Quad, 16-Bit, 2.8 GSPS, TxDAC+®
Digital-to-Analog Converter
AD9144
Data Sheet
FEATURES
TYPICAL APPLICATION CIRCUIT
Supports input data rate >1 GSPS
Proprietary low spurious and distortion design
6-carrier GSM IMD = 77 dBc at 75 MHz IF
SFDR = 82 dBc at dc IF, −9 dBFS
Flexible 8-lane JESD204B interface
Support quad or dual DAC mode at 2.8 GSPS
Multiple chip synchronization
Fixed latency
Data generator latency compensation
Selectable 1×, 2×, 4×, 8× interpolation filter
Low power architecture
Input signal power detection
Emergency stop for downstream analog circuitry protection
Transmit enable function allows extra power saving
High performance, low noise phase-locked loop (PLL) clock
multiplier
Digital inverse sinc filter
Low power: 1.6 W at 1.6 GSPS, 1.7 W at 2.0 GSPS,
full operating conditions
88-lead LFCSP with exposed pad
QUAD MOD
ADRF6720
LPF
DAC
0°/90° PHASE
SHIFTER
JESD204B
SYNCOUTx±
SYSREF±
DAC
AD9144
LO_IN
MOD_SPI
QUAD MOD
ADRF6720
QUAD
DAC
LPF
DAC
0°/90° PHASE
SHIFTER
JESD204B
SYNCOUTx±
LO_IN
MOD_SPI
CLK±
DAC
SPI
11675-001
DAC
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
PRODUCT HIGHLIGHTS
1.
GENERAL DESCRIPTION
The AD9144 is a quad, 16-bit, high dynamic range digital-toanalog converter (DAC) that provides a maximum sample rate
of 2.8 GSPS, permitting a multicarrier generation up to the
Nyquist frequency. The DAC outputs are optimized to interface
seamlessly with the ADRF6720 analog quadrature modulator
(AQM) from Analog Devices, Inc. An optional 3-wire or 4-wire
serial port interface (SPI) provides for programming/readback
of many internal parameters. Full-scale output current can be
programmed over a typical range of 13.9 mA to 27.0 mA. The
AD9144 is available in an 88-lead LFCSP.
Rev. C
2.
3.
4.
5.
6.
Greater than 1 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 ©2014–2019 Analog Devices, Inc. All rights reserved.
Technical Support
www.analog.com
AD9144
Data Sheet
TABLE OF CONTENTS
Features .............................................................................................. 1
JESD204B Setup ......................................................................... 30
Applications ....................................................................................... 1
SERDES Clocks Setup ................................................................ 32
General Description ......................................................................... 1
Equalization Mode Setup .......................................................... 32
Typical Application Circuit ............................................................. 1
Link Latency Setup ..................................................................... 32
Product Highlights ........................................................................... 1
Crossbar Setup ............................................................................ 34
Revision History ............................................................................... 3
JESD204B Serial Data Interface .................................................... 35
Functional Block Diagram .............................................................. 5
JESD204B Overview .................................................................. 35
Specifications..................................................................................... 6
Physical Layer ............................................................................. 36
DC Specifications ......................................................................... 6
Data Link Layer .......................................................................... 39
Digital Specifications ................................................................... 7
Transport Layer .......................................................................... 48
Maximum DAC Update Rate Speed Specifications by Supply ..... 8
JESD204B Test Modes ............................................................... 61
JESD204B Serial Interface Speed Specifications ...................... 8
JESD204B Error Monitoring..................................................... 62
SYSREF to DAC Clock Timing Specifications ......................... 9
Hardware Considerations ......................................................... 64
Digital Input Data Timing Specifications ................................. 9
Digital Datapath ............................................................................. 68
Latency Variation Specifications .............................................. 10
Dual Paging ................................................................................. 68
JESD204B Interface Electrical Specifications ......................... 10
Data Format ................................................................................ 68
AC Specifications........................................................................ 11
Interpolation Filters ................................................................... 68
Absolute Maximum Ratings.......................................................... 12
Digital Modulation ..................................................................... 69
Thermal Resistance .................................................................... 12
Inverse Sinc ................................................................................. 70
ESD Caution ................................................................................ 12
Digital Gain, Phase Adjust, DC Offset, and Group Delay .... 70
Pin Configuration and Function Descriptions ........................... 13
I to Q Swap .................................................................................. 71
Terminology .................................................................................... 16
NCO Alignment ......................................................................... 71
Typical Performance Characteristics ........................................... 17
Downstream Protection ............................................................ 73
Theory of Operation ...................................................................... 22
Datapath PRBS ........................................................................... 75
Serial Port Operation ..................................................................... 23
DC Test Mode ............................................................................. 75
Data Format ................................................................................ 23
Interrupt Request Operation ........................................................ 76
Serial Port Pin Descriptions ...................................................... 23
Interrupt Service Routine .......................................................... 76
Serial Port Options ..................................................................... 23
DAC Input Clock Configurations ................................................ 77
Chip Information ............................................................................ 25
Driving the CLK± Inputs .......................................................... 77
Device Setup Guide ........................................................................ 26
DAC PLL Fixed Register Writes ............................................... 77
Overview...................................................................................... 26
Clock Multiplication .................................................................. 77
Step 1: Start Up the DAC ........................................................... 26
Starting the PLL .......................................................................... 79
Step 2: Digital Datapath ............................................................. 27
Analog Outputs............................................................................... 80
Step 3: Transport Layer .............................................................. 27
Transmit DAC Operation.......................................................... 80
Step 4: Physical Layer ................................................................. 28
Device Power Dissipation.............................................................. 83
Step 5: Data Link Layer .............................................................. 28
Temperature Sensor ................................................................... 83
Step 6: Optional Error Monitoring .......................................... 29
Start-Up Sequence .......................................................................... 84
Step 7: Optional Features ........................................................... 29
Step 1: Start Up the DAC ........................................................... 84
DAC PLL Setup ........................................................................... 30
Step 2: Digital Datapath............................................................. 84
Interpolation ............................................................................... 30
Step 3: Transport Layer .............................................................. 85
Rev. C | Page 2 of 126
Data Sheet
AD9144
Step 4: Physical Layer..................................................................85
Device Configuration Register Map ......................................... 87
Step 5: Data Link Layer ..............................................................86
Device Configuration Register Descriptions .......................... 95
Step 6: Error Monitoring ............................................................86
Outline Dimensions ......................................................................125
Register Maps and Descriptions ....................................................87
Ordering Guide .........................................................................126
REVISION HISTORY
4/2019—Rev. B to Rev. C
Changes to Figure 75 ......................................................................76
Updated Outline Dimensions ......................................................125
Changes to Ordering Guide .........................................................126
3/2017—Rev. A to Rev. B
Changed 10.64 Gbps to 12.4 Gbps, 2.76 Gbps to 3.1 Gbps, and
5.52 Gbps to 6.2 Gbps ................................................... Throughout
Changes to Table 4 ............................................................................ 7
Change to Device Revision Parameter; Table 14 ........................24
Changes to Function Overview of the SERDES PLL Section ...36
Changes to Figure 38 ......................................................................37
Changes to Table 97 ........................................................................86
Changes to Table 98 ........................................................................94
6/2015—Rev. 0 to Rev. A
Changed Functional Block Diagram Section to Typical
Application Circuit Section.............................................................. 1
Changes to Figure 1........................................................................... 1
Changed Detailed Functional Block Diagram Section to
Functional Block Diagram Section ................................................. 4
Deleted Reference Voltage Parameter, Table 1 .............................. 5
Changes to Output Voltage (VOUT) Logic High Parameter,
Output Voltage (VOUT) Logic Low Parameter, and SYSREF±
Frequency Parameter, Table 2 .......................................................... 6
Changes to Table 4 ............................................................................ 7
Changes to Interpolation Parameter, Table 6 ................................ 8
Deleted Sync Off, Subclass Mode 0 Parameter, Table 7 ............... 9
Changed Junction Temperature Parameter to Operating
Junction Temperature, Table 10 ....................................................11
Changes to Terminology Section ..................................................15
Changes to Figure 26 Caption .......................................................19
Changes to Figure 29 Caption .......................................................20
Change to Device Revision Parameter, Table 14 .........................24
Changes to Step 1: Start Up the DAC Section, Table 16, and
Table 17 .............................................................................................25
Changes to Step 3: Transport Layer Section and Table 19 .........26
Changes to Table 20 and Table 21 .................................................27
Changes to Step 7: Optional Features Section .............................28
Added Table 25; Renumbered Sequentially .................................29
Changes to DAC PLL Setup Section and Table 26 ......................29
Changes to Lane0Checksum Section ...........................................30
Changes to Table 30 and Subclass 0 Section ................................31
Changes to Table 33 ........................................................................32
Changes to Table 37 ........................................................................35
Changes to Table 38 ........................................................................36
Added SERDES PLL Fixed Register Writes Section and
Table 39 ............................................................................................. 36
Changes to Figure 38 and Table 40 ............................................... 37
Changes to Figure 29 and Data Link Layer Section ................... 38
Added Figure 42; Renumbered Sequentially ............................... 39
Changes to Figure 44 ...................................................................... 40
Changes to Continuous Sync Mode (SYNCMOD = 0x2)
Section .............................................................................................. 42
Changes to Subclass 0 Section ....................................................... 43
Changes to Figure 53 ...................................................................... 50
Changes to Table 49 and Figure 54 ............................................... 51
Changes to Table 50 and Figure 55 ............................................... 52
Changes to Table 51 and Figure 56 ............................................... 53
Changes to Table 52 and Figure 57 ............................................... 54
Changes to Table 53, Table 54, and Figure 58 ............................. 55
Changes to Table 55 and Figure 59 ............................................... 56
Changes to Table 56 and Figure 60 ............................................... 57
Changes to Table 57 and Figure 61 ............................................... 58
Changes to Table 58 and Figure 62 ............................................... 59
Changes to Power Supply Recommendations Section ............... 63
Added Figure 64 .............................................................................. 64
Changes to Figure 68 ...................................................................... 66
Changes to Table 66 ........................................................................ 67
Changes to Table 70, Table 71, Table 72, and I to Q Swap
Section .............................................................................................. 70
Changes to Power Detection and Protection Section ................ 72
Changes to DC Test Mode Section ............................................... 73
Moved Figure 75 and Table 78 ...................................................... 75
Deleted Table 80; Renumbered Sequentially ............................... 76
Added DAC PLL Fixed Register Writes Section and
Table 79 ............................................................................................. 76
Changes to Clock Multiplication Section .................................... 76
Added Loop Filter Section and Charge Pump Section .............. 77
Added Temperature Tracking Section and Table 83 .................. 78
Changes to Starting the PLL Section and Figure 79 ................... 78
Changes to Transmit DAC Operation Section ............................ 79
Changes to Self Calibration Section ............................................. 81
Added Figure 86 and Figure 87 ..................................................... 81
Changes to Device Power Dissipation Section............................ 82
Changes to Table 88 and Table 89 ................................................. 83
Changes to Table 93 ........................................................................ 84
Changes to Table 94, Table 95, and Table 96 ............................... 85
Changes to Table 97 ........................................................................ 86
Changes to Table 98 ........................................................................ 94
Deleted Lookup Tables for Three Different DAC PLL Reference
Frequencies Section and Table 96 to Table 98 ........................... 122
Added Figure 89 ............................................................................ 124
Rev. C | Page 3 of 126
AD9144
Data Sheet
Updated Outline Dimensions ..................................................... 124
Changes to Ordering Guide ........................................................ 125
7/2014—Revision 0: Initial Version
Rev. C | Page 4 of 126
Data Sheet
AD9144
FUNCTIONAL BLOCK DIAGRAM
DACCLK
SERDES
PLL
HB2
HB1
OUT3+
COMPLEX
MODULATION
HB3
HB2
I-GAIN
PHASE
ADJUST
I-OFFSET
OUT2+
÷4, ÷8
HB3
FSC
HB2
HB1
OUT2–
COMPLEX
MODULATION
HB3
OUT1+
FSC
PDP0
SERDIN0±
HB1
INV SINC
SERDIN7±
NCO
MODE CONTROL
OUT3–
DACCLK
fDAC
CLOCK DATA RECOVERY
AND CLOCK FORMATTER
VTT
INV SINC
PDP1
FSC
Q-OFFSET
Q-GAIN
NCO
MODE CONTROL
I-GAIN
OUT1–
Q-OFFSET
Q-GAIN
PHASE
ADJUST
DACCLK
I-OFFSET
fDAC
HB1
PROTECT_OUT0
PROTECT_OUT1
FSC
PLL_CTRL
SERIAL
I/O PORT
POWER-ON
RESET
DAC
ALIGN
DETECT
CLK_SEL
CONFIG
REGISTERS
CLOCK DISTRIBUTION
AND
CONTROL LOGIC
DACCLK
OUT0–
REF
AND
BIAS
SYSREF
Rx
CLK
Rx
I120
SYSREF+
SYSREF–
CLK+
CLK–
11675-002
DAC PLL
TXEN1
IRQ
TXEN0
PLL_LOCK
RESET
SYNCOUT1+
SYNCOUT1–
OUT0+
÷4, ÷8
HB3
SYNCHRONIZATION
LOGIC
SDO
SDIO
SCLK
CS
SYNCOUT0+
SYNCOUT0–
HB2
Figure 2.
Rev. C | Page 5 of 126
AD9144
Data Sheet
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.
Parameter
RESOLUTION
ACCURACY
Differential Nonlinearity (DNL)
Integral Nonlinearity (INL)
MAIN DAC OUTPUTS
Gain Error
I/Q Gain Mismatch
Full-Scale Output Current
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
Test Conditions/Comments
Min
Typ
16
Max
Unit
Bits
With calibration
±1.0
±2.0
With internal reference
LSB
LSB
−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
mA
mA
mV
MΩ
pF
Based on a 4 kΩ external resistor between I120 and GND
0.2
3.0
Guaranteed
20
To within ±0.5 LSB
ns
0.04
32
ppm
ppm/°C
1.2
V
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
3.3
1.2
3.47
1.37
V
V
1.14
1.274
1.2
1.3
1.26
1.326
V
V
1.14
1.274
1.71
1.2
1.3
1.8
1.26
1.326
3.47
V
V
V
1.59
1.84
W
126
95.3
101
518.2
234
11
36
134
112.4
111
654
255
12
50
mA
mA
mA
mA
mA
mA
µA
SVDD12
IOVDD
POWER CONSUMPTION
4× Interpolation Mode,
JESD Mode 4, 8 SERDES Lanes
AVDD33
PVDD12
CVDD12
SVDD12
DVDD12
SIOVDD33
IOVDD
fDAC = 1.6 GSPS, IF = 40 MHz, NCO off, PLL on, digital gain
on, inverse sinc on, DAC FSC = 20 mA
Includes VTT
Rev. C | Page 6 of 126
Data Sheet
AD9144
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× interpolation 2 (see Table 4)
2× interpolation 3
4× interpolation
8× interpolation
1060
2120
2800
2800
MSPS
MSPS
MSPS
MSPS
1× interpolation
2× interpolation
4× interpolation
8× interpolation
1060
1060
700
350
MSPS
MSPS
MSPS
MSPS
MAXIMUM DAC UPDATE RATE 1
ADJUSTED DAC UPDATE RATE
INTERFACE 4
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
REFCLK Frequency (PLL Mode)
SYSTEM REFERENCE INPUT
(SYSREF+, SYSREF−)
Differential Peak-to-Peak
Voltage
Common-Mode Voltage
SYSREF± Frequency 5
SYSREF TO DAC CLOCK 6
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
mV
2000
fDATA/(K × S)
mV
Hz
12.4
400
Self biased input, ac-coupled
6.0 GHz ≤ fVCO ≤ 12.0 GHz
Lanes
1000
600
2800
35
400
1000
0
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. C | Page 7 of 126
ns
ns
ns
ns
AD9144
Parameter
SDO to SCLK
Data Valid Window
CS to SCLK
Setup Time
Hold Time
Data Sheet
Symbol
Test Conditions/Comments
Min
Typ
Max
Unit
tDV
25
ns
tSCS
5
ns
tHCS
2
ns
See Table 3 for detailed specifications for DAC update rate conditions.
Maximum speed for 1× interpolation is limited by the JESD interface. See Table 4 for details.
Maximum speed for 2× interpolation is limited by the JESD interface. See Table 4 for details.
4
See Table 4 for detailed specifications for JESD speed conditions.
5
K, F, and S are JESD204B transport layer parameters. See Table 44 for the full definitions.
6
See Table 5 for detailed specifications for SYSREF to DAC clock timing conditions.
1
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
Test Conditions/Comments
DVDD12, CVDD12 = 1.2 V ± 5%
DVDD12, CVDD12 = 1.2 V ± 2%
DVDD12, CVDD12 = 1.3 V ± 2%
Min
2.23
2.41
2.80
Typ
Max
Unit
GSPS
GSPS
GSPS
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. C | Page 8 of 126
Min
5.75
5.75
5.75
2.88
2.88
2.88
1.44
1.44
1.44
Typ
Max
11.4
12.0
12.4
5.98
6.06
6.2
3.0
3.04
3.1
Unit
Gbps
Gbps
Gbps
Gbps
Gbps
Gbps
Gbps
Gbps
Gbps
Data Sheet
AD9144
SYSREF 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
Fine Modulation
Coarse Modulation
fS/8
fS/4
Digital Phase Adjust
Digital Gain Adjust
Power-Up Time
Dual A Only
Dual B Only
All DACs
1
Test Conditions/Comments
Min
Register 0x011 from 0x60 to 0x00
Register 0x011 from 0x18 to 0x00
Register 0x011 from 0x7C to 0x00
PClock is the AD9144 internal processing clock and equals the lane rate ÷ 40.
Rev. C | Page 9 of 126
Typ
Max
Unit
17
PClock 1 cycles
58
137
251
484
17
20
DAC clock cycles
DAC clock cycles
DAC clock cycles
DAC clock cycles
DAC clock cycles
DAC clock cycles
8
4
12
12
DAC clock cycles
DAC clock cycles
DAC clock cycles
DAC clock cycles
60
60
60
µs
µs
µs
AD9144
Data Sheet
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
DACCLK cycles
DACCLK 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 (SYNCOUT±) 2
Output Differential Voltage
Output Offset Voltage
Output Differential Voltage
DETERMINISTIC LATENCY
Fixed
Variable
SYSREF±-to-LMFC DELAY
1
2
3
Symbol
Test Conditions/Comments
Min
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
VOD
VOS
VOD
AC-coupled, VTT = SVDD12 1
At dc
At dc
High swing mode: Register 0x2A5[0] = 1
94
−0.05
110
100
8
6
192
1.19
341
4
As measured on the input side of the ac coupling capacitor.
IEEE Standard 1596.3 LVDS compatible.
PClock is the AD9144 internal processing clock and equals the lane rate ÷ 40.
Rev. C | Page 10 of 126
Max
Unit
714
+1.85
1050
30
120
µA
µA
ps
V
mV
Ω
Ω
dB
dB
235
1.27
394
mV
V
mV
17
2
PClock 3 cycles
PClock3 cycles
DAC clock cycles
10
−4
80
Normal swing mode: Register 0x2A5[0] = 0
Typ
Data Sheet
AD9144
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. C | Page 11 of 126
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
AD9144
Data Sheet
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,
PROTECT_OUTx to Ground
LDO_BYP1
LDO_BYP2
LDO24
Ambient Operating Temperature (TA)
Operating Junction Temperature
Storage Temperature
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
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.
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), 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. C | Page 12 of 126
Unit
°C/W
Data Sheet
AD9144
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
LDO24
OUT1–
OUT1+
AVDD33
CVDD12
AVDD33
OUT2+
OUT2–
LDO24
CVDD12
LDO24
OUT3–
OUT3+
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
AD9144
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. THE EXPOSED PAD MUST BE SECURELY CONNECTED TO THE GROUND PLANE.
11675-003
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+
19
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 and DAC1. The CMOS levels are determined with respect to IOVDD.
Transmit Enable for DAC2 and DAC3. The 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.
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.
Rev. C | Page 13 of 126
AD9144
Pin No.
20
21
22
23
24
25
26
Mnemonic
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
61
DVDD12
GND
GND
PVDD12
PVDD12
PROTECT_OUT1
PROTECT_OUT0
IRQ
RESET
Data Sheet
Description
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 Protection Pin Output for DAC2 and DAC3. Pin 58 is high when power protection is in process.
Power Detection Protection Pin Output for DAC0 and DAC1. Pin 59 is high when power protection is in process.
Interrupt Request (Active Low, Open Drain).
Reset. This pin is active low. CMOS levels are determined with respect to IOVDD.
Rev. C | Page 14 of 126
Data Sheet
Pin No.
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
SDO
SDIO
SCLK
CS
IOVDD
AVDD33
OUT3+
OUT3−
LDO24
CVDD12
LDO24
OUT2−
OUT2+
AVDD33
CVDD12
AVDD33
OUT1+
OUT1−
LDO24
CVDD12
LDO24
OUT0−
OUT0+
AVDD33
I120
CVDD12
LDO_BYP2
EPAD
AD9144
Description
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.
DAC3 Positive Current Output.
DAC3 Negative Current Output.
2.4 V LDO. Requires a 1 µF capacitor to ground.
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.
DAC2 Negative Current Output.
DAC2 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.
DAC1 Positive Current Output.
DAC1 Negative Current Output.
2.4 V LDO. Requires a 1 µF capacitor to ground.
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. C | Page 15 of 126
AD9144
Data Sheet
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. C | Page 16 of 126
Data Sheet
AD9144
TYPICAL PERFORMANCE CHARACTERISTICS
0
–20
–40
–60
–80
–60
200
300
400
500
fOUT (MHz)
–100
200
300
400
500
Figure 7. Single-Tone SFDR vs. fOUT in the First Nyquist Zone
over Digital Back Off, fDAC = 983 MHz
0
fDAC = 1966MHz
fDAC = 2456MHz
MEDIAN
–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
11675-107
100
11675-104
0
0dBFS
–6dBFS
–9dBFS
–12dBFS
–20
–40
SFDR (dBc)
–60
–80
–40
–60
–100
100
200
300
400
500
fOUT (MHz)
–100
11675-305
0
0
100
200
300
400
500
fOUT (MHz)
Figure 5. Single-Tone SFDR vs. fOUT in the First Nyquist Zone,
fDAC = 1966 MHz and 2456 MHz
11675-108
–80
Figure 8. Single-Tone SFDR vs. fOUT in the First Nyquist Zone
over Digital Back Off, fDAC = 1966 MHz
0
0
IN-BAND SECOND HARMONIC
IN-BAND THIRD HARMONIC
MAX DIGITAL SPUR
–20
–40
–40
IMD3 (dBc)
–20
–60
–80
fDAC = 983MHz
fDAC = 1228MHz
fDAC = 1474MHz
–60
–80
0
100
200
300
fOUT (MHz)
400
500
11675-106
–100
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. C | Page 17 of 126
–100
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
11675-109
SFDR (dBc)
–40
–80
–100
SFDR (dBc)
0dBFS
–6dBFS
–9dBFS
–12dBFS
–20
SFDR (dBc)
SFDR (dBc)
0
fDAC = 983MHz
fDAC = 1228MHz
fDAC = 1474MHz
AD9144
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)
Figure 10. Two-Tone Third IMD (IMD3) vs. fOUT,
fDAC = 1966 MHz and 2456 MHz
0
–100
11675-110
–100
fDAC = 983MHz
fDAC = 1966MHz
0
100
200
300
400
500
fOUT (MHz)
11675-113
IMD3 (dBc)
0
Data Sheet
Figure 13. Two-Tone Third IMD (IMD3) vs. fOUT over Tone Spacing at
0 dB Back Off, fDAC = 983 MHz and 1966 MHz
–130
0dBFS
–6dBFS
–9dBFS
–12dBFS
fDAC = 983MHz
fDAC = 1228MHz
fDAC = 1474MHz
–135
–20
NSD (dBm/Hz)
IMD3 (dBc)
–140
–40
–60
–145
–150
–155
–160
–80
100
200
300
400
500
fOUT (MHz)
–170
0
100
200
300
400
500
11675-114
0
11675-111
–100
500
11675-115
–165
fOUT (MHz)
Figure 11. Two-Tone Third IMD (IMD3) vs. fOUT over Digital Back Off,
fDAC = 983 MHz, Each Tone Is at −6 dBFS
Figure 14. Single-Tone (0 dBFS) NSD vs. fOUT,
fDAC = 983 MHz, 1228 MHz, and 1474 MHz
–130
0
0dBFS
–6dBFS
–9dBFS
–12dBFS
–20
fDAC = 1966MHz
fDAC = 2456MHz
–135
NSD (dBm/Hz)
IMD3 (dBc)
–140
–40
–60
–145
–150
–155
–160
–80
–100
0
100
200
300
fOUT (MHz)
400
500
11675-112
–165
Figure 12. Two-Tone Third IMD (IMD3) vs. fOUT over Digital Back Off,
fDAC = 1966 MHz, Each Tone Is at −6 dBFS
Rev. C | Page 18 of 126
–170
0
100
200
300
400
fOUT (MHz)
Figure 15. Single-Tone (0 dBFS) NSD vs. fOUT,
fDAC = 1966 MHz and 2456 MHz
Data Sheet
–130
AD9144
–60
0dBFS
–6dBFS
–9dBFS
–12dBFS
–135
–80
PHASE NOISE (dBc/Hz)
–140
NSD (dBm/Hz)
fOUT = 30MHz
fOUT = 200MHz
fOUT = 400MHz
–145
–150
–155
PLL: OFF
PLL: ON
–100
–120
–140
–160
–160
–180
0
100
200
300
400
500
fOUT (MHz)
11675-116
–170
Figure 16. Single-Tone NSD vs. fOUT over Digital Back Off,
fDAC = 983 MHz
–130
10
100
1k
10k
100k
1M
10M
OFFSET FREQUENCY (Hz)
Figure 19. Single-Tone Phase Noise vs. Offset Frequency over fOUT,
fDAC = 2.0 GHz, PLL On and Off
0dBFS
–6dBFS
–9dBFS
–12dBFS
–135
NSD (dBm/Hz)
–140
–145
–150
–155
–160
0
100
200
300
400
500
fOUT (MHz)
11675-117
–170
Figure 17. Single-Tone NSD vs. fOUT over Digital Back Off,
fDAC = 1966 MHz
–130
Figure 20. 1C WCDMA ACLR, fOUT = 30 MHz,
fDAC = 983 MHz, 2× Interpolation, PLL Frequency = 122 MHz
PLL OFF
PLL ON
–135
fDAC = 983MHz
fDAC = 1966MHz
–140
–145
–150
–155
–160
–170
0
100
200
300
400
fOUT (MHz)
500
11675-316
–165
11675-118
NSD (dBm/Hz)
11675-315
–165
Figure 18. Single-Tone NSD (0 dBFS) vs. fOUT,
fDAC = 983 MHz and 1966 MHz, PLL On and Off
Figure 21. 1C WCDMA ACLR, fOUT = 122 MHz,
fDAC = 983 MHz, 2× Interpolation, PLL Frequency = 122 MHz
Rev. C | Page 19 of 126
11675-119
–165
Data Sheet
11675-317
11675-320
AD9144
Figure 25. 4C WCDMA ACLR, fOUT = 245 MHz,
fDAC = 1966 MHz, 4× Interpolation, PLL Frequency = 245 MHz
Figure 22. 4C WCDMA ACLR, fOUT = 30 MHz,
fDAC = 983 MHz, 2× Interpolation, PLL Frequency = 122 MHz
1800
1×
2×
4×
8×
POWER CONSUMPTION (mW)
1700
1600
1500
1400
1300
1200
Figure 23. 4C WCDMA ACLR, fOUT = 122 MHz,
fDAC = 983 MHz, 2× Interpolation, PLL Frequency = 122 MHz
1000
0
1000
500
1500
2000
2500
fDAC (MHz)
11675-326
11675-318
1100
Figure 26. Total Power Consumption vs. fDAC over Interpolation, 8 SERDES Lanes
Enabled, 4 DACs Enabled, NCO, Digital Gain, Inverse Sinc and DAC PLL Disabled
120
POWER CONSUMPTION (mW)
100
NCO
PLL (fDAC /fREF RATIO:4)
DIGITAL GAIN
INVERSE SINC
80
60
40
0
200
400
600
800
1000
1200
1400
1600
fDAC (MHz)
Figure 24. 4C WCDMA ACLR, fOUT = 30 MHz,
fDAC = 1966 MHz, 4× Interpolation, PLL Frequency = 245 MHz
Figure 27. Power Consumption vs. fDAC over Digital Functions
Rev. C | Page 20 of 126
11675-327
11675-319
20
Data Sheet
700
AD9144
2 LANES
4 LANES
8 LANES
350
1.2V SVDD12 SUPPLY
1.3V SVDD12 SUPPLY
300
SUPPLY CURRENT (mA)
500
400
300
200
1.2V SUPPLY
1.3V SUPPLY
3.3V SUPPLY
250
200
150
100
50
1
2
3
4
5
LANE RATE (Gbps)
6
7
8
0
200
11675-328
100
Figure 28. SVDD12 Current vs. Lane Rate over Number of SERDES Lanes and
Supply Voltage Setting
400
600
800
100
fDAC (MHz)
1200
1400
1600
11675-329
SVDD12 CURRENT (mA)
600
DVDD12
CVDD12
PVDD12
AVDD33
Figure 29. DVDD12, CVDD12, PVDD12, and AVDD33 Supply Current vs. fDAC
over Supply Voltage Setting, 4 DACs Enabled
Rev. C | Page 21 of 126
AD9144
Data Sheet
THEORY OF OPERATION
The AD9144 is a 16-bit, quad DAC with a SERDES interface.
Figure 2 shows a detailed functional block diagram of the
AD9144. Eight high speed serial lanes carry data at a maximum
speed of 12.4 Gbps, and a 1.06 GSPS input data rate to the DACs.
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. To add application flexibility, the quad DAC can
be configured as a dual-link device with each JESD204B link
providing data for a dual DAC pair.
The digital datapath of the AD9144 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 AD9144 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
ADRF6720 AQM. The AD9144 is 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 AD9144
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 AD9144 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. C | Page 22 of 126
Data Sheet
AD9144
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 AD9144. 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 singlepin I/O (SDIO).
SDO 62
SCLK 64
CS 65
SPI
PORT
Serial Clock (SCLK)
Figure 30. Serial Port Interface Pins
There are two phases to a communication cycle with the AD9144.
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, except for the frequency tuning word (FTW) and numerically
controlled oscillator (NCO) phase offsets, which change only when
the frequency tuning word FTW_UPDATE_REQ bit is set.
DATA FORMAT
The instruction byte contains the information shown in Table 13.
Table 13. Serial Port Instruction Word
I[15] (MSB)
R/W
I[14:0]
A[14:0]
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 ADDRINC bit. If ADDRINC is set
high (Register 0x000, Bit 5 and Bit 2), multibyte SPI writes start
on A[14:0] and increment by 1 every 8 bits sent/received. If
ADDRINC is set to 0, the address decrements by 1 every 8 bits.
SERIAL PORT PIN DESCRIPTIONS
11675-044
SDIO 63
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.
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.
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 LSBFIRST bit
(Register 0x000, Bit 6 and Bit 1). The default is MSB first
(LSBFIRST = 0).
When LSBFIRST = 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
LSBFIRST = 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
SDOACTIVE = 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 SDOACTIVE = 0, the SDO pin is unused
and the SDIO pin is used for both input and output.
Rev. C | Page 23 of 126
AD9144
Data Sheet
To prevent confusion and to ensure consistency between devices,
the chip tests the first nibble following the address phase, ignoring
the second nibble. This 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.
INSTRUCTION CYCLE
DATA TRANSFER CYCLE
CS
SDIO
R/W A14 A13
A3
A2 A1
A0 D7 N D6N D5N
D30 D20 D10 D00
11675-045
SCLK
Figure 31. Serial Register Interface Timing, MSB First, ADDRINC = 0
INSTRUCTION CYCLE
DATA TRANSFER CYCLE
CS
SCLK
SDIO
A0
A1 A2
A12 A13 A14 R/W D00 D10 D20
D4N D5 N D6 N D7N
11675-046
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 ADDRINC (Register 0x000, Bit 5 and Bit 2). When
ADDRINC is 1, the multicycle addresses are incremented.
When ADDRINC is 0, the addresses are decremented. A new
write cycle can always be initiated by bringing CS high and then
low again.
Figure 32. Serial Register Interface Timing, LSB First, ADDRINC = 1
CS
tDV
SDIO
DATA BIT n
DATA BIT n – 1
Figure 33. Timing Diagram for Serial Port Register Read
tSCS
tHCS
CS
tPWH
tPWL
tDS
SDIO
tDH
INSTRUCTION BIT 15
INSTRUCTION BIT 14
INSTRUCTION BIT 0
Figure 34. Timing Diagram for Serial Port Register Write
Rev. C | Page 24 of 126
11675-047
SCLK
11675-048
SCLK
Data Sheet
AD9144
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.
8 MSBs in Register 0x005 and 8 LSBs in Register 0x004. The product ID is 0x9144.
Register 0x006[7:4]. The product grade is 0x00.
Register 0x006[3:0]. The device revision is 0x08.
Rev. C | Page 25 of 126
AD9144
Data Sheet
DEVICE SETUP GUIDE
OVERVIEW
The sequence of steps to properly set up the AD9144 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 registers, and set
up the DAC clocks (see the Step 1: Start Up the DAC section).
Set the digital features of the AD9144 (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 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
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
STEP 1: START UP THE DAC
0x1BC
0x0D
This section describes how to set up the SPI interface, power up
necessary circuit blocks, write required configuration registers,
and set up the DAC clocks, as listed in Table 15.
0x1BE
0x02
0x1BF
0x8E
0x1C0
0x2A
0x1C1
0x2A
0x1C4
0x7E
0x08B
0x08C
0x085
Various
0x
0x
0x
0x
0x083
0x10
7.
The register writes listed in Table 15 to Table 21 give the register
writes necessary to set up the AD9144. Consider printing out
this setup guide and filling in the Value column with 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.
Table 15. Power-Up and DAC Initialization Settings
Addr.
0x000
0x000
0x011
Bit
No.
7
[6:3]
2
0x080
0x081
1
Value
0xBD
0x3C
0x
0
1
Variable
PdDACs
0
0x
0x
PdClocks
PdSysref
Description
Soft reset.
Deassert reset, set 4-wire SPI.
Power up band gap.
PdDACs = 0 if all 4 DACs are
being used. If not, see the DAC
Power-Down Setup section.
Power up master DAC.
PdClocks = 0 if all 4 DACs are
being used. If not, see the DAC
Power-Down Setup section.
PdSysref = 0x00 for Subclass 1.
PdSysref = 0x10 for Subclass 0.
See the Subclass Setup section
for details on subclass.
Description
Digital datapath configuration
Digital datapath configuration
Clock configuration
SERDES interface configuration
SERDES interface configuration
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 is 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 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.
1
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. C | Page 26 of 126
Data Sheet
AD9144
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 fine and coarse modulation,
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
Addr.
0x112
Bit No.
0x110
Variable
InterpMode
Description
Select interpolation
mode; see the
Interpolation section.
DataFmt
DataFmt = 0 if twos
complement;
DataFmt = 1 if unsigned
binary.
0x
7
1
Value
0x
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.
0x
0x300
Table 18. Digital Datapath Settings
1
Bit No. Value1 Variable
0x00
UnusedLanes
0x
6
CheckSumMode
3
DualLink
2
CurrentLink
0x450
0x
DID
0x451
0x
BID
0x452
0x
LID
0x453
See the JESD204B
Setup section for
details on these
variables.
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
[4:0]
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.
0x454
0x
Scrambling
L − 12
F − 12
0x455
0x
K − 12
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).
0x456
0x
M − 12
0x457
0x458
0x
0x
N − 12
STEP 3: TRANSPORT LAYER
Description
Power up the
interface.
See the JESD204B
Setup section.
5
Subclass
[4:0]
NP − 12
0x459
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
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
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.
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].
1
Rev. C | Page 27 of 126
AD9144
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 processing,
setting deterministic latency, and establishing the link.
Table 20. Device Configurations and Physical Layer Settings
Table 21. Data Link Layer Settings
Addr.
0x2AA
0x2AB
0x2B1
0x2B2
0x2A7
0x2AE
0x314
0x230
Bit
No.
Value1
0xB7
0x87
0xB7
0x87
0x01
0x01
0x01
0x
5
[4:2]
Halfrate
0x2
1
0x206
0x206
0x289
2
[1:0]
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]
Variable
EqMode
0x22
Description
SERDES interface termination
setting
SERDES interface termination
setting
Autotune PHY setting
Autotune PHY setting
SERDES SPI configuration
Set up CDR; see the SERDES
Clocks Setup section
SERDES PLL default
configuration
Set up CDR; see the SERDES
Clocks Setup section
Reset CDR
Release CDR reset
SERDES PLL configuration
Set 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 SERDES PLL2
See the Equalization Mode
Setup section
Required value (default)
Addr.
0x301
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
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 =
one-shot 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 polarity of
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.
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.
Rev. C | Page 28 of 126
1
Bit
No.
Data Sheet
AD9144
Table 22. Optional Features
STEP 6: OPTIONAL ERROR MONITORING
Feature
Digital
Modulation
Default
Off
STEP 7: OPTIONAL FEATURES
Inverse Sinc
On
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 Dual Paging section. Paging is particularly
important for dual specific settings like digital gain, phase adjust,
and dc offset.
Digital Gain
2.7 dB
Phase Adjust
Off
DC Offset
Off
Group Delay
0
Downstream
Protection
Off
Self Calibration
Off
For JESD204B error monitoring, see the JESD204B Error
Monitoring section. For other error checks, see the Interrupt
Request Operation section.
Rev. C | Page 29 of 126
Description
Modulates the data with a desired
carrier. See the Digital Modulation
section.
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 balance I/Q phase. See the
Phase Adjust 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.
AD9144
Data Sheet
DAC PLL SETUP
INTERPOLATION
This section explains how to select the appropriate LODivMode,
RefDivMode, and BCount in the Step 1: Start Up the DAC
section. These parameters depend on the desired DAC clock
frequency (fDACCLK) 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
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
DAC Frequency Range (MHz)
1500 to 2800
750 to 1500
420 to 750
LODivMode,
Register 0x08B[1:0]
1
2
3
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
RefDivMode,
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 = fDACCLK × 2LODivMode + 1
where 6 GHz ≤ fVCO ≤ 12 GHz.
BCount must be between 6 and 127 and is calculated based on
fDACCLK and fREF as follows:
BCount = floor((fDACCLK)/(2 × fREF/RefDivFactor))
where RefDivFactor = 2RefDivMode (see Table 24).
Table 25. VCO Control Lookup Table Reference
Register
0x1B5
Setting
0x08
0x09
0x09
Register
0x1BB
Setting
0x03
0x03
0x13
Max fDATA (MHz)
1060 (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 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, L, F, K, M, N, NP, S, HD, and
Subclass must be set the same on the transmit side.
For a summary of how a JESD204B system works and what each
parameter means, see the JESD204B Serial Data Interface section.
Available Operating Modes
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.
VCO Frequency
Range (GHz)
fVCO < 6.3
6.3 ≤ fVCO < 7.25
fVCO ≥ 7.25
Table 26. Interpolation Modes and Their Usable Bandwidth
Register
0x1C5
Setting
0x07
0x06
0x06
For more information on the DAC PLL, see the DAC Input
Clock Configurations section.
Table 27. JESD204B Operating Modes (Single-Link Only)
Parameter
M (Converter Count)
L (Lane Count)
S ((Samples per Converter) per Frame)
F ((Octets per Frame) per Lane)
0
4
8
1
1
1
4
8
2
2
Mode
2
4
4
1
2
3
4
2
1
4
Table 28. JESD204B Operating Modes (Single- or Dual-Link)
Parameter
M (Converter Count)
L (Lane Count)
S ((Samples per Converter) per Frame)
F ((Octets per Frame) per Lane)
Rev. C | Page 30 of 126
4
2
4
1
1
5
2
4
2
2
Mode
6 7
2 2
2 1
1 1
2 4
9
1
2
1
1
10
1
1
1
2
Data Sheet
AD9144
For a particular application, the number of converters to use (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 12.4 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 0,
Mode 4 and Mode 9. Other modes can use either K = 16 or K = 32.
DualLink
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-link mode are
shown in Table 28. In addition to these operating modes, the
modes in Table 28 can also be used when using single-link mode.
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 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.
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.
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
Lane0Checksum can be used for error checking purposes
to ensure that the transmitter is set up as expected. Both
CheckSumMode calculations use the fields contained in
Register 0x450 to Register 0x45A. Select whether to sum by
fields or by registers, matching the setting on the transmitter.
If CheckSumMode = 0, the summation is computed by fields.
The checksum is the lower 8 bits of the sum of the DID,
ADJCNT, BID, ADJDIR, PHADJ, LID, Scrambling, L – 1, F − 1,
K − 1, M − 1, CS, N − 1, Subclass, NP − 1, JESDVer, S − 1, HD,
and CF variables.
If CheckSumMode = 1, the summation is computed by registers.
The checksum is the sum of Register 0x450 to Register 0x45A,
Modulo 256.
DAC Power-Down Setup
As described in the Step 1: Start Up the DAC section, PdDACs
must be set to 0 if all 4 converters are being used. If fewer than
four converters are being used, the unused converters must be
powered down. Table 29 can be used to determine which DACs
are powered down based on the number of converters per link
(M) and whether the device is in DualLink mode.
Table 29. DAC Power-Down Configuration Settings
M (Converters
per link)
1
1
2
2
4
DualLink
0
1
0
1
0
DACs to Power Down
0
1
2
3
0
1
1
1
0
1
0
1
0
0
1
1
0
0
0
0
0
0
0
0
PdDACs
0b0111
0b0101
0b0011
0b0000
0b0000
PdClocks
Lanes = (2L) − 1.
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.
If both DACs in DAC Dual B (DAC2 and DAC3) are powered
down, the clock for DAC Dual B can be powered down. In this
case, PdClocks = 0x40; if not, PdClocks = 0x00.
For example, if using Mode 6 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.
Rev. C | Page 31 of 126
AD9144
Data Sheet
SERDES CLOCKS 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).
Table 30. SERDES Lane Rate Configuration Settings
Lane Rate (Gbps)
1.44 to 3.1
2.88 to 6.2
5.75 to 12.4
Halfrate
0
0
1
OvSmp
1
0
0
PLLDiv
2
1
0
Delays that are not in PClock cycles must be converted before
they are used.
Some useful internal relationships are defined as follows:
PClockPeriod = 40/LaneRate
The PClockPeriod can be used to convert from time to PClock
cycles when needed.
PClockFactor = 4/F (frames per PClock)
The PClockFactor is used to convert from units of PClock
cycles to frame clock cycles, which is needed to set LMFCDel in
Subclass 1.
Halfrate and OvSmp set how the clock detect and recover
(CDR) circuit sample. See the SERDES PLL section for an
explanation of how that circuit blocks works and the role of
PLLDiv in the block.
PClocksPerMF = K/PClockFactor (PClocks per LMFC cycle)
where PClocksPerMF is the number or PClock cycles in a
multiframe cycle.
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.
The values for PClockFactor and PClockPerMF are given per
JESD mode in Table 31 and Table 32.
Table 31. PClockFactor and PClockPerMF per LMFC
JESD Mode ID
PClockFactor
PClockPerMF (K = 32)
PClockPerMF (K = 16)
1
0
4
8
N/A1
1
2
16
8
2
2
16
8
3
1
32
16
N/A means not applicable.
Table 32. PClockFactor and PClockPerMF per LMFC
JESD Mode ID
PClockFactor
PClockPerMF (K = 32)
PClockPerMF (K = 16)
Subclass Setup
1
The AD9144 supports JESD204B Subclass 0 and Subclass 1
operation.
With Known Delays
4
4
8
N/A1
5
2
16
8
6
2
16
8
7
1
32
16
9
4
8
N/A1
10
2
16
8
N/A means not applicable.
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
(the pins can be left disconnected).
Subclass 0 still requires that all lanes arrive within the same
LMFC cycle and that the dual 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
MinDelayLane = floor(RxFixed + TxFixed + PCBFixed)
Set Subclass to 0 or 1 as desired.
MaxDelayLane = ceiling(RxFixed + RxVar + TxFixed +
TxVar + PCBFixed))
Link Delay Setup
LMFCVar and LMFCDel are used to impose delays such that all
lanes in a system arrive in the same LMFC cycle.
The unit used internally for delays is the period of the internal
processing clock (PClock), whose rate is 1/40th the lane rate.
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.
Rev. C | Page 32 of 126
Data Sheet
AD9144
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)
Note that if LMFCVar must be more than 10, the AD9144 is
unable to tolerate the variable delay in the system.
For Subclass 1
LMFCDel = ((MinDelay − 1) × PClockFactor) % K
For Subclass 0
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.
Table 33. Register Configuration and Procedure for OneShot Sync
Addr.
0x301
0x03A
Bit.
No.
Value1
0x
0x01
0x03A
0x81
0x03A
0xC1
Variable
Subclass
SYSREF±
Signal
0x300
Without Known Delays
If comprehensive delay information is not available or known,
the AD9144 can read back the link latency between the LMFCRX
and the last arriving LMFC boundary in PClock cycles. This
information is then used to calculate LMFCVar and LMFCDel.
0x
6
CheckSumMode
3
DualLink
2
CurrentLink
[1:0]
EnLinks
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 33. 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
Set to 0 to access
Link 0 status or 1
for Link 1 status
readbacks. See the
JESD204B Setup
section.
EnLinks = 3 if in
DualLink mode to
enable Link 0 and
Link 1; EnLinks = 1 if
not in DualLink mode
to enable Link 0
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, meaning the set of recorded Delay values might 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. C | Page 33 of 126
AD9144
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 AD9144 is
unable to tolerate the variable delay in the system.
For Subclass 1
LMFCDel = ((MinDelay − 1) × PClockFactor) % K
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 34. 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 obtain 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±). If instead the user wants to use
SERDIN4± as the source for Logical Lane 0, the user must write
LOGICAL_LANE0_SRC = 4.
Rev. C | Page 34 of 126
Data Sheet
AD9144
JESD204B SERIAL DATA INTERFACE
JESD204B OVERVIEW
Only certain combinations of parameters are supported. Each
supported combination is called a mode. In total, there are 10
single-link modes supported by the AD9144, as described in
Table 35. In dual-link mode, there are six supported modes, as
described in Table 36. Each of these tables shows the associated
clock rates when the lane rate is 10 Gbps.
The AD9144 has 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 (duallink mode) that share a single system reference (SYSREF±) and
device clock (CLK±).
For a particular application, the number of converters to use
(M) and the DataRate are known. The LaneRate and number of
lanes (L) can be traded off as follows:
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 35 shows the communication layers
implemented in the AD9144 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.
DataRate = (DACRate)/(InterpolationFactor)
LaneRate = (20 × DataRate × M)/L
where LaneRate must be between 1.44 Gbps and 12.4 Gbps.
Achieving and recovering synchronization of the lanes is very
important. To simplify the interface to the transmitter, the
AD9144 designates 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.
The physical layer is responsible for establishing a reliable channel
between the transmitter and the receiver, the data link layer is
responsible for unpacking the data into octets and descrambling
the data, and the transport layer receives the descrambled
JESD204B frames and converts them to DAC samples.
There are a number of JESD204B parameters (L, F, K, M, N, NP,
S, HD, and Scrambling) that define how the data is packed and
tell the device how to turn the serial data into samples. These
parameters are defined in detail in the Transport Layer section.
SYNCOUT0±
SYNCOUT1±
DATA LINK
LAYER
PHYSICAL
LAYER
SERDIN0±
TRANSPORT
LAYER
DESERIALIZER
TO
DAC
FRAME TO
SAMPLES
QBD/
DESCRAMBLER
SERDIN7±
DUAL A I DATA[15:0]
DUAL A Q DATA[15:0]
DUAL B I DATA[15:0]
DESERIALIZER
11675-004
DUAL B Q DATA[15:0]
SYSREF±
Figure 35. Functional Block Diagram of Serial Link Receiver
Table 35. Single-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 (MHz)
Frame Clock (MHz)
Sample Clock (MHz)
0
4
8
1
1
1
4
8
2
2
2
4
4
1
2
3
4
2
1
4
4
2
4
1
1
250
1000
1000
250
500
1000
250
500
500
250
250
250
250
1000
1000
Rev. C | Page 35 of 126
Mode
5
2
4
2
2
250
500
1000
6
2
2
1
2
7
2
1
1
4
9
1
2
1
1
10
1
1
1
2
250
500
500
250
250
250
250
1000
1000
250
500
500
AD9144
Data Sheet
Table 36. Dual-Link JESD204B Operating Modes for Link 0 and Link 1
Mode
Parameter
M (Converter Counts)
L (Lane Counts)
S (Samples per Converter per Frame)
F (Octets/Frame per Lane)
Example Clock for 10 Gbps Lane Rate
PClock (MHz)
Frame Clock (MHz)
Sample Clock (MHz)
4
2
4
1
1
5
2
4
2
2
6
2
2
1
2
7
2
1
1
4
9
1
2
1
1
10
1
1
1
2
250
1000
1000
250
500
1000
250
500
500
250
250
250
250
1000
1000
250
500
500
PHYSICAL LAYER
Table 37. PHY Termination Autocalibration Routine
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 36).
Address
0x2AA
0x2AB
0x2B1
0x2B2
0x2A7
0x2AE
TERMINATION
EQUALIZER
CDR
1:40
11675-006
SPI
CONTROL
FROM SERDES PLL
Figure 36. Deserializer Block Diagram
JESD204B data is input to the AD9144 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.
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 23, Pin 40, and Pin 43). 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 AD9144 complies with the JESD204B specification
regarding the receiver eye mask and is capable of capturing data
that complies with this mask. Figure 37 shows the receiver eye
mask normalized to the data rate interval with a 600 mV VTT
swing. See the JESD204B specification for more information
regarding the eye mask and permitted receiver eye opening.
The AD9144 autocalibrates the input termination to 50 Ω.
Before running the termination calibration, Register 0x2AA,
Register 0x2AB, Register 0x2B1, and Register 0x2B2 must be
written as described in Table 37 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 as shown in
Table 37.
Rev. C | Page 36 of 126
LV-OIF-11G-SR RECEIVER EYE MASK
525
55
0
–55
–525
0
0.35
0.5
0.65
TIME (UI)
Figure 37. Receiver Eye Mask
1.00
11675-007
SERDINx±
AMPLITUDE (mV)
DESERIALIZER
Value
0xB7
0x87
0xB7
0x87
0x01
0x01
Data Sheet
AD9144
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 38, then enable the SERDES PLL by writing
Register 0x280[0] to 1.
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 hit 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
This comes from 8-bit/10-bit encoding, where each byte is
represented by 10 bits.
PClockRate = ByteRate/4
SERDES PLL Fixed Register Writes
The processing clock is used for a quad-byte decoder.
To optimize the SERDES PLL across all operating conditions,
the register writes in Table 39 are recommended.
FrameRate = ByteRate/F
where F is defined as (bytes per frame) per lane.
Table 39. SERDES PLL Fixed Register Writes
PClockFactor = FrameRate/PClockRate = 4/F
where:
M is the JESD204B parameter for converters per link.
L is the JESD204B parameter for lanes per link.
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 12.4 GHz.
In the SERDES PLL, a VCO divider block divides the VCO
clock by 2 to generate a 2.825 GHz to 6.2 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 = 1/40 of the lane rate = PClockRate. This clock
is divided by a DivFactor to deliver a clock to the PFD block
that is between 35 MHz and 80 MHz. Table 38 includes the
respective SERDES_PLL_DIV_MODE register settings for each
of the desired DivFactor options available.
Register
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 38. SERDES PLL Divider Settings
LaneRate (Gbps)
1.44 to 3.1
2.88 to 6.2
5.75 to 12.4
Register
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. C | Page 37 of 126
AD9144
Data Sheet
2.825GHz TO 6.2GHz
OUTPUT
VCO
LDO
CHARGE
PUMP
I Q
PFD
80MHz
MAX
fREF
BIT RATE ÷ 40
DivFactor
(1, 2, 4)
C1
R1
UP
C2
C3
LC VCO
5.65GHz TO 12.4GHz
÷2
DOWN
÷80
R3
ALC CAL
CAL CONTROL BITS
11675-011
FO CAL
3.2mA
Figure 38. 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
6.2 GHz output from the SERDES PLL, shown in Figure 38,
is the input to the CDR.
To compensate for signal integrity distortions for each PHY
channel due to PCB trace length and impedance, the AD9144
employs an easy to use, low power equalizer on each JESD204B
channel. The AD9144 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 39 as an overlay to the JESD204B specification for
insertion loss. Figure 39 shows the equalization performance at
10.0 Gbps, near the maximum baud rate for the AD9144.
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 and
enable 2× oversampling to recover the appropriate lane rate clock.
Table 40 gives a breakdown of CDR sampling settings that must
be set dependent on the LaneRate.
Table 40. CDR Operating Modes
LaneRate (Gbps)
1.44 to 3.1
2.88 to 6.2
5.75 to 12.4
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 40 and Figure 41 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 lower power mode (shown in Figure 39).
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 39), use normal mode. At 10 Gbps operation, the EQ in
normal mode consumes about 4 mW more power per lane used
than in low power EQ mode. Note that either mode can be used
in conjunction with transmitter preemphasis to ensure
functionality and/or to optimize for power.
Rev. C | Page 38 of 126
Data Sheet
AD9144
DATA LINK LAYER
0
JESD204B SPEC ALLOWED
CHANNEL LOSS
2
EXAMPLE OF
JESD204B
COMPLIANT
CHANNE L
INSERTION LOSS (dB)
4
6
AD9144 ALLOWED
CHANNEL LOSS
(LOW POWER MODE)
8
EXAMPLE OF
AD9144
COMPATIBLE
CHANNEL (LOW
POWER MODE)
10
12
14
EXAMPLE OF
AD9144
COMPATIBLE
CHANNEL
(NORMAL MODE)
AD9144 ALLOWED
CHANNEL LOSS
(NORMAL MODE)
16
The data link layer of the AD9144 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 42 shows the link mode block
diagrams for single-link and dual-link configurations and the
interaction between the physical layer and logical layer. The
logical lanes and DACs can only be configured in sequential
order; for example in Mode 10, when in single-link mode, the
AD9144 only uses Logical Lane 0 and DAC0, and in dual-link
mode, only uses Logical Lane 0, Logical Lane 1 and DAC0,
DAC1. 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 43. The
data link layer consists of a synchronization FIFO for each lane,
a crossbar switch, a deframer, and descrambler.
18
20
22
2.5
5.0
11675-339
24
7.5
FREQUENCY (GHz)
Figure 39. Insertion Loss Allowed
0
The AD9144 can operate as a single-link or dual-link high
speed JESD204B serial data interface. When operating in duallink 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
0
1
2
3
4
5
6
7
8
9
10
FREQUENCY (GHz)
11675-010
–35
Figure 40. Insertion Loss of 50 Ω Striplines on FR4
0
–5
–15
–20
6” MICROSTRIP
10” MICROSTRIP
15” MICROSTRIP
20” MICROSTRIP
25” MICROSTRIP
30” MICROSTRIP
–25
–30
–35
–40
0
1
2
3
4
5
6
7
8
9
FREQUENCY (GHz)
10
11675-011
ATTENUATION (dB)
–10
The AD9144 decodes 8-bit/10-bit control characters, allowing
marking of the start and end of the frame and alignment between
serial lanes. Each AD9144 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 AD9144 deactivates the synchronization
request by setting the SYNCOUT0±/SYNCOUT1± signal high
at the next internal LMFC rising edge. Then, it waits 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 44).
Figure 41. Insertion Loss of 50 Ω Microstrips on FR4
Rev. C | Page 39 of 126
AD9144
Data Sheet
PHYSICAL
LAYER (PHY)
LOGICAL
LAYER
CROSSBAR
QBD
SERDIN0±/PHYSICAL LANE 0
LOGICAL LANE 0/LINK 0 LANE 0
SERDIN1±/PHYSICAL LANE 1
LOGICAL LANE 1/LINK 0 LANE 1
DAC CORE
SINGLE-LINK MODE
DAC0
LOGICAL LANE 2/LINK 0 LANE 2
SERDIN2±/PHYSICAL LANE 2
SERDIN3±/PHYSICAL LANE 3
SERDIN4±/PHYSICAL LANE 4
SERDIN5±/PHYSICAL LANE 5
CROSSBAR
REGISTER
0x308
TO
REGISTER
0x30B
DAC1
LOGICAL LANE 3/LINK 0 LANE 3
QUAD-BYTE
DEFRAMER
(QBD0)
LOGICAL LANE 4/LINK 0 LANE 4
DAC2
LOGICAL LANE 5/LINK 0 LANE 5
SERDIN6±/PHYSICAL LANE 6
LOGICAL LANE 6/LINK 0 LANE 6
SERDIN7±/PHYSICAL LANE 7
LOGICAL LANE 7/LINK 0 LANE 7
DAC3
LOGICAL LANES
PHYSICAL LANES
SERDIN0±/PHYSICAL LANE 0
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
CROSSBAR
SERDIN3±/PHYSICAL LANE 3
SERDIN4±/PHYSICAL LANE 4
SERDIN5±/PHYSICAL LANE 5
REGISTER
0x308
TO
REGISTER
0x30B
QUAD-BYTE
DEFRAMER 0
(QBD0)
DAC1
LOGICAL LANE 3/LINK 0 LANE 3
LOGICAL LANE 4/LINK 1 LANE 0
DAC2
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
QUAD-BYTE
DEFRAMER 1
(QBD1)
DAC3
11675-442
LOGICAL LANES
PHYSICAL LANES
Figure 42. Link Mode Functional Diagram
DATA LINK LAYER
SYNCOUTx±
LANE 0 DATA CLOCK
SERDIN0
FIFO
CROSS
BAR
SWITCH
LANE 7 DESERIALIZED
AND DESCRAMBLED DATA
SYSREF
SERDIN7
FIFO
LANE0 OCTETS
LANE7 OCTETS
SYSTEM CLOCK
PHASE DETECT
11675-012
LANE 7 DATA CLOCK
DESCRAMBLE
QUAD-BYTE
DEFRAMER
QBD
LANE 0 DESERIALIZED
AND DESCRAMBLED DATA
10-BIT/8-BIT DECODE
DUAL-LINK MODE
DAC0
PCLK
SPI CONTROL
Figure 43. Data Link Layer Block Diagram
Rev. C | Page 40 of 126
Data Sheet
AD9144
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
11675-013
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 44. 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 local multiframe clock
(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).
Optionally, data can be scrambled. Scrambling does not start
until the very first octet following the ILAS.
The receiver block processes and monitors the data it receives
for errors, including:
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.
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).
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 AD9144 does 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 AD9144 uses four
multiframes in the ILAS (this can be changed in Register 0x478).
If using Subclass 1, exactly four multiframes must be used.
Step 2: Initial Lane Alignment Sequence
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)
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. C | Page 41 of 126
AD9144
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 41. 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]
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 is being configured. 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 discussed 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 AD9144 provides 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
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
The first step in guaranteeing synchronization across links and
devices begins with syncing the LMFC signals. Each DAC dual
(DAC Dual A: DAC0/DAC1 and DAC Dual B: DAC2/DAC3)
has its own LMFC signal. In Subclass 0, the LMFC signals for
each of the two links are synchronized to an internal processing
clock. In Subclass 1, all LMFC signals (for all duals and devices)
are synchronized to an external SYSREF signal. All LMFC sync
registers are paged as described in the Dual Paging section.
SYSREF Signal
Write each LOGICAL_LANEy_SRC with the number (x) of the
desired physical lane (SERDINx±) from which to obtain data.
By default, all logical lanes use the corresponding physical lane
as their data source. For example, by default LOGICAL_
LANE0_SRC = 0; therefore, Logical Lane 0 obtains 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 AD9144 consists of two quad-byte deframers (QBDs). Each
deframer takes in 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 AD9144 supports 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 ( 4/SYSREF Freq.
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 keep out window (KOW) requirements.
It is possible to use ac-coupled mode without meeting the
frequency to time-constant constraint mentioned 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.
1.2V
3kΩ
SYSREF–
3kΩ
~600mV
11675-015
SYSREF+
Figure 45. SYSREF± Input Circuit
Sync Processing Modes Overview
The AD9144 supports 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 must only be used in Subclass 1 with a periodic
SYSREF± signal. In continuous mode, a phase check/alignment
occurs on every alignment edge.
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.
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. If the error window
tolerance must be set above 3, Subclass 0 with a one-shot sync is
recommended.
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, CURRERR = 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).
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)
Rev. C | Page 43 of 126
AD9144
Data Sheet
Sync Procedure
LMFC Sync IRQ
The procedure for enabling the sync is as follows:
The sync status bits (SYNCLOCK, SYNCROTATE, SYNCTRIP,
and SYNCWLIM) are available as IRQ events.
1.
2.
3.
4.
5.
6.
7.
Set Register 0x008 to 0x03 to sync the LMFC for both
duals (DAC0/DAC1 and DAC2/DAC3).
Set the desired sync processing mode. The sync processing
mode settings are listed in Table 42.
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 43.
Enable sync by writing SYNCENABLE
(Register 0x03A[7] = 1).
If in one-shot mode, arm the sync machine by writing
SYNCARM (Register 0x03A[6] = 1).
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 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
alignment edge received and phase check occurred.
This bit is sticky and can be cleared only by writing to
the SYNCCLRSTKY control bit (Register 0x03A[5]).
Table 42. Sync Processing Modes
Sync Processing Mode
One-shot
Continuous
One-shot then monitor
SYNCMODE (Register 0x03A[3:0])
0x01
0x02
0x09
Table 43. Sync Window Tolerance
Sync Error Window
Tolerance
±½ DAC clock cycles
±1 DAC clock cycles
±2 DAC clock cycles
±3 DAC clock cycles
Use Register 0x021[3:0] to enable the sync status bits for DAC
Dual A (DAC0 and DAC1), and then use Register 0x025[3:0] to
read back their statuses and reset the IRQ signals.
Use Register 0x022[3:0] to enable the sync status bits for DAC
Dual B (DAC2 and DAC3), and then use Register 0x026[3:0]
read back their statuses and 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 AD9144 supports 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 the dual DACs must be synchronized to each
other.
Minor Subclass 0 Caveats
Because the AD9144 requires an ILAS, the nonmultiple
converter single lane (NMCDA-SL) case from the JESD204A
specification is only supported 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.
ERRWINDOW (Register 0x034[2:0])
0x00
0x01
0x02
0x03
Rev. C | Page 44 of 126
Data Sheet
AD9144
Deterministic Latency Requirements
Because the LMFC is periodic, this can 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 AD9144 can achieve
proper performance with a smaller total latency. Figure 46 and
Figure 47 show a case where the link delay is larger than an
LMFC period. Note that it can be accommodated by delaying
LMFCRx.
Several key factors are required for achieving deterministic
latency in a JESD204B Subclass 1 system.
POWER CYCLE
VARIANCE
LMFC
ILAS
ALIGNED DATA
DATA
LATE ARRIVING
LMFC REFERENCE
EARLY ARRIVING
LMFC REFERENCE
Link Delay
Figure 46. Link Delay > LMFC Period Example
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 48.
POWER CYCLE
VARIANCE
LMFC
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 AD9144, this is not necessarily
the case; instead, the AD9144 uses a local LMFC for each link
(LMFCRx) that can be delayed from the SYSREF aligned LMFC.
ILAS
ALIGNED DATA
DATA
LMFCRX
LMFC REFERENCE FOR ALL POWER CYCLES
LMFC_DELAY
FRAME CLOCK
Figure 47. LMFC_DELAY to Compensate for Link Delay > LMFC
LINK DELAY = DELAYFIXED + DELAYVARIABLE
LOGIC DEVICE
(JESD204B Tx)
CHANNEL
JESD204B Rx
DSP
DAC
POWER CYCLE
VARIANCE
LMFC
DATA AT
Tx INPUT
ALIGNED DATA
AT Rx OUTPUT
11675-018
•
ILAS
DATA
ILAS
DATA
FIXED DELAY
VARIABLE
DELAY
Figure 48. JESD204B Link Delay = Fixed Delay + Variable Delay
Rev. C | Page 45 of 126
11675-019
•
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.
11675-017
•
AD9144
Data Sheet
1.
The method for setting the LMFCDel and LMFCVar 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 to K frame clock cycles, while the RBD of the AD9144
takes values from 0 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.
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.
3.
4.
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.
The example shown in Figure 49 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 2 frameclock cycles per PClock cycle, and uses K = 32
(frames/multiframe). Because PCBFixed 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
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
Rev. C | Page 101 of 126
Reset
0x0
Access
R
0x0
0x0
R
R
0x0
R
0x0
R
0x0
R
0x0
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
AD9144
Address
0x03B
Name
SYNC_STATUS
Data Sheet
Bit No.
7
Bit Name
SYNC_BUSY
[6:4]
3
RESERVED
SYNC_LOCK
2
SYNC_ROTATE
Settings
1
1
1
1
SYNC_WLIM
1
0
SYNC_TRIP
1
0x03C
SYNC_CURRERR_L
[7:4]
[3:0]
RESERVED
CURRERROR
0x03D
SYNC_CURRERR_H
7
CURRUNDER
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]
0x043
0x044
DACGAIN1_0
DACGAIN2_1
[7:0]
[7:2]
[1:0]
DACFSC_1[7:0]
RESERVED
DACFSC_2[9:8]
0x045
DACGAIN2_0
[7:0]
DACFSC_2[7:0]
0x040
DACGAIN0_1
0x041
0x042
Description
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)
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 Dual A. A 10bit twos complement value that is mapped
to analog full-scale current for DAC 0 as
shown:
01111111111 = 27.0 mA
0000000000 = 20.48 mA
1000000000 = 13.9 mA
8 LSBs of I-Channel DAC Gain Dual A.
Reserved.
2 MSBs of Q-Channel DAC Gain Dual A. A 10bit twos complement value that is mapped
to analog full-scale current for DAC 1 as
shown in Register 0x040.
01111111111 = 27.0 mA
0000000000 = 20.48 mA
1000000000 = 13.9 mA
8 LSBs of Q-Channel DAC Gain Dual A.
Reserved.
2 MSBs of I-Channel DAC Gain Dual B. A 10bit 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 I-Channel DAC Gain Dual B.
Rev. C | Page 102 of 126
Reset
0x0
Access
R
0x0
0x0
R
R
0x0
R
0x0
R
0x0
R
0x0
0x0
R
R
0x0
R
0x0
R
0x0
0x0
0x0
R
R
R/W
0x0
0x0
0x0
R/W
R
R/W
0x0
0x0
0x0
R/W
R
R/W
0x0
R/W
Data Sheet
AD9144
Address
0x046
Name
DACGAIN3_1
Bit No.
[7:2]
[1:0]
Bit Name
RESERVED
DACFSC_3[9:8]
Settings
0x047
0x050
DACGAIN3_0
NCOALIGN_MODE
[7:0]
7
DACFSC_3[7:0]
NCO_ALIGN_ARM
6
5
RESERVED
NCO_ALIGN_
MTCH
Description
Reserved.
2 MSBs of Q-Channel DAC Gain Dual B. A 10bit twos complement value that is mapped
to analog full-scale current for DAC 3 as
shown in Register 0x40.
01111111111 = 27.0 mA
0000000000 = 20.48 mA
1000000000 = 13.9 mA
8 LSBs of Q-Channel DAC Gain Dual B.
Arm NCO Align. On a rising edge, arms the
NCO align operation.
Reserved.
NCO Align Data Match.
1
0
4
NCO_ALIGN_PASS
1
0
3
NCO_ALIGN_FAIL
1
0
2
[1:0]
RESERVED
NCO_ALIGN_
MODE
00
10
01
0x051
0x052
0x053
0x054
0x060
NCOKEY_ILSB
NCOKEY_IMSB
NCOKEY_QLSB
NCOKEY_QMSB
PDP_THRES0
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
NCOKEYI[7:0]
NCOKEYI[15:8]
NCOKEYQ[7:0]
NCOKEYQ[15:8]
PDP_THRESHOLD[7:0]
0x061
PDP_THRES1
[7:5]
[4:0]
0x062
PDP_AVG_TIME
7
[6:4]
[3:0]
RESERVED
PDP_
THRESHOLD[12:8]
PDP_ENABLE
RESERVED
PDP_AVG_TIME
0x063
PDP_POWER0
[7:0]
PDP_POWER[7:0]
0x064
PDP_POWER1
[7:5]
[4:0]
RESERVED
PDP_POWER[12:8]
1
Key NCO align data match
If finished, NCO not aligned on data match
NCO Align Pass.
NCO align takes effect
Clear not taken effect yet
NCO Align Fail.
NCO reset during rotate
Not finished yet
Reserved.
NCO Align Mode.
NCO align disabled
NCO align on data key
NCO align on SYSREF
NCO Data Key for I Channel.
NCO Data Key for I Channel.
NCO Data Key for Q Channel.
NCO Data Key for Q Channel.
PDP_THRESHOLD is the average power
threshold for comparison. If the moving
average of signal power crosses this
threshold, PDP_PROTECT is set high.
Reserved.
See Register 0x60.
Enable average power calculation.
Reserved.
Can be set from 0-10. Averages across
2(9 + PDP_AVG_TIME) IQ sample pairs.
If PDP_POWER has not gone over PDP_
THRESHOLD, PDP_POWER reads back the
moving average of the signal power (I2 + Q2).
If PDP_THRESHOLD is crossed, PDP_POWER
will hold the max value until its corresponding
IRQ is cleared (0x025[7 or 0x026[7]). Only 6
data MSBs are used in calculating power.
Reserved.
See Register 0x063.
Rev. C | Page 103 of 126
Reset
0x0
0x0
Access
R
R/W
0x0
0x0
R/W
R/W
0x0
0x0
R
R
0x0
R
0x0
R
0x0
0x0
R
R/W
0x0
0x0
0x0
0x0
0x0
R/W
R/W
R/W
R/W
R/W
0x0
0x0
R
R/W
0x0
0x0
0x0
R/W
R
R/W
0x0
R
0x0
0x0
R
R
AD9144
Address
0x080
0x081
Name
CLKCFG0
SYSREF_ACTRL0
Data Sheet
Bit No.
7
Bit Name
PD_CLK01
Settings
6
PD_CLK23
5
PD_CLK_DIG
4
PD_SERDES_PCLK
3
PD_CLK_REC
[2:0]
[7:5]
4
RESERVED
RESERVED
PD_SYSREF
3
HYS_ON
2
SYSREF_RISE
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
0
RESERVED
Description
Power-Down Clock for Dual A. This bit
disables the digital and analog clocks for
Dual A.
Power-Down Clock for Dual B. This bit
disables the digital and analog clocks for
Dual B.
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.
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.
Rev. C | Page 104 of 126
Reset
0x1
Access
R/W
0x1
R/W
0x1
R/W
0x1
R/W
0x1
R/W
0x0
0x0
0x1
R
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
R
Data Sheet
AD9144
Address
0x085
Name
DACINTEGERWORD0
Bit No.
[7:0]
Bit Name
B_COUNT
0x087
DACLOOPFILT1
[7:4]
LF_C2_WORD
[3:0]
LF_C1_WORD
[7:4]
LF_R1_WORD
[3:0]
LF_C3_WORD
7
LF_BYPASS_R3
6
LF_BYPASS_R1
5
LF_BYPASS_C2
4
LF_BYPASS_C1
[3:0]
LF_R3_WORD
0x088
0x089
DACLOOPFILT2
DACLOOPFILT3
0x08A
DACCPCNTRL
[7:6]
[5:0]
RESERVED
CP_CURRENT
0x08B
DACLOGENCNTRL
[7:2]
[1:0]
RESERVED
LO_DIV_MODE
Settings
01
10
11
Description
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 DAC PLL
Fixed Register Writes 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.
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 DAC PLL Fixed Register Writes
section for more details):
DAC clock = VCO/4
DAC clock = VCO/8
DAC clock = VCO/16
Rev. C | Page 105 of 126
Reset
0x8
Access
R/W
0x8
R/W
0x8
R/W
0x8
R/W
0x8
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x8
R/W
0x0
0x20
R
R/W
0x0
0x2
R
R/W
AD9144
Address
0x08C
Name
DACLDOCNTRL1
Data Sheet
Bit No.
[7:3]
[2:0]
Bit Name
RESERVED
REF_DIV_MODE
Settings
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
[7:4]
RESERVED
3
CAL_CLK_EN
1
0x0E7
CAL_CLKDIV
1
0
0x0E8
CAL_PAGE
[2:0]
[7:4]
[3:0]
RESERVED
RESERVED
CAL_PAGE
0x0E9
CAL_CTRL
7
CAL_FIN
6
CAL_ACTIVE
1
1
Description
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 DAC PLL
Fixed Register Writes 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
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 x to include DAC x.
Readback: only one DAC at a time can be
accessed when reading back CAL_CTRL
(Register 0x0E9). Write a 1 to Bit x to read
from DAC x (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
Rev. C | Page 106 of 126
Reset
0x0
0x1
Access
R
R/W
0x2B
R/W
0x0
0x0
R
R/W
0x0
R/W
0x3
R/W
0x0
R/W
0x0
0x0
0xF
R
R
R/W
0x0
R
0x0
R
Data Sheet
Address
Name
AD9144
Bit No.
5
Bit Name
CAL_ERRHI
4
CAL_ERRLO
[3:2]
1
RESERVED
CAL_START
Settings
1
1
0
1
0
CAL_EN
0
1
0x0ED
CAL_INIT
[7:0]
CAL_INIT
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
4
PHASE_ADJ_
ENABLE
1
0
[3:2]
MODULATION_TYPE
00
01
10
11
1
SEL_SIDEBAND
0
I_TO_Q
Description
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
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
Enable Phase Compensation.
Enable phase adjust compensation
Disable phase adjust compensation
Selects Type Of Modulation Operation.
No modulation
Fine modulation (uses FTW)
fS/4 coarse modulation
fS/8 coarse modulation
Spectrum Inversion Control. Can only be
used with fine modulation. This causes the
negative sideband to be selected and is
equivalent to changing the sign of FTW.
Send I Data into Q DAC datapath. Occurs at
the end of the digital datapath prior to
entering DACs.
Rev. C | Page 107 of 126
Reset
0x0
Access
R
0x0
R
0x0
0x0
R
R/W
0x0
R/W
0xA6
R/W
0x0
R/W
0x0
0x1
R
R/W
0x0
0x1
R
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R/W
AD9144
Address
0x112
Name
INTERP_MODE
Data Sheet
Bit No.
[7:3]
[2:0]
Bit Name
RESERVED
INTERP_MODE
Settings
000
001
011
100
0x113
NCO_FTW_UPDATE
[7:2]
1
RESERVED
FTW_UPDATE_ACK
0
FTW_UPDATE_REQ
0x114
0x115
0x116
0x117
0x118
0x119
0x11A
FTW0
FTW1
FTW2
FTW3
FTW4
FTW5
NCO_PHASE_
OFFSET0
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
FTW[7:0]
FTW[15:8]
FTW[23:16]
FTW[31:24]
FTW[39:32]
FTW[47:40]
NCO_PHASE_
OFFSET[7:0]
0x11B
[7:0]
0x11C
NCO_PHASE_
OFFSET1
PHASE_ADJ0
[7:0]
NCO_PHASE_
OFFSET[15:8]
PHASE_ADJ[7:0]
0x11D
PHASE_ADJ1
0x11F
TXEN_SM_0
[7:5]
[4:0]
[7:6]
RESERVED
PHASE_ADJ[12:8]
FALL_COUNTERS
[5:4]
RISE_COUNTERS
3
2
RESERVED
PROTECT_OUT_
INVERT
0
1
[1:0]
RESERVED
Description
Reserved.
Interpolation Mode.
1× mode
2× mode
4× mode
8× mode
Reserved.
Frequency tuning word update acknowledge.
This readback is high when an FTW has been
updated.
Frequency tuning word update request from
SPI. Unlike most registers, those relating to
fine NCO modulation (Register 0x114 to
Register 0x11B) are not updated immediately
upon writing to them. Once the desired FTW
and phase offset values are written, set this
bit. These registers update on the rising edge
of this bit. It is only after this update that the
internal state matches Register 0x114 to
Register 0x11B. Confirmation that this
update has occurred can be made by reading
back bit 1 of this register and ensuring it is
set high for the update acknowledge.
NCO Frequency Tuning Word.
NCO Frequency Tuning Word.
NCO Frequency Tuning Word.
NCO Frequency Tuning Word.
NCO Frequency Tuning Word.
NCO Frequency Tuning Word.
8 LSBs of NCO Phase Offset.
NCO_PHASE_OFFSET changes the phase of
both I and Q data, and is only functional
when using NCO fine modulation. It is a 16bit twos complement number ranging from
−180 to+180 degrees in steps of .0055°.
8 MSBs of NCO Phase Offset.
8 LSBs of Phase Compensation Word. Phase
compensation changes the phase between
the I and Q data. PHASE_ADJ is a 13-bit twos
complement value. The control ranges from
−14° to +14° with 0.0035° resolution steps.
Reserved.
5 MSBs of Phase Compensation Word.
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.
Rev. C | Page 108 of 126
Reset
0x0
0x1
Access
R
R/W
0x0
0x0
R
R
0x0
R/W
0x0
0x0
0x0
0x0
0x0
0x10
0x0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
0x0
R/W
0x0
R/W
0x0
0x0
0x2
R
R/W
R/W
0x0
R/W
0x0
0x0
R
R/W
0x3
R/W
Data Sheet
AD9144
Address
0x121
Name
TXEN_RISE_
COUNT_0
Bit No.
[7:0]
Bit Name
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
Settings
0
1
0x132
0x133
0x134
DIE_TEMP0
DIE_TEMP1
DIE_TEMP_UPDATE
[7:0]
[7:0]
[7:1]
0
0x135
DC_OFFSET_CTRL
[7:1]
0
DIE_TEMP[7:0]
DIE_TEMP[15:8]
RESERVED
DIE_TEMP_
UPDATE
RESERVED
DC_OFFSET_ON
0x136
IPATH_DC_OFFSET_
1PART0
[7:0]
LSB_OFFSET_I[7:0]
0x137
IPATH_DC_OFFSET_
1PART1
[7:0]
LSB_OFFSET_I[15:8]
0x138
QPATH_DC_OFFSET_
1PART0
[7:0]
LSB_OFFSET_
Q[7:0]
0x139
QPATH_DC_OFFSET_
1PART1
[7:0]
LSB_OFFSET_
Q[15:8]
0x13A
IPATH_DC_OFFSET_
2PART
[7:5]
RESERVED
[4:0]
SIXTEENTH_
OFFSET_I
[7:5]
RESERVED
[4:0]
SIXTEENTH_
OFFSET_Q
1
x
0x13B
QPATH_DC_OFFSET_
2PART
x
0x13C
IDAC_DIG_GAIN0
[7:0]
IDAC_DIG_
GAIN[7:0]
0x13D
IDAC_DIG_GAIN1
[7:4]
[3:0]
RESERVED
IDAC_DIG_
GAIN[11:8]
Description
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
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 IPath DC Offset. LSB_OFFSET_I is a
16-bit twos complement number that is
added to incoming data.
8 MSBs of IPath DC Offset. LSB_OFFSET_I is a
16-bit twos complement number that is
added to incoming I data.
8 LSBs of QPath DC Offset. LSB_OFFSET_Q is a
16-bit twos complement number that is
added to incoming Q data.
8 MSBs of QPath DC Offset. LSB_OFFSET_Q is
a 16-bit twos complement number that is
added to incoming Q data.
Reserved.
Reset
0xF
Access
R/W
0x0
R/W
0xFF
R/W
0xFF
R/W
0x46
R/W
0x10
R/W
0x0
R/W
0x0
0x0
0x0
0x0
R
R
R
R/W
0x0
0x0
R
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R
SIXTEENTH_OFFSET_I is a 5-bit twos
complement number in 16ths of an LSB that
is added to incoming I data.
x/16 LSB DC offset
Reserved.
0x0
R/W
0x0
R
SIXTEENTH_OFFSET_Q is a 5-bit twos
complement number in 16ths of an LSB that
is added to incoming Q data.
x/16 LSB DC offset
8 LSBs of I DAC Digital Gain. IDAC_DIG_GAIN
is the digital gain of the IDAC. The digital
gain is a multiplier from 0 to 4095/2048 in
steps of 1/2048.
Reserved.
4 MSBs of I DAC Digital Gain
0x0
R/W
0xEA
R/W
0x0
0xA
R
R/W
Rev. C | Page 109 of 126
AD9144
Data Sheet
Address
0x13E
Name
QDAC_DIG_GAIN0
Bit No.
[7:0]
Bit Name
QDAC_DIG_
GAIN[7:0]
Settings
0x13F
QDAC_DIG_GAIN1
[7:4]
[3:0]
0x140
GAIN_RAMP_UP_
STEP0
[7:0]
RESERVED
QDAC_DIG_
GAIN[11:8]
GAIN_RAMP_UP_
STEP[7:0]
Description
8 LSBs of Q DAC Digital Gain.
QDAC_DIG_GAIN is the digital gain of the
QDAC. The digital gain is a multiplier from 0
to 4095/2048 in steps of 1/2048.
Reserved.
4 MSBs of Q DAC Digital Gain.
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
0x14B
PRBS
[5:0]
7
RESERVED
PRBS_GOOD_Q
0
1
6
PRBS_GOOD_I
0
1
[5:3]
2
RESERVED
PRBS_MODE
0
1
1
PRBS_RESET
0
1
0
PRBS_EN
0
1
0x14C
0x14D
0x1B0
PRBS_ERROR_I
PRBS_ERROR_Q
DACPLLT0
[7:0]
[7:0]
[7:0]
PRBS_COUNT_I
PRBS_COUNT_Q
DAC_PLL_PWR
Reset
0xEA
Access
R/W
0x0
0xA
R
R/W
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.
0x4
R/W
0x0
R
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.
0x0
R/W
0x9
R/W
0x0
R
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.
Good Data Indicator Imaginary Channel.
Incorrect sequence detected
Correct PRBS sequence detected
Good Data Indicator Real Channel.
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 Real Channel.
Error Count Value Imaginary Channel.
DAC PLL PD settings. This register must be
written to 0x00 for optimal performance.
0x0
R/W
0x0
R/W
0x0
R
0x0
0x0
R
R
0x0
R
0x0
0x0
R
R/W
0x0
R/W
0x0
R/W
0x0
0x0
0xFA
R
R
R/W
Rev. C | Page 110 of 126
Data Sheet
Address
0x1B5
Name
DACPLLT5
AD9144
Bit No.
[7:4]
Bit Name
RESERVED
[3:0]
VCO_VAR
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
0x203
GENERIC_PD
[7:2]
1
RESERVED
SPI_SYNC1_PD
Description
Must write the default value for proper
operation.
Varactor KVO Setting. See Table 83 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 83 for optimal settings based on the
fVCO being used.
VCO Bias Control. See Table 83 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 83 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
Reserved.
Power down LVDS buffer for SYNCOUT0±.
0
SPI_SYNC2_PD
Power down LVDS buffer for SYNCOUT1±.
0x0
R/W
0x206
CDR_RESET
[7:1]
0
RESERVED
SPI_CDR_RESETN
Reserved.
Resets the Digital Control Logic for All PHYs.
Hold CDR in reset
Enable CDR
Reserved.
0x0
0x1
R
R/W
0x0
R
Enables Half-Rate CDR Operation. Set to 1
when 5.75 Gbps ≤ lane rate ≤ 12.4 Gbps.
Must write the default value for proper
operation.
Enables Oversampling of the Input Data. Set
to 1 when 1.44 Gbps ≤ lane rate ≤ 3.1 Gbps.
Reserved.
Must be set to 0xFF for proper JESD interface
configuration.
0x1
R/W
0x2
R/W
0x0
R/W
0x0
0x0
R
R/W
0x1B9
DACPLLT9
[7:0]
DAC_PLL_CP1
0x1BB
DACPLLTB
[7:5]
[4:3]
RESERVED
VCO_BIAS_TCF
[2:0]
VCO_BIAS_REF
Settings
0
1
0x230
0x232
CDR_OPERATING_
MODE_REG_0
DEVICE_CONFIG_
REG_3
[7:6]
RESERVED
5
ENHALFRATE
[4:2]
RESERVED
1
CDR_OVERSAMP
0
[7:0]
RESERVED
DEVICE_CONFIG_3
Rev. C | Page 111 of 126
Reset
0x8
Access
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
0x0
0x0
R
R/W
AD9144
Address
0x268
Name
EQ_BIAS_REG
Data Sheet
Bit No.
[7:6]
Bit Name
EQ_POWER_
MODE
Settings
00
01
0x280
0x281
SERDESPLL_
ENABLE_CNTRL
PLL_STATUS
[5:0]
RESERVED
[7:3]
RESERVED
2
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
0x289
REF_CLK_DIVIDER_
LDO
[7:3]
RESERVED
2
DEVICE_CONFIG_4
[1:0]
SERDES_PLL_DIV_
MODE
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
Description
Control the Equalizer Power/Insertion Loss
Capability.
Normal mode
Low power mode
Must write the default value for proper
operation.
Reserved.
Reset
0x1
Access
R/W
0x22
R/W
0x0
R
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.
Reserved.
0x0
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
0x0
R
0x0
R/W
0x0
R/W
0x2B
R/W
0x7F
R/W
0x83
R/W
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 12.4 Gbps lane
rate
Divide by 2 for 2.88 Gbps to 6.2 Gbps lane rate
Divide by 1 for 1.44 Gbps to 3.1 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.
Rev. C | Page 112 of 126
Data Sheet
AD9144
Address
0x294
Name
SERDES_PLL_CP2
Bit No.
[7:0]
Bit Name
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
Settings
0
1
0x2A7
TERM_BLK1_
CTRLREG0
[7:1]
0
RESERVED
RCAL_TERMBLK1
0x2AA
DEVICE_CONFIG_
REG_9
DEVICE_CONFIG_
REG_10
TERM_BLK2_
CTRLREG0
[7:0]
[7:1]
DEVICE_CONFIG_
9
DEVICE_CONFIG_
10
RESERVED
0
RCAL_TERMBLK2
[7:0]
DEVICE_CONFIG_
11
DEVICE_CONFIG_
12
RESERVED
CHECKSUM_MODE
0x2AB
0x2AE
0x2B1
0x2B2
0x300
DEVICE_CONFIG_
REG_11
DEVICE_CONFIG_
REG_12
GENERAL_JRX_
CTRL_0
[7:0]
[7:0]
7
6
0
1
[5:4]
RESERVED
Description
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.
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.
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[0x450:0x45A] modulo 256).
Reserved.
Rev. C | Page 113 of 126
Reset
0xB0
Access
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
0x0
R
R/W
0xC3
R/W
0x93
R/W
0x0
R
0x0
R/W
0xC3
R/W
0x93
R/W
0x0
0x0
R
R/W
0x0
R
AD9144
Address
Name
Data Sheet
Bit No.
3
Bit Name
LINK_MODE
Settings
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
0x304
LMFC_DELAY_0
[7:5]
[4:0]
RESERVED
LMFC_DELAY_0
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
Description
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.
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.
Rev. C | Page 114 of 126
Reset
0x0
Access
R/W
0x0
R/W
0x0
R/W
0x0
0x1
R
R/W
0x0
0x0
R
R
0x0
0x0
R
R
0x0
0x0
R
R/W
0x0
0x0
R
R/W
0x0
0x6
R
R/W
0x0
0x6
R
R/W
Data Sheet
Address
0x308
Name
XBAR_LN_0_1
AD9144
Bit No.
[7:6]
[5:3]
Bit Name
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
Settings
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
[7:0]
LANE_FIFO_FULL
0x30D
FIFO_STATUS_REG_1
[7:0]
LANE_FIFO_EMPTY
0x312
SYNCB_GEN_1
[7:6]
[5:4]
RESERVED
SYNCB_ERR _DUR
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
Description
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.
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.
Rev. C | Page 115 of 126
Reset
0x0
0x1
Access
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
0x0
R
0x0
R/W
0x0
0x0
R/W
R/W
0x0
R/W
AD9144
Address
0x316
Name
PHY_PRBS_TEST_CTRL
Data Sheet
Bit No.
7
[6:4]
Bit Name
RESERVED
PHY_SRC_ERR_CNT
[3:2]
PHY_PRBS_PAT_SEL
Settings
x
00
01
10
1
PHY_TEST_START
0
1
0
PHY_TEST_RESET
0
1
0x317
0x318
0x319
0x31A
0x31B
0x31C
0x31D
0x32C
PHY_PRBS_TEST_
THRESHOLD_LOBITS
PHY_PRBS_TEST_
THRESHOLD_
MIDBITS
PHY_PRBS_TEST_
THRESHOLD_HIBITS
PHY_PRBS_TEST_
ERRCNT_LOBITS
PHY_PRBS_TEST_
ERRCNT_MIDBITS
PHY_PRBS_TEST_
ERRCNT_HIBITS
PHY_PRBS_TEST_
STATUS
SHORT_TPL_TEST_0
[7:0]
[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
[7:6]
[5:4]
RESERVED
SHORT_TPL_SP_
SEL
[3:2]
SHORT_TPL_DAC_
SEL
1
SHORT_TPL_TEST_
RESET
x
x
0
1
0
SHORT_TPL_TEST_
EN
0
1
0x32D
SHORT_TPL_TEST_1
[7:0]
SHORT_TPL_REF_
SP_LSB
Description
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
0x0
Access
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.
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 DAC x
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.
0xFF
R
0x0
0x0
R
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R/W
Rev. C | Page 116 of 126
Data Sheet
AD9144
Address
0x32E
Name
SHORT_TPL_TEST_2
Bit No.
[7:0]
Bit Name
SHORT_TPL_REF_
SP_MSB
0x32F
SHORT_TPL_TEST_3
[7:1]
0
RESERVED
SHORT_TPL_FAIL
Settings
0
1
[7:0]
0x400
DEVICE_CONFIG_
REG_13
JESD_BIT_INVERSE_
CTRL
DID_REG
0x401
BID_REG
0x333
0x334
0x402
0x403
LID0_REG
SCR_L_REG
[7:0]
DEVICE_CONFIG_
13
JESD_BIT_INVERSE
[7:0]
DID_RD
[7:4]
ADJCNT_RD
[3:0]
BID_RD
7
6
RESERVED
ADJDIR_RD
5
PHADJ_RD
[4:0]
LID0_RD
7
SCR_RD
0
1
[6:5]
[4:0]
RESERVED
L-1_RD
0
1
3
0x404
F_REG
[7:0]
F-1_RD
0
1
3
Description
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.
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
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
Rev. C | Page 117 of 126
Reset
0x0
Access
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
0x0
R
0x0
0x0
R
R
0x0
R
AD9144
Address
0x405
Name
K_REG
Data Sheet
Bit No.
[7:5]
[4:0]
Bit Name
RESERVED
K-1_RD
Settings
0x0F
0x1F
0x406
M_REG
[7:0]
M-1_RD
0
1
3
0x407
CS_N_REG
[7:6]
CS_RD
5
[4:0]
RESERVED
N-1_RD
[7:5]
SUBCLASSV_RD
[4:0]
NP-1_RD
[7:5]
JESDV_RD
0x0F
0x408
NP_REG
0x0F
0x409
S_REG
000
001
[4:0]
S-1_RD
0
1
0x40A
HD_CF_REG
7
HD_RD
0
1
0x40B
RES1_REG
[6:5]
[4:0]
RESERVED
CF_RD
[7:0]
RES1_RD
Description
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, 1, or 3.
One converter per device
Two converters per device
Four 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.
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. See
Table 35 and Table 36. 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.
Rev. C | Page 118 of 126
Reset
0x0
0x0
Access
R
R
0x0
R
0x0
R
0x0
0x0
R
R
0x0
R
0x0
R
0x0
R
0x0
R
0x0
R
0x0
0x0
R
R
0x0
R
Data Sheet
AD9144
Address
0x40C
Name
RES2_REG
Bit No.
[7:0]
Bit Name
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
0x42A
LID4_REG
0x42D
0x42E
CHECKSUM4_REG
COMPSUM4_REG
[7:5]
[4:0]
[7:0]
[7:0]
RESERVED
LID4_RD
FCHK4_RD
FCMP4_RD
0x432
LID5_REG
0x435
0x436
CHECKSUM5_REG
COMPSUM5_REG
[7:5]
[4:0]
[7:0]
[7:0]
RESERVED
LID5_RD
FCHK5_RD
FCMP5_RD
0x43A
LID6_REG
0x43D
0x43E
CHECKSUM6_REG
COMPSUM6_REG
[7:5]
[4:0]
[7:0]
[7:0]
RESERVED
LID6_RD
FCHK6_RD
FCMP6_RD
0x442
LID7_REG
0x445
0x446
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
Settings
Description
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).
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.
Rev. C | Page 119 of 126
Reset
0x0
Access
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
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
0x0
0x0
0x0
R
R
R
R
0x0
R/W
AD9144
Address
0x451
0x452
0x453
Name
ILS_BID
ILS_LID0
ILS_SCR_L
Data Sheet
Bit No.
[7:4]
Bit Name
ADJCNT
[3:0]
BID
7
6
5
RESERVED
ADJDIR
PHADJ
[4:0]
LID0
7
SCR
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
0x456
ILS_M
[7:0]
M-1
0
1
3
0x457
ILS_CS_N
[7:6]
CS
5
[4:0]
RESERVED
N-1
[7:5]
SUBCLASSV
0
0xF
0x458
ILS_NP
0
1
[4:0]
NP-1
[7:5]
JESDV
0xF
0x459
ILS_S
000
001
[4:0]
S-1
0
1
Description
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 35 and Table 36.
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 35 and Table 36.
(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
Number of Converters per Device. See Table 35
and Table 36.
One converter per link
Two converters per link
Four converters per link (single link only)
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. See
Table 35 and Table 36.
One sample per converter per frame
Two samples per converter per frame
Rev. C | Page 120 of 126
Reset
0x0
Access
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
0x1
R/W
0x0
R/W
0x0
0xF
R
R/W
0x1
R/W
0xF
R/W
0x1
R/W
0x0
R/W
Data Sheet
Address
0x45A
Name
ILS_HD_CF
AD9144
Bit No.
7
Bit Name
HD
Settings
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
0x46D
BADDISPARITY
7
RST_IRQ_DIS
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
Description
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.
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 the NIT error count for Link Lane x
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].
Rev. C | Page 121 of 126
Reset
0x1
Access
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
0x0
W
0x0
W
0x0
W
0x0
0x0
R
W
0x0
R
0x0
W
0x0
W
0x0
W
0x0
0x0
R
W
AD9144
Data Sheet
Address
0x46F
Name
UNEXPECTEDCONTROL_RB
Bit No.
[7:0]
Bit Name
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
Settings
0
1
0x471
FRAMESYNCFLG
[7:0]
FRAMESYNC
0
1
0x472
GOODCHKSUMFLG
[7:0]
GOODCHECKSUM
0
1
0x473
INITLANESYNCFLG
[7:0]
INITIALLANESYNC
0x476
CTRLREG1
[7:0]
F
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
x
Description
Unexpected Control Character Error (UCC).
Bit x is set when the UCC error count for Link
Lane x 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
SYNCOUT and SYSREF 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
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 SYNCOUT and SYSREF± Signal section
and the Deterministic Latency section.
Number of Octets per Frame. Settings of 1, 2,
and 4 are valid. See Table 35 and Table 36.
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.
4x multiframes during ILAS
Rev. C | Page 122 of 126
Reset
0x0
Access
R
0x0
W
0x0
W
0x0
W
0x0
0x0
R
W
0x0
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x1
R/W
0x0
R/W
0x0
0x0
R
R/W
0x0
0x1
R
R/W
Data Sheet
Address
0x47A
Name
IRQVECTOR_MASK
AD9144
Bit No.
7
Bit Name
BADDIS_MASK
Settings
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
5
UCC_FLAG
1
4
3
2
1
0
RESERVED
INITIALLANESYNC_
FLAG
1
BADCHECKSUM_
FLAG
1
FRAMESYNC_
FLAG
1
CODEGRPSYNC_
FLAG
1
Description
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.
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
Rev. C | Page 123 of 126
Reset
0x0
Access
W
0x0
W
0x0
W
0x0
0x0
R
W
0x0
W
0x0
W
0x0
W
0x0
R
0x0
R
0x0
R
0x0
0x0
R
R
0x0
R
0x0
R
0x0
R
AD9144
Address
0x47B
Name
SYNCASSERTIONMASK
Data Sheet
Bit No.
7
Bit Name
BADDIS_S
Settings
1
6
NIT_S
1
5
UCC_S
1
4
CMM
1
3
CMM_ENABLE
1
0
0x47C
ERRORTHRES
[2:0]
[7:0]
RESERVED
ETH
0x47D
LANEENABLE
[7:0]
LANE_ENA
0x47E
RAMP_ENA
[7:1]
0
RESERVED
ENA_RAMP_
CHECK
0
1
0x520
DIG_TEST0
[7:2]
RESERVED
0x521
DC_TEST_VALUEI0
1
0
[7:0]
0x522
DC_TEST_VALUEI1
[7:0]
0x523
DC_TEST_VALUEQ0
[7:0]
0x524
DC_TEST_VALUEQ1
[7:0]
DC_TEST_MODE
RESERVED
DC_TEST_
VALUEI[7:0]
DC_TEST_
VALUEI [15:8]
DC_TEST_
VALUEQ[7:0]
DC_TEST_
VALUEQ[15:8]
Description
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.
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 I DAC.
Reset
0x0
Access
R/W
0x0
R/W
0x0
R/W
0x0
R/W
0x1
R/W
0x0
0xFF
R
R/W
0xF
R/W
0x0
0x0
R
W
0x7
R/W
0x0
0x0
0x0
R/W
R/W
R/W
DC value MSB of DC Test Mode for I DAC.
0x0
R/W
DC value LSB of DC Test Mode for Q DAC.
0x0
R/W
DC value MSB of DC Test Mode for Q DAC.
0x0
R/W
Rev. C | Page 124 of 126
Data Sheet
AD9144
OUTLINE DIMENSIONS
12.10
12.00 SQ
11.90
0.28
0.23
0.18
0.60 MAX
0.60
MAX
88
67
66
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 88. 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
0.50
BSC
7.55
7.40 SQ
7.25
EXPOSED
PAD
0.65
0.55
0.45
22
44
TOP VIEW
0.90
0.85
0.80
PKG-004598
SEATING
PLANE
12° MAX
SIDE VIEW
0.190~0.245 REF
0.70
0.65
0.60
0.50
0.40
0.30
45
0.045
0.025
0.005
COPLANARITY
0.08
23
BOTTOM VIEW
10.50
REF
FOR PROPER CONNECTION OF
THE EXPOSED PAD, REFER TO
THE PIN CONFIGURATION AND
FUNCTION DESCRIPTIONS
SECTION OF THIS DATA SHEET.
COMPLIANT TO JEDEC STANDARDS MO-220
Figure 89. 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. C | Page 125 of 126
1.00
0.90
0.80
0.80
0.70
0.60
11-09-2018-B
11.85
11.75 SQ
11.65
AD9144
Data Sheet
ORDERING GUIDE
Model 1
AD9144BCPZ
AD9144BCPZRL
AD9144BCPAZ
AD9144BCPAZRL
AD9144-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
Package Description
88-Lead LFCSP_VQ
88-Lead LFCSP_VQ
88-Lead LFCSP_VQ (Variable Lead Length)
88-Lead LFCSP_VQ (Variable Lead Length)
FMC Evaluation Board
Z = RoHS Compliant Part.
©2014–2019 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D11675-0-4/19(C)
Rev. C | Page 126 of 126
Package Option
CP-88-6
CP-88-6
CP-88-9
CP-88-9