting the Waveform Output (OCW4x) to toggle its logical level on each Compare Match

(COM4x1:0 = 1). The waveform generated will have a maximum frequency of f_OC4= f_clkT4/4 when

OCR4C is set to three.

The general I/O port function is overridden by the Output Compare value (OC4x / OC4x) from

the Dead Time Generator, if either of the COM4x1:0 bits are set and the Data Direction Register

bits for the OC4X and OC4X pins are set as an output. If the COM4x1:0 bits are cleared, the

actual value from the port register will be visible on the port pin. The Output Compare Pin config-

urations are described in Table 15-3.

Table 15-3.

Output Compare Pin Configurations in Fast PWM Mode

COM4x1

COM4x0

OC4x Pin

Disconnected

OC4x

0

1

0

1

0

1

Disconnected

OC4x

Disconnected

OC4x

15.8.3

Phase and Frequency Correct PWM Mode

The Phase and Frequency Correct PWM Mode (PWM4x = 1 and WGM40 = 1) provides a high

resolution Phase and Frequency Correct PWM waveform generation option. The Phase and

Frequency Correct PWM mode is based on a dual-slope operation. The counter counts repeat-

edly from BOTTOM to TOP (defined as OCR4C) and then from TOP to BOTTOM. In non-

inverting Compare Output Mode the Waveform Output (OCW4x) is cleared on the Compare

152

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Match between TCNT4 and OCR4x while upcounting, and set on the Compare Match while

down-counting. In inverting Output Compare mode, the operation is inverted. In complementary

Compare Output Mode, the Waveform Output is cleared on the Compare Match and set at BOT-

TOM. The dual-slope operation has lower maximum operation frequency than single slope

operation. However, due to the symmetric feature of the dual-slope PWM modes, these modes

are preferred for motor control applications.

The timing diagram for the Phase and Frequency Correct PWM mode is shown on Figure 15-14

in which the TCNT4 value is shown as a histogram for illustrating the dual-slope operation. The

counter is incremented until the counter value matches TOP. When the counter reaches TOP, it

changes the count direction. The TCNT4 value will be equal to TOP for one timer clock cycle.

The diagram includes the Waveform Output (OCW4x) in non-inverted and inverted Compare

Output Mode. The small horizontal line marks on the TCNT4 slopes represent Compare

Matches between OCR4x and TCNT4.

Figure 15-14. Phase and Frequency Correct PWM Mode, Timing Diagram

OCnx Interrupt Flag Set

OCRnx Update

TOVn Interrupt Flag Set

TCNTn

OCWnx

(COMnx = 2)

OCWnx

(COMnx = 3)

1

2

3

Period

The Timer/Counter Overflow Flag (TOV4) is set each time the counter reaches BOTTOM. The

Interrupt Flag can be used to generate an interrupt each time the counter reaches the BOTTOM

value.

In the Phase and Frequency Correct PWM mode, the compare unit allows generation of PWM

waveforms on the OC4x pins. Setting the COM4x1:0 bits to two will produce a non-inverted

PWM and setting the COM4x1:0 to three will produce an inverted PWM output. Setting the

COM4A1:0 bits to one will enable complementary Compare Output mode and produce both the

non-inverted (OC4x) and inverted output (OC4x). The actual values will only be visible on the

port pin if the data direction for the port pin is set as output. The PWM waveform is generated by

clearing (or setting) the Waveform Output (OCW4x) at the Compare Match between OCR4x and

TCNT4 when the counter increments, and setting (or clearing) the Waveform Output at Compare

Match when the counter decrements. The PWM frequency for the output when using the Phase

and Frequency Correct PWM can be calculated by the following equation:

f

clkT4

f

= ------------

OCnxPCPWM

N

153

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The N variable represents the number of steps in dual-slope operation. The value of N equals to

the TOP value.

The extreme values for the OCR4C Register represent special cases when generating a PWM

waveform output in the Phase and Frequency Correct PWM mode. If the OCR4C is set equal to

BOTTOM, the output will be continuously low and if set equal to MAX the output will be continu-

ously high for non-inverted PWM mode. For inverted PWM the output will have the opposite

logic values.

The general I/O port function is overridden by the Output Compare value (OC4x / OC4x) from

the Dead Time Generator, if either of the COM4x1:0 bits are set and the Data Direction Register

bits for the OC4X and OC4X pins are set as an output. If the COM4x1:0 bits are cleared, the

actual value from the port register will be visible on the port pin. The configurations of the Output

Compare Pins are described in Table 15-4.

Table 15-4. Output Compare pin configurations in Phase and Frequency Correct PWM Mode

COM4x1

COM4x0

OC4x Pin

Disconnected

OC4x

0

1

0

1

0

1

Disconnected

OC4x

Disconnected

OC4x

15.8.4

PWM6 Mode

The PWM6 Mode (PWM4A = 1, WGM41 = 1 and WGM40 = x) provide PWM waveform genera-

tion option e.g. for controlling Brushless DC (BLDC) motors. In the PWM6 Mode the OCR4A

Register controls all six Output Compare waveforms as the same Waveform Output (OCW4A)

from the Waveform Generator is used for generating all waveforms. The PWM6 Mode also pro-

vides an Output Compare Override Enable Register (OC4OE) that can be used with an instant

response for disabling or enabling the Output Compare pins. If the Output Compare Override

Enable bit is cleared, the actual value from the port register will be visible on the port pin.

The PWM6 Mode provides two counter operation modes, a single-slope operation and a dual-

slope operation. If the single-slope operation is selected (the WGM40 bit is set to 0), the counter

counts from BOTTOM to TOP (defined as OCR4C) then restart from BOTTOM like in Fast PWM

Mode. The PWM waveform is generated by setting (or clearing) the Waveform Output (OCW4A)

at the Compare Match between OCR4A and TCNT4, and clearing (or setting) the Waveform

Output at the timer clock cycle the counter is cleared (changes from TOP to BOTTOM). The

Timer/Counter Overflow Flag (TOV4) is set each time the counter reaches the TOP and, if the

interrupt is enabled, the interrupt handler routine can be used for updating the compare value.

Whereas, if the dual-slope operation is selected (the WGM40 bit is set to 1), the counter counts

repeatedly from BOTTOM to TOP (defined as OCR4C) and then from TOP to BOTTOM like in

Phase and Frequency Correct PWM Mode. The PWM waveform is generated by setting (or

clearing) the Waveform Output (OCW4A) at the Compare Match between OCR4A and TCNT4

when the counter increments, and clearing (or setting) the Waveform Output at the he Compare

Match between OCR4A and TCNT4 when the counter decrements. The Timer/Counter Overflow

Flag (TOV4) is set each time the counter reaches the BOTTOM and, if the interrupt is enabled,

the interrupt handler routine can be used for updating the compare value.

The timing diagram for the PWM6 Mode in single-slope operation (WGM41 = 0) when the

COM4A1:0 bits are set to “10” is shown in Figure 15-15. The counter is incremented until the

154

7766E–AVR–04/10

ATmega16U4/ATmega32U4

counter value matches the TOP value. The counter is then cleared at the following timer clock

cycle. The TCNT4 value is in the timing diagram shown as a histogram for illustrating the single-

slope operation. The timing diagram includes Output Compare pins OC4A and OC4A, and the

corresponding Output Compare Override Enable bits (OC4OE1..OC4OE0).

Figure 15-15. PWM6 Mode, Single-slope Operation, Timing Diagram

TCNT4

OCW4A

OC4OE0

OC4A Pin

OC4OE1

OC4A Pin

OC4OE2

OC4B Pin

OC4OE3

OC4B Pin

OC4OE4

OC4D Pin

OC4OE5

OC4D Pin

The general I/O port function is overridden by the Output Compare value (OC4x / OC4x) from

the Dead Time Generator if either of the COM4x1:0 bits are set. The Output Compare pins can

also be overridden by the Output Compare Override Enable bits OC4OE5..OC4OE0. If an Over-

ride Enable bit is cleared, the actual value from the port register will be visible on the port pin

and, if the Override Enable bit is set, the Output Compare pin is allowed to be connected on the

port pin. The Output Compare Pin configurations are described in Table 15-5.

Table 15-5. Output Compare Pin configurations in PWM6 Mode

COM4A1

COM4A0

OC4A Pin (PC6)

Disconnected

OC4A Pin (PC7)

Disconnected

0

1

OC4A • OC4OE0

OC4B Pin (PB5)

Disconnected

OC4A • OC4OE1

OC4B Pin (PB6)

Disconnected

1

0

1

COM4B1

COM4B0

0

1

OC4A • OC4OE2

OC4A • OC4OE3

155

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 15-5. Output Compare Pin configurations in PWM6 Mode (Continued)

COM4A1

COM4A0

OC4A Pin (PC6)

OC4A • OC4OE2

OC4D Pin (PD6)

Disconnected

OC4A Pin (PC7)

OC4A • OC4OE3

OC4D Pin (PD7)

Disconnected

1

0

1

COM4D1

COM4D0

0

1

0

1

0

1

OC4A • OC4OE4

OC4A • OC4OE5

15.9 Timer/Counter Timing Diagrams

The Timer/Counter is a synchronous design and the timer clock (clk_T4) is therefore shown as a

clock enable signal in the following figures. The figures include information on when Interrupt

Flags are set.

Figure 15-16 contains timing data for basic Timer/Counter operation. The figure shows the count

sequence close to the MAX value in all modes other than Phase and Frequency Correct PWM

Mode. Figure 15-17 shows the same timing data, but with the prescaler enabled, in all modes

other than Phase and Frequency Correct PWM Mode. Figure 15-18 shows the setting of

OCF4A, OCF4B and OCF4D in all modes, and Figure 15-19 shows the setting of TOV4 in

Phase and Frequency Correct PWM Mode.

Figure 15-16. Timer/Counter Timing Diagram, no Prescaling

clk_PCK

clk_Tn

(clk_PCK/1)

TCNTn

TOVn

TOP - 1

TOP

BOTTOM

BOTTOM + 1

Figure 15-17. Timer/Counter Timing Diagram, with Prescaler (f_clkT4/8)

clk_PCK

clk_Tn

(clk_PCK/8)

TCNTn

TOVn

TOP - 1

TOP

BOTTOM

BOTTOM + 1

156

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 15-18. Timer/Counter Timing Diagram, Setting of OCF1x, with Prescaler (f_clkT4/8)

clk_PCK

clk_Tn

(clk_PCK/8)

TCNTn

OCRnx

OCFnx

OCRnx - 1

OCRnx

OCRnx + 1

OCRnx + 2

OCRnx Value

Figure 15-19. Timer/Counter Timing Diagram, with Prescaler (f_clkT4/8)

clk_PCK

clk_Tn

(clk_PCK/8)

TCNTn

BOTTOM + 1

BOTTOM

BOTTOM + 1

TOVn

15.10 Fault Protection Unit

The Timer/Counter4 incorporates a Fault Protection unit that can disable the PWM output pins, if

an external event is triggered. The external signal indicating an event can be applied via the

external interrupt INT0 pin or alternatively, via the analog-comparator unit. The Fault Protection

unit is illustrated by the block diagram shown in Figure 15-20. The elements of the block diagram

that are not directly a part of the Fault Protection unit are gray shaded.

Figure 15-20. Fault Protection Unit Block Diagram

FAULT_PROTECTION (Int. Req.)

ACO*

FPAC4

FPNC4

FPES4

FPEN4

Analog

Comparator

Noise

Canceler

Edge

Detector

Timer/Counter4

INT0

When the Fault Protection mode is enabled by the Fault Protection Enable (FPEN4) bit and a

change of the logic level (an event) occurs on the external interrupt pin (INT0), alternatively on

the Analog Comparator output (ACO), and this change confirms to the setting of the edge detec-

tor, a Fault Protection mode will be triggered. When a Fault Protection is triggered, the COM4x

bits are cleared, Output Comparators are disconnected from the PWM output pins and the

PORTB register bits are connected on the PWM output pins. The Fault Protection Enable

(FPEN4) is automatically cleared at the same system clock as the COM4nx bits are cleared. If

the Fault Protection Interrupt Enable bit (FPIE4) is set, a Fault Protection interrupt is generated

and the FPEN4 bit is cleared. Alternatively the FPEN4 bit can be polled by software to figure out

when the Timer/Counter has entered to Fault Protection mode.

157

7766E–AVR–04/10

ATmega16U4/ATmega32U4

15.10.1 Fault Protection Trigger Source

The main trigger source for the Fault Protection unit is the external interrupt pin (INT0). Alterna-

tively the Analog Comparator output can be used as trigger source for the Fault Protection unit.

The Analog Comparator is selected as trigger source by setting the Fault Protection Analog

Comparator (FPAC4) bit in the Timer/Counter4 Control Register (TCCR4D). Be aware that

changing trigger source can trigger a Fault Protection mode. Therefore it is recommended to

clear the FPF4 flag after changing trigger source, setting edge detector or enabling the Fault

Protection.

Both the external interrupt pin (INT0) and the Analog Comparator output (ACO) inputs are sam-

pled using the same technique as for the T0 pin (Figure 12-1 on page 89). The edge detector is

also identical. However, when the noise canceler is enabled, additional logic is inserted before

the edge detector, which increases the delay by four system clock cycles. An Input Capture can

also be triggered by software by controlling the port of the INT0 pin.

15.10.2 Noise Canceler

The noise canceler improves noise immunity by using a simple digital filtering scheme. The

noise canceler input is monitored over four samples, and all four must be equal for changing the

output that in turn is used by the edge detector.

The noise canceler is enabled by setting the Fault Protection Noise Canceler (FPNC4) bit in

Timer/Counter4 Control Register D (TCCR4D). When enabled the noise canceler introduces

additional four system clock cycles of delay from a change applied to the input. The noise can-

celer uses the system clock and is therefore not affected by the prescaler.

158

7766E–AVR–04/10

ATmega16U4/ATmega32U4

15.11 Accessing 10-Bit Registers

If 10-bit values are written to the TCNTn and OCRnA/B/C/D registers, the 10-bit registers can be

byte accessed by the AVR CPU via the 8-bit data bus using two read or write operations. The

10-bit registers have a common 2-bit Timer/Counter4 High Byte Register (TC4H) that is used for

temporary storing of the two MSBs of the 10-bit access. The same TC4H register is shared

between all 10-bit registers. Accessing the low byte triggers the 10-bit read or write operation.

When the low byte of a 10-bit register is written by the CPU, the high byte stored in the TC4H

register, and the low byte written are both copied into the 10-bit register in the same clock cycle.

When the low byte of a 10-bit register is read by the CPU, the high byte of the 10-bit register is

copied into the TC4H register in the same clock cycle as the low byte is read.

To do a 10-bit write, the high byte must be written to the TC4H register before the low byte is

written. For a 10-bit read, the low byte must be read before the high byte.

The following code examples show how to access the 10-bit timer registers assuming that no

interrupts updates the TC4H register. The same principle can be used directly for accessing the

OCRnA/B/C/C/D registers.

Assembly Code Example

...

; Set TCNTn to 0x01FF

ldir17,0x01

ldir16,0xFF

outTCnH,r17

outTCNTn,r16

; Read TCNTn into r17:r16

in r16,TCNTn

in r17,TCnH

...

C Code Example

unsigned int i;

...

/* Set TCNTn to 0x01FF */

TCnH = 0x01;

TCNTn = 0xFF;

/* Read TCNTn into i */

i = TCNTn;

i |= ((unsigned int)TCnH << 8);

...

Note:

1. The example code assumes that the part specific header file is included.

For I/O registers located in extended I/O map, “IN”, “OUT”, “SBIS”, “SBIC”, “CBI”, and “SBI”

instructions must be replaced with instructions that allow access to extended I/O. Typically

“LDS” and “STS” combined with “SBRS”, “SBRC”, “SBR”, and “CBR”.

The assembly code example returns the TCNTn value in the r17:r16 register pair.

159

7766E–AVR–04/10

ATmega16U4/ATmega32U4

It is important to notice that accessing 10-bit registers are atomic operations. If an interrupt

occurs between the two instructions accessing the 10-bit register, and the interrupt code

updates the TC4H register by accessing the same or any other of the 10-bit timer registers, then

the result of the access outside the interrupt will be corrupted. Therefore, when both the main

code and the interrupt code update the TC4H register, the main code must disable the interrupts

during the 16-bit access.

The following code examples show how to do an atomic read of the TCNTn register contents.

Reading any of the OCRnA/B/C/D registers can be done by using the same principle.

Assembly Code Example

TIM1_ReadTCNTn:

; Save global interrupt flag

in r18,SREG

; Disable interrupts

cli

; Read TCNTn into r17:r16

in r16,TCNTn

in r17,TCnH

; Restore global interrupt flag

outSREG,r18

ret

C Code Example

unsigned int TIM1_ReadTCNTn( void )

{

unsigned char sreg;

unsigned int i;

/* Save global interrupt flag */

sreg = SREG;

/* Disable interrupts */

_CLI();

/* Read TCNTn into i */

i = TCNTn;

i |= ((unsigned int)TCnH << 8);

/* Restore global interrupt flag

SREG = sreg;

return i;

}

Note:

1. The example code assumes that the part specific header file is included.

For I/O registers located in extended I/O map, “IN”, “OUT”, “SBIS”, “SBIC”, “CBI”, and “SBI”

instructions must be replaced with instructions that allow access to extended I/O. Typically

“LDS” and “STS” combined with “SBRS”, “SBRC”, “SBR”, and “CBR”.

The assembly code example returns the TCNTn value in the r17:r16 register pair.

160

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The following code examples show how to do an atomic write of the TCNTn register contents.

Writing any of the OCRnA/B/C/D registers can be done by using the same principle.

Assembly Code Example

TIM1_WriteTCNTn:

; Save global interrupt flag

in r18,SREG

; Disable interrupts

cli

; Set TCNTn to r17:r16

outTCnH,r17

outTCNTn,r16

; Restore global interrupt flag

outSREG,r18

ret

C Code Example

void TIM1_WriteTCNTn( unsigned int i )

{

unsigned char sreg;

unsigned int i;

/* Save global interrupt flag */

sreg = SREG;

/* Disable interrupts */

_CLI();

/* Set TCNTn to i */

TCnH = (i >> 8);

TCNTn = (unsigned char)i;

/* Restore global interrupt flag */

SREG = sreg;

}

Note:

1. The example code assumes that the part specific header file is included.

For I/O registers located in extended I/O map, “IN”, “OUT”, “SBIS”, “SBIC”, “CBI”, and “SBI”

instructions must be replaced with instructions that allow access to extended I/O. Typically

“LDS” and “STS” combined with “SBRS”, “SBRC”, “SBR”, and “CBR”.

The assembly code example requires that the r17:r16 register pair contains the value to be writ-

ten to TCNTn.

15.11.1 Reusing the temporary high byte register

If writing to more than one 10-bit register where the high byte is the same for all registers written,

then the high byte only needs to be written once. However, note that the same rule of atomic

operation described previously also applies in this case.

161

7766E–AVR–04/10

ATmega16U4/ATmega32U4

15.12 Register Description

15.12.1 TCCR4A – Timer/Counter4 Control Register A

Bit

7

COM4A1

R/W

6

COM4A0

R/W

5

COM4B1

R/W

4

COM4B0

R/W

3

FOC4A

W

2

FOC4B

W

1

PWM4A

R/W

0

PWM4B

R/W

0

TCCR4A

Read/Write

Initial value

0

• Bits 7,6 - COM4A1, COM4A0: Comparator A Output Mode, Bits 1 and 0

These bits control the behavior of the Waveform Output (OCW4A) and the connection of the

Output Compare pin (OC4A). If one or both of the COM4A1:0 bits are set, the OC4A output

overrides the normal port functionality of the I/O pin it is connected to. The complementary

OC4B output is connected only in PWM modes when the COM4A1:0 bits are set to “01”. Note

that the Data Direction Register (DDR) bit corresponding to the OC4A and OC4A pins must be

set in order to enable the output driver.

The function of the COM4A1:0 bits depends on the PWM4A, WGM40 and WGM41 bit settings.

Table 15-6 shows the COM4A1:0 bit functionality when the PWM4A bit is set to Normal Mode

(non-PWM).

Table 15-6. Compare Output Mode, Normal Mode (non-PWM)

COM4A1..0

OCW4A Behavior

OC4A Pin

Disconnected

Connected

OC4A Pin

00

01

10

11

Normal port operation.

Toggle on Compare Match.

Clear on Compare Match.

Set on Compare Match.

Disconnected

Table 15-7 shows the COM4A1:0 bit functionality when the PWM4A, WGM40 and WGM41 bits

are set to fast PWM mode.

Table 15-7. Compare Output Mode, Fast PWM Mode

COM4A1..0

OCW4A Behavior

OC4A

00

Normal port operation.

Disconnected

Cleared on Compare Match.

Set when TCNT4 = 0x000.

01

10

11

Connected

Disconnected

Cleared on Compare Match.

Set when TCNT4 = 0x000.

Set on Compare Match.

Cleared when TCNT4 = 0x000.

162

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 15-8 shows the COM4A1:0 bit functionality when the PWM4A, WGM40 and WGM41 bits

are set to Phase and Frequency Correct PWM Mode.

Table 15-8. Compare Output Mode, Phase and Frequency Correct PWM Mode

COM1A1..0

OCW1A Behavior

OC4A Pin

00

Normal port operation.

Disconnected

Cleared on Compare Match when up-counting.

Set on Compare Match when down-counting.

01

10

11

Connected

Disconnected

Cleared on Compare Match when up-counting.

Set on Compare Match when down-counting.

Set on Compare Match when up-counting.

Cleared on Compare Match when down-counting.

Table 15-9 shows the COM4A1:0 bit functionality when the PWM4A, WGM40 and WGM41 bits

are set to single-slope PWM6 Mode. In the PWM6 Mode the same Waveform Output (OCW4A)

is used for generating all waveforms and the Output Compare values OC4A and OC4A are con-

nected on OC4x and OC4x pins as described below.

Table 15-9. Compare Output Mode, Single-Slope PWM6 Mode

COM4A1..0

OCW4A Behavior

OC4x Pin

00

Normal port operation.

Disconnected

Cleared on Compare Match.

Set when TCNT4 = 0x000.

01

10

11

OC4A

Cleared on Compare Match.

Set when TCNT4 = 0x000.

Set on Compare Match.

Cleared when TCNT4 = 0x000.

Table 15-10 shows the COM4A1:0 bit functionality when the PWM4A, WGM40 and WGM41 bits

are set to dual-slope PWM6 Mode.I

Table 15-10. Compare Output Mode, Dual-Slope PWM6 Mode

COM4A1..0

OCW4A Behavior

OC4x Pin

00

Normal port operation.

Disconnected

Cleared on Compare Match when up-counting.

Set on Compare Match when down-counting.

01

10

11

OC4A

Cleared on Compare Match when up-counting.

Set on Compare Match when down-counting.

Set on Compare Match when up-counting.

Cleared on Compare Match when down-counting.

• Bits 5,4 - COM4B1, COM4B0: Comparator B Output Mode, Bits 1 and 0

These bits control the behavior of the Waveform Output (OCW4B) and the connection of the

Output Compare pin (OC4B). If one or both of the COM4B1:0 bits are set, the OC4B output

overrides the normal port functionality of the I/O pin it is connected to. The complementary

OC4B output is connected only in PWM modes when the COM4B1:0 bits are set to “01”. Note

163

7766E–AVR–04/10

ATmega16U4/ATmega32U4

that the Data Direction Register (DDR) bit corresponding to the OC4B pin must be set in order to

enable the output driver.

The function of the COM4B1:0 bits depends on the PWM4B and WGM40 bit settings. Table 15-

11 shows the COM4B1:0 bit functionality when the PWM4B bit is set to Normal Mode (non-

PWM).

Table 15-11. Compare Output Mode, Normal Mode (non-PWM)

COM4B1..0

OCW4B Behavior

OC4B Pin

Disconnected

Connected

OC4B Pin

00

01

10

11

Normal port operation.

Toggle on Compare Match.

Clear on Compare Match.

Set on Compare Match.

Disconnected

Table 15-12 shows the COM4B1:0 bit functionality when the PWM4B and WGM40 bits are set to

Fast PWM Mode.

Table 15-12. Compare Output Mode, Fast PWM Mode

COM4B1..0

OCW4B Behavior

OC4B Pin

00

Normal port operation.

Disconnected

Cleared on Compare Match.

Set when TCNT4 = 0x000.

01

10

11

Connected

Disconnected

Cleared on Compare Match.

Set when TCNT4 = 0x000.

Set on Compare Match.

Cleared when TCNT4 = 0x000.

Table 15-13 shows the COM4B1:0 bit functionality when the PWM4B and WGM40 bits are set to

Phase and Frequency Correct PWM Mode.

Table 15-13. Compare Output Mode, Phase and Frequency Correct PWM Mode

COM4B1..0

OCW4B Behavior

OC4B Pin

00

Normal port operation.

Disconnected

Cleared on Compare Match when up-counting.

Set on Compare Match when down-counting.

01

10

11

Connected

Disconnected

Cleared on Compare Match when up-counting.

Set on Compare Match when down-counting.

Set on Compare Match when up-counting.

Cleared on Compare Match when down-counting.

• Bit 3 - FOC4A: Force Output Compare Match 4A

The FOC4A bit is only active when the PWM4A bit specify a non-PWM mode.

Writing a logical one to this bit forces a change in the Waveform Output (OCW4A) and the Out-

put Compare pin (OC4A) according to the values already set in COM4A1 and COM4A0. If

COM4A1 and COM4A0 written in the same cycle as FOC4A, the new settings will be used. The

Force Output Compare bit can be used to change the output pin value regardless of the timer

164

7766E–AVR–04/10

ATmega16U4/ATmega32U4

value. The automatic action programmed in COM4A1 and COM4A0 takes place as if a compare

match had occurred, but no interrupt is generated. The FOC4A bit is always read as zero.

• Bit 2 - FOC4B: Force Output Compare Match 4B

The FOC4B bit is only active when the PWM4B bit specify a non-PWM mode.

Writing a logical one to this bit forces a change in the Waveform Output (OCW4B) and the Out-

put Compare pin (OC4B) according to the values already set in COM4B1 and COM4B0. If

COM4B1 and COM4B0 written in the same cycle as FOC4B, the new settings will be used. The

Force Output Compare bit can be used to change the output pin value regardless of the timer

value. The automatic action programmed in COM4B1 and COM4B0 takes place as if a compare

match had occurred, but no interrupt is generated.

The FOC4B bit is always read as zero.

• Bit 1 - PWM4A: Pulse Width Modulator A Enable

When set (one) this bit enables PWM mode based on comparator OCR4A

• Bit 0 - PWM4B: Pulse Width Modulator B Enable

When set (one) this bit enables PWM mode based on comparator OCR4B.

15.12.2 TCCR4B – Timer/Counter4 Control Register B

Bit

7

6

5

DTPS41

R/W

0

4

DTPS40

R/W

0

3

CS43

R/W

0

2

CS42

R/W

0

1

CS41

R/W

0

CS40

R/W

0

PWM4X

PSR4

TCCR4B

Read/Write

Initial value

R/W

0

R/W

0

• Bit 7 - PWM4X: PWM Inversion Mode

When this bit is set (one), the PWM Inversion Mode is selected and the Dead Time Generator

outputs, OC4x and OC4x are inverted.

• Bit 6 - PSR4: Prescaler Reset Timer/Counter4

When this bit is set (one), the Timer/Counter4 prescaler (TCNT4 is unaffected) will be reset. The

bit will be cleared by hardware after the operation is performed. Writing a zero to this bit will have

no effect. This bit will always read as zero.

• Bits 5,4 - DTPS41, DTPS40: Dead Time Prescaler Bits

The Timer/Counter4 Control Register B is a 8-bit read/write register.

The dedicated Dead Time prescaler in front of the Dead Time Generator can divide the

Timer/Counter4 clock (PCK or CK) by 1, 2, 4 or 8 providing a large range of dead times that can

be generated. The Dead Time prescaler is controlled by two bits DTPS41 and DTPS40 from the

Dead Time Prescaler register. These bits define the division factor of the Dead Time prescaler.

The division factors are given in Table 15-14.

165

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 15-14. Division factors of the Dead Time prescaler

DTPS41

DTPS40

Prescaler divides the T/C4 clock by

0

1

0

1

0

1

1x (no division)

2x

4x

8x

• Bits 3 .. 0 - CS43, CS42, CS41, CS40: Clock Select Bits 3, 2, 1, and 0

The Clock Select bits 3, 2, 1, and 0 define the prescaling source of Timer/Counter4.

Table 15-15. Timer/Counter4 Prescaler Select

CS43 CS42 CS41 CS40 Asynchronous Clocking Mode Synchronous Clocking Mode

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

T/C4 stopped

PCK

T/C4 stopped

CK

PCK/2

CK/2

PCK/4

CK/4

PCK/8

CK/8

PCK/16

CK/16

PCK/32

CK/32

PCK/64

CK/64

PCK/128

PCK/256

PCK/512

PCK/1024

PCK/2048

PCK/4096

PCK/8192

PCK/16384

CK/128

CK/256

CK/512

CK/1024

CK/2048

CK/4096

CK/8192

CK/16384

The Stop condition provides a Timer Enable/Disable function.

15.12.3 TCCR4C – Timer/Counter4 Control Register C

Bit

7

6

5

4

3

2

1

0

PWM4D

R/W

0

COM4A1S COM4A0S COM4B1S COMAB0S COM4D1

COM4D0

FOC4D

R/W

0

TCCR4C

Read/Write

Initial value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

• Bits 7,6 - COM4A1S, COM4A0S: Comparator A Output Mode, Bits 1 and 0

These bits are the shadow bits of the COM4A1 and COM4A0 bits that are described in the sec-

tion “TCCR4A – Timer/Counter4 Control Register A” on page 162.

166

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bits 5,4 - COM4B1S, COM4B0S: Comparator B Output Mode, Bits 1 and 0

These bits are the shadow bits of the COM4A1 and COM4A0 bits that are described in the sec-

tion “TCCR4A – Timer/Counter4 Control Register A” on page 162.

• Bits 3,2 - COM4D1, COM4D0: Comparator D Output Mode, Bits 1 and 0

These bits control the behavior of the Waveform Output (OCW4D) and the connection of the

Output Compare pin (OC4D). If one or both of the COM4D1:0 bits are set, the OC4D output

overrides the normal port functionality of the I/O pin it is connected to. The complementary

OC4D output is connected only in PWM modes when the COM4D1:0 bits are set to “01”. Note

that the Data Direction Register (DDR) bit corresponding to the OC4D pin must be set in order to

enable the output driver.

The function of the COM4D1:0 bits depends on the PWM4D and WGM40 bit settings. Table 15-

16 shows the COM4D1:0 bit functionality when the PWM4D bit is set to a Normal Mode (non-

PWM).

Table 15-16. Compare Output Mode, Normal Mode (non-PWM)

COM4D1..0

OCW4D Behavior

OC4D Pin

Disconnected

Connected

OC4D Pin

00

01

10

11

Normal port operation.

Toggle on Compare Match.

Clear on Compare Match.

Set on Compare Match.

Disconnected

Table 15-17 shows the COM4D1:0 bit functionality when the PWM4D and WGM40 bits are set

to Fast PWM Mode.

Table 15-17. Compare Output Mode, Fast PWM Mode

COM4D1..0

OCW4D Behavior

OC4D Pin

00

Normal port operation.

Disconnected

Cleared on Compare Match.

Set when TCNT4 = 0x000.

01

10

11

Connected

Disconnected

Cleared on Compare Match.

Set when TCNT4 = 0x000.

Set on Compare Match.

Clear when TCNT4 = 0x000.

Table 15-18 on page 168 shows the COM4D1:0 bit functionality when the PWM4D and WGM40

bits are set to Phase and Frequency Correct PWM Mode.

167

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 15-18. Compare Output Mode, Phase and Frequency Correct PWM Mode

COM4D1..0 OCW4D Behavior

OC4D Pin

00

Normal port operation.

Disconnected

Cleared on Compare Match when up-counting.

Set on Compare Match when down-counting.

01

Connected

Disconnected

Cleared on Compare Match when up-counting.

Set on Compare Match when down-counting.

10

11

Set on Compare Match when up-counting.

Cleared on Compare Match when down-counting.

• Bit 1 - FOC4D: Force Output Compare Match 4D

The FOC4D bit is only active when the PWM4D bit specify a non-PWM mode.

Writing a logical one to this bit forces a change in the Waveform Output (OCW4D) and the Out-

put Compare pin (OC4D) according to the values already set in COM4D1 and COM4D0. If

COM4D1 and COM4D0 written in the same cycle as FOC4D, the new settings will be used. The

Force Output Compare bit can be used to change the output pin value regardless of the timer

value. The automatic action programmed in COM4D1 and COM4D0 takes place as if a compare

match had occurred, but no interrupt is generated. The FOC4D bit is always read as zero.

• Bit 0 - PWM4D: Pulse Width Modulator D Enable

When set (one) this bit enables PWM mode based on comparator OCR4D.

15.12.4 TCCR4D – Timer/Counter4 Control Register D

Bit

7

6

5

FPNC4

R/W

0

4

FPES4

R/W

0

3

FPAC4

R/W

0

2

FPF4

R/W

0

1

WGM41

R/W

0

WGM40

R/W

0

FPIE4

FPEN4

TCCR4D

Read/Write

Initial value

R/W

0

R/W

0

• Bit 7 - FPIE4: Fault Protection Interrupt Enable

Setting this bit (to one) enables the Fault Protection Interrupt.

• Bit 6– FPEN4: Fault Protection Mode Enable

Setting this bit (to one) activates the Fault Protection Mode.

• Bit 5 – FPNC4: Fault Protection Noise Canceler

Setting this bit activates the Fault Protection Noise Canceler. When the noise canceler is acti-

vated, the input from the Fault Protection Pin (INT0) is filtered. The filter function requires four

successive equal valued samples of the INT0 pin for changing its output. The Fault Protection is

therefore delayed by four Oscillator cycles when the noise canceler is enabled.

• Bit 4 – FPES4: Fault Protection Edge Select

This bit selects which edge on the Fault Protection pin (INT0) is used to trigger a fault event.

When the FPES4 bit is written to zero, a falling (negative) edge is used as trigger, and when the

FPES4 bit is written to one, a rising (positive) edge will trigger the fault.

168

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 3 - FPAC4: Fault Protection Analog Comparator Enable

When written logic one, this bit enables the Fault Protection function in Timer/Counter4 to be

triggered by the Analog Comparator. The comparator output is in this case directly connected to

the Fault Protection front-end logic, making the comparator utilize the noise canceler and edge

select features of the Timer/Counter4 Fault Protection interrupt. When written logic zero, no con-

nection between the Analog Comparator and the Fault Protection function exists. To make the

comparator trigger the Timer/Counter4 Fault Protection interrupt, the FPIE4 bit in the

Timer/Counter4 Control Register D (TCCR4D) must be set.

• Bit 2- FPF4: Fault Protection Interrupt Flag

When the FPIE4 bit is set (one), the Fault Protection Interrupt is enabled. Activity on the pin will

cause an interrupt request even, if the Fault Protection pin is configured as an output. The corre-

sponding interrupt of Fault Protection Interrupt Request is executed from the Fault Protection

Interrupt Vector. The bit FPF4 is cleared by hardware when executing the corresponding inter-

rupt handling vector. Alternatively, FPF4 is cleared after a synchronization clock cycle by writing

a logical one to the flag. When the SREG I-bit, FPIE4 and FPF4 are set, the Fault Interrupt is

executed.

• Bits 1:0 - WGM41, WGM40: Waveform Generation Mode Bits

This bit associated with the PWM4x bits control the counting sequence of the counter, the

source for type of waveform generation to be used, see Table 15-19. Modes of operation sup-

ported by the Timer/Counter4 are: Normal mode (counter), Fast PWM Mode, Phase and

Frequency Correct PWM and PWM6 Modes.

Table 15-19. Waveform Generation Mode Bit Description

Update of TOV4 Flag

PWM4x

WGM41..40 Timer/Counter Mode of Operation

TOP

OCR4x at

Set on

0

1

xx

00

01

10

11

Normal

OCR4C

Immediate TOP

Fast PWM

TOP

Phase and Frequency Correct PWM

PWM6 / Single-slope

PWM6 / Dual-slope

BOTTOM

TOP

BOTTOM

TOP

BOTTOM

15.12.5 TCCR4E – Timer/Counter4 Control Register E

Bit

7

6

5

OC4OE5

R/W

4

OC4OE4

R/W

3

OC4OE3

R/W

2

OC4OE2

R/W

1

0

TLOCK4

ENHC4

OC4OE1

OC4OE0

TCCR4E

Read/Write

Initial value

R

0

R

0

R/W

0

R/W

0

• Bit 7 - TLOCK4: Register Update Lock

This bit controls the Compare registers update. When this bit is set, writing to the Compare reg-

isters will not affect the output, however the values are stored and will be updated to the

Compare registers when the TLOCK4 bit will be cleared.

Refer to Section 15.7 ”Synchronous update” on page 149 for more details.

169

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 6- ENHC4: Enhanced Compare/PWM Mode

When this bit is set, the Waveform Generation Module works in enhanced mode: the compare

registers OCR4A/B/D can welcome one more accuracy bit, while the LSB determines on which

clock edge the Compare condition is signalled and the output pin level is updated.

• Bits 5:0 – OC4OE5:OC4OE0: Output Compare Override Enable Bits

These bits are the Output Compare Override Enable bits that are used to connect or disconnect

the Output Compare Pins in PWM6 Modes with an instant response on the corresponding Out-

put Compare Pins. The actual value from the port register will be visible on the port pin, when

the Output Compare Override Enable Bit is cleared. Table 15-20 shows the Output Compare

Override Enable Bits and their corresponding Output Compare pins.

Table 15-20. Output Compare Override Enable Bits vs. Output Compare Pins

OC4OE0

OC4OE1

OC4OE2

OC4OE3

OC4OE4

OC4OE5

OC4A (PC6)

OC4A (PC7)

OC4B (PB5)

OC4B (PB6)

OC4D (PD6)

OC4D (PD7)

15.12.6 TCNT4 – Timer/Counter4

Bit

7

MSB

R/W

0

6

5

4

3

2

1

0

4

LSB

R/W

0

TCNT4

Read/Write

Initial value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

This 8-bit register contains the value of Timer/Counter4.

The Timer/Counter4 is realized as a 10-bit up/down counter with read and write access. Due to

synchronization of the CPU, Timer/Counter4 data written into Timer/Counter4 is delayed by one

and half CPU clock cycles in synchronous mode and at most one CPU clock cycles for asyn-

chronous mode. When a 10-bit accuracy is preferred, special procedures must be followed for

accessing the 10-bit TCNT4 register via the 8-bit AVR data bus. These procedures are

described in section “Accessing 10-Bit Registers” on page 159. Alternatively the Timer/Counter4

can be used as an 8-bit Timer/Counter. Note that the Timer/Counter4 always starts counting up

after writing the TCNT4 register.

15.12.7 TC4H – Timer/Counter4 High Byte

Bit

7

6

-

5

-

4

-

3

-

2

TC410

R

1

TC49

R/W

0

TC48

R/W

0

-

TC4H

Read/Write

Initial value

R

0

R

0

R

0

R

0

R

0

The temporary Timer/Counter4 register is an 2-bit read/write register.

• Bits 7:3- Res: Reserved Bits

These bits are reserved bits in the ATmega16U4/ATmega32U4 and always reads as zero.

• Bits 2- TC410: Additional MSB bits for 11-bit accesses in Enhanced PWM mode

If 10-bit accuracy is used, the Timer/Counter4 High Byte Register (TC4H) is used for temporary

storing the MSB bits (TC49, TC48) of the 10-bit accesses. The same TC4H register is shared

between all 10-bit registers within the Timer/Counter4. Note that special procedures must be fol-

lowed when accessing the 10-bit TCNT4 register via the 8-bit AVR data bus. These procedures

are described in section “Accessing 10-Bit Registers” on page 159.

170

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bits 1:0 - TC49, TC48: Two MSB bits of the 10-bit accesses

If 10-bit accuracy is used, the Timer/Counter4 High Byte Register (TC4H) is used for temporary

storing the MSB bits (TC49, TC48) of the 10-bit accesses. The same TC4H register is shared

between all 10-bit registers within the Timer/Counter4. Note that special procedures must be fol-

lowed when accessing the 10-bit TCNT4 register via the 8-bit AVR data bus. These procedures

are described in section “Accessing 10-Bit Registers” on page 159.

15.12.8 OCR4A – Timer/Counter4 Output Compare Register A

Bit

7

6

5

4

3

2

1

0

MSB

R/W

0

LSB

R/W

0

OCR4A

Read/Write

Initial value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

The output compare register A is an 8-bit read/write register.

The Timer/Counter Output Compare Register A contains data to be continuously compared with

Timer/Counter4. Actions on compare matches are specified in TCCR4A. A compare match does

only occur if Timer/Counter4 counts to the OCR4A value. A software write that sets TCNT4 and

OCR4A to the same value does not generate a compare match.

A compare match will set the compare interrupt flag OCF4A after a synchronization delay follow-

ing the compare event.

Note that, if 10-bit accuracy is used special procedures must be followed when accessing the

internal 10-bit Output Compare Registers via the 8-bit AVR data bus. These procedures are

described in section “Accessing 10-Bit Registers” on page 159.

15.12.9 OCR4B – Timer/Counter4 Output Compare Register B

Bit

7

6

5

4

3

2

1

0

MSB

R/W

0

LSB

R/W

0

OCR4B

Read/Write

Initial value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

The output compare register B is an 8-bit read/write register.

The Timer/Counter Output Compare Register B contains data to be continuously compared with

Timer/Counter4. Actions on compare matches are specified in TCCR4. A compare match does

only occur if Timer/Counter4 counts to the OCR4B value. A software write that sets TCNT4 and

OCR4B to the same value does not generate a compare match.

A compare match will set the compare interrupt flag OCF4B after a synchronization delay follow-

ing the compare event.

Note that, if 10-bit accuracy is used special procedures must be followed when accessing the

internal 10-bit Output Compare Registers via the 8-bit AVR data bus. These procedures are

described in section “Accessing 10-Bit Registers” on page 159.

15.12.10 OCR4C – Timer/Counter4 Output Compare Register C

Bit

7

6

5

4

3

2

1

0

MSB

R/W

1

LSB

R/W

1

OCR44C

Read/Write

Initial value

R/W

1

R/W

1

R/W

1

R/W

1

R/W

1

R/W

1

171

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The output compare register C is an 8-bit read/write register.

The Timer/Counter Output Compare Register C contains data to be continuously compared with

Timer/Counter4, and a compare match will clear TCNT4. This register has the same function in

Normal mode and PWM modes.

Note that, if a smaller value than three is written to the Output Compare Register C, the value is

automatically replaced by three as it is a minimum value allowed to be written to this register.

Note that, if 10-bit accuracy is used special procedures must be followed when accessing the

internal 10-bit Output Compare Registers via the 8-bit AVR data bus. These procedures are

described in section “Accessing 10-Bit Registers” on page 159.

15.12.11 OCR4D – Timer/Counter4 Output Compare Register D

Bit

7

6

5

4

3

2

1

0

MSB

R/W

0

LSB

R/W

0

OCR4D

Read/Write

Initial value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

The output compare register D is an 8-bit read/write register.

The Timer/Counter Output Compare Register D contains data to be continuously compared with

Timer/Counter4. Actions on compare matches are specified in TCCR4A. A compare match does

only occur if Timer/Counter4 counts to the OCR4D value. A software write that sets TCNT4 and

OCR4D to the same value does not generate a compare match.

A compare match will set the compare interrupt flag OCF4D after a synchronization delay follow-

ing the compare event.

Note that, if 10-bit accuracy is used special procedures must be followed when accessing the

internal 10-bit Output Compare Registers via the 8-bit AVR data bus. These procedures are

described in section “Accessing 10-Bit Registers” on page 159.

15.12.12 TIMSK4 – Timer/Counter4 Interrupt Mask Register

Bit

7

6

5

4

3

2

TOIE4

R/W

0

1

0

OCIE4D

OCIE4A

OCIE4B

TIMSK4

Read/Write

Initial value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

• Bit 7- OCIE4D: Timer/Counter4 Output Compare Interrupt Enable

When the OCIE4D bit is set (one) and the I-bit in the Status Register is set (one), the

Timer/Counter4 Compare Match D interrupt is enabled. The corresponding interrupt at vector

$010 is executed if a compare match D occurs. The Compare Flag in Timer/Counter4 is set

(one) in the Timer/Counter Interrupt Flag Register.

• Bit 6 - OCIE4A: Timer/Counter4 Output Compare Interrupt Enable

When the OCIE4A bit is set (one) and the I-bit in the Status Register is set (one), the

Timer/Counter4 Compare Match A interrupt is enabled. The corresponding interrupt at vector

$003 is executed if a compare match A occurs. The Compare Flag in Timer/Counter4 is set

(one) in the Timer/Counter Interrupt Flag Register.

172

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 5 - OCIE4B: Timer/Counter4 Output Compare Interrupt Enable

When the OCIE4B bit is set (one) and the I-bit in the Status Register is set (one), the

Timer/Counter4 Compare Match B interrupt is enabled. The corresponding interrupt at vector

$009 is executed if a compare match B occurs. The Compare Flag in Timer/Counter4 is set

(one) in the Timer/Counter Interrupt Flag Register.

• Bit 2 - TOIE4: Timer/Counter4 Overflow Interrupt Enable

When the TOIE4 bit is set (one) and the I-bit in the Status Register is set (one), the

Timer/Counter4 Overflow interrupt is enabled. The corresponding interrupt (at vector $004) is

executed if an overflow in Timer/Counter4 occurs. The Overflow Flag (Timer4) is set (one) in the

Timer/Counter Interrupt Flag Register - TIFR4.

15.12.13 TIFR4 – Timer/Counter4 Interrupt Flag Register

Bit

7

6

5

4

3

2

TOV4

R/W

0

1

0

OCF4D

OCF4A

OCF4B

R/W

0

TIFR4

Read/Write

Initial value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

• Bit 7- OCF4D: Output Compare Flag 4D

The OCF4D bit is set (one) when compare match occurs between Timer/Counter4 and the data

value in OCR4D - Output Compare Register 4D. OCF4D is cleared by hardware when executing

the corresponding interrupt handling vector. Alternatively, OCF4D is cleared, after synchroniza-

tion clock cycle, by writing a logic one to the flag. When the I-bit in SREG, OCIE4D, and OCF4D

are set (one), the Timer/Counter4 D compare match interrupt is executed.

• Bit 6 - OCF4A: Output Compare Flag 4A

The OCF4A bit is set (one) when compare match occurs between Timer/Counter4 and the data

value in OCR4A - Output Compare Register 4A. OCF4A is cleared by hardware when executing

the corresponding interrupt handling vector. Alternatively, OCF4A is cleared, after synchroniza-

tion clock cycle, by writing a logic one to the flag. When the I-bit in SREG, OCIE4A, and OCF4A

are set (one), the Timer/Counter4 A compare match interrupt is executed.

• Bit 5 - OCF4B: Output Compare Flag 4B

The OCF4B bit is set (one) when compare match occurs between Timer/Counter4 and the data

value in OCR4B - Output Compare Register 4B. OCF4B is cleared by hardware when executing

the corresponding interrupt handling vector. Alternatively, OCF4B is cleared, after synchroniza-

tion clock cycle, by writing a logic one to the flag. When the I-bit in SREG, OCIE4B, and OCF4B

are set (one), the Timer/Counter4 B compare match interrupt is executed.

• Bit 2 - TOV4: Timer/Counter4 Overflow Flag

In Normal Mode and Fast PWM Mode the TOV4 bit is set (one) each time the counter reaches

TOP at the same clock cycle when the counter is reset to BOTTOM. In Phase and Frequency

Correct PWM Mode the TOV4 bit is set (one) each time the counter reaches BOTTOM at the

same clock cycle when zero is clocked to the counter.

The bit TOV4 is cleared by hardware when executing the corresponding interrupt handling vec-

tor. Alternatively, TOV4 is cleared, after synchronization clock cycle, by writing a logical one to

the flag. When the SREG I-bit, and TOIE4 (Timer/Counter4 Overflow Interrupt Enable), and

TOV4 are set (one), the Timer/Counter4 Overflow interrupt is executed.

173

7766E–AVR–04/10

ATmega16U4/ATmega32U4

15.12.14 DT4 – Timer/Counter4 Dead Time Value

Bit

7

DT4H3

R/W

0

6

DT4H2

R/W

0

5

DT4H1

R/W

0

4

DT4H0

R/W

0

3

DT4L3

R/W

0

2

DT4L2

R/W

0

1

DT4L1

R/W

0

DT4L0

R/W

0

DT4

Read/Write

Initial value

The dead time value register is an 8-bit read/write register.

The dead time delay of all Timer/Counter4 channels are adjusted by the dead time value regis-

ter, DT4. The register consists of two fields, DT4H3..0 and DT4L3..0, one for each

complementary output. Therefore a different dead time delay can be adjusted for the rising edge

of OC4x and the rising edge of OC4x.

• Bits 7:4- DT4H3:DT4H0: Dead Time Value for OC4x Output

The dead time value for the OC1x output. The dead time delay is set as a number of the pres-

caled timer/counter clocks. The minimum dead time is zero and the maximum dead time is the

prescaled time/counter clock period multiplied by 15.

• Bits 3:0- DT4L3:DT4L0: Dead Time Value for OC4x Output

The dead time value for the OC4x output. The dead time delay is set as a number of the pres-

caled timer/counter clocks. The minimum dead time is zero and the maximum dead time is the

prescaled time/counter clock period multiplied by 15.

174

7766E–AVR–04/10

ATmega16U4/ATmega32U4

16. Output Compare Modulator (OCM1C0A)

16.1 Overview

The Output Compare Modulator (OCM) allows generation of waveforms modulated with a carrier

frequency. The modulator uses the outputs from the Output Compare Unit C of the 16-bit

Timer/Counter1 and the Output Compare Unit of the 8-bit Timer/Counter0. For more details

about these Timer/Counters see “Timer/Counter0, Timer/Counter1, and Timer/Counter3 Pres-

calers” on page 89.

Figure 16-1. Output Compare Modulator, Block Diagram

OC1C

Timer/Counter 1

OC1C /

OC0A / PB7

OC0A

Timer/Counter 0

When the modulator is enabled, the two output compare channels are modulated together as

shown in the block diagram (Figure 16-1).

16.2 Description

The Output Compare unit 1C and Output Compare unit 2 shares the PB7 port pin for output. The

outputs of the Output Compare units (OC1C and OC0A) overrides the normal PORTB7 Register

when one of them is enabled (i.e., when COMnx1:0 is not equal to zero). When both OC1C and

OC0A are enabled at the same time, the modulator is automatically enabled.

The functional equivalent schematic of the modulator is shown on Figure 16-2. The schematic

includes part of the Timer/Counter units and the port B pin 7 output driver circuit.

Figure 16-2. Output Compare Modulator, Schematic

COMA01

COMA00

Vcc

COM1C1

COM1C0

Modulator

0

1

( From Waveform Generator )

D

Q

1

0

OC1C

OC1C /

OC0A/ PB7

( From Waveform Generator )

D

Q

OC0A

D

Q

D

Q

PORTB7

DDRB7

DATABUS

175

7766E–AVR–04/10

ATmega16U4/ATmega32U4

When the modulator is enabled the type of modulation (logical AND or OR) can be selected by

the PORTB7 Register. Note that the DDRB7 controls the direction of the port independent of the

COMnx1:0 bit setting.

16.2.1

Timing Example

Figure 16-3 illustrates the modulator in action. In this example the Timer/Counter1 is set to oper-

ate in fast PWM mode (non-inverted) and Timer/Counter0 uses CTC waveform mode with toggle

Compare Output mode (COMnx1:0 = 1).

Figure 16-3. Output Compare Modulator, Timing Diagram

clk_I/O

OC1C

(FPWM Mode)

OC0A

(CTC Mode)

PB7

(PORTB7 = 0)

PB7

(PORTB7 = 1)

1

2

3

(Period)

In this example, Timer/Counter0 provides the carrier, while the modulating signal is generated

by the Output Compare unit C of the Timer/Counter1.

The resolution of the PWM signal (OC1C) is reduced by the modulation. The reduction factor is

equal to the number of system clock cycles of one period of the carrier (OC0A). In this example

the resolution is reduced by a factor of two. The reason for the reduction is illustrated in Figure

16-3 at the second and third period of the PB7 output when PORTB7 equals zero. The period 2

high time is one cycle longer than the period 3 high time, but the result on the PB7 output is

equal in both periods.

176

7766E–AVR–04/10

ATmega16U4/ATmega32U4

17. Serial Peripheral Interface – SPI

The Serial Peripheral Interface (SPI) allows high-speed synchronous data transfer between the

ATmega16U4/ATmega32U4 and peripheral devices or between several AVR devices. The

ATmega16U4/ATmega32U4 SPI includes the following features:

• Full-duplex, Three-wire Synchronous Data Transfer

• Master or Slave Operation

• LSB First or MSB First Data Transfer

• Seven Programmable Bit Rates

• End of Transmission Interrupt Flag

• Write Collision Flag Protection

• Wake-up from Idle Mode

• Double Speed (CK/2) Master SPI Mode

USART can also be used in Master SPI mode, see “USART in SPI Mode” on page 214.

The Power Reduction SPI bit, PRSPI, in “Power Reduction Register 0 - PRR0” on page 46 on

page 50 must be written to zero to enable SPI module.

Figure 17-1. SPI Block Diagram⁽¹⁾

DIVIDER

/2/4/8/16/32/64/128

Note:

1. Refer to “Pinout ATmega16U4/ATmega32U4” on page 3, and Table 10-3 on page 72 for SPI

pin placement.

The interconnection between Master and Slave CPUs with SPI is shown in Figure 17-2. The sys-

tem consists of two shift Registers, and a Master clock generator. The SPI Master initiates the

177

7766E–AVR–04/10

ATmega16U4/ATmega32U4

communication cycle when pulling low the Slave Select SS pin of the desired Slave. Master and

Slave prepare the data to be sent in their respective shift Registers, and the Master generates

the required clock pulses on the SCK line to interchange data. Data is always shifted from Mas-

ter to Slave on the Master Out – Slave In, MOSI, line, and from Slave to Master on the Master In

– Slave Out, MISO, line. After each data packet, the Master will synchronize the Slave by pulling

high the Slave Select, SS, line.

When configured as a Master, the SPI interface has no automatic control of the SS line. This

must be handled by user software before communication can start. When this is done, writing a

byte to the SPI Data Register starts the SPI clock generator, and the hardware shifts the eight

bits into the Slave. After shifting one byte, the SPI clock generator stops, setting the end of

Transmission Flag (SPIF). If the SPI Interrupt Enable bit (SPIE) in the SPCR Register is set, an

interrupt is requested. The Master may continue to shift the next byte by writing it into SPDR, or

signal the end of packet by pulling high the Slave Select, SS line. The last incoming byte will be

kept in the Buffer Register for later use.

When configured as a Slave, the SPI interface will remain sleeping with MISO tri-stated as long

as the SS pin is driven high. In this state, software may update the contents of the SPI Data

Register, SPDR, but the data will not be shifted out by incoming clock pulses on the SCK pin

until the SS pin is driven low. As one byte has been completely shifted, the end of Transmission

Flag, SPIF is set. If the SPI Interrupt Enable bit, SPIE, in the SPCR Register is set, an interrupt

is requested. The Slave may continue to place new data to be sent into SPDR before reading

the incoming data. The last incoming byte will be kept in the Buffer Register for later use.

Figure 17-2. SPI Master-slave Interconnection

SHIFT

ENABLE

The system is single buffered in the transmit direction and double buffered in the receive direc-

tion. This means that bytes to be transmitted cannot be written to the SPI Data Register before

the entire shift cycle is completed. When receiving data, however, a received character must be

read from the SPI Data Register before the next character has been completely shifted in. Oth-

erwise, the first byte is lost.

In SPI Slave mode, the control logic will sample the incoming signal of the SCK pin. To ensure

correct sampling of the clock signal, the frequency of the SPI clock should never exceed f_osc/4.

178

7766E–AVR–04/10

ATmega16U4/ATmega32U4

When the SPI is enabled, the data direction of the MOSI, MISO, SCK, and SS pins is overridden

according to Table 17-1. For more details on automatic port overrides, refer to “Alternate Port

Functions” on page 70.

Table 17-1. SPI Pin Overrides⁽¹⁾

MOSI

MISO

SCK

SS

Direction, Master SPI

User Defined

Input

Direction, Slave SPI

Input

User Defined

Input

User Defined

Input

Note:

1. See “Alternate Functions of Port B” on page 72 for a detailed description of how to define the

direction of the user defined SPI pins.

The following code examples show how to initialize the SPI as a Master and how to perform a

simple transmission. DDR_SPI in the examples must be replaced by the actual Data Direction

Register controlling the SPI pins. DD_MOSI, DD_MISO and DD_SCK must be replaced by the

actual data direction bits for these pins. E.g. if MOSI is placed on pin PB5, replace DD_MOSI

with DDB5 and DDR_SPI with DDRB.

179

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Assembly Code Example⁽¹⁾

SPI_MasterInit:

; Set MOSI and SCK output, all others input

ldi r17,(1<<DD_MOSI)|(1<<DD_SCK)

out DDR_SPI,r17

; Enable SPI, Master, set clock rate fck/16

ldi r17,(1<<SPE)|(1<<MSTR)|(1<<SPR0)

out SPCR,r17

ret

SPI_MasterTransmit:

; Start transmission of data (r16)

out SPDR,r16

Wait_Transmit:

; Wait for transmission complete

sbis SPSR,SPIF

rjmp Wait_Transmit

ret

C Code Example⁽¹⁾

void SPI_MasterInit(void)

{

/* Set MOSI and SCK output, all others input */

DDR_SPI = (1<<DD_MOSI)|(1<<DD_SCK);

/* Enable SPI, Master, set clock rate fck/16 */

SPCR = (1<<SPE)|(1<<MSTR)|(1<<SPR0);

}

void SPI_MasterTransmit(char cData)

{

/* Start transmission */

SPDR = cData;

/* Wait for transmission complete */

while(!(SPSR & (1<<SPIF)))

;

}

Note:

1. See “Code Examples” on page 8.

180

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The following code examples show how to initialize the SPI as a Slave and how to perform a

simple reception.

Assembly Code Example⁽¹⁾

SPI_SlaveInit:

; Set MISO output, all others input

ldi r17,(1<<DD_MISO)

out DDR_SPI,r17

; Enable SPI

ldi r17,(1<<SPE)

out SPCR,r17

ret

SPI_SlaveReceive:

; Wait for reception complete

sbis SPSR,SPIF

rjmp SPI_SlaveReceive

; Read received data and return

in

r16,SPDR

ret

C Code Example⁽¹⁾

void SPI_SlaveInit(void)

{

/* Set MISO output, all others input */

DDR_SPI = (1<<DD_MISO);

/* Enable SPI */

SPCR = (1<<SPE);

}

char SPI_SlaveReceive(void)

{

/* Wait for reception complete */

while(!(SPSR & (1<<SPIF)))

;

/* Return Data Register */

return SPDR;

}

Note:

1. See “Code Examples” on page 8.

17.1 SS Pin Functionality

17.1.1

Slave Mode

When the SPI is configured as a Slave, the Slave Select (SS) pin is always input. When SS is

held low, the SPI is activated, and MISO becomes an output if configured so by the user. All

181

7766E–AVR–04/10

ATmega16U4/ATmega32U4

other pins are inputs. When SS is driven high, all pins are inputs, and the SPI is passive, which

means that it will not receive incoming data. Note that the SPI logic will be reset once the SS pin

is driven high.

The SS pin is useful for packet/byte synchronization to keep the slave bit counter synchronous

with the master clock generator. When the SS pin is driven high, the SPI slave will immediately

reset the send and receive logic, and drop any partially received data in the Shift Register.

17.1.2

Master Mode

When the SPI is configured as a Master (MSTR in SPCR is set), the user can determine the

direction of the SS pin.

If SS is configured as an output, the pin is a general output pin which does not affect the SPI

system. Typically, the pin will be driving the SS pin of the SPI Slave.

If SS is configured as an input, it must be held high to ensure Master SPI operation. If the

SS pin is driven low by peripheral circuitry when the SPI is configured as a Master with the SS

pin defined as an input, the SPI system interprets this as another master selecting the SPI as a

slave and starting to send data to it. To avoid bus contention, the SPI system takes the following

actions:

1. The MSTR bit in SPCR is cleared and the SPI system becomes a Slave. As a result of

the SPI becoming a Slave, the MOSI and SCK pins become inputs.

2. The SPIF Flag in SPSR is set, and if the SPI interrupt is enabled, and the I-bit in SREG

is set, the interrupt routine will be executed.

Thus, when interrupt-driven SPI transmission is used in Master mode, and there exists a possi-

bility that SS is driven low, the interrupt should always check that the MSTR bit is still set. If the

MSTR bit has been cleared by a slave select, it must be set by the user to re-enable SPI Master

mode.

17.1.3

SPI Control Register – SPCR

Bit

7

6

5

4

3

2

1

0

SPIE

R/W

0

SPE

R/W

0

DORD

R/W

0

MSTR

R/W

0

CPOL

R/W

0

CPHA

R/W

0

SPR1

R/W

0

SPR0

R/W

0

SPCR

Read/Write

Initial Value

• Bit 7 – SPIE: SPI Interrupt Enable

This bit causes the SPI interrupt to be executed if SPIF bit in the SPSR Register is set and the if

the Global Interrupt Enable bit in SREG is set.

• Bit 6 – SPE: SPI Enable

When the SPE bit is written to one, the SPI is enabled. This bit must be set to enable any SPI

operations.

• Bit 5 – DORD: Data Order

When the DORD bit is written to one, the LSB of the data word is transmitted first.

When the DORD bit is written to zero, the MSB of the data word is transmitted first.

• Bit 4 – MSTR: Master/Slave Select

This bit selects Master SPI mode when written to one, and Slave SPI mode when written logic

zero. If SS is configured as an input and is driven low while MSTR is set, MSTR will be cleared,

182

7766E–AVR–04/10

ATmega16U4/ATmega32U4

and SPIF in SPSR will become set. The user will then have to set MSTR to re-enable SPI Mas-

ter mode.

• Bit 3 – CPOL: Clock Polarity

When this bit is written to one, SCK is high when idle. When CPOL is written to zero, SCK is low

when idle. Refer to Figure 17-3 and Figure 17-4 for an example. The CPOL functionality is sum-

marized below:

Table 17-2. CPOL Functionality

CPOL

Leading Edge

Rising

Trailing Edge

Falling

0

1

Falling

Rising

• Bit 2 – CPHA: Clock Phase

The settings of the Clock Phase bit (CPHA) determine if data is sampled on the leading (first) or

trailing (last) edge of SCK. Refer to Figure 17-3 and Figure 17-4 for an example. The CPOL

functionality is summarized below:

Table 17-3. CPHA Functionality

CPHA

Leading Edge

Sample

Trailing Edge

Setup

0

1

Setup

Sample

• Bits 1, 0 – SPR1, SPR0: SPI Clock Rate Select 1 and 0

These two bits control the SCK rate of the device configured as a Master. SPR1 and SPR0 have

no effect on the Slave. The relationship between SCK and the Oscillator Clock frequency f_oscis

shown in the following table:

Table 17-4. Relationship Between SCK and the Oscillator Frequency

SPI2X

SPR1

SPR0

SCK Frequency

f_osc/4

0

1

0

1

0

1

0

1

0

1

0

1

0

1

f_osc/16

f_osc/64

f_osc/128

f_osc/2

f_osc/8

f_osc/32

f_osc/64

17.1.4

SPI Status Register – SPSR

Bit

7

6

5

–

R

0

4

–

R

0

3

–

R

0

2

–

R

0

1

–

R

0

SPIF

R

WCOL

SPI2X

R/W

0

SPSR

Read/Write

Initial Value

R

0

183

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 7 – SPIF: SPI Interrupt Flag

When a serial transfer is complete, the SPIF Flag is set. An interrupt is generated if SPIE in

SPCR is set and global interrupts are enabled. If SS is an input and is driven low when the SPI is

in Master mode, this will also set the SPIF Flag. SPIF is cleared by hardware when executing the

corresponding interrupt handling vector. Alternatively, the SPIF bit is cleared by first reading the

SPI Status Register with SPIF set, then accessing the SPI Data Register (SPDR).

• Bit 6 – WCOL: Write COLlision Flag

The WCOL bit is set if the SPI Data Register (SPDR) is written during a data transfer. The

WCOL bit (and the SPIF bit) are cleared by first reading the SPI Status Register with WCOL set,

and then accessing the SPI Data Register.

• Bit 5..1 – Res: Reserved Bits

These bits are reserved bits in the ATmega16U4/ATmega32U4 and will always read as zero.

• Bit 0 – SPI2X: Double SPI Speed Bit

When this bit is written logic one the SPI speed (SCK Frequency) will be doubled when the SPI

is in Master mode (see Table 17-4). This means that the minimum SCK period will be two CPU

clock periods. When the SPI is configured as Slave, the SPI is only guaranteed to work at f_osc/4

or lower.

The SPI interface on the ATmega16U4/ATmega32U4 is also used for program memory and

EEPROM downloading or uploading. See page 360 for serial programming and verification.

17.1.5

SPI Data Register – SPDR

Bit

7

6

5

4

3

2

1

0

MSB

R/W

X

LSB

R/W

X

SPDR

Read/Write

Initial Value

R/W

X

R/W

X

R/W

X

R/W

X

R/W

X

R/W

X

Undefined

The SPI Data Register is a read/write register used for data transfer between the Register File

and the SPI Shift Register. Writing to the register initiates data transmission. Reading the regis-

ter causes the Shift Register Receive buffer to be read.

17.2 Data Modes

There are four combinations of SCK phase and polarity with respect to serial data, which are

determined by control bits CPHA and CPOL. The SPI data transfer formats are shown in Figure

17-3 and Figure 17-4. Data bits are shifted out and latched in on opposite edges of the SCK sig-

nal, ensuring sufficient time for data signals to stabilize. This is clearly seen by summarizing

Table 17-2 and Table 17-3, as done below:

184

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 17-5. CPOL Functionality

Leading Edge

Trailing eDge

Setup (Falling)

Sample (Falling)

Setup (Rising)

Sample (Rising)

SPI Mode

CPOL=0, CPHA=0

CPOL=0, CPHA=1

CPOL=1, CPHA=0

CPOL=1, CPHA=1

Sample (Rising)

Setup (Rising)

Sample (Falling)

Setup (Falling)

0

1

2

3

Figure 17-3. SPI Transfer Format with CPHA = 0

SCK (CPOL = 0)

mode 0

SCK (CPOL = 1)

mode 2

SAMPLE I

MOSI/MISO

CHANGE 0

MOSI PIN

CHANGE 0

MISO PIN

SS

MSB first (DORD = 0) MSB

LSB first (DORD = 1) LSB

Bit 6

Bit 1

Bit 5

Bit 2

Bit 4

Bit 3

Bit 4

Bit 2

Bit 5

Bit 1

Bit 6

LSB

MSB

Figure 17-4. SPI Transfer Format with CPHA = 1

SCK (CPOL = 0)

mode 1

SCK (CPOL = 1)

mode 3

SAMPLE I

MOSI/MISO

CHANGE 0

MOSI PIN

CHANGE 0

MISO PIN

SS

MSB first (DORD = 0)

LSB first (DORD = 1)

MSB

LSB

Bit 6

Bit 1

Bit 5

Bit 2

Bit 4

Bit 3

Bit 4

Bit 2

Bit 5

Bit 1

Bit 6

LSB

MSB

185

7766E–AVR–04/10

ATmega16U4/ATmega32U4

18. USART

The Universal Synchronous and Asynchronous serial Receiver and Transmitter (USART) is a

highly flexible serial communication device. The main features are:

• Full Duplex Operation (Independent Serial Receive and Transmit Registers)

• Asynchronous or Synchronous Operation

• Flow control CTS/RTS signals hardware management

• Master or Slave Clocked Synchronous Operation

• High Resolution Baud Rate Generator

• Supports Serial Frames with 5, 6, 7, 8, or 9 Data Bits and 1 or 2 Stop Bits

• Odd or Even Parity Generation and Parity Check Supported by Hardware

• Data OverRun Detection

• Framing Error Detection

• Noise Filtering Includes False Start Bit Detection and Digital Low Pass Filter

• Three Separate Interrupts on TX Complete, TX Data Register Empty and RX Complete

• Multi-processor Communication Mode

• Double Speed Asynchronous Communication Mode

.

18.1 Overview

A simplified block diagram of the USART Transmitter is shown in Figure 18-1 on page 187. CPU

accessible I/O Registers and I/O pins are shown in bold.

186

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 18-1. USART Block Diagram⁽¹⁾

Clock Generator

UBRR[H:L]

OSC

BAUD RATE GENERATOR

SYNC LOGIC

CONTROL

XCK

TxD

RxD

Transmitter

TX

CONTROL

UDR (Transmit)

PARITY

GENERATOR

CONTROL

TRANSMIT SHIFT REGISTER

Receiver

CLOCK

RECOVERY

RX

CONTROL

DATA

RECOVERY

CONTROL

RECEIVE SHIFT REGISTER

PARITY

CHECKER

UDR (Receive)

UCSRA

UCSRB

UCSRC

Note:

1. See “Pinout ATmega16U4/ATmega32U4” on page 3, Table 10-8 on page 77 and for USART

pin placement.

The dashed boxes in the block diagram separate the three main parts of the USART (listed from

the top): Clock Generator, Transmitter and Receiver. Control Registers are shared by all units.

The Clock Generation logic consists of synchronization logic for external clock input used by

synchronous slave operation, and the baud rate generator. The XCKn (Transfer Clock) pin is

only used by synchronous transfer mode. The Transmitter consists of a single write buffer, a

serial Shift Register, Parity Generator and Control logic for handling different serial frame for-

mats. The write buffer allows a continuous transfer of data without any delay between frames.

The Receiver is the most complex part of the USART module due to its clock and data recovery

units. The recovery units are used for asynchronous data reception. In addition to the recovery

units, the Receiver includes a Parity Checker, Control logic, a Shift Register and a two level

receive buffer (UDRn). The Receiver supports the same frame formats as the Transmitter, and

can detect Frame Error, Data OverRun and Parity Errors.

18.2 Clock Generation

The Clock Generation logic generates the base clock for the Transmitter and Receiver. The

USARTn supports four modes of clock operation: Normal asynchronous, Double Speed asyn-

chronous, Master synchronous and Slave synchronous mode. The UMSELn bit in USART

Control and Status Register C (UCSRnC) selects between asynchronous and synchronous

operation. Double Speed (asynchronous mode only) is controlled by the U2Xn found in the

187

7766E–AVR–04/10

ATmega16U4/ATmega32U4

UCSRnA Register. When using synchronous mode (UMSELn = 1), the Data Direction Register

for the XCKn pin (DDR_XCKn) controls whether the clock source is internal (Master mode) or

external (Slave mode). The XCKn pin is only active when using synchronous mode.

Figure 18-2 shows a block diagram of the clock generation logic.

Figure 18-2. Clock Generation Logic, Block Diagram

UBRR

U2X

fosc

UBRR+1

Prescaling

Down-Counter

/2

/4

/2

0

1

0

1

OSC

txclk

UMSEL

rxclk

DDR_XCK

Sync

Register

Edge

Detector

xcki

0

1

XCK

xcko

DDR_XCK

UCPOL

1

0

Signal description:

txclk

Transmitter clock (Internal Signal).

Receiver base clock (Internal Signal).

rxclk

xcki

Input from XCK pin (internal Signal). Used for synchronous slave

operation.

xcko

f_OSC

Clock output to XCK pin (Internal Signal). Used for synchronous master

operation.

XTAL pin frequency (System Clock).

18.2.1

Internal Clock Generation – The Baud Rate Generator

Internal clock generation is used for the asynchronous and the synchronous master modes of

operation. The description in this section refers to Figure 18-2.

The USART Baud Rate Register (UBRRn) and the down-counter connected to it function as a

programmable prescaler or baud rate generator. The down-counter, running at system clock

(f_osc), is loaded with the UBRRn value each time the counter has counted down to zero or when

the UBRRLn Register is written. A clock is generated each time the counter reaches zero. This

clock is the baud rate generator clock output (= f_osc/(UBRRn+1)). The Transmitter divides the

baud rate generator clock output by 2, 8 or 16 depending on mode. The baud rate generator out-

put is used directly by the Receiver’s clock and data recovery units. However, the recovery units

use a state machine that uses 2, 8 or 16 states depending on mode set by the state of the

UMSELn, U2Xn and DDR_XCKn bits.

188

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 18-1 contains equations for calculating the baud rate (in bits per second) and for calculat-

ing the UBRRn value for each mode of operation using an internally generated clock source.

Table 18-1. Equations for Calculating Baud Rate Register Setting

Equation for Calculating

Baud Rate⁽¹⁾

Equation for Calculating

UBRR Value

Operating Mode

f

OSC

UBRRn = ----------------------- – 1

16BAUD

f

OSC

Asynchronous Normal

mode (U2Xn = 0)

BAUD = -----------------------------------------

16(UBRRn + 1)

f

OSC

UBRRn = -------------------- – 1

8BAUD

Asynchronous Double

Speed mode (U2Xn =

1)

f

OSC

BAUD = --------------------------------------

8(UBRRn + 1)

f

OSC

UBRRn = -------------------- – 1

2BAUD

f

OSC

Synchronous Master

mode

BAUD = --------------------------------------

2(UBRRn + 1)

Note:

1. The baud rate is defined to be the transfer rate in bit per second (bps)

BAUD

Baud rate (in bits per second, bps)

f_OSC

System Oscillator clock frequency

UBRRn

Contents of the UBRRHn and UBRRLn Registers, (0-4095)

Some examples of UBRRn values for some system clock frequencies are found in Table 18-9 on

page 210.

18.2.2

Double Speed Operation (U2Xn)

The transfer rate can be doubled by setting the U2Xn bit in UCSRnA. Setting this bit only has

effect for the asynchronous operation. Set this bit to zero when using synchronous operation.

Setting this bit will reduce the divisor of the baud rate divider from 16 to 8, effectively doubling

the transfer rate for asynchronous communication. Note however that the Receiver will in this

case only use half the number of samples (reduced from 16 to 8) for data sampling and clock

recovery, and therefore a more accurate baud rate setting and system clock are required when

this mode is used. For the Transmitter, there are no downsides.

189

7766E–AVR–04/10

ATmega16U4/ATmega32U4

18.2.3

External Clock

External clocking is used by the synchronous slave modes of operation. The description in this

section refers to Figure 18-2 for details.

External clock input from the XCKn pin is sampled by a synchronization register to minimize the

chance of meta-stability. The output from the synchronization register must then pass through

an edge detector before it can be used by the Transmitter and Receiver. This process intro-

duces a two CPU clock period delay and therefore the maximum external XCKn clock frequency

is limited by the following equation:

f

OSC

-----------

f

<

XCK

4

Note that f_oscdepends on the stability of the system clock source. It is therefore recommended to

add some margin to avoid possible loss of data due to frequency variations.

18.2.4

Synchronous Clock Operation

When synchronous mode is used (UMSELn = 1), the XCKn pin will be used as either clock input

(Slave) or clock output (Master). The dependency between the clock edges and data sampling

or data change is the same. The basic principle is that data input (on RxDn) is sampled at the

opposite XCKn clock edge of the edge the data output (TxDn) is changed.

Figure 18-3. Synchronous Mode XCKn Timing.

UCPOL = 1

XCK

RxD / TxD

Sample

UCPOL = 0

XCK

RxD / TxD

The UCPOLn bit UCRSC selects which XCKn clock edge is used for data sampling and which is

used for data change. As Figure 18-3 shows, when UCPOLn is zero the data will be changed at

rising XCKn edge and sampled at falling XCKn edge. If UCPOLn is set, the data will be changed

at falling XCKn edge and sampled at rising XCKn edge.

18.3 Frame Formats

A serial frame is defined to be one character of data bits with synchronization bits (start and stop

bits), and optionally a parity bit for error checking. The USART accepts all 30 combinations of

the following as valid frame formats:

• 1 start bit

• 5, 6, 7, 8, or 9 data bits

• no, even or odd parity bit

• 1 or 2 stop bits

190

7766E–AVR–04/10

ATmega16U4/ATmega32U4

A frame starts with the start bit followed by the least significant data bit. Then the next data bits,

up to a total of nine, are succeeding, ending with the most significant bit. If enabled, the parity bit

is inserted after the data bits, before the stop bits. When a complete frame is transmitted, it can

be directly followed by a new frame, or the communication line can be set to an idle (high) state.

Figure 18-4 illustrates the possible combinations of the frame formats. Bits inside brackets are

optional.

Figure 18-4. Frame Formats

FRAME

(IDLE)

St

0

1

2

3

4

[5]

[6]

[7]

[8]

[P] Sp1 [Sp2] (St / IDLE)

St

Start bit, always low.

Data bits (0 to 8).

(n)

P

Parity bit. Can be odd or even.

Stop bit, always high.

Sp

IDLE

must be

No transfers on the communication line (RxDn or TxDn). An IDLE line

high.

The frame format used by the USART is set by the UCSZn2:0, UPMn1:0 and USBSn bits in

UCSRnB and UCSRnC. The Receiver and Transmitter use the same setting. Note that changing

the setting of any of these bits will corrupt all ongoing communication for both the Receiver and

Transmitter.

The USART Character SiZe (UCSZn2:0) bits select the number of data bits in the frame. The

USART Parity mode (UPMn1:0) bits enable and set the type of parity bit. The selection between

one or two stop bits is done by the USART Stop Bit Select (USBSn) bit. The Receiver ignores

the second stop bit. An FE (Frame Error) will therefore only be detected in the cases where the

first stop bit is zero.

18.3.1

Parity Bit Calculation

The parity bit is calculated by doing an exclusive-or of all the data bits. If odd parity is used, the

result of the exclusive or is inverted. The relation between the parity bit and data bits is as

follows::

P

= d

⊕ … ⊕ d ⊕ d ⊕ d ⊕ d ⊕ 0

3 2 1 0

even

n – 1

⊕ … ⊕ d ⊕ d ⊕ d ⊕ d ⊕ 1

odd

3 2 1 0

P_even

P^odd

d_n

Parity bit using even parity

Parity bit using odd parity

Data bit n of the character

If used, the parity bit is located between the last data bit and first stop bit of a serial frame.

191

7766E–AVR–04/10

ATmega16U4/ATmega32U4

18.4 USART Initialization

The USART has to be initialized before any communication can take place. The initialization pro-

cess normally consists of setting the baud rate, setting frame format and enabling the

Transmitter or the Receiver depending on the usage. For interrupt driven USART operation, the

Global Interrupt Flag should be cleared (and interrupts globally disabled) when doing the

initialization.

Before doing a re-initialization with changed baud rate or frame format, be sure that there are no

ongoing transmissions during the period the registers are changed. The TXCn Flag can be used

to check that the Transmitter has completed all transfers, and the RXC Flag can be used to

check that there are no unread data in the receive buffer. Note that the TXCn Flag must be

cleared before each transmission (before UDRn is written) if it is used for this purpose.

The following simple USART initialization code examples show one assembly and one C func-

tion that are equal in functionality. The examples assume asynchronous operation using polling

(no interrupts enabled) and a fixed frame format. The baud rate is given as a function parameter.

For the assembly code, the baud rate parameter is assumed to be stored in the r17:r16

Registers.

Assembly Code Example⁽¹⁾

USART_Init:

; Set baud rate

out UBRRHn, r17

out UBRRLn, r16

; Enable receiver and transmitter

ldi r16, (1<<RXENn)|(1<<TXENn)

out UCSRnB,r16

; Set frame format: 8data, 2stop bit

ldi r16, (1<<USBSn)|(3<<UCSZn0)

out UCSRnC,r16

ret

C Code Example⁽¹⁾

void USART_Init( unsigned int baud )

{

/* Set baud rate */

UBRRHn = (unsigned char)(baud>>8);

UBRRLn = (unsigned char)baud;

/* Enable receiver and transmitter */

UCSRnB = (1<<RXENn)|(1<<TXENn);

/* Set frame format: 8data, 2stop bit */

UCSRnC = (1<<USBSn)|(3<<UCSZn0);

}

Note:

1. See “Code Examples” on page 8.

More advanced initialization routines can be made that include frame format as parameters, dis-

able interrupts and so on. However, many applications use a fixed setting of the baud and

control registers, and for these types of applications the initialization code can be placed directly

in the main routine, or be combined with initialization code for other I/O modules.

192

7766E–AVR–04/10

ATmega16U4/ATmega32U4

18.5 Data Transmission – The USART Transmitter

The USART Transmitter is enabled by setting the Transmit Enable (TXEN) bit in the UCSRnB

Register. When the Transmitter is enabled, the normal port operation of the TxDn pin is overrid-

den by the USART and given the function as the Transmitter’s serial output. The baud rate,

mode of operation and frame format must be set up once before doing any transmissions. If syn-

chronous operation is used, the clock on the XCKn pin will be overridden and used as

transmission clock.

18.5.1

Sending Frames with 5 to 8 Data Bit

A data transmission is initiated by loading the transmit buffer with the data to be transmitted. The

CPU can load the transmit buffer by writing to the UDRn I/O location. The buffered data in the

transmit buffer will be moved to the Shift Register when the Shift Register is ready to send a new

frame. The Shift Register is loaded with new data if it is in idle state (no ongoing transmission) or

immediately after the last stop bit of the previous frame is transmitted. When the Shift Register is

loaded with new data, it will transfer one complete frame at the rate given by the Baud Register,

U2Xn bit or by XCKn depending on mode of operation.

The following code examples show a simple USART transmit function based on polling of the

Data Register Empty (UDREn) Flag. When using frames with less than eight bits, the most sig-

nificant bits written to the UDRn are ignored. The USART has to be initialized before the function

can be used. For the assembly code, the data to be sent is assumed to be stored in Register

R16.

Assembly Code Example⁽¹⁾

USART_Transmit:

; Wait for empty transmit buffer

sbis UCSRnA,UDREn

rjmp USART_Transmit

; Put data (r16) into buffer, sends the data

out UDRn,r16

ret

C Code Example⁽¹⁾

void USART_Transmit( unsigned char data )

{

/* Wait for empty transmit buffer */

while ( !( UCSRnA & (1<<UDREn)) )

;

/* Put data into buffer, sends the data */

UDRn = data;

}

Note:

1. See “Code Examples” on page 8.

The function simply waits for the transmit buffer to be empty by checking the UDREn Flag,

before loading it with new data to be transmitted. If the Data Register Empty interrupt is utilized,

the interrupt routine writes the data into the buffer.

193

7766E–AVR–04/10

ATmega16U4/ATmega32U4

18.5.2

Sending Frames with 9 Data Bit

If 9-bit characters are used (UCSZn = 7), the ninth bit must be written to the TXB8 bit in

UCSRnB before the low byte of the character is written to UDRn. The following code examples

show a transmit function that handles 9-bit characters. For the assembly code, the data to be

sent is assumed to be stored in registers R17:R16.

Assembly Code Example⁽¹⁾⁽²⁾

USART_Transmit:

; Wait for empty transmit buffer

sbis UCSRnA,UDREn

rjmp USART_Transmit

; Copy 9th bit from r17 to TXB8

cbi UCSRnB,TXB8

sbrc r17,0

sbi UCSRnB,TXB8

; Put LSB data (r16) into buffer, sends the data

out UDRn,r16

ret

C Code Example⁽¹⁾⁽²⁾

void USART_Transmit( unsigned int data )

{

/* Wait for empty transmit buffer */

while ( !( UCSRnA & (1<<UDREn))) )

;

/* Copy 9th bit to TXB8 */

UCSRnB &= ~(1<<TXB8);

if ( data & 0x0100 )

UCSRnB |= (1<<TXB8);

/* Put data into buffer, sends the data */

UDRn = data;

}

Notes: 1. These transmit functions are written to be general functions. They can be optimized if the con-

tents of the UCSRnB is static. For example, only the TXB8 bit of the UCSRnB Register is used

after initialization.

2. See “Code Examples” on page 8.

The ninth bit can be used for indicating an address frame when using multi processor communi-

cation mode or for other protocol handling as for example synchronization.

18.5.3

Transmitter Flags and Interrupts

The USART Transmitter has two flags that indicate its state: USART Data Register Empty

(UDREn) and Transmit Complete (TXCn). Both flags can be used for generating interrupts.

The Data Register Empty (UDREn) Flag indicates whether the transmit buffer is ready to receive

new data. This bit is set when the transmit buffer is empty, and cleared when the transmit buffer

contains data to be transmitted that has not yet been moved into the Shift Register. For compat-

ibility with future devices, always write this bit to zero when writing the UCSRnA Register.

194

7766E–AVR–04/10

ATmega16U4/ATmega32U4

When the Data Register Empty Interrupt Enable (UDRIEn) bit in UCSRnB is written to one, the

USART Data Register Empty Interrupt will be executed as long as UDREn is set (provided that

global interrupts are enabled). UDREn is cleared by writing UDRn. When interrupt-driven data

transmission is used, the Data Register Empty interrupt routine must either write new data to

UDRn in order to clear UDREn or disable the Data Register Empty interrupt, otherwise a new

interrupt will occur once the interrupt routine terminates.

The Transmit Complete (TXCn) Flag bit is set one when the entire frame in the Transmit Shift

Register has been shifted out and there are no new data currently present in the transmit buffer.

The TXCn Flag bit is automatically cleared when a transmit complete interrupt is executed, or it

can be cleared by writing a one to its bit location. The TXCn Flag is useful in half-duplex commu-

nication interfaces (like the RS-485 standard), where a transmitting application must enter

receive mode and free the communication bus immediately after completing the transmission.

When the Transmit Compete Interrupt Enable (TXCIEn) bit in UCSRnB is set, the USART

Transmit Complete Interrupt will be executed when the TXCn Flag becomes set (provided that

global interrupts are enabled). When the transmit complete interrupt is used, the interrupt han-

dling routine does not have to clear the TXCn Flag, this is done automatically when the interrupt

is executed.

18.5.4

18.5.5

Parity Generator

The Parity Generator calculates the parity bit for the serial frame data. When parity bit is enabled

(UPMn1 = 1), the transmitter control logic inserts the parity bit between the last data bit and the

first stop bit of the frame that is sent.

Disabling the Transmitter

The disabling of the Transmitter (setting the TXEN to zero) will not become effective until ongo-

ing and pending transmissions are completed, i.e., when the Transmit Shift Register and

Transmit Buffer Register do not contain data to be transmitted. When disabled, the Transmitter

will no longer override the TxDn pin.

18.6 Data Reception – The USART Receiver

The USART Receiver is enabled by writing the Receive Enable (RXENn) bit in the

UCSRnB Register to one. When the Receiver is enabled, the normal pin operation of the RxDn

pin is overridden by the USART and given the function as the Receiver’s serial input. The baud

rate, mode of operation and frame format must be set up once before any serial reception can

be done. If synchronous operation is used, the clock on the XCKn pin will be used as transfer

clock.

18.6.1

Receiving Frames with 5 to 8 Data Bits

The Receiver starts data reception when it detects a valid start bit. Each bit that follows the start

bit will be sampled at the baud rate or XCKn clock, and shifted into the Receive Shift Register

until the first stop bit of a frame is received. A second stop bit will be ignored by the Receiver.

When the first stop bit is received, i.e., a complete serial frame is present in the Receive Shift

Register, the contents of the Shift Register will be moved into the receive buffer. The receive

buffer can then be read by reading the UDRn I/O location.

The following code example shows a simple USART receive function based on polling of the

Receive Complete (RXCn) Flag. When using frames with less than eight bits the most significant

195

7766E–AVR–04/10

ATmega16U4/ATmega32U4

bits of the data read from the UDRn will be masked to zero. The USART has to be initialized

before the function can be used.

Assembly Code Example⁽¹⁾

USART_Receive:

; Wait for data to be received

sbis UCSRnA, RXCn

rjmp USART_Receive

; Get and return received data from buffer

in

r16, UDRn

ret

C Code Example⁽¹⁾

unsigned char USART_Receive( void )

{

/* Wait for data to be received */

while ( !(UCSRnA & (1<<RXCn)) )

;

/* Get and return received data from buffer */

return UDRn;

}

Note:

1. See “Code Examples” on page 8.

The function simply waits for data to be present in the receive buffer by checking the RXCn Flag,

before reading the buffer and returning the value.

18.6.2

Receiving Frames with 9 Data Bits

If 9-bit characters are used (UCSZn=7) the ninth bit must be read from the RXB8n bit in

UCSRnB before reading the low bits from the UDRn. This rule applies to the FEn, DORn and

UPEn Status Flags as well. Read status from UCSRnA, then data from UDRn. Reading the

UDRn I/O location will change the state of the receive buffer FIFO and consequently the TXB8n,

FEn, DORn and UPEn bits, which all are stored in the FIFO, will change.

The following code example shows a simple USART receive function that handles both nine bit

characters and the status bits.

196

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Assembly Code Example⁽¹⁾

USART_Receive:

; Wait for data to be received

sbis UCSRnA, RXCn

rjmp USART_Receive

; Get status and 9th bit, then data from buffer

in

r18, UCSRnA

r17, UCSRnB

r16, UDRn

; If error, return -1

andi r18,(1<<FEn)|(1<<DORn)|(1<<UPEn)

breq USART_ReceiveNoError

ldi r17, HIGH(-1)

ldi r16, LOW(-1)

USART_ReceiveNoError:

; Filter the 9th bit, then return

lsr r17

andi r17, 0x01

ret

C Code Example⁽¹⁾

unsigned int USART_Receive( void )

{

unsigned char status, resh, resl;

/* Wait for data to be received */

while ( !(UCSRnA & (1<<RXCn)) )

;

/* Get status and 9th bit, then data */

/* from buffer */

status = UCSRnA;

resh = UCSRnB;

resl = UDRn;

/* If error, return -1 */

if ( status & (1<<FEn)|(1<<DORn)|(1<<UPEn) )

return -1;

/* Filter the 9th bit, then return */

resh = (resh >> 1) & 0x01;

return ((resh << 8) | resl);

}

Note:

1. See “Code Examples” on page 8.

The receive function example reads all the I/O Registers into the Register File before any com-

putation is done. This gives an optimal receive buffer utilization since the buffer location read will

be free to accept new data as early as possible.

197

7766E–AVR–04/10

ATmega16U4/ATmega32U4

18.6.3

Receive Compete Flag and Interrupt

The USART Receiver has one flag that indicates the Receiver state.

The Receive Complete (RXCn) Flag indicates if there are unread data present in the receive buf-

fer. This flag is one when unread data exist in the receive buffer, and zero when the receive

buffer is empty (i.e., does not contain any unread data). If the Receiver is disabled (RXENn = 0),

the receive buffer will be flushed and consequently the RXCn bit will become zero.

When the Receive Complete Interrupt Enable (RXCIEn) in UCSRnB is set, the USART Receive

Complete interrupt will be executed as long as the RXCn Flag is set (provided that global inter-

rupts are enabled). When interrupt-driven data reception is used, the receive complete routine

must read the received data from UDRn in order to clear the RXCn Flag, otherwise a new inter-

rupt will occur once the interrupt routine terminates.

18.6.4

Receiver Error Flags

The USART Receiver has three Error Flags: Frame Error (FEn), Data OverRun (DORn) and

Parity Error (UPEn). All can be accessed by reading UCSRnA. Common for the Error Flags is

that they are located in the receive buffer together with the frame for which they indicate the

error status. Due to the buffering of the Error Flags, the UCSRnA must be read before the

receive buffer (UDRn), since reading the UDRn I/O location changes the buffer read location.

Another equality for the Error Flags is that they can not be altered by software doing a write to

the flag location. However, all flags must be set to zero when the UCSRnA is written for upward

compatibility of future USART implementations. None of the Error Flags can generate interrupts.

The Frame Error (FEn) Flag indicates the state of the first stop bit of the next readable frame

stored in the receive buffer. The FEn Flag is zero when the stop bit was correctly read (as one),

and the FEn Flag will be one when the stop bit was incorrect (zero). This flag can be used for

detecting out-of-sync conditions, detecting break conditions and protocol handling. The FEn

Flag is not affected by the setting of the USBSn bit in UCSRnC since the Receiver ignores all,

except for the first, stop bits. For compatibility with future devices, always set this bit to zero

when writing to UCSRnA.

The Data OverRun (DORn) Flag indicates data loss due to a receiver buffer full condition. A

Data OverRun occurs when the receive buffer is full (two characters), it is a new character wait-

ing in the Receive Shift Register, and a new start bit is detected. If the DORn Flag is set there

was one or more serial frame lost between the frame last read from UDRn, and the next frame

read from UDRn. For compatibility with future devices, always write this bit to zero when writing

to UCSRnA. The DORn Flag is cleared when the frame received was successfully moved from

the Shift Register to the receive buffer.

The Parity Error (UPEn) Flag indicates that the next frame in the receive buffer had a Parity

Error when received. If Parity Check is not enabled the UPEn bit will always be read zero. For

compatibility with future devices, always set this bit to zero when writing to UCSRnA. For more

details see “Parity Bit Calculation” on page 191 and “Parity Checker” on page 198.

18.6.5

Parity Checker

The Parity Checker is active when the high USART Parity mode (UPMn1) bit is set. Type of Par-

ity Check to be performed (odd or even) is selected by the UPMn0 bit. When enabled, the Parity

Checker calculates the parity of the data bits in incoming frames and compares the result with

the parity bit from the serial frame. The result of the check is stored in the receive buffer together

with the received data and stop bits. The Parity Error (UPEn) Flag can then be read by software

to check if the frame had a Parity Error.

198

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The UPEn bit is set if the next character that can be read from the receive buffer had a Parity

Error when received and the Parity Checking was enabled at that point (UPMn1 = 1). This bit is

valid until the receive buffer (UDRn) is read.

18.6.6

18.6.7

Disabling the Receiver

In contrast to the Transmitter, disabling of the Receiver will be immediate. Data from ongoing

receptions will therefore be lost. When disabled (i.e., the RXENn is set to zero) the Receiver will

no longer override the normal function of the RxDn port pin. The Receiver buffer FIFO will be

flushed when the Receiver is disabled. Remaining data in the buffer will be lost

Flushing the Receive Buffer

The receiver buffer FIFO will be flushed when the Receiver is disabled, i.e., the buffer will be

emptied of its contents. Unread data will be lost. If the buffer has to be flushed during normal

operation, due to for instance an error condition, read the UDRn I/O location until the RXCn Flag

is cleared. The following code example shows how to flush the receive buffer.

Assembly Code Example⁽¹⁾

USART_Flush:

sbis UCSRnA, RXCn

ret

in

rjmp USART_Flush

C Code Example⁽¹⁾

r16, UDRn

void USART_Flush( void )

{

unsigned char dummy;

while ( UCSRnA & (1<<RXCn) ) dummy = UDRn;

}

Note:

1. See “Code Examples” on page 8.

18.7 Asynchronous Data Reception

The USART includes a clock recovery and a data recovery unit for handling asynchronous data

reception. The clock recovery logic is used for synchronizing the internally generated baud rate

clock to the incoming asynchronous serial frames at the RxDn pin. The data recovery logic sam-

ples and low pass filters each incoming bit, thereby improving the noise immunity of the

Receiver. The asynchronous reception operational range depends on the accuracy of the inter-

nal baud rate clock, the rate of the incoming frames, and the frame size in number of bits.

18.7.1

Asynchronous Clock Recovery

The clock recovery logic synchronizes internal clock to the incoming serial frames. Figure 18-5

illustrates the sampling process of the start bit of an incoming frame. The sample rate is 16 times

the baud rate for Normal mode, and eight times the baud rate for Double Speed mode. The hor-

izontal arrows illustrate the synchronization variation due to the sampling process. Note the

larger time variation when using the Double Speed mode (U2Xn = 1) of operation. Samples

denoted zero are samples done when the RxDn line is idle (i.e., no communication activity).

199

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 18-5. Start Bit Sampling

RxD

IDLE

START

BIT 0

Sample

(U2X = 0)

0

1

2

3

2

4

5

3

6

7

4

8

9

5

10

11

6

12

13

7

14

15

8

16

1

2

3

Sample

(U2X = 1)

0

2

When the clock recovery logic detects a high (idle) to low (start) transition on the RxDn line, the

start bit detection sequence is initiated. Let sample 1 denote the first zero-sample as shown in

the figure. The clock recovery logic then uses samples 8, 9, and 10 for Normal mode, and sam-

ples 4, 5, and 6 for Double Speed mode (indicated with sample numbers inside boxes on the

figure), to decide if a valid start bit is received. If two or more of these three samples have logical

high levels (the majority wins), the start bit is rejected as a noise spike and the Receiver starts

looking for the next high to low-transition. If however, a valid start bit is detected, the clock recov-

ery logic is synchronized and the data recovery can begin. The synchronization process is

repeated for each start bit.

18.7.2

Asynchronous Data Recovery

When the receiver clock is synchronized to the start bit, the data recovery can begin. The data

recovery unit uses a state machine that has 16 states for each bit in Normal mode and eight

states for each bit in Double Speed mode. Figure 18-6 shows the sampling of the data bits and

the parity bit. Each of the samples is given a number that is equal to the state of the recovery

unit.

Figure 18-6. Sampling of Data and Parity Bit

RxD

BIT n

Sample

(U2X = 0)

1

2

3

2

4

5

3

6

7

4

8

9

5

10

11

6

12

13

7

14

15

8

16

1

Sample

(U2X = 1)

The decision of the logic level of the received bit is taken by doing a majority voting of the logic

value to the three samples in the center of the received bit. The center samples are emphasized

on the figure by having the sample number inside boxes. The majority voting process is done as

follows: If two or all three samples have high levels, the received bit is registered to be a logic 1.

If two or all three samples have low levels, the received bit is registered to be a logic 0. This

majority voting process acts as a low pass filter for the incoming signal on the RxDn pin. The

recovery process is then repeated until a complete frame is received. Including the first stop bit.

Note that the Receiver only uses the first stop bit of a frame.

Figure 18-7 shows the sampling of the stop bit and the earliest possible beginning of the start bit

of the next frame.

200

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 18-7. Stop Bit Sampling and Next Start Bit Sampling

(A)

(B)

(C)

RxD

STOP 1

Sample

(U2X = 0)

1

2

3

2

4

5

3

6

7

4

8

9

5

10

0/1 0/1 0/1

Sample

(U2X = 1)

6

0/1

The same majority voting is done to the stop bit as done for the other bits in the frame. If the stop

bit is registered to have a logic 0 value, the Frame Error (FEn) Flag will be set.

A new high to low transition indicating the start bit of a new frame can come right after the last of

the bits used for majority voting. For Normal Speed mode, the first low level sample can be at

point marked (A) in Figure 18-7. For Double Speed mode the first low level must be delayed to

(B). (C) marks a stop bit of full length. The early start bit detection influences the operational

range of the Receiver.

18.7.3

Asynchronous Operational Range

The operational range of the Receiver is dependent on the mismatch between the received bit

rate and the internally generated baud rate. If the Transmitter is sending frames at too fast or too

slow bit rates, or the internally generated baud rate of the Receiver does not have a similar (see

Table 18-2) base frequency, the Receiver will not be able to synchronize the frames to the start

bit.

The following equations can be used to calculate the ratio of the incoming data rate and internal

receiver baud rate.

(D + 1)S

S – 1 + D ⋅ S + S

(D + 2)S

(D + 1)S + S

R

= ------------------------------------------

R

= -----------------------------------

slow

fast

F

M

D

S

Sum of character size and parity size (D = 5 to 10 bit)

Samples per bit. S = 16 for Normal Speed mode and S = 8 for Double Speed

mode.

S_F

First sample number used for majority voting. S_F= 8 for normal speed and S_F= 4

for Double Speed mode.

S_M

Middle sample number used for majority voting. S_M= 9 for normal speed and

S_M= 5 for Double Speed mode.

R_slow

is the ratio of the slowest incoming data rate that can be accepted in relation to the

receiver baud rate. R_fastis the ratio of the fastest incoming data rate that can be

accepted in relation to the receiver baud rate.

Table 18-2 and Table 18-3 list the maximum receiver baud rate error that can be tolerated. Note

that Normal Speed mode has higher toleration of baud rate variations.

201

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 18-2. Recommended Maximum Receiver Baud Rate Error for Normal Speed Mode

(U2Xn = 0)

D

Recommended Max

Receiver Error (%)

# (Data+Parity Bit)

R

_slow(%)

93.20

94.12

94.81

95.36

95.81

96.17

R_fast(%)

106.67

105.79

105.11

104.58

104.14

103.78

Max Total Error (%)

+6.67/-6.8

5

6

3.0

2.5

2.0

1.5

+5.79/-5.88

+5.11/-5.19

+4.58/-4.54

+4.14/-4.19

+3.78/-3.83

7

8

9

10

Table 18-3. Recommended Maximum Receiver Baud Rate Error for Double Speed Mode

(U2Xn = 1)

D

Recommended Max

Receiver Error (%)

# (Data+Parity Bit) R_slow(%) R_fast(%) Max Total Error (%)

5

6

94.12

94.92

95.52

96.00

96.39

96.70

105.66

104.92

104,35

103.90

103.53

103.23

+5.66/-5.88

+4.92/-5.08

+4.35/-4.48

+3.90/-4.00

+3.53/-3.61

+3.23/-3.30

2.5

2.0

1.5

1.0

7

8

9

10

The recommendations of the maximum receiver baud rate error was made under the assump-

tion that the Receiver and Transmitter equally divides the maximum total error.

There are two possible sources for the receivers baud rate error. The Receiver’s system clock

(XTAL) will always have some minor instability over the supply voltage range and the tempera-

ture range. When using a crystal to generate the system clock, this is rarely a problem, but for a

resonator the system clock may differ more than 2% depending of the resonators tolerance. The

second source for the error is more controllable. The baud rate generator can not always do an

exact division of the system frequency to get the baud rate wanted. In this case an UBRR value

that gives an acceptable low error can be used if possible.

18.8 Multi-processor Communication Mode

Setting the Multi-processor Communication mode (MPCMn) bit in UCSRnA enables a filtering

function of incoming frames received by the USART Receiver. Frames that do not contain

address information will be ignored and not put into the receive buffer. This effectively reduces

the number of incoming frames that has to be handled by the CPU, in a system with multiple

MCUs that communicate via the same serial bus. The Transmitter is unaffected by the MPCMn

setting, but has to be used differently when it is a part of a system utilizing the Multi-processor

Communication mode.

If the Receiver is set up to receive frames that contain 5 to 8 data bits, then the first stop bit indi-

cates if the frame contains data or address information. If the Receiver is set up for frames with

202

7766E–AVR–04/10

ATmega16U4/ATmega32U4

nine data bits, then the ninth bit (RXB8n) is used for identifying address and data frames. When

the frame type bit (the first stop or the ninth bit) is one, the frame contains an address. When the

frame type bit is zero the frame is a data frame.

The Multi-processor Communication mode enables several slave MCUs to receive data from a

master MCU. This is done by first decoding an address frame to find out which MCU has been

addressed. If a particular slave MCU has been addressed, it will receive the following data

frames as normal, while the other slave MCUs will ignore the received frames until another

address frame is received.

18.8.1

Using MPCMn

For an MCU to act as a master MCU, it can use a 9-bit character frame format (UCSZn = 7). The

ninth bit (TXB8n) must be set when an address frame (TXB8n = 1) or cleared when a data frame

(TXB = 0) is being transmitted. The slave MCUs must in this case be set to use a 9-bit character

frame format.

The following procedure should be used to exchange data in Multi-processor Communication

mode:

1. All Slave MCUs are in Multi-processor Communication mode (MPCMn in UCSRnA is

set).

2. The Master MCU sends an address frame, and all slaves receive and read this frame.

In the Slave MCUs, the RXCn Flag in UCSRnA will be set as normal.

3. Each Slave MCU reads the UDRn Register and determines if it has been selected. If

so, it clears the MPCMn bit in UCSRnA, otherwise it waits for the next address byte and

keeps the MPCMn setting.

4. The addressed MCU will receive all data frames until a new address frame is received.

The other Slave MCUs, which still have the MPCMn bit set, will ignore the data frames.

5. When the last data frame is received by the addressed MCU, the addressed MCU sets

the MPCMn bit and waits for a new address frame from master. The process then

repeats from 2.

Using any of the 5- to 8-bit character frame formats is possible, but impractical since the

Receiver must change between using n and n+1 character frame formats. This makes full-

duplex operation difficult since the Transmitter and Receiver uses the same character size set-

ting. If 5- to 8-bit character frames are used, the Transmitter must be set to use two stop bit

(USBSn = 1) since the first stop bit is used for indicating the frame type.

Do not use Read-Modify-Write instructions (SBI and CBI) to set or clear the MPCMn bit. The

MPCMn bit shares the same I/O location as the TXCn Flag and this might accidentally be

cleared when using SBI or CBI instructions.

18.9 Hardware Flow Control

The hardware flow control can be enabled by software.

CTS: (Clear to Send)

RTS: (Request to Send)

HOST

ATmega16U4/ATm

TXD

RXD

CTS

RTS

TXD

RXD

CTS

RTS

203

7766E–AVR–04/10

ATmega16U4/ATmega32U4

18.9.1

Receiver Flow Control

The reception flow can be controlled by hardware using the RTS pin. The aim of the flow control

is to inform the external transmitter when the internal receive Fifo is full. Thus the transmitter can

stop sending characters. RTS usage and so associated flow control is enabled using RTSEN bit

in UCSRnD.

Figure 18-8. shows a reception example.

Figure 18-8. Reception Flow Control Waveform Example

FIFO

0

1

2

1

0

1

Index

RXD

RTS

CPU Read

C3

C1 C2

Figure 18-9. RTS behavior

Stop

Start

Byte0

Start

Byte1

Start

Byte2

RXD

1 additional byte may be sent

if the transmitter misses the RTS trig

RTS

Read from CPU

RTS will rise at 2/3 of the last received stop bit if the receive fifo is full.

To ensure reliable transmissions, even after a RTS rise, an extra-data can still be received and

stored in the Receive Shift Register.

18.9.2

Transmission Flow Control

The transmission flow can be controlled by hardware using the CTS pin controlled by the exter-

nal receiver. The aim of the flow control is to stop transmission when the receiver is full of data

(CTS = 1). CTS usage and so associated flow control is enabled using CTSEN bit in UCSRnD.

The CTS pin is sampled at each CPU write and at the middle of the last stop bit that is currently

being sent.

Figure 18-10. CTS behavior

Write from CPU

Stop

Start

Byte0

Start

Byte1

Start

Byte2

TXD

CTS

sample

204

7766E–AVR–04/10

ATmega16U4/ATmega32U4

18.10 USART Register Description

18.10.1 USART I/O Data Register n– UDRn

Bit

7

6

5

4

3

2

1

0

RXB[7:0]

TXB[7:0]

R/W

UDRn (Read)

UDRn (Write)

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

The USART Transmit Data Buffer Register and USART Receive Data Buffer Registers share the

same I/O address referred to as USART Data Register or UDRn. The Transmit Data Buffer Reg-

ister (TXB) will be the destination for data written to the UDRn Register location. Reading the

UDRn Register location will return the contents of the Receive Data Buffer Register (RXB).

For 5-, 6-, or 7-bit characters the upper unused bits will be ignored by the Transmitter and set to

zero by the Receiver.

The transmit buffer can only be written when the UDREn Flag in the UCSRnA Register is set.

Data written to UDRn when the UDREn Flag is not set, will be ignored by the USART Transmit-

ter. When data is written to the transmit buffer, and the Transmitter is enabled, the Transmitter

will load the data into the Transmit Shift Register when the Shift Register is empty. Then the

data will be serially transmitted on the TxDn pin.

The receive buffer consists of a two level FIFO. The FIFO will change its state whenever the

receive buffer is accessed. Due to this behavior of the receive buffer, do not use Read-Modify-

Write instructions (SBI and CBI) on this location. Be careful when using bit test instructions

(SBIC and SBIS), since these also will change the state of the FIFO.

18.10.2 USART Control and Status Register A – UCSRnA

Bit

7

6

5

4

3

2

1

0

RXCn

TXCn

UDREn

FEn

R

DORn

UPEn

U2Xn

R/W

0

MPCMn

R/W

0

UCSRnA

Read/Write

Initial Value

R

0

R/W

0

R

1

R

0

R

0

• Bit 7 – RXCn: USART Receive Complete

This flag bit is set when there are unread data in the receive buffer and cleared when the receive

buffer is empty (i.e., does not contain any unread data). If the Receiver is disabled, the receive

buffer will be flushed and consequently the RXCn bit will become zero. The RXCn Flag can be

used to generate a Receive Complete interrupt (see description of the RXCIEn bit).

• Bit 6 – TXCn: USART Transmit Complete

This flag bit is set when the entire frame in the Transmit Shift Register has been shifted out and

there are no new data currently present in the transmit buffer (UDRn). The TXCn Flag bit is auto-

matically cleared when a transmit complete interrupt is executed, or it can be cleared by writing

a one to its bit location. The TXCn Flag can generate a Transmit Complete interrupt (see

description of the TXCIEn bit).

• Bit 5 – UDREn: USART Data Register Empty

The UDREn Flag indicates if the transmit buffer (UDRn) is ready to receive new data. If UDREn

is one, the buffer is empty, and therefore ready to be written. The UDREn Flag can generate a

Data Register Empty interrupt (see description of the UDRIEn bit).

UDREn is set after a reset to indicate that the Transmitter is ready.

205

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 4 – FEn: Frame Error

This bit is set if the next character in the receive buffer had a Frame Error when received. I.e.,

when the first stop bit of the next character in the receive buffer is zero. This bit is valid until the

receive buffer (UDRn) is read. The FEn bit is zero when the stop bit of received data is one.

Always set this bit to zero when writing to UCSRnA.

• Bit 3 – DORn: Data OverRun

This bit is set if a Data OverRun condition is detected. A Data OverRun occurs when the receive

buffer is full (two characters), it is a new character waiting in the Receive Shift Register, and a

new start bit is detected. This bit is valid until the receive buffer (UDRn) is read. Always set this

bit to zero when writing to UCSRnA.

• Bit 2 – UPEn: USART Parity Error

This bit is set if the next character in the receive buffer had a Parity Error when received and the

Parity Checking was enabled at that point (UPMn1 = 1). This bit is valid until the receive buffer

(UDRn) is read. Always set this bit to zero when writing to UCSRnA.

• Bit 1 – U2Xn: Double the USART Transmission Speed

This bit only has effect for the asynchronous operation. Write this bit to zero when using syn-

chronous operation.

Writing this bit to one will reduce the divisor of the baud rate divider from 16 to 8 effectively dou-

bling the transfer rate for asynchronous communication.

• Bit 0 – MPCMn: Multi-processor Communication Mode

This bit enables the Multi-processor Communication mode. When the MPCMn bit is written to

one, all the incoming frames received by the USART Receiver that do not contain address infor-

mation will be ignored. The Transmitter is unaffected by the MPCMn setting. For more detailed

information see “Multi-processor Communication Mode” on page 202.

18.10.3 USART Control and Status Register n B – UCSRnB

Bit

7

6

5

4

3

2

1

0

RXCIEn

TXCIEn

UDRIEn

RXENn

R/W

0

TXENn

R/W

0

UCSZn2

R/W

0

RXB8n

TXB8n

R/W

0

UCSRnB

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R

0

• Bit 7 – RXCIEn: RX Complete Interrupt Enable n

Writing this bit to one enables interrupt on the RXCn Flag. A USART Receive Complete interrupt

will be generated only if the RXCIEn bit is written to one, the Global Interrupt Flag in SREG is

written to one and the RXCn bit in UCSRnA is set.

• Bit 6 – TXCIEn: TX Complete Interrupt Enable n

Writing this bit to one enables interrupt on the TXCn Flag. A USART Transmit Complete interrupt

will be generated only if the TXCIEn bit is written to one, the Global Interrupt Flag in SREG is

written to one and the TXCn bit in UCSRnA is set.

• Bit 5 – UDRIEn: USART Data Register Empty Interrupt Enable n

Writing this bit to one enables interrupt on the UDREn Flag. A Data Register Empty interrupt will

be generated only if the UDRIEn bit is written to one, the Global Interrupt Flag in SREG is written

to one and the UDREn bit in UCSRnA is set.

206

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 4 – RXENn: Receiver Enable n

Writing this bit to one enables the USART Receiver. The Receiver will override normal port oper-

ation for the RxDn pin when enabled. Disabling the Receiver will flush the receive buffer

invalidating the FEn, DORn, and UPEn Flags.

• Bit 3 – TXENn: Transmitter Enable n

Writing this bit to one enables the USART Transmitter. The Transmitter will override normal port

operation for the TxDn pin when enabled. The disabling of the Transmitter (writing TXENn to

zero) will not become effective until ongoing and pending transmissions are completed, i.e.,

when the Transmit Shift Register and Transmit Buffer Register do not contain data to be trans-

mitted. When disabled, the Transmitter will no longer override the TxDn port.

• Bit 2 – UCSZn2: Character Size n

The UCSZn2 bits combined with the UCSZn1:0 bit in UCSRnC sets the number of data bits

(Character SiZe) in a frame the Receiver and Transmitter use.

• Bit 1 – RXB8n: Receive Data Bit 8 n

RXB8n is the ninth data bit of the received character when operating with serial frames with nine

data bits. Must be read before reading the low bits from UDRn.

• Bit 0 – TXB8n: Transmit Data Bit 8 n

TXB8n is the ninth data bit in the character to be transmitted when operating with serial frames

with nine data bits. Must be written before writing the low bits to UDRn.

18.10.4 USART Control and Status Register n C – UCSRnC

Bit

7

6

5

4

3

2

1

0

UMSELn1

UMSELn0

UPMn1

UPMn0

R/W

0

USBSn

R/W

0

UCSZn1

R/W

1

UCSZn0

R/W

1

UCPOLn

UCSRnC

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

• Bits 7:6 – UMSELn1:0 USART Mode Select

These bits select the mode of operation of the USARTn as shown in Table 18-4.

Table 18-4. UMSELn Bits Settings

UMSELn1

UMSELn0

Mode

0

1

0

1

0

1

Asynchronous USART

Synchronous USART

(Reserved)

Master SPI (MSPIM)⁽¹⁾

Note:

1. See “USART in SPI Mode” on page 214 for full description of the Master SPI Mode (MSPIM)

operation

• Bits 5:4 – UPMn1:0: Parity Mode

These bits enable and set type of parity generation and check. If enabled, the Transmitter will

automatically generate and send the parity of the transmitted data bits within each frame. The

207

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Receiver will generate a parity value for the incoming data and compare it to the UPMn setting.

If a mismatch is detected, the UPEn Flag in UCSRnA will be set.

Table 18-5. UPMn Bits Settings

UPMn1

UPMn0

Parity Mode

0

1

0

1

0

1

Disabled

Reserved

Enabled, Even Parity

Enabled, Odd Parity

• Bit 3 – USBSn: Stop Bit Select

This bit selects the number of stop bits to be inserted by the Transmitter. The Receiver ignores

this setting.

Table 18-6. USBS Bit Settings

USBSn

Stop Bit(s)

1-bit

0

1

2-bit

• Bit 2:1 – UCSZn1:0: Character Size

The UCSZn1:0 bits combined with the UCSZn2 bit in UCSRnB sets the number of data bits

(Character SiZe) in a frame the Receiver and Transmitter use.

Table 18-7. UCSZn Bits Settings

UCSZn2

UCSZn1

UCSZn0

Character Size

5-bit

0

1

0

1

0

1

0

1

0

1

0

1

0

1

6-bit

7-bit

8-bit

Reserved

9-bit

• Bit 0 – UCPOLn: Clock Polarity

This bit is used for synchronous mode only. Write this bit to zero when asynchronous mode is

used. The UCPOLn bit sets the relationship between data output change and data input sample,

and the synchronous clock (XCKn).

Table 18-8. UCPOLn Bit Settings

Transmitted Data Changed (Output

of TxDn Pin)

Received Data Sampled (Input on

RxDn Pin)

UCPOLn

0

1

Rising XCKn Edge

Falling XCKn Edge

Rising XCKn Edge

208

7766E–AVR–04/10

ATmega16U4/ATmega32U4

18.10.5 USART Control and Status Register n D– UCSRnD

Bit

7

6

5

4

3

2

1

0

–

CTSEN

R/W

0

RTSEN

R/W

0

UCSRnD

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

• Bits 7:2 – Reserved bits

These bits are reserved and will be read as ‘0’. Do not set these bits.

• Bits 1 – CTSEN: UART CTS Signal Enable

Set this bit by firmware to enable the transmission flow control signal (CTS). Transmission will

be enabled only if CTS input = 0. Clear this bit to disable the transmission flow control signal.

Transmission will occur without hardware condition. Data Direction Register bit must be correctly

clear to enable the pin as an input.

• Bits 0 – RTSEN: UART RTS Signal Enable

Set this bit by firmware to enable the reception flow control signal (RTS). In this case the RTS

line will automatically rise when the FIFO is full. Clear this bit to disable the reception flow control

signal. Data Direction Register bit must be correctly set to enable the pin as an output.

18.10.6 USART Baud Rate Registers – UBRRLn and UBRRHn

Bit

15

14

13

12

11

10

9

8

–

UBRR[11:8]

UBRRHn

UBRRLn

UBRR[7:0]

7

6

5

4

3

2

1

0

Read/Write

Initial Value

R

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

• Bit 15:12 – Reserved Bits

These bits are reserved for future use. For compatibility with future devices, these bit must be

written to zero when UBRRH is written.

• Bit 11:0 – UBRR11:0: USART Baud Rate Register

This is a 12-bit register which contains the USART baud rate. The UBRRH contains the four

most significant bits, and the UBRRL contains the eight least significant bits of the USART baud

rate. Ongoing transmissions by the Transmitter and Receiver will be corrupted if the baud rate is

changed. Writing UBRRL will trigger an immediate update of the baud rate prescaler.

18.11 Examples of Baud Rate Setting

For standard crystal and resonator frequencies, the most commonly used baud rates for asyn-

chronous operation can be generated by using the UBRR settings in Table 18-9 to Table 18-12.

UBRR values which yield an actual baud rate differing less than 0.5% from the target baud rate,

are bold in the table. Higher error ratings are acceptable, but the Receiver will have less noise

209

7766E–AVR–04/10

ATmega16U4/ATmega32U4

resistance when the error ratings are high, especially for large serial frames (see “Asynchronous

Operational Range” on page 201). The error values are calculated using the following equation:

BaudRate_{Closest Match}

⎛

⎞

Error[%] = ------------------------------------------------------- – 1 • 100%

⎝

⎠

BaudRate

Table 18-9. Examples of UBRRn Settings for Commonly Used Oscillator Frequencies

f_osc= 1.0000 MHz f_osc= 1.8432 MHz

U2Xn = 0 U2Xn = 1 U2Xn = 0 U2Xn = 1

UBRR UBRR UBRR UBRR

f_osc= 2.0000 MHz

U2Xn = 0 U2Xn = 1

UBRR UBRR

Baud

Rate

(bps)

Error

0.2%

-7.0%

8.5%

-18.6%

8.5%

–

Error

0.2%

-3.5%

-7.0%

8.5%

-18.6%

8.5%

–

Error

0.0%

-25.0%

0.0%

–

Error

0.0%

–

Error

0.2%

-3.5%

-7.0%

8.5%

-18.6%

8.5%

–

Error

0.2%

2.1%

0.2%

-3.5%

-7.0%

8.5%

–

2400

25

12

6

51

25

12

8

47

23

11

7

95

47

23

15

11

7

51

25

12

8

103

51

25

16

12

8

4800

9600

14.4k

19.2k

28.8k

38.4k

57.6k

76.8k

115.2k

230.4k

250k

3

2

6

5

6

1

3

1

2

5

2

6

0

1

3

1

3

–

1

2

1

2

–

0

1

0

1

–

0

–

0

0.0%

Max. ⁽¹⁾

62.5 kbps

UBRR = 0, Error = 0.0%

125 kbps

115.2 kbps

230.4 kbps

125 kbps

250 kbps

1.

210

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 18-10. Examples of UBRRn Settings for Commonly Used Oscillator Frequencies (Continued)

f_osc= 3.6864 MHz

U2Xn = 0

f_osc= 4.0000 MHz

U2Xn = 0

f_osc= 7.3728 MHz

U2Xn = 0

Baud

Rate

(bps)

U2Xn = 1

UBRR

Error

0.0%

-7.8%

–

UBRR

191

95

47

31

23

15

11

7

Error

0.0%

-7.8%

–

UBRR

Error

0.2%

2.1%

0.2%

-3.5%

-7.0%

8.5%

0.0%

–

UBRR

207

103

51

34

25

16

12

8

Error

0.2%

-0.8%

0.2%

2.1%

0.2%

-3.5%

-7.0%

8.5%

0.0%

–

UBRR

191

95

47

31

23

15

11

7

Error

0.0%

-7.8%

–

UBRR

383

191

95

63

47

31

23

15

11

7

Error

0.0%

-7.8%

2400

4800

9600

14.4k

19.2k

28.8k

38.4k

57.6k

76.8k

115.2k

230.4k

250k

95

47

23

15

11

7

103

51

25

16

12

8

5

6

3

2

5

2

6

5

1

3

1

3

0

1

0

1

3

0

1

0

1

3

0.5M

–

0

–

0

1

1M

–

0

Max. ⁽¹⁾

230.4 kbps

UBRR = 0, Error = 0.0%

460.8 kbps

250 kbps

0.5 Mbps

460.8 kbps

921.6 kbps

1.

211

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 18-11. Examples of UBRRn Settings for Commonly Used Oscillator Frequencies (Continued)

f_osc= 8.0000 MHz f_osc= 11.0592 MHz f_osc= 14.7456 MHz

U2Xn = 0 U2Xn = 1 U2Xn = 0 U2Xn = 1 U2Xn = 0 U2Xn = 1

UBRR UBRR UBRR UBRR UBRR UBRR

Baud

Rate

(bps)

Error

0.2%

-0.8%

0.2%

2.1%

0.2%

-3.5%

-7.0%

8.5%

0.0%

–

Error

-0.1%

0.2%

0.6%

0.2%

-0.8%

0.2%

2.1%

0.2%

-3.5%

8.5%

0.0%

Error

0.0%

-7.8%

–

Error

0.0%

-7.8%

–

Error

0.0%

-7.8%

Error

0.0%

5.3%

-7.8%

2400

4800

9600

14.4k

19.2k

28.8k

38.4k

57.6k

76.8k

115.2k

230.4k

250k

207

103

51

34

25

16

12

8

416

207

103

68

51

34

25

16

12

8

287

143

71

47

35

23

17

11

8

575

287

143

95

71

47

35

23

17

11

5

383

191

95

63

47

31

23

15

11

7

767

383

191

127

95

63

47

31

23

15

7

6

3

5

1

3

2

3

1

3

2

5

3

6

0.5M

0

1

–

2

1

3

1M

–

0

–

0

1

Max. ⁽¹⁾

0.5 Mbps

UBRR = 0, Error = 0.0%

1 Mbps

691.2 kbps

1.3824 Mbps

921.6 kbps

1.8432 Mbps

1.

212

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 18-12. Examples of UBRRn Settings for Commonly Used Oscillator Frequencies (Continued)

f_osc= 16.0000 MHz f_osc= 18.4320 MHz f_osc= 20.0000 MHz

U2Xn = 0 U2Xn = 1 U2Xn = 0 U2Xn = 1 U2Xn = 0 U2Xn = 1

UBRR UBRR UBRR UBRR UBRR UBRR

Baud

Rate

(bps)

Error

-0.1%

0.2%

0.6%

0.2%

-0.8%

0.2%

2.1%

0.2%

-3.5%

8.5%

0.0%

Error

0.0%

-0.1%

0.2%

-0.1%

0.2%

0.6%

0.2%

-0.8%

0.2%

2.1%

-3.5%

0.0%

Error

0.0%

-7.8%

–

Error

0.0%

2.4%

-7.8%

–

Error

0.0%

0.2%

-0.2%

0.2%

0.9%

-1.4%

1.7%

-1.4%

8.5%

0.0%

–

Error

0.0%

0.2%

-0.2%

0.2%

-0.2%

0.2%

0.9%

-1.4%

0.0%

–

2400

4800

9600

14.4k

19.2k

28.8k

38.4k

57.6k

76.8k

115.2k

230.4k

250k

416

207

103

68

51

34

25

16

12

8

832

416

207

138

103

68

51

34

25

16

8

479

239

119

79

59

39

29

19

14

9

959

479

239

159

119

79

59

39

29

19

9

520

259

129

86

64

42

32

21

15

10

4

1041

520

259

173

129

86

64

42

32

21

3

4

10

3

7

4

8

4

9

0.5M

1

3

–

4

–

4

1M

0

1

–

Max. ⁽¹⁾

1 Mbps

UBRR = 0, Error = 0.0%

2 Mbps

1.152 Mbps

2.304 Mbps

1.25 Mbps

2.5 Mbps

1.

213

7766E–AVR–04/10

ATmega16U4/ATmega32U4

19. USART in SPI Mode

The Universal Synchronous and Asynchronous serial Receiver and Transmitter (USART) can be

set to a master SPI compliant mode of operation. The Master SPI Mode (MSPIM) has the follow-

ing features:

• Full Duplex, Three-wire Synchronous Data Transfer

• Master Operation

• Supports all four SPI Modes of Operation (Mode 0, 1, 2, and 3)

• LSB First or MSB First Data Transfer (Configurable Data Order)

• Queued Operation (Double Buffered)

• High Resolution Baud Rate Generator

• High Speed Operation (fXCKmax = fCK/2)

• Flexible Interrupt Generation

19.1 Overview

Setting both UMSELn1:0 bits to one enables the USART in MSPIM logic. In this mode of opera-

tion the SPI master control logic takes direct control over the USART resources. These

resources include the transmitter and receiver shift register and buffers, and the baud rate gen-

erator. The parity generator and checker, the data and clock recovery logic, and the RX and TX

control logic is disabled. The USART RX and TX control logic is replaced by a common SPI

transfer control logic. However, the pin control logic and interrupt generation logic is identical in

both modes of operation.

The I/O register locations are the same in both modes. However, some of the functionality of the

control registers changes when using MSPIM.

19.2 Clock Generation

The Clock Generation logic generates the base clock for the Transmitter and Receiver. For

USART MSPIM mode of operation only internal clock generation (i.e. master operation) is sup-

ported. The Data Direction Register for the XCKn pin (DDR_XCKn) must therefore be set to one

(i.e. as output) for the USART in MSPIM to operate correctly. Preferably the DDR_XCKn should

be set up before the USART in MSPIM is enabled (i.e. TXENn and RXENn bit set to one).

The internal clock generation used in MSPIM mode is identical to the USART synchronous mas-

ter mode. The baud rate or UBRRn setting can therefore be calculated using the same

equations, see Table 19-1:

Table 19-1. Equations for Calculating Baud Rate Register Setting

Equation for Calculating Baud

Rate⁽¹⁾

Equation for Calculating

UBRRn Value

Operating Mode

f

OSC

f

OSC

Synchronous Master

mode

BAUD = --------------------------------------

UBRRn = -------------------- – 1

2(UBRRn + 1)

2BAUD

214

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Note:

1. The baud rate is defined to be the transfer rate in bit per second (bps)

BAUD

f_OSC

Baud rate (in bits per second, bps)

System Oscillator clock frequency

UBRRn

Contents of the UBRRnH and UBRRnL Registers, (0-4095)

19.3 SPI Data Modes and Timing

There are four combinations of XCKn (SCK) phase and polarity with respect to serial data, which

are determined by control bits UCPHAn and UCPOLn. The data transfer timing diagrams are

shown in Figure 19-1. Data bits are shifted out and latched in on opposite edges of the XCKn

signal, ensuring sufficient time for data signals to stabilize. The UCPOLn and UCPHAn function-

ality is summarized in Table 19-2. Note that changing the setting of any of these bits will corrupt

all ongoing communication for both the Receiver and Transmitter.

Table 19-2. UCPOLn and UCPHAn Functionality-

UCPOLn

UCPHAn

SPI Mode

Leading Edge

Sample (Rising)

Setup (Rising)

Sample (Falling)

Setup (Falling)

Trailing Edge

Setup (Falling)

Sample (Falling)

Setup (Rising)

Sample (Rising)

0

1

0

1

0

1

0

1

2

3

Figure 19-1. UCPHAn and UCPOLn data transfer timing diagrams.

UCPOL=0

UCPOL=1

XCK

Data setup (TXD)

Data sample (RXD)

Data setup (TXD)

Data sample (RXD)

XCK

Data setup (TXD)

Data sample (RXD)

Data setup (TXD)

Data sample (RXD)

19.4 Frame Formats

A serial frame for the MSPIM is defined to be one character of 8 data bits. The USART in MSPIM

mode has two valid frame formats:

• 8-bit data with MSB first

• 8-bit data with LSB first

A frame starts with the least or most significant data bit. Then the next data bits, up to a total of

eight, are succeeding, ending with the most or least significant bit accordingly. When a complete

frame is transmitted, a new frame can directly follow it, or the communication line can be set to

an idle (high) state.

215

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The UDORDn bit in UCSRnC sets the frame format used by the USART in MSPIM mode. The

Receiver and Transmitter use the same setting. Note that changing the setting of any of these

bits will corrupt all ongoing communication for both the Receiver and Transmitter.

16-bit data transfer can be achieved by writing two data bytes to UDRn. A UART transmit com-

plete interrupt will then signal that the 16-bit value has been shifted out.

19.4.1

USART MSPIM Initialization

The USART in MSPIM mode has to be initialized before any communication can take place. The

initialization process normally consists of setting the baud rate, setting master mode of operation

(by setting DDR_XCKn to one), setting frame format and enabling the Transmitter and the

Receiver. Only the transmitter can operate independently. For interrupt driven USART opera-

tion, the Global Interrupt Flag should be cleared (and thus interrupts globally disabled) when

doing the initialization.

Note:

To ensure immediate initialization of the XCKn output the baud-rate register (UBRRn) must be

zero at the time the transmitter is enabled. Contrary to the normal mode USART operation the

UBRRn must then be written to the desired value after the transmitter is enabled, but before the

first transmission is started. Setting UBRRn to zero before enabling the transmitter is not neces-

sary if the initialization is done immediately after a reset since UBRRn is reset to zero.

Before doing a re-initialization with changed baud rate, data mode, or frame format, be sure that

there is no ongoing transmissions during the period the registers are changed. The TXCn Flag

can be used to check that the Transmitter has completed all transfers, and the RXCn Flag can

be used to check that there are no unread data in the receive buffer. Note that the TXCn Flag

must be cleared before each transmission (before UDRn is written) if it is used for this purpose.

The following simple USART initialization code examples show one assembly and one C func-

tion that are equal in functionality. The examples assume polling (no interrupts enabled). The

baud rate is given as a function parameter. For the assembly code, the baud rate parameter is

assumed to be stored in the r17:r16 registers.

216

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Assembly Code Example⁽¹⁾

USART_Init:

clr r18

out UBRRnH,r18

out UBRRnL,r18

; Setting the XCKn port pin as output, enables master mode.

sbi XCKn_DDR, XCKn

; Set MSPI mode of operation and SPI data mode 0.

ldi r18, (1<<UMSELn1)|(1<<UMSELn0)|(0<<UCPHAn)|(0<<UCPOLn)

out UCSRnC,r18

; Enable receiver and transmitter.

ldi r18, (1<<RXENn)|(1<<TXENn)

out UCSRnB,r18

; Set baud rate.

; IMPORTANT: The Baud Rate must be set after the transmitter is

enabled!

out UBRRnH, r17

out UBRRnL, r18

ret

C Code Example⁽¹⁾

void USART_Init( unsigned int baud )

{

UBRRn = 0;

/* Setting the XCKn port pin as output, enables master mode. */

XCKn_DDR |= (1<<XCKn);

/* Set MSPI mode of operation and SPI data mode 0. */

UCSRnC = (1<<UMSELn1)|(1<<UMSELn0)|(0<<UCPHAn)|(0<<UCPOLn);

/* Enable receiver and transmitter. */

UCSRnB = (1<<RXENn)|(1<<TXENn);

/* Set baud rate. */

/* IMPORTANT: The Baud Rate must be set after the transmitter is

enabled */

UBRRn = baud;

}

Note:

1. See “Code Examples” on page 8.

19.5 Data Transfer

Using the USART in MSPI mode requires the Transmitter to be enabled, i.e. the TXENn bit in

the UCSRnB register is set to one. When the Transmitter is enabled, the normal port operation

of the TxDn pin is overridden and given the function as the Transmitter's serial output. Enabling

the receiver is optional and is done by setting the RXENn bit in the UCSRnB register to one.

When the receiver is enabled, the normal pin operation of the RxDn pin is overridden and given

the function as the Receiver's serial input. The XCKn will in both cases be used as the transfer

clock.

217

7766E–AVR–04/10

ATmega16U4/ATmega32U4

After initialization the USART is ready for doing data transfers. A data transfer is initiated by writ-

ing to the UDRn I/O location. This is the case for both sending and receiving data since the

transmitter controls the transfer clock. The data written to UDRn is moved from the transmit buf-

fer to the shift register when the shift register is ready to send a new frame.

Note:

To keep the input buffer in sync with the number of data bytes transmitted, the UDRn register must

be read once for each byte transmitted. The input buffer operation is identical to normal USART

mode, i.e. if an overflow occurs the character last received will be lost, not the first data in the buf-

fer. This means that if four bytes are transferred, byte 1 first, then byte 2, 3, and 4, and the UDRn

is not read before all transfers are completed, then byte 3 to be received will be lost, and not byte

1.

The following code examples show a simple USART in MSPIM mode transfer function based on

polling of the Data Register Empty (UDREn) Flag and the Receive Complete (RXCn) Flag. The

USART has to be initialized before the function can be used. For the assembly code, the data to

be sent is assumed to be stored in Register R16 and the data received will be available in the

same register (R16) after the function returns.

The function simply waits for the transmit buffer to be empty by checking the UDREn Flag,

before loading it with new data to be transmitted. The function then waits for data to be present

in the receive buffer by checking the RXCn Flag, before reading the buffer and returning the

value.

Assembly Code Example⁽¹⁾

USART_MSPIM_Transfer:

; Wait for empty transmit buffer

sbis UCSRnA, UDREn

rjmp USART_MSPIM_Transfer

; Put data (r16) into buffer, sends the data

out UDRn,r16

; Wait for data to be received

USART_MSPIM_Wait_RXCn:

sbis UCSRnA, RXCn

rjmp USART_MSPIM_Wait_RXCn

; Get and return received data from buffer

in r16, UDRn

ret

C Code Example⁽¹⁾

unsigned char USART_Receive( void )

{

/* Wait for empty transmit buffer */

while ( !( UCSRnA & (1<<UDREn)) );

/* Put data into buffer, sends the data */

UDRn = data;

/* Wait for data to be received */

while ( !(UCSRnA & (1<<RXCn)) );

/* Get and return received data from buffer */

return UDRn;

}

218

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Note:

1. See “Code Examples” on page 8.

19.5.1

19.5.2

Transmitter and Receiver Flags and Interrupts

The RXCn, TXCn, and UDREn flags and corresponding interrupts in USART in MSPIM mode

are identical in function to the normal USART operation. However, the receiver error status flags

(FE, DOR, and PE) are not in use and is always read as zero.

Disabling the Transmitter or Receiver

The disabling of the transmitter or receiver in USART in MSPIM mode is identical in function to

the normal USART operation.

19.6 USART MSPIM Register Description

The following section describes the registers used for SPI operation using the USART.

19.6.1

19.6.2

USART MSPIM I/O Data Register - UDRn

The function and bit description of the USART data register (UDRn) in MSPI mode is identical to

normal USART operation. See “USART I/O Data Register n– UDRn” on page 205.

USART MSPIM Control and Status Register n A - UCSRnA

Bit

7

6

5

4

3

2

-

1

-

0

-

RXCn

TXCn

UDREn

-

UCSRnA

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R

0

R

0

R

1

R

1

R

0

• Bit 7 - RXCn: USART Receive Complete

This flag bit is set when there are unread data in the receive buffer and cleared when the receive

buffer is empty (i.e., does not contain any unread data). If the Receiver is disabled, the receive

buffer will be flushed and consequently the RXCn bit will become zero. The RXCn Flag can be

used to generate a Receive Complete interrupt (see description of the RXCIEn bit).

• Bit 6 - TXCn: USART Transmit Complete

This flag bit is set when the entire frame in the Transmit Shift Register has been shifted out and

there are no new data currently present in the transmit buffer (UDRn). The TXCn Flag bit is auto-

matically cleared when a transmit complete interrupt is executed, or it can be cleared by writing

a one to its bit location. The TXCn Flag can generate a Transmit Complete interrupt (see

description of the TXCIEn bit).

• Bit 5 - UDREn: USART Data Register Empty

The UDREn Flag indicates if the transmit buffer (UDRn) is ready to receive new data. If UDREn

is one, the buffer is empty, and therefore ready to be written. The UDREn Flag can generate a

Data Register Empty interrupt (see description of the UDRIE bit). UDREn is set after a reset to

indicate that the Transmitter is ready.

• Bit 4:0 - Reserved Bits in MSPI mode

When in MSPI mode, these bits are reserved for future use. For compatibility with future devices,

these bits must be written to zero when UCSRnA is written.

219

7766E–AVR–04/10

ATmega16U4/ATmega32U4

19.6.3

USART MSPIM Control and Status Register n B - UCSRnB

Bit

7

6

5

4

3

2

-

1

-

0

-

RXCIEn

R/W

0

TXCIEn

R/W

0

UDRIE

R/W

0

RXENn

R/W

0

TXENn

R/W

0

UCSRnB

Read/Write

Initial Value

R

1

R

1

R

0

• Bit 7 - RXCIEn: RX Complete Interrupt Enable

Writing this bit to one enables interrupt on the RXCn Flag. A USART Receive Complete interrupt

will be generated only if the RXCIEn bit is written to one, the Global Interrupt Flag in SREG is

written to one and the RXCn bit in UCSRnA is set.

• Bit 6 - TXCIEn: TX Complete Interrupt Enable

Writing this bit to one enables interrupt on the TXCn Flag. A USART Transmit Complete interrupt

will be generated only if the TXCIEn bit is written to one, the Global Interrupt Flag in SREG is

written to one and the TXCn bit in UCSRnA is set.

• Bit 5 - UDRIE: USART Data Register Empty Interrupt Enable

Writing this bit to one enables interrupt on the UDREn Flag. A Data Register Empty interrupt will

be generated only if the UDRIE bit is written to one, the Global Interrupt Flag in SREG is written

to one and the UDREn bit in UCSRnA is set.

• Bit 4 - RXENn: Receiver Enable

Writing this bit to one enables the USART Receiver in MSPIM mode. The Receiver will override

normal port operation for the RxDn pin when enabled. Disabling the Receiver will flush the

receive buffer. Only enabling the receiver in MSPI mode (i.e. setting RXENn=1 and TXENn=0)

has no meaning since it is the transmitter that controls the transfer clock and since only master

mode is supported.

• Bit 3 - TXENn: Transmitter Enable

Writing this bit to one enables the USART Transmitter. The Transmitter will override normal port

operation for the TxDn pin when enabled. The disabling of the Transmitter (writing TXENn to

zero) will not become effective until ongoing and pending transmissions are completed, i.e.,

when the Transmit Shift Register and Transmit Buffer Register do not contain data to be trans-

mitted. When disabled, the Transmitter will no longer override the TxDn port.

• Bit 2:0 - Reserved Bits in MSPI mode

When in MSPI mode, these bits are reserved for future use. For compatibility with future devices,

these bits must be written to zero when UCSRnB is written.

19.6.4

USART MSPIM Control and Status Register n C - UCSRnC

Bit

7

6

5

4

3

2

1

0

UMSELn1

UMSELn0

-

UDORDn

UCPHAn

UCPOLn

UCSRnC

Read/Write

Initial Value

R/W

0

R/W

0

R

0

R

0

R

0

R/W

1

R/W

1

R/W

0

220

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 7:6 - UMSELn1:0: USART Mode Select

These bits select the mode of operation of the USART as shown in Table 19-3. See “USART

Control and Status Register n C – UCSRnC” on page 207 for full description of the normal

USART operation. The MSPIM is enabled when both UMSELn bits are set to one. The

UDORDn, UCPHAn, and UCPOLn can be set in the same write operation where the MSPIM is

enabled.

Table 19-3. UMSELn Bits Settings

UMSELn1

UMSELn0

Mode

0

1

0

Asynchronous USART

Synchronous USART

(Reserved)

1

0

1

Master SPI (MSPIM)

• Bit 5:3 - Reserved Bits in MSPI mode

When in MSPI mode, these bits are reserved for future use. For compatibility with future devices,

these bits must be written to zero when UCSRnC is written.

• Bit 2 - UDORDn: Data Order

When set to one the LSB of the data word is transmitted first. When set to zero the MSB of the

data word is transmitted first. Refer to the Frame Formats section page 4 for details.

• Bit 1 - UCPHAn: Clock Phase

The UCPHAn bit setting determine if data is sampled on the leasing edge (first) or tailing (last)

edge of XCKn. Refer to the SPI Data Modes and Timing section page 4 for details.

• Bit 0 - UCPOLn: Clock Polarity

The UCPOLn bit sets the polarity of the XCKn clock. The combination of the UCPOLn and

UCPHAn bit settings determine the timing of the data transfer. Refer to the SPI Data Modes and

Timing section page 4 for details.

19.6.5

USART MSPIM Baud Rate Registers - UBRRnL and UBRRnH

The function and bit description of the baud rate registers in MSPI mode is identical to normal

USART operation. See “USART Baud Rate Registers – UBRRLn and UBRRHn” on page 209.

19.7 AVR USART MSPIM vs. AVR SPI

The USART in MSPIM mode is fully compatible with the AVR SPI regarding:

• Master mode timing diagram.

• The UCPOLn bit functionality is identical to the SPI CPOL bit.

• The UCPHAn bit functionality is identical to the SPI CPHA bit.

• The UDORDn bit functionality is identical to the SPI DORD bit.

However, since the USART in MSPIM mode reuses the USART resources, the use of the

USART in MSPIM mode is somewhat different compared to the SPI. In addition to differences of

the control register bits, and that only master operation is supported by the USART in MSPIM

mode, the following features differ between the two modules:

221

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• The USART in MSPIM mode includes (double) buffering of the transmitter. The SPI has no

buffer.

• The USART in MSPIM mode receiver includes an additional buffer level.

• The SPI WCOL (Write Collision) bit is not included in USART in MSPIM mode.

• The SPI double speed mode (SPI2X) bit is not included. However, the same effect is

achieved by setting UBRRn accordingly.

• Interrupt timing is not compatible.

• Pin control differs due to the master only operation of the USART in MSPIM mode.

A comparison of the USART in MSPIM mode and the SPI pins is shown in Table 19-4 on page

222.

Table 19-4. Comparison of USART in MSPIM mode and SPI pins.

USART_MSPIM

TxDn

SPI

MOSI

MISO

SCK

SS

Comment

Master Out only

RxDn

Master In only

XCKn

(Functionally identical)

Not supported by USART in MSPIM

(N/A)

222

7766E–AVR–04/10

ATmega16U4/ATmega32U4

20. 2-wire Serial Interface

20.1 Features

• Simple Yet Powerful and Flexible Communication Interface, only two Bus Lines Needed

• Both Master and Slave Operation Supported

• Device can Operate as Transmitter or Receiver

• 7-bit Address Space Allows up to 128 Different Slave Addresses

• Multi-master Arbitration Support

• Up to 400 kHz Data Transfer Speed

• Slew-rate Limited Output Drivers

• Noise Suppression Circuitry Rejects Spikes on Bus Lines

• Fully Programmable Slave Address with General Call Support

• Address Recognition Causes Wake-up When AVR is in Sleep Mode

20.2 2-wire Serial Interface Bus Definition

The 2-wire Serial Interface (TWI) is ideally suited for typical microcontroller applications. The

TWI protocol allows the systems designer to interconnect up to 128 different devices using only

two bi-directional bus lines, one for clock (SCL) and one for data (SDA). The only external hard-

ware needed to implement the bus is a single pull-up resistor for each of the TWI bus lines. All

devices connected to the bus have individual addresses, and mechanisms for resolving bus

contention are inherent in the TWI protocol.

Figure 20-1. TWI Bus Interconnection

V_CC

Device 1

Device 3

R1

R2

Device 2

Device n

........

SDA

SCL

20.2.1

TWI Terminology

The following definitions are frequently encountered in this section.

Table 20-1. TWI Terminology

Term

Description

The device that initiates and terminates a transmission. The Master also

generates the SCL clock.

Master

Slave

The device addressed by a Master.

The device placing data on the bus.

The device reading data from the bus.

Transmitter

Receiver

223

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The Power Reduction TWI bit, PRTWI bit in “Power Reduction Register 0 - PRR0” on page 46

must be written to zero to enable the 2-wire Serial Interface.

20.2.2

Electrical Interconnection

As depicted in Figure 20-1, both bus lines are connected to the positive supply voltage through

pull-up resistors. The bus drivers of all TWI-compliant devices are open-drain or open-collector.

This implements a wired-AND function which is essential to the operation of the interface. A low

level on a TWI bus line is generated when one or more TWI devices output a zero. A high level

is output when all TWI devices trim-state their outputs, allowing the pull-up resistors to pull the

line high. Note that all AVR devices connected to the TWI bus must be powered in order to allow

any bus operation.

The number of devices that can be connected to the bus is only limited by the bus capacitance

limit of 400 pF and the 7-bit slave address space. A detailed specification of the electrical char-

acteristics of the TWI is given in “SPI Timing Characteristics” on page 383. Two different sets of

specifications are presented there, one relevant for bus speeds below 100 kHz, and one valid for

bus speeds up to 400 kHz.

20.3 Data Transfer and Frame Format

20.3.1

Transferring Bits

Each data bit transferred on the TWI bus is accompanied by a pulse on the clock line. The level

of the data line must be stable when the clock line is high. The only exception to this rule is for

generating start and stop conditions.

Figure 20-2. Data Validity

SDA

SCL

Data Stable

Data Change

20.3.2

START and STOP Conditions

The Master initiates and terminates a data transmission. The transmission is initiated when the

Master issues a START condition on the bus, and it is terminated when the Master issues a

STOP condition. Between a START and a STOP condition, the bus is considered busy, and no

other master should try to seize control of the bus. A special case occurs when a new START

condition is issued between a START and STOP condition. This is referred to as a REPEATED

START condition, and is used when the Master wishes to initiate a new transfer without relin-

quishing control of the bus. After a REPEATED START, the bus is considered busy until the next

STOP. This is identical to the START behavior, and therefore START is used to describe both

START and REPEATED START for the remainder of this datasheet, unless otherwise noted. As

224

7766E–AVR–04/10

ATmega16U4/ATmega32U4

depicted below, START and STOP conditions are signalled by changing the level of the SDA

line when the SCL line is high.

Figure 20-3. START, REPEATED START and STOP conditions

SDA

SCL

START

STOP START

REPEATED START

STOP

20.3.3

Address Packet Format

All address packets transmitted on the TWI bus are 9 bits long, consisting of 7 address bits, one

READ/WRITE control bit and an acknowledge bit. If the READ/WRITE bit is set, a read opera-

tion is to be performed, otherwise a write operation should be performed. When a Slave

recognizes that it is being addressed, it should acknowledge by pulling SDA low in the ninth SCL

(ACK) cycle. If the addressed Slave is busy, or for some other reason can not service the Mas-

ter’s request, the SDA line should be left high in the ACK clock cycle. The Master can then

transmit a STOP condition, or a REPEATED START condition to initiate a new transmission. An

address packet consisting of a slave address and a READ or a WRITE bit is called SLA+R or

SLA+W, respectively.

The MSB of the address byte is transmitted first. Slave addresses can freely be allocated by the

designer, but the address 0000 000 is reserved for a general call.

When a general call is issued, all slaves should respond by pulling the SDA line low in the ACK

cycle. A general call is used when a Master wishes to transmit the same message to several

slaves in the system. When the general call address followed by a Write bit is transmitted on the

bus, all slaves set up to acknowledge the general call will pull the SDA line low in the ack cycle.

The following data packets will then be received by all the slaves that acknowledged the general

call. Note that transmitting the general call address followed by a Read bit is meaningless, as

this would cause contention if several slaves started transmitting different data.

All addresses of the format 1111 xxx should be reserved for future purposes.

Figure 20-4. Address Packet Format

Addr MSB

Addr LSB

R/W

ACK

SDA

SCL

1

2

7

8

9

START

225

7766E–AVR–04/10

ATmega16U4/ATmega32U4

20.3.4

Data Packet Format

All data packets transmitted on the TWI bus are nine bits long, consisting of one data byte and

an acknowledge bit. During a data transfer, the Master generates the clock and the START and

STOP conditions, while the Receiver is responsible for acknowledging the reception. An

Acknowledge (ACK) is signalled by the Receiver pulling the SDA line low during the ninth SCL

cycle. If the Receiver leaves the SDA line high, a NACK is signalled. When the Receiver has

received the last byte, or for some reason cannot receive any more bytes, it should inform the

Transmitter by sending a NACK after the final byte. The MSB of the data byte is transmitted first.

Figure 20-5. Data Packet Format

Data MSB

Data LSB

ACK

Aggregate

SDA

SDA from

Transmitter

SDA from

Receiver

SCL from

Master

1

2

7

8

9

STOP, REPEATED

START or Next

Data Byte

SLA+R/W

Data Byte

20.3.5

Combining Address and Data Packets into a Transmission

A transmission basically consists of a START condition, a SLA+R/W, one or more data packets

and a STOP condition. An empty message, consisting of a START followed by a STOP condi-

tion, is illegal. Note that the Wired-ANDing of the SCL line can be used to implement

handshaking between the Master and the Slave. The Slave can extend the SCL low period by

pulling the SCL line low. This is useful if the clock speed set up by the Master is too fast for the

Slave, or the Slave needs extra time for processing between the data transmissions. The Slave

extending the SCL low period will not affect the SCL high period, which is determined by the

Master. As a consequence, the Slave can reduce the TWI data transfer speed by prolonging the

SCL duty cycle.

Figure 20-6 shows a typical data transmission. Note that several data bytes can be transmitted

between the SLA+R/W and the STOP condition, depending on the software protocol imple-

mented by the application software.

Figure 20-6. Typical Data Transmission

Addr MSB

Addr LSB R/W

ACK

Data MSB

Data LSB ACK

SDA

SCL

1

2

7

8

9

1

2

7

8

9

START

SLA+R/W

Data Byte

STOP

226

7766E–AVR–04/10

ATmega16U4/ATmega32U4

20.4 Multi-master Bus Systems, Arbitration and Synchronization

The TWI protocol allows bus systems with several masters. Special concerns have been taken

in order to ensure that transmissions will proceed as normal, even if two or more masters initiate

a transmission at the same time. Two problems arise in multi-master systems:

• An algorithm must be implemented allowing only one of the masters to complete the

transmission. All other masters should cease transmission when they discover that they have

lost the selection process. This selection process is called arbitration. When a contending

master discovers that it has lost the arbitration process, it should immediately switch to Slave

mode to check whether it is being addressed by the winning master. The fact that multiple

masters have started transmission at the same time should not be detectable to the slaves,

i.e. the data being transferred on the bus must not be corrupted.

• Different masters may use different SCL frequencies. A scheme must be devised to

synchronize the serial clocks from all masters, in order to let the transmission proceed in a

lockstep fashion. This will facilitate the arbitration process.

The wired-ANDing of the bus lines is used to solve both these problems. The serial clocks from

all masters will be wired-ANDed, yielding a combined clock with a high period equal to the one

from the Master with the shortest high period. The low period of the combined clock is equal to

the low period of the Master with the longest low period. Note that all masters listen to the SCL

line, effectively starting to count their SCL high and low time-out periods when the combined

SCL line goes high or low, respectively.

Figure 20-7. SCL Synchronization Between Multiple Masters

TA _low

TA _high

SCL from

Master A

SCL from

Master B

SCL Bus

Line

TB_low

TB_high

Masters Start

Counting Low Period

Counting High Period

Arbitration is carried out by all masters continuously monitoring the SDA line after outputting

data. If the value read from the SDA line does not match the value the Master had output, it has

lost the arbitration. Note that a Master can only lose arbitration when it outputs a high SDA value

while another Master outputs a low value. The losing Master should immediately go to Slave

mode, checking if it is being addressed by the winning Master. The SDA line should be left high,

but losing masters are allowed to generate a clock signal until the end of the current data or

address packet. Arbitration will continue until only one Master remains, and this may take many

227

7766E–AVR–04/10

ATmega16U4/ATmega32U4

bits. If several masters are trying to address the same Slave, arbitration will continue into the

data packet.

Figure 20-8. Arbitration Between Two Masters

START

Master A Loses

Arbitration, SDA_ASDA

SDA from

Master A

SDA from

Master B

SDA Line

Synchronized

SCL Line

Note that arbitration is not allowed between:

• A REPEATED START condition and a data bit.

• A STOP condition and a data bit.

• A REPEATED START and a STOP condition.

It is the user software’s responsibility to ensure that these illegal arbitration conditions never

occur. This implies that in multi-master systems, all data transfers must use the same composi-

tion of SLA+R/W and data packets. In other words: All transmissions must contain the same

number of data packets, otherwise the result of the arbitration is undefined.

20.5 Overview of the TWI Module

The TWI module is comprised of several submodules, as shown in Figure 20-9. All registers

drawn in a thick line are accessible through the AVR data bus.

228

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 20-9. Overview of the TWI Module

SCL

SDA

Spike

Filter

Spike

Filter

Slew-rate

Control

Slew-rate

Control

Bus Interface Unit

Bit Rate Generator

START / STOP

Spike Suppression

Prescaler

Control

Address/Data Shift

Register (TWDR)

Bit Rate Register

(TWBR)

Arbitration detection

Ack

Address Match Unit

Control Unit

Address Register

(TWAR)

Status Register

(TWSR)

Control Register

(TWCR)

State Machine and

Status control

Address Comparator

20.5.1

20.5.2

SCL and SDA Pins

These pins interface the AVR TWI with the rest of the MCU system. The output drivers contain a

slew-rate limiter in order to conform to the TWI specification. The input stages contain a spike

suppression unit removing spikes shorter than 50 ns. Note that the internal pull-ups in the AVR

pads can be enabled by setting the PORT bits corresponding to the SCL and SDA pins, as

explained in the I/O Port section. The internal pull-ups can in some systems eliminate the need

for external ones.

Bit Rate Generator Unit

This unit controls the period of SCL when operating in a Master mode. The SCL period is con-

trolled by settings in the TWI Bit Rate Register (TWBR) and the Prescaler bits in the TWI Status

Register (TWSR). Slave operation does not depend on Bit Rate or Prescaler settings, but the

CPU clock frequency in the Slave must be at least 16 times higher than the SCL frequency. Note

that slaves may prolong the SCL low period, thereby reducing the average TWI bus clock

period. The SCL frequency is generated according to the following equation:

229

7766E–AVR–04/10

ATmega16U4/ATmega32U4

CPU Clock frequency

SCL frequency = -----------------------------------------------------------

TWPS

16 + 2(TWBR) ⋅ 4

• TWBR = Value of the TWI Bit Rate Register.

• TWPS = Value of the prescaler bits in the TWI Status Register.

Note:

TWBR should be 10 or higher if the TWI operates in Master mode. If TWBR is lower than 10, the

Master may produce an incorrect output on SDA and SCL for the reminder of the byte. The prob-

lem occurs when operating the TWI in Master mode, sending Start + SLA + R/W to a Slave (a

Slave does not need to be connected to the bus for the condition to happen).

20.5.3

Bus Interface Unit

This unit contains the Data and Address Shift Register (TWDR), a START/STOP Controller and

Arbitration detection hardware. The TWDR contains the address or data bytes to be transmitted,

or the address or data bytes received. In addition to the 8-bit TWDR, the Bus Interface Unit also

contains a register containing the (N)ACK bit to be transmitted or received. This (N)ACK Regis-

ter is not directly accessible by the application software. However, when receiving, it can be set

or cleared by manipulating the TWI Control Register (TWCR). When in Transmitter mode, the

value of the received (N)ACK bit can be determined by the value in the TWSR.

The START/STOP Controller is responsible for generation and detection of START, REPEATED

START, and STOP conditions. The START/STOP controller is able to detect START and STOP

conditions even when the AVR MCU is in one of the sleep modes, enabling the MCU to wake up

if addressed by a Master.

If the TWI has initiated a transmission as Master, the Arbitration Detection hardware continu-

ously monitors the transmission trying to determine if arbitration is in process. If the TWI has lost

an arbitration, the Control Unit is informed. Correct action can then be taken and appropriate

status codes generated.

20.5.4

Address Match Unit

The Address Match unit checks if received address bytes match the seven-bit address in the

TWI Address Register (TWAR). If the TWI General Call Recognition Enable (TWGCE) bit in the

TWAR is written to one, all incoming address bits will also be compared against the General Call

address. Upon an address match, the Control Unit is informed, allowing correct action to be

taken. The TWI may or may not acknowledge its address, depending on settings in the TWCR.

The Address Match unit is able to compare addresses even when the AVR MCU is in sleep

mode, enabling the MCU to wake up if addressed by a Master. If another interrupt (e.g., INT0)

occurs during TWI Power-down address match and wakes up the CPU, the TWI aborts opera-

tion and return to it’s idle state. If this cause any problems, ensure that TWI Address Match is the

only enabled interrupt when entering Power-down.

20.5.5

Control Unit

The Control unit monitors the TWI bus and generates responses corresponding to settings in the

TWI Control Register (TWCR). When an event requiring the attention of the application occurs

on the TWI bus, the TWI Interrupt Flag (TWINT) is asserted. In the next clock cycle, the TWI Sta-

tus Register (TWSR) is updated with a status code identifying the event. The TWSR only

contains relevant status information when the TWI Interrupt Flag is asserted. At all other times,

the TWSR contains a special status code indicating that no relevant status information is avail-

able. As long as the TWINT Flag is set, the SCL line is held low. This allows the application

software to complete its tasks before allowing the TWI transmission to continue.

230

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The TWINT Flag is set in the following situations:

• After the TWI has transmitted a START/REPEATED START condition.

• After the TWI has transmitted SLA+R/W.

• After the TWI has transmitted an address byte.

• After the TWI has lost arbitration.

• After the TWI has been addressed by own slave address or general call.

• After the TWI has received a data byte.

• After a STOP or REPEATED START has been received while still addressed as a Slave.

• When a bus error has occurred due to an illegal START or STOP condition.

20.6 TWI Register Description

20.6.1

TWI Bit Rate Register – TWBR

Bit

7

6

TWBR6

R/W

0

5

TWBR5

R/W

0

4

TWBR4

R/W

0

3

TWBR3

R/W

0

2

TWBR2

R/W

0

1

TWBR1

R/W

0

TWBR0

R/W

0

TWBR7

R/W

0

TWBR

Read/Write

Initial Value

• Bits 7..0 – TWI Bit Rate Register

TWBR selects the division factor for the bit rate generator. The bit rate generator is a frequency

divider which generates the SCL clock frequency in the Master modes. See “Bit Rate Generator

Unit” on page 229 for calculating bit rates.

20.6.2

TWI Control Register – TWCR

Bit

7

6

TWEA

R/W

0

5

TWSTA

R/W

0

4

TWSTO

R/W

0

3

2

TWEN

R/W

0

1

–

0

TWIE

R/W

0

TWINT

R/W

0

TWWC

TWCR

Read/Write

Initial Value

R

0

R

0

The TWCR is used to control the operation of the TWI. It is used to enable the TWI, to initiate a

Master access by applying a START condition to the bus, to generate a Receiver acknowledge,

to generate a stop condition, and to control halting of the bus while the data to be written to the

bus are written to the TWDR. It also indicates a write collision if data is attempted written to

TWDR while the register is inaccessible.

• Bit 7 – TWINT: TWI Interrupt Flag

This bit is set by hardware when the TWI has finished its current job and expects application

software response. If the I-bit in SREG and TWIE in TWCR are set, the MCU will jump to the

TWI Interrupt Vector. While the TWINT Flag is set, the SCL low period is stretched. The TWINT

Flag must be cleared by software by writing a logic one to it. Note that this flag is not automati-

cally cleared by hardware when executing the interrupt routine. Also note that clearing this flag

starts the operation of the TWI, so all accesses to the TWI Address Register (TWAR), TWI Sta-

tus Register (TWSR), and TWI Data Register (TWDR) must be complete before clearing this

flag.

• Bit 6 – TWEA: TWI Enable Acknowledge Bit

The TWEA bit controls the generation of the acknowledge pulse. If the TWEA bit is written to

one, the ACK pulse is generated on the TWI bus if the following conditions are met:

231

7766E–AVR–04/10

ATmega16U4/ATmega32U4

1. The device’s own slave address has been received.

2. A general call has been received, while the TWGCE bit in the TWAR is set.

3. A data byte has been received in Master Receiver or Slave Receiver mode.

By writing the TWEA bit to zero, the device can be virtually disconnected from the 2-wire Serial

Bus temporarily. Address recognition can then be resumed by writing the TWEA bit to one

again.

• Bit 5 – TWSTA: TWI START Condition Bit

The application writes the TWSTA bit to one when it desires to become a Master on the 2-wire

Serial Bus. The TWI hardware checks if the bus is available, and generates a START condition

on the bus if it is free. However, if the bus is not free, the TWI waits until a STOP condition is

detected, and then generates a new START condition to claim the bus Master status. TWSTA

must be cleared by software when the START condition has been transmitted.

• Bit 4 – TWSTO: TWI STOP Condition Bit

Writing the TWSTO bit to one in Master mode will generate a STOP condition on the 2-wire

Serial Bus. When the STOP condition is executed on the bus, the TWSTO bit is cleared auto-

matically. In Slave mode, setting the TWSTO bit can be used to recover from an error condition.

This will not generate a STOP condition, but the TWI returns to a well-defined unaddressed

Slave mode and releases the SCL and SDA lines to a high impedance state.

• Bit 3 – TWWC: TWI Write Collision Flag

The TWWC bit is set when attempting to write to the TWI Data Register – TWDR when TWINT is

low. This flag is cleared by writing the TWDR Register when TWINT is high.

• Bit 2 – TWEN: TWI Enable Bit

The TWEN bit enables TWI operation and activates the TWI interface. When TWEN is written to

one, the TWI takes control over the I/O pins connected to the SCL and SDA pins, enabling the

slew-rate limiters and spike filters. If this bit is written to zero, the TWI is switched off and all TWI

transmissions are terminated, regardless of any ongoing operation.

• Bit 1 – Res: Reserved Bit

This bit is a reserved bit and will always read as zero.

• Bit 0 – TWIE: TWI Interrupt Enable

When this bit is written to one, and the I-bit in SREG is set, the TWI interrupt request will be acti-

vated for as long as the TWINT Flag is high.

20.6.3

TWI Status Register – TWSR

Bit

7

6

TWS6

R

5

TWS5

R

4

TWS4

R

3

TWS3

R

2

–

1

TWPS1

R/W

0

TWPS0

R/W

0

TWS7

TWSR

Read/Write

Initial Value

R

1

R

0

1

• Bits 7..3 – TWS: TWI Status

These 5 bits reflect the status of the TWI logic and the 2-wire Serial Bus. The different status

codes are described later in this section. Note that the value read from TWSR contains both the

5-bit status value and the 2-bit prescaler value. The application designer should mask the pres-

232

7766E–AVR–04/10

ATmega16U4/ATmega32U4

caler bits to zero when checking the Status bits. This makes status checking independent of

prescaler setting. This approach is used in this datasheet, unless otherwise noted.

• Bit 2 – Res: Reserved Bit

This bit is reserved and will always read as zero.

• Bits 1..0 – TWPS: TWI Prescaler Bits

These bits can be read and written, and control the bit rate prescaler.

Table 20-2. TWI Bit Rate Prescaler

TWPS1

TWPS0

Prescaler Value

0

1

0

1

0

1

4

16

64

To calculate bit rates, see “Bit Rate Generator Unit” on page 229. The value of TWPS1..0 is

used in the equation.

20.6.4

TWI Data Register – TWDR

Bit

7

TWD7

R/W

1

6

TWD6

R/W

1

5

TWD5

R/W

1

4

TWD4

R/W

1

3

TWD3

R/W

1

2

TWD2

R/W

1

TWD1

R/W

1

0

TWD0

R/W

1

TWDR

Read/Write

Initial Value

In Transmit mode, TWDR contains the next byte to be transmitted. In Receive mode, the TWDR

contains the last byte received. It is writable while the TWI is not in the process of shifting a byte.

This occurs when the TWI Interrupt Flag (TWINT) is set by hardware. Note that the Data Regis-

ter cannot be initialized by the user before the first interrupt occurs. The data in TWDR remains

stable as long as TWINT is set. While data is shifted out, data on the bus is simultaneously

shifted in. TWDR always contains the last byte present on the bus, except after a wake up from

a sleep mode by the TWI interrupt. In this case, the contents of TWDR is undefined. In the case

of a lost bus arbitration, no data is lost in the transition from Master to Slave. Handling of the

ACK bit is controlled automatically by the TWI logic, the CPU cannot access the ACK bit directly.

• Bits 7..0 – TWD: TWI Data Register

These eight bits constitute the next data byte to be transmitted, or the latest data byte received

on the 2-wire Serial Bus.

20.6.5

TWI (Slave) Address Register – TWAR

Bit

7

6

5

TWA4

R/W

1

4

TWA3

R/W

1

3

TWA2

R/W

1

2

TWA1

R/W

1

TWA0

R/W

1

0

TWGCE

R/W

0

TWA6

R/W

1

TWA5

R/W

1

TWAR

Read/Write

Initial Value

The TWAR should be loaded with the 7-bit Slave address (in the seven most significant bits of

TWAR) to which the TWI will respond when programmed as a Slave Transmitter or Receiver,

and not needed in the Master modes. In multi master systems, TWAR must be set in masters

which can be addressed as Slaves by other Masters.

233

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The LSB of TWAR is used to enable recognition of the general call address (0x00). There is an

associated address comparator that looks for the slave address (or general call address if

enabled) in the received serial address. If a match is found, an interrupt request is generated.

• Bits 7..1 – TWA: TWI (Slave) Address Register

These seven bits constitute the slave address of the TWI unit.

• Bit 0 – TWGCE: TWI General Call Recognition Enable Bit

If set, this bit enables the recognition of a General Call given over the 2-wire Serial Bus.

20.6.6

TWI (Slave) Address Mask Register – TWAMR

Bit

7

6

5

4

TWAM[6:0]

R/W

3

2

1

0

–

TWAMR

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R

0

• Bits 7..1 – TWAM: TWI Address Mask

The TWAMR can be loaded with a 7-bit Slave Address mask. Each of the bits in TWAMR can

mask (disable) the corresponding address bit in the TWI Address Register (TWAR). If the mask

bit is set to one then the address match logic ignores the compare between the incoming

address bit and the corresponding bit in TWAR. Figure 20-10 shows the address match logic in

detail.

Figure 20-10. TWI Address Match Logic, Block Diagram

TWAR0

Address

Match

Address

Bit 0

TWAMR0

Address Bit Comparator 0

Address Bit Comparator 6..1

• Bit 0 – Res: Reserved Bit

This bit is reserved and will always read as zero.

20.7 Using the TWI

The AVR TWI is byte-oriented and interrupt based. Interrupts are issued after all bus events, like

reception of a byte or transmission of a START condition. Because the TWI is interrupt-based,

the application software is free to carry on other operations during a TWI byte transfer. Note that

the TWI Interrupt Enable (TWIE) bit in TWCR together with the Global Interrupt Enable bit in

SREG allow the application to decide whether or not assertion of the TWINT Flag should gener-

ate an interrupt request. If the TWIE bit is cleared, the application must poll the TWINT Flag in

order to detect actions on the TWI bus.

When the TWINT Flag is asserted, the TWI has finished an operation and awaits application

response. In this case, the TWI Status Register (TWSR) contains a value indicating the current

234

7766E–AVR–04/10

ATmega16U4/ATmega32U4

state of the TWI bus. The application software can then decide how the TWI should behave in

the next TWI bus cycle by manipulating the TWCR and TWDR Registers.

Figure 20-11 is a simple example of how the application can interface to the TWI hardware. In

this example, a Master wishes to transmit a single data byte to a Slave. This description is quite

abstract, a more detailed explanation follows later in this section. A simple code example imple-

menting the desired behavior is also presented.

Figure 20-11. Interfacing the Application to the TWI in a Typical Transmission

3. Check TWSR to see if START was

sent. Application loads SLA+W into

TWDR, and loads appropriate control

signals into TWCR, makin sure that

TWINT is written to one,

5. Check TWSR to see if SLA+W was

sent and ACK received.

Application loads data into TWDR, and

loads appropriate control signals into

TWCR, making sure that TWINT is

written to one

1. Application

writes to TWCR to

initiate

transmission of

START

7. Check TWSR to see if data was sent

and ACK received.

Application loads appropriate control

signals to send STOP into TWCR,

making sure that TWINT is written to one

and TWSTA is written to zero.

TWI bus START

SLA+W

A

Data

A

STOP

Indicates

TWINT set

4. TWINT set.

Status code indicates

SLA+W sent, ACK

received

2. TWINT set.

Status code indicates

START condition sent

6. TWINT set.

Status code indicates

data sent, ACK received

1. The first step in a TWI transmission is to transmit a START condition. This is done by

writing a specific value into TWCR, instructing the TWI hardware to transmit a START

condition. Which value to write is described later on. However, it is important that the

TWINT bit is set in the value written. Writing a one to TWINT clears the flag. The TWI

will not start any operation as long as the TWINT bit in TWCR is set. Immediately after

the application has cleared TWINT, the TWI will initiate transmission of the START

condition.

2. When the START condition has been transmitted, the TWINT Flag in TWCR is set, and

TWSR is updated with a status code indicating that the START condition has success-

fully been sent.

3. The application software should now examine the value of TWSR, to make sure that the

START condition was successfully transmitted. If TWSR indicates otherwise, the appli-

cation software might take some special action, like calling an error routine. Assuming

that the status code is as expected, the application must load SLA+W into TWDR.

Remember that TWDR is used both for address and data. After TWDR has been

loaded with the desired SLA+W, a specific value must be written to TWCR, instructing

the TWI hardware to transmit the SLA+W present in TWDR. Which value to write is

described later on. However, it is important that the TWINT bit is set in the value written.

Writing a one to TWINT clears the flag. The TWI will not start any operation as long as

the TWINT bit in TWCR is set. Immediately after the application has cleared TWINT,

the TWI will initiate transmission of the address packet.

4. When the address packet has been transmitted, the TWINT Flag in TWCR is set, and

TWSR is updated with a status code indicating that the address packet has success-

fully been sent. The status code will also reflect whether a Slave acknowledged the

packet or not.

235

7766E–AVR–04/10

ATmega16U4/ATmega32U4

5. The application software should now examine the value of TWSR, to make sure that the

address packet was successfully transmitted, and that the value of the ACK bit was as

expected. If TWSR indicates otherwise, the application software might take some spe-

cial action, like calling an error routine. Assuming that the status code is as expected,

the application must load a data packet into TWDR. Subsequently, a specific value

must be written to TWCR, instructing the TWI hardware to transmit the data packet

present in TWDR. Which value to write is described later on. However, it is important

that the TWINT bit is set in the value written. Writing a one to TWINT clears the flag.

The TWI will not start any operation as long as the TWINT bit in TWCR is set. Immedi-

ately after the application has cleared TWINT, the TWI will initiate transmission of the

data packet.

6. When the data packet has been transmitted, the TWINT Flag in TWCR is set, and

TWSR is updated with a status code indicating that the data packet has successfully

been sent. The status code will also reflect whether a Slave acknowledged the packet

or not.

7. The application software should now examine the value of TWSR, to make sure that the

data packet was successfully transmitted, and that the value of the ACK bit was as

expected. If TWSR indicates otherwise, the application software might take some spe-

cial action, like calling an error routine. Assuming that the status code is as expected,

the application must write a specific value to TWCR, instructing the TWI hardware to

transmit a STOP condition. Which value to write is described later on. However, it is

important that the TWINT bit is set in the value written. Writing a one to TWINT clears

the flag. The TWI will not start any operation as long as the TWINT bit in TWCR is set.

Immediately after the application has cleared TWINT, the TWI will initiate transmission

of the STOP condition. Note that TWINT is NOT set after a STOP condition has been

sent.

Even though this example is simple, it shows the principles involved in all TWI transmissions.

These can be summarized as follows:

• When the TWI has finished an operation and expects application response, the TWINT Flag

is set. The SCL line is pulled low until TWINT is cleared.

• When the TWINT Flag is set, the user must update all TWI Registers with the value relevant

for the next TWI bus cycle. As an example, TWDR must be loaded with the value to be

transmitted in the next bus cycle.

• After all TWI Register updates and other pending application software tasks have been

completed, TWCR is written. When writing TWCR, the TWINT bit should be set. Writing a

one to TWINT clears the flag. The TWI will then commence executing whatever operation

was specified by the TWCR setting.

In the following an assembly and C implementation of the example is given. Note that the code

below assumes that several definitions have been made, for example by using include-files.

236

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Assembly Code Example

C Example

TWCR = (1<<TWINT)|(1<<TWSTA)|

Comments

ldi r16,

(1<<TWINT)|(1<<TWSTA)|

(1<<TWEN)

1

2

Send START condition

(1<<TWEN)

out TWCR, r16

wait1:

while (!(TWCR & (1<<TWINT)))

Wait for TWINT Flag set. This

indicates that the START

condition has been transmitted

in

r16,TWCR

;

sbrs r16,TWINT

rjmp wait1

in

r16,TWSR

if ((TWSR & 0xF8) != START)

Check value of TWI Status

Register. Mask prescaler bits. If

status different from START go to

ERROR

andi r16, 0xF8

cpi r16, START

brne ERROR

ERROR();

3

4

5

ldi r16, SLA_W

out TWDR, r16

TWDR = SLA_W;

Load SLA_W into TWDR

Register. Clear TWINT bit in

TWCR to start transmission of

address

TWCR = (1<<TWINT) |

(1<<TWEN);

ldi r16, (1<<TWINT) |

(1<<TWEN)

out TWCR, r16

wait2:

while (!(TWCR & (1<<TWINT)))

Wait for TWINT Flag set. This

indicates that the SLA+W has

been transmitted, and

in

r16,TWCR

;

sbrs r16,TWINT

rjmp wait2

ACK/NACK has been received.

in

r16,TWSR

if ((TWSR & 0xF8) !=

MT_SLA_ACK)

Check value of TWI Status

Register. Mask prescaler bits. If

status different from

andi r16, 0xF8

cpi r16, MT_SLA_ACK

brne ERROR

ERROR();

MT_SLA_ACK go to ERROR

ldi r16, DATA

out TWDR, r16

TWDR = DATA;

TWCR = (1<<TWINT) |

(1<<TWEN);

Load DATA into TWDR Register.

Clear TWINT bit in TWCR to

start transmission of data

ldi r16, (1<<TWINT) |

(1<<TWEN)

out TWCR, r16

wait3:

while (!(TWCR & (1<<TWINT)))

Wait for TWINT Flag set. This

indicates that the DATA has been

transmitted, and ACK/NACK has

been received.

in

r16,TWCR

;

6

7

sbrs r16,TWINT

rjmp wait3

in

r16,TWSR

if ((TWSR & 0xF8) !=

MT_DATA_ACK)

Check value of TWI Status

Register. Mask prescaler bits. If

status different from

andi r16, 0xF8

cpi r16, MT_DATA_ACK

brne ERROR

ERROR();

MT_DATA_ACK go to ERROR

ldi r16,

(1<<TWINT)|(1<<TWEN)|

TWCR = (1<<TWINT)|(1<<TWEN)|

(1<<TWSTO);

Transmit STOP condition

(1<<TWSTO)

out TWCR, r16

237

7766E–AVR–04/10

ATmega16U4/ATmega32U4

20.8 Transmission Modes

The TWI can operate in one of four major modes. These are named Master Transmitter (MT),

Master Receiver (MR), Slave Transmitter (ST) and Slave Receiver (SR). Several of these

modes can be used in the same application. As an example, the TWI can use MT mode to write

data into a TWI EEPROM, MR mode to read the data back from the EEPROM. If other masters

are present in the system, some of these might transmit data to the TWI, and then SR mode

would be used. It is the application software that decides which modes are legal.

The following sections describe each of these modes. Possible status codes are described

along with figures detailing data transmission in each of the modes. These figures contain the

following abbreviations:

S: START condition

Rs: REPEATED START condition

R: Read bit (high level at SDA)

W: Write bit (low level at SDA)

A: Acknowledge bit (low level at SDA)

A: Not acknowledge bit (high level at SDA)

Data: 8-bit data byte

P: STOP condition

SLA: Slave Address

In Figure 20-13 to Figure 20-19, circles are used to indicate that the TWINT Flag is set. The

numbers in the circles show the status code held in TWSR, with the prescaler bits masked to

zero. At these points, actions must be taken by the application to continue or complete the TWI

transfer. The TWI transfer is suspended until the TWINT Flag is cleared by software.

When the TWINT Flag is set, the status code in TWSR is used to determine the appropriate soft-

ware action. For each status code, the required software action and details of the following serial

transfer are given in Table 20-3 to Table 20-6. Note that the prescaler bits are masked to zero in

these tables.

20.8.1

Master Transmitter Mode

In the Master Transmitter mode, a number of data bytes are transmitted to a Slave Receiver

(see Figure 20-12). In order to enter a Master mode, a START condition must be transmitted.

The format of the following address packet determines whether Master Transmitter or Master

Receiver mode is to be entered. If SLA+W is transmitted, MT mode is entered, if SLA+R is trans-

mitted, MR mode is entered. All the status codes mentioned in this section assume that the

prescaler bits are zero or are masked to zero.

238

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 20-12. Data Transfer in Master Transmitter Mode

V_CC

Device 1

MASTER

TRANSMITTER

Device 2

SLAVE

RECEIVER

Device 3

Device n

R1

R2

........

SDA

SCL

A START condition is sent by writing the following value to TWCR:

TWCR

TWINT

TWEA

TWSTA

TWSTO

TWWC

TWEN

–

TWIE

value

1

X

1

0

X

1

0

X

TWEN must be set to enable the 2-wire Serial Interface, TWSTA must be written to one to trans-

mit a START condition and TWINT must be written to one to clear the TWINT Flag. The TWI will

then test the 2-wire Serial Bus and generate a START condition as soon as the bus becomes

free. After a START condition has been transmitted, the TWINT Flag is set by hardware, and the

status code in TWSR will be 0x08 (see Table 20-3). In order to enter MT mode, SLA+W must be

transmitted. This is done by writing SLA+W to TWDR. Thereafter the TWINT bit should be

cleared (by writing it to one) to continue the transfer. This is accomplished by writing the follow-

ing value to TWCR:

TWCR

TWINT

TWEA

TWSTA

TWSTO

TWWC

TWEN

–

TWIE

value

1

X

0

X

1

0

X

When SLA+W have been transmitted and an acknowledgement bit has been received, TWINT is

set again and a number of status codes in TWSR are possible. Possible status codes in Master

mode are 0x18, 0x20, or 0x38. The appropriate action to be taken for each of these status codes

is detailed in Table 20-3.

When SLA+W has been successfully transmitted, a data packet should be transmitted. This is

done by writing the data byte to TWDR. TWDR must only be written when TWINT is high. If not,

the access will be discarded, and the Write Collision bit (TWWC) will be set in the TWCR Regis-

ter. After updating TWDR, the TWINT bit should be cleared (by writing it to one) to continue the

transfer. This is accomplished by writing the following value to TWCR:

TWCR

TWINT

TWEA

TWSTA

TWSTO

TWWC

TWEN

–

TWIE

value

1

X

0

X

1

0

X

This scheme is repeated until the last byte has been sent and the transfer is ended by generat-

ing a STOP condition or a repeated START condition. A STOP condition is generated by writing

the following value to TWCR:

TWCR

TWINT

TWEA

TWSTA

TWSTO

TWWC

TWEN

–

TWIE

value

1

X

0

1

X

1

0

X

A REPEATED START condition is generated by writing the following value to TWCR:

TWCR

TWINT

TWEA

TWSTA

TWSTO

TWWC

TWEN

–

TWIE

value

1

X

1

0

X

1

0

X

239

7766E–AVR–04/10

ATmega16U4/ATmega32U4

After a repeated START condition (state 0x10) the 2-wire Serial Interface can access the same

Slave again, or a new Slave without transmitting a STOP condition. Repeated START enables

the Master to switch between Slaves, Master Transmitter mode and Master Receiver mode with-

out losing control of the bus.

Table 20-3. Status codes for Master Transmitter Mode

Status Code

(TWSR)

Prescaler Bits

are 0

Application Software Response

To/from TWDR To TWCR

STO TWIN

Status of the 2-wire Serial Bus

and 2-wire Serial Interface

Hardware

STA

0

TWE

A

Next Action Taken by TWI Hardware

T

0x08

0x10

A START condition has been

transmitted

Load SLA+W

0

1

X

SLA+W will be transmitted;

ACK or NOT ACK will be received

A repeated START condition

has been transmitted

Load SLA+W or

Load SLA+R

0

1

X

SLA+W will be transmitted;

ACK or NOT ACK will be received

SLA+R will be transmitted;

Logic will switch to Master Receiver mode

0x18

0x20

0x28

0x30

0x38

SLA+W has been transmitted;

ACK has been received

Load data byte or

0

1

X

Data byte will be transmitted and ACK or NOT ACK will

be received

Repeated START will be transmitted

STOP condition will be transmitted and

TWSTO Flag will be reset

No TWDR action or

1

0

1

X

No TWDR action

Load data byte or

1

0

1

0

1

X

STOP condition followed by a START condition will be

transmitted and TWSTO Flag will be reset

SLA+W has been transmitted;

NOT ACK has been received

Data byte will be transmitted and ACK or NOT ACK will

be received

Repeated START will be transmitted

STOP condition will be transmitted and

TWSTO Flag will be reset

No TWDR action or

1

0

1

X

No TWDR action

Load data byte or

1

0

1

0

1

X

STOP condition followed by a START condition will be

transmitted and TWSTO Flag will be reset

Data byte has been transmit-

ted;

ACK has been received

Data byte will be transmitted and ACK or NOT ACK will

be received

Repeated START will be transmitted

STOP condition will be transmitted and

TWSTO Flag will be reset

No TWDR action or

1

0

1

X

No TWDR action

Load data byte or

1

0

1

0

1

X

STOP condition followed by a START condition will be

transmitted and TWSTO Flag will be reset

Data byte has been transmit-

ted;

NOT ACK has been received

Data byte will be transmitted and ACK or NOT ACK will

be received

Repeated START will be transmitted

STOP condition will be transmitted and

TWSTO Flag will be reset

No TWDR action or

1

0

1

X

No TWDR action

1

X

STOP condition followed by a START condition will be

transmitted and TWSTO Flag will be reset

Arbitration lost in SLA+W or

data bytes

No TWDR action or

No TWDR action

0

1

0

1

X

2-wire Serial Bus will be released and not addressed

Slave mode entered

A START condition will be transmitted when the bus

becomes free

240

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 20-13. Formats and States in the Master Transmitter Mode

MT

Successfull

transmission

to a slave

receiver

S

SLA

W

A

DATA

A

P

$08

$18

$28

Next transfer

started with a

repeated start

condition

R_S

SLA

W

R

$10

Not acknowledge

received after the

slave address

A

P

$20

MR

Not acknowledge

received after a data

byte

A

P

$30

Arbitration lost in slave

address or data byte

Other master

continues

Other master

continues

A or A

$38

A

$38

Arbitration lost and

addressed as slave

Other master

continues

To corresponding

states in slave mode

$68 $78 $B0

Any number of data bytes

and their associated acknowledge bits

From master to slave

From slave to master

DATA

A

This number (contained in TWSR) corresponds

to a defined state of the Two-Wire Serial Bus. The

prescaler bits are zero or masked to zero

n

20.8.2

Master Receiver Mode

In the Master Receiver mode, a number of data bytes are received from a Slave Transmitter

(Slave see Figure 20-14). In order to enter a Master mode, a START condition must be transmit-

ted. The format of the following address packet determines whether Master Transmitter or

Master Receiver mode is to be entered. If SLA+W is transmitted, MT mode is entered, if SLA+R

is transmitted, MR mode is entered. All the status codes mentioned in this section assume that

the prescaler bits are zero or are masked to zero.

241

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 20-14. Data Transfer in Master Receiver Mode

V_CC

Device 1

MASTER

RECEIVER

Device 2

SLAVE

TRANSMITTER

Device 3

Device n

R1

R2

........

SDA

SCL

A START condition is sent by writing the following value to TWCR:

TWCR

TWINT

TWEA

TWSTA

TWSTO

TWWC

TWEN

–

TWIE

value

1

X

1

0

X

1

0

X

TWEN must be written to one to enable the 2-wire Serial Interface, TWSTA must be written to

one to transmit a START condition and TWINT must be set to clear the TWINT Flag. The TWI

will then test the 2-wire Serial Bus and generate a START condition as soon as the bus

becomes free. After a START condition has been transmitted, the TWINT Flag is set by hard-

ware, and the status code in TWSR will be 0x08 (See Table 20-3). In order to enter MR mode,

SLA+R must be transmitted. This is done by writing SLA+R to TWDR. Thereafter the TWINT bit

should be cleared (by writing it to one) to continue the transfer. This is accomplished by writing

the following value to TWCR:

TWCR

TWINT

TWEA

TWSTA

TWSTO

TWWC

TWEN

–

TWIE

value

1

X

0

X

1

0

X

When SLA+R have been transmitted and an acknowledgement bit has been received, TWINT is

set again and a number of status codes in TWSR are possible. Possible status codes in Master

mode are 0x38, 0x40, or 0x48. The appropriate action to be taken for each of these status codes

is detailed in Table 20-4. Received data can be read from the TWDR Register when the TWINT

Flag is set high by hardware. This scheme is repeated until the last byte has been received.

After the last byte has been received, the MR should inform the ST by sending a NACK after the

last received data byte. The transfer is ended by generating a STOP condition or a repeated

START condition. A STOP condition is generated by writing the following value to TWCR:

TWCR

TWINT

TWEA

TWSTA

TWSTO

TWWC

TWEN

–

TWIE

value

1

X

0

1

X

1

0

X

A REPEATED START condition is generated by writing the following value to TWCR:

TWCR

TWINT

TWEA

TWSTA

TWSTO

TWWC

TWEN

–

TWIE

value

1

X

1

0

X

1

0

X

After a repeated START condition (state 0x10) the 2-wire Serial Interface can access the same

Slave again, or a new Slave without transmitting a STOP condition. Repeated START enables

242

7766E–AVR–04/10

ATmega16U4/ATmega32U4

the Master to switch between Slaves, Master Transmitter mode and Master Receiver mode with-

out losing control over the bus.

Table 20-4. Status codes for Master Receiver Mode

Status Code

(TWSR)

Prescaler Bits

are 0

Application Software Response

Status of the 2-wire Serial Bus

and 2-wire Serial Interface

Hardware

To TWCR

To/from TWDR

Load SLA+R

STA

0

STO

0

TWIN

T

TWE

A

Next Action Taken by TWI Hardware

0x08

0x10

A START condition has been

transmitted

1

X

SLA+R will be transmitted

ACK or NOT ACK will be received

A repeated START condition

has been transmitted

Load SLA+R or

Load SLA+W

0

1

X

SLA+R will be transmitted

ACK or NOT ACK will be received

SLA+W will be transmitted

Logic will switch to Master Transmitter mode

0x38

0x40

0x48

Arbitration lost in SLA+R or

NOT ACK bit

No TWDR action or

No TWDR action

0

1

0

1

X

2-wire Serial Bus will be released and not addressed

Slave mode will be entered

A START condition will be transmitted when the bus

becomes free

SLA+R has been transmitted;

ACK has been received

No TWDR action or

No TWDR action

0

1

0

1

Data byte will be received and NOT ACK will be

returned

Data byte will be received and ACK will be returned

SLA+R has been transmitted;

NOT ACK has been received

No TWDR action or

1

0

1

X

Repeated START will be transmitted

STOP condition will be transmitted and TWSTO Flag

will be reset

No TWDR action

1

X

STOP condition followed by a START condition will be

transmitted and TWSTO Flag will be reset

0x50

0x58

Data byte has been received;

ACK has been returned

Read data byte or

Read data byte

0

1

0

1

Data byte will be received and NOT ACK will be

returned

Data byte will be received and ACK will be returned

Data byte has been received;

NOT ACK has been returned

Read data byte or

1

0

1

X

Repeated START will be transmitted

STOP condition will be transmitted and TWSTO Flag

will be reset

Read data byte

1

X

STOP condition followed by a START condition will be

transmitted and TWSTO Flag will be reset

243

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 20-15. Formats and States in the Master Receiver Mode

MR

Successfull

reception

S

SLA

R

A

DATA

A

DATA

A

P

from a slave

receiver

$08

$40

$50

$58

Next transfer

started with a

repeated start

condition

R_S

SLA

R

$10

Not acknowledge

received after the

slave address

W

A

P

$48

MT

Arbitration lost in slave

address or data byte

Other master

continues

Other master

continues

A or A

A

$38

A

$38

Arbitration lost and

addressed as slave

Other master

continues

To corresponding

states in slave mode

$68 $78 $B0

Any number of data bytes

and their associated acknowledge bits

From master to slave

From slave to master

DATA

A

This number (contained in TWSR) corresponds

to a defined state of the Two-Wire Serial Bus. The

prescaler bits are zero or masked to zero

n

20.8.3

Slave Receiver Mode

In the Slave Receiver mode, a number of data bytes are received from a Master Transmitter

(see Figure 20-16). All the status codes mentioned in this section assume that the prescaler bits

are zero or are masked to zero.

Figure 20-16. Data transfer in Slave Receiver mode

V_CC

Device 1

SLAVE

RECEIVER

Device 2

MASTER

TRANSMITTER

Device 3

Device n

R1

R2

........

SDA

SCL

To initiate the Slave Receiver mode, TWAR and TWCR must be initialized as follows:

TWAR

TWA6

TWA5

TWA4

TWA3

TWA2

TWA1

TWA0

TWGCE

value

Device’s Own Slave Address

244

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The upper 7 bits are the address to which the 2-wire Serial Interface will respond when

addressed by a Master. If the LSB is set, the TWI will respond to the general call address (0x00),

otherwise it will ignore the general call address.

TWCR

TWINT

TWEA

TWSTA

TWSTO

TWWC

TWEN

–

TWIE

value

0

1

0

1

0

X

TWEN must be written to one to enable the TWI. The TWEA bit must be written to one to enable

the acknowledgement of the device’s own slave address or the general call address. TWSTA

and TWSTO must be written to zero.

When TWAR and TWCR have been initialized, the TWI waits until it is addressed by its own

slave address (or the general call address if enabled) followed by the data direction bit. If the

direction bit is “0” (write), the TWI will operate in SR mode, otherwise ST mode is entered. After

its own slave address and the write bit have been received, the TWINT Flag is set and a valid

status code can be read from TWSR. The status code is used to determine the appropriate soft-

ware action. The appropriate action to be taken for each status code is detailed in Table 20-5.

The Slave Receiver mode may also be entered if arbitration is lost while the TWI is in the Master

mode (see states 0x68 and 0x78).

If the TWEA bit is reset during a transfer, the TWI will return a “Not Acknowledge” (“1”) to SDA

after the next received data byte. This can be used to indicate that the Slave is not able to

receive any more bytes. While TWEA is zero, the TWI does not acknowledge its own slave

address. However, the 2-wire Serial Bus is still monitored and address recognition may resume

at any time by setting TWEA. This implies that the TWEA bit may be used to temporarily isolate

the TWI from the 2-wire Serial Bus.

In all sleep modes other than Idle mode, the clock system to the TWI is turned off. If the TWEA

bit is set, the interface can still acknowledge its own slave address or the general call address by

using the 2-wire Serial Bus clock as a clock source. The part will then wake up from sleep and

the TWI will hold the SCL clock low during the wake up and until the TWINT Flag is cleared (by

writing it to one). Further data reception will be carried out as normal, with the AVR clocks run-

ning as normal. Observe that if the AVR is set up with a long start-up time, the SCL line may be

held low for a long time, blocking other data transmissions.

Note that the 2-wire Serial Interface Data Register – TWDR does not reflect the last byte present

on the bus when waking up from these Sleep modes.

245

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 20-5. Status Codes for Slave Receiver Mode

Status Code

(TWSR)

Prescaler Bits

are 0

Application Software Response

To TWCR

STO TWIN

Status of the 2-wire Serial Bus

and 2-wire Serial Interface Hard-

ware

To/from TWDR

STA

X

TWE

A

Next Action Taken by TWI Hardware

T

0x60

0x68

0x70

0x78

Own SLA+W has been received;

ACK has been returned

No TWDR action or

0

1

0

Data byte will be received and NOT ACK will be

returned

No TWDR action

X

0

1

0

Data byte will be received and ACK will be returned

Arbitration lost in SLA+R/W as

Master; own SLA+W has been

received; ACK has been returned

No TWDR action or

Data byte will be received and NOT ACK will be

returned

No TWDR action

X

0

1

0

Data byte will be received and ACK will be returned

General call address has been

received; ACK has been returned

No TWDR action or

Data byte will be received and NOT ACK will be

returned

No TWDR action

X

0

1

0

Data byte will be received and ACK will be returned

Arbitration lost in SLA+R/W as

Master; General call address has

been received; ACK has been

returned

No TWDR action or

Data byte will be received and NOT ACK will be

returned

No TWDR action

Read data byte or

X

0

1

0

Data byte will be received and ACK will be returned

0x80

0x88

Previously addressed with own

SLA+W; data has been received;

ACK has been returned

Data byte will be received and NOT ACK will be

returned

Read data byte

X

0

1

0

Data byte will be received and ACK will be returned

Previously addressed with own

SLA+W; data has been received;

NOT ACK has been returned

Read data byte or

Switched to the not addressed Slave mode;

no recognition of own SLA or GCA

Switched to the not addressed Slave mode;

own SLA will be recognized;

GCA will be recognized if TWGCE = “1”

Switched to the not addressed Slave mode;

no recognition of own SLA or GCA;

a START condition will be transmitted when the bus

becomes free

Read data byte or

0

1

0

1

0

Read data byte

1

0

1

0

Switched to the not addressed Slave mode;

own SLA will be recognized;

GCA will be recognized if TWGCE = “1”;

a START condition will be transmitted when the bus

becomes free

0x90

0x98

Previously addressed with

general call; data has been re-

ceived; ACK has been returned

Read data byte or

X

0

Data byte will be received and NOT ACK will be

returned

Read data byte

X

0

1

0

Data byte will be received and ACK will be returned

Previously addressed with

general call; data has been

received; NOT ACK has been

returned

Read data byte or

Switched to the not addressed Slave mode;

no recognition of own SLA or GCA

Switched to the not addressed Slave mode;

own SLA will be recognized;

GCA will be recognized if TWGCE = “1”

Switched to the not addressed Slave mode;

no recognition of own SLA or GCA;

a START condition will be transmitted when the bus

becomes free

Read data byte or

0

1

0

1

0

Read data byte

No action

1

0

1

Switched to the not addressed Slave mode;

own SLA will be recognized;

GCA will be recognized if TWGCE = “1”;

a START condition will be transmitted when the bus

becomes free

0xA0

A STOP condition or repeated

START condition has been

received while still addressed as

Slave

0

1

0

1

Switched to the not addressed Slave mode;

no recognition of own SLA or GCA

Switched to the not addressed Slave mode;

own SLA will be recognized;

GCA will be recognized if TWGCE = “1”

Switched to the not addressed Slave mode;

no recognition of own SLA or GCA;

a START condition will be transmitted when the bus

becomes free

1

0

1

0

1

Switched to the not addressed Slave mode;

own SLA will be recognized;

GCA will be recognized if TWGCE = “1”;

a START condition will be transmitted when the bus

becomes free

246

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 20-17. Formats and States in the Slave Receiver Mode

Reception of the own

slave address and one or

more data bytes. All are

acknowledged

S

SLA

W

A

DATA

A

DATA

A

P or S

$60

$80

A

$A0

Last data byte received

is not acknowledged

P or S

$88

Arbitration lost as master

and addressed as slave

A

$68

A

Reception of the general call

address and one or more data

bytes

General Call

DATA

A

DATA

A

P or S

$70

$90

A

$A0

Last data byte received is

not acknowledged

P or S

$98

Arbitration lost as master and

addressed as slave by general call

A

$78

Any number of data bytes

and their associated acknowledge bits

From master to slave

From slave to master

DATA

A

This number (contained in TWSR) corresponds

to a defined state of the Two-Wire Serial Bus. The

prescaler bits are zero or masked to zero

n

20.8.4

Slave Transmitter Mode

In the Slave Transmitter mode, a number of data bytes are transmitted to a Master Receiver

(see Figure 20-18). All the status codes mentioned in this section assume that the prescaler bits

are zero or are masked to zero.

Figure 20-18. Data Transfer in Slave Transmitter Mode

V_CC

Device 1

SLAVE

TRANSMITTER

Device 2

MASTER

RECEIVER

Device 3

Device n

R1

R2

........

SDA

SCL

247

7766E–AVR–04/10

ATmega16U4/ATmega32U4

To initiate the Slave Transmitter mode, TWAR and TWCR must be initialized as follows:

TWAR

TWA6

TWA5

TWA4

TWA3

TWA2

TWA1

TWA0

TWGCE

value

Device’s Own Slave Address

The upper seven bits are the address to which the 2-wire Serial Interface will respond when

addressed by a Master. If the LSB is set, the TWI will respond to the general call address (0x00),

otherwise it will ignore the general call address.

TWCR

TWINT

TWEA

TWSTA

TWSTO

TWWC

TWEN

–

TWIE

value

0

1

0

1

0

X

TWEN must be written to one to enable the TWI. The TWEA bit must be written to one to enable

the acknowledgement of the device’s own slave address or the general call address. TWSTA

and TWSTO must be written to zero.

When TWAR and TWCR have been initialized, the TWI waits until it is addressed by its own

slave address (or the general call address if enabled) followed by the data direction bit. If the

direction bit is “1” (read), the TWI will operate in ST mode, otherwise SR mode is entered. After

its own slave address and the write bit have been received, the TWINT Flag is set and a valid

status code can be read from TWSR. The status code is used to determine the appropriate soft-

ware action. The appropriate action to be taken for each status code is detailed in Table 20-6.

The Slave Transmitter mode may also be entered if arbitration is lost while the TWI is in the

Master mode (see state 0xB0).

If the TWEA bit is written to zero during a transfer, the TWI will transmit the last byte of the trans-

fer. State 0xC0 or state 0xC8 will be entered, depending on whether the Master Receiver

transmits a NACK or ACK after the final byte. The TWI is switched to the not addressed Slave

mode, and will ignore the Master if it continues the transfer. Thus the Master Receiver receives

all “1” as serial data. State 0xC8 is entered if the Master demands additional data bytes (by

transmitting ACK), even though the Slave has transmitted the last byte (TWEA zero and expect-

ing NACK from the Master).

While TWEA is zero, the TWI does not respond to its own slave address. However, the 2-wire

Serial Bus is still monitored and address recognition may resume at any time by setting TWEA.

This implies that the TWEA bit may be used to temporarily isolate the TWI from the 2-wire Serial

Bus.

In all sleep modes other than Idle mode, the clock system to the TWI is turned off. If the TWEA

bit is set, the interface can still acknowledge its own slave address or the general call address by

using the 2-wire Serial Bus clock as a clock source. The part will then wake up from sleep and

the TWI will hold the SCL clock will low during the wake up and until the TWINT Flag is cleared

(by writing it to one). Further data transmission will be carried out as normal, with the AVR clocks

running as normal. Observe that if the AVR is set up with a long start-up time, the SCL line may

be held low for a long time, blocking other data transmissions.

Note that the 2-wire Serial Interface Data Register – TWDR does not reflect the last byte present

on the bus when waking up from these sleep modes.

248

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 20-6. Status Codes for Slave Transmitter Mode

Status Code

(TWSR)

Prescaler

Bits

Application Software Response

To TWCR

STO TWIN

Status of the 2-wire Serial Bus

and 2-wire Serial Interface Hard-

ware

To/from TWDR

STA

TWE

A

Next Action Taken by TWI Hardware

T

are 0

0xA8

0xB0

0xB8

0xC0

Own SLA+R has been received;

ACK has been returned

Load data byte or

Load data byte

X

0

1

0

1

Last data byte will be transmitted and NOT ACK should

be received

Data byte will be transmitted and ACK should be re-

ceived

1

Arbitration lost in SLA+R/W as

Master; own SLA+R has been

received; ACK has been returned

Load data byte or

Load data byte

X

0

1

0

1

Last data byte will be transmitted and NOT ACK should

be received

Data byte will be transmitted and ACK should be re-

ceived

Data byte in TWDR has been

transmitted; ACK has been

received

Load data byte or

Load data byte

X

0

1

0

1

Last data byte will be transmitted and NOT ACK should

be received

Data byte will be transmitted and ACK should be re-

ceived

Data byte in TWDR has been

transmitted; NOT ACK has been

received

No TWDR action or

0

1

0

1

Switched to the not addressed Slave mode;

no recognition of own SLA or GCA

Switched to the not addressed Slave mode;

own SLA will be recognized;

GCA will be recognized if TWGCE = “1”

Switched to the not addressed Slave mode;

no recognition of own SLA or GCA;

a START condition will be transmitted when the bus

becomes free

No TWDR action or

No TWDR action

1

0

1

0

1

Switched to the not addressed Slave mode;

own SLA will be recognized;

GCA will be recognized if TWGCE = “1”;

a START condition will be transmitted when the bus

becomes free

0xC8

Last data byte in TWDR has been

transmitted (TWEA = “0”); ACK

has been received

No TWDR action or

0

1

0

1

Switched to the not addressed Slave mode;

no recognition of own SLA or GCA

Switched to the not addressed Slave mode;

own SLA will be recognized;

GCA will be recognized if TWGCE = “1”

Switched to the not addressed Slave mode;

no recognition of own SLA or GCA;

a START condition will be transmitted when the bus

becomes free

No TWDR action or

No TWDR action

1

0

1

0

1

Switched to the not addressed Slave mode;

own SLA will be recognized;

GCA will be recognized if TWGCE = “1”;

a START condition will be transmitted when the bus

becomes free

249

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 20-19. Formats and States in the Slave Transmitter Mode

Reception of the own

slave address and one or

more data bytes

S

SLA

R

A

DATA

A

DATA

A

P or S

$A8

A

$B8

$C0

Arbitration lost as master

and addressed as slave

$B0

Last data byte transmitted.

Switched to not addressed

slave (TWEA = '0')

A

All 1's

P or S

$C8

Any number of data bytes

and their associated acknowledge bits

From master to slave

From slave to master

DATA

A

This number (contained in TWSR) corresponds

to a defined state of the Two-Wire Serial Bus. The

prescaler bits are zero or masked to zero

n

20.8.5

Miscellaneous States

There are two status codes that do not correspond to a defined TWI state, see Table 20-7.

Status 0xF8 indicates that no relevant information is available because the TWINT Flag is not

set. This occurs between other states, and when the TWI is not involved in a serial transfer.

Status 0x00 indicates that a bus error has occurred during a 2-wire Serial Bus transfer. A bus

error occurs when a START or STOP condition occurs at an illegal position in the format frame.

Examples of such illegal positions are during the serial transfer of an address byte, a data byte,

or an acknowledge bit. When a bus error occurs, TWINT is set. To recover from a bus error, the

TWSTO Flag must set and TWINT must be cleared by writing a logic one to it. This causes the

TWI to enter the not addressed Slave mode and to clear the TWSTO Flag (no other bits in

TWCR are affected). The SDA and SCL lines are released, and no STOP condition is

transmitted.

Table 20-7. Miscellaneous States

Status Code

(TWSR)

Prescaler Bits

are 0

Application Software Response

To TWCR

Status of the 2-wire Serial Bus

and 2-wire Serial Interface

Hardware

To/from TWDR

STA

STO

TWIN

T

TWE

A

Next Action Taken by TWI Hardware

Wait or proceed current transfer

0xF8

0x00

No relevant state information

available; TWINT = “0”

No TWDR action

No TWCR action

Bus error due to an illegal

START or STOP condition

0

1

X

Only the internal hardware is affected, no STOP condi-

tion is sent on the bus. In all cases, the bus is released

and TWSTO is cleared.

20.8.6

Combining Several TWI Modes

In some cases, several TWI modes must be combined in order to complete the desired action.

Consider for example reading data from a serial EEPROM. Typically, such a transfer involves

the following steps:

1. The transfer must be initiated.

2. The EEPROM must be instructed what location should be read.

3. The reading must be performed.

4. The transfer must be finished.

250

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Note that data is transmitted both from Master to Slave and vice versa. The Master must instruct

the Slave what location it wants to read, requiring the use of the MT mode. Subsequently, data

must be read from the Slave, implying the use of the MR mode. Thus, the transfer direction must

be changed. The Master must keep control of the bus during all these steps, and the steps

should be carried out as an atomical operation. If this principle is violated in a multi master sys-

tem, another Master can alter the data pointer in the EEPROM between steps 2 and 3, and the

Master will read the wrong data location. Such a change in transfer direction is accomplished by

transmitting a REPEATED START between the transmission of the address byte and reception

of the data. After a REPEATED START, the Master keeps ownership of the bus. The following

figure shows the flow in this transfer.

Figure 20-20. Combining Several TWI Modes to Access a Serial EEPROM

Master Transmitter

Master Receiver

S

SLA+W

A

ADDRESS

A

Rs

SLA+R

A

DATA

A

P

S = START

Transmitted from master to slave

Rs = REPEATED START

Transmitted from slave to master

P = STOP

20.9 Multi-master Systems and Arbitration

If multiple masters are connected to the same bus, transmissions may be initiated simultane-

ously by one or more of them. The TWI standard ensures that such situations are handled in

such a way that one of the masters will be allowed to proceed with the transfer, and that no data

will be lost in the process. An example of an arbitration situation is depicted below, where two

masters are trying to transmit data to a Slave Receiver.

Figure 20-21. An Arbitration Example

V_CC

Device 1

MASTER

TRANSMITTER

Device 3

SLAVE

RECEIVER

Device 2

MASTER

TRANSMITTER

Device n

R1

R2

........

SDA

SCL

Several different scenarios may arise during arbitration, as described below:

• Two or more masters are performing identical communication with the same Slave. In this

case, neither the Slave nor any of the masters will know about the bus contention.

• Two or more masters are accessing the same Slave with different data or direction bit. In this

case, arbitration will occur, either in the READ/WRITE bit or in the data bits. The masters

trying to output a one on SDA while another Master outputs a zero will lose the arbitration.

Losing masters will switch to not addressed Slave mode or wait until the bus is free and

transmit a new START condition, depending on application software action.

251

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Two or more masters are accessing different slaves. In this case, arbitration will occur in the

SLA bits. Masters trying to output a one on SDA while another Master outputs a zero will lose

the arbitration. Masters losing arbitration in SLA will switch to Slave mode to check if they are

being addressed by the winning Master. If addressed, they will switch to SR or ST mode,

depending on the value of the READ/WRITE bit. If they are not being addressed, they will

switch to not addressed Slave mode or wait until the bus is free and transmit a new START

condition, depending on application software action.

This is summarized in Figure 20-22. Possible status values are given in circles.

Figure 20-22. Possible Status Codes Caused by Arbitration

START

SLA

Data

STOP

Arbitration lost in SLA

Arbitration lost in Data

Own

No

38

TWI bus will be released and not addressed slave mode will be entered

A START condition will be transmitted when the bus becomes free

Address / General Call

received

Yes

Write

68/78

B0

Data byte will be received and NOT ACK will be returned

Data byte will be received and ACK will be returned

Direction

Read

Last data byte will be transmitted and NOT ACK should be received

Data byte will be transmitted and ACK should be received

252

7766E–AVR–04/10

ATmega16U4/ATmega32U4

21. USB controller

21.1 Features

• Supports full-speed and low-speed Device role

• Complies with USB Specification v2.0

• Supports ping-pong mode (dual bank)

• 832 bytes of DPRAM:

– 1 endpoint 64 bytes max (default control endpoint)

– 1 endpoints of 256 bytes max, (one or two banks)

– 5 endpoints of 64 bytes max, (one or two banks)

• Crystal-less operation for low-speed mode

21.2 Block Diagram

The USB controller provides the hardware to interface a USB link to a data flow stored in a dou-

ble port memory (DPRAM).

The USB controller requires a 48 MHz 0.25% reference clock (for Full-Speed operation), which

is the output of an internal PLL. The on-chip PLL generates the internal high frequency (48 MHz)

clock for USB interface. The PLL clock input can be configured to use external low-power crystal

oscillator, external source clock or internal RC (see Section “Crystal-less operation”, page 256).

The 48MHz clock is used to generate a 12 MHz Full-speed (or 1.5 MHz Low-Speed) bit clock

from the received USB differential data and to transmit data according to full or low speed USB

device tolerance. Clock recovery is done by a Digital Phase Locked Loop (DPLL) block, which is

compliant with the jitter specification of the USB bus.

To comply with the USB Electrical specification, USB buffers (D+ or D-) should be powered

within the 3.0 to 3.6V range. As ATmega16U4/ATmega32U4 can be powered up to 5.5V, an

internal regulator provides the USB buffers power supply.

Figure 21-1. USB controller Block Diagram overview

UVCC

AVCC

XT1

IntRC

Clock Mux

clk

8MHz

PLL

&

PLL clock

Prescaler

USB Regulator

UCAP

Div-by-2

clk

48MHz

CPU

D-

DPLL

Clock

Recovery

D+

USB

Interface

VBUS

On-Chip

USB DPRAM

253

7766E–AVR–04/10

ATmega16U4/ATmega32U4

21.3 Typical Application Implementation

Depending on the target application power supply, the ATmega16U4/ATmega32U4 requires dif-

ferent hardware typical implementations.

Figure 21-2. Operating modes versus frequency and power-supply

Max

VCC (V)

Operating Frequency (MHz)

5.5

16 MHz

4.5

USB compliant,

with internal regulator

3.6

3.4

8 MHz

USB compliant,

without internal regulator

3.0

2.7

USB not operational

2 MHz

VCC min

0

21.3.1

Bus Powered device

Figure 21-3. Typical Bus powered application with 5V I/O

UVCC

AVCC

VCC

UCAP

1µF

VBUS

UDP

UDM

UVSS

UID

VBUS

D+

Rs=22

D-

UGND

UID

XTAL1

XTAL2

GND

254

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 21-4. Typical Bus powered application with 3V I/O

External

3V Regulator

UVCC

AVCC

VCC

UCAP

1µF

VBUS

UDP

UDM

UVSS

UID

VBUS

D+

Rs=22

D-

UVSS

UGND

UID

XTAL1

XTAL2

GND

21.3.2

Self Powered device

Figure 21-5. Typical Self powered application with 3.4V to 5.5V I/O

External 3.4V - 5.5V

Power Supply

UVCC

AVCC

VCC

UCAP

1µF

VBUS

UDP

UDM

UVSS

UID

VBUS

D+

Rs=22

D-

UGND

UID

XTAL1

XTAL2

GND

255

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 21-6. Typical Self powered application with 3.0V to 3.6 I/O

External 3.0V - 3.6V

Power Supply

UVCC

AVCC

VCC

UCAP

1µF

VBUS

UDP

UDM

UVSS

UID

VBUS

D+

Rs=22

D-

UGND

UID

XTAL1

XTAL2

GND

21.4 Crystal-less operation

To reduce external components count and BOM cost, the USB module can be configured to

operate in low-speed mode with internal RC oscillator as input source clock for the PLL. The

internal RC oscillator is factory calibrated to satisfy the USB low speed frequency accuracy

within the 0°C and -40°C temperature range.

For USB full-speed operation only external crystal oscillator or external source clock can be

used.

21.5 Design guidelines

• Serial resistors on USB Data lines must have 22 Ohms value (+/- 5%).

• Traces from the input USB receptable (or from the cable connection in the case of a tethered

device) to the USB microcontroller pads should be as short as possible, and follow differential

traces routing rules (same length, as near as possible, avoid via accumulation).

• Voltage transient / ESD suppressors may also be used to prevent USB pads to be damaged

by external disturbances.

• U_capcapacitor should be 1µF (+/- 10%) for correct operation.

• A 10µF capacitor is highly recommended on VBUS line

256

7766E–AVR–04/10

ATmega16U4/ATmega32U4

21.6 General Operating Modes

21.6.1

Introduction

The USB controller is disabled and reset after an hardware reset generated by:

– Power on reset

– External reset

– Watchdog reset

– Brown out reset

– JTAG reset

But another available and optional CPU reset source is:

– USB End Of Reset

In this case, the USB controller is reset, but not disabled (so that the device remains attached).

21.6.2

Power-on and reset

The next diagram explains the USB controller main states on power-on:

Figure 21-7. USB controller states after reset

Clock stopped

FRZCLK=1

Macro off

<any other

state>

USBE=0

Reset

HW

RESET

USBE=1

USBE=0

Device

USB Controller state after an hardware reset is ‘Reset’. In this state:

• USBE is not set

• the USB controller clock is stopped in order to minimize the power consumption

(FRZCLK=1),

• the USB controller is disabled,

• the USB pad is in the suspend mode,

• the Device USB controller internal state is reset.

After setting USBE, the USB Controller enters the Device state. The controller is ‘Idle’.

The USB Controller can at any time be stopped by clearing USBE. In fact, clearing USBE acts

as an hardware reset.

21.6.3

Interrupts

Two interrupts vectors are assigned to USB interface.

257

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 21-8. USB Interrupt System

USB General

Interrupt

USB General

Interrupt Vector

USB Device

Interrupt

USB Endpoint/Pipe

Interrupt Vector

Endpoint

Interrupt

The USB hardware module distinguishes between USB General events and USB Endpoint

events that are relevant with data transfers relative to each endpoint.

Figure 21-9. USB General interrupt vector sources

USB General

Interrupt Vector

VBUSTI

USBINT.0

VBUSTE

USBCON.0

UPRSMI

UDINT.6

UPRSME

UDIEN.6

EORSMI

UDINT.5

EORSME

UDIEN.5

WAKEUPI

UDINT.4

USB General

Interrupt Vector

WAKEUPE

UDIEN.4

USB Device

Interrupt

EORSTI

UDINT.3

EORSTE

UDIEN.3

SOFI

UDINT.2

SOFE

UDIEN.2

SUSPI

UDINT.0

Asynchronous Interrupt source

(allows the CPU to wake up from power down mode)

SUSPE

UDIEN.0

Almost all these interrupts are time-relative events that will be detected only if the USB clock is

enabled (FRZCLK bit set), except for:

• VBUS plug-in detection (insert, remove)

• WAKEUP interrupt that will trigger each time a state change is detected on the data lines.

This asynchronous interrupts allow to wake-up a device that is in power-down mode, generally

after that the USB has entered the Suspend state.

258

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 21-10. USB Endpoint Interrupt vector sources

Endpoint 6

Endpoint 5

Endpoint 4

Endpoint 3

Endpoint 2

Endpoint 1

Endpoint 0

OVERFI

UESTAX.6

UNDERFI

UESTAX.5

FLERRE

UEIENX.7

NAKINI

UEINTX.6

NAKINE

UEIENX.6

NAKOUTI

UEINTX.4

TXSTPE

USB Endpoint

Interrupt Vector

UEIENX.4

RXSTPI

EPINT

UEINTX.3

UEINT.X

RXSTPE

UEIENX.3

RXOUTI

UEINTX.2

RXOUTE

UEIENX.2

STALLEDI

UEINTX.1

STALLEDE

UEIENX.1

TXINI

UEINTX.0

TXINE

UEIENX.0

Each endpoint has 8 interrupts sources associated with flags, and each source can be enabled

or not to trigger the corresponding endpoint interrupt. If, for an endpoint, at least one of the

sources is enabled to trigger interrupt, the corresponding event(s) will make the program branch

to the USB Endpoint Interrupt vector. The user may determine the source (endpoint) of the inter-

rupt by reading the UEINT register, and then handle the event detected by polling the different

flags.

21.7 Power modes

21.7.1

Idle mode

In this mode, the CPU core is halted (CPU clock stopped). The Idle mode is taken wether the

USB controller is running or not. The CPU “wakes up” on any USB interrupts.

21.7.2

Power down

In this mode, the oscillator is stopped and halts all the clocks (CPU and peripherals). The USB

controller “wakes up” when:

• the WAKEUPI interrupt is triggered

• the VBUSTI interrupt is triggered

259

7766E–AVR–04/10

ATmega16U4/ATmega32U4

21.7.3

Freeze clock

The firmware has the ability to reduce the power consumption by setting the FRZCLK bit, which

freeze the clock of USB controller. When FRZCLK is set, it is still possible to access to the fol-

lowing registers:

• USBCON, USBSTA, USBINT

• UDCON (detach, ...)

• UDINT

• UDIEN

Moreover, when FRZCLK is set, only the following interrupts may be triggered:

• WAKEUPI

• VBUSTI

21.8 Speed Control

The speed selection (Full Speed or Low Speed) depends on the D+/D- pull-up. The LSM bit in

UDCON register allows to select an internal pull up on D- (Low Speed mode) or D+ (Full Speed

mode) data lines.

Figure 21-11. Device mode Speed Selection

USB

Regulator

UCAP

DETACH

UDCON.0

LSM

UDCON.2

D+

D-

21.9 Memory management

The controller only supports the following memory allocation management.

The reservation of a Pipe or an Endpoint can only be made in the increasing order (Pipe/End-

point 0 to the last Pipe/Endpoint). The firmware shall thus configure them in the same order.

The reservation of a Pipe or an Endpoint “kⁱ” is done when its ALLOC bit is set. Then, the hard-

ware allocates the memory and inserts it between the Pipe/Endpoints “k^i-1” and “kⁱ⁺¹”. The “kⁱ⁺¹”

Pipe/Endpoint memory “slides” up and its data is lost. Note that the “kⁱ⁺²” and upper Pipe/End-

point memory does not slide.

Clearing a Pipe enable (PEN) or an Endpoint enable (EPEN) does not clear either its ALLOC bit,

or its configuration (EPSIZE/PSIZE, EPBK/PBK). To free its memory, the firmware should clear

ALLOC. Then, the “kⁱ⁺¹” Pipe/Endpoint memory automatically “slides” down. Note that the “kⁱ⁺²”

and upper Pipe/Endpoint memory does not slide.

260

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The following figure illustrates the allocation and reorganization of the USB memory in a typical

example:

Table 21-1. Allocation and reorganization USB memory flow

Free memory

5

Free memory

5

Free memory

5

4

5

4

Conflict

4

3

Lost memory

4

EPEN=0

(ALLOC=1)

3 (bigger size)

2

1

0

1

0

1

0

1

0

EPEN=1

ALLOC=1

Endpoints

activation

Free its memory

(ALLOC=0)

Endpoint

Activatation

Endpoint Disable

• First, Endpoint 0 to Endpoint 5 are configured, in the growing order. The memory of each is

reserved in the DPRAM.

• Then, the Endpoint 3 is disabled (EPEN=0), but its memory reservation is internally kept by

the controller.

• Its ALLOC bit is cleared: the Endpoint 4 “slides” down, but the Endpoint 5 does not “slide”.

• Finally, if the firmware chooses to reconfigure the Endpoint 3, with a bigger size. The

controller reserved the memory after the Endpoint 2 memory and automatically “slide” the

Endpoint 4. The Endpoint 5 does not move and a memory conflict appear, in that both

Endpoint 4 and 5 use a common area. The data of those endpoints are potentially lost.

Note that:

• the data of Endpoint 0 are never lost whatever the activation or deactivation of the higher

Endpoint. Its data is lost if it is deactivated.

• Deactivate and reactivate the same Endpoint with the same parameters does not lead to a

“slide” of the higher endpoints. For those endpoints, the data are preserved.

• CFGOK is set by hardware even in the case where there is a “conflict” in the memory

allocation.

21.10 PAD suspend

The next figures illustrates the pad behaviour:

• In the “idle” mode, the pad is put in low power consumption mode.

• In the “active” mode, the pad is working.

261

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 21-12. Pad behaviour

USBE=1

& DETACH=0

& suspend

Idle mode

USBE=0

| DETACH=1

| suspend

Active mode

The SUSPI flag indicated that a suspend state has been detected on the USB bus. This flag

automatically put the USB pad in Idle. The detection of a non-idle event sets the WAKEUPI flag

and wakes-up the USB pad.

Suspend detected = USB pad power down

Clear Suspend by software

SUSPI

WAKEUPI

Clear Resume by software

Resume = USB pad wake-up

PAD status

Power Down

Active

Moreover, the pad can also be put in the “idle” mode if the DETACH bit is set. It come back in

the active mode when the DETACH bit is cleared.

21.11 Plug-in detection

The USB connection is detected by the VBUS pad, thanks to the following architecture:

262

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 21-13. Plug-in Detection Input Block Diagram

VDD

Session_valid

VBUS

USBSTA.0

VBUSTI

USBINT.0

VBUS

VSS

Pad logic

The control logic of the VBUS pad outputs a signal regarding the VBUS voltage level:

• The “Session_valid” signal is active high when the voltage on the UVBUS pad is higher or

equal to 1.4V. If lower than 1.4V, the signal is not active.

• The VBUS status bit is set when “Session_valid” signal is active (VBUS > 1.4V).

• The VBUSTI flag is set each time the VBUS state changes.

• The USB peripheral cannot attach to the bus while VBUS bit is not set.

21.12 Registers description

21.12.1 USB general registers

Bit

7

-

6

-

5

-

4

-

3

-

2

-

1

-

0

UVREGE

R/W

UHWCON

Read/Write

Initial Value

R/W

0

R/W

0

R

0

R/W

0

R

0

R

0

R

0

• 7-1 – Reserved

These bits are reserved. Do not modify these bits.

• 0 – UVREGE: USB pad regulator Enable

Set to enable the USB pad regulator. Clear to disable the USB pad regulator.

Bit

7

USBE

R/W

0

6

-

5

FRZCLK

R/W

4

OTGPADE

R/W

3

-

2

-

1

-

0

VBUSTE

R/W

USBCON

Read/Write

Initial Value

R/W

0

R

0

R

0

R/W

0

1

0

• 7 – USBE: USB macro Enable Bit

Set to enable the USB controller. Clear to disable and reset the USB controller, to disable the

USB transceiver and to disable the USB controller clock inputs.

• 6 – Reserved

The value read from these bits is always 0. Do not set these bits.

263

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• 5 – FRZCLK: Freeze USB Clock Bit

Set to disable the clock inputs (the ”Resume Detection” is still active). This reduces the power

consumption. Clear to enable the clock inputs.

• 4 – OTGPADE: VBUS Pad Enable

Set to enable the VBUS pad. Clear to disable the VBUS pad.

Note that this bit can be set/cleared even if USBE=0. That allows the VBUS detection even if the

USB macro is disable.

• 3-1 – Reserved

The value read from these bits is always 0. Do not set these bits.

• 0 – VBUSTE: VBUS Transition Interrupt Enable Bit

Set this bit to enable the VBUS Transition interrupt generation.

Clear this bit to disable the VBUS Transition interrupt generation.

Bit

7

-

6

-

5

-

4

-

3

-

2

-

1

ID

R

1

0

VBUS

R

USBSTA

Read/Write

Initial Value

R

0

R

0

R

0

R

0

R

0

R

0

• 7-2 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 1 - ID: ID status

This bit is always read as “1”, it has been conserved for compatibility with AT90USB64/128 (in

which it indicates the value of the OTG ID pin).

• 0 – VBUS: VBus Flag

The value read from this bit indicates the state of the VBUS pin. This bit can be used in device

mode to monitor the USB bus connection state of the application. See Section 21.11, page 262

for more details.

Bit

7

-

6

-

5

-

4

-

3

-

2

-

1

-

0

VBUSTI

R/W

0

USBINT

Read/Write

Initial Value

R

0

R

0

R

0

R

0

R

0

R

0

R/W

0

7-1 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 0 – VBUSTI: IVBUS Transition Interrupt Flag

Set by hardware when a transition (high to low, low to high) has been detected on the VBUS

pad.

Shall be cleared by software.

264

7766E–AVR–04/10

ATmega16U4/ATmega32U4

21.13 USB Software Operating modes

Depending on the USB operating mode, the software should perform some the following

operations:

Power On the USB interface

• Power-On USB pads regulator

• Configure PLL interface

• Enable PLL

• Check PLL lock

• Enable USB interface

• Configure USB interface (USB speed, Endpoints configuration...)

• Wait for USB VBUS information connection

• Attach USB device

Power Off the USB interface

• Detach USB interface

• Disable USB interface

• Disable PLL

• Disable USB pad regulator

Suspending the USB interface

• Clear Suspend Bit

• Freeze USB clock

• Disable PLL

• Be sure to have interrupts enable to exit sleep mode

• Make the MCU enter sleep mode

Resuming the USB interface

• Enable PLL

• Wait PLL lock

• Unfreeze USB clock

• Clear Resume information

265

7766E–AVR–04/10

ATmega16U4/ATmega32U4

22. USB Device Operating modes

22.1 Introduction

The USB device controller supports full speed and low speed data transfers. In addition to the

default control endpoint, it provides six other endpoints, which can be configured in control, bulk,

interrupt or isochronous modes:

• Endpoint 0:programmable size FIFO up to 64 bytes, default control endpoint

• Endpoints 1 programmable size FIFO up to 256 bytes in ping-pong mode.

• Endpoints 2 to 6: programmable size FIFO up to 64 bytes in ping-pong mode.

The controller starts in the “idle” mode. In this mode, the pad consumption is reduced to the

minimum.

22.2 Power-on and reset

The next diagram explains the USB device controller main states on power-on:

Figure 22-1. USB device controller states after reset

<any

other

state>

USBE=0

Idle

Reset

USBE=1

HW

RESET

The reset state of the Device controller is:

• the macro clock is stopped in order to minimize the power consumption (FRZCLK set),

• the USB device controller internal state is reset (all the registers are reset to their default

value. Note that DETACH is set.)

• the endpoint banks are reset

• the D+ or D- pull up are not activated (mode Detach)

The D+ or D- pull-up will be activated as soon as the DETACH bit is cleared and VBUS is

present.

The macro is in the ‘Idle’ state after reset with a minimum power consumption and does not

need to have the PLL activated to enter this state.

The USB device controller can at any time be reset by clearing USBE (disable USB interface).

22.3 Endpoint reset

An endpoint can be reset at any time by setting in the UERST register the bit corresponding to

the endpoint (EPRSTx). This resets:

• the internal state machine on that endpoint,

• the Rx and Tx banks are cleared and their internal pointers are restored,

266

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• the UEINTX, UESTA0X and UESTA1X are restored to their reset value.

The data toggle field remains unchanged.

The other registers remain unchanged.

The endpoint configuration remains active and the endpoint is still enabled.

The endpoint reset may be associated with a clear of the data toggle command (RSTDT bit) as

an answer to the CLEAR_FEATURE USB command.

22.4 USB reset

When an USB reset is detected on the USB line (SE0 state with a minimum duration of 2.5µs),

the next operations are performed by the controller:

• all the endpoints are disabled

• the default control endpoint remains configured (see Section 22.3, page 266 for more

details).

If the CPU hardware reset function is activated (RSTCPU bit set in UDCON register), a reset is

generated to the CPU core without disabling the USB controller (that follows the same behavior

than after a standard USB End of Reset, and remains attached). That feature may be used to

enhance device reliability.

22.5 Endpoint selection

Prior to any operation performed by the CPU, the endpoint must first be selected. This is done

by setting the EPNUM2:0 bits (UENUM register) with the endpoint number which will be man-

aged by the CPU.

The CPU can then access to the various endpoint registers and data.

22.6 Endpoint activation

The endpoint is maintained under reset as long as the EPEN bit is not set.

The following flow must be respected in order to activate an endpoint:

267

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 22-2. Endpoint activation flow:

Endpoint

Activation

Select the endpoint

Activate the endpoint

UENUM

EPNUM=x

EPEN=1

Configure:

- the endpoint direction

- the endpoint type

UECFG0X

EPDIR

EPTYPE

...

Configure:

- the endpoint size

- the bank parametrization

Allocation and reorganization of

the memory is made on-the-fly

UECFG1X

ALLOC

EPSIZE

EPBK

Test the correct endpoint

configuration

CFGOK=1

No

Yes

Endpoint activated

ERROR

As long as the endpoint is not correctly configured (CFGOK cleared), the hardware does not

acknowledge the packets sent by the host.

CFGOK is will not be sent if the Endpoint size parameter is bigger than the DPRAM size.

A clear of EPEN acts as an endpoint reset (see Section 22.3, page 266 for more details). It also

performs the next operation:

• The configuration of the endpoint is kept (EPSIZE, EPBK, ALLOC kept)

• It resets the data toggle field.

• The DPRAM memory associated to the endpoint is still reserved.

See Section 21.9, page 260 for more details about the memory allocation/reorganization.

22.7 Address Setup

The USB device address is set up according to the USB protocol:

• the USB device, after power-up, responds at address 0

• the host sends a SETUP command (SET_ADDRESS(addr)),

• the firmware handles this request, and records that address in UADD, but keep ADDEN

cleared,

• the USB device firmware sends an IN command of 0 bytes (IN 0 Zero Length Packet),

• then, the firmware can enable the USB device address by setting ADDEN. The only accepted

address by the controller is the one stored in UADD.

ADDEN and UADD shall not be written at the same time.

UADD contains the default address 00h after a power-up or USB reset.

268

7766E–AVR–04/10

ATmega16U4/ATmega32U4

ADDEN is cleared by hardware:

• after a power-up reset,

• when an USB reset is received,

• or when the macro is disabled (USBE cleared)

When this bit is cleared, the default device address 00h is used.

22.8 Suspend, Wake-up and Resume

After a period of 3 ms during which the USB line was inactive, the controller switches to the full-

speed mode and triggers (if enabled) the SUSPI (suspend) interrupt. The firmware may then set

the FRZCLK bit.

The CPU can also, depending on software architecture, enter in the idle mode to lower again the

power consumption.

There are two ways to recover from the “Suspend” mode:

• First one is to clear the FRZCLK bit. This is possible if the CPU is not in the Idle mode.

• Second way, if the CPU is “idle”, is to enable the WAKEUPI interrupt (WAKEUPE set). Then,

as soon as an non-idle signal is seen by the controller, the WAKEUPI interrupt is triggered.

The firmware shall then clear the FRZCLK bit to restart the transfer.

There are no relationship between the SUSPI interrupt and the WAKEUPI interrupt: the WAKE-

UPI interrupt is triggered as soon as there are non-idle patterns on the data lines. Thus, the

WAKEUPI interrupt can occurs even if the controller is not in the “suspend” mode.

When the WAKEUPI interrupt is triggered, if the SUSPI interrupt bit was already set, it is cleared

by hardware.

When the SUSPI interrupt is triggered, if the WAKEUPI interrupt bit was already set, it is cleared

by hardware.

22.9 Detach

The reset value of the DETACH bit is 1.

It is possible to re-enumerate a device, simply by setting and clearing the DETACH bit (but firm-

ware must take in account a debouncing delay of some milliseconds).

• Setting DETACH will disconnect the pull-up on the D+ or D- pad (depending on full or low

speed mode selected). Then, clearing DETACH will connect the pull-up on the D+ or D- pad.

Figure 22-3. Detach a device in Full-speed:

UVREF

D +

D -

D +

D -

Detach, then

Attach

EN=1

269

7766E–AVR–04/10

ATmega16U4/ATmega32U4

22.10 Remote Wake-up

The “Remote Wake-up” (or “upstream resume”) feature is the only operation allowed to be sent

by the device on its own initiative. Anyway, to do that, the device should first have received a

DEVICE_REMOTE_WAKEUP request from the host.

• First, the USB controller must have detected the “suspend” state of the line: the remote wake-

up can only be sent when a SUSPI flag is set.

• The firmware has then the ability to set RMWKUP to send the “upstream resume” stream.

This will automatically be done by the controller after 5ms of inactivity on the USB line.

• When the controller starts to send the “upstream resume”, the UPRSMI interrupt is triggered

(if enabled). SUSPI is cleared by hardware.

• RMWKUP is cleared by hardware at the end of the “upstream resume”.

• If the controller detects a good “End Of Resume” signal from the host, an EORSMI interrupt

is triggered (if enabled).

22.11 STALL request

For each endpoint, the STALL management is performed using 2 bits:

– STALLRQ (enable stall request)

– STALLRQC (disable stall request)

– STALLEDI (stall sent interrupt)

To send a STALL handshake at the next request, the STALLRQ request bit has to be set. All fol-

lowing requests will be handshak’ed with a STALL until the STALLRQC bit is set.

Setting STALLRQC automatically clears the STALLRQ bit. The STALLRQC bit is also immedi-

ately cleared by hardware after being set by software. Thus, the firmware will never read this bit

as set.

Each time the STALL handshake is sent, the STALLEDI flag is set by the USB controller and the

EPINTx interrupt will be triggered (if enabled).

The incoming packets will be discarded (RXOUTI and RWAL will not be set).

The host will then send a command to reset the STALL: the firmware just has to set the STALL-

RQC bit and to reset the endpoint.

22.11.1 Special consideration for Control Endpoints

A SETUP request is always ACK’ed.

If a STALL request is set for a Control Endpoint and if a SETUP request occurs, the SETUP

request has to be ACK’ed and the STALLRQ request and STALLEDI sent flags are automati-

cally reset (RXSETUPI set, TXIN cleared, STALLED cleared, TXINI cleared...).

This management simplifies the enumeration process management. If a command is not sup-

ported or contains an error, the firmware set the STALL request flag and can return to the main

task, waiting for the next SETUP request.

This function is compliant with the Chapter 8 test that may send extra status for a

GET_DESCRIPTOR. The firmware sets the STALL request just after receiving the status. All

extra status will be automatically STALL’ed until the next SETUP request.

270

7766E–AVR–04/10

ATmega16U4/ATmega32U4

22.11.2 STALL handshake and Retry mechanism

The Retry mechanism has priority over the STALL handshake. A STALL handshake is sent if the

STALLRQ request bit is set and if there is no retry required.

22.12 CONTROL endpoint management

A SETUP request is always ACK’ed. When a new setup packet is received, the RXSTPI inter-

rupt is triggered (if enabled). The RXOUTI interrupt is not triggered.

The FIFOCON and RWAL fields are irrelevant with CONTROL endpoints. The firmware shall

thus never use them on that endpoints. When read, their value is always 0.

CONTROL endpoints are managed by the following bits:

• RXSTPI is set when a new SETUP is received. It shall be cleared by firmware to

acknowledge the packet and to clear the endpoint bank.

• RXOUTI is set when a new OUT data is received. It shall be cleared by firmware to

acknowledge the packet and to clear the endpoint bank.

• TXINI is set when the bank is ready to accept a new IN packet. It shall be cleared by firmware

to send the packet and to clear the endpoint bank.

22.12.1 Control Write

The next figure shows a control write transaction. During the status stage, the controller will not

necessary send a NAK at the first IN token:

• If the firmware knows the exact number of descriptor bytes that must be read, it can then

anticipate on the status stage and send a ZLP for the next IN token,

• or it can read the bytes and poll NAKINI, which tells that all the bytes have been sent by the

host, and the transaction is now in the status stage.

SETUP

DATA

STATUS

USB line

RXSTPI

RXOUTI

TXINI

SETUP

OUT

IN

NAK

HW

SW

HW

SW

HW

SW

271

7766E–AVR–04/10

ATmega16U4/ATmega32U4

22.12.2 Control Read

The next figure shows a control read transaction. The USB controller has to manage the simulta-

neous write requests from the CPU and the USB host:

SETUP

HW

DATA

STATUS

USB line

RXSTPI

RXOUTI

TXINI

IN

OUT

NAK

OUT

SW

HW

SW

HW

SW

Wr Enable

HOST

Wr Enable

CPU

A NAK handshake is always generated at the first status stage command.

When the controller detect the status stage, all the data written by the CPU are erased, and

clearing TXINI has no effects.

The firmware checks if the transmission is complete or if the reception is complete.

The OUT retry is always ack’ed. This reception:

- set the RXOUTI flag (received OUT data)

- set the TXINI flag (data sent, ready to accept new data)

software algorithm:

set transmit ready

wait (transmit complete OR Receive complete)

if receive complete, clear flag and return

if transmit complete, continue

Once the OUT status stage has been received, the USB controller waits for a SETUP request.

The SETUP request have priority over any other request and has to be ACK’ed. This means that

any other flag should be cleared and the fifo reset when a SETUP is received.

WARNING: the byte counter is reset when the OUT Zero Length Packet is received. The firm-

ware has to take care of this.

22.13 OUT endpoint management

OUT packets are sent by the host. All the data can be read by the CPU, which acknowledges or

not the bank when it is empty.

22.13.1 Overview

The Endpoint must be configured first.

Each time the current bank is full, the RXOUTI and the FIFOCON bits are set. This triggers an

interrupt if the RXOUTE bit is set. The firmware can acknowledge the USB interrupt by clearing

the RXOUTI bit. The Firmware read the data and clear the FIFOCON bit in order to free the cur-

rent bank. If the OUT Endpoint is composed of multiple banks, clearing the FIFOCON bit will

272

7766E–AVR–04/10

ATmega16U4/ATmega32U4

switch to the next bank. The RXOUTI and FIFOCON bits are then updated by hardware in accor-

dance with the status of the new bank.

RXOUTI shall always be cleared before clearing FIFOCON.

The RWAL bit always reflects the state of the current bank. This bit is set if the firmware can

read data from the bank, and cleared by hardware when the bank is empty.

Example with 1 OUT data bank

DATA

(to bank 0)

NAK

DATA

(to bank 0)

OUT

ACK

HW

OUT

ACK

HW

RXOUTI

SW

read data from CPU

BANK 0

FIFOCON

SW

read data from CPU

BANK 0

Example with 2 OUT data banks

DATA

OUT

DATA

(to bank 1)

ACK

HW

OUT

ACK

(to bank 0)

HW

RXOUTI

SW

read data from CPU

BANK 0

SW

FIFOCON

read data from CPU

BANK 1

22.13.2 Detailed description

22.13.2.1

The data are read by the CPU, following the next flow:

• When the bank is filled by the host, an endpoint interrupt (EPINTx) is triggered, if enabled

(RXOUTE set) and RXOUTI is set. The CPU can also poll RXOUTI or FIFOCON, depending

on the software architecture,

• The CPU acknowledges the interrupt by clearing RXOUTI,

• The CPU can read the number of byte (N) in the current bank (N=BYCT),

• The CPU can read the data from the current bank (“N” read of UEDATX),

• The CPU can free the bank by clearing FIFOCON when all the data is read, that is:

– after “N” read of UEDATX,

– as soon as RWAL is cleared by hardware.

If the endpoint uses 2 banks, the second one can be filled by the HOST while the current one is

being read by the CPU. Then, when the CPU clear FIFOCON, the next bank may be already

ready and RXOUTI is set immediately.

273

7766E–AVR–04/10

ATmega16U4/ATmega32U4

22.14 IN endpoint management

IN packets are sent by the USB device controller, upon an IN request from the host. All the data

can be written by the CPU, which acknowledge or not the bank when it is full.Overview

The Endpoint must be configured first.

The TXINI bit is set by hardware when the current bank becomes free. This triggers an interrupt

if the TXINE bit is set. The FIFOCON bit is set at the same time. The CPU writes into the FIFO

and clears the FIFOCON bit to allow the USB controller to send the data. If the IN Endpoint is

composed of multiple banks, this also switches to the next data bank. The TXINI and FIFOCON

bits are automatically updated by hardware regarding the status of the next bank.

TXINI shall always be cleared before clearing FIFOCON.

The RWAL bit always reflects the state of the current bank. This bit is set if the firmware can

write data to the bank, and cleared by hardware when the bank is full.

Example with 1 IN data bank

NAK

DATA

(bank 0)

IN

ACK

HW

IN

TXINI

SW

write data from CPU

BANK 0

FIFOCON

SW

write data from CPU

BANK 0

Example with 2 IN data banks

DATA

(bank 0)

DATA

(bank 1)

IN

ACK

HW

IN

ACK

TXINI

SW

write data from CPU

BANK0

write data from CPU

BANK 0

write data from CPU

BANK 1

FIFOCON

SW

22.14.1 Detailed description

The data are written by the CPU, following the next flow:

• When the bank is empty, an endpoint interrupt (EPINTx) is triggered, if enabled (TXINE set)

and TXINI is set. The CPU can also poll TXINI or FIFOCON, depending the software

architecture choice,

• The CPU acknowledges the interrupt by clearing TXINI,

• The CPU can write the data into the current bank (write in UEDATX),

• The CPU can free the bank by clearing FIFOCON when all the data are written, that is:

274

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• after “N” write into UEDATX

• as soon as RWAL is cleared by hardware.

If the endpoint uses 2 banks, the second one can be read by the HOST while the current is

being written by the CPU. Then, when the CPU clears FIFOCON, the next bank may be already

ready (free) and TXINI is set immediately.

22.14.1.1

Abort

An “abort” stage can be produced by the host in some situations:

• In a control transaction: ZLP data OUT received during a IN stage,

• In an isochronous IN transaction: ZLP data OUT received on the OUT endpoint during a IN

stage on the IN endpoint

• ...

The KILLBK bit is used to kill the last “written” bank. The best way to manage this abort is to per-

form the following operations:

Table 22-1. Abort flow

Endpoint

Abort

Clear

Disable the TXINI interrupt.

UEIENX.

TXINE

Abort is based on the fact

No

NBUSYBK

=0

that no banks are busy,

meaning that nothing has to

be sent.

Yes

Kill the last written

bank.

Endpoint

reset

KILLBK=1

Wait for the end of the

procedure.

Yes

KILLBK=1

No

Abort done

22.15 Isochronous mode

22.15.1 Underflow

An underflow can occur during IN stage if the host attempts to read a bank which is empty. In

this situation, the UNDERFI interrupt is triggered.

An underflow can also occur during OUT stage if the host send a packet while the banks are

already full. Typically, he CPU is not fast enough. The packet is lost.

It is not possible to have underflow error during OUT stage, in the CPU side, since the CPU

should read only if the bank is ready to give data (RXOUTI=1 or RWAL=1)

275

7766E–AVR–04/10

ATmega16U4/ATmega32U4

22.15.2 CRC Error

A CRC error can occur during OUT stage if the USB controller detects a bad received packet. In

this situation, the STALLEDI interrupt is triggered. This does not prevent the RXOUTI interrupt

from being triggered.

22.16 Overflow

In Control, Isochronous, Bulk or Interrupt Endpoint, an overflow can occur during OUT stage, if

the host attempts to write in a bank that is too small for the packet. In this situation, the OVERFI

interrupt is triggered (if enabled). The packet is acknowledged and the RXOUTI interrupt is also

triggered (if enabled). The bank is filled with the first bytes of the packet.

It is not possible to have overflow error during IN stage, in the CPU side, since the CPU should

write only if the bank is ready to access data (TXINI=1 or RWAL=1).

22.17 Interrupts

The next figure shows all the interrupts sources:

Figure 22-4. USB Device Controller Interrupt System

UPRSMI

UDINT.6

UPRSME

UDIEN.6

EORSMI

UDINT.5

EORSME

UDIEN.5

WAKEUPI

UDINT.4

WAKEUPE

UDIEN.4

USB Device

Interrupt

EORSTI

UDINT.3

EORSTE

UDIEN.3

SOFI

UDINT.2

SOFE

UDIEN.2

SUSPI

UDINT.0

SUSPE

UDIEN.0

There are 2 kind of interrupts: processing (i.e. their generation are part of the normal processing)

and exception (errors).

Processing interrupts are generated when:

• VBUS plug-in detection (insert, remove)(VBUSTI)

• Upstream resume(UPRSMI)

• End of resume(EORSMI)

• Wake up(WAKEUPI)

• End of reset (Speed Initialization)(EORSTI)

• Start of frame(SOFI, if FNCERR=0)

• Suspend detected after 3 ms of inactivity(SUSPI)

Exception Interrupts are generated when:

• CRC error in frame number of SOF(SOFI, FNCERR=1)

276

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 22-5. USB Device Controller Endpoint Interrupt System

Endpoint 6

Endpoint 5

Endpoint 4

Endpoint 3

Endpoint 2

Endpoint 1

Endpoint 0

OVERFI

UESTAX.6

UNDERFI

UESTAX.5

FLERRE

UEIENX.7

NAKINI

UEINTX.6

NAKINE

UEIENX.6

NAKOUTI

UEINTX.4

TXSTPE

Endpoint Interrupt

UEIENX.4

RXSTPI

EPINT

UEINTX.3

UEINT.X

TXOUTE

UEIENX.3

RXOUTI

UEINTX.2

RXOUTE

UEIENX.2

STALLEDI

UEINTX.1

STALLEDE

UEIENX.1

TXINI

UEINTX.0

TXINE

UEIENX.0

Processing interrupts are generated when:

• Ready to accept IN data(EPINTx, TXINI=1)

• Received OUT data(EPINTx, RXOUTI=1)

• Received SETUP(EPINTx, RXSTPI=1)

Exception Interrupts are generated when:

• Stalled packet(EPINTx, STALLEDI=1)

• CRC error on OUT in isochronous mode(EPINTx, STALLEDI=1)

• Overflow in isochronous mode(EPINTx, OVERFI=1)

• Underflow in isochronous mode(EPINTx, UNDERFI=1)

• NAK IN sent(EPINTx, NAKINI=1)

• NAK OUT sent(EPINTx, NAKOUTI=1)

22.18 Registers

22.18.1 USB device general registers

Bit

7

-

6

-

5

-

4

-

3

2

1

0

RSTCPU

LSM

R/W

0

RMWKUP DETACH

UDCON

Read/Write

Initial Value

R

0

R

0

R

0

R

0

R

0

R/W

0

R/W

1

277

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• 7-4 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 3 - RSTCPU - USB Reset CPU bit

Set this bit to 1 by firmware in order to reset the CPU on the detection of a USB End of Reset

signal (without disabling the USB controller and Attached state). This bit is reset when the USB

controller is disabled, but is not affected by the CPU reset generated after a USB End of Reset

(remains enabled).

• 2 - LSM - USB Device Low Speed Mode Selection

When configured USB is configured in device mode, this bit allows to select the USB the USB

Low Speed or Full Speed Mod.

Clear to select full speed mode (D+ internal pull-up will be activate with the ATTACH bit will be

set).

Set to select low speed mode (D- internal pull-up will be activate with the ATTACH bit will be

set). This bit has no effect when the USB interface is configured in HOST mode.

• 1- RMWKUP - Remote Wake-up Bit

Set to send an “upstream-resume” to the host for a remote wake-up (the SUSPI bit must be set).

Cleared by hardware when signalling finished. Clearing by software has no effect.

See Section 22.10, page 270 for more details.

• 0 - DETACH - Detach Bit

Set to physically detach de device (disconnect internal pull-up on D+ or D-).

Clear to reconnect the device. See Section 22.9, page 269 for more details.

Bit

7

-

6

5

4

3

2

1

0

UPRSMI

EORSMI WAKEUPI EORSTI

SOFI

-

SUSPI

UDINT

Read/Write

Initial Value

0

• 7 - Reserved

The value read from this bits is always 0. Do not set this bit.

• 6 - UPRSMI - Upstream Resume Interrupt Flag

Set by hardware when the USB controller is sending a resume signal called “Upstream

Resume”. This triggers an USB interrupt if UPRSME is set.

Shall be cleared by software (USB clocks must be enabled before). Setting by software has no

effect.

• 5 - EORSMI - End Of Resume Interrupt Flag

Set by hardware when the USB controller detects a good “End Of Resume” signal initiated by

the host. This triggers an USB interrupt if EORSME is set.

Shall be cleared by software. Setting by software has no effect.

278

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• 4 - WAKEUPI - Wake-up CPU Interrupt Flag

Set by hardware when the USB controller is re-activated by a filtered non-idle signal from the

lines (not by an upstream resume). This triggers an interrupt if WAKEUPE is set.

Shall be cleared by software (USB clock inputs must be enabled before). Setting by software

has no effect.

See Section 22.8, page 269 for more details.

• 3 - EORSTI - End Of Reset Interrupt Flag

Set by hardware when an “End Of Reset” has been detected by the USB controller. This triggers

an USB interrupt if EORSTE is set.

Shall be cleared by software. Setting by software has no effect.

• 2 - SOFI - Start Of Frame Interrupt Flag

Set by hardware when an USB “Start Of Frame” PID (SOF) has been detected (every 1 ms).

This triggers an USB interrupt if SOFE is set.

• 1 - Reserved

The value read from this bits is always 0. Do not set this bit

• 0 - SUSPI - Suspend Interrupt Flag

Set by hardware when an USB “Suspend” ‘idle bus for 3 frame periods: a J state for 3 ms) is

detected. This triggers an USB interrupt if SUSPE is set.

Shall be cleared by software. Setting by software has no effect.

See Section 22.8, page 269 for more details.

The interrupt bits are set even if their corresponding ‘Enable’ bits is not set.

Bit

7

6

5

4

3

2

1

0

-

UPRSME EORSME WAKEUPE EORSTE

SOFE

-

SUSPE

UDIEN

Read/Write

Initial Value

0

• 7 - Reserved

The value read from this bits is always 0. Do not set this bit.

• 6 - UPRSME - Upstream Resume Interrupt Enable Bit

Set to enable the UPRSMI interrupt.

Clear to disable the UPRSMI interrupt.

• 5 - EORSME - End Of Resume Interrupt Enable Bit

Set to enable the EORSMI interrupt.

Clear to disable the EORSMI interrupt.

• 4 - WAKEUPE - Wake-up CPU Interrupt Enable Bit

Set to enable the WAKEUPI interrupt.

Clear to disable the WAKEUPI interrupt.

279

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• 3 - EORSTE - End Of Reset Interrupt Enable Bit

Set to enable the EORSTI interrupt. This bit is set after a reset.

Clear to disable the EORSTI interrupt.

• 2 - SOFE - Start Of Frame Interrupt Enable Bit

Set to enable the SOFI interrupt.

Clear to disable the SOFI interrupt.

• 1 - Reserved

The value read from this bits is always 0. Do not set this bit

• 0 - SUSPE - Suspend Interrupt Enable Bit

Set to enable the SUSPI interrupt.

Clear to disable the SUSPI interrupt.

Bit

7

6

5

4

3

UADD6:0

R/W

2

1

0

ADDEN

UDADDR

Read/Write

W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

Initial Val-

ue

0

• 7 - ADDEN - Address Enable Bit

Set to activate the UADD (USB address).

Cleared by hardware. Clearing by software has no effect.

See Section 22.7, page 268 for more details.

• 6-0 - UADD6:0 - USB Address Bits

Load by software to configure the device address.

Bit

7

-

6

-

5

-

4

-

3

-

2

1

0

FNUM10:8

UDFNUMH

Read/Write

Initial Value

R

0

R

0

R

0

R

0

R

0

R

0

R

0

R

0

• 7-3 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 2-0 - FNUM10:8 - Frame Number Upper Value

Set by hardware. These bits are the 3 MSB of the 11-bits Frame Number information. They are

provided in the last received SOF packet. FNUM is updated if a corrupted SOF is received.

Bit

7

6

5

4

3

2

1

0

FNUM7:0

UDFNUML

Read/Write

Initial Value

R

0

R

0

R

0

R

0

R

0

R

0

R

0

R

0

280

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Frame Number Lower Value

Set by hardware. These bits are the 8 LSB of the 11-bits Frame Number information.

Bit

7

6

5

4

FNCERR

R

3

2

1

0

-

UDMFN

Read/W

rite

Initial

Value

0

• 7-5 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 4 - FNCERR -Frame Number CRC Error Flag

Set by hardware when a corrupted Frame Number in start of frame packet is received.

This bit and the SOFI interrupt are updated at the same time.

• 3-0 - Reserved

The value read from these bits is always 0. Do not set these bits.

22.18.2 USB device endpoint registers

Bit

7

-

6

-

5

-

4

-

3

-

2

1

EPNUM2:0

R/W

0

UENUM

Read/Write

Initial Value

R

0

R

0

R

0

R

0

R

0

R/W

0

R/W

0

• 7-3 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 2-0 - EPNUM2:0 Endpoint Number Bits

Load by software to select the number of the endpoint which shall be accessed by the CPU. See

Section 22.5, page 267 for more details.

EPNUM = 111b is forbidden.

Bit

7

-

6

EPRST6

R/W

5

EPRST5

R/W

4

EPRST4

R/W

3

EPRST3

R/W

2

EPRST2

R/W

1

EPRST1

R/W

0

EPRST0

R/W

UERST

Read/Write

Initial Value

R

0

• 7 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 6-0 - EPRST6:0 - Endpoint FIFO Reset Bits

Set to reset the selected endpoint FIFO prior to any other operation, upon hardware reset or

when an USB bus reset has been received. See Section 22.3, page 266 for more information

281

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Then, clear by software to complete the reset operation and start using the endpoint.

Bit

7

-

6

-

5

4

3

RSTDT

W

2

-

1

-

0

EPEN

R/W

0

STALLRQ STALLRQC

UECONX

Read/Write

Initial Value

R

0

R

0

W

0

W

0

R

0

R

0

• 7-6 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 5 - STALLRQ - STALL Request Handshake Bit

Set to request a STALL answer to the host for the next handshake.

Cleared by hardware when a new SETUP is received. Clearing by software has no effect.

See Section 22.11, page 270 for more details.

• 4 - STALLRQC - STALL Request Clear Handshake Bit

Set to disable the STALL handshake mechanism.

Cleared by hardware immediately after the set. Clearing by software has no effect.

See Section 22.11, page 270 for more details.

3

• RSTDT - Reset Data Toggle Bit

Set to automatically clear the data toggle sequence:

For OUT endpoint: the next received packet will have the data toggle 0.

For IN endpoint: the next packet to be sent will have the data toggle 0.

Cleared by hardware instantaneously. The firmware does not have to wait that the bit is cleared.

Clearing by software has no effect.

• 2 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 1 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 0 - EPEN - Endpoint Enable Bit

Set to enable the endpoint according to the device configuration. Endpoint 0 shall always be

enabled after a hardware or USB reset and participate in the device configuration.

Clear this bit to disable the endpoint. See Section 22.6, page 267 for more details.

Bit

7

6

5

-

4

-

3

-

2

-

1

-

0

EPDIR

R/W

0

EPTYPE1:0

UECFG0X

Read/Write

Initial Value

R/W

0

R/W

0

R

0

R

0

R

0

R

0

R

0

• 7-6 - EPTYPE1:0 - Endpoint Type Bits

Set this bit according to the endpoint configuration:

282

7766E–AVR–04/10

ATmega16U4/ATmega32U4

00b: Control10b: Bulk

01b: Isochronous11b: Interrupt

• 5-1 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 0 - EPDIR - Endpoint Direction Bit

Set to configure an IN direction for bulk, interrupt or isochronous endpoints.

Clear to configure an OUT direction for bulk, interrupt, isochronous or control endpoints.

Bit

7

-

6

5

EPSIZE2:0

R/W

4

3

2

1

ALLOC

R/W

0

-

EPBK1:0

UECFG1X

Read/Write

Initial Value

R

0

R/W

0

R/W

0

R/W

0

R/W

0

R

0

• 7 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 6-4 - EPSIZE2:0 - Endpoint Size Bits

Set this bit according to the endpoint size:

000b: 8 bytes100b: 128 bytes

001b: 16 bytes101b: 256 bytes

010b: 32 bytes110b: 512 bytes

011b: 64 bytes111b: Reserved. Do not use this configuration.

• 3-2 - EPBK1:0 - Endpoint Bank Bits

Set this field according to the endpoint size:

00b: One bank

01b: Double bank

1xb: Reserved. Do not use this configuration.

• 1 - ALLOC - Endpoint Allocation Bit

Set this bit to allocate the endpoint memory.

Clear to free the endpoint memory.

See Section 22.6, page 267 for more details.

• 0 - Reserved

The value read from these bits is always 0. Do not set these bits.

Bit

7

6

OVERFI

R/W

0

5

UNDERFI

R/W

4

-

3

2

1

0

CFGOK

DTSEQ1:0

NBUSYBK1:0

UESTA0X

Read/Write

Initial Value

R

0

R/W

0

R

0

R

0

R

0

R

0

283

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• 7 - CFGOK - Configuration Status Flag

Set by hardware when the endpoint X size parameter (EPSIZE) and the bank parametrization

(EPBK) are correct compared to the max FIFO capacity and the max number of allowed bank.

This bit is updated when the bit ALLOC is set.

If this bit is cleared, the user should reprogram the UECFG1X register with correct EPSIZE and

EPBK values.

• 6 - OVERFI - Overflow Error Interrupt Flag

Set by hardware when an overflow error occurs in an isochronous endpoint. An interrupt

(EPINTx) is triggered (if enabled).

See Section 22.15, page 275 for more details.

Shall be cleared by software. Setting by software has no effect.

• 5 - UNDERFI - Flow Error Interrupt Flag

Set by hardware when an underflow error occurs in an isochronous endpoint. An interrupt

(EPINTx) is triggered (if enabled).

See Section 22.15, page 275 for more details.

Shall be cleared by software. Setting by software has no effect.

• 4 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 3-2 - DTSEQ1:0 - Data Toggle Sequencing Flag

Set by hardware to indicate the PID data of the current bank:

00b Data0

01b Data1

1xb Reserved.

For OUT transfer, this value indicates the last data toggle received on the current bank.

For IN transfer, it indicates the Toggle that will be used for the next packet to be sent. This is not

relative to the current bank.

• 1-0 - NBUSYBK1:0 - Busy Bank Flag

Set by hardware to indicate the number of busy bank.

For IN endpoint, it indicates the number of busy bank(s), filled by the user, ready for IN transfer.

For OUT endpoint, it indicates the number of busy bank(s) filled by OUT transaction from the

host.

00b All banks are free

01b 1 busy bank

10b 2 busy banks

11b Reserved.

284

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Bit

7

-

6

-

5

-

4

-

3

-

2

1

0

CTRLDIR

CURRBK1:0

UESTA1X

Read/Write

Initial Value

R

0

R

0

R

0

R

0

R

0

R

0

R

0

• 7-3 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 2 - CTRLDIR - Control Direction (Flag, and bit for debug purpose)

Set by hardware after a SETUP packet, and gives the direction of the following packet:

- 1 for IN endpoint

- 0 for OUT endpoint.

Can not be set or cleared by software.

• 1-0 - CURRBK1:0 - Current Bank (all endpoints except Control endpoint) Flag

Set by hardware to indicate the number of the current bank:

00b Bank0

01b Bank1

1xb Reserved.

Can not be set or cleared by software.

Bit

7

FIFOCON

R/W

6

NAKINI

R/W

0

5

4

3

RXSTPI

R/W

0

2

1

0

TXINI

R/W

0

RWAL NAKOUTI

RXOUTI STALLEDI

UEINTX

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

• 7 - FIFOCON - FIFO Control Bit

For OUT and SETUP Endpoint:

Set by hardware when a new OUT message is stored in the current bank, at the same time than

RXOUT or RXSTP.

Clear to free the current bank and to switch to the following bank. Setting by software has no

effect.

For IN Endpoint:

Set by hardware when the current bank is free, at the same time than TXIN.

Clear to send the FIFO data and to switch the bank. Setting by software has no effect.

• 6 - NAKINI - NAK IN Received Interrupt Flag

Set by hardware when a NAK handshake has been sent in response of a IN request from the

host. This triggers an USB interrupt if NAKINE is sent.

Shall be cleared by software. Setting by software has no effect.

285

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• 5 - RWAL - Read/Write Allowed Flag

Set by hardware to signal:

- for an IN endpoint: the current bank is not full i.e. the firmware can push data into the FIFO,

- for an OUT endpoint: the current bank is not empty, i.e. the firmware can read data from the

FIFO.

The bit is never set if STALLRQ is set, or in case of error.

Cleared by hardware otherwise.

This bit shall not be used for the control endpoint.

• 4 - NAKOUTI - NAK OUT Received Interrupt Flag

Set by hardware when a NAK handshake has been sent in response of a OUT/PING request

from the host. This triggers an USB interrupt if NAKOUTE is sent.

Shall be cleared by software. Setting by software has no effect.

• 3 - RXSTPI - Received SETUP Interrupt Flag

Set by hardware to signal that the current bank contains a new valid SETUP packet. An inter-

rupt (EPINTx) is triggered (if enabled).

Shall be cleared by software to handshake the interrupt. Setting by software has no effect.

This bit is inactive (cleared) if the endpoint is an IN endpoint.

• 2 - RXOUTI / KILLBK - Received OUT Data Interrupt Flag

Set by hardware to signal that the current bank contains a new packet. An interrupt (EPINTx) is

triggered (if enabled).

Shall be cleared by software to handshake the interrupt. Setting by software has no effect.

Kill Bank IN Bit

Set this bit to kill the last written bank.

Cleared by hardware when the bank is killed. Clearing by software has no effect.

See page 275 for more details on the Abort.

• 1 - STALLEDI - STALLEDI Interrupt Flag

Set by hardware to signal that a STALL handshake has been sent, or that a CRC error has been

detected in a OUT isochronous endpoint.

Shall be cleared by software. Setting by software has no effect.

• 0 - TXINI - Transmitter Ready Interrupt Flag

Set by hardware to signal that the current bank is free and can be filled. An interrupt (EPINTx) is

triggered (if enabled).

Shall be cleared by software to handshake the interrupt. Setting by software has no effect.

This bit is inactive (cleared) if the endpoint is an OUT endpoint.

286

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Bit

7

6

5

-

4

3

2

1

0

TXINE

R/W

0

FLERRE NAKINE

NAKOUTE RXSTPE

RXOUTE STALLEDE

UEIENX

Read/Write

Initial Value

R/W

0

R/W

0

R

0

R/W

0

R/W

0

R/W

0

R/W

0

• 7 - FLERRE - Flow Error Interrupt Enable Flag

Set to enable an endpoint interrupt (EPINTx) when OVERFI or UNDERFI are sent.

Clear to disable an endpoint interrupt (EPINTx) when OVERFI or UNDERFI are sent.

• 6 - NAKINE - NAK IN Interrupt Enable Bit

Set to enable an endpoint interrupt (EPINTx) when NAKINI is set.

Clear to disable an endpoint interrupt (EPINTx) when NAKINI is set.

• 5 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 4 - NAKOUTE - NAK OUT Interrupt Enable Bit

Set to enable an endpoint interrupt (EPINTx) when NAKOUTI is set.

Clear to disable an endpoint interrupt (EPINTx) when NAKOUTI is set.

• 3 - RXSTPE - Received SETUP Interrupt Enable Flag

Set to enable an endpoint interrupt (EPINTx) when RXSTPI is sent.

Clear to disable an endpoint interrupt (EPINTx) when RXSTPI is sent.

• 2 - RXOUTE - Received OUT Data Interrupt Enable Flag

Set to enable an endpoint interrupt (EPINTx) when RXOUTI is sent.

Clear to disable an endpoint interrupt (EPINTx) when RXOUTI is sent.

• 1 - STALLEDE - Stalled Interrupt Enable Flag

Set to enable an endpoint interrupt (EPINTx) when STALLEDI is sent.

Clear to disable an endpoint interrupt (EPINTx) when STALLEDI is sent.

• 0 - TXINE - Transmitter Ready Interrupt Enable Flag

Set to enable an endpoint interrupt (EPINTx) when TXINI is sent.

Clear to disable an endpoint interrupt (EPINTx) when TXINI is sent.

Bit

7

DAT D7

R/W

0

6

DAT D6

R/W

0

5

DAT D5

R/W

0

4

DAT D4

R/W

0

3

DAT D3

R/W

0

2

DAT D2

R/W

0

1

DAT D1

R/W

0

DAT D0

R/W

0

UEDATX

Read/Write

Initial Value

• 7-0 - DAT7:0 -Data Bits

Set by the software to read/write a byte from/to the endpoint FIFO selected by EPNUM.

287

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Bit

7

6

5

4

3

2

1

0

-

BYCT D10 BYCT D9 BYCT D8 UEBCHX

Read/Write

Initial Value

R

0

R

0

R

0

R

0

R

0

R

0

R

0

R

0

• 7-3 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 2-0 - BYCT10:8 - Byte count (high) Bits

Set by hardware. This field is the MSB of the byte count of the FIFO endpoint. The LSB part is

provided by the UEBCLX register.

Bit

7

6

5

4

3

2

1

0

BYCT D7 BYCT D6 BYCT D5 BYCT D4 BYCT D3 BYCT D2 BYCT D1 BYCT D0 UEBCLX

Read/Write

Initial Value

R

0

R

0

R

0

R

0

R

0

R

0

R

0

R

0

• 7-0 - BYCT7:0 - Byte Count (low) Bits

Set by the hardware. BYCT10:0 is:

- (for IN endpoint) increased after each writing into the endpoint and decremented after each

byte sent,

- (for OUT endpoint) increased after each byte sent by the host, and decremented after each

byte read by the software.

Bit

7

-

6

5

4

3

2

1

0

EPINT D6 EPINT D5 EPINT D4 EPINT D3 EPINT D2 EPINT D1 EPINT D0 UEINT

Read/Write

Initial Value

R

0

R

0

R

0

R

0

R

0

R

0

R

0

R

0

• 7 - Reserved

The value read from these bits is always 0. Do not set these bits.

• 6-0 - EPINT6:0 - Endpoint Interrupts Bits

Set by hardware when an interrupt is triggered by the UEINTX register and if the corresponding

endpoint interrupt enable bit is set.

Cleared by hardware when the interrupt source is served.

288

7766E–AVR–04/10

ATmega16U4/ATmega32U4

23. Analog Comparator

The Analog Comparator compares the input values on the positive pin AIN+ and negative pin

AIN-. When the voltage on the positive pin AIN+ is higher than the voltage on the negative pin

AIN-, the Analog Comparator output, ACO, is set. The comparator’s output can be set to trigger

the Timer/Counter1 Input Capture function. In addition, the comparator can trigger a separate

interrupt, exclusive to the Analog Comparator. The user can select Interrupt triggering on com-

parator output rise, fall or toggle. A block diagram of the comparator and its surrounding logic is

shown in Figure 23-1. AIN+ can be connected either to the AIN0 (PE6) pin, or to the internal

Bandgap reference. AIN- can only be connected to the ADC multiplexer.

The Power Reduction ADC bit, PRADC, in “Power Reduction Register 0 - PRR0” on page 46

must be disabled by writing a logical zero to be able to use the ADC input MUX.

Figure 23-1. Analog Comparator Block Diagram⁽²⁾

BANDGAP

REFERENCE

ACBG

AIN+

AIN-

BANDGAP

REFERENCE

ACME

ADEN

ADC MULTIPLEXER

OUTPUT⁽¹⁾

Notes: 1. See Table 23-2 on page 291.

2. Refer to “Pinout ATmega16U4/ATmega32U4” on page 3 and Table 10-3 on page 72 for Ana-

log Comparator pin placement.

23.0.1

ADC Control and Status Register B – ADCSRB

Bit

7

6

5

4

–

R

0

3

-

2

1

0

–

ACME

–

ADTS2

R/W

0

ADTS1

R/W

0

ADTS0

R/W

0

ADCSRB

Read/Write

Initial Value

R

0

R/W

0

R

0

R

0

• Bit 6 – ACME: Analog Comparator Multiplexer Enable

When this bit is written logic one and the ADC is switched off (ADEN in ADCSRA is zero), the

ADC multiplexer is connected to the negative input to the Analog Comparator. When this bit is

written logic zero, the Bandgap reference is connected to the negative input of the Analog Com-

parator (See “Internal Voltage Reference” on page 54.). For a detailed description of this bit, see

“Analog Comparator Multiplexed Input” on page 291.

23.0.2

Analog Comparator Control and Status Register – ACSR

Bit

7

6

5

4

3

2

1

0

ACD

ACBG

ACO

ACI

ACIE

R/W

0

ACIC

R/W

0

ACIS1

R/W

0

ACIS0

R/W

0

ACSR

Read/Write

Initial Value

R/W

0

R/W

0

R

R/W

0

N/A

289

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 7 – ACD: Analog Comparator Disable

When this bit is written logic one, the power to the Analog Comparator is switched off. This bit

can be set at any time to turn off the Analog Comparator. This will reduce power consumption in

Active and Idle mode. When changing the ACD bit, the Analog Comparator Interrupt must be

disabled by clearing the ACIE bit in ACSR. Otherwise an interrupt can occur when the bit is

changed.

• Bit 6 – ACBG: Analog Comparator Bandgap Select

When this bit is set, a fixed bandgap reference voltage replaces the positive input to the Analog

Comparator. When this bit is cleared, AIN0 is applied to the positive input of the Analog Compar-

ator. See “Internal Voltage Reference” on page 54.

• Bit 5 – ACO: Analog Comparator Output

The output of the Analog Comparator is synchronized and then directly connected to ACO. The

synchronization introduces a delay of 1 - 2 clock cycles.

• Bit 4 – ACI: Analog Comparator Interrupt Flag

This bit is set by hardware when a comparator output event triggers the interrupt mode defined

by ACIS1 and ACIS0. The Analog Comparator interrupt routine is executed if the ACIE bit is set

and the I-bit in SREG is set. ACI is cleared by hardware when executing the corresponding inter-

rupt handling vector. Alternatively, ACI is cleared by writing a logic one to the flag.

• Bit 3 – ACIE: Analog Comparator Interrupt Enable

When the ACIE bit is written logic one and the I-bit in the Status Register is set, the Analog Com-

parator interrupt is activated. When written logic zero, the interrupt is disabled.

• Bit 2 – ACIC: Analog Comparator Input Capture Enable

When written logic one, this bit enables the input capture function in Timer/Counter1 to be trig-

gered by the Analog Comparator. The comparator output is in this case directly connected to the

input capture front-end logic, making the comparator utilize the noise canceler and edge select

features of the Timer/Counter1 Input Capture interrupt. When written logic zero, no connection

between the Analog Comparator and the input capture function exists. To make the comparator

trigger the Timer/Counter1 Input Capture interrupt, the ICIE1 bit in the Timer Interrupt Mask

Register (TIMSK1) must be set.

• Bits 1, 0 – ACIS1, ACIS0: Analog Comparator Interrupt Mode Select

These bits determine which comparator events that trigger the Analog Comparator interrupt. The

different settings are shown in Table 23-1.

Table 23-1. ACIS1/ACIS0 Settings

ACIS1

ACIS0

Interrupt Mode

0

1

0

1

0

1

Comparator Interrupt on Output Toggle.

Reserved

Comparator Interrupt on Falling Output Edge.

Comparator Interrupt on Rising Output Edge.

When changing the ACIS1/ACIS0 bits, the Analog Comparator Interrupt must be disabled by

clearing its Interrupt Enable bit in the ACSR Register. Otherwise an interrupt can occur when the

bits are changed.

290

7766E–AVR–04/10

ATmega16U4/ATmega32U4

23.1 Analog Comparator Multiplexed Input

It is possible to select any of the ADC13..0 pins to replace the negative input to the Analog Com-

parator. The ADC multiplexer is used to select this input, and consequently, the ADC must be

switched off to utilize this feature. If the Analog Comparator Multiplexer Enable bit (ACME in

ADCSRB) is set and the ADC is switched off (ADEN in ADCSRA is zero), and MUX2..0 in

ADMUX select the input pin to replace the negative input to the Analog Comparator, as shown in

Table 23-2. If ACME is cleared or ADEN is set, the Bandgap reference is applied to the negative

input to the Analog Comparator.

Table 23-2. Analog Comparator Multiplexed Input

ACME

ADEN

MUX2..0

xxx

Analog Comparator Negative Input

0

1

x

1

0

Bandgap Ref.

ADC0

xxx

000

001

ADC1

010

N/A

011

100

ADC4

ADC5

ADC6

ADC7

101

110

111

23.1.1

Digital Input Disable Register 1 – DIDR1

Bit

7

–

R

0

6

–

R

0

5

–

R

0

4

–

R

0

3

–

R

0

2

–

R

0

1

–

R

0

AIN0D

R/W

0

DIDR1

Read/Write

Initial Value

• Bit 0 – AIN0D: AIN0 Digital Input Disable

When this bit is written logic one, the digital input buffer on the AIN0 pin is disabled. The corre-

sponding PIN Register bit will always read as zero when this bit is set. When an analog signal is

applied to the AIN0 pin and the digital input from this pin is not needed, this bit should be written

logic one to reduce power consumption in the digital input buffer.

291

7766E–AVR–04/10

ATmega16U4/ATmega32U4

24. Analog to Digital Converter - ADC

24.1 Features

• 10/8-bit Resolution

• 0.5 LSB Integral Non-linearity

•

2 LSB Absolute Accuracy

• 65 - 260 µs Conversion Time

• Up to 15 kSPS at Maximum Resolution

• Twelve Multiplexed Single-Ended Input Channels

• One Differential amplifier providing gain of 1x - 10x - 40x - 200x

• Temperature sensor

• Optional Left Adjustment for ADC Result Readout

• 0 - V_CCADC Input Voltage Range

• Selectable 2.56 V ADC Reference Voltage

• Free Running or Single Conversion Mode

• ADC Start Conversion by Auto Triggering on Interrupt Sources

• Interrupt on ADC Conversion Complete

• Sleep Mode Noise Canceler

The ATmega16U4/ATmega32U4 features a 10-bit successive approximation ADC. The ADC is

connected to an 12-channel Analog Multiplexer which allows six single-ended voltage inputs

constructed from several pins of Port B, D and F. The single-ended voltage inputs refer to

0V (GND).

The device also supports 32 differential voltage input combinations, thanks to a differential

amplifier equipped with a programmable gain stage, providing amplification steps of 0 dB (1x),

10 dB (10x), 16dB (40x) or 23dB (200x) on the differential input voltage before the A/D conver-

sion. Two differential analog input channels share a common negative terminal (ADC0/ADC1),

while any other ADC input can be selected as the positive input terminal. If 1x, 10x or 40x gain is

used, 8-bit resolution can be expected. If 200x gain is used, 7-bit resolution can be expected.

The ADC contains a Sample and Hold circuit which ensures that the input voltage to the ADC is

held at a constant level during conversion. A block diagram of the ADC is shown in Figure 24-1.

The ADC has a separate analog supply voltage pin, AV_CC. AV_CCmust not differ more than

0.3V from V_CC. See the paragraph “ADC Noise Canceler” on page 301 on how to connect this

pin.

Internal reference voltages of nominally 2.56V or AV_CCare provided On-chip. The voltage refer-

ence may be externally decoupled at the AREF pin by a capacitor for better noise performance.

292

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 24-1. Analog to Digital Converter Block Schematic

ADC CONVERSION

COMPLETE IRQ

INTERRUPT

FLAGS

ADTS[3:0]

8-BIT DATA BUS

15

ADC DATA REGISTER

(ADCH/ADCL)

0

ADC MULTIPLEXER

SELECT (ADMUX)

ADC CTRL. & STATUS

REGISTER (ADCSRA)

TRIGGER

SELECT

MUX DECODER

PRESCALER

START

CONVERSION LOGIC

AVCC

INTERNAL

SAMPLE & HOLD

COMPARATOR

REFERENCE

AREF

GND

10-BIT DAC

-

+

ADHSM

BANDGAP

REFERENCE

TEMPERATURE

SENSOR

SINGLE ENDED / DIFFERENTIAL SELECTION

ADC13

ADC12

ADC11

ADC10

ADC9

ADC8

ADC7

ADC6

ADC5

ADC4

ADC1

ADC0

POS.

INPUT

MUX

ADC MULTIPLEXER

OUTPUT

DIFFERENTIAL

AMPLIFIER

+

-

NEG.

INPUT

MUX

293

7766E–AVR–04/10

ATmega16U4/ATmega32U4

24.2 Operation

The ADC converts an analog input voltage to a 10-bit digital value through successive approxi-

mation. The minimum value represents GND and the maximum value represents the voltage on

the AREF pin minus 1 LSB. Optionally, AV_CCor an internal 2.56V reference voltage may be con-

nected to the AREF pin by writing to the REFSn bits in the ADMUX Register. The internal

voltage reference may thus be decoupled by an external capacitor at the AREF pin to improve

noise immunity.

The analog input channel and differential gain are selected by writing to the MUX bits in

ADMUX. Any of the ADC input pins, as well as GND and a fixed bandgap voltage reference, can

be selected as single ended inputs to the ADC. A selection of ADC input pins can be selected as

positive and negative inputs to the differential amplifier.

The ADC is enabled by setting the ADC Enable bit, ADEN in ADCSRA. Voltage reference and

input channel selections will not go into effect until ADEN is set. The ADC does not consume

power when ADEN is cleared, so it is recommended to switch off the ADC before entering power

saving sleep modes.

The ADC generates a 10-bit result which is presented in the ADC Data Registers, ADCH and

ADCL. By default, the result is presented right adjusted, but can optionally be presented left

adjusted by setting the ADLAR bit in ADMUX.

If the result is left adjusted and no more than 8-bit precision is required, it is sufficient to read

ADCH. Otherwise, ADCL must be read first, then ADCH, to ensure that the content of the Data

Registers belongs to the same conversion. Once ADCL is read, ADC access to Data Registers

is blocked. This means that if ADCL has been read, and a conversion completes before ADCH is

read, neither register is updated and the result from the conversion is lost. When ADCH is read,

ADC access to the ADCH and ADCL Registers is re-enabled.

The ADC has its own interrupt which can be triggered when a conversion completes. The ADC

access to the Data Registers is prohibited between reading of ADCH and ADCL, the interrupt

will trigger even if the result is lost.

24.3 Starting a Conversion

A single conversion is started by writing a logical one to the ADC Start Conversion bit, ADSC.

This bit stays high as long as the conversion is in progress and will be cleared by hardware

when the conversion is completed. If a different data channel is selected while a conversion is in

progress, the ADC will finish the current conversion before performing the channel change.

Alternatively, a conversion can be triggered automatically by various sources. Auto Triggering is

enabled by setting the ADC Auto Trigger Enable bit, ADATE in ADCSRA. The trigger source is

selected by setting the ADC Trigger Select bits, ADTS in ADCSRB (See description of the ADTS

bits for a list of the trigger sources). When a positive edge occurs on the selected trigger signal,

the ADC prescaler is reset and a conversion is started. This provides a method of starting con-

versions at fixed intervals. If the trigger signal is still set when the conversion completes, a new

conversion will not be started. If another positive edge occurs on the trigger signal during con-

version, the edge will be ignored. Note that an interrupt flag will be set even if the specific

interrupt is disabled or the Global Interrupt Enable bit in SREG is cleared. A conversion can thus

be triggered without causing an interrupt. However, the interrupt flag must be cleared in order to

trigger a new conversion at the next interrupt event.

294

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 24-2. ADC Auto Trigger Logic

ADTS[2:0]

PRESCALER

CLK_ADC

START

ADATE

ADIF

SOURCE 1

.

CONVERSION

LOGIC

EDGE

DETECTOR

SOURCE n

ADSC

Using the ADC Interrupt Flag as a trigger source makes the ADC start a new conversion as soon

as the ongoing conversion has finished. The ADC then operates in Free Running mode, con-

stantly sampling and updating the ADC Data Register. The first conversion must be started by

writing a logical one to the ADSC bit in ADCSRA. In this mode the ADC will perform successive

conversions independently of whether the ADC Interrupt Flag, ADIF is cleared or not.

If Auto Triggering is enabled, single conversions can be started by writing ADSC in ADCSRA to

one. ADSC can also be used to determine if a conversion is in progress. The ADSC bit will be

read as one during a conversion, independently of how the conversion was started.

24.4 Prescaling and Conversion Timing

Figure 24-3. ADC Prescaler

ADEN

START

Reset

7-BIT ADC PRESCALER

CK

ADPS0

ADPS1

ADPS2

ADC CLOCK SOURCE

By default, the successive approximation circuitry requires an input clock frequency between 50

kHz and 200 kHz to get maximum resolution. If a lower resolution than 10 bits is needed, the

input clock frequency to the ADC can be higher than 200 kHz to get a higher sample rate. Alter-

natively, setting the ADHSM bit in ADCSRB allows an increased ADC clock frequency at the

expense of higher power consumption.

The ADC module contains a prescaler, which generates an acceptable ADC clock frequency

from any CPU frequency above 100 kHz. The prescaling is set by the ADPS bits in ADCSRA.

The prescaler starts counting from the moment the ADC is switched on by setting the ADEN bit

295

7766E–AVR–04/10

ATmega16U4/ATmega32U4

in ADCSRA. The prescaler keeps running for as long as the ADEN bit is set, and is continuously

reset when ADEN is low.

When initiating a single ended conversion by setting the ADSC bit in ADCSRA, the conversion

starts at the following rising edge of the ADC clock cycle. See “Differential Channels” on page

297 for details on differential conversion timing.

A normal conversion takes 13 ADC clock cycles. The first conversion after the ADC is switched

on (ADEN in ADCSRA is set) takes 25 ADC clock cycles in order to initialize the analog circuitry.

The actual sample-and-hold takes place 1.5 ADC clock cycles after the start of a normal conver-

sion and 13.5 ADC clock cycles after the start of an first conversion. When a conversion is

complete, the result is written to the ADC Data Registers, and ADIF is set. In Single Conversion

mode, ADSC is cleared simultaneously. The software may then set ADSC again, and a new

conversion will be initiated on the first rising ADC clock edge.

When Auto Triggering is used, the prescaler is reset when the trigger event occurs. This assures

a fixed delay from the trigger event to the start of conversion. In this mode, the sample-and-hold

takes place two ADC clock cycles after the rising edge on the trigger source signal. Three addi-

tional CPU clock cycles are used for synchronization logic.

In Free Running mode, a new conversion will be started immediately after the conversion com-

pletes, while ADSC remains high. For a summary of conversion times, see Table 24-1.

Figure 24-4. ADC Timing Diagram, First Conversion (Single Conversion Mode)

Conversion

Cycle Number

1

2

12

13

14

15

16

17

18

19

20

21

22

23

24

25

1

2

3

ADC Clock

ADEN

ADSC

ADIF

Sign and MSB of Result

ADCH

ADCL

LSB of Result

MUX

MUX and REFS

Update

Conversion

Complete

and REFS

Update

Sample & Hold

Figure 24-5. ADC Timing Diagram, Single Conversion

One Conversion

Next Conversion

1

2

3

4

5

6

7

8

9

10

11

12

13

1

2

3

Cycle Number

ADC Clock

ADSC

ADIF

ADCH

Sign and MSB of Result

LSB of Result

ADCL

Sample & Hold

Conversion

Complete

MUX and REFS

Update

MUX and REFS

Update

296

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 24-6. ADC Timing Diagram, Auto Triggered Conversion

One Conversion

Next Conversion

1

2

3

4

5

6

7

8

9

10

11

12

13

1

2

Cycle Number

ADC Clock

Trigger

Source

ADATE

ADIF

ADCH

ADCL

Sign and MSB of Result

LSB of Result

Sample &

Hold

Prescaler

Reset

Conversion

Complete

Prescaler

Reset

MUX and REFS

Update

Figure 24-7. ADC Timing Diagram, Free Running Conversion

One Conversion

Next Conversion

11

12

13

1

2

3

4

Cycle Number

ADC Clock

ADSC

ADIF

ADCH

ADCL

Sign and MSB of Result

LSB of Result

Sample & Hold

Conversion

Complete

MUX and REFS

Update

Table 24-1. ADC Conversion Time

Normal

First

Conversion

Conversion,

Single Ended

Auto Triggered

Convertion

Condition

Sample & Hold

(Cycles from Start of Convention)

14.5

1.5

13

2

Conversion Time

(Cycles)

25

13.5

24.4.1

Differential Channels

When using differential channels, certain aspects of the conversion need to be taken into

consideration.

Differential conversions are synchronized to the internal clock CK_ADC2equal to half the ADC

clock frequency. This synchronization is done automatically by the ADC interface in such a way

that the sample-and-hold occurs at a specific phase of CK_ADC2. A conversion initiated by the

user (i.e., all single conversions, and the first free running conversion) when CK_ADC2is low will

take the same amount of time as a single ended conversion (13 ADC clock cycles from the next

prescaled clock cycle). A conversion initiated by the user when CK_ADC2is high will take 14 ADC

297

7766E–AVR–04/10

ATmega16U4/ATmega32U4

clock cycles due to the synchronization mechanism. In Free Running mode, a new conversion is

initiated immediately after the previous conversion completes, and since CK_ADC2is high at this

time, all automatically started (i.e., all but the first) Free Running conversions will take 14 ADC

clock cycles.

If differential channels are used and conversions are started by Auto Triggering, the ADC must

be switched off between conversions. When Auto Triggering is used, the ADC prescaler is reset

before the conversion is started. Since the stage is dependent of a stable ADC clock prior to the

conversion, this conversion will not be valid. By disabling and then re-enabling the ADC between

each conversion (writing ADEN in ADCSRA to “0” then to “1”), only extended conversions are

performed. The result from the extended conversions will be valid. See “Prescaling and Conver-

sion Timing” on page 295 for timing details.

The gain stage is optimized for a bandwidth of 4 kHz at all gain settings. Higher frequencies may

be subjected to non-linear amplification. An external low-pass filter should be used if the input

signal contains higher frequency components than the gain stage bandwidth. Note that the ADC

clock frequency is independent of the gain stage bandwidth limitation. E.g. the ADC clock period

may be 6 µs, allowing a channel to be sampled at 12 kSPS, regardless of the bandwidth of this

channel.

24.5 Changing Channel or Reference Selection

The MUXn and REFS1:0 bits in the ADMUX Register are single buffered through a temporary

register to which the CPU has random access. This ensures that the channels and reference

selection only takes place at a safe point during the conversion. The channel and reference

selection is continuously updated until a conversion is started. Once the conversion starts, the

channel and reference selection is locked to ensure a sufficient sampling time for the ADC. Con-

tinuous updating resumes in the last ADC clock cycle before the conversion completes (ADIF in

ADCSRA is set). Note that the conversion starts on the following rising ADC clock edge after

ADSC is written. The user is thus advised not to write new channel or reference selection values

to ADMUX until one ADC clock cycle after ADSC is written.

If Auto Triggering is used, the exact time of the triggering event can be indeterministic. Special

care must be taken when updating the ADMUX Register, in order to control which conversion

will be affected by the new settings.

If both ADATE and ADEN is written to one, an interrupt event can occur at any time. If the

ADMUX Register is changed in this period, the user cannot tell if the next conversion is based

on the old or the new settings. ADMUX can be safely updated in the following ways:

a. When ADATE or ADEN is cleared.

b. During conversion, minimum one ADC clock cycle after the trigger event.

c. After a conversion, before the interrupt flag used as trigger source is cleared.

When updating ADMUX in one of these conditions, the new settings will affect the next ADC

conversion.

Special care should be taken when changing differential channels. Once a differential channel

has been selected, the stage may take as much as 125 µs to stabilize to the new value. Thus

conversions should not be started within the first 125 µs after selecting a new differential chan-

nel. Alternatively, conversion results obtained within this period should be discarded.

The same settling time should be observed for the first differential conversion after changing

ADC reference (by changing the REFS1:0 bits in ADMUX).

298

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The settling time and gain stage bandwidth is independent of the ADHSM bit setting.

24.5.1

ADC Input Channels

When changing channel selections, the user should observe the following guidelines to ensure

that the correct channel is selected:

• In Single Conversion mode, always select the channel before starting the conversion. The

channel selection may be changed one ADC clock cycle after writing one to ADSC. However,

the simplest method is to wait for the conversion to complete before changing the channel

selection.

• In Free Running mode, always select the channel before starting the first conversion. The

channel selection may be changed one ADC clock cycle after writing one to ADSC. However,

the simplest method is to wait for the first conversion to complete, and then change the

channel selection. Since the next conversion has already started automatically, the next

result will reflect the previous channel selection. Subsequent conversions will reflect the new

channel selection.

When switching to a differential gain channel, the first conversion result may have a poor accu-

racy due to the required settling time for the automatic offset cancellation circuitry. The user

should preferably disregard the first conversion result.

24.5.2

ADC Voltage Reference

The reference voltage for the ADC (V_REF) indicates the conversion range for the ADC. Single

ended channels that exceed V_REFwill result in codes close to 0x3FF. V_REFcan be selected as

either AV_CC, internal 2.56V reference, or external AREF pin.

AV_CCis connected to the ADC through a passive switch. The internal 2.56V reference is gener-

ated from the internal bandgap reference (V_BG) through an internal amplifier. In either case, the

external AREF pin is directly connected to the ADC, and the reference voltage can be made

more immune to noise by connecting a capacitor between the AREF pin and ground. V_REFcan

also be measured at the AREF pin with a high impedance voltmeter. Note that V_REFis a high

impudent source, and only a capacitive load should be connected in a system.

If the user has a fixed voltage source connected to the AREF pin, the user may not use the other

reference voltage options in the application, as they will be shorted to the external voltage. If no

external voltage is applied to the AREF pin, the user may switch between AV_CCand 2.56V as

reference selection. The first ADC conversion result after switching reference voltage source

may be inaccurate, and the user is advised to discard this result.

If differential channels are used, the selected reference should not be closer to AV_CCthan indi-

cated in Table 29-5 on page 384.

24.6 Temperature Sensor

The ATmega16U4/ATmega32U4 includes an on-chip temperature sensor, whose the value can

be read through the A/D Converter.

The temperature measurement is based on an on-chip temperature sensor that is coupled to a

single ended ADC input. MUX[5..0] bits in ADMUX register enables the temperature sensor. The

nternal 2.56V voltage reference must also be selected for the ADC voltage reference source in

he temperature sensor measurement. When the temperature sensor is enabled, the ADC con-

verter can be used in single conversion mode to measure the voltage over the temperature

sensor.

299

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The temperature sensor and its internal driver are enabled when ADMUX value selects the tem-

perature sensor as ADC input. The propagation delay of this driver is approximatively 2µS.

Therefore two successive conversions are required. The correct temperature measurement will

be the second one.

One can also reduce this timing to one conversion by setting the ADMUX during the previous

conversion. Indeed the ADMUX can be programmed to select the temperature sensor just after

the beginning of the previous conversion start event and then the driver will be enabled 2 µS

before sampling and hold phase of temperature sensor measurement.

24.6.1

Sensor Calibration

The sensor initial tolerance is large (+/-10°C), but its characteristic is linear. Thus, if the applica-

tion requires accuracy, the firmware must include a calibration stage to use the sensor for direct

temperature measurement.

Another application of this sensor may concern the Internal Calibrated RC Oscillator, whose the

frequency can be adjusted by the user through the OSCCAL register (see Section 6.5.1 ”Oscilla-

tor Calibration Register – OSCCAL” on page 32). During the production, a calibration is done at

two temperatures (+25°C and +85°C, with a tolerance of +/-10°C⁽¹⁾). At each temperature, the

temperature sensor value T_iis measured and stored in EEPROM memory⁽²⁾, and the OSCCAL

calibration value O_i(i.e. the value that should be set in OSCCAL register at this temperature to

have an accurate 8MHz output) is stored in another memory zone.

Thanks to these four values and the linear characteristics of the temperature sensor and Internal

RC Oscillator, firmware can easily recalibrate the RC Oscillator on-the-go in function of the tem-

perature sensor measure⁽³⁾(an application note describes the operation):

Figure 24-8. Linear Characterization of OSCCAL in function of T° measurement from ADC

OSCCAL

O2

O1

T1

T2

T (ADC

Notes: 1. The temperature sensor calibration values cannot be used to do accurate temperature mea-

surements since the calibration temperature during production is not accurate (+/- 10°C)

2. Be aware that if EESAVE fuse is left unprogrammed, any chip erase operation will

clear the temperature sensor calibration values contained in EEPROM memory.

3. Accuracy results after a software recalibration of OSCCAL in function of T° will be

given when device will be fully characterized.

300

7766E–AVR–04/10

ATmega16U4/ATmega32U4

24.7 ADC Noise Canceler

The ADC features a noise canceler that enables conversion during sleep mode to reduce noise

induced from the CPU core and other I/O peripherals. The noise canceler can be used with ADC

Noise Reduction and Idle mode. To make use of this feature, the following procedure should be

used:

a. Make sure that the ADC is enabled and is not busy converting. Single Conversion

mode must be selected and the ADC conversion complete interrupt must be

enabled.

b. Enter ADC Noise Reduction mode (or Idle mode). The ADC will start a conversion

once the CPU has been halted.

c. If no other interrupts occur before the ADC conversion completes, the ADC inter-

rupt will wake up the CPU and execute the ADC Conversion Complete interrupt

routine. If another interrupt wakes up the CPU before the ADC conversion is com-

plete, that interrupt will be executed, and an ADC Conversion Complete interrupt

request will be generated when the ADC conversion completes. The CPU will

remain in active mode until a new sleep command is executed.

Note that the ADC will not be automatically turned off when entering other sleep modes than Idle

mode and ADC Noise Reduction mode. The user is advised to write zero to ADEN before enter-

ing such sleep modes to avoid excessive power consumption.

If the ADC is enabled in such sleep modes and the user wants to perform differential conver-

sions, the user is advised to switch the ADC off and on after waking up from sleep to prompt an

extended conversion to get a valid result.

24.7.1

Analog Input Circuitry

The analog input circuitry for single ended channels is illustrated in Figure 24-9. An analog

source applied to ADCn is subjected to the pin capacitance and input leakage of that pin, regard-

less of whether that channel is selected as input for the ADC. When the channel is selected, the

source must drive the S/H capacitor through the series resistance (combined resistance in the

input path).

The ADC is optimized for analog signals with an output impedance of approximately 10 kΩ or

less. If such a source is used, the sampling time will be negligible. If a source with higher imped-

ance is used, the sampling time will depend on how long time the source needs to charge the

S/H capacitor, with can vary widely. The user is recommended to only use low impedance

sources with slowly varying signals, since this minimizes the required charge transfer to the S/H

capacitor.

If differential gain channels are used, the input circuitry looks somewhat different, although

source impedances of a few hundred kΩ or less is recommended.

Signal components higher than the Nyquist frequency (f_ADC/2) should not be present for either

kind of channels, to avoid distortion from unpredictable signal convolution. The user is advised

to remove high frequency components with a low-pass filter before applying the signals as

inputs to the ADC.

301

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 24-9. Analog Input Circuitry

I

IH

ADCn

1..100 kΩ

C

= 14 pF

V

S/H

I

IL

/2

CC

24.7.2

Analog Noise Canceling Techniques

Digital circuitry inside and outside the device generates EMI which might affect the accuracy of

analog measurements. If conversion accuracy is critical, the noise level can be reduced by

applying the following techniques:

a. Keep analog signal paths as short as possible. Make sure analog tracks run over

the analog ground plane, and keep them well away from high-speed switching digi-

tal tracks.

b. The AV_CCpin on the device should be connected to the digital V_CCsupply voltage

via an LC network as shown in Figure 24-10.

c. Use the ADC noise canceler function to reduce induced noise from the CPU.

d. If any ADC port pins are used as digital outputs, it is essential that these do not

switch while a conversion is in progress.

Figure 24-10. ADC Power Connections

34

VCC

GND

(ADC7) PF7

(ADC6) PF6

(ADC5) PF5

(ADC4) PF4

(ADC1) PF1

(ADC0) PF0

35

36

37

38

39

40

41

10µH

42

43

44

AREF

GND

AVCC

100nF

1

Analog Ground Plane

Note:

The same circuitry should be used for AVCC filtering on the ADC8-ADC13 side

302

7766E–AVR–04/10

ATmega16U4/ATmega32U4

24.7.3

24.7.4

Offset Compensation Schemes

The gain stage has a built-in offset cancellation circuitry that nulls the offset of differential mea-

surements as much as possible. The remaining offset in the analog path can be measured

directly by selecting the same channel for both differential inputs. This offset residue can be then

subtracted in software from the measurement results. Using this kind of software based offset

correction, offset on any channel can be reduced below one LSB.

ADC Accuracy Definitions

An n-bit single-ended ADC converts a voltage linearly between GND and V_REFin 2ⁿsteps

(LSBs). The lowest code is read as 0, and the highest code is read as 2ⁿ-1.

Several parameters describe the deviation from the ideal behavior:

• Offset: The deviation of the first transition (0x000 to 0x001) compared to the ideal transition

(at 0.5 LSB). Ideal value: 0 LSB.

Figure 24-11. Offset Error

Output Code

Ideal ADC

Actual ADC

Offset

Error

V

Input Voltage

REF

• Gain Error: After adjusting for offset, the Gain Error is found as the deviation of the last

transition (0x3FE to 0x3FF) compared to the ideal transition (at 1.5 LSB below maximum).

Ideal value: 0 LSB

303

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 24-12. Gain Error

Gain

Error

Output Code

Ideal ADC

Actual ADC

V_REF

Input Voltage

• Integral Non-linearity (INL): After adjusting for offset and gain error, the INL is the maximum

deviation of an actual transition compared to an ideal transition for any code. Ideal value: 0

LSB.

Figure 24-13. Integral Non-linearity (INL)

Output Code

Ideal ADC

Actual ADC

V_REFInput Voltage

• Differential Non-linearity (DNL): The maximum deviation of the actual code width (the interval

between two adjacent transitions) from the ideal code width (1 LSB). Ideal value: 0 LSB.

304

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 24-14. Differential Non-linearity (DNL)

Output Code

0x3FF

1 LSB

DNL

0x000

0

V

Input Voltage

REF

• Quantization Error: Due to the quantization of the input voltage into a finite number of codes,

a range of input voltages (1 LSB wide) will code to the same value. Always 0.5 LSB.

• Absolute Accuracy: The maximum deviation of an actual (unadjusted) transition compared to

an ideal transition for any code. This is the compound effect of offset, gain error, differential

error, non-linearity, and quantization error. Ideal value: 0.5 LSB.

24.8 ADC Conversion Result

After the conversion is complete (ADIF is high), the conversion result can be found in the ADC

Result Registers (ADCL, ADCH).

For single ended conversion, the result is:

V

⋅ 1023

IN

ADC = --------------------------

V

REF

where V_INis the voltage on the selected input pin and V_REFthe selected voltage reference (see

Table 24-3 on page 307 and Table 24-4 on page 308). 0x000 represents analog ground, and

0x3FF represents the selected reference voltage minus one LSB.

If differential channels are used, the result is:

(V

– V

) ⋅ GAIN ⋅ 512

NEG

POS

ADC = ------------------------------------------------------------------------

V

REF

where V_POSis the voltage on the positive input pin, V_NEGthe voltage on the negative input pin,

GAIN the selected gain factor and V_REFthe selected voltage reference. The result is presented

in two’s complement form, from 0x200 (-512d) through 0x1FF (+511d). Note that if the user

wants to perform a quick polarity check of the result, it is sufficient to read the MSB of the result

(ADC9 in ADCH). If the bit is one, the result is negative, and if this bit is zero, the result is posi-

tive. Figure 24-15 shows the decoding of the differential input range.

Table 82 shows the resulting output codes if the differential input channel pair (ADCn - ADCm) is

selected with a reference voltage of V_REF

.

305

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 24-15. Differential Measurement Range

Output Code

0x1FF

0x000

0

Differential Input

Voltage (Volts)

- V_REF

V_REF

0x3FF

0x200

Table 24-2. Correlation Between Input Voltage and Output Codes

V_ADCn

Read code

0x1FF

0x1FE

...

Corresponding decimal value

V_ADCm+ V_REF/GAIN

511

510

...

V_ADCm+ 0.999 V_REF/GAIN

V_ADCm+ 0.998 V_REF/GAIN

...

V_ADCm+ 0.001 V_REF/GAIN

V_ADCm

0x001

0x000

0x3FF

...

1

0

V_ADCm- 0.001 V_REF/GAIN

-1

...

V_ADCm- 0.999 V_REF/GAIN

0x201

0x200

-511

-512

V_ADCm- V_REF/GAIN

Example 1:

– ADMUX = 0xE9, MUX5 = 0 (ADC1 - ADC0, 10x gain, 2.56V reference, left adjusted

result)

– Voltage on ADC1 is 300 mV, voltage on ADC0 is 500 mV.

– ADCR = 512 * 10 * (300 - 500) / 2560 = -400 = 0x270

306

7766E–AVR–04/10

ATmega16U4/ATmega32U4

– ADCL will thus read 0x00, and ADCH will read 0x9C.

Writing zero to ADLAR right adjusts the result: ADCL = 0x70, ADCH = 0x02.

Example 2:

– ADMUX = 0xF0, MUX5 = 0 (ADC0 - ADC1, 1x gain, 2.56V reference, left adjusted

result)

– Voltage on ADC0 is 300 mV, voltage on ADC1 is 500 mV.

– ADCR = 512 * 1 * (300 - 500) / 2560 = -41 = 0x029.

– ADCL will thus read 0x40, and ADCH will read 0x0A.

Writing zero to ADLAR right adjusts the result: ADCL = 0x00, ADCH = 0x29.

24.9 ADC Register Description

24.9.1

ADC Multiplexer Selection Register – ADMUX

Bit

7

REFS1

R/W

0

6

REFS0

R/W

0

5

ADLAR

R/W

0

4

MUX4

R/W

0

3

MUX3

R/W

0

2

MUX2

R/W

0

1

MUX1

R/W

0

MUX0

R/W

0

ADMUX

Read/Write

Initial Value

• Bit 7:6 – REFS1:0: Reference Selection Bits

These bits select the voltage reference for the ADC, as shown in Table 24-3. If these bits are

changed during a conversion, the change will not go in effect until this conversion is complete

(ADIF in ADCSRA is set). The internal voltage reference options may not be used if an external

reference voltage is being applied to the AREF pin.

Table 24-3. Voltage Reference Selections for ADC

REFS1

REFS0 Voltage Reference Selection

0

1

0

1

0

1

AREF, Internal Vref turned off

AV_CCwith external capacitor on AREF pin

Reserved

Internal 2.56V Voltage Reference with external capacitor on AREF pin

•

Bit 5 – ADLAR: ADC Left Adjust Result

The ADLAR bit affects the presentation of the ADC conversion result in the ADC Data Register.

Write one to ADLAR to left adjust the result. Otherwise, the result is right adjusted. Changing the

ADLAR bit will affect the ADC Data Register immediately, regardless of any ongoing conver-

sions. For a complete description of this bit, see “The ADC Data Register – ADCL and ADCH” on

page 311.

• Bits 4:0 – MUX4:0: Analog Channel Selection Bits

The value of these bits selects which combination of analog inputs are connected to the ADC.

These bits also select the gain for the differential channels. See Table 24-4 for details. If these

bits are changed during a conversion, the change will not go in effect until this conversion is

complete (ADIF in ADCSRA is set).

307

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 24-4. Input Channel and Gain Selections

MUX5..0⁽¹⁾Single Ended Input Positive Differential Input Negative Differential Input

000000 ADC0

000001 ADC1

Gain

000010

000011

000100

000101

000110

000111

001000

001001

001010

001011

001100

001101

001110

001111

010000

N/A

ADC4

ADC5

ADC6

ADC7

N/A

10x

ADC1

N/A

ADC0

N/A

200x

ADC1

ADC0

N/A

ADC0

ADC1

1x

308

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 24-4. Input Channel and Gain Selections (Continued)

MUX5..0⁽¹⁾Single Ended Input Positive Differential Input Negative Differential Input

Gain

010001

010010

010011

010100

010101

010110

010111

011000

011001

011010

011011

011100

011101

011110

011111

100000

100001

100010

100011

100100

100101

100110

100111

101000

101001

101010

101011

101100

101101

101110

101111

110000

110001

110010

110011

N/A

ADC4

ADC5

ADC6

ADC7

ADC1

1x

N/A

1.1V (V_{Band Gap}

0V (GND)

ADC8

)

N/A

ADC9

ADC10

ADC11

ADC12

ADC13

N/A

ADC1

ADC0

40x

Temperature Sensor

ADC4

ADC5

ADC6

ADC7

ADC4

ADC5

ADC6

ADC7

ADC4

ADC5

ADC6

ADC7

ADC0

ADC1

ADC0

10x

40x

N/A

309

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 24-4. Input Channel and Gain Selections (Continued)

MUX5..0⁽¹⁾Single Ended Input Positive Differential Input Negative Differential Input

Gain

40x

110100

110101

110110

110111

111000

111001

111010

111011

111100

111101

111110

111111

ADC4

ADC5

ADC6

ADC7

ADC4

ADC5

ADC6

ADC7

ADC4

ADC5

ADC6

ADC7

ADC1

ADC0

ADC1

40x

200x

N/A

Note:

1. MUX5 bit make part of ADCSRB register

24.9.2

ADC Control and Status Register A – ADCSRA

Bit

7

ADEN

R/W

0

6

ADSC

R/W

0

5

ADATE

R/W

0

4

ADIF

R/W

0

3

ADIE

R/W

0

2

ADPS2

R/W

0

1

ADPS1

R/W

0

ADPS0

R/W

0

ADCSRA

Read/Write

Initial Value

• Bit 7 – ADEN: ADC Enable

Writing this bit to one enables the ADC. By writing it to zero, the ADC is turned off. Turning the

ADC off while a conversion is in progress, will terminate this conversion.

• Bit 6 – ADSC: ADC Start Conversion

In Single Conversion mode, write this bit to one to start each conversion. In Free Running mode,

write this bit to one to start the first conversion. The first conversion after ADSC has been written

after the ADC has been enabled, or if ADSC is written at the same time as the ADC is enabled,

will take 25 ADC clock cycles instead of the normal 13. This first conversion performs initializa-

tion of the ADC.

ADSC will read as one as long as a conversion is in progress. When the conversion is complete,

it returns to zero. Writing zero to this bit has no effect.

• Bit 5 – ADATE: ADC Auto Trigger Enable

When this bit is written to one, Auto Triggering of the ADC is enabled. The ADC will start a con-

version on a positive edge of the selected trigger signal. The trigger source is selected by setting

the ADC Trigger Select bits, ADTS in ADCSRB.

• Bit 4 – ADIF: ADC Interrupt Flag

This bit is set when an ADC conversion completes and the Data Registers are updated. The

ADC Conversion Complete Interrupt is executed if the ADIE bit and the I-bit in SREG are set.

ADIF is cleared by hardware when executing the corresponding interrupt handling vector. Alter-

natively, ADIF is cleared by writing a logical one to the flag. Beware that if doing a Read-Modify-

310

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Write on ADCSRA, a pending interrupt can be disabled. This also applies if the SBI and CBI

instructions are used.

• Bit 3 – ADIE: ADC Interrupt Enable

When this bit is written to one and the I-bit in SREG is set, the ADC Conversion Complete Inter-

rupt is activated.

• Bits 2:0 – ADPS2:0: ADC Prescaler Select Bits

These bits determine the division factor between the XTAL frequency and the input clock to the

ADC.

Table 24-5. ADC Prescaler Selections

ADPS2

ADPS1

ADPS0

Division Factor

0

1

0

1

0

1

0

1

0

1

0

1

0

1

2

4

8

16

32

64

128

24.9.3

The ADC Data Register – ADCL and ADCH

24.9.3.1

ADLAR = 0

Bit

15

14

13

12

11

10

9

8

–

ADC9

ADC8

ADCH

ADCL

ADC7

ADC6

ADC5

ADC4

ADC3

ADC2

ADC1

ADC0

Bit

7

R

0

6

R

0

5

R

0

4

R

0

3

R

0

2

R

0

1

R

0

R

0

Read/Write

Initial Value

0

24.9.3.2

ADLAR = 1

Bit

15

14

13

12

11

10

9

8

ADC9

ADC8

ADC7

ADC6

ADC5

ADC4

ADC3

ADC2

ADCH

ADCL

ADC1

ADC0

–

5

–

4

–

3

–

2

–

1

–

0

Bit

7

R

0

6

R

0

Read/Write

R

0

R

0

R

0

R

0

R

0

R

0

Initial Value

0

When an ADC conversion is complete, the result is found in these two registers. If differential

channels are used, the result is presented in two’s complement form.

When ADCL is read, the ADC Data Register is not updated until ADCH is read. Consequently, if

the result is left adjusted and no more than 8-bit precision (7 bit + sign bit for differential input

311

7766E–AVR–04/10

ATmega16U4/ATmega32U4

channels) is required, it is sufficient to read ADCH. Otherwise, ADCL must be read first, then

ADCH.

The ADLAR bit in ADMUX, and the MUXn bits in ADMUX affect the way the result is read from

the registers. If ADLAR is set, the result is left adjusted. If ADLAR is cleared (default), the result

is right adjusted.

• ADC9:0: ADC Conversion Result

These bits represent the result from the conversion, as detailed in “ADC Conversion Result” on

page 305.

24.9.4

ADC Control and Status Register B – ADCSRB

Bit

7

6

5

4

-

3

2

ADTS2

R/W

0

1

ADTS1

R/W

0

ADTS0

R/W

0

ADHSM

ACME

MUX5

ADTS3

ADCSRB

Read/Write

Initial Value

R/W

0

R/W

0

R

0

R

0

R

0

• Bit 7 – ADHSM: ADC High Speed Mode

Writing this bit to one enables the ADC High Speed mode. This mode enables higher conversion

rate at the expense of higher power consumption.

• Bit 5 – MUX5: Analog Channel Additional Selection Bits

This bit make part of MUX5:0 bits of ADRCSRB and ADMUX register, that select the combina-

tion of analog inputs connected to the ADC (including differential amplifier configuration).

• Bit 3:0 – ADTS3:0: ADC Auto Trigger Source

If ADATE in ADCSRA is written to one, the value of these bits selects which source will trigger

an ADC conversion. If ADATE is cleared, the ADTS3:0 settings will have no effect. A conversion

will be triggered by the rising edge of the selected interrupt flag. Note that switching from a trig-

ger source that is cleared to a trigger source that is set, will generate a positive edge on the

trigger signal. If ADEN in ADCSRA is set, this will start a conversion. Switching to Free Running

mode (ADTS[3:0]=0) will not cause a trigger event, even if the ADC Interrupt Flag is set.

Table 24-6. ADC Auto Trigger Source Selections

ADTS3

ADTS2

ADTS1

ADTS0

Trigger Source

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

1

0

Free Running mode

Analog Comparator

External Interrupt Request 0

Timer/Counter0 Compare Match

Timer/Counter0 Overflow

Timer/Counter1 Compare Match B

Timer/Counter1 Overflow

Timer/Counter1 Capture Event

Timer/Counter4 Overflow

312

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 24-6. ADC Auto Trigger Source Selections (Continued)

ADTS3

ADTS2

ADTS1

ADTS0

Trigger Source

1

0

1

0

1

Timer/Counter4 Compare Match A

Timer/Counter4 Compare Match B

Timer/Counter4 Compare Match D

24.9.5

Digital Input Disable Register 0 – DIDR0

Bit

7

ADC7D

R/W

0

6

ADC6D

R/W

0

5

4

ADC4D

R/W

0

3

2

-

1

ADC1D

R/W

0

ADC0D

R/W

0

ADC5D

R/W

0

-

DIDR0

Read/Write

Initial Value

R/W

0

• Bit 7:4, 1:0 – ADC7D..4D - ADC1D..0D : ADC7:4 - ADC1:0 Digital Input Disable

When this bit is written logic one, the digital input buffer on the corresponding ADC pin is dis-

abled. The corresponding PIN Register bit will always read as zero when this bit is set. When an

analog signal is applied to the ADC7..4 / ADC1..0 pin and the digital input from this pin is not

needed, this bit should be written logic one to reduce power consumption in the digital input

buffer.

24.9.6

Digital Input Disable Register 2 – DIDR2

Bit

7

6

5

4

3

2

1

0

ADC8D

R/W

0

-

ADC13D ADC12D ADC11D ADC10D ADC9D

DIDR2

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

• Bit 5:0 – ADC13D..ADC8D: ADC13:8 Digital Input Disable

When this bit is written logic one, the digital input buffer on the corresponding ADC pin is dis-

abled. The corresponding PIN Register bit will always read as zero when this bit is set. When an

analog signal is applied to the ADC13..8 pin and the digital input from this pin is not needed, this

bit should be written logic one to reduce power consumption in the digital input buffer.

313

7766E–AVR–04/10

ATmega16U4/ATmega32U4

25. JTAG Interface and On-chip Debug System

25.0.1

Features

• JTAG (IEEE std. 1149.1 Compliant) Interface

• Boundary-scan Capabilities According to the IEEE std. 1149.1 (JTAG) Standard

• Debugger Access to:

– All Internal Peripheral Units

– Internal and External RAM

– The Internal Register File

– Program Counter

– EEPROM and Flash Memories

• Extensive On-chip Debug Support for Break Conditions, Including

– AVR Break Instruction

– Break on Change of Program Memory Flow

– Single Step Break

– Program Memory Break Points on Single Address or Address Range

– Data Memory Break Points on Single Address or Address Range

• Programming of Flash, EEPROM, Fuses, and Lock Bits through the JTAG Interface

• On-chip Debugging Supported by AVR Studio^®

25.1 Overview

The AVR IEEE std. 1149.1 compliant JTAG interface can be used for

• Testing PCBs by using the JTAG Boundary-scan capability

• Programming the non-volatile memories, Fuses and Lock bits

• On-chip debugging

A brief description is given in the following sections. Detailed descriptions for Programming via

the JTAG interface, and using the Boundary-scan Chain can be found in the sections “Program-

ming via the JTAG Interface” on page 365 and “IEEE 1149.1 (JTAG) Boundary-scan” on page

320, respectively. The On-chip Debug support is considered being private JTAG instructions,

and distributed within ATMEL and to selected third party vendors only.

Figure 25-1 shows a block diagram of the JTAG interface and the On-chip Debug system. The

TAP Controller is a state machine controlled by the TCK and TMS signals. The TAP Controller

selects either the JTAG Instruction Register or one of several Data Registers as the scan chain

(Shift Register) between the TDI – input and TDO – output. The Instruction Register holds JTAG

instructions controlling the behavior of a Data Register.

The ID-Register, Bypass Register, and the Boundary-scan Chain are the Data Registers used

for board-level testing. The JTAG Programming Interface (actually consisting of several physical

and virtual Data Registers) is used for serial programming via the JTAG interface. The Internal

Scan Chain and Break Point Scan Chain are used for On-chip debugging only.

25.2 Test Access Port – TAP

The JTAG interface is accessed through four of the AVR’s pins. In JTAG terminology, these pins

constitute the Test Access Port – TAP. These pins are:

• TMS: Test mode select. This pin is used for navigating through the TAP-controller state

machine.

• TCK: Test Clock. JTAG operation is synchronous to TCK.

314

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• TDI: Test Data In. Serial input data to be shifted in to the Instruction Register or Data Register

(Scan Chains).

• TDO: Test Data Out. Serial output data from Instruction Register or Data Register.

The IEEE std. 1149.1 also specifies an optional TAP signal; TRST – Test ReSeT – which is not

provided.

When the JTAGEN Fuse is unprogrammed, these four TAP pins are normal port pins, and the

TAP controller is in reset. When programmed, the input TAP signals are internally pulled high

and the JTAG is enabled for Boundary-scan and programming. The device is shipped with this

fuse programmed.

For the On-chip Debug system, in addition to the JTAG interface pins, the RESET pin is moni-

tored by the debugger to be able to detect external reset sources. The debugger can also pull

the RESET pin low to reset the whole system, assuming only open collectors on the reset line

are used in the application.

Figure 25-1. Block Diagram

I/O PORT 0

DEVICE BOUNDARY

BOUNDARY SCAN CHAIN

TDI

JTAG PROGRAMMING

INTERFACE

TDO

TCK

TMS

TAP

CONTROLLER

AVR CPU

INTERNAL

SCAN

CHAIN

FLASH

MEMORY

Address

Data

PC

Instruction

INSTRUCTION

REGISTER

ID

REGISTER

BREAKPOINT

UNIT

M

U

X

FLOW CONTROL

UNIT

BYPASS

REGISTER

DIGITAL

PERIPHERAL

UNITS

ANALOG

PERIPHERIAL

UNITS

Analog inputs

BREAKPOINT

SCAN CHAIN

JTAG / AVR CORE

COMMUNICATION

INTERFACE

ADDRESS

DECODER

OCD STATUS

AND CONTROL

Control & Clock lines

I/O PORT n

315

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 25-2. TAP Controller State Diagram

1

Test-Logic-Reset

0

1

0

Run-Test/Idle

Select-DR Scan

Select-IR Scan

0

1

Capture-DR

Capture-IR

0

Shift-IR

1

Shift-DR

0

1

Exit1-DR

0

1

Exit1-IR

0

Pause-DR

1

0

Pause-IR

1

0

Exit2-DR

1

Exit2-IR

1

Update-DR

Update-IR

1

0

25.3 TAP Controller

The TAP controller is a 16-state finite state machine that controls the operation of the Boundary-

scan circuitry, JTAG programming circuitry, or On-chip Debug system. The state transitions

depicted in Figure 25-2 depend on the signal present on TMS (shown adjacent to each state

transition) at the time of the rising edge at TCK. The initial state after a Power-on Reset is Test-

Logic-Reset.

As a definition in this document, the LSB is shifted in and out first for all Shift Registers.

Assuming Run-Test/Idle is the present state, a typical scenario for using the JTAG interface is:

• At the TMS input, apply the sequence 1, 1, 0, 0 at the rising edges of TCK to enter the Shift

Instruction Register – Shift-IR state. While in this state, shift the four bits of the JTAG

instructions into the JTAG Instruction Register from the TDI input at the rising edge of TCK.

The TMS input must be held low during input of the 3 LSBs in order to remain in the Shift-IR

state. The MSB of the instruction is shifted in when this state is left by setting TMS high.

While the instruction is shifted in from the TDI pin, the captured IR-state 0x01 is shifted out on

the TDO pin. The JTAG Instruction selects a particular Data Register as path between TDI

and TDO and controls the circuitry surrounding the selected Data Register.

316

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Apply the TMS sequence 1, 1, 0 to re-enter the Run-Test/Idle state. The instruction is latched

onto the parallel output from the Shift Register path in the Update-IR state. The Exit-IR,

Pause-IR, and Exit2-IR states are only used for navigating the state machine.

• At the TMS input, apply the sequence 1, 0, 0 at the rising edges of TCK to enter the Shift

Data Register – Shift-DR state. While in this state, upload the selected Data Register

(selected by the present JTAG instruction in the JTAG Instruction Register) from the TDI input

at the rising edge of TCK. In order to remain in the Shift-DR state, the TMS input must be

held low during input of all bits except the MSB. The MSB of the data is shifted in when this

state is left by setting TMS high. While the Data Register is shifted in from the TDI pin, the

parallel inputs to the Data Register captured in the Capture-DR state is shifted out on the

TDO pin.

• Apply the TMS sequence 1, 1, 0 to re-enter the Run-Test/Idle state. If the selected Data

Register has a latched parallel-output, the latching takes place in the Update-DR state. The

Exit-DR, Pause-DR, and Exit2-DR states are only used for navigating the state machine.

As shown in the state diagram, the Run-Test/Idle state need not be entered between selecting

JTAG instruction and using Data Registers, and some JTAG instructions may select certain

functions to be performed in the Run-Test/Idle, making it unsuitable as an Idle state.

Note:

Independent of the initial state of the TAP Controller, the Test-Logic-Reset state can always be

entered by holding TMS high for five TCK clock periods.

For detailed information on the JTAG specification, refer to the literature listed in “Bibliography”

on page 319.

25.4 Using the Boundary-scan Chain

A complete description of the Boundary-scan capabilities are given in the section “IEEE 1149.1

(JTAG) Boundary-scan” on page 320.

25.5 Using the On-chip Debug System

As shown in Figure 25-1, the hardware support for On-chip Debugging consists mainly of

• A scan chain on the interface between the internal AVR CPU and the internal peripheral

units.

• Break Point unit.

• Communication interface between the CPU and JTAG system.

All read or modify/write operations needed for implementing the Debugger are done by applying

AVR instructions via the internal AVR CPU Scan Chain. The CPU sends the result to an I/O

memory mapped location which is part of the communication interface between the CPU and the

JTAG system.

The Break Point Unit implements Break on Change of Program Flow, Single Step Break, two

Program Memory Break Points, and two combined Break Points. Together, the four Break

Points can be configured as either:

• 4 single Program Memory Break Points.

• 3 Single Program Memory Break Point + 1 single Data Memory Break Point.

• 2 single Program Memory Break Points + 2 single Data Memory Break Points.

• 2 single Program Memory Break Points + 1 Program Memory Break Point with mask (“range

Break Point”).

317

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• 2 single Program Memory Break Points + 1 Data Memory Break Point with mask (“range

Break Point”).

A debugger, like the AVR Studio, may however use one or more of these resources for its inter-

nal purpose, leaving less flexibility to the end-user.

A list of the On-chip Debug specific JTAG instructions is given in “On-chip Debug Specific JTAG

Instructions” on page 318.

The JTAGEN Fuse must be programmed to enable the JTAG Test Access Port. In addition, the

OCDEN Fuse must be programmed and no Lock bits must be set for the On-chip debug system

to work. As a security feature, the On-chip debug system is disabled when either of the LB1 or

LB2 Lock bits are set. Otherwise, the On-chip debug system would have provided a back-door

into a secured device.

The AVR Studio enables the user to fully control execution of programs on an AVR device with

On-chip Debug capability, AVR In-Circuit Emulator, or the built-in AVR Instruction Set Simulator.

AVR Studio^®supports source level execution of Assembly programs assembled with ATMEL

Corporation’s AVR Assembler and C programs compiled with third party vendors’ compilers.

AVR Studio runs under Microsoft^®Windows^®95/98/2000 and Microsoft Windows NT^®.

For a full description of the AVR Studio, please refer to the AVR Studio User Guide. Only high-

lights are presented in this document.

All necessary execution commands are available in AVR Studio, both on source level and on

disassembly level. The user can execute the program, single step through the code either by

tracing into or stepping over functions, step out of functions, place the cursor on a statement and

execute until the statement is reached, stop the execution, and reset the execution target. In

addition, the user can have an unlimited number of code Break Points (using the BREAK

instruction) and up to two data memory Break Points, alternatively combined as a mask (range)

Break Point.

25.6 On-chip Debug Specific JTAG Instructions

The On-chip debug support is considered being private JTAG instructions, and distributed within

ATMEL and to selected third party vendors only. Instruction opcodes are listed for reference.

25.6.1

25.6.2

25.6.3

25.6.4

PRIVATE0; 0x8

PRIVATE1; 0x9

PRIVATE2; 0xA

PRIVATE3; 0xB

Private JTAG instruction for accessing On-chip debug system.

318

7766E–AVR–04/10

ATmega16U4/ATmega32U4

25.7 On-chip Debug Related Register in I/O Memory

25.7.1

On-chip Debug Register – OCDR

Bit

7

6

5

4

3

2

1

0

MSB/IDRD

LSB

R/W

0

OCDR

Read/Write

Initial Value

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

R/W

0

The OCDR Register provides a communication channel from the running program in the micro-

controller to the debugger. The CPU can transfer a byte to the debugger by writing to this

location. At the same time, an internal flag; I/O Debug Register Dirty – IDRD – is set to indicate

to the debugger that the register has been written. When the CPU reads the OCDR Register the

7 LSB will be from the OCDR Register, while the MSB is the IDRD bit. The debugger clears the

IDRD bit when it has read the information.

In some AVR devices, this register is shared with a standard I/O location. In this case, the OCDR

Register can only be accessed if the OCDEN Fuse is programmed, and the debugger enables

access to the OCDR Register. In all other cases, the standard I/O location is accessed.

Refer to the debugger documentation for further information on how to use this register.

25.8 Using the JTAG Programming Capabilities

Programming of AVR parts via JTAG is performed via the 4-pin JTAG port, TCK, TMS, TDI, and

TDO. These are the only pins that need to be controlled/observed to perform JTAG program-

ming (in addition to power pins). It is not required to apply 12V externally. The JTAGEN Fuse

must be programmed and the JTD bit in the MCUCR Register must be cleared to enable the

JTAG Test Access Port.

The JTAG programming capability supports:

• Flash programming and verifying.

• EEPROM programming and verifying.

• Fuse programming and verifying.

• Lock bit programming and verifying.

The Lock bit security is exactly as in parallel programming mode. If the Lock bits LB1 or LB2 are

programmed, the OCDEN Fuse cannot be programmed unless first doing a chip erase. This is a

security feature that ensures no back-door exists for reading out the content of a secured

device.

The details on programming through the JTAG interface and programming specific JTAG

instructions are given in the section “Programming via the JTAG Interface” on page 365.

25.9 Bibliography

For more information about general Boundary-scan, the following literature can be consulted:

• IEEE: IEEE Std. 1149.1-1990. IEEE Standard Test Access Port and Boundary-scan

Architecture, IEEE, 1993.

• Colin Maunder: The Board Designers Guide to Testable Logic Circuits, Addison-Wesley,

1992.

319

7766E–AVR–04/10

ATmega16U4/ATmega32U4

26. IEEE 1149.1 (JTAG) Boundary-scan

26.1 Features

• JTAG (IEEE std. 1149.1 compliant) Interface

• Boundary-scan Capabilities According to the JTAG Standard

• Full Scan of all Port Functions as well as Analog Circuitry having Off-chip Connections

• Supports the Optional IDCODE Instruction

• Additional Public AVR_RESET Instruction to Reset the AVR

26.2 System Overview

The Boundary-scan chain has the capability of driving and observing the logic levels on the digi-

tal I/O pins, as well as the boundary between digital and analog logic for analog circuitry having

off-chip connections. At system level, all ICs having JTAG capabilities are connected serially by

the TDI/TDO signals to form a long Shift Register. An external controller sets up the devices to

drive values at their output pins, and observe the input values received from other devices. The

controller compares the received data with the expected result. In this way, Boundary-scan pro-

vides a mechanism for testing interconnections and integrity of components on Printed Circuits

Boards by using the four TAP signals only.

The four IEEE 1149.1 defined mandatory JTAG instructions IDCODE, BYPASS, SAMPLE/PRE-

LOAD, and EXTEST, as well as the AVR specific public JTAG instruction AVR_RESET can be

used for testing the Printed Circuit Board. Initial scanning of the Data Register path will show the

ID-Code of the device, since IDCODE is the default JTAG instruction. It may be desirable to

have the AVR device in reset during test mode. If not reset, inputs to the device may be deter-

mined by the scan operations, and the internal software may be in an undetermined state when

exiting the test mode. Entering reset, the outputs of any port pin will instantly enter the high

impedance state, making the HIGHZ instruction redundant. If needed, the BYPASS instruction

can be issued to make the shortest possible scan chain through the device. The device can be

set in the reset state either by pulling the external RESET pin low, or issuing the AVR_RESET

instruction with appropriate setting of the Reset Data Register.

The EXTEST instruction is used for sampling external pins and loading output pins with data.

The data from the output latch will be driven out on the pins as soon as the EXTEST instruction

is loaded into the JTAG IR-Register. Therefore, the SAMPLE/PRELOAD should also be used for

setting initial values to the scan ring, to avoid damaging the board when issuing the EXTEST

instruction for the first time. SAMPLE/PRELOAD can also be used for taking a snapshot of the

external pins during normal operation of the part.

The JTAGEN Fuse must be programmed and the JTD bit in the I/O Register MCUCR must be

cleared to enable the JTAG Test Access Port.

When using the JTAG interface for Boundary-scan, using a JTAG TCK clock frequency higher

than the internal chip frequency is possible. The chip clock is not required to run.

26.3 Data Registers

The Data Registers relevant for Boundary-scan operations are:

• Bypass Register

• Device Identification Register

• Reset Register

• Boundary-scan Chain

320

7766E–AVR–04/10

ATmega16U4/ATmega32U4

26.3.1

26.3.2

Bypass Register

The Bypass Register consists of a single Shift Register stage. When the Bypass Register is

selected as path between TDI and TDO, the register is reset to 0 when leaving the Capture-DR

controller state. The Bypass Register can be used to shorten the scan chain on a system when

the other devices are to be tested.

Device Identification Register

Figure 26-1 shows the structure of the Device Identification Register.

Figure 26-1. The Format of the Device Identification Register

LSB

MSB

Bit

31

28

27

12

11

1

0

Device ID

Version

Part Number

Manufacturer ID

1

4 bits

16 bits

11 bits

1-bit

26.3.2.1

26.3.2.2

Version

Version is a 4-bit number identifying the revision of the component. The JTAG version number

follows the revision of the device. Revision A is 0x0, revision B is 0x1 and so on.

Part Number

The part number is a 16-bit code identifying the component. The JTAG Part Number for

ATmega16U4/ATmega32U4 is listed in Table 26-1.

Table 26-1. AVR JTAG Part Number

Part Number

JTAG Part Number (Hex)

AVR USB

0x9782

26.3.2.3

Manufacturer ID

The Manufacturer ID is a 11-bit code identifying the manufacturer. The JTAG manufacturer ID

for ATMEL is listed in Table 26-2.

Table 26-2. Manufacturer ID

Manufacturer

JTAG Manufacturer ID (Hex)

ATMEL

0x01F

26.3.3

Reset Register

The Reset Register is a test Data Register used to reset the part. Since the AVR tri-states Port

Pins when reset, the Reset Register can also replace the function of the unimplemented optional

JTAG instruction HIGHZ.

A high value in the Reset Register corresponds to pulling the external Reset low. The part is

reset as long as there is a high value present in the Reset Register. Depending on the fuse set-

tings for the clock options, the part will remain reset for a reset time-out period (refer to “Clock

Sources” on page 28) after releasing the Reset Register. The output from this Data Register is

not latched, so the reset will take place immediately, as shown in Figure 26-2.

321

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 26-2. Reset Register

To

TDO

From Other Internal and

External Reset Sources

From

TDI

Internal reset

D

Q

ClockDR · AVR_RESET

26.3.4

Boundary-scan Chain

The Boundary-scan Chain has the capability of driving and observing the logic levels on the dig-

ital I/O pins, as well as the boundary between digital and analog logic for analog circuitry having

off-chip connections.

See “Boundary-scan Chain” on page 324 for a complete description.

26.4 Boundary-scan Specific JTAG Instructions

The Instruction Register is 4-bit wide, supporting up to 16 instructions. Listed below are the

JTAG instructions useful for Boundary-scan operation. Note that the optional HIGHZ instruction

is not implemented, but all outputs with tri-state capability can be set in high-impedance state by

using the AVR_RESET instruction, since the initial state for all port pins is tri-state.

As a definition in this datasheet, the LSB is shifted in and out first for all Shift Registers.

The OPCODE for each instruction is shown behind the instruction name in hex format. The text

describes which Data Register is selected as path between TDI and TDO for each instruction.

26.4.1

EXTEST; 0x0

Mandatory JTAG instruction for selecting the Boundary-scan Chain as Data Register for testing

circuitry external to the AVR package. For port-pins, Pull-up Disable, Output Control, Output

Data, and Input Data are all accessible in the scan chain. For Analog circuits having off-chip

connections, the interface between the analog and the digital logic is in the scan chain. The con-

tents of the latched outputs of the Boundary-scan chain is driven out as soon as the JTAG IR-

Register is loaded with the EXTEST instruction.

The active states are:

• Capture-DR: Data on the external pins are sampled into the Boundary-scan Chain.

• Shift-DR: The Internal Scan Chain is shifted by the TCK input.

• Update-DR: Data from the scan chain is applied to output pins.

26.4.2

IDCODE; 0x1

Optional JTAG instruction selecting the 32 bit ID-Register as Data Register. The ID-Register

consists of a version number, a device number and the manufacturer code chosen by JEDEC.

This is the default instruction after power-up.

322

7766E–AVR–04/10

ATmega16U4/ATmega32U4

The active states are:

• Capture-DR: Data in the IDCODE Register is sampled into the Boundary-scan Chain.

• Shift-DR: The IDCODE scan chain is shifted by the TCK input.

26.4.3

SAMPLE_PRELOAD; 0x2

Mandatory JTAG instruction for pre-loading the output latches and taking a snap-shot of the

input/output pins without affecting the system operation. However, the output latches are not

connected to the pins. The Boundary-scan Chain is selected as Data Register.

The active states are:

• Capture-DR: Data on the external pins are sampled into the Boundary-scan Chain.

• Shift-DR: The Boundary-scan Chain is shifted by the TCK input.

• Update-DR: Data from the Boundary-scan chain is applied to the output latches. However,

the output latches are not connected to the pins.

26.4.4

AVR_RESET; 0xC

The AVR specific public JTAG instruction for forcing the AVR device into the Reset mode or

releasing the JTAG reset source. The TAP controller is not reset by this instruction. The one bit

Reset Register is selected as Data Register. Note that the reset will be active as long as there is

a logic “one” in the Reset Chain. The output from this chain is not latched.

The active states are:

• Shift-DR: The Reset Register is shifted by the TCK input.

26.4.5

BYPASS; 0xF

Mandatory JTAG instruction selecting the Bypass Register for Data Register.

The active states are:

• Capture-DR: Loads a logic “0” into the Bypass Register.

• Shift-DR: The Bypass Register cell between TDI and TDO is shifted.

26.5 Boundary-scan Related Register in I/O Memory

26.5.1

MCU Control Register – MCUCR

The MCU Control Register contains control bits for general MCU functions.

Bit

7

6

–

R

0

5

–

R

0

4

3

–

R

0

2

–

R

0

1

0

JTD

R/W

0

PUD

R/W

0

IVSEL

R/W

0

IVCE

R/W

0

MCUCR

Read/Write

Initial Value

• Bits 7 – JTD: JTAG Interface Disable

When this bit is zero, the JTAG interface is enabled if the JTAGEN Fuse is programmed. If this

bit is one, the JTAG interface is disabled. In order to avoid unintentional disabling or enabling of

the JTAG interface, a timed sequence must be followed when changing this bit: The application

software must write this bit to the desired value twice within four cycles to change its value. Note

that this bit must not be altered when using the On-chip Debug system.

323

7766E–AVR–04/10

ATmega16U4/ATmega32U4

26.5.2

MCU Status Register – MCUSR

The MCU Status Register provides information on which reset source caused an MCU reset.

Bit

7

–

R

0

6

–

R

0

5

–

R

0

4

3

2

1

0

JTRF

R/W

WDRF

R/W

BORF

R/W

EXTRF

R/W

PORF

R/W

MCUSR

Read/Write

Initial Value

See Bit Description

• Bit 4 – JTRF: JTAG Reset Flag

This bit is set if a reset is being caused by a logic one in the JTAG Reset Register selected by

the JTAG instruction AVR_RESET. This bit is reset by a Power-on Reset, or by writing a logic

zero to the flag.

26.6 Boundary-scan Chain

The Boundary-scan chain has the capability of driving and observing the logic levels on the digi-

tal I/O pins, as well as the boundary between digital and analog logic for analog circuitry having

off-chip connection.

26.6.1

Scanning the Digital Port Pins

Figure 26-3 shows the Boundary-scan Cell for a bi-directional port pin. The pull-up function is

disabled during Boundary-scan when the JTAG IC contains EXTEST or SAMPLE_PRELOAD.

The cell consists of a bi-directional pin cell that combines the three signals Output Control -

OCxn, Output Data - ODxn, and Input Data - IDxn, into only a two-stage Shift Register. The port

and pin indexes are not used in the following description

The Boundary-scan logic is not included in the figures in the datasheet. Figure 26-4 shows a

simple digital port pin as described in the section “I/O-Ports” on page 65. The Boundary-scan

details from Figure 26-3 replaces the dashed box in Figure 26-4.

When no alternate port function is present, the Input Data - ID - corresponds to the PINxn Regis-

ter value (but ID has no synchronizer), Output Data corresponds to the PORT Register, Output

Control corresponds to the Data Direction - DD Register, and the Pull-up Enable - PUExn - cor-

responds to logic expression PUD · DDxn · PORTxn.

Digital alternate port functions are connected outside the dotted box in Figure 26-4 to make the

scan chain read the actual pin value. For analog function, there is a direct connection from the

external pin to the analog circuit. There is no scan chain on the interface between the digital and

the analog circuitry, but some digital control signal to analog circuitry are turned off to avoid driv-

ing contention on the pads.

When JTAG IR contains EXTEST or SAMPLE_PRELOAD the clock is not sent out on the port

pins even if the CKOUT fuse is programmed. Even though the clock is output when the JTAG IR

contains SAMPLE_PRELOAD, the clock is not sampled by the boundary scan.

324

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 26-3. Boundary-scan Cell for Bi-directional Port Pin with Pull-up Function.

To Next Cell

ShiftDR

EXTEST

Vcc

Pull-up Enable (PUE)

0

1

Output Control (OC)

FF1

Q

LD1

0

1

0

1

D

Q

G

Output Data (OD)

0

1

FF0

Q

LD0

0

1

0

1

D

G

Q

Input Data (ID)

From Last Cell

ClockDR

UpdateDR

325

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 26-4. General Port Pin Schematic Diagram

See Boundary-scan

Description for Details!

PUExn

PUD

Q

D

DDxn

Q _CLR

WDx

RDx

RESET

OCxn

Q

D

Pxn

PORTxn

ODxn

Q _CLR

WRx

RRx

IDxn

RESET

SLEEP

SYNCHRONIZER

RPx

D

Q

D

L

Q

PINxn

Q

CLK _I/O

PUD:

PULLUP DISABLE

WDx:

RDx:

WRITE DDRx

PUExn:

OCxn:

ODxn:

IDxn:

PULLUP ENABLE for pin Pxn

OUTPUT CONTROL for pin Pxn

OUTPUT DATA to pin Pxn

INPUT DATA from pin Pxn

SLEEP CONTROL

READ DDRx

WRx:

RRx:

WRITE PORTx

READ PORTx REGISTER

READ PORTx PIN

I/O CLOCK

RPx:

SLEEP:

CLK _I/O:

26.6.2

Scanning the RESET Pin

The RESET pin accepts 5V active low logic for standard reset operation, and 12V active high

logic for High Voltage Parallel programming. An observe-only cell as shown in Figure 26-5 is

inserted for the 5V reset signal.

Figure 26-5. Observe-only Cell

To

Cell

From System Pin

To System Logic

FF1

0

1

D

Q

From

ClockDR

326

7766E–AVR–04/10

ATmega16U4/ATmega32U4

26.7 ATmega16U4/ATmega32U4 Boundary-scan Order

Table 26-3 shows the Scan order between TDI and TDO when the Boundary-scan chain is

selected as data path. Bit 0 is the LSB; the first bit scanned in, and the first bit scanned out. The

scan order follows the pin-out order as far as possible. Exceptions from the rules are the Scan

chains for the analog circuits, which constitute the most significant bits of the scan chain regard-

less of which physical pin they are connected to. In Figure 26-3, PXn. Data corresponds to FF0,

PXn. Control corresponds to FF1, PXn. Bit 4, 5, 6 and 7 of Port F is not in the scan chain, since

these pins constitute the TAP pins when the JTAG is enabled. The USB pads are not included in

the boundary-scan.

Table 26-3. ATmega16U4/ATmega32U4 Boundary-scan Order

Bit Number Signal Name

Module

88

87

86

85

84

83

82

81

80

79

78

77

76

75

74

73

72

71

70

69

68

67

66

65

64

63

62

PE6.Data

PE6.Control

Reserved

PB0.Data

PB0.Control

PB1.Data

PB1.Control

PB2.Data

PB2.Control

PB3.Data

PB3.Control

PB4.Data

PB4.Control

PB5.Data

PB5.Control

PB6.Data

PB6.Control

PB7.Data

PB7.Control

Reserved

RSTT

Port E

Port B

PORTE

Reset Logic (Observe Only)

327

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 26-3. ATmega16U4/ATmega32U4 Boundary-scan Order (Continued)

Bit Number Signal Name

Module

61

60

59

58

57

56

55

54

53

52

51

50

49

48

47

46

45

44

43

42

41

40

39

38

37

36

35

34

33

32

31

30

29

28

27

26

PD0.Data

PD0.Control

PD1.Data

PD1.Control

PD2.Data

PD2.Control

PD3.Data

PD3.Control

PD4.Data

PD4.Control

PD5.Data

PD5.Control

PD6.Data

PD6.Control

PD7.Data

PD7.Control

Reserved

Port D

Port E

Reserved

328

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 26-3. ATmega16U4/ATmega32U4 Boundary-scan Order (Continued)

Bit Number Signal Name

Module

25

24

23

22

21

20

19

18

17

16

15

14

13

12

11

10

9

PE2.Data

PE2.Control

Reserved

PF1.Data

PF1.Control

PF0.Data

PF0.Control

Port E

Reserved

8

7

6

5

4

Port F

3

2

1

0

26.8 Boundary-scan Description Language Files

Boundary-scan Description Language (BSDL) files describe Boundary-scan capable devices in

a standard format used by automated test-generation software. The order and function of bits in

the Boundary-scan Data Register are included in this description. BSDL files are available for

ATmega16U4/ATmega32U4.

329

7766E–AVR–04/10

ATmega16U4/ATmega32U4

27. Boot Loader Support – Read-While-Write Self-Programming

The Boot Loader Support provides a real Read-While-Write Self-Programming mechanism for

downloading and uploading program code by the MCU itself. This feature allows flexible applica-

tion software updates controlled by the MCU using a Flash-resident Boot Loader program. The

Boot Loader program can use any available data interface and associated protocol to read code

and write (program) that code into the Flash memory, or read the code from the program mem-

ory. The program code within the Boot Loader section has the capability to write into the entire

Flash, including the Boot Loader memory. The Boot Loader can thus even modify itself, and it

can also erase itself from the code if the feature is not needed anymore. The size of the Boot

Loader memory is configurable with fuses and the Boot Loader has two separate sets of Boot

Lock bits which can be set independently. This gives the user a unique flexibility to select differ-

ent levels of protection. General information on SPM and ELPM is provided in See “AVR CPU

Core” on page 9.

27.1 Boot Loader Features

• Read-While-Write Self-Programming

• Flexible Boot Memory Size

• High Security (Separate Boot Lock Bits for a Flexible Protection)

• Separate Fuse to Select Reset Vector

• Optimized Page⁽¹⁾Size

• Code Efficient Algorithm

• Efficient Read-Modify-Write Support

Note:

1. A page is a section in the Flash consisting of several bytes (see Table 28-11 on page 351)

used during programming. The page organization does not affect normal operation.

27.2 Application and Boot Loader Flash Sections

The Flash memory is organized in two main sections, the Application section and the Boot

Loader section (see Figure 27-2). The size of the different sections is configured by the

BOOTSZ Fuses as shown in Table 27-8 on page 344 and Figure 27-2. These two sections can

have different level of protection since they have different sets of Lock bits.

27.2.1

27.2.2

Application Section

The Application section is the section of the Flash that is used for storing the application code.

The protection level for the Application section can be selected by the application Boot Lock bits

(Boot Lock bits 0), see Table 27-2 on page 334. The Application section can never store any

Boot Loader code since the SPM instruction is disabled when executed from the Application

section.

BLS – Boot Loader Section

While the Application section is used for storing the application code, the The Boot Loader soft-

ware must be located in the BLS since the SPM instruction can initiate a programming when

executing from the BLS only. The SPM instruction can access the entire Flash, including the

BLS itself. The protection level for the Boot Loader section can be selected by the Boot Loader

Lock bits (Boot Lock bits 1), see Table 27-3 on page 334.

27.3 Read-While-Write and No Read-While-Write Flash Sections

Whether the CPU supports Read-While-Write or if the CPU is halted during a Boot Loader soft-

ware update is dependent on which address that is being programmed. In addition to the two

330

7766E–AVR–04/10

ATmega16U4/ATmega32U4

sections that are configurable by the BOOTSZ Fuses as described above, the Flash is also

divided into two fixed sections, the Read-While-Write (RWW) section and the No Read-While-

Write (NRWW) section. The limit between the RWW- and NRWW sections is given in Table 27-

1 and Figure 27-1 on page 332. The main difference between the two sections is:

• When erasing or writing a page located inside the RWW section, the NRWW section can be

read during the operation.

• When erasing or writing a page located inside the NRWW section, the CPU is halted during

the entire operation.

Note that the user software can never read any code that is located inside the RWW section dur-

ing a Boot Loader software operation. The syntax “Read-While-Write section” refers to which

section that is being programmed (erased or written), not which section that actually is being

read during a Boot Loader software update.

27.3.1

RWW – Read-While-Write Section

If a Boot Loader software update is programming a page inside the RWW section, it is possible

to read code from the Flash, but only code that is located in the NRWW section. During an on-

going programming, the software must ensure that the RWW section never is being read. If the

user software is trying to read code that is located inside the RWW section (i.e., by load program

memory, call, or jump instructions or an interrupt) during programming, the software might end

up in an unknown state. To avoid this, the interrupts should either be disabled or moved to the

Boot Loader section. The Boot Loader section is always located in the NRWW section. The

RWW Section Busy bit (RWWSB) in the Store Program Memory Control and Status Register

(SPMCSR) will be read as logical one as long as the RWW section is blocked for reading. After

a programming is completed, the RWWSB must be cleared by software before reading code

located in the RWW section. See “Store Program Memory Control and Status Register –

SPMCSR” on page 336. for details on how to clear RWWSB.

27.3.2

NRWW – No Read-While-Write Section

The code located in the NRWW section can be read when the Boot Loader software is updating

a page in the RWW section. When the Boot Loader code updates the NRWW section, the CPU

is halted during the entire Page Erase or Page Write operation.

Table 27-1. Read-While-Write Features

Which Section does the Z-

pointer Address During the

Programming?

Which Section Can

be Read During

Programming?

Is the CPU

Halted?

Read-While-Write

Supported?

RWW Section

NRWW Section

None

No

Yes

No

NRWW Section

Yes

331

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 27-1. Read-While-Write vs. No Read-While-Write

Read-While-Write

(RWW) Section

Z-pointer

Addresses NRWW

Section

Z-pointer

No Read-While-Write

(NRWW) Section

Addresses RWW

Section

CPU is Halted

During the Operation

Code Located in

NRWW Section

Can be Read During

the Operation

332

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 27-2. Memory Sections

Program Memory

BOOTSZ = '10'

Program Memory

BOOTSZ = '11'

0x0000

Application Flash Section

End RWW

Start NRWW

Application Flash Section

Boot Loader Flash Section

Application Flash Section

End Application

Start Boot Loader

Flashend

Start Boot Loader

Boot Loader Flash Section

Flashend

Program Memory

BOOTSZ = '01'

Program Memory

BOOTSZ = '00'

0x0000

Application Flash Section

End RWW, End Application

End RWW

Start NRWW, Start Boot Loader

Start NRWW

Application Flash Section

Boot Loader Flash Section

End Application

Boot Loader Flash Section

Start Boot Loader

Flashend

Note:

1. The parameters in the figure above are given in Table 27-8 on page 344.

27.4 Boot Loader Lock Bits

If no Boot Loader capability is needed, the entire Flash is available for application code. The

Boot Loader has two separate sets of Boot Lock bits which can be set independently. This gives

the user a unique flexibility to select different levels of protection.

The user can select:

• To protect the entire Flash from a software update by the MCU.

• To protect only the Boot Loader Flash section from a software update by the MCU.

• To protect only the Application Flash section from a software update by the MCU.

• Allow software update in the entire Flash.

See Table 27-2 and Table 27-3 for further details. The Boot Lock bits can be set by software and

in Serial or in Parallel Programming mode. They can only be cleared by a Chip Erase command

only. The general Write Lock (Lock Bit mode 2) does not control the programming of the Flash

memory by SPM instruction. Similarly, the general Read/Write Lock (Lock Bit mode 1) does not

control reading nor writing by (E)LPM/SPM, if it is attempted.

333

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 27-2. Boot Lock Bit0 Protection Modes (Application Section)⁽¹⁾

BLB0 Mode BLB02 BLB01 Protection

No restrictions for SPM or (E)LPM accessing the

Application section.

1

2

1

0

SPM is not allowed to write to the Application section.

SPM is not allowed to write to the Application section, and

(E)LPM executing from the Boot Loader section is not

allowed to read from the Application section. If Interrupt

Vectors are placed in the Boot Loader section, interrupts

are disabled while executing from the Application section.

3

4

0

1

(E)LPM executing from the Boot Loader section is not

allowed to read from the Application section. If Interrupt

Vectors are placed in the Boot Loader section, interrupts

are disabled while executing from the Application section.

Note:

1. “1” means unprogrammed, “0” means programmed

Table 27-3. Boot Lock Bit1 Protection Modes (Boot Loader Section)⁽¹⁾

BLB1 Mode BLB12 BLB11 Protection

No restrictions for SPM or (E)LPM accessing the Boot

Loader section.

1

2

1

0

SPM is not allowed to write to the Boot Loader section.

SPM is not allowed to write to the Boot Loader section,

and (E)LPM executing from the Application section is not

allowed to read from the Boot Loader section. If Interrupt

Vectors are placed in the Application section, interrupts

are disabled while executing from the Boot Loader section.

3

4

0

(E)LPM executing from the Application section is not

allowed to read from the Boot Loader section. If Interrupt

Vectors are placed in the Application section, interrupts

are disabled while executing from the Boot Loader section.

1

Note:

1. “1” means unprogrammed, “0” means programmed

27.5 Entering the Boot Loader Program

The bootloader can be executed with three different conditions:

27.5.1

27.5.2

Regular application conditions.

A jump or call from the application program. This may be initiated by a trigger such as a com-

mand received via USART, SPI or USB.

Boot Reset Fuse

The Boot Reset Fuse (BOOTRST) can be programmed so that the Reset Vector is pointing to

the Boot Flash start address after a reset. In this case, the Boot Loader is started after a reset.

After the application code is loaded, the program can start executing the application code. Note

that the fuses cannot be changed by the MCU itself. This means that once the Boot Reset Fuse

334

7766E–AVR–04/10

ATmega16U4/ATmega32U4

is programmed, the Reset Vector will always point to the Boot Loader Reset and the fuse can

only be changed through the serial or parallel programming interface.

Table 27-4. Boot Reset Fuse⁽¹⁾

BOOTRST

Reset Address

1

0

Reset Vector = Application Reset (address 0x0000)

Reset Vector = Boot Loader Reset (see Table 27-8 on page 344)

Note:

1. “1” means unprogrammed, “0” means programmed

27.5.3

External Hardware conditions

The Hardware Boot Enable Fuse (HWBE) can be programmed (See Table 27-5) so that upon

special hardware conditions under reset, the bootloader execution is forced after reset.

Table 27-5. Hardware Boot Enable Fuse⁽¹⁾

HWBE

Reset Address

1

0

ALE/HWB pin can not be used to force Boot Loader execution after reset

ALE/HWB pin is used during reset to force bootloader execution after reset

Note:

1. “1” means unprogrammed, “0” means programmed

When the HWBE fuse is enable the ALE/HWB pin is configured as input during reset and sam-

pled during reset rising edge. When ALE/HWB pin is ‘0’ during reset rising edge, the reset vector

will be set as the Boot Loader Reset address and the Boot Loader will be executed (See Figures

27-3).

Figure 27-3. Boot Process Description

RESET

t_SHRH

t_HHRH

ALE/HWB

HWBE ?

Ext. Hardware

Conditions ?

BOOTRST ?

Reset Vector = Application Reset

Reset Vector =Boot Lhoader Reset

335

7766E–AVR–04/10

ATmega16U4/ATmega32U4

27.5.4

Store Program Memory Control and Status Register – SPMCSR

The Store Program Memory Control and Status Register contains the control bits needed to con-

trol the Boot Loader operations.

Bit

7

6

5

4

3

2

1

0

SPMIE

R/W

0

RWWSB

SIGRD

R/W

0

RWWSRE

BLBSET

PGWRT

R/W

0

PGERS

R/W

0

SPMEN

R/W

0

SPMCSR

Read/Write

Initial Value

R

0

R/W

0

R/W

0

• Bit 7 – SPMIE: SPM Interrupt Enable

When the SPMIE bit is written to one, and the I-bit in the Status Register is set (one), the SPM

ready interrupt will be enabled. The SPM ready Interrupt will be executed as long as the SPMEN

bit in the SPMCSR Register is cleared.

• Bit 6 – RWWSB: Read-While-Write Section Busy

When a Self-Programming (Page Erase or Page Write) operation to the RWW section is initi-

ated, the RWWSB will be set (one) by hardware. When the RWWSB bit is set, the RWW section

cannot be accessed. The RWWSB bit will be cleared if the RWWSRE bit is written to one after a

Self-Programming operation is completed. Alternatively the RWWSB bit will automatically be

cleared if a page load operation is initiated.

• Bit 5 – SIGRD: Signature Row Read

If this bit is written to one at the same time as SPMEN, the next LPM instruction within three

clock cycles will read a byte from the signature row into the destination register. see “Reading

the Signature Row from Software” on page 341 for details. An SPM instruction within four cycles

after SIGRD and SPMEN are set will have no effect. This operation is reserved for future use

and should not be used.

• Bit 4 – RWWSRE: Read-While-Write Section Read Enable

When programming (Page Erase or Page Write) to the RWW section, the RWW section is

blocked for reading (the RWWSB will be set by hardware). To re-enable the RWW section, the

user software must wait until the programming is completed (SPMEN will be cleared). Then, if

the RWWSRE bit is written to one at the same time as SPMEN, the next SPM instruction within

four clock cycles re-enables the RWW section. The RWW section cannot be re-enabled while

the Flash is busy with a Page Erase or a Page Write (SPMEN is set). If the RWWSRE bit is writ-

ten while the Flash is being loaded, the Flash load operation will abort and the data loaded will

be lost.

• Bit 3 – BLBSET: Boot Lock Bit Set

If this bit is written to one at the same time as SPMEN, the next SPM instruction within four clock

cycles sets Boot Lock bits, according to the data in R0. The data in R1 and the address in the Z-

pointer are ignored. The BLBSET bit will automatically be cleared upon completion of the Lock

bit set, or if no SPM instruction is executed within four clock cycles.

An (E)LPM instruction within three cycles after BLBSET and SPMEN are set in the SPMCSR

Register, will read either the Lock bits or the Fuse bits (depending on Z0 in the Z-pointer) into the

destination register. See “Reading the Fuse and Lock Bits from Software” on page 340 for

details.

336

7766E–AVR–04/10

ATmega16U4/ATmega32U4

• Bit 2 – PGWRT: Page Write

If this bit is written to one at the same time as SPMEN, the next SPM instruction within four clock

cycles executes Page Write, with the data stored in the temporary buffer. The page address is

taken from the high part of the Z-pointer. The data in R1 and R0 are ignored. The PGWRT bit

will auto-clear upon completion of a Page Write, or if no SPM instruction is executed within four

clock cycles. The CPU is halted during the entire Page Write operation if the NRWW section is

addressed.

• Bit 1 – PGERS: Page Erase

If this bit is written to one at the same time as SPMEN, the next SPM instruction within four clock

cycles executes Page Erase. The page address is taken from the high part of the Z-pointer. The

data in R1 and R0 are ignored. The PGERS bit will auto-clear upon completion of a Page Erase,

or if no SPM instruction is executed within four clock cycles. The CPU is halted during the entire

Page Write operation if the NRWW section is addressed.

• Bit 0 – SPMEN: Store Program Memory Enable

This bit enables the SPM instruction for the next four clock cycles. If written to one together with

either RWWSRE, BLBSET, PGWRT’ or PGERS, the following SPM instruction will have a spe-

cial meaning, see description above. If only SPMEN is written, the following SPM instruction will

store the value in R1:R0 in the temporary page buffer addressed by the Z-pointer. The LSB of

the Z-pointer is ignored. The SPMEN bit will auto-clear upon completion of an SPM instruction,

or if no SPM instruction is executed within four clock cycles. During Page Erase and Page Write,

the SPMEN bit remains high until the operation is completed.

Writing any other combination than “10001”, “01001”, “00101”, “00011” or “00001” in the lower

five bits will have no effect.

Note:

Only one SPM instruction should be active at any time.

27.6 Addressing the Flash During Self-Programming

The Z-pointer is used to address the SPM commands. The Z pointer consists of the Z-registers

ZL and ZH in the register file, and RAMPZ in the I/O space. The number of bits actually used is

implementation dependent. Note that the RAMPZ register is only implemented when the pro-

gram space is larger than 64K bytes.

Bit

23

22

21

20

19

18

17

16

15

14

13

12

11

10

9

8

RAMPZ

ZH (R31)

ZL (R30)

RAMPZ7

RAMPZ6

RAMPZ5

RAMPZ4

RAMPZ3

RAMPZ2

RAMPZ1

RAMPZ0

Z15

Z7

7

Z14

Z6

6

Z13

Z5

5

Z12

Z4

4

Z11

Z3

3

Z10

Z2

2

Z9

Z1

1

Z8

Z0

0

Since the Flash is organized in pages (see Table 28-11 on page 351), the Program Counter can

be treated as having two different sections. One section, consisting of the least significant bits, is

addressing the words within a page, while the most significant bits are addressing the pages.

This is shown in Figure 27-4. Note that the Page Erase and Page Write operations are

addressed independently. Therefore it is of major importance that the Boot Loader software

addresses the same page in both the Page Erase and Page Write operation. Once a program-

ming operation is initiated, the address is latched and the Z-pointer can be used for other

operations.

The (E)LPM instruction use the Z-pointer to store the address. Since this instruction addresses

the Flash byte-by-byte, also bit Z0 of the Z-pointer is used.

337

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 27-4. Addressing the Flash During SPM⁽¹⁾

BIT 23

ZPCMSB

ZPAGEMSB

1

0

Z - POINTER

PCMSB

PAGEMSB

PROGRAM COUNTER

PCPAGE

PCWORD

PAGE ADDRESS

WITHIN THE FLASH

WORD ADDRESS

WITHIN A PAGE

PROGRAM MEMORY

PAGE

INSTRUCTION WORD

PCWORD[PAGEMSB:0]:

00

01

02

PAGEEND

Note:

1. The different variables used in Figure 27-4 are listed in Table 27-10 on page 345.

27.7 Self-Programming the Flash

The program memory is updated in a page by page fashion. Before programming a page with

the data stored in the temporary page buffer, the page must be erased. The temporary page buf-

fer is filled one word at a time using SPM and the buffer can be filled either before the Page

Erase command or between a Page Erase and a Page Write operation:

Alternative 1, fill the buffer before a Page Erase

• Fill temporary page buffer

• Perform a Page Erase

• Perform a Page Write

Alternative 2, fill the buffer after Page Erase

• Perform a Page Erase

• Fill temporary page buffer

• Perform a Page Write

If only a part of the page needs to be changed, the rest of the page must be stored (for example

in the temporary page buffer) before the erase, and then be rewritten. When using alternative 1,

the Boot Loader provides an effective Read-Modify-Write feature which allows the user software

to first read the page, do the necessary changes, and then write back the modified data. If alter-

native 2 is used, it is not possible to read the old data while loading since the page is already

erased. The temporary page buffer can be accessed in a random sequence. It is essential that

the page address used in both the Page Erase and Page Write operation is addressing the same

338

7766E–AVR–04/10

ATmega16U4/ATmega32U4

page. See “Simple Assembly Code Example for a Boot Loader” on page 342 for an assembly

code example.

27.7.1

Performing Page Erase by SPM

To execute Page Erase, set up the address in the Z-pointer, write “X0000011” to SPMCSR and

execute SPM within four clock cycles after writing SPMCSR. The data in R1 and R0 is ignored.

The page address must be written to PCPAGE in the Z-register. Other bits in the Z-pointer will

be ignored during this operation.

• Page Erase to the RWW section: The NRWW section can be read during the Page Erase.

• Page Erase to the NRWW section: The CPU is halted during the operation.

27.7.2

Filling the Temporary Buffer (Page Loading)

To write an instruction word, set up the address in the Z-pointer and data in R1:R0, write

“00000001” to SPMCSR and execute SPM within four clock cycles after writing SPMCSR. The

content of PCWORD in the Z-register is used to address the data in the temporary buffer. The

temporary buffer will auto-erase after a Page Write operation or by writing the RWWSRE bit in

SPMCSR. It is also erased after a system reset. Note that it is not possible to write more than

one time to each address without erasing the temporary buffer.

If the EEPROM is written in the middle of an SPM Page Load operation, all data loaded will be

lost.

27.7.3

Performing a Page Write

To execute Page Write, set up the address in the Z-pointer, write “X0000101” to SPMCSR and

execute SPM within four clock cycles after writing SPMCSR. The data in R1 and R0 is ignored.

The page address must be written to PCPAGE. Other bits in the Z-pointer must be written to

zero during this operation.

• Page Write to the RWW section: The NRWW section can be read during the Page Write.

• Page Write to the NRWW section: The CPU is halted during the operation.

27.7.4

27.7.5

27.7.6

Using the SPM Interrupt

If the SPM interrupt is enabled, the SPM interrupt will generate a constant interrupt when the

SPMEN bit in SPMCSR is cleared. This means that the interrupt can be used instead of polling

the SPMCSR Register in software. When using the SPM interrupt, the Interrupt Vectors should

be moved to the BLS section to avoid that an interrupt is accessing the RWW section when it is

blocked for reading. How to move the interrupts is described in “Interrupts” on page 61.

Consideration While Updating BLS

Special care must be taken if the user allows the Boot Loader section to be updated by leaving

Boot Lock bit11 unprogrammed. An accidental write to the Boot Loader itself can corrupt the

entire Boot Loader, and further software updates might be impossible. If it is not necessary to

change the Boot Loader software itself, it is recommended to program the Boot Lock bit11 to

protect the Boot Loader software from any internal software changes.

Prevent Reading the RWW Section During Self-Programming

During Self-Programming (either Page Erase or Page Write), the RWW section is always

blocked for reading. The user software itself must prevent that this section is addressed during

the self programming operation. The RWWSB in the SPMCSR will be set as long as the RWW

section is busy. During Self-Programming the Interrupt Vector table should be moved to the BLS

339

7766E–AVR–04/10

ATmega16U4/ATmega32U4

as described in “Interrupts” on page 61, or the interrupts must be disabled. Before addressing

the RWW section after the programming is completed, the user software must clear the

RWWSB by writing the RWWSRE. See “Simple Assembly Code Example for a Boot Loader” on

page 342 for an example.

27.7.7

Setting the Boot Loader Lock Bits by SPM

To set the Boot Loader Lock bits, write the desired data to R0, write “X0001001” to SPMCSR

and execute SPM within four clock cycles after writing SPMCSR. The only accessible Lock bits

are the Boot Lock bits that may prevent the Application and Boot Loader section from any soft-

ware update by the MCU.

Bit

7

6

5

4

3

2

1

0

R0

1

BLB12

BLB11

BLB02

BLB01

1

See Table 27-2 and Table 27-3 for how the different settings of the Boot Loader bits affect the

Flash access.

If bits 5..2 in R0 are cleared (zero), the corresponding Boot Lock bit will be programmed if an

SPM instruction is executed within four cycles after BLBSET and SPMEN are set in SPMCSR.

The Z-pointer is don’t care during this operation, but for future compatibility it is recommended to

load the Z-pointer with 0x0001 (same as used for reading the lO_ckbits). For future compatibility it

is also recommended to set bits 7, 6, 1, and 0 in R0 to “1” when writing the Lock bits. When pro-

gramming the Lock bits the entire Flash can be read during the operation.

27.7.8

27.7.9

EEPROM Write Prevents Writing to SPMCSR

Note that an EEPROM write operation will block all software programming to Flash. Reading the

Fuses and Lock bits from software will also be prevented during the EEPROM write operation. It

is recommended that the user checks the status bit (EEPE) in the EECR Register and verifies

that the bit is cleared before writing to the SPMCSR Register.

Reading the Fuse and Lock Bits from Software

It is possible to read both the Fuse and Lock bits from software. To read the Lock bits, load the

Z-pointer with 0x0001 and set the BLBSET and SPMEN bits in SPMCSR. When an (E)LPM

instruction is executed within three CPU cycles after the BLBSET and SPMEN bits are set in

SPMCSR, the value of the Lock bits will be loaded in the destination register. The BLBSET and

SPMEN bits will auto-clear upon completion of reading the Lock bits or if no (E)LPM instruction

is executed within three CPU cycles or no SPM instruction is executed within four CPU cycles.

When BLBSET and SPMEN are cleared, (E)LPM will work as described in the Instruction set

Manual.

Bit

Rd

7

6

5

4

3

2

1

0

–

BLB12

BLB11

BLB02

BLB01

LB2

LB1

The algorithm for reading the Fuse Low byte is similar to the one described above for reading

the Lock bits. To read the Fuse Low byte, load the Z-pointer with 0x0000 and set the BLBSET

and SPMEN bits in SPMCSR. When an (E)LPM instruction is executed within three cycles after

the BLBSET and SPMEN bits are set in the SPMCSR, the value of the Fuse Low byte (FLB) will

be loaded in the destination register as shown below. Refer to Table 28-5 on page 348 for a

detailed description and mapping of the Fuse Low byte.

Bit

Rd

7

6

5

4

3

2

1

0

FLB7

FLB6

FLB5

FLB4

FLB3

FLB2

FLB1

FLB0

340

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Similarly, when reading the Fuse High byte, load 0x0003 in the Z-pointer. When an (E)LPM

instruction is executed within three cycles after the BLBSET and SPMEN bits are set in the

SPMCSR, the value of the Fuse High byte (FHB) will be loaded in the destination register as

shown below. Refer to Table 28-4 on page 348 for detailed description and mapping of the Fuse

High byte.

Bit

Rd

7

6

5

4

3

2

1

0

FHB7

FHB6

FHB5

FHB4

FHB3

FHB2

FHB1

FHB0

When reading the Extended Fuse byte, load 0x0002 in the Z-pointer. When an (E)LPM instruc-

tion is executed within three cycles after the BLBSET and SPMEN bits are set in the SPMCSR,

the value of the Extended Fuse byte (EFB) will be loaded in the destination register as shown

below. Refer to Table 28-3 on page 347 for detailed description and mapping of the Extended

Fuse byte.

Bit

Rd

7

6

5

4

3

2

1

0

–

EFB2

EFB1

EFB0

Fuse and Lock bits that are programmed, will be read as zero. Fuse and Lock bits that are

unprogrammed, will be read as one.

27.7.10 Reading the Signature Row from Software

To read the Signature Row from software, load the Z-pointer with the signature byte address

given in Table 27-6 on page 341 and set the SIGRD and SPMEN bits in SPMCSR. When an

LPM instruction is executed within three CPU cycles after the SIGRD and SPMEN bits are set in

SPMCSR, the signature byte value will be loaded in the destination register. The SIGRD and

SPMEN bits will auto-clear upon completion of reading the Signature Row Lock bits or if no LPM

instruction is executed within three CPU cycles. When SIGRD and SPMEN are cleared, LPM will

work as described in the Instruction set Manual.

Table 27-6. Signature Row Addressing

Signature Byte

Z-Pointer Address

0x0000

Device Signature Byte 1

Device Signature Byte 2

Device Signature Byte 3

RC Oscillator Calibration Byte

0x0002

0x0004

0x0001

Note:

All other addresses are reserved for future use.

27.7.11 Preventing Flash Corruption

During periods of low V_CC, the Flash program can be corrupted because the supply voltage is

too low for the CPU and the Flash to operate properly. These issues are the same as for board

level systems using the Flash, and the same design solutions should be applied.

A Flash program corruption can be caused by two situations when the voltage is too low. First, a

regular write sequence to the Flash requires a minimum voltage to operate correctly. Secondly,

the CPU itself can execute instructions incorrectly, if the supply voltage for executing instructions

is too low.

Flash corruption can easily be avoided by following these design recommendations (one is

sufficient):

341

7766E–AVR–04/10

ATmega16U4/ATmega32U4

1. If there is no need for a Boot Loader update in the system, program the Boot Loader

Lock bits to prevent any Boot Loader software updates.

2. Keep the AVR RESET active (low) during periods of insufficient power supply voltage.

This can be done by enabling the internal Brown-out Detector (BOD) if the operating

voltage matches the detection level. If not, an external low V_CCreset protection circuit

can be used. If a reset occurs while a write operation is in progress, the write operation

will be completed provided that the power supply voltage is sufficient.

3. Keep the AVR core in Power-down sleep mode during periods of low V_CC. This will pre-

vent the CPU from attempting to decode and execute instructions, effectively protecting

the SPMCSR Register and thus the Flash from unintentional writes.

27.7.12 Programming Time for Flash when Using SPM

The calibrated RC Oscillator is used to time Flash accesses. Table 27-7 shows the typical pro-

gramming time for Flash accesses from the CPU.

Table 27-7. SPM Programming Time

Symbol

Min Programming Time Max Programming Time

Flash write (Page Erase, Page Write,

and write Lock bits by SPM)

3.7 ms 4.5 ms

27.7.13 Simple Assembly Code Example for a Boot Loader

;-the routine writes one page of data from RAM to Flash

; the first data location in RAM is pointed to by the Y pointer

; the first data location in Flash is pointed to by the Z-pointer

;-error handling is not included

;-the routine must be placed inside the Boot space

; (at least the Do_spm sub routine). Only code inside NRWW section can

; be read during Self-Programming (Page Erase and Page Write).

;-registers used: r0, r1, temp1 (r16), temp2 (r17), looplo (r24),

; loophi (r25), spmcrval (r20)

; storing and restoring of registers is not included in the routine

; register usage can be optimized at the expense of code size

;-It is assumed that either the interrupt table is moved to the Boot

; loader section or that the interrupts are disabled.

.equ PAGESIZEB = PAGESIZE*2

.org SMALLBOOTSTART

Write_page:

;PAGESIZEB is page size in BYTES, not words

; Page Erase

ldi spmcrval, (1<<PGERS) | (1<<SPMEN)

call Do_spm

; re-enable the RWW section

ldi spmcrval, (1<<RWWSRE) | (1<<SPMEN)

call Do_spm

; transfer data from RAM to Flash page buffer

ldi looplo, low(PAGESIZEB)

;init loop variable

ldi loophi, high(PAGESIZEB) ;not required for PAGESIZEB<=256

Wrloop:

ld

r0, Y+

r1, Y+

ldi spmcrval, (1<<SPMEN)

call Do_spm

adiw ZH:ZL, 2

sbiw loophi:looplo, 2

;use subi for PAGESIZEB<=256

342

7766E–AVR–04/10

ATmega16U4/ATmega32U4

brne Wrloop

; execute Page Write

subi ZL, low(PAGESIZEB)

sbci ZH, high(PAGESIZEB)

;restore pointer

;not required for PAGESIZEB<=256

ldi spmcrval, (1<<PGWRT) | (1<<SPMEN)

call Do_spm

; re-enable the RWW section

ldi spmcrval, (1<<RWWSRE) | (1<<SPMEN)

call Do_spm

; read back and check, optional

ldi looplo, low(PAGESIZEB)

;init loop variable

ldi loophi, high(PAGESIZEB) ;not required for PAGESIZEB<=256

subi YL, low(PAGESIZEB)

sbci YH, high(PAGESIZEB)

Rdloop:

;restore pointer

elpm r0, Z+

ld

r1, Y+

cpse r0, r1

jmp Error

sbiw loophi:looplo, 1

brne Rdloop

;use subi for PAGESIZEB<=256

; return to RWW section

; verify that RWW section is safe to read

Return:

in

temp1, SPMCSR

sbrs temp1, RWWSB

ret

; If RWWSB is set, the RWW section is not ready yet

; re-enable the RWW section

ldi spmcrval, (1<<RWWSRE) | (1<<SPMEN)

call Do_spm

rjmp Return

Do_spm:

; check for previous SPM complete

Wait_spm:

in

temp1, SPMCSR

sbrc temp1, SPMEN

rjmp Wait_spm

; input: spmcrval determines SPM action

; disable interrupts if enabled, store status

in

temp2, SREG

cli

; check that no EEPROM write access is present

Wait_ee:

sbic EECR, EEPE

rjmp Wait_ee

; SPM timed sequence

out SPMCSR, spmcrval

spm

; restore SREG (to enable interrupts if originally enabled)

out SREG, temp2

ret

343

7766E–AVR–04/10

ATmega16U4/ATmega32U4

27.7.14 ATmega16U4/ATmega32U4 Boot Loader Parameters

In Table 27-8 through Table 27-10, the parameters used in the description of the Self-Program-

ming are given.

Table 27-8. Boot Size Configuration (Word Addresses)⁽¹⁾

1

0

1

0

1

0

1

0

1

0

1

0

256 words

512 words

1024 words

2048 words

256 words

512 words

1024 words

2048 words

4

8

0x0000 - 0x3EFF

0x0000 - 0x3DFF

0x3F00 - 0x3FFF

0x3E00 - 0x3FFF

0x3C00 - 0x3FFF

0x3800 - 0x3FFF

0x1F00 - 0x1FFF

0x1E00 - 0x1FFF

0x1C00 - 0x1FFF

0x1800 - 0x1FFF

0x3EFF

0x3DFF

0x3BFF

0x37FF

0x1EFF

0x1DFF

0x1BFF

0x17FF

0x3F00

0x3E00

0x3C00

0x3800

0x1F00

0x1E00

0x1C00

0x1800

16 0x0000 - 0x3BFF

32 0x0000 - 0x37FF

4

8

0x0000 - 0x1EFF

0x0000 - 0x1DFF

16 0x0000 - 0x1BFF

32 0x0000 - 0x17FF

Note:

1. The different BOOTSZ Fuse configurations are shown in Figure 27-2

Table 27-9. Read-While-Write Limit (Word Addresses)⁽¹⁾

Device

Section

Pages

224

32

Address

Read-While-Write section (RWW)

No Read-While-Write section (NRWW)

Read-While-Write section (RWW)

No Read-While-Write section (NRWW)

0x0000 - 0x37FF

0x3800 - 0x3FFF

0x0000 - 0x17FF

0x1800 - 0x1FFF

ATmega32U4

97

ATmega16U4

32

Note:

1. For details about these two section, see “NRWW – No Read-While-Write Section” on page

331 and “RWW – Read-While-Write Section” on page 331.

344

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 27-10. Explanation of different variables used in Figure 27-4 and the mapping to the Z-

pointer

Corresponding

Variable

Z-value⁽¹⁾

Description

Most significant bit in the Program Counter. (The

Program Counter is 14 bits PC[13:0])

PCMSB

13

6

Most significant bit which is used to address the

words within one page (64 words in a page requires

six bits PC [5:0]).

PAGEMSB

ZPCMSB

Bit in Z-pointer that is mapped to PCMSB. Because

Z0 is not used, the ZPCMSB equals PCMSB + 1.

Z14

Z7

Bit in Z-pointer that is mapped to PCMSB. Because

Z0 is not used, the ZPAGEMSB equals PAGEMSB +

1.

ZPAGEMSB

PCPAGE

Program Counter page address: Page select, for

Page Erase and Page Write

PC[13:6]

PC[5:0]

Z14:Z7

Z6:Z1

Program Counter word address: Word select, for

filling temporary buffer (must be zero during Page

Write operation)

PCWORD

Note:

1. Z0: should be zero for all SPM commands, byte select for the (E)LPM instruction.

See “Addressing the Flash During Self-Programming” on page 337 for details about the use of Z-

pointer during Self-Programming.

345

7766E–AVR–04/10

ATmega16U4/ATmega32U4

28. Memory Programming

28.1 Program And Data Memory Lock Bits

The ATmega16U4/ATmega32U4 provides six Lock bits which can be left unprogrammed (“1”) or

can be programmed (“0”) to obtain the additional features listed in Table 28-2. The Lock bits can

only be erased to “1” with the Chip Erase command.

Table 28-1. Lock Bit Byte⁽¹⁾

Lock Bit Byte

Bit No

Description

–

Default Value

7

6

5

4

3

2

1

0

1 (unprogrammed)

0 (programmed)

1 (unprogrammed)

0 (programmed)

–

BLB12

Boot Lock bit

Lock bit

BLB11

BLB02

BLB01

LB2

LB1

Lock bit

Note:

Table 28-2. Lock Bit Protection Modes⁽¹⁾⁽²⁾

Memory Lock Bits Protection Type

1. “1” means unprogrammed, “0” means programmed

LB Mode

LB2

LB1

1

No memory lock features enabled.

Further programming of the Flash and EEPROM is

disabled in Parallel and Serial Programming mode. The

Fuse bits are locked in both Serial and Parallel

Programming mode.⁽¹⁾

2

1

0

Further programming and verification of the Flash and

EEPROM is disabled in Parallel and Serial Programming

mode. The Boot Lock bits and Fuse bits are locked in both

Serial and Parallel Programming mode.⁽¹⁾

3

BLB0 Mode BLB02 BLB01

No restrictions for SPM or (E)LPM accessing the

Application section.

1

2

1

0

SPM is not allowed to write to the Application section.

SPM is not allowed to write to the Application section, and

(E)LPM executing from the Boot Loader section is not

allowed to read from the Application section. If Interrupt

Vectors are placed in the Boot Loader section, interrupts

are disabled while executing from the Application section.

3

4

0

1

(E)LPM executing from the Boot Loader section is not

allowed to read from the Application section. If Interrupt

Vectors are placed in the Boot Loader section, interrupts

are disabled while executing from the Application section.

346

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 28-2. Lock Bit Protection Modes⁽¹⁾⁽²⁾(Continued)

Memory Lock Bits

Protection Type

BLB1 Mode BLB12 BLB11

No restrictions for SPM or (E)LPM accessing the Boot

Loader section.

1

2

1

0

SPM is not allowed to write to the Boot Loader section.

SPM is not allowed to write to the Boot Loader section,

and (E)LPM executing from the Application section is not

allowed to read from the Boot Loader section. If Interrupt

Vectors are placed in the Application section, interrupts

are disabled while executing from the Boot Loader section.

3

0

(E)LPM executing from the Application section is not

allowed to read from the Boot Loader section. If Interrupt

Vectors are placed in the Application section, interrupts

are disabled while executing from the Boot Loader section.

4

0

1

Notes: 1. Program the Fuse bits and Boot Lock bits before programming the LB1 and LB2.

2. “1” means unprogrammed, “0” means programmed

28.2 Fuse Bits

The ATmega16U4/ATmega32U4 has three bytes. Table 28-3 - Table 28-5 describe briefly the

functionality of all the fuses and how they are mapped into the Fuse bytes. Note that the fuses

are read as logical zero, “0”, if they are programmed.

Table 28-3. Extended Fuse Byte

Fuse Low Byte

Bit No

Description

Default Value

–

7

–

1

6

5

4

3

2

1

0

–

1

–

1

–

1

HWBE

Hardware Boot Enable

Brown-out Detector trigger level

0 (programmed)

1 (unprogrammed)

BODLEVEL2⁽¹⁾

BODLEVEL1⁽¹⁾

BODLEVEL0⁽¹⁾

Note:

1. See Table 8-2 on page 52 for BODLEVEL Fuse decoding.

347

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 28-4. Fuse High Byte

Fuse High Byte Bit No Description

Default Value

1 (unprogrammed, OCD

disabled)

OCDEN⁽⁴⁾

7

6

Enable OCD

Enable JTAG

0 (programmed, JTAG

enabled)

JTAGEN

Enable Serial Program and Data

Downloading

0 (programmed, SPI prog.

enabled)

SPIEN⁽¹⁾

WDTON⁽³⁾

EESAVE

5

4

3

Watchdog Timer always on

1 (unprogrammed)

EEPROM memory is preserved

through the Chip Erase

0 (programmed, EEPROM

preserved)

Select Boot Size (see Table 28-7

for details)

BOOTSZ1

BOOTSZ0

BOOTRST

2

1

0

0 (programmed)⁽²⁾

Select Boot Size (see Table 28-7

for details)

Select Bootloader Address as

Reset Vector

1 (unprogrammed, Reset

vector @0x0000)

Note:

1. The SPIEN Fuse is not accessible in serial programming mode.

2. The default value of BOOTSZ1..0 results in maximum Boot Size. See Table 27-8 on page 344

for details.

3. See “Watchdog Timer” on page 55 for details.

4. Never ship a product with the OCDEN Fuse programmed regardless of the setting of Lock bits

and JTAGEN Fuse. A programmed OCDEN Fuse enables some parts of the clock system to

be running in all sleep modes. This may increase the power consumption.

Table 28-5. Fuse Low Byte

Fuse Low Byte

Bit Nr

Description

Default Value

ATmega16U4/32U4 ATmega16U4RC/32U4RC

CKDIV8⁽³⁾

CKOUT⁽²⁾

SUT1

7

6

5

4

3

2

1

0

Divide clock by 8

Clock output

0 (programmed)

1 (unprogrammed)

0 (programmed)

1 (unprogrammed)

Select start-up time

SUT0

Select start-up time

CKSEL3

CKSEL2

CKSEL1

CKSEL0

Select Clock source 1 (unprogrammed)⁽¹⁾

Select Clock source 0 (programmed)⁽¹⁾

0 (programmed)⁽¹⁾

1 (unprogrammed)⁽¹⁾

0 (programmed)⁽¹⁾

Note:

1. The default setting of CKSEL3..0 results in Low Power Crystal Oscillator for ATmega16U4 and

ATmega32U4, and Internal RC oscillator for ATmega16U4RC and ATmega32U4RC. See

Table 6-1 on page 28 for details.

2. The CKOUT Fuse allow the system clock to be output on PORTC7. See “Clock Output Buffer”

on page 37 for details.

3. See “System Clock Prescaler” on page 37 for details.

The status of the Fuse bits is not affected by Chip Erase. Note that the Fuse bits are locked if

Lock bit1 (LB1) is programmed. Program the Fuse bits before programming the Lock bits.

348

7766E–AVR–04/10

ATmega16U4/ATmega32U4

28.2.1

Latching of Fuses

The fuse values are latched when the device enters programming mode and changes of the

fuse values will have no effect until the part leaves Programming mode. This does not apply to

the EESAVE Fuse which will take effect once it is programmed. The fuses are also latched on

Power-up in Normal mode.

28.3 Signature Bytes

All ATMEL microcontrollers have a three-byte signature code which identifies the device. This

code can be read in both serial and parallel mode, also when the device is locked. The three

bytes reside in a separate address space.

ATmega16U4 Signature Bytes:

1. 0x000: 0x1E (indicates manufactured by ATMEL).

2. 0x001: 0x94 (indicates 16KB Flash memory).

3. 0x002: 0x88 (indicates ATmega16U4 device).

ATmega32U4 Signature Bytes:

1. 0x000: 0x1E (indicates manufactured by ATMEL).

2. 0x001: 0x95 (indicates 32KB Flash memory).

3. 0x002: 0x87 (indicates ATmega32U4 device).

28.4 Calibration Byte

The ATmega16U4/ATmega32U4 has a byte calibration value for the internal RC Oscillator. This

byte resides in the high byte of address 0x000 in the signature address space. During reset, this

byte is automatically written into the OSCCAL Register to ensure correct frequency of the cali-

brated RC Oscillator.

28.5 Parallel Programming Parameters, Pin Mapping, and Commands

This section describes how to parallel program and verify Flash Program memory, EEPROM

Data memory, Memory Lock bits, and Fuse bits in the ATmega16U4/ATmega32U4. Pulses are

assumed to be at least 250 ns unless otherwise noted.

28.5.1

Signal Names

In this section, some pins of the ATmega16U4/ATmega32U4 are referenced by signal names

describing their functionality during parallel programming, see Figure 28-1 and Table 28-6. Pins

not described in the following table are referenced by pin names.

The XA1/XA0 pins determine the action executed when the XTAL1 pin is given a positive pulse.

The bit coding is shown in Table 28-9.

When pulsing WR or OE, the command loaded determines the action executed. The different

commands are shown in Table 28-10.

349

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 28-1. Parallel Programming⁽¹⁾

+5V

RDY/BSY

OE

PD1

PD2

PD3

PD4

PD5

PD6

PD7

VCC

+5V

WR

AVCC

BS1

PB7 - PB0

DATA

XA0

XA1

PAGEL

+12 V

BS2

RESET

PE6

XTAL1

GND

Note:

1. Unused Pins should be left floating.

Table 28-6. Pin Name Mapping

Signal Name in

Programming Mode Pin Name

I/O

Function

0: Device is busy programming, 1: Device is

ready for new command.

RDY/BSY

PD1

O

OE

WR

PD2

PD3

PD4

PD5

PD6

PD7

PE6

I

Output Enable (Active low).

Write Pulse (Active low).

Byte Select 1.

I

BS1

I

XA0

I

XTAL Action Bit 0

XA1

I

XTAL Action Bit 1

PAGEL

BS2

Program Memory and EEPROM data Page Load.

Byte Select 2.

I

DATA

PB7-0

I/O

Bi-directional Data bus (Output when OE is low).

Table 28-7. BS2 and BS1 Encoding

Flash /

Flash Data

EEPROM

Address

Loading /

Reading

Fuse

Programming

Reading Fuse

and Lock Bits

BS2

0

BS1

0

Low Byte

High Byte

Low Byte

High Byte

Low Byte

High Byte

Fuse Low Byte

Lock bits

0

1

Extended High

Byte

Extended Fuse

Byte

1

0

1

Reserved

Extended Byte

Reserved

Fuse High Byte

350

7766E–AVR–04/10

ATmega16U4/ATmega32U4

,

Table 28-8. Pin Values Used to Enter Programming Mode

PAGEL

XA1

Symbol

Value

Prog_enable[3]

Prog_enable[2]

Prog_enable[1]

Prog_enable[0]

0

XA0

BS1

Table 28-9. XA1 and XA0 Enoding

XA1

XA0

Action when XTAL1 is Pulsed

Load Flash or EEPROM Address (High or low address byte

determined by BS2 and BS1).

0

1

0

1

Load Data (High or Low data byte for Flash determined by BS1).

Load Command

No Action, Idle

Table 28-10. Command Byte Bit Encoding

Command Byte

1000 0000

0100 0000

0010 0000

0001 0000

0001 0001

0000 1000

0000 0100

0000 0010

0000 0011

Command Executed

Chip Erase

Write Fuse bits

Write Lock bits

Write Flash

Write EEPROM

Read Signature Bytes and Calibration byte

Read Fuse and Lock bits

Read Flash

Read EEPROM

Table 28-11. No. of Words in a Page and No. of Pages in the Flash

No. of

Flash Size

Page Size

PCWORD

Pages

PCPAGE

PCMSB

16K words (32K bytes)

128 words

PC[6:0]

128

PC[13:7]

13

351

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 28-12. No. of Words in a Page and No. of Pages in the EEPROM

No. of

Pages

EEPROM Size

Page Size

PCWORD

PCPAGE

EEAMSB

1K bytes

8 bytes

EEA[2:0]

128

EEA[9:3]

9

28.6 Parallel Programming

28.6.1

Enter Programming Mode

The following algorithm puts the device in parallel programming mode:

1. Apply 4.5 - 5.5V between V_CCand GND.

2. Set RESET to “0” and toggle XTAL1 at least six times.

3. Set the Prog_enable pins listed in Table 28-8 on page 351 to “0000” and wait at least

100 ns.

4. Apply 11.5 - 12.5V to RESET. Any activity on Prog_enable pins within 100 ns after

+12V has been applied to RESET, will cause the device to fail entering programming

mode.

5. Wait at least 50 µs before sending a new command.

28.6.2

Considerations for Efficient Programming

The loaded command and address are retained in the device during programming. For efficient

programming, the following should be considered.

• The command needs only be loaded once when writing or reading multiple memory

locations.

• Skip writing the data value 0xFF, that is the contents of the entire EEPROM (unless the

EESAVE Fuse is programmed) and Flash after a Chip Erase.

• Address high byte needs only be loaded before programming or reading a new 256 word

window in Flash or 256 byte EEPROM. This consideration also applies to Signature bytes

reading.

28.6.3

Chip Erase

The Chip Erase will erase the Flash and EEPROM⁽¹⁾memories plus Lock bits. The Lock bits are

not reset until the program memory has been completely erased. The Fuse bits are not

changed. A Chip Erase must be performed before the Flash and/or EEPROM are

reprogrammed.

Note:

1. The EEPRPOM memory is preserved during Chip Erase if the EESAVE Fuse is programmed.

Load Command “Chip Erase”

1. Set XA1, XA0 to “10”. This enables command loading.

2. Set BS1 to “0”.

3. Set DATA to “1000 0000”. This is the command for Chip Erase.

4. Give XTAL1 a positive pulse. This loads the command.

5. Give WR a negative pulse. This starts the Chip Erase. RDY/BSY goes low.

6. Wait until RDY/BSY goes high before loading a new command.

352

7766E–AVR–04/10

ATmega16U4/ATmega32U4

28.6.4

Programming the Flash

The Flash is organized in pages, see Table 28-11 on page 351. When programming the Flash,

the program data is latched into a page buffer. This allows one page of program data to be pro-

grammed simultaneously. The following procedure describes how to program the entire Flash

memory:

A. Load Command “Write Flash”

1. Set XA1, XA0 to “10”. This enables command loading.

2. Set BS1 to “0”.

3. Set DATA to “0001 0000”. This is the command for Write Flash.

4. Give XTAL1 a positive pulse. This loads the command.

B. Load Address Low byte (Address bits 7..0)

1. Set XA1, XA0 to “00”. This enables address loading.

2. Set BS2, BS1 to “00”. This selects the address low byte.

3. Set DATA = Address low byte (0x00 - 0xFF).

4. Give XTAL1 a positive pulse. This loads the address low byte.

C. Load Data Low Byte

1. Set XA1, XA0 to “01”. This enables data loading.

2. Set DATA = Data low byte (0x00 - 0xFF).

3. Give XTAL1 a positive pulse. This loads the data byte.

D. Load Data High Byte

1. Set BS1 to “1”. This selects high data byte.

2. Set XA1, XA0 to “01”. This enables data loading.

3. Set DATA = Data high byte (0x00 - 0xFF).

4. Give XTAL1 a positive pulse. This loads the data byte.

E. Latch Data

1. Set BS1 to “1”. This selects high data byte.

2. Give PAGEL a positive pulse. This latches the data bytes. (See Figure 28-3 for signal

waveforms)

F. Repeat B through E until the entire buffer is filled or until all data within the page is loaded.

While the lower bits in the address are mapped to words within the page, the higher bits address

the pages within the FLASH. This is illustrated in Figure 28-2 on page 354. Note that if less than

eight bits are required to address words in the page (pagesize < 256), the most significant bit(s)

in the address low byte are used to address the page when performing a Page Write.

G. Load Address High byte (Address bits15..8)

1. Set XA1, XA0 to “00”. This enables address loading.

2. Set BS2, BS1 to “01”. This selects the address high byte.

3. Set DATA = Address high byte (0x00 - 0xFF).

4. Give XTAL1 a positive pulse. This loads the address high byte.

H. Load Address Extended High byte (Address bits 23..16)

1. Set XA1, XA0 to “00”. This enables address loading.

2. Set BS2, BS1 to “10”. This selects the address extended high byte.

353

7766E–AVR–04/10

ATmega16U4/ATmega32U4

3. Set DATA = Address extended high byte (0x00 - 0xFF).

4. Give XTAL1 a positive pulse. This loads the address high byte.

I. Program Page

1. Set BS2, BS1 to “00”

2. Give WR a negative pulse. This starts programming of the entire page of data.

RDY/BSY goes low.

3. Wait until RDY/BSY goes high (See Figure 28-3 for signal waveforms).

J. Repeat B through I until the entire Flash is programmed or until all data has been

programmed.

K. End Page Programming

1. 1. Set XA1, XA0 to “10”. This enables command loading.

2. Set DATA to “0000 0000”. This is the command for No Operation.

3. Give XTAL1 a positive pulse. This loads the command, and the internal write signals

are reset.

Figure 28-2. Addressing the Flash Which is Organized in Pages⁽¹⁾

PCMSB

PAGEMSB

PROGRAM

COUNTER

PCPAGE

PCWORD

PAGE ADDRESS

WITHIN THE FLASH

WORD ADDRESS

WITHIN A PAGE

PROGRAM MEMORY

PAGE

INSTRUCTION WORD

PCWORD[PAGEMSB:0]:

00

01

02

PAGEEND

Note:

1. PCPAGE and PCWORD are listed in Table 28-11 on page 351.

354

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 28-3. Programming the Flash Waveforms⁽¹⁾

F

A

B

C

D

E

B

C

D

E

G

I

H

0x10

ADDR. LOW

DATA LOW

DATA HIGH

ADDR. LOW DATA LOW

DATA HIGH

ADDR. HIGH

ADDR. EXT.H

XX

DATA

XA1

XA0

BS1

BS2

XTAL1

WR

RDY/BSY

RESET +12V

OE

PAGEL

Note:

1. “XX” is don’t care. The letters refer to the programming description above.

28.6.5

Programming the EEPROM

The EEPROM is organized in pages, see Table 28-12 on page 352. When programming the

EEPROM, the program data is latched into a page buffer. This allows one page of data to be

programmed simultaneously. The programming algorithm for the EEPROM data memory is as

follows (refer to “Programming the Flash” on page 353 for details on Command, Address and

Data loading):

1. A: Load Command “0001 0001”.

2. G: Load Address High Byte (0x00 - 0xFF).

3. B: Load Address Low Byte (0x00 - 0xFF).

4. C: Load Data (0x00 - 0xFF).

5. E: Latch data (give PAGEL a positive pulse).

K: Repeat 3 through 5 until the entire buffer is filled.

L: Program EEPROM page

1. Set BS2, BS1 to “00”.

2. Give WR a negative pulse. This starts programming of the EEPROM page. RDY/BSY

goes low.

3. Wait until to RDY/BSY goes high before programming the next page (See Figure 28-4

for signal waveforms).

355

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 28-4. Programming the EEPROM Waveforms

K

A

G

B

C

E

B

C

E

L

0x11

ADDR. HIGH

ADDR. LOW

DATA

ADDR. LOW

DATA

XX

DATA

XA1

XA0

BS1

XTAL1

WR

RDY/BSY

RESET +12V

OE

PAGEL

BS2

28.6.6

Reading the Flash

The algorithm for reading the Flash memory is as follows (refer to “Programming the Flash” on

page 353 for details on Command and Address loading):

1. A: Load Command “0000 0010”.

2. H: Load Address Extended Byte (0x00- 0xFF).

3. G: Load Address High Byte (0x00 - 0xFF).

4. B: Load Address Low Byte (0x00 - 0xFF).

5. Set OE to “0”, and BS1 to “0”. The Flash word low byte can now be read at DATA.

6. Set BS to “1”. The Flash word high byte can now be read at DATA.

7. Set OE to “1”.

28.6.7

Reading the EEPROM

The algorithm for reading the EEPROM memory is as follows (refer to “Programming the Flash”

on page 353 for details on Command and Address loading):

1. A: Load Command “0000 0011”.

2. G: Load Address High Byte (0x00 - 0xFF).

3. B: Load Address Low Byte (0x00 - 0xFF).

4. Set OE to “0”, and BS1 to “0”. The EEPROM Data byte can now be read at DATA.

5. Set OE to “1”.

28.6.8

Programming the Fuse Low Bits

The algorithm for programming the Fuse Low bits is as follows (refer to “Programming the Flash”

on page 353 for details on Command and Data loading):

1. A: Load Command “0100 0000”.

2. C: Load Data Low Byte. Bit n = “0” programs and bit n = “1” erases the Fuse bit.

3. Give WR a negative pulse and wait for RDY/BSY to go high.

356

7766E–AVR–04/10

ATmega16U4/ATmega32U4

28.6.9

Programming the Fuse High Bits

The algorithm for programming the Fuse High bits is as follows (refer to “Programming the

Flash” on page 353 for details on Command and Data loading):

1. A: Load Command “0100 0000”.

2. C: Load Data Low Byte. Bit n = “0” programs and bit n = “1” erases the Fuse bit.

3. Set BS2, BS1 to “01”. This selects high data byte.

4. Give WR a negative pulse and wait for RDY/BSY to go high.

5. Set BS2, BS1 to “00”. This selects low data byte.

28.6.10 Programming the Extended Fuse Bits

The algorithm for programming the Extended Fuse bits is as follows (refer to “Programming the

Flash” on page 353 for details on Command and Data loading):

1. 1. A: Load Command “0100 0000”.

2. 2. C: Load Data Low Byte. Bit n = “0” programs and bit n = “1” erases the Fuse bit.

3. 3. Set BS2, BS1 to “10”. This selects extended data byte.

4. 4. Give WR a negative pulse and wait for RDY/BSY to go high.

5. 5. Set BS2, BS1 to “00”. This selects low data byte.

Figure 28-5. Programming the FUSES Waveforms

Write Fuse Low byte

Write Fuse high byte

Write Extended Fuse byte

A

C

A

C

A

C

0x40

DATA

XX

0x40

DATA

XX

0x40

DATA

XX

DATA

XA1

XA0

BS1

BS2

XTAL1

WR

RDY/BSY

RESET +12V

OE

PAGEL

28.6.11 Programming the Lock Bits

The algorithm for programming the Lock bits is as follows (refer to “Programming the Flash” on

page 353 for details on Command and Data loading):

1. A: Load Command “0010 0000”.

2. C: Load Data Low Byte. Bit n = “0” programs the Lock bit. If LB mode 3 is programmed

(LB1 and LB2 is programmed), it is not possible to program the Boot Lock bits by any

External Programming mode.

3. Give WR a negative pulse and wait for RDY/BSY to go high.

The Lock bits can only be cleared by executing Chip Erase.

357

7766E–AVR–04/10

ATmega16U4/ATmega32U4

28.6.12 Reading the Fuse and Lock Bits

The algorithm for reading the Fuse and Lock bits is as follows (refer to “Programming the Flash”

on page 353 for details on Command loading):

1. A: Load Command “0000 0100”.

2. Set OE to “0”, and BS2, BS1 to “00”. The status of the Fuse Low bits can now be read

at DATA (“0” means programmed).

3. Set OE to “0”, and BS2, BS1 to “11”. The status of the Fuse High bits can now be read

at DATA (“0” means programmed).

4. Set OE to “0”, and BS2, BS1 to “10”. The status of the Extended Fuse bits can now be

read at DATA (“0” means programmed).

5. Set OE to “0”, and BS2, BS1 to “01”. The status of the Lock bits can now be read at

DATA (“0” means programmed).

6. Set OE to “1”.

Figure 28-6. Mapping Between BS1, BS2 and the Fuse and Lock Bits During Read

0

Fuse Low Byte

Extended Fuse Byte

Lock Bits

0

1

0

DATA

BS2

BS1

Fuse High Byte

1

BS2

28.6.13 Reading the Signature Bytes

The algorithm for reading the Signature bytes is as follows (refer to “Programming the Flash” on

page 353 for details on Command and Address loading):

1. A: Load Command “0000 1000”.

2. B: Load Address Low Byte (0x00 - 0x02).

3. Set OE to “0”, and BS to “0”. The selected Signature byte can now be read at DATA.

4. Set OE to “1”.

28.6.14 Reading the Calibration Byte

The algorithm for reading the Calibration byte is as follows (refer to “Programming the Flash” on

page 353 for details on Command and Address loading):

1. A: Load Command “0000 1000”.

2. B: Load Address Low Byte, 0x00.

3. Set OE to “0”, and BS1 to “1”. The Calibration byte can now be read at DATA.

4. Set OE to “1”.

358

7766E–AVR–04/10

ATmega16U4/ATmega32U4

28.6.15 Parallel Programming Characteristics

Figure 28-7. Parallel Programming Timing, Including some General Timing Requirements

t_XLWL

t_XHXL

XTAL1

t_DVXH

t_XLDX

Data & Contol

(DATA, XA0/1, BS1, BS2)

t_BVPH

t_PLBXt_BVWL

t_WLBX

PAGEL

t_PHPL

t_WLWH

WR

t_PLWL

WLRL

RDY/BSY

t_WLRH

Figure 28-8. Parallel Programming Timing, Loading Sequence with Timing Requirements⁽¹⁾

LOAD DATA

LOAD ADDRESS

(LOW BYTE)

LOAD DATA

(LOW BYTE)

LOAD DATA

(HIGH BYTE)

LOAD ADDRESS

(LOW BYTE)

t_XLPH

t_XLXH

t_PLXH

XTAL1

BS1

PAGEL

DATA

ADDR0 (Low Byte)

DATA (Low Byte)

DATA (High Byte)

ADDR1 (Low Byte)

XA0

XA1

Note:

1. The timing requirements shown in Figure 28-7 (i.e., t_DVXH, t_XHXL, and t_XLDX) also apply to load-

ing operation.

Figure 28-9. Parallel Programming Timing, Reading Sequence (within the Same Page) with

Timing Requirements⁽¹⁾

LOAD ADDRESS

(LOW BYTE)

READ DATA

(LOW BYTE)

READ DATA

(HIGH BYTE)

LOAD ADDRESS

(LOW BYTE)

t_XLOL

XTAL1

BS1

t_BVDV

t_OLDV

OE

t_OHDZ

ADDR1 (Low Byte)

DATA (High Byte)

DATA

ADDR0 (Low Byte)

DATA (Low Byte)

XA0

XA1

359

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Note:

1. The timing requirements shown in Figure 28-7 (i.e., t_DVXH, t_XHXL, and t_XLDX) also apply to read-

ing operation.

Table 28-13. Parallel Programming Characteristics, V_CC= 5V 10%

Symbol

V_PP

Parameter

Min

Typ

Max

12.5

250

Units

V

Programming Enable Voltage

Programming Enable Current

Data and Control Valid before XTAL1 High

XTAL1 Low to XTAL1 High

XTAL1 Pulse Width High

Data and Control Hold after XTAL1 Low

XTAL1 Low to WR Low

11.5

I_PP

μA

ns

μs

ms

ns

t_DVXH

t_XLXH

t_XHXL

t_XLDX

t_XLWL

t_XLPH

t_PLXH

t_BVPH

t_PHPL

t_PLBX

t_WLBX

t_PLWL

t_BVWL

t_WLWH

t_WLRL

t_WLRH

t_{WLRH_CE}

t_XLOL

t_BVDV

t_OLDV

t_OHDZ

67

200

150

67

0

XTAL1 Low to PAGEL high

PAGEL low to XTAL1 high

BS1 Valid before PAGEL High

PAGEL Pulse Width High

BS1 Hold after PAGEL Low

BS2/1 Hold after WR Low

PAGEL Low to WR Low

0

150

67

150

67

150

0

BS2/1 Valid to WR Low

WR Pulse Width Low

WR Low to RDY/BSY Low

WR Low to RDY/BSY High⁽¹⁾

WR Low to RDY/BSY High for Chip Erase⁽²⁾

XTAL1 Low to OE Low

1

4.5

9

3.7

7.5

0

BS1 Valid to DATA valid

0

250

OE Low to DATA Valid

OE High to DATA Tri-stated

Notes: 1. t_WLRHis valid for the Write Flash, Write EEPROM, Write Fuse bits and Write Lock bits

commands.

2.

t_{WLRH_CE}is valid for the Chip Erase command.

28.7 Serial Downloading

Both the Flash and EEPROM memory arrays can be programmed using a serial programming

bus while RESET is pulled to GND. The serial programming interface consists of pins SCK, PDI

(input) and PDO (output). After RESET is set low, the Programming Enable instruction needs to

be executed first before program/erase operations can be executed. NOTE, in Table 28-14 on

page 361, the pin mapping for serial programming is listed. Not all packages use the SPI pins

dedicated for the internal Serial Peripheral Interface - SPI.

360

7766E–AVR–04/10

ATmega16U4/ATmega32U4

28.8 Serial Programming Pin Mapping

Table 28-14. Pin Mapping Serial Programming

Pins

Symbol

PDI

(TQFP-64)

I/O

Description

Serial Data in

Serial Data out

Serial Clock

PB2

PB3

PB1

I

O

I

PDO

SCK

Figure 28-10. Serial Programming and Verify⁽¹⁾

+1.8 - 5.5V

VCC

+1.8 - 5.5V⁽²⁾

PDI

AVCC

PDO

SCK

XTAL1

RESET

GND

Notes: 1. If the device is clocked by the internal Oscillator, it is no need to connect a clock source to the

XTAL1 pin.

2. V_CC- 0.3V < AVCC < V_CC+ 0.3V, however, AVCC should always be within 1.8 - 5.5V

When programming the EEPROM, an auto-erase cycle is built into the self-timed programming

operation (in the Serial mode ONLY) and there is no need to first execute the Chip Erase

instruction. The Chip Erase operation turns the content of every memory location in both the

Program and EEPROM arrays into 0xFF.

Depending on CKSEL Fuses, a valid clock must be present. The minimum low and high periods

for the serial clock (SCK) input are defined as follows:

Low: > 2 CPU clock cycles for f_ck< 12 MHz, 3 CPU clock cycles for f_ck>= 12 MHz

High: > 2 CPU clock cycles for f_ck< 12 MHz, 3 CPU clock cycles for f_ck>= 12 MHz

28.8.1

Serial Programming Algorithm

When writing serial data to the ATmega16U4/ATmega32U4, data is clocked on the rising edge

of SCK.

When reading data from the ATmega16U4/ATmega32U4, data is clocked on the falling edge of

SCK. See Figure 28-11 for timing details.

To program and verify the ATmega16U4/ATmega32U4 in the serial programming mode, the fol-

lowing sequence is recommended (See four byte instruction formats in Table 28-16):

361

7766E–AVR–04/10

ATmega16U4/ATmega32U4

1. Power-up sequence:

Apply power between V_CCand GND while RESET and SCK are set to “0”. In some sys-

tems, the programmer can not guarantee that SCK is held low during power-up. In this

case, RESET must be given a positive pulse of at least two CPU clock cycles duration

after SCK has been set to “0”.

2. Wait for at least 20 ms and enable serial programming by sending the Programming

Enable serial instruction to pin PDI.

3. The serial programming instructions will not work if the communication is out of syn-

chronization. When in sync. the second byte (0x53), will echo back when issuing the

third byte of the Programming Enable instruction. Whether the echo is correct or not, all

four bytes of the instruction must be transmitted. If the 0x53 did not echo back, give

RESET a positive pulse and issue a new Programming Enable command.

4. The Flash is programmed one page at a time. The memory page is loaded one byte at

a time by supplying the 7 LSB of the address and data together with the Load Program

Memory Page instruction. To ensure correct loading of the page, the data low byte must

be loaded before data high byte is applied for a given address. The Program Memory

Page is stored by loading the Write Program Memory Page instruction with the address

lines 15..8. Before issuing this command, make sure the instruction Load Extended

Address Byte has been used to define the MSB of the address. The extended address

byte is stored until the command is re-issued, i.e., the command needs only be issued

for the first page, and when crossing the 64KWord boundary. If polling (RDY/BSY) is not

used, the user must wait at least t_{WD_FLASH}before issuing the next page. (See Table 28-

15.) Accessing the serial programming interface before the Flash write operation com-

pletes can result in incorrect programming.

5. The EEPROM array is programmed one byte at a time by supplying the address and

data together with the appropriate Write instruction. An EEPROM memory location is

first automatically erased before new data is written. If polling is not used, the user must

wait at least t_{WD_EEPROM}before issuing the next byte. (See Table 28-15.) In a chip

erased device, no 0xFFs in the data file(s) need to be programmed.

6. Any memory location can be verified by using the Read instruction which returns the

content at the selected address at serial output PDO. When reading the Flash memory,

use the instruction Load Extended Address Byte to define the upper address byte,

which is not included in the Read Program Memory instruction. The extended address

byte is stored until the command is re-issued, i.e., the command needs only be issued

for the first page, and when crossing the 64KWord boundary.

7. At the end of the programming session, RESET can be set high to commence normal

operation.

8. Power-off sequence (if needed):

Set RESET to “1”.

Turn V_CCpower off.

Table 28-15. Minimum Wait Delay Before Writing the Next Flash or EEPROM Location

Symbol

Minimum Wait Delay

4.5 ms

t_{WD_FLASH}

t_{WD_EEPROM}

t_{WD_ERASE}

9.0 ms

362

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 28-11. Serial Programming Waveforms

SERIAL DATA INPUT

(MOSI)

MSB

LSB

SERIAL DATA OUTPUT

(MISO)

MSB

SERIAL CLOCK INPUT

(SCK)

SAMPLE

363

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 28-16. Serial Programming Instruction Set

Instruction Format

Byte 2 Byte 3

Instruction

Byte 1

Byte4

Operation

1010 1100 0101 0011 xxxx xxxx xxxx xxxx Enable Serial Programming after

Programming Enable

Chip Erase

RESET goes low.

1010 1100 100x xxxx xxxx xxxx xxxx xxxx Chip Erase EEPROM and Flash.

0100 1101 0000 0000 cccc cccc xxxx xxxx Defines Extended Address Byte for

Read Program Memory and Write

Load Extended Address Byte

Read Program Memory

Program Memory Page.

0010 H000 aaaa aaaa bbbb bbbb oooo oooo Read H (high or low) data o from

Program memory at word address

c:a:b.

0100 H000 xxxx xxxx xxbb bbbb iiii iiii Write H (high or low) data i to Program

Memory page at word address b. Data

low byte must be loaded before Data

high byte is applied within the same

address.

Load Program Memory Page

0100 1100 aaaa aaaa bbxx xxxx xxxx xxxx Write Program Memory Page at

address c:a:b.

Write Program Memory Page

Read EEPROM Memory

Write EEPROM Memory

1010 0000 0000 aaaa bbbb bbbb oooo oooo Read data o from EEPROM memory at

address a:b.

1100 0000 0000 aaaa bbbb bbbb iiii iiii Write data i to EEPROM memory at

address a:b.

1100 0001 0000 0000 0000 00bb iiii iiii Load data i to EEPROM memory page

Load EEPROM Memory

Page (page access)

buffer. After data is loaded, program

EEPROM page.

Write EEPROM Memory

Page (page access)

1100 0010 0000 aaaa bbbb bb00 xxxx xxxx

Write EEPROM page at address a:b.

0101 1000 0000 0000 xxxx xxxx xxoo oooo Read Lock bits. “0” = programmed, “1”

Read Lock bits

Write Lock bits

= unprogrammed. See Table 28-1 on

page 346 for details.

1010 1100 111x xxxx xxxx xxxx 11ii iiii Write Lock bits. Set bits = “0” to

program Lock bits. See Table 28-1 on

page 346 for details.

Read Signature Byte

Write Fuse bits

0011 0000 000x xxxx xxxx xxbb oooo oooo Read Signature Byte o at address b.

1010 1100 1010 0000 xxxx xxxx iiii iiii Set bits = “0” to program, “1” to

unprogram.

1010 1100 1010 1000 xxxx xxxx iiii iiii Set bits = “0” to program, “1” to

Write Fuse High bits

unprogram.

1010 1100 1010 0100 xxxx xxxx iiii iiii Set bits = “0” to program, “1” to

Write Extended Fuse Bits

unprogram. See Table 28-3 on page

347 for details.

0101 0000 0000 0000 xxxx xxxx oooo oooo Read Fuse bits. “0” = programmed, “1”

Read Fuse bits

= unprogrammed.

0101 1000 0000 1000 xxxx xxxx oooo oooo Read Fuse High bits. “0” = pro-

Read Fuse High bits

grammed, “1” = unprogrammed.

364

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 28-16. Serial Programming Instruction Set (Continued)

Instruction Format

Instruction

Byte 1

Byte 2

Byte 3

Byte4

Operation

0101 0000 0000 1000 xxxx xxxx oooo oooo Read Extended Fuse bits. “0” = pro-

grammed, “1” = unprogrammed. See

Read Extended Fuse Bits

Read Calibration Byte

Poll RDY/BSY

Table 28-3 on page 347 for details.

0011 1000 000x xxxx 0000 0000 oooo oooo Read Calibration Byte

1111 0000 0000 0000 xxxx xxxx xxxx xxxo If o = “1”, a programming operation is

still busy. Wait until this bit returns to

“0” before applying another command.

Note:

a = address high bits, b = address low bits, c = address extended bits, H = 0 - Low byte, 1 - High Byte, o = data out, i = data in,

x = don’t care

28.8.2

Serial Programming Characteristics

For characteristics of the Serial Programming module see “SPI Timing Characteristics” on page

383.

28.9 Programming via the JTAG Interface

Programming through the JTAG interface requires control of the four JTAG specific pins: TCK,

TMS, TDI, and TDO. Control of the reset and clock pins is not required.

To be able to use the JTAG interface, the JTAGEN Fuse must be programmed. The device is

default shipped with the fuse programmed. In addition, the JTD bit in MCUCSR must be cleared.

Alternatively, if the JTD bit is set, the external reset can be forced low. Then, the JTD bit will be

cleared after two chip clocks, and the JTAG pins are available for programming. This provides a

means of using the JTAG pins as normal port pins in Running mode while still allowing In-Sys-

tem Programming via the JTAG interface. Note that this technique can not be used when using

the JTAG pins for Boundary-scan or On-chip Debug. In these cases the JTAG pins must be ded-

icated for this purpose.

During programming the clock frequency of the TCK Input must be less than the maximum fre-

quency of the chip. The System Clock Prescaler can not be used to divide the TCK Clock Input

into a sufficiently low frequency.

As a definition in this datasheet, the LSB is shifted in and out first of all Shift Registers.

28.9.1

Programming Specific JTAG Instructions

The Instruction Register is 4-bit wide, supporting up to 16 instructions. The JTAG instructions

useful for programming are listed below.

The OPCODE for each instruction is shown behind the instruction name in hex format. The text

describes which Data Register is selected as path between TDI and TDO for each instruction.

The Run-Test/Idle state of the TAP controller is used to generate internal clocks. It can also be

used as an idle state between JTAG sequences. The state machine sequence for changing the

instruction word is shown in Figure 28-12.

365

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 28-12. State Machine Sequence for Changing the Instruction Word

1

Test-Logic-Reset

0

1

0

Run-Test/Idle

Select-DR Scan

Select-IR Scan

0

1

Capture-DR

Capture-IR

0

Shift-DR

0

Shift-IR

0

1

Exit1-DR

0

1

Exit1-IR

0

1

Pause-DR

1

0

Pause-IR

1

0

Exit2-DR

1

Exit2-IR

1

Update-DR

Update-IR

1

0

28.9.2

AVR_RESET (0xC)

The AVR specific public JTAG instruction for setting the AVR device in the Reset mode or taking

the device out from the Reset mode. The TAP controller is not reset by this instruction. The one

bit Reset Register is selected as Data Register. Note that the reset will be active as long as there

is a logic “one” in the Reset Chain. The output from this chain is not latched.

The active states are:

• Shift-DR: The Reset Register is shifted by the TCK input.

28.9.3

PROG_ENABLE (0x4)

The AVR specific public JTAG instruction for enabling programming via the JTAG port. The 16-

bit Programming Enable Register is selected as Data Register. The active states are the

following:

• Shift-DR: The programming enable signature is shifted into the Data Register.

• Update-DR: The programming enable signature is compared to the correct value, and

Programming mode is entered if the signature is valid.

366

7766E–AVR–04/10

ATmega16U4/ATmega32U4

28.9.4

PROG_COMMANDS (0x5)

The AVR specific public JTAG instruction for entering programming commands via the JTAG

port. The 15-bit Programming Command Register is selected as Data Register. The active

states are the following:

• Capture-DR: The result of the previous command is loaded into the Data Register.

• Shift-DR: The Data Register is shifted by the TCK input, shifting out the result of the previous

command and shifting in the new command.

• Update-DR: The programming command is applied to the Flash inputs

• Run-Test/Idle: One clock cycle is generated, executing the applied command

28.9.5

PROG_PAGELOAD (0x6)

The AVR specific public JTAG instruction to directly load the Flash data page via the JTAG port.

An 8-bit Flash Data Byte Register is selected as the Data Register. This is physically the 8 LSBs

of the Programming Command Register. The active states are the following:

• Shift-DR: The Flash Data Byte Register is shifted by the TCK input.

• Update-DR: The content of the Flash Data Byte Register is copied into a temporary register.

A write sequence is initiated that within 11 TCK cycles loads the content of the temporary

register into the Flash page buffer. The AVR automatically alternates between writing the low

and the high byte for each new Update-DR state, starting with the low byte for the first

Update-DR encountered after entering the PROG_PAGELOAD command. The Program

Counter is pre-incremented before writing the low byte, except for the first written byte. This

ensures that the first data is written to the address set up by PROG_COMMANDS, and

loading the last location in the page buffer does not make the program counter increment into

the next page.

28.9.6

PROG_PAGEREAD (0x7)

The AVR specific public JTAG instruction to directly capture the Flash content via the JTAG port.

An 8-bit Flash Data Byte Register is selected as the Data Register. This is physically the 8 LSBs

of the Programming Command Register. The active states are the following:

• Capture-DR: The content of the selected Flash byte is captured into the Flash Data Byte

Register. The AVR automatically alternates between reading the low and the high byte for

each new Capture-DR state, starting with the low byte for the first Capture-DR encountered

after entering the PROG_PAGEREAD command. The Program Counter is post-incremented

after reading each high byte, including the first read byte. This ensures that the first data is

captured from the first address set up by PROG_COMMANDS, and reading the last location

in the page makes the program counter increment into the next page.

• Shift-DR: The Flash Data Byte Register is shifted by the TCK input.

28.9.7

Data Registers

The Data Registers are selected by the JTAG instruction registers described in section “Pro-

gramming Specific JTAG Instructions” on page 365. The Data Registers relevant for

programming operations are:

• Reset Register

• Programming Enable Register

• Programming Command Register

• Flash Data Byte Register

367

7766E–AVR–04/10

ATmega16U4/ATmega32U4

28.9.8

Reset Register

The Reset Register is a Test Data Register used to reset the part during programming. It is

required to reset the part before entering Programming mode.

A high value in the Reset Register corresponds to pulling the external reset low. The part is reset

as long as there is a high value present in the Reset Register. Depending on the Fuse settings

for the clock options, the part will remain reset for a Reset Time-out period (refer to “Clock

Sources” on page 28) after releasing the Reset Register. The output from this Data Register is

not latched, so the reset will take place immediately, as shown in Figure 8-1 on page 50.

28.9.9

Programming Enable Register

The Programming Enable Register is a 16-bit register. The contents of this register is compared

to the programming enable signature, binary code 0b1010_0011_0111_0000. When the con-

tents of the register is equal to the programming enable signature, programming via the JTAG

port is enabled. The register is reset to 0 on Power-on Reset, and should always be reset when

leaving Programming mode.

Figure 28-13. Programming Enable Register

TDI

0xA370

D

Q

A

T

A

Programming Enable

=

ClockDR & PROG_ENABLE

TDO

28.9.10 Programming Command Register

The Programming Command Register is a 15-bit register. This register is used to serially shift in

programming commands, and to serially shift out the result of the previous command, if any. The

JTAG Programming Instruction Set is shown in Table 28-17. The state sequence when shifting

in the programming commands is illustrated in Figure 28-15.

368

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 28-14. Programming Command Register

TDI

S

T

R

O

B

E

S

Flash

EEPROM

Fuses

A

D

R

E

S

/

Lock Bits

D

A

T

A

TDO

369

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 28-17. JTAG Programming Instruction

Set a = address high bits, b = address low bits, c = address extended bits, H = 0 - Low byte, 1 - High Byte, o = data out,

i = data in, x = don’t care

Instruction

TDI Sequence

TDO Sequence

Notes

0100011_10000000

0110001_10000000

0110011_10000000

xxxxxxx_xxxxxxxx

1a. Chip Erase

1b. Poll for Chip Erase Complete

2a. Enter Flash Write

0110011_10000000

0100011_00010000

0001011_cccccccc

0000111_aaaaaaaa

0000011_bbbbbbbb

0010011_iiiiiiii

xxxxxox_xxxxxxxx

xxxxxxx_xxxxxxxx

(2)

2b. Load Address Extended High Byte

2c. Load Address High Byte

2d. Load Address Low Byte

2e. Load Data Low Byte

(10)

2f. Load Data High Byte

0010111_iiiiiiii

0110111_00000000

1110111_00000000

0110111_00000000

xxxxxxx_xxxxxxxx

2g. Latch Data

(1)

0110111_00000000

0110101_00000000

0110111_00000000

xxxxxxx_xxxxxxxx

2h. Write Flash Page

2i. Poll for Page Write Complete

3a. Enter Flash Read

0110111_00000000

0100011_00000010

0001011_cccccccc

0000111_aaaaaaaa

0000011_bbbbbbbb

xxxxxox_xxxxxxxx

xxxxxxx_xxxxxxxx

(2)

3b. Load Address Extended High Byte

3c. Load Address High Byte

3d. Load Address Low Byte

(10)

0110010_00000000

0110110_00000000

0110111_00000000

xxxxxxx_xxxxxxxx

xxxxxxx_oooooooo

3e. Read Data Low and High Byte

Low byte

High byte

4a. Enter EEPROM Write

4b. Load Address High Byte

4c. Load Address Low Byte

4d. Load Data Byte

0100011_00010001

0000111_aaaaaaaa

0000011_bbbbbbbb

0010011_iiiiiiii

xxxxxxx_xxxxxxxx

(10)

0110111_00000000

1110111_00000000

0110111_00000000

xxxxxxx_xxxxxxxx

4e. Latch Data

(1)

0110011_00000000

0110001_00000000

0110011_00000000

xxxxxxx_xxxxxxxx

4f. Write EEPROM Page

370

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 28-17. JTAG Programming Instruction (Continued)

Set (Continued) a = address high bits, b = address low bits, c = address extended bits, H = 0 - Low byte, 1 - High Byte,

o = data out, i = data in, x = don’t care

Instruction

TDI Sequence

TDO Sequence

Notes

4g. Poll for Page Write Complete

5a. Enter EEPROM Read

5b. Load Address High Byte

5c. Load Address Low Byte

0110011_00000000

0100011_00000011

0000111_aaaaaaaa

0000011_bbbbbbbb

xxxxxox_xxxxxxxx

xxxxxxx_xxxxxxxx

(2)

(10)

0110011_bbbbbbbb

0110010_00000000

0110011_00000000

xxxxxxx_xxxxxxxx

xxxxxxx_oooooooo

5d. Read Data Byte

6a. Enter Fuse Write

0100011_01000000

xxxxxxx_xxxxxxxx

6b. Load Data Low Byte⁽⁶⁾

0010011_iiiiiiii

(3)

(1)

0111011_00000000

0111001_00000000

0111011_00000000

xxxxxxx_xxxxxxxx

6c. Write Fuse Extended Byte

6d. Poll for Fuse Write Complete

6e. Load Data Low Byte⁽⁷⁾

0110111_00000000

xxxxxox_xxxxxxxx

(2)

(3)

0010011_iiiiiiii

xxxxxxx_xxxxxxxx

0110111_00000000

0110101_00000000

0110111_00000000

xxxxxxx_xxxxxxxx

6f. Write Fuse High Byte

(1)

6g. Poll for Fuse Write Complete

6h. Load Data Low Byte⁽⁷⁾

0110111_00000000

xxxxxox_xxxxxxxx

(2)

(3)

0010011_iiiiiiii

xxxxxxx_xxxxxxxx

0110011_00000000

0110001_00000000

0110011_00000000

xxxxxxx_xxxxxxxx

6i. Write Fuse Low Byte

(1)

6j. Poll for Fuse Write Complete

7a. Enter Lock Bit Write

7b. Load Data Byte⁽⁹⁾

0110011_00000000

0100011_00100000

0010011_11iiiiii

xxxxxox_xxxxxxxx

xxxxxxx_xxxxxxxx

(2)

(4)

0110011_00000000

0110001_00000000

0110011_00000000

xxxxxxx_xxxxxxxx

7c. Write Lock Bits

(1)

(2)

7d. Poll for Lock Bit Write complete

8a. Enter Fuse/Lock Bit Read

0110011_00000000

0100011_00000100

xxxxxox_xxxxxxxx

xxxxxxx_xxxxxxxx

0111010_00000000

0111011_00000000

xxxxxxx_xxxxxxxx

8b. Read Extended Fuse Byte⁽⁶⁾

8c. Read Fuse High Byte⁽⁷⁾

xxxxxxx_oooooooo

0111110_00000000

0111111_00000000

xxxxxxx_xxxxxxxx

xxxxxxx_oooooooo

371

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 28-17. JTAG Programming Instruction (Continued)

Set (Continued) a = address high bits, b = address low bits, c = address extended bits, H = 0 - Low byte, 1 - High Byte,

o = data out, i = data in, x = don’t care

Instruction

TDI Sequence

TDO Sequence

Notes

0110010_00000000

0110011_00000000

xxxxxxx_xxxxxxxx

8d. Read Fuse Low Byte⁽⁸⁾

xxxxxxx_oooooooo

0110110_00000000

0110111_00000000

xxxxxxx_xxxxxxxx

8e. Read Lock Bits⁽⁹⁾

(5)

xxxxxxx_xxoooooo

0111010_00000000

0111110_00000000

0110010_00000000

0110110_00000000

0110111_00000000

xxxxxxx_xxxxxxxx

xxxxxxx_oooooooo

(5)

Fuse Ext. byte

Fuse High byte

Fuse Low byte

Lock bits

8f. Read Fuses and Lock Bits

9a. Enter Signature Byte Read

9b. Load Address Byte

0100011_00001000

xxxxxxx_xxxxxxxx

0000011_bbbbbbbb

0110010_00000000

0110011_00000000

xxxxxxx_xxxxxxxx

9c. Read Signature Byte

xxxxxxx_oooooooo

10a. Enter Calibration Byte Read

10b. Load Address Byte

0100011_00001000

xxxxxxx_xxxxxxxx

0000011_bbbbbbbb

0110110_00000000

0110111_00000000

xxxxxxx_xxxxxxxx

10c. Read Calibration Byte

xxxxxxx_oooooooo

0100011_00000000

0110011_00000000

xxxxxxx_xxxxxxxx

11a. Load No Operation Command

Notes: 1. This command sequence is not required if the seven MSB are correctly set by the previous command sequence (which is

normally the case).

2. Repeat until o = “1”.

3. Set bits to “0” to program the corresponding Fuse, “1” to unprogram the Fuse.

4. Set bits to “0” to program the corresponding Lock bit, “1” to leave the Lock bit unchanged.

5. “0” = programmed, “1” = unprogrammed.

6. The bit mapping for Fuses Extended byte is listed in Table 28-3 on page 347

7. The bit mapping for Fuses High byte is listed in Table 28-4 on page 348

8. The bit mapping for Fuses Low byte is listed in Table 28-5 on page 348

9. The bit mapping for Lock bits byte is listed in Table 28-1 on page 346

10. Address bits exceeding PCMSB and EEAMSB (Table 28-11 and Table 28-12) are don’t care

11. All TDI and TDO sequences are represented by binary digits (0b...).

372

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 28-15. State Machine Sequence for Changing/Reading the Data Word

1

Test-Logic-Reset

0

1

0

Run-Test/Idle

Select-DR Scan

Select-IR Scan

0

1

Capture-DR

Capture-IR

0

Shift-IR

1

Shift-DR

0

1

Exit1-DR

0

1

Exit1-IR

0

Pause-DR

1

0

Pause-IR

1

0

Exit2-DR

1

Exit2-IR

1

Update-DR

Update-IR

1

0

28.9.11 Flash Data Byte Register

The Flash Data Byte Register provides an efficient way to load the entire Flash page buffer

before executing Page Write, or to read out/verify the content of the Flash. A state machine sets

up the control signals to the Flash and senses the strobe signals from the Flash, thus only the

data words need to be shifted in/out.

The Flash Data Byte Register actually consists of the 8-bit scan chain and a 8-bit temporary reg-

ister. During page load, the Update-DR state copies the content of the scan chain over to the

temporary register and initiates a write sequence that within 11 TCK cycles loads the content of

the temporary register into the Flash page buffer. The AVR automatically alternates between

writing the low and the high byte for each new Update-DR state, starting with the low byte for the

first Update-DR encountered after entering the PROG_PAGELOAD command. The Program

Counter is pre-incremented before writing the low byte, except for the first written byte. This

ensures that the first data is written to the address set up by PROG_COMMANDS, and loading

the last location in the page buffer does not make the Program Counter increment into the next

page.

During Page Read, the content of the selected Flash byte is captured into the Flash Data Byte

Register during the Capture-DR state. The AVR automatically alternates between reading the

low and the high byte for each new Capture-DR state, starting with the low byte for the first Cap-

373

7766E–AVR–04/10

ATmega16U4/ATmega32U4

ture-DR encountered after entering the PROG_PAGEREAD command. The Program Counter is

post-incremented after reading each high byte, including the first read byte. This ensures that

the first data is captured from the first address set up by PROG_COMMANDS, and reading the

last location in the page makes the program counter increment into the next page.

Figure 28-16. Flash Data Byte Register

STROBES

State

Machine

TDI

ADDRESS

Flash

EEPROM

Fuses

Lock Bits

D

A

T

A

TDO

The state machine controlling the Flash Data Byte Register is clocked by TCK. During normal

operation in which eight bits are shifted for each Flash byte, the clock cycles needed to navigate

through the TAP controller automatically feeds the state machine for the Flash Data Byte Regis-

ter with sufficient number of clock pulses to complete its operation transparently for the user.

However, if too few bits are shifted between each Update-DR state during page load, the TAP

controller should stay in the Run-Test/Idle state for some TCK cycles to ensure that there are at

least 11 TCK cycles between each Update-DR state.

28.9.12 Programming Algorithm

All references below of type “1a”, “1b”, and so on, refer to Table 28-17.

28.9.13 Entering Programming Mode

1. Enter JTAG instruction AVR_RESET and shift 1 in the Reset Register.

2. Enter instruction PROG_ENABLE and shift 0b1010_0011_0111_0000 in the Program-

ming Enable Register.

28.9.14 Leaving Programming Mode

1. Enter JTAG instruction PROG_COMMANDS.

2. Disable all programming instructions by using no operation instruction 11a.

3. Enter instruction PROG_ENABLE and shift 0b0000_0000_0000_0000 in the program-

ming Enable Register.

4. Enter JTAG instruction AVR_RESET and shift 0 in the Reset Register.

374

7766E–AVR–04/10

ATmega16U4/ATmega32U4

28.9.15 Performing Chip Erase

1. Enter JTAG instruction PROG_COMMANDS.

2. Start Chip Erase using programming instruction 1a.

3. Poll for Chip Erase complete using programming instruction 1b, or wait for t_{WLRH_CE}

(refer to Table 28-13 on page 360).

28.9.16 Programming the Flash

Before programming the Flash a Chip Erase must be performed, see “Performing Chip Erase”

on page 375.

1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Flash write using programming instruction 2a.

3. Load address Extended High byte using programming instruction 2b.

4. Load address High byte using programming instruction 2c.

5. Load address Low byte using programming instruction 2d.

6. Load data using programming instructions 2e, 2f and 2g.

7. Repeat steps 5 and 6 for all instruction words in the page.

8. Write the page using programming instruction 2h.

9. Poll for Flash write complete using programming instruction 2i, or wait for t_WLRH(refer to

Table 28-13 on page 360).

10. Repeat steps 3 to 9 until all data have been programmed.

A more efficient data transfer can be achieved using the PROG_PAGELOAD instruction:

1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Flash write using programming instruction 2a.

3. Load the page address using programming instructions 2b, 2c and 2d. PCWORD (refer

to Table 28-11 on page 351) is used to address within one page and must be written as

0.

4. Enter JTAG instruction PROG_PAGELOAD.

5. Load the entire page by shifting in all instruction words in the page byte-by-byte, start-

ing with the LSB of the first instruction in the page and ending with the MSB of the last

instruction in the page. Use Update-DR to copy the contents of the Flash Data Byte

Register into the Flash page location and to auto-increment the Program Counter

before each new word.

6. Enter JTAG instruction PROG_COMMANDS.

7. Write the page using programming instruction 2h.

8. Poll for Flash write complete using programming instruction 2i, or wait for t_WLRH(refer to

Table 28-13 on page 360).

9. Repeat steps 3 to 8 until all data have been programmed.

28.9.17 Reading the Flash

1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Flash read using programming instruction 3a.

3. Load address using programming instructions 3b, 3c and 3d.

4. Read data using programming instruction 3e.

5. Repeat steps 3 and 4 until all data have been read.

A more efficient data transfer can be achieved using the PROG_PAGEREAD instruction:

375

7766E–AVR–04/10

ATmega16U4/ATmega32U4

1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Flash read using programming instruction 3a.

3. Load the page address using programming instructions 3b, 3c and 3d. PCWORD (refer

to Table 28-11 on page 351) is used to address within one page and must be written as

0.

4. Enter JTAG instruction PROG_PAGEREAD.

5. Read the entire page (or Flash) by shifting out all instruction words in the page (or

Flash), starting with the LSB of the first instruction in the page (Flash) and ending with

the MSB of the last instruction in the page (Flash). The Capture-DR state both captures

the data from the Flash, and also auto-increments the program counter after each word

is read. Note that Capture-DR comes before the shift-DR state. Hence, the first byte

which is shifted out contains valid data.

6. Enter JTAG instruction PROG_COMMANDS.

7. Repeat steps 3 to 6 until all data have been read.

28.9.18 Programming the EEPROM

Before programming the EEPROM a Chip Erase must be performed, see “Performing Chip

Erase” on page 375.

1. Enter JTAG instruction PROG_COMMANDS.

2. Enable EEPROM write using programming instruction 4a.

3. Load address High byte using programming instruction 4b.

4. Load address Low byte using programming instruction 4c.

5. Load data using programming instructions 4d and 4e.

6. Repeat steps 4 and 5 for all data bytes in the page.

7. Write the data using programming instruction 4f.

8. Poll for EEPROM write complete using programming instruction 4g, or wait for t_WLRH

(refer to Table 28-13 on page 360).

9. Repeat steps 3 to 8 until all data have been programmed.

Note that the PROG_PAGELOAD instruction can not be used when programming the EEPROM.

28.9.19 Reading the EEPROM

1. Enter JTAG instruction PROG_COMMANDS.

2. Enable EEPROM read using programming instruction 5a.

3. Load address using programming instructions 5b and 5c.

4. Read data using programming instruction 5d.

5. Repeat steps 3 and 4 until all data have been read.

Note that the PROG_PAGEREAD instruction can not be used when reading the EEPROM.

28.9.20 Programming the Fuses

1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Fuse write using programming instruction 6a.

3. Load data high byte using programming instructions 6b. A bit value of “0” will program

the corresponding fuse, a “1” will unprogram the fuse.

4. Write Fuse High byte using programming instruction 6c.

5. Poll for Fuse write complete using programming instruction 6d, or wait for t_WLRH(refer to

Table 28-13 on page 360).

376

7766E–AVR–04/10

ATmega16U4/ATmega32U4

6. Load data low byte using programming instructions 6e. A “0” will program the fuse, a “1”

will unprogram the fuse.

7. Write Fuse low byte using programming instruction 6f.

8. Poll for Fuse write complete using programming instruction 6g, or wait for t_WLRH(refer to

Table 28-13 on page 360).

28.9.21 Programming the Lock Bits

1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Lock bit write using programming instruction 7a.

3. Load data using programming instructions 7b. A bit value of “0” will program the corre-

sponding lock bit, a “1” will leave the lock bit unchanged.

4. Write Lock bits using programming instruction 7c.

5. Poll for Lock bit write complete using programming instruction 7d, or wait for t_WLRH(refer

to Table 28-13 on page 360).

28.9.22 Reading the Fuses and Lock Bits

1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Fuse/Lock bit read using programming instruction 8a.

3. To read all Fuses and Lock bits, use programming instruction 8e.

To only read Fuse High byte, use programming instruction 8b.

To only read Fuse Low byte, use programming instruction 8c.

To only read Lock bits, use programming instruction 8d.

28.9.23 Reading the Signature Bytes

1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Signature byte read using programming instruction 9a.

3. Load address 0x00 using programming instruction 9b.

4. Read first signature byte using programming instruction 9c.

5. Repeat steps 3 and 4 with address 0x01 and address 0x02 to read the second and third

signature bytes, respectively.

28.9.24 Reading the Calibration Byte

1. Enter JTAG instruction PROG_COMMANDS.

2. Enable Calibration byte read using programming instruction 10a.

3. Load address 0x00 using programming instruction 10b.

4. Read the calibration byte using programming instruction 10c.

377

7766E–AVR–04/10

ATmega16U4/ATmega32U4

29. Electrical Characteristics

29.1 Absolute Maximum Ratings*

*NOTICE:

Stresses beyond those listed under “Absolute

Maximum Ratings” may cause permanent dam-

age to the device. This is a stress rating only and

functional operation of the device at these or

other conditions beyond those indicated in the

operational sections of this specification is not

implied. Exposure to absolute maximum rating

conditions for extended periods may affect

device reliability.

Operating Temperature.................................... -40°C to +85°C

Storage Temperature..................................... -65°C to +150°C

Voltage on any Pin except RESET and VBUS

with respect to Ground⁽⁸⁾.............................-0.5V to V_CC+0.5V

Voltage on RESET with respect to Ground......-0.5V to +13.0V

Voltage on VBUS with respect to Ground..........-0.5V to +6.0V

Maximum Operating Voltage ............................................ 6.0V

DC Current per I/O Pin ............................................... 40.0 mA

DC Current V_CCand GND Pins................................ 200.0 mA

29.2 DC Characteristics

T_A= -40°C to 85°C, V_CC= 2.7V to 5.5V (unless otherwise noted)

Symbol

Parameter

Condition

Min.⁽⁵⁾

Typ.

Max.⁽⁵⁾

Units

0.2V_CC-0.1V⁽¹⁾

(LVTTL)

Input Low Voltage,Except

XTAL1 and Reset pin

V_IL

V_CC= 2.7V - 5.5V

-0.5

V

Input Low Voltage,

XTAL1 pin

(1)

V_IL1

V_IL2

V

_CC= 2.7V - 5.5V

-0.5

0.1V_CC

V

Input Low Voltage,

RESET pin

(1)

V_CC= 2.7V - 5.5V

0.1V_CC

Input High Voltage,

Except XTAL1 and

RESET pins

0.2V_CC+0.9V⁽²⁾

(LVTTL)

V_IH

V_CC+ 0.5

V

Input High Voltage,

XTAL1 pin

(2)

V_IH1

V_IH2

V_OL

V_OH

I_IL

0.7V_CC

V_CC+ 0.5

V

Input High Voltage,

RESET pin

(2)

V_CC= 2.7V - 5.5V

0.9V_CC

I

_OL= 10mA, V_CC= 5V

_OL= 5mA, V_CC= 3V

0.7

0.5

Output Low Voltage⁽³⁾

,

V

I_OH= -10mA, V_CC= 5V

I_OH= -5mA, V_CC= 3V

4.2

2.3

Output High Voltage⁽⁴⁾

,

V

Input Leakage

Current I/O Pin

V_CC= 5.5V, pin low

(absolute value)

1

µA

Input Leakage

Current I/O Pin

V_CC= 5.5V, pin high

(absolute value)

I_IH

R_RST

R_PU

Reset Pull-up Resistor

I/O Pin Pull-up Resistor

30

20

60

50

kΩ

378

7766E–AVR–04/10

ATmega16U4/ATmega32U4

T_A= -40°C to 85°C, V_CC= 2.7V to 5.5V (unless otherwise noted) (Continued)

Symbol

Parameter

Condition

Min.⁽⁵⁾

Typ.

Max.⁽⁵⁾

Units

Active 4MHz, V_CC= 3V

5

mA

(ATmega16U4/ATmega32

U4)

Active 8MHz, V_CC= 5V

10

15

2

mA

(ATmega16U4/ATmega32

U4)

Power Supply Current⁽⁶⁾

Idle 4MHz, V_CC= 3V

(ATmega16U4/ATmega32

U4)

I_CC

Idle 8MHz, V_CC= 5V

6

(ATmega16U4/ATmega32

U4)

WDT enabled, V_CC= 3V,

Regulator Disabled

<10

1

12

5

µA

mV

nA

ns

Power-down mode

WDT disabled, V_CC

=

3V,Regulator Disabled

V_CC= 5V

Analog Comparator

Input Offset Voltage

V_ACIO

I_ACLK

t_ACID

Rusb

<10

40

50

V_in= V_CC/2

Analog Comparator

Input Leakage Current

V_CC= 5V

V_in= V_CC/2

-50

Analog Comparator

Propagation Delay

V_CC= 2.7V

V_CC= 4.0V

750

500

USB Series resistor

(external)

22 5%

Ω

C_UCAP=1µF 20%,

UVcc 4.0V, I≤80mA⁽⁷⁾

or

UVcc 3.4V, I≤55mA⁽⁷⁾

,

Vreg

Regulator Output Voltage

3.0

3.3

3.6

V

Note:

1. "Max" means the highest value where the pin is guaranteed to be read as low

2. "Min" means the lowest value where the pin is guaranteed to be read as high

3. Although each I/O port can sink more than the test conditions (20mA at VCC = 5V, 10mA at VCC = 3V) under steady state

conditions (non-transient), the following must be observed:

ATmega16U4/ATmega32U4:

1.)The sum of all IOL, for ports A0-A7, G2, C4-C7 should not exceed 100 mA.

2.)The sum of all IOL, for ports C0-C3, G0-G1, D0-D7 should not exceed 100 mA.

3.)The sum of all IOL, for ports G3-G5, B0-B7, E0-E7 should not exceed 100 mA.

4.)The sum of all IOL, for ports F0-F7 should not exceed 100 mA.

If IOL exceeds the test condition, VOL may exceed the related specification. Pins are not guaranteed to sink current greater

than the listed test condition.

4. Although each I/O port can source more than the test conditions (20mA at VCC = 5V, 10mA at VCC = 3V) under steady

state conditions (non-transient), the following must be observed:

ATmega16U4/ATmega32U4:

1)The sum of all IOH, for ports A0-A7, G2, C4-C7 should not exceed 100 mA.

2)The sum of all IOH, for ports C0-C3, G0-G1, D0-D7 should not exceed 100 mA.

3)The sum of all IOH, for ports G3-G5, B0-B7, E0-E7 should not exceed 100 mA.

4)The sum of all IOH, for ports F0-F7 should not exceed 100 mA.

379

7766E–AVR–04/10

ATmega16U4/ATmega32U4

5. All DC Characteristics contained in this datasheet are based on simulation and characterization of other AVR microcon-

trollers manufactured in the same process technology. These values are preliminary values representing design targets, and

will be updated after characterization of actual silicon

6. Values with “Power Reduction Register 1 - PRR1” disabled (0x00).

7. Maximum regulator output current should be reduced by the USB buffer current required when USB is active (about 25mA).

The remaining regulator output current can be used for the external application.

8. As specified on the USB Electrical chapter, the D+/D- pads can withstand voltages down to -1V applied through a 39 Ohms

resistor

29.3 External Clock Drive Waveforms

Figure 29-1. External Clock Drive Waveforms

V_IH1

V_IL1

29.4 External Clock Drive

Table 29-1. External Clock Drive

V_CC=1.8-5.5V

V_CC=2.7-5.5V

V_CC=4.5-5.5V

Symbol Parameter

Min.

Max.

Min.

Max.

Min.

Max.

Units

Oscillator

1/t_CLCL

0

2

0

8

0

16

MHz

Frequency

t_CLCL

t_CHCX

t_CLCX

t_CLCH

t_CHCL

Clock Period

High Time

Low Time

Rise Time

Fall Time

500

200

125

50

62.5

25

ns

μs

50

25

2.0

1.6

0.5

Change in period

from one clock

cycle to the next

Δt_CLCL

2

%

Note:

All DC Characteristics contained in this datasheet are based on simulation and characterization of

other AVR microcontrollers manufactured in the same process technology. These values are pre-

liminary values representing design targets, and will be updated after characterization of actual

silicon.

29.5 Maximum speed vs. V_CC

Maximum frequency is depending on V_CC.As shown in Figure 29-2, the Maximum Frequency vs.

V

_CCcurve is linear between 2.7V < V_CC< 5.5V.

380

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 29-2. Maximum Frequency vs. V_CC, ATmega16U4/ATmega32U4

16 MHz

8 MHz

Safe Operating Area

2.7V

4.5V

5.5V

29.6 2-wire Serial Interface Characteristics

Table 29-2 describes the requirements for devices connected to the 2-wire Serial Bus. The ATmega16U4/ATmega32U4 2-

wire Serial Interface meets or exceeds these requirements under the noted conditions.

Timing symbols refer to Figure 29-3.

Table 29-2. 2-wire Serial Bus Requirements

Symbol Parameter

Condition

Min

-0.5

Max

0.3 V_CC

V_CC+ 0.5

–

Units

V

Input Low-voltage

VIL

Input High-voltage

0.7 V_CC

V

VIH

Vhys

(1)

(2)

Hysteresis of Schmitt Trigger Inputs

Output Low-voltage

0.05 V_CC

V

(1)

VOL

3 mA sink current

10 pF < C_b< 400 pF⁽³⁾

0.1V_CC< V_i< 0.9V_CC

0

0.4

V

(1)

tr

(3)(2)

Rise Time for both SDA and SCL

Output Fall Time from V_IHminto V_ILmax

Spikes Suppressed by Input Filter

Input Current each I/O Pin

Capacitance for each I/O Pin

SCL Clock Frequency

20 + 0.1C_b

300

ns

µA

pF

kHz

(1)

tof

20 + 0.1C_b

250

(1)

tSP

0

-10

–

50⁽²⁾

I_i

10

C_i⁽¹⁾

10

f_SCL

f_CK⁽⁴⁾> max(16f_SCL, 250kHz)⁽⁵⁾

0

400

V_CC– 0,4V

----------------------------

3mA

f_SCL≤ 100 kHz

1000ns

C_b

-------------------

Ω

Rp

Value of Pull-up resistor

V_CC– 0,4V

----------------------------

3mA

f_SCL> 100 kHz

300ns

---------------

C_b

f_SCL≤ 100 kHz

_SCL> 100 kHz

4.0

0.6

4.7

1.3

–

µs

t_HD;STA

Hold Time (repeated) START Condition

Low Period of the SCL Clock

f

–

f_SCL≤ 100 kHz⁽⁶⁾

–

t_LOW

f_SCL> 100 kHz⁽⁷⁾

–

381

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 29-2. 2-wire Serial Bus Requirements (Continued)

Symbol Parameter

Condition

Min

4.0

0.6

4.7

0.6

0

Max

–

Units

µs

ns

µs

f_SCL≤ 100 kHz

f_SCL> 100 kHz

f_SCL≤ 100 kHz

f_SCL> 100 kHz

f_SCL≤ 100 kHz

f_SCL> 100 kHz

f_SCL≤ 100 kHz

t_HIGH

High period of the SCL clock

Set-up time for a repeated START condition

Data hold time

–

t_SU;STA

t_HD;DAT

t_SU;DAT

t_SU;STO

t_BUF

–

3.45

0.9

–

0

250

100

4.0

0.6

4.7

1.3

Data setup time

f

_SCL> 100 kHz

–

f_SCL≤ 100 kHz

f_SCL> 100 kHz

f_SCL≤ 100 kHz

f_SCL> 100 kHz

–

Setup time for STOP condition

–

Bus free time between a STOP and START

condition

–

Notes: 1. In ATmega16U4/ATmega32U4, this parameter is characterized and not 100% tested.

2. Required only for f_SCL> 100 kHz.

3. C_b= capacitance of one bus line in pF.

4. f_CK= CPU clock frequency

5. This requirement applies to all ATmega16U4/ATmega32U4 2-wire Serial Interface operation. Other devices connected to the

2-wire Serial Bus need only obey the general f_SCLrequirement.

6. The actual low period generated by the ATmega16U4/ATmega32U4 2-wire Serial Interface is (1/f_SCL- 2/f_CK), thus f_CKmust

be greater than 6 MHz for the low time requirement to be strictly met at f_SCL= 100 kHz.

7. The actual low period generated by the ATmega16U4/ATmega32U4 2-wire Serial Interface is (1/f_SCL- 2/f_CK), thus the low

time requirement will not be strictly met for f_SCL> 308 kHz when f_CK= 8 MHz. Still, ATmega16U4/ATmega32U4 devices con-

nected to the bus may communicate at full speed (400 kHz) with other ATmega16U4/ATmega32U4 devices, as well as any

other device with a proper t_LOWacceptance margin.

Figure 29-3. 2-wire Serial Bus Timing

t

HIGH

t

r

of

t

LOW

SCL

SDA

t

HD;DAT

SU;STA

HD;STA

t

SU;DAT

t

SU;STO

t

BUF

382

7766E–AVR–04/10

ATmega16U4/ATmega32U4

29.7 SPI Timing Characteristics

See Figure 29-4 and Figure 29-5 for details.

Table 29-3. SPI Timing Parameters

Description

SCK period

SCK high/low

Rise/Fall time

Setup

Mode

Master

Slave

Min

Typ

Max

1

2

See Table 17-4

50% duty cycle

3

TBD

10

4

5

Hold

10

6

Out to SCK

SCK to out

SCK to out high

SS low to out

SCK period

SCK high/low⁽¹⁾

Rise/Fall time

Setup

0.5 • t_sck

10

7

8

10

9

15

ns

10

11

12

13

14

15

16

17

18

Slave

4 • t_ck

2 • t_ck

Slave

TBD

Slave

10

t_ck

Hold

Slave

SCK to out

SCK to SS high

SS high to tri-state

SS low to SCK

Slave

15

10

Slave

20

Slave

Note:

1. In SPI Programming mode the minimum SCK high/low period is:

- 2 t_CLCLfor f_CK< 12 MHz

- 3 t_CLCLfor f_CK> 12 MHz

Figure 29-4. SPI Interface Timing Requirements (Master Mode)

SS

6

1

SCK

(CPOL = 0)

2

SCK

(CPOL = 1)

4

5

3

MISO

(Data Input)

MSB

...

LSB

7

8

MOSI

(Data Output)

MSB

...

LSB

383

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 29-5. SPI Interface Timing Requirements (Slave Mode)

SS

10

16

9

SCK

(CPOL = 0)

11

SCK

(CPOL = 1)

13

14

12

MOSI

(Data Input)

MSB

...

LSB

15

17

MISO

(Data Output)

MSB

...

LSB

X

29.8 Hardware Boot EntranceTiming Characteristics

Figure 29-6. Hardware Boot Timing Requirements

RESET

t_SHRH

t_HHRH

ALE/HWB

Table 29-4. Hardware Boot Timings

Symbol Parameter

Min

Max

HWB low Setup before Reset High

0

tSHRH

StartUpTime(

SUT) + Time

Out

HWB low Hold after Reset High

tHHRH

Delay(TOUT)

Table 29-5. ADC Characteristics

Symbol Parameter

Condition

Min

Typ

10

8

Max

Units

Single Ended Conversion

Resolution

Differential conversion, gain = 1x/10x/40x

Differential conversion, gain = 200x

Bits

8

V

_REF= 4V, V_CC= 4V, ADC clock = 200 kHz

2.0

3.0

Gain = 1x/10x/40x, V_REF= 4V, V_CC= 5V,

ADC clock = 200 kHz

2.0

TUE

Absolute accuracy

LSB

Gain = 200x, V_REF= 4V, V_CC= 5V,

ADC clock = 200 kHz

4.0

384

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table 29-5. ADC Characteristics

Symbol Parameter

Condition

Min

Typ

Max

Units

V_REF= 4V, V_CC= 4V, ADC clock = 200 kHz

0.5

1.5

Gain = 1x/10x/40x, V_REF= 4V, V_CC= 5V,

ADC clock = 200 kHz

0.3

1.5

INL

Integral Non-Linearity

LSB

Gain = 200x, V_REF= 4V, V_CC= 5V,

ADC clock = 200 kHz

0.5

0.4

0.3

1.5

0.7

1.0

V_REF= 4V, V_CC= 4V, ADC clock = 200 kHz

Gain = 1x/10x/40x, V_REF= 4V, V_CC= 5V,

ADC clock = 200 kHz

DNL

Differential Non-Linearity

LSB

Gain = 200x, V_REF= 4V, V_CC= 5V,

ADC clock = 200 kHz

0.6

-1.0

-1.5

1.0

2.5

V_REF= 4V, V_CC= 4V,ADC clock = 200 kHz

-2.5

0.0

Gain = 1x/10x/40x, V_REF= 4V, V_CC= 5V,

ADC clock = 200 kHz

-2.5

Gain Error

LSB

Gain = 200x, V_REF= 4V, V_CC= 5V,

ADC clock = 200 kHz

0.0

-2.5

-2.0

-1.8

1.5

0.0

-3.0

2.5

2.0

V_REF= 4V, V_CC= 4V, ADC clock = 200 kHz

Offset Error

V_REF= 4V, V_CC= 5V, ADC clock = 200 kHz,

Differential mode

Single Ended Conversion

Differential Conversion

2.56

AVCC

AVCC - 0.5

V_CC+ 0.3

V_REF

AVCC

V_IN

Reference Voltage

Analog Supply Voltage

Input Voltage

V

V_CC- 0.3

GND

0

Single ended channels

Differential Conversion

Single Ended Channels

Differential Channels

2.56V

AVCC

38.5

4

Input Bandwidth

kHz

V_INT

Internal Voltage Reference

Reference Input Resistance

Analog Input Resistance

2.4

2.56

32

2.8

V

R_REF

R_AIN

kΩ

MΩ

100

385

7766E–AVR–04/10

ATmega16U4/ATmega32U4

30. Typical Characteristics

The following charts show typical behavior. These figures are not tested during manufacturing.

All current consumption measurements are performed with all I/O pins configured as inputs and

with internal pull-ups enabled. A sine wave generator with rail-to-rail output is used as clock

source.

All Active- and Idle current consumption measurements are done with all bits in the PRR regis-

ters set and thus, the corresponding I/O modules are turned off. Also the Analog Comparator is

disabled during these measurements. See “Power Reduction Register” on page 45 for details.

The power consumption in Power-down mode is independent of clock selection.

The current consumption is a function of several factors such as: operating voltage, operating

frequency, loading of I/O pins, switching rate of I/O pins, code executed and ambient tempera-

ture. The dominating factors are operating voltage and frequency.

The current drawn from capacitive loaded pins may be estimated (for one pin) as C_L*V_CC*f

whereC_L= load capacitance, V_CC= operating voltage and f = average switching frequency of I/O

pin.

The parts are characterized at frequencies higher than test limits. Parts are not guaranteed to

function properly at frequencies higher than the ordering code indicates.

The difference between current consumption in Power-down mode with Watchdog Timer

enabled and Power-down mode with Watchdog Timer disabled represents the differential cur-

rent drawn by the Watchdog Timer.

30.1 Active Supply Current

Figure 30-1. ATmega16/32U4: Active Supply Current vs. Low Frequency (1MHz) and T= 25°C

1.6

5.5 V

1.4

1.2

1

5.0 V

4.5 V

4.0 V

3.6 V

0.8

0.6

0.4

0.2

0

2.7 V

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Frequency (MHz)

386

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-2. ATmega16/32U4: Active Supply Current vs. Low Frequency (1MHz) and T= 85°C

2.3

5.5 V

2.1

5.0 V

4.5 V

1.9

1.7

1.5

1.3

1.1

0.9

0.7

4.0 V

3.6 V

2.7 V

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Frequency (MHz)

Figure 30-3. ATmega16/32U4: Active Supply Current vs. Frequency (1-16 MHz) and T= -40°C

18

5.5V

5.0V

4.5V

16

14

12

10

8

4.0V

3.6V

6

2.7V

4

2

0

2

4

6

8

10

12

14

16

Frequency (MHz)

387

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-4. ATmega16/32U4: Active Supply Current vs. Frequency (1-16 MHz) and T = 25°C

16

5.5V

14

5.0V

12

4.5V

10

4.0V

8

3.6V

6

4

2.7V

2

0

2

4

6

8

10

12

14

16

Frequency (MHz)

Figure 30-5. ATmega16/32U4: Active Supply Current vs. Frequency (1-16 MHz) and T = 85°C

16

5.5V

14

5.0V

12

4.5V

10

4.0V

8

3.6V

6

4

2.7V

2

0

2

4

6

8

10

12

14

16

Frequency (MHz)

388

7766E–AVR–04/10

ATmega16U4/ATmega32U4

30.2 Idle Supply Current

Figure 30-6. ATmega16/32U4: Idle Supply Current vs. Low Frequency (1 MHz) and T = 25°C

0.5

5.5 V

0.45

0.4

5.0 V

4.5 V

0.35

0.3

4.0 V

3.6 V

0.25

0.2

2.7 V

0.15

0.1

0.05

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Frequency (MHz)

Figure 30-7. ATmega16/32U4: Idle Supply Current vs. Low Frequency (1 MHz) and T = 85°C

0.5

5.5 V

0.45

5.0 V

0.4

4.5 V

0.35

4.0 V

3.6 V

0.3

0.25

2.7 V

0.2

0.15

0.1

0.05

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Frequency (MHz)

389

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-8. ATmega16/32U4: Idle Supply Current vs. Frequency (1-16 MHz) T = 25°C

7

5.5V

6

5.0V

5

4.5V

4

4.0V

3

3.6V

2

3.3V

2.7V

1

0

2

4

6

8

10

12

14

16

Frequency (MHz)

Figure 30-9. ATmega16/32U4: Idle Supply Current vs. Frequency (1-16 MHz) T = 85°C

7

5.5V

6

5.0V

5

4.5V

4

4.0V

3

3.6V

2

2.7V

1

0

2

4

6

8

10

12

14

16

Frequency (MHz)

390

7766E–AVR–04/10

ATmega16U4/ATmega32U4

30.3 Power-down Supply Current

Figure 30-10. ATmega16/32U4: Power-Down Supply Current vs. V_CC(WDT Disabled)

3.5

85 °C

3

2.5

2

1.5

1

25 °C

-40 °C

0.5

0

2.5

2.8

3.1

3.4

3.7

4

4.3

4.6

4.9

5.2

5.5

V_CC(V)

Figure 30-11. ATmega16/32U4: Power-Down Supply Current vs. V_CC(WDT Enabled)

24

85 °C

22

20

18

16

14

12

10

8

25 °C

-40 °C

6

4

2.5

2.8

3.1

3.4

3.7

4

4.3

4.6

4.9

5.2

5.5

V

_CC(V)

391

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-12. ATmega16/32U4: Power-Down Supply Current vs. V_CC(WDT Enabled, BOD EN)

48

85 °C

45

42

39

25 °C

36

-40 °C

33

30

27

24

21

2.5

2.8

3.1

3.4

3.7

4

4.3

4.6

4.9

5.2

5.5

V

_CC(V)

30.4 Power-save Supply Current

Figure 30-13. ATmega16/32U4: Power-Save Supply Current vs. V_CC(WDT Disabled )

200

-40 °C

185

170

155

140

125

110

95

25 °C

85 °C

80

65

50

2.5

2.8

3.1

3.4

3.7

4

4.3

4.6

4.9

5.2

5.5

V

_CC(V)

392

7766E–AVR–04/10

ATmega16U4/ATmega32U4

30.5 Pin Pull-Up

Figure 30-14. ATmega16/32U4: I/O Pin Pull-up Resistor Current vs. Input Voltage (V_CC= 2.7 V)

80

70

60

50

40

30

20

25 °C

-40 °C

10

85 °C

0

0.5

1

1.5

2

2.5

3

V

_OP(V)

Figure 30-15. ATmega16/32U4: I/O Pin Pull-up Resistor Current vs. Input Voltage (V_CC= 5 V)

140

120

100

80

60

40

25 °C

85 °C

20

-40 °C

0

1

2

3

4

5

V

_OP(V)

393

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-16. ATmega16/32U4: Reset Pull-up Resistor Current vs. Reset Pin Voltage (V_CC=5V)

120

100

80

60

40

25 °C

85 °C

-40 °C

20

0

1

2

3

4

5

V_RESET(V)

30.6 Pin Driver Strength

Figure 30-17. ATmega16/32U4: I/O Pin Output Voltage vs. Sink Current (V_CC= 3 V)

4

85 °C

3.5

3

2.5

2

1.5

1

25 °C

-40 °C

0.5

0

2

4

6

8

10

12

14

16

18

20

I

_OL(mA)

394

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-18. ATmega16/32U4: I/O Pin Output Voltage vs. Sink Current (V_CC= 5 V)

1

85 °C

0.9

0.8

25 °C

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0

-40 °C

0

2

4

6

8

10

12

14

16

18

20

I_OL(mA)

Figure 30-19. ATmega16/32U4: I/O Pin Output Voltage vs. Source Current (Vcc = 3 V)

3.5

3

2.5

2

-40 °C

1.5

1

25 °C

0.5

0

85 °C

0

2

4

6

8

10

12

14

16

18

20

I_OH(mA)

395

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-20. ATmega16/32U4: I/O Pin Output Voltage vs. Source Current (V_CC= 5 V)

5.1

4.9

4.7

4.5

-40 °C

4.3

25 °C

4.1

85 °C

3.9

0

2

4

6

8

10

12

14

16

18

20

I

_OH(mA)

Figure 30-21. ATmega16/32U4: USB DP LO Pull-Up Resistor Current vs. USB Pin Voltage

2800

2400

2000

1600

1200

800

85 °C

400

25 °C

-40 °C

0

0.5

1

1.5

2

2.5

3

3.5

V_USB(V)

396

7766E–AVR–04/10

ATmega16U4/ATmega32U4

30.7 Pin Threshold and Hysteresis

Figure 30-22. ATmega16/32U4: I/O Pin Input Threshold Voltage vs. V_CC(V_IH, IO Pin read as ‘1’)

1.8

-40 °C

25 °C

85 °C

1.6

1.4

1.2

1

0.8

0.6

2.5

2.8

3.1

3.4

3.7

4

4.3

4.6

4.9

5.2

5.5

V_CC(V)

Figure 30-23. ATmega16/32U4: I/O Pin Input Threshold Voltage vs. V_CC(V_IL, IO Pin read as ‘0’)

1.8

-40 °C

25 °C

85 °C

1.6

1.4

1.2

1

0.8

0.6

2.5

2.8

3.1

3.4

3.7

4

4.3

4.6

4.9

5.2

5.5

V_CC(V)

397

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-24. ATmega16/32U4: USB Pin Input Threshold Voltage vs. V_CC(V_IH, IO Pin read as ‘1’)

2

85 °C

25 °C

1.9

1.8

-40 °C

1.7

1.6

1.5

1.4

1.3

1.2

2.7

2.8

2.9

3

3.1

3.2

3.3

3.4

3.5

3.6

3.7

3.8

V_CC(V)

Figure 30-25. ATmega16/32U4: USB Pin Input Threshold Voltage vs. V_CC(V_IL, I/O Pin read as ‘0’)

1.6

-40 °C

85 °C

25 °C

1.5

1.4

1.3

1.2

1.1

1

0.9

2.7

2.8

2.9

3

3.1

3.2

3.3

_CC(V)

3.4

3.5

3.6

3.7

3.8

V

398

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-26. ATmega16/32U4: Vbus Pin Input Threshold Voltage vs. V_CC(V_IH, IO Pin read as ‘1’)

4.6

4.58

85 °C

4.56

25 °C

4.54

4.52

4.5

4.48

4.46

4.44

4.42

4.4

-40 °C

4.38

2.5

2.8

3.1

3.4

3.7

4

4.3

4.6

4.9

5.2

5.5

V_CC(V)

Figure 30-27. ATmega16/32U4: Vbus Pin Input Threshold Voltage vs. V_CC(V_IL, I/O Pin read as ‘0’)

1.52

85 °C

25 °C

1.51

1.5

1.49

1.48

1.47

-40 °C

1.46

2.5

2.8

3.1

3.4

3.7

4

4.3

4.6

4.9

5.2

5.5

V

_CC(V)

399

7766E–AVR–04/10

ATmega16U4/ATmega32U4

30.8 BOD Threshold

Figure 30-28. ATmega16/32U4: BOD Thresholds vs. Temperature (BODLEVEL is 2.6 V)

2.8

Rising Vcc

2.78

2.76

2.74

Falling Vcc

2.72

2.7

2.68

2.66

2.64

-50 -40 -30 -20 -10

0

10

20

30

40

50

60

70

80

90

Temperature (°C)

Figure 30-29. ATmega16/32U4: BOD Thresholds vs. Temperature (BODLEVEL is 3.5 V)

3.73

Rising Vcc

3.69

Falling Vcc

3.65

3.61

3.57

3.53

3.49

3.45

-50 -40 -30 -20 -10

0

10

20

30

40

50

60

70

80

90

Temperature (°C)

400

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-30. ATmega16/32U4: BOD Thresholds vs. Temperature (BODLEVEL is 4.3 V)

4.6

4.55

Falling Vcc

Rising Vcc

4.5

4.45

4.4

4.35

4.3

4.25

4.2

4.15

4.1

-50 -40 -30 -20 -10

0

10

20

30

40

50

60

70

80

90

Temperature (°C)

Figure 30-31. ATmega16/32U4: Bandgap Voltage vs. Vcc

1.11

1.1

1.09

1.08

1.07

1.06

1.05

85 °C

25 °C

-40 °C

1.5

2

2.5

3

3.5

4

4.5

5

5.5

Vcc (V)

401

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-32. ATmega16/32U4: Bandgap Voltage vs. Temperature

1.11

1.9 V

3.0 V

4.5 V

5.0 V

5.5 V

1.1

1.09

1.08

1.07

1.06

1.05

-40 -30 -20 -10

0

10

20

30

40

50

60

70

80

90

Temperature (°C)

30.9 Internal Oscilllator Speed

Figure 30-33. ATmega16/32U4: Watchdog Oscillator Frequency vs. Temperature

124

122

120

118

116

114

112

1.9 V

3.0 V

4.0 V

4.5 V

5.5 V

-40 -30 -20 -10

0

10

20

30

40

50

60

70

80

90

Temperature (°C)

402

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-34. XXXX: Watchdog Oscillator Frequency vs. V_CC

124

122

120

118

116

114

112

110

-40 °C

25 °C

85 °C

1.5

2

2.5

3

3.5

4

4.5

5

5.5

V_CC(V)

Figure 30-35. ATmega16/32U4: Calibrated 8 MHz RC Oscillator Frequency vs. Osccal Value

14

85 °C

25 °C

-40 °C

12

10

8

6

4

2

0

16 32 48 64 80 96 112 128 144 160 176 192 208 224 240 256

OSCCAL (X1)

403

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-36. ATmega16/32U4: Calibrated 8 MHz RC Oscillator Frequency vs. Temperature

8.4

2.7 V

4.0 V

5.5 V

8.3

8.2

8.1

8

7.9

7.8

7.7

-40

-30

-20

-10

0

10

20

30

40

50

60

70

80

90

Temperature (°C)

Figure 30-37. ATmega16/32U4: Calibrated 8 MHz RC Oscillator Frequency vs. Operating Voltage

8.4

8.3

85 °C

8.2

8.1

8

25 °C

7.9

7.8

-40 °C

7.7

7.6

2.5

2.8

3.1

3.4

3.7

4

4.3

4.6

4.9

5.2

5.5

V_CC(V)

404

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-38. ATmega16/32U4: OSCCAL VALUE STEP SIZE IN % (Base frequency = 0.0 MHz)

1.4

1.2

1

0.8

0.6

0.4

85 °C

-40 °C

0.2

25 °C

0

16 32 48 64 80 96 112 128 144 160 176 192 208 224 240 256

OSCCAL (X1)

30.10 Current Consumption of Peripheral Units

Figure 30-39. ATmega16/32U4: USB Regulator Level vs. V_CC

3.5

3.4

3.3

3.2

3.1

3

-40 °C

85 °C

25 °C

2.9

2.8

2.7

2.7 2.9 3.1 3.3 3.5 3.7 3.9 4.1 4.3 4.5 4.7 4.9 5.1 5.3 5.5

_CC(V)

V

405

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-40. ATmega16/32U4: USB Regulator Level with load 75 Ω vs. V_CC

3.4

3.3

3.2

3.1

3

-40 °C

85 °C

25 °C

2.9

2.8

2.7

2.6

2.5

2.8

3.1

3.4

3.7

4

4.3

V_CC(V)

4.6

4.9

5.2

5.5

Figure 30-41. ATmega16/32U4: ADC Internal Vref vs. Vcc

2.54

2.53

2.52

2.51

2.5

85 °C

25 °C

2.49

2.48

2.47

2.46

2.45

-40 °C

2.6

2.9

3.2

3.5

3.8

4.1

4.4

4.7

5

5.3

5.6

Voltage (V)

406

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Figure 30-42. ATmega16/32U4: Internal Reference Voltage vs. Sink Current

2.52

2.51

2.5

85 °C

25 °C

2.49

2.48

2.47

2.46

2.45

2.44

2.43

-40 °C

-7

-6

-5

-4

-3

-2

-1

0

Sink current (mA)

30.11 Current Consumption in Reset and Reset Pulsewidth

Figure 30-43. ATmega16/32U4: Reset Supply Current vs. Frequency (1 - 20 MHz)

3.5

5.5V

3

2.5

2

5.0V

4.5V

4.0V

3.6V

1.5

1

2.7V

0.5

0

2

4

6

8

10

12

14

16

Frequency (MHz)

407

7766E–AVR–04/10

ATmega16U4/ATmega32U4

31. Register Summary

Address

Name

Bit 7

Bit 6

Bit 5

Bit 4

Bit 3

Bit 2

Bit 1

Bit 0

Page

(0xFF)

(0xFE)

(0xFD)

(0xFC)

(0xFB)

(0xFA)

(0xF9)

(0xF8)

(0xF7)

(0xF6)

(0xF5)

(0xF4)

(0xF3)

(0xF2)

(0xF1)

(0xF0)

(0xEF)

(0xEE)

(0xED)

(0xEC)

(0xEB)

(0xEA)

(0xE9)

(0xE8)

(0xE7)

(0xE6)

(0xE5)

(0xE4)

(0xE3)

(0xE2)

(0xE1)

(0xE0)

(0xDF)

(0xDE)

(0xDD)

(0xDC)

(0xDB)

(0xDA)

(0xD9)

(0xD8)

(0xD7)

(0xD6)

(0xD5)

(0xD4)

(0xD3)

(0xD2)

(0xD1)

(0xD0)

(0xCF)

(0xCE)

(0xCD)

(0xCC)

(0xCB)

(0xCA)

(0xC9)

(0xC8)

(0xC7)

(0xC6)

(0xC5)

(0xC4)

(0xC3)

(0xC2)

(0xC1)

(0xC0)

(0xBF)

Reserved

UEINT

-

EPINT6:0

-

UEBCHX

UEBCLX

UEDATX

UEIENX

UESTA1X

UESTA0X

UECFG1X

UECFG0X

UECONX

UERST

-

BYCT10:8

BYCT7:0

DAT7:0

RXSTPE

FLERRE

-

NAKINE

-

NAKOUTE

RXOUTE

CTRLDIR

STALLEDE

TXINE

-

CURRBK1:0

NBUSYBK1:0

ALLOC

CFGOK

OVERFI

UNDERFI

EPSIZE2:0

-

DTSEQ1:0

EPBK1:0

-

EPTYPE1:0

-

EPDIR

EPEN

-

STALLRQ

STALLRQC

RSTDT

-

EPRST6:0

UENUM

UEINTX

Reserved

UDMFN

-

EPNUM2:0

STALLEDI

FIFOCON

NAKINI

RWAL

NAKOUTI

RXSTPI

RXOUTI

TXINI

-

FNCERR

-

UDFNUMH

UDFNUML

UDADDR

UDIEN

FNUM10:8

FNUM7:0

ADDEN

UADD6:0

EORSTE

EORSTI

-

UPRSME

UPRSMI

-

EORSME

EORSMI

-

WAKEUPE

WAKEUPI

-

SOFE

SOFI

LSM

MSOFE

MSOFI

SUSPE

SUSPI

UDINT

UDCON

Reserved

USBINT

USBSTA

USBCON

UHWCON

Reserved

DT4

RSTCPU

RMWKUP

DETACH

-

ID

-

VBUSTI

VBUS

-

USBE

-

FRZCLK

-

OTGPADE

-

VBUSTE

UVREGE

-

DT4H3

DT4H2

DT4H1

DT4H0

DT4L3

DT4L2

DT4L1

DT4L0

Reserved

OCR4D

Timer/Counter4 - Output Compare Register D

Timer/Counter4 - Output Compare Register C

Timer/Counter4 - Output Compare Register B

Timer/Counter4 - Output Compare Register A

USART1 I/O Data Register

OCR4C

OCR4B

OCR4A

UDR1

UBRR1H

UBRR1L

Reserved

UCSR1C

UCSR1B

UCSR1A

CLKSTA

CLKSEL1

CLKSEL0

TCCR4E

TCCR4D

TCCR4C

TCCR4B

TCCR4A

TC4H

-

USART1 Baud Rate Register High Byte

USART1 Baud Rate Register Low Byte

-

USBS1

TXEN1

DOR1

-

UMSEL11

RXCIE1

RXC1

UMSEL10

TXCIE1

TXC1

UPM11

UDRIE1

UDRE1

-

UPM10

RXEN1

FE1

UCSZ11

UCSZ12

PE1

UCSZ10

RXB81

U2X1

UCPOL1

TXB81

MPCM1

EXTON

EXCKSEL0

CLKS

-

RCON

EXCKSEL1

-

RCCKSEL3

RCSUT1

TLOCK4

FPIE4

RCCKSEL2

RCSUT0

ENHC4

FPEN4

COM4A0S

PSR4

RCCKSEL1

EXSUT1

OC4OE5

FPNC4

COM4B1S

DTPS41

COM4B1

-

RCCKSEL0

EXSUT0

OC4OE4

FPES4

COM4B0S

DTPS40

COM4B0

-

EXCKSEL3

RCE

EXCKSEL2

EXTE

OC4OE3

FPAC4

COM4D1S

CS43

OC4OE2

FPF4

OC4OE1

WGM41

FOC4D

CS41

OC4OE0

WGM40

PWM4D

CS40

COM4A1S

PWM4X

COM4A1

-

COM4D0S

CS42

COM4A0

-

FOC4A

-

FOC4B

PWM4A

PWM4B

Timer/Counter4 High Byte

408

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Address

Name

Bit 7

Bit 6

Bit 5

Bit 4

Bit 3

Bit 2

Bit 1

Bit 0

Page

(0xBE)

(0xBD)

(0xBC)

(0xBB)

(0xBA)

(0xB9)

(0xB8)

(0xB7)

(0xB6)

(0xB5)

(0xB4)

(0xB3)

(0xB2)

(0xB1)

(0xB0)

(0xAF)

(0xAE)

(0xAD)

(0xAC)

(0xAB)

(0xAA)

(0xA9)

(0xA8)

(0xA7)

(0xA6)

(0xA5)

(0xA4)

(0xA3)

(0xA2)

(0xA1)

(0xA0)

(0x9F)

(0x9E)

(0x9D)

(0x9C)

(0x9B)

(0x9A)

(0x99)

(0x98)

(0x97)

(0x96)

(0x95)

(0x94)

(0x93)

(0x92)

(0x91)

(0x90)

(0x8F)

(0x8E)

(0x8D)

(0x8C)

(0x8B)

(0x8A)

(0x89)

(0x88)

(0x87)

(0x86)

(0x85)

(0x84)

(0x83)

(0x82)

(0x81)

(0x80)

(0x7F)

(0x7E)

(0x7D)

TCNT4

TWAMR

TWCR

Timer/Counter4 - Counter Register Low Byte

TWAM6

TWINT

TWAM5

TWEA

TWAM4

TWSTA

TWAM3

TWSTO

TWAM2

TWWC

TWAM1

TWEN

TWAM0

-

TWIE

TWDR

2-wire Serial Interface Data Register

TWAR

TWA6

TWS7

TWA5

TWS6

TWA4

TWS5

TWA3

TWS4

TWA2

TWS3

TWA1

-

TWA0

TWGCE

TWPS0

TWSR

TWPS1

TWBR

2-wire Serial Interface Bit Rate Register

Reserved

OCR3CH

OCR3CL

OCR3BH

OCR3BL

OCR3AH

OCR3AL

ICR3H

-

Timer/Counter3 - Output Compare Register C High Byte

Timer/Counter3 - Output Compare Register C Low Byte

Timer/Counter3 - Output Compare Register B High Byte

Timer/Counter3 - Output Compare Register B Low Byte

Timer/Counter3 - Output Compare Register A High Byte

Timer/Counter3 - Output Compare Register A Low Byte

Timer/Counter3 - Input Capture Register High Byte

Timer/Counter3 - Input Capture Register Low Byte

Timer/Counter3 - Counter Register High Byte

ICR3L

TCNT3H

TCNT3L

Reserved

TCCR3C

TCCR3B

TCCR3A

Reserved

OCR1CH

OCR1CL

OCR1BH

OCR1BL

OCR1AH

OCR1AL

ICR1H

Timer/Counter3 - Counter Register Low Byte

-

FOC3A

-

ICNC3

ICES3

-

WGM33

WGM32

CS32

CS31

CS30

COM3A1

COM3A0

COM3B1

COM3B0

COM3C1

COM3C0

WGM31

WGM30

-

Timer/Counter1 - Output Compare Register C High Byte

Timer/Counter1 - Output Compare Register C Low Byte

Timer/Counter1 - Output Compare Register B High Byte

Timer/Counter1 - Output Compare Register B Low Byte

Timer/Counter1 - Output Compare Register A High Byte

Timer/Counter1 - Output Compare Register A Low Byte

Timer/Counter1 - Input Capture Register High Byte

Timer/Counter1 - Input Capture Register Low Byte

Timer/Counter1 - Counter Register High Byte

ICR1L

TCNT1H

TCNT1L

Reserved

TCCR1C

TCCR1B

TCCR1A

DIDR1

Timer/Counter1 - Counter Register Low Byte

-

FOC1A

ICNC1

COM1A1

-

FOC1B

ICES1

COM1A0

-

FOC1C

-

WGM13

COM1B0

-

WGM12

CS12

CS11

WGM11

-

CS10

WGM10

AIN0D

ADC0D

ADC8D

COM1B1

-

COM1C1

COM1C0

-

DIDR0

ADC7D

-

ADC6D

-

ADC5D

ADC13D

ADC4D

ADC12D

-

ADC1D

ADC9D

DIDR2

ADC11D

ADC10D

409

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Address

Name

Bit 7

Bit 6

Bit 5

Bit 4

Bit 3

Bit 2

Bit 1

Bit 0

Page

(0x7C)

(0x7B)

ADMUX

ADCSRB

ADCSRA

ADCH

REFS1

ADHSM

ADEN

REFS0

ACME

ADSC

ADLAR

MUX5

MUX4

-

MUX3

ADTS3

ADIE

MUX2

ADTS2

ADPS2

MUX1

ADTS1

ADPS1

MUX0

ADTS0

ADPS0

(0x7A)

ADATE

ADIF

(0x79)

ADC Data Register High byte

ADC Data Register Low byte

(0x78)

ADCL

(0x77)

Reserved

TIMSK4

TIMSK3

Reserved

TIMSK1

TIMSK0

Reserved

PCMSK0

EICRB

-

(0x76)

-

(0x75)

-

(0x74)

-

(0x73)

-

(0x72)

OCIE4D

OCIE4A

OCIE4B

-

TOIE4

-

TOIE3

-

(0x71)

-

ICIE3

-

OCIE3C

OCIE3B

OCIE3A

(0x70)

-

(0x6F)

-

ICIE1

-

OCIE1C

OCIE1B

OCIE1A

TOIE1

TOIE0

-

(0x6E)

-

OCIE0B

OCIE0A

(0x6D)

-

(0x6C)

-

PCINT5

ISC61

ISC21

-

PCINT4

ISC60

ISC20

-

(0x6B)

PCINT7

PCINT6

PCINT3

PCINT2

PCINT1

PCINT0

-

(0x6A)

-

(0x69)

EICRA

ISC31

ISC30

ISC11

ISC10

ISC01

ISC00

PCIE0

RCFREQ

(0x68)

PCICR

-

(0x67)

RCCTRL

OSCCAL

PRR1

-

(0x66)

RC Oscillator Calibration Register

(0x65)

PRUSB

-

PRTIM4

PRTIM3

-

PRUSART1

(0x64)

PRR0

PRTWI

-

PRTIM0

-

PRTIM1

PRSPI

-

PRADC

(0x63)

Reserved

CLKPR

-

(0x62)

-

CLKPS1

WDP1

Z

-

(0x61)

CLKPCE

-

CLKPS3

CLKPS2

CLKPS0

(0x60)

WDTCSR

SREG

WDIF

WDIE

WDP3

WDCE

WDE

WDP2

WDP0

0x3F (0x5F)

0x3E (0x5E)

0x3D (0x5D)

0x3C (0x5C)

0x3B (0x5B)

0x3A (0x5A)

0x39 (0x59)

0x38 (0x58)

0x37 (0x57)

0x36 (0x56)

0x35 (0x55)

0x34 (0x54)

0x33 (0x53)

0x32 (0x52)

I

T

H

S

V

N

C

SPH

SP15

SP14

SP13

SP12

SP11

SP10

SP9

SP1

-

SP8

SPL

SP7

SP6

SP5

SP4

SP3

SP2

SP0

Reserved

RAMPZ

Reserved

SPMCSR

Reserved

MCUCR

MCUSR

SMCR

-

RAMPZ1

-

RAMPZ0

-

RWWSRE

-

SPMIE

RWWSB

SIGRD

BLBSET

-

PGWRT

-

PGERS

-

SPMEN

-

JTD

-

PUD

-

IVSEL

EXTRF

SM0

PDIV1

OCDR1

IVCE

PORF

SE

-

USBRF

-

JTRF

-

WDRF

SM2

PDIV3

OCDR3

BORF

SM1

PDIV2

OCDR2

-

PLLFRQ

PINMUX

OCDR7

PLLUSB

OCDR6

PLLTM1

OCDR5

PLLTM0

OCDR4

PDIV0

OCDR0

OCDR/

MONDR

0x31 (0x51)

Monitor Data Register

0x30 (0x50)

0x2F (0x4F)

0x2E (0x4E)

0x2D (0x4D)

0x2C (0x4C)

0x2B (0x4B)

0x2A (0x4A)

0x29 (0x49)

0x28 (0x48)

0x27 (0x47)

0x26 (0x46)

0x25 (0x45)

0x24 (0x44)

0x23 (0x43)

0x22 (0x42)

0x21 (0x41)

0x20 (0x40)

0x1F (0x3F)

0x1E (0x3E)

0x1D (0x3D)

0x1C (0x3C)

ACSR

Reserved

SPDR

ACD

-

ACBG

-

ACO

-

ACI

-

ACIE

-

ACIC

-

ACIS1

-

ACIS0

-

SPI Data Register

-

SPSR

SPIF

SPIE

WCOL

SPE

-

SPI2X

SPR0

SPCR

DORD

MSTR

CPOL

CPHA

SPR1

GPIOR2

GPIOR1

PLLCSR

OCR0B

OCR0A

TCNT0

TCCR0B

TCCR0A

GTCCR

EEARH

EEARL

EEDR

General Purpose I/O Register 2

General Purpose I/O Register 1

-

PINDIV

-

PLLE

PLOCK

Timer/Counter0 Output Compare Register B

Timer/Counter0 Output Compare Register A

Timer/Counter0 (8 Bit)

FOC0A

COM0A1

TSM

FOC0B

-

WGM02

CS02

CS01

CS00

COM0A0

COM0B1

COM0B0

-

WGM01

PSRASY

WGM00

-

PSRSYNC

-

EEPROM Address Register High Byte

EEPROM Address Register Low Byte

EEPROM Data Register

EECR

-

EEPM1

EEPM0

EERIE

EEMPE

EEPE

EERE

GPIOR0

EIMSK

EIFR

General Purpose I/O Register 0

-

INT6

-

INT3

INT2

INT1

INT0

INTF6

INTF3

INTF2

INTF1

INTF0

410

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Address

Name

Bit 7

Bit 6

Bit 5

Bit 4

Bit 3

Bit 2

Bit 1

Bit 0

Page

0x1B (0x3B)

0x1A (0x3A)

0x19 (0x39)

0x18 (0x38)

0x17 (0x37)

0x16 (0x36)

0x15 (0x35)

0x14 (0x34)

0x13 (0x33)

0x12 (0x32)

0x11 (0x31)

0x10 (0x30)

0x0F (0x2F)

0x0E (0x2E)

0x0D (0x2D)

0x0C (0x2C)

0x0B (0x2B)

0x0A (0x2A)

0x09 (0x29)

0x08 (0x28)

0x07 (0x27)

0x06 (0x26)

0x05 (0x25)

0x04 (0x24)

0x03 (0x23)

0x02 (0x22)

0x01 (0x21)

0x00 (0x20)

PCIFR

Reserved

TIFR4

-

PCIF0

-

OCF4D

OCF4A

OCF4B

-

TOV4

-

TIFR3

-

ICF3

-

OCF3C

OCF3B

OCF3A

TOV3

Reserved

TIFR1

-

ICF1

-

OCF1C

OCF1B

OCF1A

TOV1

TIFR0

-

OCF0B

OCF0A

TOV0

Reserved

PORTF

DDRF

-

PORTF7

DDF7

PINF7

-

PORTF6

DDF6

PINF6

PORTE6

DDE6

PINE6

PORTD6

DDD6

PIND6

PORTC6

DDC6

PINC6

PORTB6

DDB6

PINB6

-

PORTF5

PORTF4

-

PORTF1

PORTF0

DDF5

DDF4

-

DDF1

DDF0

PINF

PINF5

PINF4

-

PINF1

PINF0

PORTE

DDRE

-

PORTE2

-

DDE2

-

PINE

-

PINE2

-

PORTD

DDRD

PORTD7

DDD7

PIND7

PORTC7

DDC7

PINC7

PORTB7

DDB7

PINB7

-

PORTD5

PORTD4

PORTD3

PORTD2

PORTD1

PORTD0

DDD5

DDD4

DDD3

DDD2

DDD1

DDD0

PIND

PIND5

PIND4

PIND3

PIND2

PIND1

PIND0

PORTC

DDRC

-

PINC

-

PORTB

DDRB

PORTB5

PORTB4

PORTB3

PORTB2

PORTB1

PORTB0

DDB5

DDB4

DDB3

DDB2

DDB1

DDB0

PINB

PINB5

PINB4

PINB3

PINB2

PINB1

PINB0

Reserved

-

Note:

1. For compatibility with future devices, reserved bits should be written to zero if accessed. Reserved I/O memory addresses

should never be written.

2. I/O registers within the address range $00 - $1F are directly bit-accessible using the SBI and CBI instructions. In these reg-

isters, the value of single bits can be checked by using the SBIS and SBIC instructions.

3. Some of the status flags are cleared by writing a logical one to them. Note that the CBI and SBI instructions will operate on

all bits in the I/O register, writing a one back into any flag read as set, thus clearing the flag. The CBI and SBI instructions

work with registers 0x00 to 0x1F only.

4. When using the I/O specific commands IN and OUT, the I/O addresses $00 - $3F must be used. When addressing I/O regis-

ters as data space using LD and ST instructions, $20 must be added to these addresses. The ATmega16U4/ATmega32U4 is

a complex microcontroller with more peripheral units than can be supported within the 64 location reserved in Opcode for

the IN and OUT instructions. For the Extended I/O space from $60 - $1FF in SRAM, only the ST/STS/STD and LD/LDS/LDD

instructions can be used.

411

7766E–AVR–04/10

ATmega16U4/ATmega32U4

32. Instruction Set Summary

Mnemonics

Operands

Description

Operation

Flags

#Clocks

ARITHMETIC AND LOGIC INSTRUCTIONS

ADD

ADC

Rd, Rr

Rdl,K

Rd, Rr

Rd, K

Rd, Rr

Rd, K

Rdl,K

Rd, Rr

Rd, K

Rd, Rr

Rd, K

Rd, Rr

Rd

Add two Registers

Add with Carry two Registers

Add Immediate to Word

Subtract two Registers

Subtract Constant from Register

Subtract with Carry two Registers

Subtract with Carry Constant from Reg.

Subtract Immediate from Word

Logical AND Registers

Logical AND Register and Constant

Logical OR Registers

Rd ← Rd + Rr

Rd ← Rd + Rr + C

Rdh:Rdl ← Rdh:Rdl + K

Rd ← Rd - Rr

Z,C,N,V,H

Z,C,N,V,S

Z,C,N,V,H

Z,C,N,V,S

Z,N,V

1

2

1

2

1

2

ADIW

SUB

SUBI

SBC

Rd ← Rd - K

Rd ← Rd - Rr - C

Rd ← Rd - K - C

Rdh:Rdl ← Rdh:Rdl - K

Rd ← Rd • Rr

SBCI

SBIW

AND

ANDI

OR

Rd ← Rd • K

Z,N,V

Rd ← Rd v Rr

Z,N,V

ORI

Logical OR Register and Constant

Exclusive OR Registers

One’s Complement

Rd ← Rd v K

Z,N,V

EOR

COM

NEG

SBR

Rd ← Rd ⊕ Rr

Z,N,V

Rd ← 0xFF − Rd

Rd ← 0x00 − Rd

Rd ← Rd v K

Z,C,N,V

Z,C,N,V,H

Z,N,V

Rd

Two’s Complement

Rd,K

Rd

Set Bit(s) in Register

CBR

Clear Bit(s) in Register

Increment

Rd ← Rd • (0xFF - K)

Rd ← Rd + 1

Z,N,V

INC

Z,N,V

DEC

Rd

Decrement

Rd ← Rd − 1

Z,N,V

TST

Rd

Test for Zero or Minus

Rd ← Rd • Rd

Z,N,V

CLR

Rd

Clear Register

Rd ← Rd ⊕ Rd

Rd ← 0xFF

Z,N,V

SER

Rd

Set Register

None

MUL

Rd, Rr

Multiply Unsigned

R1:R0 ← Rd x Rr

^{R1:R0 ← (Rd x Rr)}<< 1

R1:R0 ← (Rd x Rr) << 1

Z,C

MULS

MULSU

FMUL

FMULS

FMULSU

Multiply Signed

Z,C

Multiply Signed with Unsigned

Fractional Multiply Unsigned

Fractional Multiply Signed

Fractional Multiply Signed with Unsigned

Z,C

BRANCH INSTRUCTIONS

RJMP

IJMP

k

Relative Jump

Indirect Jump to (Z)

PC ← PC + k + 1

PC ← Z

None

I

2

PC ←(EIND:Z)

EIJMP

JMP

Extended Indirect Jump to (Z)

Direct Jump

2

k

PC ← k

PC ← PC + k + 1

PC ← Z

3

RCALL

ICALL

EICALL

CALL

RET

Relative Subroutine Call

Indirect Call to (Z)

4

PC ←(EIND:Z)

Extended Indirect Call to (Z)

Direct Subroutine Call

Subroutine Return

4

k

PC ← k

5

PC ← STACK

5

RETI

Interrupt Return

PC ← STACK

5

CPSE

CP

Rd,Rr

Compare, Skip if Equal

Compare

if (Rd = Rr) PC ← PC + 2 or 3

Rd − Rr

None

Z, N,V,C,H

None

1/2/3

1

Rd,Rr

CPC

Rd,Rr

Compare with Carry

Rd − Rr − C

1

CPI

Rd,K

Compare Register with Immediate

Skip if Bit in Register Cleared

Skip if Bit in Register is Set

Skip if Bit in I/O Register Cleared

Skip if Bit in I/O Register is Set

Branch if Status Flag Set

Branch if Status Flag Cleared

Branch if Equal

Rd − K

1

SBRC

SBRS

SBIC

Rr, b

if (Rr(b)=0) PC ← PC + 2 or 3

if (Rr(b)=1) PC ← PC + 2 or 3

if (P(b)=0) PC ← PC + 2 or 3

if (P(b)=1) PC ← PC + 2 or 3

if (SREG(s) = 1) then PC←PC+k + 1

if (SREG(s) = 0) then PC←PC+k + 1

if (Z = 1) then PC ← PC + k + 1

if (Z = 0) then PC ← PC + k + 1

if (C = 1) then PC ← PC + k + 1

if (C = 0) then PC ← PC + k + 1

if (C = 1) then PC ← PC + k + 1

if (N = 1) then PC ← PC + k + 1

if (N = 0) then PC ← PC + k + 1

if (N ⊕ V= 0) then PC ← PC + k + 1

if (N ⊕ V= 1) then PC ← PC + k + 1

if (H = 1) then PC ← PC + k + 1

if (H = 0) then PC ← PC + k + 1

if (T = 1) then PC ← PC + k + 1

if (T = 0) then PC ← PC + k + 1

if (V = 1) then PC ← PC + k + 1

1/2/3

1/2

Rr, b

P, b

s, k

k

SBIS

BRBS

BRBC

BREQ

BRNE

BRCS

BRCC

BRSH

BRLO

BRMI

BRPL

BRGE

BRLT

BRHS

BRHC

BRTS

BRTC

BRVS

k

Branch if Not Equal

k

Branch if Carry Set

k

Branch if Carry Cleared

Branch if Same or Higher

Branch if Lower

k

Branch if Minus

k

Branch if Plus

k

Branch if Greater or Equal, Signed

Branch if Less Than Zero, Signed

Branch if Half Carry Flag Set

Branch if Half Carry Flag Cleared

Branch if T Flag Set

k

Branch if T Flag Cleared

Branch if Overflow Flag is Set

k

412

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Mnemonics

Operands

Description

Operation

Flags

#Clocks

BRVC

BRIE

BRID

k

Branch if Overflow Flag is Cleared

Branch if Interrupt Enabled

Branch if Interrupt Disabled

if (V = 0) then PC ← PC + k + 1

if ( I = 1) then PC ← PC + k + 1

if ( I = 0) then PC ← PC + k + 1

None

1/2

BIT AND BIT-TEST INSTRUCTIONS

SBI

CBI

P,b

Rd

s

Set Bit in I/O Register

Clear Bit in I/O Register

Logical Shift Left

I/O(P,b) ← 1

None

2

1

I/O(P,b) ← 0

None

LSL

Rd(n+1) ← Rd(n), Rd(0) ← 0

Z,C,N,V

LSR

ROL

ROR

ASR

SWAP

BSET

BCLR

BST

BLD

SEC

CLC

SEN

CLN

SEZ

CLZ

SEI

Logical Shift Right

Rd(n) ← Rd(n+1), Rd(7) ← 0

Z,C,N,V

Rotate Left Through Carry

Rotate Right Through Carry

Arithmetic Shift Right

Swap Nibbles

Rd(0)←C,Rd(n+1)← Rd(n),C←Rd(7)

Z,C,N,V

Rd(7)←C,Rd(n)← Rd(n+1),C←Rd(0)

Z,C,N,V

Rd(n) ← Rd(n+1), n=0..6

Z,C,N,V

Rd(3..0)←Rd(7..4),Rd(7..4)←Rd(3..0)

None

Flag Set

SREG(s) ← 1

SREG(s) ← 0

T ← Rr(b)

Rd(b) ← T

C ← 1

SREG(s)

s

Flag Clear

SREG(s)

Rr, b

Rd, b

Bit Store from Register to T

Bit load from T to Register

Set Carry

T

None

C

N

Z

Clear Carry

C ← 0

Set Negative Flag

N ← 1

Clear Negative Flag

Set Zero Flag

N ← 0

Z ← 1

Clear Zero Flag

Z ← 0

Z

Global Interrupt Enable

Global Interrupt Disable

Set Signed Test Flag

Clear Signed Test Flag

Set Twos Complement Overflow.

Clear Twos Complement Overflow

Set T in SREG

I ← 1

I

CLI

I ← 0

I

SES

CLS

SEV

CLV

SET

CLT

SEH

CLH

S ← 1

S

S ← 0

S

V ← 1

V

V ← 0

V

T ← 1

T

Clear T in SREG

T ← 0

T

Set Half Carry Flag in SREG

Clear Half Carry Flag in SREG

H ← 1

H

H ← 0

DATA TRANSFER INSTRUCTIONS

MOV

MOVW

LDI

Rd, Rr

Rd, K

Move Between Registers

Copy Register Word

Rd ← Rr

None

1

2

3

Rd+1:Rd ← Rr+1:Rr

Load Immediate

Rd ← K

Rd ← (X)

LD

Rd, X

Load Indirect

LD

Rd, X+

Rd, - X

Rd, Y

Load Indirect and Post-Inc.

Load Indirect and Pre-Dec.

Load Indirect

Rd ← (X), X ← X + 1

X ← X - 1, Rd ← (X)

Rd ← (Y)

LD

Rd, Y+

Rd, - Y

Rd,Y+q

Rd, Z

Load Indirect and Post-Inc.

Load Indirect and Pre-Dec.

Load Indirect with Displacement

Load Indirect

Rd ← (Y), Y ← Y + 1

Y ← Y - 1, Rd ← (Y)

Rd ← (Y + q)

LD

LDD

LD

Rd ← (Z)

LD

Rd, Z+

Rd, -Z

Rd, Z+q

Rd, k

Load Indirect and Post-Inc.

Load Indirect and Pre-Dec.

Load Indirect with Displacement

Load Direct from SRAM

Store Indirect

Rd ← (Z), Z ← Z+1

Z ← Z - 1, Rd ← (Z)

Rd ← (Z + q)

LD

LDD

LDS

ST

Rd ← (k)

X, Rr

(X) ← Rr

ST

X+, Rr

- X, Rr

Y, Rr

Store Indirect and Post-Inc.

Store Indirect and Pre-Dec.

Store Indirect

(X) ← Rr, X ← X + 1

X ← X - 1, (X) ← Rr

(Y) ← Rr

ST

Y+, Rr

- Y, Rr

Y+q,Rr

Z, Rr

Store Indirect and Post-Inc.

Store Indirect and Pre-Dec.

Store Indirect with Displacement

Store Indirect

(Y) ← Rr, Y ← Y + 1

Y ← Y - 1, (Y) ← Rr

(Y + q) ← Rr

ST

STD

ST

(Z) ← Rr

ST

Z+, Rr

-Z, Rr

Z+q,Rr

k, Rr

Store Indirect and Post-Inc.

Store Indirect and Pre-Dec.

Store Indirect with Displacement

Store Direct to SRAM

(Z) ← Rr, Z ← Z + 1

Z ← Z - 1, (Z) ← Rr

(Z + q) ← Rr

ST

STD

STS

LPM

ELPM

(k) ← Rr

Load Program Memory

Load Program Memory and Post-Inc

Extended Load Program Memory

R0 ← (Z)

Rd, Z

Rd ← (Z)

Rd, Z+

Rd ← (Z), Z ← Z+1

R0 ← (RAMPZ:Z)

Rd ← (Z)

Rd, Z

Rd, Z+

Rd ← (RAMPZ:Z), RAMPZ:Z ←RAMPZ:Z+1

413

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Mnemonics

Operands

Description

Operation

Flags

#Clocks

SPM

IN

Store Program Memory

In Port

(Z) ← R1:R0

Rd ← P

None

-

Rd, P

P, Rr

Rr

1

2

OUT

PUSH

POP

Out Port

P ← Rr

Push Register on Stack

Pop Register from Stack

STACK ← Rr

Rd ← STACK

Rd

MCU CONTROL INSTRUCTIONS

NOP

SLEEP

WDR

No Operation

Sleep

None

1

(see specific description for Sleep function)

(see specific description for WDR/timer)

For On-chip Debug Only

Watchdog Reset

Break

1

BREAK

N/A

414

7766E–AVR–04/10

ATmega16U4/ATmega32U4

33. Ordering Information

33.1 ATmega16U4

Speed (MHz)

Power Supply

Ordering Code

Default Oscillator

External XTAL

Package

Operation Range

ATmega16U4-AU

ATmega16U4RC-AU

ATmega16U4-MU

ATmega16U4RC-MU

44ML

Internal Calib. RC

External XTAL

16

2.7 - 5.5 V

Industrial (-40° to +85°C)

44PW

Internal Calib. RC

Package Type

ML, 44 - Lead, 10 x 10 mm Body Size, 1.0 mm Body Thickness

0.8 mm Lead Pitch, Thin Profile Plastic Quad Flat Package (TQFP)

44ML

44PW

PW, 44 - Lead 7.0 x 7.0 mm Body, 0.50 mm Pitch

Quad Flat No Lead Package (QFN)

415

7766E–AVR–04/10

ATmega16U4/ATmega32U4

33.2 ATmega32U4

Speed (MHz)

Power Supply

Ordering Code

Default Oscillator

External XTAL

Package

Operation Range

ATmega32U4-AU

ATmega32U4RC-AU

ATmega32U4-MU

ATmega32U4RC-MU

44ML

Internal Calib. RC

External XTAL

2.7 - 5.5 V

Industrial (-40° to +85°C)

16

44PW

Internal Calib. RC

Package Type

ML, 44 - Lead, 10 x 10 mm Body Size, 1.0 mm Body Thickness

0.8 mm Lead Pitch, Thin Profile Plastic Quad Flat Package (TQFP)

44ML

44PW

PW, 44 - Lead 7.0 x 7.0 mm Body, 0.50 mm Pitch

Quad Flat No Lead Package (QFN)

416

7766E–AVR–04/10

ATmega16U4/ATmega32U4

34. Packaging Information

34.1 TQFP44

417

7766E–AVR–04/10

ATmega16U4/ATmega32U4

34.2 QFN44

418

7766E–AVR–04/10

ATmega16U4/ATmega32U4

35. Errata

The revision letter in this section refers to the revision of the ATmega16U4/ATmega32U4

device.

35.1 ATmega16U4/ATmega32U4 Rev E

• Spike on TWI pins when TWI is enabled

• High current consumption in sleep mode

• MSB of OCR4A/B/D is write only in 11-bits enhanced PWM mode

1. Spike on TWI pins when TWI is enabled

100 ns negative spike occurs on SDA and SCL pins when TWI is enabled.

Problem Fix/work around

Enable ATmega16U4/ATmega32U4 TWI before the other nodes of the TWI network.

2. High current consumption in sleep mode

If a pending interrupt cannot wake the part up from the selected mode, the current consump-

tion will increase during sleep when executing the SLEEP instruction directly after a SEI

instruction.

Problem Fix/work around

Before entering sleep, interrupts not used to wake up the part from the sleep mode should

be disabled.

3. MSB of OCR4A/B/D is write only in 11-bits enhanced PWM mode

In the 11-bits enhanced PWM mode the MSB of OCR4A/B/D is write only. A read of

OCR4A/B/D will always return zero in the MSB position.

Problem Fix/work around

None.

35.2 ATmega16U4/ATmega32U4 Rev D

• Spike on TWI pins when TWI is enabled

• High current consumption in sleep mode

• Timer 4 11-bits enhanced PWM mode

1. Spike on TWI pins when TWI is enabled

100 ns negative spike occurs on SDA and SCL pins when TWI is enabled.

Problem Fix/work around

Enable ATmega16U4/ATmega32U4 TWI before the other nodes of the TWI network.

2. High current consumption in sleep mode

If a pending interrupt cannot wake the part up from the selected mode, the current consump-

tion will increase during sleep when executing the SLEEP instruction directly after a SEI

instruction.

Problem Fix/work around

Before entering sleep, interrupts not used to wake up the part from the sleep mode should

be disabled.

419

7766E–AVR–04/10

ATmega16U4/ATmega32U4

3. Timer 4 11-bits enhanced PWM mode

Timer 4 11-bits enhanced mode is not functional.

Problem Fix/work around

None.

35.3 ATmega16U4/ATmega32U4 Rev C

Not sampled

35.4 ATmega16U4/ATmega32U4 Rev B

• Spike on TWI pins when TWI is enabled

• High current consumption in sleep mode

• Incorrect execution of VBUSTI interrupt

• Timer 4 11-bits enhanced PWM mode

1. Spike on TWI pins when TWI is enabled

100 ns negative spike occurs on SDA and SCL pins when TWI is enabled.

Problem Fix/work around

Enable ATmega16U4/ATmega32U4 TWI before the other nodes of the TWI network.

2. High current consumption in sleep mode

If a pending interrupt cannot wake the part up from the selected mode, the current consump-

tion will increase during sleep when executing the SLEEP instruction directly after a SEI

instruction.

Problem Fix/work around

Before entering sleep, interrupts not used to wake up the part from the sleep mode should

be disabled.

3. Incorrect execution of VBUSTI interrupt

The CPU may incorrectly execute the interrupt vector related to the VBUSTI interrupt flag.

Problem fix/workaround

Do not enable this interrupt. Firmware must process this USB event by polling VBUSTI.

4. Timer 4 11-bits enhanced PWM mode

Timer 4 11-bits enhanced mode is not functional.

Problem Fix/work around

None.

420

7766E–AVR–04/10

ATmega16U4/ATmega32U4

35.5 ATmega16U4/ATmega32U4 Rev A

• Spike on TWI pins when TWI is enabled

• High current consumption in sleep mode

• Increased power consumption in power-down mode

• Internal RC oscillator start up may fail

• Internal RC oscillator calibration

• Incorrect execution of VBUSTI interrupt

• Timer 4 enhanced mode issue

1. Spike on TWI pins when TWI is enabled

100 ns negative spike occurs on SDA and SCL pins when TWI is enabled.

Problem Fix/work around

Enable ATmega16U4/ATmega32U4 TWI before the other nodes of the TWI network.

2. High current consumption in sleep mode

If a pending interrupt cannot wake the part up from the selected mode, the current consump-

tion will increase during sleep when executing the SLEEP instruction directly after a SEI

instruction.

Problem Fix/work around

Before entering sleep, interrupts not used to wake up the part from the sleep mode should

be disabled.

3. Increased power comsumption in power-down mode

The typical power consumption is increased by about 30 µA in power-down mode.

Problem Fix/work around

None.

4. Internal RC oscillator start up may fail

When the part is configured to start on internal RC oscillator, the oscillator may not start

properly after power-on.

Problem Fix/work around

Do not configure the part to start on internal RC oscillator.

5. Internal RC oscillator calibration

8 MHz frequency can be impossible to reach with internal RC even when using maximal

OSCAL value.

Problem Fix/work around

None.

6. Incorrect execution of VBUSTI interrupt

The CPU may incorrectly execute the interrupt vector related to the VBUSTI interrupt flag.

Problem fix/workaround

Do not enable this interrupt. Firmware must process this USB event by polling VBUSTI.

7. Timer 4 11-bits enhanced PWM mode

Timer 4 11-bits enhanced mode is not functional.

421

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Problem Fix/work around

None.

422

7766E–AVR–04/10

ATmega16U4/ATmega32U4

36. Datasheet Revision History for ATmega16U4/ATmega32U4

Please note that the referring page numbers in this section are referred to this document. The

referring revision in this section are referring to the document revision.

36.1 Rev. 7766E – 04/10

1.

Updated “Features” on page 1.

2.

Updated “Features” on page 253.

3.

Updated Figure 21-9 on page 258.

4.

Updated Section 21.8 on page 260.

5.

Updated “Features” on page 292.

6.

Updated “ATmega16U4/ATmega32U4 Boundary-scan Order” on page 327.

Updated Table 28-5 on page 348.

7.

8.

Updated “Electrical Characteristics” on page 378.

Updated Figure 29-2 on page 381.

9.

10.

11.

12.

Added “Typical Characteristics” on page 386.

Updated “Ordering Information” on page 415.

Updated “Errata” on page 419.

36.2 Rev. 7766D – 01/09

1.

2.

3.

4.

Updated Memory section in “Features” on page 1.

Added section “Resources” on page 8.

Added section “Data Retention” on page 8.

Updated “Ordering Information” on page 415.

36.3 Rev. 7766C – 11/08

1.

Updated Memory section in “Features” on page 1.

36.4 Rev. 7766B – 11/08

1.

2.

3.

Added ATmega16U4 device.

Created errata section and added ATmega16U4.

Updated High Speed Timer, asynchronous description Section 15. on page 139

423

7766E–AVR–04/10

ATmega16U4/ATmega32U4

36.5 Rev. 7766A – 07/08

1.

Initial revision

424

7766E–AVR–04/10

ATmega16U4/ATmega32U4

Table of Contents

1

2

Pin Configurations ................................................................................... 3

Overview ................................................................................................... 3

2.1Block Diagram ...........................................................................................................4

2.2Pin Descriptions ........................................................................................................5

3

4

About ......................................................................................................... 8

3.1Disclaimer ..................................................................................................................8

3.2Resources .................................................................................................................8

3.3Code Examples .........................................................................................................8

3.4Data Retention ..........................................................................................................8

AVR CPU Core .......................................................................................... 9

4.1Introduction ................................................................................................................9

4.2Architectural Overview ..............................................................................................9

4.3ALU – Arithmetic Logic Unit ....................................................................................10

4.4Status Register ........................................................................................................11

4.5General Purpose Register File ................................................................................12

4.6Stack Pointer ...........................................................................................................13

4.7Instruction Execution Timing ...................................................................................14

4.8Reset and Interrupt Handling ..................................................................................15

5

6

AVR ATmega16U4/ATmega32U4 Memories ........................................ 18

5.1In-System Reprogrammable Flash Program Memory .............................................18

5.2SRAM Data Memory ...............................................................................................19

5.3EEPROM Data Memory ..........................................................................................21

5.4I/O Memory ..............................................................................................................26

System Clock and Clock Options ......................................................... 27

6.1Clock Systems and their Distribution .......................................................................27

6.2Clock Sources .........................................................................................................28

6.3Low Power Crystal Oscillator ...................................................................................29

6.4Low Frequency Crystal Oscillator ............................................................................31

6.5Calibrated Internal RC Oscillator .............................................................................32

6.6External Clock .........................................................................................................33

6.7Clock Switch ............................................................................................................34

6.8Clock switch Algorithm ............................................................................................35

6.9Clock Output Buffer .................................................................................................37

i

7766E–AVR–04/10

ATmega16U4/ATmega32U4

6.10PLL ........................................................................................................................39

7

Power Management and Sleep Modes ................................................. 43

7.1Idle Mode .................................................................................................................44

7.2ADC Noise Reduction Mode ...................................................................................44

7.3Power-down Mode ..................................................................................................44

7.4Power-save Mode ...................................................................................................44

7.5Standby Mode .........................................................................................................45

7.6Extended Standby Mode .........................................................................................45

7.7Power Reduction Register .......................................................................................45

7.8Minimizing Power Consumption ..............................................................................47

8

9

System Control and Reset .................................................................... 49

8.1Internal Voltage Reference ......................................................................................54

8.2Watchdog Timer ......................................................................................................55

Interrupts ................................................................................................ 61

9.1Interrupt Vectors in ATmega16U4/ATmega32U4 ....................................................61

10 I/O-Ports .................................................................................................. 65

10.1Introduction ............................................................................................................65

10.2Ports as General Digital I/O ...................................................................................66

10.3Alternate Port Functions ........................................................................................70

10.4Register Description for I/O-Ports ..........................................................................82

11 External Interrupts ................................................................................. 85

12 Timer/Counter0, Timer/Counter1, and Timer/Counter3 Prescalers ... 89

12.1Internal Clock Source ............................................................................................89

12.2Prescaler Reset .....................................................................................................89

12.3External Clock Source ...........................................................................................89

12.4General Timer/Counter Control Register – GTCCR ..............................................90

13 8-bit Timer/Counter0 with PWM ............................................................ 91

13.1Overview ...............................................................................................................91

13.2Timer/Counter Clock Sources ...............................................................................92

13.3Counter Unit ..........................................................................................................92

13.4Output Compare Unit ............................................................................................93

13.5Compare Match Output Unit ..................................................................................95

13.6Modes of Operation ...............................................................................................96

13.7Timer/Counter Timing Diagrams .........................................................................100

ii

7766E–AVR–04/10

ATmega16U4/ATmega32U4

13.88-bit Timer/Counter Register Description ............................................................102

14 16-bit Timers/Counters (Timer/Counter1 and Timer/Counter3) ....... 108

14.1Overview .............................................................................................................108

14.2Accessing 16-bit Registers ..................................................................................110

14.3Timer/Counter Clock Sources .............................................................................113

14.4Counter Unit ........................................................................................................114

14.5Input Capture Unit ...............................................................................................115

14.6Output Compare Units .........................................................................................117

14.7Compare Match Output Unit ................................................................................119

14.8Modes of Operation .............................................................................................120

14.9Timer/Counter Timing Diagrams .........................................................................127

14.1016-bit Timer/Counter Register Description ........................................................129

15 10-bit High Speed Timer/Counter4 ..................................................... 138

15.1Features ..............................................................................................................138

15.2Overview .............................................................................................................138

15.3Counter Unit ........................................................................................................142

15.4Output Compare Unit ..........................................................................................143

15.5Dead Time Generator ..........................................................................................145

15.6Compare Match Output Unit ................................................................................146

15.7Synchronous update ...........................................................................................149

15.8Modes of Operation .............................................................................................149

15.9Timer/Counter Timing Diagrams .........................................................................156

15.10Fault Protection Unit ..........................................................................................157

15.11Accessing 10-Bit Registers ...............................................................................159

15.12Register Description ..........................................................................................162

16 Output Compare Modulator (OCM1C0A) ........................................... 175

16.1Overview .............................................................................................................175

16.2Description ..........................................................................................................175

17 Serial Peripheral Interface – SPI ......................................................... 177

17.1SS Pin Functionality ............................................................................................181

17.2Data Modes .........................................................................................................184

18 USART ................................................................................................... 186

18.1Overview .............................................................................................................186

18.2Clock Generation .................................................................................................187

iii

7766E–AVR–04/10

ATmega16U4/ATmega32U4

18.3Frame Formats ....................................................................................................190

18.4USART Initialization ............................................................................................192

18.5Data Transmission – The USART Transmitter ....................................................193

18.6Data Reception – The USART Receiver .............................................................195

18.7Asynchronous Data Reception ............................................................................199

18.8Multi-processor Communication Mode ................................................................202

18.9Hardware Flow Control ........................................................................................203

18.10USART Register Description .............................................................................205

18.11Examples of Baud Rate Setting ........................................................................209

19 USART in SPI Mode ............................................................................. 214

19.1Overview .............................................................................................................214

19.2Clock Generation .................................................................................................214

19.3SPI Data Modes and Timing ...............................................................................215

19.4Frame Formats ....................................................................................................215

19.5Data Transfer ......................................................................................................217

19.6USART MSPIM Register Description ..................................................................219

19.7AVR USART MSPIM vs. AVR SPI ......................................................................221

20 2-wire Serial Interface .......................................................................... 223

20.1Features ..............................................................................................................223

20.22-wire Serial Interface Bus Definition ..................................................................223

20.3Data Transfer and Frame Format ........................................................................224

20.4Multi-master Bus Systems, Arbitration and Synchronization ...............................227

20.5Overview of the TWI Module ...............................................................................228

20.6TWI Register Description ....................................................................................231

20.7Using the TWI ......................................................................................................234

20.8Transmission Modes ...........................................................................................238

20.9Multi-master Systems and Arbitration ..................................................................251

21 USB controller ...................................................................................... 253

21.1Features ..............................................................................................................253

21.2Block Diagram .....................................................................................................253

21.3Typical Application Implementation .....................................................................254

21.4Crystal-less operation ..........................................................................................256

21.5Design guidelines ................................................................................................256

21.6General Operating Modes ...................................................................................257

21.7Power modes ......................................................................................................259

iv

7766E–AVR–04/10

ATmega16U4/ATmega32U4

21.8Speed Control .....................................................................................................260

21.9Memory management .........................................................................................260

21.10PAD suspend ....................................................................................................261

21.11Plug-in detection ................................................................................................262

21.12Registers description .........................................................................................263

21.13USB Software Operating modes .......................................................................265

22 USB Device Operating modes ............................................................ 266

22.1Introduction ..........................................................................................................266

22.2Power-on and reset .............................................................................................266

22.3Endpoint reset .....................................................................................................266

22.4USB reset ............................................................................................................267

22.5Endpoint selection ...............................................................................................267

22.6Endpoint activation ..............................................................................................267

22.7Address Setup .....................................................................................................268

22.8Suspend, Wake-up and Resume ........................................................................269

22.9Detach .................................................................................................................269

22.10Remote Wake-up ..............................................................................................270

22.11STALL request ..................................................................................................270

22.12CONTROL endpoint management ....................................................................271

22.13OUT endpoint management ..............................................................................272

22.14IN endpoint management ..................................................................................274

22.15Isochronous mode .............................................................................................275

22.16Overflow ............................................................................................................276

22.17Interrupts ...........................................................................................................276

22.18Registers ...........................................................................................................277

23 Analog Comparator ............................................................................. 289

23.1Analog Comparator Multiplexed Input .................................................................291

24 Analog to Digital Converter - ADC ..................................................... 292

24.1Features ..............................................................................................................292

24.2Operation .............................................................................................................294

24.3Starting a Conversion ..........................................................................................294

24.4Prescaling and Conversion Timing ......................................................................295

24.5Changing Channel or Reference Selection .........................................................298

24.6Temperature Sensor ...........................................................................................299

24.7ADC Noise Canceler ...........................................................................................301

v

7766E–AVR–04/10

ATmega16U4/ATmega32U4

24.8ADC Conversion Result ......................................................................................305

24.9ADC Register Description ...................................................................................307

25 JTAG Interface and On-chip Debug System ..................................... 314

25.1Overview .............................................................................................................314

25.2Test Access Port – TAP ......................................................................................314

25.3TAP Controller .....................................................................................................316

25.4Using the Boundary-scan Chain ..........................................................................317

25.5Using the On-chip Debug System .......................................................................317

25.6On-chip Debug Specific JTAG Instructions .........................................................318

25.7On-chip Debug Related Register in I/O Memory .................................................319

25.8Using the JTAG Programming Capabilities .........................................................319

25.9Bibliography .........................................................................................................319

26 IEEE 1149.1 (JTAG) Boundary-scan ................................................... 320

26.1Features ..............................................................................................................320

26.2System Overview ................................................................................................320

26.3Data Registers .....................................................................................................320

26.4Boundary-scan Specific JTAG Instructions .........................................................322

26.5Boundary-scan Related Register in I/O Memory .................................................323

26.6Boundary-scan Chain ..........................................................................................324

26.7ATmega16U4/ATmega32U4 Boundary-scan Order ............................................327

26.8Boundary-scan Description Language Files ........................................................329

27 Boot Loader Support – Read-While-Write Self-Programming ......... 330

27.1Boot Loader Features ..........................................................................................330

27.2Application and Boot Loader Flash Sections .......................................................330

27.3Read-While-Write and No Read-While-Write Flash Sections ..............................330

27.4Boot Loader Lock Bits .........................................................................................333

27.5Entering the Boot Loader Program ......................................................................334

27.6Addressing the Flash During Self-Programming .................................................337

27.7Self-Programming the Flash ................................................................................338

28 Memory Programming ......................................................................... 346

28.1Program And Data Memory Lock Bits .................................................................346

28.2Fuse Bits .............................................................................................................347

28.3Signature Bytes ...................................................................................................349

28.4Calibration Byte ...................................................................................................349

28.5Parallel Programming Parameters, Pin Mapping, and Commands .....................349

vi

7766E–AVR–04/10

ATmega16U4/ATmega32U4

28.6Parallel Programming ..........................................................................................352

28.7Serial Downloading .............................................................................................360

28.8Serial Programming Pin Mapping ........................................................................361

28.9Programming via the JTAG Interface ..................................................................365

29 Electrical Characteristics .................................................................... 378

29.1Absolute Maximum Ratings* ...............................................................................378

29.2DC Characteristics ..............................................................................................378

29.3External Clock Drive Waveforms .........................................................................380

29.4External Clock Drive ............................................................................................380

29.5Maximum speed vs. V_{CC .............................................................................................................................380}

29.62-wire Serial Interface Characteristics .................................................................381

29.7SPI Timing Characteristics ..................................................................................383

29.8Hardware Boot EntranceTiming Characteristics ..................................................384

30 Typical Characteristics ........................................................................ 386

30.1Active Supply Current ..........................................................................................386

30.2Idle Supply Current ..............................................................................................389

30.3Power-down Supply Current ...............................................................................391

30.4Power-save Supply Current ................................................................................392

30.5Pin Pull-Up ..........................................................................................................393

30.6Pin Driver Strength ..............................................................................................394

30.7Pin Threshold and Hysteresis .............................................................................397

30.8BOD Threshold ....................................................................................................400

30.9Internal Oscilllator Speed ....................................................................................402

30.10Current Consumption of Peripheral Units ..........................................................405

30.11Current Consumption in Reset and Reset Pulsewidth ......................................407

31 Register Summary ............................................................................... 408

32 Instruction Set Summary .................................................................... 412

33 Ordering Information ........................................................................... 415

33.1ATmega16U4 ......................................................................................................415

33.2ATmega32U4 ......................................................................................................416

34 Packaging Information ........................................................................ 417

34.1TQFP44 ...............................................................................................................417

34.2QFN44 .................................................................................................................418

35 Errata ..................................................................................................... 419

vii

7766E–AVR–04/10

35.1ATmega16U4/ATmega32U4 Rev E ....................................................................419

35.2ATmega16U4/ATmega32U4 Rev D ....................................................................419

35.3ATmega16U4/ATmega32U4 Rev C ....................................................................420

35.4ATmega16U4/ATmega32U4 Rev B ....................................................................420

35.5ATmega16U4/ATmega32U4 Rev A ....................................................................421

36 Datasheet Revision History for ATmega16U4/ATmega32U4 ........... 423

36.1Rev. 7766E – 04/10 .............................................................................................423

36.2Rev. 7766D – 01/09 ............................................................................................423

36.3Rev. 7766C – 11/08 ............................................................................................423

36.4Rev. 7766B – 11/08 .............................................................................................423

36.5Rev. 7766A – 07/08 .............................................................................................424

Headquarters

International

Atmel Corporation

2325 Orchard Parkway

San Jose, CA 95131

USA

Tel: 1(408) 441-0311

Fax: 1(408) 487-2600

Atmel Asia

Atmel Europe

Le Krebs

Atmel Japan

9F, Tonetsu Shinkawa Bldg.

1-24-8 Shinkawa

Chuo-ku, Tokyo 104-0033

Japan

Tel: (81) 3-3523-3551

Fax: (81) 3-3523-7581

Unit 1-5 & 16, 19/F

BEA Tower, Millennium City 5

418 Kwun Tong Road

Kwun Tong, Kowloon

Hong Kong

8, Rue Jean-Pierre Timbaud

BP 309

78054 Saint-Quentin-en-

Yvelines Cedex

France

Tel: (852) 2245-6100

Fax: (852) 2722-1369

Tel: (33) 1-30-60-70-00

Fax: (33) 1-30-60-71-11

Product Contact

Web Site

Technical Support

Sales Contact

www.atmel.com

avr@atmel.com

www.atmel.com/contacts

Literature Requests

www.atmel.com/literature

Disclaimer: The information in this document is provided in connection with Atmel products. No license, express or implied, by estoppel or otherwise, to any

intellectual property right is granted by this document or in connection with the sale of Atmel products. EXCEPT AS SET FORTH IN ATMEL’S TERMS AND CONDI-

TIONS OF SALE LOCATED ON ATMEL’S WEB SITE, ATMEL ASSUMES NO LIABILITY WHATSOEVER AND DISCLAIMS ANY EXPRESS, IMPLIED OR STATUTORY

WARRANTY RELATING TO ITS PRODUCTS INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR

PURPOSE, OR NON-INFRINGEMENT. IN NO EVENT SHALL ATMEL BE LIABLE FOR ANY DIRECT, INDIRECT, CONSEQUENTIAL, PUNITIVE, SPECIAL OR INCIDEN-

TAL DAMAGES (INCLUDING, WITHOUT LIMITATION, DAMAGES FOR LOSS OF PROFITS, BUSINESS INTERRUPTION, OR LOSS OF INFORMATION) ARISING OUT OF

THE USE OR INABILITY TO USE THIS DOCUMENT, EVEN IF ATMEL HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. Atmel makes no

representations or warranties with respect to the accuracy or completeness of the contents of this document and reserves the right to make changes to specifications

and product descriptions at any time without notice. Atmel does not make any commitment to update the information contained herein. Unless specifically provided

otherwise, Atmel products are not suitable for, and shall not be used in, automotive applications. Atmel’s products are not intended, authorized, or warranted for use

as components in applications intended to support or sustain life.

marks or trademarks of Atmel Corporation or its subsidiaries. Other terms and product names may be trademarks of others.

7766E–AVR–04/10


型号：	ATMEGA16U4PRE
厂家：	ATMEL
描述：	Optional Boot Code Section with Independent Lock Bits 可选Boot代码区具有独立锁定位
文件：	总433页 (文件大小：8915K)
中文：	中文翻译
下载：	下载PDF数据表文档文件

ATMEGA16U4PRE [ATMEL]

相关型号：

ATMEGA16U4RC-AU

ATMEGA16U4RC-MU

ATMEGA16U4_08

ATMEGA16U4_09

ATMEGA16U4_1

ATMEGA16U4_10

ATMEGA16U4_14

ATMEGA16_06

ATMEGA16_09

ATMEGA16_091

ATMEGA16_14

ATMEGA256