PDF Data Sheet Rev. A

FEATURES
TYPICAL APPLICATION CIRCUIT
Support input data rate >2 GSPS
Proprietary low spurious and distortion design
SFDR = 82 dBc at dc IF, −9 dBFS
Flexible 8-lane JESD204B interface
Multiple chip synchronization
Fixed latency
Data generator latency compensation
Selectable 1×, 2×, 4×, or 8× interpolation filter
Low power architecture
Transmit enable function allows extra power saving and
instant control of the output status
High performance, low noise phase-locked loop (PLL) clock
multiplier
Digital inverse sinc filter
Low power: 1.42 W at 1.6 GSPS full operating conditions
88-lead LFCSP with exposed pad
QUAD MOD
LPF
ADRF6720
SYSREF±
DAC
RF OUTPUT
0°/90° PHASE
SHIFTER
JESD204B
DAC
AD9135/
AD9136
LO_IN
SYNCOUT0±
SYNCOUT1±
MOD_SPI
CLK±
DAC
SPI
12578-001
Data Sheet
Dual, 11-/16-Bit, 2.8 GSPS, TxDAC+®
Digital-to-Analog Converters
AD9135/AD9136
Figure 1.
APPLICATIONS
Wireless communications
3G/4G W-CDMA base stations
Wideband repeaters
Software defined radios
Wideband communications
Point to point
Local multipoint distribution service (LMDS) and
multichannel multipoint distribution service (MMDS)
Transmit diversity, multiple input/multiple output (MIMO)
Instrumentation
Automated test equipment
GENERAL DESCRIPTION
PRODUCT HIGHLIGHTS
The AD9135/AD9136 are dual, 11-/16-bit, high dynamic range
digital-to-analog converters (DACs) that provide a maximum
sample rate of 2800 MSPS, permitting a multicarrier generation
over a very wide bandwidth. The DAC outputs are optimized to
interface seamlessly with the ADRF6720, as well as other analog
quadrature modulators (AQMs) from Analog Devices, Inc. An
optional 3-wire or 4-wire serial port interface (SPI) provides for
programming/readback of many internal parameters. The fullscale output current can be programmed over a typical range of
13.9 mA to 27.0 mA. The AD9135/AD9136 are available in an
88-lead LFCSP.
1.
2.
3.
4.
5.
6.
Rev. A
Greater than 2 GHz, ultrawide complex signal bandwidth
enables emerging wideband and multiband wireless
applications.
Advanced low spurious and distortion design techniques
provide high quality synthesis of wideband signals from
baseband to high intermediate frequencies.
JESD204B Subclass 1 support simplifies multichip
synchronization in software and hardware design.
Fewer pins for data interface width with a serializer/
deserializer (SERDES) JESD204B eight-lane interface.
Programmable transmit enable function allows easy design
balance between power consumption and wake-up time.
Small package size with 12 mm × 12 mm footprint.
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
rights of third parties that may result from its 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 and registered trademarks 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
©2014 Analog Devices, Inc. All rights reserved.
Technical Support
www.analog.com
AD9135/AD9136
Data Sheet
TABLE OF CONTENTS
Features .............................................................................................. 1 JESD204B Setup ......................................................................... 30 Applications ....................................................................................... 1 SERDES Clocks Setup ................................................................ 32 Typical Signal Chain......................................................................... 1 Equalization Mode Setup .......................................................... 32 General Description ......................................................................... 1 Link Latency Setup ..................................................................... 32 Product Highlights ........................................................................... 1 Crossbar Setup ............................................................................ 34 Revision History ............................................................................... 3 JESD204B Serial Data Interface .................................................... 35 Functional Block Diagram .............................................................. 4 JESD204B Overview .................................................................. 35 Specifications..................................................................................... 5 Physical Layer ............................................................................. 36 DC Specifications ......................................................................... 5 Data Link Layer .......................................................................... 39 Digital Specifications ................................................................... 6 Transport Layer .......................................................................... 48 Maximum DAC Update Rate Speed Specifications by Supply ..... 7 JESD204B Test Modes ............................................................... 58 JESD204B Serial Interface Speed Specifications ...................... 7 JESD204B Error Monitoring..................................................... 59 SYSREF Signal to DAC Clock Timing Specifications.............. 8 Hardware Considerations ......................................................... 61 Digital Input Data Timing Specifications ................................. 8 Digital Datapath ............................................................................. 65 Latency Variation Specifications ................................................ 9 DAC Paging ................................................................................. 65 JESD204B Interface Electrical Specifications ........................... 9 Data Format ................................................................................ 65 AC Specifications........................................................................ 10 Interpolation Filters ................................................................... 65 Absolute Maximum Ratings.......................................................... 11 Inverse Sinc ................................................................................. 66 Thermal Resistance .................................................................... 11 Digital Gain, DC Offset, and Group Delay ............................. 66 ESD Caution ................................................................................ 11 Downstream Protection ............................................................ 68 Pin Configuration and Function Descriptions ........................... 12 Datapath PRBS ........................................................................... 69 Terminology .................................................................................... 15 DC Test Mode ............................................................................. 70 Typical Performance Characteristics ........................................... 16 Interrupt Request Operation ........................................................ 71 Theory of Operation ...................................................................... 22 Interrupt Service Routine .......................................................... 71 Serial Port Operation ..................................................................... 23 DAC Input Clock Configurations ................................................ 72 Data Format ................................................................................ 23 Driving the CLK± Inputs .......................................................... 72 Serial Port Pin Descriptions ...................................................... 23 DAC PLL Fixed Register Writes ............................................... 72 Serial Port Options ..................................................................... 23 Clock Multiplication .................................................................. 72 Chip Information ............................................................................ 25 Starting the PLL .......................................................................... 74 Device Setup Guide ........................................................................ 26 Analog Outputs............................................................................... 75 Overview...................................................................................... 26 Transmit DAC Operation.......................................................... 75 Step 1: Start Up the DAC ........................................................... 26 Device Power Dissipation.............................................................. 78 Step 2: Digital Datapath ............................................................. 27 Temperature Sensor ................................................................... 78 Step 3: Transport Layer .............................................................. 27 Start-Up Sequence .......................................................................... 79 Step 4: Physical Layer ................................................................. 28 Step 1: Start Up the DAC ........................................................... 79 Step 5: Data Link Layer .............................................................. 28 Step 2: Digital Datapath............................................................. 79 Step 6: Optional Error Monitoring .......................................... 29 Step 3: Transport Layer .............................................................. 80 Step 7: Optional Features ........................................................... 29 Step 4: Physical Layer ................................................................. 80 DAC PLL Setup ........................................................................... 30 Step 5: Data Link Layer.............................................................. 81 Interpolation ............................................................................... 30 Step 6: Error Monitoring ........................................................... 81 Rev. A | Page 2 of 117
Data Sheet
AD9135/AD9136
Register Maps and Descriptions ....................................................82 Outline Dimensions ......................................................................116 Device Configuration Register Map .........................................82 Ordering Guide .........................................................................117 Device Configuration Register Descriptions ..........................88 REVISION HISTORY
7/15—Rev. 0 to Rev. A
Changed Functional Block Diagram Section to Typical
Application Circuit Section.............................................................. 1
Changes to General Description Section ....................................... 1
Changed Detailed Functional Block Diagram Section to
Functional Block Diagram Section ................................................. 4
Changes to Offset Drift Parameter, Table 1 ................................... 5
Deleted Reference Voltage Parameter, Table 1 .............................. 5
Changed 1× Interpolation Mode Parameter to 1× Interpolation
Mode, JESD Mode 8, 8 SERDES Lanes Parameter, Table 1 ......... 5
Changes to Output Voltage (VOUT) Logic High Parameter,
Output Voltage (VOUT) Logic Low Parameter, JESD204B Serial
Interface Speed Minimum Parameter, and SYSREF± Frequency
Parameter, Table 2 ............................................................................. 6
Changes to Table 4 ............................................................................ 7
Changes to Interpolation Parameter, Table 6 ................................ 8
Changed Junction Temperature Parameter to Operating
Junction Temperature, Table 10 ....................................................11
Changes to Terminology Section ..................................................17
Changes to Figure 34 Caption and Figure 37 Caption ...............21
Changes to Device Revision Parameter, Table 14 .......................25
Changes to Overview Section, Table 15, Table 16, and
Table 17 .............................................................................................26
Changes to Step 3: Transport Layer Section and Table 19 .........27
Changes to Table 20 and Table 21 .................................................28
Changes to Step 7: Optional Features Section .............................29
Added Table 25; Renumbered Sequentially .................................30
Changes to DAC PLL Setup Section, Table 26, and Table 27 ......30
Changes to Table 28 and CurrentLink Section............................31
Added DAC Power-Down Setup Section ....................................31
Changes to Table 30 ........................................................................32
Changes to Table 32 ........................................................................33
Changes to Table 36 and Figure 44 ...............................................36
Changes to Table 37 ........................................................................37
Added SERDES PLL Fixed Register Writes Section and
Table 38 .............................................................................................37
Changes to Table 39 ........................................................................38
Changes to Figure 47 and Data Link Layer Section ...................39
Added Figure 50; Renumbered Sequentially ............................... 40
Changes to Table 45 and Table 46 ................................................. 49
Changes to Figure 61 ...................................................................... 51
Changes to Figure 62 ...................................................................... 52
Changes to Figure 63 ...................................................................... 53
Changes to Figure 64 ...................................................................... 54
Changes to Figure 65 ...................................................................... 56
Changes to Figure 66 ...................................................................... 57
Changes to Power Supply Recommendations Section ............... 61
Added Figure 68 .............................................................................. 62
Changes to Figure 72 ...................................................................... 64
Changes to Data Format Section and Table 61 ........................... 65
Changes to Figure 76 ...................................................................... 66
Changed 0x13D[7:0] to 0x13D[3:0], Table 62............................. 67
Changes to Group Delay Section .................................................. 67
Changes to DC Test Mode Section ............................................... 95
Deleted Table 70; Renumbered Sequentially ............................... 71
Moved Figure 78 and Table 68 ...................................................... 71
Added DAC PLL Fixed Register Writes Section and
Table 69 ............................................................................................. 72
Changes to Clock Multiplication Section .................................... 73
Added Loop Filter Section and Charge Pump Filter Section ...... 73
Added Temperature Tracking Section and Table 73 .................. 74
Changes to Starting the PLL Section and Figure 82 ................... 74
Changes to Transmit DAC Operation Section ............................ 75
Changes to Start-Up Sequence Section, Table 77, and
Table 78 ............................................................................................. 79
Changes to Table 80 and Table 81 ................................................. 80
Changes to Table 82 and Table 83 ................................................. 81
Changes to Table 84 ........................................................................ 82
Changes to Table 85 ........................................................................ 88
Deleted Lookup Tables for Three Different DAC PLL Reference
Frequencies Section and Table 83 to 85 ..................................... 112
Added Figure 92 ............................................................................ 116
Updated Outline Dimensions...................................................... 116
Changes to Ordering Guide ......................................................... 117
9/14—Revision 0: Initial Version
Rev. A | Page 3 of 117
AD9135/AD9136
Data Sheet
FUNCTIONAL BLOCK DIAGRAM
DACCLK
HB2
HB3
OUT1+
FSC
Q-GAIN
DACCLK
MODE CONTROL
I-GAIN
HB1
HB2
I-OFFSET
HB3
OUT0+
FSC
SERDIN0±
CLOCK DISTRIBUTION
AND
CONTROL LOGIC
PLL_CTRL
SERIAL
I/O PORT
POWER-ON
RESET
DAC
ALIGN
DETECT
CLK_SEL
CONFIG
REGISTERS
OUT0–
REF
AND
BIAS
I120
SYSREF+
SYSREF–
CLK+
CLK–
DAC PLL
12578-002
TXEN1
IRQ
TXEN0
SYSREF
Rx
CLK
Rx
DACCLK
PLL_LOCK
RESET
SYNCOUT1+
SYNCOUT1–
SYNCHRONIZATION
LOGIC
SDO
SDIO
SCLK
CS
SYNCOUT0+
SYNCOUT0–
OUT1–
Q-OFFSET
INV SINC
SERDIN7±
HB1
CLOCK DATA RECOVERY
AND CLOCK FORMATTER
VTT
INV SINC
SERDES
PLL
Figure 2.
Rev. A | Page 4 of 117
Data Sheet
AD9135/AD9136
SPECIFICATIONS
DC SPECIFICATIONS
AVDD33 = 3.3 V, SIOVDD33 = 3.3 V, IOVDD = 1.8 V, DVDD12 = 1.2 V, CVDD12 = 1.2 V, PVDD12 = 1.2 V, SVDD12 = 1.2 V, VTT = 1.2 V,
TA = −40°C to +85°C, IOUTFS = 20 mA, unless otherwise noted.
Table 1.
AD9135
Parameter
RESOLUTION
ACCURACY
Differential Nonlinearity (DNL)
Integral Nonlinearity (INL)
MAIN DAC OUTPUTS
Gain Error
I/Q Gain Mismatch
Full-Scale Output Current
(IOUTFS)
Maximum Setting
Minimum Setting
Output Compliance Range
Output Resistance
Output Capacitance
Gain DAC Monotonicity
Settling Time
MAIN DAC TEMPERATURE DRIFT
Offset
Gain
REFERENCE
Internal Reference Voltage
ANALOG SUPPLY VOLTAGES
AVDD33
PVDD12
CVDD12
DIGITAL SUPPLY VOLTAGES
SIOVDD33
VTT
DVDD12
SVDD12
IOVDD
POWER CONSUMPTION
1× Interpolation Mode,
JESD Mode 8, 8 SERDES Lanes
AVDD33
PVDD12
CVDD12
SVDD12
DVDD12
SIOVDD33
IOVDD
Test Conditions/Comments
Min
Typ
11
AD9136
Max
Min
Typ
16
Max
Unit
Bits
With calibration
±0.175
±0.35
With internal reference
±1.0
±2.0
LSB
LSB
−2.5
−0.6
+2
+5.5
+0.6
−2.5
−0.6
+2
+5.5
+0.6
% FSR
% FSR
25.5
13.1
−250
27.0
13.9
28.6
14.8
+750
25.5
13.1
−250
27.0
13.9
28.6
14.8
+750
mA
mA
mV
MΩ
pF
Based on a 4 kΩ external resistor
between I120 and GND
To within ±0.5 LSB
1.2 V nominal supply voltage
1.3 V nominal supply voltage
1.2 V nominal supply voltage
1.3 V nominal supply voltage
0.2
3.0
Guaranteed
20
0.2
3.0
Guaranteed
20
0.04
32
0.04
32
ppm
ppm/°C
1.2
1.2
V
ns
3.13
1.14
1.14
3.3
1.2
1.2
3.47
1.26
1.26
3.13
1.14
1.14
3.3
1.2
1.2
3.47
1.26
1.26
V
V
V
3.13
1.1
1.14
1.274
1.14
1.274
1.71
3.3
1.2
1.2
1.3
1.2
1.3
1.8
3.47
1.37
1.26
1.326
1.26
1.326
3.47
3.13
1.1
1.14
1.274
1.14
1.274
1.71
3.3
1.2
1.2
1.3
1.2
1.3
1.8
3.47
1.37
1.26
1.326
1.26
1.326
3.47
V
V
V
V
V
V
V
1.42
1.74
1.42
1.74
W
68
100
101
554
196
11
36
73
113.4
112
665
224
12
50
68
100
101
554
196
11
36
73
113.4
112
665
224
12
50
mA
mA
mA
mA
mA
mA
μA
fDAC = 1.6 GSPS, IF = 40 MHz, PLL on,
digital gain on, inverse sinc on, DAC
full-scale current (IOUTFS) = 20 mA
Includes VTT
Rev. A | Page 5 of 117
AD9135/AD9136
Data Sheet
DIGITAL SPECIFICATIONS
AVDD33 = 3.3 V, SIOVDD33 = 3.3 V, IOVDD = 1.8 V, DVDD12 = 1.2 V, CVDD12 = 1.2 V, PVDD12 = 1.2 V, SVDD12 = 1.2 V, VTT = 1.2 V,
TA = −40°C to +85°C, IOUTFS = 20 mA, unless otherwise noted.
Table 2.
Parameter
CMOS INPUT LOGIC LEVEL
Input Voltage (VIN) Logic
High
Low
Symbol
CMOS OUTPUT LOGIC LEVEL
Output Voltage (VOUT) Logic
High
Low
Test Conditions/Comments
Min
Typ
Max
Unit
1.8 V  IOVDD  3.3 V
1.8 V  IOVDD  3.3 V
0.7 × IOVDD
0.3 × IOVDD
V
V
1.8 V  IOVDD  3.3 V
1.8 V  IOVDD  3.3 V
0.75 × IOVDD
0.25 × IOVDD
V
V
1× interpolation2 (see Table 4)
2× interpolation2
4× interpolation3
8× interpolation3
2120
2120
2800
2800
MSPS
MSPS
MSPS
MSPS
1× interpolation
2× interpolation
4× interpolation
8× interpolation
2120
1060
700
350
MSPS
MSPS
MSPS
MSPS
MAXIMUM DAC UPDATE RATE1
ADJUSTED DAC UPDATE RATE
INTERFACE4
Number of JESD204B Lanes
JESD204B Serial Interface Speed
Minimum
Maximum
DAC CLOCK INPUT (CLK+, CLK−)
Differential Peak-to-Peak Voltage
Common-Mode Voltage
Maximum Clock Rate
REFCLK5 Frequency (PLL Mode)
SYSTEM REFERENCE INPUT
(SYSREF+, SYSREF−)
Differential Peak-to-Peak Voltage
Common-Mode Voltage
SYSREF± Frequency6
SYSREF SIGNAL TO DAC CLOCK7
Setup Time
Hold Time
Keep Out Window
SPI
Maximum Clock Rate
Minimum SCLK Pulse Width
High
Low
SDIO to SCLK
Setup Time
Hold Time
8
Per lane
Per lane, SVDD12 = 1.3 V ± 2%
1.44
Gbps
Gbps
2000
1000
mV
mV
MHz
MHz
2000
2000
fDATA/(K × S)
mV
mV
Hz
10.64
400
Self biased input, ac-coupled
6.0 GHz ≤ fVCO ≤ 12.0 GHz
Lanes
1000
600
2800
35
400
0
1000
SYSREF differential swing = 0.4 V, slew
rate = 1.3 V/ns, common modes tested:
ac-coupled, 0 V, 0.6 V, 1.25 V, 2.0 V
tSSD
tHSD
KOW
SCLK
131
119
ps
ps
ps
20
IOVDD = 1.8 V
10
tPWH
tPWL
MHz
8
12
tDS
tDH
5
2
Rev. A | Page 6 of 117
ns
ns
ns
ns
Data Sheet
Parameter
SDO to SCLK
Data Valid Window
CS to SCLK
Setup Time
Hold Time
AD9135/AD9136
Symbol
Test Conditions/Comments
Min
Typ
Max
Unit
tDV
25
ns
tSCS
5
ns
tHCS
2
ns
1
See Table 3 for detailed specifications for DAC update rate conditions.
The maximum speed for 1× and 2× interpolation is limited by the JESD204B interface with increased supply levels. See Table 4 for details.
The maximum speed for 4× and 8× interpolation is limited by the DAC core. See Table 4 for details.
4
See Table 4 for detailed specifications for JESD204B speed conditions.
5
REFCLK is the reference clock.
6
K, F, and S are JESD204B transport layer parameters. See Table 43 for the full definitions.
7
See Table 5 for detailed specifications for SYSREF signal to DAC clock timing conditions.
2
3
MAXIMUM DAC UPDATE RATE SPEED SPECIFICATIONS BY SUPPLY
AVDD33 = 3.3 V, SIOVDD33 = 3.3 V, IOVDD = 1.8 V, DVDD12 = 1.2 V, CVDD12 = 1.2 V, PVDD12 = 1.2 V, SVDD12 = 1.2 V, VTT = 1.2 V,
TA = −40°C to +85°C, IOUTFS = 20 mA, unless otherwise noted.
Table 3.
Parameter
MAXIMUM DAC UPDATE RATE
2×, 4×, and 8× Interpolation
Test Conditions/Comments
Min
Typ
Max
Unit
DVDD12, CVDD12 = 1.2 V ± 5%
DVDD12, CVDD12 = 1.2 V ± 2%
DVDD12, CVDD12 = 1.3 V ± 2%
2.23
2.41
2.80
GSPS
GSPS
GSPS
DVDD12, CVDD12 = 1.2 V ± 5%
DVDD12, CVDD12 = 1.2 V ± 2%
DVDD12, CVDD12 = 1.3 V ± 2%
1.81
1.93
2.21
GSPS
GSPS
GSPS
1× Interpolation
JESD204B SERIAL INTERFACE SPEED SPECIFICATIONS
AVDD33 = 3.3 V, SIOVDD33 = 3.3 V, IOVDD = 1.8 V, DVDD12 = 1.2 V, CVDD12 = 1.2 V, PVDD12 = 1.2 V, SVDD12 = 1.2 V, VTT = 1.2 V,
TA = −40°C to +85°C, IOUTFS = 20 mA, unless otherwise noted.
Table 4.
Parameter
HALF RATE
FULL RATE
OVERSAMPLING
Test Conditions/Comments
SVDD12 = 1.2 V ± 5%
SVDD12 = 1.2 V ± 2%
SVDD12 = 1.3 V ± 2%
SVDD12 = 1.2 V ± 5%
SVDD12 = 1.2 V ± 2%
SVDD12 = 1.3 V ± 2%
SVDD12 = 1.2 V ± 5%
SVDD12 = 1.2 V ± 2%
SVDD12 = 1.3 V ± 2%
Rev. A | Page 7 of 117
Min
5.75
5.75
5.75
2.88
2.88
2.88
1.44
1.44
1.44
Typ
Max
8.92
9.42
10.64
4.63
4.93
5.52
2.31
2.46
2.76
Unit
Gbps
Gbps
Gbps
Gbps
Gbps
Gbps
Gbps
Gbps
Gbps
AD9135/AD9136
Data Sheet
SYSREF SIGNAL TO DAC CLOCK TIMING SPECIFICATIONS
AVDD33 = 3.3 V, SIOVDD33 = 3.3 V, IOVDD = 1.8 V, DVDD12 = 1.2 V, CVDD12 = 1.2 V, PVDD12 = 1.2 V, SVDD12 = 1.2 V, VTT = 1.2 V,
TA = −40°C to +85°C, IOUTFS = 20 mA, SYSREF± common-mode voltages = 0.0 V, 0.6 V, 1.25 V, and 2.0 V, unless otherwise noted.
Table 5.
Parameter
SYSREF DIFFERENTIAL SWING = 0.4 V, SLEW RATE = 1.3 V/ns
Setup Time
Hold Time
SYSREF DIFFERENTIAL SWING = 0.7 V, SLEW RATE = 2.28 V/ns
Setup Time
Hold Time
SYSREF SWING = 1.0 V, SLEW RATE = 3.26 V/ns
Setup Time
Hold Time
Test Conditions/Comments
Min
Typ
Max
Unit
AC-coupled
DC-coupled
AC-coupled
DC-coupled
126
131
92
119
ps
ps
ps
ps
AC-coupled
DC-coupled
AC-coupled
DC-coupled
96
104
77
95
ps
ps
ps
ps
AC-coupled
DC-coupled
AC-coupled
DC-coupled
83
90
68
84
ps
ps
ps
ps
DIGITAL INPUT DATA TIMING SPECIFICATIONS
AVDD33 = 3.3 V, SIOVDD33 = 3.3 V, IOVDD = 1.8 V, DVDD12 = 1.2 V, CVDD12 = 1.2 V, PVDD12 = 1.2 V, SVDD12 = 1.2 V, VTT = 1.2 V,
TA = 25°C, IOUTFS = 20 mA, unless otherwise noted.
Table 6.
Parameter
LATENCY
Interface
Interpolation
1×
2×
4×
8×
Inverse Sinc
Digital Gain Adjust
POWER-UP TIME
1
Min
Typ
Max
Unit
17
PClock1 cycles
66
137
251
484
17
12
60
DAC clock cycles
DAC clock cycles
DAC clock cycles
DAC clock cycles
DAC clock cycles
DAC clock cycles
μs
PClock is the AD9135/AD9136 internal processing clock and equals the lane rate ÷ 40.
Rev. A | Page 8 of 117
Data Sheet
AD9135/AD9136
LATENCY VARIATION SPECIFICATIONS
AVDD33 = 3.3 V, SIOVDD33 = 3.3 V, IOVDD = 1.8 V, DVDD12 = 1.2 V, CVDD12 = 1.2 V, PVDD12 = 1.2 V, SVDD12 = 1.2 V, VTT = 1.2 V,
TA = 25°C, IOUTFS = 20 mA, unless otherwise noted.
Table 7.
Parameter
DAC LATENCY VARIATION
SYNC On
PLL Off
PLL On
Min
Typ
Max
Unit
0
1
+1
DAC clock cycles
DAC clock cycles
−1
JESD204B INTERFACE ELECTRICAL SPECIFICATIONS
AVDD33 = 3.3 V, SIOVDD33 = 3.3 V, IOVDD = 1.8 V, DVDD12 = 1.2 V, CVDD12 = 1.2 V, PVDD12 = 1.2 V, SVDD12 = 1.2 V, VTT = 1.2 V,
TA = −40°C to +85°C, IOUTFS = 20 mA, unless otherwise noted.
Table 8.
Parameter
JESD204B DATA INPUTS
Input Leakage Current
Logic High
Logic Low
Unit Interval
Common-Mode Voltage
Differential Voltage
VTT Source Impedance
Differential Impedance
Differential Return Loss
Common-Mode Return Loss
DIFFERENTIAL OUTPUTS (SYNCOUTx±)2
Output Differential Voltage
Normal Swing Mode
High Swing Mode
Output Offset Voltage
DETERMINISTIC LATENCY
Fixed
Variable
SYSREF± to LOCAL MULTIFRAME
COUNTER (LMFC) DELAY
Symbol
Test Conditions/Comments
Min
TA = 25°C
Input level = 1.2 V ± 0.25 V, VTT = 1.2 V
Input level = 0 V
UI
VRCM
R_VDIFF
ZTT
ZRDIFF
RLRDIF
RLRCM
AC-coupled, VTT = SVDD121
At dc
At dc
Typ
Max
Unit
714
+1.85
1050
30
120
μA
μA
ps
V
mV
Ω
Ω
dB
dB
235
394
1.27
mV
mV
V
17
2
PClock3 cycles
PClock3 cycles
DAC clock cycles
10
−4
94
−0.05
110
80
100
8
6
VOD
Register 0x2A5[0] = 0
Register 0x2A5[0] = 1
VOS
192
341
1.19
4
1
As measured on the input side of the ac coupling capacitor.
IEEE Standard 1596.3 LVDS compatible.
3
PClock is an AD9135/AD9136 internal processing clock and equals the lane rate ÷ 40.
2
Rev. A | Page 9 of 117
AD9135/AD9136
Data Sheet
AC SPECIFICATIONS
AVDD33 = 3.3 V, SIOVDD33 = 3.3 V, IOVDD = 1.8 V, DVDD12 = 1.2 V, CVDD12 = 1.2 V, PVDD12 = 1.2 V, SVDD12 = 1.2 V,1 VTT = 1.2 V,
TA = 25°C, IOUTFS = 20 mA, unless otherwise noted.
Table 9.
Parameter
SPURIOUS-FREE DYNAMIC RANGE (SFDR)
fDAC = 983.04 MSPS
fDAC = 983.04 MSPS
fDAC = 1966.08 MSPS
fDAC = 1966.08 MSPS
TWO-TONE INTERMODULATION DISTORTION (IMD)
fDAC =983.04 MSPS
fDAC = 983.04 MSPS
fDAC = 1966.08 MSPS
fDAC = 1966.08 MSPS
NOISE SPECTRAL DENSITY (NSD), SINGLE-TONE
fDAC = 983.04 MSPS
fDAC = 1966.08 MSPS
W-CDMA FIRST ADJACENT CHANNEL LEAKAGE
RATIO (ACLR), SINGLE-CARRIER
fDAC = 983.04 MSPS
fDAC = 983.04 MSPS
fDAC = 1966.08 MSPS
W-CDMA SECOND ACLR, SINGLE-CARRIER
fDAC = 983.04 MSPS
fDAC = 983.04 MSPS
fDAC = 1966.08 MSPS
1
Test Conditions/Comments
−9 dBFS single-tone
fOUT = 20 MHz
fOUT = 150 MHz
fOUT = 20 MHz
fOUT = 170 MHz
−9 dBFS
fOUT = 20 MHz
fOUT = 150 MHz
fOUT = 20 MHz
fOUT = 170 MHz
0 dBFS
fOUT = 150 MHz
fOUT = 150 MHz
0 dBFS
fOUT = 30 MHz
fOUT = 150 MHz
fOUT = 150 MHz
0 dBFS
fOUT = 30 MHz
fOUT = 150 MHz
fOUT = 150 MHz
SVDD12 = 1.3 V for all fDAC = 1966.08 MSPS conditions in Table 9.
Rev. A | Page 10 of 117
Min
Typ
Max
Unit
82
76
81
69
dBc
dBc
dBc
dBc
90
82
90
81
dBc
dBc
dBc
dBc
−162
−163
dBm/Hz
dBm/Hz
82
80
80
dBc
dBc
dBc
84
85
85
dBc
dBc
dBc
Data Sheet
AD9135/AD9136
ABSOLUTE MAXIMUM RATINGS
THERMAL RESISTANCE
Table 10.
Parameter
I120 to Ground
SERDINx±, VTT, SYNCOUT1±/
SYNCOUT0±, TXENx
OUTx±
SYSREF±
CLK± to Ground
RESET, IRQ, CS, SCLK, SDIO,
SDO to Ground
LDO_BYP1
LDO_BYP2
LDO24
Ambient Operating Temperature (TA)
Operating Junction Temperature
Storage Temperature Range
Rating
−0.3 V to AVDD33 + 0.3 V
−0.3 V to SIOVDD33 + 0.3 V
−0.3 V to AVDD33 + 0.3 V
GND − 0.5 V to +2.5 V
−0.3 V to PVDD12 + 0.3 V
−0.3 V to IOVDD + 0.3 V
−0.3 V to SVDD12 + 0.3 V
−0.3 V to PVDD12 + 0.3 V
−0.3 V to AVDD33 + 0.3 V
−40°C to +85°C
125°C
−65°C to +150°C
The exposed pad (EPAD) must be soldered to the ground plane
for the 88-lead LFCSP. The EPAD provides an electrical, thermal,
and mechanical connection to the board.
Typical θJA, θJB, and θJC values are specified for a 4-layer
JESD51-7 high effective thermal conductivity test board for
leaded surface-mount packages. θJA is obtained in still air
conditions (JESD51-2). Airflow increases heat dissipation,
effectively reducing θJA. θJB is obtained following double-ring
cold plate test conditions (JESD51-8). θJC is obtained with the test
case temperature monitored at the bottom of the exposed pad.
ΨJT and ΨJB are thermal characteristic parameters obtained with
θJA in still air test conditions.
Junction temperature (TJ) can be estimated using the following
equations:
TJ = TT + (ΨJT × P)
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.
or
TJ = TB + (ΨJB × P)
where:
TT is the temperature measured at the top of the package.
P is the total device power dissipation.
TB is the temperature measured at the board.
Table 11. Thermal Resistance
Package
88-Lead LFCSP1
1
θJA
22.6
θJB
5.59
θJC
1.17
ΨJT
0.1
ΨJB
5.22
The exposed pad must be securely connected to the ground plane.
ESD CAUTION
Rev. A | Page 11 of 117
Unit
°C/W
AD9135/AD9136
Data Sheet
88
87
86
85
84
83
82
81
80
79
78
77
76
75
74
73
72
71
70
69
68
67
LDO_BYP2
CVDD12
I120
AVDD33
OUT0+
OUT0–
LDO24
CVDD12
DNC
DNC
DNC
AVDD33
CVDD12
AVDD33
OUT1+
OUT1–
LDO24
CVDD12
DNC
DNC
DNC
AVDD33
PIN CONFIGURATION AND FUNCTION DESCRIPTIONS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
AD9135/AD9136
TOP VIEW
(Not to Scale)
66
65
64
63
62
61
60
59
58
57
56
55
54
53
52
51
50
49
48
47
46
45
IOVDD
CS
SCLK
SDIO
SDO
RESET
IRQ
PROTECT_OUT0
PROTECT_OUT1
PVDD12
PVDD12
GND
GND
DVDD12
SERDIN7+
SERDIN7–
SVDD12
SERDIN6+
SERDIN6–
SVDD12
VTT
SVDD12
NOTES
1. DNC = DO NOT CONNECT. DO NOT CONNECT TO THIS PIN.
2. THE EXPOSED PAD MUST BE SECURELY CONNECTED TO THE GROUND PLANE.
12578-003
SYNCOUT0+
SYNCOUT0–
VTT
SERDIN2+
SERDIN2–
SVDD12
SERDIN3+
SERDIN3–
SVDD12
SVDD12
SVDD12
LDO_BYP1
SIOVDD33
SVDD12
SERDIN4–
SERDIN4+
SVDD12
SERDIN5–
SERDIN5+
VTT
SYNCOUT1–
SYNCOUT1+
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
PVDD12
CLK+
CLK–
PVDD12
SYSREF+
SYSREF–
PVDD12
PVDD12
PVDD12
PVDD12
TXEN0
TXEN1
DVDD12
DVDD12
SERDIN0+
SERDIN0–
SVDD12
SERDIN1+
SERDIN1–
SVDD12
VTT
SVDD12
Figure 3. Pin Configuration
Table 12. Pin Function Descriptions
Pin No.
1
2
Mnemonic
PVDD12
CLK+
3
CLK−
4
5
PVDD12
SYSREF+
6
SYSREF−
7
8
9
10
11
12
13
14
15
PVDD12
PVDD12
PVDD12
PVDD12
TXEN0
TXEN1
DVDD12
DVDD12
SERDIN0+
16
SERDIN0−
17
18
SVDD12
SERDIN1+
Description
1.2 V Supply. PVDD12 provides a clean supply.
PLL Reference/Clock Input, Positive. When the PLL is used, this pin is the positive reference clock input.
When the PLL is not used, this pin is the positive device clock input. This pin is self biased and must be
ac-coupled.
PLL Reference/Clock Input, Negative. When the PLL is used, this pin is the negative reference clock input.
When the PLL is not used, this pin is the negative device clock input. This pin is self biased and must be
ac-coupled.
1.2 V Supply. PVDD12 provides a clean supply.
Positive Reference Clock for Deterministic Latency. This pin is self biased for ac coupling. It can be ac-coupled
or dc-coupled.
Negative Reference Clock for Deterministic Latency. This pin is self biased for ac coupling. It can be ac-coupled
or dc-coupled.
1.2 V Supply. PVDD12 provides a clean supply.
1.2 V Supply. PVDD12 provides a clean supply.
1.2 V Supply. PVDD12 provides a clean supply.
1.2 V Supply. PVDD12 provides a clean supply.
Transmit Enable for DAC0. CMOS levels are determined with respect to IOVDD.
Transmit Enable for DAC1. CMOS levels are determined with respect to IOVDD.
1.2 V Digital Supply.
1.2 V Digital Supply.
Serial Channel Input 0, Positive. CML compliant. SERDIN0+ is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
Serial Channel Input 0, Negative. CML compliant. SERDIN0− is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
1.2 V JESD204B Receiver Supply.
Serial Channel Input 1, Positive. CML compliant. SERDIN1+ is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
Rev. A | Page 12 of 117
Data Sheet
Pin No.
19
Mnemonic
SERDIN1−
20
21
22
23
24
25
26
SVDD12
VTT
SVDD12
SYNCOUT0+
SYNCOUT0−
VTT
SERDIN2+
27
SERDIN2−
28
29
SVDD12
SERDIN3+
30
SERDIN3−
31
32
33
34
35
36
37
SVDD12
SVDD12
SVDD12
LDO_BYP1
SIOVDD33
SVDD12
SERDIN4−
38
SERDIN4+
39
40
SVDD12
SERDIN5−
41
SERDIN5+
42
43
44
45
46
47
48
VTT
SYNCOUT1−
SYNCOUT1+
SVDD12
VTT
SVDD12
SERDIN6−
49
SERDIN6+
50
51
SVDD12
SERDIN7−
52
SERDIN7+
53
54
55
56
57
58
59
60
DVDD12
GND
GND
PVDD12
PVDD12
PROTECT_OUT1
PROTECT_OUT0
IRQ
AD9135/AD9136
Description
Serial Channel Input 1, Negative. CML compliant. SERDIN1− is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
1.2 V JESD204B Receiver Supply.
1.2 V Termination Voltage. Connect VTT to the SVDD12 supply pins.
1.2 V JESD204B Receiver Supply.
Positive LVDS Sync (Active Low) Output Signal Channel Link 0.
Negative LVDS Sync (Active Low) Output Signal Channel Link 0.
1.2 V Termination Voltage. Connect VTT to the SVDD12 supply pins.
Serial Channel Input 2, Positive. CML compliant. SERDIN2+ is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
Serial Channel Input 2, Negative. CML compliant. SERDIN2− is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
1.2 V JESD204B Receiver Supply.
Serial Channel Input 3, Positive. CML compliant. SERDIN3+ is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
Serial Channel Input 3, Negative. CML compliant. SERDIN3− is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
1.2 V JESD204B Receiver Supply.
1.2 V JESD204B Receiver Supply.
1.2 V JESD204B Receiver Supply.
LDO SERDES Bypass. This pin requires a 1 Ω resistor in series with a 1 μF capacitor to ground.
3.3 V Supply for SERDES.
1.2 V JESD204B Receiver Supply.
Serial Channel Input 4, Negative. CML compliant. SERDIN4− is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
Serial Channel Input 4, Positive. CML compliant. SERDIN4+ is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
1.2 V JESD204B Receiver Supply.
Serial Channel Input 5, Negative. CML compliant. SERDIN5− is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
Serial Channel Input 5, Positive. CML compliant. SERDIN5+ is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
1.2 V Termination Voltage. Connect VTT to the SVDD12 supply pins.
Negative LVDS Sync (Active Low) Output Signal Channel Link 1.
Positive LVDS Sync (Active Low) Output Signal Channel Link 1.
1.2 V JESD204B Receiver Supply.
1.2 V Termination Voltage. Connect VTT to the SVDD12 supply pins.
1.2 V JESD204B Receiver Supply.
Serial Channel Input 6, Negative. CML compliant. SERDIN6− is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
Serial Channel Input 6, Positive. CML compliant. SERDIN6+ is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
1.2 V JESD204B Receiver Supply.
Serial Channel Input 7, Negative. CML compliant. SERDIN7− is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
Serial Channel Input 7, Positive. CML compliant. SERDIN7+ is internally terminated to the VTT pin voltage
using a calibrated 50 Ω resistor. This pin is ac-coupled only.
1.2 V Digital Supply.
Ground. Connect GND to the ground plane.
Ground. Connect GND to the ground plane.
1.2 V Supply. PVDD12 provides a clean supply.
1.2 V Supply. PVDD12 provides a clean supply.
Power Detection and Protection Pin Output for DAC1. Pin 58 is high when power protection is in process.
Power Detection and Protection Pin Output for DAC0. Pin 59 is high when power protection is in process.
Interrupt Request (Active Low, Open Drain).
Rev. A | Page 13 of 117
AD9135/AD9136
Pin No.
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
Mnemonic
RESET
SDO
SDIO
SCLK
CS
IOVDD
AVDD33
DNC
DNC
DNC
CVDD12
LDO24
OUT1−
OUT1+
AVDD33
CVDD12
AVDD33
DNC
DNC
DNC
CVDD12
LDO24
OUT0−
OUT0+
AVDD33
I120
CVDD12
LDO_BYP2
EPAD
Data Sheet
Description
Reset. This pin is active low. CMOS levels are determined with respect to IOVDD.
Serial Port Data Output. CMOS levels are determined with respect to IOVDD.
Serial Port Data Input/Output. CMOS levels are determined with respect to IOVDD.
Serial Port Clock Input. CMOS levels are determined with respect to IOVDD.
Serial Port Chip Select. This pin is active low. CMOS levels are determined with respect to IOVDD.
IOVDD Supply for CMOS Input/Output and SPI. Operational for 1.8 V  IOVDD  3.3 V.
3.3 V Analog Supply for DAC Cores.
Do not connect to this pin.
Do not connect to this pin.
Do not connect to this pin.
1.2 V Clock Supply. Place bypass capacitors as near as possible to Pin 71.
2.4 V LDO. Requires a 1 μF capacitor to ground.
DAC1 Negative Current Output.
DAC1 Positive Current Output.
3.3 V Analog Supply for DAC Cores.
1.2 V Clock Supply. Place bypass capacitors as near as possible to Pin 76.
3.3 V Analog Supply for DAC Cores.
Do not connect to this pin.
Do not connect to this pin.
Do not connect to this pin.
1.2 V Clock Supply. Place bypass capacitors as near as possible to Pin 81.
2.4 V LDO. Requires a 1 μF capacitor to ground.
DAC0 Negative Current Output.
DAC0 Positive Current Output.
3.3 V Analog Supply for DAC Cores.
Output Current Generation Pin for DAC Full-Scale Current. Tie a 4 kΩ resistor from the I120 pin to ground.
1.2 V Clock Supply. Place bypass capacitors as near as possible to Pin 87.
LDO Clock Bypass for DAC PLL. This pin requires a 1 Ω resistor in series with a 1 μF capacitor to ground.
Exposed Pad. The exposed pad must be securely connected to the ground plane.
Rev. A | Page 14 of 117
Data Sheet
AD9135/AD9136
TERMINOLOGY
Integral Nonlinearity (INL)
INL is the maximum deviation of the actual analog output from
the ideal output, determined by a straight line drawn from zero
scale to full scale.
Differential Nonlinearity (DNL)
DNL is the measure of the variation in analog value, normalized
to full scale, associated with a 1 LSB change in digital input code.
Offset Error
Offset error is the deviation of the output current from the ideal
of 0 mA. For OUTx+, 0 mA output is expected when all inputs
are set to 0. For OUTx−, 0 mA output is expected when all
inputs are set to 1.
Gain Error
Gain error is the difference between the actual and ideal output
span. The actual span is determined by the difference between
the output when the input is at its minimum code and the
output when the input is at its maximum code.
Output Compliance Range
The output compliance range is the range of allowable voltages
at the output of a current output DAC. Operation beyond the
maximum compliance limits can cause either output stage
saturation or breakdown, resulting in nonlinear performance.
Temperature Drift
Offset drift is a measure of how far from full-scale range (FSR)
the DAC output current is at 25°C (in ppm). Gain drift is a
measure of the slope of the DAC output current across its full
ambient operating temperature range, TA (in ppm/°C).
Power Supply Rejection (PSR)
PSR is the maximum change in the full-scale output as the supplies
are varied from minimum to maximum specified voltages.
Settling Time
Settling time is the time required for the output to reach and
remain within a specified error band around its final value,
measured from the start of the output transition.
Spurious-Free Dynamic Range (SFDR)
SFDR is the difference, in decibels, between the peak amplitude
of the output signal and the peak spurious signal within the dc
to Nyquist frequency of the DAC. Typically, energy in this band
is rejected by the interpolation filters. This specification,
therefore, defines how well the interpolation filters work and
the effect of other parasitic coupling paths on the DAC output.
Signal-to-Noise Ratio (SNR)
SNR is the ratio of the rms value of the measured output signal
to the rms sum of all other spectral components below the Nyquist
frequency, excluding the first six harmonics and dc. The value
for SNR is expressed in decibels.
Interpolation Filter
If the digital inputs to the DAC are sampled at a multiple rate of
fDATA (interpolation rate), a digital filter can be constructed that
has a sharp transition band near fDATA/2. Images that typically
appear around fDAC (output data rate) can be greatly suppressed.
Adjacent Channel Leakage Ratio (ACLR)
ACLR is the ratio in decibels relative to the carrier (dBc)
between the measured power within a channel relative to its
adjacent channel.
Complex Image Rejection
In a traditional two-part upconversion, two images are created
around the second IF frequency. These images have the effect
of wasting transmitter power and system bandwidth. By placing
the real part of a second complex modulator in series with the
first complex modulator, either the upper or lower frequency
image near the second IF can be rejected.
Adjusted DAC Update Rate
The adjusted DAC update rate is defined as the DAC update
rate divided by the smallest interpolating factor. For clarity on
DACs with multiple interpolating factors, the adjusted DAC
update rate for each interpolating factor may be given.
Physical Lane
Physical Lane x refers to SERDINx±.
Logical Lane
Logical Lane x refers to physical lanes after optionally being
remapped by the crossbar block (Register 0x308 to
Register 0x30B).
Link Lane
Link Lane x refers to logical lanes considered per link. When
paging Link 0 (Register 0x300[2] = 0), Link Lane x = Logical
Lane x. When paging Link 1 (Register 0x300[2] = 1, dual link
only), Link Lane x = Logical Lane x + 4.
Rev. A | Page 15 of 117
AD9135/AD9136
Data Sheet
TYPICAL PERFORMANCE CHARACTERISTICS
0
–20
–40
–60
–80
–40
–60
100
200
300
400
500
fOUT (MHz)
12578-104
0
–100
0
MEDIAN
300
400
500
0dBFS
–6dBFS
–9dBFS
–12dBFS
–20
SFDR (dBc)
–40
–60
–80
–40
–60
100
200
300
400
500
fOUT (MHz)
12578-305
0
–100
0
100
200
300
400
500
fOUT (MHz)
12578-108
–80
–100
Figure 8. Single-Tone SFDR vs. fOUT in the First Nyquist Zone
over Digital Back Off, fDAC = 1966 MHz
Figure 5. Single-Tone SFDR vs. fOUT in the First Nyquist Zone,
fDAC = 1966 MHz and 2456 MHz
0
0
fDAC = 983MHz
fDAC = 1228MHz
fDAC = 1474MHz
IN-BAND SECOND HARMONIC
IN-BAND THIRD HARMONIC
MAX DIGITAL SPUR
–20
–20
–40
–40
IMD3 (dBc)
–60
–60
–80
–80
–100
100
200
300
fOUT (MHz)
400
500
12578-106
–100
0
Figure 6. Single-Tone Second and Third Harmonics and Maximum Digital Spur
in the First Nyquist Zone, fDAC = 1966 MHz, 0 dB Back Off
Rev. A | Page 16 of 117
0
100
200
300
400
fOUT (MHz)
Figure 9. Two-Tone Third IMD (IMD3) vs. fOUT,
fDAC = 983 MHz, 1228 MHz, and 1474 MHz
500
12578-109
SFDR (dBc)
200
Figure 7. Single-Tone SFDR vs. fOUT in the First Nyquist Zone
over Digital Back Off, fDAC = 983 MHz
fDAC = 1966MHz
fDAC = 2456MHz
–20
100
fOUT (MHz)
Figure 4. Single-Tone SFDR vs. fOUT in the First Nyquist Zone,
fDAC = 983 MHz, 1228 MHz, and 1474 MHz
0
0
12578-107
–80
–100
SFDR (dBc)
0dBFS
–6dBFS
–9dBFS
–12dBFS
–20
SFDR (dBc)
SFDR (dBc)
0
fDAC = 983MHz
fDAC = 1228MHz
fDAC = 1474MHz
Data Sheet
0
fDAC = 1966MHz
fDAC = 2456MHz
–20
–20
–40
–40
IMD3 (dBc)
–60
–80
1MHz TONE SPACING
16MHz TONE SPACING
35MHz TONE SPACING
–60
–80
0
100
200
300
400
500
fOUT (MHz)
100
200
300
400
500
fOUT (MHz)
–130
0dBFS
–6dBFS
–9dBFS
–12dBFS
–20
0
Figure 13. Two-Tone Third IMD (IMD3) vs. fOUT over Tone Spacing at 0 dB
Back Off, fDAC = 983 MHz and 1966 MHz
Figure 10. Two-Tone Third IMD (IMD3) vs. fOUT,
fDAC = 1966 MHz and 2456 MHz
0
–100
12578-110
–100
fDAC = 983MHz
fDAC = 1966MHz
12578-113
IMD3 (dBc)
0
AD9135/AD9136
fDAC = 983MHz
fDAC = 1228MHz
fDAC = 1474MHz
–135
NSD (dBm/Hz)
IMD3 (dBc)
–140
–40
–60
–145
–150
–155
–160
–80
100
200
300
400
500
fOUT (MHz)
–170
0
300
400
500
500
Figure 14. AD9136 Single-Tone (0 dBFS) NSD vs. fOUT,
fDAC = 983 MHz, 1228 MHz, and 1474 MHz
–130
0dBFS
–6dBFS
–9dBFS
–12dBFS
–20
200
fOUT (MHz)
Figure 11. Two-Tone Third IMD (IMD3) vs. fOUT over Digital Back Off,
fDAC = 983 MHz, Each Tone Is at −6 dBFS
0
100
12578-114
0
12578-111
–100
12578-421
–165
fDAC = 983MHz
fDAC = 1228MHz
fDAC = 1474MHz
–135
NSD (dBm/Hz)
IMD3 (dBc)
–140
–40
–60
–145
–150
–155
–160
–80
–100
0
100
200
300
fOUT (MHz)
400
500
12578-112
–165
Figure 12. Two-Tone Third IMD (IMD3) vs. fOUT over Digital Back Off,
fDAC = 1966 MHz, Each Tone Is at −6 dBFS
Rev. A | Page 17 of 117
–170
0
100
200
300
400
fOUT (MHz)
Figure 15. AD9135 Single-Tone (0 dBFS) NSD vs. fOUT,
fDAC = 983 MHz, 1228 MHz, and 1474 MHz
AD9135/AD9136
Data Sheet
–135
–140
–145
–145
NSD (dBm/Hz)
–140
–150
–155
–165
–165
200
300
400
500
–170
–130
fDAC = 1966MHz
fDAC = 2456MHz
–145
NSD (dBm/Hz)
–145
–150
–155
–165
–165
400
500
fOUT (MHz)
–170
12578-422
300
–130
–145
–145
NSD (dBm/Hz)
–140
–150
–155
–165
300
400
500
fOUT (MHz)
12578-116
–165
200
400
500
–155
–160
100
300
–150
–160
0
200
0dBFS
–6dBFS
–9dBFS
–12dBFS
–135
–140
–170
100
Figure 20. AD9136 Single-Tone NSD vs. fOUT over Digital Back Off,
fDAC = 1966 MHz
0dBFS
–6dBFS
–9dBFS
–12dBFS
–135
0
fOUT (MHz)
Figure 17. AD9135 Single-Tone (0 dBFS) NSD vs. fOUT,
fDAC = 1966 MHz and 2456 MHz
–130
500
–155
–160
200
400
–150
–160
100
300
0dBFS
–6dBFS
–9dBFS
–12dBFS
–135
–140
0
200
Figure 19. AD9135 Single-Tone NSD vs. fOUT over Digital Back Off,
fDAC = 983 MHz
–140
–170
100
fOUT (MHz)
–130
–135
0
12578-117
100
12578-115
0
Figure 16. AD9136 Single-Tone (0 dBFS) NSD vs. fOUT,
fDAC = 1966 MHz and 2456 MHz
NSD (dBm/Hz)
–155
–160
fOUT (MHz)
NSD (dBm/Hz)
–150
–160
–170
0dBFS
–6dBFS
–9dBFS
–12dBFS
Figure 18. AD9136 Single-Tone NSD vs. fOUT over Digital Back Off,
fDAC = 983 MHz
–170
0
100
200
300
400
500
fOUT (MHz)
Figure 21. AD9135 Single-Tone NSD vs. fOUT over Digital Back Off,
fDAC = 1966 MHz
Rev. A | Page 18 of 117
12578-424
–135
NSD (dBm/Hz)
–130
fDAC = 1966MHz
fDAC = 2456MHz
12578-423
–130
Data Sheet
–130
PLL OFF
PLL ON
–135
fDAC = 983MHz
fDAC = 1966MHz
–140
NSD (dBm/Hz)
AD9135/AD9136
–145
–150
–155
–160
–170
0
100
200
300
400
500
fOUT (MHz)
12578-118
12578-317
–165
Figure 22. AD9136 Single-Tone NSD (0 dBFS) vs. fOUT, fDAC = 983 MHz and
1966 MHz, PLL On and Off
–130
PLL OFF
PLL ON
–135
fDAC = 983MHz
fDAC = 1966MHz
–140
NSD (dBm/Hz)
Figure 25. AD9136 Four-Carrier W-CDMA ACLR, fOUT = 30 MHz,
fDAC = 983 MHz, 2× Interpolation, PLL Frequency = 122 MHz
–145
–150
–155
12578-417
–160
–170
0
100
200
300
400
500
fOUT (MHz)
12578-425
–165
Figure 23. AD9135 Single-Tone NSD (0 dBFS) vs. fOUT, fDAC = 983 MHz and
1966 MHz, PLL On and Off
–60
fOUT = 30MHz
fOUT = 200MHz
fOUT = 400MHz
–80
PHASE NOISE (dBc/Hz)
Figure 26. AD9135 Four-Carrier W-CDMA ACLR, fOUT = 30 MHz,
fDAC = 983 MHz, 2× Interpolation, PLL Frequency = 122 MHz
PLL OFF
PLL ON
–100
–120
–140
10
100
1k
10k
100k
OFFSET FREQUENCY (Hz)
1M
10M
12578-119
–180
12578-318
–160
Figure 24. AD9136 Single-Tone Phase Noise vs. Offset Frequency over fOUT,
fDAC = 2.0 GHz, PLL On and Off
Rev. A | Page 19 of 117
Figure 27. AD9136 Four-Carrier W-CDMA ACLR, fOUT = 122 MHz,
fDAC = 983 MHz, 2× Interpolation, PLL Frequency = 122 MHz
Data Sheet
12578-418
12578-320
AD9135/AD9136
Figure 31. AD9136 Four-Carrier W-CDMA ACLR, fOUT = 122 MHz,
fDAC = 1966 MHz, 4× Interpolation, PLL Frequency = 122 MHz
11675-420
12578-319
Figure 28. AD9135 Four-Carrier W-CDMA ACLR, fOUT = 122 MHz,
fDAC = 983 MHz, 2× Interpolation, PLL Frequency = 122 MHz
Figure 32. AD9135 Four-Carrier W-CDMA ACLR, fOUT = 122 MHz,
fDAC = 1966 MHz, 4× Interpolation, PLL Frequency = 122 MHz
Figure 30. AD9135 Four-Carrier W-CDMA ACLR, fOUT = 30 MHz,
fDAC = 1966 MHz, 4× Interpolation, PLL Frequency = 245 MHz
12578-433
12578-419
Figure 29. AD9136 Four-Carrier W-CDMA ACLR, fOUT = 30 MHz,
fDAC = 1966 MHz, 4× Interpolation, PLL Frequency = 245 MHz
Figure 33. AD9136 Output Performance of an Ultra Wideband (900 MHz)
QAM Signal, fDAC = 2 GHz, 1× Interpolation, Inverse sinc On,
JESD204B Mode 11
Rev. A | Page 20 of 117
Data Sheet
1300
1200
1100
1000
1500
2000
2500
Figure 34. Total Power Consumption vs. fDAC over Interpolation,
8 SERDES Lanes Enabled, 2 DACs Enabled, Digital Gain, Inverse Sinc and
DAC PLL Disabled
300
1
SUPPLY CURRENT (mA)
200
40
3
4
5
6
7
8
Figure 36. SVDD12 Current vs. Lane Rate over Number of SERDES Lanes and
Supply Voltage Setting
250
60
2
LANE RATE (Gbps)
DVDD12
CVDD12
PVDD12
AVDD33
1.2V SUPPLY
1.3V SUPPLY
3.3V SUPPLY
150
100
400
600
800
1000
1200
1400
1600
fDAC (MHz)
Figure 35. Power Consumption vs. fDAC over Digital Functions
0
400
600
800
100
fDAC (MHz)
1200
1400
1600
12578-329
50
20
12578-327
POWER CONSUMPTION (mW)
400
100
PLL (fDAC /fREF RATIO:4)
DIGITAL GAIN
INVERSE SINC
80
0
200
500
12578-328
1000
fDAC (MHz)
100
1.2V SVDD12 SUPPLY
1.3V SVDD12 SUPPLY
200
900
500
120
2 LANES
4 LANES
8 LANES
600
SVDD12 CURRENT (mA)
1400
700
1×
2×
4×
12578-326
TOTAL POWER CONSUMPTION (mW)
1500
AD9135/AD9136
Figure 37. DVDD12, CVDD12, PVDD12, and AVDD33 Supply Currents vs. fDAC
over Supply Voltage Setting, 2 DACs Enabled
Rev. A | Page 21 of 117
AD9135/AD9136
Data Sheet
THEORY OF OPERATION
The AD9135/AD9136 are 11-/16-bit, dual DACs with a SERDES
interface. Figure 2 shows a detailed functional block diagram of
the AD9135/AD9136. Eight high speed serial lanes carry data at
a maximum speed of 10.64 Gbps, and a 2120 MSPS input data rate
to each DAC. Compared to either LVDS or CMOS interfaces, the
SERDES interface simplifies pin count, board layout, and input
clock requirements to the device.
The clock for the input data is derived from the device clock
(required by the JESD204B specification). This device clock can
be sourced with a PLL reference clock used by the on-chip PLL
to generate a DAC clock or a high fidelity direct external DAC
sampling clock. The device can be configured to operate in one-,
two-, four-, or eight-lane modes, depending on the required
input data rate.
The digital datapath of the AD9135/AD9136 offers four
interpolation modes (1×, 2×, 4×, and 8×) through three half-band
filters with a maximum DAC sample rate of 2.8 GSPS. An inverse
sinc filter is provided to compensate for sinc related roll-off.
The AD9135/AD9136 DAC cores provide a fully differential
current output with a nominal full-scale current of 20 mA. The
full-scale current, IOUTFS, is user adjustable to between 13.9 mA
and 27.0 mA, typically. The differential current outputs are
complementary and are optimized for easy integration with the
Analog Devices the ADRF6720 AQM. The AD9135/AD9136
are capable of multichip synchronization that can both
synchronize multiple DACs and establish a constant and
deterministic latency (latency locking) path for the DACs.
The latency for each of the DACs remains constant from link
establishment to link establishment. An external alignment
(SYSREF±) signal makes the AD9135/AD9136 Subclass 1
compliant. Several modes of SYSREF± signal handling are
available for use in the system.
An SPI configures the various functional blocks and monitors
their statuses. The various functional blocks and the data
interface must be set up in a specific sequence for proper
operation (see the Device Setup Guide section). Simple SPI
initialization routines set up the JESD204B link and are included
in the evaluation board package. The following sections describe
the various blocks of the AD9135/AD9136 in greater detail.
Descriptions of the JESD204B interface, control parameters, and
various registers to set up and monitor the device are provided.
The recommended start-up routine reliably sets up the data link.
Rev. A | Page 22 of 117
Data Sheet
AD9135/AD9136
SERIAL PORT OPERATION
The serial port is a flexible, synchronous serial communications
port that allows easy interfacing with many industry-standard
microcontrollers and microprocessors. The serial input/output
(I/O) is compatible with most synchronous transfer formats,
including both the Motorola SPI and Intel® SSR protocols. The
interface allows read/write access to all registers that configure
the AD9135/AD9136. MSB first or LSB first transfer formats are
supported. The serial port interface can be configured as a 4-wire
interface or a 3-wire interface in which the input and output share
a single-pin I/O (SDIO).
CS 65
The serial clock pin synchronizes data to and from the device
and runs the internal state machines. The maximum frequency
of SCLK is 10 MHz. All data input is registered on the rising edge
of SCLK. All data is driven out on the falling edge of SCLK.
SPI
PORT
12578-044
SCLK 64
Figure 38. Serial Port Interface Pins
There are two phases to a communication cycle with the
AD9135/AD9136. Phase 1 is the instruction cycle (the writing
of an instruction byte into the device), coincident with the first
16 SCLK rising edges. The instruction word provides the serial
port controller with information regarding the data transfer cycle,
Phase 2 of the communication cycle. The Phase 1 instruction
word defines whether the upcoming data transfer is a read or
write, along with the starting register address for the following
data transfer.
A logic high on the CS pin followed by a logic low resets the
serial port timing to the initial state of the instruction cycle.
From this state, the next 16 rising SCLK edges represent the
instruction bits of the current I/O operation.
The remaining SCLK edges are for Phase 2 of the communication
cycle. Phase 2 is the actual data transfer between the device and
the system controller. Phase 2 of the communication cycle is a
transfer of one or more data bytes. Eight × N SCLK cycles are
needed to transfer N bytes during the transfer cycle. Registers
change immediately upon writing to the last bit of each transfer
byte.
DATA FORMAT
The instruction byte contains the information shown in Table 13.
Table 13. Serial Port Instruction Word
I[15] (MSB)
R/W
SERIAL PORT PIN DESCRIPTIONS
Serial Clock (SCLK)
SDO 62
SDIO 63
A14 to A0, Bit 14 to Bit 0 of the instruction word, determine the
register that is accessed during the data transfer portion of the
communication cycle. For multibyte transfers, A[14:0] is the
starting address. The remaining register addresses are generated
by the device based on the address increment bits. If the address
increment bits are set high (Register 0x000, Bit 5 and Bit 2),
multibyte SPI writes start at A[14:0] and increment by 1 every
8 bits sent/received. If the address increment bits are set to 0,
the address decrements by 1 every 8 bits.
I[14:0]
A[14:0]
Chip Select (CS)
An active low input starts and gates a communication cycle.
CS allows more than one device to be used on the same serial
communications lines. The SDIO pin goes to a high impedance
state when this input is high. During the communication cycle,
chip select must stay low.
Serial Data I/O (SDIO)
This pin is a bidirectional data line. In 4-wire mode, this pin
acts as the data input, and SDO acts as the data output.
SERIAL PORT OPTIONS
The serial port can support both MSB first and LSB first data
formats. This functionality is controlled by the LSB first bits
(Register 0x000, Bit 6 and Bit 1). The default is MSB first
(LSBFIRST/LSBFIRST_M = 0).
When the LSB first bits = 0 (MSB first), the instruction and data
bits must be written from MSB to LSB. R/W is followed by
A[14:0] as the instruction word, and D[7:0] is the data-word.
When the LSB first bits = 1 (LSB first), the opposite is true.
A[0:14] is followed by R/W, which is subsequently followed by
D[0:7].
The serial port supports a 3-wire or 4-wire interface. When the
SDO active bits = 1 (Register 0x000, Bit 4 and Bit 3), a 4-wire
interface with a separate input pin (SDIO) and output pin (SDO)
is used. When the SDO active bits = 0, the SDO pin is unused
and the SDIO pin is used for both input and output.
R/W, Bit 15 of the instruction word, determines whether a read
or a write data transfer occurs after the instruction word write.
Logic 1 indicates a read operation, and Logic 0 indicates a write
operation.
Rev. A | Page 23 of 117
AD9135/AD9136
Data Sheet
INSTRUCTION CYCLE
Multibyte data transfers can be performed as well. This is done
by holding the CS pin low for multiple data transfer cycles
(eight SCLKs) after the first data transfer word following the
instruction cycle. The first eight SCLKs following the
instruction cycle read from or write to the register provided in
the instruction cycle. For each additional eight SCLK cycles, the
address is either incremented or decremented and the read/write
occurs on the new register. The direction of the address can be set
using the address increment bits (Register 0x000, Bit 5 and Bit 2).
When the address increment bits is 1, the multicycle addresses
are incremented. When the address increment bits is 0, the
addresses are decremented. A new write cycle can always be
initiated by bringing CS high and then low again.
DATA TRANSFER CYCLE
CS
SDIO
R/W A14 A13
A3
A2 A1
A0 D7N D6N D5N
D30 D20 D10 D00
Figure 39. Serial Register Interface Timing, MSB First, ADDRINC = 0
INSTRUCTION CYCLE
DATA TRANSFER CYCLE
CS
SDIO
A0
A1 A2
A12 A13 A14 R/W D0 0 D10 D20
D4N D5N D6N D7N
Figure 40. Serial Register Interface Timing, LSB First, ADDRINC = 1
CS
SDIO
DATA BIT n
DATA BIT n – 1
Figure 41. Timing Diagram for Serial Port Register Read
tSCLK
CS
tPWH
tPWL
SDIO
tDH
INSTRUCTION BIT 15
INSTRUCTION BIT 14
Figure 42. Timing Diagram for Serial Port Register Write
Rev. A | Page 24 of 117
12578-047
SCLK
12578-048
SCLK
tDV
tDS
12578-046
SCLK
To prevent confusion and to ensure consistency between devices,
the chip tests the first nibble following the address phase, ignoring
the second nibble. This test is completed independently from the
LSB first bit and ensures that there are extra clock cycles
following the soft reset bits (Register 0x000, Bit 0 and Bit 7).
This only applies when writing to Register 0x000.
tDCS
12578-045
SCLK
Data Sheet
AD9135/AD9136
CHIP INFORMATION
Register 0x003 to Register 0x006 contain chip information, as shown in Table 14.
Table 14. Chip Information
Information
Chip Type
Product ID
Product Grade
Device Revision
Description
The product type is high speed DAC, which is represented by a code of 0x04 in Register 0x003.
Eight MSBs in Register 0x005 and eight LSBs in Register 0x004. The product ID is 0x9144.
Register 0x006[7:4]. The product grade is 0x6 for the AD9136 and 0x4 for the AD9135.
Register 0x006[3:0]. The device revision is 0x6.
Rev. A | Page 25 of 117
AD9135/AD9136
Data Sheet
DEVICE SETUP GUIDE
OVERVIEW
The sequence of steps to properly set up the AD9135/AD9136 is
as follows:
1.
2.
3.
4.
5.
6.
7.
Set up the SPI interface, power up necessary circuit blocks,
make the required writes to the configuration registers, and set
up the DAC clocks (see the Step 1: Start Up the DAC section).
Set the digital features of the AD9135/AD9136 (see the
Step 2: Digital Datapath section).
Set up the JESD204B links (see the Step 3: Transport Layer
section).
Set up the physical layer of the SERDES interface (see the
Step 4: Physical Layer section).
Set up the data link layer of the SERDES interface (see the
Step 5: Data Link Layer section).
Check for errors (see the Step 6: Optional Error Monitoring
section).
Optionally, enable any needed features as described in the
Step 7: Optional Features section.
The register writes listed in Table 15 to Table 21 give the register
writes necessary to set up the AD9135/AD9136. Consider printing
this setup guide and filling in the Value column with the appropriate variable values for the conditions of the desired application.
The notation 0x, shaded in gray, indicates register settings that
must be filled in by the user. To fill in the unknown register values,
select the correct settings for each variable listed in the Variable
column of Table 15 to Table 21. The Description column describes
how to set variables or provides a link to a section where this is
described. A variable is noted by concatenating multiple terms. For
example, PdDACs is a variable corresponding to the value that is
determined for Register 0x011[6:3] in the Device Setup Guide
section.
STEP 1: START UP THE DAC
This section describes how to set up the SPI interface, power up
necessary circuit blocks, write to the required configuration
registers, and set up the DAC clocks, listed in Table 15.
Table 15. Power-Up and DAC Initialization Settings
Addr.
0x000
0x000
0x011
Bit No.
7
[6:3]
2
Value1
0xBD
0x3C
0x
0
Variable
PdDACs
0x080
0
0x
PdClocks
0x081
0x
PdSysref
1
Description
Soft reset.
Deassert reset, set 4-wire SPI.
Power up band gap.
PdDACs = 0x05 to power up
DAC0/DAC1. PdDACs = 0x07
if only using DAC0.
Power up master DAC.
PdClocks = 0 if DAC0/DAC1
are being used. PdClocks =
0x40 if only using DAC0.
PdSysref = 0x00 for Subclass 1.
PdSysref = 0x10 for Subclass 0.
See the Subclass Setup section
for details on subclass.
The registers in Table 16 must be written from their default
values to be the values listed in the table for the device to work
correctly. These registers must be written after any soft reset,
hard reset, or power-up occurs.
Table 16. Required Device Configurations
Addr.
0x12D
0x146
0x2A4
0x232
0x333
Value
0x8B
0x01
0xFF
0xFF
0x01
Description
Digital datapath configuration
Digital datapath configuration
Clock configuration
SERDES interface configuration
SERDES interface configuration
If using the optional DAC PLL, also set the registers in Table 17.
Table 17. Optional DAC PLL Configuration Procedure
Addr.
0x087
Value1
0x62
0x088
0xC9
0x089
0x0E
0x08A
0x12
0x08D
0x7B
0x1B0
0x00
0x1B9
0x24
0x1BC
0x0D
0x1BE
0x02
0x1BF
0x8E
0x1C0
0x2A
0x1C1
0x2A
0x1C4
0x08B
0x08C
0x085
Various
0x7E
0x
0x
0x
0x
0x083
0x10
1
Variable
LODivMode
RefDivMode
BCount
LookUpVals
Description
Optimal DAC PLL loop filter
settings
Optimal DAC PLL loop filter
settings
Optimal DAC PLL loop filter
settings
Optimal DAC PLL charge pump
settings
Optimal DAC LDO settings for
DAC PLL
Power DAC PLL blocks when
power machine disabled
Optimal DAC PLL charge pump
settings
Optimal DAC PLL VCO control
settings
Optimal DAC PLL VCO power
control settings
Optimal DAC PLL VCO calibration
settings
Optimal DAC PLL lock counter
length setting
Optimal DAC PLL charge pump
setting
Optimal DAC PLL varactor settings
See the DAC PLL Setup section
See the DAC PLL Setup section
See the DAC PLL Setup section
See Table 25 in the DAC PLL Setup
section for the list of register
addresses and values for each.
Enable the DAC PLL2
0x denotes a register value that the user must fill in. See the Variable and
Description columns for information on selecting the appropriate register
value.
2
Verify that Register 0x084[1] reads back 1 after enabling the DAC PLL to
indicate that the DAC PLL has locked.
0x denotes a register value that the user must fill in. See the Variable and
Description columns for information on selecting the appropriate register value.
Rev. A | Page 26 of 117
Data Sheet
AD9135/AD9136
STEP 2: DIGITAL DATAPATH
Table 19. Transport Layer Settings
This section describes which interpolation filters to use and
how to set the data format being used. Additional digital
features are available, including digital gain scaling and an
inverse sinc filter used to improve pass-band flatness. Table 22
provides further details on the feature blocks available.
Addr.
0x200
0x201
Bit
No.
0x300
Value1
0x00
0x
Variable
UnusedLanes
0x
6
CheckSumMode
3
DualLink
2
CurrentLink
Table 18. Digital Datapath Settings
Addr.
0x112
Bit
No.
Variable
InterpMode
Description
Select interpolation mode;
see the Interpolation section.
0x450
0x
DID
DataFmt = 0 if twos
complement; DataFmt = 1
if unsigned binary.
0x451
0x
BID
0x denotes a register value that the user must fill in. See the Variable and
Description columns for information on selecting the appropriate register value.
0x452
0x
LID
0x110
0x
7
1
Value1
0x
DataFmt
STEP 3: TRANSPORT LAYER
0x453
This section describes how to set up the JESD204B links. The
parameters are determined by the desired JESD204B operating
mode. See the JESD204B Setup section for details.
Table 19 shows the register settings for the transport layer. If
using dual-link mode, perform writes from Register 0x300 to
Register 0x47D with CurrentLink = 0 and then repeat the same
set of register writes with CurrentLink = 1 (Register 0x200 and
Register 0x201 need only be written once).
See the JESD204B
Setup section.
See the JESD204B
Setup section.
See the JESD204B
Setup section.
Set DID to match the
device ID sent by the
transmitter.
Set BID to match the
bank ID sent by the
transmitter.
Set LID to match the
lane ID sent by the
transmitter.
0x
7
Scrambling
[4:0]
L − 12
0x454
0x
F − 12
0x455
0x
K − 12
0x456
0x
M − 12
0x457
0x458
0x
0x
N − 12
[7:5]
Subclass
NP − 12
[4:0]
0x459
See the JESD204B
Setup section.
See the JESD204B
Setup section.
See the JESD204B
Setup section.
See the JESD204B
Setup section.
See the JESD204B
Setup section.
N = 16.
See the JESD204B
Setup section.
NP = 16.
0x
[7:5]
JESDVer
[4:0]
S − 12
0x45A
JESDVer = 1 for
JESD204B, JESDVer = 0
for JESD204A.
See the JESD204B
Setup section.
0x
7
[4:0]
HD
0x45D
0
0x
CF
Lane0Checksum
0x46C
0x
Lanes
0x476
0x
F
0x47D
0x
Lanes
1
Description
Power up the interface.
See the JESD204B
Setup section.
See the JESD204B
Setup section.
CF must equal 0.
See the JESD204B
Setup section.
Deskew lanes. See the
JESD204B Setup
section.
See the JESD204B
Setup section.
Enable lanes. See the
JESD204B Setup
section.
0x denotes a register value that the user must fill in. See the Variable and
Description columns for information on selecting the correct register value.
2
This JESD204B link parameter is programmed in n − 1 notation as noted. For
example, if the setup requires L = 8 (8 lanes per link), program L − 1 or 7 into
Register 0x453[4:0].
Rev. A | Page 27 of 117
AD9135/AD9136
Data Sheet
STEP 4: PHYSICAL LAYER
STEP 5: DATA LINK LAYER
This section describes how to set up the physical layer of the
SERDES interface. In this section, the input termination settings
are configured along with the CDR sampling and SERDES PLL.
This section describes how to set up the data link layer of the
SERDES interface. This section deals with SYSREF signal
processing, setting deterministic latency, and establishing the
link.
Table 20. Device Configurations and Physical Layer Settings
Addr.
0x2AA
0x2AB
0x2B1
0x2B2
0x2A7
0x2AE
0x314
0x230
Bit
No.
Table 21. Data Link Layer Settings
Value1
0xB7
0x87
0xB7
0x87
0x01
0x01
0x01
0x
5
[4:2]
1
0x206
0x206
0x289
2
[1:0]
Halfrate
0x2
OvSmp
0x00
0x01
0x
1
PLLDiv
0x284
0x285
0x286
0x287
0x62
0xC9
0x0E
0x12
0x28A
0x28B
0x7B
0x00
0x290
0x89
0x294
0x24
0x296
0x297
0x299
0x03
0x0D
0x02
0x29A
0x8E
0x29C
0x2A
0x29F
0x78
0x2A0
0x06
0x280
0x268
0x01
0x
[7:6]
[5:0]
1
Variable
EqMode
0x22
Description
SERDES interface termination
setting
SERDES interface termination
setting
Addr.
0x301
Autotune PHY setting
Autotune PHY setting
SERDES SPI configuration
Set up the CDR; see the SERDES
Clocks Setup section
SERDES PLL default configuration
Set up the CDR; see the SERDES
Clocks Setup section
Reset the CDR
Release the CDR reset
SERDES PLL configuration
Set the CDR oversampling for
PLL; see the SERDES Clocks
Setup section
Optimal SERDES PLL loop filter
Optimal SERDES PLL loop filter
Optimal SERDES PLL loop filter
Optimal SERDES PLL charge
pump
Optimal SERDES PLL VCO LDO
Optimal SERDES PLL
configuration
Optimal SERDES PLL VCO
varactor
Optimal SERDES PLL charge
pump
Optimal SERDES PLL VCO
Optimal SERDES PLL VCO
Optimal SERDES PLL
configuration
Optimal SERDES PLL VCO
varactor
Optimal SERDES PLL charge
pump
Optimal SERDES PLL VCO
varactor
Optimal SERDES PLL VCO
varactor
Enable the SERDES PLL2
See the Equalization Mode
Setup section
Required value (default)
Value1
0x
Variable
Subclass
0x304
0x
LMFCDel
0x305
0x
LMFCDel
0x306
0x
LMFCVar
0x307
0x
LMFCVar
0x03A
0x01
0x03A
0x81
0x03A
0xC1
SYSREF±
Signal
0x308
to
0x30B
0x
XBarVals
0x334
0x
InvLanes
0x300
1
0x denotes a register value that the user must fill in. See the Variable and
Description columns for information on selecting the correct register value.
2
Verify that Register 0x281[0] reads back 1 after enabling the SERDES PLL to
indicate that the SERDES PLL has locked.
Bit
No.
0x
6
3
2
CheckSumMode
DualLink
CurrentLink
[1:0]
EnLinks
Description
See the JESD204B
Setup section.
See the Link Latency
Setup section.
See the Link Latency
section.
See the Link Latency
Setup section.
See the Link Latency
Setup section.
Set sync mode = oneshot sync; see the
Syncing LMFC Signals
section for other sync
options.
Enable the sync
machine.
Arm the sync
machine.
If Subclass = 1, ensure
that at least one
SYSREF± edge is sent
to the device.2
If remapping lanes,
set up crossbar; see
the Crossbar Setup
section.
Invert the polarity of
the desired logical
lanes. Bit x of InvLanes
must be a 1 for each
Logical Lane x to
invert.
Enable the links.
See the JESD204B
Setup section.
Set to 0 to access
Link 0 status or 1 for
Link 1 status
readbacks. See the
JESD204B Setup
section.
EnLinks = 3 if
DualLink = 1 (enables
Link 0 and Link 1);
EnLinks = 1 if
DualLink = 0 (enables
Link 0 only).
0x denotes a register value that the user must fill in. See the Variable and
Description columns for information on selecting the correct register value.
2
Verify that Register 0x03B[3] reads back 1 after sending at least one SYSREF±
edge to the device to indicate that the LMFC sync machine has properly locked.
Rev. A | Page 28 of 117
Data Sheet
AD9135/AD9136
STEP 6: OPTIONAL ERROR MONITORING
Table 22. Optional Features
For JESD204B error monitoring, see the JESD204B Error
Monitoring section. For other error checks, see the Interrupt
Request Operation section.
Feature
Inverse Sinc
Default
On
Digital Gain
2.7 dB
DC Offset
Off
Group Delay
0
Downstream
Protection
Off
Self Calibration
Off
STEP 7: OPTIONAL FEATURES
There are a number of optional features that can be enabled.
Table 22 provides links to the sections describing each feature.
These features can be enabled during the Digital Datapath
configuration step, or after the link is set up, because it is not
required to configure them for the link to be established, unlike
interpolation. Unless otherwise noted, these features are paged
as described in the DAC Paging section. Paging is particularly
important for DAC specific settings like digital gain and dc offset.
Rev. A | Page 29 of 117
Description
Improves pass-band flatness. See the
Inverse Sinc section.
Multiplies data by a factor. Can
compensate inverse sinc usage or
balance I/Q amplitude. See the Digital
Gain section.
Used to cancel LO leakage. See the
DC Offset section.
Used to control overall latency. See
the Group Delay section.
Used to protect downstream
components. See the Downstream
Protection section.
Used to improve DAC linearity. Not
paged by the dual paging register.
See the Self Calibration section.
AD9135/AD9136
Data Sheet
DAC PLL SETUP
INTERPOLATION
This section explains how to select appropriate values for
LODivMode, RefDivMode, and BCount in the Step 1: Start Up
the DAC section. These parameters depend on the desired DAC
clock frequency (fDAC) and DAC reference clock frequency (fREF).
When using the DAC PLL, the reference clock signal is applied
to the CLK± differential pins (Pin 2 and Pin 3).
The transmit path can use zero to three cascaded interpolation
filters, which each provides a 2× increase in output data rate and
a low-pass function. Table 26 shows the different interpolation
modes and the respective usable bandwidth along with the
maximum fDATA rate attainable.
Table 23. DAC PLL LODivMode Settings
LO_DIV_MODE,
Register 0x08B[1:0]
1
2
3
DAC Frequency Range (MHz)
1500 to 2800
750 to 1500
420 to 750
Table 24. DAC PLL RefDivMode Settings
DAC PLL Reference
Frequency (fREF) (MHz)
35 to 80
80 to 160
160 to 320
320 to 640
640 to 1000
Divide by
(RefDivFactor)
1
2
4
8
16
REF_DIV_MODE,
Register 0x08C[2:0]
0
1
2
3
4
The VCO frequency (fVCO) is related to the DAC clock frequency
according to the following equation:
fVCO = fDAC × 2LODivMode + 1
where 6 GHz  fVCO  12 GHz.
BCount must be between 6 and 127 and is calculated based on
fDAC and fREF as follows:
BCount = floor((fDAC)/(2 × fREF/RefDivFactor))
where RefDivFactor = 2RefDivMode (see Table 24).
Finally, to finish configuring the DAC PLL, set the VCO control
registers up as described in Table 25 based on the VCO frequency
(fVCO). Write the registers listed in the table with the corresponding
LookUpVals.
Table 25. VCO Control Lookup Table Reference
VCO Frequency
Range (GHz)
fVCO < 6.3
6.3 ≤ fVCO < 7.25
fVCO ≥ 7.25
Register
0x1B4
Setting
0x08
0x09
0x09
Register
0x1B6
Setting
0x03
0x03
0x13
Register
0x1BB
Setting
0x07
0x06
0x06
For more information on the DAC PLL, see the DAC Input
Clock Configurations section.
Table 26. Interpolation Modes and Their Usable Bandwidth
Interpolation
Mode
1× (bypass)
InterpMode
0x00
Usable
Bandwidth
0.5 × fDATA
2×
0x01
0.4 × fDATA
4×
8×
0x03
0x04
0.4 × fDATA
0.4 × fDATA
Maximum fDATA
(MSPS)
2120 (SERDES
limited)
1060 (SERDES
limited)
700
350
The usable bandwidth is defined for 1×, 2×, 4×, and 8× modes
as the frequency band over which the filters have a pass-band
ripple of less than ±0.001 dB and an image rejection of greater
than 85 dB. For more information, see the Interpolation Filters
section.
JESD204B SETUP
This section explains how to select a JESD204B operating mode
for a desired application. This section in turn defines appropriate
values for CheckSumMode, UnusedLanes, DualLink, CurrentLink,
Scrambling, L, F, K, M, N, NP, Subclass, S, HD, Lane0Checksum,
and Lanes needed for the Step 3: Transport Layer section.
Note that DualLink, Scrambling, F, K, N, NP, S, HD, and
Subclass must be set the same on the transmit side. For Mode 8,
Mode 9, and Mode 10, the number of converters (M) and the
lane count (L) on the transmit side must also match the receive
side. For Mode 11, Mode 12, and Mode 13, M and L on the
transmit side do not match the receive side. See Table 28 for
details.
For a summary of how a JESD204B system works and what each
parameter means, see the JESD204B Serial Data Interface section.
Available Operating Modes
Table 27. JESD204B Operating Modes (Single- or Dual-Link)
(Applies to Both JESD204B Tx and Rx)
Parameter
M (Converter Count)
L (Lane Count)
S ((Samples per Converter) per Frame)
F ((Octets per Frame) per Lane)
1
81
1
4
2
1
Mode
9
1
2
1
1
10
1
1
1
2
Mode 8 can only be used with 1× interpolation. Other interpolation options
are not available in this mode.
Rev. A | Page 30 of 117
Data Sheet
AD9135/AD9136
Table 28. JESD204B Operating Modes (Single-Link Only)
Parameter
M (Converter Count) (Tx Setting)
AD9135 and AD9136 M Setting1
(Rx Setting)
L (Lane Count) (Tx Setting)
AD9135 and AD9136 L Setting1
(Rx Setting)
S ((Samples per Converter) per Frame)
F ((Octets per Frame) per Lane)
112
2
1
Mode
12
2
1
13
2
1
8
4
4
2
2
1
2
1
1
1
1
2
1
Note that for Mode 11 to Mode 13, the M and L parameters programmed on
the receive side do not match the parameters on the transmit side. The
parameters on the transmit side reflect the true number of converters and
lanes per link.
2
Mode 11 can only be used with 1× interpolation. Other interpolation options
are not available in this mode.
For a particular application, the number of converters to use per
link (M) and the fDATA (DataRate) are known. The LaneRate and
number of lanes (L) can be traded off as follows:
DataRate = (DACRate)/(InterpolationFactor)
LaneRate = (20 × DataRate × M)/L
where LaneRate is between 1.44 Gbps and 10.64 Gbps.
Octets per frame per lane (F) and samples per convertor per
frame (S) define how the data is packed. If F = 1, the high density
setting must be set to one (HD = 1). Otherwise, set HD = 0.
Converter resolution and bits per sample (N and NP) must both
be set to 16. Frames per multiframe (K) must be set to 32 for
Mode 8, Mode 9, Mode 11, and Mode 12. Other modes can use
either K = 16 or K = 32.
DualLink
CurrentLink
Set CurrentLink to either 0 or 1 depending on whether Link 0
or Link 1, respectively, needs to be configured.
Lanes
Lanes is used to enable and deskew particular lanes in two
thermometer coded registers. The lanes setting for each of the
modes is given in Table 29.
Table 29. Lanes Setting per JESD Operating Mode
JESD Mode ID
Lanes
8
0x0F
9
0x03
10
0x01
11
0xFF
12
0x33
13
0x11
UnusedLanes
UnusedLanes is used to turn off unused circuit blocks to save
power. Each physical lane that is not being used (SERDINx±)
must be powered off by writing a 1 to the corresponding bit of
Register 0x201.
For example, if using Mode 9 in dual-link mode and sending
data on SERDIN0±, SERDIN1±, SERDIN4±, and SERDIN5±,
set UnusedLanes = 0xCC to power off Physical Lane 2, Lane 3,
Lane 6, and Lane 7.
CheckSumMode
CheckSumMode must match the checksum mode used on the
transmit side. If the checksum used is the sum of fields in the
link configuration table, CheckSumMode = 0. If summing the
registers containing the packed link configuration fields,
CheckSumMode = 1. For more information on the how to
calculate the two checksum modes, see the Lane0Checksum
section.
Lane0Checksum
DualLink sets up two independent JESD204B links, which
allows each link to be reset independently. If this functionality
is desired, set DualLink to 1; if a single link is desired, set
DualLink to 0. Note that Link 0 and Link 1 must have identical
parameters. The operating modes available when using dual- or
single-link mode are shown in Table 27. Additional single-link
modes that are available are shown in Table 28.
Scrambling
Scrambling is a feature that makes the spectrum of the link data
independent. This avoids spectral peaking and provides some
protection against data dependent errors caused by frequency
selective effects in the electrical interface. Set this variable to 1
if scrambling is being used, or to 0 if it is not.
Subclass
Subclass determines whether the latency of the device is
deterministic, meaning it requires an external synchronization
signal. See the Subclass Setup section for more information.
Lane0Checksum can be used for error checking purposes to
ensure that the transmitter is set up as expected.
If CheckSumMode = 0, the checksum is the lower eight bits of
the sum of the L − 1, M − 1, K − 1, N − 1, NP − 1, S − 1,
Scrambling, HD, Subclass, and JESDVer variables.
If CheckSumMode = 1, Lane0Checksum is the lower eight bits
of the sum of Register 0x450 to Register 0x45A. Select whether
to sum by fields or by registers, matching the setting on the
transmitter.
DAC Power-Down Setup
As described in the Step 1: Start Up the DAC section, PdDACs
must be set to 5 if both converters are being used either in a
single- or dual-link mode. If only one DAC is being used (M = 1
and in single-link mode), PdDACs must be set to 7.
Rev. A | Page 31 of 117
AD9135/AD9136
Data Sheet
SERDES CLOCKS SETUP
Link Delay Setup
This section describes how to select the appropriate Halfrate,
OvSmp, and PLLDiv settings in the Step 4: Physical Layer
section. These parameters depend solely on the lane rate (the
lane rate is established in the JESD204B Setup section).
LMFCVar and LMFCDel are used to impose delays such that all
lanes in a system arrive in the same LMFC cycle.
Table 30. SERDES Lane Rate Configuration Settings
Lane Rate (Gbps)
1.44 to 2.76
2.88 to 5.52
5.75 to 10.64
Halfrate
0
0
1
OvSmp
1
0
0
PLLDiv
2
1
0
The unit used internally for delays is the period of the internal
processing clock (PClock), whose rate is 1/40th the lane rate.
Delays that are not in PClock cycles must be converted before
they are used.
Some useful internal relationships are defined by
PClockPeriod = 40/LaneRate
Halfrate and OvSmp set how the clock detect and recover (CDR)
circuit samples. See the SERDES PLL section for an explanation
of how that circuit blocks works and the role of PLLDiv in the
block.
EQUALIZATION MODE SETUP
Set EqMode = 1 for a low power setting. Select this mode if the
insertion loss in the printed circuit board (PCB) is less than
12 dB. For insertion losses greater than 12 dB but less than
17.5 dB, set EqMode = 0. More details can be found in the
Equalization section.
LINK LATENCY SETUP
This section describes the steps necessary to guarantee
multichip deterministic latency in Subclass 1 and to guarantee
synchronization of links within a device in Subclass 0. Use this
section to fill in LMFCDel, LMFCVar, and Subclass in the Step 5:
Data Link Layer section. For more information, see the Syncing
LMFC Signals section.
PClockPeriod can be used to convert from time to PClock
cycles when needed.
PClockFactor = 4/F (frames per PClock)
PClockFactor is used to convert from units of PClock cycles to
FrameClock cycles, which is needed to set LMFCDel in
Subclass 1.
PClocksPerMF= K/PClockFactor (PClocks per LMFC cycle)
where PClocksPerMF is the number or PClock cycles in a
multiframe cycle.
The values for PClockFactor and PClockPerMF are given per
JESD mode in Table 31.
Table 31. PClockFactor and PClockPerMF
JESD Mode ID
PClockFactor
PClockPerMF (K = 32)
PClockPerMF (K = 16)
1
8
4
8
N/A1
9
4
8
N/A1
10
2
16
8
11
4
8
N/A1
12
4
8
N/A1
13
2
16
8
N/A means not applicable.
Subclass Setup
With Known Delays
The AD9135/AD9136 support JESD204B Subclass 0 and
Subclass 1 operation.
With information about all the system delays, LMFCVar and
LMFCDel can be calculated directly.
Subclass 1
This mode gives deterministic latency and allows links to be
synced to within ½ DAC clock periods. It requires an external
SYSREF± signal that is accurately phase aligned to the DAC clock.
Subclass 0
This mode does not require any signal on the SYSREF± pins,
which can be left disconnected.
Subclass 0 still requires that all lanes arrive within the same
LMFC cycle and that the two DACs must be synchronized to
each other; they are synchronized to an internal clock instead of
to the SYSREF± signal.
RxFixed (the fixed receiver delay in PClock cycles) and RxVar
(the variable receiver delay in PClock cycles) can be found in
Table 8. TxFixed (the fixed transmitter delay in PClock cycles)
and TxVar (the variable receiver delay in PClock cycles) can be
found in the data sheet of the transmitter used. PCBFixed (the
fixed PCB trace delay in PClock cycles) can be extracted from
software; because this is generally much smaller than a PClock
cycle, it can also be omitted. For both the PCB and transmitter
delays, convert the delays into PClock cycles.
For each lane,
Set Subclass to 0 or 1 as desired.
MinDelayLane = floor(RxFixed + TxFixed + PCBFixed)
MaxDelayLane = ceiling(RxFixed + RxVar + TxFixed +
TxVar + PCBFixed))
Rev. A | Page 32 of 117
Data Sheet
AD9135/AD9136
For safety, add a guard band of 1 PClock cycle to each end of
the link delay as in the following equations:
LMFCVar = (MaxDelay + 1) − (MinDelay − 1)
where:
MinDelay is the minimum of all MinDelayLane values across
lanes, links, and devices.
MaxDelay is the maximum of all MaxDelayLane values across
lanes, links, and devices.
Note that if LMFCVar must be more than 10, the AD9135/
AD9136 cannot tolerate the variable delay in the system.
Table 32. Register Configuration and Procedure for OneShot Sync
Addr.
0x301
0x03A
Bit.
No.
Value1
0x
Variable
Subclass
0x01
0x03A
0x81
0x03A
0xC1
SYSREF±
Signal
For Subclass 1
LMFCDel = ((MinDelay − 1) × PClockFactor) % K
For Subclass 0
0x
0x300
LMFCDel = (MinDelay − 1) % PClockPerMF
Program the same LMFCDel and LMFCVar across all links and
devices.
See the Link Delay Setup Example, with Known Delays section
for an example calculation.
6
CheckSumMode
3
DualLink
2
CurrentLink
Set to 0 to access
Link 0 status or 1
for Link 1 status
readbacks. See the
JESD204B Setup
section.
[1:0]
EnLinks
EnLinks = 3 if in
DualLink mode to
enable Link 0 and
Link 1; EnLinks = 1
if not in DualLink
mode to enable
Link 0
Without Known Delays
If comprehensive delay information is not available or known,
the AD9135/AD9136 can read back the link latency between
the local LMFC for each link (LMFCRX) and the last arriving
LMFC boundary in PClock cycles. This information is then
used to calculate LMFCVar and LMFCDel.
For each link (on each device),
1.
2.
3.
4.
5.
Power up the board.
Follow the steps in Table 15 through Table 21 of the Device
Setup Guide.
Set the subclass and perform a sync. For one-shot sync,
perform the writes in Table 32. See the Syncing LMFC
Signals section for alternate sync modes.
Record DYN_LINK_LATENCY_0 (Register 0x302) as a
value of Delay for that link and power cycle.
Record DYN_LINK_LATENCY_1 (Register 0x303) as a
value of Delay for that link and power cycle the system.
Repeat Step 1 to Step 5 twenty times for each device in the
system. Keep a single list of the Delay values across all runs and
devices.
1
Description
Set subclass
Set sync mode to
one-shot sync
Enable the sync
machine
Arm the sync
machine
If Subclass = 1,
ensure that at
least one SYSREF±
edge is sent to the
device
Enable the links
See the JESD204B
Setup section
See the JESD204B
Setup section
0x denotes a register value that the user must fill in. See the Variable and
Description columns for information on selecting the appropriate register value.
The list of Delay values is used to calculate LMFCDel and
LMFCVar, however, first some of the Delay values may need to
be remapped.
The maximum possible value for DYN_LINK_LATENCY_x
is one less than the number of PClocks in a multiframe
(PClocksPerMF). It is possible that a rollover condition may
be encountered; that is, the set of recorded Delay values
may roll over the edge of a multiframe. If so, Delay values
may be near both 0 and PClocksPerMF. If this occurs, add
PClocksPerMF to the set of values near 0.
For example, for Delay value readbacks of 6, 7, 0, and 1, the 0
and 1 Delay values must be remapped to 8 and 9, making the
new set of Delay values 6, 7, 8, and 9.
Rev. A | Page 33 of 117
AD9135/AD9136
Data Sheet
Across power cycles, links, and devices,
CROSSBAR SETUP


Register 0x308 to Register 0x30B allow arbitrary mapping of
physical lanes (SERDINx±) to logical lanes used by the SERDES
deframers.
MinDelay is the minimum of all Delay measurements
MaxDelay is the maximum of all Delay measurements
For safety, a guard band of 1 PClock cycle is added to each end
of the link delay and calculate LMFCVar and LMFCDel with
the following equation:
LMFCVar = (MaxDelay + 1) − (MinDelay − 1)
Note that if LMFCVar must be more than 10, the AD9135/
AD9136 cannot tolerate the variable delay in the system.
For Subclass 1
LMFCDel = ((MinDelay − 1) × PClockFactor)
For Subclass 0
LMFCDel = (MinDelay − 1) % PClockPerMF
Program the same LMFCDel and LMFCVar across all links and
devices.
See the Link Delay Setup Example, Without Known Delay
section for an example calculation.
Table 33. Crossbar Registers
Address
0x308
0x308
0x309
0x309
0x30A
0x30A
0x30B
0x30B
Bits
[2:0]
[5:3]
[2:0]
[5:3]
[2:0]
[5:3]
[2:0]
[5:3]
Logical Lane
LOGICAL_LANE0_SRC
LOGICAL_LANE1_SRC
LOGICAL_LANE2_SRC
LOGICAL_LANE3_SRC
LOGICAL_LANE4_SRC
LOGICAL_LANE5_SRC
LOGICAL_LANE6_SRC
LOGICAL_LANE7_SRC
Write each LOGICAL_LANEy_SRC with the number (x) of the
desired physical lane (SERDINx±) from which to receive data.
By default, all logical lanes use the corresponding physical lane
as their data source. For example, by default LOGICAL_LANE0_
SRC = 0, meaning that Logical Lane 0 receives data from
Physical Lane 0 (SERDIN0±). To use SERDIN4± as the source
for Logical Lane 0, write LOGICAL_LANE0_SRC = 4.
Rev. A | Page 34 of 117
Data Sheet
AD9135/AD9136
JESD204B SERIAL DATA INTERFACE
JESD204B OVERVIEW
The AD9135/AD9136 have eight JESD204B data ports that
receive data. The eight JESD204B ports can be configured as
part of a single JESD204B link or as part of two separate
JESD204B links (dual-link mode) that share a single system
reference (SYSREF±) and device clock (CLK±).
The JESD204B serial interface hardware consists of three layers:
the physical layer, the data link layer, and the transport layer.
These sections of the hardware are described in subsequent
sections, including information for configuring every aspect of
the interface. Figure 43 shows the communication layers
implemented in the AD9135/AD9136 serial data interface to
recover the clock and deserialize, descramble, and deframe the
data before it is sent to the digital signal processing section of
the device.
The physical layer establishes a reliable channel between the
transmitter and the receiver, the data link layer unpacks the
data into octets and descrambles the data, and the transport
layer receives the descrambled JESD204B frames and converts
them to DAC samples.
A number of JESD204B parameters (L, F, K, M, N, NP, S, HD,
and Scrambling) defines how the data is packed and instruct the
device how to turn the serial data into samples. These parameters
are defined in detail in the Transport Layer section.
Only certain combinations of parameters are supported. Each
supported combination is called a mode. In total, six modes are
supported by the AD9135/AD9136. There are three supported
single-link modes, as described in Table 35, and three modes that
can operate in either single- or dual-link mode, as described in
Table 34. These tables show the associated clock rates when the
lane rate is 10 Gbps.
For a particular application, the number of converters to use (M)
and DataRate are known. Calculate LaneRate and number of
lanes (L) as follows:
DataRate = (DACRate)/(InterpolationFactor)
LaneRate = (20 × DataRate × M)/L
where LaneRate must be between 1.44 Gbps and 10.64 Gbps.
Achieving and recovering synchronization of the lanes is very
important. To simplify the interface to the transmitter, the
AD9135/AD9136 designate a master synchronization signal for
each JESD204B link. In single-link mode, SYNCOUT0± is used as
the master signal for all lanes; in dual-link mode, SYNCOUT0± is
used as the master signal for Link 0, and SYNCOUT1± is used as
the master signal for Link 1. If any lane in a link loses
synchronization, a resynchronization request is sent to the
transmitter via the synchronization signal of the link. The
transmitter stops sending data and instead sends synchronization
characters to all lanes in that link until resynchronization is
achieved.
SYNCOUT0±
SYNCOUT1±
PHYSICAL
LAYER
SERDIN0±
DATA LINK
LAYER
DESERIALIZER
I DATA[15:0] OR [11:0]
QBD/
DESCRAMBLER
FRAME TO
SAMPLES
TO
DAC
Q DATA[15:0] OR [11:0]
DESERIALIZER
12578-004
SERDIN7±
TRANSPORT
LAYER
SYSREF±
Figure 43. Functional Block Diagram of Serial Link Receiver
Rev. A | Page 35 of 117
AD9135/AD9136
Data Sheet
Table 34. Single-Link and Dual-Link JESD204B Operating Modes
Parameter
M (Converter Counts)
L (Lane Counts)
S ((Samples per Converter) per Frame)
F ((Octets per Frame) per Lane)
Example Clocks for 10 Gbps Lane Rate
PClock Rate (MHz)
Frame Rate (MHz)
Data Rate (MHz)
8
1
4
2
1
250
1000
2000
Mode
9
1
2
1
1
10
1
1
1
2
250
1000
1000
250
500
500
The AD9135/AD9136 autocalibrate the input termination to
50 Ω. Before running the termination calibration, write to
Register 0x2AA, Register 0x2AB, Register 0x2B1, and Register
0x2B2 as described in Table 36 to guarantee proper calibration.
The termination calibration begins when Register 0x2A7[0] and
Register 0x2AE[0] transition from low to high. Register 0x2A7
controls autocalibration for PHY 0, PHY 1, PHY 6, and PHY 7.
Register 0x2AE controls autocalibration for PHY 2, PHY 3,
PHY 4, and PHY 5.
The PHY termination autocalibration routine is shown in Table 36.
Table 36. PHY Termination Autocalibration Routine
8
4
4
2
13
2
1
2
1
2
1
1
1
1
2
250
1000
2000
250
1000
1000
250
500
500
Note that for Mode 11 to Mode 13, the M and L parameters programmed on
the receive side do not match the parameters on the transmit side. The
parameters on the transmit side reflect the true number of converters and
lanes per link.
PHYSICAL LAYER
The physical layer of the JESD204B interface, hereafter referred
to as the deserializer, has eight identical channels. Each channel
consists of the terminators, an equalizer, a clock and data recovery
(CDR) circuit, and the 1:40 demux function (see Figure 45).
Value
0xB7
0x87
0xB7
0x87
0x01
0x01
Description
SERDES interface termination configuration
SERDES interface termination configuration
SERDES interface termination configuration
SERDES interface termination configuration
Autotune PHY terminations
Autotune PHY terminations
The input termination voltage of the DAC is sourced externally
via the VTT pins (Pin 21, Pin 25, Pin 42, and Pin 46). Set VTT by
connecting it to SVDD12. It is recommended that the JESD204B
inputs be ac-coupled to the JESD204B transmit device using
100 nF capacitors.
Receiver Eye Mask
The AD9135/AD9136 comply with the JESD204B specification
regarding the receiver eye mask and are capable of capturing
data that complies with this mask. Figure 44 shows the receiver
eye mask normalized to the data rate interval with a VTT swing
of 600 mV. See the JESD204B specification for more information
regarding the eye mask and permitted receiver eye opening.
LV-OIF-11G-SR RECEIVER EYE MASK
(3.125Mbps ≥ UI ≤ 12.5Gbps)
525
JESD204B data is input to the AD9135/AD9136 via the SERDINx±
1.2 V differential input pins as per the JESD204B specification.
Interface Power-Up and Input Termination
Before using the JESD204B interface, it must be powered up by
setting Register 0x200[0] = 0. In addition, each physical lane that is
not being used (SERDINx±) must be powered down. To do so,
set the corresponding Bit x for Physical Lane x in Register 0x201 to
0 if the physical lane is being used, and to 1 if it is not being used.
55
0
–55
–525
0
0.35
0.5
0.65
TIME (UI)
Figure 44. Receiver Eye Mask
DESERIALIZER
SERDINx±
TERMINATION
EQUALIZER
CDR
1:40
SPI CONTROL
FROM PLL
Figure 45. Deserializer Block Diagram
Rev. A | Page 36 of 117
12578-006
1
11
2
1
Address
0x2AA
0x2AB
0x2B1
0x2B2
0x2A7
0x2AE
1.00
12578-007
Parameter
M (Converter Count) (Tx setting)
AD9135 and AD9136 M Setting1
(Rx Setting)
L (Lane Count) (Tx setting)
AD9135 and AD9136 L Setting1
(Rx Setting)
S ((Samples per Converter) per Frame)
F ((Octets per Frame) per Lane)
Example Clocks for 10 Gbps Lane Rate
PClock Rate (MHz)
Frame Rate (MHz)
Data Rate (MHz)
Mode
12
2
1
AMPLITUDE (mV)
Table 35. Single-Link JESD204B Operating Modes
Data Sheet
AD9135/AD9136
Clock Relationships
Register 0x280 controls the synthesizer enable and recalibration.
The following clocks rates are used throughout the rest of the
JESD204B section. The relationship between any of the clocks
can be derived from the following equations:
To enable the SERDES PLL, first set the PLL divider register
according to Table 37, and then enable the SERDES PLL by
writing 1 to Register 0x280[0].
DataRate = (DACRate)/(InterpolationFactor)
Confirm that the SERDES PLL is working by reading
Register 0x281. If Register 0x281[0] = 1, the SERDES PLL has
locked. If Register 0x281[3] = 1, the SERDES PLL was successfully
calibrated. If Register 0x281[4] or Register 0x281[5] are high, the
PLL has reached the upper or lower end of its calibration band and
must be recalibrated by writing 0 and then 1 to Register 0x280[2].
LaneRate = (20 × DataRate × M)/L
ByteRate = LaneRate/10
where:
M is the JESD204B parameter for converters per link.
L is the JESD204B parameter for lanes per link.
SERDES PLL Fixed Register Writes
This relationship comes from 8-bit/10-bit encoding, where each
byte is represented by 10 bits.
To optimize the SERDES PLL across all operating conditions,
the register writes in Table 38 are recommended.
PClockRate = ByteRate/4
Table 38. SERDES PLL Fixed Register Writes
The processing clock is used for a quad-byte decoder.
FrameRate = ByteRate/F
where F is defined as bytes per frame per lane.
PClockFactor = FrameRate/PClockRate = 4/F
where F is the JESD204B parameter for octets per frame per lane.
SERDES PLL
Functional Overview of the SERDES PLL
The independent SERDES PLL uses integer-N techniques to
achieve clock synthesis. The entire SERDES PLL is integrated
on-chip, including the VCO and the loop filter. The SERDES
PLL VCO operates over the range of 5.65 GHz to 11.04 GHz.
In the SERDES PLL, a VCO divider block divides the VCO clock
by 2 to generate a 2.825 GHz to 5.52 GHz quadrature clock for the
deserializer cores. This clock is the input to the clock and data
recovery block that is described in the Clock and Data Recovery
section.
The reference clock to the SERDES PLL is always running at a
frequency, fREF, that is equal to 1/40 of the lane rate (PClockRate).
This clock is divided by the DivFactor value to deliver a clock to
the PFD block that is between 35 MHz and 80 MHz. Table 37
includes the respective SERDES_PLL_DIV_MODE register
settings for each of the desired DivFactor options available.
Address
0x284
0x285
0x286
0x287
0x28A
0x28B
0x290
0x294
0x296
0x297
0x299
0x29A
0x29C
0x29F
0x2A0
Divide by
(DivFactor)
1
2
4
Description
Optimal SERDES PLL loop filter
Optimal SERDES PLL loop filter
Optimal SERDES PLL loop filter
Optimal SERDES PLL charge pump
Optimal SERDES PLL VCO LDO
Optimal SERDES PLL configuration
Optimal SERDES PLL VCO varactor
Optimal SERDES PLL charge pump
Optimal SERDES PLL VCO
Optimal SERDES PLL VCO
Optimal SERDES PLL configuration
Optimal SERDES PLL VCO varactor
Optimal SERDES PLL charge pump
Optimal SERDES PLL VCO varactor
Optimal SERDES PLL VCO varactor
SERDES PLL IRQ
SERDES PLL lock and lost signals are available as IRQ events.
Use Register 0x01F[3:2] to enable these signals, and then use
Register 0x023[3:2] to read back their statuses and reset the IRQ
signals. See the Interrupt Request Operation section for more
information.
Table 37. SERDES PLL Divider Settings
LaneRate (Gbps)
1.44 to 2.76
2.88 to 5.52
5.75 to 10.64
Value
0x62
0xC9
0x0E
0x12
0x7B
0x00
0x89
0x24
0x03
0x0D
0x02
0x8E
0x2A
0x78
0x06
SERDES_PLL_DIV_MODE,
Register 0x289[1:0]
2
1
0
Rev. A | Page 37 of 117
AD9135/AD9136
Data Sheet
2.825GHz TO 5.52GHz
OUTPUT
VCO
LDO
CHARGE
PUMP
PFD
80MHz
REF DIVIDER MAX
N = 1, 2, 4
fREF
BIT RATE ÷ 40
DIVIDER
(1, 2, 4)
C1
R1
UP
C2
C3
I Q
LC VCO
5.65GHz TO 11.04GHz
÷2
DOWN
÷80
R3
ALC CAL
CAL CONTROL BITS
12578-011
FO CAL
3.2mA
Figure 46. SERDES PLL Synthesizer Block Diagram Including VCO Divider Block
Clock and Data Recovery
Equalization
The deserializer is equipped with a CDR circuit. Instead of
recovering the clock from the JESD204B serial lanes, the CDR
recovers the clocks from the SERDES PLL. The 2.825 GHz to
5.52 GHz output from the SERDES PLL, shown in Figure 46, is
the input to the CDR.
To compensate for signal integrity distortions for each PHY
channel due to PCB trace length and impedance, the AD9135/
AD9136 employ an easy to use, low power equalizer on each
JESD204B channel. The AD9135/AD9136 equalizers can
compensate for insertion losses far greater than required by the
JESD204B specification. The equalizers have two modes of
operation that are determined by the EQ_POWER_MODE
register setting in Register 0x268[7:6]. In low power mode
(Register 0x268[7:6] = 2b’01) and operating at the maximum
lane rate of 10 Gbps, the equalizer can compensate for up to 12 dB
of insertion loss. In normal mode (Register 0x268[7:6] = 2b’00),
the equalizer can compensate for up to 17.5 dB of insertion loss.
This performance is shown in Figure 47 as an overlay to the
JESD204B specification for insertion loss. Figure 47 shows the
equalization performance at 10.0 Gbps, near the maximum
baud rate for the AD9135/AD9136.
A CDR sampling mode must be selected to generate the lane
rate clock inside the device. If the desired lane rate is greater
than 5.65 GHz, half rate CDR operation must be used. If the
desired lane rate is less than 5.65 GHz, disable half rate operation.
If the lane rate is less than 2.825 GHz, disable half rate operation
and enable 2× oversampling to recover the appropriate lane rate
clock. Table 39 gives a breakdown of CDR sampling settings that
must be set dependent on the LaneRate.
Table 39. CDR Operating Modes
LaneRate (Gbps)
1.44 to 2.76
2.88 to 5.52
5.75 to 10.64
ENHALFRATE,
Register 0x230[5]
0
0
1
CDR_OVERSAMP,
Register 0x230[1]
1
0
0
The CDR circuit synchronizes the phase used to sample the data on
each serial lane independently. This independent phase adjustment
per serial interface ensures accurate data sampling and eases the
implementation of multiple serial interfaces on a PCB.
After configuring the CDR circuit, reset it and then release the
reset by writing 1 and then 0 to Register 0x206[0].
Power-Down Unused PHYs
Note that any unused and enabled lanes consume extra power
unnecessarily. Each lane that is not being used (SERDINx±)
must be powered off by writing a 1 to the corresponding bit of
PHY_PD (Register 0x201).
Figure 48 and Figure 49 are provided as points of reference for
hardware designers and show the insertion loss for various
lengths of well laid out stripline and microstrip transmission
lines. See the Hardware Considerations section for specific layout
recommendations for the JESD204B channel.
Low power mode is recommended if the insertion loss of the
JESD204B PCB channels is less than that of the most lossy
supported channel for low power mode (shown in Figure 47).
If the insertion loss is greater than that, but still less than that of
the most lossy supported channel for normal mode (shown in
Figure 47), use normal mode. At 10 Gbps operation, the equalizer
in normal mode consumes about 4 mW more power per lane
used than in low power equalizer mode. Note that either mode
can be used in conjunction with transmitter preemphasis to
ensure functionality and/or to optimize for power.
Rev. A | Page 38 of 117
Data Sheet
AD9135/AD9136
0
JESD204B SPEC ALLOWED
CHANNEL LOSS
2
INSERTION LOSS (dB)
4
AD9135/AD9136
ALLOWED
CHANNEL LOSS
(LOW POWER MODE)
6
8
EXAMPLE OF
AD9135/AD9136
COMPATIBLE
CHANNEL (LOW
POWER MODE)
10
12
EXAMPLE OF
AD9135/AD9136
COMPATIBLE
CHANNEL
(NORMAL MODE)
AD9135/AD9136
ALLOWED
CHANNEL LOSS
(NORMAL MODE)
14
16
DATA LINK LAYER
EXAMPLE OF
JESD204B
COMPLIANT
CHANNE L
18
20
22
24
5.0
7.5
12578-339
2.5
FREQUENCY (GHz)
Figure 47. Insertion Loss Allowed
0
The AD9135/AD9136 can operate as a single-link or dual-link
high speed JESD204B serial data interface. When operating in
dual-link mode, configure both links with the same JESD204B
parameters because they share a common device clock and system
reference. All eight lanes of the JESD204B interface handle link
layer communications such as code group synchronization,
frame alignment, and frame synchronization.
–5
ATTENUATION (dB)
–10
–15
–20
STRIPLINE = 6”
STRIPLINE = 10”
STRIPLINE = 15”
STRIPLINE = 20”
STRIPLINE = 25”
STRIPLINE = 30”
–25
–30
–40
1
2
3
4
5
6
7
8
9
10
FREQUENCY (GHz)
12578-009
–35
0
Figure 48. Insertion Loss of 50 Ω Striplines on FR4
0
–5
–15
–20
6” MICROSTRIP
10” MICROSTRIP
15” MICROSTRIP
20” MICROSTRIP
25” MICROSTRIP
30” MICROSTRIP
–30
–35
–40
0
1
2
3
4
5
6
7
8
9
FREQUENCY (GHz)
10
The AD9135/AD9136 decode 8-bit/10-bit control characters,
allowing marking of the start and end of the frame and
alignment between serial lanes. Each AD9135/AD9136 serial
interface link can issue a synchronization request by setting its
SYNCOUT0±/ SYNCOUT1± signal low. The synchronization
protocol follows Section 4.9 of the JESD204B standard. When a
stream of four consecutive /K/ symbols is received, the
AD9135/AD9136 deactivate the synchronization request by
setting the SYNCOUT0±/ SYNCOUT1± signal high at the next
internal LMFC rising edge. Then, the AD9135/AD9136 wait for
the transmitter to issue an ILAS. During the ILAS sequence, all
lanes are aligned using the /A/ to /R/ character transition as
described in the JESD204B Serial Link Establishment section.
Elastic buffers hold early arriving lane data until the alignment
character of the latest lane arrives. At this point, the buffers for
all lanes are released and all lanes are aligned (see Figure 52).
12578-010
ATTENUATION (dB)
–10
–25
The data link layer of the AD9135/AD9136 JESD204B interface
accepts the deserialized data from the PHYs and deframes and
descrambles them so that data octets are presented to the
transport layer to be put into DAC samples. Figure 50 shows
the link mode block diagrams for single-link and dual-link
configurations and the interaction between the physical layer
and logical layer. The DACs can only be configured in
sequential order; for example, in Mode 10, when in single-link
mode, the AD9135/AD9136 only uses Logical Lane 0 and
DAC0. Logical lanes must be set according to Table 29 for the
desired mode. See the Mode Configuration Maps section for
further details on each of the mode configurations supported.
The architecture of the data link layer is shown in Figure 51.
The data link layer consists of a synchronization FIFO for each
lane, a crossbar switch, a deframer, and descrambler.
Figure 49. Insertion Loss of 50 Ω Microstrips on FR4
Rev. A | Page 39 of 117
AD9135/AD9136
Data Sheet
LOGICAL LANE 0/LINK 0 LANE 0
SERDIN1±/PHYSICAL LANE 1
LOGICAL LANE 1/LINK 0 LANE 1
SERDIN2±/PHYSICAL LANE 2
SERDIN3±/PHYSICAL LANE 3
SERDIN4±/PHYSICAL LANE 4
SERDIN5±/PHYSICAL LANE 5
LOGICAL LANE 2/LINK 0 LANE 2
CROSSBAR
REGISTER
0x308
TO
REGISTER
0x30B
DAC0
LOGICAL LANE 3/LINK 0 LANE 3
LOGICAL LANE 4/LINK 0 LANE 4
LOGICAL LANE 6/LINK 0 LANE 6
SERDIN7±/PHYSICAL LANE 7
LOGICAL LANE 7/LINK 0 LANE 7
LOGICAL LANE 0/LINK 0 LANE 0
SERDIN1±/PHYSICAL LANE 1
LOGICAL LANE 1/LINK 0 LANE 1
SERDIN2±/PHYSICAL LANE 2
LOGICAL LANE 2/LINK 0 LANE 2
SERDIN5±/PHYSICAL LANE 5
DAC1
LOGICAL LANES
SERDIN0±/PHYSICAL LANE 0
SERDIN4±/PHYSICAL LANE 4
QUAD-BYTE
DEFRAMER
(QBD0)
LOGICAL LANE 5/LINK 0 LANE 5
SERDIN6±/PHYSICAL LANE 6
SERDIN3±/PHYSICAL LANE 3
DAC CORE
QBD
SERDIN0±/PHYSICAL LANE 0
PHYSICAL LANES
CROSSBAR
REGISTER
0x308
TO
REGISTER
0x30B
DAC0
QUAD-BYTE
DEFRAMER 1
(QBD1)
DAC1
LOGICAL LANE 4/LINK 1 LANE 0
LOGICAL LANE 5/LINK 1 LANE 1
SERDIN6±/PHYSICAL LANE 6
LOGICAL LANE 6/LINK 1 LANE 2
SERDIN7±/PHYSICAL LANE 7
LOGICAL LANE 7/LINK 1 LANE 3
LOGICAL LANES
12578-550
PHYSICAL LANES
QUAD-BYTE
DEFRAMER 0
(QBD0)
LOGICAL LANE 3/LINK 0 LANE 3
Figure 50. Link Mode Functional Diagram
DATA LINK LAYER
SYNCOUTx±
QUAD-BYTE
DEFRAMER
QBD
DES_DATA0
SERDIN0_CLK
SERDIN0
FIFO
CROSS
BAR
SWITCH
SYSREF
SERDIN7
FIFO
LANE 0 OCTETS
LANE 7 OCTETS
SYSTEM CLOCK
PHASE DETECT
12578-012
DES_DATA7
SERDIN7_CLK
DESCRAMBLE
DUAL-LINK MODE
LOGICAL
LAYER
CROSSBAR
10-BIT/8-BIT DECODE
SINGLE-LINK MODE
PHYSICAL
LAYER (PHY)
PCLK
SPI CONTROL
Figure 51. Data Link Layer Block Diagram
Rev. A | Page 40 of 117
Data Sheet
AD9135/AD9136
L RECEIVE LANES
(EARLIEST ARRIVAL) K K K R D D
D D A R Q C
L RECEIVE LANES
(LATEST ARRIVAL) K K K K K K K R D D
C
D D A R Q C
D D A R D D
C
D D A R D D
0 CHARACTER ELASTIC BUFFER DELAY OF LATEST ARRIVAL
4 CHARACTER ELASTIC BUFFER DELAY OF EARLIEST ARRIVAL
L ALIGNED
RECEIVE LANES K K K K K K K R D D
D D A R Q C
C
12578-013
K = K28.5 CODE GROUP SYNCHRONIZATION COMMA CHARACTER
A = K28.3 LANE ALIGNMENT SYMBOL
F = K28.7 FRAME ALIGNMENT SYMBOL
R = K28.0 START OF MULTIFRAME
Q = K28.4 START OF LINK CONFIGURATION DATA
C = JESD204B LINK CONFIGURATION PARAMETERS
D = Dx.y DATA SYMBOL
D D A R D D
Figure 52. Lane Alignment During ILAS
JESD204B Serial Link Establishment
A brief summary of the high speed serial link establishment
process for Subclass 1 is provided. See Section 5.3.3 of the
JESD204B specifications document for complete details.
After the last /A/ character of the last ILAS, multiframe data
begins streaming. The receiver adjusts the position of the /A/
character such that it aligns with the internal LMFC of the
receiver at this point.
Step 1: Code Group Synchronization
Step 3: Data Streaming
Each receiver must locate K (K28.5) characters in its input data
stream. After four consecutive K characters are detected on all
link lanes, the receiver block deasserts the SYNCOUTx± signal
to the transmitter block at the receiver LMFC edge.
In this phase, data is streamed from the transmitter block to the
receiver block.
The transmitter captures the change in the SYNCOUTx± signal,
and at a future transmitter LMFC rising edge, starts the initial
lane alignment sequence (ILAS).
The receiver block processes and monitors the data it receives
for errors, including
Step 2: Initial Lane Alignment Sequence
The main purposes of this phase are to align all the lanes of the
link and to verify the parameters of the link.
Before the link is established, write each of the link parameters
to the receiver device to designate how data is sent to the
receiver block.
The ILAS consists of four or more multiframes. The last character
of each multiframe is a multiframe alignment character, /A/.
The first, third, and fourth multiframes are populated with
predetermined data values. Note that Section 8.2 of the
JESD204B specifications document describes the data ramp that
is expected during ILAS. By default, the AD9135/AD9136 do
not require this ramp. Register 0x47E[0] can be set high to
require the data ramp. The deframer uses the final /A/ of each
lane to align the ends of the multiframes within the receiver.
The second multiframe contains an R (K28.0), Q (K28.4), and
then data corresponding to the link parameters. Additional
multiframes can be added to the ILAS if needed by the receiver.
By default, the AD9135/AD9136 use four multiframes in the ILAS
(this can be changed in Register 0x478). If using Subclass 1,
exactly four multiframes must be used.
Optionally, data can be scrambled. Scrambling does not start
until the very first octet following the ILAS.





Bad running disparity (8-bit/10-bit error)
Not in table (8-bit/10-bit error)
Unexpected control character
Bad ILAS
Interlane skew error (through character replacement)
If any of these errors exist, they are reported back to the
transmitter in one of a few ways (see the JESD204B Error
Monitoring section for details).



SYNCOUTx± signal assertion: resynchronization
(SYNCOUTx± signal pulled low) is requested at each error
for the last two errors. For the first three errors, an optional
resynchronization request can be asserted when the error
counter reaches a set error threshold.
For the first three errors, each multiframe with an error in
it causes a small pulse on SYNCOUTx±.
Errors can optionally trigger an IRQ event, which can be
sent to the transmitter.
Various test modes for verifying the link integrity can be found
in the JESD204B Test Modes section.
Rev. A | Page 41 of 117
AD9135/AD9136
Data Sheet
Lane FIFO
The FIFOs in front of the crossbar switch and deframer
synchronize the samples sent on the high speed serial data
interface with the deframer clock by adjusting the phase of the
incoming data. The FIFO absorbs timing variations between the
data source and the deframer; this allows up to two PClock
cycles of drift from the transmitter. The FIFO_STATUS_REG_0
register and FIFO_STATUS_REG_1 register (Register 0x30C
and Register 0x30D, respectively) can be monitored to identify
whether the FIFOs are full or empty.
Lane FIFO IRQ
An aggregate lane FIFO error bit is also available as an IRQ
event. Use Register 0x01F[1] to enable the FIFO error bit, and
then use Register 0x023[1] to read back its status and reset the
IRQ signal. See the Interrupt Request Operation section for
more information.
Crossbar Switch
Register 0x308 to Register 0x30B allow arbitrary mapping of
physical lanes (SERDINx±) to logical lanes used by the SERDES
deframers.
Table 40. Crossbar Registers
Address
0x308
0x308
0x309
0x309
0x30A
0x30A
0x30B
0x30B
Bits
[2:0]
[5:3]
[2:0]
[5:3]
[2:0]
[5:3]
[2:0]
[5:3]
Logical Lane
LOGICAL_LANE0_SRC
LOGICAL_LANE1_SRC
LOGICAL_LANE2_SRC
LOGICAL_LANE3_SRC
LOGICAL_LANE4_SRC
LOGICAL_LANE5_SRC
LOGICAL_LANE6_SRC
LOGICAL_LANE7_SRC
In single-link mode, Deframer 0 is used exclusively and Deframer 1
remains inactive. In dual-link mode, both QBDs are active and
must be configured separately using the LINK_PAGE bit
(Register 0x300[2]) to select which link to configure. The
LINK_MODE bit (Register 0x300[3]) is 1 for dual-link, or 0 for
single-link.
Each deframer uses the JESD204B parameters that the user has
programmed into the register map to identify how the data has
been packed and how to unpack it. The JESD204B parameters
are described in detail in the Transport Layer section; many of
the parameters are also needed in the transport layer to convert
JESD204B frames into samples.
Descrambler
The AD9135/AD9136 provide an optional descrambler block
using a self synchronous descrambler with a polynomial: 1 +
x14 + x15.
Enabling data scrambling reduces spectral peaks that are
produced when the same data octets repeat from frame to
frame. It also makes the spectrum data independent so that
possible frequency selective effects on the electrical interface do
not cause data dependent errors. Descrambling of the data is
enabled by setting the SCR bit (Register 0x453[7]) to 1.
Syncing LMFC Signals
The first step in guaranteeing synchronization across links and
devices begins with syncing the LMFC signals. Each DAC has
its own LMFC signal. In Subclass 0, the LMFC signals for each
of the two DACs are synchronized to an internal processing
clock. In Subclass 1, all LMFC signals (for all DACs and devices)
are synchronized to an external SYSREF signal. All LMFC sync
registers are paged as described in the DAC Paging section.
SYSREF Signal
Write each LOGICAL_LANEy_SRC with the number (x) of the
desired physical lane (SERDINx±) from which to receive data.
By default, all logical lanes use the corresponding physical lane
as their data source. For example, by default LOGICAL_LANE0_
SRC = 0; thus, Logical Lane 0 receives data from Physical Lane 0
(SERDIN0±). If instead the user wants to use SERDIN4± as the
source for Logical Lane 0, the user must write LOGICAL_LANE0_
SRC = 4.
Lane Inversion
Register 0x334 allows inversion of desired logical lanes, which
can be used to ease routing of the SERDINx± signals. For each
Logical Lane x, set Bit x of Register 0x334 to 1 to invert it.
Deframers
The AD9135/AD9136 consist of two quad-byte deframers (QBDs).
Each deframer receives the 8-bit/10-bit encoded data from the
deserializer (via the crossbar switch), decodes it, and descrambles it
into JESD204B frames before passing it to the transport layer to be
converted to DAC samples. The deframer processes four symbols
(or octets) per processing clock (PClock) cycle.
The SYSREF signal is a differential source synchronous input that
synchronizes the LMFC signals in both the transmitter and receiver
in a JESD204B Subclass 1 system to achieve deterministic latency.
The SYSREF signal is an active high signal that is sampled by
the device clock rising edge. It is best practice that the device clock
and SYSREF signals be generated by the same source, such as
the AD9516-1 clock generator, so that the phase alignment
between the signals is fixed. When designing for optimum
deterministic latency operation, consider the timing
distribution skew of the SYSREF signal in a multipoint link
system (multichip).
The AD9135/AD9136 support a single pulse or step, or a periodic
SYSREF± signal. The periodicity can be continuous, strobed, or
gapped periodic. The SYSREF± signal can always be dc-coupled
(with a common-mode voltage of 0 V to 2 V). When dc-coupled, a
small amount of common-mode current (<500 μA) is drawn from
the SYSREF± pins. See Figure 53 for the SYSREF± internal circuit.
Rev. A | Page 42 of 117
Data Sheet
AD9135/AD9136
To avoid this common-mode current draw, a 50% duty-cycle
periodic SYSREF± signal can be used with ac coupling capacitors.
If ac-coupled, the ac coupling capacitors combine with the
resistors shown in Figure 53 to make a high-pass filter with a
RC time constant, τ = RC. Select C such that τ > 4/SYSREF
frequency. In addition, the edge rate must be sufficiently fast—
at least 1.3 V/ns is recommended per Table 5—to meet the
SYSREF± vs. DAC clock keepout window (KOW) requirements.
Continuous mode differs from one-shot mode in two ways.
First, no SPI cycle is required to arm the device; the alignment
edge seen after continuous mode is enabled results in a phase
check. Second, a phase check (and when necessary, clock rotation)
occurs on every alignment edge in continuous mode. The one
caveat to the previous statement is that when a phase rotation cycle
is underway, subsequent alignment edges are ignored until the
logic lane is ready again.
It is possible to use ac-coupled mode without meeting the
frequency to time-constant constraint by using SYSREF±
hysteresis (Register 0x081 and Register 0x082). However, this
increases the DAC clock KOW (Table 5 does not apply) by an
amount depending on SYSREF± frequency, level of hysteresis,
capacitor choice, and edge rate.
The maximum acceptable phase error (in DAC clock cycles)
between the alignment edge and the LMFC edge is set in the
error window tolerance register. If continuous sync mode is
used with a nonzero error window tolerance, a phase check
occurs on every SYSREF± pulse, but an alignment occurs only if
the phase error is greater than the specified error window
tolerance. If the jitter of the SYSREF signal violates the KOW
specification given in Table 5 and therefore causes phase error
uncertainty, the error tolerance can be increased to avoid
constant clock rotations. Note that this means the latency is less
deterministic by the size of the window.
1.2V
2kΩ
SYSREF–
2kΩ
~800mV
12578-015
SYSREF+
Figure 53. SYSREF± Input Circuit
Sync Processing Modes Overview
The AD9135/AD9136 support various LMFC sync processing
modes. These modes are one-shot, continuous, windowed
continuous, and monitor modes. All sync processing modes
perform a phase check to see that the LMFC is phase aligned to an
alignment edge. In Subclass 1, the SYSREF pulse acts as the
alignment edge; in Subclass 0, an internal processing clock acts as
the alignment edge. If the signals are not in phase, a clock rotation
occurs to align the signals. The sync modes are described in the
following sections. See the Sync Procedure section for details on
the procedure for syncing the LMFC signals.
One-Shot Sync Mode (SYNCMODE = 0x1)
In one-shot sync mode, a phase check occurs on only the first
alignment edge that is received after the sync machine is armed. If
the phase error is larger than a specified window error tolerance, a
phase adjustment occurs. Though an LMFC synchronization
occurs only once, the SYSREF signal can still be continuous.
Continuous Sync Mode (SYNCMODE = 0x2)
Continuous mode can only be used in Subclass 1 with a periodic
SYSREF signal. In continuous mode, a phase check/alignment
occurs on every alignment edge.
For debug purposes, SYNCARM (Register 0x03A[6]) can be
used to inform the user that alignment edges are being received
in continuous mode. Because the SYNCARM bit is self cleared
after an alignment edge is received, the user can arm the sync
(SYNCARM (Register 0x03A[6]) = 1), and then read back
SYNCARM. If SYNCARM = 0, the alignment edges are being
received and phase checks are occurring. Arming the sync
machine in this mode does not affect the operation of the device.
One-Shot Then Monitor Sync Mode (SYNCMODE = 0x9)
In one-shot then monitor mode, the user can monitor the phase
error in real time. Use this sync mode with a periodic SYSREF
signal. A phase check and alignment occurs on the first alignment
edge received after the sync machine is armed. On all subsequent
alignment edges, the phase is monitored and reported, but no clock
phase adjustment occurs.
The phase error can be monitored on the SYNC_CURRERR_L
register (Register 0x03C[3:0]). Immediately after an alignment
occurs, CURRERROR = 0 indicates that there is no difference
between the alignment edge and the LMFC edge. On every
subsequent alignment edge, the phase is checked. If the
alignment is lost, the phase error is reported in the SYNC_
CURRERR_L register in DAC clock cycles. If the phase error is
beyond the selected window tolerance (Register 0x034[2:0]), one
bit of Register 0x03D[7:6] is set high depending on whether the
phase error is on the low or high side.
When an alignment occurs, snapshots of the last phase error
(Register 0x03C[3:0]) and the corresponding error flags
(Register 0x03D[7:6]) are placed into readable registers for
reference (Register 0x038 and Register 0x039, respectively).
Rev. A | Page 43 of 117
AD9135/AD9136
Data Sheet
Sync Procedure
LMFC Sync IRQ
The procedure for enabling the sync is as follows:
The sync status bits (SYNC_LOCK, SYNC_ROTATE,
SYNC_TRIP, and SYNC_WLIM) are available as IRQ events.
1.
2.
3.
4.
5.
6.
7.
Set Register 0x008 to 0x03 to sync the LMFC for both
DAC0 and DAC1.
Set the desired sync processing mode. The sync processing
mode settings are listed in Table 41.
For Subclass 1, set the error window according to the
uncertainty of the SYSREF signal relative to the DAC clock
and the tolerance of the application for deterministic
latency uncertainty. Sync window tolerance settings are
given in Table 42.
Enable sync by writing 1 to SYNCENABLE
(Register 0x03A[7]).
If in one-shot mode, arm the sync machine by writing 1 to
SYNCARM (Register 0x03A[6]).
If in Subclass 1, ensure that at least one SYSREF± pulse is
sent to the device.
Check the status by reading the following bit fields:
a) SYNC_BUSY (Register 0x03B[7]) = 0 to indicate that
the sync logic is no longer busy.
b) SYNC_LOCK (Register 0x03B[3]) = 1 to indicate that
the signals are aligned. This bit updates on every
phase check.
c) SYNC_WLIM (Register 0x03B[1]) = 0 to indicate that
the phase error is not beyond the specified error
window. This bit updates on every phase check.
d) SYNC_ROTATE (Register 0x03B[2]) = 1. If the phases
were not aligned before the sync and an alignment
occurred, this bit indicates that a clock alignment
occurred. This bit is sticky and can be cleared only by
writing to the SYNCCLRSTKY control bit
(Register 0x03A[5]).
e) SYNC_TRIP (Register 0x03B[0]) = 1 to indicate that
the alignment edge was received and a phase check
occurred. This bit is sticky and can be cleared only by
writing to the SYNCCLRSTKY control bit
(Register 0x03A[5]).
Table 41. Sync Processing Modes
Sync Processing Mode
One-shot
Continuous
One-shot then monitor
SYNCMODE (Register 0x03A[3:0])
0x01
0x02
0x09
Use Register 0x021[3:0] to enable the sync status bits for DAC0
and then use Register 0x025[3:0] to read back their statuses and
to reset the IRQ signals.
Use Register 0x022[3:0] to enable the sync status bits for DAC1
and then use Register 0x026[3:0] to read back their statuses and
to reset the IRQ signals.
See the Interrupt Request Operation section for more information.
Deterministic Latency
JESD204B systems contain various clock domains distributed
throughout each system. Data traversing from one clock
domain to a different clock domain can lead to ambiguous
delays in the JESD204B link. These ambiguities lead to
nonrepeatable latencies across the link from power cycle to
power cycle with each new link establishment. Section 6 of the
JESD204B specification addresses the issue of deterministic
latency with mechanisms defined as Subclass 1 and Subclass 2.
The AD9135/AD9136 support JESD204B Subclass 0 and
Subclass 1 operation, but not Subclass 2. Write the subclass to
Register 0x301[2:0] and once per link to Register 0x458[7:5].
Subclass 0
This mode does not require any signal on the SYSREF± pins,
which can be left disconnected.
Subclass 0 still requires that all lanes arrive within the same LMFC
cycle and that the two DACs be synchronized to each other.
Minor Subclass 0 Caveats
Because the AD9135/AD9136 require an ILAS, the nonmultiple
converter device alignment single lane (NMCDA-SL) case from
the JESD204A specification is supported only when using the
optional ILAS.
Error reporting using SYNCOUTx± is not supported when
using Subclass 0 with F = 1.
Subclass 1
This mode gives deterministic latency and allows links to be
synced to within ½ of a DAC clock period. It requires an
external SYSREF± signal that is accurately phase aligned to the
DAC clock.
Table 42. Sync Window Tolerance
Sync Error Window
Tolerance
±½ DAC clock cycles
±1 DAC clock cycles
±2 DAC clock cycles
±3 DAC clock cycles
ERRWINDOW (Register 0x034[2:0])
0x00
0x01
0x02
0x03
Rev. A | Page 44 of 117
Data Sheet
AD9135/AD9136
account for any amount of fixed delay. As a result, the LMFC
period must only be larger than the variation in the link delays,
and the AD9135/AD9136 can achieve proper performance with
a smaller total latency. Figure 54 and Figure 55 show a case
where the link delay is larger than an LMFC period. Note that it
can be accommodated by delaying LMFCRx.
DETERMINISTIC LATENCY REQUIREMENTS
Several key factors are required for achieving deterministic
latency in a JESD204B Subclass 1 system.
LMFC
ILAS
ALIGNED DATA
DATA
LATE ARRIVING
LMFC REFERENCE
EARLY ARRIVING
LMFC REFERENCE
Link Delay
Figure 54. Link Delay > LMFC Period Example
POWER CYCLE
VARIANCE
The link delay of a JESD204B system is the sum of fixed and
variable delays from the transmitter, channel, and receiver as
shown in Figure 56.
LMFC
ILAS
ALIGNED DATA
For proper functioning, all lanes on a link must be read during
the same LMFC period. Section 6.1 of the JESD204B
specification states that the LMFC period must be larger than
the maximum link delay. For the AD9135/AD9136, this is not
necessarily the case; instead, the AD9135/AD9136 use a local
LMFC for each link (LMFCRx) that can be delayed from the
SYSREF aligned LMFC. Because the LMFC is periodic, this can
DATA
LMFCRX
LMFC_DELAY_x
LMFC REFERENCE FOR ALL POWER CYCLES
FRAME CLOCK
Figure 55. LMFC_DELAY_x, to Compensate for Link Delay > LMFC
LINK DELAY = DELAYFIXED + DELAYVARIABLE
LOGIC DEVICE
(JESD204B Tx)
CHANNEL
JESD204B Rx
DSP
DAC
POWER CYCLE
VARIANCE
LMFC
ALIGNED DATA
CGS
ILAS
DATA
FIXED DELAY
VARIABLE
DELAY
Figure 56. JESD204B Link Delay = Fixed Delay + Variable Delay
Rev. A | Page 45 of 117
12578-019

POWER CYCLE
VARIANCE
12578-018

SYSREF± signal distribution skew within the system must
be less than the desired uncertainty.
SYSREF± setup and hold time requirements must be met
for each device in the system.
The total latency variation across all lanes, links, and
devices must be ≤10 PClock periods. This includes both
variable delays and the variation in fixed delays from lane
to lane, link to link, and device to device in the system.
12578-017

AD9135/AD9136
Data Sheet
1.
The method to set the LMFCDel and LMFCVar values is
described in the Link Delay Setup section.
Setting LMFCDel appropriately ensures that all the corresponding
data samples arrive in the same LMFC period. Then LMFCVar
is written into the receive buffer delay (RBD) to absorb all link
delay variation. This ensures that all data samples have arrived
before reading. By setting these to fixed values across runs and
devices, deterministic latency is achieved.
2.
The RBD described in the JESD204B specification takes values
from 1 frame clock cycle to K frame clock cycles, whereas the
RBD of the AD9135/AD9136 take values from 0 PClock cycles
to 10 PClock cycles. As a result, up to 10 PClock cycles of total
delay variation can be absorbed. Because LMFCVar is in PClock
cycles, and LMFCDel is in frame clock cycles, a conversion
between these two units is needed. The PClockFactor, or
number of frame clock cycles per PClock cycle, is equal to 4/F.
For more information on this relationship, see the Clock
Relationships section.
3.
4.
Two examples follow that show how to determine LMFCVar
and LMFCDel. After they are calculated, write LMFCDel into
both Register 0x304 and Register 0x305 for all devices in the
system, and write LMFCVar to both Register 0x306 and
Register 0x307 for all devices in the system.
5.
Link Delay Setup Example, with Known Delays
All the known system delays can be used to calculate LMFCVar
and LMFCDel as described in the Link Delay Setup section.
6.
The example shown in Figure 57 is demonstrated in the
following steps according to the procedure outlined in the Link
Delay Setup section. Note that this example is in Subclass 1 to
achieve deterministic latency, which has a PClockFactor (4/F)
of two frame clock cycles per PClock Cycle, and uses K = 32
(frames/multiframe). Because PCBFixed << PClockPeriod,
PCBFixed is negligible in this example and not included in the
calculations.
7.
Find the receiver delays using Table 8.
RxFixed = 17 PClock cycles
RxVar = 2 PClock cycles
Find the transmitter delays. The equivalent table in the
example JESD204B core (implemented on a GTH or GTX
transceiver on a Virtex-6 FPGA) states that the delay is
56 ± 2 byte clock cycles.
Because the PClockRate = ByteRate/4 as described in the
Clock Relationships section, the transmitter delays in
PClock cycles are as follows:
TxFixed = 54/4 = 13.5 PClock cycles
TxVar = 4/4 = 1 PClock cycle
Calculate MinDelayLane as follows:
MinDelayLane = floor(RxFixed + TxFixed + PCBFixed)
= floor(17 + 13.5 + 0)
= floor(30.5)
MinDelayLane = 30
Calculate MaxDelayLane as follows:
MaxDelayLane = ceiling(RxFixed + RxVar + TxFixed +
TxVar + PCBFixed))
= ceiling(17 + 2 + 13.5 + 1 + 0)
= ceiling(33.5)
MaxDelayLane = 34
Calculate LMFCVar as follows:
LMFCVar = (MaxDelay + 1) − (MinDelay − 1)
= (34 + 1) − (30 − 1) = 35 − 29
LMFCVar = 6 PClock cycles
Calculate LMFCDel as follows:
LMFCDel = ((MinDelay − 1) × PClockFactor) % K
= ((30 − 1) × 2) % 32 = (29 × 2) % 32
= 58 % 32
LMFCDel = 26 frame clock cycles
Write LMFCDel to both Register 0x304 and Register 0x305
for all devices in the system. Write LMFCVar to both
Register 0x306 and Register 0x307 for all devices in the
system.
LMFC
PCLK
FCLK
DYN_LINK_LATENCY_CNT
0
1
2
ALIGNED LANE DATA (A)
ALIGNED LANE DATA (B)
ALIGNED LANE DATA (C)
3
4
5
6
7
0
1
2
3
4
5
6
7
DATA
ILAS
ILAS
DATA
ILAS
DATA
LMFCRX
ILAS
LMFC_DELAY_x = 12
(FCLK CYCLES)
DATA
LMFC_VAR = 6
(PCLK CYCLES)
Figure 57. LMFC_DELAY_x Calculation Example
Rev. A | Page 46 of 117
12578-024
DETERMINISTICALLY
DELAYED DATA
Data Sheet
AD9135/AD9136
Link Delay Setup Example, Without Known Delay
If the system delays are not known, the AD9135/AD9136 can
read back the link latency between LMFCRX for each link and
the SYSREF± aligned LMFC. This information is then used to
calculate LMFCVar and LMFCDel, as shown in the Without
Known Delays section.
Figure 59 shows how DYN_LINK_LATENCY_x (Register 0x302
and Register 0x303) provides a readback showing the delay (in
PClock cycles) between LMFCRX and the transition from ILAS
to the first data sample. By repeatedly power cycling and taking
this measurement, the minimum and maximum delays across
power cycles can be determined and used to calculate LMFCVar
and LMFCDel.
2.
3.
4.
The example shown in Figure 59 is demonstrated in the following
steps according to the procedure outlined in the Without Known
Delays section. Note that this example is in Subclass 1 to achieve
deterministic latency, which has a PClockFactor (frame clock rate/
PClockRate) of 2 and uses K = 16; therefore, PClocksPerMF = 8.
In Figure 59, for Link A, Link B, and Link C, the system
containing the AD9135/AD9136 (including the
transmitter) is power cycled and configured 20 times. The
AD9135/AD9136 are configured as described in the Device
Setup Guide. Because the point of this exercise is to
determine LMFCDel and LMFCVar, the LMFCDel is
programmed to 0 and DYN_LINK_LATENCY_x is read
from Register 0x302 and Register 0x303 for Link 0 and
6.
SYSREF
LMFCRX
ILAS
ALIGNED DATA
DATA
12578-022
DYN_LINK_LATENCY
Figure 58. DYN_LINK_LATENCY Illustration
LMFC
PCLK
FRAME CLOCK
DYN_LINK_LATENCY_CNT
0
1
2
ALIGNED LANE DATA (A)
ALIGNED LANE DATA (B)
ALIGNED LANE DATA (C)
3
4
5
6
7
8
9
10
11
ILAS
12
13
14
15
DATA
ILAS
DATA
ILAS
DATA
LMFCRX
DETERMINISTICALLY
DELAYED DATA
ILAS
LMFC_DELAY = 6
(FCLK CYCLES)
DATA
LMFC_VAR = 7
(PCLK CYCLES)
Figure 59. Multilink Synchronization Settings, Derived Method Example
Rev. A | Page 47 of 117
12578-025
1.
5.
Link 1, respectively. The variation in the link latency over
the 20 runs is shown in Figure 59 in gray.
Link A gives readbacks of 6, 7, 0, and 1. Note that the set of
recorded delay values rolls over the edge of a multiframe at
the boundary K/PClockFactor = 8. Add PClocksPerMF = 8
to low set. Delay values range from 6 to 9.
Link B gives Delay values from 5 to 7.
Link C gives Delay values from 4 to 7.
Calculate the minimum of all Delay measurements across
all power cycles, links, and devices:
MinDelay = min(all Delay values) = 4
Calculate the maximum of all Delay measurements across
all power cycles, links, and devices:
MaxDelay = max(all Delay values) = 9
Calculate the total Delay variation (with guard band)
across all power cycles, links, and devices:
LMFCVar = (MaxDelay + 1) − (MinDelay − 1)
= (9 + 1) − (4 − 1) = 10 − 3 = 7 PClock cycles
Calculate the minimum delay in frame clock cycles (with
guard band) across all power cycles, links, and devices:
LMFCDel = ((MinDelay − 1) × PClockFactor) % K
= ((4 − 1) × 2) % 16 = (3 × 2) % 16
= 6 % 16 = 6 frame clock cycles
Write LMFCDel to both Register 0x304 and Register 0x305
for all devices in the system. Write LMFCVar to both
Register 0x306 and Register 0x307 for all devices in the
system.
AD9135/AD9136
Data Sheet
TRANSPORT LAYER
TRANSPORT LAYER
(QBD)
LANE 0 OCTETS
DELAY
BUFFER 0
F2S_0
DAC0 [15:0] OR [11:0]
DELAY
BUFFER 1
F2S_1
DAC1 [15:0] OR [11:0]
LANE 3 OCTETS
PCLK_0
SPI CONTROL
LANE 4 OCTETS
LANE 7 OCTETS
12578-026
PCLK_1
SPI CONTROL
Figure 60. Transport Layer Block Diagram
The transport layer receives the descrambled JESD204B frames
and converts them to DAC samples based on the programmed
JESD204B parameters shown in Table 43. A number of device
parameters are defined in Table 44.
Table 44. JESD204B Device Parameters
Parameter
CF
Table 43. JESD204B Transport Layer Parameters
CS
Parameter
F
K
HD
L
M
S
Description
Number of octets per frame per lane: 1, 2, or 4.
Number of frames per multiframe.
K = 32 if F = 1, K = 16 or 32 otherwise.
Number of lanes per converter device (per link), as
follows:
1, 2, 4, or 8 (single-link mode).
1, 2, or 4 (dual-link mode).
Number of converters per device (per link), as follows:
1 or 2 (single-link mode).
1 (dual-link mode).
Number of samples per converter, per frame: 1 or 2.
N
Nʹ (or NP)
Description
Number of control words per device clock per link.
Not supported, must be 0.
Number of control bits per conversion sample. Not
supported, must be 0.
High density user data format. Used when samples
must be split across lanes.
Set to 1 when F = 1, otherwise 0.
Converter resolution = 16.
Total number of bits per sample = 16.
Certain combinations of these parameters, called JESD204B
operating modes, are supported by the AD9135/AD9136. See
Table 45 and Table 46 for a list of supported modes, along with
their associated clock relationships.
Rev. A | Page 48 of 117
Data Sheet
AD9135/AD9136
Table 45. Single-Link and Dual-Link JESD204B Operating Modes
Mode
Parameter
M (Converter Count)
L (Lane Count)
S (Samples per Converter per Frame)
F (Octets per Frame per Lane)
K2 (Frames per Multiframe)
HD (High Density)
N (Converter Resolution)
NP (Bits per Sample)
Example Clocks for 10 Gbps Lane Rate
PClock Rate (MHz)
Frame Clock Rate (MHz)
Data Rate (MHz)
1
2
81
1
4
2
1
32
1
16
16
9
1
2
1
1
32
1
16
16
10
1
1
1
2
16 or 32
0
16
16
250
250
250
250
1000
1000
250
500
500
Mode 8 can only be used with 1× interpolation. Other interpolation options are not available in this mode.
K must be 32 in Mode 8 and Mode 9. It can be 16 or 32 in Mode 10.
Table 46. Single-Link JESD204B Operating Modes
Mode
Parameter
M (Converter Count)
AD9135/AD9136 M Setting2
L (Lane Count)
AD9135/AD9136 L Setting2
S (Samples per Converter per Frame)
F (Octets per Frame, per Lane)
K3 (Frames per Multiframe)
HD (High Density)
N (Converter Resolution)
NP (Bits per Sample)
Example Clocks for 10 Gbps Lane Rate
PClock Rate (MHz)
Frame Clock Rate (MHz)
Data Rate (MHz)
111
2
1
8
4
2
1
32
1
16
16
12
2
1
4
2
1
1
32
1
16
16
13
2
1
2
1
1
2
16 or 32
0
16
16
250
250
250
250
250
250
250
1000
1000
1
Mode 11 can only be used with 1× interpolation. Other interpolation options are not available in this mode.
Note that for Mode 11 through Mode 13, the M and L parameters programmed on the receive side do not match the parameters on the transmit side. The parameters
on the transmit side reflect the true number of converters and lanes per link.
3
K must be 32 in Mode 11 and Mode 12. It can be 16 or 32 in Mode 13.
2
Rev. A | Page 49 of 117
AD9135/AD9136
Data Sheet
Configuration Parameters
Data Flow Through the JESD204B Receiver
The AD9135/AD9136 modes refer to the link configuration
parameters for L, K, M, N, NP, S, and F. Table 47 provides the
description and addresses for these settings.
The link configuration parameters determine how the serial bits
on the JESD204B receiver interface are deframed and passed on
to the DACs as data samples. Figure 61 shows a detailed flow of
the data through the various hardware blocks for Mode 11 (L = 8,
M = 2, S = 2, F = 1). Simplified flow diagrams for all other modes
are shown in Figure 62 through Figure 66.
Table 47. Configuration Parameters
JESD204B
Setting
L−1
F − 11
K−1
M−1
N−1
NP − 1
S−1
HD
F1
DID
BID
LID0
JESDV
1
Description
Number of lanes − 1.
Number of ((octets per frame) per
lane) − 1.
Number of frames per multiframe − 1.
Number of converters − 1.
Converter bit resolution − 1.
Bit packing per sample − 1.
Number of ((samples per converter)
per frame) − 1.
High density format. Set to 1 if F = 1.
Leave at 0 if F ≠ 1.
F parameter, in ((octets per frame) per
lane).
Device ID. Match the device ID sent
by the transmitter.
Bank ID. Match the bank ID sent by
the transmitter.
Lane ID for Lane 0. Match the lane ID
sent by the transmitter on Logical
Lane 0.
JESD204x version. Match the version
sent by the transmitter (0x0 =
JESD204A, 0x1 = JESD204B).
Address
0x453[4:0]
0x454[7:0]
0x455[4:0]
0x456[7:0]
0x457[4:0]
0x458[4:0]
0x459[4:0]
0x45A[7]
0x476[7:0]
0x450[7:0]
0x451[3:0]
0x452[4:0]
0x459[7:5]
Single- and Dual-Link Configuration
The AD9135/AD9136 use the settings contained in Table 45
and Table 46. Mode 8 to Mode 13 can be used for single-link
operation. Mode 8 to Mode 10 can also be used for dual-link
operation.
To use dual-link mode, set LINK_MODE (Register 0x300[3]) to 1.
In dual-link mode, Link 1 must be programmed with identical
parameters to Link 0. To write to Link 1, set LINK_PAGE
(Register 0x300[2]) to 1.
If single-link mode is being used, a small amount of power can
be saved by powering down the output buffer for SYNCOUT1±,
which can be done by setting Register 0x203[0] = 1.
Checking Proper Configuration
As a convenience, the AD9135/AD9136 provide some quick
configuration checks. Register 0x030[5] is high if an illegal
LMFC_DELAY value is used. Register 0x030[3] is high if an
unsupported combination of L, M, F, or S is used. Register 0x030[2]
is high if an illegal K is used. Register 0x030[1] is high if an illegal
SUBCLASSV is used.
Deskewing and Enabling Logical Lanes
The values that need to be written in Register 0x454 and Register 0x476 are
different, F − 1 and F, respectively.
After proper configuration, the logical lanes must be deskewed and
enabled to capture data.
Set Bit x in Register 0x46C to 1 to deskew Logical Lane x and to 0 if
that logical lane is not being used. Then, set Bit x in Register 0x47D
to 1 to enable Logical Lane x and to 0 if that logical lane is not
being used.
Rev. A | Page 50 of 117
Data Sheet
AD9135/AD9136
S0
S19
J19
J8
J1
J10
DESERIALIZER
S10
S9
SERDINx±
J9
J8
J1
J0
DESERIALIZER
S0
SERDINx±
J19 J18
S19
J11 J10
S10
S9
J1
J0
DESERIALIZER
S0
S19
SERDINx±
J19
J8
J1
J10
DESERIALIZER
S10
S9
SERDINx±
J9
J8
J1
J0
DESERIALIZER
S0
SERIAL JESD204B DATA (L = 8)
SAMPLES SPLIT ACROSS LANES
(HD = 1)
D0
DAC0
D15
D0
DESCRAMBLER
DESERIALIZER
SERDINx±
J9 J8
10-BIT/8-BIT
DECODE
D15
D15
D0
DAC1
D15
D0
2 CONVERTERS
(M = 2)
2 SAMPLES PER
CONVERTER PER FRAME
(S = 2)
40 BITS PARALLEL DATA
(ENCODED AND SCRAMBLED)
16-BIT NIBBLE GROUP
(N = 16)
1 OCTET PER LANE
(F = 1)
12578-350
SERDINx±
CONV0, SMPL0
DESERIALIZER
CONV1, SMPL0
J0
NIBBLE GROUP 0
J1
NIBBLE GROUP 0
S10
S9
CONV0, SMPL1
DESERIALIZER
SERDINx±
J9 J8
LANE 3 LANE 2 LANE 1 LANE 0
OCTET 0 OCTET 0 OCTET 0 OCTET 0
S19
J11 J10
TRANSPORT
LAYER
LANE 7 LANE 6 LANE 5 LANE 4
OCTET 0 OCTET 0 OCTET 0 OCTET 0
J19 J18
DATA LINK LAYER
CONV1, SMPL1
PHYSICAL
LAYER
SERDINx±
Figure 61. JESD204B Mode 11 Data Deframing
Mode Configuration Maps
Table 48 to Table 53 contain the SPI configuration map for each
mode shown in Figure 61 through Figure 66. Figure 61 through
Figure 66 show the associated data flow through the deframing
process of the JESD204B receiver for each of the modes. Mode 8
to Mode 13 apply to single-link operation. Mode 8 to Mode 10
also apply to dual-link operation. Register 0x300 must be set
accordingly for single- or dual-link operation.
For additional details regarding all the SPI registers, see the
Register Maps and Descriptions section.
Rev. A | Page 51 of 117
AD9135/AD9136
Data Sheet
Table 48. SPI Configuration Map—Register Settings for JESD204B Parameters for Mode 8
LANE 0,
OCTET 0
J0
J1
SERDINx±
J7
LANE 2,
OCTET 0
J6
J8
J9
SERDINx±
J1
SERDINx±
LANE 1,
OCTET 0
J6
J15 J14
1 OCTET PER LANE
(F = 1)
16-BIT NIBBLE GROUP
(N = 16)
2 SAMPLES PER
CONVERTER PER FRAME
(S = 2)
J7
SERIAL JESD204B DATA (L = 4)
SAMPLES SPLIT ACROSS LANES
(HD = 1)
J9
J8
J0
Description
Register 0x453[7] = 0 or 1: scrambling disabled or enabled, Register 0x453[4:0] = 0x3: L = 4 lanes per link
Register 0x454[7:0] = 0x00: F = 1 octet per frame
Register 0x455[4:0] = 0x1F: K = 32 frames per multiframe
Register 0x456[7:0] = 0x00: M = 1 converter per link
Register 0x457[7:6] = 0x0: always set CS = 0; Register 0x457[4:0] = 0x0F: N = 16, always set to 16-bit resolution
Register 0x458[7:5] = 0x0 or 0x1: Subclass 0 or Subclass 1; Register 0x458[4:0] = 0xF: NP = 16 bits per sample
Register 0x459[7:5] = 0x1: set to JESD204B version; Register 0x459[4:0] = 0x1: S = 2 (samples per converter) per frame
Register 0x45A[7] = 1: HD = 1; Register 0x45A[4:0] = 0x00: always set CF = 0
Register 0x46C[7:0] = 0x0F: deskew Link Lane 0 to Link Lane 3
Register 0x476[7:0] = 0x01: F = 1 octet per frame
Register 0x47D[7:0] = 0x0F: enable Link Lane 0 to Link Lane 3
SERDINx±
LANE 3,
OCTET 0
NIBBLE GROUP 0
CONVERTER 0, SAMPLE 0
CONVERTER 0, SAMPLE 1
D15 ... D0 (0)
D15 ... D0 (1)
1 CONVERTER
(M = 1)
DAC0
12578-352
Setting
0x03 or 0x83
0x00
0x1F
0x00
0x0F
0x0F or 0x2F
0x21
0x80
0x0F
0x01
0x0F
J15 J14
Address
0x453
0x454
0x455
0x456
0x457
0x458
0x459
0x45A
0x46C
0x476
0x47D
Figure 62. JESD204B Mode 8 Data Deframing
Rev. A | Page 52 of 117
Data Sheet
AD9135/AD9136
Table 49. SPI Configuration Map—Register Settings for JESD204B Parameters for Mode 9
J0
J1
J8
SERDINx±
LANE 0,
OCTET 0
LANE 1,
OCTET 0
NIBBLE GROUP 0
CONVERTER 0, SAMPLE 0
1 CONVERTER
(M = 1)
D15 ... D0
DAC0
Figure 63. JESD204B Mode 9 Data Deframing
Rev. A | Page 53 of 117
12578-035
1 OCTET PER LANE
(F = 1)
16-BIT NIBBLE GROUP
(N = 16)
1 SAMPLE PER
CONVERTER PER FRAME
(S = 1)
J9
SERIAL JESD204B DATA (L = 2)
SAMPLES SPLIT ACROSS LANES
(HD = 1)
J11 J10
Description
Register 0x453[7] = 0 or 1: scrambling disabled or enabled, Register 0x453[4:0] = 0x1: L = 2 lanes per link
Register 0x454[7:0] = 0x00: F = 1 octet per frame
Register 0x455[4:0] = 0x1F: K = 32 frames per multiframe
Register 0x456[7:0] = 0x00: M = 1 converter per link
Register 0x457[7:6] = 0x0: always set CS = 0; Register 0x457[4:0] = 0x0F: N = 16, always set to 16-bit resolution
Register 0x458[7:5] = 0x0 or 0x1: Subclass 0 or Subclass 1; Register 0x458[4:0] = 0xF: NP = 16 bits per sample
Register 0x459[7:5] = 0x1: set to JESD204B version; Register 0x459[4:0] = 0x0: S = 1 (sample per converter) per frame
Register 0x45A[7] = 1: HD = 1; Register 0x45A[4:0] = 0x00: always set CF = 0
Register 0x46C[7:0] = 0x0F: deskew Link Lane 0 and Link Lane 1
Register 0x476[7:0] = 0x01: F = 1 octet per frame
Register 0x47D[7:0] = 0x03: enable Link Lane 0 and Link Lane 1
J19 J18
Setting
0x01 or 0x81
0x00
0x1F
0x00
0x0F
0x0F or 0x2F
0x20
0x80
0x03
0x01
0x03
SERDINx±
Address
0x453
0x454
0x455
0x456
0x457
0x458
0x459
0x45A
0x46C
0x476
0x47D
AD9135/AD9136
Data Sheet
Table 50. SPI Configuration Map—Register Settings for JESD204B Parameters for Mode 10
2 OCTETS PER LANE
(F = 2)
16-BIT NIBBLE GROUP
(N = 16)
1 SAMPLE PER
CONVERTER PER FRAME
(S = 1)
LANE 0,
OCTET 0
LANE 1,
OCTET 0
NIBBLE GROUP 0
CONVERTER 0, SAMPLE 0
D15 ... D0
1 CONVERTER
(M = 1)
DAC0
Figure 64. JESD204B Mode 10 Data Deframing
Rev. A | Page 54 of 117
12578-036
SERIAL JESD204B DATA (L = 1)
SAMPLES SPLIT ACROSS LANES
(HD = 0)
J1
J0
Description
Register 0x453[7] = 0 or 1: scrambling disabled or enabled, Register 0x453[4:0] = 0x0: L = 1 lane per link
Register 0x454[7:0] = 0x01: F = 2 octets per frame
Register 0x455[4:0] = 0x0F or 0x1F: K = 16 or 32 frames per multiframe
Register 0x456[7:0] = 0x00: M = 1 converter per link
Register 0x457[7:6] = 0x0: always set CS = 0; Register 0x457[4:0] = 0x0F: N = 16, always set to 16-bit resolution
Register 0x458[7:5] = 0x0 or 0x1: Subclass 0 or Subclass 1; Register 0x458[4:0] = 0xF: NP = 16 bits per sample
Register 0x459[7:5] = 0x1: set to JESD204B version; Register 0x459[4:0] = 0x0: S = 1 (sample per converter) per frame
Register 0x45A[7] = 0: HD = 0; Register 0x45A[4:0] = 0x00: always set CF = 0
Register 0x46C[7:0] = 0x0F: deskew Link Lane 0
Register 0x476[7:0] = 0x02: F = 2 octets per frame
Register 0x47D[7:0] = 0x01: enable Link Lane 0
J19 J18
Setting
0x00 or 0x80
0x01
0x0F or 0x1F
0x00
0x0F
0x0F or 0x2F
0x20
0x00
0x01
0x02
0x01
SERDINx±
Address
0x453
0x454
0x455
0x456
0x457
0x458
0x459
0x45A
0x46C
0x476
0x47D
Data Sheet
AD9135/AD9136
Table 51. SPI Configuration Map—Register Settings for JESD204B Parameters for Mode 11
Address
0x453
Setting
0x03 or 0x83
0x454
0x455
0x456
0x457
0x458
0x459
0x45A
0x46C
0x476
0x47D
0x00
0x1F
0x00
0x0F
0x0F or 0x2F
0x21
0x80
0xFF
0x01
0xFF
1
Description
Register 0x453[7] = 0 or 1: scrambling disabled or enabled, Register 0x453[4:0] = 0x3: L = 4 lanes per link (L = 8 on
transmit side)1
Register 0x454[7:0] = 0x00: F = 1 octet per frame
Register 0x455[4:0] = 0x1F: K = 32 frames per multiframe
Register 0x456[7:0] = 0x00: M = 1 converter per link (M = 2 on transmit side)1
Register 0x457[7:6] = 0x0: always set CS = 0; Register 0x457[4:0] = 0x0F: N = 16, always set to 16-bit resolution
Register 0x458[7:5] = 0x0 or 0x1: Subclass 0 or Subclass 1; Register 0x458[4:0] = 0xF: NP = 16 bits per sample
Register 0x459[7:5] = 0x1: set to JESD204B version; Register 0x459[4:0] = 0x1: S = 2 (samples per converter) per frame
Register 0x45A[7] = 1: HD = 1; Register 0x45A[4:0] = 0x00: always set CF = 0
Register 0x46C[7:0] = 0x0F: deskew Link Lane 0 to Link Lane 7
Register 0x476[7:0] = 0x01: F = 1 octet per frame
Register 0x47D[7:0] = 0x0F: enable Link Lane 0 to Link Lane 7
Note that for Mode 11 through Mode 13, the M and L parameters programmed on the receive side do not match the parameters on the transmit side. The parameters
on the transmit side reflect the true number of converters and lanes per link.
See Figure 61 for an illustration of the AD9135/AD9136 JESD204B Mode 11 data deframing process.
Rev. A | Page 55 of 117
AD9135/AD9136
Data Sheet
Table 52. SPI Configuration Map—Register Settings for JESD204B Parameters for Mode 12
Address
0x453
Setting
0x01 or 0x81
0x454
0x455
0x456
0x457
0x458
0x459
0x45A
0x46C
0x476
0x47D
0x00
0x1F
0x00
0x0F
0x0F or 0x2F
0x20
0x80
0x33
0x01
0x33
2 CONVERTERS
(M = 2)
LANE 1,
OCTET 0
J1
J9
J8
SERDINx±
J0
J11 J10
SERDINx±
J19 J18
J1
J8
SERDINx±
J0
J11 J10
LANE 0,
OCTET 0
LANE 2,
OCTET 0
LANE 3,
OCTET 0
NIBBLE GROUP 0
NIBBLE GROUP 1
CONVERTER 0, SAMPLE 0
CONVERTER 1, SAMPLE 0
D15 ... D0
D15 ... D0
DAC0
DAC1
12578-354
1 OCTET PER LANE
(F = 1)
16-BIT NIBBLE GROUP
(N = 16)
1 SAMPLE PER
CONVERTER PER FRAME
(S = 1)
J9
SERIAL JESD204B DATA (L = 4)
SAMPLES SPLIT ACROSS LANES
(HD = 1)
SERDINx±
Note that for Mode 11 through Mode 13, the M and L parameters programmed on the receive side do not match the parameters on the transmit side. The parameters
on the transmit side reflect the true number of converters and lanes per link.
J19 J18
1
Description
Register 0x453[7] = 0 or 1: scrambling disabled or enabled, Register 0x453[4:0] = 0x1: L = 2 lanes per link (L = 4 on
transmit side)1
Register 0x454[7:0] = 0x00: F = 1 octet per frame
Register 0x455[4:0] = 0x1F: K = 32 frames per multiframe
Register 0x456[7:0] = 0x00: M = 1 converter per link (M = 2 on transmit side)1
Register 0x457[7:6] = 0x0: always set CS = 0; Register 0x457[4:0] = 0x0F: N = 16, always set to 16-bit resolution
Register 0x458[7:5] = 0x0 or 0x1: Subclass 0 or Subclass 1; Register 0x458[4:0] = 0xF: NP = 16 bits per sample
Register 0x459[7:5] = 0x1: set to JESD204B version; Register 0x459[4:0] = 0x0: S = 1 (sample per converter) per frame
Register 0x45A[7] = 1: HD = 1; Register 0x45A[4:0] = 0x00: always set CF = 0
Register 0x46C[7:0] = 0x0F: deskew Link Lane 0, Link Lane 1, Link Lane 4, and Link Lane 5
Register 0x476[7:0] = 0x01: F = 1 octet per frame
Register 0x47D[7:0] = 0x03: enable Link Lane 0, Link Lane 1, Link Lane 4, and Link Lane 5
Figure 65. JESD204B Mode 12 Data Deframing
Rev. A | Page 56 of 117
Data Sheet
AD9135/AD9136
Table 53. SPI Configuration Map—Register Settings for JESD204B Parameters for Mode 13
Address
0x453
Setting
0x00 or 0x80
0x454
0x455
0x456
0x457
0x458
0x459
0x45A
0x46C
0x476
0x47D
0x01
0x0F or 0x1F
0x00
0x0F
0x0F or 0x2F
0x20
0x00
0x11
0x02
0x11
J0
J1
J19 J18
DAC0
D0
D1
D4
D5
D6
D7
D8
D9
D10
D11
D12
D13
D14
D15
D0
D1
D2
D3
D4
D5
D6
D7
D8
D9
CONVERTER 1, SAMPLE 0
D10
CONVERTER 0, SAMPLE 0
D2
LANE 1, OCTET 1
D3
J1
LANE 1, OCTET 0
NIBBLE GROUP 1
D11
D12
D13
LANE 0, OCTET 1
NIBBLE GROUP 0
DAC1
12578-033
2 CONVERTERS
(M = 2)
LANE 0, OCTET 0
D14
2 OCTETS PER LANE
(F = 2)
16-BIT NIBBLE GROUP
(N = 16)
1 SAMPLE PER
CONVERTER PER FRAME
(S = 1)
SERDINx±
SERIAL JESD204B DATA (L = 2)
SAMPLES NOT SPLIT
ACROSS LANES
(HD = 0)
J19 J18
SERDINx±
J0
Note that for Mode 11 through Mode 13, the M and L parameters programmed on the receive side do not match the parameters on the transmit side. The parameters
on the transmit side reflect the true number of converters and lanes per link.
D15
1
Description
Register 0x453[7] = 0 or 1: scrambling disabled or enabled, Register 0x453[4:0] = 0x0: L = 1 lane per link (L = 2 on
transmit side)1
Register 0x454[7:0] = 0x01: F = 2 octets per frame
Register 0x455[4:0] = 0x0F or 0x1F: K = 16 or 32 frames per multiframe
Register 0x456[7:0] = 0x00: M = 1 converter per link (M = 2 on transmit side)1
Register 0x457[7:6] = 0x0: always set CS = 0; Register 0x457[4:0] = 0x0F: N = 16, always set to 16-bit resolution
Register 0x458[7:5] = 0x0 or 0x1: Subclass 0 or Subclass 1; Register 0x458[4:0] = 0xF: NP = 16 bits per sample
Register 0x459[7:5] = 0x1: set to JESD204B version; Register 0x459[4:0] = 0x0: S = 1 (sample per converter) per frame
Register 0x45A[7] = 0: HD = 0; Register 0x45A[4:0] = 0x00: always set CF = 0
Register 0x46C[7:0] = 0x0F: deskew Link Lane 0 and Link Lane 4
Register 0x476[7:0] = 0x02: F = 2 octets per frame
Register 0x47D[7:0] = 0x01: enable Link Lane 0 and Link Lane 4
Figure 66. JESD204B Mode 13 Data Deframing
Rev. A | Page 57 of 117
AD9135/AD9136
Data Sheet
JESD204B TEST MODES
Transport Layer Testing
PHY PRBS Testing
The JESD204B receiver in the AD9135/AD9136 supports the
short transport layer (STPL) test as described in the JESD204B
standard. This test can be used to verify the data mapping
between the JESD204B transmitter and receiver. To perform
this test, this function must be implemented in the logic device
and enabled there. Before running the test on the receiver side,
the link must be established and running without errors (see the
Device Setup Guide).
The JESD204B receiver on the AD9135/AD9136 includes a
PRBS pattern checker on the back end of its physical layer.
This functionality enables bit error rate (BER) testing of each
physical lane of the JESD204B link. The PHY PRBS pattern
checker does not require that the JESD204B link be established.
The pattern checker can synchronize with a PRBS7, PRBS15,
or PRBS31 data pattern. PRBS pattern verification can be
performed on multiple lanes at once. The error counts for
failing lanes are reported for one JESD204B lane at a time. The
process for performing PRBS testing on the AD9135/AD9136 is
as follows:
1.
2.
3.
4.
5.
6.
7.
8.
9.
Start sending a PRBS7, PRBS15, or PRBS31 pattern from
the JESD204B transmitter.
Select and write the appropriate PRBS pattern to
Register 0x316[3:2], as shown in Table 54.
Enable the PHY test for all lanes being tested by writing to
PHY_TEST_EN (Register 0x315). Each bit of Register 0x315
enables the PRBS test for the corresponding lane. For example,
writing a 1 to Bit 0 enables the PRBS test for Physical Lane 0.
Toggle PHY_TEST_RESET (Register 0x316[0]) from 0 to 1
then back to 0.
Set PHY_PRBS_ERROR_THRESHOLD (Register 0x317 to
Register 0x319) as desired.
Write a 0 and then a 1 to PHY_TEST_START
(Register 0x316[1]). The rising edge of PHY_TEST_START
starts the test.
Wait 500 ms.
Stop the test by writing PHY_TEST_START
(Register 0x316[1]) = 0.
Read the PRBS test results.
a. Each bit of PHY_PRBS_PASS (Register 0x31D)
corresponds to one SERDES lane: 0 is fail, 1 is pass.
b. The number of PRBS errors seen on each failing lane
can be read by writing the lane number to check (0 to 7)
in the PHY_SRC_ERR_CNT (Register 0x316[6:4]) and
reading the PHY_PRBS_ERR_CNT (Register 0x31A
to Register 0x31C). The maximum error count is 224 − 1.
If all bits of Register 0x31A to Register 0x31C are high,
the maximum error count on the selected lane has
been exceeded.
Table 54. PHY PRBS Pattern Selection
PHY_PRBS_PAT_SEL Setting,
(Register 0x316[3:2])
0b00 (default)
0b01
0b10
PRBS Pattern
PRBS7
PRBS15
PRBS31
The STPL test ensures that each sample from each converter is
mapped appropriately according to the number of converters
(M) and the number of samples per converter (S). As specified
in the JESD204B standard, the converter manufacturer specifies
what test samples are transmitted. Each sample must have a
unique value. For example, if M = 2 and S = 2, there are 4
unique samples transmitted repeatedly until the test is stopped.
The expected sample must be programmed into the device and
the expected sample is compared to the received sample one
sample at a time until all have been tested. The process to
perform this test on the AD9135/AD9136 is described as follows:
1.
2.
3.
4.
5.
6.
7.
8.
Synchronize JESD204B link.
Enable the STPL test at the JESD204B transmitter.
Select Converter 0 Sample 0 for testing. Write
SHORT_TPL_DAC_SEL (Register 0x32C[3:2]) = 0 and
SHORT_TPL_SP_SEL (Register 0x32C[5:4]) = 0.
Set the expected test sample for Converter 0, Sample 0.
Program the expected 11-/16-bit test sample into the
SHORT_TPL_REF_SP_x registers (Register 0x32E and
Register 0x32D).
Enable the STPL test. Write SHORT_TPL_TEST_EN
(Register 0x32C[0]) = 1.
Toggle the STPL reset. SHORT_TPL_TEST_RESET
(Register 0x32C[1]) from 0 to 1 then back to 0.
Check for failures. Read SHORT_TPL_FAIL
(Register 0x32F[0]): 0 is pass, 1 is fail.
Repeat Step 3 to Step 7 for each sample of each converter,
Conv0Sample0 through ConvM − 1SampleS − 1.
Repeated CGS and ILAS Test
As per Section 5.3.3.8.2 of the JESD204B specification, the
AD9135/AD9136 can check that a constant stream of /K28.5/
characters is being received, or that CGS followed by a constant
stream of ILAS is being received.
To run a repeated CGS test, send a constant stream of /K28.5/
characters to the AD9135/AD9136 SERDES inputs. Next, set up
the device and enable the links as described in the Device Setup
Guide section. Ensure that the /K28.5/ characters are being
received by verifying that the SYNCOUTx± has been deasserted
and that CGS has passed for all enabled link lanes by reading
Register 0x470. Program Register 0x300[2] = 0 to monitor the
status of lanes on Link 0, and Register 0x300[2] = 1 to monitor
the status of lanes on Link 1 for dual-link mode.
Rev. A | Page 58 of 117
Data Sheet
AD9135/AD9136
To run the CGS followed by a repeated ILAS sequence test, follow
the Device Setup Guide section; however, before performing the
last write (enabling the links), enable the ILAS test mode by
writing a 1 to Register 0x477[7]. Then, enable the links. When
the device recognizes four CGS characters on each lane, it
deasserts SYNCOUTx±. At this point, the transmitter starts
sending a repeated ILAS sequence.
Read Register 0x473 to verify that initial lane synchronization has
passed for all enabled link lanes. Program Register 0x300[2] = 0
to monitor the status of lanes on Link 0, and Register 0x300[2] = 1
to monitor the status of lanes on Link 1 for dual-link mode.
JESD204B ERROR MONITORING
Check for Error Count Over Threshold
In addition to reading the error count per lane and error type as
described in the Checking Error Counts section, the user can
check a register to see if the error count for a given error type
has reached a programmable threshold.
The same error threshold is used for the three error types
(disparity, not in table, and unexpected control character). The
error counters are on a per error type basis. To use this feature,
complete the following steps:
1.
2.
Disparity, Not in Table, and Unexpected Control
Character Errors
As per Section 7.6 of the JESD204B specification, the
AD9135/AD9136 can detect disparity errors, not in table errors,
and unexpected control character errors, and can optionally
issue a sync request and reinitialize the link when errors occur.
Note that the disparity error counter counts all characters with
invalid disparity, regardless of whether they are in the 8-bit/10-bit
decoding table. This is a minor deviation from the JESD204B
specification, which only counts disparity errors when they are
in the 8-bit/10-bit decoding table.
Checking Error Counts
The error count can be checked for disparity errors, not in table
errors, and unexpected control character errors. The error
counts are on a per lane and per error type basis. Note that the
lane select and counter select are programmed into Register 0x46B
and the error count is read back from the same address. To
check the error count, complete the following steps:
1.
2.
Select the desired link lane and error type of the counter to
view. Write these to Register 0x46B according to Table 55.
To select a link lane, first select a link (Register 0x300[2] =
0 to select Link 0 or Register 0x300[2] = 1 to select Link 1
(dual link only)).
Note that when using Link 1, Link Lane x refers to Logical
Lane x + 4.
Read the error count from Register 0x46B. Note the
maximum error count is equal to the error threshold set in
Register 0x47C.
Table 55. Error Counters
Addr.
0x46B
Bits
[6:4]
Variable
LaneSel
[1:0]
CntrSel
Description
LaneSel = x to monitor the error count
of Link Lane x. See the notes on link
lane in Step 1 of the Checking Error
Counts section.
CntrSel = 0b00 for bad running
disparity counter.
CntrSel = 0b01 for not in table error
counter.
CntrSel = 0b10 for unexpected control
character counter.
Program the desired error count threshold into
ERRORTHRES (Register 0x47C).
Read back the error status for each error type to see if the
error count has reached the error threshold.
Disparity errors are reported in Register 0x46D.
Not in table errors are reported in Register 0x46E.
Unexpected control character errors are reported in
Register 0x46F.
Error Counter and IRQ Control
The user can write to Register 0x46D and Register 0x46F to
reset or disable the error counts and to reset the IRQ for a given
lane. Note that these are the same registers that are used to
report error count over threshold (see the Check for Error
Count Over Threshold section); therefore, the readback is not
the value that was written. For each error type,
1.
2.
3.
Select the link lane to access. To select a link lane, first
select a link (Register 0x300[2] = 0 to select Link 0,
Register 0x300[2] = 1 to select Link 1 (dual link only)).
Note that when using Link 1, Link Lane x refers to
Logical Lane x + 4.
Decide whether to reset the IRQ, disable the error count,
and/or reset the error count for the given lane and error type.
Write the link lane and desired reset or disable action to
Register 0x46D to Register 0x46F according to Table 56.
Table 56. Error Counter and IRQ Control: Disparity
(Register 0x46D), Not In Table (Register 0x46E), Unexpected
Control Character (Register 0x46F)
Bits
7
Variable
RstIRQ
6
Disable_ErrCnt
5
RstErrCntr
[2:0]
LaneAddr
Rev. A | Page 59 of 117
Description
RstIRQ = 1 to reset IRQ for the lane
selected in Bits[2:0].
Disable_ErrCnt = 1 to disable the error
count for the lane selected in Bits[2:0].
RsteErrCntr = 1 to reset the error count
for the lane selected in Bits[2:0].
LaneAddr = x to monitor the error
count of Link Lane x. See the notes on
link lane in Step 1 of the Checking Error
Counts section.
AD9135/AD9136
Data Sheet
Monitoring Errors via SYNCOUTx±
Table 58. Sync Assertion Mask
When one or more disparity, not in table, or unexpected control
character error occurs, the error is reported on the SYNCOUTx±
pins as per Section 7.6 of the JESD204B specification. The
JESD204B specification states that the SYNCOUTx± signal is
asserted for exactly two frame periods when an error occurs. For
the AD9135/AD9136, the width of the SYNCOUTx± pulse can be
programmed to ½, 1, or 2 PClock cycles. The settings to achieve a
SYNCOUTx± pulse of 2 frame clock cycles are given in Table 57.
Addr.
0x47B
Table 57. Setting SYNCOUTx± Error Pulse Duration
JESD Mode
IDs
8, 9, 11, 12
10, 13
1
PClockFactor
(Frames/PClock)
4
2
5
UCC_S
CGS, Frame Sync, Checksum, and ILAS Monitoring
Register 0x470 to Register 0x473 can be monitored to verify
that each stage of the JESD204B link establishment has
occurred. Program Register 0x300[2] = 0 to monitor the status
of the lanes on Link 0, and Register 0x300[2] = 1 to monitor the
status of the lanes on Link 1.
See the Error Counter and IRQ Control section for information
on resetting the IRQ. See the Interrupt Request Operation
section for more information on IRQs.
Errors Requiring Reinitializing
A link reinitialization automatically occurs when four invalid
disparity characters are received as per Section 7.1 of the
JESD204B specification. When a link reinitialization occurs, the
resync request is five frames and nine octets long.
The user can optionally reinitialize the link when the error
count for disparity errors, not in table errors, or unexpected
control characters reaches a programmable error threshold. The
process to enable the reinitialization feature for certain error
types is as follows:
4.
NIT_S
0 (default)
1
For disparity, not in table, and unexpected control character
errors, error count over the threshold events are available as
IRQ events. Enable these events by writing to Register 0x47A[7:5].
The IRQ event status can be read at the same address
(Register 0x47A[7:5]) after the IRQs are enabled.
3.
6
Description
Set to 1 to assert SYNCOUTx±
if the disparity error count
reaches the threshold
Set to 1 to assert SYNCOUTx±
if the not in table error count
reaches the threshold
Set to 1 to assert SYNCOUTx±
if the unexpected control
character count reaches the
threshold
SYNCB_ERR_DUR
Disparity, Not in Table, and Unexpected Control
Character IRQs
2.
Bit Name
BADDIS_S
(Register 0x312[5:4]) Setting1
These register settings assert the SYNCOUTx± signal for 2 frame clock cycle
pulse widths.
1.
Bit No.
7
Set THRESHOLD_MASK_EN (Register 0x477[3]) = 1.
Note that when this bit is set, unmasked errors do not
saturate at either threshold or maximum value.
Enable the sync assertion mask for each type of error by
writing to the SYNCASSERTIONMASK register
(Register 0x47B[7:5]) according to Table 58.
Program the desired error counter threshold into
ERRORTHRES (Register 0x47C).
For each error type enabled in the SYNCASSERTIONMASK
register, if the error counter on any lane reaches the
programmed threshold, SYNCOUTx± falls, issuing a sync
request. Note that all error counts are reset when a link
reinitialization occurs. The IRQ does not reset and must be
reset manually.
Bit x of CODEGRPSYNCFLAG (Register 0x470) is high if Link
Lane x received at least four K28.5 characters and passed code
group synchronization.
Bit x of FRAMESYNCFLAG (Register 0x471) is high if Link
Lane x completed initial frame synchronization.
Bit x of GOODCHKSUMFLG (Register 0x472) is high if the
checksum sent over the lane matches the sum of the JESD204B
parameters sent over the lane during ILAS for Link Lane x. The
parameters can be added either by summing the individual fields
in registers or summing the packed register. If Register 0x300[6] =
0 (default), the calculated checksums are the lower eight bits of
the sum of the following fields: DID, BID, LID, SCR, L − 1, F − 1,
K − 1, M − 1, N − 1, SUBCLASSV, NP − 1, JESDV, S − 1, and HD.
If Register 0x300[6] = 1, the calculated checksums are the lower
eight bits of the sum of Register 0x400 to Register 0x40C and LID.
Bit x of INITLANESYNCFLG (Register 0x473) is high if Link
Lane x passed the initial lane alignment sequence.
CGS, Frame Sync, Checksum, and ILAS IRQs
Fail signals for CGS, frame sync, checksum, and ILAS are available
as IRQ events. Enable them by writing to Register 0x47A[3:0].
The IRQ event status can be read at the same address
(Register 0x47A[3:0]) after the IRQs are enabled. Write a 1 to
Register 0x470[7] to reset the CGS IRQ. Write a 1 to Register 0x471
to reset the frame sync IRQ. Write a 1 to Register 0x472 to reset
the checksum IRQ. Write a 1 to Register 0x473 to reset the ILAS
IRQ.
See the Interrupt Request Operation section for more information.
Rev. A | Page 60 of 117
Data Sheet
AD9135/AD9136
Configuration Mismatch IRQ
Power and Ground Planes
The AD9135/AD9136 have a configuration mismatch flag that
is available as an IRQ event. Use Register 0x47B[3] to enable the
mismatch flag (it is enabled by default), and then use
Register 0x47B[4] to read back its status and reset the IRQ
signal. See the Interrupt Request Operation section for more
information.
Solid ground planes are recommended to avoid ground loops
and to provide a solid, uninterrupted ground reference for the
high speed transmission lines that require controlled impedances.
Do not use segmented power planes as a reference for controlled
impedances unless the entire length of the controlled impedance
trace traverses across only a single segmented plane. These and
additional guidelines for the topology of high speed transmission
lines are described in the JESD204B Serial Interface Inputs
(SERDIN0± to SERDIN7±) section.
The configuration mismatch event flag is high when the link
configuration settings (in Register 0x450 to Register 0x45D) do
not match the JESD204B transmitted settings (Register 0x400 to
Register 0x40D). All these registers are paged per link (in
Register 0x300). For Mode 11 through Mode 13, the
configuration mismatch flag is high because the values for the
M and L parameters sent over the link do not match the
parameters programmed to Register 0x453 and Register 0x456.
Table 59. Power Supplies
Note that this function is different from the good checksum
flags in Register 0x472. The good checksum flags ensure that
the transmitted checksum matches a calculated checksum based
on the transmitted settings. The configuration mismatch event
ensures that the transmitted settings match the configured settings.
Supply Domain
DVDD121
PVDD122
SVDD123
CVDD121
IOVDD
VTT4
SIOVDD33
AVDD33
HARDWARE CONSIDERATIONS
1
Power Supply Recommendations
The power supply domains are described in Table 59. The
power supplies can be grouped into separate PCB domains as
show in Figure 67. All the AD9135/AD9136 supply domains
must remain as noise free as possible for the best operation.
Power supply noise has a frequency component that affects
performance, and is specified in terms of V rms. Figure 68
shows the recommended power supply components.
Voltage (V)
1.2
1.2
1.2
1.2
1.8
1.2
3.3
3.3
Circuitry
Digital core
DAC PLL
JESD204B receiver interface
DAC clocking
SPI interface
VTT
Sync LVDS transmit
DAC
This supply requires a 1.3 V supply when operating at maximum DAC sample
rates. See Table 3 for details.
2
This supply can be combined with CVDD12 on the same regulator with a
separate supply filter network and sufficient bypass capacitors near the pins.
3
This supply requires a 1.3 V supply when operating at maximum interface
rates. See Table 4 for details.
4
This supply can be connected to SVDD12 and does not need separate circuitry.
An LC filter on the output of the power supply is recommended
to attenuate the noise, and must be placed as close to the
AD9135/AD9136 as possible. An effective filter is shown in
Figure 67. This filter scheme reduces high frequency noise
components. Each of the power supply pins of the AD9135/
AD9136 must also have a 0.1 μF capacitor connected to the
ground plane, as shown in Figure 67. Place the capacitor as close
to the supply pin as possible. Adjacent power pins can share a
bypass capacitor. Connect the ground pins of the AD9135/
AD9136 to the ground plane using vias.
Rev. A | Page 61 of 117
AD9135/AD9136
Data Sheet
1.8V FPGA VCCIO
OR
OTHER SYSTEM SUPPLY
10µH
IOVDD
66
10µF
1.2V
LINEAR
REGULATOR
DVDD12
13
10µF
14
53
3.3V
LINEAR
REGULATOR
SIOVDD33
35
10µF
AVDD33
67
75
1.2V
LINEAR
REGULATOR
SVDD12
17
10µF
77
20
85
22
28
1.2V
LINEAR
REGULATOR
10µH
31
CVDD12
71
32
10µF
76
33
81
36
87
39
1
45
10µH
PVDD12
10µF
47
4
50
7
VTT
8
21
10µF
9
25
10
42
56
46
57
12578-039
NOTES
1. UNLABELED CAPACITORS ARE 0.1µF, CLOSE TO DEVICE PIN(S), WITH MINIMUM DISTANCE
AND VIAS BETWEEN CAPACITORS AND PIN(S).
Figure 67. JESD204B Interface PCB Power Domain Recommendation
AD9135/AD9136
1.8V
POWER
INPUT
+12V
ADP1741
ADP2119
ADP1741
ADP1753
BUCK
1.2MHz/600kHz
800mA
3.8V
ADM7154-3.3
1.2V
1.2V
SVDD12 + VTT
DVDD12
CVDD12 + PVDD12
3.3V
AVDD33
3.3V
IOVDD + SIOVDD33
ADP2370
ADM7160-3.3
Figure 68. Power Supply Connections
Rev. A | Page 62 of 117
12578-568
+3.3V
STEP-DOWN DDC
1.2MHz, 2A
1.2V
Data Sheet
AD9135/AD9136
JESD204B Serial Interface Inputs (SERDIN0± to SERDIN7±)
Return Loss
When considering the layout of the JESD204B serial interface
transmission lines, there are many factors to consider to
maintain optimal link performance. Among these factors are
insertion loss, return loss, signal skew, and the topology of the
differential traces.
The JESD204B specification limits the amount of return loss
allowed in a converter device and a logic device, but does
not specify return loss for the channel. However, every effort
must be made to maintain a continuous impedance on the
transmission line between the transmitting logic device and the
AD9135/AD9136. As mentioned in the Insertion Loss section,
minimizing the use of vias, or eliminating them altogether,
reduces one of the primary sources for impedance mismatches
on a transmission line. Maintain a solid reference beneath
(for microstrip) or above and below (for stripline) the
differential traces to ensure continuity in the impedance of
the transmission line. If the stripline technique is used, follow
the guidelines listed in the Insertion Loss section to minimize
impedance mismatches and stub effects.
Insertion Loss
The JESD204B specification limits the amount of insertion loss
allowed in the transmission channel (see Figure 47). The
AD9135/AD9136 equalization circuitry allows significantly
more loss in the channel than is required by the JESD204B
specification. It is still important that the designer of the PCB
minimize the amount of insertion loss by adhering to the
following guidelines:


Keep the differential traces short by placing the
AD9135/AD9136 as near to the transmitting logic device
as possible and routing the trace as directly as possible
between the devices.
Route the differential pairs on a single plane using a solid
ground plane as a reference.
Use a PCB material with a low dielectric constant (<4) to
minimize loss, if possible.
When choosing between the stripline and microstrip
techniques, keep in mind the following considerations: stripline
has less loss (see Figure 48 and Figure 49) and emits less EMI,
but requires the use of vias that can add complexity to the task
of controlling the impedance; whereas microstrip is easier to
implement if the component placement and density allow
routing on the top layer and eases the task of controlling the
impedance.
If using the top layer of the PCB is problematic or the advantages
of stripline are desirable, follow these recommendations:


Minimize the number of vias.
If possible, use blind vias to eliminate via stub effects and
use micro vias to minimize via inductance.
If using standard vias, use the maximum via length to
minimize the stub size. For example, on an 8-layer board,
use Layer 7 for the stripline pair (see Figure 69).
For each via pair, place a pair of ground vias adjacent to them
to minimize the impedance discontinuity (see Figure 69).
LAYER 1
LAYER 2
ADD GROUND VIAS
y
STANDARD VIA
LAYER 6
MINIMIZE STUB EFFECT
There are many sources for signal skew, but the two sources to
consider when laying out a PCB are interconnect skew within a
single JESD204B link and skew between multiple JESD204B
links. In each case, keeping the channel lengths matched to
within 15 mm is adequate for operating the JESD204B link at
speeds of up to 10.64 Gbps. Managing the interconnect skew
within a single link is fairly straightforward. Managing multiple
links across multiple devices is more complex. However, follow
the 15 mm guideline for length matching.
Topology
Structure the differential SERDINx± pairs to achieve 50 Ω to
ground for each half of the pair. Stripline vs. microstrip tradeoffs are described in the Insertion Loss section. In either case,
it is important to keep these transmission lines separated from
potential noise sources such as high speed digital signals and
noisy supplies. If using stripline differential traces, route them
using a coplanar method, with both traces on the same layer.
Although this does not offer more noise immunity than the
broadside routing method (traces routed on adjacent layers),
it is easier to route and manufacture so that the impedance
continuity is maintained. An illustration of broadside vs.
coplanar differential routing techniques is shown in Figure 70.
Tx DIFF A
GND
LAYER 7
Signal Skew
y
DIFF+
LAYER 5
LAYER 8
y
DIFF–
LAYER 3
LAYER 4
GND
12578-040


Another primary source for impedance mismatch is at either
end of the transmission line, where care must be taken to match
the impedance of the termination to that of the transmission
line. The AD9135/AD9136 handle this internally with a
calibrated termination scheme for the receiving end of the line.
See the Interface Power-Up and Input Termination section for
details on this circuit and the calibration routine.
Figure 69. Minimizing Stub Effect and Adding Ground Vias for Differential
Stripline Traces
Tx DIFF B
Tx ACTIVE
BROADSIDE DIFFERENTIAL Tx LINES
Tx
DIFF A
Tx
DIFF B
Tx
ACTIVE
COPLANAR DIFFERENTIAL Tx LINES
Figure 70. Broadside vs. Coplanar Differential Stripline Routing Techniques
Rev. A | Page 63 of 117
12578-041

AD9135/AD9136
Data Sheet
SYNCOUTx±, SYSREF±, and CLK± Signals
When considering the trace width vs. copper weight and
thickness, the speed of the interface must be considered. At
multigigabit speeds, the skin effect of the conducting material
confines the current flow to the surface. Maximize the surface
area of the conductor by making the trace width made wider to
reduce the losses. Additionally, loosely couple differential traces
to accommodate the wider trace widths. This helps reduce the
crosstalk and minimize the impedance mismatch when the
traces must separate to accommodate components, vias,
connectors, or other routing obstacles. Tightly coupled vs.
loosely coupled differential traces are shown in Figure 71.
TIGHTLY COUPLED
DIFFERENTIAL Tx LINES
Tx
DIFF A
Tx
DIFF B
LOOSELY COUPLED
DIFFERENTIAL Tx LINES
It is important to keep similar trace lengths for the CLK± and
SYSREF± signals from the clock source to each of the devices
on either end of the JESD204B links (see Figure 72). If using a
clock chip that can tightly control the phase of CLK± and
SYSREF±, the trace length matching requirements are greatly
reduced.
Figure 71. Tightly Coupled vs. Loosely Coupled Differential Traces
AC Coupling Capacitors
The AD9135/AD9136 require that the JESD204B input signals
be ac-coupled to the source. These capacitors must be 100 nF
and placed as close as possible to the transmitting logic device.
To minimize the impedance mismatch at the pads, select the
package size of the capacitor so that the pad size on the PCB
matches the trace width as closely as possible.
LANE 0
LANE 1
Tx
DEVICE
Rx
DEVICE
LANE N – 1
LANE N
SYSREF
DEVICE CLOCK
SYSREF
CLOCK SOURCE
(AD9516-1, ADCLK925)
SYSREF TRACE LENGTH
DEVICE CLOCK TRACE LENGTH
DEVICE CLOCK
SYSREF TRACE LENGTH
DEVICE CLOCK TRACE LENGTH
Figure 72. SYSREF Signal and Device Clock Trace Length
Rev. A | Page 64 of 117
12578-043
Tx
DIFF B
Separate the SYNCOUTx± signal from other noisy signals,
because noise on the SYNCOUTx± may be interpreted as a
request for K characters. The SYNCOUTx± signal has two
modes of operation available for use. Register 0x2A5[0] defaults
to 0, which sets the SYNCOUTx± swing to normal swing mode.
When this bit is set to 1, the SYNCOUTx± swing is configured
for high swing mode. For more details, see Table 8.
12578-042
Tx
DIFF A
The SYNCOUTx± and SYSREF± signals on the AD9135/
AD9136 are low speed LVDS differential signals. Use controlled
impedance traces routed with 100 Ω differential impedance and
50 Ω to ground when routing these signals. As with the
SERDIN0± to SERDIN7± data pairs, it is important to keep
these signals separated from potential noise sources such as
high speed digital signals and noisy supplies.
Data Sheet
AD9135/AD9136
DIGITAL DATAPATH
INV
SINC
DIGITAL GAIN,
DC OFFSET,
AND
GROUP DELAY
ADJUSTMENT
INTERPOLATION FILTERS
12578-049
INTERPOLATION
MODES
1×, 2×, 4×, 8×
Figure 73. Block Diagram of Digital Datapath
The block diagram in Figure 73 shows the functionality of the
digital datapath (all blocks can be bypassed). The digital processing
includes three half-band interpolation filters, an inverse sinc filter,
and gain, offset, and group delay adjustment blocks.
Note that the pipeline delay changes when digital datapath
functions are enabled/disabled. If fixed DAC pipeline latency
is desired, do not reconfigure these functions after initial
configuration.
DAC PAGING
Digital datapath registers are paged to allow configuration of
either DAC independently or both simultaneously. Table 60 shows
how to use the DAC paging bits.
The transmit path contains three half-band interpolation filters,
which each provide a 2× increase in output data rate and a lowpass function. The filters can be cascaded to provide a 4× or 8×
interpolation ratio. Table 61 shows how to select each available
interpolation mode, their usable bandwidths, and their
maximum data rates. Note that fDATA = fDAC/Interpolation Factor.
Interpolation mode is paged as described in the DAC Paging
section. Register 0x030[0] is high if an unsupported
interpolation mode is selected.
Table 61. Interpolation Modes and Usable Bandwidth
Interpolation
Mode
1× (bypass)1
2×
4×
8×
INTERP_MODE,
Reg. 0x112[2:0]
0x00
0x01
0x03
0x04
Usable
Bandwidth
0.5 × fDATA
0.4 × fDATA
0.4 × fDATA
0.4 × fDATA
Maximum
fDATA (MHz)
21202
10602
700
350
1
Mode 8 and Mode 11 can only use 1× interpolation. 2×, 4×, and 8×
interpolation are only available in Mode 9, Mode 10, Mode 12, and Mode 13.
2
The maximum speed for 1× and 2× interpolation is limited by the JESD204B
interface. See Table 4 for the appropriate supply levels.
Table 60. Paging Modes
DACs Paged
DAC0
DAC1
DAC0 and DAC1
Filter Performance
DATA FORMAT
BINARY_FORMAT (Register 0x110[7]), paged as described in
the DAC Paging section) controls the expected input data
format. By default it is 0, which means the input data must be in
twos complement. It can also be set to 1, which means the input
data is in offset binary. For the AD9136, 0x0000 is negative full
scale and 0xFFFF is positive full scale. For the AD9135, 0x0000
is negative full scale and 0xFFE0 is positive full scale.
Though the AD9135 is an 11-bit resolution DAC at the output, the
input to the part must still be 16-bits wide for proper 8-bit/10-bit
decoding. The AD9135 uses a 16-bit datapath that is truncated
to 11 bits before going into the DAC core. Either 11-bit zeropadded data or full 16-bit data can be sent into the device, with
the latter having slightly better spectral performance by minimizing
quantization error along the datapath.
The usable bandwidth (as shown in Table 61) is defined as the
frequency band over which the filters have a pass-band ripple of
less than ±0.001 dB and an image rejection of greater than 85 dB.
0
2×
4×
8×
–20
–40
–60
–80
–100
0
0.2
0.4
0.6
0.8
FREQUENCY (×fDAC )
1.0
12578-368
Several functions are paged by DAC, such as input data format,
downstream protection, interpolation, inverse sinc, digital gain,
dc offset, group delay, datapath PRBS, and LMFC sync.
The interpolation filters interpolate between existing data in
such a way that they minimize changes in the incoming data
while suppressing the creation of interpolation images. This is
shown for each filter in Figure 74.
MAGNITUDE (dB)
DAC_PAGE, Register 0x008[1:0]
1
2
3 (default)
Figure 74. All Band Responses of Interpolation Filters
Filter Performance Beyond Specified Bandwidth
The interpolation filters are specified to 0.4 × fDATA (with pass
band). The filters can be used slightly beyond this ratio at the
expense of increased pass-band ripple and decreased interpolation
image rejection.
Rev. A | Page 65 of 117
AD9135/AD9136
Data Sheet
0
–0.1
70
–0.2
60
–0.3
50
–0.4
40
–0.5
30
20
40
IMAGE REJECTION
PASS-BAND RIPPLE
41
42
43
44
45
MAXIMUM PASS-BAND RIPPLE (dB)
80
DIGITAL GAIN, DC OFFSET, AND GROUP DELAY
–0.6
12578-369
MINIMUM INTERPOLATION IMAGE REJECTION (dB)
90
BANDWIDTH (% fDATA )
Digital gain and dc offset (as described in the Digital Gain
section and DC Offset section) allow compensation of imbalances
in the I and Q paths due to analog mismatches between DAC I/Q
outputs, quadrature modulator I/Q baseband inputs, and
DAC/modulator interface I/Q paths. These imbalances can
cause the two following issues:


Figure 75. Interpolation Filter Performance Beyond Specified Bandwidth
Figure 75 shows the performance of the interpolation filters
beyond 0.4 × fDATA. Note that the ripple increases much slower
than the image rejection decreases. This means that if the
application can tolerate degraded image rejection from the
interpolation filters, more bandwidth can be used.
INVERSE SINC
The AD9135/AD9136 provide a digital inverse sinc filter to
compensate the DAC roll-off over frequency. The filter is enabled
by setting the INVSINC_ENABLE bit (Register 0x111[7]; paged
as described in the DAC Paging section) and is enabled by default.
The inverse sinc (sinc−1) filter is a seven-tap FIR filter. Figure 76
shows the frequency response of sin(x)/x roll-off, the inverse
sinc filter, and the composite response. The composite response
has less than ±0.05 dB pass-band ripple up to a frequency of
0.4 × fDAC. To provide the necessary peaking at the upper end of
the pass band, the inverse sinc filter shown has an intrinsic
insertion loss of about 3.8 dB; in many cases, this can be partially
compensated as described in the Digital Gain section.
Group delay allows adjustment of the delay through the DAC,
which can be used to adjust digital predistortion (DPD) loop delay.
Digital Gain
Digital gain can be used to independently adjust the digital
signal magnitude being fed into each DAC. This is useful to
balance the gain between I and Q channels of a DAC or to
cancel out the insertion loss of the inverse sinc filter. Digital
gain must be enabled when using the blanking state machine
(see the Downstream Protection section). If digital gain is
disabled, TXENx must be tied high.
Digital gain is enabled by setting the DIG_GAIN_ENABLE bit
(Register 0x111[5], paged as described in the DAC Paging
section). In addition to enabling the function the amount of
digital gain (GainCode) desired must be programmed. By
default, digital gain is enabled and GainCode is 0xAEA.
0 ≤ Gain ≤ 4095/2048
−∞ dB ≤ dBGain ≤ 6.018 dB
Gain = GainCode × (1/2048)
1
dBGain = 20 × log10(Gain)
SIN(x)/x ROLL-OFF
SINC–1 FILTER RESPONSE
COMPOSITE RESPONSE
0
An unwanted sideband signal to appear at the quadrature
modulator output with significant energy. Tuning the
quadrature gain adjust values can optimize image rejection
in single sideband radios or can optimize the error vector
magnitude (EVM) in zero IF (ZIF) architectures.
The I/Q mismatch can cause LO leakage through a
modulator, which can be tuned out using dc offset.
GainCode = 2048 × Gain = 2048 × 10dBGain/20
MAGNITUDE (dB)
where GainCode is a 12-bit unsigned binary number.
–1
The I/Q digital gain is set as shown in Table 62 and paged as
described in the DAC Paging section.
–2
–3
–5
0
0.05
0.10
0.15
0.20
0.25
0.30
0.35
FREQUENCY (× fIN) (Hz)
0.40
0.45
0.50
12578-058
–4
The default GainCode value (0xAEA = 2.7 dB) is appropriate to
counteract the insertion loss of the inverse sinc filter without
causing digital clipping when using 2× interpolation. This value
can be read from of Figure 76 at 0.25 × fDAC, because that is the
Nyquist rate when using a 2× interpolation. The recommended
GainCode values for 4× and 8× interpolation are 0xBB3 (3.3 dB)
and 0xBF8 (3.5 dB), respectively.
Figure 76. Responses of sin(x)/x Roll-Off, the Sinc−1 Filter, and the Composite
of the Two-Input Signal and Protection
Rev. A | Page 66 of 117
Data Sheet
AD9135/AD9136
Table 63. DC Offset Registers
Table 62. Digital Gain Registers
Register
0x111[5]
0x13C[7:0]
0x13D[3:0]
Bit Name
DIG_GAIN_ENABLE
DAC_DIG_GAIN[7:0]
DAC_DIG_GAIN[11:8]
Description
Set to 1 to enable digital gain
LSB gain code
MSB gain code
Register
0x135[0]
0x136[7:0]
0x137[7:0]
0x13A[4:0]
Bit Name
DC_OFFSET_ON
LSB_OFFSET[7:0]
LSB_OFFSET[15:8]
SIXTEENTH_OFFSET
Description
Set to 1 to enable dc offset
LSB dc offset code
MSB dc offset code
Sub-LSB dc offset code
DC Offset
The dc offset feature is used to individually offset the data into
the I or Q DAC. This feature can be used to cancel LO leakage.
The offset is programmed as a 16-bit twos complement number
in LSBs, plus a 5-bit twos complement number in 16ths of an LSB,
as shown in Table 63. DC offset is paged as described in the
DAC Paging section.
−215 ≤ LSBs Offset < 215
−16 ≤ Sixteenths Offset  15
where
LSBs Offset is the value of Register 0x136 and Register 0x137.
Sixteenths Offset is the value of Register 0x13A.
Group Delay
Group delay can be used to delay the I or Q channels. This can
be useful, for example, for DPD loop delay adjustment.
−4 ≤ DAC Clock Cycles ≤ 3.5
Group Delay = (DAC Clock Cycles × 2) + 8
where Group Delay is a 4-bit twos complement number.
Write the GroupDelay to the GROUP_DLY register
(Register 0x014[3:0]). This feature is paged as described in the
DAC Paging section.
Rev. A | Page 67 of 117
AD9135/AD9136
Data Sheet
DOWNSTREAM PROTECTION
FILTER
AND
MODULATION
DATA
FROM LMFC
SYNC LOGIC
TXENx
DIGITAL
GAIN
BSM
Tx ENSM
DATA TO DACs
BSM_PROTECT
Tx_PROTECT
1
0
1
0
PROTECT_OUTx
Tx_PROTECT_OUT
SPI_PROTECT
PROTECT_OUT_INVERT
1
SPI_PROTECT_OUT
PROTECT OUTx GENERATION
12578-372
0
Figure 77. Downstream Protection Block Diagram
The AD9135/AD9136 have several blocks designed to protect the
power amplifier (PA) of the system, as well as other downstream
blocks. The AD9135/AD9136 consist of a blanking state machine
(BSM) and a transmit enable state machine (Tx ENSM).
after the BSM settles (see the Blanking State Machine (BSM)
section). If the masks are enabled, an additional delay is
imposed; the output is not valid until the BSM settles and the
DACs fully power on (nominally an additional ~35 μs).
The Tx ENSM is a block that controls delay between TXENx
and the Tx_PROTECT signal. The Tx_PROTECT signal is used
as an input to the BSM and its inverse can optionally be routed
externally. Optionally, the Tx ENSM can also power down its
associated DAC.
The Tx ENSM is configured as shown in Table 64 and is paged
as described in the DAC Paging section.
The BSM gently ramps data entering the DAC and flushes the
datapath. The BSM is activated by the Tx_PROTECT signal or
automatically by the LMFC sync logic during a rotation. For
proper function, digital gain must be enabled; tie TXENx high
if disabling digital gain.
Finally, some simple logic takes the outputs from each of those
blocks and uses them to generate a desired PROTECT_OUTx
signal on an external pin. This signal can be used to
enable/disable downstream components, such as a PA.
Transmit Enable State Machine
The Tx ENSM is a simple block that controls the delay between
the TXENx signal and the TX_PROTECT signal. This signal is
used as an input to the BSM and its inverse can be routed to an
external pin (PROTECT_OUTx) to turn downstream
components on or off as desired.
Table 64. Tx ENSM Registers
Addr.
0x11F
Bit No.
[7:6]
Bit Name
FALL_COUNTERS
[5:4]
RISE_COUNTERS
0x121
[7:0]
RISE_COUNT_0
0x122
[7:0]
RISE_COUNT_1
0x123
[7:0]
FALL_COUNT_0
0x124
[7:0]
FALL_COUNT_1
The TXENx signal can be used to power down the associated
DAC. If DAC0_MASK (Register 0x012[0]) = 1, a falling edge of
TXENx causes DAC0 to power down. If DAC1_MASK
(Register 0x012[1]) = 1, a falling edge of TXENx causes DAC1
to power down. On a rising edge of TXENx, without
DAC0_MASK and DAC1_MASK enabled, the output is valid
Rev. A | Page 68 of 117
Description
Number of fall counters
to use (1 to 2).
Number of rise
counters to use (0 to 2).
Delay TX_PROTECT rise
from TXENx rising edge
by 32 × RISE_COUNT_0
DAC clock cycles.
Delay TX_PROTECT rise
from TXENx rising edge
by 32 × RISE_COUNT_1
DAC clock cycles.
Delay TX_PROTECT rise
from TXENx rising edge
by 32 × FALL_COUNT_0
DAC clock cycles. Must
be at least 0x12.
Delay TX_PROTECT rise
from TXENx rising edge
by 32 × FALL_COUNT_1
DAC clock cycles.
Data Sheet
AD9135/AD9136
Blanking State Machine (BSM)
Table 66. PROTECT_OUTx Registers
The BSM gently ramps data entering the DAC and flushes the
datapath.
On a falling edge of TX_PROTECT (the TXENx signal delayed
by the Tx ENSM), the datapath holds the latest data value and
the digital gain gently ramps from its set value to 0. At the same
time, the datapath is flushed with zeros.
On a rising edge of TX_PROTECT, the TXENx signal is
delayed by the Tx ENSM; data is allowed to flow through the
datapath again and the digital gain gently ramps the data from 0
up to the set digital gain.
Both of these functions are also triggered automatically by the
LMFC sync logic during a rotation to prevent glitching on the
output.
Ramping
For proper ramping, digital gain must be enabled; tie TXENx
high if disabling digital gain.
The step size to use when ramping gain to 0 or its assigned value
can be controlled via the GAIN_RAMP_DOWN_STEP[11:0]
registers (Register 0x142 and Register 0x143) and the
GAIN_RAMP_UP_STEP[11:0] registers (Register 0x140 and
Register 0x141). These registers are paged as described in the
DAC Paging section.
The current BSM state can be read back as shown in Table 65.
Table 65. Blanking State Machine Ramping Readbacks
Register
0x147[7:6]
Value
0b00
0b01
0b10
0b11
Description
Data is being held at midscale.
Ramping gain to 0. Data ramping to
midscale.
Ramping gain to assigned value. Data
ramping to normal amplitude.
Data at normal amplitude.
Blanking State Machine IRQ
Blanking completion is available as an IRQ event.
Use Register 0x021[5] to enable blanking completion for DAC0
and then use Register 0x025[5] to read back its status and reset
the IRQ signal.
Use Register 0x022[5] to enable blanking completion for DAC1
and then use Register 0x026[5] to read back its status and reset
the IRQ signal.
See the Interrupt Request Operation section for more information.
PROTECT_OUTx Generation
Register 0x013 controls which signals are OR’ed into the external
PROTECT_OUTx signal. Register 0x11F[2] can be used to invert
the PROTECT_OUTx signal. By default, PROTECT_OUTx is
high when the output is valid. Register 0x013 and Register 0x11F
are paged as described in the DAC Paging section.
Reg.
0x013
0x11F
Bit
No.
5
Bit Name
TX_PROTECT_OUT
3
SPI_PROTECT_OUT
2
2
SPI_PROTECT
PROTECT_OUT_INVERT
Description
1: Tx ENSM triggers
PROTECT_OUT
1: SPI_PROTECT
triggers PROTECT_OUT
Sets SPI_PROTECT
Inverts PROTECT_OUTx
DATAPATH PRBS
The datapath PRBS can be used to verify that the AD9135/
AD9136 datapath is receiving and correctly decoding data. The
datapath PRBS verifies that the JESD204B parameters of the
transmitter and receiver match, that the lanes of the receiver are
mapped appropriately, that the lanes have been appropriately
inverted, if necessary, and in general that the start-up routine
has been implemented correctly. The datapath PRBS test is
designed to support input data rates of up to 1060 MHz.
The datapath PRBS is paged as described in the DAC Paging
section. To run the datapath PRBS test, complete the following
steps:
1.
Set up the device in the desired operating mode. See the
Device Setup Guide section for details on setting up the
device.
2. Send the PRBS7 or PRBS15 data.
3. Write Register 0x14B[2] = 0 for PRBS7 or 1 for PRBS15.
4. Write Register 0x14B[1:0] = 0b11 to enable and reset the
PRBS test.
5. Write Register 0x14B[1:0] = 0b01 to enable the PRBS test
and release reset.
6. Wait 500 ms.
7. Check the status by checking the IRQ for DAC0 and DAC1
PRBS as described in the Datapath PRBS IRQ section.
8. If there are failures, set Register 0x008 = 0x01 to view the
status of DAC0. Set Register 0x008 = 0x02 to view the
status of DAC1.
9. Read Register 0x14B[6]. Bit 6 is 0 if the selected DAC has
any errors. This must match the IRQ.
10. Read Register 0x14C to read the error count of the selected
DAC.
Note that the PRBS processes 32 bits at a time, and compares
the 32 new bits to the previous set of 32 bits. It detects (and
reports) only 1 error in every group of 32 bits; therefore, the
error count partly depends on when the errors are seen. For
example,



Rev. A | Page 69 of 117
Bits: 32 good, 31 good, 1 bad; 32 good (2 errors)
Bits: 32 good, 22 good, 10 bad; 32 good (2 errors)
Bits: 32 good, 31 good, 1 bad; 31 good, 1 bad; 32 good
(3 errors)
AD9135/AD9136
Data Sheet
Datapath PRBS IRQ
DC TEST MODE
The PRBS fail signals for each DAC are available as IRQ events.
Use Register 0x020, Bit 2 and Bit 0, to enable the fail signals,
and then use Register 0x024, Bit 2 and Bit 0, to read back their
statuses and reset the IRQ signals. See the Interrupt Request
Operation section for more information.
As a convenience, the AD9135/AD9136 provide a dc test mode,
which is enabled by setting Register 0x520[1] to 1 and clearing
Register 0x146[0] to 0. When this mode is enabled, the datapath
is given 0 (midscale) for its data. Register 0x146[0] must be set
to 1 for all other modes of operation.
In conjunction with dc offset, this test mode can provide the
desired dc data to the DACs.
Rev. A | Page 70 of 117
Data Sheet
AD9135/AD9136
INTERRUPT REQUEST OPERATION
The AD9135/AD9136 provide an interrupt request output
signal on Pin 60 (IRQ) that can be used to notify an external host
processor of significant device events. On assertion of the
interrupt, query the device to determine the precise event that
occurred. The IRQ pin is an open-drain, active low output. Pull
the IRQ pin high external to the device. This pin can be tied to
the interrupt pins of other devices with open-drain outputs to
wire; OR these pins together.
Table 67. IRQ Register Block Details
Figure 78 shows a simplified block diagram of how the IRQ
blocks works. If IRQ_EN is low, the INTERRUPT_SOURCE
signal is set to 0. If IRQ_EN is high, any rising edge of an event
causes the INTERRUPT_SOURCE signal to be set high. If any
INTERRUPT_SOURCE signal is high, the IRQ pin is pulled
low. INTERRUPT_SOURCE can be reset to 0 by either an
IRQ_RESET signal or a DEVICE_RESET.
Register Block
0x01F to 0x026
Event
Reported
Per chip
0x46D to 0x46F; 0x470
to 0x473; 0x47A
0x47B[4]
Per link and
lane
Per link
EVENT_STATUS
INTERRUPT_SOURCE if
IRQ is enabled, if not, it
is EVENT
INTERRUPT_SOURCE if
IRQ is enabled, if not, 0
INTERRUPT_SOURCE if
IRQ is enabled, if not, 0
INTERRUPT SERVICE ROUTINE
Interrupt request management begins by selecting the set of event
flags that require host intervention or monitoring. Enable the
events that require host action so that the host is notified when
they occur. For events requiring host intervention upon IRQ
activation, run the following routine to clear an interrupt request:
Depending on the STATUS_MODE signal, the EVENT_STATUS
bit reads back event or INTERRUPT_SOURCE. The AD9135/
AD9136 have several IRQ register blocks, which can monitor
up to 75 events (depending on device configuration). Certain
details vary by IRQ register block as described in Table 67.
Table 68 shows which registers the IRQ_EN, IRQ_RESET, and
STATUS_MODE signals in Figure 78 originate from, as well as
the address where EVENT_STATUS is read back.
1.
2.
3.
4.
5.
6.
7.
Read the status of the event flag bits that are being monitored.
Disable the interrupt by writing 0 to IRQ_EN.
Read the event source. For Register 0x01F to
Register 0x026, EVENT_STATUS has a live readback. For
other events, see their registers.
Perform any actions required to clear the cause of the EVENT.
In many cases, no specific actions may be required.
Verify that the event source is functioning as expected.
Clear the interrupt by writing 1 to IRQ_RESET.
Enable the interrupt by writing 1 to IRQ_EN.
0
1
EVENT_FLAG
IRQ
INTERRUPT_
SOURCE
INTERRUPT_ENABLE
OTHER
INTERRUPT
SOURCES
EVENT_FLAG_SOURCE
12578-060
WRITE_1_TO_EVENT_FLAG
DEVICE_RESET
Figure 78. Simplified Schematic of IRQ Circuitry
Table 68. IRQ Register Block Address of IRQ Signal Details
Register Block
0x01F to 0x026
0x46D to 0x46F
IRQ_EN
0x01F to 0x022; R/W per chip
0x47A; W per link
0x470 to 0x473
0x47B[4]
0x47A; W per link
0x47B[3]; R/W per link; 1 by
default
1
Address of IRQ Signals1
IRQ_RESET
STATUS_MODE
0x023 to 0x026; W per chip
STATUS_MODE = IRQ_EN
N/A, STATUS_MODE = 1
0x46D to 0x46F; W per link
and lane
0x470 to 0x473; W per link
N/A, STATUS_MODE = 1
0x47B[4]; W per link
N/A, STATUS_MODE = 1
N/A means not applicable.
Rev. A | Page 71 of 117
EVENT_STATUS
0x023 to 0x026; R per chip
0x47A; R per link
0x47A; R per link
0x47B[4]; R per link
AD9135/AD9136
Data Sheet
DAC INPUT CLOCK CONFIGURATIONS
The AD9135/AD9136 DAC sample clock (DACCLK) can be
sourced directly through CLK± (Pin 2 and Pin 3) or by clock
multiplication through the CLK± differential input. Clock
multiplication employs the on-chip PLL that accepts a reference
clock operating at a submultiple of the desired DACCLK rate.
The PLL then multiplies the reference clock up to the desired
DACCLK frequency, which is used to generate all the internal
clocks required by the DAC. The clock multiplier provides a
high quality clock that meets the performance requirements of
most applications. Using the on-chip clock multiplier removes
the burden of generating and distributing the high speed
DACCLK.
The second mode bypasses the clock multiplier circuitry and
allows DACCLK to be sourced directly to the DAC core. This
mode allows the user to source a very high quality clock directly to
the DAC core.
DRIVING THE CLK± INPUTS
The CLK± differential input circuitry is shown in Figure 79 as
a simplified circuit diagram of the input. The on-chip clock
receiver has a differential input impedance of 10 kΩ. It is self
biased to a common-mode voltage of about 600 mV. The inputs
can be driven by differential PECL or LVDS drivers with accoupling between the clock source and the receiver.
CLK+
To optimize the PLL across all operating conditions, the register
writes in Table 69 are recommended. These writes properly set
up the DAC PLL, including the loop filter and the charge pump.
Table 69. DAC PLL Fixed Register Writes
Register
Address
0x087
0x088
0x089
0x08A
0x08D
0x1B0
Register
Value
0x62
0xC9
0x0E
0x12
0x7B
0x00
0x1B9
0x1BC
0x1BE
0x24
0x0D
0x02
0x1BF
0x8E
0x1C0
0x2A
0x1C1
0x1C4
0x2A
0x7E
Description
Optimal DAC PLL loop filter settings
Optimal DAC PLL loop filter settings
Optimal DAC PLL loop filter settings
Optimal DAC PLL charge pump settings
Optimal DAC LDO settings for DAC PLL
Power DAC PLL blocks when power
machine disabled
Optimal DAC PLL charge pump settings
Optimal DAC PLL VCO control settings
Optimal DAC PLL VCO power control
settings
Optimal DAC PLL VCO calibration
settings
Optimal DAC PLL lock counter length
setting
Optimal DAC PLL charge pump setting
Optimal DAC PLL varactor settings
CLOCK MULTIPLICATION
5kΩ
The on-chip PLL clock multiplier circuit can be used to generate
the DAC sample rate clock from a lower frequency reference clock.
The PLL is integrated on-chip, including the VCO and the loop
filter. The VCO operates over the frequency range of 6 GHz to
12 GHz.
600mV
12578-061
5kΩ
CLK–
DAC PLL FIXED REGISTER WRITES
Figure 79. Clock Receiver Input Simplified Equivalent Circuit
The minimum input drive level to the differential clock input is
400 mV p-p differential. The optimal performance is achieved
when the clock input signal is between 800 mV p-p differential
and 1000 mV p-p differential. Whether using the on-chip clock
multiplier or sourcing the DACCLK directly (the CLK± pins are
used in both cases), the input clock signal to the device must
have low jitter and fast edge rates to optimize the DAC noise
performance. Direct clocking with a low noise clock produces
the lowest noise spectral density at the DAC outputs.
The clocks and clock receiver are powered down by default. The
clocks must be enabled by writing to Register 0x080. To enable
all clocks on the device, write Register 0x080 = 0x00.
Register 0x080, Bit 7 powers up the clocks for DAC0. Bit 6
powers up the clocks for DAC1. Bit 5 powers up the digital
clocks; Bit 4 powers up the SERDES clocks; and Bit 3 powers up
the clock receiver.
The PLL configuration parameters must be programmed before
the PLL is enabled. Step by step instructions on how to program the
PLL can be found in the Starting the PLL section. The functional
block diagram of the clock multiplier is shown in Figure 82.
The clock multiplication circuit generates the DAC sampling
clock from the REFCLK input, which is fed in on the CLK±
differential pins (Pin 2 and Pin 3). The frequency of the
REFCLK input is referred to as fREF.
The REFCLK input is divided by the variable RefDivFactor.
Select the RefDivFactor variable to ensure that the frequency
into the phase frequency detector (PFD) block is between
35 MHz and 80 MHz. The valid values for RefDivFactor are
1, 2, 4, 8, 16, or 32. Each RefDivFactor value maps to the
appropriate REF_DIV_MODE register control according to
Table 70. The REF_DIV_MODE register is programmed
through Register 0x08C[2:0].
Rev. A | Page 72 of 117
Data Sheet
AD9135/AD9136
Table 70. Mapping of RefDivFactor to REF_DIV_MODE
Divide by
(RefDivFactor)
1
2
4
8
16
REF_DIV_MODE,
Reg. 0x08C[2:0]
0
1
2
3
4
The range of fREF is 35 MHz to 1 GHz, and the output frequency
of the PLL is 420 MHz to 2.8 GHz. Use the following equations to
determine the RefDivFactor:
35 MHz 
f REF
 80 MHz
RefDivFactor
(1)
where:
fREF is the reference frequency on the CLK± input pins.
RefDivFactor is the reference divider division ratio.
Table 72. Common Frequency Examples
Frequency
(MHz)
368.64
184.32
307.2
122.88
61.44
491.52
245.76
fDAC
(MHz)
1474.56
1474.56
1228.88
983.04
983.04
1966.08
1966,08
f DACCLK
f REF

RefDivFactor
2  BCount
(2)
TO VCO
R1
C2
C3
C1
TO VCO LDO
Figure 80. Loop Filter
The PFD compares fREF/RefDivRate to fDAC/(2 × BCount) and
pulses the charge pump up or down to control the frequency of
the VCO. A low noise VCO is tunable over an octave with an
oscillation range of 6 GHz to 12 GHz.
The clock multiplication circuit operates such that the VCO
outputs a frequency, fVCO.
(3)
and from Equation 2, the DAC sample clock frequency, fDAC, is
equal to
Charge Pump
The charge pump current is 6-bit programmable and varies
from 0.1 mA to 6.4 mA in 0.1 mA steps. The charge pump
current is programmed into Register 0x08A for the DAC PLL
as shown in the DAC PLL Fixed Register Writes section. The
charge pump calibration must be run one time during chip
initialization to reduce reference spurs. This calibration is on
by default.
(4)
UP
TO LOOP FILTER
DOWN
Table 71. DAC VCO Divider Selection
Divide by
(LODivFactor)
4
8
16
LO_DIV_MODE,
Register 0x08B[1:0]
1
2
3
Rev. A | Page 73 of 117
CHARGE PUMP CURRENT = 0.1mA TO 6.4mA
Figure 81. Charge Pump
12578-063
The LODivFactor is chosen to keep fVCO in the operating range
between 6 GHz and 12 GHz. The valid values for LODivFactor
are 4, 8, and 16. Each LODivFactor maps to a LO_DIV_MODE
value. The LO_DIV_MODE (Register 0x08B[1:0]) is
programmed as described in Table 71.
DAC Frequency
Range (MHz)
>1500
750 to 1500
420 to 750
BCount
16
16
16
8
8
16
16
R3
FROM CHARGE PUMP
The BCount value is programmed with Bits[7:0] of
Register 0x085. It is programmable from 6 to 127.
f REF
RefDivFactor
LODivFactor
8
8
8
8
8
4
4
The RF PLL filter is fully integrated on-chip and is a standard
passive third-order filter with five 4-bit programmable
components (see Figure 80). The C1, C2, C3, R1, and R3 filter
components are programmed with Register 0x087 through
Register 0x089, as described in the DAC PLL Fixed Register
Writes section.
where:
fDAC is the DAC sample clock.
BCount is the feedback loop divider ratio.
f DACCLK  2  BCount 
RefDivFactor
8
4
8
2
1
8
4
Loop Filter
The BCount value is the divide ratio of the loop divider. It is set
to divide the fDAC to frequency match the fREF/RefDivFactor.
Select BCount so that the following equation is true:
fVCO = fDAC × LODivFactor
fVCO
(MHz)
11796.48
11796.48
9831.04
7864.35
7864.35
7864.35
7864.35
12578-062
DAC Reference
Frequency Range (MHz)
35 to 80
80 to 160
160 to 320
320 to 640
640 to 1000
Table 72 lists some common frequency examples for the
RefDivFactor, LODivFactor, and BCount values that are needed
to configure the PLL properly.
AD9135/AD9136
Data Sheet
Charge pump calibration is run during the first power-up of the
PLL, and the coefficient of the calibration is held for all subsequent
starts. The PLL is enabled by writing 0x10 into Register 0x083,
but the configuration registers must be programmed before the
PLL is enabled. The calibration tries to match the up and down
current, which minimizes the spurs at the reference frequency
that appears at the DAC output. The charge pump calibration
takes 64 reference clock cycles. Bit 5 in Register 0x084 notifies
the user that the charge pump calibration is completed and is
valid.
STARTING THE PLL
The programming sequence for the DAC PLL is as follows:
1.
2.
3.
4.
Temperature Tracking
When properly configured, the device automatically selects one of
the 512 VCO bands. The PLL settings selected by the device
ensure that the PLL remains locked over the full −40°C to +85°C
operating temperature range of the device without further
adjustment. The PLL remains locked over the full temperature
range, even if the temperature during initialization is at one of
the temperature extremes. Confirm the PLL lock bit to ensure
that the calibration completed properly. The PLL lock bit is Bit 1
of Register 0x084.
5.
6.
7.
Register 0x084[5] notifies the user that the DAC PLL calibration is
completed and is valid.
To properly configure temperature tracking, follow the settings
in the DAC PLL Fixed Register Writes section and the fVCO
dependent SPI writes shown in Table 73.
Register 0x084[1] notifies the user that the PLL has locked.
Register 0x084[7] and Register 0x084[6] notify the user that the
DAC PLL has reached the upper or lower edge of its operating
band, respectively. If either of these bits are high, recalibrate the
DAC PLL by setting Register 0x083[7] to 0 and then 1.
Table 73. VCO Control Lookup Table Reference
4-BIT
PROGRAMMABLE,
INTEGRATED
LOOP FILTER
PFD
80MHz
MAX
÷2
÷4
÷8
÷16
The DAC PLL lock and lost signals are available as IRQ events.
Use Register 0x01F[5:4] to enable these signals, and then use
Register 0x023[5:4] to read back their statuses and reset the IRQ
signals. See the Interrupt Request Operation section for more
information.
C1
RETIMER
UP
C2
C3
R1
DOWN
VCO
LDO
LC VCO
6GHz
TO
12GHz
÷2
R3
÷2
÷2
I
Q
÷2
I
Q
I
Q
ALC CAL
FO CAL
CAL CONTROL BITS
0.1mA TO 6.4mA
MUX/SELECTABLE BUFFERS
LODivFactor =
4, 8, 16
÷2
B COUNTER
BCount (INTEGER FEEDBACK DIVIDER)
RANGE = 6 TO 127
MAXIMUM FREQUENCY = 1.6GHz
Figure 82. Device Clock PLL Block Diagram
Rev. A | Page 74 of 117
DAC CLOCK
420MHz TO 2.8GHz
12578-064
35MHz
TO 1GHz
DAC PLL IRQ
<750MHz
CHARGE
PUMP
RefDivFactor =
1, 2, 4, 8, 16
fREF
Register
0x1BB
Setting
0x07
0x06
0x06
750MHz TO 1.5GHz
Register
0x1B6
Setting
0x03
0x03
0x13
>1.5GHz
Register
0x1B4
Setting
0x08
0x09
0x09
VCO Frequency
Range (GHz)
fVCO < 6.3
6.3 ≤ fVCO < 7.25
fVCO ≥ 7.25
Program the registers in the DAC PLL Fixed Register
Writes section.
Determine the VCO frequency based on the DAC
frequency requirements.
Determine the VCO divider ratio to achieve the desired
DAC frequency. Program the VCO divider ratio in
Register 0x08B[1:0].
Determine the BCount ratio to achieve the desired PLL
reference frequency (35 MHz to 80 MHz). Program the
BCount ratio in Register 0x085[7:0].
Determine the reference divider ratio to achieve the
desired PLL reference frequency. Program the reference
divider ratio in Register 0x08C[2:0].
Based on the fVCO found in Step 2, write the temperature
tracking registers as shown in Table 73.
Enable the DAC PLL synthesizer by setting Register 0x083[4]
to 1.
Data Sheet
AD9135/AD9136
ANALOG OUTPUTS
TRANSMIT DAC OPERATION
28
Figure 83 shows a simplified block diagram of the transmit path
DACs. The DAC core consists of a current source array, a switch
core, digital control logic, and full-scale output current control.
The DAC full-scale output current (IOUTFS) is nominally 20.48 mA.
The output currents from the OUTx± pins are complementary,
meaning that the sum of the two currents always equals the fullscale current of the DAC. The digital input code to the DAC
determines the effective differential current delivered to the load.
26
FSC (mA)
18
16
OUT1+
12
0
OUT1–
256
384
512
640
768
896
1024
Figure 84. DAC Full-Scale Current (IOUTFS) vs. Gain DAC Code
DAC0
DAC0
FULL-SCALE
ADJUST
OUT0+
Transmit DAC Transfer Function
OUT0–
The output currents from the OUTx+ and OUTx− pins are
complementary, meaning that the sum of the positive and negative
currents always equals the full-scale current of the DAC. The
digital input code to the DAC determines the effective differential
current delivered to the load. OUTx± provides the maximum
output current when all bits are high for binary data. The
output currents vs. DACCODE for the DAC outputs using
binary format are expressed as
12578-371
4kΩ
Figure 83. Simplified Block Diagram of DAC Core
The DAC has a 1.2 V band gap reference. A 4 kΩ external resistor,
RSET, must be connected from the I120 pin to the ground plane.
This resistor, along with the reference control amplifier, sets up
the correct internal bias currents for the DAC. Because the fullscale current is inversely proportional to this resistor, the
tolerance of RSET is reflected in the full-scale output amplitude.
DACFSC_x (where x is either 0 or 1, corresponding to DAC0 or
DAC1) is a 10-bit twos complement value that controls the fullscale current of each of the four DAC outputs. These values are
stored in Register 0x040 to Register 0x041 and Register 0x044
to Register 0x045 as shown in Table 74.
The typical full-scale current for each DAC is given by
I OUTP 
Table 74. DAC Full-Scale Current Registers
Description
DAC0 MSB gain code
DAC0 LSB gain code
DAC1 MSB gain code
DAC1 LSB gain code
(5)
(6)
where DACCODEBIN is the 11-/16-bit input to the DAC in
unsigned binary. DACCODEBIN has a range of 0 to 2N − 1.
If the data format is twos complement, the output currents are
expressed as
I OUTP 
For nominal values of VREF (1.2 V), RSET (4 kΩ), and DACFSC_x
(0, which is midscale in twos complement), the full-scale current of
the DAC is nominally 20.48 mA. The DAC full-scale current
can be adjusted from 13.9 mA to 27.0 mA, by programming the
appropriate DACFSC_x values in Register 0x040, Register 0x041,
and Register 0x044, and Register 0x045. Analog output full-scale
current vs. gain DAC code is shown in Figure 84.
DACCODEBIN
 I OUTFS
2N – 1
IOUTN = IOUTFS − IOUTP
IOUTFS = 20.45 + (DACFSC_x × 6.55 mA)/2(10 − 1)
Value
DACFSC_0[9:8]
DACFSC_0[7:0]
DACFSC_1[9:8]
DACFSC_1[7:0]
128
GAIN DAC CODE
CURRENT
SCALING
Address
0x040[1:0]
0x041[7:0]
0x044[1:0]
0x045[7:0]
20
14
DAC1
FULL-SCALE
ADJUST
DAC1
I120
22
12578-066
1.2V
24
DACCODETWOS  2N – 1
2N – 1
IOUTN = IOUTFS − IOUTP
 IOUTFS
(7)
(8)
where DACCODETWOS is the 11-/16-bit input to the DAC in twos
complement. DACCODETWOS has a range of −2N − 1 to 2N − 1 − 1.
Powering Down Unused DACs
Power down any unused DAC outputs to avoid burning excess
power. The DAC power downs are located in Register 0x011.
Register 0x011, Bit 6 corresponds to DAC0, and Bit 4 to DAC1.
Write a 1 to each bit to power down the appropriate DACs.
Register 0x011, Bit 7 and Bit 2, must stay low to enable the band
gap and DAC master bias, respectively.
Rev. A | Page 75 of 117
AD9135/AD9136
Data Sheet
–40
Self Calibration
4
–80
–100
0
50
100
200
250
300
Figure 87. Pre-Calibration and Post-Calibration, HD2 and HD4
–60
CALIBRATION OFF
CALIBRATION ON
–65
–75
–80
–85
0
–90
–1
–100
–3
0
50
100
150
200
250
–4
0
10k
20k
30k
40k
50k
60k
70k
DAC GAIN CODE
12578-088
fOUT (MHz)
Figure 88. Pre-Calibration and Post-Calibration, IMD2
To calibrate, follow the routine in Table 75.
Figure 85. Pre-Calibration and Post-Calibration, INL
4
Table 75. Device Self Calibration Procedure
CALIBRATION OFF
CALIBRATION ON
Addr.
0x0E7
3
2
Bit
[7:0]
SPI Data
Byte
0x38
3
2
1
0
[7:0]
[7:0]
[7:0]
[7:0]
0
0b0 or 0b1
0
0b0 or 0b1
0xA2
0x01
0x03
0x30
0x0E8
1
0
–2
20k
30k
40k
50k
60k
DAC GAIN CODE
Figure 86. Pre-Calibration and Post-Calibration, DNL
70k
12578-089
–1
10k
300
12578-096
–95
–2
0
150
fOUT (MHz)
12578-095
–90
IMD2 (dBc)
1
INL (LSB)
–70
–70
2
DNL (LSB)
SECOND HARMONIC
FOURTH HARMONIC
–60
CALIBRATION OFF
CALIBRATION ON
3
CALIBRATION OFF
CALIBRATION ON
–50
SFDR (dBc)
The AD9135/AD9136 have a self calibration feature that
improves the DAC dc and ac linearity in zero or low IF
applications. The performance improvement includes the
INL/DNL, second and fourth harmonic distortions (HD2 and
HD4), and second-order intermodulation distortion (IMD2) of
the device. Figure 85 and Figure 86 show the typical DAC INL
and DNL before and after the calibration. Figure 87 and Figure 88
show the calibration effect on the HD2, HD4, and IMD2
performance. The improvement from calibration decreases with
the DAC output frequency. For improvement in HD2 and HD4,
it is recommended to run the calibration routine when the
desired output frequency is below 100 MHz. For improvement
in IMD2, it is recommended to run the routine when the desired
output frequency is below 200 MHz. A single run of the routine is
sufficient to obtain the desired performance for both ac and dc
performance.
0x0ED
0x0E9
0x0E9
0x0E7
Description
Use highest comparator speed
and set calibration clock divider
Select DACs to calibrate
Set this bit to 0
1 if DAC1 is enabled
Set this bit to 0
1 if DAC0 is enabled
Configure initial value
Enable calibration
Start calibration
Disable calibration clock
For each DAC that is calibrated, verify the calibration status
by writing a 1 in the corresponding bit of CAL_PAGE
(Register 0x0E8) and reading Register 0x0E9. If the calibration
completed correctly, CAL_FIN (Register 0x0E9[7]) = 1 to
indicate that calibration is complete, and Register 0x0E9[6:4] = 0
to indicate that no errors have occurred.
Rev. A | Page 76 of 117
Data Sheet
AD9135/AD9136
–60
The post-calibration result is a function of operating
temperature. A set of calibration coefficients obtained at one
temperature may not be the optimal setting for a different
temperature. Figure 89 and Figure 90 show the typical
temperature drift effect after a single run calibration.
–75
–80
–85
–90
–95
–100
SFDR (dBc)
–70
–80
50
100
150
fOUT (MHz)
200
250
300
12578-378
–90
0
50
100
150
200
250
300
Figure 90. Post-Calibration IMD2 over Temperature, Calibrated at 25°C
SECOND HARMONIC
FOURTH HARMONIC
–60
–100
0
fOUT (MHz)
–40°C
+25°C
+85°C
–50
–70
Figure 89. Post-Calibration HD2 and HD4 over Temperature, Calibrated at 25°C
Rev. A | Page 77 of 117
12578-379
–40
–65
IMD2 (dBc)
For optimal performance, run the calibration again when the
operating temperature changes significantly. Note that it is
recommended to power down the DAC outputs when running
the calibration routine. If continuous transmission is required in
the system, running the calibration again during the operation
may not be an option. In this case, it is recommended to perform a
calibration at the average temperature of the operating temperature
range and to use the same set of coefficients during the operation.
This results in the best overall performance over temperature.
–40°C
+25°C
+85°C
AD9135/AD9136
Data Sheet
DEVICE POWER DISSIPATION
The AD9135/AD9136 have eight supply rails, AVDD33,
DVDD12, SVDD12, SIOVDD33, CVDD12, IOVDD, VTT, and
PVDD12, which can be driven from five regulators to achieve
optimum performance, as shown in Figure 67.
The AVDD33 supply powers the DAC core circuitry. The power
dissipation of the AVDD33 supply rail is independent of the
digital operating mode and sample rate. The current drawn
from the AVDD33 supply rail is typically 68 mA (225 mW)
when the full-scale current of DAC0 and DAC1 are set to the
nominal value of 20.48 mA.
PVDD12 powers the DAC PLLs and varies depending on the
DAC sample rate. CVDD12 can be combined with the PVDD12
regulator, but requires proper bypass capacitor networks near
the pins. CVDD12 powers the clock tree, and the current varies
directly with the DAC sample rate. DVDD12 powers the DSP
core, and the current draw depends on the number of DSP
functions and the DAC sample rate used. SVDD12 supplies the
SERDES lanes and associated circuitry including the equalizers,
SERDES PLL, PHY, and up to the input of the DSP. The current
depends on the number lanes and the lane bit rate. IOVDD
powers the SPI circuit and draws a very small current.
TEMPERATURE SENSOR
The AD9135/AD9136 have a band gap temperature sensor for
monitoring the temperature changes of the AD9135/AD9136.
The temperature must be calibrated against a known temperature
to remove the device-to-device variation on the band gap circuit
used to sense the temperature.
To monitor temperature change, the user must take a reading at
a known ambient temperature for a single-point calibration of
each AD9135/AD9136 device.
Tx = TREF + 7.3 × (CODE_X − CODE_REF)/1000
where:
CODE_X is the readback code at the unknown temperature, Tx.
CODE_REF is the readback code at the calibrated temperature,
TREF.
To use the temperature sensor, it must be enabled by setting
Register 0x12F[0] to 1. The user must write a 1 to Register 0x134[0]
before reading back the die temperature from Register 0x132
and Register 0x133.
SIOVDD33 powers the equalizers for the SERDES lanes. The
VTT termination voltage draws a very small current of <5 mA.
Rev. A | Page 78 of 117
Data Sheet
AD9135/AD9136
START-UP SEQUENCE
Table 76 through Table 83 show the register writes needed to set up
the AD9135/AD9136 with fDAC = 1474.56 MHz, 1× interpolation,
and the DAC PLL enabled with a 368.64 MHz reference clock.
The JESD204B interface is configured in Mode 11, single-link
mode, Subclass 1, and scrambling is enabled with all eight SERDES
lanes running at 7.3728 Gbps, inputting twos complement
formatted data. No remapping of lanes with the crossbar is used
in this example.
The sequence of steps to properly start up the AD9135/AD9136
is as follows:
1.
2.
3.
4.
5.
6.
Set up the SPI interface, power up necessary circuit blocks,
make required writes to the configuration register, and set up
the DAC clocks (see the Step 1: Start Up the DAC section).
Set the digital features of the AD9135/AD9136 (see the
Step 2: Digital Datapath section).
Set up the JESD204B links (see the Step 3: Transport Layer
section).
Set up the physical layer of the SERDES interface (see the
Step 4: Physical Layer section).
Set up the data link layer of the SERDES interface. This
procedure is for quick startup or debug only and does not
guarantee deterministic latency (see the Step 5: Data Link
Layer section).
Check for errors on Link 0 and Link 1 (see the Step 6:
Error Monitoring section).
These steps are outlined in detail in the following sections in
tables that list the required register write and read commands.
Configure the DAC PLL
Table 78. Configure DAC PLL
Command
W
Address
0x087
Value
0x62
W
0x088
0xC9
W
0x089
0x0E
W
0x08A
0x12
W
0x08D
0x7B
W
0x1B0
0x00
W
0x1B9
0x24
W
0x1BC
0x0D
W
0x1BE
0x02
W
0x1BF
0x8E
W
0x1C0
0x2A
W
0x1C1
0x2A
W
0x1C4
0x7E
W
0x08B
0x02
W
0x08C
0x03
W
0x085
0x10
W
0x1B5
0x09
W
0x1BB
0x13
W
0x1C5
0x06
W
R
0x083
0x084
0x10
0x01
STEP 1: START UP THE DAC
Power-Up and DAC Initialization
Table 76. Power-Up and DAC Initialization
Command
W
W
W
Address
0x000
0x000
0x011
Value
0xBD
0x3C
0x28
W
W
0x080
0x081
0x00
0x00
Description
Soft reset
Deassert reset, set 4-wire SPI
Enable reference, DAC channels,
and master DAC
Power up all clocks
Power up SYSREF± receiver,
disable hysteresis
Required Device Configurations
Table 77. Required Device Configurations
Command
W
W
W
W
W
Address
0x12D
0x146
0x2A4
0x232
0x333
Value
0x8B
0x01
0xFF
0xFF
0x01
Description
Digital datapath configuration
Digital datapath configuration
Clock configuration
SERDES interface configuration
SERDES interface configuration
Description
Optimal DAC PLL loop filter
settings
Optimal DAC PLL loop filter
settings
Optimal DAC PLL loop filter
settings
Optimal DAC PLL charge pump
settings
Optimal DAC LDO settings for
DAC PLL
Power DAC PLL blocks when
power machine disabled
Optimal DAC PLL charge pump
settings
Optimal DAC PLL VCO control
settings
Optimal DAC PLL VCO power
control settings
Optimal DAC PLL VCO
calibration settings
Optimal DAC PLL lock counter
length setting
Optimal DAC PLL charge pump
setting
Optimal DAC PLL varactor
settings
Set the VCO LO divider to 8 so
that 6 GHz ≤ fVCO = fDAC ×
2(LODivMode + 1) ≤ 12 GHz
Set the reference clock divider
to 8 so that the reference clock
into the PLL is less than 80 MHz
Set the B counter to 16 to
divide the DAC clock down to
2× the reference clock
PLL lookup value from Table 25
for fVCO  7.25 GHz
PLL lookup value from Table 25
for fVCO  7.25 GHz
PLL lookup value from Table 25
for fVCO  7.25 GHz
Enable DAC PLL
Verify that Bit 1 reads back
high for PLL locked
STEP 2: DIGITAL DATAPATH
Table 79. Digital Datapath
Command
W
W
Rev. A | Page 79 of 117
Address
0x112
0x110
Value
0x00
0x00
Description
Set the interpolation to 1×
Set twos complement data
format
AD9135/AD9136
Data Sheet
STEP 3: TRANSPORT LAYER
STEP 4: PHYSICAL LAYER
Table 80. Link 0 Transport Layer
Table 81. Physical Layer
Command
W
W
W
Address
0x200
0x201
0x300
Value
0x00
0x00
0x08
Command
W
Address
0x2AA
Value
0xB7
W
0x2AB
0x87
W
0x450
0x00
W
0x2B1
0xB7
W
0x451
0x00
W
0x2B2
0x87
W
0x452
0x00
W
0x453
0x83
W
W
W
W
0x2A7
0x2AE
0x314
0x230
0x01
0x01
0x01
0x28
W
W
W
0x454
0x455
0x456
0x00
0x1F
0x00
W
W
W
0x206
0x206
0x289
0x00
0x01
0x04
W
W
0x457
0x458
0x0F
0x2F
W
0x459
0x21
W
W
W
W
0x284
0x285
0x286
0x287
0x62
0xC9
0x0E
0x12
W
W
W
W
W
0x45A
0x45D
0x46C
0x476
0x47D
0x80
0x45
0xFF
0x01
0xFF
W
W
0x28A
0x28B
0x7B
0x00
W
0x290
0x89
W
0x294
0x24
W
W
W
0x296
0x297
0x299
0x03
0x0D
0x02
W
0x29A
0x8E
W
0x29C
0x2A
W
0x29F
0x78
W
0x2A0
0x06
W
R
0x280
0x281
0x01
0x01
W
0x268
0x62
1
Description
Power up the interface
Enable all lanes
Bit 3 = 0 for single link, Bit 2 = 0
to access Link 0 registers
Set the device ID to match Tx
(0x00 in this example)
Set the bank ID to match Tx
(0x00 in this example)
Set the lane ID to match Tx
(0x00 in this example)
Set descrambling and L to 4
(in n − 1 notation) (L = 8 on
transmit side)1
Set F = 1 (in n − 1 notation)
Set K = 32 (in n − 1 notation)
Set M to 1 (in n − 1 notation)
(M = 2 on transmit side)1
Set N = 16 (in n − 1 notation)
Set Subclass 1 and NP = 16
(in n − 1 notation)
Set JESD204B Version and S = 2
(in n − 1 notation)
Set HD = 1
Set checksum for Lane 0
Deskew Lane 0 to Lane 7
Set F (not in n − 1 notation)
Enable Lane 0 to Lane 7
Note that for Mode 11 through Mode 13, the M and L the parameters
programmed on the receive side do not match the parameters on the
transmit side. The parameters on the transmit side reflect the true number of
converters and lanes per link.
Rev. A | Page 80 of 117
Description
SERDES interface termination
setting
SERDES interface termination
setting
SERDES interface termination
setting
SERDES interface termination
setting
Autotune PHY setting
Autotune PHY setting
SERDES SPI configuration
Configure CDRs in half rate
mode
Resets CDR logic
Release CDR logic reset
Configure PLL divider to 1 along
with PLL required configuration
Optimal SERDES PLL loop filter
Optimal SERDES PLL loop filter
Optimal SERDES PLL loop filter
Optimal SERDES PLL charge
pump
Optimal SERDES PLL VCO LDO
Optimal SERDES PLL
configuration
Optimal SERDES PLL VCO
varactor
Optimal SERDES PLL charge
pump
Optimal SERDES PLL VCO
Optimal SERDES PLL VCO
Optimal SERDES PLL
configuration
Optimal SERDES PLL VCO
varactor
Optimal SERDES PLL charge
pump
Optimal SERDES PLL VCO
varactor
Optimal SERDES PLL VCO
varactor
Enable SERDES PLL
Verify that Bit 0 reads back high
for SERDES PLL lock
Set EQ mode to low power
Data Sheet
AD9135/AD9136
STEP 5: DATA LINK LAYER
STEP 6: ERROR MONITORING
Note that this procedure does not guarantee deterministic latency.
Link 0 Checks
Table 82. Data Link Layer—Does Not Guarantee Deterministic
Latency
Command
W
W
W
W
Address
0x301
0x304
0x305
0x306
Value
0x01
0x00
0x00
0x0A
W
0x307
0x0A
W
W
W
SYSREF±
Signal
W
0x03A
0x03A
0x03A
0x01
0x81
0xC1
0x300
0x01
Description
Set subclass = 1
Set the LMFC delay setting to 0
Set the LMFC delay setting to 0
Set the LMFC receive buffer
delay to 10
Set the LMFC receive buffer
delay to 10
Set sync mode to one-shot sync
Enable the sync machine
Arm the sync machine
Ensure that at least one SYSREF±
edge is sent to the device
Bit 0 = 1 to enable Link 0,
Bit 2 = 0 to access Link 0
Confirm that the registers in Table 83 read back as noted and
that system tasks are completed as described.
Table 83. Link 0 Checks
Command
R
Address
0x470
Value
0xFF
SYNCOUT0±
Signal
SERDINx±
Signals
R
0x471
0xFF
R
R
0x472
0x473
0xFF
0xFF
Rev. A | Page 81 of 117
Description
Acknowledge that four
consecutive K28.5 characters
have been detected on Lane
0 to Lane 3.
Confirm that SYNCOUT0± is
high.
Apply ILAS and data to
SERDES input pins.
Check for frame sync on all
lanes.
Check for good checksum.
Check for ILAS.
AD9135/AD9136
Data Sheet
REGISTER MAPS AND DESCRIPTIONS
In the following tables, register addresses (Reg. column) and reset (Reset column) values are hexadecimal and in the read/write (R/W)
column, R means read only, W means write only, R/W means read/write, and N/A means not applicable. All values in the register address
and reset columns are hexadecimal numbers.
DEVICE CONFIGURATION REGISTER MAP
Table 84. Device Configuration Register Map
Reg.
Name
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Reset
R/W
0x000
SPI_INTFCONFA
SOFT
RESET_M
LSBFIRST_
M
ADDRINC_M
SDOACTIVE_M
SDOACTIVE
ADDRINC
LSBFIRST
SOFTRESET
0x00
R/W
0x003
CHIPTYPE
CHIPTYPE
0x04
R
0x004
PRODIDL
PRODIDL
0x44
R
0x005
PRODIDH
0x006
CHIPGRADE
0x008
SPI_PAGEINDX
0x00A
SCRATCH_PAD
0x011
PWRCNTRL0
PRODIDH
PROD_GRADE
DEV_REVISION
RESERVED
DAC_PAGE
SCRATCHPAD
PD_BG
PD_DAC_0 RESERVED
PD_DAC_1
0x012
TXENMASK
0x013
PWRCNTRL3
0x014
GROUP_DLY
0x01F
IRQEN_
STATUSMODE0
0x020
IRQEN_
STATUSMODE1
0x021
IRQEN_
STATUSMODE2
IRQEN_
SMODE_
PDPERR0
RESERVED
IRQEN_
SMODE_
BLNKDONE0
RESERVED
0x022
IRQEN_
STATUSMODE3
IRQEN_
SMODE_
PDPERR1
RESERVED
IRQEN_
SMODE_
BLNKDONE1
RESERVED
0x023
IRQ_STATUS0
CALPASS
CALFAIL
DACPLLLOST DACPLLLOCK
RESERVED
PD_DACM
RESERVED
RESERVED
TX_PROTECT_ RESERVED
OUT
DAC1_MASK
RESERVED
IRQEN_
SMODE_
CALPASS
IRQEN_
SMODE_
CALFAIL
R
R
0x03
R/W
0x00
R/W
0x7C
R/W
DAC0_MASK 0x00
R/W
RESERVED
SPI_PROTECT_OUT SPI_PROTECT
0x91
0x46/
0x66
RESERVED
GROUP_DLY
0x20
R/W
0x88
R/W
IRQEN_SMODE_
SERPLLLOCK
RESERVED
IRQEN_
SMODE_
LANEFIFOERR
0x00
R/W
IRQEN_SMODE_
PRBS1
RESERVED
IRQEN_
SMODE_
PRBS0
0x00
R/W
IRQEN_SMODE_
SYNC_LOCK0
IRQEN_SMODE_
SYNC_ROTATE0
IRQEN_
SMODE_
SYNC_
WLIM0
0x00
IRQEN_
SMODE_
SYNC_TRIP0
R/W
IRQEN_SMODE_
SYNC_LOCK1
IRQEN_SMODE_
SYNC_ROTATE1
IRQEN_
SMODE_
SYNC_
WLIM1
0x00
IRQEN_
SMODE_
SYNC_TRIP1
R/W
SERPLLLOST
SERPLLLOCK
LANEFIFOERR
RESERVED
0x00
R
IRQEN_
IRQEN_SMODE_ IRQEN_SMODE_
SMODE_
DACPLLLOCK
SERPLLLOST
DACPLLLOST
RESERVED
0x024
IRQ_STATUS1
PRBS1
RESERVED
PRBS0
0x00
R
0x025
IRQ_STATUS2
PDPERR0
RESERVED
BLNKDONE0
RESERVED
RESERVED
SYNC_LOCK0
SYNC_ROTATE0
SYNC_
WLIM0
SYNC_TRIP0 0x00
R
0x026
IRQ_STATUS3
PDPERR1
RESERVED
BLNKDONE1
RESERVED
SYNC_LOCK1
SYNC_ROTATE1
SYNC_
WLIM1
SYNC_TRIP1 0x00
R
0x030
JESD_CHECKS
ERR_JESDBAD
ERR_KUNSUPP
ERR_
SUBCLASS
ERR_INTSUPP 0x00
R
0x034
SYNC_
ERRWINDOW
0x038
SYNC_LASTERR_L
0x039
SYNC_LASTERR_H LASTUNDER
LASTOVER
0x03A
SYNC_CONTROL
SYNCENABLE
SYNCARM
0x03B
SYNC_STATUS
SYNC_
BUSY
0x03C
SYNC_CURRERR_L
0x03D
SYNC_CURRERR_H CURRUNDER
RESERVED
ERR_DLYOVER ERR_WINLIMIT
RESERVED
ERRWINDOW
RESERVED
LASTERROR
RESERVED
SYNCCLRSTKY SYNCCLRLAST
RESERVED
SYNCMODE
SYNC_LOCK
SYNC_
ROTATE
RESERVED
SYNC_WLIM
CURRERROR
CURROVER
RESERVED
Rev. A | Page 82 of 117
SYNC_
TRIP
0x00
R/W
0x00
R
0x00
R
0x00
R/W
0x00
R
0x00
R
0x00
R
Data Sheet
Reg.
Name
0x040
DACGAIN0_1
0x041
DACGAIN0_0
0x044
DACGAIN1_1
0x045
DACGAIN1_0
0x080
CLKCFG0
AD9135/AD9136
Bit 7
Bit 6
PD_CLK0
0x083
DACPLLCNTRL
RECAL_
DACPLL
0x084
DACPLLSTATUS
DACPLL_
OVERRANGE_H
0x085
DACINTEGERWORD0
0x087
DACLOOPFILT1
0x08A
DACCPCNTRL
0x08B
DACLOGENCNTRL
0x08C
DACLDOCNTRL1
Bit 1
Bit 0
DACFSC_0[9:8]
DACFSC_1[9:8]
DACFSC_1[7:0]
SYSREF_ACTRL0
DACLOOPFILT2
Bit 2
RESERVED
SYSREF_ACTRL1
DACLOOPFILT3
Bit 3
DACFSC_0[7:0]
0x082
0x089
Bit 4
RESERVED
0x081
0x088
Bit 5
PD_CLK1
PD_CLK_DIG PD_SERDES_
PCLK
RESERVED
PD_CLK_REC
PD_SYSREF
RESERVED
HYS_ON
SYSREF_RISE
HYS_CNTRL1
DACPLL_
OVERRANGE_L
LF_R1_WORD
LF_
LF_BYPASS_
BYPASS_R1 C2
LF_BYPASS_C1
RESERVED
R
0x08
R/W
LF_C1_WORD
0x88
R/W
LF_C3_WORD
0x88
R/W
LF_R3_WORD
0x08
R/W
LO_DIV_MODE
RESERVED
DACLDOCNTRL2
CAL_CTRL_
GLOBAL
0x0E7
CAL_CLKDIV
RESERVED
0x0E8
CAL_PAGE
RESERVED
0x0E9
CAL_CTRL
REF_DIV_MODE
DAC_LDO
RESERVED
CAL_FIN
CAL_
ACTIVE
CAL_START_
AVG
CAL_CLK_EN
CAL_EN_
AVG
RESERVED
CAL_PAGE
CAL_ERRHI
CAL_ERRLO
R/W
0x00
RESERVED
CP_CURRENT
0x0E2
R/W
0xF8
R/W
RESERVED
0x08D
R/W
0x00
0x00
B_COUNT
LF_
BYPASS_
R3
0x00
R/W
DACPLL_
LOCK
LF_C2_WORD
R/W
R/W
RESERVED
DACPLL_
CAL_VALID
0x00
0x10
RESERVED
ENABLE_
DACPLL
R/W
R/W
0x00
HYS_CNTRL0
RESERVED
Reset
0x00
RESERVED
CAL_START
CAL_EN
CAL_INIT
0x20
R/W
0x02
R/W
0x01
R/W
0x2B
R/W
0x00
R/W
0x30
R/W
0x0F
R/W
0x00
R/W
A6
R/W
00
R/W
0xA0
R/W
0x01
R/W
0x83
R/W
0x0ED
CAL_INIT
0x110
DATA_FORMAT
BINARY_
FORMAT
0x111
DATAPATH_CTRL
INVSINC_
ENABLE
0x112
INTERP_MODE
0x11F
TXEN_SM_0
0x121
TXEN_RISE_
COUNT_0
RISE_COUNT_0
0x0F
R/W
0x122
TXEN_RISE_
COUNT_1
RISE_COUNT_1
0x00
R/W
0x123
TXEN_FALL_
COUNT_0
FALL_COUNT_0
0xFF
R/W
0x124
TXEN_FALL_
COUNT_1
FALL_COUNT_1
0xFF
R/W
0x12D
DEVICE_CONFIG_
REG_0
DEVICE_CONFIG_0
0x46
R/W
0x12F
DIE_TEMP_CTRL0
0x20
R/W
0x132
DIE_TEMP0
DIE_TEMP[7:0]
0x00
R
0x133
DIE_TEMP1
DIE_TEMP[15:8]
0x00
R
0x134
DIE_TEMP_
UPDATE
0x00
R/W
RESERVED
RESERVED
RESERVED
DIG_GAIN_
ENABLE
RESERVED
FALL_COUNTERS
INTERP_MODE
RISE_COUNTERS
RESERVED
RESERVED
RESERVED
Rev. A | Page 83 of 117
PROTECT_OUT_
INVERT
RESERVED
AUXADC_
ENABLE
DIE_TEMP_
UPDATE
AD9135/AD9136
Reg.
Name
0x135
DC_OFFSET_CTRL
0x136
DAC_DC_
OFFSET_1PART0
0x137
DAC_DC_
OFFSET_1PART1
0x13A
DAC_DC_
OFFSET_2PART
0x13C
DAC_DIG_GAIN0
0x13D
DAC_DIG_GAIN1
0x140
GAIN_RAMP_UP_
STEP0
0x141
GAIN_RAMP_
UP_STEP1
0x142
GAIN_RAMP_
DOWN_STEP0
0x143
GAIN_RAMP_
DOWN_STEP1
0x146
DEVICE_CONFIG_
REG_1
0x147
BSM_STAT
0x14B
PRBS
0x14C
PRBS_ERROR
0x1B0
DACPLLT0
0x1B5
DACPLLT5
0x1B9
DACPLLT9
0x1BB
DACPLLTB
0x1BC
DACPLLTC
0x1BE
DACPLLTE
Data Sheet
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Reset
R/W
DC_OFFSET_ 0x00
ON
R/W
LSB_OFFSET[7:0]
0x00
R/W
LSB_OFFSET[15:8]
0x00
R/W
0x00
R/W
RESERVED
RESERVED
Bit 0
SIXTEENTH_OFFSET
DAC_DIG_GAIN[7:0]
0xEA
R/W
0x0A
R/W
0x04
R/W
0x00
R/W
0x09
R/W
0x00
R/W
0x00
R/W
0x00
R
0x10
R/W
PRBS_COUNT
0x00
R
DAC_PLL_PWR
0xFA
R/W
0x83
R/W
0x34
R/W
0x0C
R/W
DAC_PLL_VCO_CTRL
0x00
R/W
DAC_PLL_VCO_PWR
0x00
R/W
RESERVED
DAC_DIG_GAIN[11:8]
GAIN_RAMP_UP_STEP[7:0]
RESERVED
GAIN_RAMP_UP_STEP[11:8]
GAIN_RAMP_DOWN_STEP[7:0]
RESERVED
GAIN_RAMP_DOWN_STEP[11:8]
DEVICE_CONFIG_1
SOFTBLANKRB
RESERVED
RESERVED PRBS_
GOOD
RESERVED
PRBS_MODE
RESERVED
PRBS_RESET
PRBS_EN
VCO_VAR
DAC_PLL_CP1
RESERVED
VCO_BIAS_TCF
VCO_BIAS_REF
0x1BF
DACPLLTF
DAC_PLL_VCOCAL
0x8D
R/W
0x1C0
DACPLLT10
DAC_PLL_LOCK_CNTR
0x2E
R/W
0x1C1
DACPLLT11
DAC_PLL_CP2
0x24
R/W
0x1C4
DACPLLT17
DAC_PLL_VAR1
0x33
R/W
0x1C5
DACPLLT18
DAC_PLL_VAR2
0x08
R/W
0x200
MASTER_PD
0x01
R/W
0x201
PHY_PD
0x203
GENERIC_PD
0x206
CDR_RESET
0x230
CDR_OPERATING_
MODE_REG_0
0x232
DEVICE_CONFIG_
REG_3
0x268
EQ_BIAS_REG
0x280
SERDESPLL_
ENABLE_CNTRL
0x281
PLL_STATUS
0x284
LOOP_FILTER_1
LOOP_FILTER_1
0x77
R/W
0x285
LOOP_FILTER_2
LOOP_FILTER_2
0x87
R/W
0x286
LOOP_FILTER_3
LOOP_FILTER_3
0x08
R/W
RESERVED
SPI_PD_
MASTER
SPI_PD_PHY
RESERVED
SPI_
SYNC1_PD
RESERVED
RESERVED
ENHALFRATE
RESERVED
CDR_
OVERSAMP
0x00
R/W
SPI_
SYNC2_PD
0x00
R/W
SPI_CDR_
RESETN
0x01
R/W
RESERVED
0x28
R/W
0x0
R/W
0x62
R/W
0x00
R/W
DEVICE_CONFIG_3
EQ_POWER_MODE
RESERVED
RESERVED
RESERVED
RECAL_
SERDESPLL
SERDES_PLL_ SERDES_PLL_
OVERRANGE_ OVERRANGE_L
H
SERDES_PLL_CAL_
VALID_RB
Rev. A | Page 84 of 117
RESERVED
RESERVED
ENABLE_
SERDESPLL
SERDES_PLL_ 0x00
LOCK_RB
R
Data Sheet
Reg.
Name
0x287
SERDES_PLL_CP1
0x289
REF_CLK_
DIVIDER_LDO
0x28A
VCO_LDO
0x28B
SERDES_PLL_PD1
0x290
0x294
AD9135/AD9136
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Reset
R/W
0x3F
R/W
0x00
R/W
SERDES_PLL_VCO_LDO
0x2B
R/W
SERDES_PLL_PD1
0x7F
R/W
SERDESPLL_VAR1
SERDES_PLL_VAR1
0x83
R/W
SERDES_PLL_CP2
SERDES_PLL_CP2
0xB0
R/W
0x296
SERDESPLL_VCO1
SERDES_PLL_VCO1
0x0C
R/W
0x297
SERDESPLL_VCO2
SERDES_PLL_VCO2
0x00
R/W
0x299
SERDES_PLL_PD2
SERDES_PLL_PD2
0x00
R/W
0x29A
SERDESPLL_VAR2
SERDES_PLL_VAR2
0xFE
R/W
SERDES_PLL_CP1
RESERVED
SERDES_PLL_DIV_MODE
DEVICE_
CONFIG_4
0x29C
SERDES_PLL_CP3
SERDES_PLL_CP3
0x17
R/W
0x29F
SERDESPLL_VAR3
SERDES_PLL_VAR3
0x33
R/W
0x2A0
SERDESPLL_VAR4
SERDES_PLL_VAR4
0x08
R/W
0x2A4
DEVICE_CONFIG_
REG_8
DEVICE_CONFIG_8
0x4B
R/W
0x2A5
SYNCOUTB_
SWING
RESERVED
SYNCOUTB_
SWING_MD
0x00
R/W
0x2A7
TERM_BLK1_
CTRLREG0
RESERVED
RCAL_
TERMBLK1
0x00
R/W
0x2AA DEVICE_CONFIG_
REG_9
DEVICE_CONFIG_9
0xC3
R/W
0x2AB
DEVICE_CONFIG_
REG_10
DEVICE_CONFIG_10
0x93
R/W
0x2AE
TERM_BLK2_
CTRLREG0
0x00
R/W
0x2B1
DEVICE_CONFIG_
REG_11
DEVICE_CONFIG_11
0xC3
R/W
0x2B2
DEVICE_CONFIG_
REG_12
DEVICE_CONFIG_12
0x93
R/W
0x300
GENERAL_JRX_
CTRL_0
0x00
R/W
0x301
GENERAL_JRX_
CTRL_1
0x01
R/W
0x302
DYN_LINK_
LATENCY_0
RESERVED
DYN_LINK_LATENCY_0
0x00
R
0x303
DYN_LINK_
LATENCY_1
RESERVED
DYN_LINK_LATENCY_1
0x00
R
0x304
LMFC_DELAY_0
RESERVED
LMFC_DELAY_0
0x00
R/W
0x305
LMFC_DELAY_1
RESERVED
LMFC_DELAY_1
0x00
R/W
0x306
LMFC_VAR_0
RESERVED
LMFC_VAR_0
0x06
R/W
0x307
LMFC_VAR_1
RESERVED
LMFC_VAR_1
0x06
R/W
0x308
XBAR_LN_0_1
RESERVED
LOGICAL_LANE0_SRC
0x08
R/W
0x309
XBAR_LN_2_3
RESERVED
LOGICAL_LANE3_SRC
LOGICAL_LANE2_SRC
0x1A
R/W
0x30A
XBAR_LN_4_5
RESERVED
LOGICAL_LANE5_SRC
LOGICAL_LANE4_SRC
0x2C
R/W
0x30B
XBAR_LN_6_7
RESERVED
LOGICAL_LANE7_SRC
LOGICAL_LANE6_SRC
0x3E
R/W
0x30C
FIFO_STATUS_
REG_0
LANE_FIFO_FULL
0x00
R
0x30D
FIFO_STATUS_
REG_1
LANE_FIFO_EMPTY
0x00
R
RESERVED
RESERVED CHECKSUM
_MODE
RESERVED
RCAL_
TERMBLK2
LINK_MODE
LINK_PAGE
RESERVED
RESERVED
LINK_EN
SUBCLASSV_LOCAL
LOGICAL_LANE1_SRC
0x312
SYNCB_GEN_1
0x00
R/W
0x314
SERDES_SPI_REG
SYNCB_ERR_DUR
SERDES_SPI_CONFIG
RESERVED
0x00
R/W
0x315
PHY_PRBS_TEST_
EN
PHY_TEST_EN
0x00
R/W
Rev. A | Page 85 of 117
AD9135/AD9136
Data Sheet
Reg.
Name
0x316
PHY_PRBS_TEST_ RESERVED
CTRL
Bit 7
Bit 6
0x317
PHY_PRBS_TEST_
THRESHOLD_
LOBITS
0x318
Bit 5
Bit 4
Bit 3
Bit 1
Bit 0
Reset
R/W
PHY_TEST_
START
PHY_TEST_
RESET
0x00
R/W
PHY_PRBS_THRESHOLD[7:0]
0x00
R/W
PHY_PRBS_TEST_
THRESHOLD_
MIDBITS
PHY_PRBS_THRESHOLD[15:8]
0x00
R/W
0x319
PHY_PRBS_TEST_
THRESHOLD_
HIBITS
PHY_PRBS_THRESHOLD[23:16]
0x00
R/W
0x31A
PHY_PRBS_TEST_
ERRCNT_LOBITS
PHY_PRBS_ERR_CNT[7:0]
0x00
R
0x31B
PHY_PRBS_TEST_
ERRCNT_MIDBITS
PHY_PRBS_ERR_CNT[15:8]
0x00
R
0x31C
PHY_PRBS_TEST_
ERRCNT_HIBITS
PHY_PRBS_ERR_CNT[23:16]
0x00
R
0x31D
PHY_PRBS_TEST_
STATUS
PHY_PRBS_PASS
0xFF
R
0x32C
SHORT_TPL_
TEST_0
0x32D
SHORT_TPL_
TEST_1
0x32E
PHY_SRC_ERR_CNT
RESERVED
Bit 2
PHY_PRBS_PAT_SEL
SHORT_TPL_SP_SEL
SHORT_TPL_DAC_SEL
SHORT_TPL_ 0x00
TEST_EN
R/W
SHORT_TPL_REF_SP_LSB
0x00
R/W
SHORT_TPL_
TEST_2
SHORT_TPL_REF_SP_MSB
0x00
R/W
0x32F
SHORT_TPL_
TEST_3
RESERVED
0x00
R
0x333
DEVICE_CONFIG_
REG_13
DEVICE_CONFIG_13
0x00
R/W
0x334
JESD_BIT_
INVERSE_CTRL
JESD_BIT_INVERSE
0x00
R/W
0x400
DID_REG
DID_RD
0x00
R
0x401
BID_REG
0x00
R
0x402
LID0_REG
RESERVED ADJDIR_RD PHADJ_RD
LID0_RD
0x00
R
0x403
SCR_L_REG
SCR_RD
L-1_RD
0x00
R
0x404
F_REG
0x00
R
0x405
K_REG
0x406
M_REG
SHORT_
TPL_FAIL
ADJCNT_RD
BID_RD
RESERVED
F-1_RD
RESERVED
K-1_RD
M-1_RD
R
0x00
R
CS_N_REG
0x408
NP_REG
0x409
S_REG
0x40A
HD_CF_REG
0x40B
RES1_REG
0x40C
RES2_REG
RES2_RD
0x00
R
0x40D
CHECKSUM_REG
FCHK0_RD
0x00
R
0x40E
COMPSUM0_REG
FCMP0_RD
0x00
R
0x412
LID1_REG
0x415
CHECKSUM1_REG
COMPSUM1_REG
0x41A
LID2_REG
0x41D
CHECKSUM2_REG
0x41E
COMPSUM2_REG
0x422
LID3_REG
0x425
CHECKSUM3_REG
HD_RD
RESERVED
0x00
0x407
0x416
CS_RD
SHORT_TPL_
TEST_RESET
N-1_RD
0x00
R
SUBCLASSV_RD
NP-1_RD
0x00
R
JESDV_RD
S-1_RD
0x00
R
CF_RD
0x00
R
0x00
R
RESERVED
RES1_RD
RESERVED
LID1_RD
FCHK1_RD
FCMP1_RD
RESERVED
LID2_RD
FCHK2_RD
FCMP2_RD
RESERVED
LID3_RD
FCHK3_RD
Rev. A | Page 86 of 117
0x00
R
0x00
R
0x00
R
0x00
R
0x00
R
0x00
R
0x00
R
0x00
R
Data Sheet
Reg.
Name
0x426
COMPSUM3_REG
0x42A
LID4_REG
AD9135/AD9136
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
FCMP3_RD
RESERVED
LID4_RD
Reset
R/W
0x00
R
0x00
R
0x42D
CHECKSUM4_REG
FCHK4_RD
0x00
R
0x42E
COMPSUM4_REG
FCMP4_RD
0x00
R
0x432
LID5_REG
0x00
R
0x435
CHECKSUM5_REG
FCHK5_RD
0x00
R
0x436
COMPSUM5_REG
FCMP5_RD
0x00
R
0x43A
LID6_REG
0x00
R
0x43D
CHECKSUM6_REG
FCHK6_RD
0x00
R
0x43E
COMPSUM6_REG
FCMP6_RD
0x00
R
0x442
LID7_REG
0x00
R
0x445
CHECKSUM7_REG
FCHK7_RD
0x00
R
0x446
COMPSUM7_REG
FCMP7_RD
0x00
R
0x450
ILS_DID
0x451
ILS_BID
0x452
ILS_LID0
RESERVED ADJDIR
0x453
ILS_SCR_L
SCR
0x454
ILS_F
0x455
ILS_K
0x456
ILS_M
0x457
ILS_CS_N
0x458
ILS_NP
0x459
ILS_S
0x45A
ILS_HD_CF
RESERVED
LID5_RD
RESERVED
LID6_RD
RESERVED
LID7_RD
DID
ADJCNT
BID
PHADJ
RESERVED
0x00
R/W
L-1
0x83
R/W
0x00
R/W
0x1F
R/W
0x01
R/W
0x0F
R/W
K-1
M-1
CS
HD
RESERVED
R/W
R/W
LID0
F-1
RESERVED
0x00
0x00
N-1
SUBCLASSV
NP-1
0x2F
R/W
JESDV
S-1
0x20
R/W
CF
0x80
R/W
RESERVED
0x45B
ILS_RES1
RES1
0x00
R/W
0x45C
ILS_RES2
RES2
0x00
R/W
FCHK0
0x45
R/W
READERRORCNTR
0x00
R
0x00
R/W
0x45D
ILS_CHECKSUM
0x46B
ERRCNTRMON_RB
0x46B
ERRCNTRMON
0x46C
LANEDESKEW
0x46D
BADDISPARITY_RB
0x46D
BADDISPARITY
0x46E
NIT_RB
0x46E
NIT_W
0x46F
UNEXPECTEDCONTROL_RB
0x46F
UNEXPECTEDCONTROL_W
0x470
CODEGRPSYNCFLG
RESERVED
LANESEL
RESERVED
CNTRSEL
LANEDESKEW
0x0F
R/W
BADDIS
0x00
R
0x00
R/W
0x00
R
0x00
R/W
0x00
R
0x00
R/W
CODEGRPSYNC
0x00
R/W
RST_IRQ_ DISABLE_ RST_ERR_
DIS
ERR_CNTR_ CNTR_DIS
DIS
RESERVED
RST_IRQ_ DISABLE_ RST_ERR_
NIT
ERR_CNTR_ CNTR_NIT
NIT
RESERVED
LANE_ADDR_DIS
NIT
LANE_ADDR_NIT
UCC
RST_IRQ_ DISABLE_ RST_ERR_
UCC
ERR_CNTR_ CNTR_UCC
UCC
RESERVED
LANE_ADDR_UCC
0x471
FRAMESYNCFLG
FRAMESYNC
0x00
R/W
0x472
GOODCHKSUMFLG
GOODCHECKSUM
0x00
R/W
0x473
INITLANESYNCFLG
INITIALLANESYNC
0x00
R/W
0x476
CTRLREG1
0x477
CTRLREG2
0x478
KVAL
F
ILAS_
MODE
RESERVED
THRESHOLD_
MASK_EN
KSYNC
Rev. A | Page 87 of 117
RESERVED
0x01
R/W
0x00
R/W
0x01
R/W
AD9135/AD9136
Bit 7
Data Sheet
Reg.
Name
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Reset
R/W
0x47A
IRQVECTOR_MASK BADDIS_
MASK
NIT_MASK
UCC_
MASK
RESERVED
INITIALLANESYNC_
MASK
BADCHECKSUM FRAMESYNC_
_MASK
MASK
0x00
CODEGRPSYNC_MASK
R/W
0x47A
IRQVECTOR_FLAG BADDIS_
FLAG
NIT_FLAG
UCC_FLAG
RESERVED
INITIALLANESYNC_
FLAG
BADCHECKSUM FRAMESYNC_
_FLAG
FLAG
CODEGRPSYNC_FLAG
0x47B
SYNCASSERTIONMASK
UCC_S
CMM
CMM_ENABLE
0x47C
ERRORTHRES
0x47D
LANEENABLE
0x47E
RAMP_ENA
0x520
DIG_TEST0
0x521
DC_TEST_VALUE0
0x522
DC_TEST_VALUE1
BADDIS_S NIT_S
Bit 1
Bit 0
RESERVED
ETH
LANE_ENA
0x00
R
0x008
R/W
0xFF
R/W
0x0F
R/W
ENA_RAMP_ 0x00
CHECK
R/W
RESERVED
0x1C
R/W
DC_TEST_VALUE[7:0]
0x00
R/W
DC_TEST_VALUE[15:8]
0x00
R/W
RESERVED
RESERVED
DC_TEST_
MODE
DEVICE CONFIGURATION REGISTER DESCRIPTIONS
Table 85. Device Configuration Register Descriptions
Address
0x000
Name
SPI_INTFCONFA
Bit No.
7
6
5
4
3
2
Bit Name
SOFTRESET_M
LSBFIRST_M
ADDRINC_M
SDOACTIVE_M
SDOACTIVE
ADDRINC
Settings
1
0
1
LSBFIRST
1
0
0
SOFTRESET
1
0x003
CHIPTYPE
[7:0]
CHIPTYPE
0x004
0x005
0x006
PRODIDL
PRODIDH
CHIPGRADE
[7:0]
[7:0]
[7:4]
PRODIDL
PRODIDH
PROD_GRADE
0x008
SPI_PAGEINDX
[3:0]
[7:2]
[1:0]
DEV_REVISION
RESERVED
DAC_PAGE
0b01
0b10
0b11
Description
Soft Reset (Mirror).
LSB First (Mirror).
Address Increment (Mirror).
SDO Active (Mirror).
SDO Active.
Address Increment. Controls whether
addresses are incremented or decremented
during multibyte data transfers.
Addresses are incremented during multibyte
data transfers
Addresses are decremented during
multibyte data transfers
LSB First. Controls whether input and output
data are oriented as LSB first or MSB first.
Shift LSB in first
Shift MSB in first
Soft Reset. Setting this bit initiates a reset.
This bit is autoclearing after the soft reset is
complete.
Assert soft reset
The product type is “High Speed DAC”, which
is represented by a code of 0x04.
Product Identification Low.
Product Identification High.
Product Grade.
AD9136
AD9135
Device Revision.
Reserved.
DAC Paging. Selects which DAC is accessed
and written to when changing digital features,
such as digital gain, dc offset. This paging
affects Register 0x013 to Register 0x014,
Register 0x034 to Register 0x03D,
Register 0x110 to Register 0x124, and
Register 0x135 to Register 0x14C.
Read and write DAC0
Read and write DAC1
Write both DACs; read DAC0
Rev. A | Page 88 of 117
Reset
0x0
0x0
0x0
0x0
0x0
0x0
Access
R
R
R
R
R/W
R/W
0x0
R/W
0x0
R/W
0x4
R
0x44
0x91
R
R
R
R
R
R
R
R/W
0x6
0x4
0x6
0x0
0x3
Data Sheet
AD9135/AD9136
Address
0x00A
Name
SCRATCH_PAD
Bit No.
[7:0]
Bit Name
SCRATCHPAD
Settings
0x011
PWRCNTRL0
7
PD_BG
6
PD_DAC_0
5
4
RESERVED
PD_DAC_1
3
2
RESERVED
PD_DACM
[1:0]
[7:2]
1
RESERVED
RESERVED
DAC1_MASK
0
DAC0_MASK
[7:6]
5
4
3
RESERVED
TX_PROTECT_OUT
RESERVED
SPI_PROTECT_
OUT
SPI_PROTECT
RESERVED
RESERVED
GROUP_DLY
1
1
1
1
0x012
TXENMASK
1
1
0x013
PWRCNTRL3
0x014
GROUP_DLY
0x01F
IRQEN_
STATUSMODE0
2
[1:0]
[7:4]
[3:0]
7
IRQEN_SMODE_
CALPASS
1
1
1
0
6
IRQEN_SMODE_
CALFAIL
1
0
5
IRQEN_SMODE_
DACPLLLOST
1
0
Description
This register does not affect any functions in
the part and can be used for testing SPI
communication with the part. Any value
written to this register will be read back to
reflect the change unless a reset or powercycle occurs.
Reference Power-Down. Powers down the
band gap reference for the entire chip.
Circuits will not be provided with bias
currents.
Power down reference
Powers Down DAC0. Powers down the
I-channel DAC.
Powers down DAC0
Reserved.
Powers Down DAC1. Powers down the
Q-channel DAC.
Powers down DAC1
Reserved.
Powers Down the DAC Master Bias. The
master bias cell provides currents and DAC
full-scale adjustments to the four DACs. With
the DAC master bias powered down, the
DACs are inoperative.
Powers down the DAC master bias
Reserved.
Reserved.
DAC1 TXEN1 Mask. Power down DAC1 on a
falling edge of TXEN1.
If TXEN1 is low, power down DAC1
DAC0 TXEN0 Mask. Power down DAC0 on a
falling edge of TXEN0.
If TXEN0 is low, power down DAC0
Reserved.
TX_PROTECT triggers PROTECT_OUTx.
Reserved.
SPI_PROTECT triggers PROTECT_OUTx.
Reset
0x00
Access
R/W
0x0
R/W
0x1
R/W
0x0
0x1
R
R/W
0x0
0x1
R
R/W
0x0
0x0
0x0
R
R
R/W
0x0
R/W
0x0
0x1
0x0
0x0
R
R/W
R
R/W
SPI_PROTECT
Reserved.
Reserved.
Group Delay Control. Delays the selected
DAC channel output per the paging register.
0 = minimum delay. 15 = maximum delay.
The range of the delay is −4 to +3.5 DAC
clock periods, and the resolution is 1/2 DAC
clock period.
Calibration Pass Detection Status Mode.
If CALPASS goes high, it latches and pulls IRQ
low
CALPASS shows current status
Calibration Fail Detection Status Mode.
If CALFAIL goes high, it latches and pulls IRQ
low
CALFAIL shows current status
DAC PLL Lost Detection Status Mode.
If DACPLLLOST goes high, it latches and pulls
IRQ low
DACPLLLOST shows current status
0x0
0x0
0x8
0x8
R/W
R
R
R/W
0x0
R/W
0x0
R/W
0x0
R/W
Rev. A | Page 89 of 117
AD9135/AD9136
Address
Name
Data Sheet
Bit No.
4
Bit Name
IRQEN_SMODE_
DACPLLLOCK
Settings
1
0
3
IRQEN_SMODE_
SERPLLLOST
1
0
2
IRQEN_SMODE_
SERPLLLOCK
1
0
1
IRQEN_SMODE_
LANEFIFOERR
1
0
0x020
IRQEN_
STATUSMODE1
0
[7:3]
RESERVED
RESERVED
2
IRQEN_SMODE_
PRBS1
1
0
1
0
RESERVED
IRQEN_SMODE_
PRBS0
0
IRQEN_
STATUSMODE2
7
IRQEN_SMODE_
PDPERR0
1
0
6
5
RESERVED
IRQEN_SMODE_
BLNKDONE0
1
0
4
3
RESERVED
IRQEN_SMODE_
SYNC_LOCK0
1
0
2
IRQEN_SMODE_
SYNC_ROTATE0
1
0
1
IRQEN_SMODE_
SYNC_WLIM0
1
0
0
IRQEN_SMODE_
SYNC_TRIP0
DAC1 PRBS Error Status Mode.
1
0
Reset
0x0
Access
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
0x0
R
R
0x0
R/W
0x0
0x0
R/W
R/W
0x0
R/W
0x0
0x0
R
R/W
0x0
0x0
R
R/W
0x0
R/W
0x0
R/W
0x0
R/W
If PRBS1 goes high, it latches and pulls IRQ
low
PRBS1 shows current status
DAC0 PRBS Error Status Mode.
1
0x021
Description
DAC PLL Lock Detection Status Mode.
If DACPLLLOCK goes high, it latches and
pulls IRQ low
DACPLLLOCK shows current status
SERDES PLL Lost Detection Status Mode.
If SERPLLLOST goes high, it latches and pulls
IRQ low
SERPLLLOST shows current status
SERDES PLL Lock Detection Status Mode.
If SERPLLLOCK goes high, it latches and pulls
IRQ low
SERPLLLOCK shows current status
Lane FIFO Error Detection Status Mode.
If LANEFIFOERR goes high, latches and pulls
IRQ low
LANEFIFOERR shows current status
Reserved.
Reserved.
If PRBS0 goes high, it latches and pulls IRQ
low
PRBS0 shows current status
DAC0 PDP Error.
If PDPERR0 goes high, it latches and pulls IRQ
low
PDPERR0 shows current status
Reserved.
DAC0 Blanking Done Status Mode.
If BLNKDONE0 goes high, it latches and pulls
IRQ low
BLNKDONE0 shows current status
Reserved.
DAC0 Alignment Locked Status Mode.
If SYNC_LOCK0 goes high, it latches and pulls
IRQ low
SYNC_LOCK0 shows current status
DAC0 Alignment Rotate Status Mode.
If SYNC_ROTATE0 goes high, it latches and
pulls IRQ low
SYNC_ROTATE0 shows current status
DAC0 Outside Window Status Mode.
If SYNC_WLIM0 goes high, it latches and
pulls IRQ low
SYNC_WLIM0 shows current status
DAC0 Alignment Tripped Status Mode.
If SYNC_TRIP0 goes high, it latches and pulls
IRQ low
SYNC_TRIP0 shows current status
Rev. A | Page 90 of 117
Data Sheet
Address
0x022
Name
IRQEN_
STATUSMODE3
AD9135/AD9136
Bit No.
7
Bit Name
IRQEN_SMODE_
PDPERR1
Settings
1
0
6
5
RESERVED
IRQEN_SMODE_
BLNKDONE1
1
0
4
3
RESERVED
IRQEN_SMODE_
SYNC_LOCK1
1
0
2
IRQEN_SMODE_
SYNC_ROTATE1
1
0
1
IRQEN_SMODE_
SYNC_WLIM1
1
0
0
IRQEN_SMODE_
SYNC_TRIP1
1
0
0x023
IRQ_STATUS0
7
CALPASS
6
CALFAIL
5
DACPLLLOST
4
DACPLLLOCK
3
SERPLLLOST
1
1
1
1
1
Description
DAC1 PDP Error.
If PDPERR1 goes high, it latches and pulls IRQ
low
PDPERR1 shows current status
Reserved.
DAC1 Blanking Done Status Mode.
If BLNKDONE1 goes high, it latches and pulls
IRQ low
BLNKDONE1 shows current status
Reserved.
DAC1 Alignment Locked Status Mode.
If SYNC_LOCK1 goes high, it latches and pulls
IRQ low
SYNC_LOCK1 shows current status
DAC1 Alignment Rotate Status Mode.
If SYNC_ROTATE1 goes high, it latches and
pulls IRQ low
SYNC_ROTATE1 shows current status
DAC1 Outside Window Status Mode.
If SYNC_WLIM1 goes high, it latches and
pulls IRQ low
SYNC_WLIM1 shows current status
DAC1 Alignment Tripped Status Mode.
If SYNC_TRIP1 goes high, it latches and pulls
IRQ low
SYNC_TRIP1 shows current status
Calibration Pass Status. If
IRQEN_SMODE_CALPASS is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
Calibration passed
Calibration Fail Detection Status. If
IRQEN_SMODE_CALFAIL is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
Calibration failed
DAC PLL Lost Status. If
IRQEN_SMODE_DACPLLLOST is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
DAC PLL lock was lost
DAC PLL Lock Status. If
IRQEN_SMODE_DACPLLLOCK is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
DAC PLL locked
SERDES PLL Lost Status. If
IRQEN_SMODE_SERPLLLOST is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
SERDES PLL lock was lost
Rev. A | Page 91 of 117
Reset
0x0
Access
R/W
0x0
0x0
R
R/W
0x0
0x0
R
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R
0x0
R
0x0
R
0x0
R
0x0
R
AD9135/AD9136
Address
Name
Data Sheet
Bit No.
2
Bit Name
SERPLLLOCK
1
LANEFIFOERR
0
[7:3]
2
RESERVED
RESERVED
PRBS1
1
0
RESERVED
PRBS0
7
PDPERR0
6
5
RESERVED
BLNKDONE0
4
3
RESERVED
SYNC_LOCK0
2
SYNC_ROTATE0
Settings
1
1
0x024
IRQ_STATUS1
1
1
0x025
IRQ_STATUS2
1
1
1
1
Description
SERDES PLL Lock Status. If
IRQEN_SMODE_SERPLLLOCK is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
SERDES PLL locked
Lane FIFO Error Status. If
IRQEN_SMODE_LANEFIFOERR is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low.
A lane FIFO error occurs when there is a full
or empty condition on any of the FIFOs
between the deserializer block and the core
digital. This error requires a link disable and
reenable to remove it. The status of the lane
FIFOs can be found in Register 0x30C (FIFO
full), and Register 0x30D (FIFO empty).
Lane FIFO error
Reserved.
Reserved.
DAC1 PRBS Error Status. If
IRQEN_SMODE_PRBS1 is low, this bit shows
current status. If not, this bit latches on a
rising edge and pull IRQ low. When latched,
write a 1 to clear this bit.
DAC1 failed PRBS
Reserved.
DAC0 PRBS Error Status. If
IRQEN_SMODE_PRBS0 is low, this bit shows
current status. If not, this bit latches on a
rising edge and pull IRQ low. When latched,
write a 1 to clear this bit.
DAC0 failed PRBS
DAC0 PDP Error. If IRQEN_SMODE_PAERR0 is
low, this bit shows current status. If not, this
bit latches on a rising edge and pull IRQ low.
When latched, write a 1 to clear this bit.
Data into DAC0 over power threshold
Reserved.
DAC0 Blanking Done Status. If
IRQEN_SMODE_BLNKDONE0 is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
DAC0 blanking done
Reserved
DAC0 LMFC Alignment Locked Status. If
IRQEN_SMODE_SYNC_LOCK0 is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
DAC0 LMFC alignment locked
DAC0 LMFC Alignment Rotate Status. If
IRQEN_SMODE_SYNC_ROTATE0 is low, this
bit shows current status. If not, this bit
latches on a rising edge and pull IRQ low.
When latched, write a 1 to clear this bit.
DAC0 LMFC alignment rotated
Rev. A | Page 92 of 117
Reset
0x0
Access
R
0x0
R
0x0
0x0
0x0
R
R
R
0x0
0x0
R
R
0x0
R
0x0
0x0
R
R
0x0
0x0
R
R
0x0
R
Data Sheet
Address
Name
AD9135/AD9136
Bit No.
1
Bit Name
SYNC_WLIM0
0
SYNC_TRIP0
7
PDPERR1
6
5
RESERVED
BLNKDONE1
4
3
RESERVED
SYNC_LOCK1
2
SYNC_ROTATE1
1
SYNC_WLIM1
0
SYNC_TRIP1
[7:6]
5
RESERVED
ERR_DLYOVER
4
ERR_WINLIMIT
Settings
1
1
0x026
IRQ_STATUS3
1
1
1
1
1
1
0x030
JESD_CHECKS
1
1
3
ERR_JESDBAD
1
2
ERR_KUNSUPP
1
Description
DAC0 Outside Window Status. If
IRQEN_SMODE_SYNC_WLIM0 is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
DAC0 LMFC phase outside of window
DAC0 LMFC Alignment Tripped Status. If
IRQEN_SMODE_SYNC_TRIP0 is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
DAC0 LMFC alignment tripped
DAC1 PDP Error. If IRQ_SMODE_PDPERR1 is
low, this bit shows current status. If not, this
bit latches on a rising edge and pull IRQ low.
When latched, write a 1 to clear this bit.
Data into DAC1 over power threshold
Reserved.
DAC1 Blanking Done Status. If
IRQEN_SMODE_BLNKDONE1 is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
DAC1 blanking done
Reserved.
DAC1 LMFC Alignment Locked Status. If
IRQEN_SMODE_SYNC_LOCK1 is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
DAC1 LMFC alignment locked
DAC1 LMFC Alignment Rotate Status. If
IRQEN_SMODE_SYNC_ROTATE1 is low, this
bit shows current status. If not, this bit
latches on a rising edge and pull IRQ low.
When latched, write a 1 to clear this bit.
DAC1 LMFC alignment rotated
DAC1 Outside Window Status. If
IRQEN_SMODE_SYNC_WLIM1 is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
DAC1 LMFC phase outside of window
DAC1 LMFC Alignment Tripped Status. If
IRQEN_SMODE_SYNC_TRIP1 is low, this bit
shows current status. If not, this bit latches
on a rising edge and pull IRQ low. When
latched, write a 1 to clear this bit.
DAC1 LMFC alignment tripped
Reserved.
Error: LMFC_Delay > JESD_K Parameter.
LMFC_Delay > JESD_K
Unsupported Window Limit.
Unsupported SYSREF window limit
Unsupported M/L/S/F Selection.
This JESD combination is not supported
Unsupported K Values. 16 and 32 are
supported.
K value unsupported
Rev. A | Page 93 of 117
Reset
0x0
Access
R
0x0
R
0x0
R
0x0
0x0
R
R
0x0
0x0
R
R
0x0
R
0x0
R
0x0
R
0x0
0x0
R
R
0x0
R
0x0
R
0x0
R
AD9135/AD9136
Address
Name
Data Sheet
Bit No.
1
Bit Name
ERR_SUBCLASS
0
ERR_INTSUPP
Settings
1
1
0x034
SYNC_ERRWINDOW
[7:2]
[1:0]
RESERVED
ERRWINDOW
0x038
SYNC_LASTERR_L
[7:4]
[3:0]
RESERVED
LASTERROR
0x039
SYNC_LASTERR_H
7
LASTUNDER
1
6
LASTOVER
1
0x03A
SYNC_CONTROL
[5:0]
7
RESERVED
SYNCENABLE
1
0
6
SYNCARM
1
5
SYNCCLRSTKY
4
SYNCCLRLAST
[3:0]
SYNCMODE
0b0001
0b0010
0b1000
0b1001
0x03B
SYNC_STATUS
7
SYNC_BUSY
1
[6:4]
3
RESERVED
SYNC_LOCK
2
SYNC_ROTATE
1
1
1
SYNC_WLIM
1
0
SYNC_TRIP
1
Description
Unsupported Subclass Value. 0 and 1 are
supported.
Unsupported subclass value
Unsupported Interpolation Rate Factor. 1, 2,
4, 8 are supported.
Unsupported interpolation rate factor
Reserved.
LMFC Sync Error Window. The error window
allows the SYSREF sample phase to vary
within the confines of the window without
triggering a clock adjustment. This is useful if
SYSREF cannot be guaranteed to always
arrive in the same period of the device clock
associated with the target phase.
Error window tolerance = ± ERRWINDOW
Reserved.
LMFC Sync Last Alignment Error. 4-bit twos
complement value that represents the phase
error (in number of DAC clock cycles) when the
clocks were last adjusted.
LMFC Sync Last Error Under Flag.
Last phase error was beyond lower window
tolerance boundary
LMFC Sync Last Error Over Flag.
Last phase error was beyond upper window
tolerance boundary
Reserved.
LMFC Sync Logic Enable.
Enable sync logic
Disable sync logic
LMFC Sync Arming Strobe.
Sync one-shot armed
LMFC Sync Sticky Bit Clear. On a rising edge,
this bit clears SYNC_ROTATE and SYNC_TRIP.
LMFC Sync Clear Last Error. On a rising edge,
this bit clears LASTERROR, LASTUNDER,
LASTOVER.
LMFC Sync Mode.
Sync one-shot mode
Sync continuous mode
Sync monitor only mode
Sync one-shot, then monitor
LMFC Sync Machine Busy.
Sync logic SM is busy
Reserved.
LMFC Sync Alignment Locked.
Sync logic aligned within window
LMFC Sync Rotated.
Sync logic rotated with SYSREF (sticky)
LMFC Sync Alignment Limit Range.
Phase error outside window threshold
LMFC Sync Tripped After Arming.
Sync received SYSREF pulse (sticky)
Rev. A | Page 94 of 117
Reset
0x0
Access
R
0x0
R
0x0
0x0
R
R/W
0x0
R
R
0x0
R
0x0
R
0x0
0x0
R
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R
0x0
0x0
R
R
0x0
R
0x0
R
0x0
R
Data Sheet
AD9135/AD9136
Address
0x03C
Name
SYNC_CURRERR_L
Bit No.
[7:4]
[3:0]
Bit Name
RESERVED
CURRERROR
0x03D
SYNC_CURRERR_H
7
CURRUNDER
Settings
1
6
CURROVER
1
[5:0]
[7:2]
[1:0]
RESERVED
RESERVED
DACFSC_0[9:8]
DACGAIN0_0
DACGAIN1_1
[7:0]
[7:2]
[1:0]
DACFSC_0[7:0]
RESERVED
DACFSC_1[9:8]
DACGAIN1_0
CLKCFG0
[7:0]
7
DACFSC_1[7:0]
PD_CLK0
6
PD_CLK1
5
PD_CLK_DIG
4
PD_SERDES_PCLK
3
PD_CLK_REC
[2:0]
RESERVED
0x040
DACGAIN0_1
0x041
0x044
0x045
0x080
Description
Reserved.
LMFC Sync Alignment Error. 4-bit twos
complement value that represents the phase
error in number of DAC clock cycles (that is,
number of DAC clocks between LMFC edge and
SYSREF edge).
When an adjustment of the clocks is made
on any given SYSREF, the value of the phase
error is placed into SYNC_ LASTERR, and
SYNC_CURRERR is forced to 0.
LMFC Sync Current Error Under Flag.
Current phase error is beyond lower window
tolerance boundary
LMFC Sync Current Error Over Flag.
Current phase error is beyond upper window
tolerance boundary
Reserved.
Reserved.
2 MSBs of I-Channel DAC Gain DAC0. A 10-bit
twos complement value that is mapped to
analog full-scale current for DAC0 as shown:
01111111111 = 27.0 mA
0000000000 = 20.48 mA
1000000000 = 13.9 mA
8 LSBs of I-Channel DAC Gain DAC0.
Reserved.
2 MSBs of Q-Channel DAC Gain DAC1. A
10-bit twos complement value that is
mapped to analog full-scale current for DAC
as shown in Register 0x040.
01111111111 = 27.0 mA
0000000000 = 20.48 mA
1000000000 = 13.9 mA
8 LSBs of Q-Channel DAC Gain DAC1.
Power-Down Clock for DAC0. This bit
disables the digital and analog clocks for
DAC0.
Power-Down Clock for DAC1. This bit
disables the digital and analog clocks for
DAC1.
Power-Down Clocks to all DACs. This bit
disables the digital and analog clocks for
both duals. This includes all reference clocks,
PCLK, DAC clocks, and digital clocks.
Serdes PLL Clock Power-Down. This bit
disables the reference clock to the SERDES
PLL, which is needed to have an operational
serial interface.
Clock Receiver Power-Down. This bit powers
down the analog DAC clock receiver block.
With this bit set, clocks are not passed to
internal nets.
Reserved.
Rev. A | Page 95 of 117
Reset
0x0
0x0
Access
R
R
0x0
R
0x0
R
0x0
0x0
0x0
R
R
R/W
0x0
0x0
0x0
R/W
R
R/W
0x0
0x1
R/W
R/W
0x1
R/W
0x1
R/W
0x1
R/W
0x1
R/W
0x0
R
AD9135/AD9136
Address
0x081
Name
SYSREF_ACTRL0
Data Sheet
Bit No.
[7:5]
4
Bit Name
RESERVED
PD_SYSREF
3
HYS_ON
2
SYSREF_RISE
Settings
0
1
0x082
0x083
0x084
SYSREF_ACTRL1
DACPLLCNTRL
DACPLLSTATUS
[1:0]
HYS_CNTRL1
[7:0]
7
HYS_CNTRL0
RECAL_DACPLL
[6:5]
4
RESERVED
ENABLE_DACPLL
[3:0]
7
RESERVED
DACPLL_
OVERRANGE_H
6
DACPLL_
OVERRANGE_L
5
DACPLL_CAL_
VALID
[4:2]
1
RESERVED
DACPLL_LOCK
0x085
DACINTEGERWORD0
0
[7:0]
RESERVED
B_COUNT
0x087
DACLOOPFILT1
[7:4]
LF_C2_WORD
[3:0]
LF_C1_WORD
[7:4]
LF_R1_WORD
[3:0]
LF_C3_WORD
0x088
DACLOOPFILT2
Description
Reserved.
Power-Down SYSREF Buffer. This bit powers
down the SYSREF receiver. For Subclass 1
operation to work, this buffer must be enabled.
Hysteresis Enabled. This bit enables the
programmable hysteresis control for the
SYSREF receiver. Using hysteresis gives some
noise resistance, but delays the SYSREF±
edge an amount depending on HYS_CNTRL
and the SYSREF± edge rate. The SYSREF±
KOW is not guaranteed when using
hysteresis.
Select DAC Clock Edge to Sample SYSREF.
Use falling edge of DAC clock to sample
SYSREF for alignment
Use rising edge of DAC clock to sample
SYSREF for alignment
Hysteresis Control Bits[9:8]. HYS_CNTRL is a
10-bit thermometer-coded number. Each bit
set adds 10 mV of differential hysteresis to
the SYSREF receiver.
Hysteresis Control Bits[7:0].
Recalibrate DAC PLL. On a rising edge of this bit,
recalibrate the DAC PLL.
Reserved.
Synthesizer Enable. This bit enables and
calibrates the DAC PLL.
Reserved.
DAC PLL High Overrange. This bit indicates
that the DAC PLL hit the upper edge of its
operating band. Recalibrate.
DAC PLL Low Overrange. This bit indicates
that the DAC PLL hit the lower edge of its
operating band. Recalibrate.
DAC PLL Calibration Valid. This bit indicates
that the DAC PLL has been successfully
calibrated.
Reserved.
DAC PLL Lock Bit. This bit is set high by the PLL
when it has achieved lock.
Reserved.
Integer Division Word. This bit controls the
integer feedback divider for the DAC PLL.
Determine the frequency of the DAC clock by
the following equations (see the Clock
Multiplication section for more details):
fDAC = fREF/(REF_DIVRATE) × 2 × B_COUNT
fVCO = fREF/(REF_DIVRATE) × 2 × B_COUNT ×
LO_DIV_MODE
Minimum value is 6.
C2 Control Word. Set this control to 0x6 for
optimal performance.
C1 Control Word. Set this control to 0x2 for
optimal performance.
R1 Control Word. Set this control to 0xC for
optimal performance.
C3 Control Word. Set this control to 0x9 for
optimal performance.
Rev. A | Page 96 of 117
Reset
0x0
0x1
Access
R
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
0x0
R/W
R/W
0x0
0x0
R
R/W
0x0
0x0
R
R
0x0
R
0x0
R
0x0
0x0
R
R
0x0
0x8
R
R/W
0x8
R/W
0x8
R/W
0x8
R/W
0x8
R/W
Data Sheet
Address
0x089
Name
DACLOOPFILT3
AD9135/AD9136
Bit No.
7
Bit Name
LF_BYPASS_R3
6
LF_BYPASS_R1
5
LF_BYPASS_C2
4
LF_BYPASS_C1
[3:0]
LF_R3_WORD
0x08A
DACCPCNTRL
[7:6]
[5:0]
RESERVED
CP_CURRENT
0x08B
DACLOGENCNTRL
[7:2]
[1:0]
RESERVED
LO_DIV_MODE
Settings
01
10
11
0x08C
DACLDOCNTRL1
[7:3]
[2:0]
RESERVED
REF_DIV_MODE
000
001
010
011
100
0x08D
DACLDOCNTRL2
[7:0]
DAC_LDO
0x0E2
CAL_CTRL_GLOBAL
[7:2]
1
RESERVED
CAL_START_AVG
0
CAL_EN_AVG
1
Description
Bypass R3 Resistor. When this bit is set,
bypass the R3 capacitor (set to 0 pF) when
R3_WORD is set to 0. Set this control to 0x0 for
optimal performance.
Bypass R1 Resistor. When this bit is set,
bypass the R1 capacitor (set to 0 pF) when
R1_WORD is set to 0. Set this control to 0x0 for
optimal performance.
Bypass C2 Capacitor. When this bit is set,
bypass the C2 capacitor (set to 0 pF) when
C2_WORD is set to 0. Set this control to 0x0 for
optimal performance.
Bypass C1 Capacitor. When this bit is set,
bypass the C1 capacitor (set to 0 pF) when
C1_WORD is set to 0. Set this control to 0x0 for
optimal performance.
R3 Control Word. Set this control to 0xE for
optimal performance.
Reserved.
Charge Pump Current Control. Set this control
to 0x12 for optimal performance.
Reserved.
This range controls the RF clock divider
between the VCO and DAC clock rates. The
options are 4×, 8×, or 16× division. Choose the
LO_DIV_MODE so that 6 GHz < fVCO < 12 GHz
(see the Clock Multiplication section for more
details):
DAC clock = VCO/4
DAC clock = VCO/8
DAC clock = VCO/16
Reserved.
Reference Clock Division Ratio. This field
controls the amount of division that is done
to the input clock at the CLK+/CLK− pins
before it is presented to the PLL as a reference
clock. The reference clock frequency must be
between 35 MHz and 80 MHz, but the
CLK+/CLK− input frequency can range from
35 MHz to 1 GHz. The user sets this division
to achieve a 35 MHz to 80 MHz PLL reference
frequency. For more details see the Clock
Multiplication section.
1
2
4
8
16
DAC PLL LDO setting. This register must be
written to 0x7B for optimal performance.
Reserved.
Averaged Calibration Start. On rising edge,
calibrate the DACs. Only use if calibrating all
DACs.
Averaged Calibration Enable. Set prior to
starting calibration with CAL_START_AVG.
While this bit is set, calibration can be
performed, and the results are applied.
Enable averaged calibration
Rev. A | Page 97 of 117
Reset
0x0
Access
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x8
R/W
0x0
0x20
R
R/W
0x0
0x2
R
R/W
0x0
0x1
R
R/W
0x2B
R/W
0x0
0x0
R
R/W
0x0
R/W
AD9135/AD9136
Address
0x0E7
Name
CAL_CLKDIV
Data Sheet
Bit No.
[7:4]
Bit Name
RESERVED
3
CAL_CLK_EN
Settings
1
0
0x0E8
CAL_PAGE
[2:0]
[7:4]
[3:0]
RESERVED
RESERVED
CAL_PAGE
0x0E9
CAL_CTRL
7
CAL_FIN
6
CAL_ACTIVE
5
CAL_ERRHI
4
CAL_ERRLO
[3:2]
1
RESERVED
CAL_START
1
1
1
1
0
1
0
CAL_EN
0
1
Description
Must write the default value for proper
operation.
Enable Self Calibration Clock.
Enable calibration clock
Disable calibration clock
Reserved.
Reserved.
DAC Calibration Paging. Selects which of the
DACs are being accessed for calibration or
calibration readback. This paging affects
Register 0x0E9 and Register 0x0ED.
Calibration: any number of DACs can be
accessed simultaneously to write and
calibrate. Write a 1 to Bit 0 to include DAC0.
Write a 1 to Bit 2 to include DAC1.
Readback: only one DAC at a time can be
accessed when reading back CAL_CTRL
(Register 0x0E9). Write a 1 to Bit 0 to read
from DAC0 or write a 1 to Bit 2 to read from
DAC1 (the other bits must be 0).
Calibration finished. This bit is high when the
calibration has completed. If the calibration
completes and either CAL_ERRHI or CAL_
ERRLO is high, then the calibration cannot be
considered valid and are considered a timeout
event.
Calibration ran and is finished
Calibration Active. This bit is high while the
calibration is in progress.
Calibration is running
SAR Data Error: Too High. This bit is set at the
end of a calibration cycle if any of the calibration DACs has overranged to the high side.
This typically means that the algorithm adjusts
the calibration preset of the calibration DACs
and runs another cycle.
Data saturated high
SAR Data Error: Too Low. This bit is set at the
end of a calibration cycle if any of the calibration DACs has overranged to the low side.
This typically means that the algorithm adjusts
the calibration preset of the calibration DACs
and runs another cycle.
Data saturated low
Reserved.
Calibration Start. The rising edge of this bit
kicks off a calibration sequence for the DACs
that have been selected in the CAL_INDX
register.
Normal operation
Start calibration state machine
Calibration Enable. Enable the calibration
DAC of the converter. Enable to calibration
engine and machines. Prepare for a calibration
start. For calibration coefficients to be applied
to the calibrated DACs, this bit must be high.
Do not use calibration DACs
Use calibration DACs
Rev. A | Page 98 of 117
Reset
0x3
Access
R/W
0x0
R/W
0x0
0x0
0xF
R
R
R/W
0x0
R
0x0
R
0x0
R
0x0
R
0x0
0x0
R
R/W
0x0
R/W
Data Sheet
AD9135/AD9136
Address
0x0ED
Name
CAL_INIT
Bit No.
[7:0]
Bit Name
CAL_INIT
Settings
0x110
DATA_FORMAT
7
BINARY_FORMAT
0
1
0x111
DATAPATH_CTRL
[6:0]
7
RESERVED
INVSINC_ENABLE
1
0
6
5
RESERVED
DIG_GAIN_ENABLE
1
0
0x112
INTERP_MODE
[4:0]
[7:3]
[2:0]
RESERVED
RESERVED
INTERP_MODE
000
001
011
100
0x11F
TXEN_SM_0
[7:6]
FALL_COUNTERS
[5:4]
RISE_COUNTERS
3
2
RESERVED
PROTECT_OUT_
INVERT
0
1
[1:0]
RESERVED
0x121
TXEN_RISE_COUNT_0
[7:0]
RISE_COUNT_0
0x122
TXEN_RISE_COUNT_1
[7:0]
RISE_COUNT_1
0x123
TXEN_FALL_
COUNT_0
[7:0]
FALL_COUNT_0
0x124
TXEN_FALL_
COUNT_1
[7:0]
FALL_COUNT_1
0x12D
DEVICE_CONFIG_
REG_0
DIE_TEMP_CTRL0
[7:0]
DEVICE_CONFIG_0
[7:1]
RESERVED
0
AUXADC_ENABLE
0x12F
0
1
Description
Initialize Calibration. Must be written to 0xA2
before starting calibration or averaged
calibration.
Binary or Twos Complementary Format on
the Data Bus.
Input data is twos complement
Input data is offset binary
Reserved.
Enable Inverse Sinc Filter.
Enable inverse sinc filter
Disable inverse sinc filter
Reserved.
Enable Digital Gain.
Enable digital gain function
Disable digital gain function
Reserved
Reserved.
Interpolation Mode.
1× mode
2× mode
4× mode
8× mode
Fall Counters. The number of counters to use
to delay TX_PROTECT fall from TXENx falling
edge. Must be set to 1 or 2.
Rise Counters. The number of counters to
use to delay TX_PROTECT rise from TXENx
rising edge.
Reserved.
PROTECT_OUTx Invert.
PROTECT_OUTx is high when output is valid.
Suitable for enabling downstream
components during transmission
PROTECT_OUTx is high when output is
invalid. Suitable for disabling downstream
components when not transmitting
Must write the default value for proper
operation.
First counter used to delay TX_PROTECT rise
from TXENx rising edge. Delays by 32 ×
RISE_COUNT_0 DAC clock cycles.
Second counter used to delay TX_PROTECT
rise from TXENx rising edge. Delays by 32 ×
RISE_COUNT_1 DAC clock cycles.
First counter used to delay TX_PROTECT fall
from TXENx falling edge. Delays by 32 ×
FALL_COUNT_0 DAC clock cycles. Must be
set to a minimum of 0x12.
Second counter used to delay TX_PROTECT
fall from TXENx falling edge. Delays by 32 ×
FALL_COUNT_1 DAC clock cycles.
Must be set to 0x8B for proper digital
datapath configuration.
Must write the default value for proper
operation.
Enables the AUX ADC Block.
AUX ADC disable
AUX ADC enable
Rev. A | Page 99 of 117
Reset
0xA6
Access
R/W
0x0
R/W
0x0
0x1
R
R/W
0x0
0x1
R
R/W
0x0
0x0
0x1
R
R
R/W
0x2
R/W
0x0
R/W
0x0
0x0
R
R/W
0x3
R/W
0xF
R/W
0x0
R/W
0xFF
R/W
0xFF
R/W
0x46
R/W
0x10
R/W
0x0
R/W
AD9135/AD9136
Data Sheet
Address
0x132
0x133
0x134
Name
DIE_TEMP0
DIE_TEMP1
DIE_TEMP_UPDATE
Bit No.
[7:0]
[7:0]
[7:1]
0
0x135
DC_OFFSET_CTRL
[7:1]
0
Bit Name
DIE_TEMP[7:0]
DIE_TEMP[15:8]
RESERVED
DIE_TEMP_
UPDATE
RESERVED
DC_OFFSET_ON
0x136
DAC_DC_OFFSET_1
PART0
[7:0]
LSB_OFFSET[7:0]
0x137
DAC_DC_OFFSET_
1PART1
[7:0]
LSB_OFFSET[15:8]
0x13A
DAC_DC_OFFSET_
2PART
[7:5]
[4:0]
RESERVED
SIXTEENTH_
OFFSET
0x13C
DAC_DIG_GAIN0
[7:0]
DAC_DIG_
GAIN[7:0]
0x13D
DAC_DIG_GAIN1
[7:4]
[3:0]
0x140
GAIN_RAMP_UP_
STEP0
[7:0]
RESERVED
DAC_DIG_
GAIN[11:8]
GAIN_RAMP_UP_
STEP[7:0]
Settings
1
x
0x0
0xFFF
0x141
0x142
GAIN_RAMP_UP_
STEP1
GAIN_RAMP_DOWN_
STEP0
[7:4]
RESERVED
[3:0]
GAIN_RAMP_UP_
STEP[11:8]
GAIN_RAMP_
DOWN_STEP[7:0]
[7:0]
0
0xFFF
0x143
0x146
0x147
GAIN_RAMP_
DOWN_STEP1
DEVICE_CONFIG_
REG_1
BSM_STAT
[7:4]
RESERVED
[3:0]
[7:0]
GAIN_RAMP_
DOWN_STEP[11:8]
DEVICE_CONFIG_1
[7:6]
SOFTBLANKRB
00
01
10
11
[5:0]
RESERVED
Description
Aux ADC Readback Value.
Aux ADC Readback Value.
Reserved.
Die Temperature Update. On a rising edge, a
new temperature code is generated.
Reserved.
DC Offset On.
Enables dc offset module
8 LSBs of DC Offset. LSB_OFFSET is a 16-bit
twos complement number that is added to
incoming data. Applies to the DAC selected
by DAC_PAGE (Register 0x008 [1:0]).
8 MSBs of DC Offset. LSB_OFFSET is a 16-bit
twos complement number that is added to
incoming data. Applies to the DAC selected
by DAC_PAGE (Register 0x008 [1:0]).
Reserved.
SIXTEENTH_OFFSET is a 5-bit twos
complement number in 16ths of an LSB that
is added to incoming I data.
x/16 LSB DC offset
8 LSBs of DAC Digital Gain. DAC_DIG_GAIN is
the digital gain of the DAC selected by
DAC_PAGE (Register 0x008 [1:0]). The digital
gain is a multiplier from 0 to 4095/2048 in
steps of 1/2048.
Reserved.
4 MSBs of DAC Digital Gain
8 LSBs of Gain Ramp Up Step.
GAIN_RAMP_UP_STEP controls the amplitude
step size of the BSM’s ramping feature when
the gain is being ramped to its assigned value.
Smallest ramp up step size
Largest ramp up step size
Reserved.
4 MSBs of Gain Ramp Up Step. See Register
0x140 for description.
8 LSBs of Gain Ramp Down Step.
GAIN_RAMP_DOWN_STEP controls the
amplitude step size of the BSM’s ramping
feature when the gain is being ramped to zero.
Smallest ramp down step size
Largest ramp down step size
Reserved.
4 MSBs of Gain Ramp Down Step. See
Register 0x142 for description.
Must be set to 0x01 for proper digital
datapath configuration.
Blanking State.
Data is fully blanked
Ramping from data process to full blanking
Ramping from fully blanked to data process
Data is being processed
Reserved.
Rev. A | Page 100 of 117
Reset
0x0
0x0
0x0
0x0
Access
R
R
R
R/W
0x0
0x0
R
R/W
0x0
R/W
0x0
R/W
0x0
0x0
R
R/W
0xEA
R/W
0x0
0xA
R
R/W
0x4
R/W
0x0
R
0x0
R/W
0x9
R/W
0x0
R
0x0
R/W
0x0
R/W
0x0
R
0x0
R
Data Sheet
Address
0x14B
Name
PRBS
AD9135/AD9136
Bit No.
7
6
Bit Name
RESERVED
PRBS_GOOD
Settings
0
1
[5:3]
2
RESERVED
PRBS_MODE
0
1
1
PRBS_RESET
0
1
0
PRBS_EN
0
1
0x14C
0x1B0
PRBS_ERROR
DACPLLT0
[7:0]
[7:0]
PRBS_COUNT
DAC_PLL_PWR
0x1B5
DACPLLT5
[7:4]
RESERVED
[3:0]
VCO_VAR
0x1B9
DACPLLT9
[7:0]
DAC_PLL_CP1
0x1BB
DACPLLTB
[7:5]
[4:3]
RESERVED
VCO_BIAS_TCF
[2:0]
VCO_BIAS_REF
0x1BC
DACPLLTC
[7:0]
DAC_PLL_VCO_
CTRL
0x1BE
DACPLLTE
[7:0]
DAC_PLL_VCO_
PWR
0x1BF
DACPLLTF
[7:0]
DAC_PLL_VCOCAL
0x1C0
DACPLLT10
[7:0]
0x1C1
DACPLLT11
[7:0]
DAC_PLL_LOCK_
CNTR
DAC_PLL_CP2
0x1C4
DACPLLT17
[7:0]
DAC_PLL_VAR1
0x1C5
DACPLLT18
[7:0]
DAC_PLL_VAR2
0x200
MASTER_PD
[7:1]
0
RESERVED
SPI_PD_MASTER
0x201
PHY_PD
[7:0]
SPI_PD_PHY
Description
Reserved.
Good Data Indicator.
Incorrect sequence detected
Correct PRBS sequence detected
Reserved.
Polynomial Select
7-bit: x7 + x6 + 1
15-bit: x15 + x14 + 1
Reset Error Counters.
Normal operation
Reset counters
Enable PRBS Checker.
Disable
Enable
Error Count Value.
DAC PLL PD settings. This register must be
written to 0x00 for optimal performance.
Must write the default value for proper
operation.
Varactor KVO Setting. See Table 73 for
optimal settings based on the fVCO being
used.
DAC PLL Charge Pump settings. This register
must be written to 0x24 for optimal
performance.
Reserved.
Temperature Coefficient for VCO Bias. See
Table 73 for optimal settings based on the
fVCO being used.
VCO Bias Control. See Table 73 for optimal
settings based on the fVCO being used.
DAC PLL VCO control settings. This register
must be written to 0x0D for optimal
performance.
DAC PLL VCO power control settings. This
register must be written to 0x02 for optimal
performance.
DAC PLL VCO calibration settings. This
register must be written to 0x8E for optimal
performance.
This register must be written to 0x2A for
optimal performance.
This register must be written to0x2A for
optimal performance.
DAC PLL Varactor setting. Must be set to
0x7E for proper DAC PLL configuration.
DAC PLL Varactor setting. See Table 73 for
optimal settings based on the fVCO being
used.
Reserved.
Power Down the Entire JESD Receiver Analog
(All Eight Channels Plus Bias).
SPI Override to Power Down the Individual
PHYs.
Set Bit x to power down the corresponding
SERDINx± PHY
Rev. A | Page 101 of 117
Reset
0x0
0x0
Access
R
R
0x0
0x0
R
R/W
0x0
R/W
0x0
R/W
0x0
0xFA
R
R/W
0x8
R/W
0x3
R/W
0x34
R/W
0x0
0x1
R
R/W
0x4
R/W
0x00
R/W
0x00
R/W
0x8D
R/W
0x2E
R/W
0x24
R/W
0x33
R/W
0x08
R/W
0x0
0x1
R
R/W
0x0
R/W
AD9135/AD9136
Address
0x203
Name
GENERIC_PD
0x206
CDR_RESET
Data Sheet
Bit No.
[7:2]
1
Bit Name
RESERVED
SPI_SYNC1_PD
Settings
Description
Reserved.
Power down LVDS buffer for SYNCOUT0±.
0x232
0x268
CDR_OPERATING_
MODE_REG_0
DEVICE_CONFIG_
REG_3
EQ_BIAS_REG
SPI_SYNC2_PD
Power down LVDS buffer for SYNCOUT1±.
0x0
R/W
[7:1]
0
RESERVED
SPI_CDR_RESETN
Reserved.
Resets the Digital Control Logic for All PHYs.
Hold CDR in reset
Enable CDR
Reserved.
Enables Half-Rate CDR Operation. Set to 1
when 5.75 Gbps ≤ lane rate ≤ 10.64.
Must write the default value for proper
operation.
Enables Oversampling of the Input Data. Set
to 1 when 1.44 Gbps ≤ lane rate ≤ 2.76 Gbps.
Reserved.
Must be set to 0xFF for proper JESD interface
configuration.
Control the Equalizer Power/Insertion Loss
Capability.
Normal mode
Low power mode
Must write the default value for proper
operation.
Reserved.
Recalibrate SERDES PLL. On a rising edge,
recalibrate the SERDES PLL.
Reserved.
Enable the SERDES PLL. Setting this bit
enables and calibrates the SERDES PLL.
Reserved.
SERDES PLL High Overrange. This bit
indicates that the SERDES PLL hit the lower
edge of its operating band. Recalibrate.
SERDES PLL Low Overrange. This bit
indicates that the SERDES PLL hit the lower
edge of its operating band. Recalibrate.
SERDES PLL Calibration Valid. This bit
indicates that the SERDES PLL has been
successfully calibrated.
Reserved.
SERDES PLL Lock. This bit is set high by the PLL
when it has achieved lock.
SERDES PLL loop filter setting. This register
must be written to 0x62 for optimal
performance.
SERDES PLL loop filter setting. This register
must be written to 0xC9 for optimal
performance.
SERDES PLL loop filter setting. This register
must be written to 0x0E for optimal
performance.
SERDES PLL charge pump setting. This
register must be written to 0x12 for optimal
performance.
0x0
0x1
R
R/W
0x0
0x1
R
R/W
0x2
R/W
0x0
R/W
0x0
0x0
R
R/W
0x1
R/W
0x22
R/W
0x0
0x0
R
R/W
0x0
0x0
R
R/W
0x0
0x0
R
R
0x0
R
0x0
R
0x0
0x0
R
R
0x77
R/W
0x87
R/W
0x08
R/W
0x3F
R/W
[7:6]
5
RESERVED
ENHALFRATE
[4:2]
RESERVED
1
CDR_OVERSAMP
0
[7:0]
RESERVED
DEVICE_CONFIG_3
[7:6]
EQ_POWER_
MODE
00
01
0x280
0x281
SERDESPLL_
ENABLE_CNTRL
PLL_STATUS
Access
R
R/W
0
0
1
0x230
Reset
0x0
0x0
[5:0]
RESERVED
[7:3]
2
RESERVED
RECAL_SERDESPLL
1
0
RESERVED
ENABLE_
SERDESPLL
RESERVED
SERDES_PLL_
OVERRANGE_H
[7:6]
5
4
SERDES_PLL_
OVERRANGE_L
3
SERDES_PLL_CAL_
VALID_RB
[2:1]
0
0x284
LOOP_FILTER_1
[7:0]
RESERVED
SERDES_PLL_
LOCK_RB
LOOP_FILTER_1
0x285
LOOP_FILTER_2
[7:0]
LOOP_FILTER_2
0x286
LOOP_FILTER_3
[7:0]
LOOP_FILTER_3
0x287
SERDES_PLL_CP1
[7:0]
SERDES_PLL_CP1
Rev. A | Page 102 of 117
Data Sheet
Address
0x289
Name
REF_CLK_DIVIDER_
LDO
AD9135/AD9136
Bit No.
[7:3]
2
Bit Name
RESERVED
DEVICE_CONFIG_4
[1:0]
SERDES_PLL_DIV_
MODE
Settings
00
01
10
0x28A
VCO_LDO
[7:0]
SERDES_PLL_
VCO_LDO
0x28B
SERDES_PLL_PD1
[7:0]
SERDES_PLL_PD1
0x290
SERDESPLL_VAR1
[7:0]
SERDES_PLL_VAR1
0x294
SERDES_PLL_CP2
[7:0]
SERDES_PLL_CP2
0x296
SERDESPLL_VCO1
[7:0]
0x297
SERDESPLL_VCO2
[7:0]
0x299
SERDES_PLL_PD2
[7:0]
SERDES_PLL_
VCO1
SERDES_PLL_
VCO2
SERDES_PLL_PD2
0x29A
SERDESPLL_VAR2
[7:0]
SERDES_PLL_VAR2
0x29C
SERDES_PLL_CP3
[7:0]
SERDES_PLL_CP3
0x29F
SERDESPLL_VAR3
[7:0]
SERDES_PLL_VAR3
0x2A0
SERDESPLL_VAR4
[7:0]
SERDES_PLL_VAR4
0x2A4
DEVICE_CONFIG_
REG_8
SYNCOUTB_SWING
[7:0]
DEVICE_CONFIG_8
[7:1]
0
RESERVED
SYNCOUTB_
SWING_MD
0x2A5
0
1
0x2A7
0x2AA
0x2AB
TERM_BLK1_
CTRLREG0
DEVICE_CONFIG_
REG_9
DEVICE_CONFIG_
REG_10
[7:1]
RESERVED
0
RCAL_TERMBLK1
[7:0]
DEVICE_CONFIG_
9
DEVICE_CONFIG_
10
[7:0]
Description
Reserved.
Must be set to 1 for proper SERDES PLL
configuration.
SERDES PLL Reference Clock Division Factor.
This field controls the division of the SERDES
PLL reference clock before it is fed into the
SERDES PLL Phase Frequency Detector (PFD).
It must be set so fREF/DivFactor is between
35 MHz and 80 MHz.
Divide by 4 for 5.75 Gbps to 10.64 Gbps lane
rate
Divide by 2 for 2.88 Gbps to 5.52 Gbps lane
rate
Divide by 1 for 1.44 Gbps to 2.76 Gbps lane rate
SERDES PLL VCO LDO setting. This register
must be written to 0x7B for optimal
performance.
SERDES PLL PD setting. This register must be
written to 0x00 for optimal performance.
SERDES PLL Varactor setting. This register
must be written to 0x89 for optimal
performance.
SERDES PLL Charge Pump setting. This
register must be set to 0x24 for optimal
performance.
SERDES PLL VCO setting. This register must
be set to 0x03 for optimal performance.
SERDES PLL VCO setting. This register must
be set to 0x0D for optimal performance.
SERDES PLL PD setting. This register must be
set to 0x02 for optimal performance.
SERDES PLL Varactor setting. This register
must be set to 0x8E for optimal performance.
SERDES PLL Charge Pump setting. Must be
set to 0x2A for proper SERDES PLL
configuration.
SERDES PLL Varactor setting. Must be set to
0x78 for proper SERDES PLL configuration.
SERDES PLL Varactor setting. This register
must be set to 0x06 for optimal performance.
Must be set to 0xFF for proper clock
configuration.
Reserved.
SYNCOUTx± Swing Mode. Sets the output
differential swing mode for the SYNCOUTx±
pins. See Table 8 for details.
Normal Swing Mode
High Swing Mode
Reserved.
Reset
0x0
0x0
Access
R
R/W
0x0
R/W
0x2B
R/W
0x7F
R/W
0x83
R/W
0xB0
R/W
0x0C
R/W
0x00
R/W
0x00
R/W
0xFE
R/W
0x17
R/W
0x33
R/W
0x08
R/W
0x4B
R/W
0x0
0x0
R
R/W
0x0
R
Termination Calibration. The rising edge of
this bit calibrates PHY0, PHY1, PHY6, and PHY7
terminations to 50 Ω.
Must be set to 0xB7 for proper JESD interface
termination configuration.
Must be set to 0x87 for proper JESD interface
termination configuration.
0x0
R/W
0xC3
R/W
0x93
R/W
Rev. A | Page 103 of 117
AD9135/AD9136
Data Sheet
Address
0x2AE
Name
TERM_BLK2_
CTRLREG0
Bit No.
[7:1]
0
Bit Name
RESERVED
RCAL_TERMBLK2
0x2B1
DEVICE_CONFIG_
REG_11
DEVICE_CONFIG_
REG_12
GENERAL_JRX_
CTRL_0
[7:0]
DEVICE_CONFIG_
11
DEVICE_CONFIG_
12
RESERVED
CHECKSUM_MODE
0x2B2
0x300
[7:0]
7
6
Settings
0
1
[5:4]
3
RESERVED
LINK_MODE
0
1
2
LINK_PAGE
0
1
[1:0]
LINK_EN
0b00
0b01
0b10
0b11
0x301
GENERAL_JRX_CTRL_1
[7:3]
[2:0]
RESERVED
SUBCLASSV_
LOCAL
000
001
0x302
DYN_LINK_LATENCY_0
[7:5]
[4:0]
RESERVED
DYN_LINK_
LATENCY_0
0x303
DYN_LINK_LATENCY_1
[7:5]
[4:0]
RESERVED
DYN_LINK_
LATENCY_1
Description
Reserved.
Terminal Calibration. The rising edge of this
bit calibrates PHY2, PHY3, PHY4 and PHY5
terminations to 50 Ω.
Must be set to 0xB7 for proper JESD interface
termination configuration.
Must be set to 0x87 for proper JESD interface
termination configuration.
Reserved.
Checksum Mode. This bit controls the locally
generated JESD204B link parameter checksum
method. The value is stored in the FCMP
registers (Register 0x40E, Register 0x416,
Register 0x41E, Register 0x426, Register
0x42E, Register 0x436, Register 0x43E, and
Register 0x446).
Checksum is calculated by summing the
individual fields in the link configuration
table as defined in Section 8.3, Table 20 of
the JESD204B standard
Checksum is calculated by summing the registers containing the packed link configuration
fields (Σ[0x400:0x40A] modulo 256).
Reserved.
Link Mode. This register selects either singlelink or dual-link mode.
Single-link mode
Dual-link mode
Link Paging. Selects which link’s register map
is used. This paging affects Registers 0x401
to 0x47E.
Use Link 0 register map
Use Link 1 register map
Link Enable. These bits bring up the JESD204B
receiver digital circuitry: Bit 0 for Link 0 and
Bit 1 for Link 1. Enable the link only after the
following has occurred: all JESD204B parameters are set, the DAC PLL is enabled and
locked (Register 0x084[1] = 1), and the
JESD204B PHY is enabled (Register 0x200 =
0x00) and calibrated (Register 0x281[2] = 0).
Disable both JESD Link 1 and JESD Link 0
Disable JESD Link 1, enable JESD Link 0
Enable JESD Link 1, disable JESD Link 0
Enable both JESD Link 1 and JESD Link 0
Reserved.
JESD204B Subclass.
Subclass 0
Subclass 1
Reserved.
Dynamic Link Latency: Link 0. Latency
between the LMFCRx for Link 0 and the last
arriving LMFC boundary in units of PCLK
cycles. See the Deterministic Latency section.
Reserved.
Dynamic Link Latency: Link 1. Latency
between the LMFCRx for Link 1 and the last
arriving LMFC boundary in units of PCLK
cycles. See the Deterministic Latency section.
Rev. A | Page 104 of 117
Reset
0x0
0x0
Access
R
R/W
0xC3
R/W
0x93
R/W
0x0
0x0
R
R/W
0x0
0x0
R
R/W
0x0
R/W
0x0
R/W
0x0
0x1
R
R/W
0x0
0x0
R
R
0x0
0x0
R
R
Data Sheet
AD9135/AD9136
Address
0x304
Name
LMFC_DELAY_0
Bit No.
[7:5]
[4:0]
Bit Name
RESERVED
LMFC_DELAY_0
Settings
0x305
LMFC_DELAY_1
[7:5]
[4:0]
RESERVED
LMFC_DELAY_1
0x306
LMFC_VAR_0
[7:5]
[4:0]
RESERVED
LMFC_VAR_0
0x307
LMFC_VAR_1
[7:5]
[4:0]
RESERVED
LMFC_VAR_1
0x308
XBAR_LN_0_1
[7:6]
[5:3]
RESERVED
LOGICAL_LANE1_
SRC
[2:0]
LOGICAL_LANE0_
SRC
[7:6]
[5:3]
RESERVED
LOGICAL_LANE3_
SRC
[2:0]
LOGICAL_LANE2_
SRC
[7:6]
[5:3]
RESERVED
LOGICAL_LANE5_
SRC
[2:0]
LOGICAL_LANE4_
SRC
[7:6]
[5:3]
RESERVED
LOGICAL_LANE7_
SRC
[2:0]
LOGICAL_LANE6_
SRC
[7:0]
LANE_FIFO_FULL
x
x
0x309
XBAR_LN_2_3
x
x
0x30A
XBAR_LN_4_5
x
x
0x30B
XBAR_LN_6_7
x
x
0x30C
FIFO_STATUS_REG_0
Description
Reserved.
LMFC Delay: Link 0 Delay from the LMFC to
LMFCRx for Link 0. In units of frame clock
cycles for subclass 1 and PCLK cycles for
subclass 0. See the Deterministic Latency
section.
Reserved.
LMFC Delay: Link 1. Delay from the LMFC to
LMFCRx for Link 1. In units of frame clock
cycles for subclass 1 and PCLK cycles for
subclass 0. See the Deterministic Latency
section.
Reserved.
Variable Delay Buffer: Link 0. Sets when data is
read from a buffer to be consistent across links
and power cycles. In units of PCLK cycles. See
the Deterministic Latency section.
This setting must not be more than 10.
Reserved.
Variable Delay Buffer: Link 1. Sets when data is
read from a buffer to be consistent across links
and power cycles. In units of PCLK cycles. See
the Deterministic Latency section.
This setting must not be more than 10.
Reserved.
Logical Lane 1 Source. Selects a physical lane
to be mapped onto Logical Lane 1.
Data is from SERDINx
Logical Lane 0 Source. Selects a physical lane
to be mapped onto Logical Lane 0.
Data is from SERDINx
Reserved.
Logical Lane 3 Source. Selects a physical lane
to be mapped onto Logical Lane 3.
Data is from SERDINx
Logical Lane 2 source. Selects a physical lane
to be mapped onto Logical Lane 2.
Data is from SERDINx
Reserved.
Logical Lane 5 Source. Selects a physical lane
to be mapped onto Logical Lane 5.
Data is from SERDINx
Logical Lane 4 Source. Selects a physical lane
to be mapped onto Logical Lane 4.
Data is from SERDINx
Reserved.
Logical Lane 7 Source. Selects a physical lane
to be mapped onto Logical Lane 7.
Data is from SERDINx
Logical Lane 6 Source. Selects a physical lane
to be mapped onto Logical Lane 6.
Data is from SERDINx
FIFO Full Flags for Each Logical Lane. A full
FIFO indicates an error in the JESD204B
configuration or with a system clock.
If the FIFO for Lane x is full, Bit x in this
register will be high.
Rev. A | Page 105 of 117
Reset
0x0
0x0
Access
R
R/W
0x0
0x0
R
R/W
0x0
0x6
R
R/W
0x0
0x6
R
R/W
0x0
0x1
R
R/W
0x0
R/W
0x0
0x3
R
R/W
0x2
R/W
0x0
0x5
R
R/W
0x4
R/W
0x0
0x7
R
R/W
0x6
R/W
0x0
R
AD9135/AD9136
Data Sheet
Address
0x30D
Name
FIFO_STATUS_REG_1
Bit No.
[7:0]
Bit Name
LANE_FIFO_EMPTY
0x312
SYNCB_GEN_1
[7:6]
[5:4]
RESERVED
SYNCB_ERR_DUR
Settings
0
1
2
0x314
SERDES_SPI_REG
[3:0]
[7:0]
0x315
PHY_PRBS_TEST_EN
[7:0]
RESERVED
SERDES_SPI_
CONFIG
PHY_TEST_EN
0x316
PHY_PRBS_TEST_CTRL
7
[6:4]
RESERVED
PHY_SRC_ERR_CNT
[3:2]
PHY_PRBS_PAT_SEL
x
00
01
10
1
PHY_TEST_START
0
1
0
PHY_TEST_RESET
0
1
0x317
0x318
0x319
0x31A
0x31B
0x31C
0x31D
PHY_PRBS_TEST_
THRESHOLD_LOBITS
PHY_PRBS_TEST_
THRESHOLD_
MIDBITS
PHY_PRBS_TEST_
THRESHOLD_HIBITS
PHY_PRBS_TEST_
ERRCNT_LOBITS
[7:0]
PHY_PRBS_TEST_
ERRCNT_MIDBITS
PHY_PRBS_TEST_
ERRCNT_HIBITS
PHY_PRBS_TEST_
STATUS
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
PHY_PRBS_
THRESHOLD[7:0]
PHY_PRBS_
THRESHOLD[15:8]
PHY_PRBS_
THRESHOLD[23:16]
PHY_PRBS_ERR_
CNT[7:0]
PHY_PRBS_ERR_
CNT[15:8]
PHY_PRBS_ERR_
CNT[23:16]
PHY_PRBS_PASS
Description
FIFO Empty Flags for Each Logical Lane. An
empty FIFO indicates an error in the JESD204B
configuration or with a system clock.
If the FIFO for Logical Lane x is empty, Bit x in
this register will be high.
Reserved.
Duration of SYNCOUTx± Low for Error. The
duration applies to both SYNCOUT0 and
SYNCOUT1. A sync error is asserted at the
end of a multiframe whenever one or more
disparity, not in table or unexpected control
character errors are encountered.
½ PCLK cycle
1 PCLK cycle
2 PCLK cycles
Reserved.
SERDES SPI Configuration. Must be written to
0x01 as part of the Physical Layer setup step.
PHY Test Enable. Enables the PHY BER test.
Set Bit x to enable the PHY test for Lane x.
Reserved.
PHY Error Count Source. Selects which PHY
errors are being reported in Register 0x31A
to Register 0x31C.
Report Lane x error count
PHY PRBS Pattern Select. Selects the PRBS
pattern for PHY BER test.
PRBS7
PRBS15
PRBS31
PHY PRBS Test Start. Starts and stops the PHY
PRBS test.
Test stopped
Test in progress
PHY PRBS Test Reset. Resets the PHY PRBS
test state machine and error counters.
Enable PHY PRBS test state machine
Hold PHY PRBS test state machine in reset
8 LSBs of PHY PRBS Error Threshold.
Reset
0x0
Access
R
0x0
R/W
0x0
0x0
R/W
R/W
0x0
R/W
0x0
0x0
R
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R/W
8 ISBs of PHY PRBS Error Threshold.
0x0
R/W
8 MSBs of PHY PRBS Error Threshold.
0x0
R/W
8 LSBs of PHY PRBS Error Count.
Reported PHY BERT error count from lane
selected using Register 0x316[6:4].
8 ISBs of PHY PRBS Error Count.
0x0
R
0x0
R
8 MSBs of PHY PRBS Error Count.
0x0
R
PHY PRBS Test Pass/Fail.
Bit x corresponds to PHY PRBS pass/fail for
Physical Lane x.
The bit is set to 1 while the error count for
Physical Lane x is less than
PHY_PRBS_THRESHOLD.
0xFF
R
Rev. A | Page 106 of 117
Data Sheet
Address
0x32C
Name
SHORT_TPL_TEST_0
AD9135/AD9136
Bit No.
[7:6]
[5:4]
Bit Name
RESERVED
SHORT_TPL_SP_
SEL
[3:2]
SHORT_TPL_DAC_
SEL
Settings
x
0
2
1
SHORT_TPL_TEST_
RESET
0
1
0
SHORT_TPL_TEST_
EN
0
1
0x32D
SHORT_TPL_TEST_1
[7:0]
SHORT_TPL_REF_
SP_LSB
0x32E
SHORT_TPL_TEST_2
[7:0]
SHORT_TPL_REF_
SP_MSB
0x32F
SHORT_TPL_TEST_3
[7:1]
0
RESERVED
SHORT_TPL_FAIL
0
1
0x333
[7:0]
DEVICE_CONFIG_
13
JESD_BIT_INVERSE
0x400
DEVICE_CONFIG_
REG_13
JESD_BIT_INVERSE_
CTRL
DID_REG
[7:0]
DID_RD
0x401
BID_REG
[7:4]
ADJCNT_RD
[3:0]
BID_RD
7
6
RESERVED
ADJDIR_RD
5
PHADJ_RD
[4:0]
LID0_RD
0x334
0x402
LID0_REG
[7:0]
Description
Reserved.
Short Transport Layer Sample Select. Selects
which sample to check from the DAC
selected via Bits[3:2].
Sample x
Short Transport Layer Test DAC Select.
Selects which DAC to sample.
Sample from DAC0
Sample from DAC1
Short Transport Layer Test Reset. Resets the
result of short transport layer test.
Not reset
Reset
Short Transport Layer Test Enable. See the
Subclass 0 section for details on how to
perform this test.
Disable
Enable
Short Transport Layer Test Reference, Sample
LSB. This is the lower eight bits of the
expected DAC sample. It is used to compare
with the received DAC sample at the output
of the JESD204B receiver.
Short Transport Layer Test Reference, Sample
MSB. This is the upper eight bits of the
expected DAC sample. It is used to compare
with the received DAC sample at the output
of the JESD204B receiver.
Reserved.
Short Transport Layer Test Fail. This bit shows
whether the selected DAC sample matches
the reference sample. If they match, it is a
test pass, otherwise it is a test fail.
Test pass
Test fail
Must be set to 0x01 for proper JESD interface
configuration.
Logical Lane Invert. Set Bit x high to invert
the JESD deserialized data on Logical Lane x.
Device Identification Number. Link
information received on Link Lane 0 as
specified in Section 8.3 of JESD204B.
Adjustment Resolution to DAC LMFC. Link
information received on Link Lane 0 as
specified in Section 8.3 of JESD204B.
Must be 0.
Bank Identification: Extension to DID. Link
information received on Link Lane 0 as
specified in Section 8.3 of JESD204B.
Reserved.
Direction to Adjust DAC LMFC. Link information
received on Link Lane 0 as specified in
Section 8.3 of JESD204B. Must be 0.
Phase Adjustment Request to DAC Link
information received on Link Lane 0 as
specified in Section 8.3 of JESD204B. Must be 0.
Lane Identification for Lane 0. Link
information received on Link Lane 0 as
specified in Section 8.3 of JESD204B.
Rev. A | Page 107 of 117
Reset
0x0
0x0
Access
R
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
0x0
R
R
00
R/W
0x0
R/W
0x0
R
0x0
R
0x0
R
0x0
0x0
R
R
0x0
R
0x0
R
AD9135/AD9136
Address
0x403
Name
SCR_L_REG
Data Sheet
Bit No.
7
Bit Name
SCR_RD
Settings
0
1
[6:5]
[4:0]
RESERVED
L-1_RD
0
1
3
7
0x404
F_REG
[7:0]
F-1_RD
0
1
3
0x405
K_REG
[7:5]
[4:0]
RESERVED
K-1_RD
0x0F
0x1F
0x406
M_REG
[7:0]
M-1_RD
0
1
0x407
CS_N_REG
[7:6]
CS_RD
5
[4:0]
RESERVED
N-1_RD
[7:5]
SUBCLASSV_RD
[4:0]
NP-1_RD
0x0F
0x408
NP_REG
0x0F
Description
Transmit Scrambling Status.
Link information received on Link Lane 0 as
specified in Section 8.3 of JESD204B.
Scrambling is disabled
Scrambling is enabled
Reserved.
Number of Lanes per Converter Device. Link
information received on Link Lane 0 as
specified in Section 8.3 of JESD204B.
One lane per converter
Two lanes per converter
Four lanes per converter
Eight lanes per converter (single link only)
Number of Octets per Frame. Settings of 1, 2
and 4 octets per frame are valid. Link
information received on Link Lane 0 as
specified in Section 8.3 of JESD204B.
(One octet per frame) per lane
(Two octets per frame) per lane
(Four octets per frame) per lane
Reserved.
Number of Frames per Multiframe. Settings
of 16 or 32 are valid. Link information
received on Link Lane 0 as specified in
Section 8.3 of JESD204B.
16 frames per multiframe
32 frames per multiframe
Number of converters per device. Link
information received on Link Lane 0 as
specified in Section 8.3 of JESD204B. Must be
0 or 1.
One converter per device
Two converters per device
Number of Control Bits per Sample. Link
information received on Link Lane 0 as
specified in Section 8.3 of JESD204B. CS must
be 0.
Reserved.
Converter Resolution. Link information
received on Link Lane 0 as specified in
Section 8.3 of JESD204B. Converter
resolution must be 16.
Converter resolution of 16
Device Subclass Version. Link information
received on Link Lane 0 as specified in
Section 8.3 of JESD204B.
Total Number of Bits per Sample. Link
information received on Link Lane 0 as
specified in Section 8.3 of JESD204B. Must be
16 bits per sample.
16 bits per sample.
Rev. A | Page 108 of 117
Reset
0x0
Access
R
0x0
0x0
R
R
0x0
R
0x0
0x0
R
R
0x0
R
0x0
R
0x0
0x0
R
R
0x0
R
0x0
R
Data Sheet
Address
0x409
Name
S_REG
AD9135/AD9136
Bit No.
[7:5]
Bit Name
JESDV_RD
Settings
000
001
[4:0]
S-1_RD
0
1
0x40A
HD_CF_REG
7
HD_RD
0
1
[6:5]
[4:0]
RESERVED
CF_RD
0x40B
RES1_REG
[7:0]
RES1_RD
0x40C
RES2_REG
[7:0]
RES2_RD
0x40D
CHECKSUM_REG
[7:0]
FCHK0_RD
0x40E
COMPSUM0_REG
[7:0]
FCMP0_RD
0x412
LID1_REG
[7:5]
[4:0]
RESERVED
LID1_RD
0x415
CHECKSUM1_REG
[7:0]
FCHK1_RD
0x416
COMPSUM1_REG
[7:0]
FCMP1_RD
0x41A
LID2_REG
0x41D
0x41E
CHECKSUM2_REG
COMPSUM2_REG
[7:5]
[4:0]
[7:0]
[7:0]
RESERVED
LID2_RD
FCHK2_RD
FCMP2_RD
0x422
LID3_REG
0x425
0x426
CHECKSUM3_REG
COMPSUM3_REG
[7:5]
[4:0]
[7:0]
[7:0]
RESERVED
LID3_RD
FCHK3_RD
FCMP3_RD
Description
JESD204 Version. Link information received
on Link Lane 0 as specified in Section 8.3 of
JESD204B.
JESD204A
JESD204B
Number of Samples per Converter per Frame
Cycle. Settings of one and two are valid. Link
information received on Link Lane 0 as
specified in Section 8.3 of JESD204B.
One sample per converter per frame
Two samples per converter per frame
High Density Format. See Section 5.1.3 of the
JESD294B standard. Link information
received on Link Lane 0 as specified in
Section 8.3 of JESD204B.
Low density mode
High density mode: link information received
on Lane 0 as specified in Section 8.3 of
JESD204B
Reserved.
Number of Control Words per Frame Clock
Period per Link. Link information received on
Link Lane 0 as specified in Section 8.3 of
JESD204B. Bits[4:0] must be 0.
Reserved Field 1. Link information received on
Link Lane 0 as specified in Section 8.3 of
JESD204B.
Reserved Field 2. Link information received on
Link Lane 0 as specified in Section 8.3 of
JESD204B.
Checksum for Link Lane 0. Link information
received on Link Lane 0 as specified in
Section 8.3 of JESD204B.
Computed Checksum for Link Lane 0. The
JESD204B receiver computes the checksum
of the link information received on Lane 0 as
specified in Section 8.3 of JESD204B. The
computation method is set by the
CHECKSUM_MODE bit (Address 0x300[6])
and must match the likewise calculated
checksum in Register 0x40D.
Reserved.
Lane Identification for Link Lane 1.Link
information received on Lane 0 as specified
in section 8.3 of JESD204B.
Checksum for Link Lane 1. Link information
received on Lane 0 as specified in Section 8.3
of JESD204B.
Computed Checksum for Link Lane 1. See the
description for Register 0x40E.
Reserved.
Lane Identification for Link Lane 2.
Checksum for Link Lane 2.
Computed Checksum for Link Lane 2 (see the
description for Register 0x40E).
Reserved.
Lane Identification for Link Lane 3.
Checksum for Link Lane 3.
Computed Checksum for Link Lane 3 (see the
description for Register 0x40E).
Rev. A | Page 109 of 117
Reset
0x0
Access
R
0x0
R
0x0
R
0x0
0x0
R
R
0x0
R
0x0
R
0x0
R
0x0
R
0x0
0x0
R
R
0x0
R
0x0
R
0x0
0x0
0x0
0x0
R
R
R
R
0x0
0x0
0x0
0x0
R
R
R
R
AD9135/AD9136
Address
0x42A
Name
LID4_REG
0x42D
0x42E
CHECKSUM4_REG
COMPSUM4_REG
0x432
LID5_REG
0x435
0x436
CHECKSUM5_REG
COMPSUM5_REG
0x43A
LID6_REG
0x43D
0x43E
CHECKSUM6_REG
COMPSUM6_REG
0x442
LID7_REG
0x445
0x446
Data Sheet
Bit No.
[7:5]
[4:0]
[7:0]
[7:0]
Bit Name
RESERVED
LID4_RD
FCHK4_RD
FCMP4_RD
[7:5]
[4:0]
[7:0]
[7:0]
RESERVED
LID5_RD
FCHK5_RD
FCMP5_RD
[7:5]
[4:0]
[7:0]
[7:0]
RESERVED
LID6_RD
FCHK6_RD
FCMP6_RD
CHECKSUM7_REG
COMPSUM7_REG
[7:5]
[4:0]
[7:0]
[7:0]
RESERVED
LID7_RD
FCHK7_RD
FCMP7_RD
0x450
ILS_DID
[7:0]
DID
0x451
ILS_BID
[7:4]
ADJCNT
[3:0]
BID
7
6
5
RESERVED
ADJDIR
PHADJ
[4:0]
LID0
7
SCR
0x452
0x453
ILS_LID0
ILS_SCR_L
Settings
0
1
[6:5]
[4:0]
RESERVED
L-1
0
1
3
7
0x454
ILS_F
[7:0]
F-1
0
1
3
0x455
ILS_K
[7:5]
[4:0]
RESERVED
K-1
0x0F
0x1F
Description
Reserved.
Lane Identification for Link Lane 4.
Checksum for Link Lane 4.
Computed Checksum for Link Lane 4 (see the
description for Register 0x40E).
Reserved.
Lane Identification for Link Lane 5.
Checksum for Link Lane 5.
Computed Checksum for Link Lane 5 (see the
description for Register 0x40E).
Reserved.
Lane Identification for Link Lane 6.
Checksum for Link Lane 6.
Computed Checksum for Link Lane 6 (see the
description for Register 0x40E).
Reserved.
Lane Identification for Link Lane 7.
Checksum for Link Lane 7.
Computed Checksum for Link Lane 7 (see the
description for Register 0x40E).
Device Identification Number. Link information
received on Link Lane 0 as specified in
Section 8.3 of JESD204B. Must be set to value
read in Register 0x400.
Adjustment Resolution to DAC LMFC Must
be set to 0.
Bank Identification: Extension to DID Must be
set to value read in Register 0x401[3:0].
Reserved.
Direction to Adjust DAC LMFC. Must be set to 0.
Phase Adjustment Request to DAC. Must be
set to 0.
Lane Identification for Link Lane 0. Must be set
to the value read in Register 0x402[4:0].
Receiver Descrambling Enable.
Descrambling is disabled
Descrambling is enabled
Reserved.
Number of Lanes per Converter Device. See
Table 34 and Table 35.
One lane per converter
Two lanes per converter
Four lanes per converter
Eight lanes per converter (single link only)
Number of Octets per Lane per Frame. Settings
of 1, 2, and 4 (octets per lane) per frame are
valid. See Table 34 and Table 35.
(One octet per lane) per frame
(Two octets per lane) per frame
(Four octets per lane) per frame
Reserved.
Number of Frames per Multiframe. Settings
of 16 or 32 are valid. Must be set to 32 when
F = 1 (Register 0x476).
16 frames per multiframe
32 frames per multiframe
Rev. A | Page 110 of 117
Reset
0x0
0x0
0x0
0x0
Access
R
R
R
R
0x0
0x0
0x0
0x0
R
R
R
R
0x0
0x0
0x0
0x0
R
R
R
R
0x0
0x0
0x0
0x0
R
R
R
R
0x0
R/W
0x0
R/W
0x0
R/W
0x0
0x0
0x0
R
R/W
R/W
0x0
R/W
0x1
R/W
0x0
0x3
R
R/W
0x0
R/W
0x0
0x1F
R
R/W
Data Sheet
Address
0x456
Name
ILS_M
AD9135/AD9136
Bit No.
[7:0]
Bit Name
M-1
Settings
0
1
0x457
ILS_CS_N
[7:6]
CS
5
[4:0]
RESERVED
N-1
[7:5]
SUBCLASSV
0
0x0F
0x458
ILS_NP
0
1
[4:0]
NP-1
0xF
0x459
ILS_S
[7:5]
JESDV
000
001
[4:0]
S-1
0
1
0x45A
ILS_HD_CF
7
HD
0
1
[6:5]
[4:0]
RESERVED
CF
0x45B
0x45C
0x45D
ILS_RES1
ILS_RES2
ILS_CHECKSUM
[7:0]
[7:0]
[7:0]
RES1
RES2
FCHK0
0x46B
ERRCNTRMON_RB
[7:0]
READERRORCNTR
0x46B
ERRCNTRMON
7
[6:4]
RESERVED
LANESEL
[3:2]
[1:0]
RESERVED
CNTRSEL
x
00
01
10
0x46C
0x46D
LANEDESKEW
BADDISPARITY_RB
[7:0]
[7:0]
LANEDESKEW
BADDIS
Description
Number of Converters per Device. See Table 34
and Table 35.
One converter per link
Two converters per link
Number of Control Bits per Sample. Must be
set to 0. Control bits are not supported.
Zero control bits per sample
Reserved.
Converter Resolution. Must be set to 16 bits
of resolution.
Converter resolution of 16.
Device Subclass Version.
Subclass 0
Subclass 1
Total Number of Bits per Sample. Must be set
to 16 bits per sample.
16 bits per sample.
JESD204 Version.
JESD204A
JESD204B
Number of Samples per Converter per Frame
Cycle. Settings of one and two are valid.
One sample per converter per frame
Two samples per converter per frame
High Density Format. If F = 1, HD must be set
to 1. Otherwise, HD must be set to 0. See
Section 5.1.3 of JESD204B standard.
Low density mode
High density mode
Reserved.
Number of Control Words per Frame Clock
Period per Link. Must be set to 0. Control bits
are not supported.
Reserved Field 1.
Reserved Field 2.
Checksum for Link Lane 0. Calculated
checksum. Calculation depends on 0x300[6].
Read JESD204B Error Counter. After selecting
the lane and error counter by writing to
LANESEL and CNTRSEL (both in this same
register), the selected error counter is read
back here.
Reserved.
Link Lane select for JESD204B error counter.
Selects the lane whose errors are read back
in this register.
Selects Link Lane x
Reserved.
JESD204B Error Counter Select. Selects the
type of error that are read back in this register.
BADDISCNTR: bad running disparity counter
NITCNTR: not in table error counter
UCCCNTR: Unexpected control character counter
Lane Deskew. Setting Bit x deskews Link Lane x
Bad Disparity Character Error (BADDIS). Bit x
is set when the bad disparity error count for
Link Lane x reaches the threshold in
Register 0x47C.
Rev. A | Page 111 of 117
Reset
0x1
Access
R/W
0x0
R/W
0x0
0xF
R
R/W
0x1
R/W
0xF
R/W
0x1
R/W
0x0
R/W
0x1
R/W
0x0
0x0
R
R/W
0x0
0x0
0x45
R/W
R/W
R/W
0x0
R
0x0
0x0
R
W
0x0
0x0
R
W
0xF
0x0
R/W
R
AD9135/AD9136
Address
0x46D
Name
BADDISPARITY
Data Sheet
Bit No.
7
Bit Name
RST_IRQ_DIS
Settings
6
DISABLE_ERR_
CNTR_DIS
5
RST_ERR_CNTR_DIS
[4:3]
[2:0]
RESERVED
LANE_ADDR_DIS
0x46E
NIT_RB
[7:0]
NIT
0x46E
NIT_W
7
RST_IRQ_NIT
6
DISABLE_ERR_
CNTR_NIT
5
RST_ERR_CNTR_NIT
[4:3]
[2:0]
RESERVED
LANE_ADDR_NIT
0x46F
UNEXPECTEDCONTROL_RB
[7:0]
UCC
0x46F
UNEXPECTEDCONTROL_W
7
RST_IRQ_UCC
6
DISABLE_ERR_
CNTR_UCC
5
[4:3]
[2:0]
RST_ERR_CNTR_
UCC
RESERVED
LANE_ADDR_UCC
[7:0]
CODEGRPSYNC
0x470
CODEGRPSYNCFLG
0
1
0x471
FRAMESYNCFLG
[7:0]
FRAMESYNC
0
1
0x472
GOODCHKSUMFLG
[7:0]
GOODCHECKSUM
0
1
Description
BADDIS IRQ Reset. Reset BADDIS IRQ for lane
selected via Bits[2:0] by writing 1 to this bit.
BADDIS Error Counter Disable. Disable the
BADDIS error counter for lane selected via
Bits[2:0] by writing 1 to this bit.
BADDIS Error Counter Reset. Reset BADDIS
error counter for lane selected via Bits[2:0] by
writing 1 to this bit.
Reserved.
Link Lane Address for Functions Described in
Bits[7:5].
Not in table Character Error (NIT). Bit x is set
when Link Lane x’s NIT error count reaches
the threshold in Register 0x47C.
IRQ Reset. Reset IRQ for lane selected via
Bits[2:0] by writing 1 to this bit.
Disable Error Counter. Disable the error
counter for lane selected via Bits[2:0] by
writing 1 to this bit.
Reset Error Counter. Reset error counter for lane
selected via Bits[2:0] by writing 1 to this bit.
Reserved.
Link Lane Address for Functions Described in
Bits[7:5].
Unexpected Control Character Error (UCC).
Bit x is set when Link Lane x’s UCC error
count reaches the threshold in Register 0x47C.
IRQ Reset. Reset IRQ for lane selected via
Bits[2:0] by writing 1 to this bit.
Disable Error Counter. Disable the error
counter for lane selected via Bits[2:0] by
writing 1 to this bit.
Reset Error Counter. Reset error counter for
lane selected via Bits[2:0] by writing 1 to this bit.
Reserved.
Link Lane Address for Functions Described in
Bits[7:5].
Code Group Sync Flag (from Each Instantiated
Lane). Writing 1 to Bit 7 resets the IRQ. The
associated IRQ flag is located in Register
0x47A[0]. A loss of CODEGRPSYNC triggers
sync request assertion. See the SYNCOUTx±,
SYSREF±, and CLK± Signals section and the
Deterministic Latency section.
Synchronization is lost
Synchronization is achieved
Frame Sync Flag (from Each Instantiated
Lane). This register indicates the live status
for each lane. Writing 1 to Bit 7 resets the
IRQ. A loss of frame sync automatically
initiates a synchronization sequence.
Synchronization is lost
Synchronization is achieved
Good Checksum Flag (from Each Instantiated
Lane). Writing 1 to Bit 7 resets the IRQ.
The associated IRQ flag is located in
Register 0x47A[2].
Last computed checksum is not correct
Last computed checksum is correct
Rev. A | Page 112 of 117
Reset
0x0
Access
W
0x0
W
0x0
W
0x0
0x0
R
W
0x0
R
0x0
W
0x0
W
0x0
W
0x0
0x0
R
W
0x0
R
0x0
W
0x0
W
0x0
W
0x0
0x0
R
W
0x0
R/W
0x0
R/W
0x0
R/W
Data Sheet
AD9135/AD9136
Address
0x473
Name
INITLANESYNCFLG
Bit No.
[7:0]
Bit Name
INITIALLANESYNC
0x476
CTRLREG1
[7:0]
F
Settings
1
2
4
0x477
CTRLREG2
7
ILAS_MODE
1
0
[6:4]
3
0x478
KVAL
[2:0]
[7:0]
RESERVED
THRESHOLD_
MASK_EN
RESERVED
KSYNC
0x47A
IRQVECTOR_MASK
7
BADDIS_MASK
x
1
6
NIT_MASK
1
5
UCC_MASK
1
4
3
2
1
0
0x47A
IRQVECTOR_FLAG
7
RESERVED
INITIALLANESYNC_
MASK
1
BADCHECKSUM_
MASK
1
FRAMESYNC_
MASK
1
CODEGRPSYNC_
MASK
1
BADDIS_FLAG
1
6
NIT_FLAG
1
Description
Initial Lane Sync Flag (from Each Instantiated
Lane). Writing 1 to Bit 7 resets the IRQ. The
associated IRQ flag is located in Register
0x47A[3]. Loss of synchronization is also
reported on SYNCOUT1± or SYNCOUT0±. See
the SYNCOUTx±, SYSREF±, and CLK± Signals
section and the Deterministic Latency section.
Number of Octets per Frame. Settings of 1, 2,
and 4 are valid. See Table 34 and Table 35.
One octet per frame
Two octets per frame
Four octets per frame
ILAS Test Mode. Defined in Section 5.3.3.8 of
JESD204B specification.
JESD204B receiver is constantly receiving
ILAS frames
Normal link operation
Reserved.
Threshold Mask Enable. Set this bit if using
SYNC_ASSERTION_MASK (Register 0x47B[7:5]).
Reserved.
Number of K Multiframes During ILAS
(Divided by Four). Sets the number of
multiframes to send initial lane alignment
sequence. Cannot be set to 0.
4× multiframes during ILAS
Bad Disparity Mask.
If the bad disparity count reaches
ERRORTHRESH on any lane, IRQ is pulled low.
Not in table Mask.
If the not in table character count reaches
ERRORTHRESH on any lane, IRQ is pulled low.
Unexpected Control Character Mask.
If the unexpected control character count
reaches ERRORTHRESH on any lane, IRQ is
pulled low.
Reserved.
Initial Lane Sync Mask.
If initial lane sync (0x473) fails on any lane,
IRQ is pulled low.
Bad Checksum Mask.
If there is a bad checksum (0x472) on any
lane, IRQ is pulled low.
Frame Sync Mask
If frame sync (0x471) fails on any lane, IRQ is
pulled low.
Code Group Sync Machine Mask.
If code group sync (0x470) fails on any lane,
IRQ is pulled low.
Bad Disparity Error Count.
Bad disparity character count reached
ERRORTHRESH (0x47C) on at least one lane.
Read Register 0x46D to determine which
lanes are in error.
Not in table Error Count
Not in table character count reached
ERRORTHRESH (0x47C) on at least one lane.
Read Register 0x46E to determine which
lanes are in error.
Rev. A | Page 113 of 117
Reset
0x0
Access
R/W
0x1
R/W
0x0
R/W
0x0
0x0
R
R/W
0x0
0x1
R
R/W
0x0
W
0x0
W
0x0
W
0x0
0x0
R
W
0x0
W
0x0
W
0x0
W
0x0
R
0x0
R
AD9135/AD9136
Address
Name
Data Sheet
Bit No.
5
Bit Name
UCC_FLAG
Settings
1
4
3
2
1
0
0x47B
SYNCASSERTIONMASK
7
RESERVED
INITIALLANESYNC_
FLAG
1
BADCHECKSUM_
FLAG
1
FRAMESYNC_
FLAG
1
CODEGRPSYNC_
FLAG
1
BADDIS_S
1
6
NIT_S
1
5
UCC_S
1
4
CMM
1
3
CMM_ENABLE
1
0
0x47C
ERRORTHRES
[2:0]
[7:0]
0x47D
LANEENABLE
[7:0]
RESERVED
ETH
LANE_ENA
Description
Unexpected Control Character Error Count
Unexpected control character count reached
ERRORTHRESH (0x47C) on at least one lane.
Read Register 0x46F to determine which
lanes are in error.
Reserved.
Initial Lane Sync Flag.
Initial lane sync failed on at least one lane.
Read Register 0x473 to determine which
lanes are in error
Bad Checksum Flag.
Bad checksum on at least one lane. Read
Register 0x472 to determine which lanes are in
error.
Frame Sync Flag.
Frame sync failed on at least one lane. Read
Register 0x471 to determine which lanes are
in error.
Code Group Sync Flag.
Code group sync failed on at least one lane.
Read Register 0x470 to determine which
lanes are in error
Bad Disparity Error on Sync.
Asserts a sync request on SYNCOUTx± when
the bad disparity character count reaches the
threshold in Register 0x47C
Not in table Error on Sync.
Asserts a sync request on SYNCOUTx± when
the not in table character count reaches the
threshold in Register 0x47C
Unexpected Control Character Error on Sync.
Asserts a sync request on SYNCOUTx± when
the unexpected control character count
reaches the threshold in Register 0x47C
Configuration Mismatch IRQ. If
CMM_ENABLE is high, this bit latches on a
rising edge and pull IRQ low. When latched,
write a 1 to clear this bit. If CMM_ENABLE is
low, this bit is non-functional.
Link Lane 0 configuration registers (Register
0x450 to Register 0x45D) do not match the
JESD204B transmit settings (Register 0x400
to Register 0x40D)
Configuration Mismatch IRQ Enable.
Enables IRQ generation if a configuration
mismatch is detected
Configuration mismatch IRQ disabled
Reserved.
Error Threshold. Bad disparity, not in table,
and unexpected control character errors are
counted and compared to the error
threshold value. When the count reaches the
threshold, either an IRQ is generated or the
SYNCOUTx± signal is asserted per the mask
register settings, or both. Function is
performed in all lanes.
Lane Enable. Setting Bit x enables Link Lane x.
This register must be programmed before
receiving the code group pattern for proper
operation.
Rev. A | Page 114 of 117
Reset
0x0
Access
R
0x0
0x0
R
R
0x0
R
0x0
R
0x0
R
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x1
R/W
0x0
0xFF
R
R/W
0xF
R/W
Data Sheet
Address
0x47E
Name
RAMP_ENA
AD9135/AD9136
Bit No.
[7:1]
0
Bit Name
RESERVED
ENA_RAMP_
CHECK
Settings
0
1
0x520
DIG_TEST0
[7:2]
RESERVED
0x521
DC_TEST_VALUE0
1
0
[7:0]
0x522
DC_TEST_VALUE1
[7:0]
DC_TEST_MODE
RESERVED
DC_TEST_
VALUE[7:0]
DC_TEST_
VALUE[15:8]
Description
Reserved.
Enable Ramp Checking at the Beginning of
ILAS.
Disable ramp checking at beginning of ILAS;
ILAS data need not be a ramp
Enable ramp checking; ILAS data needs to be
a ramp starting at 00-01-02; otherwise, the
ramp ILAS fails and the device does not start up
Must write default value for proper
operation.
DC Test Mode
Reserved.
DC Value LSB of DC Test Mode for DAC0 and
DAC1.
DC value MSB of DC Test Mode for DAC0 and
DAC1.
Rev. A | Page 115 of 117
Reset
0x0
0x0
Access
R
W
0x7
R/W
0x0
0x0
0x0
R/W
R/W
R/W
0x0
R/W
AD9135/AD9136
Data Sheet
OUTLINE DIMENSIONS
12.10
12.00 SQ
11.90
0.28
0.23
0.18
0.60 MAX
0.60
MAX
67
66
88
PIN 1
INDICATOR
1
PIN 1
INDICATOR
11.85
11.75 SQ
11.65
0.50
BSC
0.50
0.40
0.30
22
23
45
44
TOP VIEW
BOTTOM VIEW
10.50
REF
0.70
0.65
0.60
FOR PROPER CONNECTION OF
THE EXPOSED PAD, REFER TO
THE PIN CONFIGURATION AND
FUNCTION DESCRIPTIONS
SECTION OF THIS DATA SHEET.
0.045
0.025
0.005
COPLANARITY
0.08
0.20 REF
SEATING
PLANE
08-10-2012-A
12° MAX
0.90
0.85
0.80
7.55
7.40 SQ
7.25
EXPOSED PAD
COMPLIANT TO JEDEC STANDARDS MO-220-VRRD
Figure 91. 88-Lead Lead Frame Chip Scale Package [LFCSP_VQ]
12 mm × 12 mm Body, Very Thin Quad
(CP-88-6)
Dimensions shown in millimeters
12.10
12.00 SQ
11.90
0.30
0.25
0.20
0.60 MAX
0.60
MAX
67
88
66
1
PIN 1
INDICATOR
PIN 1
INDICATOR
11.85
11.75 SQ
11.65
0.50
BSC
7.55
7.40 SQ
5.25
EXPOSED
PAD
0.65
0.55
0.45
22
44
0.90
0.85
0.80
PKG-004598
SEATING
PLANE
0.70
0.65
0.60
12° MAX
0.190~0.245 REF
0.50
0.40
0.30
45
23
BOTTOM VIEW
10.50
REF
0.045
0.025
0.005
COPLANARITY
0.08
COMPLIANT TO JEDEC STANDARDS MO-220
Figure 92. 88-Lead Lead Frame Chip Scale Package [LFCSP_VQ] (Variable Lead Length)
12 mm × 12 mm Body, Very Thin Quad
(CP-88-9)
Dimensions shown in millimeters
Rev. A | Page 116 of 117
08-04-2014-A
TOP VIEW
1.00
0.90
0.80
0.80
0.70
0.60
Data Sheet
AD9135/AD9136
ORDERING GUIDE
Model1
AD9135BCPZ
AD9135BCPZRL
AD9135BCPAZ
AD9135BCPAZRL
AD9136BCPZ
AD9136BCPZRL
AD9136BCPAZ
AD9136BCPAZRL
AD9136-EBZ
AD9136-FMC-EBZ
AD9135-EBZ
AD9135-FMC-EBZ
1
Temperature Range
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
Package Description
88-Lead LFCSP_VQ
88-Lead LFCSP_VQ
88-Lead LFCSP_VQ (Variable Lead Length)
88-Lead LFCSP_VQ (Variable Lead Length)
88-Lead LFCSP_VQ
88-Lead LFCSP_VQ
88-Lead LFCSP_VQ (Variable Lead Length)
88-Lead LFCSP_VQ (Variable Lead Length)
DPG3 Evaluation Board
FMC Evaluation Board
DPG3 Evaluation Board
FMC Evaluation Board
Z = RoHS Compliant Part.
©2014–2015 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D12578-0-7/15(A)
Rev. A | Page 117 of 117
Package Option
CP-88-6
CP-88-6
CP-88-9
CP-88-9
CP-88-6
CP-88-6
CP-88-9
CP-88-9