AD9655 [ADI]

Dual, 16-Bit, 125 MSPS Serial LVDS, 1.8 V Analog-to-Digital Converter;
AD9655
型号: AD9655
厂家: ADI    ADI
描述:

Dual, 16-Bit, 125 MSPS Serial LVDS, 1.8 V Analog-to-Digital Converter

文件: 总38页 (文件大小:1550K)
中文:  中文翻译
下载:  下载PDF数据表文档文件
Dual, 16-Bit, 125 MSPS Serial LVDS,  
1.8 V Analog-to-Digital Converter  
Data Sheet  
AD9655  
FUNCTIONAL BLOCK DIAGRAM  
FEATURES  
1.8 V supply operation  
AVDD  
AGND  
DRVDD  
Low power: approximately 150 mW/channel at 125 MSPS,  
2V p-p input range (typical)  
SNR/SFDR at 69.5 MHz  
AD9655  
D0A+  
D0A–  
D1A+  
D1A–  
D0B+  
D0B–  
D1B+  
D1B–  
DCO+  
DCO–  
FCO+  
FCO–  
16  
VINA+  
VINA–  
16-BIT  
PIPELINE  
ADC  
77.5 dBFS/88 dBc, 2.0 V p-p input range (typical)  
79.3 dBFS/84 dBc, 2.8 V p-p input range (typical)  
Linearity  
DNL = 0.7 LSB; INL = 4.0 LSB (typical, 2.0 V p-p input span)  
DNL = 0.7 LSB; INL = 3.4 LSB (typical, 2.8 V p-p input span)  
Serial LVDS, two data lanes per ADC channel  
500 MHz full power analog bandwidth  
Serial port control  
16  
16  
VCM  
VINB+  
VINB–  
16-BIT  
PIPELINE  
ADC  
16  
REFERENCE  
SERIAL PORT  
INTERFACE  
1 TO 8  
CLOCK DIVIDER  
SCLK/ SDIO/ CSB  
DFS PDWN  
CLK+ CLK–  
Full chip and individual channel power-down modes  
Flexible bit orientation  
Built-in and custom digital test pattern generation  
Clock divider  
Figure 1.  
Programmable output clock and data alignment  
Standby mode  
APPLICATIONS  
Communications  
Diversity radio systems  
Multimode digital receivers  
GSM, EDGE, W-CDMA, LT E, CDMA2000, WiMAX, TD-SCDMA  
I/Q demodulation systems  
Smart antenna systems  
Broadband data applications  
Battery-powered instruments  
Handheld scope meters  
Portable medical imaging and ultrasound  
Radar/LIDAR  
Individual channel power-down is supported. The AD9655  
typically consumes less than 2 mW in serial port interface (SPI)  
power-down mode. The available digital test pat-terns include  
built-in deterministic and pseudorandom patterns, along with  
custom user-defined test patterns entered via the SPI.  
GENERAL DESCRIPTION  
The AD9655 is available in an RoHS-compliant, 32-lead LFCSP.  
It is specified over the industrial temperature range of −40°C to  
+85°C. This device is protected by a U.S. patent.  
The AD9655 is a dual, 16-bit, 125 MSPS analog-to-digital  
converter (ADC) with an on-chip sample-and-hold circuit  
designed for low cost, low power, small size, and ease of use.  
The product operates at a conversion rate of up to 125 MSPS  
and is optimized for outstanding dynamic performance and low  
power in applications where a small package size is critical.  
PRODUCT HIGHLIGHTS  
1. Small Footprint.  
Two ADCs are contained in a small, space-saving package.  
2. Pin Compatible.  
The ADC requires a single 1.8 V power supply and an LVPECL-/  
CMOS-/LVDS-compatible sample rate clock for full performance  
operation. External reference or driver components are not  
required for many applications.  
The AD9655 is pin compatible to the AD9645 14-bit and  
AD9635 12-bit dual ADCs.  
3. Ease of Use.  
A DCO operates at frequencies of up to 500 MHz and  
supports double data rate (DDR) operation.  
4. User Flexibility.  
The ADC automatically multiplies the sample rate clock for the  
appropriate LVDS serial data rate. A data clock output (DCO)  
for capturing data on the output and a frame clock output (FCO)  
for signaling a new output byte are provided.  
The SPI control offers a wide range of flexible features to  
meet specific system requirements.  
Rev. 0  
Document Feedback  
Information furnished by Analog Devices is believed to be accurate and reliable. However, no  
responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other  
rightsof third parties that may result fromits use. Specifications subject to change without notice. No  
license is granted by implication or otherwise under any patent or patent rights of Analog Devices.  
Trademarks andregisteredtrademarks are the property of their respective owners.  
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.  
Tel: 781.329.4700  
Technical Support  
©2015 Analog Devices, Inc. All rights reserved.  
www.analog.com  
 
 
 
 
 
AD9655* PRODUCT PAGE QUICK LINKS  
Last Content Update: 02/23/2017  
COMPARABLE PARTS  
View a parametric search of comparable parts.  
DESIGN RESOURCES  
AD9655 Material Declaration  
PCN-PDN Information  
Quality And Reliability  
Symbols and Footprints  
EVALUATION KITS  
AD9655 Evaluation Board  
DOCUMENTATION  
Data Sheet  
DISCUSSIONS  
View all AD9655 EngineerZone Discussions.  
AD9655: Dual, 16-Bit, 125 MSPS Serial LVDS, 1.8 V Analog-  
to-Digital Converter Data Sheet  
SAMPLE AND BUY  
Visit the product page to see pricing options.  
User Guides  
AD9655/AD9645/AD9635 Evaluation Documentation  
TECHNICAL SUPPORT  
Submit a technical question or find your regional support  
number.  
TOOLS AND SIMULATIONS  
AD9655 S Parameter  
DOCUMENT FEEDBACK  
Submit feedback for this data sheet.  
This page is dynamically generated by Analog Devices, Inc., and inserted into this data sheet. A dynamic change to the content on this page will not  
trigger a change to either the revision number or the content of the product data sheet. This dynamic page may be frequently modified.  
AD9655  
Data Sheet  
TABLE OF CONTENTS  
Features .............................................................................................. 1  
Clock Input Considerations...................................................... 22  
Power Dissipation and Power-Down Mode ........................... 23  
Digital Outputs and Timing ..................................................... 24  
Output Test Modes..................................................................... 27  
Serial Port Interface (SPI).............................................................. 28  
Configuration Using the SPI..................................................... 28  
Hardware Interface..................................................................... 29  
Configuration Without the SPI ................................................ 29  
SPI Accessible Features.............................................................. 29  
Memory Map .................................................................................. 30  
Reading the Memory Map Register Table............................... 30  
Memory Map Register Table..................................................... 31  
Memory Map Register Descriptions........................................ 34  
Applications Information .............................................................. 36  
Design Guidelines ...................................................................... 36  
Power and Ground Guidelines................................................. 36  
Clock Stability Considerations ................................................. 36  
Exposed Pad Thermal Heat Slug Recommendations............ 36  
VCM............................................................................................. 36  
Reference Bypassing................................................................... 36  
SPI Port ........................................................................................ 36  
Outline Dimensions....................................................................... 37  
Ordering Guide .......................................................................... 37  
Applications....................................................................................... 1  
General Description......................................................................... 1  
Functional Block Diagram .............................................................. 1  
Product Highlights ........................................................................... 1  
Revision History ............................................................................... 2  
Specifications..................................................................................... 3  
DC Specifications ......................................................................... 3  
AC Specifications.......................................................................... 5  
Digital Specifications ................................................................... 7  
Switching Specifications .............................................................. 8  
Timing Specifications .................................................................. 8  
Absolute Maximum Ratings.......................................................... 10  
Thermal Resistance .................................................................... 10  
ESD Caution................................................................................ 10  
Pin Configuration and Function Descriptions........................... 11  
Typical Performance Characteristics ........................................... 12  
VREF = 1.0 V ................................................................................. 12  
VREF = 1.4 V ................................................................................. 15  
Equivalent Circuits ......................................................................... 18  
Theory of Operation ...................................................................... 19  
Analog Input Considerations.................................................... 19  
Voltage Reference ....................................................................... 20  
REVISION HISTORY  
1/15—Revision 0: Initial Version  
Rev. 0 | Page 2 of 37  
 
Data Sheet  
AD9655  
SPECIFICATIONS  
DC SPECIFICATIONS  
AVDD = 1.8 V, DRVDD = 1.8 V, 2 V p-p full-scale differential input mode, internal reference voltage (VREF) = 1.0 V, input amplitude  
(AIN) = −1.0 dBFS, 125 MSPS, unless otherwise noted.  
Table 1.  
Parameter1  
Temperature  
Min  
Typ  
Max  
Unit  
RESOLUTION  
16  
Bits  
ACCURACY  
No Missing Codes  
Offset Error  
Offset Matching  
Gain Error  
Full  
Guaranteed2  
25°C  
25°C  
25°C  
25°C  
25°C  
25°C  
0.2  
0.1  
3.4  
0.4  
0.7  
4.0  
% FSR  
% FSR  
% FSR  
% FSR  
LSB  
Gain Matching  
Differential Nonlinearity (DNL)  
Integral Nonlinearity (INL)  
TEMPERATURE DRIFT  
Gain Error  
LSB  
Full  
Full  
−23  
0.9  
ppm/°C  
ppm/°C  
Offset Error  
INTERNAL VOLTAGE REFERENCE  
Output Voltage (1 V Mode)  
Load Regulation at 1.0 mA (VREF = 1 V)  
Input Resistance  
25°C  
25°C  
25°C  
1.0  
2.9  
7.5  
V
mV  
kΩ  
INPUT-REFERRED NOISE  
VREF = 1.0 V  
25°C  
2.7  
LSB rms  
ANALOG INPUTS  
Differential Input Voltage (VREF = 1 V)  
Common-Mode Voltage  
Common-Mode Range  
Differential Input Resistance  
Differential Input Capacitance  
POWER SUPPLY  
Full  
Full  
25°C  
25°C  
25°C  
2
0.9  
V p-p  
V
V
kΩ  
pF  
0.5  
1.2  
1.9  
6.6  
AVDD  
Full  
Full  
25°C  
25°C  
25°C  
1.7  
1.7  
1.8  
1.8  
93  
73  
62  
1.9  
1.9  
V
V
mA  
mA  
mA  
DRVDD  
3
IAVDD  
IDRVDD (ANSI-644 Mode)3  
IDRVDD (Reduced Range Mode)3  
TOTAL POWER CONSUMPTION  
Sine Wave Input (Two Channels, Including Output Drivers ANSI-644 Mode)  
Sine Wave Input (Two Channels, Including Output Drivers Reduced Range Mode)  
Power-Down  
Standby4  
25°C  
25°C  
25°C  
25°C  
299  
279  
2
mW  
mW  
mW  
mW  
142  
1 See the AN-835 Application Note, Understanding High Speed ADC Testing and Evaluation, for definitions and for details on how these tests were completed.  
2 No missing codes guaranteed if Register 0x18 = 0x04 (default, no digital scaling of the output).  
3 Measured with a low input frequency, −1 dBFS sine wave on both channels, DDR operation, and two-lane operation.  
4 Standby mode can be controlled via the SPI.  
Rev. 0 | Page 3 of 37  
 
 
AD9655  
Data Sheet  
AVDD = 1.8 V, DRVDD = 1.8 V, 2.8 V p-p full-scale differential input mode, VREF = 1.4 V, A IN = −1.0 dBFS, 125 MSPS, unless otherwise noted.  
Table 2.  
Parameter1  
Temperature Min  
Typ Max  
Unit  
RESOLUTION  
16  
Bits  
ACCURACY  
No Missing Codes  
Offset Error  
Offset Matching  
Gain Error  
Gain Matching  
Differential Nonlinearity (DNL)  
Full  
Guaranteed2  
−0.12 +0.2 +0.48 % FSR  
Full  
Full  
Full  
Full  
Full  
25°C  
Full  
25°C  
−0.2  
−2.4  
−1.2  
−0.99  
+0.1 +0.33 % FSR  
+2.8 +8.2  
+0.4 +1.9  
% FSR  
% FSR  
+1.43 LSB  
0.7  
3.4  
LSB  
LSB  
LSB  
Integral Nonlinearity (INL)  
−8.5  
1.37  
+8.5  
TEMPERATURE DRIFT  
Gain Error  
Offset Error  
Full  
Full  
−66  
0.9  
ppm/°C  
ppm/°C  
INTERNAL VOLTAGE REFERENCE  
Output Voltage (1.4 V Mode)  
Load Regulation at 1.0 mA (VREF = 1.4 V)  
Input Resistance  
Full  
25°C  
25°C  
1.38 1.41  
186  
7.5  
V
mV  
kΩ  
INPUT-REFERRED NOISE  
VREF = 1.4 V  
25°C  
2
LSB rms  
ANALOG INPUTS  
Differential Input Voltage (VREF = 1.4 V)  
Common-Mode Voltage  
Common-Mode Range  
Differential Input Resistance  
Differential Input Capacitance  
POWER SUPPLY  
Full  
Full  
25°C  
25°C  
25°C  
2.8  
0.9  
1.0  
1.9  
6.6  
V p-p  
V
V
kΩ  
pF  
0.7  
AVDD  
Full  
Full  
Full  
Full  
Full  
1.7  
1.7  
1.8  
1.8  
101  
73  
1.9  
1.9  
111  
79  
V
V
mA  
mA  
mA  
DRVDD  
3
IAVDD  
IDRVDD (ANSI-644 Mode)3  
IDRVDD (Reduced Range Mode)3  
62  
68  
TOTAL POWER CONSUMPTION  
Sine Wave Input (Two Channels, Including Output Drivers ANSI-644 Mode)  
Sine Wave Input (Two Channels, Including Output Drivers Reduced Range Mode)  
Power-Down  
Standby4  
Full  
Full  
Full  
Full  
313  
293  
2
342  
322  
4
mW  
mW  
mW  
mW  
155  
172  
1 See the AN-835 Application Note, Understanding High Speed ADC Testing and Evaluation, for definitions and for details on how these tests were completed.  
2 No missing codes guaranteed if Register 0x18 = 0x04 (default, no digital scaling of the output).  
3 Measured with a low input frequency, −1 dBFS sine wave on both channels, DDR operation, and two-lane operation.  
4 Standby mode can be controlled via the SPI.  
Rev. 0 | Page 4 of 37  
Data Sheet  
AD9655  
AC SPECIFICATIONS  
AVDD = 1.8 V, DRVDD = 1.8 V, 2 V p-p full-scale differential input mode, VREF = 1.0 V, AIN = −1.0 dBFS, 125 MSPS, unless otherwise noted.  
Table 3.  
Parameter1  
Temperature  
Min  
Typ  
Max  
Unit  
SIGNAL-TO-NOISE RATIO (SNR)  
fIN = 9.7 MHz  
fIN = 19.7 MHz  
fIN = 69.5 MHz  
fIN = 100.5 MHz  
fIN = 139.5 MHz  
fIN = 301 MHz  
25°C  
25°C  
25°C  
25°C  
25°C  
25°C  
77.9  
77.9  
77.5  
76.6  
75.6  
71.0  
dBFS  
dBFS  
dBFS  
dBFS  
dBFS  
dBFS  
SIGNAL-TO-NOISE-AND-DISTORTION RATIO (SINAD)  
fIN = 9.7 MHz  
fIN = 19.7 MHz  
fIN = 69.5 MHz  
fIN = 100.5 MHz  
fIN = 139.5 MHz  
fIN = 301 MHz  
25°C  
25°C  
25°C  
25°C  
25°C  
25°C  
77.5  
77.1  
77.1  
76.5  
75.2  
68.0  
dBFS  
dBFS  
dBFS  
dBFS  
dBFS  
dBFS  
EFFECTIVE NUMBER OF BITS (ENOB)  
fIN = 9.7 MHz  
fIN = 19.7 MHz  
fIN = 69.5 MHz  
fIN = 100.5 MHz  
fIN = 139.5 MHz  
fIN = 301 MHz  
25°C  
25°C  
25°C  
25°C  
25°C  
25°C  
12.6  
12.5  
12.5  
12.4  
12.2  
11  
Bits  
Bits  
Bits  
Bits  
Bits  
Bits  
SPURIOUS-FREE DYNAMIC RANGE (SFDR)  
fIN = 9.7 MHz  
fIN = 19.7 MHz  
fIN = 69.5 MHz  
fIN = 100.5 MHz  
fIN = 139.5 MHz  
fIN = 301 MHz  
25°C  
25°C  
25°C  
25°C  
25°C  
25°C  
88  
86  
88  
91  
85  
70  
dBc  
dBc  
dBc  
dBc  
dBc  
dBc  
WORST HARMONIC (SECOND OR THIRD)  
fIN = 9.7 MHz  
fIN = 19.7 MHz  
fIN = 69.5 MHz  
fIN = 100.5 MHz  
fIN = 139.5 MHz  
fIN = 301 MHz  
25°C  
25°C  
25°C  
25°C  
25°C  
25°C  
−88  
−86  
−88  
−91  
−85  
−70  
dBc  
dBc  
dBc  
dBc  
dBc  
dBc  
WORST OTHER (EXCLUDING SECOND OR THIRD HARMONIC)  
fIN = 9.7 MHz  
fIN = 19.7 MHz  
fIN = 69.5 MHz  
fIN = 100.5 MHz  
fIN = 139.5 MHz  
fIN = 301 MHz  
25°C  
25°C  
25°C  
25°C  
25°C  
25°C  
−95  
−99  
−92  
−91  
−89  
−80  
dBc  
dBc  
dBc  
dBc  
dBc  
dBc  
TWO-TONE INTERMODULATION DISTORTION (IMD)—AIN1 AND AIN2 = −7.0 dBFS  
fIN1 = 100.1 MHz, fIN2 = 102.1 MHz  
CROSSTALK2  
25°C  
25°C  
25°C  
25°C  
90  
dBc  
dB  
−104  
−100  
500  
CROSSTALK (OVERRANGE CONDITION)3  
dB  
ANALOG INPUT BANDWIDTH, FULL POWER  
MHz  
1 See the AN-835 Application Note, Understanding High Speed ADC Testing and Evaluation, for definitions and for details on how these tests were completed.  
2 Crosstalk is measured at 69.5 MHz with −1.0 dBFS analog input on one channel and no input on the adjacent channel. Measurements are taken using a less dense  
board to demonstrate the AD9655 crosstalk performance, not board limitations.  
3 Overrange condition is specified as being 3 dB above the full-scale input range.  
Rev. 0 | Page 5 of 37  
 
AD9655  
Data Sheet  
AVDD = 1.8 V, DRVDD = 1.8 V, 2.8 V p-p full-scale differential input mode, VREF = 1.4 V, A IN = −1.0 dBFS, 125 MSPS, unless otherwise noted.  
Table 4.  
Parameter1  
Temperature  
Min  
Typ  
Max  
Unit  
SIGNAL-TO-NOISE RATIO (SNR)  
fIN = 9.7 MHz  
fIN = 19.7 MHz  
fIN = 69.5 MHz  
fIN = 100.5 MHz  
fIN = 139.5 MHz  
fIN = 301 MHz  
25°C  
25°C  
Full  
25°C  
25°C  
25°C  
79.6  
79.4  
79.3  
78.0  
76.5  
55.0  
dBFS  
dBFS  
dBFS  
dBFS  
dBFS  
dBFS  
78.0  
SIGNAL-TO-NOISE-AND-DISTORTION RATIO (SINAD)  
fIN = 9.7 MHz  
fIN = 19.7 MHz  
fIN = 69.5 MHz  
fIN = 100.5 MHz  
fIN = 139.5 MHz  
fIN = 301 MHz  
25°C  
25°C  
Full  
25°C  
25°C  
25°C  
79.1  
78.3  
77.8  
77.0  
75.8  
54.8  
dBFS  
dBFS  
dBFS  
dBFS  
dBFS  
dBFS  
77.2  
12.5  
80  
EFFECTIVE NUMBER OF BITS (ENOB)  
fIN = 9.7 MHz  
fIN = 19.7 MHz  
fIN = 69.5 MHz  
fIN = 100.5 MHz  
fIN = 139.5 MHz  
fIN = 301 MHz  
25°C  
25°C  
Full  
25°C  
25°C  
25°C  
12.8  
12.7  
12.6  
12.5  
12.3  
8.8  
Bits  
Bits  
Bits  
Bits  
Bits  
Bits  
SPURIOUS-FREE DYNAMIC RANGE (SFDR)  
fIN = 9.7 MHz  
fIN = 19.7 MHz  
fIN = 69.5 MHz  
fIN = 100.5 MHz  
fIN = 139.5 MHz  
fIN = 301 MHz  
25°C  
25°C  
Full  
25°C  
25°C  
25°C  
88  
85  
84  
83  
82  
68  
dBc  
dBc  
dBc  
dBc  
dBc  
dBc  
WORST HARMONIC (SECOND OR THIRD)  
fIN = 9.7 MHz  
fIN = 19.7 MHz  
fIN = 69.5 MHz  
fIN = 100.5 MHz  
fIN = 139.5 MHz  
fIN = 301 MHz  
25°C  
25°C  
Full  
25°C  
25°C  
25°C  
−88  
−85  
−84  
−83  
−82  
−68  
dBc  
dBc  
dBc  
dBc  
dBc  
dBc  
−80  
−85  
WORST OTHER (EXCLUDING SECOND OR THIRD HARMONIC)  
fIN = 9.7 MHz  
fIN = 19.7 MHz  
fIN = 69.5 MHz  
fIN = 100.5 MHz  
fIN = 139.5 MHz  
fIN = 301 MHz  
25°C  
25°C  
Full  
25°C  
25°C  
25°C  
−97  
−98  
−93  
−91  
−89  
−72  
dBc  
dBc  
dBc  
dBc  
dBc  
dBc  
TWO-TONE INTERMODULATION DISTORTION (IMD)—AIN1 AND AIN2 = −7.0 dBFS  
fIN1 = 100.1 MHz, fIN2 = 102.1 MHz  
CROSSTALK2  
25°C  
25°C  
25°C  
25°C  
85  
dBc  
dB  
−104  
−103  
500  
CROSSTALK (OVERRANGE CONDITION)3  
dB  
ANALOG INPUT BANDWIDTH, FULL POWER  
MHz  
1 See the AN-835 Application Note, Understanding High Speed ADC Testing and Evaluation, for definitions and for details on how these tests were completed.  
2 Crosstalk is measured at 69.5 MHz with −1.0 dBFS analog input on one channel and no input on the adjacent channel. Measurements are taken using a less dense  
board to demonstrate AD9655 crosstalk performance, not board limitations.  
3 Overrange condition is specified as being 3 dB above the full-scale input range.  
Rev. 0 | Page 6 of 37  
Data Sheet  
AD9655  
DIGITAL SPECIFICATIONS  
AVDD = 1.8 V and DRVDD = 1.8 V, unless otherwise noted.  
Table 5.  
Parameter1  
Temperature Min  
Typ  
Max  
Unit  
CLOCK INPUTS (CLK+, CLK−)  
Logic Compliance  
Differential Input Voltage2  
Input Voltage Range  
CMOS/LVDS/LVPECL  
3.6  
Full  
Full  
0.2  
V p-p  
V
AGND − 0.2  
AVDD + 0.2  
Input Common-Mode Voltage  
Input Resistance (Differential)  
Input Capacitance  
Full  
25°C  
25°C  
0.9  
15  
4
V
kΩ  
pF  
LOGIC INPUTS (SCLK/DFS)  
Logic 1 Voltage  
Logic 0 Voltage  
Full  
Full  
1.2  
0
AVDD + 0.2  
0.8  
V
V
Input Resistance  
Input Capacitance  
25°C  
25°C  
30  
2
kΩ  
pF  
LOGIC INPUT (CSB)  
Logic 1 Voltage  
Logic 0 Voltage  
Full  
Full  
1.2  
0
AVDD + 0.2  
0.8  
V
V
Input Resistance  
Input Capacitance  
25°C  
25°C  
26  
2
kΩ  
pF  
LOGIC INPUT (SDIO/PDWN)  
Logic 1 Voltage  
Logic 0 Voltage  
Full  
Full  
1.2  
0
AVDD + 0.2  
0.8  
V
V
Input Resistance  
Input Capacitance  
25°C  
25°C  
26  
5
kΩ  
pF  
LOGIC OUTPUT (SDIO/PDWN)3  
Logic 1 Voltage (IOH = 800 μA)  
Logic 0 Voltage (IOL = 50 μA)  
DIGITAL OUTPUTS (D0x , D1x ), ANSI-644  
Logic Compliance  
Differential Output Voltage (VOD)  
Output Offset Voltage (VOS)  
Output Coding (Default)  
DIGITAL OUTPUTS (D0x , D1x ), LOW POWER, REDUCED SIGNAL OPTION  
Logic Compliance  
Full  
Full  
1.79  
V
V
0.05  
LVDS  
350  
1.22  
Full  
Full  
298  
1.15  
400  
1.30  
mV  
V
Twos complement  
LVDS  
Differential Output Voltage (VOD)  
Output Offset Voltage (VOS)  
Output Coding (Default)  
Full  
Full  
170  
1.15  
200  
1.22  
Twos complement  
231  
1.30  
mV  
V
1 See the AN-835 Application Note, Understanding High Speed ADC Testing and Evaluation, for definitions and for details on how these tests were completed.  
2 Specified for LVDS and LVPECL only.  
3 Specified for 13 SDIO/PDWN pins sharing the same connection.  
Rev. 0 | Page 7 of 37  
 
AD9655  
Data Sheet  
SWITCHING SPECIFICATIONS  
AVDD = 1.8 V and DRVDD = 1.8 V, unless otherwise noted.  
Table 6.  
Parameter1, 2  
Temperature  
Min  
Typ  
Max  
Unit  
CLOCK3  
Input Clock Rate  
Conversion Rate  
Full  
Full  
Full  
Full  
20  
20  
4.00  
4.00  
1000  
125  
MHz  
MSPS  
ns  
Clock Pulse Width High (tEH)  
Clock Pulse Width Low (tEL)  
OUTPUT PARAMETERS3  
Propagation Delay (tPD)4  
Rise Time (tR)5 (20% to 80%)  
Fall Time (tF)5 (20% to 80%)  
FCO Propagation Delay (tFCO  
DCO Propagation Delay (tCPD  
DCO to Data Delay (tDATA  
FCO to DCO Delay (tFRAME  
Data to Data Skew  
ns  
Full  
Full  
Full  
Full  
Full  
Full  
Full  
Full  
25°C  
25°C  
Full  
(tSAMPLE/4) + 5  
(tSAMPLE/4) + 5  
(tSAMPLE/4) + 6.1  
170  
160  
(tSAMPLE/4) + 6.1  
tFCO + (tSAMPLE/16) + 0.2  
(tSAMPLE/4) + 7  
(tSAMPLE/4) + 7  
ns  
ps  
ps  
ns  
ns  
ps  
ps  
ps  
4
)
4
)
4, 6  
)
(tSAMPLE/16) − 500  
(tSAMPLE/16) + 10  
(tSAMPLE/16) + 100  
(tSAMPLE/16) + 330  
80  
4, 7  
)
37  
250  
250  
16  
Wake-Up Time (Standby)  
Wake-Up Time (Power-Down)8  
Pipeline Latency  
ns  
ms  
Clock cycles  
APERTURE  
Aperture Delay (tA)9  
25°C  
25°C  
25°C  
1
80  
1
ns  
fs rms  
Clock cycles  
Aperture Uncertainty (Jitter, tJ5, 9  
)
Out-of-Range Recovery Time  
1 See the AN-835 Application Note, Understanding High Speed ADC Testing and Evaluation, for definitions and for details on how these tests were completed.  
2 Measured on standard FR-4 material.  
3 The Output parameters can be adjusted via the SPI. The conversion rate is the clock rate after the divider. Valid for 2-lane operation.  
4 tSAMPLE = tEH + tEL = 1/fS. tCPD, tDATA and tFRAME are adjustable with SPI Register 0x16.  
5 This term does not appear in the Timing Diagrams section, which includes Figure 2 and Figure 3.  
6 tDATA is the time from DCO rise or fall to output data rise or fall.  
7 tFRAME is the time from FCO rise to DCO rise.  
8 Wake-up time from power-down is defined as the time required to return to normal operation from SPI power-down mode. The value of 250 ms assumes a sample  
rate of 125 MSPS. About 31 × 106 sample clock cycles are required.  
9 tA and tJ are with Register 0x09 = 0x04 (default, duty cycle stabilizer and clock divider are bypassed).  
TIMING SPECIFICATIONS  
Table 7.  
Parameter  
Description  
Limit  
Unit  
SPI TIMING REQUIREMENTS  
tDS  
tDH  
tCLK  
tS  
See Figure 68, unless otherwise noted  
Setup time between the data and the rising edge of SCLK  
Hold time between the data and the rising edge of SCLK  
Period of the SCLK  
Setup time between CSB and SCLK  
Hold time between CSB and SCLK  
SCLK pulse width high  
SCLK pulse width low  
Time required for the SDIO pin to switch from an input to an output relative to the  
SCLK falling edge  
Time required for the SDIO pin to switch from an output to an input relative to the  
SCLK rising edge  
4
2
40  
2
2
10  
10  
10  
ns min  
ns min  
ns min  
ns min  
ns min  
ns min  
ns min  
ns min  
tH  
tHIGH  
tLOW  
tEN_SDIO  
1
1
tDIS_SDIO  
10  
ns min  
1 This parameter is not shown in Figure 68.  
Rev. 0 | Page 8 of 37  
 
 
 
Data Sheet  
AD9655  
Timing Diagrams  
N – 1  
VINx±  
N
N + 1  
tA  
tEH  
tEL  
CLK–  
CLK+  
tCPD  
DCO–  
DDR  
DCO+  
DCO–  
SDR  
DCO+  
tFCO  
tFRAME  
FCO–  
FCO+  
tPD  
tDATA  
D0A–  
BITWISE  
MODE  
D14  
D12  
D10  
D08  
D06  
D04  
D02  
LSB  
D14  
D12  
D10  
D08  
D06  
D04  
D02  
LSB  
N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 16 N – 16 N – 16 N – 16 N – 16 N – 16 N – 16 N – 16  
D0A+  
D1A–  
D1A+  
MSB  
D13  
D11  
D09  
D07  
D05  
D03  
D01  
MSB  
D13  
D11  
D09  
D07  
D05  
D03  
D01  
N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 16 N – 16 N – 16 N – 16 N – 16 N – 16 N – 16 N – 16  
FCO–  
FCO+  
D0A–  
BYTEWISE  
D07  
D06  
D05  
D04  
D03  
D2  
D01  
LSB  
D07  
D06  
D05  
D04  
D03  
D02  
D01  
LSB  
N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 16 N – 16 N – 16 N – 16 N – 16 N – 16 N – 16 N – 16  
MODE  
D0A+  
D1A–  
D1A+  
MSB  
D14  
D13  
D12  
D11  
D10  
D09  
D08  
MSB  
D14  
D13  
D12  
D11  
D10  
D09  
D08  
N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 16 N – 16 N – 16 N – 16 N – 16 N – 16 N – 16 N – 16  
Figure 2. 16-Bit DDR/Single Data Rate (SDR), Two-Lane, 1× Frame Mode (Default)  
N – 1  
VINx±  
tA  
N
tEH  
tEL  
CLK–  
CLK+  
tCPD  
DCO–  
DCO+  
tFCO  
tFRAME  
FCO–  
FCO+  
D0x–  
D0x+  
tDATA  
tPD  
MSB  
D14  
D13  
D12  
D11  
D10  
D9  
D8  
D7  
D6  
D5  
D4  
D3  
D2  
D1  
LSB  
MSB  
D14  
D13  
N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 17 N – 16 N – 16 N – 16  
Figure 3. Wordwise DDR, One-Lane, 1× Frame, 16-Bit Output Mode  
Rev. 0 | Page 9 of 37  
 
 
 
AD9655  
Data Sheet  
ABSOLUTE MAXIMUM RATINGS  
THERMAL RESISTANCE  
Table 8.  
The exposed pad is the only ground connection for the chip.  
The exposed pad must be soldered to the AGND plane of the  
circuit board. Soldering the exposed pad to the board also  
increases the reliability of the solder joints and maximizes the  
thermal capability of the package.  
Parameter  
Rating  
Electrical  
AVDD to AGND  
DRVDD to AGND  
Digital Outputs (D0x , D1x , DCO ,  
FCO ) to AGND  
CLK+, CLK− to AGND  
VINx+, VINx− to AGND  
SCLK/DFS, SDIO/PDWN, CSB to AGND  
RBIAS to AGND  
VREF to AGND  
VCM to AGND  
−0.3 V to +2.0 V  
−0.3 V to +2.0 V  
−0.3 V to +2.0 V  
Table 9. Thermal Resistance  
−0.3 V to +2.0 V  
−0.3 V to +2.0 V  
−0.3 V to +2.0 V  
−0.3 V to +2.0 V  
−0.3 V to +2.0 V  
−0.3 V to +2.0 V  
Airflow  
Velocity  
Package Type (m/sec)  
1, 2  
1, 3  
1, 4  
1, 2  
θJA  
θJC  
3.1  
θJB  
20.7  
ΨJT  
0.3  
Unit  
°C/W  
°C/W  
°C/W  
32-Lead LFCSP,  
5 mm × 5 mm  
0
1.0  
2.5  
37.1  
32.4  
29.1  
N/A5 N/A5 0.5  
N/A5 N/A5 0.8  
Environmental  
1 Per JEDEC 51-7, plus JEDEC 51-5 2S2P test board.  
2 Per JEDEC JESD51-2 (still air) or JEDEC JESD51-6 (moving air).  
3 Per MIL-Std 883, Method 1012.1.  
Operating Temperature Range (Ambient)  
Maximum Junction Temperature  
Lead Temperature (Soldering, 10 sec)  
Storage Temperature Range (Ambient)  
−40°C to +85°C  
150°C  
300°C  
4 Per JEDEC JESD51-8 (still air).  
5 N/A means not applicable.  
−65°C to +150°C  
Typical θJA is specified for a 4-layer printed circuit board (PCB)  
with a solid ground plane. As shown in Table 9, airflow improves  
heat dissipation, which reduces θJA. In addition, metal in direct  
contact with the package leads from metal traces, through holes,  
ground, and power planes reduces θJA.  
Stresses at or above those listed under Absolute Maximum  
Ratings may cause permanent damage to the product. This is a  
stress rating only; functional operation of the product at these  
or any other conditions above those indicated in the operational  
section of this specification is not implied. Operation beyond  
the maximum operating conditions for extended periods may  
affect product reliability.  
ESD CAUTION  
Rev. 0 | Page 10 of 37  
 
 
 
 
Data Sheet  
AD9655  
PIN CONFIGURATION AND FUNCTION DESCRIPTIONS  
AVDD  
CLK+  
CLK–  
1
2
3
4
5
6
7
8
24 AVDD  
23  
22 VCM  
RBIAS  
AD9655  
SDIO/PDWN  
SCLK/DFS  
DRVDD  
D1B–  
21 VREF  
TOP VIEW  
20  
19  
CSB  
DRVDD  
(Not to Scale)  
18 D0A+  
17 D0A–  
D1B+  
NOTES  
1. THE EXPOSED PAD IS THE ONLY GROUND CONNECTION  
ON THE CHIP. IT MUST BE SOLDERED TO THE ANALOG GROUND  
OF THE PCB TO ENSURE PROPER FUNCTIONALITY AND HEAT  
DISSIPATION, NOISE, AND MECHANICAL STRENGTH BENEFITS.  
Figure 4. Pin Configuration, Top View  
Table 10. Pin Function Descriptions  
Pin No.  
Mnemonic  
Description  
0, Exposed Pad  
AGND,  
Exposed Pad  
Exposed Pad. The exposed pad is the only ground connection on the chip. It must be soldered to the  
analog ground of the PCB to ensure proper functionality and heat dissipation, noise, and mechanical  
strength benefits.  
1, 24, 25, 28, 29, 32 AVDD  
1.8 V Supply Pins for the ADC Core Domain.  
2, 3  
4
CLK+, CLK−  
SDIO/PDWN  
Differential Encode Clock. These pins are PECL-, LVDS-, or 1.8 V CMOS-compatible inputs.  
SPI Data Input/Output (SDIO). This pin is a bidirectional SPI data input/output with a 31 kΩ internal  
pull-down resistor.  
Non-SPI Mode Power-Down (PDWN). This pin provides static control of chip power-down, and has a 31 kΩ  
internal pull-down resistor.  
5
SCLK/DFS  
SPI Clock Input in SPI Mode (SCLK). This pin has a 30 kΩ internal pull-down resistor.  
Non-SPI Mode Data Format Select (DFS). This provides static control of the data output format. This  
pin has a 30 kΩ internal pull-down resistor. Pull DFS high for a twos complement output; pull DFS  
low for an offset binary output.  
6, 19  
7, 8  
DRVDD  
1.8 V Supply Pins for Output Driver Domain.  
Channel B Lane 1 Digital Outputs.  
Channel B Lane 0 Digital Outputs.  
Data Clock Outputs.  
Frame Clock Outputs.  
Channel A Lane 1 Digital Outputs.  
Channel A Lane 0 Digital Outputs.  
SPI Chip Select. Active low enable; this pin has a 15 kΩ internal pull-up resistor.  
D1B−, D1B+  
D0B−, D0B+  
DCO−, DCO+  
FCO−, FCO+  
D1A−, D1A+  
D0A−, D0A+  
CSB  
9, 10  
11, 12  
13, 14  
15, 16  
17, 18  
20  
21  
VREF  
1.0 V to 1.4 V Voltage Reference Output. Bypass this pin to ground with a 1.0 µF capacitor in parallel  
with a 0.1 µF capacitor; this pin internally provides reference voltage to the ADC. This pin can be  
disabled via Register 0x114 if external VREF is desired.  
22  
VCM  
Analog Output Voltage at Mid AVDD Supply. Bypass this pin to ground with a 0.1 µF capacitor; this  
pin can be used to set the common mode of the analog inputs externally.  
23  
26, 27  
30, 31  
RBIAS  
VINA−, VINA+  
VINB+, VINB−  
Sets Analog Current Bias. Connect this pin to a 10.0 kΩ (1% tolerance) resistor to ground.  
Channel A ADC Analog Inputs.  
Channel B ADC Analog Inputs.  
Rev. 0 | Page 11 of 37  
 
AD9655  
Data Sheet  
TYPICAL PERFORMANCE CHARACTERISTICS  
VREF = 1.0 V  
0
0
–20  
A
f
= –1dBFS  
A
f
= –1dBFS  
= 100.5MHz  
IN  
= 9.7MHz  
IN  
IN  
IN  
SNR = 77.9dBFS  
SINAD = 76.6dBc  
SFDR = 89.7dBc  
SNR = 76.6dBFS  
SINAD = 75.5dBc  
SFDR = 91.1dBc  
–20  
–40  
–40  
–60  
–60  
–80  
–80  
–100  
–120  
–140  
–100  
–120  
–140  
0
10  
20  
30  
40  
50  
60  
0
10  
20  
30  
40  
50  
60  
FREQUENCY (MHz)  
FREQUENCY (MHz)  
Figure 5. Single-Tone 32k FFT with fIN = 9.7 MHz, fSAMPLE = 125 MSPS,  
Figure 8. Single-Tone 32k FFT with fIN = 100.5 MHz, fSAMPLE = 125 MSPS, VREF = 1.0 V  
VREF = 1.0 V  
0
–20  
0
A
= –1dBFS  
IN  
= 19.7MHz  
A
f
= –1dBFS  
IN  
= 139.5MHz  
f
IN  
IN  
SNR = 78dBFS  
SINAD = 76.1dBc  
SFDR = 85.9dBc  
SNR = 75.7dBFS  
SINAD = 74.3dBc  
SFDR = 85.5dBc  
–20  
–40  
–40  
–60  
–60  
–80  
–80  
–100  
–120  
–140  
–100  
–120  
–140  
0
10  
20  
30  
40  
50  
60  
0
10  
20  
30  
40  
50  
60  
FREQUENCY (MHz)  
FREQUENCY (MHz)  
Figure 6. Single-Tone 32k FFT with fIN = 19.7 MHz, fSAMPLE = 125 MSPS, VREF = 1.0 V  
Figure 9. Single-Tone 32k FFT with fIN = 139.5 MHz, fSAMPLE = 125 MSPS, VREF = 1.0 V  
0
0
A
f
= –1dBFS  
A
f
= –1dBFS  
IN  
= 69.5MHz  
IN  
= 301MHz  
IN  
IN  
SNR = 77.5dBFS  
SINAD = 76.0dBc  
SFDR = 86.6dBc  
–20  
–40  
SNR = 71.2dBFS  
SINAD = 67dBc  
SFDR = 70.3dBc  
–20  
–40  
–60  
–60  
–80  
–80  
–100  
–120  
–140  
–100  
–120  
–140  
0
10  
20  
30  
40  
50  
60  
0
10  
20  
30  
40  
50  
60  
FREQUENCY (MHz)  
FREQUENCY (MHz)  
Figure 10. Single-Tone 32k FFT with fIN = 301 MHz, fSAMPLE = 125 MSPS, VREF = 1.0 V  
Figure 7. Single-Tone 32k FFT with fIN = 69.5 MHz, fSAMPLE = 125 MSPS, VREF = 1.0 V  
Rev. 0 | Page 12 of 37  
 
 
Data Sheet  
AD9655  
120  
100  
90  
80  
70  
60  
50  
40  
30  
20  
10  
0
SFDR (dBFS)  
SNR (dBFS)  
SFDR (dBc)  
SNR (dBFS)  
100  
80  
60  
SFDR (dBc)  
SNR (dB)  
40  
20  
0
–20  
–90  
–10  
0
100  
200  
300  
400  
500  
–70  
–50  
–30  
INPUT FREQUENCY (MHz)  
INPUT AMPLITUDE (dBFS)  
Figure 14. SNR, SFDR vs. Input Frequency (fIN); fSAMPLE = 125 MSPS, VREF = 1.0 V  
Figure 11. SNR, SFDR vs. Input Amplitude; fIN = 9.7 MHz, fSAMPLE = 125 MSPS,  
REF = 1.0 V  
V
95  
0
–20  
A
f
= –7dBFS  
= 100.1MHz, 102.1MHz  
IN  
IN  
SFDR (dBc)  
IMD2 = –89.6dBc  
IMD3 = –90.2dBc  
SFDR = 89.6dBc  
90  
–40  
85  
80  
–60  
2f1  
+
f2  
f1 + 2f2  
–80  
f1 + V  
2f2 f1  
2
f1 f2  
f1 f2  
–100  
–120  
–140  
SNR (dBFS)  
75  
70  
–40  
–20  
0
20  
40  
60  
80  
0
10  
20  
30  
40  
50  
60  
TEMPERATURE (°C)  
FREQUENCY (MHz)  
Figure 12. Two-Tone 32k FFT with fIN1 = 100.1 MHz and fIN2 = 102.1 MHz,  
SAMPLE = 125 MSPS, VREF = 1.0 V  
Figure 15. SNR, SFDR vs. Temperature; fIN = 9.7 MHz, fSAMPLE = 125 MSPS,  
f
VREF = 1.0 V  
0
5
4
–SFDR (dBc)  
IMD3 (dBc)  
–20  
3
–40  
–60  
2
1
0
–80  
–1  
–2  
–3  
–4  
–5  
–SFDR (dBFS)  
IMD3 (dBFS)  
–100  
–120  
–140  
–90  
–80  
–70  
–60  
–50  
–40  
–30  
–20  
–10  
0
10000  
20000  
30000  
40000  
50000  
60000  
INPUT AMPLITUDE (dBFS)  
OUTPUT CODE  
Figure 13. Two-Tone SFDR, IMD3 vs. Input Amplitude (AIN) with  
IN1 = 100.1 MHz and fIN2 = 102.1 MHz, fSAMPLE = 125 MSPS, VREF = 1.0 V  
Figure 16. INL; fIN = 9.7 MHz, fSAMPLE = 125 MSPS, VREF = 1.0 V  
f
Rev. 0 | Page 13 of 37  
AD9655  
Data Sheet  
0.8  
120  
100  
80  
60  
40  
20  
0
SFDR (dBc)  
SNR (dBFS)  
0.6  
0.4  
0.2  
0
–0.2  
–0.4  
–0.6  
–0.8  
0
10000  
20000  
30000  
40000  
50000  
60000  
60  
80  
20  
40  
100  
120  
OUTPUT CODE  
SAMPLE RATE (MSPS)  
Figure 19. SNR, SFDR vs. Sample Rate; fIN = 9.7 MHz, VREF = 1.0 V  
Figure 17. DNL; fIN = 9.7 MHz, fSAMPLE = 125 MSPS, VREF = 1.0 V  
120  
350k  
2.7LSB rms  
300k  
250k  
200k  
150k  
100k  
50k  
100  
80  
60  
40  
20  
0
SFDR (dBc)  
SNR (dBFS)  
0
60  
80  
20  
40  
100  
120  
SAMPLE RATE (MSPS)  
OUTPUT CODE  
Figure 20. SNR, SFDR vs. Sample Rate; fIN = 69.5 MHz, VREF = 1.0 V,  
Clock Divider = 4  
Figure 18. Input Referred Noise Histogram; fSAMPLE = 125 MSPS, VREF = 1.0 V  
Rev. 0 | Page 14 of 37  
Data Sheet  
AD9655  
VREF = 1.4 V  
0
0
–20  
A
f
= –1dBFS  
= 9.7MHz  
A
f
= –1dBFS  
IN  
IN  
= 100.5MHz  
IN  
IN  
SNR = 79.6dBFS  
SINAD = 78dBc  
SFDR = 88.1dBc  
SNR = 78dBFS  
SINAD = 76dBc  
SFDR = 83.5dBc  
–20  
–40  
–40  
–60  
–60  
–80  
–80  
–100  
–120  
–140  
–100  
–120  
–140  
0
10  
20  
30  
40  
50  
60  
0
10  
20  
30  
40  
50  
60  
FREQUENCY (MHz)  
FREQUENCY (MHz)  
Figure 21. Single-Tone 32k FFT with fIN = 9.7 MHz, fSAMPLE = 125 MSPS,  
Figure 24. Single-Tone 32k FFT with fIN = 100.5 MHz, fSAMPLE = 125 MSPS,  
VREF = 1.4 V  
VREF = 1.4 V  
0
–20  
0
–20  
A
f
= –1dBFS  
= 19.7MHz  
A
f
= –1dBFS  
IN  
IN  
= 139.5MHz  
IN  
IN  
SNR = 79.4dBFS  
SINAD = 77.2dBc  
SFDR = 85.1dBc  
SNR = 76.6dBFS  
SINAD = 74.8dBc  
SFDR = 82.4dBc  
–40  
–40  
–60  
–60  
–80  
–80  
–100  
–120  
–140  
–100  
–120  
–140  
0
10  
20  
30  
40  
50  
60  
0
10  
20  
30  
40  
50  
60  
FREQUENCY (MHz)  
FREQUENCY (MHz)  
Figure 22. Single-Tone 32k FFT with fIN = 19.7 MHz, fSAMPLE = 125 MSPS,  
Figure 25. Single-Tone 32k FFT with fIN = 139.5 MHz, fSAMPLE = 125 MSPS,  
VREF = 1.4 V  
VREF = 1.4 V  
0
–20  
0
–20  
A
f
= –1dBFS  
= 69.5MHz  
A
f
= –1dBFS  
IN  
IN  
= 301MHz  
IN  
IN  
SNR = 79.4dBFS  
SINAD = 76.6dBc  
SFDR = 83.7dBc  
SNR = 55dBFS  
SINAD = 53.8dBc  
SFDR = 71.6dBc  
–40  
–40  
–60  
–60  
–80  
–80  
–100  
–120  
–140  
–100  
–120  
–140  
0
10  
20  
30  
40  
50  
60  
0
10  
20  
30  
40  
50  
60  
FREQUENCY (MHz)  
FREQUENCY (MHz)  
Figure 23. Single-Tone 32k FFT with fIN = 69.5 MHz, fSAMPLE = 125 MSPS,  
REF = 1.4 V  
Figure 26. Single-Tone 32k FFT with fIN = 301 MHz, fSAMPLE = 125 MSPS,  
REF = 1.4 V  
V
V
Rev. 0 | Page 15 of 37  
 
AD9655  
Data Sheet  
120  
100  
80  
60  
40  
20  
0
100  
90  
80  
70  
60  
50  
40  
30  
20  
10  
0
SFDR (dBFS)  
SNR (dBFS)  
SFDR (dBc)  
SNR (dBFS)  
SFDR (dBc)  
SNR (dB)  
–20  
–90  
–70  
–50  
–30  
–10  
0
100  
200  
300  
400  
500  
INPUT AMPLITUDE (dBFS)  
INPUT FREQUENCY(MHz)  
Figure 27. SNR, SFDR vs. Input Amplitude; fIN = 9.7 MHz, fSAMPLE = 125 MSPS,  
REF = 1.4 V  
Figure 30. SNR, SFDR vs. Input Frequency (fIN); fSAMPLE = 125 MSPS, VREF = 1.4 V  
V
95  
0
A
f
= –7dBFS  
IN  
= 101.1MHz, 102.1MHz  
IN  
IMD2 = –88.36dBc  
IMD3 = –85dBc  
SFDR = 85dBc  
–20  
90  
–40  
SFDR (dBc)  
85  
–60  
2f1  
+
f2  
f1 + 2f2  
2f2 f1  
–80  
f1  
+
f2  
80  
2f1 f2  
SNR (dBFS)  
f1 f2  
–100  
–120  
–140  
75  
70  
–40  
0
10  
20  
30  
40  
50  
60  
–20  
0
20  
40  
60  
80  
FREQUENCY (MHz)  
TEMPERATURE (°C)  
Figure 28. Two-Tone 32k FFT with fIN1 = 100.1 MHz and fIN2 = 102.1 MHz,  
SAMPLE = 125 MSPS, VREF = 1.4 V  
Figure 31. SNR, SFDR vs. Temperature; fIN = 9.7 MHz, fSAMPLE = 125 MSPS,  
f
VREF = 1.4 V  
0
4
3
–20  
–SFDR (dBc)  
2
–40  
–60  
1
IMD3 (dBc)  
–SFDR (dBFS)  
IMD3 (dBFS)  
0
–80  
–1  
–2  
–3  
–4  
–100  
–120  
–140  
–90  
–80  
–70  
–60  
–50  
–40  
–30  
–20  
–10  
0
10000  
20000  
30000  
40000  
50000  
60000  
INPUT AMPLITUDE (dBFS)  
OUTPUT CODE  
Figure 29. Two-Tone SFDR, IMD3 vs. Input Amplitude (AIN) with  
IN1 = 100.1 MHz and fIN2 = 102.1 MHz, fSAMPLE = 125 MSPS, VREF = 1.4 V  
Figure 32. INL; fIN = 9.7 MHz, fSAMPLE = 125 MSPS, VREF = 1.4 V  
f
Rev. 0 | Page 16 of 37  
Data Sheet  
AD9655  
120  
100  
80  
60  
40  
20  
0
0.8  
SFDR (dBc)  
SNR (dBFS)  
0.6  
0.4  
0.2  
0
–0.2  
–0.4  
–0.6  
–0.8  
20  
40  
60  
80  
100  
120  
0
10000  
20000  
30000  
40000  
50000  
60000  
SAMPLE RATE (MSPS)  
OUTPUT CODE  
Figure 33. DNL; fIN = 9.7 MHz, fSAMPLE = 125 MSPS, VREF = 1.4 V  
Figure 35. SNR, SFDR vs. Sample Rate; fIN = 9.7 MHz, VREF = 1.4 V  
450k  
120  
2.0LSB rms  
400k  
350k  
300k  
250k  
200k  
150k  
100k  
50k  
100  
SFDR (dBc)  
80  
SNR (dBFS)  
60  
40  
20  
0
0
20  
40  
60  
80  
100  
120  
SAMPLE RATE (MSPS)  
OUTPUT CODE  
Figure 34. Input Referred Noise Histogram; fSAMPLE = 125 MSPS, VREF = 1.4 V  
Figure 36. SNR, SFDR vs. Sample Rate; fIN = 69.5 MHz, VREF = 1.4 V,  
Clock Divider = 4  
Rev. 0 | Page 17 of 37  
AD9655  
Data Sheet  
EQUIVALENT CIRCUITS  
DRVDD  
AVDD  
400  
VINx±  
SCLK/DFS  
30kΩ  
Figure 37. Equivalent Analog Input Circuit  
Figure 41. Equivalent SCLK/DFS Input Circuit  
AVDD  
10  
CLK+  
AVDD  
15kΩ  
15kΩ  
0.9V  
AVDD  
400  
RBIAS  
AND VCM  
10Ω  
CLK–  
Figure 42. Equivalent RBIAS and VCM Circuit  
Figure 38. Equivalent Clock Input Circuit  
DRVDD  
DRVDD  
400  
SDIO/PDWN  
15k  
31kΩ  
400Ω  
CSB  
Figure 39. Equivalent SDIO/PDWN Input Circuit  
Figure 43. Equivalent CSB Input Circuit  
DRVDD  
AVDD  
V
V
D0x–, D1x–  
D0x+, D1x+  
10  
400Ω  
V
V
VREF  
7.5kΩ  
Figure 40. Equivalent Digital Output Circuit  
Figure 44. Equivalent VREF Circuit  
Rev. 0 | Page 18 of 37  
 
 
Data Sheet  
AD9655  
THEORY OF OPERATION  
The AD9655 is a multistage, pipelined ADC. Each stage  
provides sufficient overlap to correct for flash errors in the  
preceding stage. The quantized outputs from each stage are  
combined into a final 16-bit result in the digital correction  
logic. The serializer transmits this converted data in a 16-bit  
output. The pipelined architecture permits the first stage to  
operate with a new input sample while the remaining stages  
operate with preceding samples. Sampling occurs on the rising  
edge of the clock.  
A small resistor in series with each input can help reduce the  
peak transient current injected from the output stage of the  
driving source. A differential capacitor, two single-ended  
capacitors, or a combination of these capacitors can be placed  
on the inputs to provide a matching passive network. This  
ultimately creates a low-pass filter at the input to limit unwanted  
broadband noise. See the AN-742 Application Note, the AN-827  
Application Note, and the Analog Dialogue article Transformer-  
Coupled Front-End for Wideband A/D Converters” (Volume 39,  
April 2005) for more information. In general, the precise values  
depend on the application.  
Each stage of the pipeline, excluding the last, consists of a low  
resolution flash ADC connected to a switched capacitor DAC  
and an interstage residue amplifier, for example, a multiplying  
digital-to-analog converter (MDAC). The residue amplifier  
magnifies the difference between the reconstructed DAC output  
and the flash input for the next stage in the pipeline. One bit of  
redundancy in each stage facilitates digital correction of flash  
errors. The last stage consists of a flash ADC.  
Input Common Mode  
The AD9655 analog inputs are not internally dc-biased.  
Therefore, in ac-coupled applications, the user must provide  
this bias externally. Setting the device so that VCM = AVDD/2 is  
recommended for optimum performance, but the device can  
function over a wider range with reasonable performance, as  
shown in Figure 46 and Figure 47.  
The output staging block aligns the data, corrects errors, and  
passes the data to the output buffers. Then, the data is serialized  
and aligned to the frame and data clocks.  
100  
SFDR (dBc)  
90  
ANALOG INPUT CONSIDERATIONS  
80  
SNR (dBFS)  
The analog input to the AD9655 is a differential switched  
capacitor circuit designed for processing differential input  
signals. This circuit can support a wide common-mode range  
while maintaining excellent performance. By using an input  
common-mode voltage of midsupply, users can minimize signal  
dependent errors and achieve optimum performance.  
70  
60  
50  
40  
30  
20  
10  
H
0.7  
0.8  
0.5  
0.6  
0.9  
(V)  
1.0  
1.1  
1.2  
C
PAR  
V
CM  
H
H
VINx+  
VINx–  
Figure 46. SNR, SFDR vs. Input Common-Mode Voltage (VCM), fIN = 9.7 MHz,  
fSAMPLE = 125 MSPS, VREF = 1.0 V  
C
SAMPLE  
S
S
S
S
100  
SFDR (dBc)  
C
SAMPLE  
90  
C
PAR  
80  
SNR (dBFS)  
H
70  
60  
50  
40  
30  
20  
10  
Figure 45. Switched Capacitor Input Circuit  
The clock signal alternately switches the input circuit between  
sample mode and hold mode (see Figure 45). When the input  
circuit switches to sample mode, the signal source must be  
capable of charging the sample capacitors and settling within  
one half of a clock cycle.  
0.70  
0.75  
0.80  
0.85  
(V)  
0.90  
0.95  
1.00  
V
CM  
Figure 47. SNR, SFDR vs. Input Common-Mode Voltage (VCM), fIN = 9.7 MHz,  
fSAMPLE = 125 MSPS, VREF = 1.4 V  
Rev. 0 | Page 19 of 37  
 
 
 
 
 
AD9655  
Data Sheet  
An on-chip, common-mode dc voltage reference is included in  
the design and is available from the VCM pin. The VCM pin  
must be bypassed to ground by a 0.1 µF capacitor, as described  
in the Applications Information section. VCM error vs. load  
current is shown in Figure 48.  
Figure 52 show the typical drift characteristics of the internal  
reference.  
The internal buffer generates the positive and negative full-scale  
references for the ADC core. A digital reset using Register 0x08  
must follow any programmatic change in internal analog VREF  
.
0
0
–2  
–4  
–10  
–20  
–30  
–40  
–50  
–6  
–8  
–10  
–12  
–14  
–16  
–18  
INTERNAL V  
2.0  
= 1V  
REF  
–60  
0
0.5  
1.0  
1.5  
2.0  
2.5  
1.0  
0
0.5  
1.5  
2.5  
LOAD CURRENT (mA)  
LOAD CURRENT (mA)  
Figure 48. VCM Error vs. Load Current  
Figure 49. VREF Error vs. Load Current, VREF = 1.0 V  
Differential Input Configurations  
0
–10  
–20  
–30  
–40  
–50  
–60  
–70  
There are several ways to drive the AD9655 either actively or  
passively. However, optimum performance is achieved by  
driving the analog inputs differentially. Using a differential  
double balun configuration to drive the AD9655 provides  
excellent performance and a flexible interface to the ADC for  
baseband applications (see Figure 53).  
INTERNAL V  
= 1.4V  
REF  
For applications where SNR is a key parameter, differential  
transformer coupling is the recommended input configuration (see  
Figure 54) because the noise performance of most amplifiers is not  
adequate to achieve the true performance of the AD9655.  
Regardless of the configuration, the value of the shunt capacitor,  
C, is dependent on the input frequency and may need to be  
reduced or removed.  
1.0  
LOAD CURRENT (mA)  
0
0.5  
1.5  
2.0  
2.5  
Figure 50. VREF Error vs. Load Current, VREF = 1.4 V  
It is not recommended to drive the AD9655 inputs single-ended.  
0.004  
0.002  
0
VOLTAGE REFERENCE  
A stable and accurate 1.0 V to 1.4 V dc voltage reference is built  
into the AD9655. Externally bypass the VREF pin to ground  
with a low ESR, 1.0 μF capacitor in parallel with a low ESR,  
0.1 μF ceramic capacitor.  
–0.002  
–0.004  
–0.006  
–0.008  
Maximum SNR performance is achieved by setting the ADC to  
the largest span in a differential configuration. In the case of the  
AD9655, the largest input span available is 2.8 V p-p, which is  
achieved by setting VREF to 1.4 V.  
INTERNAL V  
REF  
= 1V  
80  
If the internal reference of the AD9655 is used to drive multiple  
converters to improve gain matching, the loading of the reference  
by the other converters must be considered. Figure 49 and  
Figure 50 show how the internal reference voltage is affected by  
loading. It is recommended that VREF not be used to drive the  
reference voltage of other devices at 1.4 V. Figure 51 and  
0
–40  
–20  
20  
40  
60  
TEMPERATURE (°C)  
Figure 51. Typical VREF Drift, VREF = 1.0 V  
Rev. 0 | Page 20 of 37  
 
 
 
 
 
Data Sheet  
AD9655  
0.003  
0.002  
0.001  
0
–0.001  
–0.002  
–0.003  
–0.004  
–0.005  
–0.006  
–0.007  
INTERNAL V  
= 1.4V  
80  
REF  
–40  
–20  
0
20  
40  
60  
TEMPERATURE (°C)  
Figure 52. Typical VREF Drift, VREF = 1.4 V  
1
0.1µF  
C
R
C1  
0.1µF  
VINx+  
33  
33Ω  
33Ω  
2V p-p  
C
10pF  
ADC  
0.1µF  
C
R
VINx–  
VCM  
33Ω  
ET1-1-I3  
1
C1  
R
200Ω  
0.1µF  
C
0.1µF  
1
C1 IS OPTIONAL  
Figure 53. Differential Double Balun Input Configuration for Baseband Applications  
ADT1-1WT  
1:1 Z RATIO  
1
C1  
R
VINx+  
VINx–  
33  
2Vp-p  
49.9ꢀ  
ADC  
C
10pF  
R
VCM  
33ꢀ  
1
C1  
200ꢀ  
0.1µF  
0.1μF  
1
C1 IS OPTIONAL  
Figure 54. Differential Transformer Coupled Configuration for Baseband Applications  
Rev. 0 | Page 21 of 37  
 
 
 
AD9655  
Data Sheet  
If a low jitter clock source is not available, another option is to  
ac couple a differential PECL signal to the sample clock input  
pins, as shown in Figure 57. The AD9510/AD9511/AD9512/  
AD9513/AD9514/AD9515/AD9516/AD9517 clock drivers, noted  
as AD951x in Figure 57, Figure 58, and Figure 59, offer excellent  
jitter performance.  
CLOCK INPUT CONSIDERATIONS  
For optimum performance, clock the AD9655 sample clock inputs,  
CLK+ and CLK−, with a differential signal. The signal is typically  
ac-coupled into the CLK+ and CLK− pins via a transformer or  
capacitors. These pins are biased internally (see Figure 38) and  
require no external bias.  
Clock Input Options  
0.1µF  
0.1µF  
CLOCK  
INPUT  
The AD9655 has a flexible clock input structure. The clock  
input can be a CMOS, LVDS, LVPECL, or sine wave signal.  
Regardless of the type of signal used, clock source jitter is an  
important consideration, as described in the Jitter  
Considerations section.  
CLK+  
AD951x  
PECL DRIVER  
100Ω  
ADC  
0.1µF  
0.1µF  
CLOCK  
INPUT  
CLK–  
240Ω  
240Ω  
50kΩ  
50kΩ  
Figure 57. Differential PECL Sample Clock (Up to 1 GHz)  
Figure 55 and Figure 56 show two preferred methods for  
clocking the AD9655 at clock rates up to 1 GHz prior to the  
internal clock divider. A low jitter clock source is converted  
from a single-ended signal to a differential signal using either a  
radio frequency (RF) transformer or an RF balun.  
A third option is to ac couple a differential LVDS signal to the  
sample clock input pins, as shown in Figure 58. The AD9510/  
AD9511/AD9512/AD9513/AD9514/AD9515/AD9516/AD9517  
clock drivers offer excellent jitter performance.  
®
Mini-Circuits  
ADT1-1WT, 1:1 Z  
0.1µF  
0.1µF  
0.1µF  
0.1µF  
CLOCK  
INPUT  
CLK+  
XFMR  
CLOCK  
INPUT  
CLK+  
AD951x  
LVDS DRIVER  
100Ω  
100Ω  
ADC  
50Ω  
ADC  
0.1µF  
0.1µF  
0.1µF  
CLOCK  
INPUT  
CLK–  
CLK–  
SCHOTTKY  
DIODES:  
HSMS2822  
50kΩ  
50kΩ  
0.1µF  
Figure 58. Differential LVDS Sample Clock (Up to 1 GHz)  
Figure 55. Transformer Coupled Differential Clock (Up to 200 MHz)  
In some applications, it may be acceptable to drive the sample  
clock inputs with a single-ended 1.8 V CMOS signal. In such  
applications, drive the CLK+ pin directly from a CMOS gate,  
and bypass the CLK− pin to ground with a 0.1 μF capacitor  
(see Figure 59).  
0.1µF  
0.1µF  
0.1µF  
CLOCK  
INPUT  
CLK+  
50Ω  
ADC  
0.1µF  
CLK–  
SCHOTTKY  
DIODES:  
HSMS2822  
V
CC  
OPTIONAL  
100Ω  
0.1µF  
1
0.1µF  
1kΩ  
1kΩ  
AD951x  
CMOS DRIVER  
CLOCK  
INPUT  
CLK+  
Figure 56. Balun-Coupled Differential Clock (Up to 1 GHz)  
50Ω  
ADC  
The RF balun configuration is recommended for clock  
frequencies between 125 MHz and 1 GHz, and the RF  
transformer configuration is recommended for clock  
frequencies from 20 MHz to 200 MHz. The antiparallel  
Schottky diodes across the transformer/balun secondary  
winding limit clock excursions into the AD9655 to  
approximately 0.8 V p-p differential.  
CLK–  
0.1µF  
1
50Ω RESISTOR IS OPTIONAL.  
Figure 59. Single-Ended 1.8 V CMOS Input Clock (Up to 200 MHz)  
Input Clock Divider  
The AD9655 contains an input clock divider that can divide the  
input clock by integer values from 1 to 8. To achieve a given  
sample rate, the frequency of the externally applied clock must  
be multiplied by the divide value. The increased rate of the  
external clock normally results in lower clock jitter, which is  
beneficial for intermediate frequency (IF) undersampling  
applications.  
This limit helps prevent the large voltage swings of the clock  
from feeding through to other portions of the AD9655 while  
preserving the fast rise and fall times of the signal that are  
critical to achieving low jitter performance. However, the diode  
capacitance comes into play at frequencies above 500 MHz.  
Care must be taken when choosing the appropriate signal  
limiting diode.  
Rev. 0 | Page 22 of 37  
 
 
 
 
 
 
Data Sheet  
AD9655  
Clock Duty Cycle  
Treat the clock input as an analog signal when aperture jitter  
may affect the dynamic range of the AD9655. Separate clock  
driver power supplies from the ADC output driver supplies to  
avoid modulating the clock signal with digital noise. Low jitter,  
crystal controlled oscillators are the best clock sources. If the  
clock is generated from another type of source (by gating,  
dividing, or other methods), it is recommended that the clock  
be retimed by the original clock as the last step.  
Typical high speed ADCs use both clock edges to generate a  
variety of internal timing signals and, as a result, may be sensitive  
to the clock duty cycle. Commonly, a 5% tolerance is required  
on the clock duty cycle to maintain dynamic performance  
characteristics.  
The AD9655 contains a duty cycle stabilizer (DCS) that retimes  
the nonsampling (falling) edge, providing an internal clock  
signal with a nominal 50% duty cycle. This allows the user to  
provide a wide range of clock input duty cycles without affecting  
the performance of the AD9655. Noise and distortion performance  
are nearly flat for a wide range of duty cycles with the DCS on.  
See the AN-501 Application Note and the AN-756 Application  
Note for more information about jitter performance as it relates  
to ADCs.  
POWER DISSIPATION AND POWER-DOWN MODE  
Jitter in the rising edge of the clock is still of concern and is not  
easily reduced by the internal stabilization circuit. The duty  
cycle control loop does not function for clock rates of less than  
20 MHz, nominally. The loop has a time constant associated  
with it that must be considered in applications in which the  
clock rate can change dynamically. A wait time of 1.5 µs to 5 µs  
is required after a dynamic clock frequency increase or decrease  
before the DCS loop is relocked to the input signal.  
As shown in Figure 61, the power dissipated by the AD9655 is  
proportional to its sample rate. The AD9655 is placed in power-  
down mode either by the SPI port or by asserting the SDIO/  
PDWN pin high when in non-SPI mode. In this state, the ADC  
typically dissipates 2 m W. D u ri ng power-down, the output drivers  
are placed in a high impedance state. In non-SPI mode, asserting  
the SDIO/PDWN pin low returns the AD9655 to its normal  
operating mode. Note that SDIO/PDWN is referenced to the  
digital output driver supply (DRVDD) and must not exceed that  
supply voltage.  
Jitter Considerations  
High speed, high resolution ADCs are sensitive to the quality of  
the clock input. The following equation shows how SNR  
degrades at a given input frequency (fA) due only to aperture  
jitter (tJ):  
0.32  
0.30  
0.28  
0.26  
1
SNR Degradation = 20 log  
10   
V
= 1.4V  
REF  
2π× fA ×tJ  
0.24  
0.22  
0.20  
0.18  
0.16  
V
= 1.0V  
REF  
In this equation, the rms aperture jitter represents the root sum  
square of all jitter sources, including the clock input, analog  
input signal, and ADC aperture jitter specifications. IF under-  
sampling applications are particularly sensitive to jitter. The  
effect of jitter alone on SNR, with no other noise contributors,  
is shown in Figure 60.  
20  
40  
60  
80  
100  
120  
SAMPLE RATE (MSPS)  
130  
RMS CLOCK JITTER REQUIREMENT  
Figure 61. Total Power Dissipation vs. Sample Rate (fSAMPLE) for fIN = 9.7 MHz,  
REF = 1.0 V and VREF = 1.4 V  
120  
110  
V
The AD9655 achieves low power dissipation in power-down  
mode by shutting down the reference, reference buffer, biasing  
networks, and clock. When using the SPI port interface, the  
user can place the ADC in power-down mode or standby mode.  
Standby mode allows the user to keep the internal reference  
circuitry powered when faster wake-up times are required. Do  
not invoke standby mode while foreground calibration is in  
progress. Foreground calibration is invoked automatically at  
power-up, and by executing a digital reset with Register 0x08.  
Completion is indicated by the contents of Register 0x107 =  
0x00. See the Memory Map section for more details on using  
these features.  
16 BITS  
100  
90  
80  
70  
60  
50  
40  
30  
14 BITS  
12 BITS  
10 BITS  
8 BITS  
0.125ps  
0.25ps  
0.5ps  
1.0ps  
2.0ps  
1
10  
100  
1000  
ANALOG INPUT FREQUENCY (MHz)  
Figure 60. Ideal SNR vs. Analog Input Frequency and Jitter  
Rev. 0 | Page 23 of 37  
 
 
 
 
AD9655  
Data Sheet  
DIGITAL OUTPUTS AND TIMING  
The AD9655 differential outputs conform to the ANSI-644  
LVDS standard on default power-up. This default setting can be  
changed to a low power, reduced signal option (similar to the  
IEEE 1596.3 standard) via the SPI. The LVDS driver current is  
derived on chip and sets the output current at each output equal  
to a nominal 3.5 mA. A 100 Ω differential termination resistor  
placed at the LVDS receiver inputs results in a nominal 350 mV  
swing (or 700 mV p-p differential) at the receiver.  
When operating in reduced range mode, the output current  
reduces to 2 mA. This results in a 200 mV swing (or 400 mV p-p  
differential) across a 100 Ω termination at the receiver.  
D0 500mV/DIV  
D1 500mV/DIV  
DCO 500mV/DIV  
FCO 500mV/DIV  
4ns/DIV  
The LVDS outputs facilitate interfacing with LVDS receivers in  
custom ASICs and FPGAs for superior switching performance  
in noisy environments. Single point-to-point net topologies are  
recommended with a 100 Ω termination resistor placed as close  
as possible to the receiver. Timing errors may result if there is  
no far-end receiver termination, or if there is poor differential  
trace routing. To avoid such timing errors, ensure that the trace  
length is less than 24 inches and that the differential output  
traces are close together and at equal lengths.  
Figure 62. LVDS Output Timing Example in ANSI-644 Mode (Default)  
Figure 62 shows an example of the FCO and data stream with  
proper trace length and position.  
Figure 63 shows the LVDS output timing example in reduced  
range mode.  
D0 400mV/DIV  
D1 400mV/DIV  
DCO 400mV/DIV  
FCO 400mV/DIV  
4ns/DIV  
Figure 63. LVDS Output Timing Example in Reduced Range Mode  
Rev. 0 | Page 24 of 37  
 
 
 
Data Sheet  
AD9655  
Figure 64. Data Eye for LVDS Outputs in ANSI-644 Mode with Trace Lengths  
of Less Than 24 Inches (Approximate 6 Inch Trace Length Result Shown) on  
Standard FR-4 Material, External 100 Ω Far-End Termination Only  
Figure 66. Data Eye for LVDS Outputs in ANSI-644 Mode with Trace Lengths  
Greater Than 24 Inches (Approximate 36 Inch Trace Length Result Shown) on  
Standard FR-4 Material, External 100 Ω Far-End Termination Only  
Figure 67. TIE Jitter Histogram for Trace Lengths Greater Than 24 Inches  
(Approximate 36 Inch Trace Length Result Shown)  
Figure 65. TIE Jitter Histogram for Trace Lengths Less Than 24 Inches  
(Approximate 6 Inch Trace Length Result Shown)  
Two output clocks assist in capturing data from the AD9655.  
The DCO clocks the output data and is equal to 4× the sample  
clock (CLK) rate for the default mode of operation. Data is  
clocked out of the AD9655 and must be captured on the rising  
and falling edges of the DCO that supports DDR capturing. The  
FCO signals the start of a new output byte and is equal to the  
sample clock rate in 1× frame mode. See the Timing Diagrams  
section for more information.  
Figure 64 shows an example of the LVDS output data eye using  
the ANSI-644 standard (default), and Figure 65 shows a time  
interval error (TIE) jitter histogram with trace lengths of less  
than 24 inches on standard FR-4 material.  
Figure 66 shows an example of the LVDS output data eye using  
the ANSI-644 standard (default), and Figure 67 shows a time  
interval error (TIE) jitter histogram with trace lengths greater  
than 24 inches on standard FR-4 material. Note that the TIE  
jitter histogram reflects the decrease of the data eye opening as  
the edge deviates from the ideal position. It is the responsibility  
of the user to determine if the waveforms meet the timing budget  
of the design.  
When the SPI is used, the DCO phase can be adjusted in  
approximately 60° increments relative to one data cycle (30°  
relative to one DCO cycle). This allows the user to refine system  
timing margins if required. The example DCO+ and DCO−  
timing, as shown in Figure 2, is 180° relative to one data cycle  
(90° relative to one DCO cycle).  
The format of the output data is twos complement by default.  
See Table 11 for an example of the output coding format. To  
change the output data format to offset binary, see the  
Memory Map section.  
In default mode, as shown in Figure 2, the MSB is first in the  
data output serial stream. This can be inverted by using the SPI  
so that the LSB is first in the data output serial stream.  
Data from each ADC is serialized and provided on a separate  
channel in two lanes in DDR mode. The data rate for each serial  
stream is equal to (16 bits × the sample clock rate)/2 lanes, with  
a maximum of 1 Gbps/lane (16 bits × 125 MSPS)/(2 lanes) =  
1 Gbps/lane). The minimum conversion rate is 20 MSPS.  
Rev. 0 | Page 25 of 37  
 
 
 
 
AD9655  
Data Sheet  
Table 11. Digital Output Coding  
Input (V)  
Condition (V)  
<−VREF − 0.5 LSB  
− VREF  
Offset Binary Output Mode  
0000 0000 0000 0000  
0000 0000 0000 0000  
1000 0000 0000 0000  
1111 1111 1111 1111  
1111 1111 1111 1111  
Twos Complement Mode  
VINx+ − VINx−  
VINx+ − VINx−  
VINx+ − VINx−  
VINx+ − VINx−  
VINx+ − VINx−  
1000 0000 0000 0000  
1000 0000 0000 0000  
0000 0000 0000 0000  
0111 1111 1111 1111  
0111 1111 1111 1111  
0 V  
+ VREF − 1.0 LSB  
>+ VREF − 0.5 LSB  
Table 12. Flexible Output Test Modes  
Output Test Mode  
Bit Sequence  
Digital Output  
Word 1  
Digital Output  
Word 2  
Subject to Data  
Pattern Name  
Off (default)  
Format Select  
Not applicable  
Yes  
Notes  
0000  
Not applicable  
Not applicable  
Not applicable  
0001  
Midscale short  
12-bit: 1000 0000  
0000  
Offset binary code shown  
16-bit: 1000 0000  
0000 0000  
0010  
0011  
0100  
+Full-scale short  
−Full-scale short  
Checkerboard  
12-bit: 1111 1111  
1111  
16-bit: 1111 1111  
1111 1111  
Not applicable  
Not applicable  
Yes  
Yes  
No  
Offset binary code shown  
Offset binary code shown  
12-bit: 0000 0000  
0000  
16-bit: 0000 0000  
0000 0000  
12-bit: 1010 1010  
1010  
12-bit: 0101 0101  
0101  
16-bit: 1010 1010  
1010 1010  
16-bit: 0101 0101  
0101 0101  
0101  
0110  
0111  
PN sequence long1  
PN sequence short1  
Not applicable  
Not applicable  
Not applicable  
Not applicable  
Yes  
Yes  
No  
PN23  
ITU 0.150  
x23 + x18 + 1  
PN9  
ITU 0.150  
x9 + x5 + 1  
One-/zero-word toggle 12-bit: 1111 1111  
1111  
12-bit: 0000 0000  
0000  
16-bit: 111 1111  
1111 1111  
16-bit: 0000 0000  
0000 0000  
1000  
1001  
User input  
Register 0x19 and  
Register 0x1A  
Register 0x1B and  
Register 0x1C  
No  
No  
1-/0-bit toggle  
12-bit: 1010 1010  
1010  
16-bit: 1010 1010  
1010 1010  
Not applicable  
Not applicable  
Not applicable  
Not applicable  
1010  
1011  
1100  
1× sync  
12-bit: 0000 0000  
1111  
16-bit: 0000 0000  
1111 1111  
No  
No  
No  
One bit high  
Mixed frequency  
12-bit: 1000 0000  
0000  
16-bit: 1000 0000  
0000 0000  
Pattern associated with  
the external pin  
12-bit: 1010 0001  
1001  
16-bit: 1010 0001  
1001 1100  
1 All test mode options except PN sequence short and PN sequence long can support 12-bit to 16-bit word lengths to verify data capture to the receiver.  
Rev. 0 | Page 26 of 37  
 
 
Data Sheet  
AD9655  
There are 12 digital output test pattern options available that  
can be initiated through the SPI. This is a useful feature when  
validating receiver capture and timing. See Table 12 for the  
available output bit sequencing options. Some test patterns have  
two serial sequential words and can be alternated in various  
ways, depending on the test pattern chosen.  
Table 14. Power-Down Mode Pin Settings  
SDIO/PDWN Pin Voltage  
AGND (Default)  
DRVDD  
Device Mode  
Run device, normal operation  
Power down device  
SCLK/DFS Pin  
Note that some patterns do not adhere to the data format select  
option. In addition, custom user defined test patterns can be  
assigned in Register 0x19, Register 0x1A, Register 0x1B, and  
Register 0x1C.  
The SCLK/DFS pin is used for output format selection in  
applications that do not require SPI mode operation. This pin  
determines the digital output format when the CSB pin is held high  
during device power-up. When SCLK/DFS is tied to DRVDD, the  
ADC output format is twos complement; when SCLK/DFS is  
tied to AGND, the ADC output format is offset binary.  
The pseudorandom number (PN) sequence short pattern  
produces a pseudorandom bit sequence that repeats itself every  
29 − 1 or 511 bits. A description of the PN sequence and how it  
is generated can be found in Section 5.1 of the ITU-T 0.150  
(05/96) standard. The seed value is all 1s (see Table 13 for the  
initial values). The output is a parallel representation of the  
serial PN9 sequence in MSB first format. The first output word  
is the first 16 bits of the PN9 sequence in MSB aligned form.  
Table 15. Digital Output Format  
SCLK/DFS Voltage  
Output Format  
AGND  
Offset binary  
DRVDD  
Twos complement  
CSB Pin  
Table 13. PN Sequence  
Tie the CSB pin to DRVDD for applications that do not require  
SPI mode operation. By tying CSB high, all SCLK and SDIO  
information is ignored.  
Initial  
Value  
Next Three Output Samples  
(MSB First), Twos Complement  
Sequence  
PN Sequence Short 0x7F83 0x5F17, 0xB209, 0xCED1  
PN Sequence Long 0x7FFF 0x7E00, 0x807C, 0x801F  
RBIAS Pin  
To set the internal core bias current of the ADC, place a  
10.0 kΩ, 1% tolerance resistor to ground at the RBIAS pin.  
The PN sequence long pattern produces a pseudorandom bit  
sequence that repeats itself every 223 − 1 or 8,388,607 bits. A  
description of the PN sequence and how it is generated can be  
found in Section 5.6 of the ITU-T 0.150 (05/96) standard. The  
seed value is all 1s (see Table 13 for the initial values) and the  
AD9655 inverts the bit stream in relation to the ITU standard.  
The output is a parallel representation of the serial PN23  
sequence in MSB first format. The first output word is the first  
16 bits of the PN23 sequence in MSB aligned form.  
OUTPUT TEST MODES  
The output test options are described in Table 12 and are  
controlled by the output test mode bits at Register 0x0D. When  
an output test mode is enabled, the analog section of the ADC is  
disconnected from the digital back-end blocks and the test  
pattern is run through the output formatting block. Some of the  
test patterns are subject to output formatting, and some are not.  
The PN generators from the PN sequence tests can be reset by  
setting Bit 4 or Bit 5 of Register 0x0D. These tests can be  
performed with or without an analog signal (if present, the  
analog signal is ignored), but they do require an encode clock.  
For general information, see the AN-877 Application Note,  
Interfacing to High Speed ADCs via SPI.  
See the Memory Map section for information on how to change  
these additional digital output timing features through the SPI.  
SDIO/PDWN Pin  
For applications that do not require SPI mode operation, the  
CSB pin is tied to DRVDD, and the SDIO/PDWN pin controls  
the power-down mode according to Table 14.  
Rev. 0 | Page 27 of 37  
 
 
 
 
AD9655  
Data Sheet  
SERIAL PORT INTERFACE (SPI)  
The AD9655 SPI allows the user to configure the converter for  
specific functions or operations through a structured register  
space provided inside the ADC. The SPI offers the user added  
flexibility and customization, depending on the application.  
Addresses are accessed via the serial port and can be written to  
or read from via the port. Memory is organized into bytes that  
can be further divided into fields, which are documented in the  
Memory Map section. Information specific to the AD9655 is  
contained in this data sheet, and takes precedence over the the  
AN-877 Application Note, Interfacing to High Speed ADCs via SPI,  
which provides general information.  
The falling edge of CSB, in conjunction with the rising edge of  
SCLK/DFS, determines the start of the framing. An example of  
the serial timing is shown in Figure 68. See Table 7 for definitions  
of the timing parameters.  
Other modes involving the CSB pin are available. CSB can be  
held low indefinitely, which permanently enables SPI mode; this  
is called streaming. CSB can stall high between bytes to allow  
for additional external timing. When the CSB pin is tied high at  
power-up, SPI functions are placed in high impedance mode.  
This mode turns on the secondary functions of the SPI pins.  
Note that, when SPI mode is entered, that is, CSB is taken low,  
these secondary functions cannot be invoked without power  
cycling the device.  
CONFIGURATION USING THE SPI  
Three pins define the SPI of this ADC: the SCLK/DFS pin, the  
SDIO/PDWN pin, and the CSB pin (see Table 16). SCLK/DFS  
(a serial clock when CSB is low) synchronizes the read and  
write data presented from and to the ADC. SDIO/PDWN  
(serial data input/output when CSB is low) is a dual-purpose  
pin that allows data to be sent to and read from the internal  
ADC memory map registers. CSB (chip select bar) is an active  
low control that enables or disables the SPI read and write cycles.  
During the instruction phase of an SPI operation, a 16-bit  
instruction is transmitted. Data follows the instruction phase,  
and the length of this data is determined by the W0 and W1 bits  
(see Figure 68).  
In addition to word length, the instruction phase determines  
whether the serial frame is a read or write operation, allowing  
the serial port to be used both to program the chip and to read  
the contents of the on-chip memory. The first bit of the first  
byte in a multibyte serial data transfer frame indicates whether a  
read command or a write command is issued. If the instruction  
is a readback operation, performing a readback causes the serial  
data input/output SDIO/PDWN pin to change direction from  
an input to an output at the appropriate point in the serial  
frame.  
Table 16. Serial Port Interface Pins  
Pin  
Function  
SCLK/DFS  
Serial clock when CSB is low. The serial shift clock  
input, which synchronizes serial interface reads  
and writes.  
SDIO/PDWN Serial data input/output when CSB is low. A dual-  
purpose pin that typically serves as an input or an  
output, depending on the instruction sent and  
the relative position in the timing frame.  
All data is composed of 8-bit words. Data can be sent in MSB  
first mode or in LSB first mode. MSB first mode is the default  
on power-up and can be changed via the SPI port configuration  
register. For more information about this and other features,  
see the AN-877 Application Note, Interfacing to High Speed  
ADCs via SPI.  
CSB  
Chip select bar. An active low control that enables  
the SPI mode read and write cycles.  
tHIGH  
tDS  
tCLK  
tH  
tS  
tDH  
tLOW  
CSB  
SCLK/DFS DON’T CARE  
SDIO/PDWN DON’T CARE  
DON’T CARE  
R/W  
W1  
W0  
A12  
A11  
A10  
A9  
A8  
A7  
D5  
D4  
D3  
D2  
D1  
D0  
DON’T CARE  
Figure 68. Serial Port Interface Timing Diagram  
Rev. 0 | Page 28 of 37  
 
 
 
 
Data Sheet  
AD9655  
HARDWARE INTERFACE  
CONFIGURATION WITHOUT THE SPI  
The pins described in Table 16 comprise the physical interface  
between the user programming device and the serial port of the  
AD9655. The SCLK/DFS pin and the CSB pin function as inputs  
when using the SPI interface. The SDIO/PDWN pin is bidirec-  
tional, functioning as an input during write phases and as an  
output during readback.  
In applications that do not interface to the SPI control registers,  
the SCLK/DFS pin and the SDIO/PDWN pin serve as standalone  
CMOS-compatible control pins. When the device is powered  
up, it is assumed that the user intends to use the pins as static  
control lines for the output data format and power-down  
feature control. In this mode, connect CSB to DRVDD, which  
disables the serial port interface.  
The SPI interface is flexible enough to be controlled by either  
FPGAs or microcontrollers. One method for SPI configuration  
is described in detail in the AN-812 Application Note, Micro-  
controller-Based Serial Port Interface (SPI) Boot Circuit.  
SPI ACCESSIBLE FEATURES  
Table 17 provides a brief description of the general features  
accessible via the SPI. These features are described in general in  
the AN-877 Application Note, Interfacing to High Speed ADCs  
via SPI. The AD9655 device-specific features are described in  
detail following Table 18, the external memory map register table.  
It is recommended that the SPI port not be active during  
periods when the full dynamic performance of the converter is  
required. Because the SCLK/DFS signal, the CSB signal, and the  
SDIO/PDWN signal are typically asynchronous to the ADC clock,  
noise from these signals can degrade converter performance. If the  
on-board SPI bus is used for other devices, it may be necessary to  
provide buffers between this bus and the AD9655 to prevent these  
signals from transitioning at the converter inputs during critical  
sampling periods.  
Table 17. Features Accessible Using the SPI  
Feature Name  
Description  
Power Mode  
Allows the user to set either power-down  
mode or standby mode  
Clock  
Allows the user to access the DCS, set the  
clock divider, and set the clock divider phase  
The SCLK/DFS and SDIO/PDWN pins serve a dual function  
when the SPI interface is not in use. When the pins are strapped  
to DRVDD or ground during device power-on, they are  
associated with a specific function. Table 14 and Table 15  
describe the strappable functions supported on the AD9655.  
Offset  
Test I/O  
Allows the user to digitally adjust the converter  
offset  
Allows the user to set test modes to have  
known data on output bits  
Allows the user to set the output mode  
Allows the user to set the output clock polarity  
Output Mode  
Output Phase  
ADC Resolution  
Allows power consumption scaling with  
respect to sample rate  
Rev. 0 | Page 29 of 37  
 
 
 
 
AD9655  
Data Sheet  
MEMORY MAP  
Default Values  
READING THE MEMORY MAP REGISTER TABLE  
After the AD9655 is soft reset by Register 0x00, critical registers  
are loaded with default values. The default values for the  
registers are given in the memory map register table, Table 18.  
Each row in the memory map register table (see Table 18) has  
eight bit locations. The memory map is roughly divided into  
three sections: the chip configuration registers (Address 0x00 to  
Address 0x02); the device index register (Address 0x05); and the  
global ADC function registers, including setup, control, and test  
registers (Address 0x08 and beyond).  
Logic Levels  
An explanation of logic level terminology is as follows:  
“Bit is set” is synonymous with “bit is set to Logic 1” or  
“writing Logic 1 for the bit.”  
The memory map register table lists the default hexadecimal  
value for each hexadecimal address shown. The column with  
the heading Bit 7 (MSB) contains the most significant bit of the  
default hexadecimal value given. For example, Address 0x05,  
the device index register, has a hexadecimal default value of  
0x33. This means that in Address 0x05, Bits[7:6] = 00, Bits[5:4]  
= 11, Bits[3:2] = 00, and Bits[1:0] = 11 (in binary). This setting  
is the default channel index setting. The default value results in  
both ADC channels receiving the next write command. For  
more information on this function and others, see the AN-877  
Application Note, Interfacing to High Speed ADCs via SPI. This  
application note documents the functions controlled by  
Register 0x00 to Register 0xFF. Specific register functions for  
the AD9655 are documented in the Memory Map Register  
Descriptions section.  
“Clear a bit” is synonymous with “bit is set to Logic 0” or  
“writing Logic 0 for the bit.”  
Channel Specific Registers  
Some channel setup functions can be programmed individually  
for each channel. In these cases, channel address locations are  
internally duplicated for each channel. These registers and bits  
are designated in Table 18 as local. These local registers and bits  
can be accessed by setting the appropriate data channel bits  
(Channel A or Channel B) and the clock channel DCO bit (Bit 5)  
and clock channel FCO bit (Bit 4) in Register 0x05. If all the bits  
are set, the subsequent write affects the registers of both channels  
and the DCO/FCO clock channels. In a read cycle, only set one  
channel (Channel A or Channel B) to read one of the two registers.  
If all the bits are set during an SPI read cycle, the device returns the  
value for Channel A. Registers and bits that are designated as global  
in Table 18 affect the entire device and channel features for which  
independent settings are not allowed between channels. The set-  
tings in Register 0x05 do not affect the global registers and bits.  
Open Locations  
All address and bit locations not included in Table 18 are not  
currently supported for this device. Write unused bits of a valid  
address location with 0s. Writing to these locations is required  
only when part of an address location is open (for example,  
Address 0x05). If the entire address location is open or not listed  
in Table 18 (for example, Address 0x13), this address location must  
not be written.  
Rev. 0 | Page 30 of 37  
 
 
Data Sheet  
AD9655  
When Bit 5 in Register 0x00 is set high, the SPI enters a soft  
reset, where all of the user registers revert to their default values  
and Bit 2 is automatically cleared.  
MEMORY MAP REGISTER TABLE  
The AD9655 uses a 3-wire interface and 16-bit addressing;  
therefore, Bit 0 and Bit 7 in Register 0x00 are set to 0, and Bit 3  
and Bit 4 are set to 1.  
Table 18.  
Default  
Addr.  
(Hex)  
Bit 7  
(MSB)  
Bit 0  
(LSB)  
Value  
(Hex)  
Parameter Name  
Bit 6  
Bit 5  
Bit 4  
Bit 3  
Bit 2  
Bit 1  
Comments  
Chip Configuration Registers  
0x00  
SPI port  
configuration  
0 = SDO  
active  
LSB first  
Soft reset  
1 = 16-bit  
address  
1 =  
16-bit  
address  
Soft reset  
LSB first  
0 = SDO  
active  
0x18  
Nibbles are  
mirrored to  
allow a given  
register value  
to perform  
the same  
function for  
either MSB  
first or LSB  
first mode.  
0x01  
0x02  
Chip ID (global)  
8-bit chip ID, Bits[7:0]  
0xC2 = AD9655, dual, 16-bit, 125 MSPS, serial LVDS  
0xC2  
0x62  
Unique chip  
ID used to  
differentiate  
devices;  
read only.  
Chip grade  
(global)  
Open  
Speed grade ID, Bits[6:4]  
110 = 125 MSPS  
Revision, Bits[3:0]  
Unique speed  
grade ID used  
to differentiate  
graded  
devices;  
read only.  
Device Index Register  
0x05 Device index  
Open  
Clock  
channel  
DCO  
Clock channel  
FCO  
Open  
Data  
Channel B  
Data  
Channel A  
0x33  
Bits are set to  
determine  
which device  
on chip  
receives the  
next write  
command.  
Default is all  
devices on  
chip.  
Global ADC Function Registers  
0x08  
Power modes  
(global)  
Open  
Power mode  
0x00  
0x04  
Determines  
various  
generic  
modes of chip  
operation.  
00 = chip run  
01 = full power-down  
10 = standby  
11 = digital reset  
0x09  
Clock (global)  
Open  
DCS  
bypass  
0 =  
inactive  
1 =  
active  
Open  
DCS  
0 = off  
1 = on  
DCS controls.  
Bit 2 = 1  
bypasses the  
clock divider  
as well as  
DCS.  
0x0B  
Clock divide  
(global)  
Open  
Clock divide ratio, Bits[2:0]  
000 = divide by 1  
001 = divide by 2  
010 = divide by 3  
011 = divide by 4  
100 = divide by 5  
101 = divide by 6  
110 = divide by 7  
111 = divide by 8  
0x00  
Setting  
Register 0x09,  
Bit 2 = 1  
bypasses the  
clock divider  
as well as  
DCS.  
0x0C  
Enhancement  
control  
Open  
Chop  
Open  
0x00  
Enables/  
disables  
chop mode.  
mode  
0 = off  
1 = on  
Rev. 0 | Page 31 of 37  
 
 
AD9655  
Data Sheet  
Default  
Value  
(Hex)  
Addr.  
(Hex)  
Bit 7  
(MSB)  
Bit 0  
(LSB)  
Parameter Name  
Bit 6  
Bit 5  
Bit 4  
Bit 3  
Bit 2  
Bit 1  
Comments  
0x0D  
Test mode  
(local except for  
PN sequence  
resets)  
User input test mode  
00 = single  
Reset PN  
long  
gen-  
Reset PN short  
generator  
Output test mode, Bits[3:0] (local)  
0000 = off (default)  
0x00  
When set, the  
test data is  
placed on the  
output pins in  
place of  
01 = alternate  
0001 = midscale short  
0010 = positive FS  
0011 = negative FS  
10 = single once  
11 = alternate once  
(Bits[7:6] affect user  
input, test mode only,  
Bits[3:0] = 1000)  
erator  
0100 = alternating checkerboard  
0101 = PN23 sequence  
0110 = PN9 sequence  
0111 = one-/zero-word toggle  
1000 = user input  
normal data.  
1001 = 1-/0-bit toggle  
1010 = 1× sync  
1011 = one bit high  
1100 = mixed bit frequency  
0x10  
0x14  
Offset adjust  
(local)  
8-bit device offset adjustment, Bits[7:0] (local);  
offset adjust in LSBs from −128 to +127 (twos complement format)  
0x00  
0x01  
Device offset  
trim.  
0
0
0
Output mode  
0
LVDS-  
ANSI/  
LVDS-  
IEEE  
Output  
invert  
(local)  
Open  
Output  
format  
0 = offset  
binary  
Configures  
the outputs  
and format of  
the data.  
option  
0 = LVDS-  
ANSI  
1 = twos  
comple-  
ment  
1 =  
(global)  
LVDS-  
IEEE  
reduced  
range  
link  
(global);  
see  
Table 19  
0x16  
Output phase  
Open  
Input clock phase adjust, Bits[6:4]  
(value is number of input clock cycles  
of phase delay); see Table 20  
Output clock phase adjust, Bits[3:0]  
(0000 through 1011); see Table 21  
0x03  
On devices  
using global  
clock divide,  
Bits[6:4]  
determine  
which phase  
of the divider  
output  
supplies the  
output clock.  
Internal  
latching is  
unaffected.  
0x18  
VREF  
Open  
Internal VREF adjustment digital scaling,  
Bits[2:0]  
0x04  
Digitally  
adjusts full-  
scale input  
voltage. Does  
not affect  
analog input  
swing. Values  
shown are for  
000 = 1.0 V p-p (1.4 V p-p)  
001 = 1.14 V p-p (1.6 V p-p)  
010 = 1.33 V p-p (1.86 V p-p)  
011 = 1.6 V p-p (2.24 V p-p)  
100 = 2.0 V p-p (2.8 V p-p)  
VREF = 1.0 V  
(1.4 V).  
0x19  
0x1A  
0x1B  
0x1C  
USER_PATT1_LSB  
(global)  
B7  
B6  
B5  
B4  
B3  
B2  
B1  
B9  
B1  
B9  
B0  
B8  
B0  
B8  
0x00  
0x00  
0x00  
0x00  
User Defined  
Pattern 1  
(8 LSBs).  
USER_PATT1_MSB B15  
(global)  
B14  
B6  
B13  
B5  
B12  
B4  
B11  
B3  
B10  
B2  
User Defined  
Pattern 1  
(8 MSBs).  
USER_PATT2_LSB  
(global)  
B7  
User Defined  
Pattern 2  
(8 LSBs).  
USER_PATT2_MSB B15  
(global)  
B14  
B13  
B12  
B11  
B10  
User Defined  
Pattern 2  
(8 MSBs).  
Rev. 0 | Page 32 of 37  
Data Sheet  
AD9655  
Default  
Value  
(Hex)  
Addr.  
(Hex)  
Bit 7  
(MSB)  
Bit 0  
(LSB)  
Parameter Name  
Bit 6  
Bit 5  
Bit 4  
Bit 3  
Bit 2  
Bit 1  
Comments  
0x21  
Serial output  
data control  
(global)  
LVDS  
SDR/DDR one-lane/two-lane,  
bitwise/bytewise, Bits[6:4]  
000 = SDR two-lane, bitwise  
001 = SDR two-lane, bytewise  
010 = DDR two-lane, bitwise  
0
0 = 1×  
frame  
(default)  
1 = 2×  
frame  
Serial output  
number of bits  
00 = 16 bits (default)  
0x30  
Serial stream  
control.  
output  
0 = MSB  
first  
(default)  
1 = LSB  
first  
01 = 14 bits  
10 = 12 bits  
11 = 10 bits  
011 = DDR two-lane, bytewise (default)  
100 = DDR one-lane, wordwise  
0x22  
Serial channel  
status (local)  
Open  
Channel  
output  
reset  
Channel  
power-  
down  
0x00  
Used with  
Register 0x05  
to power  
down  
individual  
sections of a  
converter.  
0x101 User Input/Output  
Control 2  
Open  
Disable  
SDIO  
pull-down  
0x00  
0x00  
0x00  
Disables SDIO  
pull-down  
resistor.  
0x102 User Input/Output  
Control 3  
Open  
VCM  
power-  
down  
0
0
0
VCM control.  
0x107 Foreground  
calibrations status  
Open  
Bit 1 = 1  
during  
Bit 0 = 1  
during  
Read only.  
foreground foreground  
calibration,  
Bit 1 = 0  
after  
calibration,  
Bit 0 = 0  
after  
calibration  
is complete  
calibration  
is complete  
0x112 Clock monitor  
control  
Open  
0
Recovery mode  
000 = recovery off (default)  
101 = recovery on  
1
1
1
0x07  
0x00  
Select clock  
recovery  
mode.  
0x114 VREF control  
Open  
Drive  
VREF  
external  
0 = no  
1 = yes  
VREF  
Bits[1:0] set  
the internal  
reference  
voltage. Bit 2  
disables the  
reference if  
an external  
source is  
00 = 1.0 V  
01 = 1.2 V  
10 = 1.3 V  
11 = 1.4 V  
desired.  
Rev. 0 | Page 33 of 37  
AD9655  
Data Sheet  
During a digital reset (Bits[1:0] = 11), a foreground calibration is  
invoked, all the digital datapath clocks and the outputs (where  
applicable) on the chip are reset, except the SPI port. Note that the  
SPI is always left under control of the user; that is, it is never  
automatically disabled or in reset (except by power-on reset).  
MEMORY MAP REGISTER DESCRIPTIONS  
For general information about the functions controlled in  
Register 0x00 to Register 0xFF, see the AN-877 Application Note,  
Interfacing to High Speed ADCs via SPI.  
Device Index (Register 0x05)  
Enhancement Control (Register 0x0C)  
Bits[7:3]—Open  
There are certain features in the map that can be set indepen-  
dently for each channel, whereas other features apply globally  
to all channels (depending on context), regardless of which  
channel is selected. Bits[1:0] in Register 0x05 can be used to  
select which individual data channel is affected by the next SPI  
action. The output clock channels can be selected in Register 0x05,  
as well. A smaller subset of the independent feature list can be  
applied to those devices.  
Bit 2—Chop Mode  
For applications sensitive to offset voltages and other low  
frequency noise, such as homodyne or direct conversion  
receivers, chopping in the first stage of the AD9655 is an  
available feature, enabled by setting Bit 2. In the frequency  
domain, chopping translates offsets and other low frequency  
noise to fCLK/2, where it can be filtered.  
Refer to the following sequence of SPI writes for an example of  
how to use Register 0x5 to power down Channel B while keeping  
Channel A active:  
Bits[1:0]—Open  
Output Mode (Register 0x14)  
Bit 7—0  
1. SPI_Write (0x05, 0x02)—designates Channel B to be  
affected by the next local SPI register instruction.  
2. SPI_Write (0x22, 0x01)—powers down the Channel  
previously designated.  
Bit 6—LVDS-ANSI/LVDS-IEEE Option  
Setting this bit selects the LVDS-IEEE (reduced range) option.  
3. SPI_Write (0x05, 0x31)—designates DCO, FCO and  
Channel A to be affected by future local SPI register  
instruction (optional).  
4. SPI_Write (0x08, 0x03)—digital reset.  
5. SPI_Write (0x08, 0x00)—takes the device back to  
normal operation.  
The default setting is LVDS-ANSI. When LVDS-ANSI or the  
LVDS-IEEE reduced range link is selected, the driver current is  
automatically selected to give the proper output swing.  
Table 19. LVDS-ANSI/LVDS-IEEE Options  
Output  
Mode, Bit 6  
Output Mode  
Output Driver Current  
Power Modes (Register 0x08)  
Bits[7:2]—Open  
0
LVDS-ANSI  
Automatically selected to  
give proper swing  
1
LVDS-IEEE reduced Automatically selected to  
range link give proper swing  
Bits[1:0]—Power Mode  
In normal operation (Bits[1:0] = 00), both ADC channels are  
active.  
Bits[5:3]—000  
In power-down mode (Bits[1:0] = 01), the digital datapath clocks  
are disabled while the digital datapath is reset. Outputs are disabled.  
Bit 2—Output Invert  
Setting this bit inverts the output bit stream.  
Bit 1—Open  
In standby mode (Bits[1:0] = 10), the digital datapath clocks  
and the outputs are disabled. Do not invoke standby mode  
while foreground calibration is in progress. Foreground  
calibration is invoked automatically at power-up and by  
executing a digital reset with Register 0x08. Foreground  
calibration completion is indicated by the contents of  
Register 0x107 = 0x00.  
Bit 0—Output Format  
By default, this bit is set to send the data output in twos  
complement format. Clearing this bit to 0 changes the output  
mode to offset binary.  
Rev. 0 | Page 34 of 37  
 
 
Data Sheet  
AD9655  
Output Phase (Register 0x16)  
Bit 7—Open  
available in the AD9655. Note that, in single data rate (SDR)  
mode, the DCO frequency is double that of the frequency in  
DDR mode for a given sample rate. In SDR mode, to stay within  
the capability of the DCO LVDS driver, the ADC sample rate  
must be reduced to ≤62.5 MSPS to keep the DCO frequency at  
≤500 MHz.  
Bits[6:4]—Input Clock Phase Adjust  
When the clock divider (Register 0x0B) is used, the applied  
clock is at a higher frequency than the internal sampling clock.  
Bits[6:4] determine at which phase of the external clock sampling  
occurs. This is applicable only when the clock divider is used.  
Selecting a value for Bits[6:4] greater than Register 0x0B,  
Bits[2:0] is prohibited. See Table 20 for details.  
User I/O Control 2 (Register 0x101)  
Bits[7:1]—Open  
Bit 0—Disable SDIO Pull-Down  
Bit 0 can be set to disable the internal 31 kΩ pull-down resistor  
on the SDIO/PWDN pin, which limits the loading when many  
devices are connected to the SPI bus.  
Table 20. Input Clock Phase Adjust Options  
Input Clock Phase Adjust,  
Bits[6:4]  
Number of Input Clock Cycles  
of Phase Delay  
000 (Default)  
001  
010  
011  
100  
101  
110  
111  
0
1
2
3
4
5
6
7
User I/O Control 3 (Register 0x102)  
Bits[7:4]—Open  
Bit 3—VCM Power-Down  
Bit 3 can be set high to power down the internal VCM generator.  
This feature is used when applying an external reference.  
Bits[2:0]—000  
Clock Monitor Control (Register 0x112)  
Bit 7—Open  
Bits[3:0]—Output Clock Phase Adjust  
See Table 21 for details.  
Bit 6—0 (Reserved)  
Table 21. Output Clock Phase Adjust Options  
Bits[5:3]—Recovery Mode  
DCO Phase Adjustment  
By default (Bits[5:3] = 000), recovery mode is off. In this condition,  
the AD9655 does not automatically recover from a state change or  
corruption due to a clock glitch or irregularity. With recovery mode  
off, a digital reset (Register 0x08 = 0x03, then Register 0x08 = 0x00)  
is needed to restore the AD9655 to proper operation in case of  
disruption due to clock instability.  
Output Clock (DCO),  
Phase Adjust, Bits[3:0]  
(Approximate Degrees Relative to  
the D0x /D1x Edge)  
0000  
0001  
0010  
0011 (Default)  
0100  
0101  
0110  
0111  
1000  
1001  
1010  
1011 through 1111  
0
60  
120  
180  
240  
300  
360  
420  
480  
540  
600  
660  
With clock recovery mode on (Bits[5:3] = 101), AD9655  
automatically recovers from corruption due to a clock glitch or  
irregularity. After the corruption is autodetected, 31 × 106 clock  
cycles are needed to restore proper operation.  
Bits[2:0]— Recovery Mode Setup  
Recovery mode is set up for correct operation by default  
(Bits[2:0] = 111).  
VREF Control (Register 0x114)  
Serial Output Data Control (Register 0x21)  
This register adjusts the internal analog VREF value. A digital reset  
The serial output data control register to programs the AD9655  
in various output data modes, depending on the data capture  
solution. Table 22 describes the various serialization options  
using Register 0x08 must follow any change in analog VREF  
.
Table 22. SPI Register 0x21 Options  
Serialization Options Selected  
Register 0x21  
Contents  
Serial Output Number  
of Bits (SONB)  
Frame Mode  
Serial Data Mode  
DCO Multiplier  
4 × fS  
4 × fS  
8 × fS  
8 × fS  
Timing Diagram  
0x30  
0x20  
0x10  
0x00  
0x40  
16-bit  
16-bit  
16-bit  
16-bit  
16-bit  
1×  
1×  
1×  
1×  
1×  
DDR two-lane bytewise  
DDR two-lane bitwise  
SDR two-lane bytewise  
SDR two-lane bitwise  
DDR one-lane wordwise  
See Figure 2 (default setting)  
See Figure 2  
See Figure 2  
See Figure 2  
See Figure 3  
8 × fS  
Rev. 0 | Page 35 of 37  
 
 
 
AD9655  
Data Sheet  
APPLICATIONS INFORMATION  
DESIGN GUIDELINES  
EXPOSED PAD THERMAL HEAT SLUG  
RECOMMENDATIONS  
Before starting design and layout of the AD9655 as a system, it  
is recommended that the designer become familiar with these  
guidelines, which describe the special circuit connections and  
layout requirements needed for certain pins.  
It is required that the exposed pad on the underside of the ADC be  
connected to analog ground (AGND) to achieve the best electrical  
and thermal performance of the AD9655. It is recommended  
that an exposed continuous copper plane on the PCB mate to  
the AD9655 exposed pad, Pin 0. It is also recommended that  
this copper plane have several vias to achieve the lowest possible  
resistive thermal path for heat dissipation to flow through the  
bottom of the PCB. Solder-fill or plug these vias.  
POWER AND GROUND GUIDELINES  
When connecting power to the AD9655, it is recommended  
that two separate 1.8 V supplies be used. Use one supply for  
analog (AVDD); use a separate supply for the digital outputs  
(DRVDD). For both AVDD and DRVDD, use several different  
bypass capacitor values to cover both high and low frequencies.  
Place these capacitors close to the point of entry at the PCB  
level and close to the pins of the device, with minimal trace length.  
To maximize the coverage and adhesion between the ADC and  
PCB, partition the continuous copper plane by overlaying a  
silkscreen or solder mask on the PCB into several uniform sections.  
This provides several tie points between the ADC and PCB  
during the reflow process, whereas using one continuous plane  
with no partitions only guarantees one tie point. See Figure 69  
for a PCB layout example. For detailed information on  
packaging and the PCB layout of chip scale packages, see the  
AN-772 Application Note, A Design and Manufacturing Guide  
for the Lead Frame Chip Scale Package (LFCSP).  
A single PCB ground plane is sufficient when using the  
AD9655. With proper bypassing and smart partitioning of the  
PCB analog, digital, and clock sections, optimum performance  
is easily achieved.  
CLOCK STABILITY CONSIDERATIONS  
When powered on, the AD9655 enters an initialization phase  
during which an internal state machine sets up the biases and  
the registers for proper operation. During the initialization  
process, the AD9655 needs a stable clock. If the ADC clock  
source is not present or not stable during ADC power-up, it  
disrupts the state machine and causes the ADC to start up in an  
unknown state. To correct this, reinvoke an initialization  
sequence after the ADC clock is stable by issuing a digital reset  
via Register 0x08. In the default configuration (internal VREF, ac-  
coupled input) where VREF and VCM are supplied by the ADC, a  
stable clock during power-up is sufficient. WhenVCM is supplied  
by an external source, this, too, must be stable at power-up;  
otherwise, a subsequent digital reset via Register 0x08 is  
needed. Clock instability during normal operation may also  
necessitate digital reset to restore proper operation.  
SILKSCREEN PARTITION  
PIN 1 INDICATOR  
Figure 69. Typical PCB Layout  
VCM  
Bypass the VCM pin to ground with a 0.1 μF capacitor.  
REFERENCE BYPASSING  
Externally bypass the VREF pin to ground with a low ESR,  
1.0 μF capacitor in parallel with a low ESR, 0.1 μF ceramic  
capacitor.  
The pseudo code sequence for a digital reset is as follows:  
SPI PORT  
1. SPI_Write (0x08, 0x03)—digital reset.  
2. SPI_Write (0x08, 0x00)—normal operation.  
The SPI port must not be active during periods when the full  
dynamic performance of the converter is required. Because the  
SCLK, CSB, and SDIO signals are typically asynchronous to the  
ADC clock, noise from these signals can degrade converter  
performance. If the on-board SPI bus is used for other devices,  
it may be necessary to provide buffers between this bus and the  
AD9655 to prevent these signals from transitioning at the  
converter inputs during critical sampling periods.  
Rev. 0 | Page 36 of 37  
 
 
 
 
 
 
 
 
 
Data Sheet  
AD9655  
OUTLINE DIMENSIONS  
5.10  
5.00 SQ  
4.90  
0.30  
0.25  
0.18  
PIN 1  
INDICATOR  
PIN 1  
INDICATOR  
25  
32  
24  
1
0.50  
BSC  
*
3.75  
EXPOSED  
PAD  
3.60 SQ  
3.55  
17  
8
16  
9
0.50  
0.40  
0.30  
0.25 MIN  
TOP VIEW  
BOTTOM VIEW  
FOR PROPER CONNECTION OF  
THE EXPOSED PAD, REFER TO  
THE PIN CONFIGURATION AND  
FUNCTION DESCRIPTIONS  
0.80  
0.75  
0.70  
0.05 MAX  
0.02 NOM  
SECTION OF THIS DATA SHEET.  
COPLANARITY  
0.08  
0.20 REF  
SEATING  
PLANE  
*
COMPLIANT TO JEDEC STANDARDS MO-220-WHHD-5  
WITH THE EXCEPTION OF THE EXPOSED PAD DIMENSION.  
Figure 70. 32-Lead Lead Frame Chip Scale Package [LFCSP_WQ]  
5 mm × 5 mm Body, Very Very Thin Quad  
(CP-32-12)  
Dimensions shown in millimeters  
ORDERING GUIDE  
Model1  
Temperature Range  
−40°C to +85°C  
−40°C to +85°C  
Package Description  
Package Option  
CP-32-12  
CP-32-12  
AD9655BCPZ-125  
AD9655BCPZRL7-125  
AD9655-125EBZ  
32-Lead Lead Frame Chip Scale Package (LFCSP_WQ)  
32-Lead Lead Frame Chip Scale Package (LFCSP_WQ)  
Evaluation Board  
1 Z = RoHS Compliant Part.  
©2015 Analog Devices, Inc. All rights reserved. Trademarks and  
registered trademarks are the property of their respective owners.  
D12737-0-1/15(0)  
Rev. 0 | Page 37 of 37  
 
 

相关型号:

AD9655-125EBZ

Dual, 16-Bit, 125 MSPS Serial LVDS, 1.8 V Analog-to-Digital Converter
ADI

AD9655BCPZ-125

Dual, 16-Bit, 125 MSPS Serial LVDS, 1.8 V Analog-to-Digital Converter
ADI

AD9655BCPZRL7-125

Dual, 16-Bit, 125 MSPS Serial LVDS, 1.8 V Analog-to-Digital Converter
ADI

AD9655_17

Dual, 16-Bit, 125 MSPS Serial LVDS, 1.8 V Analog-to-Digital Converter
ADI

AD9656

Quad, 16-Bit, 125 MSPS 1.8 V Analog-to-Digital Converter
ADI

AD9656BCPZ-125

Quad, 16-Bit, 125 MSPS 1.8 V Analog-to-Digital Converter
ADI

AD9656BCPZRL7-125

Quad, 16-Bit, 125 MSPS 1.8 V Analog-to-Digital Converter
ADI

AD9656EBZ

Quad, 16-Bit, 125 MSPS 1.8 V Analog-to-Digital Converter
ADI

AD9660

200 MHz Laser Diode Driver with Light Power Control
ADI

AD96606

200 MHz Laser Diode Driver with Light Power Control
ADI

AD9660KR

200 MHz Laser Diode Driver with Light Power Control
ADI

AD9660KR-REEL

200 MHz Laser Diode Driver with Light Power Control
ADI