EM351_12 [SILICON]
High-Performance, Integrated ZigBee/802.15.4 System-on-Chip; 高性能，集成的ZigBee / 802.15.4系统级芯片
| 型号： | EM351_12 |
| 厂家： | 
SILICON |
| 描述： | High-Performance, Integrated ZigBee/802.15.4 System-on-Chip |
| 文件： | 总244页 (文件大小：3868K) |
| 中文： | 中文翻译 | 下载： | 下载PDF数据表文档文件 |
EM351 / EM357
High-Performance, Integrated ZigBee/802.15.4 System-on-Chip
Complete System-on-Chip
Exceptional RF Performance
.
.
32-bit ARM® Cortex-M3 processor
Normal mode link budget up to 103 dB;
•
•
configurable up to 110 dB
2.4 GHz IEEE 802.15.4-2003 transceiver & lower
MAC
•
-100 dBm normal RX sensitivity;
configurable to -102 dBm
(1% PER, 20 byte packet)
•
128 or 192 kB flash, with optional read
protection
•
+3 dB normal mode output power;
configurable up to +8 dBm
•
•
12 kB RAM memory
•
•
•
AES128 encryption accelerator
Robust Wi-Fi and Bluetooth coexistence
Flexible ADC, UART/SPI/TWI serial
communications, and general purpose timers
Innovative network and processor debug
.
.
Packet Trace Port for non-intrusive
packet trace with Ember development
tools
•
24 highly configurable GPIOs with Schmitt
trigger inputs
Industry-leading ARM® Cortex-M3 processor
•
.
.
Serial Wire/JTAG interface
•
•
Leading 32-bit processing performance
Highly efficient Thumb-2 instruction set
Operation at 6, 12, or 24 MHz
•
•
•
•
Standard ARM debug capabilities: Flash
Patch & Breakpoint; Data Watchpoint &
Trace; Instrumentation Trace Macrocell
Flexible Nested Vectored Interrupt Controller
Application Flexibility
Low power consumption, advanced management
Single voltage operation: 2.1-3.6 V
with internal 1.8 V and 1.25 V regulators
•
•
•
Rx Current (w/ CPU): 26 mA
•
•
•
Optional 32.768 kHz crystal for higher
timer accuracy
Tx Current (w/ CPU, +3 dBm TX): 31 mA
Low deep sleep current, with retained RAM and
GPIO: 400 nA without/800 nA with sleep timer
Low external component count with
single 24 MHz crystal
Low-frequency internal RC oscillator for low-
power sleep timing
•
•
Support for external power amplifier
Small 7x7 mm 48-pin QFN package
•
•
High-frequency internal RC oscillator for fast
(110 µsec) processor start-up from sleep
TX_ACTIVE
PA select
Data
RAM
12 kB
Program
RF_TX_ALT_P,N
Flash
128/192 kB
SYNTH
PA
DAC
MAC
PA
2 level
Interrupt
+
RF_P,N
nd
®
TM
Baseband
ARM Cortex -M3
LNA
ADC
IF
CPU with NVIC
and MPU
controller
Packet Trace
Bias
OSCA
OSCB
CPU debug
HF crystal
OSC
Encryption
acclerator
TPIU/ITM/
FPB/DWT
Calibration
ADC
Internal HF
RC-OSC
General
purpose
VDD_CORE
VREG_OUT
Always
Powered
1.25V
Regulator
timers
SWCLK,
JTCK
Serial
GPIO
Domain
Wire and
1.8V
Regulator
JTAG
debug
registers
Watchdog
General
nRESET
Purpose
ADC
POR
UART/
SPI/TWI
Silicon Laboratories Inc.
400 West Cesar Chavez
Austin, TX 78701
Chip
manager
Sleep
timer
LF crystal
OSC
Internal LF
RC-OSC
GPIO multiplexor switch
Tel: 1+(512) 416-8500
Fax: 1+(512) 416-9669
Toll Free: 1+(877) 444-3032
www.silabs.com
PA[7:0], PB[7:0], PC[7:0]
120-035X-000 Rev. 1.2
Final
EM351 / EM357
General Description
The EM351 and EM357 are fully integrated System-on-Chips that integrate a 2.4 GHz, IEEE 802.15.4-2003-
compliant transceiver, 32-bit ARM® Cortex-M3 microprocessor, flash and RAM memory, and peripherals of
use to designers of ZigBee-based systems.
The transceiver uses an efficient architecture that exceeds the dynamic range requirements imposed by the
IEEE 802.15.4-2003 standard by over 15 dB. The integrated receive channel filtering allows for robust co-
existence with other communication standards in the 2.4 GHz spectrum, such as IEEE 802.11-2007 and
Bluetooth. The integrated regulator, VCO, loop filter, and power amplifier keep the external component count
low. An optional high performance radio mode (boost mode) is software-selectable to boost dynamic range.
The integrated 32-bit ARM® Cortex-M3 microprocessor is highly optimized for high performance, low power
consumption, and efficient memory utilization. Including an integrated MPU, it supports two different modes
of operation—privileged mode and user mode. This architecture could allow for separation of the networking
stack from the application code, and prevents unwanted modification of restricted areas of memory and
registers resulting in increased stability and reliability of deployed solutions.
The EM351 has 128 kB of embedded flash memory and the EM357 has 192 kB of embedded flash memory. Both
chips have 12 kB of integrated RAM for data and program storage. The Ember software for the EM35x employs
an effective wear-leveling algorithm that optimizes the lifetime of the embedded flash.
To maintain the strict timing requirements imposed by the ZigBee and IEEE 802.15.4-2003 standards, the
EM35x integrates a number of MAC functions, AES128 encryption accelerator, and automatic CRC handling into
the hardware. The MAC hardware handles automatic ACK transmission and reception, automatic backoff
delay, and clear channel assessment for transmission, as well as automatic filtering of received packets. The
Ember Packet Trace Interface is also integrated with the MAC, allowing complete, non-intrusive capture of all
packets to and from the EM35x with Ember development tools.
The EM35x offers a number of advanced power management features that enable long battery life. A high-
frequency internal RC oscillator allows the processor core to begin code execution quickly upon waking.
Various deep sleep modes are available with less than 1 µA power consumption while retaining RAM contents.
To support user-defined applications, on-chip peripherals include UART, SPI, TWI, ADC, and general-purpose
timers, as well as up to 24 GPIOs. Additionally, an integrated voltage regulator, power-on-reset circuit, and
sleep timer are available.
Finally, the EM35x utilizes standard Serial Wire and JTAG interfaces for powerful software debugging and
programming of the ARM Cortex-M3 core. The EM35x integrates the standard ARM system debug components:
Flash Patch and Breakpoint (FPB), Data Watchpoint and Trace (DWT), and Instrumentation Trace Macrocell
(ITM).
Target applications for the EM35x include the following:
Smart Energy
•
•
•
•
•
Building automation and control
Home automation and control
Security and monitoring
General ZigBee wireless sensor networking
This technical data sheet details the EM35x features available to customers using it with Ember software.
Final
120-035X-000 Rev. 1.2
EM351 / EM357
Contents
6.2.3 Reset Generation Module 6-5
1
2
Pin Assignments
1-1
2-1
2-1
6.3 Clocks
6-6
6-8
6-8
6-9
6-9
Electrical Characteristics
2.1 Absolute Maximum Ratings
6.3.1 High-Frequency Internal RC
Oscillator (OSCHF)
6.3.2 High-Frequency Crystal
Oscillator (OSC24M)
6.3.3 Low-Frequency Internal RC
Oscillator (OSCRC)
6.3.4 Low-Frequency Crystal
Oscillator (OSC32K)
2.2 Recommended Operating
Conditions
2-1
2-2
2-2
2-6
2.3 Environmental Characteristics
2.4 DC Electrical Characteristics
2.5 Digital I/O Specifications
6.3.5 Clock Switching
6.4 System Timers
6-10
2.6 Non-RF System Electrical
Characteristics
6-11
6-11
6-11
6-11
2-8
6.4.1 Watchdog Timer
6.4.2 Sleep Timer
6.4.3 Event Timer
2.7 RF Electrical Characteristics
2.7.1 Receive
2-8
2-8
2-10
2-12
2.7.2 Transmit
2.7.3 Synthesizer
6.5 Power Management
6.5.1 Wake Sources
6-11
6-12
6-13
3
4
Top-Level Functional Description 3-1
6.5.2 Basic Sleep Modes
6.5.3 Further options for deep
sleep
Radio Module
4-1
6-14
4.1 Receive (Rx) Path
4.1.1 Rx Baseband
4.1.2 RSSI and CCA
4-1
4-1
4-1
6.5.4 Use of debugger with sleep
modes
6-14
6.5.5 Registers
6-15
6-15
6.6 Security Accelerator
4.2 Transmit (Tx) Path
4.2.1 Tx Baseband
4-1
4-1
7
GPIO (General Purpose Input /
Output)
4.2.2 TX_ACTIVE and
nTX_ACTIVE Signals
7-1
4-1
4-2
4-2
4-2
4-2
7.1 GPIO Ports
7-1
7-2
7-3
7-4
7-4
4.3 Calibration
7.2 Configuration
7.3 Forced Functions
7.4 Reset
4.4 Integrated MAC Module
4.5 Packet Trace Interface (PTI)
4.6 Random Number Generator
7.5 Boot Configuration
5
ARM® Cortex™-M3 and Memory
Modules
5.1 ARM® Cortex™-M3 Microprocessor 5-1
7.6 GPIO Modes
7-5
7-5
7-5
7-6
7-6
5-1
7.6.1 Analog Mode
7.6.2 Input Mode
7.6.3 Output Mode
7.6.4 Alternate Output Mode
5.2 Embedded Memory
5.2.1 Flash Memory
5.2.2 RAM
5-2
5-4
5-7
5-8
7.7 Wake Monitoring
7-6
7-7
7-8
5.2.3 Registers
7.8 External Interrupts
7.9 Debug Control and Status
5.3 Memory Protection Unit
5-8
6-1
6-2
6
System Modules
7.10 GPIO Signal Assignment Summary 7-8
7.11 Registers
7-10
8-1
8-1
6.1 Power domains
6.1.1 Internally regulated
power
6.1.2 Externally regulated
power
8
Serial Controllers
8.1 Overview
6-2
6-2
8.2 Configuration
8.2.1 Registers
8-2
8-3
8-6
8-6
6.2 Resets
6-3
6-3
6-5
6.2.1 Reset Sources
6.2.2 Reset Recording
8.3 SPI - Master Mode
8.3.1 GPIO Usage
Final
120-035X-000 Rev. 1.2
EM351 / EM357
8.3.2 Set Up and Configuration 8-7
9.5 Registers
9-31
8.3.3 Operation
8.3.4 Interrupts
8.3.5 Registers
8-8
8-9
8-10
10 ADC (Analog to Digital Converter) 10-1
10.1 Setup and Configuration
10.1.1 GPIO Usage
10-1
10-2
10-2
10-2
10-3
8.4 SPI - Slave Mode
8.4.1 GPIO Usage
8-14
8-14
10.1.2 Voltage Reference
10.1.3 Offset/Gain Correction
10.1.4 DMA
8.4.2 Set Up and Configuration 8-15
8.4.3 Operation
8.4.4 DMA
8.4.5 Interrupts
8.4.6 Registers
8-16
8-17
8-17
8-18
10.1.5 ADC Configuration
Register
10-3
10-5
10-6
10-7
10-8
10-13
10.2 Interrupts
10.3 Operation
8.5 TWI - Two Wire serial Interfaces 8-18
8.5.1 GPIO Usage 8-18
8.5.2 Set Up and Configuration 8-18
10.4 Calibration
10.5 ADC Key Parameters
10.6 Registers
8.5.3 Constructing Frames
8.5.4 Interrupts
8.5.5 Registers
8-19
8-21
8-22
11 Interrupt System
11-1
11.1 Nested Vectored Interrupt
Controller (NVIC)
8.6 UART - Universal Asynchronous
Receiver / Transmitter
11-1
11-3
11-6
11-6
11-7
8-24
8-24
11.2 Event Manager
11.3 Non-maskable Interrupt (NMI)
11.4 Faults
8.6.1 GPIO Usage
8.6.2 Set Up and Configuration 8-25
8.6.3 FIFOs
8.6.4 RTS/CTS Flow control
8.6.5 DMA
8-27
8-27
8-28
8-28
8-29
11.5 Registers
12 Trace Port Interface Unit (TPIU) 12-1
13 Instrumentation Trace
8.6.6 Interrupts
8.6.7 Registers
8.7 DMA Channels
8-32
8-34
Macrocell (ITM)
13-1
8.7.1 Registers
14 Data Watchpoint and
Trace (DWT)
9
General Purpose Timers (TIM1 and
TIM2)
14-1
9-1
15 Flash Patch and Breakpoint (FPB) 15-1
9.1 Introduction
9.2 GPIO Usage
9-1
9-3
16 Integrated Voltage Regulator
16-1
17 Serial Wire and JTAG
(SWJ) Interface
9.3 Timer Functional Description
9.3.1 Time-Base Unit
9-3
9-3
9-4
9-9
17-1
18-1
19-1
9.3.2 Counter Modes
9.3.3 Clock Selection
9.3.4 Capture/Compare
Channels
9.3.5 Input Capture Mode
9.3.6 PWM Input Mode
9.3.7 Forced Output Mode
9.3.8 Output Compare Mode
9.3.9 PWM Mode
9.3.10 One-Pulse Mode
9.3.11 Encoder Interface Mode 9-20
9.3.12 Timer Input XOR Function 9-22
18 Typical Application
19 Mechanical Details
9-12
9-13
9-14
9-15
9-15
9-16
9-19
19.1 QFN48 Footprint
Recommendations
19-1
19-3
19.2 Solder Temperature Profile
20 Part Marking
20-1
21-1
22-1
23-1
A-1
21 Ordering Information
22 Shipping Box Label
23 Revision History
9.3.13 Timers and External
Appendix A Register Address Table
Trigger Synchronization 9-22
9.3.14 Timer Synchronization
9.3.15 Timer Signal Descriptions 9-29
9.4 Interrupts 9-30
9-25
Appendix B Abbreviations and
Acronyms
B-1
C-1
Appendix C References
Final
120-035X-000 Rev. 1.2
EM351 / EM357
1
Pin Assignments
Figure 1-1. EM35x Pin Assignments
48 47 46 45 44 43 42 41 40 39 38 37
VDD_24MHZ
1
2
36
35
34
33
32
31
30
29
28
27
26
25
PB0, VREF, IRQA, TRACECLK, TIM1CLK, TIM2MSK
PC4, JTMS, SWDIO
49
GND
VDD_VCO
RF_P
3
PC3, JTDI
RF_N
4
PC2, JTDO, SWO
VDD_RF
5
SWCLK, JTCK
RF_TX_ALT_P
RF_TX_ALT_N
VDD_IF
6
PB2, SC1MISO, SC1MOSI, SC1SCL, SC1RXD, TIM2C2
PB1, SC1MISO, SC1MOSI, SC1SDA, SC1TXD, TIM2C1
PA6, TIM1C3
EM35x
7
8
NC
9
VDD_PADS
VDD_PADSA
PC5, TX_ACTIVE
nRESET
10
11
12
PA5, ADC5, PTI_DATA, nBOOTMODE, TRACEDATA3
PA4, ADC4, PTI_EN, TRACEDATA2
PA3, SC2nSSEL, TRACECLK, TIM2C2
13 14 15 16 17 18 19 20 21 22 23 24
Refer to Chapter 7, GPIO for details about selecting GPIO pin functions.
1-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 1-1. EM35x Pin Descriptions
Pin # Signal
Direction
Power
Power
I/O
Description
1
VDD_24MHZ
1.8 V high-frequency oscillator supply
1.8 V VCO supply
2
VDD_VCO
RF_P
3
Differential (with RF_N) receiver input/transmitter output
Differential (with RF_P) receiver input/transmitter output
1.8 V RF supply (LNA and PA)
4
RF_N
I/O
5
VDD_RF
RF_TX_ALT_P
RF_TX_ALT_N
VDD_IF
Power
O
6
Differential (with RF_TX_ALT_N) transmitter output (optional)
Differential (with RF_TX_ALT_P) transmitter output (optional)
1.8 V IF supply (mixers and filters)
7
O
8
Power
9
NC
Do not connect
10
11
VDD_PADSA
PC5
Power
I/O
Analog pad supply (1.8 V)
Digital I/O
TX_ACTIVE
O
Logic-level control for external Rx/Tx switch. The EM35x baseband
controls TX_ACTIVE and drives it high (VDD_PADS) when in Tx mode.
Select alternate output function with GPIO_PCCFGH[7:4]
12
13
nRESET
PC6
I
Active low chip reset (internal pull-up)
Digital I/O
I/O
I/O
OSC32B
32.768 kHz crystal oscillator
Select analog function with GPIO_PCCFGH[11:8]
nTX_ACTIVE
O
Inverted TX_ACTIVE signal (see PC5)
Select alternate output function with GPIO_PCCFGH[11:8]
14
PC7
I/O
I/O
Digital I/O
OSC32A
32.768 kHz crystal oscillator
Select analog function with GPIO_PCCFGH[15:12]
OSC32_EXT
VREG_OUT
VDD_PADS
VDD_CORE
PA7
I
Digital 32.768 kHz clock input source
Regulator output (1.8 V while awake, 0 V during deep sleep)
Pads supply (2.1-3.6 V)
15
16
17
18
Power
Power
Power
1.25 V digital core supply decoupling
I/O
Digital I/O
High
Disable REG_EN with GPIO_DBGCFG[4]
current
TIM1C4
O
Timer 1 Channel 4 output
Enable timer output with TIM1_CCER
Select alternate output function with GPIO_PACFGH[15:12]
Disable REG_EN with GPIO_DBGCFG[4]
TIM1C4
REG_EN
PB3
I
Timer 1 Channel 4 input
Cannot be remapped
O
External regulator open drain output
Enabled after reset
19
I/O
Digital I/O
1-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Pin # Signal
Direction
Description
TIM2C3
O
Timer 2 channel 3 output
Enable remap with TIM2_OR[6]
Enable timer output in TIM2_CCER
(see also Pin 22)
Select alternate output function with GPIO_PBCFGL[15:12]
TIM2C3
I
I
Timer 2 channel 3 input
Enable remap with TIM2_OR[6]
(see also Pin 22)
SC1nCTS
UART CTS handshake of Serial Controller 1
Enable with SC1_UARTCFG[5]
Select UART with SC1_MODE
SC1SCLK
O
SPI master clock of Serial Controller 1
Either disable timer output in TIM2_CCER,
or disable remap with TIM2_OR[6]
Enable master with SC1_SPICFG[4]
Select SPI with SC1_MODE
Select alternate output function with GPIO_PBCFGL[15:12]
SC1SCLK
PB4
I
SPI slave clock of Serial Controller 1
Enable slave with SC1_SPICFG[4]
Select SPI with SC1_MODE
20
I/O
O
Digital I/O
TIM2C4
Timer 2 channel 4 output
Enable remap with TIM2_OR[7]
(see also Pin 24)
Enable timer output in TIM2_CCER
Select alternate output function with GPIO_PBCFGH[3:0]
TIM2C4
I
Timer 2 channel 4 input
Enable remap with TIM2_OR[7]
(see also Pin 24)
SC1nRTS
O
UART RTS handshake of Serial Controller 1
Either disable timer output in TIM2_CCER,
or disable remap with TIM2_OR[7]
Enable with SC1_UARTCFG[5]
Select UART with SC1_MODE
Select alternate output function with GPIO_PBCFGH[3:0]
SC1nSSEL
PA0
I
SPI slave select of Serial Controller 1
Enable slave with SC1_SPICFG[4]
Select SPI with SC1_MODE
21
I/O
O
Digital I/O
TIM2C1
Timer 2 channel 1 output
Disable remap with TIM2_OR[4]
(see also Pin 30)
Enable timer output in TIM2_CCER
Select alternate output function with GPIO_PACFGL[3:0]
TIM2C1
I
Timer 2 channel 1 input
Disable remap with TIM2_OR[4]
(see also Pin 30)
1-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Pin # Signal
Direction
Description
SC2MOSI
O
SPI master data out of Serial Controller 2
Either disable timer output in TIM2_CCER,
or enable remap with TIM2_OR[4]
Enable master with SC2_SPICFG[4]
Select SPI with SC2_MODE
Select alternate output function with GPIO_PACFGL[3:0]
SC2MOSI
I
SPI slave data in of Serial Controller 2
Enable slave with SC2_SPICFG[4]
Select SPI with SC2_MODE
22
PA1
I/O
O
Digital I/O
TIM2C3
Timer 2 channel 3 output
Disable remap with TIM2_OR[6]
(see also Pin 19)
Enable timer output in TIM2_CCER
Select alternate output function with GPIO_PACFGL[7:4]
TIM2C3
I
Timer 2 channel 3 input
Disable remap with TIM2_OR[6]
(see also Pin 19)
SC2SDA
I/O
TWI data of Serial Controller 2
Either disable timer output in TIM2_CCER,
or enable remap with TIM2_OR[6]
Select TWI with SC2_MODE
Select alternate open-drain output function with GPIO_PACFGL[7:4]
SC2MISO
O
SPI slave data out of Serial Controller 2
Either disable timer output in TIM2_CCER,
or enable remap with TIM2_OR[6]
Enable slave with SC2_SPICFG[4]
Select SPI with SC2_MODE
Select alternate output function with GPIO_PACFGL[7:4]
SC2MISO
I
SPI master data in of Serial Controller 2
Enable slave with SC2_SPICFG[4]
Select SPI with SC2_MODE
23
24
VDD_PADS
PA2
Power
I/O
Pads supply (2.1-3.6 V)
Digital I/O
TIM2C4
O
Timer 2 channel 4 output
Disable remap with TIM2_OR[7]
(see also Pin 20)
Enable timer output in TIM2_CCER
Select alternate output function with GPIO_PACFGL[11:8]
TIM2C4
I
Timer 2 channel 4 input
Disable remap with TIM2_OR[7]
(see also Pin 20)
SC2SCL
I/O
TWI clock of Serial Controller 2
Either disable timer output in TIM2_CCER,
or enable remap with TIM2_OR[7]
Select TWI with SC2_MODE
Select alternate open-drain output function with GPIO_PACFGL[11:8]
1-4
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Pin # Signal
Direction
Description
SC2SCLK
O
SPI master clock of Serial Controller 2
Either disable timer output in TIM2_CCER,
or enable remap with TIM2_OR[7]
Enable master with SC2_SPICFG[4]
Select SPI with SC2_MODE
Select alternate output function with GPIO_PACFGL[11:8]
SC2SCLK
I
SPI slave clock of Serial Controller 2
Enable slave with SC2_SPICFG[4]
Select SPI with SC2_MODE
25
PA3
I/O
I
Digital I/O
SC2nSSEL
SPI slave select of Serial Controller 2
Enable slave with SC2_SPICFG[4]
Select SPI with SC2_MODE
TRACECLK
O
Synchronous CPU trace clock
Either disable timer output in TIM2_CCER,
or enable remap with TIM2_OR[5]
(see also Pin 36)
Enable trace interface in ARM core
Select alternate output function with GPIO_PACFGL[15:12]
TIM2C2
O
I
Timer 2 channel 2 output
Disable remap with TIM2_OR[5]
Enable timer output in TIM2_CCER
Select alternate output function with GPIO_PACFGL[15:12]
(see also Pin 31)
TIM2C2
Timer 2 channel 2 input
Disable remap with TIM2_OR[5]
(see also Pin 31)
26
PA4
I/O
Digital I/O
ADC4
Analog
ADC Input 4
Select analog function with GPIO_PACFGH[3:0]
PTI_EN
O
O
Frame signal of Packet Trace Interface (PTI)
Disable trace interface in ARM core
Enable PTI in Ember software
Select alternate output function with GPIO_PACFGH[3:0]
TRACEDATA2
Synchronous CPU trace data bit 2
Select 4-wire synchronous trace interface in ARM core
Enable trace interface in ARM core
Select alternate output function with GPIO_PACFGH[3:0]
27
PA5
I/O
Digital I/O
ADC5
Analog
ADC Input 5
Select analog function with GPIO_PACFGH[7:4]
PTI_DATA
O
I
Data signal of Packet Trace Interface (PTI)
Disable trace interface in ARM core
Enable PTI in Ember software
Select alternate output function with GPIO_PACFGH[7:4]
nBOOTMODE
Activate FIB monitor instead of main program or bootloader when
coming out of reset.
Signal is active during and immediately after a reset on nRESET. See
Section 7.5, Boot Configuration, in Chapter 7, GPIO.
1-5
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Pin # Signal
TRACEDATA3
Direction
Description
O
Synchronous CPU trace data bit 3
Select 4-wire synchronous trace interface in ARM core
Enable trace interface in ARM core
Select alternate output function with GPIO_PACFGH[7:4]
28
29
VDD_PADS
PA6
Power
Pads supply (2.1-3.6 V)
Digital I/O
I/O
High
current
TIM1C3
TIM1C3
O
I
Timer 1 channel 3 output
Enable timer output in TIM1_CCER
Select alternate output function with GPIO_PACFGH[11:8]
Timer 1 channel 3 input
Cannot be remapped
30
PB1
I/O
O
Digital I/O
SC1MISO
SPI slave data out of Serial Controller 1
Either disable timer output in TIM2_CCER,
or disable remap with TIM2_OR[4]
Select SPI with SC1_MODE
Select slave with SC1_SPICR
Select alternate output function with GPIO_PBCFGL[7:4]
SC1MOSI
O
SPI master data out of Serial Controller 1
Either disable timer output in TIM2_CCER,
or disable remap with TIM2_OR[4]
Select SPI with SC1_MODE
Select master with SC1_SPICR
Select alternate output function with GPIO_PBCFGL[7:4]
SC1SDA
SC1TXD
I/O
O
TWI data of Serial Controller 1
Either disable timer output in TIM2_CCER,
or disable remap with TIM2_OR[4]
Select TWI with SC1_MODE
Select alternate open-drain output function with GPIO_PBCFGL[7:4]
UART transmit data of Serial Controller 1
Either disable timer output in TIM2_CCER,
or disable remap with TIM2_OR[4]
Select UART with SC1_MODE
Select alternate output function with GPIO_PBCFGL[7:4]
TIM2C1
O
I
Timer 2 channel 1 output
Enable remap with TIM2_OR[4]
Enable timer output in TIM2_CCER
Select alternate output function with GPIO_PACFGL[7:4]
(see also Pin 21)
TIM2C1
Timer 2 channel 1 input
Disable remap with TIM2_OR[4]
(see also Pin 21)
31
PB2
I/O
I
Digital I/O
SC1MISO
SPI master data in of Serial Controller 1
Select SPI with SC1_MODE
Select master with SC1_SPICR
1-6
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Pin # Signal
Direction
Description
SC1MOSI
I
SPI slave data in of Serial Controller 1
Select SPI with SC1_MODE
Select slave with SC1_SPICR
SC1SCL
SC1RXD
I/O
TWI clock of Serial Controller 1
Either disable timer output in TIM2_CCER,
or disable remap with TIM2_OR[5]
Select TWI with SC1_MODE
Select alternate open-drain output function with GPIO_PBCFGL[11:8]
I
UART receive data of Serial Controller 1
Select UART with SC1_MODE
TIM2C2
O
Timer 2 channel 2 output
Enable remap with TIM2_OR[5]
(see also Pin 25)
Enable timer output in TIM2_CCER
Select alternate output function with GPIO_PBCFGL[11:8]
TIM2C2
I
Timer 2 channel 2 input
Enable remap with TIM2_OR[5]
(see also Pin 25)
32
33
SWCLK
JTCK
I/O
I
Serial Wire clock input/output with debugger
Selected when in Serial Wire mode (see JTMS description, Pin 35)
JTAG clock input from debugger
Selected when in JTAG mode (default mode, see JTMS description,
Pin 35)
Internal pull-down is enabled
PC2
I/O
O
Digital I/O
Enable with GPIO_DBGCFG[5]
JTDO
JTAG data out to debugger
Selected when in JTAG mode (default mode, see JTMS description,
Pin 35)
SWO
O
Serial Wire Output asynchronous trace output to debugger
Select asynchronous trace interface in ARM core
Enable trace interface in ARM core
Select alternate output function with GPIO_PCCFGL[11:8]
Enable Serial Wire mode (see JTMS description, Pin 35)
Internal pull-up is enabled
34
35
PC3
I/O
I
Digital I/O
Either Enable with GPIO_DBGCFG[5],
or enable Serial Wire mode (see JTMS description)
JTDI
JTAG data in from debugger
Selected when in JTAG mode (default mode, see JTMS description,
Pin 35)
Internal pull-up is enabled
PC4
I/O
Digital I/O
Enable with GPIO_DBGCFG[5]
1-7
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Pin # Signal
Direction
Description
JTMS
I
JTAG mode select from debugger
Selected when in JTAG mode (default mode)
JTAG mode is enabled after power-up or by forcing nRESET low
Select Serial Wire mode using the ARM-defined protocol through a
debugger
Internal pull-up is enabled
SWDIO
I/O
Serial Wire bidirectional data to/from debugger
Enable Serial Wire mode (see JTMS description)
Select Serial Wire mode using the ARM-defined protocol through a
debugger
Internal pull-up is enabled
36
PB0
I/O
Digital I/O
VREF
Analog O
ADC reference output
Enable analog function with GPIO_PBCFGL[3:0]
VREF
Analog I
ADC reference input
Enable analog function with GPIO_PBCFGL[3:0]
Enable reference output with an Ember system function
IRQA
I
External interrupt source A
TRACECLK
O
Synchronous CPU trace clock
Enable trace interface in ARM core
(see also Pin 25)
Select alternate output function with GPIO_PBCFGL[3:0]
TIM1CLK
TIM2MSK
VDD_PADS
PC1
I
Timer 1 external clock input
Timer 2 external clock mask input
Pads supply (2.1-3.6 V)
Digital I/O
I
37
38
Power
I/O
ADC3
Analog
ADC Input 3
Enable analog function with GPIO_PCCFGL[7:4]
SWO
O
Serial Wire Output asynchronous trace output to debugger
Select asynchronous trace interface in ARM core
Enable trace interface in ARM core
(see also Pin 33)
Select alternate output function with GPIO_PCCFGL[7:4]
TRACEDATA0
O
Synchronous CPU trace data bit 0
Select 1-, 2- or 4-wire synchronous trace interface in ARM core
Enable trace interface in ARM core
Select alternate output function with GPIO_PCCFGL[7:4]
39
40
VDD_MEM
PC0
Power
1.8 V supply (flash, RAM)
I/O
Digital I/O
High
current
Either enable with GPIO_DBGCFG[5],
or enable Serial Wire mode (see JTMS description, Pin 35) and disable
TRACEDATA1
JRST
I
I
JTAG reset input from debugger
Selected when in JTAG mode (default mode, see JTMS description) and
TRACEDATA1 is disabled
Internal pull-up is enabled
IRQD1
Default external interrupt source D
1-8
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Pin # Signal
TRACEDATA1
Direction
Description
O
Synchronous CPU trace data bit 1
Select 2- or 4-wire synchronous trace interface in ARM core
Enable trace interface in ARM core
Select alternate output function with GPIO_PCCFGL[3:0]
41
42
43
PB7
I/O
High
current
Digital I/O
ADC2
Analog
ADC Input 2
Enable analog function with GPIO_PBCFGH[15:12]
IRQC1
I
Default external interrupt source C
TIM1C2
O
Timer 1 channel 2 output
Enable timer output in TIM1_CCER
Select alternate output function with GPIO_PBCFGH[15:12]
TIM1C2
PB6
I
Timer 1 channel 2 input
Cannot be remapped
I/O
High
current
Digital I/O
ADC1
Analog
ADC Input 1
Enable analog function with GPIO_PBCFGH[11:8]
IRQB
I
External interrupt source B
TIM1C1
O
Timer 1 channel 1 output
Enable timer output in TIM1_CCER
Select alternate output function with GPIO_PBCFGH[11:8]
TIM1C1
I
Timer 1 channel 1 input
Cannot be remapped
PB5
I/O
Digital I/O
ADC0
Analog
ADC Input 0
Enable analog function with GPIO_PBCFGH[7:4]
TIM2CLK
TIM1MSK
VDD_CORE
VDD_PRE
VDD_SYNTH
OSCB
I
Timer 2 external clock input
Timer 1 external clock mask input
1.25 V digital core supply decoupling
1.8 V prescaler supply
I
44
45
46
47
Power
Power
Power
I/O
1.8 V synthesizer supply
24 MHz crystal oscillator or left open when using external clock input on
OSCA
48
OSCA
I/O
24 MHz crystal oscillator or external clock input.
(An external clock input should only be used for test and debug
purposes. If used in this manner, the external clock input should be a
1.8 V, 50% duty cycle, square wave.)
49
GND
Ground
Ground supply pad in the bottom center of the package forms Pin 49.
See the various Silicon Labs EM35x Reference Design documentation for
PCB considerations.
1IRQC and IRQD external interrupts can be mapped to any digital I/O pin using the GPIO_IRQSEL and GPIO_IRQDSEL registers.
1-9
120-035X-000 Rev. 1.2
Final
EM351 / EM357
2 Electrical Characteristics
2.1 Absolute Maximum Ratings
Table 2-1 lists the absolute maximum ratings for the EM35x.
Table 2-1. Absolute Maximum Ratings
Test Conditions
Parameter
Min.
Max.
Unit
Regulator input voltage (VDD_PADS)
-0.3
+3.6
V
Analog, Memory and Core voltage
(VDD_24MHZ, VDD_VCO, VDD_RF, VDD_IF,
VDD_PADSA, VDD_MEM, VDD_PRE,
VDD_SYNTH, VDD_CORE)
-0.3
+2.0
V
Voltage on RF_P,N; RF_TX_ALT_P,N
-0.3
+3.6
+15
V
RF Input Power
RX signal into a lossless balun
dBm
(for max level for correct packet reception
see Table 2-7)
Voltage on any GPIO (PA[7:0], PB[7:0],
PC[7:0]), SWCLK, nRESET, VREG_OUT
-0.3
-0.3
VDD_PADS
+0.3
V
V
Voltage on any GPIO pin (PA4, PA5, PB5, PB6,
PB7, PC1), when used as an input to the
general purpose ADC
2.0
Voltage on OSCA, OSCB, NC
-0.3 VDD_PADSA
+0.3
V
Storage temperature
-40
+140
°C
2.2 Recommended Operating Conditions
Table 2-2 lists the rated operating conditions of the EM35x.
Table 2-2. Operating Conditions
Test Conditions
Parameter
Min.
Typ.
Max.
Unit
Regulator input voltage (VDD_PADS)
2.1
3.6
V
Analog and memory input voltage
(VDD_24MHZ, VDD_VCO, VDD_RF, VDD_IF,
VDD_PADSA, VDD_MEM, VDD_PRE,
VDD_SYNTH)
1.7
1.8
1.9
V
Core input voltage when supplied from
internal regulator (VDD_CORE)
1.18
1.18
-40
1.25
1.32
1.9
V
V
Core input voltage when supplied externally
(VDD_CORE)
Operating temperature range
+85
°C
2-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
2.3 Environmental Characteristics
Table 2-3 lists the rated environmental characteristics of the EM35x.
Table 2-3. Environmental Characteristics
Parameter
Test Conditions
Min.
Typ.
Max.
Unit
kV
ESD (human body model)
On any pin
±2
ESD (charged device model)
ESD (charged device model)
Moisture Sensitivity Level (MSL)
Non-RF pins
RF pins
±400
±225
V
V
MSL2
2.4 DC Electrical Characteristics
Table 2-4 lists the DC electrical characteristics of the EM35x.
Table 2-4. DC Characteristics
Test Conditions
Parameter
Min.
Typ.
Max.
Unit
Regulator input voltage (VDD_PADS)
2.1
3.6
V
Power supply range (VDD_MEM)
Regulator output or external input
Regulator output
1.7
1.8
1.9
V
V
Power supply range (VDD_CORE)
1.18
1.25
1.32
Deep Sleep Current
Quiescent current, internal RC oscillator
disabled
-40°C, VDD_PADS=3.6 V
+25°C, VDD_PADS=3.6 V
+85°C, VDD_PADS=3.6 V
-40°C, VDD_PADS=3.6 V
+25°C, VDD_PADS=3.6 V
+85°C, VDD_PADS=3.6 V
-40°C, VDD_PADS=3.6 V
+25°C, VDD_PADS=3.6 V
+85°C, VDD_PADS=3.6 V
-40°C, VDD_PADS=3.6 V
+25°C, VDD_PADS=3.6 V
+85°C, VDD_PADS=3.6 V
With no debugger activity
0.4
0.4
0.7
0.7
0.7
1.1
0.8
1.0
1.5
1.1
1.3
1.8
300
µA
µA
µA
µA
µA
µA
µA
µA
µA
µA
µA
µA
µA
Quiescent current, including internal RC
oscillator
Quiescent current, including 32.768 kHz
oscillator
Quiescent current, including internal RC
oscillator and 32.768 kHz oscillator
Simulated deep sleep (debug mode) current
2-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Parameter
Test Conditions
Min.
Typ.
Max.
Unit
Reset Current
Quiescent current, nRESET asserted
Typ at 25°C/3.0 V
Max at 85°C/3.6 V
1.2
2.0
mA
Processor and Peripheral Currents
ARM® CortexTM-M3, RAM, and flash memory
25°C, 1.8 V memory and 1.25 V core
6.5
7.5
3.0
2.0
mA
mA
mA
mA
ARM® CortexTM-M3 running at 12 MHz from
crystal oscillator
Radio and all peripherals off
ARM® CortexTM-M3, RAM, and flash memory
25°C, 1.8 V memory and 1.25 V core
ARM® CortexTM-M3 running at 24 MHz from
crystal oscillator
Radio and all peripherals off
ARM® CortexTM-M3, RAM, and flash memory
sleep current
25°C, 1.8 V memory and 1.25 V core
ARM® CortexTM-M3 sleeping, CPU clock set to
12 MHz from the crystal oscillator
Radio and all peripherals off
ARM® CortexTM-M3, RAM, and flash memory
sleep current
25°C, 1.8 V memory and 1.25 V core
ARM® CortexTM-M3 sleeping, CPU clock set to
6 MHz from the high frequency RC oscillator
Radio and all peripherals off
Serial controller current
For each controller at maximum data rate
For each timer at maximum clock rate
At maximum sample rate, DMA enabled
0.2
0.25
1.1
mA
mA
mA
General purpose timer current
General purpose ADC current
Rx Current
Radio receiver, MAC, and baseband
ARM® CortexTM-M3 sleeping, CPU clock set to
12 MHz
22.0
25.5
mA
mA
Total Rx current ( = IRadio receiver, MAC and baseband,
25°C, VDD_PADS=3.0 V
ARM® CortexTM-M3 running at 12 MHz
+ IRAM, and Flash memory )
CPU
25°C, VDD_PADS=3.0 V
ARM® CortexTM-M3 running at 24 MHz
26.5
27.5
28.5
mA
mA
mA
Boost mode total Rx current ( = IRadio receiver, MAC 25°C, VDD_PADS=3.0 V
and baseband, CPU+ IRAM, and Flash memory )
ARM® CortexTM-M3 running at 12 MHz
25°C, VDD_PADS=3.0 V
ARM® CortexTM-M3 running at 24 MHz
2-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Parameter
Test Conditions
Min.
Typ.
Max.
Unit
Tx Current
Radio transmitter, MAC, and baseband
25°C and 1.8 V core; max. power out
(+3 dBm typical)
26.0
mA
ARM® CortexTM-M3 sleeping, CPU clock set to
12 MHz
Total Tx current ( = IRadio transmitter, MAC and baseband, 25°C, VDD_PADS=3.0 V; maximum power
42.5
mA
+ IRAM, and Flash memory )
setting (+8 dBm); ARM® CortexTM-M3 running
at 12 MHz
CPU
25°C, VDD_PADS=3.0 V; +3 dBm power
30.0
27.5
21.5
43.5
mA
mA
mA
mA
setting; ARM® CortexTM-M3 running at 12 MHz
25°C, VDD_PADS=3.0 V; 0dBm power setting;
ARM® CortexTM-M3 running at 12 MHz
25°C, VDD_PADS=3.0 V; minimum power
setting; ARM® CortexTM-M3 running at 12 MHz
25°C, VDD_PADS=3.0 V; maximum power
setting (+8 dBm); ARM® CortexTM-M3 running
at 24 MHz
25°C, VDD_PADS=3.0 V; +3 dBm power
31.0
28.5
22.5
mA
mA
mA
setting; ARM® CortexTM-M3 running at 24 MHz
25°C, VDD_PADS=3.0 V; 0 dBm power
setting; ARM® CortexTM-M3 running at 24 MHz
25°C, VDD_PADS=3.0 V; minimum power
setting; ARM® CortexTM-M3 running at 24 MHz
2-4
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 2-1 shows the variation of current in transmit mode (with the ARM® CortexTM-M3 running at 12 MHz).
Figure 2-1. Transmit Power Consumption
2-5
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 2-2 shows typical output power against power setting on the Silicon Labs reference design.
Figure 2-2. Transmit Output Power
2.5 Digital I/O Specifications
Table 2-5 lists the digital I/O specifications for the EM35x. The digital I/O power (named VDD_PADS) comes
from three dedicated pins (Pins 23, 28 and 37). The voltage applied to these pins sets the I/O voltage.
Table 2-5. Digital I/O specifications
Parameter
Test Conditions
Min.
Typ.
Max.
Unit
Voltage supply (Regulator Input voltage)
2.1
3.6
V
Low Schmitt switching threshold
High Schmitt switching threshold
VSWIL
0.42 x
VDD_PADS
0.50 x
VDD_PADS
V
V
Schmitt input threshold going from high to
low
VSWIH
0.62 x
0.80 x
VDD_PADS
VDD_PADS
Schmitt input threshold going from low to
high
Input current for logic 0
Input current for logic 1
IIL
-0.5
+0.5
μA
μA
IIH
2-6
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Parameter
Test Conditions
Min.
Typ.
Max.
Unit
Input pull-up resistor value
RIPU
24
29
34
kΩ
Input pull-down resistor value
Output voltage for logic 0
RIPD
VOL
24
0
29
34
kΩ
0.18 x
V
VDD_PADS
(IOL = 4 mA for standard pads, 8 mA for
high current pads)
Output voltage for logic 1
VOH
0.82 x
VDD_PADS
VDD_PADS
4
V
(IOH = 4 mA for standard pads, 8 mA for
high current pads)
Output source current (standard current
pad)
IOHS
mA
Output sink current (standard current pad)
IOLS
4
8
mA
mA
Output source current
IOHH
high current pad: PA6, PA7, PB6, PB7, PC0
Output sink current
high current pad: PA6, PA7, PB6, PB7, PC0
IOLH
8
mA
mA
Total output current (for I/O Pads)
IOH + IOL
40
Table 2-6 lists the nRESET pin specifications for the EM35x. The digital I/O power (named VDD_PADS) comes
from three dedicated pins (pins 23, 28 and 37). The voltage applied to these pins sets the I/O voltage.
Table 2-6. nReset pin specifications
Parameter
Test Conditions
Min.
Typ.
Max.
Unit
Low Schmitt switching threshold
VSWIL
0.42 x
0.50 x
V
VDD_PADS
VDD_PADS
Schmitt input threshold going from high to
low
High Schmitt switching threshold
VSWIH
0.62 x
0.80 x
V
VDD_PADS
VDD_PADS
Schmitt input threshold going from low to
high
Input current for logic 0
Input current for logic 1
Input pull-up resistor value
IIL
-0.5
+0.5
34
μA
μA
kΩ
IIH
RIPU
24
12
29
Pull-up value while the chip is not reset
Input pull-up resistor value
RIPURESET
14.5
17
kΩ
Pull-up value while the chip is reset
2-7
120-035X-000 Rev. 1.2
Final
EM351 / EM357
2.6 Non-RF System Electrical Characteristics
Table 2-6 lists the non-RF system level characteristics for the EM35x.
Table 2-7. Non-RF System Specifications
Test Conditions
Parameter
Min.
Typ.
Max.
Unit
System wake time from deep sleep
From wakeup event to first ARM® CortexTM-M3
instruction running from 6 MHz internal RC
clock
110
µs
Includes supply ramp time and oscillator
startup time
Shutdown time going into deep sleep
From last ARM® CortexTM-M3 instruction to
deep sleep mode
5
µs
2.7 RF Electrical Characteristics
2.7.1 Receive
Table 2-7 lists the key parameters of the integrated IEEE 802.15.4-2003 receiver on the EM35x.
Note: Receive measurements were collected with the EM35x Ceramic Balun Reference Design (Version A0) at
2440 MHz. The Typical number indicates one standard deviation above the mean, measured at room
temperature (25°C). The Min and Max numbers were measured over process corners at room temperature
Table 2-8. Receive Characteristics
Parameter
Test Conditions
Min.
Typ.
Max.
Unit
Frequency range
2400
2500
MHz
Sensitivity (boost mode)
1% PER, 20 byte packet defined by IEEE
802.15.4-2003
-102
-100
35
-96
dBm
dBm
dB
Sensitivity
1% PER, 20 byte packet defined by IEEE
802.15.4-2003
-94
High-side adjacent channel rejection
Low-side adjacent channel rejection
2nd high-side adjacent channel rejection
2nd low-side adjacent channel rejection
High-side adjacent channel rejection
Low-side adjacent channel rejection
IEEE 802.15.4-2003 interferer signal, wanted
IEEE 802.15.4-2003 signal at -82 dBm
IEEE 802.15.4-2003 interferer signal, wanted
IEEE 802.15.4-2003 signal at -82 dBm
35
dB
IEEE 802.15.4-2003 interferer signal, wanted
IEEE 802.15.4-2003 signal at -82 dBm
46
dB
IEEE 802.15.4-2003 interferer signal, wanted
IEEE 802.15.4-2003 signal at -82 dBm
46
dB
Filtered IEEE 802.15.4-2003 interferer signal,
wanted IEEE 802.15.4-2003 signal at -82 dBm
39
dB
Filtered IEEE 802.15.4-2003 interferer signal,
wanted IEEE 802.15.4-2003 signal at -82 dBm
47
dB
2-8
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Parameter
Test Conditions
Min.
Typ.
Max.
Unit
2nd high-side adjacent channel rejection
Filtered IEEE 802.15.4-2003 interferer signal,
wanted IEEE 802.15.4-2003 signal at -82 dBm
49
dB
2nd low-side adjacent channel rejection
High-side adjacent channel rejection
Low-side adjacent channel rejection
2nd high-side adjacent channel rejection
2nd low-side adjacent channel rejection
Channel rejection for all other channels
Filtered IEEE 802.15.4-2003 interferer signal,
wanted IEEE 802.15.4-2003 signal at -82 dBm
49
44
47
59
59
40
36
dB
dB
CW interferer signal, wanted IEEE 802.15.4-
2003 signal at -82 dBm
CW interferer signal, wanted IEEE 802.15.4-
2003 signal at -82 dBm
dB
CW interferer signal, wanted IEEE 802.15.4-
2003 signal at -82 dBm
dB
CW interferer signal, wanted IEEE 802.15.4-
2003 signal at -82 dBm
dB
IEEE 802.15.4-2003 interferer signal, wanted
IEEE 802.15.4-2003 signal at -82 dBm
dB
802.11g rejection centered at +12 MHz
or -13 MHz
IEEE 802.15.4-2003 interferer signal, wanted
IEEE 802.15.4-2003 signal at -82 dBm
dB
Maximum input signal level for correct
operation
0
dBm
dBc
ppm
Co-channel rejection
IEEE 802.15.4-2003 interferer signal, wanted
IEEE 802.15.4-2003 signal at -82 dBm
-6
Relative frequency error
-120
-120
+120
+120
(50% greater than the 2x40 ppm required by
IEEE 802.15.4-2003)
Relative timing error
ppm
(50% greater than the 2x40 ppm required by
IEEE 802.15.4-2003)
Linear RSSI range
RSSI Range
As defined by IEEE 802.15.4-2003
40
dB
-90
-40
dBm
2-9
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 2-3 shows the variation of receive sensitivity with temperature for boost mode and normal mode for a
typical chip.
Figure 2-3. Receive sensitivity vs temperature
2.7.2 Transmit
Table 2-8 lists the key parameters of the integrated IEEE 802.15.4-2003 transmitter on the EM35x.
Note: Transmit measurements were collected with the EM35x Ceramic Balun Reference Design (Version A0) at
2440 MHz. The Typical number indicates one standard deviation below the mean, measured at room
temperature (25°C). The Min and Max numbers were measured over process corners at room temperature. In
terms of impedance, this reference design presents a 3n3 inductor in parallel with a 100:50 Ω balun to the RF
pins.
2-10
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 2-9. Transmit Characteristics
Parameter
Test Conditions
Min.
Typ.
Max.
Unit
Maximum output power (boost mode)
At highest boost mode power setting (+8)
8
dBm
Maximum output power
At highest normal mode power setting (+3)
At lowest power setting
1
5
-55
5
dBm
dBm
%
Minimum output power
Error vector magnitude (Offset-EVM)
As defined by IEEE 802.15.4-2003, which sets
a 35% maximum
15
Carrier frequency error
PSD mask relative
-40
+40
ppm
dB
3.5 MHz away
3.5 MHz away
-20
-30
PSD mask absolute
dBm
Figure 2-4 shows the variation of transmit power with temperature for maximum boost mode power, and
normal mode for a typical chip.
Figure 2-4. Transmit power vs temperature
2-11
120-035X-000 Rev. 1.2
Final
EM351 / EM357
2.7.3 Synthesizer
Table 2-9 lists the key parameters of the integrated synthesizer on the EM35x.
Table 2-10. Synthesizer Characteristics
Parameter
Test Conditions
Min.
2400
Typ.
Max.
Unit
Frequency range
2500
MHz
Frequency resolution
Lock time
11.7
kHz
μs
From off
100
100
Relock time
Channel change or Rx/Tx turnaround (IEEE
802.15.4-2003 defines 192 μs turnaround
time)
μs
Phase noise at 100 kHz offset
Phase noise at 1 MHz offset
Phase noise at 4 MHz offset
Phase noise at 10 MHz offset
-75
dBc/Hz
dBc/Hz
dBc/Hz
dBc/Hz
-100
-108
-114
2-12
120-035X-000 Rev. 1.2
Final
EM351 / EM357
3 Top-Level Functional Description
Figure 3-1 shows a detailed block diagram of the EM35x.
Figure 3-1. EM35x Block Diagram
TX_ACTIVE
Data
RAM
Program
Flash
PA select
PA
RF_TX_ALT_P,N
RF_P,N
12 kB
128/192 kB
SYNTH
DAC
ADC
MAC
+
Baseband
PA
LNA
ARM® CortexTM-M3
CPU with NVIC
and MPU
2
nd level
IF
Interrupt
controller
Packet Trace
Bias
CPU debug
TPIU/ITM/
FPB/DWT
Encryption
acclerator
OSCA
OSCB
HF crystal
OSC
Internal HF
RC-OSC
Calibration
ADC
General
purpose
timers
Always
Powered
Domain
1.25V
Regulator
VDD_CORE
VREG_OUT
Serial
Wire and
JTAG
SWCLK,
JTCK
1.8V
Regulator
GPIO
registers
Watchdog
debug
General
Purpose
ADC
nRESET
POR
UART/
SPI/TWI
Chip
manager
Sleep
timer
LF crystal
OSC
Internal LF
RC-OSC
GPIO multiplexor switch
PA[7:0], PB[7:0], PC[7:0]
The EM35x radio receiver is a low-IF, super-heterodyne receiver. The architecture has been chosen to
optimize co-existence with other devices in the 2.4 GHz band (namely Wi-Fi and Bluetooth), and to minimize
power consumption. The receiver uses differential signal paths to reduce sensitivity to noise interference.
Following RF amplification, the signal is downconverted by an image-rejecting mixer, filtered, and then
digitized by an ADC.
The digital section of the receiver uses a coherent demodulator to generate symbols for the hardware-based
MAC. The digital receiver also contains the analog radio calibration routines, and controls the gain within the
receiver path.
The radio transmitter uses an efficient architecture in which the data stream directly modulates the VCO
frequency. An integrated PA provides the output power. Digital logic controls Tx path and output power
calibration. If the EM35x is to be used with an external PA, use the TX_ACTIVE or nTX_ACTIVE signal to control
the timing of the external switching logic.
The integrated 4.8 GHz VCO and loop filter minimize off-chip circuitry. Only a 24 MHz crystal with its loading
capacitors is required to establish the PLL local oscillator signal.
The MAC interfaces the on-chip RAM to the Rx and Tx baseband modules. The MAC provides hardware-based
IEEE 802.15.4-2003 packet-level filtering. It supplies an accurate symbol time base that minimizes the
synchronization effort of the Ember software and meets the protocol timing requirements. In addition, it
provides timer and synchronization assistance for the IEEE 802.15.4-2003 CSMA-CA algorithm.
The EM35x integrates hardware support for a packet trace module, which allows robust packet-based debug.
This element is a critical component of Ember Desktop, the Ember development environment, and provides
advanced network debug capability when used with the Ember Debug Adapter (ISA3).
3-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
The EM35x integrates an ARM® CortexTM-M3 microprocessor, revision r1p1. This industry-leading core provides
32-bit performance and is very power-efficient. It has excellent code density using the ARM® Thumb-2
instruction set. The processor can be operated at 12 MHz or 24 MHz when using the high-frequency crystal
oscillator, or at 6 MHz or 12 MHz when using the high-frequency internal RC oscillator.
The EM351 has 128 kB of flash memory and the EM357 has 192 kB of flash memory. Both chips have 12 kB of
RAM on-chip, and the ARM configurable memory protection unit (MPU).
The EM35x implements both the ARM Serial Wire and JTAG debug interfaces. These interfaces provide real
time, non-intrusive programming and debugging capabilities. Serial Wire and JTAG provide the same
functionality, but are mutually exclusive. The Serial Wire interface uses two pins; the JTAG interface uses
five. Serial Wire is preferred, since it uses fewer pins.
The EM35x contains 24 GPIO pins shared with other peripheral or alternate functions. Because of flexible
routing within the EM35x, external devices can use the alternate functions on a variety of different GPIOs. The
integrated serial controller SC1 can be configured for SPI (master or slave), TWI (master-only), or UART
operation, and the serial controller SC2 can be configured for SPI (master or slave) or TWI (master-only)
operation.
The EM35x has a general purpose ADC which can sample analog signals from six GPIO pins in single-ended or
differential modes. It can also sample the 1.8 V regulated supply VDD_PADSA, the voltage reference VREF, and
GND. The ADC has one voltage range: 0 V to 1.2 V (normal). The ADC has a DMA mode to capture samples and
automatically transfer them into RAM. The integrated voltage reference for the ADC, VREF, can be made
available to external circuitry. An external voltage reference can also be driven into the ADC. The regulator
input voltage, VDD_PADS, cannot be measured using the general purpose ADC, but it can be measured through
Ember software.
The EM35x contains four oscillators: a high-frequency 24 MHz external crystal oscillator, a high-frequency
12 MHz internal RC oscillator, an optional low-frequency 32.768 kHz external crystal oscillator, and a low-
frequency 10 kHz internal RC oscillator.
The EM35x has an ultra low power, deep sleep state with a choice of clocking modes. The sleep timer can be
clocked with either the external 32.768 kHz crystal oscillator or with a 1 kHz clock derived from the internal
10 kHz RC oscillator. Alternatively, all clocks can be disabled for the lowest power mode. In the lowest power
mode, only external events on GPIO pins will wake up the chip. The EM35x has a fast startup time (typically
110 µs) from deep sleep to the execution of the first ARM® CortexTM-M3 instruction.
The EM35x contains three power domains. The always-on high voltage supply powers the GPIO pads and
critical chip functions. Regulated low voltage supplies power the rest of the chip. The low voltage supplies are
disabled during deep sleep to reduce power consumption. Integrated voltage regulators generate regulated
1.25 V and 1.8 V voltages from an unregulated supply voltage. The 1.8 V regulator output is decoupled and
routed externally to supply analog blocks, RAM, and flash memories. The 1.25 V regulator output is decoupled
externally and supplies the core logic.
Note: The EM35x is not pin-compatible with the previous generation chip, the EM250, except for the RF
section of the chip. Pins 1-11 and 45-48 are compatible, to ease migration to the EM35x.
3-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
4
Radio Module
The radio module consists of an analog front end and digital baseband as shown in Figure 3-1, EM35x Block
Diagram in Chapter 3, Top Level Functional Description.
4.1
Receive (Rx) Path
The Rx path uses a low-IF, super-heterodyne receiver that rejects the image frequency using complex mixing
and polyphase filtering. In the analog domain, the input RF signal from the antenna is first amplified and
mixed down to a 4 MHz IF frequency. The mixers’ output is filtered, combined, and amplified before being
sampled by a 12 MSPS ADC. The digitized signal is then demodulated in the digital baseband. The filtering
within the Rx path improves the EM35x’s co-existence with other 2.4 GHz transceivers such as Zigbee/
802.15.4-2003, IEEE 802.11-2007, and Bluetooth radios. The digital baseband also provides gain control of the
Rx path, both to enable the reception of small and large wanted signals and to tolerate large interferers.
4.1.1 Rx Baseband
The EM35x Rx digital baseband implements a coherent demodulator for optimal performance. The baseband
demodulates the O-QPSK signal at the chip level and synchronizes with the IEEE 802.15.4-2003-defined
preamble. An automatic gain control (AGC) module adjusts the analog gain continuously every ¼ symbol until
the preamble is detected. Once detected, the gain is fixed for the remainder of the packet. The baseband
despreads the demodulated data into 4-bit symbols. These symbols are buffered and passed to the hardware-
based MAC module for packet assembly and filtering.
In addition, the Rx baseband provides the calibration and control interface to the analog Rx modules,
including the LNA, Rx baseband filter, and modulation modules. The Ember software includes calibration
algorithms that use this interface to reduce the effects of silicon process and temperature variation.
4.1.2 RSSI and CCA
The EM35x calculates the RSSI over every 8-symbol period as well as at the end of a received packet. The
linear range of RSSI is specified to be at least 40 dB over temperature. At room temperature, the linear range
is approximately 60 dB (-90 dBm to -30 dBm input signal).
The EM35x Rx baseband provides support for the IEEE 802.15.4-2003 RSSI CCA method. Clear channel reports
busy medium if RSSI exceeds its threshold.
4.2
Transmit (Tx) Path
The EM35x Tx path produces an O-QPSK-modulated signal using the analog front end and digital baseband. The
area- and power-efficient Tx architecture uses a two-point modulation scheme to modulate the RF signal
generated by the synthesizer. The modulated RF signal is fed to the integrated PA and then out of the EM35x.
4.2.1 Tx Baseband
The EM35x Tx baseband in the digital domain spreads the 4-bit symbol into its IEEE 802.15.4-2003-defined 32-
chip sequence. It also provides the interface for the Ember software to calibrate the Tx module to reduce
silicon process, temperature, and voltage variations.
4.2.2 TX_ACTIVE and nTX_ACTIVE Signals
For applications requiring an external PA, two signals are provided called TX_ACTIVE and nTX_ACTIVE. These
signals are the inverse of each other. They can be used for external PA power management and RF switching
logic. In transmit mode the Tx baseband drives TX_ACTIVE high, as described in Table 7-5, GPIO Signal
Assignments. In receive mode the TX_ACTIVE signal is low. TX_ACTIVE is the alternate function of PC5, and
4-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
nTX_ACTIVE is the alternate function of PC6. See Chapter 7 GPIO for details of the alternate GPIO functions.
The digital I/O that provide these signals have a 4 mA output sink and source capability.
4.3
4.4
Calibration
The Ember software calibrates the radio using dedicated hardware resources.
Integrated MAC Module
The EM35x integrates most of the IEEE 802.15.4-2003 MAC requirements in hardware. This allows the ARM®
CortexTM-M3 CPU to provide greater bandwidth to application and network operations. In addition, the
hardware acts as a first-line filter for unwanted packets. The EM35x MAC uses a DMA interface to RAM to
further reduce the overall ARM® CortexTM-M3 CPU interaction when transmitting or receiving packets.
When a packet is ready for transmission, the Ember software configures the Tx MAC DMA by indicating the
packet buffer RAM location. The MAC waits for the backoff period, then switches the baseband to Tx mode
and performs channel assessment. When the channel is clear the MAC reads data from the RAM buffer,
calculates the CRC, and provides 4-bit symbols to the baseband. When the final byte has been read and sent
to the baseband, the CRC remainder is read and transmitted.
The MAC is in Rx mode most of the time. In Rx mode various format and address filters keep unwanted packets
from using excessive RAM buffers, and prevent the CPU from being unnecessarily interrupted. When the
reception of a packet begins, the MAC reads 4-bit symbols from the baseband and calculates the CRC. It then
assembles the received data for storage in a RAM buffer. Rx MAC DMA provides direct access to RAM. Once the
packet has been received additional data, which provides statistical information on the packet to the Ember
software, is appended to the end of the packet in the RAM buffer space.
The primary features of the MAC are:
CRC generation, appending, and checking
.
Hardware timers and interrupts to achieve the MAC symbol timing
.
Automatic preamble and SFD pre-pending on Tx packets
.
Address recognition and packet filtering on Rx packets
.
Automatic acknowledgement transmission
.
Automatic transmission of packets from memory
.
Automatic transmission after backoff time if channel is clear (CCA)
.
Automatic acknowledgement checking
.
Time stamping received and transmitted messages
.
Attaching packet information to received packets (LQI, RSSI, gain, time stamp, and packet status)
.
IEEE 802.15.4-2003 timing and slotted/unslotted timing
.
4.5
4.6
Packet Trace Interface (PTI)
The EM35x integrates a true PHY-level PTI for effective network-level debugging. It monitors all the PHY Tx
and Rx packets between the MAC and baseband modules without affecting their normal operation. It cannot
be used to inject packets into the PHY/MAC interface. This 500 kbps asynchronous interface comprises the
frame signal (PTI_EN, PA4) and the data signal (PTI_DATA, PA5). PTI is supported by the Ember development
tools.
Random Number Generator
Thermal noise in the analog circuitry is digitized to provide entropy for a true random number generator
(TRNG). The TRNG produces 16-bit uniformly distributed numbers. The Ember software uses the TRNG to seed
a pseudo random number generator (PRNG). The TRNG is also used directly for cryptographic key generation.
4-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
5 ARM® Cortex™-M3 and Memory Modules
This chapter discusses the ARM® CortexTM-M3 Microprocessor, and reviews the EM35x’s flash and RAM memory
modules as well as the Memory Protection Unit (MPU).
5.1 ARM® Cortex™-M3 Microprocessor
The EM35x integrates the ARM® CortexTM-M3 microprocessor, revision r1p1, developed by ARM Ltd., making the
EM35x a true System-on-Chip solution. The ARM® CortexTM-M3 is an advanced 32-bit modified Harvard
architecture processor that has separate internal program and data buses, but presents a unified program and
data address space to software. The word width is 32 bits for both the program and data sides. The ARM®
CortexTM-M3 allows unaligned word and half-word data accesses to support efficiently-packed data structures.
The ARM® CortexTM-M3 clock speed is configurable to 6 MHz, 12 MHz, or 24 MHz. For normal operation 24 MHz
is preferred over 12 MHz due to improved performance for all applications and improved duty cycling for
applications using sleep modes. The 6 MHz operation can only be used when radio operations are not required
since the radio requires an accurate 12 MHz clock.
The ARM® CortexTM-M3 in the EM35x has also been enhanced to support two separate memory protection
levels. Basic protection is available without using the MPU, but normal operation uses the MPU. The MPU
allows for protecting unimplemented areas of the memory map to prevent common software bugs from
interfering with software operation. The architecture could also allow for separation of the networking stack
from the application code using a fine granularity RAM protection module. Errant writes are captured and
details are reported to the developer to assist in tracking down and fixing issues.
5-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
5.2 Embedded Memory
Figure 5-1 shows the EM351 ARM® CortexTM-M3 memory map and Figure 5-2 shows the EM357 ARM® CortexTM-M3
memory map.
Figure 5-1. EM351 ARM® CortexTM-M3 Memory Map
0xE00FFFFF
ROM table
0xE00FF000
Not used
0xE0042000
0xFFFFFFFF
Not used
0xE0041000
Not used
TPIU
0xE0040000
Private periph bus (external)
0xE003FFFF
Not used
Private periph bus (internal)
0xE000F000
0xE0000000
NVIC
0xDFFFFFFF
0xE000E000
Not used
0xE0003000
FPB
0xE0002000
DWT
0xE0001000
Not used
ITM
0xE0000000
0x42002XXX
Register bit band
alias region
0xA0000000
0x9FFFFFFF
mapped onto System
interface
(not used)
0x42000000
0x40000XXX
Not used
Registers
mapped onto System
interface
0x40000000
0x22002000
0x60000000
0x5FFFFFFF
RAM bit band
alias region
mapped onto System
interface
Peripheral
(not used)
0x40000000
0x3FFFFFFF
0x22000000
0x20002FFF
RAM (12kB)
mapped onto System
interface
RAM
0x20000000
0x08040FFF
0x20000000
0x1FFFFFFF
Customer Info Block (2kB)
Fixed Info Block (2kB)
0x08040800
0x080407FF
0x08040000
Flash
0x0801FFFF
0x00000000
Main Flash Block (128kB)
Upper mapping
(Boot mode)
0x08000000
0x0001FFFF
Optional boot mode
maps Fixed Info Block
to the start of memory
Main Flash Block (128kB)
Lower mapping
0x000007FF
0x00000000
(Normal Mode)
Fixed Info Block (2kB)
5-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 5-2. EM357 ARM® CortexTM-M3 Memory Map
0xE00FFFFF
0xE00FF000
ROM table
Not used
Not used
TPIU
0xE0042000
0xE0041000
0xE0040000
0xFFFFFFFF
Not used
Private periph bus (external)
Private periph bus (internal)
0xE003FFFF
0xE000F000
Not used
NVIC
0xE0000000
0xDFFFFFFF
0xE000E000
0xE0003000
0xE0002000
0xE0001000
0xE0000000
Not used
FPB
DWT
Not used
ITM
0x42002XXX
Register bit band
alias region
0xA0000000
0x9FFFFFFF
mapped onto System
interface
(not used)
0x42000000
0x40000XXX
Not used
Registers
mapped onto System
interface
0x40000000
0x22002000
0x60000000
0x5FFFFFFF
RAM bit band
alias region
mapped onto System
interface
Peripheral
(not used)
0x40000000
0x3FFFFFFF
0x22000000
0x20002FFF
RAM (12kB)
mapped onto System
interface
RAM
0x20000000
0x08040FFF
0x20000000
0x1FFFFFFF
Customer Info Block (2kB)
Fixed Info Block (2kB)
0x08040800
0x080407FF
0x08040000
Flash
0x0802FFFF
0x00000000
Main Flash Block (192kB)
Upper mapping
(Boot mode)
0x08000000
0x0002FFFF
Optional boot mode
maps Fixed Info Block
to the start of memory
Main Flash Block (192kB)
Lower mapping
0x000007FF
0x00000000
(Normal Mode)
Fixed Info Block (2kB)
5-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
5.2.1
Flash Memory
5.2.1.1
Flash Overview
The EM351 provides a total of 132 kB of flash memory and the EM357 provides a total of 196 kB of flash
memory. The flash memory is provided in three separate blocks:
Main Flash Block (MFB)
.
Fixed Information Block (FIB)
.
Customer Information Block (CIB)
.
The MFB is divided into 2048-byte pages. The EM351 has 64 pages and the EM357 has 96 pages. The CIB is a
single 2048-byte page. The FIB is a single 2048-byte page. The smallest erasable unit is one page and the
smallest writable unit is an aligned 16-bit half-word. The flash is rated to have a guaranteed 20,000
write/erase cycles. The flash cell has been qualified for a data retention time of >100 years at room
temperature.
Flash may be programmed either through the Serial Wire/JTAG interface or through bootloader software.
Programming flash through Serial Wire/JTAG requires the assistance of RAM-based utility code. Programming
through a bootloader requires Ember software for over-the-air loading or serial link loading.
5.2.1.2
Main Flash Block
The start of the MFB is mapped to both address 0x00000000 and address 0x08000000 in normal boot mode, but
is mapped only to address 0x08000000 in FIB monitor mode (see also section 7.5, Boot Configuration in
Chapter 7, GPIO). Consequently, it is recommended that software intended to execute from the MFB is
designed to operate from the upper address, 0x08000000, since this address mapping is always available in all
modes.
The MFB stores all program instructions and constant data. A small portion of the MFB is devoted to non-
volatile token storage using the Ember Simulated EEPROM system.
5.2.1.3
Fixed Information Block
The 2 kB FIB is used to store fixed manufacturing data including serial numbers and calibration values. The
start of the FIB is mapped to address 0x08040000. This block can only be programmed during production by
Silicon Labs.
The FIB also contains a monitor program, which is a serial-link-only way of performing low-level memory
accesses. In FIB monitor mode (see section 7.5, Boot Configuration in Chapter 7, GPIO), the start of the FIB is
mapped to both address 0x00000000 and address 0x08040000 so the monitor may be executed out of reset.
5.2.1.4
Customer Information Block
The 2048 byte CIB can be used to store customer data. The start of the CIB is mapped to address 0x08040800.
The CIB cannot be executed.
The first eight half-words of the CIB are dedicated to special storage called option bytes. An option byte is a
16 bit quantity of flash where the lower 8 bits contain the data and the upper 8 contain the inverse of the
lower 8 bits. The upper 8 bits are automatically generated by hardware and cannot be written to by the user,
see Table 5-1.
The option byte hardware also verifies the inverse of each option byte when exiting from reset and generates
an error, which prevents the CPU from executing code, if a discrepancy is found. All of this is transparent to
the user.
5-4
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 5-1. Option Byte Storage
Address
bits [15:8]
bits [7:0]
Notes
0x08040800
Inverse Option Byte 0
Option Byte 0
Configures flash read protection
Reserved
0x08040802
0x08040804
0x08040806
0x08040808
0x0804080A
0x0804080C
0x0804080E
Inverse Option Byte 1
Inverse Option Byte 2
Inverse Option Byte 3
Inverse Option Byte 4
Inverse Option Byte 5
Inverse Option Byte 6
Inverse Option Byte 7
Option Byte 1
Option Byte 2
Option Byte 3
Option Byte 4
Option byte 5
Option Byte 6
Option Byte 7
Available for customer use1
Available for customer use1
Configures flash write protection
Configures flash write protection
Configures flash write protection2
Reserved
1 Option bytes 2 and 3 do not link to any specific hardware functionality other than the option byte loader. Therefore, they
are best used for storing data that requires a hardware verification of the data integrity.
2 Option byte 6 is reserved/unused in the EM351 due to the smaller flash size.
Table 5-2 shows the mapping of the option bytes that are used for read and write protection of the flash. Each
bit of the flash write protection option bytes protects a 4 page region of the main flash block. The EM351 has
16 regions and therefore option bytes 4 and 5 control flash write protection (option byte 6 is
reserved/unused). The EM357 has 24 regions and therefore option bytes 4, 5, and 6 control flash write
protection. These write protection bits are active low, and therefore the erased state of 0xFF disables write
protection. Like read protection, write protection only takes effect after a reset. Write protection not only
prevents a write to the region, but also prevents page erasure.
Option byte 0 controls flash read protection. When option byte 0 is set to 0xA5, read protection is disabled.
All other values, including the erased state 0xFF, enable read protection when coming out of reset. The
internal state of read protection (active versus disabled) can only be changed by applying a full chip reset. If a
debugger is connected to the EM35x, the intrusion state is latched. Read protection is combined with this
latched intrusion signal. When both read protection and intrusion are set, all flash is disconnected from the
internal bus. As a side effect, the CPU cannot execute code since all flash is disconnected from the bus. This
functionality prevents a debug tool from being able to read the contents of any flash. The only means of
clearing the intrusion signal is to disconnect the debugger and reset the entire chip using the nRESET pin. By
requiring a chip reset, a debugger cannot install or execute malicious code that could allow the contents of
the flash to be read.
The only way to disable read protection is to program option byte 0 with the value 0xA5. Option byte 0 must
be erased before it can be programmed. Erasing option byte 0 while read protection is active automatically
mass-erases the main flash block. By automatically erasing main flash, a debugger cannot disable read
protection and readout the contents of main flash without destroying its contents.
Note: When read protection is active, the bottom four flash pages, addresses 0x08000000 to 0x08001FFF, are
automatically write-protected. Write protecting the bottom four flash pages of main flash prevents an
attacker from reprogramming the reset vector and executing arbitrary code.
In general, if read protection is active then write protection should also be active. This prevents an attacker
from reprogramming flash with malicious code that could readout the flash after the debugger is
disconnected. Even though read protection automatically protects the reset vector, the same technique of
reprogramming flash could be performed at an address outside the bottom four flash pages. To obtain fully
protected flash, both read protection and write protection should be active.
5-5
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 5-2. Option Byte Write Protection Bit Map
Option Byte
Bit
Notes
Option Byte 0
bit [7:0]
Read protection of all flash (MFB, FIB, CIB)
Option Byte 1
Option Byte 2
Option Byte 3
Option Byte 4
bit [7:0]
bit [7:0]
bit [7:0]
bit [0]
bit [1]
bit [2]
bit [3]
bit [4]
bit [5]
bit [6]
bit [7]
bit [0]
bit [1]
bit [2]
bit [3]
bit [4]
bit [5]
bit [6]
bit [7]
bit [0]
bit [1]
bit [2]
bit [3]
bit [4]
bit [5]
bit [6]
bit [7]
bit [7:0]
Reserved for Silicon Labs use
Available for customer use
Available for customer use
Write protection of address range 0x08000000 – 0x08001FFF
Write protection of address range 0x08002000 – 0x08003FFF
Write protection of address range 0x08004000 – 0x08005FFF
Write protection of address range 0x08006000 – 0x08007FFF
Write protection of address range 0x08008000 – 0x08009FFF
Write protection of address range 0x0800A000 – 0x0800BFFF
Write protection of address range 0x0800C000 – 0x0800DFFF
Write protection of address range 0x0800E000 – 0x0800FFFF
Write protection of address range 0x08010000 – 0x08011FFF
Write protection of address range 0x08012000 – 0x08013FFF
Write protection of address range 0x08014000 – 0x08015FFF
Write protection of address range 0x08016000 – 0x08017FFF
Write protection of address range 0x08018000 – 0x08019FFF
Write protection of address range 0x0801A000 – 0x0801BFFF
Write protection of address range 0x0801C000 – 0x0801DFFF
Write protection of address range 0x0801E000 – 0x0801FFFF
Write protection of address range 0x08020000 – 0x08021FFF
Write protection of address range 0x08022000 – 0x08023FFF
Write protection of address range 0x08024000 – 0x08025FFF
Write protection of address range 0x08026000 – 0x08027FFF
Write protection of address range 0x08028000 – 0x08029FFF
Write protection of address range 0x0802A000 – 0x0802BFFF
Write protection of address range 0x0802C000 – 0x0802DFFF
Write protection of address range 0x0802E000 – 0x0802FFFF
Reserved for Silicon Labs use
Option Byte 5
Option Byte 61
Option Byte 7
1 Option byte 6 is reserved/unused in the EM351 due to the smaller flash size.
5-6
120-035X-000 Rev. 1.2
Final
EM351 / EM357
5.2.1.5
Simulated EEPROM
Ember software reserves 8 kB of the main flash block as a simulated EEPROM storage area for stack and
customer tokens. The simulated EEPROM storage area implements a wear-leveling algorithm to extend the
number of simulated EEPROM write cycles beyond the physical limit of 20,000 write cycles for which each
flash cell is qualified.
5.2.2
RAM
RAM Overview
5.2.2.1
The EM35x has 12 kB of static RAM on-chip. The start of RAM is mapped to address 0x20000000. Although the
ARM® CortexTM-M3 allows bit band accesses to this address region, the standard MPU configuration does not
permit use of the bit-band feature.
The RAM is physically connected to the AHB System bus and is therefore accessible to both the ARM® CortexTM-
M3 microprocessor and the debugger. The RAM can be accessed for both instruction and data fetches as bytes,
half words, or words. The standard MPU configuration does not permit execution from the RAM, but for special
purposes the MPU may be disabled. To the bus, the RAM appears as 32-bit wide memory and in most situations
has zero wait state read or write access. In the higher CPU clock mode the RAM requires two wait states. This
is handled by hardware transparent to the user application with no configuration required.
5.2.2.2
Direct Memory Access (DMA) to RAM
Several of the peripherals are equipped with DMA controllers allowing them to transfer data into and out of
RAM autonomously. This applies to the radio (802.15.4-2003 MAC), general purpose ADC, and both serial
controllers. In the case of the serial controllers, the DMA is full duplex so that a read and a write to RAM may
be requested at the same time. Thus there are six DMA channels in total. See Chapter 8, Section 8.7 and
Chapter 10, Section 10.1.4 for a description of how to configure the serial controllers and ADC for DMA
operation. The DMA channels do not use AHB system bus bandwidth as they access the RAM directly.
The EM35x integrates a DMA arbiter that ensures fair access to the microprocessor as well as the peripherals
through a fixed priority scheme appropriate to the memory bandwidth requirements of each master. The
priority scheme is as follows, with the top peripheral being the highest priority:
1. General Purpose ADC
2. Serial Controller 2 Receive
3. Serial Controller 2 Transmit
4. MAC
5. Serial Controller 1 Receive
6. Serial Controller 1 Transmit
5.2.2.3
RAM Memory Protection
The EM35x integrates two memory protection mechanisms. The first memory protection mechanism is through
the ARM® CortexTM-M3 Memory Protection Unit (MPU) described in the Memory Protection Unit section. The
MPU may be used to protect any area of memory. MPU configuration is normally handled by Ember software.
The second memory protection mechanism is through a fine granularity RAM protection module. This allows
segmentation of the RAM into 32-byte blocks where any block can be marked as write protected. An attempt
to write to a protected RAM block using a user mode write results in a bus error being signaled on the AHB
System bus. A privileged mode write is allowed at any time and reads are allowed in either mode. The main
purpose of this fine granularity RAM protection module is to notify the software of erroneous writes to system
areas of memory. RAM protection is configured using a group of registers that provide a bit map. Each bit in
the map represents a 32-byte block of RAM. When the bit is set the block is write-protected.
The fine granularity RAM memory protection mechanism is also available to the peripheral DMA controllers. A
register bit enables protection from DMA writes to protected memory. If a DMA write is made to a protected
location in RAM, a management interrupt is generated. At the same time the faulting address and the
5-7
120-035X-000 Rev. 1.2
Final
EM351 / EM357
identification of the peripheral is captured for later debugging. Note that only peripherals capable of writing
data to RAM, such as received packet data or a received serial port character, can generate this interrupt.
5.2.3
Registers
Appendix A, Register Address Table provides a short description of all application-accessible registers within
the EM35x. Complete descriptions are provided at the end of each applicable peripheral’s description. The
registers are mapped to the system address space starting at address 0x40000000. These registers allow for
the control and configuration of the various peripherals and modules. The CPU only performs word-aligned
accesses on the system bus. The CPU performs a word aligned read-modify-write for all byte, half-word, and
unaligned writes and a word-aligned read for all reads. Silicon Labs recommends accessing all peripheral
registers using word-aligned addressing.
As with the RAM, the peripheral registers fall within an address range that allows for bit-band access by the
ARM® CortexTM-M3, but the standard MPU configuration does not allow access to this alias address range.
5.3 Memory Protection Unit
The EM35x includes the ARM® CortexTM-M3 Memory Protection Unit, or MPU. The MPU controls access rights
and characteristics of up to eight address regions, each of which may be divided into eight equal sub-regions.
Refer to the ARM® CortexTM-M3 Technical Reference Manual (DDI 0337A) for a detailed description of the MPU.
Ember software configures the MPU in a standard configuration and application software should not modify it.
The configuration is designed for optimal detection of illegal instruction or data accesses. If an illegal access
is attempted, the MPU captures information about the access type, the address being accessed, and the
location of the offending software. This simplifies software debugging and increases the reliability of deployed
devices. As a consequence of this MPU configuration, accessing RAM and register bit-band address alias regions
is not permitted, and generates a bus fault if attempted.
5-8
120-035X-000 Rev. 1.2
Final
EM351 / EM357
6 System Modules
System modules encompass power domains, resets, clocks, system timers, power management, and
encryption. Figure 6-1 shows these modules and how they interact.
Figure 6-1. System Module Block Diagram
OSCRC
DIV10
CLK1K
OSC32A
OSC32B
CLK32K
OSC32K
Wakeup Recording
deep sleep
REG_EN
wakeup
Power Management
watchdog
Sleep Timer
Watchdog
always-on supply
VDD_PADS
POR HV
POR HV
VREG_1V25
VREG_1V8
recomended
connections for
internal regulator
VREG_OUT
VDD_MEM
mem supply
POR LVmem
core supply
POR LV
VDD_CORE
nRESET
External
Regulator
POR LVcore
optional
connections for
external regulator
Reset Filter
SWJ
JRST
CDBGRSTREQ
registers
registers
PRESETHV
always-on domain
PRESETLV
AHB-AP
RAM
ARM®
Cortex-M3
CPU
FLITF
Flash
ARM®
Cortex-M3
Debug
OSCHF
OSCA
OSCB
SYSCLK
Security Accelerator
OSC24M
mem domain
core domain
6-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
6.1 Power domains
The EM35x contains three power domains:
An “always-on domain” containing all logic and analog cells required to manage the EM35x’s power modes,
including the GPIO controller and sleep timer. This domain must remain powered.
.
.
.
A “core domain” containing the CPU, Nested Vectored Interrupt Controller (NVIC), and peripherals. To
save power, this domain can be powered down using a mode called deep sleep.
A “memory domain” containing the RAM and flash memories. This domain is managed by the power
management controller. When in deep sleep, the RAM portion of this domain is powered from the always-
on domain supply to retain the RAM contents while the regulators are disabled. During deep sleep the flash
portion is completely powered down.
6.1.1
Internally regulated power
The preferred and recommended power configuration is to use the internal regulated power supplies to
provide power to the core and memory domains. The internal regulators (VREG_1V25 and VREG_1V8) generate
nominal 1.25 V and 1.8 V supplies. The 1.25 V supply is internally routed to the core domain and to an
external pin. The 1.8 V supply is routed to an external pin where it can be externally routed back into the chip
to supply the memory domain. The internal regulators are described in Chapter 16, Integrated Voltage
Regulator.
When using the internal regulators, the always-on domain must be powered between 2.1 V and 3.6 V at all
four VDD_PADS pins.
When using the internal regulators, the VREG_1V8 regulator output pin (VREG_OUT) must be connected to the
VDD_MEM, VDD_PADSA, VDD_VCO, VDD_RF, VDD_IF, VDD_PRE, and VDD_SYNTH pins.
When using the internal regulators, the VREG_1V25 regulator output and supply requires a connection
between both VDD_CORE pins.
6.1.2
Externally regulated power
Optionally, the on-chip regulators may be left unused, and the core and memory domains may instead be
powered from external supplies. For simplicity, the voltage for the core domain can be raised to nominal
1.8 V, requiring only one external regulator, or the core domain can be powered from the on-chip regulators
while the other domains are powered externally. Note that if the core domain is powered at a higher voltage
(1.8 V instead of 1.25 V) then power consumption increases. A regulator enable signal, REG_EN, is provided for
control of external regulators. This is an open-drain signal that requires an external pull-up resistor. If REG_EN
is not required to control external regulators it can be disabled (see section 7.3, Forced Functions in Chapter
7, GPIO).
Using an external regulator requires the always-on domain to be powered between 2.1 V and 3.6 V at all four
VDD_PADS pins.
When using an external regulator, the VREG_1V8 regulator output pin (VREG_OUT) must be left unconnected.
When using an external regulator, this external nominal 1.8 V supply has to be connected to both VDD_CORE
pins and to the VDD_MEM, VDD_PADSA, VDD_VCO, VDD_RF, VDD_IF, VDD_PRE and VDD_SYNTH pins.
6-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
6.2 Resets
The EM35x resets are generated from a number of sources. Each of these reset sources feeds into central reset
detection logic that causes various parts of the system to be reset depending on the state of the system and
the nature of the reset event.
6.2.1
Reset Sources
6.2.1.1
Power-On-Resets (POR HV and POR LV)
The EM35x measures the voltage levels supplied to the three power domains. If a supply voltage drops below a
low threshold, then a reset is applied. The reset is released if the supply voltage rises above a high threshold.
There are three detection circuits for power-on-reset as follows:
POR HV monitors the always-on domain supply voltage. Thresholds are given in Table 6-1.
.
POR LVcore monitors the core domain supply voltage. Thresholds are given in Table 6-2.
.
POR LVmem monitors the memory supply voltage. Thresholds are given in Table 6-3.
.
Table 6-1. POR HV Thresholds
Parameter
Test conditions
Min
Typ
Max
Unit
Always-on domain release
0.62
0.95
1.20
V
Always-on domain assert
Supply rise time
0.45
0.65
0.85
250
V
From 0.5 V to 1.7 V
µs
Table 6-2. POR LVcore Thresholds
Test conditions
Parameter
Min
Typ
Max
Unit
1.25 V domain release
0.9
0.8
1.0
0.9
1.1
V
1.25 V domain assert
1.0
V
Table 6-3 POR LVmem Thresholds
Test conditions
Parameter
Min
Typ
Max
1.65
1.54
Unit
V
1.8 V domain release
1.8 V domain assert
1.35
1.26
1.5
1.4
V
The POR LVcore and POR LVmem reset sources are merged to provide a single reset source, POR LV, to the
Reset Generation module, since the detection of either event needs to reset the same system modules.
6-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
6.2.1.2
nRESET Pin
A single active low pin, nRESET, is provided to reset the system. This pin has a Schmitt triggered input.
To afford good noise immunity and resistance to switch bounce, the pin is filtered with the Reset Filter
module and generates the pin reset source, nRESET, to the Reset Generation module. Table 6-4 contains the
specification for the filter.
Table 6-4. Reset Filter Specification for nRESET
Parameter
Min
Typ
Max
Unit
Reset filter time constant
2.1
12.0
16.0
µs
Reset pulse width to guarantee a reset
26.0
0
µs
µs
Reset pulse width guaranteed not to cause a reset
1.0
6.2.1.3
Watchdog Reset
The EM35x contains a watchdog timer (see also the Watchdog Timer section) that is clocked by the internal
1 kHz timing reference. When the timer expires it generates the reset source WATCHDOG_RESET to the Reset
Generation module.
6.2.1.4
Software Reset
The ARM® CortexTM-M3 CPU can initiate a reset under software control. This is indicated with the reset source
SYSRESETREQ to the Reset Generation module.
6.2.1.5
Option Byte Error
The flash memory controller contains a state machine that reads configuration information from the
information blocks in the flash at system start time. An error check is performed on the option bytes that are
read from flash and, if the check fails, an error is signaled that provides the reset source OPT_BYTE_ERROR to
the Reset Generation module.
If an option byte error is detected, the system restarts and the read and check process is repeated. If the
error is detected again the process is repeated but stops on the 3rd failure. The system is then placed into an
emulated deep sleep where recovery is possible. In this state, flash memory readout protection is forced
active to prevent secure applications from being compromised.
6.2.1.6
Debug Reset
The Serial Wire/JTAG Interface (SWJ) provides access to the SWJ Debug Port (SWJ-DP) registers. By setting
the register bit CDBGRSTREQ in the SWJ-DP, the reset source CDBGRSTREQ is provided to the Reset
Generation module.
6.2.1.7
JRST
One of the EM35x’s pins can function as the JTAG reset, conforming to the requirements of the JTAG
standard. This input acts independently of all other reset sources and, when asserted, does not reset any on-
chip hardware except for the JTAG TAP. If the EM35x is in the Serial Wire mode or if the SWJ is disabled, this
input has no effect.
6-4
120-035X-000 Rev. 1.2
Final
EM351 / EM357
6.2.1.8
Deep Sleep Reset
The Power Management module informs the Reset Generation module of entry into and exit from the deep
sleep states. The deep sleep reset is applied in the following states: before entry into deep sleep, while
removing power from the memory and core domain, while in deep sleep, while waking from deep sleep, and
while reapplying power until reliable power levels have been detect by POR LV.
The Power Management module allows a special emulated deep sleep state that retains memory and core
domain power while in deep sleep.
6.2.2
Reset Recording
The EM35x records the last reset condition that generated a restart to the system. The reset conditions
recorded are:
POR HV
always-on domain power supply failure
.
.
.
.
.
.
.
POR LV
core domain (POR LVcore) or memory domain (POR LVmem) power supply failure
pin reset asserted
nRESET
watchdog
watchdog timer expired
software reset by SYSERSETREQ from ARM® CortexTM-M3 CPU
SYSRESETREQ
deep sleep wakeup
option byte error
wake-up from deep sleep
error check failed when reading option bytes from flash
Note: While CPU Lockup is shown as a reset condition in software, CPU Lockup is not specifically a reset
event. CPU Lockup is set to indicate that the CPU entered an unrecoverable exception. Execution stops but a
reset is not applied. This is so that a debugger can interpret the cause of the error. Silicon Labs recommends
that in a live application (in other words, no debugger attached) the watchdog be enabled by default so that
the EM35x can be restarted.
6.2.3
Reset Generation Module
The Reset Generation module responds to reset sources and generates the following reset signals:
PORESET
SYSRESET
Reset of the ARM® CortexTM-M3 CPU and ARM® CortexTM-M3 System Debug components
(Flash Patch and Breakpoint, Data Watchpoint and Trace, Instrumentation Trace
Macrocell, Nested Vectored Interrupt Controller). ARM defines PORESET as the region
that is reset when power is applied.
Reset of the ARM® CortexTM-M3 CPU without resetting the Core Debug and System Debug
components, so that a live system can be reset without disturbing the debug
configuration.
.
.
DAPRESET
PRESETHV
Reset to the SWJ’s AHB Access Port (AHB-AP)
.
.
Peripheral reset for always-on power domain, for peripherals that are required to retain
their configuration across a deep sleep cycle
PRESETLV
Peripheral reset for core power domain, for peripherals that are not required to retain
their configuration across a deep sleep cycle
.
6-5
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 6-5 shows which reset sources generate certain resets.
Table 6-5. Generated Resets
Reset Generation Module Output
Reset Source
PORESET
SYSRESET
DAPRESET
PRESETHV
PRESETLV
POR HV
X
X
X
X
X
POR LV (due to waking from
normal deep sleep)
X
X
X
X
X
X
POR LV (not due to waking from
normal deep sleep)
X
X
X
X
nRESET
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Watchdog
SYSRESETREQ
Option byte error
Normal deep sleep
Emulated deep sleep
Debug reset
X
X
X
6.3 Clocks
The EM35x integrates four oscillators:
12 MHz RC oscillator
.
24 MHz crystal oscillator
.
10 kHz RC oscillator
.
.
32.768 kHz crystal oscillator
Figure 6-2 shows a block diagram of the clocks in the EM35x. This simplified view shows all the clock sources
and the general areas of the chip to which they are routed.
6-6
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 6-2. Clocks Block Diagram
OSC24M_CTRL[0]
12MHz
RC
Failover monitor
(selects RC when
XTAL fails)
OSCHF
SYSCLK
/2
PCLK
oscillator
OSC24M
24MHz
XTAL
ADC
OSC24M_CTRL[1]
SigmaDelta
Produces 6MHz
or 1MHz
10kHz
RC
/N
OSCRC
CLK1K
ADC_CFG[2]
(nominal 10)
CPU_CLKSEL[0]
oscillator
OSC32K
32kHz
XTAL
FLITF
bus Flash
32kHz
digital in
SLEEPTMR_CLKEN[0]
FCLK
CPU
bus RAM
RAM CTRL
Watchdog
counter
SysTick
counter
Sleep Timer
counter
ST_CSR[2]
/(2^N)
SLEEPTMR_CFG[7:4]
MAC Timer
counter
SLEEPTMR_CFG[0]
TIMx
counter
SCx
RATEGEN
TIMxCLK
digital in
SCxSCLK
digital i/o
TIMx_SMCR[2:0]
TIMx_OR[1:0]
DEBUG_EMCR[24]
AND
TRACECLK
digital out
/2
6-7
120-035X-000 Rev. 1.2
Final
EM351 / EM357
6.3.1
High-Frequency Internal RC Oscillator (OSCHF)
The high-frequency RC oscillator (OSCHF) is used as the default system clock source when power is applied to
the core domain. The nominal frequency coming out of reset is 12 MHz and Ember software calibrates this
clock to 12 MHz. Table 6-6 contains the specification for the high frequency RC oscillator.
Most peripherals, excluding the radio peripheral, are fully functional using the OSCHF clock source.
Application software must be aware that peripherals are clocked at different speeds depending on whether
OSCHF or OSC24M is being used. Since the frequency step of OSCHF is 0.3 MHz and the high-frequency crystal
oscillator is used for calibration, the calibrated accuracy of OSCHF is ±150 kHz ±40 ppm. The UART and ADC
peripherals may not be usable due to the lower accuracy of the OSCHF frequency.
Table 6-6. High-Frequency RC Oscillator Specification
Parameter
Test conditions
Min
Typ
Max
Unit
Frequency at reset
6
12
20
MHz
Frequency Steps
Duty cycle
0.3
MHz
%
40
60
5
Supply dependence
Change in supply = 0.1 V
%
Test at supply changes: 1.8 V to 1.7 V
6.3.2
High-Frequency Crystal Oscillator (OSC24M)
The high-frequency crystal oscillator (OSC24M) requires an external 24 MHz crystal with an accuracy of
±40 ppm. Based upon the application’s bill of materials and current consumption requirements, the external
crystal may cover a range of ESR requirements. Table 6-7 contains the specification for the high frequency
crystal oscillator.
The crystal oscillator has a software-programmable bias circuit to minimize current consumption. Ember
software configures the bias circuit for minimum current consumption.
All peripherals including the radio peripheral are fully functional using the OSC24M clock source. Application
software must be aware that peripherals are clocked at different speeds depending on whether OSCHF or
OSC24M is being used.
If the 24 MHz crystal fails, a hardware failover mechanism forces the system to switch back to the high-
frequency RC oscillator as the main clock source, and a non-maskable interrupt (NMI) is signaled to the ARM®
CortexTM-M3 NVIC.
Table 6-7. High-Frequency Crystal Oscillator Specification
Parameter
Frequency
Accuracy
Test conditions
Min
Typ
Max
Unit
MHz
ppm
24
-40
40
+40
60
1
Duty cycle
%
Start-up time at max bias
Start up time at optimal bias
Current consumption
ms
ms
μA
2
200
300
6-8
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Parameter
Test conditions
Min
Typ
Max
Unit
Current consumption at max bias
1
mA
Crystal with high ESR
Load capacitance
100
10
7
Ω
pF
pF
µW
Ω
Crystal capacitance
Crystal power dissipation
Crystal with low ESR
Load capacitance
200
60
18
7
pF
pF
mW
Crystal capacitance
Crystal power dissipation
1
6.3.3
Low-Frequency Internal RC Oscillator (OSCRC)
A low-frequency RC oscillator (OSCRC) is provided as an internal timing reference. The nominal frequency
coming out of reset is 10 kHz, and Ember software calibrates this clock to 10 kHz. From the tuned 10 kHz
oscillator (OSCRC) Ember software calibrates a fractional-N divider to produce a 1 kHz reference clock, CLK1K.
Table 6-8 contains the specification for the low frequency RC oscillator.
Table 6-8. Low-Frequency RC Oscillator Specification
Parameter
Test conditions
Min Typ Max Unit
Nominal Frequency
After trimming
9
10
0.5
1
11
kHz
kHz
%
Analog trim step size
Supply dependence
For a voltage drop from 3.6 V to 3.1 V or 2.6 V to 2.1 V
(without re-calibration)
Temperature
dependence
Frequency variation with temperature for a change
from -40 oC to +85oC
2
%
(without re-calibration)
6.3.4
Low-Frequency Crystal Oscillator (OSC32K)
A low-frequency 32.768 kHz crystal oscillator (OSC32K) is provided as an optional timing reference for on-chip
timers. This oscillator is designed for use with an external watch crystal. When using the 32.768 kHz crystal,
you must connect it to GPIO PC6 and PC7, and must configure these two GPIOs for analog input. Alternatively,
when PC7 is configured as a digital input, PC7 can accept an external digital clock input instead of a
32.786 kHz crystal. The digital clock input signal must be a 1 V peak-to-peak sine wave with a DC bias of
0.5 V. Refer to Chapter 7, GPIO for GPIO configuration details. Using the low-frequency oscillator, crystal or
digital clock, is enabled through Ember software.
6-9
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 6-9 contains the specification for the low frequency crystal oscillator.
Table 6-9. Low-Frequency Crystal Oscillator Specification
Parameter
Test conditions
Min
Typ
Max
Unit
Frequency
32.768
kHz
Accuracy
At 25ºC
-20
+20
ppm
pF
pF
kΩ
s
Load capacitance OSC32A
Load capacitance OSC32B
Crystal ESR
27
18
100
2
Start-up time
Current consumption
At 25°C, VDD_PADS=3.0 V
0.5
μA
6.3.5
Clock Switching
The EM35x has two switching mechanisms for the main system clock, providing four clock modes. Table 6-10
shows these clock modes and how they affect the internal clocks.
The register bit OSC24M_CTRL_OSC24M_SEL in the OSC24M_CTRL register switches between the high-
frequency RC oscillator (OSCHF) and the high-frequency crystal oscillator (OSC24M) as the main system clock
(SYSCLK). The peripheral clock (PCLK) is always half the frequency of SYSCLK.
The register bit CPU_CLKSEL_FIELD in the CPU_CLKSEL register switches between PCLK and SYSCLK to produce
the ARM® CortexTM-M3 CPU clock (FCLK). The default and preferred mode of operation is to run the CPU at the
higher PCLK frequency, 24 MHz, to give higher processing performance for all applications and improved duty
cycling for applications using sleep modes.
In addition to these modes, further automatic control is invoked by hardware when flash programming is
enabled. To ensure accuracy of the flash controller’s timers, the FCLK frequency is forced to 12 MHz during
flash programming and erase operations.
Table 6-10. System Clock Modes
FCLK
OSC24M_CTRL_ CPU_CLKSEL_FI
SYSCLK
PCLK
Flash Program/Erase
Inactive
Flash Program/Erase
Active
OSC24M_SEL
ELD
0 (OSCHF)
0 (OSCHF)
1 (OSC24M)
1 (OSC24M)
0 (Normal CPU)
1 (Fast CPU)
12 MHz
12 MHz
24 MHz
24 MHz
6 MHz
6 MHz
6 MHz
12 MHz
12 MHz
24 MHz
12 MHz
12 MHz
12 MHz
12 MHz
0 (Normal CPU)
1 (Fast CPU)
12 MHz
12 MHz
6-10
120-035X-000 Rev. 1.2
Final
EM351 / EM357
6.4 System Timers
6.4.1
Watchdog Timer
The EM35x integrates a watchdog timer which can be enabled to provide protection against software crashes
and ARM® CortexTM-M3 CPU lockup. By default, it is disabled at power up of the always-on power domain. The
watchdog timer uses the calibrated 1 kHz clock (CLK1K) as its reference and provides a nominal 2.048 s
timeout. A low water mark interrupt occurs at 1.792 s and triggers an NMI to the ARM® CortexTM-M3 NVIC as an
early warning. When the watchdog is enabled, the timer must be periodically reset before it expires. The
watchdog timer is paused when the debugger halts the ARM® CortexTM-M3. Additionally, the Ember software
that implements deep sleep functionality disables the watchdog when entering deep sleep and restores the
watchdog, if it was enabled, when exiting deep sleep.
Ember software provides an API for enabling, resetting, and disabling the watchdog timer.
6.4.2
Sleep Timer
The EM35x integrates a 32-bit timer dedicated to system timing and waking from sleep at specific times. The
sleep timer can use either the calibrated 1 kHz reference (CLK1K), or the 32 kHz crystal clock (CLK32K). The
default clock source is the internal 1 kHz clock.
The sleep timer has a prescaler, a divider of the form 2^N, where N can be programmed from 1 to 2^15. This
divider allows for very long periods of sleep to be timed. Ember software’s default configuration is to use the
prescaler to always produce a 1024 Hz sleep timer tick. The timer provides two compare outputs and wrap
detection, all of which can be used to generate an interrupt or a wake up event.
While it is possible to do so, by default the sleep timer is not paused when the debugger halts the ARM®
CortexTM-M3. Silicon Labs does not advise pausing the sleep timer when the debugger halts the CPU.
To save current during deep sleep, the low-frequency internal RC oscillator (OSCRC) can be turned off. If
OSCRC is turned off during deep sleep and a low-frequency 32.768 kHz crystal oscillator is not being used,
then the sleep timer will not operate during deep sleep and sleep timer wake events cannot be used to wake
up the EM35x.
Ember software provides the system timer software API for interacting with the sleep timer as well as using
the sleep timer and RC oscillator during deep sleep.
Note: Because the system timer software module handles all interactions with the sleep timer, the module
will return the correct value in all situations. In the situation where the chip performs a deep sleep that
maintains the system time and is woken up from an external event (that is, not a sleep timer event), the deep
sleep module in the Ember software delays until the next sleep timer clock tick (up to 1 ms) to guarantee that
the sleep timer updates correctly.
6.4.3
Event Timer
The SysTick timer is an ARM® standard system timer in the NVIC. The SysTick timer can be clocked from either
the FCLK (the clock going into the CPU) or the Sleep Timer clock. FCLK is either the SYSCLK or PCLK as
selected by CPU_CLKSEL register (see the Clock Switching section).
6.5 Power Management
The EM35x’s power management system is designed to achieve the lowest deep sleep current consumption
possible while still providing flexible wakeup sources, timer activity, and debugger operation. The EM35x has
four main sleep modes:
Idle Sleep: Puts the CPU into an idle state where execution is suspended until any interrupt occurs. All
power domains remain fully powered and nothing is reset.
.
6-11
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Deep Sleep 1: The primary deep sleep state. In this state, the core power domain is fully powered down
and the sleep timer is active.
.
.
.
Deep Sleep 2: The same as Deep Sleep 1 except that the sleep timer is inactive to save power. In this
mode the sleep timer cannot wake up the EM35x.
Deep Sleep 0 (also known as Emulated Deep Sleep): The chip emulates a true deep sleep without powering
down the core domain. Instead, the core domain remains powered and all peripherals except the system
debug components (ITM, DWT, FPB, NVIC) are held in reset. The purpose of this sleep state is to allow
EM35x software to perform a deep sleep cycle while maintaining debug configuration such as breakpoints.
CSYSPWRUPREQ, CDBGPWRUPREQ, and the corresponding CSYSPWRUPACK and CDBGPWRUPACK are bits in the
debug port’s CTRL/STAT register in the SWJ. For further information on these bits and the operation of the
SWJ-DP please refer to the ARM Debug Interface v5 Architecture Specification (ARM IHI 0031A).
For further power savings when not in deep sleep, the ADC, Timer 1, Timer 2, Serial Controller 1, and Serial
Controller 2 peripherals can be individually disabled through the PERIPHERAL_DISABLE register. Disabling a
peripheral saves power by stopping the clock feeding that peripheral. A peripheral should only be disabled
through the PERIPHERAL_DISABLE register when the peripheral is idle and disabled through the peripheral's
own configuration registers, otherwise undefined behavior may occur. When a peripheral is disabled through
the PERIPHERAL_DISABLE register, all registers associated with that peripheral ignore all subsequent writes,
and subsequent reads return the value seen in the register at the moment the peripheral is disabled.
6.5.1
Wake Sources
When in deep sleep the EM35x can be returned to the running state in a number of ways, and the wake
sources are split depending on deep sleep 1 or deep sleep 2.
The following wake sources are available in both deep sleep 1 and 2.
Wake on GPIO activity: Wake due to change of state on any GPIO.
.
Wake on serial controller 1: Wake due to a change of state on GPIO Pin PB2.
.
Wake on serial controller 2: Wake due to a change of state on GPIO Pin PA2.
.
.
Wake on IRQD: Wake due to a change of state on IRQD. Since IRQD can be configured to point to any GPIO,
this wake source is another means of waking on any GPIO activity.
Wake on setting of CDBGPWRUPREQ: Wake due to setting the CDBGPWRUPREQ bit in the debug port in the
SWJ.
.
Wake on setting of CSYSPWRUPREQ: Wake due to setting the CSYSPWRUPREQ bit in the debug port in the
SWJ.
.
The following sources are only available in deep sleep 1 since the sleep timer is not active in deep sleep 2.
Wake on sleep timer compare A.
.
Wake on sleep timer compare B.
.
Wake on sleep timer wrap.
.
The following source is only available in deep sleep 0 since the SWJ is required to write a memory mapped
register to set this wake source and the SWJ only has access to some registers in deep sleep 0.
Wake on write to the WAKE_CORE register bit.
.
The Wakeup Recording module monitors all possible wakeup sources. More than one wakeup source may be
recorded because events are continually being recorded (not just in deep-sleep) and another event may
happen between the first wake event and when the EM35x wakes up.
6-12
120-035X-000 Rev. 1.2
Final
EM351 / EM357
6.5.2
Basic Sleep Modes
The power management state diagram in Figure 6-3 shows the basic operation of the power management
controller.
Figure 6-3. Power Management State Diagram
CDBGPWRUPREQ set
EMULATED
DEEP SLEEP
DEEP SLEEP
CDBGPWRUPREQ cleared
1
=
=
0
Wake up event
resets the processor
(
CDBGPWRUPREQ
CSYSPWRUPREQ
&
)
Deep sleep requested
(WFI instruction with SLEEP_DEEP=1)
PRE-DEEP
SLEEP
RUNNING
CSYSPWRUPREQ
IDLE SLEEP
&
INHIBIT
Interrupt
In normal operation an application may request one of two low power modes through program execution:
Idle Sleep is achieved by executing a WFI instruction while the SLEEPDEEP bit in the Cortex System Control
.
register (SCS_SCR) is clear. This puts the CPU into an idle state where execution is suspended until an
interrupt occurs. This is indicated by the state at the bottom of the diagram. Power is maintained to the
core logic of the EM35x during the Idle Sleeping state.
Deep sleep is achieved by executing a WFI instruction with the SLEEPDEEP bit in SCS_SCR set. This triggers
.
the state transitions around the main loop of the diagram, resulting in powering down the EM35x’s core
logic, and leaving only the always-on domain powered. Wake up is triggered when one of the pre-
determined events occurs.
If a deep sleep is requested the EM35x first enters a pre-deep sleep state. This state prevents any section of
the chip from being powered off or reset until the SWJ goes idle (by clearing CSYSPWRUPREQ). This pre-deep
sleep state ensures debug operations are not interrupted.
In the deep sleep state the EM35x waits for a wake up event which will return it to the running state. In
powering up the core logic the ARM® CortexTM-M3 is put through a reset cycle and Ember software restores the
stack and application state to the point where deep sleep was invoked.
6-13
120-035X-000 Rev. 1.2
Final
EM351 / EM357
6.5.3
Further options for deep sleep
By default the low-frequency internal RC oscillator (OSCRC) is running during deep sleep (known as deep
sleep 1).
To conserver power, OSCRC can be turned of during deep sleep. This mode is known as deep sleep 2. Since the
OSCRC is disabled, the sleep timer and watchdog timer do not function and cannot wake the chip unless the
low-frequency 32.768 kHz crystal oscillator is used. Non-timer based wake sources continue to function. Once
a wake event does occur, OSCRC is restarted and comes back up.
6.5.4
Use of debugger with sleep modes
The debugger communicates with the EM35x using the SWJ.
When the debugger is logically connected, the CDBGPWRUPREQ bit in the debug port in the SWJ is set, and
the EM35x will only enter deep sleep 0 (the Emulated Deep Sleep state). The CDBGPWRUPREQ bit indicates
that a debug tool is logically connected to the chip and therefore debug state may be in the system debug
components. To maintain the debug state in the system debug components only deep sleep 0 may be used,
since deep sleep 0 will not cause a power cycle or reset of the core domain. The CSYSPWRUPREQ bit in the
debug port in the SWJ indicates that a debugger wants to access memory actively in the EM35x. Therefore,
whenever the CSYSPWRUPREQ bit is set while the EM35x is awake, the EM35x cannot enter deep sleep until
this bit is cleared. This ensures the EM35x does not disrupt debug communication into memory.
Clearing both CSYSPWRUPREQ and CDBGPWRUPREQ allows the EM35x to achieve a true deep sleep state (deep
sleep 1 or 2). Both of these signals also operate as wake sources, so that when a debugger logically connects
to the EM35x and begins accessing the chip, the EM35x automatically comes out of deep sleep. When the
debugger initiates access while the EM35x is in deep sleep, the SWJ intelligently holds off the debugger for a
brief period of time until the EM35x is properly powered and ready.
Note: The SWJ-DP signals CSYSPWRUPREQ and CDBGPWRUPREQ are only reset by a power-on-reset or a
debugger. Physically connecting or disconnecting a debugger from the chip will not alter the state of these
signals. A debugger must logically communicate with the SWJ-DP to set or clear these two signals.
For more information regarding the SWJ and the interaction of debuggers with deep sleep, contact customer
support for Application Notes and ARM® CoreSightTM documentation.
6-14
120-035X-000 Rev. 1.2
Final
EM351 / EM357
6.5.5
Registers
PERIPHERAL_DISABLE
Peripheral Disable Register
Address: 0x40004038 Reset: 0x0
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
0
22
0
21
0
20
19
18
17
16
0
0
0
0
0
15
0
14
0
13
0
12
11
10
9
8
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
PERIDIS_RSVD
PERIDIS_ADC
PERIDIS_TIM2
PERIDIS_TIM1
PERIDIS_SC1
PERIDIS_SC2
Bitname
Bitfield
Access
Description
PERIDIS_RSVD
[5]
RW
Reserved: this bit can change during normal operation. When writing to
PERIPHERAL_DISABLE, the value of this bit must be preserved.
PERIDIS_ADC
PERIDIS_TIM2
PERIDIS_TIM1
PERIDIS_SC1
PERIDIS_SC2
[4]
[3]
[2]
[1]
[0]
RW
RW
RW
RW
RW
Disable the clock to the ADC peripheral.
Disable the clock to the TIM2 peripheral.
Disable the clock to the TIM1 peripheral.
Disable the clock to the SC1 peripheral.
Disable the clock to the SC2 peripheral.
6.6 Security Accelerator
The EM35x contains a hardware AES encryption engine accessible from the ARM® CortexTM-M3. NIST-based
CCM, CCM*, CBC-MAC, and CTR modes are implemented in hardware. These modes are described in the IEEE
802.15.4-2003 specification, with the exception of CCM*, which is described in the ZigBee Security Services
Specification 1.0.
6-15
120-035X-000 Rev. 1.2
Final
EM351 / EM357
7
GPIO (General Purpose Input / Output)
The EM35x has 24 multi-purpose GPIO pins, which may be individually configured as:
General purpose output
.
General purpose open-drain output
.
Alternate output controlled by a peripheral device
.
Alternate open-drain output controlled by a peripheral device
.
Analog
.
General purpose input
.
General purpose input with pull-up or pull-down resistor
.
The basic structure of a single GPIO is illustrated in Figure 7-1.
Figure 7-1. GPIO Block Diagram
GPIO_PxCFGH/L
VDD_PADS
GPIO_PxSET
P-MOS
Output control
GPIO_PxOUT
GPIO_PxCLR
VDD_PADS
(push pull,
open drain, or
disabled)
VDD_PADS
Protection
N-MOS
GND
diode
Alternate output
PIN
Alternate input
Protection
diode
GPIO_PxIN
Analog
functions
GND
GND
Schmitt trigger
Wake detection
GPIO_PxWAKE
A Schmitt trigger converts the GPIO pin voltage to a digital input value. The digital input signal is then always
routed to the GPIO_PxIN register; to the alternate inputs of associated peripheral devices; to wake detection
logic if wake detection is enabled; and, for certain pins, to interrupt generation logic. Configuring a pin in
analog mode disconnects the digital input from the pin and applies a high logic level to the input of the
Schmitt trigger.
Only one device at a time can control a GPIO output. The output is controlled in normal output mode by the
GPIO_PxOUT register and in alternate output mode by a peripheral device. When in input mode or analog
mode, digital output is disabled.
7.1
GPIO Ports
The 24 GPIO pins are grouped into three ports: PA, PB, and PC. Individual GPIOs within a port are numbered 0
to 7 according to their bit positions within the GPIO registers.
Note: Because GPIO port registers’ functions are identical, the notation Px is used here to refer to PA, PB, or
PC. For example, GPIO_PxIN refers to the registers GPIO_PAIN, GPIO_PBIN, and GPIO_PCIN.
7-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Each of the three GPIO ports has the following registers whose low-order eight bits correspond to the port’s
eight GPIO pins:
GPIO_PxIN (input data register) returns the pin level (unless in analog mode).
.
GPIO_PxOUT (output data register) controls the output level in normal output mode.
.
GPIO_PxCLR (clear output data register) clears bits in GPIO_PxOUT.
.
GPIO_PxSET (set output data register) sets bits in GPIO_PxOUT.
.
GPIO_PxWAKE (wake monitor register) specifies the pins that can wake the EM35x.
.
In addition to these registers, each port has a pair of configuration registers, GPIO_PxCFGH and GPIO_PxCFGL.
These registers specify the basic operating mode for the port’s pins. GPIO_PxCFGL configures the pins Px[3:0]
and GPIO_PxCFGH configures the pins Px[7:4]. For brevity, the notation GPIO_PxCFGH/L refers to the pair of
configuration registers.
Five GPIO pins (PA6, PA7, PB6, PB7 and PC0) can sink and source higher current than standard GPIO outputs.
Refer to Table 2-5, Digital I/O Specifications in Chapter 2, Electrical Characteristics, for more information.
7.2
Configuration
Each pin has a 4-bit configuration value in the GPIO_PxCFGH/L register. The various GPIO modes and their
4-bit configuration values are shown in Table 7-1.
Table 7-1. GPIO Configuration Modes
GPIO Mode
GPIO_PxCFGH/L Description
Analog
0x0
0x4
0x8
Analog input or output. When in analog mode, the digital input
(GPIO_PxIN) always reads 1.
Input (floating)
Digital input without an internal pull up or pull down. Output is
disabled.
Input (pull-up or
pull-down)
Digital input with an internal pull up or pull down. A set bit in
GPIO_PxOUT selects pull up and a cleared bit selects pull down.
Output is disabled.
Output (push-
pull)
0x1
0x5
0x9
0xD
Push-pull output. GPIO_PxOUT controls the output.
Output (open-
drain)
Open-drain output. GPIO_PxOUT controls the output. If a pull up is
required, it must be external.
Alternate Output
(push-pull)
Push-pull output. An onboard peripheral controls the output.
Alternate Output
(open-drain)
Open-drain output. An onboard peripheral controls the output. If a
pull up is required, it must be external.
If a GPIO has two peripherals that can be the source of alternate output mode data, then other registers in
addition to GPIO_PxCFGH/L determine which peripheral controls the output.
Several GPIOs share an alternate output with Timer 2 and the Serial Controllers. Bits in Timer 2’s TIM2_OR
register control routing Timer 2 outputs to different GPIOs. Bits in Timer 2’s TIM2_CCER register enable Timer
2 outputs. When Timer 2 outputs are enabled they override Serial Controller outputs. Table 7-2 indicates the
GPIO mapping for Timer 2 outputs depending on the bits in the register TIM2_OR. Refer to Chapter 9, General
Purpose Timers for complete information on timer configuration.
7-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 7-2. Timer 2 Output Configuration Controls
GPIO Mapping Selected by TIM2_OR Bit
Timer 2 Output
Option Register Bit
0
1
TIM2C1
TIM2_OR[4]
PA0
PB1
TIM2C2
TIM2C3
TIM2C4
TIM2_OR[5]
TIM2_OR[6]
TIM2_OR[7]
PA3
PA1
PA2
PB2
PB3
PB4
For outputs assigned to the serial controllers, the serial interface mode registers (SCx_MODE) determine how
the GPIO pins are used.
The alternate outputs of PA4 and PA5 can either provide packet trace data (PTI_EN and PTI_DATA), or
synchronous CPU trace data (TRACEDATA2 and TRACEDATA3). The selection of packet trace or CPU trace is
made through the Ember software.
If a GPIO does not have an associated peripheral in alternate output mode, its output is set to 0.
7.3
Forced Functions
For some GPIOs the GPIO_PxCFGH/L configuration will be overridden. These functions are forced when the
EM35x is reset and remain forced until software overrides the forced functions. Table 7-3 shows the GPIOs
that have different functions forced on them regardless of the GPIO_PxCFGH/L registers.
Table 7-3. GPIO Forced Functions
GPIO Forced Mode
Forced Signal
PA7
PC0
PC2
PC3
Open-drain output
Input with pull up
Push-pull output
Input with pull up
REG_EN
JRST
JTDO
JDTI
PC41 Input with pull up
JTMS
PC41 Bidirectional (push-pull output or floating input) controlled by debugger interface SWDIO
1 The choice of PC4’s forced signal is controlled by an external debug tool. JTMS is forced when the SWJ is in JTAG mode
and SWDIO is forced when the SWJ is in Serial Wire mode.
PA7 is forced to be the regulator enable signal, REG_EN. If an external regulator is used and controlled
through REG_EN, PA7’s forced functionality must not be overridden. If an external regulator is not used,
REG_EN may be disabled and PA7 may be reclaimed as a normal GPIO. Disabling REG_EN is done by clearing
the bit GPIO_EXTREGEN in the GPIO_DBGCFG register.
PC0, PC2, PC3, and PC4 are forced to be the Serial Wire and JTAG (SWJ) Interface. When the EM35x resets,
these four GPIOs are forced to operate in JTAG mode. Switching the debug interface between JTAG mode and
Serial Wire mode can only be accomplished by the external debug tool and cannot be affected by software
executing on the EM35x. Due to the fact that Serial Wire mode can only be invoked by an external debug tool
7-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
and JTAG mode is forced when the EM35x resets, a designer must treat all four debug GPIOs as working in
unison even though the Serial Wire interface only uses one of the GPIO, PC4.
Note: An application must disable all debug SWJ debug functionality to reclaim any of the four GPIOs: PC0,
PC2, PC3, and PC4. Disabling SWJ debug functionality prevents external debug tools from operating, including
flash programming and high-level debug tools.
Disabling the SWJ debugger interface is accomplished by setting the GPIO_DEBUGDIS bit in the GPIO_DBGCFG
register. When this bit is set, all debugger-related pins (PC0, PC2, PC3, PC4) behave as standard GPIOs. If the
SWJ debugger interface is already active, the bit GPIO_DEBUGDIS cannot be set. When GPIO_DEBUGDIS is set,
the SWJ debugger interface can be reclaimed by activating the SWJ while the EM35x is held in reset. If the
SWJ debugger interface is forced active in this manner, the bit GPIO_FORCEDBG is set in the GPIO_DBGSTAT
register. The SWJ debugger interface is defined as active when the CDBGPWRUPREQ signal, a bit in the debug
port’s CRTL/STAT register in the SWJ, is set high by an external debug tool.
7.4
Reset
A full chip reset is one due to power on (low or high voltage), the nRESET pin, the watchdog, or the
SYSRESETREQ bit. A full chip reset affects the GPIO configuration as follows:
The GPIO_PxCFGH/L configurations of all pins are configured as floating inputs.
.
.
The GPIO_EXTREGEN bit is set in the GPIO_DBGCFG register, which overrides the normal configuration for
PA7.
The GPIO_DEBUGDIS bit in the GPIO_DBGCFG register is cleared, allowing Serial Wire/JTAG access to
override the normal configuration of PC0, PC2, PC3, and PC4.
.
7.5
Boot Configuration
nBOOTMODE is a special alternate function of PA5 that is active only during a pin reset (nRESET) or a power-
on-reset of the always-powered domain (POR HV). If nBOOTMODE is asserted (pulled or driven low) when
coming out of reset, the processor starts executing an embedded serial-link-only monitor instead of its normal
program.
While in reset and during the subsequent power-on-reset startup delay (512 OSCHF clocks), PA5 is
automatically configured as an input with a pull-up resistor. At the end of this time, the EM35x samples
nBOOTMODE: a high level selects normal boot mode, and a low level selects the embedded monitor. Figure 7-2
shows the timing parameters for invoking monitor mode from a pin (nRESET) reset. Because OSCHF is running
uncalibrated during the reset sequence, the time for 512 OSCHF clocks may vary as indicated.
Figure 7-2. nBOOTMODE and nRESET Timing
26 μsec min
.
.
.
. . .
nRESET
512 clocks;
26 μsec min – 85 μsec max
OSCHF
.
.
.
.
.
.
. . .
nBOOTMODE Sampled;
FIB Monitor mode entered
nBOOTMODE
.
. .
nBOOTMODE Sampled by
FIB Monitor code
7-4
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Timing for a power-on-reset is similar except that OSCHF does not begin oscillating until up to 70 µsec after
both core and HV supplies are valid. Combined with the maximum 250 µsec allowed for HV to ramp from 0.5 V
to 1.7 V, an additional 320 µsec may be added to the 512 OSCHF clocks until nBOOTMODE is sampled.
If the monitor mode is selected (nBOOTMODE is low after 512 clocks), the FIB monitor software begins
execution. In order to filter out inadvertent jumps into the monitor, the FIB monitor re-samples the
nBOOTMODE signal after a 3 ms delay. If the signal is still low, then the device stays in monitor mode. If the
signal is high, then monitor mode is exited and the normal program begins execution. In summary, the
nBOOTMODE signal must be held low for 4 ms in order to properly invoke the FIB monitor.
After nBOOTMODE has been sampled, PA5 is configured as a floating input like the other GPIO configurations.
The GPIO_BOOTMODE bit in the GPIO_DBGSTAT register captures the state of nBOOTMODE so that software
may act on this signal if required.
Note: To avoid inadvertently asserting nBOOTMODE, PA5’s capacitive load may not exceed 250 pF.
7.6
GPIO Modes
7.6.1 Analog Mode
Analog mode enables analog functions, and disconnects a pin from the digital input and output logic. Only the
following GPIO pins have analog functions:
PA4, PA5, PB5, PB6, PB7, and PC1 can be analog inputs to the ADC.
.
.
PB0 can be an external analog voltage reference input to the ADC, or it can output the internal analog
voltage reference from the ADC. The Ember software selects an internal or external voltage reference.
PC6 and PC7 can connect to an optional 32.768 kHz crystal.
.
Note: When an external timing source is required, a 32.768 kHz crystal is commonly connected to PC6 and
PC7. Alternatively, when PC7 is configured as a digital input, PC7 can accept a digital external clock input.
When configured in analog mode:
The output drivers are disabled.
.
The internal pull-up and pull-down resistors are disabled.
.
The Schmitt trigger input is connected to a high logic level.
.
.
Reading GPIO_PxIN returns a constant 1.
7.6.2 Input Mode
Input mode is used both for general purpose input and for on-chip peripheral inputs. Input floating mode
disables the internal pull-up and pull-down resistors, leaving the pin in a high-impedance state. Input pull-up
or pull-down mode enables either an internal pull-up or pull-down resistor based on the GPIO_PxOUT register.
Setting a bit to 0 in GPIO_PxOUT enables the pull-down and setting a bit to 1 enables the pull up.
When configured in input mode:
The output drivers are disabled.
.
An internal pull-up or pull-down resistor may be activated depending on GPIO_PxCFGH/L and GPIO_PxOUT.
.
The Schmitt trigger input is connected to the pin.
.
Reading GPIO_PxIN returns the input at the pin.
.
The input is also available to on-chip peripherals.
.
7-5
120-035X-000 Rev. 1.2
Final
EM351 / EM357
7.6.3 Output Mode
Output mode provides a general purpose output under direct software control. Regardless of whether an
output is configured as push-pull or open-drain, the GPIO’s bit in the GPIO_PxOUT register controls the
output. The GPIO_PxSET and GPIO_PxCLR registers can atomically set and clear bits within GPIO_PxOUT
register. These set and clear registers simplify software using the output port because they eliminate the need
to disable interrupts to perform an atomic read-modify-write operation of GPIO_PxOUT.
When configured in output mode:
The output drivers are enabled and are controlled by the value written to GPIO_PxOUT:
.
In open-drain mode: 0 activates the N-MOS current sink; 1 tri-states the pin.
•
In push-pull mode: 0 activates the N-MOS current sink; 1 activates the P-MOS current source.
•
The internal pull-up and pull-down resistors are disabled.
.
The Schmitt trigger input is connected to the pin.
.
Reading GPIO_PxIN returns the input at the pin.
.
.
Reading GPIO_PxOUT returns the last value written to the register.
Note: Depending on configuration and usage, GPIO_PxOUT and GPIO_PxIN may not have the same value.
7.6.4 Alternate Output Mode
In this mode, the output is controlled by an on-chip peripheral instead of GPIO_PxOUT and may be configured
as either push-pull or open-drain. Most peripherals require a particular output type – TWI requires an open-
drain driver, for example – but since using a peripheral does not by itself configure a pin, the GPIO_PxCFGH/L
registers must be configured properly for a peripheral’s particular needs. As described in the Configuration
section, when more than one peripheral can be the source of output data, registers in addition to
GPIO_PxCFGH/L determine which to use.
When configured in alternate output mode:
The output drivers are enabled and are controlled by the output of an on-chip peripheral:
.
In open-drain mode: 0 activates the N-MOS current sink; 1 tri-states the pin.
•
In push-pull mode: 0 activates the N-MOS current sink; 1 activates the P-MOS current source.
•
The internal pull-up and pull-down resistors are disabled.
.
The Schmitt trigger input is connected to the pin.
.
Reading GPIO_PxIN returns the input to the pin.
.
Note: Depending on configuration and usage, GPIO_PxOUT and GPIO_PxIN may not have the same value.
7.7
Wake Monitoring
The GPIO_PxWAKE registers specify which GPIOs are monitored to wake the processor. If a GPIO’s wake enable
bit is set in GPIO_PxWAKE, then a change in the logic value of that GPIO causes the EM35x to wake from deep
sleep. The logic values of all GPIOs are captured by hardware upon entering sleep. If any GPIO’s logic value
changes while in sleep and that GPIO’s GPIO_PxWAKE bit is set, then the EM35x wakes from deep sleep.
(There is no mechanism for selecting a specific rising-edge, falling-edge, or level on a GPIO: any change in
logic value triggers a wake event.) Hardware records the fact that GPIO activity caused a wake event, but not
which specific GPIO was responsible. Instead, the Ember software reads the state of the GPIOs on waking to
determine this.
The register GPIO_WAKEFILT contains bits to enable digital filtering of the external wakeup event sources: the
GPIO pins, SC1 activity, SC2 activity, and IRQD. The digital filter operates by taking samples based on the
(nominal) 10 kHz RC oscillator. If three samples in a row all have the same logic value, and this sampled logic
value is different from the logic value seen upon entering sleep, the filter outputs a wakeup event.
7-6
120-035X-000 Rev. 1.2
Final
EM351 / EM357
In order to use GPIO pins to wake the EM35x from deep sleep, the GPIO_WAKE bit in the WAKE_SEL register
must be set. Waking up from GPIO activity does not work with pins configured for analog mode since the
digital logic input is always set to 1 when in analog mode. Refer to Chapter 6, System Modules, for information
on the EM35x’s power management and sleep modes.
7.8
External Interrupts
The EM35x can use up to four external interrupt sources (IRQA, IRQB, IRQC, and IRQD), each with its own top-
level NVIC interrupt vector. Since these external interrupt sources connect to the standard GPIO input path,
an external interrupt pin may simultaneously be used by a peripheral device or even configured as an output.
Analog mode is the only GPIO configuration that is not compatible with using a pin as an external interrupt.
External interrupts have individual triggering and filtering options selected using the registers GPIO_INTCFGA,
GPIO_INTCFGB, GPIO_INTCFGC, and GPIO_INTCFGD. The bit field GPIO_INTMOD of the GPIO_INTCFGx register
enables IRQx’s second-level interrupt and selects the triggering mode: 0 is disabled; 1 for rising edge; 2 for
falling edge; 3 for both edges; 4 for active high level; 5 for active low level. The minimum width needed to
latch an unfiltered external interrupt in both level- and edge-triggered mode is 80 ns. With the digital filter
enabled (the GPIO_INTFILT bit in the GPIO_INTCFGx register is set), the minimum width needed is 450 ns.
The register INT_GPIOFLAG is the second-level interrupt flag register that indicates pending external
interrupts. Writing 1 to a bit in the INT_GPIOFLAG register clears the flag while writing 0 has no effect. If the
interrupt is level-triggered, the flag bit is set again immediately after being cleared if its input is still in the
active state.
Two of the four external interrupts, IRQA and IRQB, have fixed pin assignments. The other two external
interrupts, IRQC and IRQD, can use any GPIO pin. The GPIO_IRQCSEL and GPIO_IRQDSEL registers specify the
GPIO pins assigned to IRQC and IRQD, respectively. Table 7-4 shows how the GPIO_IRQCSEL and GPIO_IRQDSEL
register values select the GPIO pin used for the external interrupt.
Table 7-4. IRQC/D GPIO Selection
GPIO_IRQxSEL
GPIO
GPIO_IRQxSEL
GPIO
GPIO_IRQxSEL
GPIO
0
PA0
8
PB0
16
PC0
1
2
3
4
5
6
7
PA1
PA2
PA3
PA4
PA5
PA6
PA7
9
PB1
PB2
PB3
PB4
PB5
PB6
PB7
17
18
19
20
21
22
23
PC1
PC2
PC3
PC4
PC5
PC6
PC7
10
11
12
13
14
15
In some cases, it may be useful to assign IRQC or IRQD to an input also in use by a peripheral, for example to
generate an interrupt from the slave select signal (nSSEL) in an SPI slave mode interface.
Refer to Chapter 11, Interrupt System, for further information regarding the EM35x interrupt system.
7-7
120-035X-000 Rev. 1.2
Final
EM351 / EM357
7.9
Debug Control and Status
Two GPIO registers are largely concerned with debugger functions. GPIO_DBGCFG can disable debugger
operation, but has other miscellaneous control bits as well. GPIO_DBGSTAT, a read-only register, returns
status related to debugger activity (GPIO_FORCEDBG and GPIO_SWEN), as well a flag (GPIO_BOOTMODE)
indicating whether nBOOTMODE was asserted at the last power-on or nRESET-based reset.
7.10 GPIO Signal Assignment Summary
The GPIO signal assignments are shown in Table 7-5.
7-8
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 7-5. GPIO Signal Assignments
Input
GPIO Analog Alternate Output
Output Current
Drive
PA0
PA1
PA2
PA3
PA4
PA5
PA6
PA7
PB0
PB1
PB2
PB3
PB4
PB5
PB6
PB7
PC0
PC1
PC2
PC3
PC4
PC5
PC6
PC7
TIM2C11, SC2MOSI
TIM2C31, SC2MISO, SC2SDA
TIM2C41, SC2SCLK, SC2SCL
TIM2C21, TRACECLK
PTI_EN, TRACEDATA2
PTI_DATA, TRACEDATA3
TIM1C3
TIM2C11, SC2MOSI
Standard
Standard
Standard
Standard
Standard
Standard
High
TIM2C31, SC2MISO, SC2SDA
TIM2C41, SC2SCLK
TIM2C21, SC2nSSEL
ADC4
ADC5
nBOOTMODE2
TIM1C3
TIM1C4, REG_EN3
TIM1C4
High
VREF
TRACECLK
TIM1CLK, TIM2MSK, IRQA
Standard
Standard
TIM2C14, SC1TXD, SC1MOSI, SC1MISO, SC1SDA TIM2C14, SC1SDA
TIM2C24, SC1SCLK
TIM2C24, SC1MISO, SC1MOSI, SC1SCL, SC1RXD
Standard
Standard
Standard
Standard
High
TIM2C34, SC1SCLK
TIM2C44, SC1nRTS
TIM2C34, SC1SCLK, SC1nCTS
TIM2C44, SC1nSSEL
TIM2CLK, TIM1MSK
TIM1C1, IRQB
TIM1C2
ADC0
ADC1
ADC2
TIM1C1
TIM1C2
High
TRACEDATA1
TRACEDATA0, SWO
JTDO6, SWO
JRST5
High
ADC3
Standard
Standard
Standard
Standard
Standard
Standard
Standard
JTDI5
SWDIO7
SWDIO7, JTMS7
TX_ACTIVE
OSC32B nTX_ACTIVE
OSC32A
OSC32_EXT
1 Default signal assignment (not remapped).
2 Overrides during reset as an input with pull up.
3 Overrides after reset as an open-drain output.
4 Alternate signal assignment (remapped).
5 Overrides in JTAG mode as a input with pull up.
6 Overrides in JTAG mode as a push-pull output.
7 Overrides in Serial Wire mode as either a push-pull output, or a floating input, controlled by the debugger.
7-9
120-035X-000 Rev. 1.2
Final
EM351 / EM357
7.11 Registers
GPIO_PxCFGL
GPIO_PACFGL
Port A Configuration Register (Low)
Address: 0x4000B000 Reset: 0x4444
Address: 0x4000B400 Reset: 0x4444
Address: 0x4000B800 Reset: 0x4444
GPIO_PBCFGL
Port B Configuration Register (Low)
GPIO_PCCFGL
Port C Configuration Register (Low)
Substitute A, B, or C for x in the following detail description.
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
Px3_CFG
Px1_CFG
Px2_CFG
Px0_CFG
7
6
5
4
3
2
1
0
Bitname
Bitfield
Access
Description
Px3_CFG
[15:12]
RW
GPIO configuration control.
0x0: Analog, input or output (GPIO_PxIN always reads 1).
0x1: Output, push-pull (GPIO_PxOUT controls the output).
0x4: Input, floating.
0x5: Output, open-drain (GPIO_PxOUT controls the output).
0x8: Input, pulled up or down (selected by GPIO_PxOUT: 0 = pull-down, 1 = pull-up).
0x9: Alternate output, push-pull (peripheral controls the output).
0xD: Alternate output, open-drain (peripheral controls the output).
Px2_CFG
Px1_CFG
Px0_CFG
[11:8]
[7:4]
[3:0]
RW
RW
RW
GPIO configuration control: see Px3_CFG above.
GPIO configuration control: see Px3_CFG above.
GPIO configuration control: see Px3_CFG above.
7-10
120-035X-000 Rev. 1.2
Final
EM351 / EM357
GPIO_PxCFGH
GPIO_PACFGH
Port A Configuration Register (High)
Address: 0x4000B004 Reset: 0x4444
Address: 0x4000B404 Reset: 0x4444
Address: 0x4000B804 Reset: 0x4444
GPIO_PBCFGH
Port B Configuration Register (High)
GPIO_PCCFGH
Port C Configuration Register (High)
Substitute A, B, or C for x in the following detail description.
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
Px7_CFG
Px5_CFG
Px6_CFG
Px4_CFG
7
6
5
4
3
2
1
0
Bitname
Bitfield
Access
Description
Px7_CFG
[15:12]
RW
GPIO configuration control.
0x0: Analog, input or output (GPIO_PxIN always reads 1).
0x1: Output, push-pull (GPIO_PxOUT controls the output).
0x4: Input, floating.
0x5: Output, open-drain (GPIO_PxOUT controls the output).
0x8: Input, pulled up or down (selected by GPIO_PxOUT: 0 = pull-down, 1 = pull-up).
0x9: Alternate output, push-pull (peripheral controls the output).
0xD: Alternate output, open-drain (peripheral controls the output).
Px6_CFG
Px5_CFG
Px4_CFG
[11:8]
[7:4]
[3:0]
RW
RW
RW
GPIO configuration control: see Px7_CFG above.
GPIO configuration control: see Px7_CFG above.
GPIO configuration control: see Px7_CFG above.
7-11
120-035X-000 Rev. 1.2
Final
EM351 / EM357
GPIO_PxIN
GPIO_PAIN
Port A Input Data Register
Address: 0x4000B008 Reset: 0x0
Address: 0x4000B408 Reset: 0x0
Address: 0x4000B808 Reset: 0x0
GPIO_PBIN
Port B Input Data Register
GPIO_PCIN
Port C Input Data Register
Substitute A, B, or C for x in the following detail description.
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
0
14
0
13
0
12
0
11
0
10
0
9
8
0
0
7
6
5
4
3
2
1
0
Px7
Px6
Px5
Px4
Px3
Px2
Px1
Px0
Bitname
Bitfield
[7]
Access
Description
Px7
Px6
Px5
Px4
Px3
Px2
Px1
Px0
RW
RW
RW
RW
RW
RW
RW
RW
Input level at pin Px7.
Input level at pin Px6.
Input level at pin Px5.
Input level at pin Px4.
Input level at pin Px3.
Input level at pin Px2.
Input level at pin Px1.
Input level at pin Px0.
[6]
[5]
[4]
[3]
[2]
[1]
[0]
7-12
120-035X-000 Rev. 1.2
Final
EM351 / EM357
GPIO_PxOUT
GPIO_PAOUT
Port A Output Data Register
Address: 0x4000B00C Reset: 0x0
Address: 0x4000B40C Reset: 0x0
Address: 0x4000B80C Reset: 0x0
GPIO_PBOUT
Port B Output Data Register
GPIO_PCOUT
Port C Output Data Register
Substitute A, B, or C for x in the following detail description.
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
0
14
0
13
0
12
0
11
0
10
0
9
8
0
0
7
6
5
4
3
2
1
0
Px7
Px6
Px5
Px4
Px3
Px2
Px1
Px0
Bitname
Bitfield
[7]
Access
Description
Px7
Px6
Px5
Px4
Px3
Px2
Px1
Px0
RW
RW
RW
RW
RW
RW
RW
RW
Output data for Px7.
Output data for Px6.
Output data for Px5.
Output data for Px4.
Output data for Px3.
Output data for Px2.
Output data for Px1.
Output data for Px0.
[6]
[5]
[4]
[3]
[2]
[1]
[0]
7-13
120-035X-000 Rev. 1.2
Final
EM351 / EM357
GPIO_PxCLR
GPIO_PACLR
Port A Output Clear Register
Address: 0x4000B014 Reset: 0x0
Address: 0x4000B414 Reset: 0x0
Address: 0x4000B814 Reset: 0x0
GPIO_PBCLR
Port B Output Clear Register
GPIO_PCCLR
Port C Output Clear Register
Substitute A, B, or C for x in the following detail description.
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
0
14
0
13
0
12
0
11
0
10
0
9
8
0
0
7
6
5
4
3
2
1
0
Px7
Px6
Px5
Px4
Px3
Px2
Px1
Px0
Bitname
Bitfield
[7]
Access
Description
Px7
Px6
Px5
Px4
Px3
Px2
Px1
Px0
W
W
W
W
W
W
W
W
Write 1 to clear the output data bit for Px7 (writing 0 has no effect).
Write 1 to clear the output data bit for Px6 (writing 0 has no effect).
Write 1 to clear the output data bit for Px5 (writing 0 has no effect).
Write 1 to clear the output data bit for Px4 (writing 0 has no effect).
Write 1 to clear the output data bit for Px3 (writing 0 has no effect).
Write 1 to clear the output data bit for Px2 (writing 0 has no effect).
Write 1 to clear the output data bit for Px1 (writing 0 has no effect).
Write 1 to clear the output data bit for Px0 (writing 0 has no effect).
[6]
[5]
[4]
[3]
[2]
[1]
[0]
7-14
120-035X-000 Rev. 1.2
Final
EM351 / EM357
GPIO_PxSET
GPIO_PASET
Port A Output Set Register
Address: 0x4000B010 Reset: 0x0
Address: 0x4000B410 Reset: 0x0
Address: 0x4000B810 Reset: 0x0
GPIO_PBSET
Port B Output Set Register
GPIO_PCSET
Port C Output Set Register
Substitute A, B, or C for x in the following detail description.
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
GPIO_PXSETRSVD
7
6
5
4
3
2
1
0
Px7
Px6
Px5
Px4
Px3
Px2
Px1
Px0
Bitname
Bitfield
[15:8]
[7]
Access
Description
Reserved: these bits must be set to 0.
GPIO_PXSETRSVD
W
W
W
W
W
W
W
W
W
Px7
Px6
Px5
Px4
Px3
Px2
Px1
Px0
Write 1 to set the output data bit for Px7 (writing 0 has no effect).
Write 1 to set the output data bit for Px6 (writing 0 has no effect).
Write 1 to set the output data bit for Px5 (writing 0 has no effect).
Write 1 to set the output data bit for Px4 (writing 0 has no effect).
Write 1 to set the output data bit for Px3 (writing 0 has no effect).
Write 1 to set the output data bit for Px2 (writing 0 has no effect).
Write 1 to set the output data bit for Px1 (writing 0 has no effect).
Write 1 to set the output data bit for Px0 (writing 0 has no effect).
[6]
[5]
[4]
[3]
[2]
[1]
[0]
7-15
120-035X-000 Rev. 1.2
Final
EM351 / EM357
GPIO_PxWAKE
GPIO_PAWAKE
Port A Wakeup Monitor Register
Address: 0x4000BC08 Reset: 0x0
Address: 0x4000BC0C Reset: 0x0
Address: 0x4000BC10 Reset: 0x0
GPIO_PBWAKE
Port B Wakeup Monitor Register
GPIO_PCWAKE
Port C Wakeup Monitor Register
Substitute A, B, or C for x in the following detail description.
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
0
14
0
13
0
12
0
11
0
10
0
9
8
0
0
7
6
5
4
3
2
1
0
Px7
Px6
Px5
Px4
Px3
Px2
Px1
Px0
Bitname
Bitfield
[7]
Access
Description
Px7
Px6
Px5
Px4
Px3
Px2
Px1
Px0
RW
RW
RW
RW
RW
RW
RW
RW
Write 1 to enable wakeup monitoring of Px7.
Write 1 to enable wakeup monitoring of Px6.
Write 1 to enable wakeup monitoring of Px5.
Write 1 to enable wakeup monitoring of Px4.
Write 1 to enable wakeup monitoring of Px3.
Write 1 to enable wakeup monitoring of Px2.
Write 1 to enable wakeup monitoring of Px1.
Write 1 to enable wakeup monitoring of Px0.
[6]
[5]
[4]
[3]
[2]
[1]
[0]
7-16
120-035X-000 Rev. 1.2
Final
EM351 / EM357
GPIO_WAKEFILT
GPIO Wakeup Filtering Register
Address: 0x4000BC1C Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
IRQD_WAKE_FILTER SC2_WAKE_FILTER
SC1_WAKE_FILTER GPIO_WAKE_FILTER
Bitname
Bitfield
[3]
Access
Description
IRQD_WAKE_FILTER
SC2_WAKE_FILTER
SC1_WAKE_FILTER
GPIO_WAKE_FILTER
RW
RW
RW
RW
Enable filter on GPIO wakeup source IRQD.
Enable filter on GPIO wakeup source SC2 (PA2).
Enable filter on GPIO wakeup source SC1 (PB2).
[2]
[1]
[0]
Enable filter on GPIO wakeup sources enabled by the GPIO_PnWAKE registers.
7-17
120-035X-000 Rev. 1.2
Final
EM351 / EM357
GPIO_IRQxSEL
GPIO_IRQCSEL
Interrupt C Select Register
Address: 0x4000BC14 Reset: 0xF
Address: 0x4000BC18 Reset: 0x10
GPIO_IRQDSEL
Interrupt D Select Register
Substitute C or D in the detailed description below.
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
22
21
20
19
18
17
16
0
0
0
0
0
0
0
0
15
0
14
0
13
0
12
0
11
0
10
9
0
1
8
0
0
0
2
7
6
5
4
3
0
0
0
SEL_GPIO
Bitname
SEL_GPIO
Bitfield
Access
RW
Description
[4:0]
Pin assigned to IRQx.
0x00: PA0.
0x01: PA1.
0x02: PA2.
0x03: PA3.
0x04: PA4.
0x05: PA5.
0x06: PA6.
0x07: PA7.
0x08: PB0.
0x09: PB1.
0x0A: PB2.
0x0B: PB3.
0x0C: PB4.
0x0D: PB5.
0x0E: PB6.
0x0F: PB7.
0x10: PC0.
0x11: PC1.
0x12: PC2.
0x13: PC3.
0x14: PC4.
0x15: PC5.
0x16: PC6.
0x17: PC7.
0x18 - 0x1F: Reserved.
7-18
120-035X-000 Rev. 1.2
Final
EM351 / EM357
GPIO_INTCFGx
GPIO_INTCFGA
GPIO Interrupt A Configuration Register
Address: 0x4000A860 Reset: 0x0
Address: 0x4000A864 Reset: 0x0
Address: 0x4000A868 Reset: 0x0
Address: 0x4000A86C Reset: 0x0
GPIO_INTCFGB
GPIO Interrupt B Configuration Register
GPIO_INTCFGC
GPIO Interrupt C Configuration Register
GPIO_INTCFGD
GPIO Interrupt D Configuration Register
Substitute A, B, C, or D for x in the following detail description.
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
0
14
0
13
0
12
0
11
0
10
0
9
8
0
GPIO_INTFILT
7
6
5
4
3
2
1
0
GPIO_INTMOD
0
0
0
0
0
Bitname
Bitfield
Access
Description
GPIO_INTFILT
GPIO_INTMOD
[8]
RW
RW
Set this bit to enable digital filtering on IRQx.
[7:5]
IRQx triggering mode.
0x0: Disabled.
0x1: Rising edge triggered.
0x2: Falling edge triggered.
0x3: Rising and falling edge triggered.
0x4: Active high level triggered.
0x5: Active low level triggered.
0x6, 0x7: Reserved.
7-19
120-035X-000 Rev. 1.2
Final
EM351 / EM357
INT_GPIOFLAG
GPIO Interrupt Flag Register
Address: 0x4000A814 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
INT_IRQDFLAG
INT_IRQCFLAG
INT_IRQBFLAG
INT_IRQAFLAG
Bitname
Bitfield
[3]
Access
Description
INT_IRQDFLAG
INT_IRQCFLAG
INT_IRQBFLAG
INT_IRQAFLAG
RW
RW
RW
RW
IRQD interrupt pending. Write 1 to clear IRQD interrupt (writing 0 has no effect).
IRQC interrupt pending. Write 1 to clear IRQC interrupt (writing 0 has no effect).
IRQB interrupt pending. Write 1 to clear IRQB interrupt (writing 0 has no effect).
IRQA interrupt pending. Write 1 to clear IRQA interrupt (writing 0 has no effect).
[2]
[1]
[0]
7-20
120-035X-000 Rev. 1.2
Final
EM351 / EM357
GPIO_DBGCFG
GPIO Debug Configuration Register
Address: 0x4000BC00 Reset: 0x10
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
GPIO_DEBUGDIS
GPIO_EXTREGEN
GPIO_DBGCFGRSVD
0
0
0
Bitname
Bitfield
Access
Description
GPIO_DEBUGDIS
GPIO_EXTREGEN
GPIO_DBGCFGRSVD
[5]
RW
Disable debug interface override of normal GPIO configuration.
0: Permit debug interface to be active.
1: Disable debug interface (if it is not already active).
[4]
[3]
RW
RW
Enable REG_EN override of PA7's normal GPIO configuration.
0: Disable override.
1: Enable override.
Reserved: this bit can change during normal operation. When writing to GPIO_DBGCFG,
the value of this bit must be preserved.
7-21
120-035X-000 Rev. 1.2
Final
EM351 / EM357
GPIO_DBGSTAT
GPIO Debug Status Register
Address: 0x4000BC04 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
GPIO_BOOTMODE
0
GPIO_FORCEDBG
GPIO_SWEN
Bitname
Bitfield
Access
Description
GPIO_BOOTMODE
GPIO_FORCEDBG
GPIO_SWEN
[3]
R
R
R
The state of the nBOOTMODE signal sampled at the end of reset.
0: nBOOTMODE was not asserted (it read high).
1: nBOOTMODE was asserted (it read low).
[1]
[0]
Status of debugger interface.
0: Debugger interface not forced active.
1: Debugger interface forced active by debugger cable.
Status of Serial Wire interface.
0: Not enabled by SWJ-D P.
1: Enabled by SWJ-D P.
7-22
120-035X-000 Rev. 1.2
Final
EM351 / EM357
8
Serial Controllers
8.1
Overview
The EM35x has two serial controllers, SC1 and SC2, which provide several options for full-duplex synchronous
and asynchronous serial communications.
SPI (Serial Peripheral Interface), master or slave
.
TWI (Two Wire serial Interface), master only
.
UART (Universal Asynchronous Receiver/Transmitter), SC1 only
.
.
Receive and transmit FIFOs and DMA channels, SPI and UART modes
Receive and transmit FIFOs allow faster data speeds using byte-at-a-time interrupts. For the highest SPI and
UART speeds, dedicated receive and transmit DMA channels reduce CPU loading and extend the allowable time
to service a serial controller interrupt. Polled operation is also possible using direct access to the serial data
registers. Figure 8-1 shows the components of the serial controllers.
Note: The notation SCx means that either SC1 or SC2 may be substituted to form the name of a specific
register or field within a register.
Figure 8-1. Serial Controller Block Diagram
SCx Interrupt
INT_SCxCFG
OFF
0
INT_SCxFLAG
SC1_UARTPER/FRAC
Baud Generator
SC1
only
TXD
UART
SC1_UARTSTAT
SC1_UARTCFG
RXD
nRTS
nCTS
UART
Controller
1
SCx_MODE
SPI Slave
Controller
MISO
SCx_SPISTAT
SCx_SPICFG
SPI
MOSI
SCLK
nSSEL
2
SPI Master
Controller
SCx_RATELIN/EXP
Clock Generator
TWI
3
SCL
SDA
SCx_TWISTAT
SCx_TWICTRL1
SCx_TWICTRL2
TWI Master
Controller
SCx_DATA
TX-FIFO
SCx TX DMA
SCx_DMACTRL
SCx_RXCNTA/B
SCx_RXCNTSAVED
SCx_TXCNT
channel
DMA
SCx_TX/RXBEGA/B
SCx_TX/RXENDA/B
Controller
SCx RX DMA
channel
SCx_DMASTAT
SCx_RXERRA/B
RX-FIFO
8-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
8.2
Configuration
Before using a serial controller, configure and initialize it as follows:
Set up the parameters specific to the operating mode (master/slave for SPI, baud rate for UART, etc.).
.
.
Configure the GPIO pins used by the serial controller as shown in Table 8-1 and Table 8-2. Section 2 in
Chapter 7, GPIO shows how to configure GPIO pins.
If using DMA, set up the DMA and buffers. This is described fully in section 8.7.
.
.
If using interrupts, select edge- or level-triggered interrupts with the SCx_INTMODE register, enable
the desired second-level interrupt sources in the INT_SCxCFG register, and finally enable the top-level
SCx interrupt in the NVIC.
Write the serial interface operating mode — SPI, TWI, or UART — to the SCx_MODE register.
.
Table 8-1. SC1 GPIO Usage and Configuration
PB1
PB2
PB3
PB4
SPI - Master
SC1MOSI
Alternate Output
(push-pull)
SC1MISO
Input
SC1SCLK
Alternate Output
(push-pull)
(not used)
SPI - Slave
TWI - Master
UART
SC1MISO
Alternate Output
(push-pull)
SC1MOSI
Input
SC1SCLK
Input
SC1nSSEL
Input
SC1SDA
Alternate Output
(open-drain)
SC1SCL
Alternate Output
(open-drain)
(not used)
(not used)
TXD
RXD
Input
nCTS
nRTS
Alternate Output (push-
pull)1
Alternate Output
(push-pull)
Input1
1 used if RTS/CTS hardware flow control is enabled.
Table 8-2. SC2 GPIO Usage and Configuration
PA0
PA1
PA2
PA3
SPI - Master
SPI - Slave
TWI - Master
SC2MOSI
Alternate Output
(push-pull)
SC2MISO
Input
SC2SCLK
Alternate Output
(push-pull)
(not used)
SC2MOSI
Input
SC2MISO
Alternate Output
(push-pull)
SC2SCLK
Input
SC2nSSEL
Input
(not used)
SC2SDA
SC2SCL
(not used)
Alternate Output
(open-drain)
Alternate Output
(open-drain)
8-2
Final
120-035X-000 Rev. 1.2
EM351 / EM357
8.2.1 Registers
SCx_MODE
SC1_MODE
Serial Mode Register
Address: 0x4000C854 Reset: 0x0
Address: 0x4000C054 Reset: 0x0
SC2_MODE
Serial Mode Register
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
0
14
0
13
0
12
0
11
0
10
0
9
8
0
0
7
6
5
4
3
2
1
0
0
0
0
0
0
0
SC_MODE
Bitname
Bitfield
Access
RW
Description
SC_MODE
[1:0]
Serial controller mode.
0: Disabled.
1: UART mode (valid only for SC1).
2: SPI mode.
3: TWI mode.
8-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
INT_SCxFLAG
INT_SC1FLAG
Serial Controller 1 Interrupt Flag Register
Address: 0x4000A808 Reset: 0x0
Address: 0x4000A80C Reset: 0x0
INT_SC2FLAG
Serial Controller 2 Interrupt Flag Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
INT_SC1PARERR
INT_SC1FRMERR
INT_SCTXULDB
INT_SCTXULDA
INT_SCRXULDB
INT_SCRXULDA
INT_SCNAK
7
6
5
4
3
2
1
0
INT_SCCMDFIN
INT_SCTXFIN
INT_SCRXFIN
INT_SCTXUND
INT_SCRXOVF
INT_SCTXIDLE
INT_SCTXFREE
INT_SCRXVAL
Bitname
Bitfield
Access
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
Description
INT_SC1PARERR
INT_SC1FRMERR
INT_SCTXULDB
INT_SCTXULDA
INT_SCRXULDB
INT_SCRXULDA
INT_SCNAK
[14]
[13]
[12]
[11]
[10]
[9]
Parity error received (UART) interrupt pending.
Frame error received (UART) interrupt pending.
DMA transmit buffer B unloaded interrupt pending.
DMA transmit buffer A unloaded interrupt pending.
DMA receive buffer B unloaded interrupt pending.
DMA receive buffer A unloaded interrupt pending.
NACK received (TWI) interrupt pending.
[8]
INT_SCCMDFIN
INT_SCTXFIN
[7]
START/STOP command complete (TWI) interrupt pending.
Transmit operation complete (TWI) interrupt pending.
Receive operation complete (TWI) interrupt pending.
Transmit buffer underrun interrupt pending.
Receive buffer overrun interrupt pending.
Transmitter idle interrupt pending.
[6]
INT_SCRXFIN
[5]
INT_SCTXUND
INT_SCRXOVF
INT_SCTXIDLE
INT_SCTXFREE
INT_SCRXVAL
[4]
[3]
[2]
[1]
Transmit buffer free interrupt pending.
[0]
Receive buffer has data interrupt pending.
8-4
120-035X-000 Rev. 1.2
Final
EM351 / EM357
INT_SCxCFG
INT_SC1CFG
Serial Controller 1 Interrupt Configuration Register
Address: 0x4000A848 Reset: 0x0
Address: 0x4000A84C Reset: 0x0
INT_SC2CFG
Serial Controller 2 Interrupt Configuration Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
INT_SC1PARERR
INT_SC1FRMERR
INT_SCTXULDB
INT_SCTXULDA
INT_SCRXULDB
INT_SCRXULDA
INT_SCNAK
7
6
5
4
3
2
1
0
INT_SCCMDFIN
INT_SCTXFIN
INT_SCRXFIN
INT_SCTXUND
INT_SCRXOVF
INT_SCTXIDLE
INT_SCTXFREE
INT_SCRXVAL
Bitname
Bitfield
Access
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
Description
INT_SC1PARERR
INT_SC1FRMERR
INT_SCTXULDB
INT_SCTXULDA
INT_SCRXULDB
INT_SCRXULDA
INT_SCNAK
[14]
[13]
[12]
[11]
[10]
[9]
Parity error received (UART) interrupt enable.
Frame error received (UART) interrupt enable.
DMA transmit buffer B unloaded interrupt enable.
DMA transmit buffer A unloaded interrupt enable.
DMA receive buffer B unloaded interrupt enable.
DMA receive buffer A unloaded interrupt enable.
NACK received (TWI) interrupt enable.
[8]
INT_SCCMDFIN
INT_SCTXFIN
[7]
START/STOP command complete (TWI) interrupt enable.
Transmit operation complete (TWI) interrupt enable.
Receive operation complete (TWI) interrupt enable.
Transmit buffer underrun interrupt enable.
Receive buffer overrun interrupt enable.
[6]
INT_SCRXFIN
[5]
INT_SCTXUND
INT_SCRXOVF
INT_SCTXIDLE
INT_SCTXFREE
INT_SCRXVAL
[4]
[3]
[2]
Transmitter idle interrupt enable.
[1]
Transmit buffer free interrupt enable.
[0]
Receive buffer has data interrupt enable.
8-5
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_INTMODE
SC1_INTMODE
Serial Controller 1 Interrupt Mode Register
Address: 0x4000A854 Reset: 0x0
Address: 0x4000A858 Reset: 0x0
SC2_INTMODE
Serial Controller 2 Interrupt Mode Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
0
SC_TXIDLELEVEL
SC_TXFREELEVEL
SC_RXVALLEVEL
Bitname
Bitfield
[2]
Access
Description
SC_TXIDLELEVEL
SC_TXFREELEVEL
SC_RXVALLEVEL
RW
RW
RW
Transmitter idle interrupt mode - 0: edge triggered, 1: level triggered.
Transmit buffer free interrupt mode - 0: edge triggered, 1: level triggered.
[1]
[0]
Receive buffer has data interrupt mode - 0: edge triggered, 1: level triggered.
8.3
SPI - Master Mode
The SPI master controller has the following features:
Full duplex operation
.
Programmable clock frequency (12 MHz max.)
.
Programmable clock polarity and phase
.
Selectable data shift direction (either LSB or MSB first)
.
Receive and transmit FIFOs
.
Receive and transmit DMA channels
.
8.3.1 GPIO Usage
The SPI master controller uses the three signals:
MOSI (Master Out, Slave In) – outputs serial data from the master
.
MISO (Master In, Slave Out) – inputs serial data from a slave
.
SCLK (Serial Clock) – outputs the serial clock used by MOSI and MISO
.
8-6
120-035X-000 Rev. 1.2
Final
EM351 / EM357
The GPIO pins used for these signals are shown in Table 8-3. Additional outputs may be needed to drive the
nSSEL signals on slave devices.
Table 8-3. SPI Master GPIO Usage
MOSI
MISO
SCLK
Direction
Output
Input
Output
GPIO Configuration
Alternate Output
(push-pull)
Input
Alternate Output
(push-pull)
SC1 pin
SC2 pin
PB1
PA0
PB2
PA1
PB3
PA2
8.3.2 Set Up and Configuration
Both serial controllers, SC1 and SC2, support SPI master mode. SPI master mode is enabled by the following
register settings:
The serial controller mode register (SCx_MODE) is 2.
.
.
The SC_SPIMST bit in the SPI configuration register (SCx_SPICFG) is 1.
The SPI serial clock (SCLK) is produced by a programmable clock generator. The serial clock is produced by
dividing down 12 MHz according to this equation:
12MHz
rate =
(LIN +1)* 2EXP
EXP is the value written to the SCx_RATEEXP register and LIN is the value written to the SCx_RATELIN register.
EXP and LIN can both be zero so the SPI master mode clock may be 12 Mbps.
The SPI master controller supports various frame formats depending upon the clock polarity (SC_SPIPOL),
clock phase (SC_SPIPHA), and direction of data (SC_SPIORD) (see Table 8-4). The bits SC_SPIPOL, SC_SPIPHA,
and SC_SPIORD are defined within the SCx_SPICFG register.
8-7
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 8-4. SPI Master Mode Formats
SCx_SPICFG
SC_SPIxxx1
MST ORD PHA POL
Frame Formats
1
1
1
1
1
0
0
0
0
1
0
0
1
1
-
0
1
0
1
-
SCLKout
MOSIout
MISOin
TX[7]
RX[7]
TX[6]
RX[6]
TX[5]
RX[5]
TX[4]
RX[4]
TX[3]
RX[3]
TX[2]
RX[2]
TX[1]
RX[1]
TX[0]
RX[0]
SCLKout
MOSIout
MISOin
TX[7]
RX[7]
TX[6]
RX[6]
TX[5]
RX[5]
TX[4]
RX[4]
TX[3]
RX[3]
TX[2]
TX[1]
TX[0]
RX[2]
RX[1]
RX[0]
SCLKout
MOSIout
MISOin
TX[7]
RX[7]
TX[6]
RX[6]
TX[5]
RX[5]
TX[4]
RX[4]
TX[3]
RX[3]
TX[2]
RX[2]
TX[1]
RX[1]
TX[0]
RX[0]
SCLKout
MOSIout
MISOin
TX[7]
RX[7]
TX[6]
RX[6]
TX[5]
RX[5]
TX[4]
RX[4]
TX[3]
RX[3]
TX[2]
RX[2]
TX[1]
RX[1]
TX[0]
RX[0]
Same as above except data is sent LSB first instead of MSB first
1 The notation xxx means that the corresponding column header below is inserted to form the field name.
8.3.3 Operation
Characters transmitted and received by the SPI master controller are buffered in transmit and receive FIFOs
that are both 4 entries deep. When software writes a character to the SCx_DATA register, the character is
pushed onto the transmit FIFO. Similarly, when software reads from the SCx_DATA register, the character
returned is pulled from the receive FIFO. If the transmit and receive DMA channels are used, they also write to
and read from the transmit and receive FIFOs.
When the transmit FIFO and the serializer are both empty, writing a character to the transmit FIFO clears the
SC_SPITXIDLE bit in the SCx_SPISTAT register. This indicates that some characters have not yet been
transmitted. If characters are written to the transmit FIFO until it is full, the SC_SPITXFREE bit in the
SCx_SPISTAT register is cleared. Shifting out a character to the MOSI pin sets the SC_SPITXFREE bit in the
SCx_SPISTAT register. When the transmit FIFO empties and the last character has been shifted out, the
SC_SPITXIDLE bit in the SCx_SPISTAT register is set.
Characters received are stored in the receive FIFO. Receiving characters sets the SC_SPIRXVAL bit in the
SCx_SPISTAT register, indicating that characters can be read from the receive FIFO. Characters received while
the receive FIFO is full are dropped, and the SC_SPIRXOVF bit in the SCx_SPISTAT register is set. The receive
FIFO hardware generates the INT_SCRXOVF interrupt, but the DMA register will not indicate the error
condition until the receive FIFO is drained. Once the DMA marks a receive error, two conditions will clear the
error indication: setting the appropriate SC_TX/RXDMARST bit in the SCx_DMACTRL register, or loading the
appropriate DMA buffer after it has unloaded.
To receive a character, you must transmit a character. If a long stream of receive characters is expected, a
long sequence of dummy transmit characters must be generated. To avoid software or transmit DMA initiating
these transfers and consuming unnecessary bandwidth, the SPI serializer can be instructed to retransmit the
8-8
120-035X-000 Rev. 1.2
Final
EM351 / EM357
last transmitted character or to transmit a busy token (0xFF), which is determined by the SC_SPIRPT bit in the
SCx_SPICFG register. This functionality can only be enabled or disabled when the transmit FIFO is empty and
the transmit serializer is idle, indicated by a cleared SC_SPITXIDLE bit in the SCx_SPISTAT register. Refer to
the register description of SCx_SPICFG for more detailed information about SC_SPIRPT.
Every time an automatic character transmission starts, a transmit underrun is detected as there is no data in
transmit FIFO, and the INT_SCTXUND bit in the INT_SC2FLAG register is set. After automatic character
transmission is disabled, no more new characters are received. The receive FIFO holds characters just
received.
Note: The Receive DMA complete event does not always mean the receive FIFO is empty.
The DMA Channels section describes how to configure and use the serial receive and transmit DMA channels.
8.3.4 Interrupts
SPI master controller second-level interrupts are generated by the following events:
Transmit FIFO empty and last character shifted out (depending on SCx_INTMODE, either the 0 to 1
transition or the high level of SC_SPITXIDLE)
.
Transmit FIFO changed from full to not full (depending on SCx_INTMODE, either the 0 to 1 transition or the
high level of SC_SPITXFREE)
.
Receive FIFO changed from empty to not empty (depending on SCx_INTMODE, either the 0 to 1 transition
or the high level of SC_SPIRXVAL)
.
Transmit DMA buffer A/B complete (1 to 0 transition of SC_TXACTA/B)
.
Receive DMA buffer A/B complete (1 to 0 transition of SC_RXACTA/B)
.
Received and lost character while receive FIFO was full (receive overrun error)
.
.
Transmitted character while transmit FIFO was empty (transmit underrun error)
To enable CPU interrupts, set the desired interrupt bits in the second-level INT_SCxCFG register, and enable
the top-level SCx interrupt in the NVIC by writing the INT_SCx bit in the INT_CFGSET register.
8-9
120-035X-000 Rev. 1.2
Final
EM351 / EM357
8.3.5 Registers
SCx_DATA
SC1_DATA
Serial Data Register
Address: 0x4000C83C Reset: 0x0
Address: 0x4000C03C Reset: 0x0
SC2_DATA
Serial Data Register
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
0
14
0
13
0
12
0
11
0
10
0
9
8
0
0
7
6
5
4
3
2
1
0
SC_DATA
Bitname
Bitfield
Access
RW
Description
SC_DATA
[7:0]
Transmit and receive data register. Writing to this register adds a byte to the transmit
FIFO. Reading from this register takes the next byte from the receive FIFO and clears the
overrun error bit if it was set.
In UART mode (SC1 only), reading from this register loads the UART status register with
the parity and frame error status of the next byte in the FIFO, and clears these bits if the
FIFO is now empty.
8-10
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_SPICFG
SC1_SPICFG
SPI Configuration Register
Address: 0x4000C858 Reset: 0x0
Address: 0x4000C058 Reset: 0x0
SC2_SPICFG
SPI Configuration Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
SC_SPIRXDRV
SC_SPIMST
SC_SPIRPT
SC_SPIORD
SC_SPIPHA
SC_SPIPOL
Bitname
Bitfield
Access
Description
SC_SPIRXDRV
[5]
RW
Receiver-driven mode selection bit (SPI master mode only). Clear this bit to initiate
transactions when transmit data is available. Set this bit to initiate transactions when the
receive buffer (FIFO or DMA) has space.
SC_SPIMST
SC_SPIRPT
[4]
[3]
RW
RW
Set this bit to put the SPI in master mode, clear this bit to put the SPI in slave mode.
This bit controls behavior when the transmit serializer must send a byte and there is no
data already available in/to the serializer. The conditions for sending this “busy” token
are transmit buffer underrun condition when using DMA in master or slave mode, empty
FIFO in slave mode, and the busy token will always be sent as the first byte every time
nSSEL is asserted while operating in slave mode. Clear this bit to send the BUSY token
(0xFF) and set this bit to repeat the last byte. Changes to this bit take effect when the
transmit FIFO is empty and the transmit serializer is idle. Note that when the chip comes
out of reset, if SC_SPIRPT is set before any data has been transmitted and no data is
available (in the FIFO), the “last byte” that will be transmitted after the padding byte is
0x00 due to the FIFO having been reset to 0x00.
SC_SPIORD
[2]
RW
This bit specifies the bit order in which SPI data is transmitted and received.
0: Most significant bit first.
1: Least significant bit first.
SC_SPIPHA
SC_SPIPOL
[1]
[0]
RW
RW
Clock phase configuration: clear this bit to sample on the leading (first edge) and set this
bit to sample on the second edge.
Clock polarity configuration: clear this bit for a rising leading edge and set this bit for a
falling leading edge.
8-11
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_SPISTAT
SC1_SPISTAT
SPI Status Register
Address: 0x4000C840 Reset: 0x0
Address: 0x4000C040 Reset: 0x0
SC2_SPISTAT
SPI Status Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
SC_SPITXIDLE
SC_SPITXFREE
SC_SPIRXVAL
SC_SPIRXOVF
Bitname
Bitfield
[3]
Access
Description
SC_SPITXIDLE
SC_SPITXFREE
SC_SPIRXVAL
SC_SPIRXOVF
R
R
R
R
This bit is set when both the transmit FIFO and the transmit serializer are empty.
This bit is set when the transmit FIFO has space to accept at least one byte.
This bit is set when the receive FIFO contains at least one byte.
[2]
[1]
[0]
This bit is set if a byte is received when the receive FIFO is full. This bit is cleared by
reading the data register.
8-12
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_RATELIN
SC1_RATELIN
Serial Clock Linear Prescaler Register
Address: 0x4000C860 Reset: 0x0
Address: 0x4000C060 Reset: 0x0
SC2_RATELIN
Serial Clock Linear Prescaler Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
SC_RATELIN
Bitname
SC_RATELIN
Bitfield
Access
RW
Description
[3:0]
The linear component (LIN) of the clock rate in the equation:
rate = 12MHz / ( (LIN + 1) * (2^EXP) )
8-13
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_RATEEXP
SC1_RATEEXP
Serial Clock Exponential Prescaler Register
Address: 0x4000C864 Reset: 0x0
Address: 0x4000C064 Reset: 0x0
SC2_RATEEXP
Serial Clock Exponential Prescaler Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
SC_RATEEXP
Bitname
SC_RATEEXP
Bitfield
Access
RW
Description
[3:0]
The exponential component (EXP) of the clock rate in the equation:
rate = 12MHz / ( (LIN + 1) * (2^EXP) )
8.4
SPI - Slave Mode
Both SC1 and SC2 SPI controllers include a SPI slave controller with these features:
Full duplex operation
.
Up to 5 Mbps data transfer rate
.
Programmable clock polarity and clock phase
.
Selectable data shift direction (either LSB or MSB first)
.
Slave select input
.
8.4.1 GPIO Usage
The SPI slave controller uses four signals:
MOSI (Master Out, Slave In) – inputs serial data from the master
.
MISO (Master In, Slave Out) – outputs serial data to the master
.
SCLK (Serial Clock) – clocks data transfers on MOSI and MISO
.
.
nSSEL (Slave Select) – enables serial communication with the slave
Note: The SPI slave controller does not tri-state the MISO signal when slave select is deasserted.
8-14
120-035X-000 Rev. 1.2
Final
EM351 / EM357
The GPIO pins that can be assigned to these signals are shown in Table 8-5.
Table 8-5. SPI Slave GPIO Usage
MOSI
MISO
SCLK
nSSEL
Direction
Input
Output
Input
Input
GPIO Configuration
Input
Alternate Output
(push-pull)
Input
Input
SC1 pin
SC2 pin
PB2
PA0
PB1
PA1
PB3
PA2
PB4
PA3
8.4.2 Set Up and Configuration
Both serial controllers, SC1 and SC2, support SPI slave mode. SPI slave mode is enabled by the following
register settings:
The serial controller mode register, SCx_MODE, is 2
.
.
The SC_SPIMST bit in the SPI configuration register, SCx_SPICFG, is 0
The SPI slave controller receives its clock from an external SPI master device and supports rates up to 5 Mbps.
8-15
120-035X-000 Rev. 1.2
Final
EM351 / EM357
The SPI slave controller supports various frame formats depending upon the clock polarity (SC_SPIPOL), clock
phase (SC_SPIPHA), and direction of data (SC_SPIORD) (see Table 8-6). The SC_SPIPOL, SC_SPIPHA, and
SC_SPIORD bits are defined within the SCx_SPICFG registers.
Table 8-6. SPI Slave Formats
SCx_SPICFG
SC_SPIxxx1
MST ORD PHA POL Frame Format
0
0
0
0
0
0
0
0
0
1
0
0
1
1
-
0
1
0
1
-
nSSEL
SCLKin
MOSIin
MISOout
RX[7]
TX[7]
RX[6]
TX[6]
RX[5]
TX[5]
RX[4]
TX[4]
RX[3]
TX[3]
RX[2]
TX[2]
RX[1]
TX[1]
RX[0]
TX[0]
SCLKin
MOSIin
MISOout
RX[7]
TX[7]
RX[6]
TX[6]
RX[5]
TX[5]
RX[4]
TX[4]
RX[3]
TX[3]
RX[2]
TX[2]
RX[1]
TX[1]
RX[0]
TX[0]
nSSEL
SCLKin
MOSIin
MISOout
RX[7]
TX[7]
RX[6]
TX[6]
RX[5]
TX[5]
RX[4]
TX[4]
RX[3]
TX[3]
RX[2]
TX[2]
RX[1]
TX[1]
RX[0]
TX[0]
nSSEL
SCLKin
MOSIin
MISOout
RX[7]
TX[7]
RX[6]
TX[6]
RX[5]
TX[5]
RX[4]
TX[4]
RX[3]
TX[3]
RX[2]
TX[2]
RX[1]
TX[1]
RX[0]
TX[0]
Same as above except LSB first instead of MSB first
1 The notation xxx means that the corresponding column header below is inserted to form the field name.
8.4.3 Operation
When the slave select (nSSEL) signal is asserted by the master, SPI transmit data is driven to the output pin
MISO, and SPI data is received from the input pin MOSI. The nSSEL pin has to be asserted to enable the
transmit serializer to drive data to the output signal MISO. A falling edge on nSSEL resets the SPI slave shift
registers.
Note: The SPI slave controller does not tri-state the MISO signal when slave select is deasserted.
Characters transmitted and received by the SPI slave controller are buffered in the transmit and receive FIFOs
that are both 4 entries deep. When software writes a character to the SCx_DATA register, it is pushed onto
the transmit FIFO. Similarly, when software reads from the SCx_DATA register, the character returned is
pulled from the receive FIFO. If the transmit and receive DMA channels are used, the DMA channels also write
to and read from the transmit and receive FIFOs.
Characters received are stored in the receive FIFO. Receiving characters sets the SC_SPIRXVAL bit in the
SCx_SPISTAT register, to indicate that characters can be read from the receive FIFO. Characters received
8-16
120-035X-000 Rev. 1.2
Final
EM351 / EM357
while the receive FIFO is full are dropped, and the SC_SPIRXOVF bit in the SCx_SPISTAT register is set. The
receive FIFO hardware generates the INT_SCRXOVF interrupt, but the DMA register will not indicate the error
condition until the receive FIFO is drained. Once the DMA marks a receive error, two conditions will clear the
error indication: setting the appropriate SC_TX/RXDMARST bit in the SCx_DMACTRL register, or loading the
appropriate DMA buffer after it has unloaded.
Receiving a character causes the serial transmission of a character pulled from the transmit FIFO. When the
transmit FIFO is empty, a transmit underrun is detected (no data in transmit FIFO) and the INT_SCTXUND bit in
the INT_SCxFLAG register is set. Because no character is available for serialization, the SPI serializer
retransmits the last transmitted character or a busy token (0xFF), determined by the SC_SPIRPT bit in the
SCx_SPICFG register. Refer to the register description of SCx_SPICFG for more detailed information about
SC_SPIRPT.
When the transmit FIFO and the serializer are both empty, writing a character to the transmit FIFO clears the
SC_SPITXIDLE bit in the SCx_SPISTAT register. This indicates that not all characters have been transmitted. If
characters are written to the transmit FIFO until it is full, the SC_SPITXFREE bit in the SCx_SPISTAT register is
cleared. Shifting out a transmit character to the MISO pin causes the SC_SPITXFREE bit in the SCx_SPISTAT
register to get set. When the transmit FIFO empties and the last character has been shifted out, the
SC_SPITXIDLE bit in the SCx_SPISTAT register is set.
The SPI slave controller must guarantee that there is time to move new transmit data from the transmit FIFO
into the hardware serializer. To provide sufficient time, the SPI slave controller inserts a byte of padding at
the start of every new string of transmit data defined by every time nSSEL is asserted. This byte is inserted
as if this byte was placed there by software. The value of the byte of padding is always 0xFF.
8.4.4 DMA
The DMA Channels section describes how to configure and use the serial receive and transmit DMA channels.
When using the receive DMA channel and nSSEL transitions to the high (deasserted) state, the active buffer’s
receive DMA count register (SCx_RXCNTA/B) is saved in the SCx_RXCNTSAVED register. SCx_RXCNTSAVED is
only written the first time nSSEL goes high after a buffer has been loaded. Subsequent rising edges set a status
bit but are otherwise ignored. The 3-bit field SC_RXSSEL in the SCx_DMASTAT register records what, if
anything, was saved to the SCx_RXCNTSAVED register, and whether or not another rising edge occurred on
nSSEL.
8.4.5 Interrupts
SPI slave controller second-level interrupts are generated on the following events:
Transmit FIFO empty and last character shifted out (depending on SCx_INTMODE, either the 0 to 1
transition or the high level of SC_SPITXIDLE)
.
Transmit FIFO changed from full to not full (depending on SCx_INTMODE, either the 0 to 1 transition or the
high level of SC_SPITXFREE)
.
Receive FIFO changed from empty to not empty (depending on SCx_INTMODE, either the 0 to 1 transition
or the high level of SC_SPIRXVAL)
.
Transmit DMA buffer A/B complete (1 to 0 transition of SC_TXACTA/B)
.
Receive DMA buffer A/B complete (1 to 0 transition of SC_RXACTA/B)
.
Received and lost character while receive FIFO was full (receive overrun error)
.
.
Transmitted character while transmit FIFO was empty (transmit underrun error)
To enable CPU interrupts, set desired interrupt bits in the second-level INT_SCxCFG register, and also enable
the top-level SCx interrupt in the NVIC by writing the INT_SCx bit in the INT_CFGSET register.
8-17
120-035X-000 Rev. 1.2
Final
EM351 / EM357
8.4.6 Registers
Refer to Registers (in the SPI Master Mode section) for a description of the SCx_DATA, SCx_SPICFG, and
SCx_SPISTAT registers.
8.5
TWI - Two Wire serial Interfaces
Both EM35x serial controllers SC1 and SC2 include a Two Wire serial Interface (TWI) master controller with the
following features:
Uses only two bidirectional GPIO pins
.
Programmable clock frequency (up to 400 kHz)
.
Supports both 7-bit and 10-bit addressing
Compatible with Philips’ I2C-bus slave devices
.
.
8.5.1 GPIO Usage
The TWI master controller uses just two signals:
SDA (Serial Data) – bidirectional serial data
.
.
SCL (Serial Clock) – bidirectional serial clock
Table 8-7 lists the GPIO pins used by the SC1 and SC2 TWI master controllers. Because the pins are configured
as open-drain outputs, they require external pull-up resistors.
Table 8-7. TWI Master GPIO Usage
SDA
SCL
Direction
Input / Output
Input / Output
GPIO Configuration
Alternate Output
(open drain)
Alternate Output
(open drain)
SC1 pin
SC2 pin
PB1
PA1
PB2
PA2
8.5.2 Set Up and Configuration
The TWI controller is enabled by writing 3 to the SCx_MODE register. The TWI controller operates only in
master mode and supports both Standard (100 kbps) and Fast (400 kbps) TWI modes. Address arbitration is not
implemented, so multiple master applications are not supported.
The TWI master controller’s serial clock (SCL) is produced by a programmable clock generator. SCL is
produced by dividing down 12 MHz according to this equation:
12MHz
rate =
(LIN +1)* 2EXP
8-18
120-035X-000 Rev. 1.2
Final
EM351 / EM357
EXP is the value written to the SCx_RATEEXP register and LIN is the value written to the SCx_RATELIN register.
Table 8-8 shows the rate settings for Standard-Mode TWI (100 kbps) and Fast-Mode TWI (400 kbps) operation.
Table 8-8. TWI Clock Rate Programming
Clock Rate
SCx_RATELIN
SCx_RATEEXP
100 kbps
14
3
375 kbps
400 kbps
15
14
1
1
Note: At 400 kbps, the Philips I2C Bus specification requires the minimum low period of SCL to be 1.3 µs, but
on the EM35x it is 1.25 µs. If a slave device requires strict compliance with SCL timing, the clock rate must be
lowered to 375 kbps.
The EM35x supports clock stretching. The slave device can hold SCL low on any received or transmitted data
bit. This inhibits further data transfers until SCL is allowed to go high again.
8.5.3 Constructing Frames
The TWI master controller supports generating various frame segments by means of the SC_TWISTART,
SC_TWISTOP, SC_TWISEND, and SC_TWIRECV bits in the SCx_TWICTRL1 registers. Table 8-9 summarizes these
frames.
8-19
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 8-9. TWI Master Frame Segments
SCx_TWICTRL1
SC_TWIxxxx1
START SEND RECV STOP Frame Segments
TWI start segment
TWI re-start segment - after transmit or frame with NACK
SCLoutSLAVE
1
0
0
0
SCLoutSLAVE
SCLout
SCLout
SDAout
SDAout
SDAoutSLAVE
SDAoutSLAVE
TWI transmit segment - after (re-)start frame
0
1
0
0
SCLoutSLAVE
SCLout
TX[7]
TX[6]
TX[5]
TX[4]
TX[3]
TX[2]
TX[1]
TX[0]
SDAout
(N)ACK
(N)ACK
(N)ACK
SDAoutSLAVE
TWI transmit segment – after transmit with ACK
SCLoutSLAVE
SCLout
TX[7]
TX[6]
TX[5]
TX[4]
TX[3]
TX[2]
TX[1]
TX[0]
SDAout
SDAoutSLAVE
TWI receive segment – transmit with ACK
0
0
1
0
SCLoutSLAVE
SCLout
SDAout
RX[7]
RX[6]
RX[5]
RX[4]
RX[3]
RX[2]
RX[1]
RX[0]
SDAoutSLAVE
TWI receive segment - after receive with ACK
SCLoutSLAVE
SCLout
(N)ACK
SDAout
RX[7]
RX[6]
RX[5]
RX[4]
RX[3]
RX[2]
RX[1]
RX[0]
SDAoutSLAVE
TWI stop segment - after frame with NACK or stop
SCLoutSLAVE
0
0
0
0
0
0
1
0
SCLout
SDAout
SDAoutSLAVE
No pending frame segment
Illegal
1
-
-
1
1
-
-
1
1
-
-
-
1
1
1
-
1 The notation xxx means that the corresponding column header below is inserted to form the field name.
8-20
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Full TWI frames have to be constructed by software from individual TWI segments. All necessary segment
transitions are shown in Figure 8-2. ACK or NACK generation of a TWI receive frame segment is determined
with the SC_TWIACK bit in the SCx_TWICTRL2 register.
Figure 8-2. TWI Segment Transitions
IDLE
START Segment
STOP Segment
TRANSMIT Segment
NO
received ACK ?
YES
RECEIVE Segment
with NACK
RECEIVE Segment
with ACK
Generation of a 7-bit address is accomplished with one transmit segment. The upper 7 bits of the transmitted
character contain the 7-bit address. The remaining lower bit contains the command type (“read” or “write”).
Generation of a 10-bit address is accomplished with two transmit segments. The upper 5 bits of the first
transmit character must be set to 0x1E. The next 2 bits are for the 2 most significant bits of the 10-bit
address. The remaining lower bit contains the command type (“read” or “write”). The second transmit
segment is for the remaining 8 bits of the 10-bit address.
Transmitted and received characters are accessed through the SCx_DATA register.
To initiate (re)start and stop segments, set the SC_TWISTART or SC_TWISTOP bit in the SCx_TWICTRL1
register, then wait until the bit is clear. Alternatively, the SC_TWICMDFIN bit in the SCx_TWISTAT can be used
for waiting.
To initiate a transmit segment, write the data to the SCx_DATA data register, then set the SC_TWISEND bit in
the SCx_TWICTRL1 register, and finally wait until the bit is clear. Alternatively the SC_TWITXFIN bit in the
SCx_TWISTAT register can be used for waiting.
To initiate a receive segment, set the SC_TWIRECV bit in the SCx_TWICTRL1 register, wait until it is clear, and
then read from the SCx_DATA register. Alternatively, the SC_TWIRXFIN bit in the SCx_TWISTAT register can be
used for waiting. Now the SC_TWIRXNAK bit in the SCx_TWISTAT register indicates if a NACK or ACK was
received from a TWI slave device.
8.5.4 Interrupts
TWI master controller interrupts are generated on the following events:
Bus command (SC_TWISTART/SC_TWISTOP) completed (0 to 1 transition of SC_TWICMDFIN)
.
.
Character transmitted and slave device responded with NACK
8-21
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Character transmitted (0 to 1 transition of SC_TWITXFIN)
Character received (0 to 1 transition of SC_TWIRXFIN)
.
.
.
.
Received and lost character while receive FIFO was full (receive overrun error)
Transmitted character while transmit FIFO was empty (transmit underrun error)
To enable CPU interrupts, set the desired interrupt bits in the second-level INT_SCxCFG register, and enable
the top-level SCx interrupt in the NVIC by writing the INT_SCx bit in the INT_CFGSET register.
8.5.5 Registers
Refer to Registers (in the SPI Master Mode section) for a description of the SCx_DATA, SCx_RATELIN, and
SCx_RATEEXP registers.
SCx_TWISTAT
SC1_TWISTAT
TWI Status Register
Address: 0x4000C844 Reset: 0x0
Address: 0x4000C044 Reset: 0x0
SC2_TWISTAT
TWI Status Register
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
0
22
0
21
0
20
0
19
18
17
16
0
0
0
0
15
0
14
0
13
0
12
0
11
10
9
8
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
SC_TWICMDFIN
SC_TWIRXFIN
SC_TWITXFIN
SC_TWIRXNAK
Bitname
Bitfield
Access
Description
SC_TWICMDFIN
[3]
R
This bit is set when a START or STOP command completes. It clears on the next TWI bus
activity.
SC_TWIRXFIN
SC_TWITXFIN
SC_TWIRXNAK
[2]
[1]
[0]
R
R
R
This bit is set when a byte is received. It clears on the next TWI bus activity.
This bit is set when a byte is transmitted. It clears on the next TWI bus activity.
This bit is set when a NACK is received from the slave. It clears on the next TWI bus
activity.
8-22
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_TWICTRL1
SC1_TWICTRL1
TWI Control Register 1
Address: 0x4000C84C Reset: 0x0
Address: 0x4000C04C Reset: 0x0
SC2_TWICTRL1
TWI Control Register 1
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
SC_TWISTOP
SC_TWISTART
SC_TWISEND
SC_TWIRECV
Bitname
Bitfield
[3]
Access
Description
SC_TWISTOP
SC_TWISTART
RW
RW
Setting this bit sends the STOP command. It clears when the command completes.
[2]
Setting this bit sends the START or repeated START command. It clears when the
command completes.
SC_TWISEND
SC_TWIRECV
[1]
[0]
RW
RW
Setting this bit transmits a byte. It clears when the command completes.
Setting this bit receives a byte. It clears when the command completes.
8-23
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_TWICTRL2
SC1_TWICTRL2
TWI Control Register 2
Address: 0x4000C850 Reset: 0x0
Address: 0x4000C050 Reset: 0x0
SC2_TWICTRL2
TWI Control Register 2
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
SC_TWIACK
Bitname
SC_TWIACK
Bitfield
Access
RW
Description
[0]
Setting this bit signals ACK after a received byte. Clearing this bit signals NACK after a
received byte.
8.6
UART - Universal Asynchronous Receiver / Transmitter
The SC1 UART is enabled by writing 1 to SC1_MODE. The SC2 serial controller does not include UART functions.
The UART supports the following features:
Flexible baud rate clock (300 bps to 921.6 kbps)
.
Data bits (7 or 8)
.
Parity bits (none, odd, or even)
.
Stop bits (1 or 2)
.
False start bit and noise filtering
.
Receive and transmit FIFOs
.
Optional RTS/CTS flow control
.
Receive and transmit DMA channels
.
8.6.1 GPIO Usage
The UART uses two signals to transmit and receive serial data:
TXD (Transmitted Data) – serial data sent by the EM35x
.
.
RXD (Received Data) – serial data received by the EM35x
If RTS/CTS flow control is enabled, these two signals are also used:
nRTS (Request To Send) – indicates the EM35x is able to receive data
.
.
nCTS (Clear To Send) – inhibits sending data from the EM35x if not asserted
8-24
120-035X-000 Rev. 1.2
Final
EM351 / EM357
The GPIO pins assigned to these signals are shown in Table 8-10.
Table 8-10. UART GPIO Usage
TXD
RXD
nCTS1
nRTS1
Direction
Output
Input
Input
Input
Output
GPIO Configuration
Alternate
Input
Alternate
Output (push-pull)
Output (push-pull)
SC1 pin
PB1
PB2
PB3
PB4
1 only used if RTS/CTS hardware flow control is enabled.
8.6.2 Set Up and Configuration
The UART baud rate clock is produced by a programmable baud generator starting from the 24 Hz clock:
24MHz
baud =
2N + F
The integer portion of the divisor, N, is written to the SC1_UARTPER register and the fractional part, F, to the
SC1_UARTFRAC register. Table 8-11 shows the values used to generate some common baud rates and their
associated clock frequency error. The UART requires an internal clock that is at least eight times the baud
rate clock, so the minimum allowable setting for SC1_UARTPER is 8.
Table 8-11. UART Baud Rate Divisors for Common Baud Rates
Baud Rate
(bits/sec)
SC1_UARTPER
SC1_UARTFRAC
Baud Rate Error (%)
300
40000
5000
2500
1250
625
312
208
104
52
0
0
0
0
0
1
1
0
0
0
0
0
0
2400
4800
0
9600
0
19200
38400
57600
115200
230400
460800
921600
0
0
- 0.08
+ 0.16
+ 0.16
+ 0.16
+ 0.16
26
13
The UART can miss bytes when the inter-byte gap is long or there is a baud rate mismatch between receiver
and transmitter. The UART may detect a parity and/or framing error on the corrupted byte, but there will not
necessarily be any error detected.
8-25
120-035X-000 Rev. 1.2
Final
EM351 / EM357
The UART is best operated in systems where the other side of the communication link also uses a crystal as its
timing reference, and baud rates should be selected to minimize the baud rate mismatch to the crystal
tolerance. Additionally, UART protocols should contain some form of error checking (for example CRC) at the
packet level to detect, and retry in the event of errors. Since the probability of corruption is low, there would
only be a small effect on UART throughput due to retries.
Errors may occur when:
106
baud∗Ferror)
Tgap ≥
(
where
Tgap = inter-byte gap in seconds
baud = baud rate in bps
Ferror = relative frequency error in ppm
For example, if the baud rate tolerance between receive and transmit is 200 ppm (reasonable if both sides are
derived from a crystal), and the baud rate is 115200 bps, then errors will not occur until the inter-byte gap
exceeds 43 ms. If the gap is exceeded then the chance of an error is essentially random, with a probability of
approximately P = baud / 24e6. At 115200 bps, the probability of corruption is 0.5%.
The UART character frame format is determined by four bits in the SC1_UARTCFG register:
SC_UART8BIT specifies the number of data bits in received and transmitted characters. If this bit is clear,
characters have 7 data bits; if set, characters have 8 data bits.
.
SC_UART2STP selects the number of stop bits in transmitted characters. (Only one stop bit is required in
received characters.) If this bit is clear, characters are transmitted with one stop bit; if set, characters are
.
transmitted with two stop bits.
SC_UARTPAR controls whether or not received and transmitted characters include a parity bit. If
SC_UARTPAR is clear, characters do not contain a parity bit, otherwise, characters do contain a parity bit.
.
SC_UARTODD specifies whether transmitted and received parity bits contain odd or even parity. If this bit
.
is clear, the parity bit is even, and if set, the parity bit is odd. Even parity is the exclusive-or of all of the
data bits, and odd parity is the inverse of the even parity value. SC_UARTODD has no effect if SC_UARTPAR
is clear.
A UART character frame contains, in sequence:
The start bit
.
The least significant data bit
.
The remaining data bits
.
If parity is enabled, the parity bit
.
The stop bit, or bits, if 2 stop bits are selected.
.
Figure 8-3 shows the UART character frame format, with optional bits indicated. Depending on the options
chosen for the character frame, the length of a character frame ranges from 9 to 12 bit times.
Note that asynchronous serial data may have arbitrarily long idle periods between characters. When idle,
serial data (TXD or RXD) is held in the high state. Serial data transitions to the low state in the start bit at the
beginning of a character frame.
8-26
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 8-3. UART Character Frame Format
UART Character Frame Format
(optional sections are in italics)
Next
Start Bit
or
TXD
or
RXD
Start
Bit
Data
Bit 0
Data
Bit 1
Data
Bit 2
Data
Bit 3
Data
Bit 4
Data
Bit 5
Data
Bit 6
Parity
Bit
Stop
Bit
Stop
Bit
Data
Bit 7
Idle time
IdleTime
8.6.3 FIFOs
Characters transmitted and received by the UART are buffered in the transmit and receive FIFOs that are both
4 entries deep (see Figure 8-4). When software writes a character to the SC1_DATA register, it is pushed onto
the transmit FIFO. Similarly, when software reads from the SC1_DATA register, the character returned is
pulled from the receive FIFO. If the transmit and receive DMA channels are used, the DMA channels also write
to and read from the transmit and receive FIFOs.
Figure 8-4. UART FIFOs
Receive Shift Register
Transmit Shift Register
TXD
RXD
Parity/Frame Errors
SC1_DATA (read)
SC1_DATA (write)
SC1_UARTSTAT
CPU and DMA
Channel Access
8.6.4 RTS/CTS Flow control
RTS/CTS flow control, also called hardware flow control, uses two signals (nRTS and nCTS) in addition to
received and transmitted data (see Figure 8-5). Flow control is used by a data receiver to prevent buffer
overflow, by signaling an external device when it is and is not allowed to transmit.
Figure 8-5. RTS/CTS Flow Control Connections
EM350
Other Device
RXD
TXD
UART Receiver
UART Transmitter
nRTS
nCTS
TXD
RXD
UART Transmitter
UART Receiver
nCTS
nRTS
The UART RTS/CTS flow control options are selected by the SC_UARTFLOW and SC_UARTAUTO bits in the
SC1_UARTCFG register (see Table 8-12). Whenever the SC_UARTFLOW bit is set, the UART will not start
transmitting a character unless nCTS is low (asserted). If nCTS transitions to the high state (deasserts) while a
character is being transmitted, transmission of that character continues until it is complete.
8-27
120-035X-000 Rev. 1.2
Final
EM351 / EM357
If the SC_UARTAUTO bit is set, nRTS is controlled automatically by hardware: nRTS is put into the low state
(asserted) when the receive FIFO has room for at least two characters, otherwise is it in the high state
(unasserted). If SC_UARTAUTO is clear, software controls the nRTS output by setting or clearing the
SC_UARTRTS bit in the SC1_UARTCFG register. Software control of nRTS is useful if the external serial device
cannot stop transmitting characters promptly when nRTS is set to the high state (deasserted).
Table 8-12. UART RTS/CTS Flow Control Configurations
SC1_UARTCFG
SC_UARTxxx1
FLOW AUTO RTS Pins Used
Operating Mode
0
1
-
-
TXD, RXD
No RTS/CTS flow control
0
0/1
TXD, RXD, Flow control using RTS/CTS with software control of nRTS:
nCTS, nRTS nRTS controlled by SC_UARTRTS bit in SC1_UARTCFG register
1
1
-
TXD, RXD, Flow control using RTS/CTS with hardware control of nRTS:
nCTS, nRTS nRTS is asserted if room for at least 2 characters in receive FIFO
1 The notation xxx means that the corresponding column header below is inserted to form the field name.
8.6.5 DMA
The DMA Channels section describes how to configure and use the serial receive and transmit DMA channels.
The receive DMA channel has special provisions to record UART receive errors. When the DMA channel
transfers a character from the receive FIFO to a buffer in memory, it checks the stored parity and frame error
status flags. When an error is flagged, the SC1_RXERRA/B register is updated, marking the offset to the first
received character with a parity or frame error. Similarly if a receive overrun error occurs, the SC1_RXERRA/B
registers mark the error offset. The receive FIFO hardware generates the INT_SCRXOVF interrupt and DMA
status register indicates the error immediately, but in this case the error offset is 4 characters ahead of the
actual overflow at the input to the receive FIFO. Two conditions will clear the error indication: setting the
appropriate SC_RXDMARST bit in the SC1_DMACTRL register, or loading the appropriate DMA buffer after it has
unloaded.
8.6.6 Interrupts
UART interrupts are generated on the following events:
Transmit FIFO empty and last character shifted out (depending on SCx_INTMODE, either the 0 to 1
transition or the high level of SC_UARTTXIDLE)
.
Transmit FIFO changed from full to not full (depending on SCx_INTMODE, either the 0 to 1 transition or the
high level of SC_UARTTXFREE)
.
Receive FIFO changed from empty to not empty (depending on SCx_INTMODE, either the 0 to 1 transition
or the high level of SC_UARTRXVAL)
.
Transmit DMA buffer A/B complete (1 to 0 transition of SC_TXACTA/B)
.
Receive DMA buffer A/B complete (1 to 0 transition of SC_RXACTA/B)
.
Character received with parity error
.
Character received with frame error
.
Character received and lost when receive FIFO was full (receive overrun error)
.
To enable CPU interrupts, set the desired interrupt bits in the second-level INT_SCxCFG register, and enable
the top-level SCx interrupt in the NVIC by writing the INT_SCx bit in the INT_CFGSET register.
8-28
120-035X-000 Rev. 1.2
Final
EM351 / EM357
8.6.7 Registers
Refer to Registers (in the SPI Master Mode section) for a description of the SCx_DATA register.
SC1_UARTSTAT
UART Status Register
Address: 0x4000C848 Reset: 0x40
31
0
30
29
0
28
0
27
0
26
0
25
0
24
0
0
22
0
23
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
SC_UARTTXIDLE
SC_UARTPARERR
SC_UARTFRMERR
SC_UARTRXOVF
SC_UARTTXFREE
SC_UARTRXVAL
SC_UARTCTS
Bitname
Bitfield
Access
Description
SC_UARTTXIDLE
SC_UARTPARERR
[6]
[5]
R
R
This bit is set when both the transmit FIFO and the transmit serializer are empty.
This bit is set when the byte in the data register was received with a parity error. This bit
is updated when the data register is read, and is cleared if the receive FIFO is empty.
SC_UARTFRMERR
SC_UARTRXOVF
[4]
[3]
R
R
This bit is set when the byte in the data register was received with a frame error. This bit
is updated when the data register is read, and is cleared if the receive FIFO is empty.
This bit is set when the receive FIFO has been overrun. This occurs if a byte is received
when the receive FIFO is full. This bit is cleared by reading the data register.
SC_UARTTXFREE
SC_UARTRXVAL
SC_UARTCTS
[2]
[1]
[0]
R
R
R
This bit is set when the transmit FIFO has space for at least one byte.
This bit is set when the receive FIFO contains at least one byte.
This bit shows the logical state (not voltage level) of the nCTS input:
0: nCTS is deasserted (pin is high, 'XOFF', RS232 negative voltage); the UART is inhibited
from starting to transmit a byte.
1: nCTS is asserted (pin is low, 'XON', RS232 positive voltage); the UART may transmit.
8-29
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SC1_UARTCFG
UART Configuration Register
Address: 0x4000C85C Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
SC_UARTAUTO
SC_UARTFLOW
SC_UARTODD
SC_UARTPAR
SC_UART2STP
SC_UART8BIT
SC_UARTRTS
Bitname
Bitfield
Access
Description
SC_UARTAUTO
[6]
RW
Set this bit to enable automatic nRTS control by hardware (SC_UARTFLOW must also be
set). When automatic control is enabled, nRTS will be deasserted when the receive FIFO
has space for only one more byte (inhibits transmission from the other device) and will be
asserted if it has space for more than one byte (enables transmission from the other
device). The SC_UARTRTS bit in this register has no effect if this bit is set.
SC_UARTFLOW
SC_UARTODD
[5]
[4]
RW
RW
Set this bit to enable using nRTS/nCTS flow control signals. Clear this bit to disable the
signals. When this bit is clear, the UART transmitter will not be inhibited by nCTS.
If parity is enabled, specifies the kind of parity.
0: Even parity.
1: Odd parity.
SC_UARTPAR
SC_UART2STP
SC_UART8BIT
SC_UARTRTS
[3]
[2]
[1]
[0]
RW
RW
RW
RW
Specifies whether to use parity bits.
0: Don't use parity.
1: Use parity.
Number of stop bits transmitted.
0: 1 stop bit.
1: 2 stop bits.
Number of data bits.
0: 7 data bits.
1: 8 data bits.
nRTS is an output to control the flow of serial data sent to the EM35x from another
device. This bit directly controls the output at the nRTS pin (SC_UARTFLOW must be set
and SC_UARTAUTO must be cleared). When this bit is set, nRTS is asserted (pin is low,
'XON', RS232 positive voltage); the other device's transmission is enabled. When this bit is
cleared, nRTS is deasserted (pin is high, 'XOFF', RS232 negative voltage), the other
device's transmission is inhibited.
8-30
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SC1_UARTPER
UART Baud Rate Period Register
Address: 0x4000C868 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
SC_UARTPER
SC_UARTPER
7
6
5
4
3
2
1
0
Bitname
SC_UARTPER
Bitfield
Access
RW
Description
[15:0]
The integer part of baud rate period (N) in the equation:
rate = 24MHz / ( (2 * N) + F )
8-31
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SC1_UARTFRAC
UART Baud Rate Fractional Period Register
Address: 0x4000C86C Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
SC_UARTFRAC
Bitname
SC_UARTFRAC
Bitfield
Access
RW
Description
[0]
The fractional part of the baud rate period (F) in the equation:
rate = 24MHz / ( (2 * N) + F )
8.7
DMA Channels
The EM35x serial DMA channels enable efficient, high-speed operation of the SPI and UART controllers by
reducing the load on the CPU as well as decreasing the frequency of interrupts that it must service. The
transmit and receive DMA channels can transfer data between the transmit and receive FIFOs and the DMA
buffers in main memory as quickly as it can be transmitted or received. Once software defines, configures,
and activates the DMA, it only needs to handle an interrupt when a transmit buffer has been emptied or a
receive buffer has been filled. The DMA channels each support two memory buffers, labeled A and B, and can
alternate (“ping-pong”) between them automatically to allow continuous communication without critical
interrupt timing.
Note: DMA memory buffer terminology
load - make a buffer available for the DMA channel to use
.
pending – a buffer loaded but not yet active
.
active - the buffer that will be used for the next DMA transfer
.
unload – DMA channel action when it has finished with a buffer
.
idle – a buffer that has not been loaded, or has been unloaded
.
To use a DMA channel, software should follow these steps:
Reset the DMA channel by setting the SC_TXDMARST (or SC_RXDMARST) bit in the SCx_DMACTRL register.
.
.
Set up the DMA buffers. The two DMA buffers, A and B, are defined by writing the start address to
SCx_TXBEGA/B (or SCx_RXBEGA/B) and the (inclusive) end address to SCx_TXENDA/B (or SCx_RXENDA/B).
Note that DMA buffers must be in RAM.
Configure and initialize SCx for the desired operating mode.
.
.
Enable second-level interrupts triggered when DMA buffers unload by setting the INT_SCTXULDA/B (or
INT_SCRXULDA/B) bits in the INT_SCxFLAG register.
Enable top-level NVIC interrupts by setting the INT_SCx bit in the INT_CFGSET register.
.
.
Start the DMA by loading the DMA buffers by setting the SC_TXLODA/B (or SC_RXLODA/B) bits in the
SCx_DMACTRL register.
8-32
120-035X-000 Rev. 1.2
Final
EM351 / EM357
A DMA buffer’s end address, SCx_TXENDA/B (or SCx_RXENDA/B), can be written while the buffer is loaded or
active. This is useful for receiving messages that contain an initial byte count, since it allows software to set
the buffer end address at the last byte of the message.
As the DMA channel transfers data between the transmit or receive FIFO and a memory buffer, the DMA count
register contains the byte offset from the start of the buffer to the address of the next byte that will be
written or read. A transmit DMA channel has a single DMA count register (SCx_TXCNT) that applies to
whichever transmit buffer is active, but a receive DMA channel has two DMA count registers (SCx_RXCNTA/B),
one for each receive buffer. The DMA count register contents are preserved until the corresponding buffer, or
either buffer in the case of the transmit DMA count, is loaded, or until the DMA is reset.
The receive DMA count register may be written while the corresponding buffer is loaded. If the buffer is not
loaded, writing the DMA count register also loads the buffer while preserving the count value written. This
feature can simplify handling UART receive errors.
The DMA channel stops using a buffer and unloads it when the following is true:
(DMA buffer start address + DMA buffer count) > DMA buffer end address
Typically a transmit buffer is unloaded after all its data has been sent, and a receive buffer is unloaded after
it is filled with data, but writing to the buffer end address or buffer count registers can also cause a buffer to
unload early.
Serial controller DMA channels include additional features specific to the SPI and UART operation and are
described in those sections.
8-33
120-035X-000 Rev. 1.2
Final
EM351 / EM357
8.7.1 Registers
SCx_DMACTRL
SC1_DMACTRL
Serial DMA Control Register
Address: 0x4000C830 Reset: 0x0
Address: 0x4000C030 Reset: 0x0
SC2_DMACTRL
Serial DMA Control Register
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
0
22
0
21
20
19
18
17
16
0
0
0
0
0
0
15
0
14
0
13
12
11
10
9
8
0
0
0
3
0
0
1
0
7
6
5
4
2
0
0
0
SC_TXDMARST
SC_RXDMARST
SC_TXLODB
SC_TXLODA
SC_RXLODB
SC_RXLODA
Bitname
Bitfield
[5]
Access
W
Description
SC_TXDMARST
SC_RXDMARST
SC_TXLODB
Setting this bit resets the transmit DMA. The bit clears automatically.
Setting this bit resets the receive DMA. The bit clears automatically.
[4]
W
[3]
RW
Setting this bit loads DMA transmit buffer B addresses and allows the DMA controller to
start processing transmit buffer B. If both buffer A and B are loaded simultaneously,
buffer A will be used first. This bit is cleared when DMA completes. Writing a zero to this
bit has no effect.
Reading this bit returns DMA buffer status:
0: DMA processing is complete or idle.
1: DMA processing is active or pending.
SC_TXLODA
SC_RXLODB
SC_RXLODA
[2]
[1]
[0]
RW
RW
RW
Setting this bit loads DMA transmit buffer A addresses and allows the DMA controller to
start processing transmit buffer A. If both buffer A and B are loaded simultaneously,
buffer A will be used first. This bit is cleared when DMA completes. Writing a zero to this
bit has no effect.
Reading this bit returns DMA buffer status:
0: DMA processing is complete or idle.
1: DMA processing is active or pending.
Setting this bit loads DMA receive buffer B addresses and allows the DMA controller to
start processing receive buffer B. If both buffer A and B are loaded simultaneously, buffer
A will be used first. This bit is cleared when DMA completes. Writing a zero to this bit has
no effect.
Reading this bit returns DMA buffer status:
0: DMA processing is complete or idle.
1: DMA processing is active or pending.
Setting this bit loads DMA receive buffer A addresses and allows the DMA controller to
start processing receive buffer A. If both buffer A and B are loaded simultaneously, buffer
A will be used first. This bit is cleared when DMA completes. Writing a zero to this bit has
no effect.
Reading this bit returns DMA buffer status:
0: DMA processing is complete or idle.
1: DMA processing is active or pending.
8-34
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_DMASTAT
SC1_DMASTAT
Serial DMA Status Register
Address: 0x4000C82C Reset: 0x0
Address: 0x4000C02C Reset: 0x0
SC2_DMASTAT
Serial DMA Status Register
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
22
21
20
0
19
18
0
17
16
0
0
0
0
11
0
0
15
14
13
12
10
9
8
0
7
0
6
0
5
SC_RXSSEL
3
SC_RXFRMB
1
SC_RXFRMA
0
4
2
SC_RXPARB
SC_RXPARA
SC_RXOVFB
SC_RXOVFA
SC_TXACTB
SC_TXACTA
SC_RXACTB
SC_RXACTA
Bitname
Bitfield
Access
Description
SC_RXSSEL
[12:10]
R
Status of the receive count saved in SCx_RXCNTSAVED (SPI slave mode) when nSSEL
deasserts. Cleared when a receive buffer is loaded and when the receive DMA is reset.
0: No count was saved because nSSEL did not deassert.
2: Buffer A's count was saved, nSSEL deasserted once.
3: Buffer B's count was saved, nSSEL deasserted once.
6: Buffer A's count was saved, nSSEL deasserted more than once.
7: Buffer B's count was saved, nSSEL deasserted more than once.
1, 4, 5: Reserved.
SC_RXFRMB
SC_RXFRMA
SC_RXPARB
SC_RXPARA
SC_RXOVFB
[9]
[8]
[7]
[6]
[5]
R
R
R
R
R
This bit is set when DMA receive buffer B reads a byte with a frame error from the receive
FIFO. It is cleared the next time buffer B is loaded or when the receive DMA is reset. (SC1
in UART mode only)
This bit is set when DMA receive buffer A reads a byte with a frame error from the receive
FIFO. It is cleared the next time buffer A is loaded or when the receive DMA is reset. (SC1
in UART mode only)
This bit is set when DMA receive buffer B reads a byte with a parity error from the receive
FIFO. It is cleared the next time buffer B is loaded or when the receive DMA is reset. (SC1
in UART mode only)
This bit is set when DMA receive buffer A reads a byte with a parity error from the receive
FIFO. It is cleared the next time buffer A is loaded or when the receive DMA is reset. (SC1
in UART mode only)
This bit is set when DMA receive buffer B was passed an overrun error from the receive
FIFO. Neither receive buffer was capable of accepting any more bytes (unloaded), and
the FIFO filled up. Buffer B was the next buffer to load, and when it drained the FIFO the
overrun error was passed up to the DMA and flagged with this bit. Cleared the next time
buffer B is loaded and when the receive DMA is reset.
SC_RXOVFA
[4]
R
This bit is set when DMA receive buffer A was passed an overrun error from the receive
FIFO. Neither receive buffer was capable of accepting any more bytes (unloaded), and
the FIFO filled up. Buffer A was the next buffer to load, and when it drained the FIFO the
overrun error was passed up to the DMA and flagged with this bit. Cleared the next time
buffer A is loaded and when the receive DMA is reset.
SC_TXACTB
SC_TXACTA
SC_RXACTB
SC_RXACTA
[3]
[2]
[1]
[0]
R
R
R
R
This bit is set when DMA transmit buffer B is active.
This bit is set when DMA transmit buffer A is active.
This bit is set when DMA receive buffer B is active.
This bit is set when DMA receive buffer A is active.
8-35
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_TXBEGA
SC1_TXBEGA
Transmit DMA Begin Address Register A
Address: 0x4000C810 Reset: 0x20000000
Address: 0x4000C010 Reset: 0x20000000
SC2_TXBEGA
Transmit DMA Begin Address Register A
31
0
30
0
29
1
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_TXBEGA
7
6
5
4
3
2
1
0
SC_TXBEGA
Bitname
SC_TXBEGA
Bitfield
Access
RW
Description
[13:0]
DMA transmit buffer A start address.
8-36
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_TXBEGB
SC1_TXBEGB
Transmit DMA Begin Address Register B
Address: 0x4000C818 Reset: 0x20000000
Address: 0x4000C018 Reset: 0x20000000
SC2_TXBEGB
Transmit DMA Begin Address Register B
31
0
30
0
29
1
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_TXBEGB
7
6
5
4
3
2
1
0
SC_TXBEGB
Bitname
SC_TXBEGB
Bitfield
Access
RW
Description
[13:0]
DMA transmit buffer B start address.
8-37
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_TXENDA
SC1_TXENDA
Transmit DMA End Address Register A
Address: 0x4000C814 Reset: 0x20000000
Address: 0x4000C014 Reset: 0x20000000
SC2_TXENDA
Transmit DMA End Address Register A
31
0
30
0
29
1
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_TXENDA
7
6
5
4
3
2
1
0
SC_TXENDA
Bitname
SC_TXENDA
Bitfield
Access
RW
Description
Address of the last byte that will be read from the DMA transmit buffer A.
[13:0]
8-38
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_TXENDB
SC1_TXENDB
Transmit DMA End Address Register B
Address: 0x4000C81C Reset: 0x20000000
Address: 0x4000C01C Reset: 0x20000000
SC2_TXENDB
Transmit DMA End Address Register B
31
0
30
0
29
1
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_TXENDB
7
6
5
4
3
2
1
0
SC_TXENDB
Bitname
SC_TXENDB
Bitfield
Access
RW
Description
Address of the last byte that will be read from the DMA transmit buffer B.
[13:0]
8-39
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_TXCNT
SC1_TXCNT
Transmit DMA Count Register
Address: 0x4000C828 Reset: 0x0
Address: 0x4000C028 Reset: 0x0
SC2_TXCNT
Transmit DMA Count Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_TXCNT
7
6
5
4
3
2
1
0
SC_TXCNT
Bitname
SC_TXCNT
Bitfield
Access
Description
[13:0]
R
The offset from the start of the active DMA transmit buffer from which the next byte will
be read. This register is set to zero when the buffer is loaded and when the DMA is reset.
8-40
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_RXBEGA
SC1_RXBEGA
Receive DMA Begin Address Register A
Address: 0x4000C800 Reset: 0x20000000
Address: 0x4000C000 Reset: 0x20000000
SC2_RXBEGA
Receive DMA Begin Address Register A
31
0
30
0
29
1
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_RXBEGA
7
6
5
4
3
2
1
0
SC_RXBEGA
Bitname
SC_RXBEGA
Bitfield
Access
RW
Description
[13:0]
DMA receive buffer A start address.
8-41
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_RXBEGB
SC1_RXBEGB
Receive DMA Begin Address Register B
Address: 0x4000C808 Reset: 0x20000000
Address: 0x4000C008 Reset: 0x20000000
SC2_RXBEGB
Receive DMA Begin Address Register B
31
0
30
0
29
1
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_RXBEGB
7
6
5
4
3
2
1
0
SC_RXBEGB
Bitname
SC_RXBEGB
Bitfield
Access
RW
Description
[13:0]
DMA receive buffer B start address.
8-42
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_RXENDA
SC1_RXENDA
Receive DMA End Address Register A
Address: 0x4000C804 Reset: 0x20000000
SC2_RXENDA
Receive DMA End Address Register Address: 0x4000C004 Reset: 0x20000000
31
0
30
0
29
1
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_RXENDA
7
6
5
4
3
2
1
0
SC_RXENDA
Bitname
SC_RXENDA
Bitfield
Access
RW
Description
Address of the last byte that will be written in the DMA receive buffer A.
[13:0]
8-43
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_RXENDB
SC1_RXENDB
Receive DMA End Address Register B
Address: 0x4000C80C Reset: 0x20000000
Address: 0x4000C00C Reset: 0x20000000
SC2_RXENDB
Receive DMA End Address Register B
31
0
30
0
29
1
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_RXENDB
7
6
5
4
3
2
1
0
SC_RXENDB
Bitname
SC_RXENDB
Bitfield
Access
RW
Description
Address of the last byte that will be written in the DMA receive buffer B.
[13:0]
8-44
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_RXCNTA
SC1_RXCNTA
Receive DMA Count Register A
Address: 0x4000C820 Reset: 0x0
Address: 0x4000C020 Reset: 0x0
SC2_RXCNTA
Receive DMA Count Register A
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_RXCNTA
7
6
5
4
3
2
1
0
SC_RXCNTA
Bitname
SC_RXCNTA
Bitfield
Access
RW
Description
[13:0]
The offset from the start of DMA receive buffer A at which the next byte will be written.
This register is set to zero when the buffer is loaded and when the DMA is reset. If this
register is written when the buffer is not loaded, the buffer is loaded.
8-45
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_RXCNTB
SC1_RXCNTB
Receive DMA Count Register B
Address: 0x4000C824 Reset: 0x0
Address: 0x4000C024 Reset: 0x0
SC2_RXCNTB
Receive DMA Count Register B
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_RXCNTB
7
6
5
4
3
2
1
0
SC_RXCNTB
Bitname
SC_RXCNTB
Bitfield
Access
RW
Description
[13:0]
The offset from the start of DMA receive buffer B at which the next byte will be written.
This register is set to zero when the buffer is loaded and when the DMA is reset. If this
register is written when the buffer is not loaded, the buffer is loaded.
8-46
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_RXCNTSAVED
SC1_RXCNTSAVED
Saved Receive DMA Count Register
Address: 0x4000C870 Reset: 0x0
Address: 0x4000C070 Reset: 0x0
SC2_RXCNTSAVED
Saved Receive DMA Count Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_RXCNTSAVED
7
6
5
4
3
2
1
0
SC_RXCNTSAVED
Bitname
SC_RXCNTSAVED
Bitfield
Access
Description
[13:0]
R
Receive DMA count saved in SPI slave mode when nSSEL deasserts. The count is only saved
the first time nSSEL deasserts.
8-47
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_RXERRA
SC1_RXERRA
DMA First Receive Error Register A
Address: 0x4000C834 Reset: 0x0
Address: 0x4000C034 Reset: 0x0
SC2_RXERRA
DMA First Receive Error Register A
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_RXERRA
7
6
5
4
3
2
1
0
SC_RXERRA
Bitname
SC_RXERRA
Bitfield
Access
Description
[13:0]
R
The offset from the start of DMA receive buffer A of the first byte received with a parity,
frame, or overflow error. Note that an overflow error occurs at the input to the receive
FIFO, so this offset is 4 bytes before the overflow position. If there is no error, it reads
zero. This register will not be updated by subsequent errors until the buffer unloads and
is reloaded, or the receive DMA is reset.
8-48
120-035X-000 Rev. 1.2
Final
EM351 / EM357
SCx_RXERRB
SC1_RXERRB
DMA First Receive Error Register B
Address: 0x4000C838 Reset: 0x0
Address: 0x4000C038 Reset: 0x0
SC2_RXERRB
DMA First Receive Error Register B
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
SC_RXERRB
7
6
5
4
3
2
1
0
SC_RXERRB
Bitname
SC_RXERRB
Bitfield
Access
Description
[13:0]
R
The offset from the start of DMA receive buffer B of the first byte received with a parity,
frame, or overflow error. Note that an overflow error occurs at the input to the receive
FIFO, so this offset is 4 bytes before the overflow position. If there is no error, it reads
zero. This register will not be updated by subsequent errors until the buffer unloads and
is reloaded, or the receive DMA is reset.
8-49
120-035X-000 Rev. 1.2
Final
EM351 / EM357
9 General Purpose Timers (TIM1 and TIM2)
9.1 Introduction
Each of the EM35x’s two general-purpose timers consists of a 16-bit auto-reload counter driven by a
programmable prescaler. They may be used for a variety of purposes, including measuring the pulse lengths of
input signals (input capture) or generating output waveforms (output compare and PWM). Pulse lengths and
waveform periods can be modulated from a few microseconds to several milliseconds using the timer
prescaler. The timers are completely independent, and do not share any resources. They can be synchronized
together as described in the Timer Synchronization section.
The two general-purpose timers, TIM1 and TIM2, have the following features:
16-bit up, down, or up/down auto-reload counter.
.
Programmable prescaler to divide the counter clock by any power of two from 1 through 32768.
.
4 independent channels for:
.
Input capture
•
Output compare
•
PWM generation (edge- and center-aligned mode)
•
One-pulse mode output
•
Synchronization circuit to control the timer with external signals and to interconnect the timers.
.
.
Flexible clock source selection:
Peripheral clock (PCLK at 6 or 12 MHz)
•
•
•
32.768 kHz external clock (if available)
1 kHz clock
GPIO input
•
Interrupt generation on the following events:
.
Update: counter overflow/underflow, counter initialization (software or internal/external trigger)
•
Trigger event (counter start, stop, initialization or count by internal/external trigger)
•
•
•
Input capture
Output compare
Supports incremental (quadrature) encoders and Hall sensors for positioning applications.
.
.
Trigger input for external clock or cycle-by-cycle current management.
Figure 9-1 shows an overview of a timer’s internal structure.
Note: Because the two timers are identical, the notation TIMx refers to either TIM1 or TIM2. For example,
TIMx_PSC refers to both TIM1_PSC and TIM2_PSC. Similarly, “y” refers to any of the four channels of a given
timer, so for example, OCy refers to OC1, OC2, OC3, and OC4.
9-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 9-1. General-Purpose Timer Block Diagram
Note: The internal signals shown in Figure 9-1 are described in the Timer Signal Descriptions section, and are
used throughout the text to describe how the timer components are interconnected.
9-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
9.2 GPIO Usage
The timers can optionally use GPIOs in the PA and PB ports for external inputs or outputs. As with all EM35x
digital inputs, a GPIO used as a timer input can be shared with other uses of the same pin. Available timer
inputs include an external timer clock, a clock mask, and four input channels. Any GPIO used as a timer output
must be configured as an alternate output and is controlled only by the timer.
Many of the GPIOs that can be assigned as timer outputs can also be used by another on-chip peripheral such
as a serial controller. Using a GPIO as a timer output takes precedence over another peripheral function, as
long as the channel is configured as an output in the TIMx_CCMR1 register and is enabled in the TIMx_CCER
register.
The GPIOs that can be used by Timer 1 are fixed, but the GPIOs that can be used as Timer 2 channels can be
mapped to either of two pins, as shown in Table 9-1. The Timer 2 Option Register (TIM2_OR) has four single
bit fields (TIM_REMAPCy) that control whether a Timer 2 channel is mapped to its default GPIO in port PA, or
remapped to a GPIO in PB.
Table 9-1 specifies the pins that may be assigned to Timer 1 and Timer 2 functions.
Table 9-1. Timer GPIO Usage
Signal
(direction)
TIMxC1
(in or out) (in or out)
TIMxC2
TIMxC3
(in or out)
TIMxC4
(in or out)
TIMxCLK
(in)
TIMxMSK
(in)
Timer 1
PB6
PA0
PB7
PA3
PA6
PA1
PA7
PA2
PB0
PB5
PB5
PB0
Timer 2
(TIM_REMAPCy = 0)
Timer 2
PB1
PB2
PB3
PB4
PB5
PB0
(TIM_REMAPCy = 1)
The TIMxCLK and TIMxMSK inputs can be used only in the external clock modes; refer to the External Clock
Source Mode 1 and External Clock Source Mode 2 sections for details concerning their use.
9.3 Timer Functional Description
9.3.1 Time-Base Unit
The main block of the general purpose timer is a 16-bit counter with its related auto-reload register. The
counter can count up, down, or alternate up and down. The counter clock can be divided by a prescaler.
The counter, the auto-reload register, and the prescaler register can be written to or read by software. This is
true even when the counter is running.
The time-base unit includes:
Counter Register (TIMx_CNT)
.
Prescaler Register (TIMx_PSC)
.
Auto-Reload Register (TIMx_ARR)
.
Some timer registers cannot be directly accessed by software, which instead reads and writes a “buffer
register”. The internal registers actually used for timer operations are called “shadow registers”.
The auto-reload register is buffered. Writing to or reading from the auto-reload register accesses the buffer
register. The contents of the buffer register are transferred into the shadow register permanently or at each
update event (UEV), depending on the auto-reload buffer enable bit (TIM_ARBE) in the TIMx_CR1 register. The
9-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
UEV is generated when both the counter reaches the overflow (or underflow when down-counting) and when
the TIM_UDIS bit equals 0 in the TIMx_CR1 register. It can also be generated by software. UEV generation is
described in detail for each configuration.
The counter is clocked by the prescaler output CK_CNT, which is enabled only when the counter enable bit
(TIM_CEN) in the TIMx_CR1 register is set. Refer also to the slave mode controller description in the Timers
and External Trigger Synchronization section to get more details on counter enabling.
Note that the actual counter enable signal CNT_EN is set one clock cycle after TIM_CEN.
Note: When the EM35x enters debug mode and the ARM® CortexTM-M3 core is halted, the counters continue to
run normally.
9.3.1.1
Prescaler
The prescaler can divide the counter clock frequency by power of two from 1 through 32768. It is based on a
16-bit counter controlled through the 4-bit TIM_PSCEXP bit field in the TIMx_PSC register. The factor by which
the counter is divided is two raised to the power TIM_PSCEXP (2TIM_PSCEXP).
It can be changed on the fly as this control register is buffered. The new prescaler ratio is used starting at the
next UEV.
Figure 9-2 gives an example of the counter behavior when the prescaler ratio is changed on the fly.
Figure 9-2. Counter Timing Diagram with Prescaler Division Change from 1 to 4
9.3.2 Counter Modes
9.3.2.1
Up-Counting Mode
In up-counting mode, the counter counts from 0 to the auto-reload value (contents of the TIMx_ARR register),
then restarts from 0 and generates a counter overflow event.
A UEV can be generated at each counter overflow, by setting the TIM_UG bit in the TIMx_EGR register, or by
using the slave mode controller.
Software can disable the UEV by setting the TIM_UDIS bit in the TIMx_CR1 register, to avoid updating the
shadow registers while writing new values in the buffer registers. No UEV will occur until the TIM_UDIS bit is
written to 0. Both the counter and the prescaler counter restart from 0, but the prescale rate does not
change. In addition, if the TIM_URS bit in the TIMx_CR1 register is set, setting the TIM_UG bit generates a UEV
but without setting the INT_TIMUIF flag. Thus no interrupt request is sent. This avoids generating both update
and capture interrupts when clearing the counter on the capture event.
9-4
120-035X-000 Rev. 1.2
Final
EM351 / EM357
When a UEV occurs, the update flag (the INT_TIMUIF bit in the INT_TIMxFLAG register) is set (unless TIM_URS
is 1) and the following registers are updated:
The buffer of the prescaler is reloaded with the buffer value (contents of the TIMx_PSC register).
.
.
The auto-reload shadow register is updated with the buffer value (TIMx_ARR).
Figure 9-3, Figure 9-4, Figure 9-5, and Figure 9-6 show some examples of the counter behavior for different
clock frequencies when TIMx_ARR = 0x36.
Figure 9-3. Counter Timing Diagram, Internal Clock Divided by 1
Figure 9-4. Counter Timing Diagram, Internal Clock Divided by 4
Figure 9-5. Counter Timing Diagram, Update Event when TIM_ARBE = 0 (TIMx_ARR not buffered)
9-5
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 9-6. Counter Timing Diagram, Update Event when TIM_ARBE = 1 (TIMx_ARR buffered)
9.3.2.2
Down-Counting Mode
In down-counting mode, the counter counts from the auto-reload value (contents of the TIMx_ARR register)
down to 0, then restarts from the auto-reload value and generates a counter underflow event.
A UEV can be generated at each counter underflow, by setting the TIM_UG bit in the TIMx_EGR register, or by
using the slave mode controller. Software can disable the UEV by setting the TIM_UDIS bit in the TIMx_CR1
register, to avoid updating the shadow registers while writing new values in the buffer registers. No UEV
occurs until the TIM_UDIS bit is written to 0. However, the counter restarts from the current auto-reload
value, whereas the prescaler’s counter restarts from 0, but the prescale rate doesn’t change.
In addition, if the TIM_URS bit in the TIMx_CR1 register is set, setting the TIM_UG bit generates a UEV, but
without setting the INT_TIMUIF flag. Thus no interrupt request is sent. This avoids generating both update and
capture interrupts when clearing the counter on the capture event.
When a UEV occurs, the update flag (the INT_TIMUIF bit in the INT_TIMxFLAG register) is set (unless TIM_URS
is 1) and the following registers are updated:
The prescaler shadow register is reloaded with the buffer value (contents of the TIMx_PSC register).
.
.
The auto-reload active register is updated with the buffer value (contents of the TIMx_ARR register). The
auto-reload is updated before the counter is reloaded, so that the next period is the expected one.
Figure 9-7 and Figure 9-8 show some examples of the counter behavior for different clock frequencies when
TIMx_ARR = 0x36.
9-6
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 9-7. Counter Timing Diagram, Internal Clock Divided by 1
Figure 9-8. Counter Timing Diagram, Internal Clock Divided by 4
9.3.2.3
Center-Aligned Mode (Up/Down Counting)
In center-aligned mode, the counter counts from 0 to the auto-reload value (contents of the TIMx_ARR
register) – 1 and generates a counter overflow event, then counts from the autoreload value down to 1 and
generates a counter underflow event. Then it restarts counting from 0.
In this mode, the direction bit (TIM_DIR in the TIMx_CR1 register) cannot be written. It is updated by hardware
and gives the current direction of the counter.
The UEV can be generated at each counter overflow and at each counter underflow. Setting the TIM_UG bit in
the TIMx_EGR register by software or by using the slave mode controller also generates a UEV. In this case,
the both the counter and the prescaler’s counter restart counting from 0.
Software can disable the UEV by setting the TIM_UDIS bit in the TIMx_CR1 register. This avoids updating the
shadow registers while writing new values in the buffer registers. Then no UEV occurs until the TIM_UDIS bit
has been written to 0. However, the counter continues counting up and down, based on the current auto-
reload value.
In addition, if the TIM_URS bit in the TIMx_CR1 register is set, setting the TIM_UG bit generates a UEV, but
without setting the INT_TIMUIF flag. Thus no interrupt request is sent. This avoids generating both update and
capture interrupt when clearing the counter on the capture event.
When a UEV occurs, the update flag (the INT_TIMUIF bit in the INT_TIMxFLAG register) is set (unless TIM_URS
is 1) and the following registers are updated:
The prescaler shadow register is reloaded with the buffer value (contents of the TIMx_PSC register).
.
9-7
120-035X-000 Rev. 1.2
Final
EM351 / EM357
The auto-reload active register is updated with the buffer value (contents of the TIMx_ARR register). If the
update source is a counter overflow, the auto-reload is updated before the counter is reloaded, so that the
next period is the expected one. The counter is loaded with the new value.
.
Figure 9-9, Figure 9-10, and Figure 9-11 show some examples of the counter behavior for different clock
frequencies.
Figure 9-9. Counter Timing Diagram, Internal Clock Divided by 1, TIMx_ARR = 0x6
Figure 9-10. Counter Timing Diagram, Update Event with TIM_ARBE = 1 (counter underflow)
9-8
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 9-11. Counter Timing Diagram, Update Event with TIM_ARBE = 1 (counter overflow)
9.3.3 Clock Selection
The counter clock can be provided by the following clock sources:
Internal clock (PCLK)
.
External clock mode 1: external input pin (TIy)
.
External clock mode 2: external trigger input (ETR)
.
.
Internal trigger input (ITR0): using the other timer as prescaler. Refer to the section Using One Timer as
Prescaler for the Other Timer for more details.
9.3.3.1
Internal Clock Source (CK_INT)
The internal clock is selected when the slave mode controller is disabled (TIM_SMS = 000 in the TIMx_SMCR
register). In this mode, the TIM_CEN, TIM_DIR (in the TIMx_CR1 register), and TIM_UG bits (in the TIMx_EGR
register) are actual control bits and can be changed only by software, except for TIM_UG, which remains
cleared automatically. As soon as the TIM_CEN bit is written to 1, the prescaler is clocked by the internal
clock CK_INT.
Figure 9-12 shows the behavior of the control circuit and the up-counter in normal mode, without prescaling.
Figure 9-12. Control Circuit in Normal Mode, Internal Clock Divided by 1
9.3.3.2
External Clock Source Mode 1
This mode is selected when TIM_SMS = 111 in the TIMx_SMCR register. The counter can count at each rising or
falling edge on a selected input. Figure 9-13 shows the registers and signals used in the example that follows.
9-9
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 9-13. TI2 External Clock Connection Example
For example, to configure the up-counter to count in response to a rising edge on the TI2 input, use the
following procedure:
Configure channel 2 to detect rising edges on the TI2 input: Write TIM_CC2S = 01 in the TIMx_CCMR1
register.
.
Configure the input filter duration: Write the TIM_IC2F bits in the TIMx_CCMR1 register (if no filter is
needed, keep TIM_IC2F = 0000).
.
Note: The capture prescaler is not used for triggering, so it does not need to be configured.
Select rising edge polarity: Write TIM_CC2P = 0 in the TIMx_CCER register.
.
Configure the timer in external clock mode 1: Write TIM_SMS = 111 in the TIMx_SMCR register.
.
Select TI2 as the input source: Write TIM_TS = 110 in the TIMx_SMCR register.
.
.
Enable the counter: Write TIM_CEN = 1 in the TIMx_CR1 register.
When a rising edge occurs on TI2, the counter counts once and the INT_TIMTIF flag is set. The delay between
the rising edge on TI2 and the actual clock of the counter is due to the resynchronization circuit on the TI2
input. The relationship between rising edges on TI2 and the resulting counter clocks is shown in Figure 9-14.
Figure 9-14. Control Circuit in External Clock Mode 1
9.3.3.3
External Clock Source Mode 2
This mode is selected by writing TIM_ECE = 1 in the TIMx_SMCR register. The counter can count at each rising
or falling edge on the external trigger input ETR.
The TIM_EXTRIGSEL bits in the TIMx_OR register select a clock signal that drives ETR, as shown in Table 9-2.
9-10
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 9-2. TIM_EXTRIGSEL Clock Signal Selection
TIM_EXTRIGSEL bits Clock Signal Selection
00
PCLK (peripheral clock). When running from the 24 MHz crystal oscillator, the PCLK
frequency is 12 MHz. When the 12 MHz RC oscillator is in use, the frequency is 6 MHz.
01
10
11
Calibrated 1 kHz internal RC oscillator
Optional 32.786 kHz clock
TIMxCLK pin. If the TIM_CLKMSKEN bit in the TIMx_OR register is set, this signal is
AND’ed with the TIMxMSK pin providing a gated clock input.
Figure 9-15 gives an overview of the external trigger input block.
Figure 9-15. External Trigger Input Block
For example, to configure the up-counter to count each 2 rising edges on ETR, use the following procedure:
As no filter is needed in this example, write TIM_ETF = 0000 in the TIMx_SMCR register.
.
Set the prescaler: Write TIM_ETPS = 01 in the TIMx_SMCR register.
.
Select rising edge detection on ETR: WriteTIM_ETP = 0 in the TIMx_SMCR register.
.
Enable external clock mode 2: Write TIM_ECE = 1 in the TIMx_SMCR register.
.
Enable the counter: Write TIM_CEN = 1 in the TIMx_CR1 register.
.
The counter counts once each 2 ETR rising edges. The delay between the rising edge on ETR and the actual
clock of the counter is due to the resynchronization circuit on the ETRP signal.
Figure 9-16 illustrates counting every 2 rising edges of ETR using external clock mode 2.
9-11
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 9-16. Control Circuit in External Clock Mode 2
9.3.4 Capture/Compare Channels
Each capture/compare channel is built around a capture/compare register including a shadow register, an
input stage for capture with digital filter, multiplexing and prescaler, and an output stage with comparator
and output control.
Figure 9-17 gives an overview of the input stage of one capture/compare channel. The input stage samples the
corresponding TIy input to generate a filtered signal (TIyF). Then an edge detector with polarity selection
generates a signal (TIyFPy) which can be used either as trigger input by the slave mode controller or as the
capture command. It is prescaled before the capture register (ICyPS).
Figure 9-17. Capture/Compare Channel (Example: Channel 1 Input Stage)
The output stage generates an intermediate reference signal, OCyREF, which is only used internally. OCyREF is
always active high, but it may be inverted to create the output signal, OCy, that controls a GPIO output.
Figure 9-18 shows the basic elements of a capture/compare channel.
9-12
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 9-18. Capture/Compare Channel 1 Main Circuit
Figure 9-19 show details of the output stage of a capture/compare channel.
Figure 9-19. Output Stage of Capture/Compare Channel (Channel 1)
The capture/compare block is made of a buffer register and a shadow register. Writes and reads always access
the buffer register.
In capture mode, captures are first written to the shadow register, then copied into the buffer register.
In compare mode, the content of the buffer register is copied into the shadow register which is compared to
the counter.
9.3.5 Input Capture Mode
In input capture mode, a capture/compare register (TIMx_CCRy) latches the value of the counter after a
transition is detected by the corresponding ICy signal. When a capture occurs, the corresponding INT_TIMCCyIF
flag in the INT_TIMxFLAG register is set, and an interrupt request is sent if enabled.
9-13
120-035X-000 Rev. 1.2
Final
EM351 / EM357
If a capture occurs when the INT_TIMCCyIF flag is already high, then the missed capture flag INT_TIMMISSCCyIF
in the INT_TIMxMISS register is set. INT_TIMCCyIF can be cleared by software writing a 1 to its bit or reading
the captured data stored in the TIMx_CCRy register. To clear the INT_TIMMISSCCyIF bit, write a 1 to it.
The following example shows how to capture the counter value in the TIMx_CCR1 when the TI1 input rises.
Select the active input: TIMx_CCR1 must be linked to the TI1 input, so write the TIM_CC1S bits to 01 in the
TIMx_CCMR1 register. As soon as TIM_CC1S becomes different from 00, the channel is configured in input
and the TIMx_CCR1 register becomes read-only.
.
Program the required input filter duration with respect to the signal connected to the timer, when the
.
input is one of the TIy (ICyF bits in the TIMx_CCMR1 register). Consider a situation in which, when toggling,
the input signal is unstable during at most 5 internal clock cycles. The filter duration must be longer than
these 5 clock cycles. The transition on TI1 can be validated when 8 consecutive samples with the new level
have been detected (sampled at PCLK frequency). To do this, write the TIM_IC1F bits to 0011 in the
TIMx_CCMR1 register.
Select the edge of the active transition on the TI1 channel: Write the TIM_CC1P bit to 0 in the TIMx_CCER
register (rising edge in this case).
.
Program the input prescaler: In this example, the capture is to be performed at each valid transition, so
the prescaler is disabled (write the TIM_IC1PSC bits to 00 in the TIMx_CCMR1 register).
.
Enable capture from the counter into the capture register: Set the TIM_CC1E bit in the TIMx_CCER
register.
.
If needed, enable the related interrupt request by setting the INT_TIMCC1IF bit in the INT_TIMxCFG
register.
.
When an input capture occurs:
.
The TIMx_CCR1 register gets the value of the counter on the active transition.
•
INT_TIMCC1IF flag is set (capture/compare interrupt flag). The missed capture/compare flag
INT_TIMMISSCC1IF in INT_TIMxMISS is also set if another capture occurs before the INT_TIMCC1IF
•
flag is cleared.
An interrupt may be generated if enabled by the INT_TIMCC1IF bit.
•
To detect missed captures reliably, read captured data in TIMxCCRy before checking the missed
capture/compare flag. This sequence avoids missing a capture that could happen after reading the flag and
before reading the data.
Note: Software can generate IC interrupt requests by setting the corresponding TIM_CCyG bit in the TIMx_EGR
register.
9.3.6 PWM Input Mode
This mode is a particular case of input capture mode. The procedure is the same except:
Two ICy signals are mapped on the same TIy input.
.
These two ICy signals are active on edges with opposite polarity.
.
One of the two TIyFP signals is selected as trigger input and the slave mode controller is configured in
reset mode.
.
For example, to measure the period in the TIMx_CCR1 register and the duty cycle in the TIMx_CCR2 register of
the PWM applied on TI1, use the following procedure depending on CK_INT frequency and prescaler value:
Select the active input for TIMx_CCR1: write the TIM_CC1S bits to 01 in the TIMx_CCMR1 register (TI1
selected).
.
Select the active polarity for TI1FP1, used both for capture in the TIMx_CCR1 and counter clear, by writing
the TIM_CC1P bit to 0 (active on rising edge).
.
9-14
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Select the active input for TIMx_CCR2by writing the TIM_CC2S bits to 10 in the TIMx_CCMR1 register (TI1
selected).
.
.
.
.
.
Select the active polarity for TI1FP2 (used for capture in the TIMx_CCR2) by writing the TIM_CC2P bit to 1
(active on falling edge).
Select the valid trigger input by writing the TIM_TS bits to 101 in the TIMx_SMCR register (TI1FP1
selected).
Configure the slave mode controller in reset mode by writing the TIM_SMS bits to 100 in the TIMx_SMCR
register.
Enable the captures by writing the TIM_CC1E and TIM_CC2E bits to 1 in the TIMx_CCER register.
Figure 9-20 illustrates this example.
Figure 9-20. PWM Input Mode Timing
9.3.7 Forced Output Mode
In output mode (CCyS bits = 00 in the TIMx_CCMR1 register), software can force each output compare signal
(OCyREF and then OCy) to an active or inactive level independently of any comparison between the output
compare register and the counter.
To force an output compare signal (OCyREF/OCy) to its active level, write 101 in the TIM_OCyM bits in the
corresponding TIMx_CCMR1 register. OCyREF is forced high (OCyREF is always active high) and OCy gets the
opposite value to the TIM_CCyP polarity bit. For example, TIM_CCyP = 0 defines OCy as active high, so when
OCyREF is active, OCy is also set to a high level.
The OCyREF signal can be forced low by writing the TIM_OCyM bits to 100 in the TIMx_CCMR1 register.
The comparison between the TIMx_CCRy shadow register and the counter is still performed and allows the
INT_TIMxCCRyIF flag to be set. Interrupt requests can be sent accordingly. This is described in the output
compare mode section.
9.3.8 Output Compare Mode
This mode is used to control an output waveform or to indicate when a period of time has elapsed.
When a match is found between the capture/compare register and the counter, the output compare function:
Assigns the corresponding output pin to a programmable value defined by the output compare mode (the
.
TIM_OCyM bits in the TIMx_CCMR1 register) and the output polarity (the TIM_CCyP bit in the TIMx_CCER
register). The output can be frozen (TIM_OCyM = 000), be set active (TIM_OCyM = 001), be set inactive
(TIM_OCyM = 010), or can toggle (TIM_OCyM = 011) on the match.
9-15
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Sets a flag in the interrupt flag register (the INT_TIMCCyIF bit in the INT_TIMxFLAG register).
.
.
Generates an interrupt if the corresponding interrupt mask is set (the TIM_CCyIF bit in the INT_TIMxCFG
register).
The TIMx_CCRy registers can be programmed with or without buffer registers using the TIM_OCyBE bit in the
TIMx_CCMR1 register.
In output compare mode, the UEV has no effect on OCyREF or the OCy output. The timing resolution is one
count of the counter. Output compare mode can also be used to output a single pulse (in one pulse mode).
Procedure:
1. Select the counter clock (internal, external, and prescaler).
2. Write the desired data in the TIMx_ARR and TIMx_CCRy registers.
3. Set the INT_TIMCCyIF bit in INT_TIMxCFG if an interrupt request is to be generated.
4. Select the output mode. For example, you must write TIM_OCyM = 011, TIM_OCyBE = 0, TIM_CCyP = 0 and
TIM_CCyE = 1 to toggle the OCy output pin when TIMx_CNT matches TIMx_CCRy, TIMx_CCRy buffer is not
used, OCy is enabled and active high.
5. Enable the counter: Set the TIM_CEN bit in the TIMx_CR1 register.
To control the output waveform, software can update the TIMx_CCRy register at any time, provided that the
buffer register is not enabled (TIM_OCyBE = 0). Otherwise TIMx_CCRy shadow register is updated only at the
next UEV. An example is given in Figure 9-21.
Figure 9-21. Output Compare Mode, Toggle on OC1
9.3.9 PWM Mode
Pulse width modulation mode allows you to generate a signal with a frequency determined by the value of the
TIMx_ARR register, and a duty cycle determined by the value of the TIMx_CCRy register.
PWM mode can be selected independently on each channel (one PWM per OCy output) by writing 110 (PWM
mode 1) or 111 (PWM mode 2) in the TIM_OCyM bits in the TIMx_CCMR1 register. The corresponding buffer
register must be enabled by setting the TIM_OCyBE bit in the TIMx_CCMR1 register. Finally, in up-counting or
center-aligned mode the auto-reload buffer register must be enabled by setting the TIM_ARBE bit in the
TIMx_CR1 register.
9-16
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Because the buffer registers are only transferred to the shadow registers when a UEV occurs, before starting
the counter initialize all the registers by setting the TIM_UG bit in the TIMx_EGR register.
OCy polarity is software programmable using the TIM_CCyP bit in the TIMx_CCER register. It can be
programmed as active high or active low. OCy output is enabled by the TIM_CCyE bit in the TIMx_CCER
register. Refer to the TIMx_CCER register description in the Registers section for more details.
In PWM mode (1 or 2), TIMx_CNT and TIMx_CCRy are always compared to determine whether
TIMx_CCRy ≤ TIMx_CNT or TIMx_CNT ≤ TIMx_CCRy, depending on the direction of the counter. The OCyREF
signal is asserted only:
When the result of the comparison changes, or
.
.
When the output compare mode (TIM_OCyM bits in the TIMx_CCMR1 register) switches from the “frozen”
configuration (no comparison, TIM_OCyM = 000) to one of the PWM modes (TIM_OCyM = 110 or 111).
This allows software to force a PWM output to a particular state while the timer is running.
The timer is able to generate PWM in edge-aligned mode or center-aligned mode depending on the TIM_CMS
bits in the TIMx_CR1 register.
9.3.9.1
PWM Edge-Aligned Mode: Up-Counting Configuration
Up-counting is active when the TIM_DIR bit in the TIMx_CR1 register is low. Refer to the section Up-Counting
Mode.
The following example uses PWM mode 1. The reference PWM signal OCyREF is high as long as
TIMx_CNT < TIMx_CCRy, otherwise it becomes low. If the compare value in TIMx_CCRy is greater than the
auto-reload value in TIMx_ARR, then OCyREF is held at 1. If the compare value is 0, then OCyREF is held at 0.
Figure 9-22 shows some edge-aligned PWM waveforms in an example, where TIMx_ARR = 8.
Figure 9-22. Edge-Aligned PWM Waveforms (ARR = 8)
9.3.9.2
PWM Edge-Aligned Mode: Down-Counting Configuration
Down-counting is active when the TIM_DIR bit in the TIMx_CR1 register is high. Refer to the Down-Counting
Mode section for more information.
In PWM mode 1, the reference signal OCyREF is low as long as TIMx_CNT > TIMx_CCRy, otherwise it becomes
high. If the compare value in TIMx_CCRy is greater than the auto-reload value in TIMx_ARR, then OCyREF is
held at 1. Zero-percent PWM is not possible in this mode.
9-17
120-035X-000 Rev. 1.2
Final
EM351 / EM357
9.3.9.3
PWM Center-Aligned Mode
Center-aligned mode is active except when the TIM_CMS bits in the TIMx_CR1 register are 00 (all
configurations where TIM_CMS is non-zero have the same effect on the OCyREF/OCy signals). The compare flag
is set when the counter counts up, when it counts down, or when it counts up and down, depending on the
TIM_CMS bits configuration. The direction bit (TIM_DIR) in the TIMx_CR1 register is updated by hardware and
must not be changed by software. Refer to the Center-Aligned Mode (Up/Down Counting) section for more
information.
Figure 9-23 shows some center-aligned PWM waveforms in an example where:
TIMx_ARR = 8
.
PWM mode is the PWM mode 1
.
The output compare flag is set when the counter counts down corresponding to the center-aligned mode 1
selected for TIM_CMS = 01 in the TIMx_CR1 register
.
Figure 9-23. Center-Aligned PWM Waveforms (ARR = 8)
9-18
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Hints on using center-aligned mode:
When starting in center-aligned mode, the current up-down configuration is used. This means that the
counter counts up or down depending on the value written in the TIM_DIR bit in the TIMx_CR1 register. The
TIM_DIR and TIM_CMS bits must not be changed at the same time by the software.
.
.
Writing to the counter while running in center-aligned mode is not recommended as it can lead to
unexpected results. In particular:
The direction is not updated when the value written to the counter that is greater than the auto-
reload value (TIMx_CNT > TIMx_ARR). For example, if the counter was counting up, it continues to
count up.
•
The direction is updated when 0 or the TIMx_ARR value is written to the counter, but no UEV is
generated.
•
The safest way to use center-aligned mode is to generate an update by software (setting the TIM_UG bit in
the TIMx_EGR register) just before starting the counter, and not to write the counter while it is running.
.
9.3.10 One-Pulse Mode
One-pulse mode (OPM) is a special case of the previous modes. It allows the counter to be started in response
to a stimulus and to generate a pulse with a programmable length after a programmable delay.
Starting the counter can be controlled through the slave mode controller. Generating the waveform can be
done in output compare mode or PWM mode. Select OPM by setting the TIM_OPM bit in the TIMx_CR1 register.
This makes the counter stop automatically at the next UEV.
A pulse can be correctly generated only if the compare value is different from the counter initial value.
Before starting (when the timer is waiting for the trigger), the configuration must be:
In up-counting: TIMx_CNT < TIMx_CCRy ≤ TIMx_ARR (in particular, 0 < TIMx_CCRy),
In down-counting: TIMx_CNT > TIMx_CCRy.
For example, to generate a positive pulse on OC1 with a length of tPULSE and after a delay of tDELAY as soon as a
rising edge is detected on the TI2 input pin, using TI2FP2 as trigger 1:
Map TI2FP2 on TI2: Write TIM_IC2S = 01 in the TIMx_CCMR1 register.
.
TI2FP2 must detect a rising edge. Write TIM_CC2P = 0 in the TIMx_CCER register.
.
Configure TI2FP2 as trigger for the slave mode controller (TRGI): Write TIM_TS = 110 in the TIMx_SMCR
register.
.
Use TI2FP2 to start the counter: Write TIM_SMS to 110 in the TIMx_SMCR register (trigger mode).
.
.
The OPM waveform is defined: Write the compare registers, taking into account the clock frequency and
the counter prescaler.
The tDELAY is defined by the value written in the TIMx_CCR1 register.
•
The tPULSE is defined by the difference between the auto-reload value and the compare value
(TIMx_ARR - TIMx_CCR1).
•
To build a waveform with a transition from 0 to 1 when a compare match occurs and a transition from 1 to
0 when the counter reaches the auto-reload value:
.
Enable PWM mode 2: Write TIM_OC1M = 111 in the TIMx_CCMR1 register.
•
Optionally, enable the buffer registers: Write TIM_OC1BE = 1 in the TIMx_CCMR1 register and
•
TIM_ARBE in the TIMx_CR1 register. In this case, also write the compare value in the TIMx_CCR1
register, the auto-reload value in the TIMx_ARR register, generate an update by setting the TIM_UG
bit, and wait for external trigger event on TI2. TIM_CC1P is written to 0 in this example.
In the example, the TIM_DIR and TIM_CMS bits in the TIMx_CR1 register should be low.
Since only one pulse is desired, software should set the TIM_OPM bit in the TIMx_CR1 register to stop the
counter at the next UEV (when the counter rolls over from the auto-reload value back to 0).
9-19
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 9-24 illustrates this example.
Figure 9-24. Example of One Pulse Mode
9.3.10.1 A Special Case: OCy Fast Enable
In one-pulse mode, the edge detection on the TIy input sets the TIM_CEN bit, which enables the counter. Then
the comparison between the counter and the compare value toggles the output. However, several clock cycles
are needed for this operation, and it limits the minimum delay (tDELAY min) achievable.
To output a waveform with the minimum delay, set the TIM_OCyFE bit in the TIMx_CCMR1 register. Then
OCyREF and OCy are forced in response to the stimulus, without taking the comparison into account. Its new
level is the same as if a compare match had occurred. TIM_OCyFE acts only if the channel is configured in
PWM mode 1 or 2.
9.3.11 Encoder Interface Mode
To select encoder interface mode, write TIM_SMS = 001 in the TIMx_SMCR register to count only TI2 edges,
TIM_SMS = 010 to count only TI1 edges, and TIM_SMS = 011 to count both TI1 and TI2 edges.
Select the TI1 and TI2 polarity by programming the TIM_CC1P and TIM_CC2P bits in the TIMx_CCER register. If
needed, program the input filter as well.
The two inputs TI1 and TI2 are used to interface to an incremental encoder (see Table 9-3). Assuming that it is
enabled (the TIM_CEN bit in the TIMx_CR1 register = 1), the counter is clocked by each valid transition on
TI1FP1 or TI2FP2 (TI1 and TI2 after input filter and polarity selection, TI1FP1 = TI1 if not filtered and not
inverted, TI2FP2 = TI2 if not filtered and not inverted.) The timer input logic evaluates the sequence of the
two inputs’ values, and from this generates both count pulses and the direction signal. Depending on the
sequence, the counter counts up or down, and hardware modifies the TIM_DIR bit in the TIMx_CR1 register
accordingly. The TIM_DIR bit is calculated at each transition on any input (TI1 or TI2), whether the counter is
counting on TI1 only, TI2 only, or both TI1 and TI2.
Encoder interface mode acts simply as an external clock with direction selection. This means that the counter
counts continuously between 0 and the auto-reload value in the TIMx_ARR register (0 to TIMx_ARR or
TIMx_ARR down to 0 depending on the direction), so TIMx_ARR must be configured before starting. In the
same way, the capture, compare, prescaler, and trigger output features continue to work as normal.
In this mode the counter is modified automatically following the speed and the direction of the incremental
encoder, and therefore its contents always represent the encoder’s position. The count direction corresponds
to the rotation direction of the connected sensor. Table 9-3 summarizes the possible combinations, assuming
TI1 and TI2 do not switch at the same time.
9-20
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 9-3. Counting Direction versus Encoder Signals
Active Edges
Level on
TI1FP1 Signal
TI2FP2 Signal
Opposite Signal
(TI1FP1 for TI2,
TI2FP2 for TI1)
Rising
Falling
Rising
Falling
Counting on TI1
only
High
Low
High
Low
High
Low
Down
Up
Up
Down
No Count
No Count
Up
No Count
No Count
Down
Up
Counting on TI2
only
No Count
No Count
Down
No Count
No Count
Up
Down
Up
Counting on TI1
and TI2
Down
Up
Up
Down
Down
An external incremental encoder can be connected directly to the MCU without external interface logic.
However, comparators are normally used to convert an encoder’s differential outputs to digital signals, and
this greatly increases noise immunity. If a third encoder output indicates the mechanical zero (or index)
position, it may be connected to an external interrupt input and can trigger a counter reset.
Figure 9-25 gives an example of counter operation, showing count signal generation and direction control. It
also shows how input jitter is compensated for when both inputs are used for counting. This might occur if the
sensor is positioned near one of the switching points. This example assumes the following configuration:
TIM_CC1S = 01 (TIMx_CCMR1 register, IC1FP1 mapped on TI1).
.
TIM_CC2S = 01 (TIMx_CCMR2 register, IC2FP2 mapped on TI2).
.
TIM_CC1P = 0 (TIMx_CCER register, IC1FP1 non-inverted, IC1FP1 = TI1).
.
TIM_CC2P = 0 (TIMx_CCER register, IC2FP2 non-inverted, IC2FP2 = TI2).
.
TIM_SMS = 011 (TIMx_SMCR register, both inputs are active on both rising and falling edges).
.
TIM_CEN = 1 (TIMx_CR1 register, counter is enabled).
.
Figure 9-25. Example of Counter Operation in Encoder Interface Mode
9-21
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 9-26 gives an example of counter behavior when IC1FP1 polarity is inverted (same configuration as
above except TIM_CC1P = 1).
Figure 9-26. Example of Encoder Interface Mode with IC1FP1 Polarity Inverted
The timer configured in encoder interface mode provides information on a sensor’s current position. To obtain
dynamic information (speed, acceleration/deceleration), measure the period between two encoder events
using a second timer configured in capture mode. The output of the encoder that indicates the mechanical
zero can be used for this purpose. Depending on the time between two events, the counter can also be read
at regular times. Do this by latching the counter value into a third input capture register. (In this case the
capture signal must be periodic and can be generated by another timer).
9.3.12 Timer Input XOR Function
The TIM_TI1S bit in the TIM1_CR2 register allows the input filter of channel 1 to be connected to the output of
a XOR gate that combines the three input pins TIMxC2 to TIMxC4.
The XOR output can be used with all the timer input functions such as trigger or input capture. It is especially
useful to interface to Hall effect sensors.
9.3.13 Timers and External Trigger Synchronization
The timers can be synchronized with an external trigger in several modes: reset mode, gated mode, and
trigger mode.
9.3.13.1 Slave Mode: Reset Mode
Reset mode reinitializes the counter and its prescaler in response to an event on a trigger input. Moreover, if
the TIM_URS bit in the TIMx_CR1 register is low, a UEV is generated. Then all the buffered registers
(TIMx_ARR, TIMx_CCRy) are updated.
In the following example, the up-counter is cleared in response to a rising edge on the TI1 input:
Configure the channel 1 to detect rising edges on TI1:
.
Configure the input filter duration. In this example, no filter is required so TIM_IC1F = 0000.
•
The capture prescaler is not used for triggering, so it is not configured.
•
The TIM_CC1S bits select the input capture source only, TIM_CC1S = 01 in the TIMx_CCMR1 register.
•
•
Write TIM_CC1P = 0 in the TIMx_CCER register to validate the polarity, and detect rising edges
only.
Configure the timer in reset mode: Write TIM_SMS = 100 in the TIMx_SMCR register.
.
Select TI1 as the input source by writing TIM_TS = 101 in the TIMx_SMCR register.
.
Start the counter: Write TIM_CEN = 1 in the TIMx_CR1 register.
.
9-22
120-035X-000 Rev. 1.2
Final
EM351 / EM357
The counter starts counting on the internal clock, then behaves normally until the TI1 rising edge. When TI1
rises, the counter is cleared and restarts from 0. In the meantime, the trigger flag is set (the INT_TIMTIF bit in
the INT_TIMxFLAG register) and an interrupt request can be sent if enabled (depending on the INT_TIMTIF bit
in the INT_TIMxCFG register).
Figure 9-27 shows this behavior when the auto-reload register TIMx_ARR = 0x36. The delay between the rising
edge on TI1 and the actual reset of the counter is due to the resynchronization circuit on the TI1 input.
Figure 9-27. Control Circuit in Reset Mode
9.3.13.2 Slave Mode: Gated Mode
In gated mode the counter is enabled depending on the level of a selected input.
In the following example, the up-counter counts only when the TI1 input is low:
Configure channel 1 to detect low levels on TI1:
.
Configure the input filter duration. In this example, no filter is required, so TIM_IC1F = 0000.
•
The capture prescaler is not used for triggering, so it is not configured.
•
The TIM_CC1S bits select the input capture source only, TIM_CC1S = 01 in the TIMx_CCMR1 register.
•
Write TIM_CC1P = 1 in the TIMx_CCER register to validate the polarity (and detect low level only).
•
Configure the timer in gated mode: Write TIM_SMS = 101 in the TIMx_SMCR register.
.
Select TI1 as the input source by writing TIM_TS = 101 in the TIMx_SMCR register.
.
Enable the counter: Write TIM_CEN = 1 in the TIMx_CR1 register. In gated mode, the counter does not start
if TIM_CEN = 0, regardless of the trigger input level.
.
The counter starts counting on the internal clock as long as TI1 is low and stops as soon as TI1 becomes high.
The INT_TIMTIF flag in the INT_TIMxFLAG register is set when the counter starts and when it stops. The delay
between the rising edge on TI1 and the actual stop of the counter is due to the resynchronization circuit on
the TI1 input.
Figure 9-28 shows the counter in gated mode with counting enabled when TI1 is low.
Figure 9-28. Control Circuit in Gated Mode
9-23
120-035X-000 Rev. 1.2
Final
EM351 / EM357
9.3.13.3 Slave Mode: Trigger Mode
In trigger mode the counter starts in response to an event on a selected input.
In the following example, the up-counter starts in response to a rising edge on the TI2 input:
Configure channel 2 to detect rising edges on TI2:
.
Configure the input filter duration. In this example, no filter is required so TIM_IC2F = 0000.
•
The capture prescaler is not used for triggering, so it is not configured.
•
The TIM_CC2S bits select the input capture source only, TIM_CC2S = 01 in the TIMx_CCMR1 register.
•
•
Write TIM_CC2P = 0 in the TIMx_CCER register to validate the polarity and detect high level only.
Configure the timer in trigger mode: Write TIM_SMS = 110 in the TIMx_SMCR register.
.
.
Select TI2 as the input source by writing TIM_TS = 110 in the TIMx_SMCR register.
When a rising edge occurs on TI2, the counter starts counting on the internal clock and the INT_TIMTIF flag is
set. The delay between the rising edge on TI2 and the actual start of the counter is due to the
resynchronization circuit on the TI2 input.
Figure 9-29 illustrates the example in which the counter is started by a rising edge on TI2.
Figure 9-29. Control Circuit in Trigger Mode
9.3.13.4 Slave Mode: External Clock Mode 2 +Trigger Mode
External clock mode 2 can be used in combination with another slave mode (except external clock mode 1 and
encoder mode). In this case, the ETR signal is used as external clock input, and another input can be selected
as trigger input when operating in reset mode, gated mode or trigger mode. It is not recommended to select
ETR as TRGI through the TIM_TS bits of TIMx_SMCR register.
In the following example, shown in Figure 9-30, the up-counter is incremented at each rising edge of the ETR
signal as soon as a rising edge of TI1 occurs:
Configure the external trigger input circuit: Program the TIMx_SMCR register as follows:
.
TIM_ETF = 0000: no filter.
•
TIM_ETPS = 00: prescaler disabled.
•
TIM_ETP = 0: detection of rising edges on ETR and TIM_ECE = 1 to enable the external clock
mode 2.
•
Configure the channel 1 to detect rising edges on TI, as follows:
.
TIM_IC1F = 0000: no filter.
•
The capture prescaler is not used for triggering and does not need to be configured.
•
•
TIM_CC1S = 01 in the TIMx_CCMR1 register to select only the input capture source.
TIM_CC1P = 0 in the TIMx_CCER register to validate the polarity (and detect rising edge only).
•
Configure the timer in trigger mode: WriteTIM_SMS = 110 in the TIMx_SMCR register.
.
.
Select TI1 as the input source by writing TIM_TS = 101 in the TIMx_SMCR register.
9-24
120-035X-000 Rev. 1.2
Final
EM351 / EM357
A rising edge on TI1 enables the counter and sets the INT_TIMTIF flag. The counter then counts on ETR rising
edges. The delay between the rising edge of the ETR signal and the actual reset of the counter is due to the
resynchronization circuit on ETRP input.
Figure 9-30. Control circuit in External Clock Mode 2 + Trigger Mode
9.3.14 Timer Synchronization
The two timers can be linked together internally for timer synchronization or chaining. A timer configured in
master mode can reset, start, stop or clock the counter of the other timer configured in slave mode.
Figure 9-31 presents an overview of the trigger selection and the master mode selection blocks.
9.3.14.1 Using One Timer as Prescaler for the Other Timer
For example, to configure Timer 1 to act as a prescaler for Timer 2:
Configure Timer 1 in master mode so that it outputs a periodic trigger signal on each UEV. Writing
TIM_MMS = 010 in the TIM1_CR2 register causes a rising edge to be output on TRGO each time a UEV is
generated.
.
To connect the TRGO output of Timer 1 to Timer 2, configure Timer 2 in slave mode using ITR0 as an
internal trigger. Write TIM_TS = 100 in the TIM2_SMCR register.
.
Put the slave mode controller in external clock mode 1: Write TIM_SMS = 111 in the TIM2_SMCR register.
This causes Timer 2 to be clocked by the rising edge of the periodic Timer 1 trigger signal, which
.
corresponds to the Timer 1 counter overflow.
Finally, enable both timers: Set their respective TIM_CEN bits in the TIMx_CR1 register.
.
Note: If OCy is selected on Timer 1 as trigger output (TIM_MMS = 1xx), its rising edge is used to clock the
counter of Timer 2.
Figure 9-31. Master/Slave Timer Example
9-25
120-035X-000 Rev. 1.2
Final
EM351 / EM357
9.3.14.2 Using One Timer to Enable the Other Timer
In this example, shown in Figure 9-32, the enable of Timer 2 is controlled with the output compare 1 of
Timer 1. Timer 2 counts on the divided internal clock only when OC1REF of Timer 1 is high. Both counter clock
frequencies are divided by 3 by the prescaler compared to CK_INT (fCK_CNT = fCK_INT /3).
Configure Timer 1 in master mode to send its Output Compare Reference (OC1REF) signal as trigger
output: Write TIM_MMS = 100 in the TIM1_CR2 register.
.
Configure the Timer 1 OC1REF waveform (TIM1_CCMR1 register).
.
Configure Timer 2 to get the input trigger from Timer 1: Write TIM_TS = 000 in the TIM2_SMCR register.
.
Configure Timer 2 in gated mode: Write TIM_SMS = 101 in the TIM2_SMCR register.
.
Enable Timer 2: Write 1 in the TIM_CEN bit in the TIM2_CR1 register.
.
Start Timer 1: Write 1 in the TIM_CEN bit in the TIM1_CR1 register.
.
Note: The counter 2 clock is not synchronized with counter 1, this mode only affects the Timer 2 counter
enable signal.
Figure 9-32. Gating Timer 2 with OC1REF of Timer 1
In the example in Figure 9-32, the Timer 2 counter and prescaler are not initialized before being started. So
they start counting from their current value. It is possible to start from a given value by resetting both timers
before starting Timer 1, then writing the desired value in the timer counters. The timers can easily be reset
by software using the TIM_UG bit in the TIMx_EGR registers.
The next example, illustrated in Figure 9-33, synchronizes Timer 1 and Timer 2. Timer 1 is the master and
starts from 0. Timer 2 is the slave and starts from 0xE7. The prescaler ratio is the same for both timers. Timer
2 stops when Timer 1 is disabled by writing 0 to the TIM_CEN bit in the TIM1_CR1 register:
Configure Timer 1 in master mode to send its Output Compare Reference (OC1REF) signal as trigger
output: Write TIM_MMS = 100 in the TIM1_CR2 register)
.
Configure the Timer 1 OC1REF waveform (TIM1_CCMR1 register).
.
Configure Timer 2 to get the input trigger from Timer 1: Write TIM_TS = 000 in the TIM2_SMCR register.
.
Configure Timer 2 in gated mode: Write TIM_SMS = 101 in the TIM2_SMCR register.
.
Reset Timer 1: Write 1 in the TIM_UG bit (TIM1_EGR register.
.
Reset Timer 2 by writing 1 in the TIM_UG bit (TIM2_EGR register).
.
Initialize Timer 2 to 0xE7: Write 0xE7 in the Timer 2 counter (TIM2_CNTL).
.
Enable Timer 2: Write 1 in the TIM_CEN bit in the TIM2_CR1 register.
.
Start Timer 1: Write 1 in the TIM_CEN bit in the TIM1_CR1 register.
.
Stop Timer 1: Write 0 in the TIM_CEN bit in the TIM1_CR1 register.
.
9-26
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 9-33. Gating Timer 2 with Enable of Timer 1
9.3.14.3 Using One Timer to Start the Other Timer
In this example (see Figure 9-34), the enable of Timer 2 is set with the UEV of Timer 1. Timer 2 starts counting
from its current value (which can be non-zero) on the divided internal clock as soon as Timer 1 generates the
UEV.
When Timer 2 receives the trigger signal its TIM_CEN bit is automatically set and the counter counts until 0 is
written to the TIM_CEN bit in the TIM2_CR1 register. Both counter clock frequencies are divided by 3 by the
prescaler compared to CK_INT (fCK_CNT = fCK_INT/3).
Configure Timer 1 in master mode to send its UEV as trigger output: WriteTIM_MMS = 010 in the TIM1_CR2
register.
.
Configure the Timer 1 period (TIM1_ARR register).
.
Configure Timer 2 to get the input trigger from Timer 1: Write TIM_TS = 000 in the TIM2_SMCR register.
.
Configure Timer 2 in trigger mode. Write TIM_SMS = 110 in the TIM2_SMCR register.
.
.
Start Timer 1: Write 1 in the TIM_CEN bit in theTIM1_CR1 register.
Figure 9-34. Triggering Timer 2 with Update of Timer 1
As in the previous example, both counters can be initialized before starting counting. Figure 9-35 shows the
behavior with the same configuration shown in Figure 9-34, but in trigger mode instead of gated mode
(TIM_SMS = 110 in the TIM2_SMCR register).
9-27
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 9-35. Triggering Timer 2 with Enable of Timer 1
9.3.14.4 Starting both Timers Synchronously in Response to an External Trigger
This example sets the enable of Timer 1 when its TI1 input rises, and the enable of Timer 2 with the enable of
Timer 1. To ensure the counters are aligned, Timer 1 must be configured in master/slave mode (slave with
respect to TI1, master with respect to Timer 2):
Configure Timer 1 in master mode to send its Enable as trigger output: Write TIM_MMS = 001 in the
TIM1_CR2 register.
.
Configure Timer 1 slave mode to get the input trigger from TI1: Write TIM_TS = 100 in the TIM1_SMCR
register.
.
Configure Timer 1 in trigger mode: Write TIM_SMS = 110 in the TIM1_SMCR register.
.
Configure the Timer 1 in master/slave mode: Write TIM_MSM = 1 in the TIM1_SMCR register.
.
Configure Timer 2 to get the input trigger from Timer 1: Write TIM_TS = 000 in the TIM2_SMCR register.
.
.
Configure Timer 2 in trigger mode: Write TIM_SMS = 110 in the TIM2_SMCR register.
When a rising edge occurs on TI1 (Timer 1), both counters start counting synchronously on the internal clock
and both timers’ INT_TIMTIF flags are set. Figure 9-36 shows this in operation.
Note: In this example both timers are initialized before starting by setting their respective TIM_UG bits. Both
counters starts from 0, but an offset can be inserted between them by writing any of the counter registers
(TIMx_CNT). The master/slave mode inserts a delay between CNT_EN and CK_PSC on Timer 1.
9-28
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 9-36. Triggering Timer 1 and 2 with Timer 1 TI1 Input
9.3.15 Timer Signal Descriptions
Table 9-4. Timer Signal Descriptions
Signal
Internal/ Description
External
CK_INT
Internal Internal clock source: connects to EM35x peripheral clock (PCLK) in internal clock mode.
CK_PSC Internal Input to the clock prescaler.
ETR
Internal External trigger input (used in external timer mode 2): a clock selected by
TIM_EXTRIGSEL in TIMx_OR.
ETRF
ETRP
ICy
Internal External trigger: ETRP after filtering.
Internal External trigger: ETR after polarity selection, edge detection and prescaling.
External Input capture or clock: TIy after filtering and edge detection.
ICyPS
Internal Input capture signal after filtering, edge detection and prescaling: input to the capture
register.
ITR0
OCy
Internal Internal trigger input: connected to the other timer’s output, TRGO.
External Output compare: TIMxCy when used as an output. Same as OCyREF but includes possible
polarity inversion.
OCyREF Internal Output compare reference: always active high, but may be inverted to produce OCy.
PCLK
External Peripheral clock connects to CK_INT and used to clock input filtering. Its frequency is
12 MHz if using the 24 MHz crystal oscillator and 6 MHz if using the 12 MHz RC oscillator.
TIy
Internal Timer input: TIMxCy when used as a timer input.
TIyFPy
TIMxCy
Internal Timer input after filtering and polarity selection.
Internal Timer channel at a GPIO pin: can be a capture input (ICy) or a compare output (OCy).
TIMxCLK External Clock input (if selected) to the external trigger signal (ETR).
TIMxMSK External Clock mask (if enabled) AND’ed with the other timer’s TIMxCLK signal.
TRGI
Internal Trigger input for slave mode controller.
9-29
120-035X-000 Rev. 1.2
Final
EM351 / EM357
9.4 Interrupts
Each timer has its own top-level NVIC interrupt. Writing 1 to the INT_TIMx bit in the INT_CFGSET register
enables the TIMx interrupt, and writing 1 to the INT_TIMx bit in the INT_CFGCLR register disables it. Chapter
11, Interrupt System describes the interrupt system in detail.
Several kinds of timer events can generate a timer interrupt, and each has a status flag in the INT_TIMxFLAG
register to identify the reason(s) for the interrupt:
INT_TIMTIF – set by a rising edge on an external trigger, either edge in gated mode
.
INT_TIMCCRyIF – set by a channel y input capture or output compare event
.
INT_TIMUIF – set by a UEV
.
Clear bits in INT_TIMxFLAG by writing a 1 to their bit position. When a channel is in capture mode, reading the
TIMx_CCRy register will also clear the INT_TIMCCRyIF bit.
The INT_TIMxCFG register controls whether or not the INT_TIMxFLAG bits actually request a top-level NVIC
timer interrupt. Only the events whose bits are set to 1 in INT_TIMxCFG can do so.
If an input capture or output compare event occurs and its INT_TIMMISSCCyIF is already set, the corresponding
capture/compare missed flag is set in the INT_TMRxMISS register. Clear a bit in the INT_TMRxMISS register by
writing a 1 to it.
9-30
120-035X-000 Rev. 1.2
Final
EM351 / EM357
9.5 Registers
TIMx_CR1
TIM1_CR1
Timer 1 Control Register 1
Address: 0x4000E000 Reset: 0x0
Address: 0x4000F000 Reset: 0x0
TIM2_CR1
Timer 2 Control Register 1
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
22
0
21
0
20
19
18
17
16
0
0
0
0
0
0
15
14
0
13
0
12
11
10
9
8
0
7
0
4
0
3
0
2
0
1
0
0
6
5
TIM_ARBE
TIM_CMS
TIM_DIR
TIM_OPM
TIM_URS
TIM_UDIS
TIM_CEN
Bitname
Bitfield
Access
Description
TIM_ARBE
[7]
RW
Auto-Reload Buffer Enable.
0: TIMx_ARR register is not buffered.
1: TIMx_ARR register is buffered.
TIM_CMS
[6:5]
RW
Center-aligned Mode Selection.
00: Edge-aligned mode. The counter counts up or down depending on the direction bit
(TIM_DIR).
01: Center-aligned mode 1. The counter counts up and down alternatively.
Output compare interrupt flags of configured output channels (TIM_CCyS=00 in
TIMx_CCMRy register) are set only when the counter is counting down.
10: Center-aligned mode 2. The counter counts up and down alternatively.
Output compare interrupt flags of configured output channels (TIM_CCyS=00 in
TIMx_CCMRy register) are set only when the counter is counting up.
11: Center-aligned mode 3. The counter counts up and down alternatively.
Output compare interrupt flags of configured output channels (TIM_CCyS=00 in
TIMx_CCMRy register) are set both when the counter is counting up or down.
Note: Software may not switch from edge-aligned mode to center-aligned mode when the
counter is enabled (TIM_CEN=1).
TIM_DIR
TIM_OPM
TIM_URS
[4]
[3]
[2]
RW
RW
RW
Direction.
0: Counter used as up-counter.
1: Counter used as down-counter.
One Pulse Mode.
0: Counter does not stop counting at the next UEV.
1: Counter stops counting at the next UEV (and clears the bit TIM_CEN).
Update Request Source.
0: When enabled, update interrupt requests are sent as soon as registers are updated
(counter overflow/underflow, setting the TIM_UG bit, or update generation through the
slave mode controller).
1: When enabled, update interrupt requests are sent only when the counter reaches
overflow or underflow.
TIM_UDIS
[1]
RW
Update Disable.
0: A UEV is generated as soon as a counter overflow occurs, a software update is
generated, or a hardware reset is generated by the slave mode controller. Shadow
registers are then loaded with their buffer register values.
1: A UEV is not generated and shadow registers keep their value (TIMx_ARR, TIMx_PSC,
TIMx_CCRy). The counter and the prescaler are reinitialized if the TIM_UG bit is set or if a
hardware reset is received from the slave mode controller.
9-31
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Bitname
Bitfield
Access
Description
TIM_CEN
[0]
RW
Counter Enable.
0: Counter disabled.
1: Counter enabled.
Note: External clock, gated mode and encoder mode can work only if the TIM_CEN bit has
been previously set by software. Trigger mode sets the TIM_CEN bit automatically through
hardware.
9-32
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIMx_CR2
TIM1_CR2
Timer 1 Control Register 2
Address: 0x4000E004 Reset: 0x0
Address: 0x4000F004 Reset: 0x0
TIM2_CR2
Timer 2 Control Register 2
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
TIM_TI1S
TIM_MMS
0
0
0
0
Bitname
Bitfield
Access
Description
TIM_TI1S
[7]
RW
TI1 Selection.
0: TI1M (input of the digital filter) is connected to TI1 input.
1: TI1M is connected to the TI_HALL inputs (XOR combination).
TIM_MMS
[6:4]
RW
Master Mode Selection.
This selects the information to be sent in master mode to a slave timer for
synchronization using the trigger output (TRGO).
000: Reset - the TIM_UG bit in the TMRx_EGR register is trigger output.
If the reset is generated by the trigger input (slave mode controller configured in reset
mode), then the signal on TRGO is delayed compared to the actual reset.
001: Enable - counter enable signal CNT_EN is trigger output.
This mode is used to start both timers at the same time or to control a window in which a
slave timer is enabled. The counter enable signal is generated by either the TIM_CEN
control bit or the trigger input when configured in gated mode. When the counter enable
signal is controlled by the trigger input there is a delay on TRGO except if the
master/slave mode is selected (see the TIM_MSM bit description in TMRx_SMCR register).
010: Update - UEV is trigger output.
This mode allows a master timer to be a prescaler for a slave timer.
011: Compare Pulse.
The trigger output sends a positive pulse when the TIM_CC1IF flag is to be set (even if it
was already high) as soon as a capture or a compare match occurs.
100: Compare - OC1REF signal is trigger output.
101: Compare - OC2REF signal is trigger output.
110: Compare - OC3REF signal is trigger output.
111: Compare - OC4REF signal is trigger output.
9-33
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIMx_SMCR
TIM1_SMCR
Timer 1 Slave Mode Control Register
Address: 0x4000E008 Reset: 0x0
Address: 0x4000F008 Reset: 0x0
TIM2_SMCR
Timer 2 Slave Mode Control Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
TIM_ETP
TIM_ECE
TIM_ETPS
TIM_ETF
7
6
5
4
3
2
1
0
TIM_MSM
TIM_TS
0
TIM_SMS
Bitname
Bitfield
Access
Description
External Trigger Polarity.
TIM_ETP
[15]
RW
This bit selects whether ETR or the inverse of ETR is used for trigger operations.
0: ETR is non-inverted, active at a high level or rising edge.
1: ETR is inverted, active at a low level or falling edge.
TIM_ECE
[14]
RW
External Clock Enable.
This bit enables external clock mode 2.
0: External clock mode 2 disabled.
1: External clock mode 2 enabled. The counter is clocked by any active edge on the ETRF
signal.
Note 1: Setting the TIM_ECE bit has the same effect as selecting external clock mode 1
with TRGI connected to ETRF (TIM_SMS=111 and TIM_TS=111).
Note 2: It is possible to use this mode simultaneously with the following slave modes:
reset mode, gated mode and trigger mode. TRGI must not be connected to ETRF in this
case (the TIM_TS bits must not be 111).
Note 3: If external clock mode 1 and external clock mode 2 are enabled at the same
time, the external clock input will be ETRF.
TIM_ETPS
[13:12]
RW
External Trigger Prescaler.
External trigger signal ETRP frequency must be at most 1/4 of CK frequency. A prescaler
can be enabled to reduce ETRP frequency. It is useful with fast external clocks.
00: ETRP prescaler off.
01: Divide ETRP frequency by 2.
10: Divide ETRP frequency by 4.
11: Divide ETRP frequency by 8.
9-34
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Bitname
Bitfield
Access
Description
TIM_ETF
[11:8]
RW
External Trigger Filter.
This defines the frequency used to sample the ETRP signal, Fsampling, and the length of
the digital filter applied to ETRP. The digital filter is made of an event counter in which N
events are needed to validate a transition on the output:
0000: Fsampling=PCLK, no filtering.
0001: Fsampling=PCLK, N=2.
0010: Fsampling=PCLK, N=4.
0011: Fsampling=PCLK, N=8.
0100: Fsampling=PCLK/2, N=6.
0101: Fsampling=PCLK/2, N=8.
0110: Fsampling=PCLK/4, N=6.
0111: Fsampling=PCLK/4, N=8.
1000: Fsampling=PCLK/8, N=6.
1001: Fsampling=PCLK/8, N=8.
1010: Fsampling=PCLK/16, N=5.
1011: Fsampling=PCLK/16, N=6.
1100: Fsampling=PCLK/16, N=8.
1101: Fsampling=PCLK/32, N=5.
1110: Fsampling=PCLK/32, N=6.
1111: Fsampling=PCLK/32, N=8.
Note: PCLK is 12 MHz when the EM35x is using the 24 MHz crystal oscillator, and 6 MHz if
using the 12 MHz RC oscillator.
TIM_MSM
TIM_TS
[7]
RW
RW
Master/Slave Mode.
0: No action.
1: The effect of an event on the trigger input (TRGI) is delayed to allow exact
synchronization between the current timer and the slave (through TRGO). It is useful for
synchronizing timers on a single external event.
[6:4]
Trigger Selection.
This bit field selects the trigger input used to synchronize the counter.
000 : Internal Trigger 0 (ITR0).
100 : TI1 Edge Detector (TI1F_ED).
101 : Filtered Timer Input 1 (TI1FP1).
110 : Filtered Timer Input 2 (TI2FP2).
111 : External Trigger input (ETRF).
Note: These bits must be changed only when they are not used (when TIM_SMS=000) to
avoid detecting spurious edges during the transition.
TIM_SMS
[2:0]
RW
Slave Mode Selection.
When external signals are selected the active edge of the trigger signal (TRGI) is linked to
the polarity selected on the external input.
000: Slave mode disabled.
If TIM_CEN = 1 then the prescaler is clocked directly by the internal clock.
001: Encoder mode 1. Counter counts up/down on TI1FP1 edge depending on TI2FP2
level.
010: Encoder mode 2. Counter counts up/down on TI2FP2 edge depending on TI1FP1
level.
011: Encoder mode 3. Counter counts up/down on both TI1FP1 and TI2FP2 edges
depending on the level of the other input.
100: Reset Mode. Rising edge of the selected trigger signal (TRGI) >reinitializes the
counter and generates an update of the registers.
101: Gated Mode. The counter clock is enabled when the trigger signal (TRGI) is high. The
counter stops (but is not reset) as soon as the trigger becomes low. Both starting and
stopping the counter are controlled.
110: Trigger Mode. The counter starts at a rising edge of the trigger TRGI (but it is not
reset). Only starting the counter is controlled.
111: External Clock Mode 1. Rising edges of the selected trigger (TRGI) clock the counter.
Note: Gated mode must not be used if TI1F_ED is selected as the trigger input
(TIM_TS=100). TI1F_ED outputs 1 pulse for each transition on TI1F, whereas gated mode
checks the level of the trigger signal.
9-35
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIMx_EGR
TIM1_EGR
Timer 1 Event Generation Register
Address: 0x4000E014 Reset: 0x0
Address: 0x4000F014 Reset: 0x0
TIM2_EGR
Timer 2 Event Generation Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
TIM_TG
0
TIM_CC4G
TIM_CC3G
TIM_CC2G
TIM_CC1G
TIM_UG
Bitname
Bitfield
Access
Description
TIM_TG
[6]
W
Trigger Generation.
0: Does nothing.
1: Sets the TIM_TIF flag in the INT_TIMxFLAG register.
TIM_CC4G
[4]
[3]
[2]
[1]
[0]
W
Capture/Compare 4 Generation.
0: Does nothing.
1: If CC4 configured as output channel:
The TIM_CC4IF flag is set.
If CC4 configured as input channel:
The TIM_CC4IF flag is set.
The INT_TIMMISSCC4IF flag is set if the TIM_CC4IF flag was already high.
The current value of the counter is captured in TMRx_CCR4 register.
TIM_CC3G
TIM_CC2G
TIM_CC1G
TIM_UG
W
W
W
W
Capture/Compare 3 Generation.
0: Does nothing.
1: If CC3 configured as output channel:
The TIM_CC3IF flag is set.
If CC3 configured as input channel:
The TIM_CC3IF flag is set.
The INT_TIMMISSCC3IF flag is set if the TIM_CC3IF flag was already high.
The current value of the counter is captured in TMRx_CCR3 register.
Capture/Compare 2 Generation.
0: Does nothing.
1: If CC2 configured as output channel:
The TIM_CC2IF flag is set.
If CC2 configured as input channel:
The TIM_CC2IF flag is set.
The INT_TIMMISSCC2IF flag is set if the TIM_CC2IF flag was already high.
The current value of the counter is captured in TMRx_CCR2 register.
Capture/Compare 1 Generation.
0: Does nothing.
1: If CC1 configured as output channel:
The TIM_CC1IF flag is set.
If CC1 configured as input channel:
The TIM_CC1IF flag is set.
The INT_TIMMISSCC1IF flag is set if the TIM_CC1IF flag was already high.
The current value of the counter is captured in TMRx_CCR1 register.
Update Generation.
0: Does nothing.
1: Re-initializes the counter and generates an update of the registers. This also clears the
prescaler counter but the prescaler ratio is not affected. The counter is cleared if center-
aligned mode is selected or if TIM_DIR=0 (up-counting), otherwise it takes the auto-reload
value (TMR1_ARR) if TIM_DIR=1 (down-counting).
9-36
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIMx_CCMR1
TIM1_CCMR1
Timer 1 Capture/Compare Mode Register 1
Address: 0x4000E018 Reset: 0x0
Address: 0x4000F018 Reset: 0x0
TIM2_CCMR1
Timer 2 Capture/Compare Mode Register 1
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
TIM_OC2M
TIM_OC2BE
TIM_OC2FE
TIM_CC2S
TIM_CC1S
TIM_IC2F
TIM_IC1F
TIM_IC2PSC
7
6
5
4
3
2
1
0
0
TIM_OC1M
TIM_OC1BE
TIM_OC1FE
TIM_IC1PSC
Timer channels can be programmed as inputs (capture mode) or outputs (compare mode). The direction of channel y is defined by TIM_CCyS in this
register.
The other bits in this register have different functions in input and in output modes. The TIM_OC* fields only apply to a channel configured as an
output (TIM_CCyS = 0), and the TIM_IC* fields only apply to a channel configured as an input (TIM_CCyS > 0).
Bitname
Bitfield
Access
Description
TIM_OC2M
[14:12]
RW
Output Compare 2 Mode. (Applies only if TIM_CC2S = 0.)
Define the behavior of the output reference signal OC2REF from which OC2 derives.
OC2REF is active high whereas OC2''s active level depends on the TIM_CC2P bit.
000: Frozen - The comparison between the output compare register TIMx_CCR2 and the
counter TIMx_CNT has no effect on the outputs.
001: Set OC2REF to active on match. The OC2REF signal is forced high when the counter
TIMx_CNT matches the capture/compare register 2 (TIMx_CCR2)
010: Set OC2REF to inactive on match. OC2REF signal is forced low when the counter
TIMx_CNT matches the capture/compare register 2 (TIMx_CCR2).
011: Toggle - OC2REF toggles when TIMx_CNT = TIMx_CCR2.
100: Force OC2REF inactive.
101: Force OC2REF active.
110: PWM mode 1 - In up-counting, OC2REF is active as long as TIMx_CNT < TIMx_CCR2,
otherwise OC2REF is inactive. In down-counting, OC2REF is inactive if
TIMx_CNT > TIMx_CCR2, otherwise OC2REF is active.
111: PWM mode 2 - In up-counting, OC2REF is inactive if TIMx_CNT < TIMx_CCR2,
otherwise OC2REF is active. In down-counting, OC2REF is active if TIMx_CNT > TIMx_CCR2,
otherwise it is inactive.
Note: In PWM mode 1 or 2, the OC2REF level changes only when the result of the
comparison changes or when the output compare mode switches from “frozen” mode to
“PWM” mode.
TIM_OC2BE
[11]
RW
Output Compare 2 Buffer Enable. (Applies only if TIM_CC2S = 0.)
0: Buffer register for TIMx_CCR2 is disabled. TIMx_CCR2 can be written at anytime, the
new value is used by the shadow register immediately.
1: Buffer register for TIMx_CCR2 is enabled. Read/write operations access the buffer
register. TIMx_CCR2 buffer value is loaded in the shadow register at each UEV.
Note: The PWM mode can be used without enabling the buffer register only in one pulse
mode (TIM_OPM bit set in the TIMx_CR2 register), otherwise the behavior is undefined.
9-37
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Bitname
Bitfield
Access
Description
TIM_OC2FE
[10]
RW
Output Compare 2 Fast Enable. (Applies only if TIM_CC2S = 0.)
This bit speeds the effect of an event on the trigger in input on the OC2 output.
0: OC2 behaves normally depending on the counter and TIM_CCR2 values even when the
trigger is ON. The minimum delay to activate OC2 when an edge occurs on the trigger
input is 5 clock cycles.
1: An active edge on the trigger input acts like a compare match on the OC2 output. OC2
is set to the compare level independently from the result of the comparison. Delay to
sample the trigger input and to activate OC2 output is reduced to 3 clock cycles.
TIM_OC2FE acts only if the channel is configured in PWM 1 or PWM 2 mode.
TIM_IC2F
[15:12]
RW
Input Capture 1 Filter. (Applies only if TIM_CC2S > 0.)
This defines the frequency used to sample the TI2 input, Fsampling, and the length of the
digital filter applied to TI2. The digital filter requires N consecutive samples in the same
state before being output.
0000: Fsampling=PCLK, no filtering.
0001: Fsampling=PCLK, N=2.
0010: Fsampling=PCLK, N=4.
0011: Fsampling=PCLK, N=8.
0100: Fsampling=PCLK/2, N=6.
0101: Fsampling=PCLK/2, N=8.
0110: Fsampling=PCLK/4, N=6.
0111: Fsampling=PCLK/4, N=8.
1000: Fsampling=PCLK/8, N=6.
1001: Fsampling=PCLK/8, N=8.
1010: Fsampling=PCLK/16, N=5.
1011: Fsampling=PCLK/16, N=6.
1100: Fsampling=PCLK/16, N=8.
1101: Fsampling=PCLK/32, N=5.
1110: Fsampling=PCLK/32, N=6.
1111: Fsampling=PCLK/32, N=8.
Note: PCLK is 12 MHz when using the 24 MHz crystal oscillator, and 6 MHz using the 12 MHz
RC oscillator.
TIM_IC2PSC
TIM_CC2S
[11:10]
[9:8]
RW
RW
Input Capture 1 Prescaler. (Applies only if TIM_CC2S > 0.)
00: No prescaling, capture each time an edge is detected on the capture input.
01: Capture once every 2 events.
10: Capture once every 4 events.
11: Capture once every 8 events.
Capture / Compare 2 Selection.
This configures the channel as an output or an input. If an input, it selects the input
source.
00: Channel is an output.
01: Channel is an input and is mapped to TI2.
10: Channel is an input and is mapped to TI1.
11: Channel is an input and is mapped to TRGI. This mode requires an internal trigger
input selected by the TIM_TS bit in the TIMx_SMCR register.
Note: TIM_CC2S may be written only when the channel is off (TIM_CC2E = 0 in the
TIMx_CCER register).
TIM_OC1M
TIM_OC1BE
TIM_OC1FE
TIM_IC1F
[6:4]
[3
RW
RW
RW
RW
RW
Output Compare 1 Mode. (Applies only if TIM_CC1S = 0.)
See TIM_OC2M description above.
Output Compare 1 Buffer Enable. (Applies only if TIM_CC1S = 0.)
See TIM_OC2BE description above.
[2]
Output Compare 1 Fast Enable. (Applies only if TIM_CC1S = 0.)
See TIM_OC2FE description above.
[7:4]
[3:2]
Input Capture 1 Filter. (Applies only if TIM_CC1S > 0.)
See TIM_IC2F description above.
TIM_IC1PSC
Input Capture 1 Prescaler. (Applies only if TIM_CC1S > 0.)
See TIM_IC2PSC description above.
9-38
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Bitname
Bitfield
Access
Description
TIM_CC1S
[1:0]
RW
Capture / Compare 1 Selection.
This configures the channel as an output or an input. If an input, it selects the input
source.
00: Channel is an output.
01: Channel is an input and is mapped to TI1.
10: Channel is an input and is mapped to TI2.
11: Channel is an input and is mapped to TRGI. This requires an internal trigger input
selected by the TIM_TS bit in the TIM_SMCR register.
Note: TIM_CC1S may be written only when the channel is off (TIM_CC1E = 0 in the
TIMx_CCER register).
9-39
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIMx_CCMR2
TIM1_CCMR2
Timer 1 Capture/Compare Mode Register 2
Address: 0x4000E01C Reset: 0x0
Address: 0x4000F01C Reset: 0x0
TIM2_CCMR2
Timer 2 Capture/Compare Mode Register 2
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
TIM_OC4M
TIM_OC4BE
TIM_OC4FE
TIM_CC4S
TIM_CC3S
TIM_IC4F
TIM_IC3F
TIM_IC4PSC
7
6
5
4
3
2
1
0
0
TIM_OC3M
TIM_OC3BE
TIM_OC3FE
TIM_IC3PSC
Timer channels can be programmed as inputs (capture mode) or outputs (compare mode). The direction of channel y is defined by TIM_CCyS in this
register.
The other bits in this register have different functions in input and in output modes. The TIM_OC* fields only apply to a channel configured as an
output (TIM_CCyS = 0), and the TIM_IC* fields only apply to a channel configured as an input (TIM_CCyS > 0).
Bitname
Bitfield
Access
Description
TIM_OC4M
[14:12]
RW
Output Compare 4 Mode. (Applies only if TIM_CC4S = 0.)
Define the behavior of the output reference signal OC4REF from which OC4 derives.
OC4REF is active high whereas OC4’s active level depends on the TIM_CC4P bit.
000: Frozen - The comparison between the output compare register TIMx_CCR4 and the
counter TIMx_CNT has no effect on the outputs.
001: Set OC4REF to active on match. The OC4REF signal is forced high when the counter
TIMx_CNT matches the capture/compare register 4 (TIMx_CCR4)
010: Set OC4REF to inactive on match. OC4REF signal is forced low when the counter
TIMx_CNT matches the capture/compare register 4 (TIMx_CCR4).
011: Toggle - OC4REF toggles when TIMx_CNT = TIMx_CCR4.
100: Force OC4REF inactive.
101: Force OC4REF active.
110: PWM mode 1 - In up-counting, OC4REF is active as long as TIMx_CNT < TIMx_CCR4,
otherwise OC4REF is inactive. In down-counting, OC4REF is inactive if
TIMx_CNT > TIMx_CCR4, otherwise OC4REF is active.
111: PWM mode 2 - In up-counting, OC4REF is inactive if TIMx_CNT < TIMx_CCR4,
otherwise OC4REF is active. In down-counting, OC4REF is active if TIMx_CNT > TIMx_CCR4,
otherwise it is inactive.
Note: In PWM mode 1 or 2, the OC4REF level changes only when the result of the
comparison changes or when the output compare mode switches from “frozen” mode to
“PWM” mode.
TIM_OC4BE
[11]
RW
Output Compare 4 Buffer Enable. (Applies only if TIM_CC4S = 0.)
0: Buffer register for TIMx_CCR4 is disabled. TIMx_CCR4 can be written at anytime, the
new value is used by the shadow register immediately.
1: Buffer register for TIMx_CCR4 is enabled. Read/write operations access the buffer
register. TIMx_CCR4 buffer value is loaded in the shadow register at each UEV.
Note: The PWM mode can be used without enabling the buffer register only in one pulse
mode (TIM_OPM bit set in the TIMx_CR2 register), otherwise the behavior is undefined.
9-40
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Bitname
Bitfield
Access
Description
TIM_OC4FE
[10]
RW
Output Compare 4 Fast Enable. (Applies only if TIM_CC4S = 0.)
This bit speeds the effect of an event on the trigger in input on the OC4 output.
0: OC4 behaves normally depending on the counter and TIM_CCR4 values even when the
trigger is ON. The minimum delay to activate OC4 when an edge occurs on the trigger
input is 5 clock cycles.
1: An active edge on the trigger input acts like a compare match on the OC4 output. OC4
is set to the compare level independently from the result of the comparison. Delay to
sample the trigger input and to activate OC4 output is reduced to 3 clock cycles.
TIM_OC4FE acts only if the channel is configured in PWM 1 or PWM 2 mode.
TIM_IC4F
[15:12]
RW
Input Capture 4 Filter. (Applies only if TIM_CC4S > 0.)
This defines the frequency used to sample the TI4 input, Fsampling, and the length of the
digital filter applied to TI4. The digital filter requires N consecutive samples in the same
state before being output.
0000: Fsampling=PCLK, no filtering.
0001: Fsampling=PCLK, N=2.
0010: Fsampling=PCLK, N=4.
0011: Fsampling=PCLK, N=8.
0100: Fsampling=PCLK/2, N=6.
0101: Fsampling=PCLK/2, N=8.
0110: Fsampling=PCLK/4, N=6.
0111: Fsampling=PCLK/4, N=8.
1000: Fsampling=PCLK/8, N=6.
1001: Fsampling=PCLK/8, N=8.
1010: Fsampling=PCLK/16, N=5.
1011: Fsampling=PCLK/16, N=6.
1100: Fsampling=PCLK/16, N=8.
1101: Fsampling=PCLK/32, N=5.
1110: Fsampling=PCLK/32, N=6.
1111: Fsampling=PCLK/32, N=8.
Note: PCLK is 12 MHz when using the 24 MHz crystal oscillator, and 6 MHz using the 12 MHz
RC oscillator.
TIM_IC4PSC
TIM_CC4S
[11:10]
[9:8]
RW
RW
Input Capture 4 Prescaler. (Applies only if TIM_CC4S > 0.)
00: No prescaling, capture each time an edge is detected on the capture input.
01: Capture once every 2 events.
10: Capture once every 4 events.
11: Capture once every 8 events.
Capture / Compare 4 Selection.
This configures the channel as an output or an input. If an input, it selects the input
source.
00: Channel is an output.
01: Channel is an input and is mapped to TI4.
10: Channel is an input and is mapped to TI3.
11: Channel is an input and is mapped to TRGI. This mode requires an internal trigger
input selected by the TIM_TS bit in the TIMx_SMCR register.
Note: TIM_CC4S may be written only when the channel is off (TIM_CC4E = 0 in the
TIMx_CCER register).
TIM_OC3M
TIM_OC3BE
TIM_OC3FE
TIM_IC3F
[6:4]
[3
RW
RW
RW
RW
RW
Output Compare 3 Mode. (Applies only if TIM_CC3S = 0.)
See TIM_OC4M description above.
Output Compare 3 Buffer Enable. (Applies only if TIM_CC3S = 0.)
See TIM_OC4BE description above.
[2]
Output Compare 3 Fast Enable. (Applies only if TIM_CC3S = 0.)
See TIM_OC4FE description above.
[7:4]
[3:2]
Input Capture 3 Filter. (Applies only if TIM_CC3S > 0.)
See TIM_IC4F description above.
TIM_IC3PSC
Input Capture 3 Prescaler. (Applies only if TIM_CC3S > 0.)
See TIM_IC4PSC description above.
9-41
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Bitname
Bitfield
Access
Description
TIM_CC3S
[1:0]
RW
Capture / Compare 3 Selection.
This configures the channel as an output or an input. If an input, it selects the input
source.
00: Channel is an output.
01: Channel is an input and is mapped to TI3.
10: Channel is an input and is mapped to TI4.
11: Channel is an input and is mapped to TRGI. This requires an internal trigger input
selected by the TIM_TS bit in the TIM_SMCR register.
Note: TIM_CC3S may be written only when the channel is off (TIM_CC3E = 0 in the
TIMx_CCER register).
9-42
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIMx_CCER
TIM1_CCER
Timer 1 Capture/Compare Enable Register
Address: 0x4000E020 Reset: 0x0
Address: 0x4000F020 Reset: 0x0
TIM2_CCER
Timer 2 Capture/Compare Enable Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
TIM_CC4P
TIM_CC4E
0
0
TIM_CC3P
TIM_CC3E
7
6
5
4
3
2
1
0
0
0
TIM_CC2P
TIM_CC2E
0
0
TIM_CC1P
TIM_CC1E
Bitname
TIM_CC4P
Bitfield
Access
RW
Description
[13]
Capture/Compare 4 output Polarity.
If CC4 is configured as an output channel:
0: OC4 is active high.
1: OC4 is active low.
If CC4 configured as an input channel:
0: IC4 is not inverted. Capture occurs on a rising edge of IC4. When used as an external
trigger, IC4 is not inverted.
1: IC4 is inverted. Capture occurs on a falling edge of IC4. When used as an external
trigger, IC4 is inverted.
TIM_CC4E
[12]
RW
Capture/Compare 4 output Enable.
If CC4 is configured as an output channel:
0: OC4 is disabled.
1: OC4 is enabled.
If CC4 configured as an input channel:
0: Capture is disabled.
1: Capture is enabled.
TIM_CC3P
TIM_CC3E
TIM_CC2P
TIM_CC2E
TIM_CC1P
TIM_CC1E
[9]
[8]
[5]
[4]
[1]
[0]
RW
RW
RW
RW
RW
RW
Refer to the CC4P description above.
Refer to the CC4E description above.
Refer to the CC4P description above.
Refer to the CC43 description above.
Refer to the CC4P description above.
Refer to the CC4E description above.
9-43
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIMx_CNT
TIM1_CNT
Timer 1 Counter Register
Address: 0x4000E024 Reset: 0x0
Address: 0x4000F024 Reset: 0x0
TIM2_CNT
Timer 2 Counter Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
TIM_CNT
TIM_CNT
7
6
5
4
3
2
1
0
Bitname
TIM_CNT
Bitfield
Access
RW
Description
[15:0]
Counter value.
9-44
Final
120-035X-000 Rev. 1.2
EM351 / EM357
TIMx_PSC
TIM1_PSC
Timer 1 Prescaler Register
Address: 0x4000E028 Reset: 0x0
Address: 0x4000F028 Reset: 0x0
TIM2_PSC
Timer 2 Prescaler Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
TIM_PSC
Bitname
TIM_PSC
Bitfield
Access
RW
Description
[3:0]
The prescaler divides the internal timer clock frequency. The counter clock frequency
CK_CNT is equal to fCK_PSC / (2 ^ TIM_PSC). Clock division factors can range from 1
through 32768. The division factor is loaded into the shadow prescaler register at each
UEV (including when the counter is cleared through TIM_UG bit of TMR1_EGR register or
through the trigger controller when configured in reset mode).
9-45
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIMx_ARR
TIM1_ARR
Timer 1 Auto-Reload Register
Address: 0x4000E02C Reset: 0xFFFF
Address: 0x4000F02C Reset: 0xFFFF
TIM2_ARR
Timer 2 Auto-Reload Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
TIM_ARR
TIM_ARR
7
6
5
4
3
2
1
0
Bitname
TIM_ARR
Bitfield
Access
RW
Description
[15:0]
TIM_ARR is the value to be loaded in the shadow auto-reload register.
The auto-reload register is buffered. Writing or reading the auto-reload register accesses
the buffer register. The content of the buffer register is transfered in the shadow register
permanently or at each UEV, depending on the auto-reload buffer enable bit (TIM_ARBE)
in TMRx_CR1 register. The UEV is sent when the counter reaches the overflow point (or
underflow point when down-counting) and if the TIM_UDIS bit equals 0 in the TMRx_CR1
register. It can also be generated by software. The counter is blocked while the auto-
reload value is 0.
9-46
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIMx_CCR1
TIM1_CCR1
Timer 1 Capture/Compare Register 1
Address: 0x4000E034 Reset: 0x0
Address: 0x4000F034 Reset: 0x0
TIM2_CCR1
Timer 2 Capture/Compare Register 1
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
TIM_CCR
TIM_CCR
7
6
5
4
3
2
1
0
Bitname
TIM_CCR
Bitfield
Access
RW
Description
[15:0]
If the CC1 channel is configured as an output (TIM_CC1S = 0):
TIM_CCR1 is the buffer value to be loaded in the actual capture/compare 1 register. It is
loaded permanently if the preload feature is not selected in the TMR1_CCMR1 register
(bit OC1PE). Otherwise the buffer value is copied to the shadow capture/compare 1
register when an UEV occurs. The active capture/compare register contains the value to
be compared to the counter TMR1_CNT and signaled on the OC1 output.
If the CC1 channel is configured as an input (TIM_CC1S is not 0):
CCR1 is the counter value transferred by the last input capture 1 event (IC1).
9-47
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIMx_CCR2
TIM1_CCR2
Timer 1 Capture/Compare Register 2
Address: 0x4000E038 Reset: 0x0
Address: 0x4000F038 Reset: 0x0
TIM2_CCR2
Timer 2 Capture/Compare Register 2
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
TIM_CCR
TIM_CCR
7
6
5
4
3
2
1
0
Bitname
TIM_CCR
Bitfield
Access
RW
Description
[15:0]
See description in the TIMx_CCR1 register.
9-48
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIMx_CCR3
TIM1_CCR3
Timer 1 Capture/Compare Register 3
Address: 0x4000E03C Reset: 0x0
Address: 0x4000F03C Reset: 0x0
TIM2_CCR3
Timer 2 Capture/Compare Register 3
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
TIM_CCR
TIM_CCR
7
6
5
4
3
2
1
0
Bitname
TIM_CCR
Bitfield
Access
RW
Description
[15:0]
See description in the TIMx_CCR1 register.
9-49
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIMx_CCR4
TIM1_CCR4
Timer 1 Capture/Compare Register 4
Address: 0x4000E040 Reset: 0x0
Address: 0x4000F040 Reset: 0x0
TIM2_CCR4
Timer 2 Capture/Compare Register 4
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
TIM_CCR
TIM_CCR
7
6
5
4
3
2
1
0
Bitname
TIM_CCR
Bitfield
Access
RW
Description
[15:0]
See description in the TIMx_CCR1 register.
9-50
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIM1_OR
Timer 1 Option Register
Address: 0x4000E050 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
TIM_ORRSVD
TIM_CLKMSKEN
TIM_EXTRIGSEL
Bitname
Bitfield
[3]
Access
Description
TIM_ORRSVD
RW
RW
Reserved: this bit must always be set to 0.
TIM_CLKMSKEN
[2]
Enables TIM1MSK when TIM1CLK is selected as the external trigger: 0 = TIM1MSK not used,
1 = TIM1CLK is ANDed with the TIM1MSK input.
TIM_EXTRIGSEL
[1:0]
RW
Selects the external trigger used in external clock mode 2: 0 = PCLK, 1 = calibrated 1 kHz
clock, 2 = 32 kHz reference clock (if available), 3 = TIM1CLK pin.
9-51
120-035X-000 Rev. 1.2
Final
EM351 / EM357
TIM2_OR
Timer 2 Option Register
Address: 0x4000F050 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
TIM_REMAPC4
TIM_REMAPC3
TIM_REMAPC2
TIM_REMAPC1
TIM_ORRSVD
TIM_CLKMSKEN
TIM_EXTRIGSEL
Bitname
Bitfield
Access
RW
Description
TIM_REMAPC4
TIM_REMAPC3
TIM_REMAPC2
TIM_REMAPC1
TIM_ORRSVD
[7]
[6]
[5]
[4]
[3]
[2]
Selects the GPIO used for TIM2C4: 0 = PA2, 1 = PB4.
Selects the GPIO used for TIM2C3: 0 = PA1, 1 = PB3.
Selects the GPIO used for TIM2C2: 0 = PA3, 1 = PB2.
Selects the GPIO used for TIM2C1: 0 = PA0, 1 = PB1.
Reserved: this bit must always be set to 0.
RW
RW
RW
RW
TIM_CLKMSKEN
RW
Enables TIM2MSK when TIM2CLK is selected as the external trigger: 0 = TIM2MSK not used,
1 = TIM2CLK is ANDed with the TIM2MSK input.
TIM_EXTRIGSEL
[1:0]
RW
Selects the external trigger used in external clock mode 2: 0 = PCLK, 1 = calibrated 1 kHz
clock, 2 = 32 kHz reference clock (if available), 3 = TIM2CLK pin.
9-52
120-035X-000 Rev. 1.2
Final
EM351 / EM357
INT_TIMxCFG
INT_TIM1CFG
Timer 1 Interrupt Configuration Register
Address: 0x4000A840 Reset: 0x0
Address: 0x4000A844 Reset: 0x0
INT_TIM2CFG
Timer 2 Interrupt Configuration Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
INT_TIMTIF
0
INT_TIMCC4IF
INT_TIMCC3IF
INT_TIMCC2IF
INT_TIMCC1IF
INT_TIMUIF
Bitname
Bitfield
Access
Description
INT_TIMTIF
[6]
[4]
[3]
[2]
[1]
[0]
RW
RW
RW
RW
RW
RW
Trigger interrupt enable.
INT_TIMCC4IF
INT_TIMCC3IF
INT_TIMCC2IF
INT_TIMCC1IF
INT_TIMUIF
Capture or compare 4 interrupt enable.
Capture or compare 3 interrupt enable.
Capture or compare 2 interrupt enable.
Capture or compare 1 interrupt enable.
Update interrupt enable.
9-53
120-035X-000 Rev. 1.2
Final
EM351 / EM357
INT_TIMxFLAG
INT_TIM1FLAG
Timer 1 Interrupt Flag Register
Address: 0x4000A800 Reset: 0x0
Address: 0x4000A804 Reset: 0x0
INT_TIM2FLAG
Timer 2 Interrupt Flag Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
INT_TIMRSVD
0
7
6
5
4
3
2
1
0
0
INT_TIMTIF
0
INT_TIMCC4IF
INT_TIMCC3IF
INT_TIMCC2IF
INT_TIMCC1IF
INT_TIMUIF
Bitname
Bitfield
Access
Description
INT_TIMRSVD
INT_TIMTIF
[12:9]
[6]
R
May change during normal operation.
Trigger interrupt.
RW
RW
RW
RW
RW
RW
INT_TIMCC4IF
INT_TIMCC3IF
INT_TIMCC2IF
INT_TIMCC1IF
INT_TIMUIF
[4]
Capture or compare 4 interrupt pending.
Capture or compare 3 interrupt pending.
Capture or compare 2 interrupt pending.
Capture or compare 1 interrupt pending.
Update interrupt pending.
[3]
[2]
[1]
[0]
9-54
120-035X-000 Rev. 1.2
Final
EM351 / EM357
INT_TIMxMISS
INT_TIM1MISS
Timer 1 Missed Interrupt Register
Address: 0x4000A818 Reset: 0x0
Address: 0x4000A81C Reset: 0x0
INT_TIM2MISS
Timer 2 Missed Interrupts Register
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
INT_TIMMISSCC4IF
INT_TIMMISSCC3IF
INT_TIMMISSCC2IF
INT_TIMMISSCC1IF
0
7
6
5
4
3
2
1
0
0
INT_TIMMISSRSVD
Bitname
Bitfield
[12]
Access
Description
INT_TIMMISSCC4IF
INT_TIMMISSCC3IF
INT_TIMMISSCC2IF
INT_TIMMISSCC1IF
INT_TIMMISSRSVD
RW
RW
RW
RW
R
Capture or compare 4 interrupt missed.
Capture or compare 3 interrupt missed.
Capture or compare 2 interrupt missed.
Capture or compare 1 interrupt missed.
May change during normal operation.
[11]
[10]
[9]
[6:0]
9-55
120-035X-000 Rev. 1.2
Final
EM351 / EM357
10 ADC (Analog to Digital Converter)
The EM35x ADC is a first-order sigma-delta converter with the following features:
Resolution of up to 14 bits
.
.
.
.
.
.
.
.
Sample times as fast as 5.33 µs (188 kHz)
Differential and single-ended conversions from six external and four internal sources
One voltage range (differential): -VREF to +VREF
Choice of internal or external VREF
internal VREF may be output to PB0 or external VREF may be derived from PB0
Digital offset and gain correction
Dedicated DMA channel with one-shot and continuous operating modes
Figure 10-1 shows the basic ADC structure.
Figure 10-1. ADC Block Diagram
P input
GPIO
MUX
VDD_PADSA/2
VREF
`
VREF/2
GND
Delta
Sigma
ADC
Offset and
Gain
Correction
ADC_DATA
register
or DMA
Decimator
N input
GPIO
MUX
1MHz
6MHz
VDD_PADSA/2
VREF
`
VREF/2
GND
Sample clock
While the ADC Module supports both single-ended and differential inputs, the ADC input stage always operates
in differential mode. Single-ended conversions are performed by connecting one of the differential inputs to
VREF/2 while fully differential operation uses two external inputs.
Note: The regulator input voltage, VDD_PADS, cannot be measured using the ADC, but it can be measured
through Ember software.
10.1
Setup and Configuration
To use the ADC follow this procedure, described in more detail in the next sections:
Configure any GPIO pins to be used by the ADC in analog mode.
.
Configure the voltage reference (internal or external).
.
Set the offset and gain values.
.
.
If using DMA, reset the ADC DMA, define the DMA buffer, and start the DMA in the proper transfer mode.
10-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
If interrupts will be used, configure the top-level and second-level ADC interrupt bits.
.
.
Write the ADC configuration register to define the inputs, sample time, and start the conversions.
10.1.1 GPIO Usage
A GPIO pin used by the ADC as an input or voltage reference must be configured in analog mode by writing 0
to its 4-bit field in the proper GPIO_PxCFGH/L register. Note that a GPIO pin in analog mode cannot be used
for any digital functions, and GPIO_PxIN always reads it as 1. Only certain pins can be configured in analog
mode. These are listed in Table 10-1.
Table 10-1. ADC GPIO Pin Usage
Analog Signal
GPIO
Configuration control
ADC0 input
PB5
GPIO_PBCFGH[7:4]
ADC1 input
PB6
PB7
PC1
PA4
PA5
PB0
GPIO_PBCFGH[11:8]
GPIO_PBCFGH[15:12]
GPIO_PCCFGL[7:4]
GPIO_PACFGH[3:0]
GPIO_PACFGH[7:4]
GPIO_PBCFGL[3:0]
ADC2 input
ADC3 input
ADC4 input
ADC5 input
VREF input or output
See Chapter 7, GPIO for more information about how to configure GPIO.
10.1.2 Voltage Reference
The ADC voltage reference (VREF), may be internally generated or externally sourced from PB0. If internally
generated, it may optionally be output on PB0. To output the internal VREF on PB0, the ADC must be enabled
(ADC_ENABLE bit set in the ADC_CFG register) and PB0 must be configured in analog mode.
To use an external reference, the Ember software must be called after reset and after waking from deep
sleep. PB0 must also be configured in analog mode using GPIO_PBCFGH[3:0]. See the Ember software
documentation for more information on using an external reference.
10.1.3 Offset/Gain Correction
When a conversion is complete, the 16-bit converted data is processed in several steps by offset/gain
correction hardware:
1. The initial signed ADC conversion result is added to the 16-bit signed (two’s complement) value of the ADC
offset register (ADC_OFFSET).
2. The offset-corrected data is multiplied by the 16-bit ADC gain register, ADC_GAIN, to produce a 16-bit
signed result. If the product is greater than 0x7FFF (32767), or less than 0x8000 (-32768), it is limited to
that value and the INT_ADCSAT bit is set in the INT_ADCFLAG register.
3. The offset/gain corrected value is divided by two to produce the final result.
ADC_GAIN is an unsigned scaled 16-bit value: ADC_GAIN[15] is the integer part of the gain factor and
ADC_GAIN[14:0] is the fractional part. As a result, ADC_GAIN values can represent gain factors from 0 through
(2 – 2-15). Although ADC_GAIN can represent a much greater range, its purpose is to correct small gain error,
and in practice is loaded with values within a range of about 0.95 to 1.05.
10-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Reset initializes the offset to zero (ADC_OFFSET = 0) and gain factor to one (ADC_GAIN = 0x8000).
10.1.4 DMA
The ADC DMA channel writes converted data, which incorporates the offset/gain correction, into a DMA buffer
in RAM.
The ADC DMA buffer is defined by two registers:
ADC_DMABEG is the start address of the buffer and must be even.
.
.
ADC_DMASIZE specifies the size of the buffer in 16-bit samples, or half its length in bytes.
To prepare the DMA channel for operation, reset it by writing the ADC_DMARST bit in the ADC_DMACFG
register, then start the DMA in either linear or auto wrap mode by setting the ADC_DMALOAD bit in the
ADC_DMACFG register. The ADC_DMAAUTOWRAP bit in the ADC_DMACFG register selects the DMA mode: 0 for
linear mode, 1 for auto wrap mode.
In linear mode the DMA writes to the buffer until the number of samples given by ADC_DMASIZE has been
.
output. The DMA then stops and sets the INT_ADCULDFULL bit in the INT_ADCFLAG register. If another ADC
conversion completes before the DMA is reset or the ADC is disabled, the INT_ADCOVF bit in the
INT_ADCFLAG register is set.
In auto wrap mode the DMA writes to the buffer until it reaches the end, then resets its pointer to the
start of the buffer and continues writing samples. The DMA transfers continue until the ADC is disabled or
the DMA is reset.
.
When the DMA fills the lower and upper halves of the buffer, it sets the INT_ADCULDHALF and
INT_ADCULDFULL bits, respectively, in the INT_ADCFLAG register. The current location to which the DMA is
writing can also be determined by reading the ADC_DMACUR register.
10.1.5 ADC Configuration Register
The ADC configuration register (ADC_CFG) sets up most of the ADC operating parameters.
10.1.5.1 Input
The analog input of the ADC can be chosen from various sources. The analog input is configured with the
ADC_MUXP and ADC_MUXN bits within the ADC_CFG register. Table 10-2 shows the possible input selections.
10-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 10-2. ADC Inputs
ADC_MUXn1 Analog source at ADC
GPIO pin
Purpose
0
1
ADC0
PB5
ADC1
PB6
PB7
PC1
PA4
PA5
2
ADC2
3
ADC3
4
ADC4
5
ADC5
6
No connection
No connection
GND
7
8
Internal connection
Internal connection
Internal connection
Internal connection
Calibration
9
VREF/2
Calibration
10
11
12
13
14
15
VREF
Calibration
VDD_PADSA/2
No connection
No connection
No connection
No connection
Supply monitoring and calibration
1Denotes bits ADC_MUXP or ADC_MUXN in register ADC_CFG.
Table 10-3 shows the typical configurations of ADC inputs.
Table 10-3. Typical ADC Input Configurations
ADC P input
ADC0
ADC N input
VREF/2
VREF/2
VREF/2
VREF/2
VREF/2
VREF/2
ADC0
ADC_MUXP
ADC_MUXN
Purpose
0
1
9
9
9
9
9
9
0
2
4
9
9
9
Single-ended
Single-ended
Single-ended
Single-ended
Single-ended
Single-ended
Differential
Differential
Differential
Calibration
Calibration
Calibration
ADC1
ADC2
2
ADC3
3
ADC4
4
ADC5
5
ADC1
1
ADC3
ADC2
3
ADC5
ADC4
5
GND
VREF/2
VREF/2
VREF/2
8
VREF
10
11
VDD_PADSA/2
10-4
120-035X-000 Rev. 1.2
Final
EM351 / EM357
10.1.5.2 Input Range
The single-ended input range is fixed as 0 V to VREF and the differential input range is fixed as -VREF to
+VREF.
10.1.5.3 Sample Time
ADC sample time is programmed by selecting the sampling clock and the clocks per sample.
The sampling clock may be either 1 MHz or 6 MHz. If the ADC_1MHZCLK bit in the ADC_CFG register is
clear, the 6 MHz clock is used; if it is set, the 1 MHz clock is selected. The 6 MHz sample clock offers faster
conversion times but the ADC resolution is lower than that achieved with the 1 MHz clock.
.
The number of clocks per sample is determined by the ADC_PERIOD bits in the ADC_CFG register.
.
ADC_PERIOD values select from 32 to 4096 sampling clocks in powers of two. Longer sample times produce
more significant bits. Regardless of the sample time, converted samples are always 16-bits in size with the
significant bits left-aligned within the value.
Table 10-4 shows the options for ADC sample times and the significant bits in the conversion results.
Table 10-4. ADC Sample Times
Sample Time (µs)
1 MHz clock 6 MHz clock
Sample Frequency (kHz)
Sample
Clocks
ADC_PERIOD
Significant Bits
1 MHz clock
6 MHz clock
0
1
2
3
4
5
6
7
32
64
32
64
5.33
10.7
21.3
42.7
85.3
170
31.3
188
7
15.6
7.81
93.8
46.9
23.4
11.7
5.86
2.93
1.47
8
128
256
512
1024
2048
4096
128
256
512
1024
2048
4096
9
3.91
10
11
12
13
14
1.95
0.977
0.488
0.244
341
682
Note: ADC sample timing is the same whether the EM35x is using the 24 MHz crystal oscillator or the 12 MHz
high-speed RC oscillator. This facilitates using the ADC soon after the CPU wakes from deep sleep, before
switching to the crystal oscillator.
10.2
Interrupts
The ADC has its own top-level interrupt in the NVIC. The ADC interrupt is enabled by writing the INT_ADC bit
to the INT_CFGSET register, and cleared by writing the INT_ADC bit to the INT_CFGCLR register. Chapter 11,
Interrupt System, describes the interrupt system in detail.
Five kinds of ADC events can generate an ADC interrupt, and each has a bit flag in the INT_ADCFLAG register
to identify the reason(s) for the interrupt:
INT_ADCOVF – an ADC conversion result was ready but the DMA was disabled (DMA buffer overflow).
.
.
INT_ADCSAT– the gain correction multiplication exceeded the limits for a signed 16-bit number (gain
saturation).
10-5
120-035X-000 Rev. 1.2
Final
EM351 / EM357
INT_ADCULDFULL – the DMA wrote to the last location in the buffer (DMA buffer full).
.
.
INT_ADCULDHALF – the DMA wrote to the last location of the first half of the DMA buffer (DMA buffer half
full).
INT_ADCDATA – there is data ready in the ADC_DATA register.
.
Bits in INT_ADCFLAG register may be cleared by writing a 1 to their position. Writing 0 to any bit in the
INT_ADCFLAG register is ineffectual.
The INT_ADCCFG register controls whether or not INT_ADCFLAG register bits actually propagate the ADC
interrupt to the NVIC. Only the events whose bits are 1 in the INT_ADCCFG register can do so.
For non-interrupt (polled) ADC operation set the INT_ADCCFG register to zero, and read the bit flags in the
INT_ADCFLAG register to determine the ADC status.
Note: When making changes to the ADC configuration it is best to disable the DMA beforehand. If this isn’t
done it can be difficult to determine at which point the sampled data in the DMA buffer switched from the old
configuration to the new configuration. However, since the ADC will be left running, if it completes a
conversion after the DMA is disabled, the INT_ADCOVF flag will be set. To prevent these unwanted DMA buffer
overflow indications, clear the INT_ADCOVF flag immediately after enabling the DMA, preferably with
interrupts off. Disabling the ADC in addition to the DMA is often undesirable because of the additional analog
startup time when it is re-enabled.
10.3
Operation
Setting the ADC_EN bit in the ADC_CFG register enables the ADC. Once the ADC is enabled, it performs
conversions continuously until it is disabled. If the ADC had previously been disabled, a 21 µs analog startup
delay is automatically imposed before the ADC starts conversions. The delay timing is performed in hardware
and is simply added to the time until the first conversion result is output.
When the ADC is first enabled, and/or if any change is made to ADC_CFG after it is enabled, the time until a
result is output is double the normal sample time. This is because the ADC’s internal design requires it to
discard the first conversion after startup or a configuration change. This is done automatically and is hidden
from software. Switching the system clock between OSCHF and OSC24M also causes the ADC to go through this
startup cycle. If the ADC was newly enabled, the analog delay time is added to the doubled sample time.
If the DMA is running when the ADC_CFG register is modified, the DMA does not stop, so the DMA buffer may
contain conversion results from both the old and new configurations.
The following procedure illustrates a simple polled method of using the ADC without DMA. This assumes that
any GPIOs and the voltage reference have already been configured.
1. Disable all ADC interrupts: Write 0 to the INT_ADCCFG register.
2. Write the desired offset and gain correction values to the ADC_OFFSET and ADC_GAIN registers.
3. Write the desired conversion configuration, with the ADC_EN bit set, to ADC_CFG register.
4. Clear the ADC data flag: Write the INT_ADCDATA bit to INT_ADCFLAG register.
5. Wait until the INT_ADCDATA bit is set in INT_ADCFLAG register, then read the result, as a 16-bit signed
variable, from the ADC_DATA register.
The following procedure illustrates a simple polled method of using the ADC with DMA. After completing the
procedure, the latest conversion results are available in the location written to by the DMA. This assumes that
any GPIOs and the voltage reference have already been configured.
1. Allocate a 16-bit signed variable, for example analogData, to receive the ADC output.
(Make sure that analogData is half-word aligned – that is, at an even address.)
2. Disable all ADC interrupts: Write 0 to the INT_ADCCFG register.
10-6
120-035X-000 Rev. 1.2
Final
EM351 / EM357
3. Set up the DMA to output conversion results to the variable, analogData.
Reset the DMA: Set the ADC_DMARST bit in ADC_DMACFG register.
Define a one sample buffer: Write analogData’s address to the ADC_DMABEG register and set the
ADC_DMASIZE register to 1.
4. Write the desired offset and gain correction values to the ADC_OFFSET and ADC_GAIN registers.
5. Start the ADC and the DMA.
Write the desired conversion configuration, with the ADC_EN bit set, to the ADC_CFG register.
Clear the ADC buffer full flag: Write the INT_ADCULDFULL bit to the INT_ADCFLAG register.
Start the DMA in auto wrap mode: Set the ADC_DMAAUTOWRAP and ADC_DMALOAD bits in the
ADC_DMACFG register.
6. Wait until the INT_ADCULDFULL bit is set in the INT_ADCFLAG register, then read the result from
analogData.
To convert multiple inputs using this approach, repeat steps 4 through 6, loading the desired input
configurations to the ADC_CFG register in step 5. If the inputs can use the same offset/gain correction, just
repeat steps 5 and 6.
10.4
Calibration
Sampling of internal connections GND, VREF/2, and VREF allow for offset and gain calibration of the ADC in
applications where absolute accuracy is important. Offset error is calculated from the minimum input and gain
error is calculated from the full scale input range. Correction using VREF is recommended because VREF is
calibrated by the Ember software against VDD_PADSA. The VDD_PADSA regulator is factory-trimmed to 1.80 V
± 20 mV. If better absolute accuracy is required, the ADC can be configured to use an external reference. The
ADC calibrates as a single-ended measurement. Differential signals require correction of both their inputs.
The following steps outline the calibration procedure
Calibrate VREF against VDD_PADSA.
.
.
Determine the ADC gain by sampling independently VREF and GND. Gain is calculated from the slope of
these two measurements.
Apply gain correction.
.
Determine the ADC offset by sampling GND.
.
Apply offset correction.
.
Table 10-5 shows the equations used to calculate the gain and offset correction values.
Table 10-5. ADC Gain and offset correction equations
Calibration
Correction value
Gain
16384
(NVREF − NGND
32768×
)
Offset ( after applying gain correction )
2×(57344 − NGND
)
Equation notes
The ADC output is two’s complement. All N are therefore 16-bit two’s complement numbers.
Offset is a 16-bit two’s complement number.
.
.
.
Gain is a 16-bit number representing a gain of 0 to 65535/32768 in 1/32768 steps. The default value is
32768, corresponding to a gain of 1.
NGND is a sampling of ground. Due to the ADC's internal design, VGND does not yield the minimum 16 bit
two’s complement value 32768 as the conversion result. Instead, VGND yields a value close to 57344 when
.
10-7
120-035X-000 Rev. 1.2
Final
EM351 / EM357
the input buffer is not selected. VGND cannot be measured when the input buffer is enabled because it is
outside the buffer’s input range.
NVREF is a sampling of VREF. Due to the ADC's internal design, VREF does not yield the maximum positive
16-bit two’s complement 32767 as the conversion result. Instead, VREF yields a value close to 8192.
.
NVREF/2 is a sampling of VREF/2. VREF/2 yields a value close to 0.
.
.
Offset correction is affected by the gain correction value. Offset correction is calculated after gain
correction has been applied.
10.5
ADC Key Parameters
Table 10-6 describes the key ADC parameters measured at 25°C and VDD_PADS at 3.0 V, for a sampling clock
of 1 MHz. The single-ended measurements were done at finput = 7.7% fNyquist; 0 dBFS level (where full-scale is a
1.2 V p-p swing). The differential measurements were done at finput = 7.7% fNyquist; -6 dBFS level (where full-
scale is a 2.4 V p-p swing) and a common mode voltage of 0.6 V.
Table 10-6. ADC Module Key Parameters for 1 MHz sampling
Parameter
Performance
ADC_PERIOD
0
1
2
3
4
5
6
7
Conversion Time (µs)
Nyquist Freq (kHz)
3 dB Cut-off (kHz)
INL (codes peak)
32
64
128
256
512
977
1024
488
2048
244
4096
122
15.6k
7.81k
3.91k
1.95k
9.43k
0.083
4.71k
0.092
2.36k
0.163
1.18k
0.306
589
295
147
73.7
0.624
1.229
2.451
4.926
0.047
0.028
0.051
0.035
0.093
0.038
0.176
0.044
0.362
0.074
0.719
0.113
1.435
0.184
2.848
0.333
INL (codes RMS)
DNL (codes peak)
0.008
5.6
0.009
7.0
0.011
8.6
0.014
10.1
0.019
11.5
0.029
12.6
0.048
13.0
0.079
13.2
DNL (codes RMS)
ENOB
(from single-cycle test)
10-8
Final
120-035X-000 Rev. 1.2
EM351 / EM357
Parameter
Performance
SNR (dB)
35
35
44
44
53
53
62
62
70
71
75
77
77
79
77
80
Single-Ended
Differential
SINAD (dB)
35
35
44
44
53
53
61
62
67
70
69
75
70
76
70
76
Single-Ended
Differential
SDFR (dB)
59
60
68
69
72
77
72
80
72
81
72
81
72
81
73
81
Single-Ended
Differential
THD (dB)
-45
-45
-54
-54
-62
-63
-67
-71
-69
-75
-69
-76
-69
-76
-69
-76
Single-Ended
Differential
ENOB (from SNR)
Single-Ended
Differential
5.6
5.6
7.1
7.1
8.6
8.6
10.0
10.1
11.3
11.4
12.2
12.5
12.4
12.9
12.5
12.9
ENOB (from SINAD)
Single-Ended
5.5
5.6
7.0
7.0
8.5
8.5
9.9
10.9
11.3
11.2
12.1
11.3
12.3
11.3
12.4
Differential
10.0
Equivalent ADC Bits
7
8
9
10
11
12
13
14
[15:9]
[15:8]
[15:7]
[15:6]
[15:5]
[15:4]
[15:3]
[15:2]
Note: INL and DNL are referenced to a LSB of the Equivalent ADC Bits shown in the last row of Table 10-6.
ENOB (effective number of bits) can be calculated from either SNR (signal to non-harmonic noise ratio) or
SINAD (signal-to-noise and distortion ratio).
Table 10-7 describes the key ADC parameters measured at 25°C and VDD_PADS at 3.0 V, for a sampling rate of
6 MHz. The single-ended measurements were done at finput = 7.7% fNyquist; 0 dBFS level (where full-scale is a
1.2 V p-p swing). The differential measurements were done at finput = 7.7% fNyquist; -6 dBFS level (where full-
scale is a 2.4 V p-p swing) and a common mode voltage of 0.6 V.
Table 10-7. ADC Module Key Parameters for 6 MHz sampling
Parameter
Performance
ADC_PERIOD
0
1
2
3
4
5
6
7
Conversion Time (µs)
Nyquist Freq (kHz)
3 dB Cut-off (kHz)
INL (codes peak)
5.33
93.8k
10.7
46.9k
21.3
23.4k
42.7
11.7k
85.3
5.86k
171
341
683
732
2.93k
1.47k
56.6k
0.084
28.3k
0.084
14.1k
0.15
7.07k
0.274
3.54k
0.518
1.77k
1.057
884
442
2.106
4.174
0.046
0.026
0.044
0.023
0.076
0.044
0.147
0.052
0.292
0.096
0.58
1.14
2.352
0.371
INL (codes RMS)
DNL (codes peak)
0.119
0.196
10-9
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Parameter
Performance
0.007
5.6
0.009
7.0
0.013
8.5
0.015
0.024
0.03
12.6
0.05
13.1
0.082
13.2
DNL (codes RMS)
10.0
11.4
ENOB
(from single-cycle test)
SNR (dB)
Single-Ended
Differential
35
35
44
44
53
53
62
62
70
71
75
77
76
79
77
80
SINAD (dB)
Single-Ended
Differential
35
35
44
44
53
53
62
62
68
70
71
75
71
77
71
77
SDFR (dB)
Single-Ended
Differential
60
60
68
69
75
77
75
80
75
80
75
80
75
80
75
80
THD (dB)
Single-Ended
Differential
-45
-45
-54
-54
-63
-63
-68
-71
-70
-76
-70
-77
-70
-78
-70
-78
ENOB (from SNR)
Single-Ended
Differential
5.6
5.6
7.1
7.1
8.6
8.6
10.0
10.1
11.4
11.5
12.1
12.5
12.4
12.9
12.5
13.0
ENOB (from SINAD)
Single-Ended
5.5
5.6
7.0
7.1
8.5
8.6
9.9
11.0
11.4
11.4
12.4
11.5
12.8
11.5
13.0
Differential
10.1
Equivalent ADC Bits
7
8
9
10
11
12
13
14
[15:9]
[15:8]
[15:7]
[15:6]
[15:5]
[15:4]
[15:3]
[15:2]
Note: INL and DNL are referenced to a LSB of the Equivalent ADC Bits shown in the last row of Table 10-7.
ENOB (effective number of bits) can be calculated from either SNR (signal to non-harmonic noise ratio) or
SINAD (signal-to-noise and distortion ratio).
Table 10-8 describes the key ADC parameters measured at 25°C and VDD_PADS at 3.0 V, for a sampling clock
of 6 MHz. The single-ended measurements were done at finput = 7.7% fNyquist; level = 1.2 V p-p swing centered
on 1.5 V. The differential measurements were done at finput = 7.7% fNyquist, level = 1.2 V p-p swing and a
common mode voltage of 1.5 V.
Table 10-8. ADC Module Key Parameters for input buffer enabled and 6 MHz sampling
Parameter
Performance
ADC_PERIOD
0
1
2
3
4
5
6
7
Conversion Time (µs)
Nyquist Freq (kHz)
3 dB Cut-off (kHz)
32
64
128
256
512
1024
2.93k
1.77k
2048
1.47k
884
4096
732
442
93.8k
56.6k
46.9k
28.3k
23.4k
14.1k
11.7k
7.07k
5.86k
3.54k
10-10
Final
120-035X-000 Rev. 1.2
EM351 / EM357
Parameter
Performance
0.055
0.028
0.028
0.01
0.032
0.017
0.017
0.006
5.0
0.038
0.02
0.02
0.006
6.6
0.07
0.04
0.04
0.007
8.1
0.123
0.261
0.167
0.167
0.013
10.7
0.522
0.326
0.326
0.023
11.3
1.028
0.65
INL (codes peak)
0.077
0.077
0.008
9.5
INL (codes RMS)
DNL (codes peak)
DNL (codes RMS)
0.65
0.038
11.6
3.6
ENOB
(from single-cycle test)
SNR (dB)
Single-Ended
Differential
23
23
32
32
41
41
50
50
59
59
65
66
67
69
68
71
SINAD (dB)
Single-Ended
Differential
23
23
32
32
41
41
50
50
58
59
64
66
66
69
66
71
SDFR (dB)
Single-Ended
Differential
48
48
56
57
65
65
72
74
72
82
72
88
73
88
73
88
THD (dB)
Single-Ended
Differential
-33
-33
-42
-42
-51
-51
-59
-60
-66
-69
-68
-76
-68
-80
-68
-82
ENOB (from SNR)
Single-Ended
Differential
3.6
3.6
5.1
5.1
6.6
6.6
8.1
8.1
9.5
9.5
10.5
10.7
10.9
11.3
11
11.5
ENOB (from SINAD)
Single-Ended
3.6
3.6
5.0
5.1
6.5
6.6
8.0
8.0
9.4
9.5
10.3
10.6
10.7
11.3
10.7
11.4
Differential
Equivalent ADC Bits
7
8
9
10
11
12
13
14
[15:9]
[15:8]
[15:7]
[15:6]
[15:5]
[15:4]
[15:3]
[15:2]
INL and DNL are referenced to a LSB of the Equivalent ADC Bits shown in the last row of Table 10-6. ENOB
(effective number of bits) can be calculated from either SNR (signal to non-harmonic noise ratio) or SINAD
(signal-to-noise and distortion ratio).
Table 10-9 lists other specifications for the ADC not covered in Table 10-6 and Table 10-7.
Table 10-9. ADC Specifications
Parameter
Min.
Typ.
Max.
Units
VREF
1.17
1.2
1.23
V
VREF output current
1
mA
10-11
Final
120-035X-000 Rev. 1.2
EM351 / EM357
Parameter
Min.
Typ.
Max.
Units
VREF load capacitance
10
nF
External VREF voltage range
External VREF input impedance
Minimum input voltage
1.1
1
1.2
1.3
V
MΩ
V
0
Maximum input voltage
Single-ended signal range
Differential signal range
Common mode range
VREF
VREF
+VREF
VREF
10
V
0
-VREF
0
V
V
V
Input referred ADC offset
-10
mV
Input Impedance
1 MHz sample clock
6 MHz sample clock
Not sampling
1
MΩ
0.5
10
Note: The signal-ended ADC measurements are limited in their range and only guaranteed for accuracy within
the limits shown in this table. The ADC's internal design allows for measurements outside of this range
(±200 mV), but the accuracy of such measurements is not guaranteed. The maximum input voltage is of more
interest to the differential sampling where a differential measurement might be small, but a common mode
can push the actual input voltage on one of the signals towards the upper voltage limit.
10-12
Final
120-035X-000 Rev. 1.2
EM351 / EM357
10.6
Registers
ADC_DATA
ADC Data Register
Address: 0x4000D000 Reset: 0x00000000
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
ADC_DATA_FIELD
ADC_DATA_FIELD
7
6
5
4
3
2
1
0
Bitname
Bitfield
Access
Description
ADC_DATA_FIELD
[15:0]
R
ADC conversion result. The result is a signed 2’s complement value. The significant bits
of the value begin at bit 15 regardless of the sample period used.
10-13
Final
120-035X-000 Rev. 1.2
EM351 / EM357
ADC_CFG
ADC Configuration Register
Address: 0x4000D004 Reset: 0x00001800
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
ADC_PERIOD
ADC_MUXP
ADC_CFGRSVD2
7
6
5
4
3
2
1
0
ADC_MUXP
ADC_MUXN
ADC_1MHZCLK
ADC_CFGRSVD
ADC_ENABLE
Bitname
Bitfield
Access
Description
ADC_PERIOD
[15:13]
RW
ADC sample time in clocks and the equivalent significant bits in the conversion.
0: 32 clocks (7 bits).
1: 64 clocks (8 bits).
2: 128 clocks (9 bits).
3: 256 clocks (10 bits).
4: 512 clocks (11 bits).
5: 1024 clocks (12 bits).
6: 2048 clocks (13 bits).
7: 4096 clocks (14 bits).
Reserved: these bits must be set to 0.
ADC_CFGRSVD2
ADC_MUXP
[12:11]
[10:7]
RW
RW
Input selection for the P channel.
0x0: PB5 pin.
0x1: PB6 pin.
0x2: PB7 pin.
0x3: PC1 pin.
0x4: PA4 pin.
0x5: PA5 pin.
0x8: GND (0V) (not for high voltage range).
0x9: VREF/2 (0.6V).
0xA: VREF (1.2V).
0xB: VDD_PADSA/2 (0.9V) (not for high voltage range).
0x6, 0x7, 0xC-0xF: reserved.
ADC_MUXN
[6:3]
RW
Input selection for the N channel.
Refer to ADC_MUXP above for choices.
ADC_1MHZCLK
ADC_CFGRSVD
ADC_ENABLE
[2]
[1]
[0]
RW
RW
RW
Select ADC clock: 0 = 6 MHz, 1 = 1 MHz.
Reserved: this bit must always be set to 0.
Enable the ADC: write 1 to enable continuous conversions, write 0 to stop.
When the ADC is started the first conversion takes twice the usual number of clocks plus
21 microseconds. If anything in this register is modified while the ADC is running, the next
conversion takes twice the usual number of clocks.
10-14
Final
120-035X-000 Rev. 1.2
EM351 / EM357
ADC_OFFSET
ADC Offset Register
Address: 0x4000D008 Reset: 0x0000
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
ADC_OFFSET_FIELD
7
6
5
4
3
2
1
0
ADC_OFFSET_FIELD
Bitname
Bitfield
Access
RW
Description
ADC_OFFSET_FIELD
[15:0]
16-bit signed offset added to the basic ADC conversion result before gain correction is
applied.
10-15
Final
120-035X-000 Rev. 1.2
EM351 / EM357
ADC_GAIN
ADC Gain Register
Address: 0x4000D00C Reset: 0x8000
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
ADC_GAIN_FIELD
ADC_GAIN_FIELD
7
6
5
4
3
2
1
0
Bitname
Bitfield
Access
RW
Description
ADC_GAIN_FIELD
[15:0]
Gain factor that is multiplied by the offset-corrected ADC result to produce the output
value. The gain is a 16-bit unsigned scaled integer value with a binary decimal point
between bits 15 and 14. It can represent values from 0 to (almost) 2. The reset value is a
gain factor of 1.
10-16
Final
120-035X-000 Rev. 1.2
EM351 / EM357
ADC_DMACFG
ADC DMA Configuration Register
Address: 0x4000D010 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
ADC_DMARST
0
0
ADC_DMAAUTOWRA
P
ADC_DMALOAD
Bitname
Bitfield
[4]
Access
Description
ADC_DMARST
W
Write 1 to reset the ADC DMA. This bit auto-clears.
ADC_DMAAUTOWRAP
[1]
RW
Selects DMA mode.
0: Linear mode, the DMA stops when the buffer is full.
1: Auto-wrap mode, the DMA output wraps back to the start when the buffer is full.
ADC_DMALOAD
[0]
RW
Loads the DMA buffer.
Write 1 to start DMA (writing 0 has no effect). Cleared when DMA starts or is reset.
10-17
Final
120-035X-000 Rev. 1.2
EM351 / EM357
ADC_DMASTAT
ADC DMA Status Register
Address: 0x4000D014 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
0
0
ADC_DMAOVF
ADC_DMAACT
Bitname
Bitfield
Access
Description
ADC_DMAOVF
[1]
R
DMA overflow: occurs when an ADC result is ready and the DMA is not active. Cleared by
DMA reset.
ADC_DMAACT
[0]
R
DMA status: reads 1 if DMA is active.
10-18
Final
120-035X-000 Rev. 1.2
EM351 / EM357
ADC_DMABEG
ADC DMA Begin Address Register
Address: 0x4000D018 Reset: 0x20000000
31
0
30
0
29
1
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
ADC_DMABEG
7
6
5
4
3
2
1
0
ADC_DMABEG
Bitname
ADC_DMABEG
Bitfield
Access
RW
Description
[13:0]
ADC buffer start address. Caution: this must be an even address - the least significant bit
of this register is fixed at zero by hardware.
10-19
Final
120-035X-000 Rev. 1.2
EM351 / EM357
ADC_DMASIZE
ADC DMA Buffer Size Register
Address: 0x4000D01C Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
ADC_DMASIZE_FIELD
7
6
5
4
3
2
1
0
ADC_DMASIZE_FIELD
Bitname
ADC_DMASIZE_FIELD
Bitfield
Access
RW
Description
[12:0]
ADC buffer size. This is the number of 16-bit ADC conversion results the buffer can hold,
not its length in bytes. (The length in bytes is twice this value.)
10-20
Final
120-035X-000 Rev. 1.2
EM351 / EM357
ADC_DMACUR
ADC DMA Current Address Register
Address: 0x4000D020 Reset: 0x20000000
31
0
30
0
29
1
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
ADC_DMACUR_FIELD
7
6
5
4
3
2
1
0
ADC_DMACUR_FIELD
0
Bitname
ADC_DMACUR_FIELD
Bitfield
Access
Description
[13:1]
R
Current DMA address: the location that will be written next by the DMA.
10-21
Final
120-035X-000 Rev. 1.2
EM351 / EM357
ADC_DMACNT
ADC DMA Count Register
Address: 0x4000D024 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
ADC_DMACNT_FIELD
7
6
5
4
3
2
1
0
ADC_DMACNT_FIELD
Bitname
ADC_DMACNT_FIELD
Bitfield
Access
Description
DMA count: the number of 16-bit conversion results that have been written to the buffer.
[12:0]
R
10-22
Final
120-035X-000 Rev. 1.2
EM351 / EM357
INT_ADCFLAG
ADC Interrupt Flag Register
Address: 0x4000A810 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
INT_ADCOVF
INT_ADCSAT
INT_ADCULDFULL
INT_ADCULDHALF
INT_ADCFLAGRSVD
Bitname
Bitfield
[4]
Access
Description
INT_ADCOVF
RW
RW
RW
RW
RW
DMA buffer overflow interrupt pending.
Gain correction saturation interrupt pending.
DMA buffer full interrupt pending.
INT_ADCSAT
[3]
INT_ADCULDFULL
INT_ADCULDHALF
INT_ADCDATA
[2]
[1]
DMA buffer half full interrupt pending.
ADC_DATA register has data interrupt pending.
[0]
10-23
Final
120-035X-000 Rev. 1.2
EM351 / EM357
INT_ADCCFG
ADC Interrupt Configuration Register
Address: 0x4000A850 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
INT_ADCOVF
INT_ADCSAT
INT_ADCULDFULL
INT_ADCULDHALF
INT_ADCCFGRSVD
Bitname
Bitfield
[4]
Access
Description
INT_ADCOVF
RW
RW
RW
RW
RW
DMA buffer overflow interrupt enable.
Gain correction saturation interrupt enable.
DMA buffer full interrupt enable.
INT_ADCSAT
[3]
INT_ADCULDFULL
INT_ADCULDHALF
INT_ADCDATA
[2]
[1]
DMA buffer half full interrupt enable.
ADC_DATA register has data interrupt enable.
[0]
10-24
Final
120-035X-000 Rev. 1.2
EM351 / EM357
11 Interrupt System
The EM35x’s interrupt system is composed of two parts: a standard ARM® CortexTM-M3 Nested Vectored
Interrupt Controller (NVIC) that provides top-level interrupts, and a proprietary Event Manager (EM) that
provides second-level interrupts. The NVIC and EM provide a simple hierarchy. All second-level interrupts from
the EM feed into top-level interrupts in the NVIC. This two-level hierarchy allows for both fine granular control
of interrupt sources and coarse granular control over entire peripherals, while allowing peripherals to have
their own interrupt vector.
The Nested Vectored Interrupt Controller (NVIC) section provides a description of the NVIC and an overview of
the exception table (ARM nomenclature refers to interrupts as exceptions). The Event Manager section
provides a more detailed description of the Event Manager including a table of all top-level peripheral
interrupts and their second-level interrupt sources.
In practice, top-level peripheral interrupts are only used to enable or disable interrupts for an entire
peripheral. Second-level interrupts originate from hardware sources, and therefore are the main focus of
applications using interrupts.
11.1 Nested Vectored Interrupt Controller (NVIC)
The ARM® CortexTM-M3 Nested Vectored Interrupt Controller (NVIC) facilitates low-latency exception and
interrupt handling. The NVIC and the processor core interface are closely coupled, which enables low-latency
interrupt processing and efficient processing of late-arriving interrupts. The NVIC also maintains knowledge of
the stacked (nested) interrupts to enable tail-chaining of interrupts.
The ARM® CortexTM-M3 NVIC contains 10 standard interrupts that are related to chip and CPU operation and
management. In addition to the 10 standard interrupts, it contains 17 individually vectored peripheral
interrupts specific to the EM35x.
The NVIC defines a list of exceptions. These exceptions include not only traditional peripheral interrupts, but
also more specialized events such as faults and CPU reset. In the ARM® CortexTM-M3 NVIC, a CPU reset event is
considered an exception of the highest priority, and the stack pointer is loaded from the first position in the
NVIC exception table. The NVIC exception table defines all exceptions and their position, including peripheral
interrupts. The position of each exception is important since it directly translates to the location of a 32-bit
interrupt vector for each interrupt, and defines the hardware priority of exceptions. Each exception in the
table is a 32-bit address that is loaded into the program counter when that exception occurs. Table 11-1 lists
the entire exception table. Exceptions 0 (stack pointer) through 15 (SysTick) are part of the standard ARM®
CortexTM-M3 NVIC, while exceptions 16 (Timer 1) through 32 (Debug) are the peripheral interrupts specific to
the EM35x peripherals. The peripheral interrupts are listed in greater detail in Table 11-2.
11-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 11-1. NVIC Exception Table
Description
Exception
Position
-
0
1
Stack top is loaded from first entry of vector table on reset.
Reset
Invoked on power up and warm reset. On first instruction, drops to
lowest priority (Thread mode). Asynchronous.
NMI
2
3
Cannot be stopped or preempted by any exception but reset.
Asynchronous.
Hard Fault
All classes of fault, when the fault cannot activate because of priority
or the Configurable Fault handler has been disabled. Synchronous.
Memory Fault
Bus Fault
4
5
MPU mismatch, including access violation and no match. Synchronous.
Pre-fetch, memory access, and other address/memory-related faults.
Synchronous when precise and asynchronous when imprecise.
Usage Fault
6
Usage fault, such as ‘undefined instruction executed’ or ‘illegal state
transition attempt’. Synchronous.
-
7-10
11
Reserved.
SVCall
System service call with SVC instruction. Synchronous.
Debug Monitor
12
Debug monitor, when not halting. Synchronous, but only active when
enabled. It does not activate if lower priority than the current
activation.
-
13
14
Reserved.
PendSV
Pendable request for system service. Asynchronous and only pended by
software.
SysTick
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
System tick timer has fired. Asynchronous.
Timer 1 peripheral interrupt.
Timer 2 peripheral interrupt.
Management peripheral interrupt.
Baseband peripheral interrupt.
Sleep Timer peripheral interrupt.
Serial Controller 1 peripheral interrupt.
Serial Controller 2 peripheral interrupt.
Security peripheral interrupt.
MAC Timer peripheral interrupt.
MAC Transmit peripheral interrupt.
MAC Receive peripheral interrupt.
ADC peripheral interrupt.
Timer 1
Timer 2
Management
Baseband
Sleep Timer
Serial Controller 1
Serial Controller 2
Security
MAC Timer
MAC Transmit
MAC Receive
ADC
IRQA
IRQA peripheral interrupt.
IRQB
IRQB peripheral interrupt.
IRQC
IRQC peripheral interrupt.
IRQD
IRQD peripheral interrupt.
Debug
Debug peripheral interrupt.
11-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
The NVIC also contains a software-configurable interrupt prioritization mechanism. The Reset, NMI, and Hard
Fault exceptions, in that order, are always the highest priority, and are not software-configurable. All other
exceptions can be assigned a 5-bit priority number, with low values representing higher priority. If any
exceptions have the same software-configurable priority, then the NVIC uses the hardware-defined priority.
The hardware-defined priority number is the same as the position of the exception in the exception table. For
example, if IRQA and IRQB both fire at the same time and have the same software-defined priority, the NVIC
handles IRQA, with priority number 28, first because it has a higher hardware priority than IRQB with priority
number 29.
The top-level interrupts are controlled through five ARM® CortexTM-M3NVIC registers: INT_CFGSET,
INT_CFGCLR, INT_PENDSET, INT_PENDCLR, and INT_ACTIVE. Writing 0 into any bit in any of these five register
is ineffective.
INT_CFGSET - Writing 1 to a bit in INT_CFGSET enables that top-level interrupt.
.
INT_CFGCLR - Writing 1 to a bit in INT_CFGCLR disables that top-level interrupt.
.
INT_PENDSET - Writing 1 to a bit in INT_PENDSET triggers that top-level interrupt.
.
INT_PENDCLR - Writing 1 to a bit in INT_PENDCLR clears that top-level interrupt.
.
INT_ACTIVE cannot be written to and is used for indicating which interrupts are currently active.
.
INT_PENDSET and INT_PENDCLR set and clear a simple latch; INT_CFGSET and INT_CFGCLR set and clear a
mask on the output of the latch. Interrupts may be pended and cleared at any time, but any pended interrupt
will not be taken unless the corresponding mask (INT_CFGSET) is set, which allows that interrupt to
propagate. If an INT_CFGSET bit is set and the corresponding INT_PENDSET bit is set, then the interrupt will
propagate and be taken. If INT_CFGSET is set after INT_PENDSET is set, then the interrupt will also propagate
and be taken. Interrupt flags (signals) from the top-level interrupts are level-sensitive.
The second-level interrupt registers, which provide control of the second-level Event Manager peripheral
interrupts, are described in the Event Manager section.
For further information on the NVIC and ARM® CortexTM-M3 exceptions, refer to the ARM® CortexTM-M3
Technical Reference Manual and the ARM ARMv7-M Architecture Reference Manual.
11.2 Event Manager
While the standard ARM® CortexTM-M3 Nested Vectored Interrupt Controller provides top-level interrupts into
the CPU, the proprietary Event Manager provides second-level interrupts. The Event Manager takes a large
variety of hardware interrupt sources from the peripherals and merges them into a smaller group of interrupts
in the NVIC. Effectively, all second-level interrupts from a peripheral are “OR’d” together into a single
interrupt in the NVIC. In addition, the Event Manager provides missed indicators for the top-level peripheral
interrupts with the register INT_MISS.
The description of each peripheral’s interrupt configuration and flag registers can be found in the chapters of
this datasheet describing each peripheral. Figure 11-1 shows the Peripheral Interrupts Block Diagram.
11-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 11-1. Peripheral Interrupts Block Diagram
interrupts into NVIC/CPU
AND
peripheral interrupt instance
read
Q
latch
S
R
OR
OR
write 1
write 1
INT_CFGCLR
INT_CFGSET
INT_periphCFG
AND
read
Q
latch
AND
S
R
read
write 1 INT_PENDCLR
write 1 INT_PENDSET
OR
Q
latch
INT_periphFLAG
S
R
read
Q
latch
write 1
S
R
write 1
INT_MISS
source interrupt events
interrupts from all peripherals
Given a peripheral, ‘periph’, the Event Manager registers (INT_periphCFG and INT_periphFLAG) follow the
form:
INT_periphCFG enables and disables second-level interrupts. Writing 1 to a bit in the INT_periphCFG
.
register enables the second-level interrupt. Writing 0 to a bit in the INT_periphCFG register disables it.
The INT_periphCFG register behaves like a mask, and is responsible for allowing the INT_periphFLAG bits
to propagate into the top-level NVIC interrupts.
INT_periphFLAG indicates second-level interrupts that have occurred. Writing 1 to a bit in a
.
INT_periphFLAG register clears the second-level interrupt. Writing 0 to any bit in the INT_periphFLAG
register is ineffective. The INT_periphFLAG register is always active and may be set or cleared at any time,
meaning if any second-level interrupt occurs, then the corresponding bit in the INT_periphFLAG register is
set regardless of the state of INT_periphCFG.
If a bit in the INT_periphCFG register is set after the corresponding bit in the INT_periphFLAG register is set
then the second-level interrupt propagates into the top-level interrupts. The interrupt flags (signals) from the
second-level interrupts into the top-level interrupts are level-sensitive. If a top-level NVIC interrupt is driven
by a second-level EM interrupt, then the top-level NVIC interrupt cannot be cleared until all second-level EM
interrupts are cleared.
The INT_periphFLAG register bits are designed to remain set if the second-level interrupt event re-occurs at
the same moment as the INT_periphFLAG register bit is being cleared. This ensures the re-occurring second-
level interrupt event is not missed.
If another enabled second-level interrupt event of the same type occurs before the first interrupt event is
cleared, the second interrupt event is lost because no counting or queuing is used. However, this condition is
detected and stored in the top-level INT_MISS register to facilitate software detection of such problems. The
INT_MISS register is “acknowledged” in the same way as the INT_periphFLAG register—by writing a 1 into the
corresponding bit to be cleared.
Table 11-2 provides a map of all peripheral interrupts. This map lists the top-level NVIC Interrupt bits and, if
there is one, the corresponding second-level EM Interrupt register bits that feed the top-level interrupts.
11-4
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Table 11-2. NVIC and EM Peripheral Interrupt Map
NVIC Interrupt
(top-level)
EM Interrupt
(second-level)
NVIC Interrupt
(top-level)
EM Interrupt
(second-level)
16
15
14
13
12
11
INT_DEBUG
INT_IRQD
INT_IRQC
INT_IRQB
INT_IRQA
INT_ADC
5
INT_SC1
INT_SC1FLAG register
14
13
12
11
10
9
INT_SC1PARERR
INT_SC1FRMERR
INT_SCTXULDB
INT_SCTXULDA
INT_SCRXULDB
INT_SCRXULDA
INT_SCNAK
INT_ADCFLAG register
4
3
2
1
0
INT_ADCOVF
INT_ADCSAT
8
INT_ADCULDFULL
INT_ADCULDHALF
INT_ADCDATA
7
INT_SCCDMFIN
INT_SCTXFIN
INT_SCRXFIN
INT_SCTXUND
INT_SCRXOVF
INT_SCTXIDLE
INT_SCTXFREE
INT_SCRXVAL
6
5
10
9
INT_MACRX
INT_MACTX
INT_MACTMR
INT_SEC
4
3
8
2
7
1
6
INT_SC2
INT_SC2FLAG register
0
12
11
10
9
INT_SCTXULDB
INT_SCTXULDA
INT_SCRXULDB
INT_SCRXULDA
INT_SCNAK
4
3
2
1
INT_SLEEPTMR
INT_BB
INT_MGMT
INT_TMR2
INT_TMR2FLAG register
8
6
4
3
2
1
0
INT_TMRTIF
7
INT_SCCDMFIN
INT_SCTXFIN
INT_SCRXFIN
INT_SCTXUND
INT_SCRXOVF
INT_SCTXIDLE
INT_SCTXFREE
INT_SCRXVAL
INT_TMRCC4IF
INT_TMRCC3IF
INT_TMRCC2IF
INT_TMRCC1IF
INT_TMRUIF
6
5
4
3
2
0
INT_TMR1
INT_TMR1FLAG register
1
6
4
3
2
1
0
INT_TMRTIF
0
INT_TMRCC4IF
INT_TMRCC3IF
INT_TMRCC2IF
INT_TMRCC1IF
INT_TMRUIF
11-5
120-035X-000 Rev. 1.2
Final
EM351 / EM357
11.3 Non-maskable Interrupt (NMI)
The non-maskable interrupt (NMI) is a special case. Despite being one of the 10 standard ARM® CortexTM-
M3 NVIC interrupts, it is sourced from the Event Manager like a peripheral interrupt. The NMI has two second-
level sources; failure of the 24 MHz crystal and watchdog low water mark.
1. Failure of the 24MHz crystal: If the EM35x’s main clock, SYSCLK, is operating from the 24 MHz crystal and
the crystal fails, the EM35x detects the failure and automatically switches to the internal 12 MHz RC clock.
When this failure detection and switch has occurred, the EM35x triggers the CLK24M_FAIL second-level
interrupt, which then triggers the NMI.
2. Watchdog low water mark: If the EM35x’s watchdog is active and the watchdog counter has not been
reset for nominally 1.792 seconds, the watchdog triggers the WATCHDOG_INT second-level interrupt,
which then triggers the NMI.
11.4 Faults
Four of the exceptions in the NVIC are faults: Hard Fault, Memory Fault, Bus Fault, and Usage Fault. Of these,
three (Hard Fault, Memory Fault, and Usage Fault) are standard ARM® CortexTM-M3 exceptions.
The Bus Fault, though, is derived from EM35x-specific sources. The Bus Fault sources are recorded in the
SCS_AFSR register. Note that it is possible for one access to set multiple SCS_AFSR bits. Also note that MPU
configurations could prevent most of these bus fault accesses from occurring, with the advantage that illegal
writes are made precise faults. The four bus faults are:
WRONGSIZE – Generated by an 8-bit or 16-bit read or write of an APB peripheral register. This fault can
also result from an unaligned 32-bit access.
.
PROTECTED – Generated by a user mode (unprivileged) write to a system APB or AHB peripheral or
protected RAM (see Chapter 5, Section 5.2.2.3).
.
RESERVED – Generated by a read or write to an address within an APB peripheral’s 4 kB block range, but
the address is above the last physical register in that block range. Also generated by a read or write to an
.
address above the top of RAM or flash.
MISSED – Generated by a second SCS_AFSR fault. In practice, this bit is not seen since a second fault also
generates a hard fault, and the hard fault preempts the bus fault.
.
11-6
120-035X-000 Rev. 1.2
Final
EM351 / EM357
11.5 Registers
INT_CFGSET
Top-Level Set Interrupts Configuration Register
Address: 0xE000E100 Reset: 0x0
31
30
29
28
27
26
25
24
0
0
0
0
0
0
0
0
23
22
21
20
19
18
17
16
INT_DEBUG
8
0
15
0
14
0
13
0
0
11
0
10
0
12
INT_IRQA
4
9
INT_IRQD
7
INT_IRQC
6
INT_IRQB
5
INT_ADC
3
INT_MACRX
2
INT_MACTX
1
INT_MACTMR
0
INT_SEC
INT_SC2
INT_SC1
INT_SLEEPTMR
INT_BB
INT_MGMT
INT_TIM2
INT_TIM1
Bitname
Bitfield
[16]
[15]
[14]
[13]
[12]
[11]
[10]
[9]
Access
Description
INT_DEBUG
INT_IRQD
INT_IRQC
INT_IRQB
INT_IRQA
INT_ADC
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
Write 1 to enable debug interrupt. (Writing 0 has no effect.)
Write 1 to enable IRQD interrupt. (Writing 0 has no effect.)
Write 1 to enable IRQC interrupt. (Writing 0 has no effect.)
Write 1 to enable IRQB interrupt. (Writing 0 has no effect.)
Write 1 to enable IRQA interrupt. (Writing 0 has no effect.)
Write 1 to enable ADC interrupt. (Writing 0 has no effect.)
Write 1 to enable MAC receive interrupt. (Writing 0 has no effect.)
Write 1 to enable MAC transmit interrupt. (Writing 0 has no effect.)
Write 1 to enable MAC timer interrupt. (Writing 0 has no effect.)
Write 1 to enable security interrupt. (Writing 0 has no effect.)
Write 1 to enable serial controller 2 interrupt. (Writing 0 has no effect.)
Write 1 to enable serial controller 1 interrupt. (Writing 0 has no effect.)
Write 1 to enable sleep timer interrupt. (Writing 0 has no effect.)
Write 1 to enable baseband interrupt. (Writing 0 has no effect.)
Write 1 to enable management interrupt. (Writing 0 has no effect.)
Write 1 to enable timer 2 interrupt. (Writing 0 has no effect.)
Write 1 to enable timer 1 interrupt. (Writing 0 has no effect.)
INT_MACRX
INT_MACTX
INT_MACTMR
INT_SEC
[8]
[7]
INT_SC2
[6]
INT_SC1
[5]
INT_SLEEPTMR
INT_BB
[4]
[3]
INT_MGMT
INT_TIM2
INT_TIM1
[2]
[1]
[0]
11-7
120-035X-000 Rev. 1.2
Final
EM351 / EM357
INT_CFGCLR
Top-Level Clear Interrupts Configuration Register
Address: 0xE000E180 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
16
23
0
22
0
21
0
20
0
19
0
18
0
17
0
INT_DEBUG
15
14
13
12
11
10
9
8
INT_IRQD
INT_IRQC
INT_IRQB
INT_IRQA
INT_ADC
INT_MACRX
INT_MACTX
INT_MACTMR
7
6
5
4
3
2
1
0
INT_SEC
INT_SC2
INT_SC1
INT_SLEEPTMR
INT_BB
INT_MGMT
INT_TIM2
INT_TIM1
Bitname
Bitfield
[16]
[15]
[14]
[13]
[12]
[11]
[10]
[9]
Access
Description
INT_DEBUG
INT_IRQD
INT_IRQC
INT_IRQB
INT_IRQA
INT_ADC
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
Write 1 to disable debug interrupt. (Writing 0 has no effect.)
Write 1 to disable IRQD interrupt. (Writing 0 has no effect.)
Write 1 to disable IRQC interrupt. (Writing 0 has no effect.)
Write 1 to disable IRQB interrupt. (Writing 0 has no effect.)
Write 1 to disable IRQA interrupt. (Writing 0 has no effect.)
Write 1 to disable ADC interrupt. (Writing 0 has no effect.)
Write 1 to disable MAC receive interrupt. (Writing 0 has no effect.)
Write 1 to disable MAC transmit interrupt. (Writing 0 has no effect.)
Write 1 to disable MAC timer interrupt. (Writing 0 has no effect.)
Write 1 to disable security interrupt. (Writing 0 has no effect.)
Write 1 to disable serial controller 2 interrupt. (Writing 0 has no effect.)
Write 1 to disable serial controller 1 interrupt. (Writing 0 has no effect.)
Write 1 to disable sleep timer interrupt. (Writing 0 has no effect.)
Write 1 to disable baseband interrupt. (Writing 0 has no effect.)
Write 1 to disable management interrupt. (Writing 0 has no effect.)
Write 1 to disable timer 2 interrupt. (Writing 0 has no effect.)
Write 1 to disable timer 1 interrupt. (Writing 0 has no effect.)
INT_MACRX
INT_MACTX
INT_MACTMR
INT_SEC
[8]
[7]
INT_SC2
[6]
INT_SC1
[5]
INT_SLEEPTMR
INT_BB
[4]
[3]
INT_MGMT
INT_TIM2
INT_TIM1
[2]
[1]
[0]
11-8
120-035X-000 Rev. 1.2
Final
EM351 / EM357
INT_PENDSET
Top-Level Set Interrupts Pending Register
Address: 0xE000E200 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
16
23
0
22
0
21
0
20
0
19
0
18
0
17
0
INT_DEBUG
15
14
13
12
11
10
9
8
INT_IRQD
INT_IRQC
INT_IRQB
INT_IRQA
INT_ADC
INT_MACRX
INT_MACTX
INT_MACTMR
7
6
5
4
3
2
1
0
INT_SEC
INT_SC2
INT_SC1
INT_SLEEPTMR
INT_BB
INT_MGMT
INT_TIM2
INT_TIM1
Bitname
Bitfield
[16]
[15]
[14]
[13]
[12]
[11]
[10]
[9]
Access
Description
INT_DEBUG
INT_IRQD
INT_IRQC
INT_IRQB
INT_IRQA
INT_ADC
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
Write 1 to pend debug interrupt. (Writing 0 has no effect.)
Write 1 to pend IRQD interrupt. (Writing 0 has no effect.)
Write 1 to pend IRQC interrupt. (Writing 0 has no effect.).
Write 1 to pend IRQB interrupt. (Writing 0 has no effect.)
Write 1 to pend IRQA interrupt. (Writing 0 has no effect.)
Write 1 to pend ADC interrupt. (Writing 0 has no effect.)
INT_MACRX
INT_MACTX
INT_MACTMR
INT_SEC
Write 1 to pend MAC receive interrupt. (Writing 0 has no effect.)
Write 1 to pend MAC transmit interrupt. (Writing 0 has no effect.)
Write 1 to pend MAC timer interrupt. (Writing 0 has no effect.)
Write 1 to pend security interrupt. (Writing 0 has no effect.)
Write 1 to pend serial controller 2 interrupt. (Writing 0 has no effect.)
Write 1 to pend serial controller 1 interrupt. (Writing 0 has no effect.)
Write 1 to pend sleep timer interrupt. (Writing 0 has no effect.)
Write 1 to pend baseband interrupt. (Writing 0 has no effect.)
Write 1 to pend management interrupt. (Writing 0 has no effect.)
Write 1 to pend timer 2 interrupt. (Writing 0 has no effect.)
Write 1 to pend timer 1 interrupt. (Writing 0 has no effect.)
[8]
[7]
INT_SC2
[6]
INT_SC1
[5]
INT_SLEEPTMR
INT_BB
[4]
[3]
INT_MGMT
INT_TIM2
INT_TIM1
[2]
[1]
[0]
11-9
120-035X-000 Rev. 1.2
Final
EM351 / EM357
INT_PENDCLR
Top-Level Clear Interrupts Pending Register
Address: 0xE000E280 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
16
23
0
22
0
21
0
20
0
19
0
18
0
17
0
INT_DEBUG
15
14
13
12
11
10
9
8
INT_IRQD
INT_IRQC
INT_IRQB
INT_IRQA
INT_ADC
INT_MACRX
INT_MACTX
INT_MACTMR
7
6
5
4
3
2
1
0
INT_SEC
INT_SC2
INT_SC1
INT_SLEEPTMR
INT_BB
INT_MGMT
INT_TIM2
INT_TIM1
Bitname
Bitfield
[16]
[15]
[14]
[13]
[12]
[11]
[10]
[9]
Access
Description
INT_DEBUG
INT_IRQD
INT_IRQC
INT_IRQB
INT_IRQA
INT_ADC
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
Write 1 to unpend debug interrupt. (Writing 0 has no effect.)
Write 1 to unpend IRQD interrupt. (Writing 0 has no effect.)
Write 1 to unpend IRQC interrupt. (Writing 0 has no effect.)
Write 1 to unpend IRQB interrupt. (Writing 0 has no effect.)
Write 1 to unpend IRQA interrupt. (Writing 0 has no effect.)
Write 1 to unpend ADC interrupt. (Writing 0 has no effect.)
Write 1 to unpend MAC receive interrupt. (Writing 0 has no effect.)
Write 1 to unpend MAC transmit interrupt. (Writing 0 has no effect.)
Write 1 to unpend MAC timer interrupt. (Writing 0 has no effect.)
Write 1 to unpend security interrupt. (Writing 0 has no effect.)
Write 1 to unpend serial controller 2 interrupt. (Writing 0 has no effect.)
Write 1 to unpend serial controller 1 interrupt. (Writing 0 has no effect.)
Write 1 to unpend sleep timer interrupt. (Writing 0 has no effect.)
Write 1 to unpend baseband interrupt. (Writing 0 has no effect.)
Write 1 to unpend management interrupt. (Writing 0 has no effect.)
Write 1 to unpend timer 2 interrupt. (Writing 0 has no effect.)
Write 1 to unpend timer 1 interrupt. (Writing 0 has no effect.)
INT_MACRX
INT_MACTX
INT_MACTMR
INT_SEC
[8]
[7]
INT_SC2
[6]
INT_SC1
[5]
INT_SLEEPTMR
INT_BB
[4]
[3]
INT_MGMT
INT_TIM2
INT_TIM1
[2]
[1]
[0]
11-10
Final
120-035X-000 Rev. 1.2
EM351 / EM357
INT_ACTIVE
Top-Level Active Interrupts Register
Address: 0xE000E300 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
16
23
0
22
0
21
0
20
0
19
0
18
0
17
0
INT_DEBUG
15
14
13
12
11
10
9
8
INT_IRQD
INT_IRQC
INT_IRQB
INT_IRQA
INT_ADC
INT_MACRX
INT_MACTX
INT_MACTMR
7
6
5
4
3
2
1
0
INT_SEC
INT_SC2
INT_SC1
INT_SLEEPTMR
INT_BB
INT_MGMT
INT_TIM2
INT_TIM1
Bitname
Bitfield
[16]
[15]
[14]
[13]
[12]
[11]
[10]
[9]
Access
Description
INT_DEBUG
INT_IRQD
INT_IRQC
INT_IRQB
INT_IRQA
INT_ADC
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
Debug interrupt active.
IRQD interrupt active.
IRQC interrupt active.
IRQB interrupt active.
IRQA interrupt active.
ADC interrupt active.
INT_MACRX
INT_MACTX
INT_MACTMR
INT_SEC
MAC receive interrupt active.
MAC transmit interrupt active.
MAC timer interrupt active.
Security interrupt active.
Serial controller 2 interrupt active.
Serial controller 1 interrupt active.
Sleep timer interrupt active.
Baseband interrupt active.
Management interrupt active.
Timer 2 interrupt active.
Timer 1 interrupt active.
[8]
[7]
INT_SC2
[6]
INT_SC1
[5]
INT_SLEEPTMR
INT_BB
[4]
[3]
INT_MGMT
INT_TIM2
INT_TIM1
[2]
[1]
[0]
11-11
Final
120-035X-000 Rev. 1.2
EM351 / EM357
INT_MISS
Top-Level Missed Interrupts Register
Address: 0x4000A820 Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
INT_MISSIRQD
INT_MISSIRQC
INT_MISSIRQB
INT_MISSIRQA
INT_MISSADC
INT_MISSMACRX
INT_MISSMACTX
INT_MISSMACTMR
7
6
5
4
3
2
1
0
INT_MISSSEC
INT_MISSSC2
INT_MISSSC1
INT_MISSSLEEP
INT_MISSBB
INT_MISSMGMT
0
0
Bitname
Bitfield
Access
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
Description
INT_MISSIRQD
INT_MISSIRQC
INT_MISSIRQB
INT_MISSIRQA
INT_MISSADC
INT_MISSMACRX
INT_MISSMACTX
INT_MISSMACTMR
INT_MISSSEC
[15]
[14]
[13]
[12]
[11]
[10]
[9]
IRQD interrupt missed.
IRQC interrupt missed.
IRQB interrupt missed.
IRQA interrupt missed.
ADC interrupt missed.
MAC receive interrupt missed.
MAC transmit interrupt missed.
MAC Timer interrupt missed.
Security interrupt missed.
Serial controller 2 interrupt missed.
Serial controller 1 interrupt missed.
Sleep timer interrupt missed.
Baseband interrupt missed.
Management interrupt missed.
[8]
[7]
INT_MISSSC2
[6]
INT_MISSSC1
[5]
INT_MISSSLEEP
INT_MISSBB
[4]
[3]
INT_MISSMGMT
[2]
11-12
Final
120-035X-000 Rev. 1.2
EM351 / EM357
SCS_AFSR
Auxiliary Fault Status Register
Address: 0xE000ED3C Reset: 0x0
31
0
30
0
29
0
28
0
27
0
26
0
25
0
24
0
23
0
22
0
21
0
20
0
19
0
18
0
17
0
16
0
15
14
13
12
11
10
9
8
0
0
0
0
0
0
0
0
7
6
5
4
3
2
1
0
0
0
0
0
WRONGSIZE
PROTECTED
RESERVED
MISSED
Bitname
Bitfield
Access
Description
WRONGSIZE
PROTECTED
RESERVED
[3]
RW
RW
RW
A bus fault resulted from an 8-bit or 16-bit read or write of an APB peripheral register.
This fault can also result from an unaligned 32-bit access.
[2]
[1]
A bus fault resulted from a user mode (unprivileged) write to a system APB or AHB
peripheral or protected RAM.
A bus fault resulted from a read or write to an address within an APB peripheral's 4 kB
block range, but above the last physical register in that block. Can also result from a read
or write to an address above the top of RAM or flash.
MISSED
[0]
RW
A bus fault occurred when a bit was already set in this register.
11-13
Final
120-035X-000 Rev. 1.2
EM351 / EM357
12 Trace Port Interface Unit (TPIU)
The EM35x integrates the standard ARM® Trace Port Interface Unit (TPIU). The TPIU receives a data stream
from the on-chip trace data generated by the standard ARM® Instrument Trace Macrocell (ITM), buffers the
data in a FIFO, formats the data, and serializes the data to be sent off chip through alternate functions of the
GPIO. Since the primary function of the TPIU is to provide a bridge between on-chip ARM system debug
components and external GPIO, the TPIU itself does not generate data. Figure 12-1 illustrates the three
primary components of the TPIU.
Figure 12-1. TPIU Block Diagram
SWO
TRACECLK
TRACEDATA0
TRACEDATA1
TRACEDATA2
TRACEDATA3
Asynchronous
FIFO
Trace Out
(serializer)
ITM
Formatter
The TPIU is composed of:
Asynchronous FIFO: The asynchronous FIFO receives a data stream generated by the ITM and enables the
trace data to be sent off chip at a speed that is not dependent on the speed of the data source.
.
.
Formatter: The formatter inserts source ID signals into the data packet stream so that trace data can be
re-associated with its trace source. Since the EM35x has only one trace source, the ITM, it is not necessary
to use the formatter and therefore the formatter only adds overhead into the data stream. Since certain
modes of the TPIU automatically enable the formatter, these modes should be avoided whenever possible.
Trace Out: The trace out block serializes the data and sends it off chip by the proper alternate output
GPIO functions.
.
The five pins available to the TPIU are:
SWO
.
TRACECLK
.
TRACEDATA0
.
TRACEDATA1
.
TRACEDATA2
.
TRACEDATA3
.
Since these pins are alternate outputs of GPIO, refer to Chapter 1, Pin Assignments and Chapter 7, GPIO for
complete pin descriptions and configurations.
Note: The SWO alternate output is mirrored on GPIO PC1 and PC2.
Note: GPIO PC1 shares both the SWO and TRACEDATA0 alternate outputs. This is possible because SWO and
TRACEDATA0 are mutually exclusive and only one may be selected at a time in the trace out block.
The Ember software utilizes the TPIU for efficiently outputting debug data. Altering the TPIU configuration
may conflict with Ember debug output.
For further information on the TPIU, contact Silicon Labs support for the ARM® CortexTM-M3 Technical
Reference Manual, the ARM® CoreSightTM Components Technical Reference Manual, the ARM® v7-M
Architecture Reference Manual, and the ARM® v7-M Architecture Application Level Reference Manual.
12-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
13 Instrumentation Trace Macrocell (ITM)
The EM35x integrates the standard ARM® Instrumentation Trace Macrocell (ITM). The ITM is an application-
driven trace source that supports printf style debugging to trace software events and emits diagnostic system
information from the ARM® Data Watchpoint and Trace (DWT). Software using the ITM generates Software
Instrumentation Trace (SWIT). In addition, the ITM provides coarse-grained timestamp functionality. The ITM
emits trace information as packets, and these packets are sent to the Trace Port Interface Unit (TPIU). Three
sources can generate packets. If multiple sources generate packets at the same time, the ITM arbitrates the
order in which the packets are output. The three sources, in decreasing order of priority, are:
Software trace. Software can write directly to ITM stimulus registers, emitting packets.
.
Hardware trace. The DWT generates packets that the ITM emits.
.
Time stamping. Timestamps are emitted relative to packets and the ITM contains a 21-bit counter to
generate the timestamps.
.
The Ember software utilizes the ITM for efficiently generating debug data. Altering the ITM configuration may
conflict with Ember debug output.
For further information on the ITM, contact Silicon Labs support for the ARM® CortexTM-M3 Technical
Reference Manual, the ARM® CoreSightTM Components Technical Reference Manual, the ARM® v7-M
Architecture Reference Manual, and the ARM® v7-M Architecture Application Level Reference Manual.
13-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
14 Data Watchpoint and Trace (DWT)
The EM35x integrates the standard ARM® Data Watchpoint and Trace (DWT). The DWT provides hardware
support for profiling and debugging functionality. The DWT offers the following features:
PC sampling
.
.
Comparators to support:
Watchpoints – enters debug state
•
Data tracing
•
Cycle count matched PC sampling
•
Exception trace support
.
.
Instruction cycle count calculation support
Apart from exception tracing, DWT functionality is counter- or comparator-based. Watchpoint and data trace
support use a set of compare, mask, and function registers. DWT-generated events result in one of two
actions:
Generation of a hardware event packet. Packets are generated and combined with software events and
timestamp packets for transmission through the ITM/TPIU.
.
A core halt – entry to debug state.
.
When exception tracing is enabled, the DWT emits an exception trace packet under the following conditions:
Exception entry (from thread mode or pre-emption of a thread or handler).
.
Exception exit when exiting a handler.
.
Exception return when re-entering a pre-empted thread or handler code sequence.
.
The DWT is designed for use with advanced profiling and debug tools, available from multiple vendors.
Altering DWT configuration may conflict with the operation of advanced profiling and debug tools.
For further information on the DWT, contact Silicon Labs support for the ARM® CortexTM-M3 Technical
Reference Manual, the ARM® CoreSightTM Components Technical Reference Manual, the ARM® v7-M
Architecture Reference Manual, and the ARM® v7-M Architecture Application Level Reference Manual.
14-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
15 Flash Patch and Breakpoint (FPB)
The EM35x integrates the standard ARM® Flash Patch and Breakpoint (FPB). The FPB implements hardware
breakpoints. The FPB also provides support for remapping of specific instruction or literal locations from flash
memory to an address in RAM memory. The FPB contains:
Two literal comparators for matching against literal loads from flash space, and remapping to a
corresponding RAM space.
.
Six instruction comparators for matching against instruction fetches from flash space, and remapping to a
corresponding RAM space. Alternatively, the comparators can be individually configured to return a
breakpoint instruction to the processor core on a match, implementing hardware breakpoint capability.
.
The FPB contains a global enable, but also individual enables for the eight comparators. If the comparison for
an entry matches, the address is remapped to the address defined in the remap register plus and offset
corresponding to the comparator that matched. Alternately, the address is remapped to a breakpoint
instruction. The comparison happens on the fly, but the result of the comparison occurs too late to stop the
original instruction fetch or literal load taking place from the flash space. The processor ignores this
transaction, however, and only the remapped transaction is used.
Memory Protection Unit (MPU) lookups are performed for the original address, not the remapped address.
Unaligned literal accesses are not remapped. The original access to the bus takes place in this case.
The FPB is designed for use with advanced debug tools, available from multiple vendors. Altering FPB
configuration may conflict with the operation of advanced debug tools.
For further information on the FPB, contact Silicon Labs support for the ARM® CortexTM-M3 Technical
Reference Manual, the ARM® CoreSightTM Components Technical Reference Manual, the ARM® v7-M
Architecture Reference Manual, and the ARM® v7-M Architecture Application Level Reference Manual.
15-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
16 Integrated Voltage Regulator
The EM35x integrates two low dropout regulators to provide 1.8 V and 1.25 V power supplies, as detailed in
Table 16-1. The 1V8 regulator supplies the analog and memories, and the 1V25 regulator supplies the digital
core. In deep sleep the voltage regulators are disabled.
When enabled, the 1V8 regulator steps down the pads supply voltage (VDD_PADS) from a nominal 3.0 V to
1.8 V. The regulator output pin (VREG_OUT) must be decoupled externally with a suitable capacitor.
VREG_OUT should be connected to the 1.8 V supply pins VDDA, VDD_RF, VDD_VCO, VDD_SYNTH, VDD_IF, and
VDD_MEM. The 1V8 regulator can supply a maximum of 50 mA.
When enabled, the 1V25 regulator steps down VDD_PADS to 1.25 V. The regulator output pin (VDD_CORE,
Pin 17) must be decoupled externally with a suitable capacitor. It should connect to the other VDD_CORE pin
(Pin 44). The 1V25 regulator can supply a maximum of 10 mA.
The regulators are controlled by the digital portion of the chip as described in Chapter 6, System Modules.
An example of decoupling capacitors and PCB layout can be found in the application notes (see the various
Silicon Labs EM35x reference design documentation).
Table 16-1. Integrated Voltage Regulator Specifications
Spec Point
Min.
2.1
Typ.
Max.
3.6
Units Comments
Supply range for regulator
1V8 regulator output
V
V
VDD_PADS
-5%
-5%
1.8
+5%
+5%
Regulator output after initialization
Regulator output after reset
1V8 regulator output after
reset
1.75
1V25 regulator output
-5%
-5%
1.25
1.45
+5%
+5%
V
Regulator output after initialization
Regulator output after reset
1V25 regulator output
after reset
1V8 regulator capacitor
2.2
µF
Low ESR tantalum capacitor
ESR greater than 2 Ω
ESR less than 10 Ω
de-coupling less than 100 nF ceramic
1V25 regulator capacitor
1.0
µF
Ceramic capacitor (0603)
Regulator output current
1V8 regulator output
current
0
0
50
10
mA
1V25 regulator output
current
mA
Regulator output current
No load current
600
200
µA
No load current (bandgap and regulators)
Short circuit current limit
1V8 regulator current
limit
mA
1V25 regulator current
limit
25
50
50
mA
µs
Short circuit current limit
1V8 regulator start-up
time
0 V to POR threshold
2.2 µF capacitor
1V25 regulator start-up
time
µs
0 V to POR threshold
1.0 µF capacitor
16-1
Final
120-035X-000 Rev. 1.2
EM351 / EM357
An external 1.8 V regulator may replace both internal regulators. The EM35x can control external regulators
during deep sleep using open-drain GPIO PA7, as described in Chapter 7, GPIO. The EM35x drives PA7 low
during deep sleep to disable the external regulator and an external pull-up is required to release this signal to
indicate that supply voltage should be provided. Current consumption increases approximately 2 mA when
using an external regulator. When using an external regulator the internal regulators should be disabled
through Ember software. The always-on domain needs to be minimally powered at 2.1 V, and cannot be
powered from the external 1.8 V regulator.
16-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
17 Serial Wire and JTAG (SWJ) Interface
The EM35x includes a standard Serial Wire and JTAG (SWJ) Interface. The SWJ is the primary debug and
programming interface of the EM35x. The SWJ gives debug tools access to the internal buses of the EM35x,
and allows for non-intrusive memory and register access as well as CPU halt-step style debugging. Therefore,
any design implementing the EM35x should make the SWJ signals readily available.
Serial Wire is an ARM® standard, bidirectional, two-wire protocol designed to replace JTAG, and provides all
the normal JTAG debug and test functionality. JTAG is a standard five-wire protocol providing debug and test
functionality. In addition, the two Serial Wire signals (SWDIO and SWCLK) are overlaid on two of the JTAG
signals (JTMS and JTCK). This keeps the design compact and allows debug tools to switch between Serial Wire
and JTAG as needed, without changing pin connections.
While Serial Wire and JTAG offer the same debug and test functionality, Silicon Labs recommends Serial Wire.
Serial Wire uses only two pins instead of five, and offers a simple communication protocol, high performance
data rates, low power, built-in error detection, and protection from glitches.
The ARM® CoreSightTM Debug Access Port (DAP) comprises the Serial Wire and JTAG Interface (SWJ). As
illustrated in Figure 17-1, the DAP includes two primary components: a debug port (the SWJ-DP) and an access
port (the AHB-AP). The SWJ-DP provides external debug access, while the AHB-AP provides internal bus
access. An external debug tool connected to the EM35x’s debug pins communicates with the SWJ-DP. The
SWJ-DP then communicates with the AHB-AP. Finally, the AHB-AP communicates on the internal bus.
Figure 17-1. SWJ Block Diagram
SWJ-DAP
SWJ-DP
SW
interface
SWJ-DP
select
Control and
AP interface
pins
AHB-AP
AHB
JTAG
interface
Serial Wire and JTAG share five pins:
JRST
.
JTDO
.
JTDI
.
SWDIO/JTMS
.
SWCLK/JTCK
.
Note: The SWJ pins are forced functions, and their corresponding GPIO_PxCFGH/L configurations are
overridden when the EM35x resets. An application must disable all debug SWJ debug functionality to reclaim
any of the four SWJ GPIOs: PC0, PC2, PC3, and PC4.
Since these pins can be repurposed, refer to Chapter 1, Pin Assignments, and Section 7.3, Forced Functions, in
Chapter 7, GPIO, for complete pin descriptions and configurations.
For further information on the SWJ, contact customer support for Application Notes and ARM® CoreSightTM
documentation.
17-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
18 Typical Application
Figure 18-1 illustrates the typical application circuit, and Table 18-1 contains an example Bill of Materials
(BOM) for the off-chip components required by the EM35x.
Note: The circuit shown in Figure 18-1 is for example purposes only, and the BOM is for budgetary quotes
only. For a complete reference design, please download one of the latest Silicon Labs hardware reference
designs from the Silicon Labs website (www.silabs.com/zigbee-support).
The Balun provides an impedance transformation from the antenna to the EM35x for both Tx and Rx modes.
L1 tunes the impedance presented to the RF port for maximum transmit power and receive sensitivity.
The harmonic filter (L2, L3, C5, C6 and C9) provides additional suppression of the second harmonic, which
increases the margin over the FCC limit.
The 24 MHz crystal Y1 with loading capacitors is required and provides the high-frequency crystal oscillator
source for the EM35x’s main system clock. The 32.768 kHz crystal with loading capacitors generates a highly
accurate low-frequency crystal oscillator for use with peripherals, but it is not mandatory as the low-
frequency internal RC oscillator can be used.
Loading capacitance and ESR (C1 and R3) provides stability for the internal 1.8 V regulator.
Loading capacitance C2 provides stability for the internal 1.25 V regulator, no ESR is required because it is
contained within the chip.
Resistor R1 reduces the operating voltage of the flash memory, this reduces current consumption and
improves sensitivity by 1 dB when compared to not using it.
Various decoupling capacitors are required, these should be placed as close to their corresponding pins as
possible. For values and locations see one of the latest reference designs.
An antenna matched to 50 Ω is required.
18-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 18-1. Typical Application Circuit
C3
Y1
C4
VBRD
R1
1
2
3
36
VDD_24MHz
PB0
35
Antenna
VDD_VCO
RF_P
PC4
34
PC3
Ceramic
Balun
33
PC2
EM35x
L1
32
JTCK
L3
L2
4
31
RF_N
PB2
5
6
7
8
30
VDD_RF
PB1
RF_TX_ALT_P
RF_TX_ALT_N
VDD_IF
29
PA6
C9
C5 C6
28
VDD_PADS
9
NC
Harmonic
Filter
27
PA5
10
11
12
26
VDD_PADSA
PC5
PA4
25
PA3
49
nRESET
GND
C7
C2
Optional
Y2
C8
PC2
PC0
PC3
JTCK
PC4
Programing and
debug interface
(these pins should be
routed to test points)
R3
C1
nReset
PA4
PA5
18-2
Final
120-035X-000 Rev. 1.2
EM351 / EM357
Table 18-1 contains a typical Bill of Materials for the application circuit shown in Figure 18-1. The information
within this table should be used for a rough cost analysis. For a more detailed BOM, please refer to one of
Silicon Labs’ EM35x-based reference designs at the Silicon Labs website (www.silabs.com/zigbee-support).
Table 18-1. Bill of Materials for Figure 18-1
Item Qty Reference
Description
Manufacturer
1
1
1
1
2
1
2
1
1
2
1
1
1
C2
CAPACITOR, 1 µF, 6.3 V, X5R, 10%, 0402
<not specified>
2
C1
CAPACITOR, 2.2 µF, 10 V, X5R, 10%, 0603
CAPACITOR, 22 pF, ±5%, 50 V, NPO, 0402
CAPACITOR, 18 pF, ±5%, 50 V, NPO, 0402
CAPACITOR, 33 pF, ±5%, 50 V, NPO, 0402
CAPACITOR, 1 pF, ±0.25 pF, 50 V, 0402, NPO
CAPACITOR, 1.8pF, ±0.25 pF, 50 V, 0402, NPO
INDUCTOR, 5.1 nH, ±0.3 nH, 0402 MULTILAYER
INDUCTOR, 2.7 nH, ±0.3 nH, 0402, MULTILAYER
RESISTOR, 10 Ω, 5%, 0402
<not specified>
<not specified>
<not specified>
<not specified>
<not specified>
3
C7
4
C3,C4
C8
5
6
C5, C9
C6
7
8
L1
Murata LQG15HS5N1
Murata LQG15HS2N7
<not specified>
9
L2, L3
R1
10
11
12
R3
RESISTOR, 1 Ω, 5%, 0402
<not specified>
U1
EM35x SINGLE-CHIP ZIGBEE/802.15.4-2003
SOLUTION
Ember EM35x
13
14
15
1
1
1
Y1
CRYSTAL, 24.000 MHz, ± 25 ppm STABILITY OVER ILSI, Abracon, KDS, Epson
-40 to +85ºC, 18 pF
Y2 (Optional)
BLN1
CRYSTAL, 32.768 kHz, ±20 ppm INITIAL
Abracon, KDS, Epson
TOLERANCE AT +25ºC, 12.5 pF
BALUN, CERAMIC 50/200 Ω
Wurth 748421245
Johanson 2450BL15B100E
Murata LDB212G4010C
16
1
ANT1
ANTENNA
Johanson
2450AT18B100E
18-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
19 Mechanical Details
The EM35x package is a plastic 48-pin QFN that is 7 mm x 7 mm x 0.90 mm. Figure 19-1 illustrates the package
drawing.
Figure 19-1. Package Drawing
Top View
Bottom View
D2
e
Detail A
D
3
Pin 1
Common Dimensions (mm)
E1 E2
E
Sym.
Minimum
Nominal
0.90
Maximum
A
0.80
1.00
0.05
0.23
7.12
A1
A2
D
0
0.18
0.20
7.00
5.5
6.82
D1
D2
E
5.275
6.82
5.30
7.00
5.5
5.325
7.12
Exposed
Die
Pad
E1
E2
L
5.275
0.48
5.30
0.50
0.25
0.50
5.325
0.52
b
0.225
0.475
0.275
0.525
D1
b
e
R
b min / 2
Edge View
A
Detail B
Notes:
1. JEDEC ref. MO - 220
2. All dimensions are in millimeters
3. Pin 1 orientation identified by chamfer on corner of
exposed die pad
Detail A
4. Dimension ‘e’ represents the terminal pitch
Detail B
0.3000
0.3000
R
19.1 QFN48 Footprint Recommendations
Figure 19-2 demonstrates the IPC-7351 recommended PCB Footprint for the EM35x (QFN50P700X700X90-49N).
A ground pad in the bottom center of the package forms a 49th pin.
A 3 x 3 array of non-thermal vias should connect the EM35x decal center shown in Figure 19-2 to the PCB
ground plane through the ground pad. In order to properly solder the EM35x to the footprint, the Paste Mask
layer should have a 3 x 3 array of circular openings at 1.015 mm diameter spaced approximately 1.625 mm
(center to center) apart, as shown in Figure 19-3. This will cause an evenly distributed solder flow and
coplanar attachment to the PCB. The solder mask layer (illustrated in Figure 19-4) should be the same as the
copper layer for the EM35x footprint.
For more information on the package footprint, please refer to the appropriate EM35x Reference Design.
19-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Figure 19-2. PCB Footprint for the EM35x
X2
Via Drill DIA = 0.254mm
b2
b0
MIN TYP
MAX
a1
C1
C2
X1
X2
Y1
Y2
E
a0
a1
a2
6.85
6.85
E
0.30
5.35
0.90
5.35
0.50
1.80
1.625
0.75
b1
C2
Y2
b0
b1
b2
1.80
1.625
0.75
* Dimensions in mm
* Dimensions are for
Figures 19-2, 19-3
and 19-4
Y1
a2
X1
a0
C1
Figure 19-3. Paste Mask Dimensions
Figure 19-4. Solder Mask Dimensions
X2
a1
a0
DIA = 1.01mm
b0
b1
b2
E
C2
C2
Y2
Y1
E
Y1
X1
a2
C1
X1
C1
19-2
Final
120-035X-000 Rev. 1.2
EM351 / EM357
19.2 Solder Temperature Profile
Figure 19-5 illustrates the solder temperature profile for the EM35x. This temperature profile is similar for
other RoHS compliant packages, but manufacturing lines should be programmed with this profile in order to
guarantee proper solder connection to the PCB.
Figure 19-5. EM35x Reflow Profile
Temperature
Tpeak
Ramp-up
Ramp-down
TL
Tsoak
max
Tsoak
Tsoak
min
t
soak
25C
time
t
peak
t
25C to tpeak
Table 19-1 contains the temperature profile parameters.
Table 19-1. Solder Reflow Parameters
Parameter
Value
Average Ramp Up Rate (from Tsoakmax to Tpeak)
Minimum Soak Temperature (Tsoakmin
3°C per second max
150°C
)
Maximum Soak Temperature (Tsoakmax
TL
)
200°C
217°C
Time above TL
60 – 150 seconds
260 + 0°C
Tpeak
Time within 5°C of Tpeak
Ramp Down Rate
Time from 25°C to Tpeak
20 – 40 seconds
6°C per second max
8 minutes, max
19-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
20 Part Marking
Figure 20-1 shows the part marking for the EM300 Series. The circle in the top corner indicates Pin 1. Pins are
numbered counter-clockwise from Pin 1 with 12 pins per package edge.
Figure 20-1. Part Marking for EM35X
EM357
ZZZZZZ.ZZ
YYWW M
where:
ZZZZZZ.ZZ defines the production lot code.
•
•
YYWW defines the year and week assembled.
M defines the package assembly location (if there is no letter on the package, then the package was
assembled in South Wales)
•
o
o
o
W indicates South Wales
C indicates China
M indicates Malaysia
20-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
21 Ordering Information
Use the following part numbers to order the EM357:
EM357-RTR Reel contains 2000 units / reel
EM357-RTY Sample tray
•
•
Use the following part numbers to order the EM351:
EM351-RTR Reel contains 2000 units / reel
•
EM351-RTY Sample tray
•
The EM300 Series package is RoHS-compliant. It conforms to the European Court of Justice decision regarding
the Deca-BDE exemption of the RoHS Directive. It is PFOS-compliant in accordance with European Directive
2006/122/EC*1 released in December 2006. The EM357-RTR and EM351-RTR reel conforms to EIA Specification
481.
To order parts, contact Silicon Labs at 1+(877) 444-3032, or find a sales office or distributor on our web-
site, www.silabs.com.
21-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
22 Shipping Box Label
Silicon Labs includes the following information on each tape and reel box label (EM357-RTR or EM351-RTR):
Package
•
Device Type
•
Quantity (Bar coded)
•
•
Box ID (Bar coded)
Lot Number (Bar coded)
•
Date Code (Bar coded)
•
Figure 22-1 depicts the label position on the box. As shown in this figure, there can be up to two date codes in
a single tape and reel.
Figure 22-1. Contents Label
PACKAGE 48 LEAD QFN
DEVICE
BOX QTY
EM357-RTR
2000
BOX Id
XXXX-YYYYYY
LOT No
LOT No
AAAAAA.B.CC
DDDDDD.E.FF
QTY
QTY
DATE YYWW
DATE YYWW
22-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
23 Revision History
Revision
Location
Description of Change
1.2
Chapter 21
Removed Table 21-1 and Figure 21-1.
Update to table.
1.1
M
Table 2-3
Table 1-1, pin 48
Section 6.1.2 and Chapter 16
Update to table.
Always-on domain power range change.
Chapter 9, TIM1_OR and TIM2_OR Change TIM1_EXTRIGSEL to TIM_EXTRIGSEL.
Appendix C
Datasheet
Update to references.
L
Rebranding to Silicon Labs.
K
Chapter 1
Chapter 5
Section 7.5
Adjusted naming and clarified description of FIB monitor.
Chapter 20
Added Malaysia to package assembly locations.
Register reference corrected.
Minor corrections.
J
I
Section 9.3.5
Register tables TIM_CC2S,
TIMxCCMR1, TIMx_CCMR2,
TIMx_CCER
Table 6-9, Table 18-1
Chapter 7
Updates to tables.
Update to Figure 7-2 and associated text.
Chapter 8
Update to SCx_SPICFG register to clarify SPI slave transmit
padding.
Section 8.3.2
Update 12 MHz SPI Master operation.
Updates to tables.
H
G
Table 2-4, Table 6-9
Section 8.4.1 and 8.4.3
Chapter 21
New notes.
Update with RTY information.
Removes references to high voltage mode.
Table 2-1
Chapter 3
Chapter 10
Table 2-4
Updates to Table 2-4.
Section 6.4
Section 6.4.2
Section 8.6.2
Figure 2-1
Table 2-8
Better describes Ember software capabilities.
New note.
Adds information on conditions when UART can miss bytes.
Caption corrected, scaling updated.
EVM clarified.
F
23-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Revision
Location
Table 1-1
Table 2-7
Description of Change
Pin 48 updated.
E
Updates to Table 2-7.
Chapter 2
Table 6-8
Changes to ST comments and reset levels.
Supply dependence parameter now typical, not max.
Reset language clarified.
Chapter 6
Section 10.5.3.
Section 10.1.5
10.4
New section.
Paragraph deleted.
Calibration clarifications.
23-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Appendix A Register Address Table
BLOCK
CM_LV
40004000 - 40004038 CM_LV
Address
40004038
Name
Type
Reset
Description
Peripheral Disable Register
PERIPHERAL_DISABLE
RW
0
BLOCK
INTERRUPTS
Name
4000A000 - 4000AFFF Interrupts
Address
Type
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
Reset
Description
4000A800
4000A804
4000A808
4000A80C
4000A810
4000A814
4000A818
4000A81C
4000A820
4000A840
4000A844
4000A848
4000A84C
4000A850
4000A854
4000A858
4000A860
4000A864
4000A868
4000A86C
INT_TIM1FLAG
INT_TIM2FLAG
INT_SC1FLAG
INT_SC2FLAG
INT_ADCFLAG
INT_GPIOFLAG
INT_TIM1MISS
INT_TIM2MISS
INT_MISS
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Timer 1 Interrupt Flag Register
Timer 2 Interrupt Flag Register
Serial Controller 1 Interrupt Flag Register
Serial Controller 2 Interrupt Flag Register
ADC Interrupt Flag Register
GPIO Interrupt Flag Register
Timer 1 Missed Interrupt Register
Timer 2 Missed Interrupts Register
Top-Level Missed Interrupts Register
Timer 1 Interrupt Configuration Register
Timer 2 Interrupt Configuration Register
Serial Controller 1 Interrupt Configuration Register
Serial Controller 2 Interrupt Configuration Register
ADC Interrupt Configuration Register
Serial Controller 1 Interrupt Mode Register
Serial Controller 2 Interrupt Mode Register
GPIO Interrupt A Configuration Register
GPIO Interrupt B Configuration Register
GPIO Interrupt C Configuration Register
GPIO Interrupt D Configuration Register
INT_TIM1CFG
INT_TIM2CFG
INT_SC1CFG
INT_SC2CFG
INT_ADCCFG
SC1_INTMODE
SC2_INTMODE
GPIO_INTCFGA
GPIO_INTCFGB
GPIO_INTCFGC
GPIO_INTCFGD
A-1
Final
120-035X-000 Rev. 1.2
EM351 / EM357
BLOCK
GPIO
4000B000 - 4000BFFF General Purpose IO
Address
Name
Type
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
R
Reset
Description
4000B000
4000B004
4000B008
4000B00C
4000B010
4000B014
4000B400
4000B404
4000B408
4000B40C
4000B410
4000B414
4000B800
4000B804
4000B808
4000B80C
4000B810
4000B814
4000BC00
4000BC04
4000BC08
4000BC0C
4000BC10
4000BC14
4000BC18
4000BC1C
GPIO_PACFGL
GPIO_PACFGH
GPIO_PAIN
4444
Port A Configuration Register (Low)
Port A Configuration Register (High)
Port A Input Data Register
4444
0
GPIO_PAOUT
GPIO_PASET
GPIO_PACLR
GPIO_PBCFGL
GPIO_PBCFGH
GPIO_PBIN
0
Port A Output Data Register
Port A Output Set Register
0
0
Port A Output Clear Register
Port B Configuration Register (Low)
Port B Configuration Register (High)
Port B Input Data Register
4444
4444
0
0
GPIO_PBOUT
GPIO_PBSET
GPIO_PBCLR
GPIO_PCCFGL
GPIO_PCCFGH
GPIO_PCIN
Port B Output Data Register
Port B Output Set Register
0
0
Port B Output Clear Register
Port C Configuration Register (Low)
Port C Configuration Register (High)
Port C Input Data Register
4444
4444
0
GPIO_PCOUT
GPIO_PCSET
GPIO_PCCLR
GPIO_DBGCFG
GPIO_DBGSTAT
GPIO_PAWAKE
GPIO_PBWAKE
GPIO_PCWAKE
GPIO_IRQCSEL
GPIO_IRQDSEL
GPIO_WAKEFILT
0
Port C Output Data Register
Port C Output Set Register
0
0
Port C Output Clear Register
GPIO Debug Configuration Register
GPIO Debug Status Register
Port A Wakeup Monitor Register
Port B Wakeup Monitor Register
Port C Wakeup Monitor Register
Interrupt C Select Register
10
0
RW
RW
RW
RW
RW
RW
0
0
0
F
10
0
Interrupt D Select Register
GPIO Wakeup Filtering Register
A-2
Final
120-035X-000 Rev. 1.2
EM351 / EM357
BLOCK
SERIAL
4000C000 - 4000CFFF Serial Controllers
Address
Name
Type
RW
RW
RW
RW
RW
RW
RW
RW
R
Reset
Description
4000C000
4000C004
4000C008
4000C00C
4000C010
4000C014
4000C018
4000C01C
4000C020
4000C024
4000C028
4000C02C
4000C030
4000C034
4000C038
4000C03C
4000C040
4000C044
4000C04C
4000C050
4000C054
4000C058
4000C060
4000C064
4000C070
4000C800
4000C804
4000C808
4000C80C
4000C810
4000C814
4000C818
4000C81C
4000C820
4000C824
4000C828
4000C82C
4000C830
SC2_RXBEGA
SC2_RXENDA
SC2_RXBEGB
SC2_RXENDB
SC2_TXBEGA
SC2_TXENDA
SC2_TXBEGB
SC2_TXENDB
SC2_RXCNTA
SC2_RXCNTB
SC2_TXCNT
SC2_DMASTAT
SC2_DMACTRL
SC2_RXERRA
SC2_RXERRB
SC2_DATA
20000000
Receive DMA Begin Address Register A
Receive DMA End Address Register A
Receive DMA Begin Address Register B
Receive DMA End Address Register B
Transmit DMA Begin Address Register A
Transmit DMA End Address Register A
Transmit DMA Begin Address Register B
Transmit DMA End Address Register B
Receive DMA Count Register A
Receive DMA Count Register B
Transmit DMA Count Register
20000000
20000000
20000000
20000000
20000000
20000000
20000000
0
R
0
R
0
R
0
Serial DMA Status Register
RW
R
0
Serial DMA Control Register
0
DMA First Receive Error Register A
DMA First Receive Error Register B
Serial Data Register
R
0
RW
R
0
SC2_SPISTAT
SC2_TWISTAT
SC2_TWICTRL1
SC2_TWICTRL2
SC2_MODE
0
SPI Status Register
R
0
TWI Status Register
RW
RW
RW
RW
RW
RW
R
0
TWI Control Register 1
0
TWI Control Register 2
0
Serial Mode Register
SC2_SPICFG
SC2_RATELIN
SC2_RATEEXP
SC2_RXCNTSAVED
SC1_RXBEGA
SC1_RXENDA
SC1_RXBEGB
SC1_RXENDB
SC1_TXBEGA
SC1_TXENDA
SC1_TXBEGB
SC1_TXENDB
SC1_RXCNTA
SC1_RXCNTB
SC1_TXCNT
SC1_DMASTAT
SC1_DMACTRL
0
SPI Configuration Register
0
Serial Clock Linear Prescaler Register
Serial Clock Exponential Prescaler Register
Saved Receive DMA Count Register
Receive DMA Begin Address Register A
Receive DMA End Address Register A
Receive DMA Begin Address Register B
Receive DMA End Address Register B
Transmit DMA Begin Address Register A
Transmit DMA End Address Register A
Transmit DMA Begin Address Register B
Transmit DMA End Address Register B
Receive DMA Count Register A
Receive DMA Count Register B
Transmit DMA Count Register
0
0
RW
RW
RW
RW
RW
RW
RW
RW
R
20000000
20000000
20000000
20000000
20000000
20000000
20000000
20000000
0
R
0
R
0
R
0
Serial DMA Status Register
RW
0
Serial DMA Control Register
A-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
BLOCK
SERIAL
4000C000 - 4000CFFF Serial Controllers
Address
Name
Type
R
Reset
Description
4000C834
4000C838
4000C83C
4000C840
4000C844
4000C848
4000C84C
4000C850
4000C854
4000C858
4000C85C
4000C860
4000C864
4000C868
4000C86C
4000C870
SC1_RXERRA
SC1_RXERRB
SC1_DATA
0
0
0
0
0
40
0
0
0
0
0
0
0
0
0
0
DMA First Receive Error Register A
DMA First Receive Error Register B
Serial Data Register
R
RW
R
SC1_SPISTAT
SC1_TWISTAT
SC1_UARTSTAT
SC1_TWICTRL1
SC1_TWICTRL2
SC1_MODE
SPI Status Register
R
TWI Status Register
R
UART Status Register
RW
RW
RW
RW
RW
RW
RW
RW
RW
R
TWI Control Register 1
TWI Control Register 2
Serial Mode Register
SC1_SPICFG
SC1_UARTCFG
SC1_RATELIN
SC1_RATEEXP
SC1_UARTPER
SC1_UARTFRAC
SC1_RXCNTSAVED
SPI Configuration Register
UART Configuration Register
Serial Clock Linear Prescaler Register
Serial Clock Exponential Prescaler Register
UART Baud Rate Period Register
UART Baud Rate Fractional Period Register
Saved Receive DMA Count Register
BLOCK
ADC
4000D000 - 4000DFFF Analog to Digital Converter
Address
Name
Type
R
Reset
Description
4000D000
4000D004
4000D008
4000D00C
4000D010
4000D014
4000D018
4000D01C
4000D020
4000D024
ADC_DATA
0
ADC Data Register
ADC_CFG
RW
RW
RW
RW
R
00001800
ADC Configuration Register
ADC Offset Register
ADC_OFFSET
ADC_GAIN
0000
8000
ADC Gain Register
ADC_DMACFG
ADC_DMASTAT
ADC_DMABEG
ADC_DMASIZE
ADC_DMACUR
ADC_DMACNT
0
ADC DMA Configuration Register
ADC DMA Status Register
ADC DMA Begin Address Register
ADC DMA Buffer Size Register
ADC DMA Current Address Register
ADC DMA Count Register
0
RW
RW
R
20000000
0
20000000
0
R
A-4
120-035X-000 Rev. 1.2
Final
EM351 / EM357
BLOCK
TIM1
4000E000 - 4000EFFF General Purpose Timer 1
Address
Name
Type
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
Reset
Description
4000E000
4000E004
4000E008
4000E014
4000E018
4000E01C
4000E020
4000E024
4000E028
4000E02C
4000E034
4000E038
4000E03C
4000E040
4000E050
TIM1_CR1
TIM1_CR2
TIM1_SMCR
TIM1_EGR
TIM1_CCMR1
TIM1_CCMR2
TIM1_CCER
TIM1_CNT
TIM1_PSC
TIM1_ARR
TIM1_CCR1
TIM1_CCR2
TIM1_CCR3
TIM1_CCR4
TIM1_OR
0
0
Timer 1 Control Register 1
Timer 1 Control Register 2
0
Timer 1 Slave Mode Control Register
Timer 1 Event Generation Register
Timer 1 Capture/Compare Mode Register 1
Timer 1 Capture/Compare Mode Register 2
Timer 1 Capture/Compare Enable Register
Timer 1 Counter Register
0
0
0
0
0
0
Timer 1 Prescaler Register
FFFF
0
Timer 1 Auto-Reload Register
Timer 1 Capture/Compare Register 1
Timer 1 Capture/Compare Register 2
Timer 1 Capture/Compare Register 3
Timer 1 Capture/Compare Register 4
Timer 1 Option Register
0
0
0
0
BLOCK
TIM2
4000F000 - 4000FFFF General Purpose Timer 2
Address
Name
Type
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
RW
Reset
Description
4000F000
4000F004
4000F008
4000F014
4000F018
4000F01C
4000F020
4000F024
4000F028
4000F02C
4000F034
4000F038
4000F03C
4000F040
4000F050
TIM2_CR1
TIM2_CR2
TIM2_SMCR
TIM2_EGR
TIM2_CCMR1
TIM2_CCMR2
TIM2_CCER
TIM2_CNT
TIM2_PSC
TIM2_ARR
TIM2_CCR1
TIM2_CCR2
TIM2_CCR3
TIM2_CCR4
TIM2_OR
0
0
Timer 2 Control Register 1
Timer 2 Control Register 2
0
Timer 2 Slave Mode Control Register
Timer 2 Event Generation Register
Timer 2 Capture/Compare Mode Register 1
Timer 2 Capture/Compare Mode Register 2
Timer 2 Capture/Compare Enable Register
Timer 2 Counter Register
0
0
0
0
0
0
Timer 2 Prescaler Register
FFFF
0
Timer 2 Auto-Reload Register
Timer 2 Capture/Compare Register 1
Timer 2 Capture/Compare Register 2
Timer 2 Capture/Compare Register 3
Timer 2 Capture/Compare Register 4
Timer 2 Option Register
0
0
0
0
A-5
Final
120-035X-000 Rev. 1.2
EM351 / EM357
BLOCK
NVIC
E000E000 - E000EFFF Nested Vectored Interrupt Controller
Address
E000E100
E000E180
E000E200
E000E280
E000E300
E000ED3C
Name
Type
RW
RW
RW
RW
R
Reset
Description
INT_CFGSET
INT_CFGCLR
INT_PENDSET
INT_PENDCLR
INT_ACTIVE
SCS_AFSR
0
0
0
0
0
0
Top-Level Set Interrupts Configuration Register
Top-Level Clear Interrupts Configuration Register
Top-Level Set Interrupts Pending Register
Top-Level Clear Interrupts Pending Register
Top-Level Active Interrupts Register
RW
Auxiliary Fault Status Register
A-6
Final
120-035X-000 Rev. 1.2
EM351 / EM357
Appendix B Abbreviations and Acronyms
Acronym/Abbreviation
Meaning
ACK
Acknowledgement
ADC
Analog to Digital Converter
Advanced Encryption Standard
Automatic Gain Control
AES
AGC
AHB
Advanced High Speed Bus
Advanced Peripheral Bus
APB
CBC-MAC
CCA
Cipher Block Chaining—Message Authentication Code
Clear Channel Assessment
CCM
CCM*
CIB
Counter with CBC-MAC Mode for AES encryption
Improved Counter with CBC-MAC Mode for AES encryption
Customer Information Block
1 kHz Clock
CLK1K
CLK32K
CPU
32.768 kHz Crystal Clock
Central Processing Unit
CRC
Cyclic Redundancy Check
CSMA-CA
CTR
Carrier Sense Multiple Access-Collision Avoidance
Counter Mode
CTS
Clear to Send
DNL
Differential Non-Linearity
DMA
DWT
EEPROM
EM
Direct Memory Access
Data Watchpoint and Trace
Electrically Erasable Programmable Read Only Memory
Event Manager
ENOB
ESD
effective number of bits
Electro Static Discharge
ESR
Equivalent Series Resistance
External Trigger Input
ETR
FCLK
FIB
ARM® CortexTM-M3 CPU Clock
Fixed Information Block
FIFO
First-in, First-out
B-1
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Acronym/Abbreviation
Meaning
FPB
Flash Patch and Breakpoint
GPIO
HF
General Purpose I/O (pins)
High Frequency
I2C
Inter-Integrated Circuit
Integrated Development Environment
Intermediate Frequency
Institute of Electrical and Electronics Engineers
Integral Non-linearity
IDE
IF
IEEE
INL
ITM
Instrumentation Trace Macrocell
Joint Test Action Group
Low Frequency
JTAG
LF
LNA
Low Noise Amplifier
LQI
Link Quality Indicator
LSB
Least significant bit
MAC
MFB
MISO
MOS
MOSI
MPU
MSB
Medium Access Control
Main Flash Block
Master in, slave out
Metal Oxide Semiconductor (P-channel or N-channel)
Master out, slave in
Memory Protection Unit
Most significant bit
MSL
Moisture Sensitivity Level
NACK
NIST
NMI
Negative Acknowledge
National Institute of Standards and Technology
Non-Maskable Interrupt
NVIC
OPM
O-QPSK
OSC24M
OSC32K
OSCHF
Nested Vectored Interrupt Controller
One-Pulse Mode
Offset-Quadrature Phase Shift Keying
High Frequency Crystal Oscillator
Low-Frequency 32.768 kHz Oscillator
High-Frequency Internal RC Oscillator
B-2
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Acronym/Abbreviation
Meaning
OSCRC
Low-Frequency RC Oscillator
PA
Power Amplifier
PCLK
PER
PHY
PLL
Peripheral clock
Packet Error Rate
Physical Layer
Phase-Locked Loop
POR
PRNG
PSD
Power-On-Reset
Pseudo Random Number Generator
Power Spectral Density
Packet Trace Interface
Pulse Width Modulation
Quad Flat Pack
PTI
PWM
QFN
RAM
RC
Random Access Memory
Resistive/Capacitive
RF
Radio Frequency
RMS
RoHS
RSSI
RTS
Root Mean Square
Restriction of Hazardous Substances
Receive Signal Strength Indicator
Request to Send
Rx
Receive
SYSCLK
SDFR
SFD
System clock
Spurious Free Dynamic Range
Start Frame Delimiter
Signal-to-noise and distortion ratio
Serial Peripheral Interface
Serial Wire and JTAG Interface
Total Harmonic Distortion
True random number generator
Two Wire serial interface
Transmit
SINAD
SPI
SWJ
THD
TRNG
TWI
Tx
UART
Universal Asynchronous Receiver/Transmitter
B-3
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Acronym/Abbreviation
Meaning
UEV
Update event
VCO
Voltage Controlled Oscillator
Abbreviation
Meaning
dB
decibel
dBc
decibels relative to the carrier
dBm
GHz
kB
decibels relative to 1 mW
GigaHerz
Kilobyte
kbps
kHz
kΩ
kilobits/second
kiloherz
kiloOhm
kV
kiloVolt
mA
Mbps
MHz
MΩ
MSPS
µA
milliAmpere
Megabits per second
megaherz
megaOhm
Megasamples per second
microAmpere
microsecond
nanohenry
µsec
nH
ns
nanoseconds
Ohm
Ω
pF
picofarad
ppm
V
part per million
Volt
B-4
120-035X-000 Rev. 1.2
Final
EM351 / EM357
Appendix C References
ZigBee Specification (www.zigbee.org; ZigBee Document 053474) (ZigBee Alliance membership required)
.
.
ZigBee-PRO Stack Profile (www.zigbee.org; ZigBee Document 074855) (ZigBee Alliance membership
required)
ZigBee Stack Profile (www.zigbee.org; ZigBee Document 064321) (ZigBee Alliance membership required)
.
.
Bluetooth Core Specification v2.1
(http://www.bluetooth.org/docman/handlers/downloaddoc.ashx?doc_id=241363)
IEEE 802.15.4-2003 (http://standards.ieee.org/getieee802/download/802.15.4-2003.pdf)
IEEE 802.11g (standards.ieee.org/getieee802/download/802.11g-2003.pdf)
.
.
.
ARM® Cortex-M3 reference manual
(http://infocenter.arm.com/help/topic/com.arm.doc.subset.cortexm.m3/index.html#cortexm3)
Copyright © 2012 Silicon Laboratories
The information in this document is believed to be accurate in all respects at the time of publication but is subject to
change without notice. Silicon Laboratories assumes no responsibility for errors and omissions, and disclaims responsibility
for any consequences resulting from the use of information included herein. Additionally, Silicon Laboratories assumes no
responsibility for the functioning of undescribed features or parameters. Silicon Laboratories reserves the right to make
changes without further notice. Silicon Laboratories makes no warranty, representation or guarantee regarding the
suitability of its products for any particular purpose, nor does Silicon Laboratories assume any liability arising out of the
application or use of any product or circuit, and specifically disclaims any and all liability, including without limitation
consequential or incidental damages. Silicon Laboratories products are not designed, intended, or authorized for use in
applications intended to support or sustain life, or for any other application in which the failure of the Silicon Laboratories
product could create a situation where personal injury or death may occur. Should Buyer purchase or use Silicon
Laboratories products for any such unintended or unauthorized application, Buyer shall indemnify and hold Silicon
Laboratories harmless against all claims and damages.
Silicon Laboratories, Silicon Labs, and Ember are registered trademarks of Silicon Laboratories Inc.
Other products or brandnames mentioned herein are trademarks or registered trademarks of their respective holders.
C-1
120-035X-000 Rev. 1.2
Final
相关型号：
©2020 ICPDF网 联系我们和版权申明