AD9166-FMC-EBZ [ADI]
DC to 9 GHz Vector Signal Generator;型号: | AD9166-FMC-EBZ |
厂家: | ADI |
描述: | DC to 9 GHz Vector Signal Generator |
文件: | 总138页 (文件大小:2524K) |
中文: | 中文翻译 | 下载: | 下载PDF数据表文档文件 |
DC to 9 GHz Vector Signal Generator
AD9166
Data Sheet
form generation without the need for external bias tees or similar
circuitry, which makes the AD9166 uniquely suited for the most
demanding high speed ultrawideband RF transmit applications.
FEATURES
DC-coupled, 50 Ω matched output
Up to 4.3 dBm output power, −9.5 dBm at 9 GHz
DAC core update rate: 12.0 GSPS (guaranteed minimum) in
2× NRZ mode
The various filter stages enable the AD9166 to be configured for
lower data rates, while maintaining higher DAC clock rates to
ease the filtering requirements and reduce the overall system size,
weight, and power.
Wide analog bandwidth
DC to 9.0 GHz in 2× NRZ mode (12.0 GSPS DAC update rate)
1.0 GHz to 8.0 GHz in mix mode (6.0 GSPS DAC update rate)
DC to 4.5 GHz in NRZ mode (6.0 GSPS DAC update rate)
Power dissipation of 4.88 W in 2× NRZ mode (10 GSPS DAC
update rate)
The data interface receiver consists of up to eight JESD204B
SERDES lanes, each capable of carrying up to 12.5 Gbps. To
enable maximum flexibility, the receiver is fully configurable
according to the data rate, number of SERDES lanes, and lane
mapping required by the JESD204B transmitter.
Bypassable datapath interpolation
2×, 3×, 4×, 6×, 8×, 12×, 16×, 24×
In 2× nonreturn-to-zero (NRZ) mode of operation (with FIR85
enabled), the AD9166 can reconstruct RF carriers from true dc
to the edge of the third Nyquist zone, or an analog bandwidth of
true dc up to 9 GHz.
Instantaneous (complex) signal bandwidth
2.25 GHz with device clock at 5 GHz (2× interpolation)
1.8 GHz with device clock at 6 GHz (3× interpolation)
Fast frequency hopping
Integrated BiCMOS buffer amplifier
In mix mode, the AD9166 can reconstruct RF carriers in the
second and third Nyquist zones while consuming lower power
and maintaining a performance comparable to 2× NRZ mode.
APPLICATIONS
Instrumentation: automated test equipment, electronic test
and measurement, arbitrary waveform generators
Electronic warfare: radars, jammers
Broadband communications systems
Local oscillator drivers
In baseband modes, such as return-to-zero (RZ) and 1× NRZ,
the AD9166 is ideal to reconstruct RF carriers from true dc to
the edge of the first Nyquist zone while consuming lower power
compared to 2× NRZ mode.
The quadrature DDS block can be configured as a digital
upconverter to upconvert I/Q data samples to the desired
location across the spectrum, in all three Nyquist zones.
GENERAL DESCRIPTION
The AD91661 is a high performance, wideband, on-chip vector
signal generator composed of a high speed JESD204B serializer/
deserializer (SERDES) interface, a flexible 16-bit digital datapath, a
inphase/quadrature (I/Q) digital-to-analog converter (DAC)
core, and an integrated differential to single-ended output
buffer amplifier, matched to a 50 Ω load up to 10 GHz.
The DDS also consists of a bank of 32 numerically controlled
oscillators (NCOs), each with its own 32-bit phase accumulator.
When combined with a 100 MHz serial peripheral interface (SPI),
the DDS allows a phase coherent FFH, with a phase settling
time as low as 300 ns.
The DAC core is based on a quad-switch architecture, which is
configurable to increase the effective DAC core update rate of
up to 12.8 GSPS from a 6.4 GHz DAC sampling clock, with an
analog output bandwidth of true dc to 9.0 GHz, typically. The
digital datapath includes multiple interpolation filter stages, a
direct digital synthesizer (DDS) block with multiple
numerically controlled oscillators (NCOs) supporting fast
frequency hopping (FFH), and additional FIR85 and inverse
sinc filter stages to allow flexible spectrum planning.
The AD9166 is configured using a common SPI interface that
monitors the status of all registers. The AD9166 is offered in a
324-ball, 15 mm × 15 mm, 0.8 mm pitch BGA_ED package.
PRODUCT HIGHLIGHTS
1. High dynamic range and signal reconstruction bandwidth
supports RF signal synthesis of up to 9 GHz.
2. Fully supports zero IF and other dc-coupled applications.
3. Up to an eight-lane JESD204B SERDES interface, with
various features to allow flexibility when interfacing to a
JESD204B transmitter.
The differential to single-ended buffer eliminates the need for a
wideband balun, and supports the full analog output bandwidth
of the DAC core. DC coupling the output allows baseband wave-
1 Protected by U.S. Patents 6,842,132 and 7,796,971.
Rev. 0
Document Feedback
Information furnished by Analog Devices is believed to be accurate and reliable. However, no
responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other
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 registeredtrademarks arethe property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781.329.4700
Technical Support
©2020 Analog Devices, Inc. All rights reserved.
www.analog.com
AD9166
Data Sheet
TABLE OF CONTENTS
Features .............................................................................................. 1
JESD204B Overview .................................................................. 29
Physical Layer ............................................................................. 31
Data Link Layer .......................................................................... 34
Transport Layer .......................................................................... 42
JESD204B Test Modes ............................................................... 44
JESD204B Error Monitoring..................................................... 46
Hardware Considerations ......................................................... 48
Main Digital Datapath ................................................................... 49
Data Format ................................................................................ 49
Interpolation Filters ................................................................... 49
Digital Modulation..................................................................... 52
Inverse Sinc ................................................................................. 55
Downstream Protection ............................................................ 55
Datapath PRBS ........................................................................... 56
Datapath PRBS IRQ................................................................... 56
Interrupt Request Operation ........................................................ 57
Interrupt Service Routine.......................................................... 57
Applications Information.............................................................. 58
Hardware Considerations ......................................................... 58
Analog Interface Considerations.................................................. 61
Analog Modes of Operation ..................................................... 61
Clock Input.................................................................................. 62
Shuffle Mode............................................................................... 63
Voltage Reference and Full-Scale Current (FSC)................... 63
Analog Output............................................................................ 64
Temperature Sensors.................................................................. 65
Start-Up Sequence.......................................................................... 67
Register Summary: DAC ............................................................... 70
Register Details: DAC Register Map............................................ 79
Register Summary: Amplifier..................................................... 135
Register Details: Amplifier Register Map.................................. 136
Outline Dimensions..................................................................... 138
Ordering Guide ........................................................................ 138
Applications....................................................................................... 1
General Description......................................................................... 1
Product Highlights ........................................................................... 1
Revision History ............................................................................... 2
Functional Block Diagram .............................................................. 3
Specifications..................................................................................... 4
DC Specifications ......................................................................... 4
Power Supply DC Specifications ................................................ 5
Device Input Clock Rate and DAC Update Rate Specifications
......................................................................................................... 7
JESD204B Interface Specifications............................................. 8
Input Data Rate and Bandwidth Specifications........................ 9
Pipeline Delay and Latency Uncertainty Specifications.......... 9
AC Specifications........................................................................ 10
CMOS Pin Specifications .......................................................... 11
Timing Specifications ................................................................ 12
Absolute Maximum Ratings.......................................................... 14
Reflow Profile.............................................................................. 14
Thermal Management ............................................................... 14
Thermal Resistance .................................................................... 14
ESD Caution................................................................................ 14
Pin Configuration and Function Descriptions........................... 15
Typical Performance Characteristics ........................................... 18
AC Performance (2× NRZ (FIR85) Mode)............................. 18
LTE Performance (2× NRZ (FIR85) Mode) ........................... 23
802.11AC Performance (2× NRZ (FIR85) Mode) ................. 24
Terminology .................................................................................... 25
Theory of Operation ...................................................................... 26
Serial Port Operation ..................................................................... 27
Data Format ................................................................................ 27
Serial Port Pin Descriptions...................................................... 27
Serial Port Options..................................................................... 28
JESD204B Serial Data Interface.................................................... 29
REVISION HISTORY
7/2020—Revision 0: Initial Version
Rev. 0 | Page 2 of 138
Data Sheet
AD9166
FUNCTIONAL BLOCK DIAGRAM
ISET
VREF
RESET
IRQ
SDIO
SDO
CS_AMP/CS_DAC
SCLK
VREF
NRZ RZ MIX
SPI
AD9166
I
CM
FSC
SERDIN0±
FIR85
WIDEBAND
AMP
HB
2×
INV
SINC
NCO
SERDIN7±
JESD
DAC
CORE
RFOUT
SYNCOUT±
SYSREF±
HB
2×
CLOCK
DISTRIBUTION
HB
2×,
4×,
8×
TO JESD
TO DATAPATH
HB
3×
TX_ENABLE
CLK±
NOTES
1. FSC IS FULL-SCALE CURRENT.
2. I
IS THE INPUT COMMON-MODE CURRENT OF THE BUFFER AMPLIFIER.
CM
Figure 1.
Rev. 0 | Page 3 of 138
AD9166
Data Sheet
SPECIFICATIONS
DC SPECIFICATIONS
DAC_2P5_AN = 2.5 V, DAC_1P2_AN = DAC_1P2_CLK = 1.2 V, DAC_N1P2_AN = −1.2 V, DAC_1P2_DIG = 1.2 V, VDD_IO = 2.5 V,
DAC_1P2_SER = 1.2 V, DAC_3P3_SYNC = 3.3 V, AMP_5V_IN = 5.0 V, AMP_3P3_OUT = 3.3 V, AMP_3P3 = 3.3 V, AMP_N5 = −5.0 V,
DAC output full-scale current (IOUTFS) = 40 mA, and TA = −40°C to +85°C, unless otherwise noted. 50 Ω matched output.
Table 1.
Parameter
Test Conditions/Comments
Min
Typ
Max
Unit
RESOLUTION
16
Bits
DAC ANALOG OUTPUT
Power-Up Delay
Gain Error (with Internal Reference)
From DAC output off to enabled
10
−1.7
ns
%
Full-Scale Output Current (IOUTFS
)
Minimum
Maximum
DAC reference current setting resistor (RSET) = 9.76 kΩ
RSET = 9.76 kΩ
7.37
35.8
8
8.57
mA
mA
38.76 41.3
AMPLIFIER ANALOG OUTPUT
Maximum Full-Scale Power
Measured with full-scale output current set to its typical
maximum
DC
9 GHz
4.3
−9.5
dBm
dBm
FIR85 enabled (2× NRZ)
DEVICE CLOCK INPUT (CLK+, CLK−)
Differential Input Power
Common-Mode Voltage
Input Impedance1
Load resistance (RLOAD) = 90 Ω differential on chip
AC-coupled
3 GSPS input clock
−20
0
0.6
90
+10
dBm
V
Ω
Maximum Input Frequency (fCLK
TEMPERATURE SENSOR
Amplifier Sensor Accuracy2
DAC Sensor Accuracy3
ANALOG SUPPLY VOLTAGES
DAC_2P5_AN
DAC_1P2_AN4
DAC_1P2_CLK4
DAC_N1P2_AN
AMP_5V_IN
)
6400
MHz
See Table 3 for more details
°C
°C
±5
±5
2.375
1.14
2.5
1.2
1.2
−1.2
5
3.3
−5
3.3
2.625
1.326
1.326
−1.14
5.25
3.465
−4.75
3.465
V
V
V
V
V
V
V
V
1.14
−1.26
4.75
3.135
−5.25
3.135
AMP_3P3_OUT
AMP_N5
AMP_3P3
DIGITAL SUPPLY VOLTAGES
DAC_1P2_DIG
VDD_IO5
1.14
1.71
1.2
2.5
1.326
3.465
V
V
SERDES SUPPLY VOLTAGES
DAC_1P2_SER
DAC_3P3_SYNC
1.14
3.135
1.2
3.3
1.326
3.465
V
V
1 See the Clock Input section for more details.
2 The temperature sensor of the amplifier is a more accurate representation of TJ, but requires one-point calibration.
3 Do not use the DAC temperature sensor reading to monitor TJ. Use as a reference only.
4 For the lowest noise performance, use a separate power supply filter network for the DAC_1P2_CLK and the DAC_1P2_AN pins.
5 VDD_IO can range from 1.8 V to 3.3 V, with ±5% tolerance.
Rev. 0 | Page 4 of 138
Data Sheet
AD9166
POWER SUPPLY DC SPECIFICATIONS
IOUTFS = 40 mA, TA = −40°C to +85°C, unless otherwise noted. FIR85 is the finite impulse response with 85 dB digital attenuation.
Table 2.
Parameter
Test Conditions/Comments
Min
Typ
Max
Unit
8 LANES, 2× INTERPOLATION (80%), 3 GSPS
Analog Supply Currents
DAC_2P5_AN
NCO on, FIR85 on
53.5
1
239
57
111
255
mA
μA
DAC_1P2_AN
DAC_1P2_CLK
DAC_N1P2_AN
AMP_5V_IN
AMP_3P3_OUT
AMP_N5
mA
mA
mA
mA
mA
mA
−117 −111.1
169.9
182
70.7
209
23
65.1
187.8
21.2
AMP_3P3
Digital Supply Currents
DAC_1P2_DIG
545
2.5
567.4
9.1
611
2.7
627
11
mA
mA
mA
mA
VDD_IO1
VDD_IO = 2.5 V
DAC_1P2_SER
DAC_3P3_SYNC
8 LANES, 6× INTERPOLATION (80%), 3 GSPS
Analog Supply Currents
DAC_2P5_AN
NCO on, FIR85 on
53.5
1.4
mA
μA
DAC_1P2_AN
DAC_1P2_CLK
DAC_N1P2_AN
AMP_5V_IN
AMP_3P3_OUT
AMP_N5
AMP_3P3
238.1
−111.1
169.7
65.0
195.1
21.1
mA
mA
mA
mA
mA
mA
Digital Supply Currents
DAC_1P2_DIG
632.4
0.025
614.2
9.2
mA
mA
mA
mA
VDD_IO1
VDD_IO = 2.5 V
DAC_1P2_SER
DAC_3P3_SYNC
NCO ONLY MODE, 5 GSPS
Analog Supply Currents
DAC_2P5_AN
47.6
0
359
63
109
382
mA
μA
DAC_1P2_AN
DAC_1P2_CLK
DAC_N1P2_AN
AMP_5V_IN
AMP_3P3_OUT
AMP_N5
mA
mA
mA
mA
mA
mA
−120 −104.9
169.7
182
71
216
23
65.1
194.9
21.1
AMP_3P3
Digital Supply Currents
DAC_1P2_DIG
446.9
2.5
3.0
493
2.7
8.5
mA
mA
mA
mA
VDD_IO1
VDD_IO = 2.5 V
DAC_1P2_SER
DAC_3P3_SYNC
8 LANES, 4× INTERPOLATION (80%), 5 GSPS
Analog Supply Currents
DAC_2P5_AN
0.34
0.44
NCO on, FIR85 off (unless otherwise noted for this test)
55.5
0.1
358.9
59
123
382
mA
μA
mA
mA
mA
DAC_1P2_AN
DAC_1P2_CLK
DAC_N1P2_AN
AMP_5V_IN
−126 −120.0
166.8
178
Rev. 0 | Page 5 of 138
AD9166
Data Sheet
Parameter
Test Conditions/Comments
Min
Typ
Max
71
207
23
Unit
mA
mA
mA
AMP_3P3_OUT
AMP_N5
AMP_3P3
65.1
185.9
21.3
Digital Supply Currents
VDD_IO1
DAC_1P2_DIG
VDD_IO = 2.5 V
2.5
2.7
769
819
1044 mA
586
11
mA
mA
mA
NCO on, FIR85 off
NCO off, FIR85 on
NCO on, FIR85 on
705.1
749.1
962.7
541.6
9.2
DAC_1P2_SER
DAC_3P3_SYNC
mA
mA
8 LANES, 4× INTERPOLATION (80%), 5.8 GSPS
Analog Supply Currents
DAC_2P5_AN
NCO on, FIR85 on
53.5
0
57
68
mA
μA
DAC_1P2_AN
DAC_1P2_CLK
DAC_N1P2_AN
AMP_5V_IN
AMP_3P3_OUT
AMP_N5
AMP_3P3
406
430
mA
mA
mA
mA
mA
mA
−117 −111.1
169.6
182
71
216
23
65.0
194.5
21.2
Digital Supply Currents
VDD_IO1
VDD_IO = 2.5 V
2.5
2.7
mA
DAC_1P2_DIG
DAC_1P2_SER
DAC_3P3_SYNC
1090
575.5
9.0
1200 mA
622
11
mA
mA
8 LANES, 3× INTERPOLATION (80%), 4.5 GSPS
Analog Supply Currents
DAC_2P5_AN
NCO on, FIR85 on
53.5
0
330.5
57
68
352
mA
μA
DAC_1P2_AN
DAC_1P2_CLK
DAC_N1P2_AN
AMP_5V_IN
AMP_3P3_OUT
AMP_N5
AMP_3P3
mA
mA
mA
mA
mA
mA
−117 −111.1
169.7
182
71
216
23
65.0
195.0
21.2
Digital Supply Currents
VDD_IO1
VDD_IO = 2.5 V
2.5
mA
DAC_1P2_DIG
DAC_1P2_SER
DAC_3P3_SYNC
1025.1
579.4
9.2
1115 mA
626
11
mA
mA
POWER DISSIPATION
Amplifier, Standalone
DAC, Standalone, 3 GSPS
2× NRZ Mode, 6×, FIR85 Enabled, NCO On
NRZ Mode, 24×, FIR85 Disabled, NCO On
DAC, Standalone, 5 GSPS
NRZ Mode, 8×, FIR85 Disabled, NCO On
NRZ Mode, 16×, FIR85 Disabled, NCO On
DAC, Standalone, 10 GSPS
2× NRZ Mode, 6×, FIR85 Enabled, NCO On
Total, Amplifier and DAC, 10 GSPS
2.33
2.43
W
Using 80%, 3× filter, eight-lane JESD204B
Using 80%, 2× filter, one-lane JESD204B
2.0
1.2
2.21
1.31
W
W
Using 80%, 2× filter, eight-lane JESD204B
Using 80%, 2× filter, eight-lane JESD204B
2.08
1.99
2.30
2.18
W
W
Using 80%, 3× filter, eight-lane JESD204B
Using 80%, 3× filter, eight-lane JESD204B
2.55
4.88
2.85
W
W
1 VDD_IO can range from 1.8 V to 3.3 V, with ±5% tolerance.
Rev. 0 | Page 6 of 138
Data Sheet
AD9166
DEVICE INPUT CLOCK RATE AND DAC UPDATE RATE SPECIFICATIONS
DAC_2P5_AN = 2.5 V, DAC_1P2_AN = DAC_1P2_CLK = 1.2 V, DAC_N1P2_AN = −1.2 V, DAC_1P2_DIG = 1.2 V, VDD_IO = 2.5 V,
DAC_1P2_SER = 1.2 V, DAC_3P3_SYNC = 3.3 V, AMP_5V_IN = 5.0 V, AMP_3P3_OUT = 3.3 V, AMP_3P3 = 3.3 V, AMP_N5 = −5.0 V,
IOUTFS = 40 mA, TA = −40°C to +85°C, unless otherwise noted.
Maximum guaranteed speed using the temperature and voltage conditions shown in Table 3, where DAC_1P2_x includes DAC_1P2_AN,
DAC_1P2_CLK, DAC_1P2_DIG, and DAC_1P2_SER. Any device clock speed over 5.1 GHz requires a maximum junction temperature not
exceeding 105°C to avoid damage to the device. See Table 11 for details on maximum junction temperature permitted for certain clock speeds.
Table 3.
Parameter
Test Conditions/Comments1
Min
Typ
Max
Unit
MAXIMUM INPUT CLOCK RATE (fCLK
DAC_1P2_x = 1.2 V ± 5%
)
TJ_DAC_MAX = 25°C
TJ_DAC_MAX = 85°C
TJ_DAC_MAX = 105°C
TJ_DAC_MAX = 25°C
TJ_DAC_MAX = 85°C
TJ_DAC_MAX = 105°C
TJ_DAC_MAX = 25°C
TJ_DAC_MAX = 85°C
T_DAC_MAX = 105°C
6.0
5.6
5.4
6.1
5.8
5.6
6.4
6.2
6.0
GHz
GHz
GHz
GHz
GHz
GHz
GHz
GHz
GHz
DAC_1P2_x = 1.2 V ± 2%
DAC_1P2_x = 1.3 V ± 2%
DAC UPDATE RATE (fDAC
)
Minimum
Maximum
1.5
GSPS
GSPS
GSPS
GSPS
DAC_1P2_x = 1.3 V ± 2%
DAC_1P2_x = 1.3 V ± 2%, FIR85 (2× NRZ) enabled
DAC_1P2_x = 1.3 V ± 2%
6
12
6
6.4
12.8
6.4
Adjusted2
1 TJ_DAC_MAX is the maximum junction temperature measured using the DAC temperature sensor.
2 The adjusted DAC update rate is calculated as follows: when FIR85 is disabled, fDAC is divided by the minimum required interpolation factor. For the AD9166, the
minimum interpolation factor is 1. Therefore, with fDAC = 6.0 GSPS, fDAC adjusted = 6.0 GSPS. When FIR85 is enabled, which puts the device into 2× NRZ mode, fDAC = 2 ×
fCLK, and the minimum interpolation is 2× (interpolation value). Thus, for the AD9166, with FIR85 enabled and fCLK = 6 GHz, fDAC = 12.0 GSPS, minimum interpolation = 2×, and the
adjusted DAC update rate = 6.0 GSPS.
Rev. 0 | Page 7 of 138
AD9166
Data Sheet
JESD204B INTERFACE SPECIFICATIONS
DAC_2P5_AN = 2.5 V, DAC_1P2_AN = DAC_1P2_CLK = 1.2 V, DAC_N1P2_AN = −1.2 V, DAC_1P2_DIG = 1.2 V, VDD_IO = 2.5 V,
DAC_1P2_SER = 1.2 V, DAC_3P3_SYNC = 3.3 V, AMP_5V_IN = 5.0 V, AMP_3P3_OUT = 3.3 V, AMP_3P3 = 3.3 V, AMP_N5 = −5.0 V,
IOUTFS = 40 mA, TA = −40°C to +85°C, unless otherwise noted. VTT is the termination voltage.
Table 4.
Parameter
Symbol Test Conditions/Comments
Guaranteed operating range per each lane
Min
Typ Max
Unit
SERIAL INTERFACE SPEED
Half Rate
Full Rate
Oversampling
2× Oversampling
6
3
12.5
6.25
3.125
Gbps
Gbps
Gbps
1.5
0.750
1.5625 Gbps
JESD204B DATA INPUTS
Input Leakage Current
Logic High
TA = 25°C
Input level = 1.2 V ± 0.25 V, VTT = 1.2 V
Input level = 0 V
10
−4
μA
μA
Logic Low
Unit Interval
UI
80
−0.05
110
1333
+1.85
1050
30
ps
V
mV
Ω
Ω
dB
dB
Common-Mode Voltage
Differential Voltage
VTT Source Impedance
Differential Impedance
Differential Return Loss
Common-Mode Return Loss
SYSREF± INPUT
VRCM
VDIFF
ZTT
ZRDIFF
RLRDIF
RLRCM
AC-coupled, VTT = DAC_1P2_SER1
At dc
At dc
80
100 120
8
6
Differential Impedance
DIFFERENTIAL OUTPUTS (SYNCOUT±)2
Output Differential Voltage
Output Offset Voltage
121
Ω
Driving 100 Ω differential load
VOD
VOS
350
1.15
420 450
1.2
mV
V
1.27
1 As measured on the input side of the ac coupling capacitor.
2 IEEE Standard 1596.3 LVDS compatible.
Rev. 0 | Page 8 of 138
Data Sheet
AD9166
INPUT DATA RATE AND BANDWIDTH SPECIFICATIONS
DAC_2P5_AN = 2.5 V, DAC_1P2_AN = DAC_1P2_CLK = 1.2 V, DAC_N1P2_AN = −1.2 V, DAC_1P2_DIG = 1.2 V, VDD_IO = 2.5 V,
DAC_1P2_SER = 1.2 V, DAC_3P3_SYNC = 3.3 V, AMP_5V_IN = 5.0 V, AMP_3P3_OUT = 3.3 V, AMP_3P3 = 3.3 V, AMP_N5 = −5.0 V,
IOUTFS = 40 mA, TA = −40°C to +85°C, unless otherwise noted.
Table 5.
Parameter
Test Conditions/Comments
Min
Typ
Max
Unit
INPUT DATA RATE
Complex1
Real
Interpolation > 1×
Interpolation = 1×
0.15
0.3
2.5
5.0
GSPS
GSPS
INSTANTANEOUS SIGNAL BANDWIDTH2
Complex
fCLK = 5 GHz, interpolation = 2×
fCLK = 6 GHz, interpolation = 3×
fCLK = 5 GHz, interpolation = 1×
2.25
1.8
2.5
GHz
GHz
GHz
Real
ANALOG BANDWIDTH
2x NRZ (FIR85 Enabled)
Minimum
Maximum
Mix Mode (FIR85 Disabled)
Minimum
fDAC = 12.0 GSPS
fDAC = 6.0 GSPS
fDAC = 6.0 GSPS
DC
9.0
GHz
GHz
1.0
8.0
GHz
GHz
Maximum
NRZ (FIR85 Disabled)
Minimum
Maximum3
DC
4.5
GHz
GHz
1 The complex data rate is the combined rate for both I and Q.
2 Interpolation filter bandwidth set to 90%.
3 Limited by the available output power due to sinc roll-off. See Figure 88 for more details.
PIPELINE DELAY AND LATENCY UNCERTAINTY SPECIFICATIONS
DAC_2P5_AN = 2.5 V, DAC_1P2_AN = DAC_1P2_CLK = 1.2 V, DAC_N1P2_AN = −1.2 V, DAC_1P2_DIG = 1.2 V, VDD_IO = 2.5 V,
DAC_1P2_SER = 1.2 V, DAC_3P3_SYNC = 3.3 V, AMP_5V_IN = 5.0 V, AMP_3P3_OUT = 3.3 V, AMP_3P3 = 3.3 V, AMP_N5 = −5.0 V,
IOUTFS = 40 mA, TA = −40°C to +85°C, unless otherwise noted.
Table 6.
Parameter1
Test Conditions/Comments
Min Typ Max Unit
JESD204B LINK LATENCY
Fixed
Variable
12
2
PCLK2 cycles
PCLK2 cycles
PCLK2 cycle
fCLK cycles
JESD204B TO DATAPATH INTERFACE LATENCY
DATAPATH PIPELINE DELAY3
SYSREF± to LOCAL MULTIFRAME CLOCKS (LMFC) DELAY
DETERMINISTIC LATENCY UNCERTAINTY
JED204B Subclass 0
1
NCO only, FIR85 off, inverse sinc off
JED204B Subclass 1
48
4
fCLK cycles
32
fCLK cycles
fCLK cycles
JED204B Subclass 14
4
1 The total latency through the device is calculated as follows:
Total Latency = Fixed Latency + Variable Latency + Interface Latency + Datapath Pipeline Delay.
2 PCLK is the internal processing clock for the AD9166 and equals the lane rate ÷ 40.
3 See Table 33 for pipeline delay (latency) values across different datapath configurations.
4 The SYSREF± signal input is sampled at a rate of fCLK/4, which leads to up to 4 fCLK cycles of deterministic latency uncertainty, provided that the setup and hold times for
sampling SYSREF± are met according to Table 10. The deterministic latency uncertainty can be further improved, using Register 0x037 and Register 0x038 to read the
exact clock cycle that was used to sample SYSREF±. See the SYSREF± Signal section for more details.
Rev. 0 | Page 9 of 138
AD9166
Data Sheet
AC SPECIFICATIONS
DAC_2P5_AN = 2.5 V, DAC_1P2_AN = DAC_1P2_CLK = 1.2 V, DAC_N1P2_AN = −1.2 V, DAC_1P2_DIG = 1.2 V, VDD_IO = 2.5 V,
DAC_1P2_SER = 1.2 V, DAC_3P3_SYNC = 3.3 V, AMP_5V_IN = 5.0 V, AMP_3P3_OUT = 3.3 V, AMP_3P3 = 3.3 V, AMP_N5 = −5.0 V.
IOUTFS = 20 mA, digital scale = 0 dBFS, fDAC = 12.0 GSPS, FIR85 enabled, TA = 25°C, unless otherwise noted. fOUT is output frequency.
Table 7.
Parameter
Test Conditions/Comments
Min
Typ
Max
Unit
SPURIOUS-FREE DYNAMIC RANGE (SFDR)1
Single Tone
fOUT = 51 MHz
fOUT = 451 MHz
fOUT = 1051 MHz
fOUT = 2051 MHz
fOUT = 4051 MHz
fOUT = 6051 MHz
fOUT = 9051 MHz
−83
−66
−54
−46
−38
−42
−35
dBc
dBc
dBc
dBc
dBc
dBc
dBc
Single Tone, IOUTFS = 40 mA
fOUT = 51 MHz
fOUT = 451 MHz
fOUT = 1051 MHz
fOUT = 2051 MHz
fOUT = 4051 MHz
fOUT = 6051 MHz
fOUT = 9051 MHz
−69
−55
−43
−33
−25
−29
−20
dBc
dBc
dBc
dBc
dBc
dBc
dBc
ADJACENT CHANNEL LEAKAGE RATIO (ACLR)
Single-Carrier Long-Term Evolution (LTE)
fOUT = 849 MHz
First adjacent channel, −6 dBFS
−70
−70
−71
−71
−69
−67
dBc
dBc
dBc
dBc
dBc
dBc
fOUT = 1865 MHz
fOUT = 2150 MHz
fOUT = 2680 MHz
fOUT = 3380 MHz
fOUT = 3680 MHz
Single-Carrier IEEE 802.11AC
fOUT = 5160 MHz
fOUT = 5865 MHz
First adjacent channel
−60
−59
dBc
dBc
INTERMODULATION DISTORTION (IMD)
Two-Tone Test
fOUT = 51 MHz
fOUT = 451 MHz
fOUT = 1051 MHz
fOUT = 2051 MHz
fOUT = 4051 MHz
fOUT = 6051 MHz
fOUT = 9051 MHz
−78
−65
−59
−51
−37
−55
−43
dBc
dBc
dBc
dBc
dBc
dBc
dBc
Two-Tone Test, IOUTFS = 40 mA
fOUT = 51 MHz
fOUT = 451 MHz
fOUT = 1051 MHz
fOUT = 2051 MHz
fOUT = 4051 MHz
fOUT = 6051 MHz
fOUT = 9051 MHz
−75
−60
−55
−49
−31
−38
−32
dBc
dBc
dBc
dBc
dBc
dBc
dBc
Rev. 0 | Page 10 of 138
Data Sheet
AD9166
Parameter
Test Conditions/Comments
Min
Typ
Max
Unit
NOISE SPECTRAL DENSITY (NSD)
fOUT = 537 MHz
Single tone, IOUTFS = 40 mA
−157
−157
−157
−154
−157
−153
−150
dBc/Hz
dBc/Hz
dBc/Hz
dBc/Hz
dBc/Hz
dBc/Hz
dBc/Hz
fOUT = 1044 MHz
fOUT = 2062 MHz
fOUT = 3791 MHz
fOUT = 4095 MHz
fOUT = 5011 MHz
fOUT = 5926 MHz
SINGLE SIDEBAND PHASE NOISE AT OFFSET
1 kHz Offset
10 kHz Offset
100 kHz Offset
1 MHz Offset
10 MHz Offset
fOUT = 3600 MHz, fDAC = 12,042.24 MSPS
−110.2
−134.8
−140.4
−149.0
−154.0
dBc/Hz
dBc/Hz
dBc/Hz
dBc/Hz
dBc/Hz
1 See the Clock Input section for more details on optimizing SFDR and reducing the image of the fundamental with clock input tuning.
CMOS PIN SPECIFICATIONS
DAC_2P5_AN = 2.5 V, DAC_1P2_AN = DAC_1P2_CLK = 1.2 V, DAC_N1P2_AN = −1.2 V, DAC_1P2_DIG = 1.2 V, VDD_IO = 2.5 V,
DAC_1P2_SER = 1.2 V, DAC_3P3_SYNC =3.3 V, AMP_5V_IN = 5.0 V, AMP_3P3_OUT = 3.3 V, AMP_3P3 = 3.3V, AMP_N5 = −5.0 V,
CS_AMP
CS_DAC
IOUTFS = 40 mA, TA = −40°C to +85°C, unless otherwise noted.
refers to
and
.
CS_x
Table 8.
Parameter
Symbol
Test Comments/Conditions
Min
Typ Max
Unit
INPUTS (SDIO, SCLK, CS_x, RESET, TX_ENABLE)
Voltage Input
High
VIH
VIL
1.8 V ≤ VDD_IO ≤ 2.5 V
1.8 V ≤ VDD_IO ≤ 2.5 V
0.7 × VDD_IO
V
V
Low
Current Input
High
0.3 × VDD_IO
IIH
IIL
75
μA
μA
Low
−150
OUTPUTS (SDIO, SDO)
Voltage Output
High
VOH
VOL
1.8 V ≤ VDD_IO ≤ 3.3 V
1.8 V ≤ VDD_IO ≤ 3.3 V
0.8 × VDD_IO
V
V
Low
0.2 × VDD_IO
Current Output
High
Low
IOH
IOL
4
4
mA
mA
Rev. 0 | Page 11 of 138
AD9166
Data Sheet
TIMING SPECIFICATIONS
Serial Port
DAC_2P5_AN = 2.5 V, DAC_1P2_AN = DAC_1P2_CLK = 1.2 V, DAC_N1P2_AN = −1.2 V, DAC_1P2_DIG = 1.2 V, VDD_IO = 2.5 V,
DAC_1P2_SER = 1.2 V, DAC_3P3_SYNC =3.3 V, AMP_5V_IN = 5.0 V, AMP_3P3_OUT = 3.3 V, AMP_3P3 = 3.3V, AMP_N5 = −5.0 V,
CS_AMP
CS_DAC
IOUTFS = 40 mA, TA = −40°C to +85°C, unless otherwise noted.
refers to
and
.
CS_x
Table 9.
Parameter
Symbol
Test Comments/Conditions
Min
Typ
Max
Unit
WRITE OPERATION
See Figure 47
Maximum SCLK Clock Rate
SCLK Clock High
SCLK Clock Low
SDIO to SCLK Setup Time
SCLK to SDIO Hold Time
CS_x to SCLK Setup Time
SCLK to CS_x Hold Time
fSCLK, 1/tSCLK
100
2.1
4.3
2.6
3.5
9
MHz
ns
ns
ns
ns
tPWH
tPWL
tDS
tDH
tS
SCLK = 20 MHz
SCLK = 20 MHz
2
1.5
2.53
6.7
ns
tH
9
ns
READ OPERATION
See Figure 46
SCLK Clock Rate
SCLK Clock High
SCLK Clock Low
SDIO to SCLK Setup Time
SCLK to SDIO Hold Time
CS_x to SCLK Setup Time
SCLK to SDIO (or SDO) Data Valid Time
CS_x to SDIO (or SDO) Output Valid to High-Z
fSCLK, 1/tSCLK
20
MHz
ns
ns
ns
ns
tPWH
tPWL
tDS
tDH
tS
Not shown in Figure 46
Not shown in Figure 46
Not shown in Figure 46
Not shown in Figure 46
Not shown in Figure 46
20
20
10
5
10
ns
tDV
12
21
ns
ns
Not shown in Figure 46
Rev. 0 | Page 12 of 138
Data Sheet
AD9166
SYSREF
DAC_2P5_AN = 2.5 V, DAC_1P2_AN = DAC_1P2_CLK = 1.2 V, DAC_N1P2_AN = −1.2 V, DAC_1P2_DIG = 1.2 V, VDD_IO = 2.5 V,
DAC_1P2_SER = 1.2 V, DAC_3P3_SYNC = 3.3 V, AMP_5V_IN = 5.0 V, AMP_3P3_OUT = 3.3 V, AMP_3P3 = 3.3 V, AMP_N5 = −5.0 V,
IOUTFS = 40 mA, TA = −40°C to +85°C, unless otherwise noted.
Table 10.
Parameter
Test Conditions/Comments
Min
Typ
Max1 Unit
2
SYSREF±
Differential Swing = 1.0 V
Minimum Setup Time, tSYSS
AC-coupled
65
45
68
19
5
ps
ps
ps
ps
ps
ps
DC-coupled, common-mode voltage = 0 V
DC-coupled, common-mode voltage = 1.25 V
AC-coupled
DC-coupled, common-mode voltage = 0 V
DC-coupled, common-mode voltage = 1.25 V
Minimum Hold Time, tSYSH
51
1 The maximum setup and hold times can be inferred from the data sheet for the 11 mm × 11 mm variant of the AD9164, under the assumption that the variations due
to difference in device laminate between the AD9166 and the AD9164 are minimal.
2 The SYSREF± pulse must have a duration longer than the device sample and hold time, plus four additional device clock cycles. For more information, refer to the
SYSREF± Signal section.
tSYSS
tSYSH
SYSREF+
CLK+
MIN 4 CLOCK EDGES
Figure 2. SYSREF to Device Clock Timing Diagram (Only SYSREF+ and CLK+ Shown)
Rev. 0 | Page 13 of 138
AD9166
Data Sheet
ABSOLUTE MAXIMUM RATINGS
REFLOW PROFILE
Table 11.
Parameter
Rating
The AD9166 reflow profile is in accordance with the JEDEC
JESD204B criteria for Pb-free devices. The maximum reflow
temperature is 260°C.
Supply Pins
DAC_1P2_AN, DAC_1P2_CLK,
DAC_1P2_DIG, DAC_1P2_SER
to GND
−0.3 V to +1.326 V
THERMAL MANAGEMENT
DAC_2P5_AN to GND
DAC_N1P2_AN to GND
VDD_IO, DAC_3P3_SYNC,
AMP_3P3_OUT, AMP_3P3 to
GND
AMP_5V_IN to GND
AMP_N5 to GND
−0.3 V to +2.625 V
−1.26 V to +0.3 V
−0.3 V to +3.465 V
The AD9166 is a high-power device that can dissipate as much as
4.88 W depending on the user application and configuration.
Due to the high power density of the AD9166, thermal
management is required to avoid exceeding the maximum
junction temperatures specified in Table 11, especially at
elevated ambient temperatures in still air.
−0.3 V to +5.25 V
−5.25 V to +0.3 V
Input/Output Pins
THERMAL RESISTANCE
RESET IRQ CS_AMP CS_DAC
,
−0.3 V to VDD_IO + 0.3 V
,
,
,
Thermal performance is directly linked to printed circuit board
(PCB) design and operating environment. Careful attention to
PCB thermal design is required.
SCLK, SDIO, SDO to GND
SYNCOUT±
−0.3 V to DAC_3P3_SYNC + 0.3 V
−0.3 V to DAC_1P2_SER + 0.3 V
− 0.5 V to +2.5 V
−0.3 V to DAC_1P2_CLK + 0.3 V
−0.3 V to DAC_2P5_AN + 0.3 V
SERDINx±
SYSREF±
CLK± to GND
ISET, VREF to DAC_VBGNEG
Junction Temperature1
θJA is the natural convection junction to ambient thermal
resistance measured in a one cubic foot sealed enclosure. θJC is
the junction to case thermal resistance.
Table 12. Thermal Resistance
DAC Core (TJ_DAC
fCLK > 5.1 GHz
fCLK ≤ 5.1 GHz
)
105°C
110°C
Package Type
θJA
θJC
Unit
BP-324-11
25.1
8.7
°C/W
Amplifier (TJ_AMP
Peak Reflow
Storage Temperature Range
)
105°C
260°C
−65°C to +150°C
1 Thermal resistance values specified are simulated based on JEDEC
specifications in compliance with JESD51-12.
1 Some operating modes of the device may cause the device to approach or
exceed the maximum junction temperature during operation at supported
ambient temperatures. Removal of heat from the device may require
additional measures such as active airflow, heat sinks, or other measures.
ESD CAUTION
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.
Rev. 0 | Page 14 of 138
Data Sheet
AD9166
PIN CONFIGURATION AND FUNCTION DESCRIPTIONS
TOP VIEW
(Not to Scale)
3
4
5
7
8
9
10 11 12 13 14 15 16 17 18
1
2
6
A
B
C
D
E
F
GND
GND
GND
GND
GND
GND
GND
GND RFOUT GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
VREF
GND
GND
GND
AMP_
VBG
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
CS_AMP GND
GND
DAC_ DAC_
N1P2
_AN
AMP_
1P8_
BYPASS
DAC_
DAC_
DAC_
DAC_
2P5_AN 2P5_AN
DAC_
GND
N1P2
_AN
GND
ISET
2P5_AN 2P5_AN 2P5_AN
DAC_
N1P2
_AN
DAC_ DAC_
N1P2
_AN
DAC_
N1P2
_AN
AMP_
1P8_
BYPASS
AMP_
3P3
AMP_
N5
AMP_
N5
DAC_
DAC_ DAC_
DAC_
DAC_
N1P2
_AN
GND
CLK+
CLK–
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
2P5_AN 2P5_AN 1P2_AN 1P2_AN 2P5_AN
DAC_
N1P2
_AN
DAC_ DAC_
N1P2
_AN
DAC_
N1P2
_AN
DAC_ DAC_
N1P2
_AN
DAC_ DAC_
1P2_
CLK
AMP_ AMP_
3P3_
OUT
AMP_
3P3
N1P2
_AN
AMP_N5 AMP_N5 AMP_N5 VDD_IO
N1P2
_AN
1P2_
CLK
3P3_
OUT
DAC_
1P2_
CLK
DAC_ AMP_
1P2_
CLK
AMP_ AMP_
DAC_
1P2_
CLK
DAC_
1P2_
CLK
DAC_
1P2_
CLK
DAC_ DAC_
1P2_
CLK
5V_
IN
5V_
IN
5V_
IN
1P2_
CLK
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
DAC_
DAC_
DAC_
DAC_
DAC_
DAC_
DAC_
GND
GND
GND
GND
IRQ
GND
GND
GND
GND
GND
DNC
GND
GND
DNC
GND
GND
GND
GND
G
H
J
2P5_AN 2P5_AN 2P5_AN 2P5_AN 2P5_AN 2P5_AN 2P5_AN
DAC_
VBG
NEG
DAC_
2P5_AN
DAC_
2P5_AN
SYS
REF+
GND
GND
GND
GND
GND
SCLK
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
DAC_
2P5_AN
SYS
REF–
GND
GND
GND
SDIO
GND
GND
SDO
DNC
GND
GND
GND
GND
GND
GND
K
L
DAC_
1P2_
SER
DAC_
1P2_
SER
SERDIN
0+
SERDIN
7+
GND
GND
GND
DAC_
1P2_
SER
DAC_
1P2_
SER
DAC_
1P2_
SER
DAC_
DAC_
1P2_
SER
TX_
ENABLE
SERDIN
0–
SERDIN
7–
RESET
1P2_ VDD_IO CS_DAC
SER
GND
GND
GND
GND
VDD_IO
GND
GND
GND
GND
M
N
P
R
T
DAC_
1P2_
SER
DAC_
1P2_
SER
DAC_
1P2_
DIG
DAC_
1P2_
DIG
DAC_ DAC_
DAC_
1P2_
DIG
DAC_
1P2_
DIG
DAC_
1P2_
DIG
DAC_
1P2_
DIG
DAC_
1P2_
DIG
GND
GND
1P2_
DIG
1P2_
DIG
GND
DAC_ DAC_
1P2_ 1P2_
SER SERDES SYNC
DAC_
3P3_
DAC_
1P2_
SER
DAC_
1P2_
SER
DAC_
1P2_
SER
DAC_
3P3_
SYNC
SERDIN
6+
SERDIN
1+
SYNC
OUT–
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
DNC
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
DAC_
1P2_
SER
DAC_ DAC_
1P2_
SER
SERPLL_
LDO_
BYPASS
DAC_
1P2_
SER
DAC_
1P2_
SER
DAC_
1P2_
SER
SERDIN
6–
SERDIN
1–
SYNC
OUT+
1P2_
SER
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
GND
U
SERDIN SERDIN
5+ 5–
SERDIN SERDIN
4+ 4–
SERDIN SERDIN
3– 3+
SERDIN SERDIN
2– 2+
V
DNC = DO NOT CONNECT. LEAVE THESE PINS FLOATING.
3.3V ANALOG SUPPLY, BUFFER OUTPUT
–5V ANALOG SUPPLY, BUFFER REFERENCE
5V ANALOG SUPPLY, BUFFER INPUT
2.5V ANALOG SUPPLY, DAC OUTPUT
–1.2V ANALOG SUPPLY, DAC OUTPUT
1.2V ANALOG SUPPLY, DAC CLOCK
1.2V ANALOG SUPPLY, DAC MIXED-SIGNAL
CMOS I/O
1.2V SERDES SUPPLY
1.2V DIGITAL SUPPLY
I/O PINS SUPPLY (1.8V TO 3.3V)
3.3V SYNCOUT+/SYNCOUT– SUPPLY
SERDES LANE x
SYSTEM REFERENCE POSITIVE AND NEGATIVE
RF SIGNALS
BYPASS NODE
GND REFERENCE
REFERENCE NODE
DNC
Figure 3. Pin Configuration
Table 13. Pin Function Descriptions
Pin No.
Mnemonic
Description
A1 to A8, A10 to A18, B1 to B12, B15 to B18, C1, C2, C8
to C10, C12, C15, C18, D1, D2, D9, D10, D18, E2, E18,
F2 to F4, F10, F11, F17, F18, G1 to G9, G17, G18, H2
to H6, H8 to H10, H14 to H18, J2 to J7, J10, J11, J15
to J18, K1 to K12, K14 to K18, L2, L3, L6 to L14, L16,
L17, M2, M3, M9, M14, M16, M17, N1 to N3, N5, N16
to N18, P2 to P4, P9, P11, P16, P17, R2 to R4, R8, R9,
R11, R16, R17, T1 to T18, U1 to U18, V1 to V3, V6, V9,
V10, V13, V16 to V18
GND
Ground.
A9
RFOUT
Device RF Output. Internally matched to a 50 Ω single-
ended load impedance.
Rev. 0 | Page 15 of 138
AD9166
Data Sheet
Pin No.
Mnemonic
Description
B13
AMP_VBG
Amplifier Band Gap Voltage. Connect Pin B13 to a 0.1 ꢀF
capacitor to ground, and a 1 kΩ resistor in series with a
1 ꢀF capacitor to ground. For information about the
voltage measured at this pin, VBGA, see the Amplifier
Junction Temperature Sensor section.
B14
CS_AMP
Amplifier Serial Port Chip Select (Active Low) Input.
CMOS levels on Pin B14 are determined with respect to
VDD_IO.
C3, C4, D3 to D5, D17, E3 to E5, E15 to E17
C5 to C7, C13, C14, D12, D13, D16, G10 to G16, H11,
H13, J9
DAC_N1P2_AN
DAC_2P5_AN
−1.2 V Analog Supply Voltage.
2.5 V Analog Supply Voltage.
C11, D11
AMP_1P8_BYPASS
Bypass Node for Internal1.8 V Analog Supply. Short
Pin C11 and Pin D11 and connect a 1 ꢀF capacitor to
ground.
C16
C17
ISET
DAC Reference Current. Connect Pin C16 with a 9.76 kΩ
resistor (RSET) to DAC_N1P2_AN.
DAC 1.2 V Reference Input/Output. Connect Pin C17 with
a 1 ꢀF capacitor to ground.
VREF
D6, E6
D7, D8, E9, E10, E11
D14, D15
AMP_3P3
AMP_N5
DAC_1P2_AN
CLK+, CLK−
3.3 V Analog Supply Voltage.
−5 V Analog Supply Voltage.
1.2 V Analog Supply Voltage.
Positive and Negative Device Clock Inputs. When FIR85 is
disabled, the input frequency to these pins (fCLK) is the
E1, F1
DAC clock frequency (fDAC). When FIR85 is enabled, fDAC
2× fCLK
=
.
E7, E8
AMP_3P3_OUT
VDD_IO
3.3 V Analog Supply Voltage for the Output Stage of the
Amplifier.
Supply Voltage for CMOS Input/Output and SPI.
Operational for 1.8 V to 3.3 V plus tolerance (see Table 1
for details).
E12, M7, M12
E13, E14, F5, F6, F12 to F16
F7, F8, F9
DAC_1P2_CLK
AMP_5V_IN
1.2 V Clock Supply Voltage.
5 V Analog Supply Voltage for the Input Stage of the
Amplifier. Pin F7 to Pin F9 internally supply the full-scale
current to the output stage of the DAC.
H1, J1
SYSREF+, SYSREF−
System Reference Positive and Negative Inputs. The H1
and J1 pins are self biased for ac coupling. They both can
be either ac-coupled or dc-coupled.
H7, J8, K13, P8
DNC
Do Not Connect. Do not connect these pins. Leave the
DNC pins floating.
H12
J12
DAC_VBGNEG
SDIO
DAC Band Gap Voltage. Connect Pin H12 with a 0.1 ꢀF
capacitor to DAC_N1P2_AN.
Serial Port Data Input/Output. CMOS levels on Pin J12 are
determined with respect to VDD_IO. See the Serial Data
I/O (SDIO) section for details.
J13
J14
SDO
Serial Port Data Output. CMOS levels on Pin J13 are
determined with respect to VDD_IO.
Serial Port Data Clock. CMOS levels on Pin J14 are
determined with respect to VDD_IO. See the Serial Clock
(SCLK) section for details.
SCLK
L1, M1
SERDIN7+, SERDIN7− SERDES Lane 7 Negative and Positive Inputs.
L4, L15, M4, M8, M10, M11, M15, N4, N15, P5, P6, P10,
P13, P14, R5 to R7, R13 to R15
DAC_1P2_SER
1.2 V SERDES Digital Supply.
L5
IRQ
Interrupt Request Output (Active Low, Open Drain).
L18, M18
M5
SERDIN0+, SERDIN0− SERDES Lane 1 Positive and Negative Inputs.
TX_ENABLE
Transmit Enable Input. Pin M5 can be used instead of the
DAC output bias power-down bits in Register 0x040,
Bits[1:0], to enable the DAC output. CMOS levels are
determined with respect to VDD_IO.
Rev. 0 | Page 16 of 138
Data Sheet
AD9166
Pin No.
Mnemonic
Description
M6
RESET
Reset (Active Low) Input. CMOS levels on Pin M6 are
determined with respect to VDD_IO.
M13
CS_DAC
DAC Serial Port Chip Select (Active Low) Input. CMOS
levels on Pin M13 are determined with respect to
VDD_IO.
N6 to N14
DAC_1P2_DIG
1.2 V Digital Supply Voltage for the Digital Signal
Processing (DSP) Blocks of the DAC.
P1, R1
SERDIN6+, SERDIN6− SERDES Lane 6 Negative and Positive Inputs.
P7, P15
P12, R12
DAC_3P3_SYNC
SYNCOUT−,
SYNCOUT+
3.3 V SERDES Sync Supply Voltage.
Negative and Positive LVDS Sync (Active Low) Output
Signals.
P18, R18
R10
SERDIN1+, SERDIN1− SERDES Lane 1 Positive and Negative Inputs.
SERPLL_LDO_BYPASS SERDES PLL Supply Voltage Bypass. Connect this pin with
a 1 Ω resistor in series with a 1 μF capacitor to ground.
V15, V14
V12, V11
V7, V8
SERDIN2+, SERDIN2− SERDES Lane 2 Positive and Negative Inputs.
SERDIN3+, SERDIN3− SERDES Lane 3 Positive and Negative Inputs.
SERDIN4+, SERDIN4− SERDES Lane 4 Negative and Positive Inputs.
SERDIN5+, SERDIN5− SERDES Lane 5 Negative and Positive Inputs.
V4, V5
Rev. 0 | Page 17 of 138
AD9166
Data Sheet
TYPICAL PERFORMANCE CHARACTERISTICS
AC PERFORMANCE (2× NRZ (FIR85) MODE)
IOUTFS = 20 mA, fCLK = 6.0 GHz, FIR85 enabled (fDAC = 2 × fCLK), interpolation = 4, nominal supplies, TA = 25°C, unless otherwise noted.
When data is transmitted across a JESD204B link: if fCLK ≤ 5.0 GHz, then interpolation = 2×; if fCLK > 5.0 GHz, then interpolation = 4×.
10
10
4.5dBm
0.9dBm
0
0
–10
–20
–40
–40
–50
–60
–60
–80
–90
–100
–10
–20
–40
–40
–50
–60
–60
–80
–90
–100
0
0
0
2000
4000
6000
8000
10000
12000
12000
12000
0
0
0
2000
4000
6000
8000
10000
12000
12000
12000
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 4. Single-Tone Spectrum at fOUT = 71 MHz
Figure 7. Single-Tone Spectrum at fOUT = 5032 MHz
10
0
10
0
3.6dBm
–4.2dBm
–10
–20
–40
–40
–50
–60
–60
–80
–90
–100
–10
–20
–40
–40
–50
–60
–60
–80
–90
–100
2000
4000
6000
8000
10000
2000
4000
6000
8000
10000
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 5. Single-Tone Spectrum at fOUT = 1875 MHz
Figure 8. Single-Tone Spectrum at fOUT = 7738 MHz
10
0
10
0
1.2dBm
–10.2dBm
–10
–20
–40
–40
–50
–60
–60
–80
–90
–100
–10
–20
–40
–40
–50
–60
–60
–80
–90
–100
2000
4000
6000
8000
10000
2000
4000
6000
8000
10000
FREQUENCY (MHz)
FREQUENCY (MHz)
Figure 6. Single-Tone Spectrum at fOUT = 3679 MHz
Figure 9. Single-Tone Spectrum at fOUT = 9222 MHz
Rev. 0 | Page 18 of 138
Data Sheet
AD9166
10
0
FOLDED SECOND HARMONIC
FOLDED THIRD HARMONIC
0
–20
–10
–20
–30
–40
–60
–80
–40
I
–100
–120
= 20mA, DIGITAL SCALE = 0dB
= 40mA, DIGITAL SCALE = 0dB
= 40mA, DIGITAL SCALE = –6dB
I
I
I
= 20mA, DIGITAL SCALE = 0dB
= 40mA, DIGITAL SCALE = 0dB
= 40mA, DIGITAL SCALE = –6dB
OUTFS
OUTFS
OUTFS
OUTFS
OUTFS
OUTFS
I
I
–50
0
2000
4000
6000
8000
10000
12000
0
0
0
2000
4000
6000
fOUT (MHz)
8000
10000
12000
fOUT (MHz)
Figure 10. Single-Tone Output Power vs. fOUT, over Digital Scale and IOUTFS
Figure 13. SFDR vs. fOUT over Digital Scale and IOUTFS
Folded Second and Third Harmonics
,
1.0
0
I
I
I
= 40mA
= 20mA
= 10mA
+
–
fOUT
fOUT
OUTFS
OUTFS
OUTFS
0.8
0.6
–20
–40
0.4
0.2
0
–60
–0.2
–0.4
–0.6
–0.8
–1.0
–80
–100
–120
I
I
I
= 20mA, DIGITAL SCALE = 0dB
= 40mA, DIGITAL SCALE = 0dB
= 40mA, DIGITAL SCALE = –6dB
OUTFS
OUTFS
OUTFS
0
–10
–20
–30
–40
–50
–60
2000
4000
6000
fOUT (MHz)
8000
10000
12000
DIGITAL SCALE (dBFS)
Figure 14. fCLK fOUT Spurious Output Power vs. fOUT
,
Figure 11. Output Power Error vs. Digital Scale, over IOUTFS, fOUT = 503 MHz
over Digital Scale and IOUTFS
0
0
IMD3
IMD5
IMD7
REAL SECOND HARMONIC
REAL THIRD HARMONIC
–20
–40
–20
–40
–60
–60
–80
–80
–100
–120
–100
I
I
I
= 20mA, DIGITAL SCALE = 0dB
= 40mA, DIGITAL SCALE = 0dB
= 40mA, DIGITAL SCALE = –6dB
I
I
I
= 20mA, DIGITAL SCALE = 0dB
= 40mA, DIGITAL SCALE = 0dB
= 40mA, DIGITAL SCALE = –6dB
OUTFS
OUTFS
OUTFS
OUTFS
OUTFS
OUTFS
–120
2000
4000
6000
fOUT (MHz)
8000
10000
12000
0
2000
4000
6000
fOUT (MHz)
8000
10000
12000
Figure 15. Worst Case Third-Order, Fifth-Order, and Seventh-Order
Intermodulation (IMD3, IMD5, IMD7) vs. fOUT, over Digital Scale and IOUTFS
Figure 12. SFDR vs. fOUT over Digital Scale and IOUTFS
Real Second and Third Harmonics
,
Rev. 0 | Page 19 of 138
AD9166
Data Sheet
0
0
HIGH
LOW
FOLDED SECOND HARMONIC
FOLDED THIRD HARMONIC
–20
–20
–40
–40
–60
–60
–80
–80
–100
–100
–120
fDAC = 6.0 GSPS
I
I
I
= 20mA, DIGITAL SCALE = 0dB
= 40mA, DIGITAL SCALE = 0dB
= 40mA, DIGITAL SCALE = –6dB
OUTFS
OUTFS
OUTFS
fDAC = 10.0 GSPS
fDAC = 12.0 GSPS
–120
0
2000
4000
6000
fOUT (MHz)
8000
10000
12000
0
2000
4000
6000
8000
10000
12000
fOUT (MHz)
Figure 16. Second-Order Intermodulation (IMD2) vs. fOUT, over Digital Scale
and IOUTFS
Figure 19. SFDR vs. fOUT over fDAC, Folded Second and Third Harmonics
0
0
IMD3
IMD5
IMD7
–5
–20
–10
–15
–40
–60
–80
–20
–25
–30
fDAC = 6.0 GSPS
–100
fDAC = 6.0 GSPS
–35
fDAC = 10.0 GSPS
fDAC = 10.0 GSPS
fDAC = 12.0 GSPS
fDAC = 12.0 GSPS
–40
–120
0
2000
4000
6000
8000
10000
12000
0
2000
4000
6000
8000
10000
12000
fOUT (MHz)
fOUT (MHz)
Figure 17. Single-Tone Output Power vs. fOUT over fDAC
Figure 20. Worst Case IMD3, IMD5, IMD7 vs. fOUT over fDAC
0
0
REAL SECOND HARMONIC
REAL THIRD HARMONIC
HIGH
LOW
–20
–40
–20
–40
–60
–60
–80
–80
–100
–120
fDAC = 6.0 GSPS
–100
–120
fDAC = 6.0 GSPS
fDAC = 10.0 GSPS
fDAC = 12.0 GSPS
fDAC = 10.0 GSPS
fDAC = 12.0 GSPS
0
2000
4000
6000
8000
10000
12000
0
2000
4000
6000
8000
10000
12000
fOUT (MHz)
fOUT (MHz)
Figure 18. SFDR vs. fOUT over fDAC, Real Second and Third Harmonics
Figure 21. IMD2 vs. fOUT over fDAC
Rev. 0 | Page 20 of 138
Data Sheet
AD9166
0
0
IMD3
IMD5
IMD7
–5
–20
–10
–15
–40
–60
–20
–25
–30
–35
–40
–80
–100
–120
T
T
T
= +105°C
= +60°C
= +5°C
T
T
T
= +105°C
= +60°C
= +5°C
J
J
J
J
J
J
–45
–50
0
2000
4000
6000
8000
10000
12000
0
2000
4000
6000
8000
10000
12000
fOUT (MHz)
fOUT (MHz)
Figure 22. Single-Tone Output Power vs. fOUT, over Temperature
Figure 25. Worst Case IMD3, IMD5, IMD7 vs. fOUT over Temperature
0
0
REAL SECOND HARMONIC
REAL THIRD HARMONIC
HIGH
LOW
–20
–40
–20
–40
–60
–60
–80
–80
–100
–100
T
T
T
= +105°C
= +60°C
= +5°C
T
T
T
= +105°C
= +60°C
= +5°C
J
J
J
J
J
J
–120
–120
0
2000
4000
6000
8000
10000
12000
0
2000
4000
6000
8000
10000
12000
fOUT (MHz)
fOUT (MHz)
Figure 23. SFDR vs. fOUT over Temperature, Real Second and Third Harmonics
Figure 26. IMD2 vs. fOUT over Temperature
0
–135
–140
–145
–150
–155
–160
fDAC = 6.0 GSPS
fDAC = 10.0 GSPS
fDAC = 12.0 GSPS
FOLDED SECOND HARMONIC
FOLDED THIRD HARMONIC
–20
–40
–60
–80
–100
T
T
T
= +105°C
= +60°C
= +5°C
J
J
J
–120
0
2000
4000
6000
8000
10000
12000
0
1000
2000
3000
4000
5000
6000
fOUT (MHz)
fOUT (MHz)
Figure 24. SFDR vs. fOUT over Temperature, Folded Second and Third
Harmonics
Figure 27. Single-Tone NSD vs. fOUT over fCLK, IOUTFS = 40 mA, NSD Measured at
10% Offset from fOUT
Rev. 0 | Page 21 of 138
AD9166
Data Sheet
–135
–140
–145
–150
–155
–60
DIGITAL SCALE = 0
DIGITAL SCALE = –6
900MHz
1800MHz
3600MHz
4500MHz
7200MHz
CLOCK SOURCE
–80
–100
–120
–140
–160
–180
CLOCK SOURCE:
R&S SMA100B
ULTRALOW PHASE NOISE OPTION
–160
0
10
100
1k
10k
100k
1M
10M
100M
1G
1000
2000
3000
4000
5000
6000
FREQUENCY OFFSET (Hz)
fOUT (MHz)
Figure 30. Single Sideband Phase Noise vs. Frequency Offset over fOUT
,
Figure 28. Single-Tone NSD vs. fOUT over Digital Scale, IOUTFS = 40 mA,
NSD Measured at 10% Offset from fOUT
f
DAC = 12,042.24 MSPS
–60
–135
+10dBm
+6dBm
0dBm
–6dBm
–12dBm
T
T
T
= +105°C
= +60°C
= –5°C
J
J
J
–80
–100
–120
–140
–160
–180
–140
–145
–150
–155
–160
CLOCK SOURCE:
R&S SMA100B
ULTRALOW PHASE NOISE OPTION
10
100
1k
10k
100k
1M
10M
100M
1G
0
1000
2000
3000
4000
5000
6000
FREQUENCY OFFSET (Hz)
fOUT (MHz)
Figure 31. Single Sideband Phase Noise vs. Frequency Offset over Clock
Power, fDAC = 12,042.24 MSPS, fOUT = 3.6 GHz
Figure 29. Single-Tone NSD vs. fOUT over Temperature, IOUTFS = 40 mA,
NSD Measured at 10% Offset from fOUT
Rev. 0 | Page 22 of 138
Data Sheet
AD9166
LTE PERFORMANCE (2× NRZ (FIR85) MODE)
IOUTFS = 20 mA, fCLK = 6021.12 MHz, FIR85 enabled (fDAC = 2× fCLK), nominal supplies, TA = 25°C, unless otherwise noted. When data is
transmitted across a JESD204B link: if fCLK ≤ 5.0 GHz, then interpolation = 2×; if fCLK > 5.0 GHz, then interpolation = 4×.
Figure 35. 20 MHz LTE Carrier ACLR at 2685.0 MHz
Figure 32. 20 MHz LTE Carrier ACLR at 889.0 MHz
Figure 36. 20 MHz LTE Carrier ACLR at 3695.0 MHz
Figure 33. 20 MHz LTE Carrier ACLR at 1875.0 MHz
–64
–66
–68
–70
–72
–74
–76
FIRST ACLR
SECOND ACLR
THIRD ACLR
FOURTH ACLR
FIFTH ACLR
500
1000
1500
2000
2500
3000
3500
4000
fOUT (MHz)
Figure 37. Worst Case 20 MHz LTE Carrier ACLR vs. fOUT
Figure 34. 20 MHz LTE Carrier ACLR at 2165.0 MHz
Rev. 0 | Page 23 of 138
AD9166
Data Sheet
802.11AC PERFORMANCE (2× NRZ (FIR85) MODE)
IOUTFS = 20 mA, fCLK = 6021.12 MHz, FIR85 enabled (fDAC = 2× fCLK), nominal supplies, TA = 25°C, unless otherwise noted. When data is
transmitted across a JESD204B link: if fCLK ≤ 5.0 GHz, then interpolation = 2×; if fCLK > 5.0 GHz, then interpolation = 4×.
–45
FIRST ACLR
SECOND ACLR
THIRD ACLR
–50
–55
–60
–65
–70
5200
5300
5400
5500
5600
5700
5800
fOUT (MHz)
Figure 38. 20 MHz 802.11AC ACLR at 5825.0 MHz
Figure 41. Worst Case 80 MHz 802.11AC ACLR vs. fOUT
–45
–30
–35
–40
–45
–50
–55
FIRST ACLR
SECOND ACLR
THIRD ACLR
BANDWIDTH = 80MHz, STANDARD: 802.11AC
–50
–55
–60
–65
–70
0
1
2
3
4
5
6
7
8
5100
5200
5300
5400
5500
5600
5700
5800
5900
FREQUENCY (Hz)
fOUT (MHz)
Figure 42. EVM vs. Frequency, 80 MHz Channel, Swept Across First, Second,
and Third Nyquist, fDAC = 11796.48 MSPS
Figure 39. Worst Case 20 MHz 802.11AC ACLR vs. fOUT
Figure 40. 80 MHz 802.11AC ACLR at 5530.0 MHz
Rev. 0 | Page 24 of 138
Data Sheet
AD9166
TERMINOLOGY
Offset Error
Signal-to-Noise Ratio (SNR)
Offset error is the deviation of the DAC output current from the
ideal of 0 mA.
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.
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.
Error Vector Magnitude (EVM)
EVM defines the average deviation of a modulation symbol
from its ideal location within a decision boundary. Typically,
EVM is quoted as the rms average of all error vector
magnitudes between the received symbols and their ideal
locations, for a given modulation order. For example, EVM for a
quadrature phase shift keying (QPSK) signal is the average of
the EVM across four decision boundaries. EVM is measured
using a baseband signal that is a pseudorandom binary
sequence (PRBS) of a statistically significant length.
Temperature Drift
Temperature drift is specified as the maximum change from the
ambient (25°C) value to the value at either TMIN or TMAX. For offset
and gain drift, the drift is reported in ppm of full-scale range
(FSR) per degree Celsius. For reference drift, the drift is reported
in ppm per degree Celsius.
Settling Time
Interpolation Filter
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.
If the digital inputs to the DAC are sampled at a multiple rate of
the interpolation rate (fDATA), a digital filter can be constructed that
has a sharp transition band near fDATA/2. Images that typically
appear around the output data rate (fDAC) can be greatly suppressed.
Spurious-Free Dynamic Range (SFDR)
SFDR is the difference, in decibels relative to carrier (dBc),
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.
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.
Adjusted DAC Update Rate
The adjusted DAC update rate is 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 can be given.
x-Order Intermodulation Distortion (IMDx)
IMDx (where x is 2, 3, 5, or 7 for second-order, third-order,
fifth-order, or seventh-order intermodulation distortion) is the
difference, in decibels relative to carrier (dBc), between the peak
amplitude of the output signal and the peak intermodulation
product of a specific x-order within the dc to Nyquist frequency
of the DAC. The signal is composed of two continuous wave
tones. If multiple IMDx products are present, the IMDx that is
located nearest to the signal and containing the highest power is
chosen to calculate the difference. This specification defines the
linearity of the analog output stage.
Physical Lane
Physical Lane x refers to SERDINx , where x represents 0 to 7.
Logical Lane
Logical Lane x (where x represents 0 to 7) 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 in the link, where
x represents 0 to 7.
Rev. 0 | Page 25 of 138
AD9166
Data Sheet
THEORY OF OPERATION
The AD9166 is a high performance, wideband, transmit
subsystem, composed of a high speed JESD204B SERDES
interface, a flexible 16-bit digital datapath, a I/Q DAC core,
along with an integrated differential to single-ended buffer
amplifier that is matched to a 50 Ω load at dc to 10 GHz.
digital datapath. The 100 MHz speed of the SPI write interface
enables rapid updating of the frequency tuning word (FTW) of
the NCO.
In addition to the main 48-bit NCO, the AD9166 also offers an
FFH NCO for selected DDS applications. The FFH NCO consists
of 32, 32-bit NCOs, each with its own phase accumulator, a
FTW select register to select one of the NCOs, and a phase
coherent hopping mode. Together, these elements enable phase
coherent FFH. With the FTW select register and the 100 MHz
SPI, dwell times as fast as 260 ns can be achieved.
The AD9166 DAC core uses the patented quad-switch
architecture, which enable DAC decoder settings that can extend
the output frequency range into the second and third Nyquist
zones with mix mode , RZ mode, and 2× NRZ mode (with
FIR85 enabled). The output can cover a range from 0 Hz to
more than 9 GHz in 2× NRZ mode. Mix mode can be used to
access 1.5 GHz to around 9 GHz at a reduced device power
consumption when compared to 2× NRZ. The NCO can then
shift a signal of up to 1.8 GHz instantaneous bandwidth to the
The differential core output is buffered and converted to a
single-ended output. The buffer is designed using a proprietary
BiCMOS process, which greatly improves the spectral response
of the core at higher operating frequencies. The improved
spectral response is essential for applications where extra wide
signal bandwidth and spectral flatness and purity are required.
Its output has impedance match to 50 Ω, up to 10 GHz, which
eases impedance matching concerns in wideband applications.
The differential to single-ended buffer eliminates the need for
an expensive, wideband balun, and supports the full operating
range of the DAC core, from true dc to 9 GHz. DC coupling
also allows baseband waveform generation, eliminating the
need for external bias tees or similar circuitry.
desired fOUT
.
Figure 1 shows a functional block diagram of the AD9166. Eight
high speed serial lanes carry data at a maximum speed of
12.5 Gbps, and either a 5 GSPS real input or a 2.5 GSPS complex
input data rate to the digital datapath. 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 is
sourced with a high fidelity, direct, external device sampling
clock. The performance of the DAC core can be optimized by using
on-chip adjustments to the device clock input accessible through
the SPI port. The SERDES interface can be configured to operate in
one-lane, two-lane, three-lane, four-lane, six-lane, or eight-lane
mode, depending on the required input data rate.
The AD9166 is capable of multichip synchronization that can both
synchronize multiple subsystems and establish a constant and
deterministic latency (latency locking) path to the subsystem
output. The latency for each of the subsystems remains constant
to within several device clock cycles from link establishment to
link establishment. An external alignment (SYSREF+ or
SYSREF−) signal makes the AD9166 Subclass 1 compliant.
Several modes of SYSREF signal handling are available for use
in the system.
The digital datapath of the AD9166 offers a bypass (1×) mode
and several interpolation modes (2×, 3×, 4×, 6×, 8×, 12×, 16×,
and 24×) through either an initial half-band (2×) or third-band
(3×) filter with programmable 80% or 90% bandwidth, and
three subsequent half-band filters (all 90%) with a maximum
DAC core sample rate of 6.0 GSPS. An inverse sinc filter is
provided to compensate for sinc related roll-off. An additional
half-band filter, FIR85, takes advantage of the quad-switch
architecture to interpolate on the falling edge of the clock, and
effectively double the DAC update rate in 2× NRZ mode. A 48-bit
programmable modulus numerically controlled oscillator (NCO) is
provided to enable digital frequency shifts of signals with near
infinite precision. The NCO can be operated alone in NCO only
mode or with digital data from the SERDES interface and
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
Start-Up Sequence section). Simple SPI initialization routines set
up the JESD204B link and are included in the evaluation board
package. This data sheet describes the various blocks of the
AD9166 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. 0 | Page 26 of 138
Data Sheet
AD9166
SERIAL PORT OPERATION
The AD9166 includes two separate SPI controllers, one for the
DAC and one for the amplifier. Either the DAC or the amplifier
can be addressed using the same SDIO, SDO, and SCLK pins,
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 FTW and NCO phase offsets, which change only
when the frequency tuning word bit, FTW_LOAD_REQ, is set.
CS_AMP
while asserting the corresponding chip select pin,
CS_DAC CS_AMP CS_DAC
cannot be asserted
or
.
and
simultaneously to address both the DAC and the amplifier
during the same communication cycle, as described in the Chip
Select (CS_AMP and CS_DAC) section.
DATA FORMAT
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 AD9166. MSB first or LSB first transfer formats are supported.
The serial port interface can be configured as a 4-wire interface
or a 3-wire interface in which the input and output share a single-
pin I/O (SDIO).
The instruction byte contains the information listed in Table 14.
Table 14. Serial Port Instruction Word
I15 (MSB)
I[14:0]
R/W
A[14:0]
W
R/ , Bit I15 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.
J12
SIO
A14 to A0, Bit I14 to Bit I0 of the instruction word, determine
the register that is accessed during the data transfer portion of
the communication cycle. For multibyte transfers, A[14:0] is the
starting address. The remaining register addresses are generated
by the device based on the address increment bit. If the address
increment bits are set high (Register 0x000, Bit 5 and Bit 2), multi-
byte SPI writes start on A[14:0] and increment by 1 for every
eight bits sent or received. If the address increment bits are set
to 0, the address decrements by 1 every eight bits.
SDO J13
J14
B14
M13
SCLK
CS_AMP
CS_DAC
SPI
PORT
Figure 43. Serial Port Interface Pins
There are two phases to a communication cycle with the AD9166.
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.
SERIAL PORT PIN DESCRIPTIONS
Serial Clock (SCLK)
The serial clock pin synchronizes data to and from the device
and runs the internal state machines. The maximum frequency
of SCLK is 100 MHz. All data input is registered on the rising
edge of SCLK. All data is driven out on the falling edge of SCLK.
A logic high on the 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.
Chip Select (
CS_AMP
and
)
CS_DAC
The AD9166 includes two chip select pins, one for the DAC
CS_DAC CS_AMP
(
) and one for the buffer amplifier (
),
pin must be
CS_AMP
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 FTW and NCO phase offsets, which change
only when the frequency tuning word FTW_LOAD_REQ bit is set.
CS_x CS_x
hereafter referred to as
asserted to address the particular silicon die.
CS_DAC
. The correct
and
cannot be asserted simultaneously.
An active low input starts and gates a communication cycle.
CS_x
allows more than one device to be used on the same serial
communications line. The SDIO pin goes to a high impedance
state when this input is high. During the communication cycle,
the chip select must stay low.
CS_AMP CS_DAC
or
A logic high on
followed by a logic low
Serial Data I/O (SDIO)
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 SDIO pin is a bidirectional data line. In 4-wire mode, the
SDIO pin acts as the data input and the SDO pin acts as the data
output.
Rev. 0 | Page 27 of 138
AD9166
Data Sheet
When ADDRINC or ADDRINC_M is 1, the multicycle
addresses are incremented. When ADDRINC or ADDRINC_M
is 0, the addresses are decremented. A new write cycle can
SERIAL PORT OPTIONS
The serial port can support both MSB first and LSB first data
formats. This functionality is controlled by the LSB first bit
(Register 0x000, Bit 6 and Bit 1). The default is MSB first (LSB
bit = 0).
CS_x
always be initiated by bringing
high and then low again.
To prevent confusion and to ensure consistency between
devices, the chip tests the first nibble following the address
phase, ignoring the second nibble. This test is completed
independently from the LSB first bits and ensures that there are
extra clock cycles following the soft reset bits (Register 0x000,
Bit 0 and Bit 7). This test of the first nibble only applies when
writing to Register 0x000.
When the LSB first bits = 0 (MSB first), the instruction and data
bits must be written from MSB to LSB. Read/write (R/ ) is
followed by the instruction word, A[14:0], and D[7:0], the data-
word. When the LSB first bits = 1 (LSB first), the opposite is
true. A[0:14] is followed by R/ , which is subsequently
W
W
followed by D[0:7].
INSTRUCTION CYCLE
DATA TRANSFER CYCLE
The serial port supports a 3-wire or 4-wire interface. When the
SDO active bits = 1 (Register 0x000, Bit 4 and Bit 3), a 4-wire
interface with a separate input pin (SDIO) and output pin (SDO) is
used. When the SDO active bits = 0, the SDO pin is unused and
the SDIO pin is used for both the input and the output.
CS_x
SCLK
SDIO
R/W A14 A13
A3 A2 A1 A0 D7N D6N D5N
D30 D20 D10 D00
Multibyte Data Transfers
Figure 44. Serial Register Interface Timing, MSB First, Register 0x000, Bit 5
and Bit 2 = 0
CS_AMP
Multibyte data transfers can be performed by holding
CS_DAC
INSTRUCTION CYCLE
DATA TRANSFER CYCLE
or
low for multiple data transfer cycles (eight SCLK
CS_x
cycles) after the first data transfer word following the instruction
cycle. The first eight SCLK cycles 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 is set using
ADDRINC or ADDRINC_M (Register 0x000, Bit 5 and Bit 2).
SCLK
SDIO
A0 A1 A2
A12 A13 A14 R/W D00 D10 D20
D4N D5N D6N D7N
Figure 45. Serial Register Interface Timing, LSB First, Register 0x000, Bit 5 and
Bit 2 = 1
CS_x
SCLK
tDV
SDIO
tS
DATA BIT n
DATA BIT n – 1
Figure 46. Timing Diagram for Serial Port Register Read
tH
CS_x
tPWH
tPWL
SCLK
tDS
tDH
SDIO
INSTRUCTION BIT 15 INSTRUCTION BIT 14
INSTRUCTION BIT 0
Figure 47. Timing Diagram for Serial Port Register Write
Rev. 0 | Page 28 of 138
Data Sheet
AD9166
JESD204B SERIAL DATA INTERFACE
lane count to be used. In addition, the interpolation rate and
JESD204B OVERVIEW
number of lanes can be used to define the rest of the configura-
tion needed to set up the AD9166. The interpolation rate and
the number of lanes are selected in Register 0x110.
The AD9166 has eight JESD204B data ports that receive data.
The eight JESD204B ports can be configured as part of a single
JESD204B link that uses a single system reference (SYSREF ) and
device clock (CLK ).
The AD9166 has a single DAC output. However, for the
purposes of the complex signal processing on chip, whenever
interpolation is used, the converter count is defined as M = 2.
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 48 shows the communication layers
implemented in the AD9166 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.
For a particular application, the number of converters to use
(M) and the data rate variable (DataRate) are known. The lane
rate variable (LaneRate) and number of lanes (L) can be traded
off as follows:
DataRate = (DACRate)/(InterpolationFactor)
LaneRate = (20 × DataRate × M)/L
The physical layer establishes 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.
The transport layer receives the descrambled JESD204B frames
and converts them to DAC samples.
where LaneRate must be between 750 Mbps and 12.5 Gbps.
Achieving and recovering synchronization of the lanes is very
important. To simplify the interface to the transmitter, the
AD9166 designates a master synchronization signal for each
SYNCOUT−
SYNCOUT+
JESD204B link. The
and
pins are used
Various JESD204B parameters (L, F, K, M, N, NP, S, HD) define
how the data is packed and tell the device how to turn the serial
data into samples. These parameters are defined in the Transport
Layer section. The AD9166 also has a descrambling option (see
the Descrambler section for more information).
as the master signal for all lanes. 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 various combinations of JESD204B parameters that are
supported depend solely on the number of lanes. Thus, a
unique set of parameters can be determined by selecting the
SYNCOUT+/SYNCOUT–
PHYSICAL
LAYER
DATA LINK
LAYER
TRANSPORT
LAYER
I DATA[15:0]
SERDIN0±
DESERIALIZER
DESERIALIZER
TO DAC
DSP BLOCK
QBD/
DESCRAMBLER
FRAME TO
SAMPLES
SERDIN7±
SYSREF±
Q DATA[15:0]
Figure 48. Functional Block Diagram of Serial Link Receiver
Table 15. Single-Link JESD204B Operating Modes
Number of Lanes (L)
Parameter
Parametric Symbol
1
1
2
4
1
2
2
2
2
1
3
3
2
4
3
4
4
2
1
1
6
6
2
2
3
8
Lane Count
Converter Count
Octets per Frame per Lane
Samples per Converter per Frame
L
M
F
8
1 (real), 2 (complex)
1
4 (real), 2 (complex)
S
Rev. 0 | Page 29 of 138
AD9166
Data Sheet
Table 16. Data Structure per Lane for JESD204B Operating Modes1
JESD204B Operating Mode
Lane No.
Lane 0
Lane 1
Lane 2
Lane 3
Lane 4
Lane 5
Lane 6
Lane 7
Lane 0
Lane 1
Lane 2
Lane 3
Lane 4
Lane 5
Lane 6
Lane 7
Lane 0
Lane 1
Lane 2
Lane 3
Lane 4
Lane 5
Lane 0
Lane 1
Lane 2
Lane 3
Lane 0
Lane 1
Lane 2
Lane 0
Lane 1
Lane 0
Frame 0
Frame 1
Frame 2
Frame 3
L = 8, M = 1, F = 1, S = 4
M0S0[15:8]
M0S0[7:0]
M0S1[15:8]
M0S1[7:0]
M0S2[15:8]
M0S2[7:0]
M0S3[15:8]
M0S3[7:0]
M0S0[15:8]
M0S0[7:0]
M0S1[15:8]
M0S1[7:0]
M1S0[15:8]
M1S0[7:0]
M1S1[15:8]
M1S1[7:0]
M0S0[15:8]
M0S1[15:8]
M0S2[15:8]
M1S0[15:8]
M1S1[15:8]
M1S2[15:8]
M0S0[15:8]
M0S0[7:0]
M1S0[15:8]
M1S0[7:0]
M0S0[15:8]
M0S2[15:8]
M1S1[15:8]
M0S0[15:8]
M1S0[15:8]
M0S0[15:8]
L = 8, M = 2, F = 1, S = 2
L = 6, M = 2, F = 2, S = 3
M0S0[7:0]
M0S1[7:0]
M0S2[7:0]
M1S0[7:0]
M1S1[7:0]
M1S2[7:0]
L = 4, M = 2, F = 1, S = 1
L = 3, M = 2, F = 4, S = 3
M0S0[7:0]
M0S2[7:0]
M1S1[7:0]
M0S0[7:0]
M1S0[7:0]
M0S0[7:0]
M0S1[15:8]
M1S0[15:8]
M1S2[15:8]
M0S1[7:0]
M1S0[7:0]
M1S2[7:0]
L = 2, M = 2, F = 2, S = 1
L = 1, M = 2, F = 4, S = 1
M1S0[15:8]
M1S0[7:0]
1 Mx is the converter number and Sy is the sample number. For example, M0S0 means Converter 0, Sample 0. Blank cells are not applicable.
Rev. 0 | Page 30 of 138
Data Sheet
AD9166
use the TERM_BLKx_CTRLREG1 registers (Register 0x2A8 and
Register 0x2AF). Set the registers as follows:
PHYSICAL LAYER
The physical layer of the JESD204B interface, hereafter referred
to as the deserializer, has eight identical channels. Each channel
consists of the terminators, an equalizer, a clock and data recovery
(CDR) circuit, and the 1:40 demux function (see Figure 49).
Default setting: set to 0xXXX0XXXX. The termination
block autocalibrates the termination values.
Overwrite setting: set to 0xXXX1XXXX to overwrite the
autocalibration with the termination values in Bits[3:1] of
Register 0x2A8 and Register 0x2AF.
DESERIALIZER
SERDINx±
TERMINATION
EQUALIZER
CDR
1:40
Individual offsets from the autocalibration value for each lane
are programmed in Bits[3:0] of Register 0x2BB to Register 0x2C2.
The value is a signed magnitude, with Bit 3 as the sign bit. The
total range of the termination resistor value is about 94 Ω to 120 Ω,
with approximately 3.5% increments across the range (for example,
smaller steps at the bottom of the range than at the top).
SPI
CONTROL
FROM SERDES PLL
Figure 49. Deserializer Block Diagram
Receiver Eye Mask
JESD204B data is input to the AD9166 via the SERDINx 1.2 V
differential input pins as per the JESD204B specification.
The AD9166 complies with the JESD204B specification regarding
the receiver eye mask and is capable of capturing data that
complies with this mask. Figure 50 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.
Interface Power-Up and Input Termination
Before using the JESD204B interface, it must be powered up by
setting Register 0x200, Bit 0 = 0. In addition, each physical lane
(PHY) 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 as follows:
LV-OIF-11G-SR RECEIVER EYE MASK
525
Set the bit to 0 when the physical lane is used.
Set the bit to 1 when the physical lane is not used.
55
0
The AD9166 autocalibrates the input termination to 50 Ω.
Before running the termination calibration, Register 0x2A7 and
Register 0x2AE must be written as described in Table 17 to
guarantee proper calibration. The termination calibration begins
when Register 0x2A7, Bit 0 and Register 0x2AE, Bit 0 transition
from low to high.
–55
–525
0
0.35 0.5 0.65
TIME (UI)
1.00
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.
Figure 50. Receiver Eye Mask for 600 mV VTT Swing
Clock Relationships
The following clock rates are used throughout the remainder of the
JESD204B Serial Data Interface section. The relationship between
any of the clocks can be derived from the following equations:
The PHY x termination autocalibration routine is listed in Table 17.
Table 17. PHYx Termination Autocalibration Routine
Address Value Description
DataRate = (DACRate)/(InterpolationFactor)
LaneRate = (20 × DataRate × M)/L
ByteRate = LaneRate/10
0x2A7
0x01
Autocalibrate PHY 0, PHY 1, PHY 6, and
PHY 7 terminations
0x2AE
0x01
Autocalibrate PHY 2, PHY 3, PHY 4, and
PHY 5 terminations
where:
M is the JESD204B parameter for converters per link.
L is the JESD204B parameter for lanes per link.
The input termination voltage of the DAC is sourced externally
through the DAC_1P2_SER pins. It is recommended that the
JESD204B inputs be ac-coupled to the JESD204B transmit device
using 100 nF capacitors.
This relationship comes from 8-bit/10-bit encoding, where each
byte is represented by 10 bits.
PCLK Rate = ByteRate/4
The calibration code of the termination can be read from
Bits[3:0] in Register 0x2AC (PHY 0, PHY 1, PHY 6, and PHY 7)
and Register 0x2B3 (PHY 2, PHY 3, PHY 4, and PHY 5). If
needed, the termination values can be adjusted or set using
several registers. To override the autocalibrated termination values,
The processing clock is used for a quad-byte decoder.
FrameRate = ByteRate/F
where F is JESD204B parameter for octets per frame per lane.
PCLK Factor = FrameRate/PCLK Rate = 4/F
Rev. 0 | Page 31 of 138
AD9166
Data Sheet
To generate the lane rate clock inside the device, a CDR
sampling mode must be selected as follows:
SERDES PLL
Functional Overview of the SERDES PLL
For a lane rate greater than 6.25 Gbps, use half rate CDR.
For a lane rate between 3 Gbps and 6.25 Gbps, disable half
rate operation.
For a lane rate less than 3 Gbps, disable full rate and enable
2× oversampling to recover the appropriate lane rate clock.
The independent SERDES phase-locked loop (PLL) uses
Integer N techniques to achieve clock synthesis. The entire
SERDES PLL is integrated on chip, including the voltage
controlled oscillator (VCO) and the loop filter. The SERDES
PLL VCO operates over the range of 6 GHz to 12.5 GHz.
In the SERDES PLL, a VCO divider block divides the VCO
clock by 2 to generate a 3 GHz to 6.25 GHz quadrature clock for
the deserializer cores. This clock is the input to the CDR block
that is described in the Clock and Data Recovery section.
Table 19 lists the CDR sampling settings that must be set
depending on the lane rate value.
Table 19. CDR Operating Modes
SPI_ENHALFRATE
(Register 0x230,
Bit 5)
SPI_DIVISION_RATE
(Register 0x230,
Bits[2:1])
The reference clock to the SERDES PLL is always running at a
frequency, fREF, that is equal to 1/40 of the lane rate (PCLK rate).
The fREF frequency is divided by an integer factor, set by SERDES_
PLL_DIV_FACTOR, to deliver a clock to the phase frequency
detector (PFD) block, fPFD, that is between 35 MHz and 80 MHz.
Table 18 includes the respective SERDES_PLL_DIV_FACTOR
register settings for each of the desired PLL_REF_CLK_RATE
options available.
Lane Rate
(Gbps)
0.750 to 1.5625 0 (full rate)
0b10 (divide by 4)
0b01 (divide by 2)
0b00 (no divide)
0b00 (no divide)
1.5 to 3.125
3 to 6.25
0 (full rate)
0 (full rate)
1 (half rate)
6 to 12.5
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.
Table 18. SERDES PLL Divider Settings
PLL_REF_CLK_RATE
(Register 0x084,
Bits[5:4])
Lane Rate
(Gbps)
SERDES_PLL_DIV_FACTOR
(Register 0x289, Bits[1:0])
After configuring the CDR circuit, reset it and then release the
reset by writing 0 and then writing 1 to Register 0x206, Bit 0.
0.750 to 1.5625 0b01 = 2×
0b10 = ÷1
0b10 = ÷1
0b01 = ÷2
0b00 = ÷4
1.5 to 3.125
3 to 6.25
0b00 = 1×
0b00 = 1×
0b00 = 1×
In some clocking configuration, it may be necessary to reset the
CDR after the JESD204B transmitter begins sending /K/
characters as part of the JESD204B serial link establishment, so
that the CDR restarts its search loop and aligns the clocks
correctly (see the JESD204B Serial Link Establishment section).
6 to 12.5
SERDES PLL Enable and Recalibration
Register 0x280 controls the synthesizer enable and recalibration.
Power-Down Unused PHYs
To enable the SERDES PLL, first set the PLL divider register
(see Table 18). Then enable the SERDES PLL by writing
Register 0x280, Bit 0 = 1. If a recalibration is needed, write
Register 0x280, Bit 2 = 0b1 and then reset the bit to 0b0. The
rising edge of the bit causes a recalibration to begin.
Unused lanes that are left enabled consume extra power
unnecessarily. Each lane that is not in use (SERDINx ) must be
powered off by writing a 1 to the corresponding bit of PHY_PD
(Register 0x201).
Equalization
Confirm that the SERDES PLL is working by reading
Register 0x281. If Register 0x281, Bit 0 = 1, the SERDES PLL
has locked. If Register 0x281, Bit 3 = 1, the SERDES PLL
calibration has completed. If Register 0x281, Bit 4 or Bit 5 is
high, the PLL reaches the lower or upper end of its calibration
band and must be recalibrated by writing 0 and then 1 to
Register 0x280, Bit 2.
To compensate for signal integrity distortions for each PHY
channel due to PCB trace length and impedance, the AD9166
employs an easy to use, low power equalizer on each JESD204B
channel. The AD9166 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,
Bits[7:6]. In low power mode (Register 0x268, Bits[7:6] = 0b01)
and operating at the maximum lane rate of 12.5 Gbps, the
equalizer can compensate for up to 11.5 dB of insertion loss. In
normal mode (Register 0x268, Bits[7:6] = 0b00), the equalizer
can compensate for up to 17.2 dB of insertion loss. This perfor-
mance is shown in Figure 51 as an overlay to the JESD204B
specification for insertion loss. Figure 51 shows the equalization
performance at 12.5 Gbps, near the maximum baud rate for the
AD9166.
Clock and Data Recovery (CDR)
The deserializer is equipped with a CDR circuit. Instead of
recovering the clock directly from the JESD204B serial lanes, the
CDR circuit continuously aligns the phase of the sampling
clocks for each SERDES lane with the incoming bit stream from
the JESD204B transmitter. The sampling clocks are derived
from the SERDES PLL. The 3 GHz to 6.25 GHz sampling clocks
are derived from the SERDES PLL, as shown in Figure 54, at the
input to the CDR.
Rev. 0 | Page 32 of 138
Data Sheet
AD9166
0
–5
Figure 52 and Figure 53 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, respectively. See the Hardware Considerations section for
specific layout recommendations for the JESD204B channel.
–10
–15
–20
–25
–30
–35
–40
Low power mode is recommended if the insertion loss of the
JESD204B PCB channels is less than that of the most lossy
supported channel for low power mode (shown in Figure 51). 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 51), use normal mode. At 12.5 Gbps operation, the
equalizer in normal mode consumes about 4 mW more power
per lane than in low power equalizer mode. Note that either mode
can be used in conjunction with transmitter preemphasis to
ensure functionality and/or optimize for power.
STRIPLINE = 6"
STRIPLINE = 10"
STRIPLINE = 15"
STRIPLINE = 20"
STRIPLINE = 25"
STRIPLINE = 30"
0
1
2
3
4
5
6
7
8
9
10
FREQUENCY (GHz)
Figure 52. Insertion Loss of 50 Ω Striplines on FR4
0
–5
0
JESD204B SPEC ALLOWED
CHANNEL LOSS
EXAMPLE OF
JESD204B
COMPLIANT
CHANNEL
2
4
–10
–15
–20
–25
–30
–35
–40
6
EXAMPLE OF
AD9166
COMPATIBLE
CHANNEL (LOW
POWER MODE)
8
10
12
14
16
18
20
22
24
AD9166 ALLOWED
CHANNEL LOSS
(LOW POWER MODE)
AD9166 ALLOWED
CHANNEL LOSS
(NORMAL MODE)
STRIPLINE = 6"
STRIPLINE = 10"
STRIPLINE = 15"
STRIPLINE = 20"
STRIPLINE = 25"
STRIPLINE = 30"
EXAMPLE OF
AD9166
COMPATIBLE
CHANNEL
(NORMAL MODE)
3.125
6.250
9.375
FREQUENCY (GHz)
FREQUENCY (GHz)
Figure 53. Insertion Loss of 50 Ω Microstrips on FR4
Figure 51. Insertion Loss Allowed
MODE
DIVIDE (N)
20
HALF RATE
FULL RATE, NO DIV 40
ENABLE HALF RATE
DIVISION RATE
REG 0x230
FULL RATE, DIV 2
FULL RATE, DIV 4
80
160
SAMPLE CLOCK
I, Q TO CDR
INTERPOLATION
JESD LANES
REG 0x110
CDR OVERSAMP
REG 0x289
VALID RANGE
3GHz TO 6.25GHz
JESD LANE CLOCK
(SAME RATE AS PCLK)
CDR
÷N
PLL REF CLOCK
VALID RANGE
35MHz TO 80MHz
PCLK
÷4
÷4, ÷2,
OR ÷1
CP
LF
÷2
÷8
DAC CLOCK
GENERATOR
PLL_REF_CLK_RATE
÷6 TO ÷127,
DEFAULT: 10
1×, 2×, 4×
REG 0x084
Figure 54. SERDES PLL Synthesizer Block Diagram Including VCO Divider Block
Rev. 0 | Page 33 of 138
AD9166
Data Sheet
between serial lanes. Each AD9166 serial interface link can issue
DATA LINK LAYER
SYNCOUT
a synchronization request by setting its
signal low.
The data link layer of the AD9166 JESD204B interface accepts
the deserialized data from the PHYs and deframes, and descram-
bles them so that data octets are presented to the transport layer to
be put into DAC samples. The architecture of the data link layer
is shown in Figure 55. The data link layer consists of a synchroni-
zation FIFO for each lane, a crossbar switch, a deframer, and a
descrambler.
The synchronization protocol follows Section 4.9 of the JESD204B
standard. When a stream of four consecutive /K/ symbols is
received, the AD9166 deactivates the synchronization request
SYNCOUT
by setting the
signal high at the next internal local
multiframe clock (LMFC) rising edge. Then, AD9166 waits for
the transmitter to issue an initial lane alignment sequence
(ILAS). During the ILAS, 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 56).
The AD9166 can operate as a single-link high speed JESD204B
serial data interface. All eight lanes of the JESD204B interface
handle link layer communications such as code group synchroniza-
tion (CGS), frame alignment, and frame synchronization.
The AD9166 decodes 8-bit/10-bit control characters, allowing
marking of the start and end of the frame and alignment
DATA LINK LAYER
SYNCOUT+/SYNCOUT–
QUAD-BYTE
DEFRAMER
LANE 0 DESERIALIZED
AND DESCRAMBLED DATA
QBD
LANE 0 OCTETS
SERDIN0±
LANE 0 DATA CLOCK
FIFO
CROSS-
BAR
SWITCH
LANE 7 DESERIALIZED
AND DESCRAMBLED DATA
SERDIN7±
LANE 7 OCTETS
LANE 7 DATA CLOCK
FIFO
SYSTEM CLOCK
PHASE DETECT
SYSREF±
PCLK
SPI CONTROL
Figure 55. Data Link Layer Block Diagram
L RECEIVE LANES
K
K
K
K
K
K
R
K
D
K
D
K
D
D
A
R
Q
C
C
D
C
D
A
R
D
D
D
D
(EARLIEST ARRIVAL)
L RECEIVE LANES
(LATEST ARRIVAL)
K
R D D
D
D
A
R
Q
C
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
D D A R D D
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 = JESD204x LINK CONFIGURATION PARAMETERS
D = Dx.y DATA SYMBOL
Figure 56. Lane Alignment During ILAS
Rev. 0 | Page 34 of 138
Data Sheet
AD9166
If any of these errors exist, they are reported back to the
JESD204B Serial Link Establishment
transmitter in one of the following ways (see the JESD204B
Error Monitoring section for details):
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.
SYNCOUT
signal assertion: resynchronization
SYNCOUT
(
signal pulled low) is requested at each error
Step 1: Code Group Synchronization
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
Each receiver must locate /K/ (K28.5) characters in its input bit
stream. After four consecutive /K/ characters are detected on all
SYNCOUT
link lanes, the receiver block deasserts the
to the transmitter block at the receiver LMFC edge.
SYNCOUT
signal
SYNCOUT
.
The transmitter captures the change in the
signal
Errors can optionally trigger an interrupt request (IRQ)
event, which can be sent to the transmitter.
and, at a future transmitter LMFC rising edge, starts the ILAS.
Step 2: Initial Lane Alignment Sequence
For more information about the various test modes for
verifying the link integrity, see the JESD204B Test Modes
section.
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.
Lane First In/First Out (FIFO)
The FIFOs in front of the crossbar switch and deframer synchro-
nize 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 to allow up to two PCLK 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.
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. The AD9166 does not require this 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/ (K.28.0), /Q/ (K.28.4), and then data corresponding to the
link parameters. Additional multiframes can be added to the
ILAS if needed by the receiver. By default, the AD9166 uses four
multiframes in the ILAS (this can be changed in Register 0x478). If
using Subclass 1, exactly four multiframes must be used.
Lane FIFO Interrupt Request (IRQ)
An aggregate lane FIFO overflow/underflow error bit is also
available as an IRQ event. Use Register 0x020, Bit 2 to enable
the FIFO overflow/underflow error bit, and then use
Register 0x024, Bit 2 to read back its status and reset the IRQ
signal. See the Interrupt Request Operation section for more
information.
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.
Crossbar Switch
Register 0x308 to Register 0x30B allow arbitrary mapping of
physical lanes (SERDINx ) to logical lanes used by the SERDES
deframers.
Step 3: Data Streaming
In this phase, data is streamed from the transmitter block to the
receiver block. Optionally, data can be scrambled. Scrambling
does not start until the very first octet following the ILAS.
Table 20. 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
SRC_LANE0
SRC_LANE1
SRC_LANE2
SRC_LANE3
SRC_LANE4
SRC_LANE5
SRC_LANE6
SRC_LANE7
The receiver block processes and monitors the data it receives
for errors, including the following:
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)
Write each SRC_LANEx 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, SRC_LANE0 = 0.
Rev. 0 | Page 35 of 138
AD9166
Data Sheet
Therefore, Logical Lane 0 obtains data from Physical Lane 0
(SERDIN0 ). To use SERDIN4 as the source for Logical Lane 0
instead, the user must write SRC_LANE0 = 4 (decimal).
Because SYSREF is sampled with fCLK ÷ 4, there is a four-fCLK cycle
ambiguity between the SYSREF edge and fCLK. The phase of the
fCLK ÷ 4 clock used to sample SYSREF is stored in Register 0x037,
Bits[7:0] and Register 0x038, Bits[3:0] as a thermometer code. This
value determines which fCLK cycle the SYSREF edge corresponds
to, which is used to compensate for the cycle ambiguity and
improve deterministic latency uncertainty. The compensation must
be performed outside the AD9166 by delaying or advancing the
data samples, fCLK, or the SYSREF signal to be sampled. After
compensation, the deterministic latency uncertainty can be
improved to 0 fCLK cycles between device resets, as long as the
sample and hold times for SYSREF are met across the device
operating conditions.
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.
Deframer
The AD9166 consists of one quad-byte deframer (QBD). The
QBD accepts 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 (PCLK) cycle.
As an indication whether setup and hold times for SYSREF were
met, monitor the values in SYNC_LMFC_STATx (Register 0x034
and Register 0x035) after a SYSREF edge is sampled, resetting
the register before each reading by writing 0x0 to Register 0x34.
The SYNC_LMFC_STATx value must be constant across
multiple readings. Refer to the Sync Procedure section for more
details.
The deframer uses the JESD204B parameters that the user has
programmed into the register map to identify how the data is
packed, and unpacks it. The JESD204B parameters are
described in detail in the Transport Layer section. Many of the
parameters are also needed in the transport layer to convert
JESD204B frames into samples.
The AD9166 supports 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 1.25 V). When dc-coupled, a small amount of common-
mode current (<500 μA) is drawn from the SYSREF pins. See
Figure 57 for the SYSREF internal circuit.
Descrambler
The AD9166 provides an optional descrambler block using a
self synchronous descrambler with the following polynomial:
1 + x14 + x15
Enabling data scrambling reduces spectral peaks that are
produced when the same data octets repeat from frame to
frame. Data scrambling 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, Bit 7) to 1.
To avoid this common-mode current draw, use a 50% duty cycle
periodic SYSREF signal with ac coupling capacitors. If ac-coupled,
the ac coupling capacitors combine with the resistors shown in
Figure 57 to make a high-pass filter with an RC time constant of
τ = RC. Select C such that τ > 4/SYSREF frequency. In
addition, the edge rate must be sufficiently fast to meet the
SYSREF vs. device clock (fCLK) keep out window requirements.
Synchronizing LMFC Signals
It is possible to use ac-coupled mode without meeting the
frequency to time constant constraints (τ = RC and τ > 4/SYSREF
frequency) by using SYSREF hysteresis (Register 0x088 and
Register 0x089). However, using hysteresis increases the fCLK
keep out window (the setup and hold specifications in Table 10
do not apply) by an amount depending on the SYSREF
frequency, level of hysteresis, capacitor choice, and edge rate.
The first step to ensuring synchronization across links and
devices begins with synchronizing the LMFC signals. In
Subclass 0, the LMFC signal is synchronized to an internal
processing clock. In Subclass 1, LMFC signals are synchronized
to an external SYSREF signal.
SYSREF Signal
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.
3kΩ
SYSREF+
50Ω
19kΩ
The SYSREF signal is sampled by a divide by 4 version of the
device clock (fCLK). For fixed phase alignment between signals,
generate the device clock and SYSREF signals from the same
source, such as the HMC7044 clock generator. When designing
for optimum deterministic latency operation, consider the
timing distribution skew of the SYSREF signal in a multipoint
link system (multichip).
19kΩ
50Ω
3kΩ
SYSREF−
Figure 57. SYSREF Input Circuit
Rev. 0 | Page 36 of 138
Data Sheet
AD9166
3. Optionally, read back the SYSREF count to check whether
the SYSREF pulses are being received.
Sync Processing Modes Overview
The AD9166 supports several LMFC sync processing modes.
These modes are one shot, continuous, and monitor modes. All
sync processing modes perform a phase check to confirm that the
LMFC is phase aligned to an alignment edge. In Subclass 1, the
SYSREF signal acts as the alignment edge. In Subclass 0, an
internal processing clock (PCLK) acts as the alignment edge.
a. Set Register 0x036 = 0. Writing anything to
SYSREF_COUNT resets the count.
b. Set Register 0x034 = 0. Writing anything to
SYNC_LMFC_STAT0 saves the data for readback and
registers the count.
c. Read SYSREF_COUNT from the value from
Register 0x036.
The sync modes are described in the following sections (One-Shot
Sync Mode (SYNC_MODE = Register 0x03A, Bits[1:0] = 0b10),
Continuous Sync Mode (SYNC_MODE = Register 0x03A,
Bits[1:0] = 0b01), Monitor Sync Mode (SYNC_MODE = Register
0x03A, Bits[1:0]) = 0b00), and Sync Procedure). See the Sync
Procedure section for details on the procedure for syncing the
LMFC signals.
4. Perform a one-shot sync.
a. Set Register 0x03A = 0x00. Clear one shot mode if
already enabled.
b. Set Register 0x03A = 0x02. Enable one-shot sync
mode. The state machine enters monitor mode after a
sync occurs.
5. Optionally, read back the SYNC_LMFC_STATx registers to
verify that the synchronization completed correctly.
a. Set Register 0x034 = 0. Register 0x034 must be written
to read the value.
One-Shot Sync Mode (SYNC_MODE = Register 0x03A,
Bits[1:0] = 0b10)
In one-shot sync mode, a phase check occurs on only the first
alignment edge that is received after the sync machine is armed.
After the phase is aligned on the first edge, the AD9166 transitions
to monitor mode. Though an LMFC synchronization occurs only
once, the SYSREF signal can still be continuous. In this case,
the phase is monitored and reported, but no clock phase
adjustment occurs.
b. Read Register 0x035 and Register 0x034 to find the
value of SYNC_LMFC_STATx. It is recommended to
set SYNC_LMFC_STATx to 0 but it can be set to 4, or
a LMFC period in fCLK − 4, due to jitter.
6. Optionally, read back the SYSREF_PHASEx registers to
identify which phase of the divide by 4 was used to sample
SYSREF . Read Register 0x038 and Register 0x037 as
thermometer code. The MSBs of Register 0x037, Bits[7:4]
normally show the thermometer code value.
7. Turn the link on (Register 0x300, Bit 0 = 1).
8. Read back Register 0x302 (dynamic link latency).
9. Repeat reestablishment of the link several times (Step 1 to
Step 7) and note the dynamic link latency values. Based on the
noted values, program the LMFC delay (Register 0x304) and
the LMFC variable (Register 0x306), and then restart the link.
Continuous Sync Mode (SYNC_MODE = Register 0x03A,
Bits[1:0] = 0b01)
Continuous sync mode must be used in Subclass 1 only with a
periodic SYSREF signal. In continuous sync mode, a phase
check/alignment occurs on every alignment edge.
Continuous sync mode differs from one-shot sync mode in two
ways. First, no SPI cycle is required to arm the device. The
alignment edge seen after continuous sync mode is enabled
results in a phase check. Second, a phase check occurs on every
alignment edge in continuous sync mode.
Table 21. Sync Processing Modes
Sync Processing Mode
No synchronization
One shot
SYNC_MODE (Register 0x03A, Bits[1:0])
Monitor Sync Mode (SYNC_MODE = Register 0x03A,
Bits[1:0]) = 0b00)
0b00
0b10
0b01
Monitor sync mode allows the user to monitor the phase error in
real time. Use this sync mode with a periodic SYSREF signal. The
phase is monitored and reported, but no clock phase
adjustment occurs.
Continuous
Table 22. SYSREF Jitter Window Tolerance
SYSREF Jitter Window
Tolerance (fCLK Cycles)
SYSREF_JITTER_WINDOW
(Register 0x039, Bits[5:0])1
When an alignment request (SYSREF edge) occurs, snapshots
of the last phase error are placed into readable registers for
reference (Register 0x037 and Register 0x038, Bits[3:0]), and, if
appropriate, the IRQ_SYSREF_JITTER interrupt is set.
±±
±4
±8
±12
±16
±20
±24
±28
0x00
0x04
0x08
0x0C
0x10
0x14
0x18
0x1C
Sync Procedure
The procedure for enabling the sync is as follows:
1. Set up the DAC; the SERDES PLL locks it and enables the
CDR (see the Start-Up Sequence section).
2. Set Register 0x039 (SYSREF jitter window). A minimum
of four fCLK cycles is recommended. See Table 22 for
settings.
1 The two least significant digits are ignored because the SYSREF± signal is
sampled with a divide by 4 version of fCLK. As a result, the jitter window is set
by the fCLK ÷ 4 clock rather than fCLK. It is recommended that at least a four-
device clock SYSREF± jitter window be chosen.
Rev. 0 | Page 37 of 138
AD9166
Data Sheet
Deterministic Latency
Subclass 1
JESD204B systems contain various clock domains distributed
throughout its 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.
Subclass 1 mode gives deterministic latency and allows latency
to stay repeatable within a specified number of device clock
(fCLK) periods between synchronization events, as specified in
Table 4. This mode requires an external SYSREF signal that is
accurately phase aligned to fCLK
.
Deterministic Latency Requirements
Several key factors are required for achieving deterministic
latency in a JESD204B Subclass 1 system.
The AD9166 supports JESD204B Subclass 0 and Subclass 1
operation, but not Subclass 2. Write the subclass to Register 0x458,
Bits[7:5].
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 PCLK periods, which includes both
variable delays and the variation in fixed delays from lane
to lane, link to link, and device to device in the system.
Subclass 0
The Subclass 0 mode gives deterministic latency to within
32 fCLK cycles. 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.
Link Delay
The link delay of a JESD204B system is the sum of the fixed and
variable delays from the transmitter, channel, and receiver as
shown in Figure 58.
LINK DELAY = FIXED DELAY + VARIABLE DELAY
LOGIC DEVICE
(JESD204B Tx)
CHANNEL
JESD204B Rx
DSP
DAC
POWER CYCLE
VARIANCE
LMFC
DATA AT
Tx INPUT
ILAS
DATA
ALIGNED DATA
AT Rx OUTPUT
ILAS
FIXED DELAY
DATA
VARIABLE
DELAY
Figure 58. JESD204B Link Delay = Fixed Delay + Variable Delay
Rev. 0 | Page 38 of 138
Data Sheet
AD9166
For proper functioning, all lanes on a link must be read during
the same LMFC period. Section 6.1 of the JESD204B specifica-
tion states that the LMFC period must be larger than the maximum
link delay. For the AD9166, this is not necessarily the case.
Instead, the AD9166 uses a local LMFC for each link (LMFCRx)
that can be delayed from the SYSREF aligned LMFC.
The method to select the LMFCDel (Register 0x304), and
LMFCVar (Register 0x306) variables is described in the Link
Delay Setup Example, with Known Delays 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 to absorb all link delay
variation. This write ensures that all data samples have arrived
before reading. By setting these to fixed values across runs and
devices, deterministic latency is achieved.
Because the LMFC is periodic, this delay 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 AD9166 can
achieve proper performance with a smaller total latency.
The receive buffer delay described in the JESD204B specification
takes values from one frame clock cycle to K frame clock cycles,
and the receive buffer delay of the AD9166 takes values from 0
PCLK cycles to 10 PCLK cycles. As a result, up to 10 PCLK
cycles of total delay variation can be absorbed. LMFCVar and
LMFCDel are both in PCLK cycles. The PCLK factor, or
number of frame clock cycles per PCLK cycle, is equal to 4/F.
For more information on this relationship, see the Clock
Relationships section.
Figure 59 and Figure 60 show a case where the link delay is
greater than an LMFC period. Note that it can be accommodated
by delaying LMFCRx.
POWER CYCLE
VARIANCE
LMFC
ILAS
DATA
ALIGNED DATA
EARLY ARRIVING
LMFC REFERENCE
LATE ARRIVING
LMFC REFERENCE
Two examples follow that show how to determine LMFCVar
and LMFCDel. After they are calculated, write LMFCDel into
Register 0x304 for all devices in the system, and write LMFCVar
to Register 0x306 for all devices in the system.
Figure 59. Link Delay > LMFC Period Example
POWER CYCLE
VARIANCE
LMFC
ILAS
DATA
ALIGNED DATA
LMFC
RX
LMFC REFERENCE FOR ALL POWER CYCLES
LMFC_DELAY_0
FRAME CLOCK
Figure 60. LMFC_DELAY_0 to Compensate for Link Delay > LMFC Period
Rev. 0 | Page 39 of 138
AD9166
Data Sheet
4. Calculate MinDelayLane as follows:
Link Delay Setup Example, with Known Delays
MinDelayLane = floor(RxFixed + TxFixed + PCBFixed)
= floor(12 + 13.5 + 0)
= floor(25.5)
MinDelayLane = 25
All the known system delays can be used to calculate LMFCVar
and LMFCDel. The example shown in Figure 61 is demonstrated
in the following steps.
Note that this example is in Subclass 1 to achieve deterministic
latency, which has a PCLK factor (4/F) of two frame clock
cycles per PCLK cycle, and uses K = 32 (frames/multiframe).
Because PCBFixed << PCLK Period, PCBFixed is negligible in
this example and not included in the calculations.
5. Calculate MaxDelayLane as follows:
MaxDelayLane = ceiling(RxFixed + RxVar + TxFixed +
TxVar + PCBFixed))
= ceiling(12 + 2 + 13.5 + 1 + 0)
= ceiling(28.5)
1. Find the receiver delays using Table 6.
RxFixed = 12 PCLK cycles
MaxDelayLane = 29
6. Calculate LMFCVar as follows:
LMFCVar = (MaxDelay + 1) − (MinDelay − 1)
= (29 + 1) − (25 − 1) = 30 − 24
LMFCVar = 6 PCLK cycles
7. Calculate LMFCDel as follows:
LMFCDel = (MinDelay − 1) % (K/PClockFactor)
= ((30 − 1)) % (32/2)
= 29 % 16
LMFCDel = 13 PCLK cycles
8. Write LMFCDel to Register 0x304 for all devices in the
system. Write LMFCVar to Register 0x306 for all devices in
the system.
RxVar = 2 PCLK cycles
2. Find the transmitter delays. The equivalent table in the
example JESD204B core (implemented on a GTH or
gigabit transceiver (GTX) on a Virtex-6 FPGA) states that
the delay is 56 2 byte clock cycles.
3. Because the PCLK Rate = ByteRate/4 as described in the
Clock Relationships section, the transmitter delays in
PCLK cycles are calculated as follows:
TxFixed = 54/4 = 13.5 PCLK cycles
TxVar = 4/4 = 1 PCLK cycle
LMFC
PCLK
FRAME CLOCK
DATA AT Tx FRAMER
ILAS
DATA
ALIGNED LANE DATA
AT Rx DEFRAMER OUTPUT
Tx VAR Rx VAR
DATA
ILAS
DELAY
DELAY
PCB FIXED
DELAY
LMFC
RX
LMFC DELAY = 26 FRAME CLOCK CYCLES
TOTAL VARIABLE
LATENCY = 4
PCLK CYCLES
TOTAL FIXED LATENCY = 30 PCLK CYCLES
Figure 61. LMFC Delay Calculation Example
Rev. 0 | Page 40 of 138
Data Sheet
AD9166
Link Delay Setup Example, Without Known Delay
Link B gives delay values from 5 to 7.
Link C gives delay values from 4 to 7.
If the system delays are not known, the AD9166 can read back
the link latency between LMFCRX for each link and the
SYSREF aligned LMFC. This information is then used to
calculate LMFCVar and LMFCDel. Figure 62 shows how
DYN_LINK_LATENCY_0 (Register 0x302) provides a
readback showing the delay (in PCLK cycles) between the
LMFCRX and the transition from the ILAS to the first data
sample. By repeatedly power cycling and taking this measurement,
the minimum and maximum delays across power cycles can be
determined to calculate LMFCVar and LMFCDel.
The example shown in Figure 62 is demonstrated in the
following steps. Note that this example is in Subclass 1 to
achieve deterministic latency, which has a PCLK factor (frame
rate ÷ PCLK rate) of 4 and uses K = 32. Therefore, PCLK cycles
per multiframe = 8.
1. Calculate the minimum of all delay measurements across
all power cycles, links, and devices as follows:
MinDelay = min(all Delay values) = 4
2. Calculate the maximum of all delay measurements across
In Figure 62, the AD9166 is configured as described in the Sync
Procedure section. Because the purpose of this exercise is to
determine LMFCDel and LMFCVar, the LMFCDel value is
programmed to 0 and the DYN_LINK_LATENCY_0 value is
read from Register 0x302. For Link A, Link B, and Link C, the
system containing the AD9166 (including the transmitter) is
power cycled and configured 20 times.
all power cycles, links, and devices as follows:
MaxDelay = max(all Delay values) = 9
3. Calculate the total delay variation (with guard band) across
all power cycles, links, and devices as follows:
LMFCVar = (MaxDelay + 1) − (MinDelay − 1)
= (9 + 1) − (4 − 1) = 10 − 3 = 7 PCLK cycles
4. Calculate the minimum delay in PCLK cycles (with guard
band) across all power cycles, links, and devices as follows:
The variation in the link latency over the 20 runs is shown in
Figure 62, described as follows:
LMFCDel
= (MinDelay − 1) % (K/PCLK Factor)
= (4 − 1) % 32/4
Link A gives readbacks of 6, 7, 0, and 1. Note that the set of
recorded delay values rolls over the edge of a multiframe at
the boundary of K/PCLK factor = 8. Add the number of
PCLK cycles per multiframe = 8 to the readback values of 0
and 1 because they rolled over the edge of the multiframe.
Delay values range from 6 to 9.
= 3 % 8 = 3 PCLK cycles
5. Write LMFCDel to Register 0x304 for all devices in the system.
6. Write LMFCVar to Register 0x306 for all devices in the system.
LMFC
PCLK
FRAME CLOCK
DYN_LINK_LATENCY_0 COUNT
ALIGNED DATA (LINK A)
ALIGNED DATA (LINK B)
ALIGNED DATA (LINK C)
0
1
2
3
4
5
6
7
0
1
2
3
4
5
6
7
ILAS
DATA
ILAS
DATA
DATA
ILAS
LMFC
RX
DETERMINISTICALLY
DELAYED DATA
ILAS
DATA
LMFC_DELAY_0 = 6
(FRAME CLOCK CYCLES)
LMFC_VAR_0 = 7
(PCLK CYCLES)
Figure 62. Multilink Synchronization Settings, Derived Method Example
SYSREF±
LMFC
RX
ILAS
DATA
ALIGNED DATA
DYN_LINK_LATENCY_0
Figure 63. DYN_LINK_LATENCY_0
Rev. 0 | Page 41 of 138
AD9166
Data Sheet
TRANSPORT LAYER
TRANSPORT LAYER
(QBD)
LANE 0 OCTETS
DAC A_I0[15:0]
DAC A_Q0[15:0]
DELAY
BUFFER 0
F2S_0
LANE 3 OCTETS
PCLK_0
SPI CONTROL
LANE 4 OCTETS
DAC B_I0[15:0]
DAC B_Q0[15:0]
PCLK_0
TO
PCLK_1
FIFO
DELAY
BUFFER 1
F2S_1
LANE 7 OCTETS
PCLK_1
SPI CONTROL
NOTES
F2S_0 AND F2S_1 ARE FRAME TO SAMPLE CONVERSION BLOCK 0 AND BLOCK 1.
Figure 64. Transport Layer Block Diagram
The transport layer receives the descrambled JESD204B frames
and converts them to DAC samples based on the programmed
JESD204B parameters shown in Table 23. The device parameters
are defined in Table 24.
Table 24. JESD204B Device Parameters
Parameter Description
CF
CS
HD
Number of control words per device clock per link.
Not supported, must be 0.
Number of control bits per conversion sample. Not
supported, must be 0.
High density user data format. Used when samples
must be split across lanes. Set to1 always, even when
F does not equal 1. Otherwise, a link configuration
error triggers and the IRQ_ILAS flag is set.
Table 23. JESD204B Transport Layer Parameters
Parameter Description
F
K
L
Number of octets per frame per lane: 1, 2, or 4
Number of frames per multiframe: K = 32
Number of lanes per converter device (per link), as
follows: 4 or 8
N
Converter resolution = 16.
M
S
Number of converters per device (per link), as
follows: 1 or 2 (1 is used for real data mode; 2 is
used for complex data modes)
N' (or NP)
Total number of bits per sample = 16.
Number of samples per converter, per frame: 1 or 2
Rev. 0 | Page 42 of 138
Data Sheet
AD9166
Certain combinations of these JESD2014B parameters are
supported by the AD9166, as shown in Table 26 (JESD204B
interpolation rate and number of lanes), Table 25 (fixed values),
and Table 27 (supported and unsupported interpolation rates).
Table 25 lists JESD204B parameters that have fixed values.
Table 25. JESD204B Parameters with Fixed Values
Parameter
Value
32
16
16
0
K
N
NP
CF
HD
CS
See Table 26 for a list of supported interpolation rates and the
number of lanes that is supported for each rate. Table 26 lists the
JESD204B parameters for each of the interpolation and number
of lanes configuration, and gives an example lane rate for a
5 GHz device clock (fCLK).
1
0
Table 26. JESD204B Parameters for Interpolation Rate and Number of Lanes
Interpolation Rate
No. of Lanes
M
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
F
1
2
1
2
1
4
1
2
1
4
1
2
1
2
4
1
2
1
2
4
1
2
1
4
2
4
1
2
1
4
2
4
1
2
1
S
4
3
2
3
2
3
1
3
2
3
1
3
2
1
3
1
3
2
1
3
1
3
2
1
1
3
1
3
2
1
1
3
1
3
2
PCLK Period (fCLK Cycles)
LMFC Period (fCLK Cycles)
Lane Rate at fCLK = 5 GHz
1
2
2
3
3
4
4
4
4
6
6
6
6
8
8
8
8
6
8
6
8
3
4
6
8
3
4
6
8
2
3
4
6
8
2
3
4
6
8
1
2
3
4
6
8
1
2
3
4
6
8
16
12
16
18
24
12
16
24
32
18
24
36
48
16
24
32
48
64
24
36
48
72
96
16
32
48
64
96
128
24
48
72
96
144
192
128
192
128
288
192
384
128
384
256
576
192
576
384
256
768
256
768
512
384
1152
384
1152
768
512
512
1536
512
1536
1024
768
768
2304
768
2304
1536
12.5
16.661
12.5
11.11
8.33
16.661
12.5
8.33
6.25
11.11
8.33
5.55
4.16
12.5
8.33
6.25
4.16
3.12
8.33
5.55
4.16
2.77
2.08
12.5
6.25
4.16
3.12
2.08
1.56
8.33
4.16
2.77
2.08
1.38
1.04
8
8
12
12
12
12
12
16
16
16
16
16
16
24
24
24
24
24
24
1 Maximum lane rate is 12.5 GHz. These modes must be run with the DAC rate below 3.75 GHz.
Rev. 0 | Page 43 of 138
AD9166
Data Sheet
A value of yes in Table 27 means the interpolation rate is supported
for the number of lanes. A blank cell means it is not supported.
Data Flow Through the JESD204B Receiver
The link configuration parameters determine how the serial bits
on the JESD204B receiver interface are deframed and passed on
to the DACs as data samples.
Table 27. Interpolation Rates and Number of Lanes
Interpolation
8
6
4
3
2
1
Deskewing and Enabling Logical Lanes
1×
2×
3×
4×
6×
8×
12×
16×
24×
Yes1
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes1
Yes
Yes
Yes
Yes
Yes
Yes
Yes
After proper configuration, the logical lanes are automatically
deskewed. All logical lanes are enabled or not based on the lane
number setting in Register 0x110, Bits[7:4]. The physical lanes
are all powered up by default.
Yes
Yes
Yes
Yes
Yes
Yes
Yes1
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
To disable power to physical lanes that are not being used, set Bit x
in Register 0x201 to 1 to disable Physical Lane x, and keep it at 0
to enable it.
Yes
Yes
JESD204B TEST MODES
PHY PRBS Testing
1 These modes restrict the maximum device clock rate to 5 GHz.
Configuration Parameters
The JESD204B receiver on the AD9166 includes a pseudorandom
binary sequence (PRBS) pattern checker on the back end of its
physical layer. This functionality enables bit error rate (BER)
testing of each physical lane of the JESD204B link. The PHY
PRBS pattern checker does not require that the JESD204B link be
established. The pattern checker can synchronize with a PRBS7,
PRBS15, or PRBS31 data pattern. The PRBS pattern can be
verified on multiple lanes simultaneously. The error counts for
failing lanes are reported for one JESD204B lane at a time.
The AD9166 modes refer to the link configuration parameters
for L, K, M, N, NP, S, and F. Table 28 provides the description
and addresses for these settings.
Table 28. Configuration Parameters
JESD204B
Setting
Description
Address
L − 1
Number of lanes minus 1.
Register 0x453,
Bits[4:0]
F − 1
K − 1
M − 1
N − 1
NP − 1
S − 1
HD
Number of ((octets per frame) per
lane) minus 1.
Number of frames per multiframe
minus 1.
Number of converters minus 1.
Converter bit resolution minus 1.
Bit packing per sample minus 1.
Register 0x454,
Bits[7:0]
Register 0x455,
Bits[4:0]
Register 0x456,
Bits[7:0]
Register 0x457,
Bits[4:0]
Register 0x458,
Bits[4:0]
Register 0x459,
Bits[4:0]
Register
0x45A, Bit 7
Table 29. PHY PRBS Pattern Selection
PHY_PRBS_PAT_SEL Setting
(Register 0x316, Bits[3:2])
PRBS Pattern
PRBS7
PRBS15
0b00 (default)
0b01
0b10
PRBS31
The process for performing PRBS testing on the AD9166 is as
follows, using Table 29 for reference.
Number of ((samples per
converter) per frame) minus 1.
High density format. Set to 1 if F =
1. Leave at 0 if F ≠ 1.
1. Start sending a PRBS7, PRBS15, or PRBS31 pattern from
the JESD204B transmitter.
2. Select and write the appropriate PRBS pattern to
Register 0x316, Bits[3:2], as shown in Table 29.
3. Enable the PHY test for all lanes being tested by writing to
PHY_TEST_EN (Register 0x315). Each bit of Register 0x315
enables the PRBS test for the corresponding lane. For example,
writing a 1 to Bit 0 enables the PRBS test for Physical Lane 0.
4. Toggle PHY_TEST_RESET (Register 0x316, Bit 0) from 0
to 1 then back to 0.
DID
Device ID. Match the device ID
sent by the transmitter.
Register 0x450,
Bits[7:0]
BID
Bank ID. Match the bank ID sent by Register 0x451,
the transmitter. Bits[7:0]
Lane ID for Lane 0. Match the Lane Register 0x452,
LID0
ID sent by the transmitter on
Logical Lane 0.
Bits[4:0]
JESDV
JESD204x version. Match the
version sent by the transmitter
(0x0 = JESD204A, 0x1 =
JESD204B).
Register 0x459,
Bits[7:5]
5. Set PHY_PRBS_TEST_THRESHOLD_xBITS (Bits[23:0],
Register 0x319 to Register 0x317) as desired.
6. Write a 0 and then a 1 to PHY_TEST_START (Register 0x316,
Bit 1). The rising edge of PHY_TEST_START starts the test.
a. (Optional) In some cases, it may be necessary to
repeat Step 4 at this point. Toggle PHY_TEST_RESET
(Register 0x316, Bit 0) from 0 to 1, then back to 0.
7. Wait 500 ms.
8. Stop the test by writing PHY_TEST_START
(Register 0x316, Bit 1) = 0.
Rev. 0 | Page 44 of 138
Data Sheet
AD9166
9. Read the PRBS test results.
1. Synchronize the JESD204B link.
a. Each bit of PHY_PRBS_PASS (Register 0x31D)
corresponds to one SERDES lane (0 = fail, 1 = pass).
b. The number of PRBS errors seen on each failing lane
can be read by writing the lane number (0 to 7) to PHY_
SRC_ERR_CNT (Register 0x316, Bits[6:4]) and reading
the PHY_PRBS_ERR_CNT_xBITS (Register 0x31C to
Register 0x31A). The maximum error count is 224 − 1. If
all bits of Register 0x31C to Register 0x31A are high, the
maximum error count on the selected lane is exceeded.
2. Enable the short transport layer at the JESD204B
transmitter.
3. Depending on JESD204B case, there may be up to two
DACs, and each frame may contain up to four DAC
samples. Configure the SHORT_TPL_REF_SP_MSB bits
(Register 0x32E) and SHORT_TPL_REF_SP_LSB bits
(Register 0x32D) to match one of the samples for one
converter within one frame.
4. Set SHORT_TPL_SP_SEL (Register 0x32C, Bits[7:4]) to
select the sample within one frame for the selected
converter according to Table 30.
Transport Layer Testing
The JESD204B receiver in the AD9166 supports the short
transport layer test as described in the JESD204B standard. This
test can be used to verify the data mapping between the
JESD204B transmitter and receiver. To perform the short
transport layer test, this function must be implemented in the
logic device and enabled there. Before running the test on the
receiver side, the link must be established and running without
errors.
5. Set SHORT_TPL_TEST_EN (Register 0x32C, Bit 0) to 1.
6. Set SHORT_TPL_TEST_RESET (Register 0x32C, Bit 1) to
1, then back to 0.
7. Wait for the desired time. The desired time is calculated as
1/(sample rate × BER). For example, given a bit error rate
of BER = 1 × 10−10 and a sample rate = 1 GSPS, the desired
time = 10 sec. Then, set SHORT_TPL_TEST_EN to 0.
8. Read the test result at SHORT_TPL_FAIL (Register 0x32F,
Bit 0).
9. Choose another sample for the same or another converter
to continue with the test, until all samples for both
converters from one frame are verified. (Note that the
converter count is M = 2 for all interpolator modes on the
AD9166 to enable complex signal processing.)
The short transport layer test ensures that each sample from
each converter is mapped appropriately according to the
number of converters (M) and the number of samples per
converter (S). As specified in the JESD204B standard, the
converter manufacturer specifies which test samples are
transmitted. Each sample must have a unique value. For
example, if M = 2 and S = 2, four unique samples are
transmitted repeatedly until the test is stopped.
Consult Table 30 for a guide to the test sample alignment. Note
that the sample order for 1×, eight-lane mode has Sample 1 and
Sample 2 swapped. Also, the short transport layer test for the
three-lane and six-lane options is not functional and always fails.
The expected sample must be programmed into the device and
the expected sample is compared to the received sample, one
sample at a time until all samples are tested. The process for
performing this test on the AD9166 is described as follows:
Table 30. Short Transport Layer Test Samples Assignment1
JESD204x Mode
Required Samples from JESD204x Tx
Samples Assignment
1× Eight-Lane (L = 8, M = 1, F = 1, S = 4)
Send four samples: M0S0, M0S1, M0S2,
M0S3, and repeat
SP0: M0S0, SP4: M0S0, SP8: M0S0, SP12: M0S0
SP1: M0S2, SP5: M0S2, SP9: M0S2, SP13: M0S2
SP2: M0S1, SP6: M0S1, SP10: M0S1, SP14: M0S1
SP3: M0S3, SP7: M0S3, SP11: M0S3, SP15: M0S3
SP0: M0S0, SP4: M0S0, SP8: M0S0, SP12: M0S0
SP1: M1S0, SP5: M1S0, SP9: M1S0, SP13: M1S0
SP2: M0S1, SP6: M0S1, SP10: M0S1, SP14: M0S1
SP3: M1S1, SP7: M1S1, SP11: M1S1, SP15: M1S1
2× Eight-Lane (L = 8, M = 2, F = 1, S = 2)
3× Eight-Lane (L = 8, M = 2, F = 1, S = 2)
4× Eight-Lane (L = 8, M = 2, F = 1, S = 2)
6× Eight-Lane (L = 8, M = 2, F = 1, S = 2)
8× Eight-Lane (L = 8, M = 2, F = 1, S = 2)
12× Eight-Lane e (L = 8, M = 2, F = 1, S = 2)
16× Eight-Lane (L = 8, M = 2, F = 1, S = 2)
24× Eight-Lane (L = 8, M = 2, F = 1, S = 2)
2× Six-Lane (L = 6, M = 2, F = 2, S = 3)
3× Six-Lane (L = 6, M = 2, F = 2, S = 3)
4× Six-Lane (L = 6, M = 2, F = 2, S = 3)
6× Six-Lane (L = 6, M = 2, F = 2, S = 3)
Send four samples: M0S0, M0S1, M1S0,
M1S1, and repeat
Send six samples: M0S0, M0S1, M0S2,
M1S0, M1S1, M1S2, and repeat
Test hardware is not functional; short transport
layer always fails
Rev. 0 | Page 45 of 138
AD9166
Data Sheet
JESD204x Mode
Required Samples from JESD204x Tx
Samples Assignment
8× Six-Lane (L = 6, M = 2, F = 2, S = 3)
12× Six-Lane (L = 6, M = 2, F = 2, S = 3)
16× Six-Lane (L = 6, M = 2, F = 2, S = 3)
24× Six-Lane (L = 6, M = 2, F = 2, S = 3)
4× Three-Lane (L = 3, M = 2, F = 4, S = 3)
6× Three-Lane (L = 3, M = 2, F = 4, S = 3)
8× Three-Lane (L = 3, M = 2, F = 4, S = 3)
12× Three-Lane (L = 3, M = 2, F = 4, S = 3)
16× Three-Lane (L = 3, M = 2, F = 4, S = 3)
24× Three-Lane (L = 3, M = 2, F = 4, S = 3)
4× Four-Lane (L = 4, M = 2, F = 1, S = 1)
6× Four-Lane (L = 4, M = 2, F = 1, S = 1)
8× Four-Lane (L = 4, M = 2, F = 1, S = 1)
12× Four-Lane (L = 4, M = 2, F = 1, S = 1)
16× Four-Lane (L = 4, M = 2, F = 1, S = 1)
24× Four-Lane (L = 4, M = 2, F = 1, S = 1)
8× Two-Lane (L = 2, M = 2, F = 2, S = 1)
12× Two-Lane (L = 2, M = 2, F = 2, S = 1)
16× Two-Lane (L = 2, M = 2, F = 2, S = 1)
24× Two-Lane (L = 2, M = 2, F = 2, S = 1)
16× One-Lane (L = 1, M = 2, F = 4, S = 1)
24× One-Lane (L = 1, M = 2, F = 4, S = 1)
Send two samples: M0S0, M1S0, repeat
SP0: M0S0, SP4: M0S0, SP8: M0S0, SP12: M0S0
SP1: M1S0, SP5: M1S0, SP9: M1S0, SP13: M1S0
SP2: M0S0, SP6: M0S0, SP10: M0S0, SP14: M0S0
SP3: M1S0, SP7: M1S0, SP11: M1S0, SP15: M1S0
1 Mx is the converter number and Sy is the sample number. For example, M0S0 means Converter 0, Sample 0. SPx is the sample pattern word number. For example, SP0
means Sample Pattern Word 0.
Repeated CGS and ILAS Test
JESD204B ERROR MONITORING
As per Section 5.3.3.8.2 of the JESD204B specification, the
AD9166 can check that a constant stream of /K28.5/ characters
is being received, or that code group synchronization (CGS)
followed by a constant stream of ILAS is being received.
Disparity, Not in Table, and Unexpected Control (K)
Character Errors
As per Section 7.6 of the JESD204B specification, the AD9166
can detect disparity errors, not in table (NIT) errors, and
unexpected control (K) character errors, and can optionally
issue a sync request and reinitialize the link when errors occur.
To run a repeated CGS test, send a constant stream of /K28.5/
characters to the AD9166 SERDES inputs. Next, set up the
device and enable the links. Ensure that the /K28.5/ characters are
Note that the disparity error counter counts all characters with
invalid disparity, regardless of whether they are in the 8-bit/10-bit
decoding table. This error counting method is a minor deviation
from the JESD204B specification, which only counts disparity
errors when they are in the 8-bit/10-bit decoding table.
SYNCOUT
being received by verifying that
is deasserted and
that CGS has passed for all enabled link lanes by reading
Register 0x470.
To run the CGS followed by a repeated ILAS sequence test,
follow the procedure to set up the links. However, before
performing the last write (enabling the links), enable the ILAS
test mode by writing a 1 to Register 0x477, Bit 7. Then, enable the
links. When the device recognizes four CGS characters on each
Several other interpretations of the JESD204B specification are
noted in this section. When three NIT errors are injected to one
lane and QUAL_RDERR (Register 0x476, Bit 4) = 1, the readback
values of the bad disparity error (BDE) count register is 1.
Reporting of disparity errors that occur at the same character
position of a NIT error is disabled. No such disabling is
performed for the disparity errors in the characters after a NIT
error. Therefore, it is expected behavior that a NIT error may
result in a BDE error.
SYNCOUT
lane, it desserts the
. At this point, the transmitter
starts sending a repeated ILAS sequence.
Read Register 0x473 to verify that initial lane synchronization has
passed for all enabled link lanes.
A resync is triggered when four NIT errors are injected with
Register 0x476, Bit 4 = 1. When this bit is set, the error counter
does not distinguish between a concurrent invalid symbol with the
wrong running disparity but is in the 8-bit/10-bit decoding
table, and a NIT error. Thus, a resync can be triggered when
Rev. 0 | Page 46 of 138
Data Sheet
AD9166
four NIT errors are injected because they are not distinguished
from disparity errors.
Error Counter and Interrupt Request Control
For error counter and interrupt request control, follow these
steps:
Checking Error Counts
The error count can be checked for BDEs, NIT errors, and
unexpected K (UEK) character errors. The error counts are on a
per lane and per error type basis. Each error type and lane has a
register dedicated to it. To check the error count, the following
steps must be performed:
1. Enable the interrupts. Enable the JESD204B interrupts. The
interrupts for the UEK, NIT, and BDE error counters are in
Register 0x4B8, Bits[7:5]. There are other interrupts to
monitor when bringing up the link, such as interlane
deskewing, initial lane sync, good check sum, frame sync,
code group sync (Register 0x4B8, Bits[4:0]), and
configuration mismatch (Register 0x4B9, Bit 0). These bits
are off by default but can be enabled by writing 0b1 to the
corresponding bit.
2. Read the JESD204B interrupt status. The interrupt status
bits are in Register 0x4BA, Bits[7:0] and Register 0x4BB,
Bit 0, with the status bit position corresponding to the
enable bit position.
1. Choose and enable the type of errors to monitor by
selecting them in Register 0x480, Bits[5:3] to Register 0x487,
Bits[5:3]. UEK, BDE, and NIT error monitoring can be
selected for each lane by writing a 1 to the appropriate bit,
as described in the register map. These bits are enabled by
default.
2. The corresponding error counter reset bits are in
Register 0x480, Bits[2:0] to Register 0x487, Bits[2:0].
Write a 0 to the corresponding bit to reset that error
counter.
3. Registers 0x488, Bits[2:0] to Register 0x48F, Bits[2:0] have
the terminal count hold indicator for each error counter. If
this flag is enabled, when the terminal error count of 0xFF
is reached, the counter ceases counting and holds that value
until reset. Otherwise, it wraps to 0x00 and continues
counting. Select the desired behavior, and program the
corresponding register bits per lane.
3. It is recommended to enable all interrupts that are planned
to be used prior to bringing up the JESD204B link. When
the link is up, the interrupts can be reset and then used to
monitor the link status.
SYNCOUT
Monitoring Errors via
As per the JESD204B specifications, when one or more BDE,
NIT, or unexpected control character (including UEK) errors
SYNCOUT
occur, the error is reported on the
pin by asserting
signal for exactly two frame periods. For the
SYNCOUT
SYNCOUT
the
AD9166, the width of the
to ½, 1, or 2 PCLK cycles. The settings to achieve a
pulse of two frame clock cycles are listed in Table 31.
Check for Error Count Overthreshold
pulse can be programmed
SYNCOUT
To check for the error count over threshold, follow these steps:
1. Define the error counter threshold. The error counter
threshold can be set to a user defined value in Register 0x47C,
or remain at the default value of 0xFF. When the error
SYNCOUT
Table 31. Setting
PCLK Factor (Frames
F1 per PCLK)
Error Pulse Duration
SYNC_ERR_DUR (Register 0x312,
SYNCOUT
threshold is reached, an IRQ is generated,
is
Bits[7:4]) Setting2
asserted, or both, depending on the mask register settings.
This single error threshold is used for all three types of
errors (UEK, NIT, and BDE).
1
2
4
4
2
1
0 (default)
1
2
SYNCOUT
2. Set the SYNC_ASSERT_MASK bits. The
1 F is a link configuration parameter (see Table 28).
assertion behavior is set in Register 0x47D, Bits[2:0]. By
default, when any error counter of any lane is equal to the
2
SYNCOUT±
These register settings assert the
pulse widths.
signal for two frame clock cycle
SYNCOUT
threshold, it asserts
0b111).
(Register 0x47D, Bits[2:0] =
UEK, NIT, and BDE IRQs
For UEK, NIT, and BDE errors, error count overthreshold events
are available as IRQ events. Enable these events by writing to
Register 0x4B8, Bits[7:5]. After the IRQs are enabled, the IRQ event
status can be read at Register 0x4BA, Bits[7:5].
3. Read the indicator for error count reached. Each error
counter has an indicator for when a terminal count is reached,
per lane. This indicator is set to 1 when the terminal count
of an error counter for a particular lane has been reached.
These status bits are located in Register 0x490, Bits[2:0] to
Register 0x497, Bits[2:0]. These registers also indicate
whether a particular lane is active by setting Bit 3 = 0b1.
See the Interrupt Request Operation section for more information
on IRQs and see the Error Counter and Interrupt Request
Control section for information on resetting the IRQ.
Rev. 0 | Page 47 of 138
AD9166
Data Sheet
Bit x of GOOD_CHECKSUM (Register 0x472) is high if the
checksum sent over the lane matches the sum of the JESD204B
parameters sent over the lane during ILAS for Link Lane x. The
parameters can be added either by summing the individual fields
in registers or summing the packed register. If Register 0x300,
Bit 6 = 0 (default), the calculated checksums are the lower eight
bits of the sum of the following fields: DID, BID, LIDx, SCR, L − 1,
F − 1, K − 1, M − 1, N − 1, SUBCLASSV, NP − 1, JESDV, S − 1,
and HD. If Register 0x300, Bit 6 = 1, the calculated checksums
are the lower eight bits of the sum of Register 0x400 to
Errors Requiring Reinitializing
A link reinitialization automatically occurs when four invalid
disparity characters are received per the JESD204B specification.
When a link reinitialization occurs, the resync request is five
frames and nine octets long.
The user can optionally reinitialize the link when the error
count for disparity errors, NIT errors, or UEK character errors
reaches a programmable error threshold. The process to enable
the reinitialization feature for certain error types is as follows:
Register 0x40C and LIDx, where x refers to a Link Lane x.
1. Choose and enable which errors to monitor by selecting
them in Register 0x480, Bits[5:3] to Register 0x487,
Bits[5:3]. UEK, BDE, and NIT error monitoring can be
selected for each lane by writing a 1 to the appropriate bit,
as described in Table 46. These bits are enabled by default.
2. Enable the sync assertion mask for each type of error by
writing to SYNC_ASSERT_MASK (Register 0x47D,
Bits[2:0]) according to Table 32.
3. Program the desired error counter threshold into
ERRORTHRES (Register 0x47C).
4. For each error type enabled in the SYNC_ASSERT_MASK
register, if the error counter on any lane reaches the
Bit x of INIT_LANE_SYNC (Register 0x473) is high if Link
Lane x passed the initial lane alignment sequence.
CGS, Frame Sync, Checksum, and ILAS IRQs
Fail signals for CGS, frame sync, checksum, and ILAS are available
as IRQ events. Enable them by writing to Register 0x4B8,
Bits[3:0]. The IRQ event status can be read at Register 0x4BA,
Bits[3:0] after the IRQs are enabled.
Write a 1 to Register 0x4BA, Bit 0 to reset the CGS IRQ.
Write a 1 to Register 0x4BA, Bit 1 to reset the frame sync IRQ.
Write a 1 to Register 0x4BA, Bit 2 to reset the checksum IRQ.
Write a 1 to Register 0x4BA, Bit 3 to reset the ILAS IRQ.
SYNCOUT
programmed threshold,
falls, issuing a sync
request. Note that all error counts are reset when a link
reinitialization occurs. The IRQ does not reset and must be
reset manually.
See the Interrupt Request Operation section for more information.
Configuration Mismatch IRQ
Table 32. Sync Assertion Mask (SYNC_ASSERT_MASK,
Address 0x47D)
Bit No. Bit Name Description
The AD9166 has a configuration mismatch flag that is available
as an IRQ event. Use Register 0x4B9, Bit 0 to enable the mismatch
flag (it is enabled by default), and then use Register 0x4BB, Bit 0
to read back its status and reset the IRQ signal. See the Interrupt
Request Operation section for more information.
2
UEK
Set to 1 to assert SYNCOUT± if the UEK
character error count reaches the
threshold
The configuration mismatch event flag is high when the link
configuration settings (in Register 0x450 to Register 0x45D) do
not match the JESD204B transmitted settings (Register 0x400 to
Register 0x40D).
1
0
NIT
Set to 1 to assert SYNCOUT± if the NIT error
count reaches the threshold
Set to 1 to assert SYNCOUT± if the disparity
error count reaches the threshold
BDE
This function is different from the good checksum flags in
Register 0x472. The good checksum flags ensure that the transmit-
ted checksum matches a calculated checksum based on the
transmitted settings. The configuration mismatch event ensures
that the transmitted settings match the configured settings.
CGS, Frame Sync, Checksum, and ILAS Monitoring
Monitor Register 0x470 to Register 0x473 to verify that each
stage of the JESD204B link establishment has occurred.
Bit x of CODE_GRP_SYNC (Register 0x470) is high if Link
Lane x received at least four K28.5 characters and passed CGS.
HARDWARE CONSIDERATIONS
Bit x of FRAME_SYNC (Register 0x471) is high if Link Lane x
completed initial frame synchronization.
See the Applications Information section for information on
hardware considerations.
Rev. 0 | Page 48 of 138
Data Sheet
AD9166
MAIN DIGITAL DATAPATH
HB
2×
INV
SINC
NCO
HB
2×
JESD
HB
2×,
4×,
8×
HB
3×
Figure 65. Block Diagram of the Main Digital Datapath
The block diagram in Figure 65 shows the functionality of the
main digital datapath. The digital processing includes an input
interpolation block with choice of bypass 1×, 2×, or 3× interpo-
lation, three additional 2× half-band interpolation filters, a final
2× NRZ mode interpolator filter, FIR85, that can be bypassed,
and a quadrature modulator that consists of a 48-bit NCO and
an inverse sinc block.
DATA FORMAT
The input data format for all modes on the AD9166 is 16-bit,
twos complement. The digital datapath and the DAC decoder
operate in twos complement format. The DAC is a current
steering DAC and cannot represent 0. The DAC must either
source or sink current. As a result, when the 0 of twos
complement is represented in the DAC, it is a +1, and all the
positive values thereafter are shifted by +1. This mapping error
introduces a ½ LSB shift in the DAC output. The leakage can
become apparent when using the NCO to shift a signal that is
above or below 0 Hz when synthesized. The NCO frequency is
seen as a small spur at the NCO frequency tuning word.
All of the interpolation filters accept I and Q data streams as a
complex data stream. Similarly, the quadrature modulator and
inverse sinc function also accept input data as a complex data
stream. Thus, any use of the digital datapath functions requires
the input data to be a complex data stream.
In bypass mode (1× interpolation), the input data stream is
expected to be real data.
To avoid the NCO frequency leakage, operate the DAC with a
slight digital backoff of one or several codes, and then add 1 to
all values in the data stream. These actions remove the NCO
frequency leakage but cause a ½ LSB dc offset. This small dc
offset is benign to the DAC and does not affect most applications
because the DAC output is ac-coupled through dc blocking
capacitors.
Table 33. Pipeline Delay (Latency) for Various Datapath
Configurations
FIR85 Filter
Inverse
Pipeline Delay1
NCO (fCLK Cycles)
Mode
On
Bandwidth Sinc
NCO only
1× (Bypass) No
1× (Bypass) No
No
N/A2
N/A2
N/A2
80%
90%
80%
80%
80%
80%
80%
90%
80%
80%
80%
80%
80%
80%
No
No
Yes
No
No
No
Yes
Yes
Yes
No
No
No
No
No
No
No
No
Yes
No
No
No
No
No
No
No
Yes
No
No
No
No
No
No
No
No
48
113
137
155
176
202
185
239
279
168
202
308
332
602
674
1188
1272
INTERPOLATION FILTERS
The main digital path contains five half-band interpolation
filters, plus a final half-band interpolation filter that is used in
2× NRZ mode. The filters are cascaded, as shown in Figure 65.
2×
2×
2×
2×
2×
2×
3×
3×
4×
6×
8×
12×
16×
24×
No
No
Yes
No
Yes
Yes
No
No
No
No
No
No
No
No
The first pair of filters is a 2× (HB2) or 3× (HB3) filter. Each of
these filters has two options for bandwidth, 80% or 90%. The
80% filters are lower power than the 90% filters. The filters
default to the lower power 80% bandwidth. To select the filter
bandwidth as 90%, program the FILT_BW bit in the
DATAPATH_CFG register to 1 (Register 0x111, Bit 4 = 0b1).
Following the first pair of filters is a series of 2× half-band
filters, each of which halves the usable bandwidth of the
previous one. HB4 has 45%, HB5 has 22.5%, and HB6 has
11.25% of the fDATA bandwidth.
The final half-band filter, FIR85, is used in 2× NRZ mode.
FIR85 is clocked at the 2 × fCLK rate and has a usable bandwidth
of 45% of the fCLK rate. The FIR85 filter is a complex filter, and
therefore the bandwidth is centered at 0 Hz. The FIR85 filter is
used in conjunction with the complex interpolation modes to
allow doubling the DAC update rate, thus moving the image
spur further from the signal. Table 34 shows how to select each
available interpolation mode, their usable bandwidths, and their
1 The pipeline delay given is a representative number, and may vary by a cycle
or two based on the internal handoff timing conditions at startup.
2 N/A means not applicable.
The pipeline delay changes based on the digital datapath
functions that are selected. See Table 33 for examples of the
pipeline delay per block. These delays are in addition to the
JESD204B latency.
Rev. 0 | Page 49 of 138
AD9166
Data Sheet
maximum data rates. Calculate the available signal bandwidth,
BWSIGNAL, as follows:
Filter Performance Beyond Specified Bandwidth
Some of the interpolation filters are specified to 0.4 × fDATA (with
a pass band). The filters can be used slightly beyond this ratio at
the expense of increased pass-band ripple and decreased interpo-
lation image rejection.
BWSIGNAL = BWFILT × (fCLK/InterpolationFactor)
where BWFILT is the interpolator filter bandwidth.
Filter Performance
90
80
70
60
50
40
30
20
0
The interpolation filters interpolate between existing data in
such a way that they minimize changes in the incoming data
while suppressing the creation of interpolation images. This
datapath is shown for each filter in Figure 66.
–0.1
–0.2
–0.3
–0.4
–0.5
–0.6
The usable bandwidth (as shown in Table 34) is defined as the
frequency band over which the filters have a pass-band ripple of
less than 0.001 dB and an image rejection of greater than 85 dB.
A conceptual drawing that shows the relative bandwidth of each
of the filters is shown in Figure 66. The maximum pass band
amplitude of all filters is the same. They are different in the
illustration to improve understanding.
IMAGE REJECTION
PASS-BAND RIPPLE
40
41
42
43
44
45
1×
2×
3×
4×
6×
8×
12×
16×
24×
FIR85
BANDWIDTH (% fDATA
)
Figure 67. Interpolation Filter Performance Beyond Specified Bandwidth for
the 80% Filters
Figure 67 shows the performance of the interpolation filters
beyond 0.4 × fDATA. The ripple increases more slowly than the
image rejection decreases. This means that if the application can
tolerate degraded image rejection from the interpolation filters,
more bandwidth can be used.
Most of the filters are specified to 0.45 × fDATA (with pass band).
Figure 68 to Figure 75 show the filter response for each of the
interpolator filters on the AD9166.
FREQUENCY (MHz)
Figure 66. All Band Responses of Interpolation Filters
Table 34. Interpolation Modes and Usable Bandwidth
Interpolation Mode
INTERP_MODE, Register 0x110, Bits[3:0] Available Signal Bandwidth1 Maximum fDATA (MHz)
2
1× (Bypass)
2×
3×
4×
6×
8×
12×
16×
24×
0x00
0x01
0x02
0x03
0x04
0x05
0x06
0x07
0x08
fDAC/2
fDAC
Bandwidth × fDATA/2
Bandwidth × fDATA/2
Bandwidth × fDATA/2
Bandwidth × fDATA/2
Bandwidth × fDATA/2
Bandwidth × fDATA/2
Bandwidth × fDATA/2
Bandwidth × fDATA/2
fDAC/22
fDAC/3
fDAC/4
fDAC/6
fDAC/8
fDAC/12
fDAC/16
fDAC/24
2× NRZ (Register 0x111, Bit 0 = 1) Any combination3
0.45 × fCLK
fCLK (real) or fCLK/2 (complex)2
4
1 The data rate (fDATA) for all interpolator modes is a complex data rate, meaning both I data and Q data run at that rate. The available signal bandwidth is the data rate
multiplied by the bandwidth of the initial 2× or 3× interpolator filters, which can be set to bandwidth = 80% or bandwidth = 90%. This bandwidth is centered at 0 Hz.
2 The maximum speed for 1× and 2× interpolation is limited by the JESD204B interface.
3 The 2× NRZ filter, FIR85, can be used with any of the interpolator combinations.
4 The bandwidth of the FIR85 filter is centered at 0 Hz.
Rev. 0 | Page 50 of 138
Data Sheet
AD9166
20
20
0
0
–20
–40
–60
–80
–100
–120
–140
–160
–20
–40
–60
–80
–100
–120
–140
NORMALIZED FREQUENCY (Rad/Sample)
NORMALIZED FREQUENCY (Rad/Sample)
Figure 68. First 2× Half-Band 80% Filter Response
Figure 71. 3× Third-Band 90% Filter Response
20
0
20
0
–20
–20
–40
–60
–80
–100
–120
–140
–40
–60
–80
–100
–120
–140
–160
NORMALIZED FREQUENCY (Rad/Sample)
NORMALIZED FREQUENCY (Rad/Sample)
Figure 69. First 2× Half-Band 90% Filter Response
Figure 72. Second 2× Half-Band 45% Filter Response
20
0
20
0
–20
–20
–40
–40
–60
–60
–80
–80
–100
–120
–140
–160
–100
–120
–140
–160
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
NORMALIZED FREQUENCY (Rad/Sample)
NORMALIZED FREQUENCY (Rad/Sample)
Figure 73. Third 2× Half-Band 22.5% Filter Response
Figure 70. 3× Third-Band 80% Filter Response
Rev. 0 | Page 51 of 138
AD9166
Data Sheet
20
48-Bit Dual Modulus NCO
0
This modulation mode uses an NCO, a phase shifter, and a
complex modulator to modulate the signal by a programmable
carrier signal as shown in Figure 76. This configuration allows
output signals to be placed anywhere in the output spectrum
with very fine frequency resolution.
–20
–40
–60
–80
–100
–120
–140
The NCO produces a quadrature carrier to translate the input
signal to a new center frequency. A quadrature carrier is a pair of
sinusoidal waveforms of the same frequency, offset 90° from
each other. The frequency of the quadrature carrier is set via an
FTW. The quadrature carrier is mixed with the I and Q data and
then summed into the I and Q datapaths, as shown in Figure 76.
NORMALIZED FREQUENCY (Rad/Sample)
Integer NCO Mode
Figure 74. Fourth 2× Half-Band 11.25% Filter Response
The main 48-bit NCO can be used as an integer NCO by using
the following formula to create the FTW:
20
0
−fCLK/2 ≤ fCARRIER < + fCLK /2
FTW = (fCARRIER/fCLK) × 248
where:
–20
–40
–60
–80
–100
–120
–140
f
CARRIER is the carrier frequency.
FTW is a 48-bit, twos complement number.
When in 2× NRZ mode (FIR85 enabled with Register 0x111,
Bit 0 = 1), the frequency tuning word is calculated as
0 ≤ fCARRIER < fCLK
FTW = (fCARRIER/fCLK) × 248
where FTW is a 48-bit binary number.
NORMALIZED FREQUENCY (Rad/Sample)
Figure 75. FIR85 2× Half-Band 45% Filter Response
This method to calculate the FTW for 2× NRZ mode allows the
tone to correctly move from 0 Hz toward fCLK when the FTW is
incremented. Use this method only for 2× NRZ mode. If the
desired tone is placed between fCLK/2 and fCLK and the FIR85
enable bit is set to 0b0 without adjusting the FTW, the tone may
appear to change location to a new frequency.
DIGITAL MODULATION
The AD9166 has digital modulation features to modulate the
baseband quadrature signal to the desired DAC output
frequency.
The AD9166 is equipped with several NCO modes. The default
NCO is a 48-bit, integer NCO. The A/B ratio of the dual-modulus
NCO allows the output frequency to be synthesized with very
fine precision. NCO mode is selected, as shown in Table 35.
The FTW is set as shown in Table 36.
Table 36. NCO FTW Registers
Address
0x114
0x115
0x116
0x117
0x118
0x119
Value
Description
FTW[7:0]
8 LSBs of FTW
Table 35. Modulation Mode Selection
Modulation Type
FTW[15:8]
FTW[23:16]
FTW[31:24]
FTW[39:32]
FTW[47:40]
Next 8 bits of FTW
Next 8 bits of FTW
Next 8 bits of FTW
Next 8 bits of FTW
8 MSBs of FTW
Register 0x111, Register 0x111,
Modulation Mode
None
48-Bit Integer NCO
48-Bit Dual-Modulus NCO 0b1
32-Bit FFH NCO1
0b1
Bit 6
0b0
0b1
Bit 2
0b0
0b0
0b1
0b1
1 The FFH NCOs are enabled by writing a nonzero word to their frequency
tuning word registers when the main 48-bit NCO is enabled (see the Fast
Frequency Hopping (FFH) section). The modulus can be enabled or disabled.
If the modulus is enabled, the same modulus ratio applies to all the NCOs.
Rev. 0 | Page 52 of 138
Data Sheet
AD9166
Programmable Modulus Example
Unlike other registers, the FTW registers are not updated
immediately upon writing. Instead, the FTW registers update on
the rising edge of FTW_LOAD_REQ (Register 0x113, Bit 0).
After an update request, FTW_LOAD_ACK (Register 0x113,
Bit 1) must be high to acknowledge that the frequency tuning
word has updated.
Consider the case in which fCLK = 2500 MHz and the desired
value of fCARRIER is 250 MHz. This scenario synthesizes an output
frequency that is not a power of two submultiple of the sample
rate, namely fCARRIER = (1/10) fCLK, which is not possible with a
typical accumulator-based DDS.
The SEL_SIDEBAND bit (Register 0x111, Bit 1 = 0b1) is a conven-
ience bit that can be set to use the lower sideband modulation
result, which is equivalent to flipping the sign of the frequency
tuning word.
The frequency ratio, fCARRIER/fCLK, leads directly to M and N,
which are determined by reducing the fraction
(250,000,000/2,500,000,000) to its lowest terms, that is,
M/N = 250,000,000/2,500,000,000 = 1/10
Therefore, M = 1 and N = 10.
INTERPOLATION
I DATA
After calculation, X = 28,147,497,671,065, A = 3, and B = 5.
Programming these values into the registers for X, A, and B
(X is programmed in Register 0x114 to Register 0x119, B is
programmed in Register 0x124 to Register 0x129, and A is
programmed in Register 0x12A to Register 0x12F) causes the
NCO to produce an output frequency of exactly 250 MHz given
a 2500 MHz sampling clock. For more details, refer to the AN-953
Application Note on the Analog Devices, Inc., website.
COS(ωn + θ)
ω
FTW[47:0]
NCO_PHASE_OFFSET
[15:0]
π
NCO
OUT_I
θ
SIN(ωn + θ)
OUT_Q
–
+
–1
SEL_SIDEBAND
1
0
NCO Reset
Resetting the NCO can be useful when determining the start time
and phase of the NCO. The NCO can be reset by several differ-
ent methods, including a SPI write, using the TX_ENABLE pin,
or by the SYSREF signal. Due to internal timing variations
from device to device, these methods achieve an accuracy of
6 fCLK cycles.
INTERPOLATION
Q DATA
Figure 76. NCO Modulator Block Diagram
Modulus NCO Mode (Direct Digital Synthesis (DDS))
The main 48-bit NCO can also be used in a dual-modulus
mode to create fractional frequencies beyond the 48-bit
accuracy. The modulus mode is enabled by programming the
MODULUS_EN bit in the DATAPATH_CFG register to 1
(Register 0x111, Bit 2 = 0b1).
Program Register 0x800, Bits[7:6] to 0b01 to set the NCO in phase
discontinuous switching mode via a write to the SPI port. Then,
any time the frequency tuning word is updated, the NCO phase
accumulator resets and the NCO begins counting at the new
frequency tuning word.
The frequency ratio for the programmable modulus DDS is
very similar to that of the typical accumulator-based DDS. The
only difference is that N is not required to be a power of two for
the programmable modulus, but can be an arbitrary integer. In
practice, hardware constraints place limits on the range of values
for N. As a result, the modulus extends the use of the NCO to
applications that require exact rational frequency synthesis. The
underlying function of the programmable modulus technique is
to alter the accumulator modulus.
Fast Frequency Hopping (FFH)
To support fast frequency hopping, the AD9166 has several
features in the NCO block. There are two implementations of
the NCO function. The main 48-bit NCO is a general-purpose
NCO and supports some of the fast frequency hopping modes,
whereas the fast frequency hopping NCO is specifically designed
to support several different fast frequency hopping modes.
Main NCO Frequency Hopping
Implementation of the programmable modulus function within
the AD9166 is such that the fraction, M/N, is expressible using
the following equation. Note that the form of the equation
implies a compound FTW with X representing the integer part
and A/B representing the fractional part.
In the main 48-bit NCO, the mode of updating the frequency
tuning word can be changed from requiring a write to the
FTW_LOAD_REQ bit (Register 0x113, Bit 0) to an automatic
update mode. In the automatic update mode, the frequency
tuning word is updated as soon as the chosen frequency tuning
word is written.
A
B
X
248
fCARRIER
fDAC
M
N
To set the automatic frequency tuning word update mode,
write the appropriate word to the FTW_REQ_MODE bits
(Register 0x113, Bits[6:4]), choosing the particular frequency
tuning word that causes the automatic update.
where:
X is programmed in Register 0x114 to Register 0x119.
A is programmed in Register 0x12A to Register 0x12F.
B is programmed in Register 0x124 to Register 0x129.
Rev. 0 | Page 53 of 138
AD9166
Data Sheet
For example, if relatively coarse frequency steps are needed, it
may be sufficient to write a single word to the MSB byte of the
frequency tuning word, and therefore the FTW_REQ_MODE
bits can be programmed to 110 (Register 0x113, Bits[6:4] =
0b110). Then, each time the most significant byte, FTW5, is
written, the NCO frequency tuning word is automatically updated.
words, then select the phase coherent switch mode to start them
at the same time.
To conserve power, each of the 31 additional NCOs and phase
accumulators is enabled only when a frequency tuning word is
programmed into its register. To power down a particular NCO
and phase accumulator, program all zeros to the FTW register
for a given NCO. All NCO frequency tuning words have a
default value of 0x0. The main 48-bit NCO, which is FTW0 in
the fast frequency hopping NCO, is enabled by the NCO_EN bit
in the DATAPATH_CFG register (Register 0x111, Bit 6 = 0b1).
The FTW_REQ_MODE bits can be configured to use any
frequency tuning word as the automatic update trigger word. This
configuration provides convenience when choosing the order in
which to program the FTW registers.
The speed of the SPI port write function is a minimum of
100 MHz (see Table 9). Thus, the NCO frequency tuning word
can be updated in as little as 240 ns with a one-register write in
automatic update mode.
To ensure that there is no residual power consumption or
possible residual spurious from one of the 32-bit NCOs after
powering it up and then powering it down, the suggested method
to power down the additional NCO is to first program the
frequency tuning word to 0x0001, and then program it to 0x0000.
This procedure ensures that the phase accumulator is flushed of
residual values prior to receiving the all 0s word, which powers
down the output but not the accumulator. The accumulator is
powered down with the NCO_EN bit in Register 0x111, Bit 6.
FFH NCO
The fast frequency hopping NCO is implemented as the main
48-bit NCO with an additional 31, 32-bit NCOs, with an
associated bank of 31 frequency tuning words. These frequency
tuning words can be preloaded into the hopping frequency
register bank. Any of the 32 frequency tuning words can be
selected by a one-register write to the HOPF_SEL bits in the
HOPF_CTRL register (Register 0x800, Bits[4:0]). The manner in
which the NCO transitions to the new frequency is determined
by the hopping frequency change mode selection.
NCO Only Mode
The AD9166 is capable of operating in a mode with only the
modulus NCO enabled, to function as a DDS, without a
JESD204B link. In this mode, a single-tone sine wave is
generated by the NCO, mixed with dc data samples to set the
tone amplitude, and sent to the DAC output. Thus, NCO only
mode is sometimes referred to as dc test mode. The dc data
samples are generated internally, without the need for a
functional JESD204B link to provide the data samples. All of the
features described in the Digital Modulation section are
available in the NCO only mode. It is not necessary to bring up
the JESD204B link in this mode.
The fast frequency hopping NCO supports several modes for
transitioning the phase of the NCO output as the output is
hopped to a new frequency: phase continuous switching, phase
discontinuous switching, and phase coherent switching. The
NCO frequency change modes are listed in Table 37.
Table 37. NCO Frequency Change Mode
Register 0x800, Bits[7:6] Description
NCO only mode is a useful option, for example, to generate a
sine wave for setting up a transmitter radio signal chain without
the need for a digital data source from a JESD204B transmitter.
The NCO only mode can also be used in applications where a
sine wave is all that is needed, such as in a local oscillator (LO)
application, as an LO to a mixing stage.
0b00
0b01
Phase continuous switch
Phase discontinuous switch (reset
NCO accumulator)
0b10
Phase coherent switch
In phase continuous switching, the frequency tuning word of
the NCO is updated and the phase accumulator continues to
accumulate to the new frequency.
To enable the NCO only mode, program the DC_TEST_EN bit
in Register 0x150, Bit 1 = 0b1. Then, program a dc value into
the twos complement dc test data word in Register 0x14E (MSB)
and Register 0x14F (LSB). The default value is 0x0000 (zero
amplitude), and a typical value to program is 0x7FFF for a full-
scale tone. The final step is to program the interpolation value
to 1× bypass mode by selecting INTERP_MODE = 0b0000 in
Register 0x110, Bits[3:0]. This step is necessary because the dc
test value is only available in the bypass path and is not
accessible in the complex datapath.
In phase discontinuous mode, the frequency tuning word of the
NCO is updated and the phase accumulator is reset, making an
instantaneous jump to the new frequency.
In phase coherent mode, the bank of additional 31 phase
accumulators is enabled, one each to shadow each FTW in the
hopping frequency register bank.
Upon enabling the phase coherent switching mode (Register 0x800,
Bits[7:6] = 0b10), all 32 NCO phase accumulators begin
counting simultaneously, and all continue counting regardless
of which individual NCO output is currently being used in the
digital datapath. In this way, the frequency of an individual
NCO can be chosen and is always phase coherent to Time 0.
Therefore, it is recommended to preload all frequency tuning
When DC_TEST_EN = 1, the data source of the digital datapath is
the dc test data word. This means that the JESD204B link can be
brought up and data can be transferred to the device over the link,
but the data is not presented to the DAC when DC_TEST_EN = 1.
Connection to the SERDES data source is only achieved when
Rev. 0 | Page 54 of 138
Data Sheet
AD9166
DC_TEST_EN = 0. The DC_TEST_EN bit can be toggled to
switch the NCO input between SERDES data and dc data while the
device is playing a signal (hot switching), but because switching to
the SERDES datapath normally requires synchronizing a
JESD204B link and/or setting the interpolation value, hot
switching the DC_TEST_EN bit is not normally practical.
DOWNSTREAM PROTECTION
The AD9166 has several features designed to protect the power
amplifier of the system, as well as other downstream blocks.
These features consist of a control signal from the LMFC sync
logic and a transmit enable function. The protection mechanism
in each case is the blanking of data that is passed to the DAC
decoder. The differences lie in the location in the datapath and
slight variations of functionality.
INVERSE SINC
The AD9166 provides a digital inverse sinc filter to compensate
the DAC roll-off over frequency. The filter is enabled by setting
the INVSINC_EN bit (Register 0x111, Bit 7) and is disabled by
default.
The inverse sinc (sinc−1) filter is a seven-tap FIR filter. Figure 77
shows the frequency response of sin(x)/x roll-off, the inverse
sinc filter, and the composite response. The composite response
has less than 0.05 dB pass-band ripple up to a frequency of
0.4 × fCLK. When 2× NRZ mode is enabled, the inverse sinc filter
The JESD204B serial link has several flags and quality measures
to indicate the serial link is active and running error free. If any
of these measures flags an issue, a signal from the LMFC sync logic
is sent to a mux that stops data from flowing to the DAC
decoder and replaces it with 0s.
There are several transmit enable features, including a TX_
ENABLE register that can be used to squelch data at several
points in the datapath or configure the TX_ENABLE pin to do
likewise.
operates to 0.4 × 2× fCLK
.
To provide the necessary peaking at the upper end of the pass
band without the risk of clipping, the inverse sinc filter has an
intrinsic insertion loss of about 3.8 dB (shown in Figure 77).
Transmit Enable
The transmit enable feature can be configured either as a SPI
controlled function or a pin controlled function, and can be
used for several different purposes. The SPI controlled function
has less accurate timing due to its reliance on a microcontroller
to program it. Therefore, it is typically used as a preventative
measure at power-up or when configuring the device.
1
SIN(x)/x ROLL-OFF
–1
SINC FILTER RESPONSE
COMPOSITE RESPONSE
0
–1
–2
–3
–4
–5
The SPI controlled TX_ENABLE function can be used to zero
the input to the digital datapath or to zero the output from the
digital datapath, as shown in Figure 78. If the input to the
digital datapath is zeroed, any filtering that is selected filters the
0 signal, causing a gradual ramp-down of energy in the digital
datapath. If the digital datapath is bypassed, as in 1× interpolation
mode, the data at the input to the DAC immediately drops to zero.
0
0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
FREQUENCY (× fDAC
)
Figure 77. Responses of Sin(x)/x Roll-Off, the Sinc−1 Filter, and the Composite
of the Two
DATA
TO DAC
0
MAIN
0
DIGITAL
0
PATH
FROM LMFC
SYNC LOGIC
FROM REG
0x03F[7]
FROM REG
0x03F[6]
TX_ENABLE
TX_ENABLE
FROM REG
0x03F[2]
FROM REG
0x03F[1]
Figure 78. Downstream Protection Block Diagram
Rev. 0 | Page 55 of 138
AD9166
Data Sheet
The TX_ENABLE pin can be used for more accurate timing
when enabling or disabling the DAC output. The effect of the
TX_ENABLE pin can be configured by the TX_ENABLE
register (Register 0x03F) as is used for the SPI controlled func-
tions, and the pin can be configured to have the same effects as
the SPI controlled function, namely to zero the input to the
digital datapath or to zero the output from the digital datapath.
In addition, the TX_ENABLE pin can also be configured to
ramp down (or up) the full-scale current of the DAC. The ramp
down reduces the output power of the DAC by about 20 dB
from full scale to the minimum output current.
DATAPATH PRBS
The datapath PRBS can verify the AD9166 datapath receives
and correctly decodes data. The datapath PRBS verifies the
JESD204B parameters of the transmitter and receiver match, the
lanes of the receiver are mapped appropriately, the lanes are
appropriately inverted, and, if necessary, the start-up routine is
correctly implemented.
To run the datapath PRBS test, complete the following steps:
1. Set up the device in the desired operating mode using the
start-up sequence.
2. Send PRBS7 or PRBS15 data.
3. Write Register 0x14B. Bit 2 = 0 for PRBS7 or Bit 2 = 1 for
PRBS15.
The TX_ENABLE pin can also be programmed to reset the
NCO phase accumulator. See Table 38 for a description of the
settings available for the TX_ENABLE register function.
4. Write Register 0x14B, Bits[1:0] = 0b11 to enable and reset
Table 38. TX_ENABLE Settings
the PRBS test.
5. Write Register 0x14B, Bits[1:0] = 0b01 to enable the PRBS
test and release reset.
6. Wait 500 ms.
7. Check the status of the PRBS by checking the IRQ for the I
and Q path PRBS as described in the Datapath PRBS IRQ
section.
8. Read Register 0x14B, Bits[7:6]. Bit 6 is 0 if the I channel
has any errors. Bit 7 is 0 if the Q channel has any errors.
9. Read Register 0x14C to read the error count for the I channel.
10. Read Register 0x14D to read the error count for the Q
channel. The PRBS processes 32 bits at a time, and compares
the 32 new bits to the previous set of 32 bits. The PRBS
detects and reports only one error in every group of 32 bits.
Therefore, the error count partly depends on when the
errors are seen.
Register 0x03F Setting Description
Bit 7
0
1
SPI control: zero data to the DAC
SPI control: allow data to pass to
the DAC
Bit 6
0
1
SPI control: zero data at input to
the datapath
SPI control: allow data to enter the
datapath
Bits[5:4]
Bit 3
N/A1
Reserved
Use SPI writes to reset the NCO2
Use TX_ENABLE to reset the NCO
0
1
0
Bit 2
Bit 1
Bit 0
Use SPI control to zero data to the
DAC
Use TX_ENABLE pin to zero data to
the DAC
1
0
1
0
1
Use SPI control to zero data at the
input to the datapath
Use TX_ENABLE pin to zero data at
input to the datapath
For example, see the following sequences:
Bits: 32 good; 31 good, 1 bad; 32 good (one error)
Bits: 32 good; 22 good, 10 bad; 32 good (one error)
Bits: 32 good; 31 good, 1 bad; 31 good, 1 bad; 32 good (two
errors)
Use SPI registers to control the full-
scale current
Use TX_ENABLE pin to control the
full-scale current
DATAPATH PRBS IRQ
1 N/A means not applicable.
The PRBS fail signals for the I and Q path are available as IRQ
events. Use Register 0x020, Bits[1:0] to enable the fail signals,
and then use Register 0x024, Bits[1:0] to read back the status
and reset the IRQ signals. See the Interrupt Request Operation
section for more information.
2. Use SPI writes to reset the NCO if resetting the NCO is desired. Register
0x800, Bits[7:6] determine whether the NCO is reset. See Table 37 for more
details.
Rev. 0 | Page 56 of 138
Data Sheet
AD9166
INTERRUPT REQUEST OPERATION
IRQ
) on
The AD9166 provides an interrupt request output signal (
Table 39. IRQ Register Block Details
Ball L5 that can be used to notify an external host processor of
significant device events. On assertion of the interrupt, query the
Event
Register Block
Reported
EVENT_STATUS
IRQ
pin high,
device to determine the precise event that occurred. The
0x020, 0x024
Per chip
IRQ
INTERRUPT_SOURCE when
IRQ
is enabled; when
is
is the event signal
IRQ
IRQ
pin is an open-drain, active low output. Pull the
IRQ
IRQ
disabled,
external to the device. The
pin can be tied to the interrupt
pins of other devices with open-drain outputs to wire-OR these
pins together.
0x4B8 to 0x4BB;
0x470 to 0x473
Per link and
lane
INTERRUPT_SOURCE when
IRQ
is enabled; when
disabled, 0
is
Figure 79 shows a simplified block diagram of how the IRQ blocks
work. When the IRQ_EN signal is low, the INTERRUPT_
SOURCE signal is set to 0. When IRQ_EN is high, any rising
edge of the event signal causes the INTERRUPT_SOURCE
signal to be set high. If any INTERRUPT_SOURCE signal is
INTERRUPT SERVICE ROUTINE
Interrupt request management starts by selecting the set of event
flags that require host intervention or monitoring. Enable the
events that require host action so that the host is notified when
such events occur. For events requiring host intervention upon
IRQ
high, the
pin is pulled low. INTERRUPT_SOURCE can be
IRQ
activation, run the following routine to clear an interrupt
reset to 0 by either an IRQ_RESET signal or a DEVICE_RESET
signal.
request:
1. Read the status of the event flag bits that are being monitored.
2. Disable the interrupt by writing 0 to IRQ_EN.
3. Read the event source.
4. Perform any actions that may be required to clear the cause
of the event. In many cases, no specific actions may be
required.
5. Verify that the event source is functioning as expected.
6. Clear the interrupt by writing 1 to the IRQ_RESET signal.
7. Enable the interrupt by writing 1 to the IRQ_EN signal.
Depending on the STATUS_MODE signal, the EVENT_STATUS
bit reads back an event signal or INTERRUPT_SOURCE signal.
The AD9166 has several interrupt register blocks (IRQ) that can
monitor up to 75 events (depending on device configuration).
Certain details vary by IRQ register block as described in Table 39.
Table 40 shows the source registers of the IRQ_EN, IRQ_RESET,
and STATUS_MODE signals in Figure 79, as well as the address
where EVENT_STATUS is read back.
0
EVENT_STATUS
1
STATUS_MODE
IRQ
IRQ_EN
INTERRUPT_SOURCE
0
1
OTHER
INTERRUPT
SOURCES
EVENT
IRQ_EN
IRQ_RESET
DEVICE_RESET
IRQ
Figure 79. Simplified Schematic of
Circuitry
Table 40. IRQ Register Block Address of IRQ Signal Details
Address of IRQ Signals1
STATUS_MODE2
STATUS_MODE = IRQ_EN 0x024; R per chip
Register Block IRQ_EN
IRQ_RESET
0x024; W per chip
0x4BA, 0x4BB; W per error type N/A, STATUS_MODE = 1
EVENT_STATUS
0x020, 0x024
0x020; R/W per chip
0x4B8 to 0x4BB 0x4B8, 0x4B9; W per error type
0x470 to 0x473 0x470 to 0x473; W per error type
0x4BA, 0x4BB; R per chip
0x470 to 0x473; R per link
0x470 to 0x473; W per link
N/A, STATUS_MODE = 1
1 R is read; W is write; and R/W is read/write.
2 N/A means not applicable.
Rev. 0 | Page 57 of 138
AD9166
Data Sheet
APPLICATIONS INFORMATION
HARDWARE CONSIDERATIONS
Power Supply Recommendations
or burst signals such as global system for mobile communications
(GSM), time division multiple access (TDMA), or other signals
that have an on and off time domain response. Therefore, the
power supply must be able to supply current quickly to
accommodate burst signals. Because the amount of current
variation depends on the signals used, it is recommended to
perform lab testing first to establish ranges. A typical variation
can be several hundred milliamperes over a short time period,
requiring more than 220 μF of bulk capacitance with low ESR.
All AD9166 supply domains must remain as noise free as possible
for optimal operation. Power supply noise carries frequency
content that may adversely affect performance, and it may
appear in the output spectrum of the device.
An inductor/capacitor (LC) filter on the output of the power
supply is recommended to attenuate the noise, and must be placed
as close to the AD9166 as possible.
Power Sequencing
The DAC_1P2_CLK supply is the most noise sensitive supply on
the device, where phase noise and other spectral content are
modulated directly onto the output signal. Other analog supplies
are sensitive as well. DAC_2P5_AN and DAC_N1P2_AN are
the DAC output rails and modulate noise onto the DAC output
and into the amplifier input. AMP_5V_IN, AMP_3P3_OUT,
and AMP_N5 are the analog rails for the amplifier input and
output and may similarly modulate noise onto the output.
The AD9166 requires power sequencing to avoid damage to
internal circuitry. A board design with the AD9166 must
include a power sequencer chip, such as the LTC2928, to ensure
that the domains power up in the correct order. To minimize
current transients during startup, separate high power,
noncritical power domains across the sequence.
Perform the power-up sequence in the following order, and
ensure that the supplies in each group power up and settle
together:
It is highly recommended that DAC_1P2_CLK be supplied by
itself with an ultralow noise regulator such as the ADM7154 or
ADP1761, with high power supply rejection ratio (PSRR)
specifications to filter any unwanted switching supply noise,
and achieve optimal phase noise performance.
1. DAC_1P2_DIG, AMP_3P3, AMP_3P3_OUT, VDD_IO,
DAC_3P3_SYNC, AMP_N5, DAC_1P2_AN,
DAC_1P2_CLK
2. DAC_N1P2_AN, DAC_1P2_SER
3. AMP_5V_IN, DAC_2P5_AN (can be derived from
AMP_5V_IN)
Noisier regulators impose phase noise onto the DAC output, the
amplifier input and output, and ultimately onto the output
signal at the RFOUT pin.
Within each group, monitor the supply with the longest settling
time to ensure that all the supplies settle to their target voltage
before sequencing to the next group.
The DAC_1P2_AN supply can be connected to the digital
DAC_1P2_DIG supply with a separate filter network.
Connect DAC_1P2_SER, the 1.2V supply for the JESD204B
circuitry, to a separate regulator.
There are no requirements for a power-down sequence.
Power and Ground Planes
The AMP_3P3 supply can be connected to the AMP_3P3_OUT
RF output supply with a separate filter network.
Solid ground planes are recommended to avoid ground loops
and to provide a solid, uninterrupted ground reference for the
high speed transmission lines that require controlled impedances.
For high frequency filtering, stack power planes between ground
layers. Stacking also adds extra filtering and isolation between
power supply domains (in addition to the decoupling capacitors).
The VDD_IO supply can be connected to either the AMP_3P3
or the AMP_3P3_OUT supply with a separate filter network.
VDD_IO can also be powered separately from a system
controller (for example, a microcontroller), 1.8 V to 3.3 V
supply. The power supply sequencing requirement must be met,
where AMP_3P3 must come up with or before VDD_IO.
Do not use segmented power planes as a reference for controlled
impedances unless the entire length of the controlled impedance
trace traverses across only a single segmented plane. These and
additional guidelines for the topology of high speed transmission
lines are described in the JESD204B Serial Interface Inputs
(SERDIN0 to SERDIN7 ) section.
Take note of the maximum power consumption numbers given
in Table 2 to ensure the power supply design can tolerate tempera-
ture and IC process variation extremes. The amount of current
drawn is dependent on the chosen use cases. Specifications are
provided for several use cases to illustrate examples and
contributions from individual blocks, and to assist in calculating
the maximum required current per supply.
For some applications, where highest performance and higher
output frequencies are required, the choice of PCB materials
significantly impacts results. For example, materials such as
polyimide or materials from the Rogers Corporation can be
used to improve tolerance to high temperatures and improve
performance. Materials from Rogers such as the RO43xx series,
or from Isola, such as Tachyon, are typically used for the top
Another consideration for the power supply design is peak
current handling capability. The AD9166 draws more current in
the main digital supply (DAC_1P2_DIG) when synthesizing a
signal with significant amplitude variations, such as a
modulated signal with high peak to average power ratio (PAPR)
Rev. 0 | Page 58 of 138
Data Sheet
AD9166
LAYER 1
LAYER 2
LAYER 3
LAYER 4
LAYER 5
LAYER 6
LAYER 7
LAYER 8
layer, and possibly for the bottom. Three layers are used in some
board designs to allow the top layer to reference one of two
planes: one for differential traces and one for single-ended
traces. In this method, the trace widths are kept well within the
manufacturing tolerances of PCB vendors. However, this may
not be practical in some designs.
GND
ADD GROUND VIAS
DIFF–
y
y
y
STANDARD VIA
DIFF+
GND
JESD204B Serial Interface Inputs (SERDIN0 to
SERDIN7 )
MINIMIZE STUB EFFECT
Figure 80. Minimizing Stub Effect and Adding Ground Vias for Differential
Stripline Traces
When considering the layout of the JESD204B serial interface
transmission lines, there are many factors to consider to
maintain optimal link performance. Among these factors are
insertion loss, return loss, signal skew, and the topology of the
differential traces.
Return Loss
The JESD204B specification limits the amount of return loss
allowed in a converter device and a logic device, but does not
specify return loss for the channel. However, every effort must
be made to maintain a continuous impedance on the transmis-
sion line between the transmitting logic device and the AD9166.
Minimizing the use of vias or eliminating vias reduces one of
the primary sources for impedance mismatches on a transmission
line (see the Insertion Loss section). Maintain a solid reference
beneath (for microstrip) or above and below (for stripline) the
differential traces to ensure continuity in the impedance of the
transmission line. If the stripline technique is used, follow the
guidelines listed in the Insertion Loss section to minimize
impedance mismatches and stub effects.
Insertion Loss
The JESD204B specification limits the amount of insertion loss
allowed in the transmission channel (see Figure 51). The AD9166
equalization circuitry allows significantly more loss in the channel
than is required by the JESD204B specification. It is still important
that the designer of the PCB minimize the amount of insertion
loss by adhering to the following guidelines:
Keep the differential traces short by placing the AD9166 as
near to the transmitting logic device as possible and routing
the trace as directly as possible between the devices.
Route the differential pairs on a single plane using a solid
ground plane as a reference. To avoid vias being used in the
SERDES lanes, route the SERDES lanes on the same layer as
the AD9166.
Another primary source for impedance mismatch is at either
end of the transmission line, where care must be taken to match
the impedance of the termination to that of the transmission
line. The AD9166 handles this internally with a calibrated
termination scheme for the receiving end of the line. See the
Interface Power-Up and Input Termination section for details on
this circuit and the calibration routine.
Use a PCB material with a low dielectric constant (<4) to
minimize loss, if possible.
When choosing between the stripline and microstrip techniques,
keep in mind the following considerations: stripline has less loss
(see Figure 52 and Figure 53) and emits less EMI, but requires
the use of vias that can add complexity to the task of controlling
the impedance. Microstrip is easier to implement (if the
component placement and density allow routing on the top
layer) and eases the task of controlling the impedance.
Signal Skew
There are many sources for signal skew, but the two sources to
consider when laying out a PCB are interconnect skew within a
single JESD204B link and skew between multiple JESD204B
links. In each case, keeping the channel lengths matched to within
12.5 mm is adequate for operating the JESD204B link at speeds
of up to 12.5 Gbps. This amount of channel length match is equiv-
alent to about 85% UI on the AD9166 evaluation board.
If using the top layer of the PCB is problematic or the advantages
of stripline are desirable, follow these recommendations:
Minimize the number of vias.
If possible, use blind vias to eliminate via stub effects and
use microvias to minimize via inductance.
If using standard vias, use the maximum via length to
minimize the stub size. For example, on an 8-layer board,
use Layer 7 for the stripline pair (see Figure 80).
For each via pair, place a pair of ground vias adjacent to them
to minimize the impedance discontinuity (see Figure 80).
Managing the interconnect skew within a single link is fairly
straightforward. Managing multiple links across multiple devices
is more complex. However, follow the 12.5 mm guideline for
length matching. The AD9166 can handle more skew than the
85% UI due to the six-PCLK cycle buffer in the JESD204B
receiver, but matching the channel lengths as close as possible
is still recommended.
Rev. 0 | Page 59 of 138
AD9166
Data Sheet
Topology
Tx
Tx
Tx
Tx
DIFF A
DIFF B
DIFF A
DIFF B
Structure the differential SERDINx pairs to achieve 50 Ω to
ground for each half of the pair. Stripline vs. microstrip trade-
offs are described in the Insertion Loss section. In either case, it
is important to keep these transmission lines separated from
potential noise sources, such as high speed digital signals and
noisy supplies.
TIGHTLY COUPLED
DIFFERENTIAL Tx LINES
LOOSELY COUPLED
DIFFERENTIAL Tx LINES
Figure 82. Tightly Coupled vs. Loosely Coupled Differential Traces
AC Coupling Capacitors
If using stripline differential traces, route them using a coplanar
method, with both traces on the same layer. Although this
method does not offer more noise immunity than the broadside
routing method (traces routed on adjacent layers), it is easier to
route and manufacture so that the impedance continuity is
maintained. An illustration of broadside vs. coplanar is shown
in Figure 81.
The AD9166 requires that the JESD204B input signals be
ac-coupled to the source. These capacitors must be 100 nF and
placed as close as possible to the transmitting logic device. To
minimize the impedance mismatch at the pads, select the
package size of the capacitor so that the pad size on the PCB
matches the trace width as closely as possible.
SYNCOUT
, SYSREF , and CLK Signals
Tx DIFF A
SYNCOUT
Tx
Tx
Tx
The
and SYSREF signals on the AD9166 are low
DIFF A
DIFF B ACTIVE
Tx DIFF B
Tx ACTIVE
speed LVDS differential signals. Use controlled impedance traces
routed with 100 Ω differential impedance and 50 Ω to ground
when routing these signals. As with the SERDIN0 to SERDIN7
data pairs, it is important to keep these signals separated from
potential noise sources, such as high speed digital signals and
noisy supplies.
BROADSIDE DIFFERENTIAL Tx LINES
COPLANAR DIFFERENTIAL Tx LINES
Figure 81. Broadside vs. Coplanar Differential Stripline Routing Techniques
When considering the trace width vs. copper weight and
thickness, the speed of the interface must be considered. At
multigigabit speeds, the skin effect of the conducting material
confines the current flow to the surface. To reduce losses,
maximize the surface area of the conductor by making the trace
width wider. Additionally, loosely couple differential traces to
accommodate the wider trace widths. This coupling helps
reduce the crosstalk and minimize the impedance mismatch
when the traces must separate to accommodate components,
vias, connectors, or other routing obstacles. Tightly coupled vs.
loosely coupled differential traces are shown in Figure 82.
SYNCOUT
SYNCOUT
Separate the
noise on the
signal from other noisy signals, because
may be interpreted as a request for
/K/ characters.
It is important to keep similar trace lengths for the CLK and
SYSREF signals from the clock source to each of the devices
on either end of the JESD204B links (see Figure 83). If using a
clock chip that can tightly control the phase of CLK and
SYSREF , the trace length matching requirements are greatly
reduced.
LANE 0
LANE 1
Tx
Rx
DEVICE
DEVICE
LANE N – 1
LANE N
SYSREF±
SYSREF±
CLOCK SOURCE
(AD9516-1, ADCLK925)
DEVICE CLOCK
DEVICE CLOCK
SYSREF± TRACE LENGTH
SYSREF± TRACE LENGTH
DEVICE CLOCK TRACE LENGTH
DEVICE CLOCK TRACE LENGTH
Figure 83. SYSREF Signal and Device Clock Trace Length
Rev. 0 | Page 60 of 138
Data Sheet
AD9166
ANALOG INTERFACE CONSIDERATIONS
simply the complement of the rising edge sample value. In 2×
NRZ mode, both the rising edge and falling edge clock new data
samples. See the 2× NRZ Mode section for more details.
ANALOG MODES OF OPERATION
The AD9166 DAC core uses the quad-switch architecture shown in
Figure 84. Only one pair of switches is enabled during a half-clock
cycle, thus requiring each pair to be clocked on alternative clock
edges. A key benefit of the quad-switch architecture is that it masks
the code dependent glitches that occur in the conventional two-
switch DAC architecture.
When mix mode is used, the output is effectively chopped at the
DAC sample rate. This chopping has the effect of reducing the
power of the fundamental signal while increasing the power of
the images centered around the DAC sample rate, thus improving
the dynamic range of these images.
CLK±
IOUTP
IOUTN
CLK
V
V
V
V
1
2
3
4
G
G
G
INPUT
DATA
D
D
D
D
D
D
D
D
D
D
10
1
2
3
4
5
6
7
8
9
DAC CLOCK
LATCHES
V
1
V
2
V
3
V 4
G
G
G
G
D3
–D8
D2
D4
–D7
–D9
DATA INPUT
G
D1
D5
–D6
–D10
FOUR-SWITCH
DAC OUTPUT
t
(
fS MIX MODE)
D6
–D5
D10
–D1
–D2
–D4
D7
D9
V
SSA
–D3
D8
NOTES
IOUTP AND IOUTN ARE THE POSITIVE AND NEGATIVE OUTPUT
Figure 86. Mix Mode Waveform
CURRENTS, V x IS THE GATE VOLTAGE, AND V
IS THE ANALOG
G
SSA
SUPPLY VOLTAGE.
This ability to change modes provides the user the flexibility to
place a carrier anywhere in the first three Nyquist zones, depending
on the operating mode selected. Switching between baseband
and mix mode reshapes the sinc roll-off inherent at the DAC
output. In baseband mode, the sinc null appears at fS because the
same sample latched on the rising clock edge is also latched again
on the falling clock edge, thus resulting in the same ubiquitous sinc
response of a traditional DAC. In mix mode, the complement
sample of the rising edge is latched on the falling edge, therefore
pushing the sinc null to 2 × fS. Figure 87 shows the ideal frequency
response of the three modes with the sinc roll-off included.
Figure 84. Quad-Switch Architecture
In dual-switch architecture, when a switch transition occurs and
Data 1 and Data 2 (D1 and D2 in Figure 85) are in different
states, a glitch occurs. However, if D1 and D2 happen to be at the
same state, the switch transitions and no glitches occur. This
code dependent glitching causes an increased amount of
distortion in the DAC core. In quad-switch architecture (no
matter what the codes are), there are always two switches that
are transitioning at each half-clock cycle, thus eliminating the
code dependent glitches but creating a constant glitch at 2 × fDAC
in the process. For this reason, a significant clock spur at 2 × fDAC
is evident in the output spectrum of the core.
FIRST
NYQUIST ZONE
SECOND
NYQUIST ZONE
THIRD
NYQUIST ZONE
0
–5
MIX MODE
INPUT
DATA
D
D
D
D
D
D
D
D
D
D
1
2
3
4
5
6
7
8
9
10
DAC CLOCK
RZ MODE
–10
–15
–20
–25
–30
–35
D
D
D
3
D
D
t
t
1
2
4
5
TWO-SWITCH
DAC OUTPUT
NORMAL
MODE
D
D
D
D
D
6
7
8
9
10
FOUR-SWITCH
DAC OUTPUT
(NORMAL MODE)
D
D
D
D
D
10
6
7
8
9
D
D
D
D
D
5
1
2
3
4
Figure 85. Two-Switch and Quad-Switch DAC Waveforms
0fS
0.25fS
0.50fS
0.75fS
1.00fS
1.25fS
1.50fS
FREQUENCY (Hz)
As a consequence of the quad-switch architecture enabling updates
on each half-clock cycle, it is possible to operate the DAC core at
2× the expected DAC update rate (fS) if new data samples are
latched into the DAC core on both the rising and falling edge of the
device clock (fCLK). This notion serves as the basis when operating
the AD9166 in mix mode, RZ mode, or 2× NRZ mode. In each
case, the DAC core is presented with new data samples on each
clock edge. In RZ mode, the rising edge clocks data and the
falling edge clocks zero. In mix mode, the falling edge sample is
Figure 87. Sinc Roll-Off for NRZ, RZ, and Mix Mode Operation
The quad-switch can be configured via SPI (Register 0x152,
Bits[1:0]) to operate in either NRZ mode (0b00), RZ mode
(0b10), or mix mode (0b01).
Rev. 0 | Page 61 of 138
AD9166
Data Sheet
Higher clock input level results in improved phase noise (phase
jitter) performance, because the higher swing results in a higher
slew rate (faster rise time).
2× NRZ Mode
The AD9166 has an additional mode that allows doubling the
DAC sample rate. The 2× NRZ mode is implemented using the
FIR85 2× interpolation filter, which provides new samples to
the DAC core on both rising and falling edges of the device
clock. As a result, the analog bandwidth in 2× NRZ mode also
doubles in comparison to NRZ mode, as illustrated in Figure 88.
DUTY CYCLE
RESTORER
TO DAC
AND DLL
CLK+
CLK–
CROSS
CONTROL
Because the DAC sample rate (fDAC) is now twice the device
clock (fCLK), the energy in the image frequency at fCLK – fOUT
appears at 2 × fCLK – fOUT. Assuming the differential device clock
has perfect phase and amplitude balance, the image at fCLK – fOUT
can be eliminated altogether. Phase imbalance can be compensated
in the device clock receiver of the AD9166. See the Clock Input
section for more details.
5kΩ
5kΩ
16µA
40kΩ
1.25V
Figure 89. Clock Input
The quality of the clock source, as well as its interface to the
AD9166 clock input, directly impacts ac performance. Select the
phase noise and spur characteristics of the clock source to meet
the target application requirements. Phase noise and spurs at a
given frequency offset on the clock source are directly translated
to the output signal. Specifically, it can be shown that the phase
noise characteristics of a reconstructed output sine wave are
related to the clock source by 20 × log10 (fOUT/fDAC) when the
internal device clock path contribution is negligible.
Because the FIR85 interpolator is sampling at a higher rate, it
also consumes more power when compared to a similar NRZ
mode, RZ mode, or mix mode.
0
NRZ MODE
–3
–6
2× NRZ MODE
MIX MODE
RZ MODE
–9
–12
–15
–18
–21
–24
–27
–30
–33
–36
Figure 90 shows a clock source based on the ADF4372 low phase
noise/jitter PLL. The ADF4372 can provide output frequencies
from 62.5 MHz up to 16,000 MHz using the RF16x port and up
to 8,000 MHz using the RF8x or RFAUX8x ports.
Adjusting Clock Duty-Cycle and Differential Phase
Imbalance
0
0.4fCLK
0.8fCLK
1.2fCLK
1.6fCLK
2.0fCLK
The clock control registers exist at Register 0x082 through
Register 0x084. CLK_DUTY (Register 0x082) can be used to
enable duty cycle correction (Bit 7), enable duty cycle offset
control (Bit 6), and set the duty cycle offset (Bits[4:0]). The duty
cycle offset word is a signed magnitude word, with Bit 4 as the
sign bit (1 is negative) and Bits[3:0] as the magnitude. The duty
cycle adjusts across a range of approximately 3%. Recommended
settings for this register are listed in the Start-Up Sequence section.
FREQUENCY (Hz)
Figure 88. Sinc Roll-Off for 2× NRZ, NRZ, RZ, and Mix Mode Operation
CLOCK INPUT
The AD9166 contains a low jitter, differential clock receiver that
is capable of interfacing directly to a differential or single-ended
clock source. Because the input is self biased with a nominal
impedance of 90 Ω, it is recommended that the clock source be
ac-coupled to the CLK input pins. The nominal differential input
is 1 V p-p, but the clock receiver can operate with a span that
ranges from 250 mV p-p to 2.0 V p-p.
V
OUT
AD9166
7.4nH
ADF4372
100pF
100pF
CLK+
CLK–
OUTPUT
STAGE
PLL
VCO
fREF
7.4nH
2GHz TO 6GHz
0dBm
V
OUT
Figure 90. Possible Signal Chain for CLK Input
Rev. 0 | Page 62 of 138
Data Sheet
AD9166
The clock receiver can compensate for phase imbalance of the
CLK+ and CLK− inputs (CLK_PHASE_TUNE register at
Address 0x07F). The register value is a signed binary, with the
MSB acting as the sign bit. Each additional step increase adds
20 fF of capacitance to either the CLK+ or the CLK− input. See
Table 41 for more details. Compensating for phase imbalance
improves the image rejection of the DAC.
Delay Locked Loop (DLL)
The CLK input goes to a high frequency DLL to ensure
reliable locking of the internal DAC sample clock to the input
clock. The DLL is configured and enabled as part of the
recommended start-up sequence. The DLL control registers are
located at Register 0x090 through Register 0x09B. The DLL
settings are determined during product characterization and are
given in the recommended start-up sequence (see the Start-Up
Sequence section). It is not normally necessary to change these
values, and the product characterization data is valid only with the
recommended settings.
Table 41. CLK Phase Adjust Values
Register 0x07F,
Bits[5:0]
Capacitance at
CLK+ (fF)
Capacitance at
CLK− (fF
000000
000001
000010
…
0
0
0
0
…
1 × 20
2 × 20
…
SHUFFLE MODE
The spurious performance of the AD9166 can be improved with
a feature called shuffle mode. Shuffle mode uses proprietary
technology to spread the energy of spurious signals across the
DAC output as random noise. Shuffle mode is enabled by
programming Register 0x151, Bit 2 = 0b1. Because shuffle mode
is implemented using the MSB current sources of the DAC, it is
most effective when the DAC is operated with a small amount
of digital backoff relative to 0 dBFS.
011111
100000
100001
100010
…
31 × 20
0
0
0
…
0
0
0
1 × 20
2 × 20
…
111111
31 × 20
The improvement in performance depends on the phase
balance of the external components as well as on the internal
clock path. Process variations may result in varying phase
balance across devices of the same population. Thus, if higher
levels of image rejection are desired, it may be beneficial to
calibrate each device independently, installed in the target
system. Performing this calibration has shown significant
improvements in some cases, in particular when less expensive
baluns are used.
The amount of noise rise at the output of the DAC core caused
by shuffle mode is directly related to the power in the affected
spurious signals. Because the AD9166 has a wideband buffer
amplifier at the output of the DAC core, the increase in noise
spectral density is not apparent with or without shuffle.
Shuffle mode improves the spurious performance related to
clock and foldback spurs, but does not affect real harmonics
generated at the DAC output.
VOLTAGE REFERENCE AND FULL-SCALE CURRENT
(FSC)
Figure 91 shows how adjusting the clock phase, duty cycle, and
cross control can help compensate for phase and amplitude
imbalance at the CLK pins.
The full-scale current at the DAC output, IOUTFS, controls the
maximum output current swing out of the DAC and into the
input of the buffer amplifier. When adjusting IOUTFS, the input
common-mode current of the buffer amplifier, ICM, must be
adjusted to match the IOUTFS value of the DAC.
–20
–30
–40
–50
–60
–70
–80
–90
PHASE 0,
CROSS 6
I
CM is set using the digital control bits (AMP_ICM, Amplifier
Register 0x18), and it must be set to the nearest corresponding
value that matches IOUTFS. For more details, see the Adjusting ICM
to Match IOUTFS section.
I
OUTFS is set through a combination of digital control bits and the
PHASE 28,
CROSS 10
reference current, ISET, as shown in Figure 92.
AD9166
ANA_FULL_SCALE_CURRENT [9:0]
0
1000
2000
3000
4000
5000
6000
V
BG
1.2V
fOUT (MHz)
Figure 91. Performance Improvement from Tuning the Clock Input
VREF
ISET
DAC
–
+
CURRENT
SCALING
1µF
I
OUTFS
ISET
9.6kΩ
DAC_N1P2_AN
VSS
Figure 92. Voltage Reference Circuit
Rev. 0 | Page 63 of 138
AD9166
Data Sheet
ISET is obtained by forcing the band gap voltage of the DAC
across an external 9.76 kꢀ RSET resistor from the ISET pin to
DAC_N1P2_AN. The 1.2 V nominal band gap voltage (VBG at
VREF) results in a 125 ꢁA reference current, ISET, through the
9.76 kꢀ resistor. ISET is then internally amplified to set the
maximum IOUTFS value, IOUTFS_MAX, which can be controlled
digitally. Although RSET can be adjusted to a higher value to
limit IOUTFS_MAX, the suggested value is 9.76 kꢀ. Further
adjustments to IOUTFS must be performed digitally.
The ICM value can be adjusted digitally over a 6.4 mA to 30.4 mA
range by the AMP_ICM bits in amplifier Register 0x18. The
following equation relates ICM to the AMP_ICM bits:
I
CM = 24 mA × (AMP_ICM/15) + 6.4 mA
To minimize potential stress on the DAC output stage, adjust
IOUTFS and ICM sequentially, as part of the same SPI write
sequence.
ANALOG OUTPUT
The IOUTFS_MAX setting is related to the external resistor by the
following equation:
The AD9166 output is a single-ended, 50 Ω, internally terminated
output with a bipolar output stage for ease of interface with
broadband 50 Ω environments. The equivalent output circuit is
shown in Figure 93, and the equivalent lumped element model
is shown in Figure 94. The output stage is internally biased and
terminated, with no external bias or termination components
needed, and can be connected directly to downstream devices
that present a 50 Ω ground referenced load. The maximum
output voltage swing corresponds to +IOUTFS to −IOUTFS, and can
be adjusted by modifying the IOUTFS of the DAC, which results in
a maximum output power into a 50 Ω load near 4 dBm.
AMP_3P3_OUT
I
OUTFS_MAX = 1.2 V/RSET × 320
where:
1.2 V is the nominal band gap voltage.
R
SET is the external resistor value in kΩ.
320 is a gain constant.
Note that the following constraints apply when configuring the
voltage reference circuit:
Both the 9.76 kΩ resistor and 1 μF bypass capacitor are
required for proper operation.
Adjusting the DAC output full-scale current, IOUTFS, from
its default setting of 40 mA must be performed digitally.
The AD9166 is not a multiplying DAC. Modulation of the
reference current, ISET, with an ac signal is not supported.
The band gap voltage appearing at the VREF pin must be
buffered for use with external circuitry because it has a high
output impedance.
5OΩ
RFOUT
BIAS
OUTPUT FROM
PREVIOUS STAGE
TRIM CURRENT
AMP_N5
An external reference can be used to overdrive the internal
reference by connecting it to the VREF pin.
Figure 93. Equivalent Output Circuit to Pin RFOUT
310pH
1.2pF
4.2Ω
The IOUTFS value can be adjusted digitally over an 8 mA to
40 mA range by ANA_FULL_SCALE_CURRENT, Bits[9:0]
(Register 0x042, Bits[7:0] and Register 0x041, Bits[1:0]). The
following equation relates IOUTFS to the ANA_FULL_
tD = 15.3ps
50Ω
RFOUT
49.6Ω
800pH
300fF
SCALE_CURRENT bits, which can be set from 0 to 1023:
NOTES
tD IS THE DELAY TIME.
IOUTFS = 32 mA × (ANA_FULL_SCALE_CURRENT/1023) + 8 mA
Figure 94. Equivalent Lumped Element Model of the AD9166 Output at
Pin RFOUT
Note that the default value of 0x3FF generates 40 mA full scale,
and this value is used for most of the characterization presented
in this data sheet, unless noted otherwise.
The output dc offset voltage (VOS) at the RFOUT pin can be
adjusted from its nominal value by writing to the VOUT_TRIM
bits in Amplifier Register 0x19. The full adjustment range is
between 350 mV and −250 mV relative to the nominal VOS when
VOUT_TRIM = 0x6A and the offset voltage adjustment
(VOS_ADJ) = 0.0 V.
Adjusting ICM to Match IOUTFS
When adjusting IOUTFS, the buffer amplifier ICM value must be
adjusted to correspond to IOUTFS, which helps minimize the
output common-mode voltage offset of the DAC to within
acceptable levels to maintain performance.
V
OS_ADJ is defined by the following equation:
The relationship between IOUTFS and the ideal ICM value, ICM_IDEAL
,
VOS_ADJ = 0.6 V × VOUT_TRIM/255 − 0.25
can be described by the following equation:
where:
I
CM_IDEAL = IOUTFS/2 + 3.8 mA
V
OS_ADJ is the voltage adjustment from the nominal value.
With ICM_IDEAL known, ICM at the amplifier input stage must be
set to the nearest value, such that the error between ICM and
ICM_IDEAL is minimized.
VOUT_TRIM is the bit value set in Amplifier Register 0x19.
To assist in high frequency PCB design and impedance matching
the AD9166 output, a 50 Ω normalized Smith chart is provided,
Rev. 0 | Page 64 of 138
Data Sheet
AD9166
showing the simulated output return loss (S22) of the amplifier
output (see Figure 95), along with the equivalent lump element
model shown in Figure 94.
To minimize the error due to self heating during calibration, the
AD9166 must be in a low power state with only the temperature
measurement circuitry powered on.
8.91GHz
9.44 +J62.3
Amplifier Junction Temperature Sensor
7.5GHz
6.81 +J39.39
J1
The junction temperature sensor reading of the amplifier is
derived from the following transfer functions:
10.59GHz
20.94 +J107.23
J2
J0.5
5.62GHz
V
ADC = VBGA × ((CODE_x + 4)/255)
MEAS = 318.75 × VADC + TOS
where:
(2)
(3)
7.77 +J17.45
T
J0.2
J5
V
ADC is the ADC input voltage reading sampled from the
0.01GHz
49.6 –J0.18
temperature sensor.
V
0
∞
BGA is the band gap voltage of the amplifier, with a nominal
0.2
0.5
1
2
5
value of 1.09 V.
CODE_x is the ADC readback code at the unknown
4.22GHz
11.2 +J3.91
–J0.2
–J5
temperature, TMEAS
OS is the offset of the transfer function to be determined by
calibration
MEAS is a temperature measurement directly proportional to the
junction temperature of the amplifier (TJ_AMP).
BGA can be measured at the device pin, and is nominally 1.09 V
.
2.24GHz
23.91 –J12.44
T
–J0.5
–J2
T
–J1
Figure 95. S22 vs. Frequency, Showing Output Impedance Relative to ZO = 50 Ω
V
with an uncertainty of 30 mV over process variations in
production, supply voltage variation across the specified range,
and operating temperature range. The error due to VBGA
uncertainty is 2.75% over process, voltage, and temperature
(PVT) and the full range of the temperature sensor. Most of the
error is due to shifts in the manufacturing process, where the
remaining variations over supply voltage and operating
temperature account for only ~0.3% VBGA uncertainty.
TEMPERATURE SENSORS
The AD9166 has two junction temperature sensors, a DAC
sensor and an amplifier sensor. The amplifier sensor monitors
the temperature changes of the amplifier, and is located near the
junction of the buffer amplifier. This sensor represents the
junction temperature of the amplifier (TJ_AMP). The DAC sensor
can help monitor the temperature changes of the DAC core,
together with the digital interface inside the AD9166, and is
located near the DAC core. This sensor represents the junction
temperature of the DAC (TJ_DAC).
TOS is derived by applying a known temperature, TREF, to the
device case and recording the code, CODE_REF, after the
reading has settled to a constant value within the measurement
uncertainty. During calibration, the AD9166 must be retained
in a low power state so that the error due to internal power
dissipation and self heating is minimized.
As the ambient temperature rises, the amplifier typically reaches
its thermal limit before the DAC core. Therefore, TJ,AMP
determines the actual maximum safe operating temperature for
both the amplifier and the DAC core inside the AD9166.
Equation 2 and Equation 3 can be combined to solve for TOS
after deriving TREF and CODE_REF during calibration.
To avoid damage to the amplifier, TJ,AMP must never exceed the
limit defined in the Absolute Maximum Ratings section.
T
REF = 318.75 × VBGA × ((CODE_REF + 4)/255) + TOS
OS = TREF − (318.75 × VBGA × (CODE_REF + 4)/255)
(4)
(5)
Before use, the sensors must be calibrated to remove the device
to device variations on the band gap and the circuitry that
senses the temperature. The temperature must be calibrated
against a known temperature reference to determine either the
slope or the intercept. This type of calibration is typically
referred to as one-point calibration.
T
where:
REF is the calibrated temperature at which the temperature sensor
T
is read.
CODE_REF is the readback code at the measured temperature,
TREF
.
The measured temperature of either sensor, TMEAS, is generally
related to the temperature code by the following formula:
T
MEAS = M × CODE_x + TOFFSET
where CODE_x is the readback code at the unknown
temperature, TMEAS
(1)
.
Rev. 0 | Page 65 of 138
AD9166
Data Sheet
With TOS known, TJ_AMP can be calculated from subsequent
readback code, CODE_x:
has settled to a constant value within the measurement
uncertainty. Use the following equation to calculate M:
TREF = M × (CODE_REF/1000) – 190
M = (TREF + 190)/(CODE_REF/1000)
where:
TREF is the calibrated temperature at which the temperature sensor
is read.
T
MEAS = 318.75 × VBGA × ((CODE_x + 4)/255) + TOS
J,AMP = TMEAS + 5.5
(6)
(7)
(9)
T
where:
J_AMP is the junction temperature of the amplifier.
CODE_x is the ADC readback code at the unknown
temperature, TJ_AMP
T
.
CODE_REF is the readback code at the measured temperature,
TREF
.
Equation 7 is only valid if the temperature sensor is properly
calibrated. To calibrate the sensor, reset the AD9166 before
Rearranging Equation 8 and plugging it into Equation 1 yields a
transfer function that directly relates CODE_REF and TREF to
the sensor reading, as follows:
enabling the sensor so that minimal amount of power is drawn
by the AD9166 and self heating due to internal power dissipation
is minimized.
190 = M × (CODE_REF/1000) – TREF
To enable the temperature sensor, make sure its sampling ADC
is powered up (Amplifier Register 0x10, Bit 1 = 0b0). Set
ST_ADC_CLKF_0 = 0b1 (Amplifier Register 0x1B, Bit 0). Wait
at least 17 ADC clock cycles. ADC sampling clock rate can be
adjusted using ST_ADC_CLKF_1 (Amplifier Register 0x1B,
Bit 1), although the default 2 MHz clock setting is recommended.
While the ADC is sampling the measurement, the ADC_EOC
bit stays low until the conversion is complete, at which point
ADC_EOC becomes high (ADC_EOC = 0b1). The ADC code
can read from the ADC_CODE bits (Amplifier Register 0x1D),
where the ADC_CODE represents CODE_x in Equation 6.
T
MEAS = M × (CODE_x − CODE_REF)/1000 + TREF
where:
CODE_x is the readback code at the unknown temperature,
TMEAS
CODE_REF is the readback code at the calibrated temperature,
TREF
.
.
Equivalently, Equation 2 can be used directly, after M is derived
from the one-point calibration procedure of Equation 1,
described in the Temperature Sensors section, as follows:
TMEAS = M × (CODE_x/1000) − 190
DAC Junction Temperature Sensor
TJ_DAC = TMEAS
The DAC temperature sensor reading is derived from the
following transfer function:
To calibrate the sensor, reset the AD9166 before enabling the
sensor so that minimal amount of power is drawn by the
AD9166 and self heating is minimized.
TMEAS = M × (CODE_x/1000) − 190
(8)
where::
To enable the sensor, set Register 0x135 to 0xA1. The user must
write 0b1 to Register 0x134, Bit 0, before reading back the die
temperature from Register 0x132 (LSB) and Register 0x133 (MSB).
CODE_x is the readback code at the unknown temperature,
TMEAS
.
M is the slope of the transfer function to be determined by
calibration.
M is derived by applying a known temperature, TREF, to the
sensor and recording the code, CODE_REF, after the reading
Rev. 0 | Page 66 of 138
Data Sheet
AD9166
START-UP SEQUENCE
Several steps are required to program the AD9166 to the proper
operating state after the device is powered up according to the
recommended power-up sequence (see the Power Sequencing
section).
that optimize the performance of the DAC and the device clock
DLL (see Table 42). Run this sequence whenever the DAC is
powered down or reset.
The configure JESD204B sequence configures the SERDES
block and then brings up the links (see Table 43). First, run the
configure DAC start-up sequence, then run the configure
JESD204B sequence.
The start-up sequence is divided into several steps, and is listed
in Table 42, Table 43, and Table 44, along with an explanation of
the purpose of each step. Private registers are reserved but
must be written for proper operation. Blank cells in Table 42 to
Table 44 mean that the value depends on the result as described
in the Description column.
Follow the configure NCO sequence if using the NCO (see
Table 44). Note that the NCO can be used in NCO only mode
or in conjunction with synthesized data from the SERDES data
interface. Only one mode can be used at a time and this mode is
selected in the second step in Table 44. The configure DAC
start-up sequence is run first, then the configure NCO sequence.
The AD9166 is calibrated at the factory as part of the automatic
test program. The configure DAC start-up sequence loads the
factory calibration coefficients and configures some parameters
Table 42. Configure DAC Start-Up Sequence After Power-Up
Register
0x000
0x0D2
0x0D2
0x606
0x607
0x604
Value Description
R/W
0x18
0x52
Configure the device for 4-wire serial port operation (optional: leave at the default of 3-wire SPI).
Reset internal calibration registers (private).
W
W
W
W
W
W
R
0xD2 Clear the reset bit for the internal calibration registers (private).
0x02
0x00
0x01
N/A1
Configure the nonvolatile random access memory (NVRAM) (private).
Configure the NVRAM (private).
Load the NVRAM. Loads factory calibration factors from the NVRAM (private).
Optional. Read CHIP_TYPE, PROD_ID, Bits[15:0], PROD_GRADE, and DEV_REVISION from
Register 0x003, Register 0x004, Register 0x005, and Register 0x006.
0x003, 0x004,
0x005, 0x006
0x604, Bit 1
0b1
Optional. Read the boot loader pass bit in Register 0x604, Bit 1 = 0b1 to indicate a completed boot
load (private)
R
0x058
0x090
0x080
0x040
0x09E
0x091
0x092, Bit 0
0x03
0x1E
0x00
0x00
0x85
0xE9
0b1
Enable the band gap reference (private).
Power up the device clock DLL.
Enable the clock receiver.
Enable the DAC bias circuits.
Configure DAC analog parameters (private).
Enable the device clock DLL.
W
W
W
W
W
W
R
Check DLL_STATUS; set Register 0x092, Bit 0 = 1 to indicate the DLL is locked to the device clock
input.
0x0E8
0x152, Bits[1:0]
0x20
Enable calibration factors (private).
Configure the DAC decode mode (0b00 = NRZ, 0b01 = mix mode, or 0b10 = RZ).
W
W
1 N/A means not applicable.
Rev. 0 | Page 67 of 138
AD9166
Data Sheet
Table 43. Configure JESD204B Start-Up Sequence
Register
0x300
0x480
0x481
0x482
0x483
0x484
0x485
0x486
0x487
0x110
0x111
Value Description
R/W
W
W
W
W
W
W
W
W
W
W
W
0x00
0x38
0x38
0x38
0x38
0x38
0x38
0x38
0x38
Ensure the SERDES links are disabled before configuring them.
Enable SERDES error counters.
Enable SERDES error counters.
Enable SERDES error counters.
Enable SERDES error counters.
Enable SERDES error counters.
Enable SERDES error counters.
Enable SERDES error counters.
Enable SERDES error counters.
Configure number of lanes (Bits[7:4]) and interpolation rate (Bits[3:0]).
Configure the datapath options for Bit 7 (INVSINC_EN), Bit 6 (NCO_EN), Bit 4 (FILT_BW), Bit 2
(MODULUS_EN), Bit 1 (SEL_SIDEBAND), and Bit 0 (FIR85_FILT_EN). See the Register Summary section for
details on the options. Set the reserved bits (Bit 5 and Bit 3) to 0b0.
0x230
Configure the CDR block according to Table 19 for both half rate enable and the divider.
Set up the SERDES PLL divider based on the conditions shown in Table 18.
Set up the PLL reference clock rate based on the conditions shown in Table 18.
Enable the JESD204B block (disable master SERDES power-down).
Soft reset the JESD204B quad-byte deframer.
Optional. Enable scrambling on SERDES lanes.
Set the subclass type: 0b000 = Subclass 0, 0b001 = Subclass 1.
Set the JESD204x version to JESD204B.
Program the calculated checksum value for Lane 0 from values in Register 0x450 to Register 0x45C.
Bring the JESD204B quad-byte deframer out of reset.
Set any bits to 1 to power down the appropriate physical lane.
Optional. Calibrate SERDES PHY Termination Block 1 (PHY 0, PHY 1, PHY 6, PHY 7).
Optional. Calibrate SERDES PHY Termination Block 2 (PHY 2, PHY 3, PHY 4, PHY 5).
Override defaults in the SERDES PLL settings (private).
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
R
0x289, Bits[1:0]
0x084, Bits[5:4]
0x200
0x00
0x09
0b1
0x475
0x453, Bit 7
0x458, Bits[7:5]
0x459, Bits[7:5]
0x45D
0x475
0x201, Bits[7:0]
0x2A7
0x2AE
0x29E
0x280
0x281, Bit 0
0b1
0x01
0x01
0x01
0x1F
0x03
0b1
Enable the SERDES PLL.
Read back Register 0x281 until Bit 0 = 1 to indicate the SERDES PLL is locked.
Prior to enabling the link, be sure that the JESD204B transmitter is enabled and ready to begin link
synchronization.
0x206
0x206
0x300
0x00
0x01
0x01
Reset the CDR to realign the sampling clock with the incoming data.
Bring the CDR out of reset.
Enable the JESD204B receiver to begin link synchronization. When SYNCOUT± is asserted, the JESD204B
transmitter begins CGS by sending /K/ characters.
Read the CGS status for all lanes.
Read the frame sync status for all lanes.
Read the good checksum status for all lanes.
Read the initial lane sync status for all lanes.
Clear the datapath interrupts.
Clear the SERDES interrupts.
Clear the SERDES interrupt.
Optional. Enable the interrupts.
Optional. Enable JESD204B interrupts.
W
W
W
0x470
0x471
0x472
0x473
0x024
0x4BA
0x4BB
0x020
0x4B8
0x4B9
0xFF
0xFF
0xFF
0xFF
0x1F
0xFF
0x01
0x0F
0xFF
0x01
R
R
R
R
W
W
W
W
W
W
Optional. Enable JESD204B interrupts.
Rev. 0 | Page 68 of 138
Data Sheet
AD9166
Table 44. Configure NCO Sequence
Register
Value Description
0x80 (Optional). Perform this write if NCO only mode is desired.
Configure NCO_EN (Bit 6) = 0b1. Configure other datapath options for Bit 7 (INVSINC_EN), Bit 4 (FILT_BW),
R/W
0x110
0x111, Bit 6 0b1
W
W
Bit 2 (MODULUS_EN), Bit 1 (SEL_SIDEBAND), and Bit 0 (FIR85_FILT_EN). See the Register Summary section for
details on the options. Set the reserved bits (Bit 5 and Bit 3) to 0b0.
0x150, Bit 1
0x14E
0x14F
Configure the DC_TEST_EN bit: 0b0 = NCO operation with data interface; 0b1 = NCO only mode.
Write amplitude value for tone amplitude in NCO only mode (Bits[15:8]).
Write amplitude value for tone amplitude in NCO only mode (Bits[7:0]).
Ensure the frequency tuning word write request is low.
Write FTW, Bits[47:40].
Write FTW, Bits[39:32].
Write FTW, Bits[31:24].
Write FTW, Bits[23:16].
Write FTW, Bits[15:8].
W
W
W
W
W
W
W
W
W
W
W
0x113
0x119
0x118
0x117
0x116
0x115
0x114
0x113
0x00
0x01
Write FTW, Bits[7:0].
Load the FTW to the NCO.
Rev. 0 | Page 69 of 138
AD9166
Data Sheet
REGISTER SUMMARY: DAC
Table 45. DAC Register Summary
Reg.
Name
Bits Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Reset RW
0x000 SPI_INTFCONFA [7:0] SOFTRESET_ LSBFIRST_ ADDRINC_M SDOACTIVE_ SDOACTIVE ADDRINC
LSBFIRST
SOFTRESET 0x00 R/W
M
0x001 SPI_INTFCONFB [7:0] SINGLEINS
0x002 SPI_DEVCONF [7:0]
M
M
CS
Reserved
SOFTRESET1
CUSTOPMODE
SOFTRESET0 Reserved
SYSOPMODE
0x00 R/W
0x00 R/W
STALL
DEVSTATUS
0x003 SPI_CHIPTYPE [7:0]
CHIP_TYPE
0x00
0x00
0x00
0x00
R
R
R
R
0x004 SPI_PRODIDL
[7:0]
PROD_ID[7:0]
PROD_ID[15:8]
0x005 SPI_PRODIDH [7:0]
0x006 SPI_CHIPGRADE [7:0]
PROD_GRADE
Reserved
DEV_REVISION
0x020 IRQ_ENABLE
[7:0]
EN_SYSREF_ EN_DATA_ EN_LANE_ FIFO EN_PRBSQ EN_PRBSI 0x00 R/W
JITTER
READY
0x024 IRQ_STATUS
[7:0]
Reserved
Reserved
IRQ_
SYSREF_
JITTER
IRQ_
DATA_
READY
IRQ_LANE_
FIFO
IRQ_
PRBSQ
IRQ_PRBSI 0x00 R/W
0x031 SYNC_LMFC_
DELAY_FRAME
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
SYNC_LMFC_DELAY_SET_FRM
SYNC_LMFC_DELAY_SET[7:0]
SYNC_LMFC_DELAY_SET[11:8]
SYNC_LMFC_DELAY_STAT[7:0]
SYNC_LMFC_DELAY_STAT[11:8]
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x032 SYNC_LMFC_
DELAY0
0x033 SYNC_LMFC_
DELAY1
Reserved
0x034 SYNC_LMFC_
STAT0
0x035 SYNC_LMFC_
STAT1
Reserved
Reserved
0x036 SYSREF_COUNT [7:0]
0x037 SYSREF_PHASE0 [7:0]
0x038 SYSREF_PHASE1 [7:0]
SYSREF_COUNT
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
SYSREF_PHASE[7:0]
SYSREF_PHASE[11:8]
SYSREF_JITTER_WINDOW
0x039 SYSREF_JITTER_ [7:0]
WINDOW
Reserved
0x03A SYNC_CTRL
0x03F TX_ENABLE
[7:0]
Reserved
SYNC_MODE
TXEN_ TXEN_
DATAPATH_ DAC_FSC
PRE
0x00 R/W
0xC0 R/W
[7:0] SPI_
SPI_
Reserved
TXEN_NCO TXEN_
_RESET
DATAPATH_
POST
DATAPATH_ DATAPATH
POST _PRE
0x040 ANA_DAC_
BIAS_PD
[7:0]
[7:0]
Reserved
Reserved
ANA_DAC_ ANA_DAC_ 0x03 R/W
BIAS_PD1 BIAS_PD0
0x041 ANA_FSC0
ANA_FULL_SCALE_ 0x03 R/W
CURRENT[1:0]
0x042 ANA_FSC1
[7:0]
[7:0]
ANA_FULL_SCALE_CURRENT[9:2]
CLK_PHASE_TUNE
0xFF R/W
0x00 R/W
0x07F CLK_PHASE_
TUNE
Reserved
0x080 CLK_PD
[7:0]
Reserved
DACCLK_PD 0x01 R/W
0x80 R/W
0x082 CLK_DUTY
[7:0] CLK_DUTY_ CLK_
EN
DUTY_
OFFSET_EN
CLK_ DUTY_
BOOST_EN
CLK_DUTY_PRG
0x083 CLK_CRS_CTRL [7:0] CLK_CRS_EN
Reserved
CLK_CRS_ADJ
0x80 R/W
0x084 PLL_REF_CLK_ [7:0]
PD
Reserved
PLL_REF_CLK_RATE
Reserved
Reserved
Reserved
PLL_REF_ 0x00 R/W
CLK_PD
0x088 SYSREF_CTRL0 [7:0]
0x089 SYSREF_CTRL1 [7:0]
HYS_ON
HYS_CNTRL[7:0]
DLL_FINE_ DLL_FINE_ DLL_ COARSE_ DLL_
DC_EN XC_EN DC_EN
SYSREF_RISE
HYS_CNTRL[9:8]
0x00 R/W
0x00 R/W
0x090 DLL_PD
[7:0]
DLL_CLK_ 0x1F R/W
COARSE_ PD
XC_EN
DLL_MODE
0x091 DLL_CTRL
0x092 DLL_STATUS
[7:0] DLL_TRACK_ DLL_
ERR
SEARCH_
ERR
DLL_SLOPE
Reserved
DLL_SEARCH
DLL_
ENABLE
0xF0 R/W
0x00 R/W
[7:0]
DLL_FAIL
DLL_LOST DLL_
LOCKED
0x093 DLL_GB
[7:0]
[7:0]
Reserved
DLL_GUARD
0x00 R/W
0x00 R/W
0x094 DLL_COARSE
Reserved
DLL_COARSE
Rev. 0 | Page 70 of 138
Data Sheet
AD9166
Reg.
Name
Bits Bit 7
[7:0]
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
DLL_PHS
Bit 1
Bit 0
Reset RW
0x80 R/W
0x08 R/W
0x00 R/W
0x095 DLL_FINE
0x096 DLL_PHASE
0x097 DLL_BW
0x098 DLL_READ
DLL_FINE
[7:0]
Reserved
Reserved
[7:0]
DLL_FILT_BW
DLL_COARSE_RB
DLL_FINE_RB
DLL_WEIGHT
[7:0]
Reserved
DLL_READ 0x00 R/W
0x099 DLL_COARSE_ [7:0]
RB
Reserved
0x00
R
0x09A DLL_FINE_RB
[7:0]
0x00
0x00
R
R
0x09B DLL_PHASE_RB [7:0]
Reserved
DLL_PHS_RB
0x09D DIG_CLK_
INVERT
[7:0]
Reserved
INV_DIG_CLK DIG_CLK_ DIG_CLK_ 0x03 R/W
DC_EN XC_EN
0x0A0 DLL_CLK_
DEBUG
[7:0] DLL_TEST_EN
Reserved
DLL_TEST_DIV
0x00 R/W
0x110 INTERP_MODE [7:0]
JESD_LANES
INTERP_MODE
0x81 R/W
0x00 R/W
0x111 DATAPATH_
CFG
[7:0] INVSINC_EN NCO_EN Reserved
FILT_BW
Reserved MODULUS_EN SEL_
FIR85_
SIDEBAND FILT_EN
FTW_ FTW_LOAD 0x00 R/W
LOAD_ACK _REQ
0x113 FTW_UPDATE [7:0] Reserved
FTW_REQ_MODE
Reserved FTW_LOAD_
SYSREF
0x114 FTW0
0x115 FTW1
0x116 FTW2
0x117 FTW3
0x118 FTW4
0x119 FTW5
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
FTW[7:0]
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
FTW[15:8]
FTW[23:16]
FTW[31:24]
FTW[39:32]
FTW[47:40]
0x11C PHASE_
OFFSET0
NCO_PHASE_OFFSET[7:0]
NCO_PHASE_OFFSET[15:8]
ACC_MODULUS[7:0]
0x11D PHASE_
OFFSET1
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x124 ACC_
MODULUS0
0x125 ACC_
MODULUS1
0x126 ACC_
MODULUS2
0x127 ACC_
MODULUS3
0x128 ACC_
MODULUS4
0x129 ACC_
MODULUS5
ACC_MODULUS[15:8]
ACC_MODULUS[23:16]
ACC_MODULUS[31:24]
ACC_MODULUS[39:32]
ACC_MODULUS[47:40]
0x12A ACC_DELTA0
0x12B ACC_DELTA1
0x12C ACC_DELTA2
0x12D ACC_DELTA3
0x12E ACC_DELTA4
0x12F ACC_DELTA5
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
ACC_DELTA[7:0]
ACC_DELTA[15:8]
ACC_DELTA[23:16]
ACC_DELTA[31:24]
ACC_DELTA[39:32]
ACC_DELTA[47:40]
TEMP_SENS_OUT[7:0]
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
R
0x132 TEMP_SENS_
LSB
0x133 TEMP_SENS_
MSB
[7:0]
[7:0]
TEMP_SENS_OUT[15:8]
Reserved
R
0x134 TEMP_SENS_
UPDATE
TEMP_
SENS_
0x00 R/W
UPDATE
0x135 TEMP_SENS_
CTRL
[7:0] TEMP_SENS_
FAST
Reserved
TEMP_
SENS_
R/W
ENABLE
0x14B PRBS
[7:0] PRBS_GOOD_ PRBS_
Reserved
PRBS_INV_Q PRBS_INV_I PRBS_MODE
PRBS_RESET PRBS_EN
0x10 R/W
Q
GOOD_I
0x14C PRBS_ERROR_I [7:0]
0x14D PRBS_ERROR_Q [7:0]
PRBS_COUNT_I
PRBS_COUNT_Q
0x00
0x00
R
R
0x14E TEST_DC_
DATA1
[7:0]
DC_TEST_DATA[15:8]
0x00 R/W
Rev. 0 | Page 71 of 138
AD9166
Data Sheet
Reg.
Name
Bits Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Reset RW
0x14F TEST_DC_
DATA0
[7:0]
DC_TEST_DATA[7:0]
0x00 R/W
0x150 DIG_TEST
[7:0]
Reserved
DC_TEST_ Reserved
EN
0x00 R/W
0x151 DECODE_CTRL [7:0]
0x152 DECODE_MODE [7:0]
0x1DF SPI_STRENGTH [7:0]
Reserved
Reserved
Shuffle
Reserved
0x01 R/W
0x00 R/W
0x0F R/W
0x01 R/W
DECODE_MODE
Reserved
SPIDRV
0x200 MASTER_PD
[7:0]
Reserved
SPI_PD_PHY
SPI_PD_
MASTER
0x201 PHY_PD
0x203 GENERIC_PD
[7:0]
[7:0]
0x00 R/W
0x00 R/W
Reserved
Reserved
Reserved
SPI_
Reserved
SYNC1_PD
0x206 CDR_RESET
[7:0]
[7:0]
SPI_CDR_ 0x01 R/W
RESET
0x230 CDR_
OPERATING_
MODE_REG_0
Reserved
SPI_
SPI_DIVISION_RATE
Reserved
0x28 R/W
ENHALFRATE
0x250 EQ_CONFIG_
PHY_0_1
[7:0]
[7:0]
[7:0]
[7:0]
SPI_EQ_CONFIG1
SPI_EQ_CONFIG3
SPI_EQ_CONFIG5
SPI_EQ_CONFIG7
SPI_EQ_CONFIG0
SPI_EQ_CONFIG2
SPI_EQ_CONFIG4
SPI_EQ_CONFIG6
0x88 R/W
0x88 R/W
0x88 R/W
0x88 R/W
0x251 EQ_CONFIG_
PHY_2_3
0x252 EQ_CONFIG_
PHY_4_5
0x253 EQ_CONFIG_
PHY_6_7
0x268 EQ_BIAS_REG [7:0]
EQ_POWER_MODE
Reserved
SPI_RECAL_
SYNTH
0x62 R/W
0x00 R/W
0x280 SYNTH_
ENABLE_CNTRL
[7:0]
Reserved
Reserved
SPI_
ENABLE_
SYNTH
0x281 PLL_STATUS
[7:0]
Reserved
SPI_CP_
OVER_
SPI_CP_
OVER_
SPI_CP_
CAL_
Reserved
SPI_PLL_
LOCK_RB
0x00
R
RANGE_
HIGH_RB
RANGE_
LOW_RB
VALID_RB
0x289 REF_CLK_
DIVIDER_LDO
[7:0]
[7:0]
Reserved
SERDES_PLL_DIV_
FACTOR
0x04 R/W
0x2A7 TERM_BLK1_
CTRLREG0
Reserved
SPI_I_TUNE 0x00 R/W
_R_CAL_
TERMBLK1
0x2A8 TERM_BLK1_
CTRLREG1
[7:0]
[7:0]
[7:0]
SPI_I_SERIALIZER_RTRIM_TERMBLK1
0x00 R/W
0x2AC TERM_BLK1_
RD_REG0
Reserved
SPI_O_RCAL_CODE_TERMBLK1
0x00
R
0x2AE TERM_BLK2_
CTRLREG0
Reserved
SPI_I_TUNE 0x00 R/W
_R_CAL_
TERMBLK2
0x2AF TERM_BLK2_
CTRLREG1
[7:0]
[7:0]
SPI_I_SERIALIZER_RTRIM_TERMBLK2
0x00 R/W
0x2B3 TERM_BLK2_
RD_REG0
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
SPI_O_RCAL_CODE_TERMBLK2
0x00
R
0x2BB TERM_OFFSET_ [7:0]
0
TERM_OFFSET_0
TERM_OFFSET_1
TERM_OFFSET_2
TERM_OFFSET_3
TERM_OFFSET_4
TERM_OFFSET_5
TERM_OFFSET_6
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x2BC TERM_OFFSET_ [7:0]
1
0x2BD TERM_OFFSET_ [7:0]
2
0x2BE TERM_OFFSET_ [7:0]
3
0x2BF TERM_OFFSET_ [7:0]
4
0x2C0 TERM_OFFSET_ [7:0]
5
0x2C1 TERM_OFFSET_ [7:0]
6
Rev. 0 | Page 72 of 138
Data Sheet
AD9166
Reg.
Name
Bits Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Reset RW
0x2C2 TERM_OFFSET_ [7:0]
7
Reserved
TERM_OFFSET_7
0x00 R/W
0x300 GENERAL_JRX_ [7:0] Reserved
CTRL_0
CHECKSUM
_MODE
Reserved
LINK_EN
0x00 R/W
0x302 DYN_LINK_
LATENCY_0
[7:0]
Reserved
DYN_LINK_LATENCY_0
0x00
R
0x304 LMFC_DELAY_0 [7:0]
0x306 LMFC_VAR_0 [7:0]
Reserved
Reserved
LMFC_DELAY_0
LMFC_VAR_0
0x00 R/W
0x1F R/W
0x08 R/W
0x1A R/W
0x2C R/W
0x3E R/W
0x308 XBAR_LN_0_1 [7:0]
0x309 XBAR_LN_2_3 [7:0]
0x30A XBAR_LN_4_5 [7:0]
0x30B XBAR_LN_6_7 [7:0]
Reserved
SRC_LANE1
SRC_LANE3
SRC_LANE5
SRC_LANE7
SRC_LANE0
Reserved
Reserved
Reserved
SRC_LANE2
SRC_LANE4
SRC_LANE6
0x30C FIFO_STATUS_ [7:0]
REG_0
LANE_FIFO_FULL
0x00
R
0x30D FIFO_STATUS_ [7:0]
REG_1
LANE_FIFO_EMPTY
0x00
R
0x311 SYNC
[7:0]
Reserved
EOMF_MASK_0 Reserved
EOF_
MASK_0
0x00 R/W
0x00 R/W
_GEN_0
0x312 SYNC
0x313 SYNC
[7:0]
[7:0]
[7:0]
SYNC
_ERR_DUR
SYNC
_SYNCREQ_DUR
_GEN_1
LMFC_PERIOD
PHY_TEST_EN
0x00
R
_GEN_3
0x315 PHY_PRBS_
TEST_EN
0x00 R/W
0x316 PHY_PRBS_
TEST_CTRL
[7:0] Reserved
[7:0]
PHY_SRC_ERR_CNT
PHY_PRBS_PAT_SEL
PHY_PRBS_THRESHOLD_LOBITS
PHY_TEST_ PHY_TEST_ 0x00 R/W
START RESET
0x317 PHY_PRBS_
TEST_
0x00 R/W
0x00 R/W
0x00 R/W
THRESHOLD_
LOBITS
0x318 PHY_PRBS_
TEST_
[7:0]
[7:0]
PHY_PRBS_THRESHOLD_MIDBITS
PHY_PRBS_THRESHOLD_HIBITS
THRESHOLD_
MIDBITS
0x319 PHY_PRBS_
TEST_
THRESHOLD_
HIBITS
0x31A PHY_PRBS_
TEST_ERRCNT_
LOBITS
[7:0]
[7:0]
[7:0]
PHY_PRBS_ERR_CNT_LOBITS
PHY_PRBS_ERR_CNT_MIDBITS
PHY_PRBS_ERR_CNT_HIBITS
0x00
0x00
0x00
0xFF
R
R
R
R
0x31B PHY_PRBS_
TEST_ERRCNT_
MIDBITS
0x31C PHY_PRBS_
TEST_ERRCNT_
HIBITS
0x31D PHY_PRBS_
TEST_STATUS
[7:0]
[7:0]
PHY_PRBS_PASS
0x31E PHY_DATA_
SNAPSHOT_
CTRL
Reserved
PHY_GRAB_LANE_SEL
PHY_GRAB PHY_GRAB 0x00 R/W
_MODE
_DATA
0x31F PHY_SNAPSHOT [7:0]
_DATA_BYTE0
PHY_SNAPSHOT_DATA_BYTE0
PHY_SNAPSHOT_DATA_BYTE1
PHY_SNAPSHOT_DATA_BYTE2
PHY_SNAPSHOT_DATA_BYTE3
PHY_SNAPSHOT_DATA_BYTE4
0x00
0x00
0x00
0x00
0x00
R
R
R
R
R
0x320 PHY_SNAPSHOT [7:0]
_DATA_BYTE1
0x321 PHY_SNAPSHOT [7:0]
_DATA_BYTE2
0x322 PHY_SNAPSHOT [7:0]
_DATA_BYTE3
0x323 PHY_SNAPSHOT [7:0]
_DATA_BYTE4
0x32C SHORT_TPL_
TEST_0
[7:0]
SHORT_TPL_SP_SEL
SHORT_TPL_M_SEL
SHORT_
TPL_TEST_ TPL_TEST_
RESET EN
SHORT_
0x00 R/W
Rev. 0 | Page 73 of 138
AD9166
Data Sheet
Reg.
Name
Bits Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Reset RW
0x32D SHORT_TPL_
TEST_1
[7:0]
SHORT_TPL_REF_SP_LSB
SHORT_TPL_REF_SP_MSB
Reserved
0x00 R/W
0x32E SHORT_TPL_
TEST_2
[7:0]
[7:0]
[7:0]
0x00 R/W
0x32F SHORT_TPL_
TEST_3
SHORT_
TPL_FAIL
0x00
R
0x334 JESD_BIT_
INVERSE_CTRL
JESD_BIT_INVERSE
0x00 R/W
0x400 DID_REG
0x401 BID_REG
0x402 LID0_REG
[7:0]
DID_RD
BID_RD
0x00
0x00
0x00
R
R
R
[7:0]
[7:0] Reserved
ADJDIR_
RD
PHADJ_RD
Reserved
LL_LID0
0x403 SCR_L_REG
0x404 F_REG
[7:0] SCR_RD
[7:0]
Reserved
L_RD
K_RD
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
R
R
R
R
R
R
R
R
R
R
R
F_RD
0x405 K_REG
[7:0]
Reserved
CS_RD
0x406 M_REG
[7:0]
M_RD
0x407 CS_N_REG
0x408 NP_REG
0x409 S_REG
[7:0]
N_RD
NP_RD
S_RD
[7:0]
SUBCLASSV_RD
JESDV_RD
[7:0]
0x40A HD_CF_REG
0x40B RES1_REG
0x40C RES2_REG
[7:0] HD_RD
[7:0]
Reserved
CF_RD
RES1_RD
RES2_RD
LL_FCHK0
[7:0]
0x40D CHECKSUM0_ [7:0]
REG
0x40E COMPSUM0_
REG
[7:0]
LL_FCMP0
0x00
R
0x412 LID1_REG
[7:0]
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
LL_LID1
LL_LID2
LL_LID3
LL_LID4
LL_LID5
LL_LID6
LL_LID7
0x00
0x00
R
R
0x415 CHECKSUM1_ [7:0]
REG
LL_FCHK1
LL_FCMP1
0x416 COMPSUM1_
REG
[7:0]
0x00
R
0x41A LID2_REG
[7:0]
0x00
0x00
R
R
0x41D CHECKSUM2_ [7:0]
REG
LL_FCHK2
LL_FCMP2
0x41E COMPSUM2_
REG
[7:0]
0x00
R
0x422 LID3_REG
[7:0]
0x00
0x00
R
R
0x425 CHECKSUM3_ [7:0]
REG
LL_FCHK3
LL_FCMP3
0x426 COMPSUM3_
REG
[7:0]
0x00
R
0x42A LID4_REG
[7:0]
0x00
0x00
R
R
0x42D CHECKSUM4_ [7:0]
REG
LL_FCHK4
LL_FCMP4
0x42E COMPSUM4_
REG
[7:0]
0x00
R
0x432 LID5_REG
[7:0]
0x00
0x00
R
R
0x435 CHECKSUM5_ [7:0]
REG
LL_FCHK5
LL_FCMP5
0x436 COMPSUM5_
REG
[7:0]
0x00
R
0x43A LID6_REG
[7:0]
0x00
0x00
R
R
0x43D CHECKSUM6_ [7:0]
REG
LL_FCHK6
LL_FCMP6
0x43E COMPSUM6_
REG
[7:0]
0x00
R
0x442 LID7_REG
[7:0]
0x00
0x00
R
R
0x445 CHECKSUM7_ [7:0]
REG
LL_FCHK7
LL_FCMP7
0x446 COMPSUM7_
REG
[7:0]
0x00
R
Rev. 0 | Page 74 of 138
Data Sheet
AD9166
Reg.
Name
Bits Bit 7
[7:0]
Bit 6
Bit 5
Bit 4
Bit 3
DID
Bit 2
Bit 1
Bit 0
Reset RW
0x00 R/W
0x00 R/W
0x00 R/W
0x87 R/W
0x450 ILS_DID
0x451 ILS_BID
0x452 ILS_LID0
0x453 ILS_SCR_L
0x454 ILS_F
[7:0]
BID
[7:0] Reserved
[7:0] SCR
[7:0]
ADJDIR
PHADJ
LID0
L
Reserved
F
0x00
R
0x455 ILS_K
[7:0]
Reserved
K
0x1F R/W
0x456 ILS_M
[7:0]
M
0x01
0x0F
R
R
0x457 ILS_CS_N
0x458 ILS_NP
0x459 ILS_S
[7:0]
CS
Reserved
N
NP
S
[7:0]
SUBCLASSV
JESDV
0x0F R/W
0x01 R/W
[7:0]
0x45A ILS_HD_CF
0x45B ILS_RES1
0x45C ILS_RES2
[7:0] HD
[7:0]
Reserved
CF
0x80
R
RES1
RES2
0x00 R/W
0x00 R/W
0x00 R/W
[7:0]
0x45D ILS_CHECKSUM [7:0]
FCHK0
0x46C LANE_DESKEW [7:0] ILD7
0x46D BAD_DISPARITY [7:0] BDE7
0x46E NOT_IN_TABLE [7:0] NIT7
ILS6
ILD5
BDE5
NIT5
UEK5
ILD4
BDE4
NIT4
UEK4
ILD3
ILD2
ILD1
BDE1
NIT1
UEK1
ILD0
BDE0
NIT0
UEK0
0x00
0x00
0x00
0x00
R
R
R
R
BDE6
NIT6
UEK6
BDE3
NIT3
BDE2
NIT2
0x46F UNEXPECTED_ [7:0] UEK7
KCHAR
UEK3
UEK2
0x470 CODE_GRP_
SYNC
[7:0] CGS7
CGS6
CGS5
CGS4
CGS3
CGS2
CGS1
CGS0
0x00
R
0x471 FRAME_SYNC
[7:0] FS7
FS6
FS5
FS4
FS3
FS2
FS1
FS0
0x00
0x00
R
R
0x472 GOOD_
CHECKSUM
[7:0] CKS7
CKS6
CKS5
CKS4
CKS3
CKS2
CKS1
CKS0
0x473 INIT_LANE_
SYNC
[7:0] ILS7
[7:0] RX_DIS
[7:0]
ILS6
ILS5
ILS4
ILS3
ILS2
ILS1
ILS0
0x00
R
0x475 CTRLREG0
CHAR_
REPL_DIS
Reserved
SOFTRST
FORCESYNCREQ Reserved
REPL_FRM_ 0x01 R/W
ENA
0x476 CTRLREG1
Reserved
QUAL_
RDERR
DEL_SCR CGS_SEL
NO_ILAS
Reserved
FCHK_N
0x14 R/W
0x477 CTRLREG2
0x478 KVAL
[7:0] ILS_MODE
Reserved REPDATATEST QUETESTERR AR_ECNTR
0x00 R/W
0x01 R/W
0xFF R/W
0x07 R/W
[7:0]
[7:0]
KSYNC
0x47C ERRORTHRES
ETH
0x47D SYNC_ASSERT_ [7:0]
MASK
Reserved
SYNC_ASSERT_MASK
0x480 ECNT_CTRL0
0x481 ECNT_CTRL1
0x482 ECNT_CTRL2
0x483 ECNT_CTRL3
0x484 ECNT_CTRL4
0x485 ECNT_CTRL5
0x486 ECNT_CTRL6
0x487 ECNT_CTRL7
0x488 ECNT_TCH0
0x489 ECNT_TCH1
0x48A ECNT_TCH2
0x48B ECNT_TCH3
0x48C ECNT_TCH4
0x48D ECNT_TCH5
0x48E ECNT_TCH6
0x48F ECNT_TCH7
0x490 ECNT_STAT0
0x491 ECNT_STAT1
0x492 ECNT_STAT2
0x493 ECNT_STAT3
0x494 ECNT_STAT4
0x495 ECNT_STAT5
0x496 ECNT_STAT6
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
ECNT_ENA0
ECNT_ENA1
ECNT_ENA2
ECNT_ENA3
ECNT_ENA4
ECNT_ENA5
ECNT_ENA6
ECNT_ENA7
ECNT_RST0
ECNT_RST1
ECNT_RST2
ECNT_RST3
ECNT_RST4
ECNT_RST5
ECNT_RST6
ECNT_RST7
ECNT_TCH0
ECNT_TCH1
ECNT_TCH2
ECNT_TCH3
ECNT_TCH4
ECNT_TCH5
ECNT_TCH6
ECNT_TCH7
ECNT_TCR0
ECNT_TCR1
ECNT_TCR2
ECNT_TCR3
ECNT_TCR4
ECNT_TCR5
ECNT_TCR6
0x3F R/W
0x3F R/W
0x3F R/W
0x3F R/W
0x3F R/W
0x3F R/W
0x3F R/W
0x3F R/W
0x07 R/W
0x07 R/W
0x07 R/W
0x07 R/W
0x07 R/W
0x07 R/W
0x07 R/W
0x07 R/W
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
LANE_ENA0
LANE_ENA1
LANE_ENA2
LANE_ENA3
LANE_ENA4
LANE_ENA5
LANE_ENA6
0x00
0x00
0x00
0x00
0x00
0x00
0x00
R
R
R
R
R
R
R
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Rev. 0 | Page 75 of 138
AD9166
Data Sheet
Reg.
Name
Bits Bit 7
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
[7:0]
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Reset RW
0x497 ECNT_STAT7
0x498 BD_CNT0
0x499 BD_CNT1
0x49A BD_CNT2
0x49B BD_CNT3
0x49C BD_CNT4
0x49D BD_CNT5
0x49E BD_CNT6
0x49F BD_CNT7
0x4A0 NIT_CNT0
0x4A1 NIT_CNT1
0x4A2 NIT_CNT2
0x4A3 NIT_CNT3
0x4A4 NIT_CNT4
0x4A5 NIT_CNT5
0x4A6 NIT_CNT6
0x4A7 NIT_CNT7
0x4A8 UEK_CNT0
0x4A9 UEK_CNT1
0x4AA UEK_CNT2
0x4AB UEK_CNT3
0x4AC UEK_CNT4
0x4AD UEK_CNT5
0x4AE UEK_CNT6
0x4AF UEK_CNT7
Reserved
LANE_ENA7
ECNT_TCR7
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
0x00
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
BD_CNT0
BD_CNT1
BD_CNT2
BD_CNT3
BD_CNT4
BD_CNT5
BD_CNT6
BD_CNT7
NIT_CNT0
NIT_CNT1
NIT_CNT2
NIT_CNT3
NIT_CNT4
NIT_CNT5
NIT_CNT6
NIT_CNT7
UEK_CNT0
UEK_CNT1
UEK_CNT2
UEK_CNT3
UEK_CNT4
UEK_CNT5
UEK_CNT6
UEK_CNT7
ILS0
0x4B0 LINK_STATUS0 [7:0] BDE0
0x4B1 LINK_STATUS1 [7:0] BDE1
0x4B2 LINK_STATUS2 [7:0] BDE2
0x4B3 LINK_STATUS3 [7:0] BDE3
0x4B4 LINK_STATUS4 [7:0] BDE4
0x4B5 LINK_STATUS5 [7:0] BDE5
0x4B6 LINK_STATUS6 [7:0] BDE6
0x4B7 LINK_STATUS7 [7:0] BDE7
NIT0
NIT1
NIT2
NIT3
NIT4
NIT5
NIT6
NIT7
EN_NIT
UEK0
UEK1
UEK2
UEK3
UEK4
UEK5
UEK6
UEK7
ILD0
ILD1
ILD2
ILD3
ILD4
ILD5
ILD6
ILD7
EN_ILD
CKS0
CKS1
CKS2
CKS3
CKS4
CKS5
CKS6
CKS7
EN_CKS
FS0
FS1
FS2
FS3
FS4
FS5
FS6
FS7
EN_FS
CGS0
CGS1
CGS2
CGS3
CGS4
CGS5
CGS6
CGS7
EN_CGS
ILS1
ILS2
ILS3
ILS4
ILS5
ILS6
ILS7
0x4B8 JESD_IRQ_
ENABLEA
[7:0] EN_BDE
EN_UEK
EN_ILS
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x4B9 JESD_IRQ_
ENABLEB
[7:0]
Reserved
IRQ_ILD
Reserved
EN_ILAS
IRQ_CGS
IRQ_ILAS
0x4BA JESD_IRQ_
STATUSA
[7:0] IRQ_BDE
[7:0]
IRQ_NIT
IRQ_UEK
IRQ_ILS
IRQ_CKS
IRQ_FS
0x4BB JESD_IRQ_
STATUSB
0x800 HOPF_CTRL
[7:0]
HOPF_MODE
Reserved
HOPF_SEL
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x806 HOPF_FTW1_0 [7:0]
0x807 HOPF_FTW1_1 [7:0]
0x808 HOPF_FTW1_2 [7:0]
0x809 HOPF_FTW1_3 [7:0]
0x80A HOPF_FTW2_0 [7:0]
0x80B HOPF_FTW2_1 [7:0]
0x80C HOPF_FTW2_2 [7:0]
0x80D HOPF_FTW2_3 [7:0]
0x80E HOPF_FTW3_0 [7:0]
0x80F HOPF_FTW3_1 [7:0]
0x810 HOPF_FTW3_2 [7:0]
0x811 HOPF_FTW3_3 [7:0]
0x812 HOPF_FTW4_0 [7:0]
0x813 HOPF_FTW4_1 [7:0]
0x814 HOPF_FTW4_2 [7:0]
HOPF_FTW1[7:0]
HOPF_FTW1[15:8]
HOPF_FTW1[23:16]
HOPF_FTW1[31:24]
HOPF_FTW2[7:0]
HOPF_FTW2[15:8]
HOPF_FTW2[23:16]
HOPF_FTW2[31:24]
HOPF_FTW3[7:0]
HOPF_FTW3[15:8]
HOPF_FTW3[23:16]
HOPF_FTW3[31:24]
HOPF_FTW4[7:0]
HOPF_FTW4[15:8]
HOPF_FTW4[23:16]
Rev. 0 | Page 76 of 138
Data Sheet
AD9166
Reg.
Name
Bits Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Reset RW
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x815 HOPF_FTW4_3 [7:0]
0x816 HOPF_FTW5_0 [7:0]
0x817 HOPF_FTW5_1 [7:0]
0x818 HOPF_FTW5_2 [7:0]
0x819 HOPF_FTW5_3 [7:0]
0x81A HOPF_FTW6_0 [7:0]
0x81B HOPF_FTW6_1 [7:0]
0x81C HOPF_FTW6_2 [7:0]
0x81D HOPF_FTW6_3 [7:0]
0x81E HOPF_FTW7_0 [7:0]
0x81F HOPF_FTW7_1 [7:0]
0x820 HOPF_FTW7_2 [7:0]
0x821 HOPF_FTW7_3 [7:0]
0x822 HOPF_FTW8_0 [7:0]
0x823 HOPF_FTW8_1 [7:0]
0x824 HOPF_FTW8_2 [7:0]
0x825 HOPF_FTW8_3 [7:0]
0x826 HOPF_FTW9_0 [7:0]
0x827 HOPF_FTW9_1 [7:0]
0x828 HOPF_FTW9_2 [7:0]
0x829 HOPF_FTW9_3 [7:0]
0x82A HOPF_FTW10_0 [7:0]
0x82B HOPF_FTW10_1 [7:0]
0x82C HOPF_FTW10_2 [7:0]
0x82D HOPF_FTW10_3 [7:0]
0x82E HOPF_FTW11_0 [7:0]
0x82F HOPF_FTW11_1 [7:0]
0x830 HOPF_FTW11_2 [7:0]
0x831 HOPF_FTW11_3 [7:0]
0x832 HOPF_FTW12_0 [7:0]
0x833 HOPF_FTW12_1 [7:0]
0x834 HOPF_FTW12_2 [7:0]
0x835 HOPF_FTW12_3 [7:0]
0x836 HOPF_FTW13_0 [7:0]
0x837 HOPF_FTW13_1 [7:0]
0x838 HOPF_FTW13_2 [7:0]
0x839 HOPF_FTW13_3 [7:0]
0x83A HOPF_FTW14_0 [7:0]
0x83B HOPF_FTW14_1 [7:0]
0x83C HOPF_FTW14_2 [7:0]
0x83D HOPF_FTW14_3 [7:0]
0x83E HOPF_FTW15_0 [7:0]
0x83F HOPF_FTW15_1 [7:0]
0x840 HOPF_FTW15_2 [7:0]
0x841 HOPF_FTW15_3 [7:0]
0x842 HOPF_FTW16_0 [7:0]
0x843 HOPF_FTW16_1 [7:0]
0x844 HOPF_FTW16_2 [7:0]
0x845 HOPF_FTW16_3 [7:0]
0x846 HOPF_FTW17_0 [7:0]
0x847 HOPF_FTW17_1 [7:0]
0x848 HOPF_FTW17_2 [7:0]
0x849 HOPF_FTW17_3 [7:0]
0x84A HOPF_FTW18_0 [7:0]
0x84B HOPF_FTW18_1 [7:0]
0x84C HOPF_FTW18_2 [7:0]
0x84D HOPF_FTW18_3 [7:0]
HOPF_FTW4[31:24]
HOPF_FTW5[7:0]
HOPF_FTW5[15:8]
HOPF_FTW5[23:16]
HOPF_FTW5[31:24]
HOPF_FTW6[7:0]
HOPF_FTW6[15:8]
HOPF_FTW6[23:16]
HOPF_FTW6[31:24]
HOPF_FTW7[7:0]
HOPF_FTW7[15:8]
HOPF_FTW7[23:16]
HOPF_FTW7[31:24]
HOPF_FTW8[7:0]
HOPF_FTW8[15:8]
HOPF_FTW8[23:16]
HOPF_FTW8[31:24]
HOPF_FTW9[7:0]
HOPF_FTW9[15:8]
HOPF_FTW9[23:16]
HOPF_FTW9[31:24]
HOPF_FTW10[7:0]
HOPF_FTW10[15:8]
HOPF_FTW10[23:16]
HOPF_FTW10[31:24]
HOPF_FTW11[7:0]
HOPF_FTW11[15:8]
HOPF_FTW11[23:16]
HOPF_FTW11[31:24]
HOPF_FTW12[7:0]
HOPF_FTW12[15:8]
HOPF_FTW12[23:16]
HOPF_FTW12[31:24]
HOPF_FTW13[7:0]
HOPF_FTW13[15:8]
HOPF_FTW13[23:16]
HOPF_FTW13[31:24]
HOPF_FTW14[7:0]
HOPF_FTW14[15:8]
HOPF_FTW14[23:16]
HOPF_FTW14[31:24]
HOPF_FTW15[7:0]
HOPF_FTW15[15:8]
HOPF_FTW15[23:16]
HOPF_FTW15[31:24]
HOPF_FTW16[7:0]
HOPF_FTW16[15:8]
HOPF_FTW16[23:16]
HOPF_FTW16[31:24]
HOPF_FTW17[7:0]
HOPF_FTW17[15:8]
HOPF_FTW17[23:16]
HOPF_FTW17[31:24]
HOPF_FTW18[7:0]
HOPF_FTW18[15:8]
HOPF_FTW18[23:16]
HOPF_FTW18[31:24]
Rev. 0 | Page 77 of 138
AD9166
Data Sheet
Reg.
Name
Bits Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Reset RW
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x00 R/W
0x84E HOPF_FTW19_0 [7:0]
0x84F HOPF_FTW19_1 [7:0]
0x850 HOPF_FTW19_2 [7:0]
0x851 HOPF_FTW19_3 [7:0]
0x852 HOPF_FTW20_0 [7:0]
0x853 HOPF_FTW20_1 [7:0]
0x854 HOPF_FTW20_2 [7:0]
0x855 HOPF_FTW20_3 [7:0]
0x856 HOPF_FTW21_0 [7:0]
0x857 HOPF_FTW21_1 [7:0]
0x858 HOPF_FTW21_2 [7:0]
0x859 HOPF_FTW21_3 [7:0]
0x85A HOPF_FTW22_0 [7:0]
0x85B HOPF_FTW22_1 [7:0]
0x85C HOPF_FTW22_2 [7:0]
0x85D HOPF_FTW22_3 [7:0]
0x85E HOPF_FTW23_0 [7:0]
0x85F HOPF_FTW23_1 [7:0]
0x860 HOPF_FTW23_2 [7:0]
0x861 HOPF_FTW23_3 [7:0]
0x862 HOPF_FTW24_0 [7:0]
0x863 HOPF_FTW24_1 [7:0]
0x864 HOPF_FTW24_2 [7:0]
0x865 HOPF_FTW24_3 [7:0]
0x866 HOPF_FTW25_0 [7:0]
0x867 HOPF_FTW25_1 [7:0]
0x868 HOPF_FTW25_2 [7:0]
0x869 HOPF_FTW25_3 [7:0]
0x86A HOPF_FTW26_0 [7:0]
0x86B HOPF_FTW26_1 [7:0]
0x86C HOPF_FTW26_2 [7:0]
0x86D HOPF_FTW26_3 [7:0]
0x86E HOPF_FTW27_0 [7:0]
0x86F HOPF_FTW27_1 [7:0]
0x870 HOPF_FTW27_2 [7:0]
0x871 HOPF_FTW27_3 [7:0]
0x872 HOPF_FTW28_0 [7:0]
0x873 HOPF_FTW28_1 [7:0]
0x874 HOPF_FTW28_2 [7:0]
0x875 HOPF_FTW28_3 [7:0]
0x876 HOPF_FTW29_0 [7:0]
0x877 HOPF_FTW29_1 [7:0]
0x878 HOPF_FTW29_2 [7:0]
0x879 HOPF_FTW29_3 [7:0]
0x87A HOPF_FTW30_0 [7:0]
0x87B HOPF_FTW30_1 [7:0]
0x87C HOPF_FTW30_2 [7:0]
0x87D HOPF_FTW30_3 [7:0]
0x87E HOPF_FTW31_0 [7:0]
0x87F HOPF_FTW31_1 [7:0]
0x880 HOPF_FTW31_2 [7:0]
0x881 HOPF_FTW31_3 [7:0]
HOPF_FTW19[7:0]
HOPF_FTW19[15:8]
HOPF_FTW19[23:16]
HOPF_FTW19[31:24]
HOPF_FTW20[7:0]
HOPF_FTW20[15:8]
HOPF_FTW20[23:16]
HOPF_FTW20[31:24]
HOPF_FTW21[7:0]
HOPF_FTW21[15:8]
HOPF_FTW21[23:16]
HOPF_FTW21[31:24]
HOPF_FTW22[7:0]
HOPF_FTW22[15:8]
HOPF_FTW22[23:16]
HOPF_FTW22[31:24]
HOPF_FTW23[7:0]
HOPF_FTW23[15:8]
HOPF_FTW23[23:16]
HOPF_FTW23[31:24]
HOPF_FTW24[7:0]
HOPF_FTW24[15:8]
HOPF_FTW24[23:16]
HOPF_FTW24[31:24]
HOPF_FTW25[7:0]
HOPF_FTW25[15:8]
HOPF_FTW25[23:16]
HOPF_FTW25[31:24]
HOPF_FTW26[7:0]
HOPF_FTW26[15:8]
HOPF_FTW26[23:16]
HOPF_FTW26[31:24]
HOPF_FTW27[7:0]
HOPF_FTW27[15:8]
HOPF_FTW27[23:16]
HOPF_FTW27[31:24]
HOPF_FTW28[7:0]
HOPF_FTW28[15:8]
HOPF_FTW28[23:16]
HOPF_FTW28[31:24]
HOPF_FTW29[7:0]
HOPF_FTW29[15:8]
HOPF_FTW29[23:16]
HOPF_FTW29[31:24]
HOPF_FTW30[7:0]
HOPF_FTW30[15:8]
HOPF_FTW30[23:16]
HOPF_FTW30[31:24]
HOPF_FTW31[7:0]
HOPF_FTW31[15:8]
HOPF_FTW31[23:16]
HOPF_FTW31[31:24]
Rev. 0 | Page 78 of 138
Data Sheet
AD9166
REGISTER DETAILS: DAC REGISTER MAP
Table 46. Register Details
Hex.
Bit
Addr.
Register Name
No. Bit Name
Settings
Description
Reset Access
0x000
SPI_INTFCONFA
7
6
5
4
3
2
SOFTRESET_M
Soft reset (mirror). Set this to
mirror Bit 0.
0x0
0x0
0x0
R
LSBFIRST_M
ADDRINC_M
SDOACTIVE_M
SDOACTIVE
ADDRINC
LSB first (mirror). Set this to
mirror Bit 1.
R
Address increment (mirror).
Set this to mirror Bit 2.
R
SDO active (mirror). Set this to 0x0
mirror Bit 3.
R
SDO active. Enables 4-wire SPI 0x0
bus mode.
R/W
R/W
Address increment. When set, 0x0
causes incrementing
streaming addresses.
Otherwise, descending
addresses are generated.
1 Streaming addresses are
incremented.
0 Streaming addresses are
decremented.
1
0
LSBFIRST
LSB first. When set, causes
input and output data to be
oriented as LSB first. If this bit
is clear, data is oriented as
MSB first.
0x0
R/W
R/W
1 Shift LSB in first.
0 Shift MSB in first.
SOFTRESET
Soft reset. This bit automatically 0x0
clears to 0 after performing a
reset operation. Setting this
bit initiates a reset. This bit is
autoclearing after the soft
reset is complete.
1 Pulse the soft reset line.
0 Reset the soft reset line.
0x001
SPI_INTFCONFB
7
6
SINGLEINS
Single instruction.
1 Perform single transfers.
0 Perform multiple transfers.
0x0
0x0
R/W
R/W
CS
STALL
CS_x
stalling.
0
1
CS_x
CS_x
Disable
Enable
stalling.
stalling.
[5:3] Reserved
Reserved.
0x0
0x0
R/W
R/W
2
1
0
SOFTRESET1
Soft Reset 1. This bit
automatically clears to 0 after
performing a reset operation.
1 Pulse the Soft Reset 1 line.
0 Pulse the Soft Reset 1 line.
SOFTRESET0
Reserved
Soft Reset 0. This bit
automatically clears to 0 after
performing a reset operation.
0x0
0x0
R/W
1 Pulse the Soft Reset 0 line.
0 Pulse the Soft Reset 0 line.
Reserved.
R
Rev. 0 | Page 79 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x002
SPI_DEVCONF
[7:4] DEVSTATUS
[3:2] CUSTOPMODE
[1:0] SYSOPMODE
Device status.
0x0
0x0
0x0
R/W
R/W
R/W
Customer operating mode.
System operating mode.
0 Normal operation.
1 Low power operation.
2 Medium power standby.
3 Low power sleep.
Chip type.
0x003
0x004
0x005
0x006
SPI_CHIPTYPE
SPI_PRODIDL
SPI_PRODIDH
SPI_CHIPGRADE
[7:0] CHIP_TYPE
0x0
0x0
0x0
0x0
0x0
0x0
R
[7:0] PROD_ID[7:0]
[7:0] PROD_ID[15:8]
[7:4] PROD_GRADE
[3:0] DEV_REVISION
[7:5] Reserved
Product ID.
R
Product ID.
R
Product grade.
R
Device revision.
R
0x020
IRQ_ENABLE
Reserved.
R
4
EN_SYSREF_JITTER
Enable SYSREF± jitter interrupt. 0x0
0 Disable interrupt.
R/W
1 Enable interrupt.
3
EN_DATA_READY
Enable JESD204x receiver
ready (JRX_DATA_READY) low
interrupt.
0x0
R/W
0 Disable interrupt.
1 Enable interrupt.
2
1
0
EN_LANE_FIFO
EN_PRBSQ
Enable lane FIFO overflow/
underflow interrupt.
0 Disable interrupt.
1 Enable interrupt.
0x0
0x0
0x0
R/W
R/W
R/W
Enable PRBS imaginary error
interrupt.
0 Disable interrupt.
1 Enable interrupt.
EN_PRBSI
Enable PRBS real error
interrupt.
0 Disable interrupt.
1 Enable interrupt.
Reserved.
0x024
IRQ_STATUS
[7:5] Reserved
0x0
0x0
R
4
IRQ_SYSREF_JITTER
SYSREF± jitter is too big.
Writing 1 clears the status.
R/W
3
IRQ_DATA_READY
IRQ_LANE_FIFO
JRX_DATA_READY is low.
Writing 1 clears the status.
0 No warning.
0x0
R/W
R/W
1 Warning detected.
2
Lane FIFO overflow/under-
flow. Writing 1 clears the
status.
0x0
0 No warning.
1 Warning detected.
1
0
IRQ_PRBSQ
IRQ_PRBSI
PRBS imaginary error. Writing 0x0
1 clears the status.
0 No warning.
R/W
R/W
1 Warning detected.
PRBS real error. Writing 1
clears the status.
0x0
0 No warning.
1 Warning detected.
Reserved.
0x031
SYNC_LMFC_DELAY_FRAME
[7:5] Reserved
0x0
R
[4:0] SYNC_LMFC_DELAY_SET_FRM
Desired delay from rising edge 0x0
of SYSREF± input to rising
R/W
edge of LMFC in frames.
Rev. 0 | Page 80 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x032
SYNC_LMFC_DELAY0
SYNC_LMFC_DELAY1
[7:0] SYNC_LMFC_DELAY_SET[7:0]
Desired delay from rising edge 0x0
of SYSREF± input to rising
edge of LMFC in clock units.
R/W
0x033
0x034
[7:4] Reserved
Reserved.
0x0
R
[3:0] SYNC_LMFC_DELAY_SET[11:8]
Desired delay from rising edge 0x0
of SYSREF± input to rising
edge of LMFC in clock units.
R/W
SYNC_LMFC_STAT0
SYNC_LMFC_STAT1
[7:0] SYNC_LMFC_DELAY_STAT[7:0]
Measured delay from rising
edge of SYSREF± input to
rising edge of LMFC in device
clock units (2 LSBs are always
zero). A write to SYNC_LMFC_
STATx or SYSREF_PHASEx
saves the data for readback.
0x0
R/W
0x035
0x036
[7:4] Reserved
Reserved.
0x0
0x0
R
[3:0] SYNC_LMFC_DELAY_STAT[11:8]
Measured delay from rising
edge of SYSREF± input to
rising edge of LMFC in device
clock units (2 LSBs are always
zero). A write to SYNC_LMFC_
STATx or SYSREF_PHASEx
saves the data for readback.
R/W
SYSREF_COUNT
[7:0] SYSREF_COUNT
Count of SYSREF± signals
received. A write resets the
count. A write to SYNC_
LMFC_STATx or SYSREF_
PHASEx saves the data for
readback.
0x0
0x0
R/W
R/W
0x037
0x038
SYSREF_PHASE0
SYSREF_PHASE1
[7:0] SYSREF_PHASE[7:0]
Phase of measured SYSREF±
event. Thermometer encoded.
A write to SYNC_LMFC_STATx
or SYSREF_PHASEx saves the
data for readback.
[7:4] Reserved
Reserved.
0x0
0x0
R
[3:0] SYSREF_PHASE[11:8]
Phase of measured SYSREF±
event. Thermometer encoded.
A write to SYNC_LMFC_STATx
or SYSREF_PHASEx saves the
data for readback.
R/W
0x039
0x03A
SYSREF_JITTER_WINDOW
[7:6] Reserved
Reserved.
0x0
0x0
R
[5:0] SYSREF_JITTER_WINDOW
Amount of jitter allowed on
the SYSREF± input. SYSREF±
jitter variations bigger than
this triggers an interrupt. Units
are in device clock cycles. The
bottom two bits are ignored.
R/W
SYNC_CTRL
[7:2] Reserved
Reserved.
0x0
0x0
R
[1:0] SYNC_MODE
Synchronization mode.
R/W
00 Do not perform
synchronization; monitor
SYSREF± to LMFC delay only.
01 Perform continuous
synchronization of LMFC on
every SYSREF±.
10 Perform a single
synchronization on the next
SYSREF±, then switch to
monitor mode.
Rev. 0 | Page 81 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x03F
TX_ENABLE
7
SPI_DATAPATH_POST
SPI control of the data at the
output of the datapath.
0x1
R/W
0 Disable or zero the data from
the datapath into the DAC.
1 Use the data from the
datapath to drive the DAC.
6
SPI_DATAPATH_PRE
SPI control of the data at the
input of the datapath.
0x1
R/W
0 Disable or zero the data
feeding into the datapath.
1 Use the data from the
JESD204B lanes to drive into
the datapath.
[5:4] Reserved
Reserved.
0x0
R
3
TXEN_NCO_RESET
Allows TX_ENABLE to control 0x0
the DDS NCO reset.
R/W
0 Use the SPI (HOPF_MODE bits
to control the DDS NCO reset.
1 Use the TX_ENABLE pin to
control the DDS NCO reset.
2
TXEN_DATAPATH_POST
TXEN_DATAPATH_PRE
TXEN_DAC_FSC
Allows TX_ENABLE to control 0x0
the data at the output of the
datapath.
R/W
R/W
R/W
0 Use the SPI (Bit SPI_DATAPATH_
POST) for control.
1 Use the TX_ENABLE pin for
control.
1
0
Allows TX_ENABLE to control 0x0
the data at the input of the
datapath.
0 Use the SPI (Bit SPI_DATAPATH
_PRE) for control.
1 Use the TX_ENABLE pin for
control.
Allows TX_ENABLE to control 0x0
the DAC full-scale current.
0 Use SPI Register ANA_FSC0
and ANA_FSC1 for control.
1 Use the TX_ENABLE pin for
control.
0x040
0x041
ANA_DAC_BIAS_PD
[7:2] Reserved
Reserved.
0x0
0x1
R
1
ANA_DAC_BIAS_PD1
Powers down the DAC core
bias circuits. A 1 powers down
the DAC core bias circuits.
R/W
0
ANA_DAC_BIAS_PD0
Powers down the DAC core
bias circuits. A 1 powers down
the DAC core bias circuits.
0x1
0x0
R/W
ANA_FSC0
[7:2] Reserved
Reserved.
R
[1:0] ANA_FULL_SCALE_CURRENT
DAC full-scale current. Analog 0x3
full-scale current adjustment
(IOUTFS).
R/W
IOUTFS = 32 mA ×
(ANA_FULL_SCALE_CURRENT/
1023) + 8 mA
0x042
ANA_FSC1
[7:0] ANA_FULL_SCALE_CURRENT[9:2]
DAC full-scale current. Analog 0xFF R/W
full-scale current adjustment
(IOUTFS).
IOUTFS = 32 mA ×
(ANA_FULL_SCALE_CURRENT/
1023) + 8 mA
Rev. 0 | Page 82 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x07F
CLK_PHASE_TUNE
[7:6] Reserved
Reserved.
0x0
R
[5:0] CLK_PHASE_TUNE
Fine tuning of the clock input 0x0
phase balance. Adds
R/W
capacitance to the CLK+ or
the CLK− input to phase shift
the differential input at CLK±.
The register is coded as signed
binary. Added nominal
capacitance =
CLK_PHASE_TUNE × 20 fF
Added Nominal Capacitance
At CLK+
0x0 0
At CLK−
0
0x1 20
0x2 40
… …
0x1F 620
0x20 0
0x21 0
0x22 0
… …
0
0
…
0
0
20
40
…
620
0x3F 0
0x080
0x082
CLK_PD
[7:1] Reserved
Reserved.
0x0
0x1
R
0
DACCLK_PD
Device clock power-down.
Powers down the device input
clock circuitry.
R/W
0 Power up.
1 Power down.
CLK_DUTY
7
6
5
CLK_DUTY_EN
Enable duty cycle control.
Enable duty cycle offset.
0x1
0x0
R/W
R/W
R/W
CLK_DUTY_OFFSET_EN
CLK_DUTY_BOOST_EN
Enable duty cycle range boost. 0x0
Extends range to ±5% at cost
of 1 dB to 2 dB worse phase
noise.
[4:0] CLK_DUTY_PRG
Program the duty cycle offset. 0x0
5-bit signed magnitude field,
with the MSB as the sign bit
and the four LSBs as the
R/W
magnitude from 0 to 15. A
larger magnitude skews duty
cycle to a greater amount.
Range is ±3%.
0x083
0x084
CLK_CRS_CTRL
7
CLK_CRS_EN
Enable clock cross control
adjustment.
0x1
R/W
[6:4] Reserved
Reserved.
0x0
0x0
R
[3:0] CLK_CRS_ADJ
Program the clock crossing
point.
R/W
PLL_REF_CLK_PD
[7:6] Reserved
Reserved.
0x0
0x0
R
[5:4] PLL_REF_CLK_RATE
PLL reference clock rate
multiplier.
R/W
00 Normal rate (1×) PLL reference
clock.
01 Double rate (2×) PLL reference
clock.
10 Quadruple rate (4×) PLL
reference clock.
11 Disable the PLL reference
clock.
[3:1] Reserved
Reserved.
0x0
R
Rev. 0 | Page 83 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0
PLL_REF_CLK_PD
PLL reference clock power-
down.
0x0
R/W
0 Enable the PLL reference
clock.
1 Power down the PLL reference
clock.
0x088
SYSREF_CTRL0
[7:4] Reserved
Reserved.
0x0
0x0
R
3
HYS_ON
SYSREF± hysteresis enable.
This bit enables the
programmable hysteresis
control for the SYSREF±
receiver.
R/W
2
SYSREF_RISE
Use SYSREF± rising edge.
0x0
0x0
R/W
R/W
[1:0] HYS_CNTRL[9:8]
[7:0] HYS_CNTRL[7:0]
[7:5] Reserved
Controls the amount of
hysteresis in the SYSREF±
receiver. Each of the 10 bits
adds 10 mV of differential
hysteresis to the receiver
input.
0x089
0x090
SYSREF_CTRL1
DLL_PD
Controls the amount of
hysteresis in the SYSREF±
receiver. Each of the 10 bits
adds 10 mV of differential
hysteresis to the receiver
input.
0x0
R/W
Reserved.
0x0
0x1
R
4
3
2
1
0
DLL_FINE_DC_EN
DLL_FINE_XC_EN
DLL_COARSE_DC_EN
DLL_COARSE_XC_EN
DLL_CLK_PD
Fine delay line duty cycle
correction enable.
R/W
Fine delay line cross control
enable.
0x1
0x1
R/W
R/W
R/W
R/W
Coarse delay line duty cycle
correction enable.
Coarse delay line cross control 0x1
enable.
Powers down DLL and digital 0x1
clock generator.
0 Power up DLL controller.
1 Power down DLL controller.
0x091
DLL_CTRL
7
6
5
DLL_TRACK_ERR
DLL_SEARCH_ERR
DLL_SLOPE
Track error behavior.
0 Continue on error.
1 Restart on error.
Search error behavior.
0 Stop on error.
1 Retry on error.
Desired slope.
0 Negative slope.
1 Positive slope.
Search direction.
0x1
0x1
0x1
0x2
R/W
R/W
R/W
R/W
[4:3] DLL_SEARCH
00 Search down from initial point
only.
01 Search up from initial point
only.
10 Search up and down from
initial point.
[2:1] DLL_MODE
Controller mode.
00 Search then track.
01 Track only.
0x0
R/W
10 Search only.
Rev. 0 | Page 84 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0
DLL_ENABLE
Controller enable.
0x0
R/W
0 Disable DLL controller: use
static SPI settings.
1 Enable DLL controller: use
controller with feedback loop.
0x092
DLL_STATUS
[7:3] Reserved
Reserved.
0x0
R
R
2
1
0
DLL_FAIL
The device clock DLL failed to 0x0
lock.
DLL_LOST
DLL_LOCKED
The device clock DLL has lost 0x0
lock.
R/W
R
The device clock DLL has
achieved lock.
0x0
0x093
0x094
DLL_GB
[7:4] Reserved
[3:0] DLL_GUARD
[7:6] Reserved
[5:0] DLL_COARSE
[7:0] DLL_FINE
[7:5] Reserved
[4:0] DLL_PHS
Reserved.
0x0
0x0
0x0
0x0
R
Search guard band.
Reserved.
R/W
R
DLL_COARSE
Coarse delay line setpoint.
Fine delay line setpoint.
Reserved.
R/W
0x095
0x096
DLL_FINE
0x80 R/W
DLL_PHASE
0x0
0x8
R
Desired phase.
R/W
0 Minimum allowed phase.
16 Maximum allowed phase.
Reserved.
0x097
DLL_BW
[7:5] Reserved
0x0
0x0
R
[4:2] DLL_FILT_BW
Phase measurement filter
bandwidth.
R/W
[1:0] DLL_WEIGHT
[7:1] Reserved
Tracking speed.
Reserved.
0x0
0x0
R/W
R
0x098
0x099
DLL_READ
0
DLL_READ
Read request: 0 to 1 transition 0x0
updates the coarse, fine, and
phase readback values.
R/W
DLL_COARSE_RB
[7:6] Reserved
Reserved.
0x0
0x0
0x0
0x0
0x0
0x0
0x0
R
[5:0] DLL_COARSE_RB
[7:0] DLL_FINE_RB
[7:5] Reserved
Coarse delay line readback.
Fine delay line readback.
Reserved.
R
0x09A
0x09B
DLL_FINE_RB
R
DLL_PHASE_RB
R
[4:0] DLL_PHS_RB
[7:3] Reserved
Phase readback.
R
0x09D
DIG_CLK_INVERT
Reserved.
R
2
INV_DIG_CLK
Invert digital clock from DLL.
0 Normal polarity.
1 Inverted polarity.
R/W
1
0
7
DIG_CLK_DC_EN
DIG_CLK_XC_EN
DLL_TEST_EN
Digital clock duty cycle
correction enable.
Digital clock cross control
enable.
0x1
0x1
R/W
R/W
0x0A0
0x110
DLL_CLK_DEBUG
INTERP_MODE
DLL clock output test enable. 0x0
R/W
R
[6:2] Reserved
Reserved.
0x0
0x0
0x8
[1:0] DLL_TEST_DIV
[7:4] JESD_LANES
DLL clock output divide.
R/W
R/W
Number of JESD204B lanes.
For proper operation of the
JESD204B data link, this signal
must only be programmed
while the QBD is held in soft
reset (Register 0x475, Bit 3),
and must not be changed
during normal operation.
Rev. 0 | Page 85 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
[3:0] INTERP_MODE
Interpolation mode. For
proper operation of the
0x1
R/W
JESD204B data link, this signal
must only be programmed
while the QBD is held in soft
reset (Register 0x475, Bit 3),
and must not be changed
during normal operation.
0000 1× (bypass).
0001 2×.
0010 3×.
0011 4×.
0100 6×.
0101 8×.
0110 12×.
0111 16×.
1000 24×.
0x111
DATAPATH_CFG
7
6
INVSINC_EN
NCO_EN
Inverse sinc filter enable.
0x0
0x0
R/W
R/W
0 Disable inverse sinc filter.
1 Enable inverse sinc filter.
Modulation enable.
0 Disable NCO.
1 Enable NCO.
5
4
Reserved
FILT_BW
Reserved.
0x0
0x0
R
Datapath filter bandwidth.
0 Filter bandwidth is 80%.
1 Filter bandwidth is 90%.
Reserved.
R/W
3
2
Reserved
0x0
0x0
R
MODULUS_EN
Modulus DDS enable.
0 Disable modulus DDS.
1 Enable modulus DDS.
R/W
1
SEL_SIDEBAND
Selects upper or lower
sideband from modulation
result.
0x0
R/W
0 Use upper sideband.
1 Use lower sideband = spectral
flip.
0
7
FIR85_FILT_EN
Reserved
FIR85 filter enable.
Reserved.
0x0
0x0
0x0
R/W
R
0x113
FTW_UPDATE
[6:4] FTW_REQ_MODE
Frequency tuning word (FTW)
automatic update mode.
R/W
000 No automatic requests are
generated when the FTW
registers are written.
001 Automatically generate
FTW_LOAD_REQ after FTW0 is
written.
010 Automatically generate
FTW_LOAD_REQ after FTW1 is
written.
011 Automatically generate
FTW_LOAD_REQ after FTW2 is
written.
100 Automatically generate
FTW_LOAD_REQ after FTW3 is
written.
101 Automatically generate
FTW_LOAD_REQ after FTW4 is
written.
Rev. 0 | Page 86 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
110 Automatically generate
FTW_LOAD_REQ after FTW5 is
written.
3
2
Reserved
Reserved.
0x0
R
FTW_LOAD_SYSREF
FTW load and reset from rising 0x0
edge of SYSREF±.
R/W
1
FTW_LOAD_ACK
FTW_LOAD_REQ
Frequency tuning word
update acknowledge.
0 FTW is not loaded.
1 FTW is loaded.
0x0
R
0
Frequency tuning word
update request from SPI.
0x0
R/W
0 Clear FTW_LOAD_ACK.
1 0 to 1 transition loads the
FTW.
0x114
0x115
0x116
0x117
0x118
0x119
FTW0
FTW1
FTW2
FTW3
FTW4
FTW5
[7:0] FTW[7:0]
NCO frequency tuning word. 0x0
R/W
R/W
R/W
R/W
R/W
R/W
This is X in the equation fOUT
fCLK × (M/N) = fCLK × ((X +
A/B)/248).
=
[7:0] FTW[15:8]
[7:0] FTW[23:16]
[7:0] FTW[31:24]
[7:0] FTW[39:32]
[7:0] FTW[47:40]
NCO frequency tuning word. 0x0
This is X in the equation fOUT
fCLK × (M/N) = fCLK × ((X +
A/B)/248).
=
NCO frequency tuning word. 0x0
This is X in the equation fOUT
fCLK × (M/N) = fCLK × ((X +
A/B)/248).
=
NCO frequency tuning word. 0x0
This is X in the equation fOUT
fCLK × (M/N) = fCLK × ((X +
A/B)/248).
=
NCO frequency tuning word. 0x0
This is X in the equation fOUT
CLK × (M/N) = fCLK × ((X +
A/B)/248).
NCO frequency tuning word. 0x0
=
f
This is X in the equation fOUT
fCLK × (M/N) = fCLK × ((X +
A/B)/248).
=
0x11C
0x11D
0x124
PHASE_OFFSET0
PHASE_OFFSET1
ACC_MODULUS0
[7:0] NCO_PHASE_OFFSET[7:0]
[7:0] NCO_PHASE_OFFSET[15:8]
[7:0] ACC_MODULUS[7:0]
NCO phase offset.
NCO phase offset.
0x0
0x0
R/W
R/W
R/W
DDS modulus. This is B in the 0x0
equation fOUT = fCLK × (M/N) = fCLK
× ((X + A/B)/248). This modulus
value is used for all NCO FTWs.
0x125
0x126
ACC_MODULUS1
ACC_MODULUS2
[7:0] ACC_MODULUS[15:8]
[7:0] ACC_MODULUS[23:16]
DDS modulus. This is B in the 0x0
equation fOUT = fCLK × (M/N) = fCLK
× ((X + A/B)/248). This modulus
value is used for all NCO FTWs.
R/W
R/W
DDS modulus. This is B in the 0x0
equation fOUT = fCLK × (M/N) =
fCLK × ((X + A/B)/248). This
modulus value is used for all
NCO FTWs.
0x127
ACC_MODULUS3
[7:0] ACC_MODULUS[31:24]
DDS modulus. This is B in the 0x0
equation fOUT = fCLK × (M/N) =
fCLK × ((X + A/B)/248). This
modulus value is used for all
NCO FTWs.
R/W
Rev. 0 | Page 87 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x128
ACC_MODULUS4
ACC_MODULUS5
ACC_DELTA0
[7:0] ACC_MODULUS[39:32]
[7:0] ACC_MODULUS[47:40]
[7:0] ACC_DELTA[7:0]
DDS modulus. This is B in the 0x0
equation fOUT = fCLK × (M/N) =
R/W
R/W
R/W
f
CLK × ((X + A/B)/248). This
modulus value is used for all
NCO FTWs.
0x129
0x12A
DDS modulus. This is B in the 0x0
equation fOUT = fCLK × (M/N) =
fCLK × ((X + A/B)/248). This
modulus value is used for all
NCO FTWs.
DDS delta. This is A in the
equation fOUT = fCLK × (M/N) =
fCLK × ((X + A/B)/248). This
modulus value is used for all
NCO FTWs. Note this delta
value is used for all NCO FTWs.
0x0
0x0
0x0
0x12B
0x12C
ACC_DELTA1
ACC_DELTA2
[7:0] ACC_DELTA[15:8]
[7:0] ACC_DELTA[23:16]
DDS delta. This is A in the
equation fOUT = fCLK × (M/N) =
fCLK × ((X + A/B)/248). This
modulus value is used for all
NCO FTWs. Note this delta
value is used for all NCO FTWs.
R/W
R/W
DDS delta. This is A in the
equation fOUT = fCLK × (M/N) =
f
CLK × ((X + A/B)/248). Note this
modulus value is used for all
NCO FTWs. Note this delta
value is used for all NCO FTWs.
0x12D
0x12E
ACC_DELTA3
ACC_DELTA4
[7:0] ACC_DELTA[31:24]
[7:0] ACC_DELTA[39:32]
DDS delta. This is A in the
0x0
0x0
R/W
R/W
equation fOUT = fCLK × (M/N) =
f
CLK × ((X + A/B)/248). This delta
value is used for all NCO FTWs.
DDS Delta. This is A in the
equation fOUT = fCLK × (M/N) =
fCLK × ((X + A/B)/248). This
modulus value is used for all
NCO FTWs. Note this delta
value is used for all NCO FTWs.
0x12F
ACC_DELTA5
[7:0] ACC_DELTA[47:40]
DDS delta. This is A in the
equation fOUT = fCLK × (M/N) =
fCLK × ((X + A/B)/248). This
0x0
R/W
modulus value is used for all
NCO FTWs. Note this delta
value is used for all NCO FTWs.
0x132
0x133
0x134
TEMP_SENS_LSB
[7:0] TEMP_SENS_OUT[7:0]
[7:0] TEMP_SENS_OUT[15:8]
[7:1] Reserved
Output of the temperature
sensor ADC.
0x0
0x0
R
R
TEMP_SENS_MSB
TEMP_SENS_UPDATE
Output of the temperature
sensor ADC.
Reserved.
0x0
0x0
R
0
TEMP_SENS_UPDATE
Set to 1 to update the
temperature sensor reading
with a new value.
R/W
0x135
TEMP_SENS_CTRL
7
TEMP_SENS_FAST
A 1 sets the temperature
0x0
R/W
sensor digital filter bandwidth
wider for faster settling time.
[6:1] Reserved
TEMP_SENS_ENABLE
Reserved.
0x10 R/W
0x0 R/W
0
Set to 1 to enable the
temperature sensor.
Rev. 0 | Page 88 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x14B
PRBS
7
PRBS_GOOD_Q
Good data indicator,
imaginary channel.
0x0
R
0 Incorrect sequence detected.
1 Correct PRBS sequence
detected.
6
PRBS_GOOD_I
Good data indicator, real
channel.
0x0
R
0 Incorrect sequence detected.
1 Correct PRBS sequence
detected.
5
4
Reserved
Reserved.
0x0
0x1
R
PRBS_INV_Q
Data inversion, imaginary
channel.
R/W
0 Expect normal data.
1 Expect inverted data.
Data inversion, real channel.
0 Expect normal data.
1 Expect inverted data.
Polynomial select.
0 7-bit: x7 + x6 + 1.
1 15-bit: x15 + x14 + 1.
Reset error counters.
0 Normal operation.
1 Reset counters.
3
2
1
0
PRBS_INV_I
PRBS_MODE
PRBS_RESET
PRBS_EN
0x0
0x0
0x0
0x0
R/W
R/W
R/W
R/W
Enable PRBS checker.
0 Disable.
1 Enable.
0x14C
0x14D
PRBS_ERROR_I
PRBS_ERROR_Q
[7:0] PRBS_COUNT_I
[7:0] PRBS_COUNT_Q
Error count value real channel. 0x0
R
R
Error count value imaginary
channel.
0x0
0x14E
0x14F
0x150
TEST_DC_DATA1
TEST_DC_DATA0
DIG_TEST
[7:0] DC_TEST_DATA[15:8]
[7:0] DC_TEST_DATA[7:0]
[7:2] Reserved
DC test data.
DC test data.
0x0
0x0
0x0
0x0
R/W
R/W
R
Reserved.
1
DC_TEST_EN
DC data test mode enable.
1 DC test mode enable.
0 DC test mode disable.
Reserved.
R/W
0
Reserved
0x0
0x0
R/W
R/W
R/W
0x151
DECODE_CTRL
[7:3] Reserved
Shuffle
Reserved.
2
Shuffle mode. Enables shuffle 0x0
mode for improved spurious
performance.
0 Disable MSB shuffling (use
thermometer encoding).
1 Enable MSB shuffling.
[1:0] Reserved
Reserved.
0x0
0x0
0x0
R/W
R
0x152
0x1DF
DECODE_MODE
SPI_STRENGTH
[7:2] Reserved
Reserved.
[1:0] DECODE_MODE
Decode mode.
R/W
00 NRZ mode (first Nyquist).
01 Mix mode (second Nyquist).
10 Return to zero.
11 Reserved.
[7:4] Reserved
[3:0] SPIDRV
Reserved.
0x0
0xF
R
Slew and drive strength for
CMOS SPI outputs. Slew =
Bits[1:0], drive = Bits[3:2].
R/W
Rev. 0 | Page 89 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x200
MASTER_PD
[7:1] Reserved
Reserved.
0x0
0x1
R
0
SPI_PD_MASTER
Powers down the entire
JESD204B Rx analog (all eight
channels and bias).
R/W
0x201
PHY_PD
[7:0] SPI_PD_PHY
SPI override to power down
the individual PHYs.
0x0
R/W
Bit 0 controls SERDIN0± PHY.
Bit 1 controls SERDIN1± PHY.
Bit 2 controls SERDIN2± PHY.
Bit 3 controls SERDIN3± PHY.
Bit 4 controls SERDIN4± PHY.
Bit 5 controls SERDIN5± PHY.
Bit 6 controls SERDIN6± PHY.
Bit 7 controls SERDIN7± PHY.
Reserved.
0x203
0x206
0x230
GENERIC_PD
CDR_RESET
[7:2] Reserved
0x0
R
1
SPI_SYNC1_PD
Powers down LVDS buffer for 0x0
the sync request signal,
R/W
SYNCOUT±
.
0
Reserved
Reserved.
Reserved.
0x0
0x0
R/W
R
[7:1] Reserved
0
RESET
SPI_CDR_
Resets the digital control logic 0x1
for all PHYs.
0 CDR logic is reset.
R/W
1 CDR logic is operational.
CDR_OPERATING_MODE_
REG_0
[7:6] Reserved
SPI_ENHALFRATE
Reserved.
0x0
0x1
R/W
R/W
5
Enables half rate CDR
operation, must be enabled
for data rates > 6 Gbps.
0 Disables CDR half rate
operation, data rate ≤ 6 Gbps.
1 Enables CDR half rate
operation, data rate > 6 Gbps.
[4:3] Reserved
Reserved.
0x1
0x0
R/W
R/W
[2:1] SPI_DIVISION_RATE
Enables oversampling of the
input data.
00 No division. Data rate > 3 Gbps.
01 Division by 2. 1.5 Gbps < data
rate ≤ 3 Gbps.
10 Division by 4. 750 Mbps < data
rate ≤ 1.5 Gbps.
0
Reserved
Reserved.
0x0
R/W
R/W
0x250
EQ_CONFIG_PHY_0_1
[7:4] SPI_EQ_CONFIG1
Controls equalizer boost level. 0x8
0000 Manual mode (SPI configured
values used).
0001 Boost level = 1.
0010 Boost level = 2.
0011 Boost level = 3.
0100 Boost level = 4.
0101 Boost level = 5.
0110 Boost level = 6.
0111 Boost level = 7.
1000 Boost level = 8.
1001 Boost level = 9.
1010 Boost level = 10.
1011 Boost level = 11.
1100 Boost level = 12.
Rev. 0 | Page 90 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
1101 Boost level = 13.
1110 Boost level = 14.
1111 Boost level = 15.
[3:0] SPI_EQ_CONFIG0
[7:4] SPI_EQ_CONFIG3
[3:0] SPI_EQ_CONFIG2
Controls equalizer boost level. 0x8
0000 Manual mode (SPI configured
values used).
R/W
R/W
R/W
0001 Boost level = 1.
0010 Boost level = 2.
0011 Boost level = 3.
0100 Boost level = 4.
0101 Boost level = 5.
0110 Boost level = 6.
0111 Boost level = 7.
1000 Boost level = 8.
1001 Boost level = 9.
1010 Boost level = 10.
1011 Boost level = 11.
1100 Boost level = 12.
1101 Boost level = 13.
1110 Boost level = 14.
1111 Boost level = 15.
Controls equalizer boost level. 0x8
0x251
EQ_CONFIG_PHY_2_3
0000 Manual mode (SPI configured
values used).
0001 Boost level = 1.
0010 Boost level = 2.
0011 Boost level = 3.
0100 Boost level = 4.
0101 Boost level = 5.
0110 Boost level = 6.
0111 Boost level = 7.
1000 Boost level = 8.
1001 Boost level = 9.
1010 Boost level = 10.
1011 Boost level = 11.
1100 Boost level = 12.
1101 Boost level = 13.
1110 Boost level = 14.
1111 Boost level = 15.
Controls equalizer boost level. 0x8
0000 Manual mode (SPI configured
values used).
0001 Boost level = 1.
0010 Boost level = 2.
0011 Boost level = 3.
0100 Boost level = 4.
0101 Boost level = 5.
0110 Boost level = 6.
0111 Boost level = 7.
1000 Boost level = 8.
1001 Boost level = 9.
1010 Boost level = 10.
1011 Boost level = 11.
1100 Boost level = 12.
Rev. 0 | Page 91 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
1101 Boost Level = 13.
1110 Boost level = 14.
1111 Boost level = 15.
0x252
EQ_CONFIG_PHY_4_5
[7:4] SPI_EQ_CONFIG5
[3:0] SPI_EQ_CONFIG4
[7:4] SPI_EQ_CONFIG7
Controls equalizer boost level. 0x8
0000 Manual mode (SPI configured
values used).
R/W
R/W
R/W
0001 Boost level = 1.
0010 Boost level = 2.
0011 Boost level = 3.
0100 Boost level = 4.
0101 Boost level = 5.
0110 Boost level = 6.
0111 Boost level = 7.
1000 Boost level = 8.
1001 Boost level = 9.
1010 Boost level = 10.
1011 Boost level = 11.
1100 Boost level = 12.
1101 Boost level = 13.
1110 Boost level = 14.
1111 Boost level = 15.
Controls equalizer boost level. 0x8
0000 Manual mode (SPI configured
values used).
0001 Boost level = 1.
0010 Boost level = 2.
0011 Boost level = 3.
0100 Boost level = 4.
0101 Boost level = 5.
0110 Boost level = 6.
0111 Boost level = 7.
1000 Boost level = 8.
1001 Boost level = 9.
1010 Boost level = 10.
1011 Boost level = 11.
1100 Boost level = 12.
1101 Boost level = 13.
1110 Boost level = 14.
1111 Boost level = 15.
Controls equalizer boost level. 0x8
0x253
EQ_CONFIG_PHY_6_7
0000 Manual mode (SPI configured
values used).
0001 Boost level = 1.
0010 Boost level = 2.
0011 Boost level = 3.
0100 Boost level = 4.
0101 Boost level = 5.
0110 Boost level = 6.
0111 Boost level = 7.
1000 Boost level = 8.
1001 Boost level = 9.
1010 Boost level = 10.
1011 Boost level = 11.
1100 Boost level = 12.
Rev. 0 | Page 92 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
1101 Boost level = 13.
1110 Boost level = 14.
1111 Boost level = 15.
[3:0] SPI_EQ_CONFIG6
Controls equalizer boost level. 0x8
R/W
0000 Manual mode (SPI configured
values used).
0001 Boost level = 1.
0010 Boost level = 2.
0011 Boost level = 3.
0100 Boost level = 4.
0101 Boost level = 5.
0110 Boost level = 6.
0111 Boost level = 7.
1000 Boost level = 8.
1001 Boost level = 9.
1010 Boost level = 10.
1011 Boost level = 11.
1100 Boost level = 12.
1101 Boost level = 13.
1110 Boost level = 14.
1111 Boost level = 15.
0x268
0x280
EQ_BIAS_REG
[7:6] EQ_POWER_MODE
Controls the equalizer power 0x1
mode/insertion loss capability.
00 Normal mode.
R/W
01 Low power mode.
[5:0] Reserved
[7:3] Reserved
Reserved.
Reserved.
0x4
0x0
0x0
R/W
R
SYNTH_ENABLE_CNTRL
2
SPI_RECAL_SYNTH
Set this bit high to rerun all of
the SERDES PLL calibration
routines. Set this bit low again
to allow additional recalibra-
tions. Rising edge causes the
calibration.
R/W
1
0
Reserved
Reserved.
0x0
R/W
R/W
SPI_ENABLE_SYNTH
Enable the SERDES PLL. Setting 0x0
this bit turns on all currents
and proceeds to calibrate the
PLL. Make sure reference clock
and division ratios are correct
before enabling this bit.
0x281
PLL_STATUS
[7:6] Reserved
Reserved.
0x0
0x0
R
R
5
4
SPI_CP_OVER_RANGE_HIGH_RB
If set, the SERDES PLL CP
output is above the valid
operating range.
0 Charge pump output is within
operating range.
1 Charge pump output is above
operating range.
SPI_CP_OVER_RANGE_LOW_RB
If set, the SERDES PLL CP
output is below the valid
operating range.
0x0
R
0 Charge pump output is within
operating range.
1 Charge pump output is below
operating range.
Rev. 0 | Page 93 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
3
SPI_CP_CAL_VALID_RB
This bit tells the user if the
charge pump calibration has
completed and is valid.
0x0
R
0 Charge pump calibration is
not valid.
1 Charge pump calibration is
valid.
[2:1] Reserved
Reserved.
0x0
R
R
0
SPI_PLL_LOCK_RB
If set, the SERDES synthesizer 0x0
is locked.
0 PLL is not locked.
1 PLL is locked.
0x289
REF_CLK_DIVIDER_LDO
[7:2] Reserved
Reserved.
0x0
0x0
R
[1:0] SERDES_PLL_DIV_FACTOR
SERDES PLL reference clock
division factor. This field
R/W
controls the division of the
SERDES PLL reference clock
before it is fed into the
SERDES PLL PFD. It must be
set so that fREF/division factor is
between 35 MHz and 80 MHz.
00 Divide by 4 for lane rate
between 6 Gbps and
12.5 Gbps.
01 Divide by 2 for lane rate
between 3 Gbps and 6 Gbps.
10 Divide by 1 for lane rate
between 1.5 Gbps and 3 Gbps.
0x2A7
0x2A8
TERM_BLK1_CTRLREG0
TERM_BLK1_CTRLREG1
[7:1] Reserved
Reserved.
0x0
R
0
SPI_I_TUNE_R_CAL_TERMBLK1
Rising edge of this bit starts a 0x0
termination calibration
routine.
R/W
[7:0] SPI_I_SERIALIZER_RTRIM_
TERMBLK1
SPI override for termination
value for PHY 0, PHY 1, PHY 6,
and PHY 7. Value options are
as follows:
0x0
R/W
XXX0XXXX Automatically calibrate
termination value.
XXX1000X Force 000 as termination value.
XXX1001X Force 001 as termination value.
XXX1010X Force 010 as termination value.
XXX1011X Force 011 as termination value.
XXX1100X Force 100 as termination value.
XXX1101X Force 101 as termination value.
XXX1110X Force 110 as termination value.
XXX1111X Force 111 as termination value.
XXX1000X Force 000 as termination value.
Reserved.
0x2AC
0x2AE
TERM_BLK1_RD_REG0
TERM_BLK2_CTRLREG0
[7:4] Reserved
0x0
R
R
[3:0] SPI_O_RCAL_CODE_TERMBLK1
Readback of calibration code 0x0
for PHY 0, PHY 1, PHY 6, and
PHY 7.
[7:1] Reserved
Reserved.
0x0
R
0
SPI_I_TUNE_R_CAL_TERMBLK2
Rising edge of this bit starts a 0x0
termination calibration
routine.
R/W
Rev. 0 | Page 94 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x2AF
TERM_BLK2_CTRLREG1
[7:0] SPI_I_SERIALIZER_RTRIM_
TERMBLK2
SPI override for termination
value for PHY 2, PHY 3, PHY 4,
and PHY 5. Value options are
as follows:
0x0
R/W
XXX0XXXX Automatically calibrate
termination value.
XXX1000X Force 000 as termination value.
XXX1001X Force 001 as termination value.
XXX1010X Force 010 as termination value.
XXX1011X Force 011 as termination value.
XXX1100X Force 100 as termination value.
XXX1101X Force 101 as termination value.
XXX1110X Force 110 as termination value.
XXX1111X Force 111 as termination value.
XXX1000X Force 000 as termination value.
Reserved.
0x2B3
0x2BB
TERM_BLK2_RD_REG0
TERM_OFFSET_0
[7:4] Reserved
0x0
R
R
[3:0] SPI_O_RCAL_CODE_TERMBLK2
Readback of calibration code 0x0
for PHY 2, PHY 3, PHY 4, and
PHY 5.
[7:4] Reserved
Reserved.
0x0
0x0
R
[3:0] TERM_OFFSET_0
Add or subtract from the
termination calibration value
of Physical Lane 0. 4-bit signed
magnitude value that adds to
or subtracts from the
R/W
termination value. Bit 3 is the
sign bit, and Bits[2:0] are the
magnitude bits.
0x2BC
0x2BD
0x2BE
TERM_OFFSET_1
TERM_OFFSET_2
TERM_OFFSET_3
[7:4] Reserved
Reserved.
0x0
0x0
R
[3:0] TERM_OFFSET_1
Add or subtract from the
termination calibration value
of Physical Lane 1. 4-bit signed
magnitude value that adds to
or subtracts from the
termination value. Bit 3 is the
sign bit, and Bits[2:0] are the
magnitude bits.
R/W
[7:4] Reserved
Reserved.
0x0
0x0
R
[3:0] TERM_OFFSET_2
Add or subtract from the
termination calibration value
of Physical Lane 2. 4-bit signed
magnitude value that adds to
or subtracts from the
termination value. Bit 3 is the
sign bit, and Bits[2:0] are the
magnitude bits.
R/W
[7:4] Reserved
Reserved.
0x0
0x0
R
[3:0] TERM_OFFSET_3
Add or subtract from the
termination calibration value
of Physical Lane 3. 4-bit signed
magnitude value that adds to
or subtracts from the
R/W
termination value. Bit 3 is the
sign bit, and Bits[2:0] are the
magnitude bits.
Rev. 0 | Page 95 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x2BF
TERM_OFFSET_4
TERM_OFFSET_5
TERM_OFFSET_6
TERM_OFFSET_7
GENERAL_JRX_CTRL_0
[7:4] Reserved
Reserved.
0x0
0x0
R
[3:0] TERM_OFFSET_4
Add or subtract from the
termination calibration value
of Physical Lane 4. 4-bit signed
magnitude value that adds to
or subtracts from the
termination value. Bit 3 is the
sign bit, and Bits[2:0] are the
magnitude bits.
R/W
0x2C0
0x2C1
0x2C2
0x300
[7:4] Reserved
Reserved.
0x0
0x0
R
[3:0] TERM_OFFSET_5
Add or subtract from the
termination calibration value
of Physical Lane 5. 4-bit signed
magnitude value that adds to
or subtracts from the
termination value. Bit 3 is the
sign bit, and Bits[2:0] are the
magnitude bits.
R/W
[7:4] Reserved
Reserved.
0x0
0x0
R
[3:0] TERM_OFFSET_6
Add or subtract from the
termination calibration value
of Physical Lane 6. 4-bit signed
magnitude value that adds to
or subtracts from the
termination value. Bit 3 is the
sign bit, and Bits[2:0] are the
magnitude bits.
R/W
[7:4] Reserved
Reserved.
0x0
0x0
R
[3:0] TERM_OFFSET_7
Add or subtract from the
termination calibration value
of Physical Lane 7. 4-bit signed
magnitude value that adds to
or subtracts from the
termination value. Bit 3 is the
sign bit, and Bits[2:0] are the
magnitude bits.
R/W
7
6
Reserved
Reserved.
0x0
0x0
R
CHECKSUM_MODE
JESD204B link parameter
R/W
checksum calculation method.
0 Checksum is sum of fields.
1 Checksum is sum of octets.
Reserved.
[5:1] Reserved
LINK_EN
0x0
0x0
R
0
This bit brings up the
JESD204B receiver when all
link parameters are
programmed and all clocks
are ready.
R/W
0x302
0x304
DYN_LINK_LATENCY_0
LMFC_DELAY_0
[7:5] Reserved
Reserved.
0x0
R
R
[4:0] DYN_LINK_LATENCY_0
Measurement of the JESD204B 0x0
link delay (in PCLK units).
Link 0 dynamic link latency.
Latency between current
deframer LMFC and the global
LMFC.
[7:5] Reserved
Reserved.
0x0
0x0
R
[4:0] LMFC_DELAY_0
Fixed part of the JESD204B
link delay (in PCLK units).
Delay in frame clock cycles for
global LMFC for Link 0.
R/W
Rev. 0 | Page 96 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x306
LMFC_VAR_0
[7:5] Reserved
Reserved.
0x0
R
[4:0] LMFC_VAR_0
Variable part of the JESD204B 0x1F R/W
link delay (in PCLK units).
Location in Rx LMFC where
JESD204B words are read out
from buffer. This setting must
not be more than 10 PCLKs.
0x308
0x309
0x30A
XBAR_LN_0_1
[7:6] Reserved
Reserved.
0x0
R
[5:3] SRC_LANE1
Select data from SERDINx±, for 0x1
Logical Lane 1.
R/W
000 Data is from SERDIN0±.
001 Data is from SERDIN1±.
010 Data is from SERDIN2±.
011 Data is from SERDIN3±.
100 Data is from SERDIN4±.
101 Data is from SERDIN5±.
110 Data is from SERDIN6±.
111 Data is from SERDIN7±.
[2:0] SRC_LANE0
Select data from SERDINx± for 0x0
Logical Lane 0.
R/W
000 Data is from SERDIN0±.
001 Data is from SERDIN1±.
010 Data is from SERDIN2±.
011 Data is from SERDIN3±.
100 Data is from SERDIN4±.
101 Data is from SERDIN5±.
110 Data is from SERDIN6±.
111 Data is from SERDIN7±.
XBAR_LN_2_3
[7:6] Reserved
Reserved.
0x0
R
[5:3] SRC_LANE3
Select data from SERDINx± for 0x3
Logical Lane 3.
R/W
000 Data is from SERDIN0±.
001 Data is from SERDIN1±.
010 Data is from SERDIN2±.
011 Data is from SERDIN3±.
100 Data is from SERDIN4±.
101 Data is from SERDIN5±.
110 Data is from SERDIN6±.
111 Data is from SERDIN7±.
[2:0] SRC_LANE2
Select data from SERDINx± for 0x2
Logical Lane 2.
R/W
000 Data is from SERDIN0±.
001 Data is from SERDIN1±.
010 Data is from SERDIN2±.
011 Data is from SERDIN3±.
100 Data is from SERDIN4±.
101 Data is from SERDIN5±.
110 Data is from SERDIN6±.
111 Data is from SERDIN7±.
XBAR_LN_4_5
[7:6] Reserved
Reserved.
0x0
R
[5:3] SRC_LANE5
Select data from SERDINx± for 0x5
Logical Lane 5.
R/W
000 Data is from SERDIN0±.
001 Data is from SERDIN1±.
010 Data is from SERDIN2±.
011 Data is from SERDIN3±.
100 Data is from SERDIN4±.
Rev. 0 | Page 97 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
101 Data is from SERDIN5±.
110 Data is from SERDIN6±.
111 Data is from SERDIN7±.
[2:0] SRC_LANE4
Select data from SERDINx± for 0x4
Logical Lane 4.
R/W
000 Data is from SERDIN0±.
001 Data is from SERDIN1±.
010 Data is from SERDIN2±.
011 Data is from SERDIN3±.
100 Data is from SERDIN4±.
101 Data is from SERDIN5±.
110 Data is from SERDIN6±.
111 Data is from SERDIN7±.
0x30B
XBAR_LN_6_7
[7:6] Reserved
Reserved.
0x0
R
[5:3] SRC_LANE7
Select data from SERDINx± for 0x7
Logical Lane 7.
R/W
000 Data is from SERDIN0±.
001 Data is from SERDIN1±.
010 Data is from SERDIN2±.
011 Data is from SERDIN3±.
100 Data is from SERDIN4±.
101 Data is from SERDIN5±.
110 Data is from SERDIN6±.
111 Data is from SERDIN7±.
[2:0] SRC_LANE6
Select data from SERDINx± for 0x6
Logical Lane 6.
R/W
000 Data is from SERDIN0±.
001 Data is from SERDIN1±.
010 Data is from SERDIN2±.
011 Data is from SERDIN3±.
100 Data is from SERDIN4±.
101 Data is from SERDIN5±.
110 Data is from SERDIN6±.
111 Data is from SERDIN7±.
0x30C
FIFO_STATUS_REG_0
[7:0] LANE_FIFO_FULL
Bit 0 corresponds to FIFO full 0x0
flag for data from SERDIN0±.
R
Bit 1 corresponds to FIFO full
flag for data from SERDIN1±.
Bit 2 corresponds to FIFO full
flag for data from SERDIN2±.
Bit 3 corresponds to FIFO full
flag for data from SERDIN3±.
Bit 4 corresponds to FIFO full
flag for data from SERDIN4±.
Bit 5 corresponds to FIFO full
flag for data from SERDIN5±.
Bit 6 corresponds to FIFO full
flag for data from SERDIN6±.
Bit 7 corresponds to FIFO full
flag for data from SERDIN7±.
Rev. 0 | Page 98 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
FIFO_STATUS_REG_1
No. Bit Name
Settings
Description
Reset Access
0x30D
[7:0] LANE_FIFO_EMPTY
Bit 0 corresponds to FIFO
empty flag for data from
SERDIN0±.
0x0
R
Bit 1 corresponds to FIFO
empty flag for data from
SERDIN1±.
Bit 2 corresponds to FIFO
empty flag for data from
SERDIN2±.
Bit 3 corresponds to FIFO
empty flag for data from
SERDIN3±.
Bit 4 corresponds to FIFO
empty flag for data from
SERDIN4±.
Bit 5 corresponds to FIFO
empty flag for data from
SERDIN5±.
Bit 6 corresponds to FIFO
empty flag for data from
SERDIN6±.
Bit 7 corresponds to FIFO
empty flag for data from
SERDIN7±.
0x311
SYNC
[7:3] Reserved
Reserved.
0x0
0x0
R
_GEN_0
2
EOMF_MASK_0
Mask end of multiframe
R/W
(EOMF) flag, based on output
from QBD Lane 0. Controls
SYNCOUT±
whether
is
asserted in response to loss of
multiframe sync.
0
1
SYNCOUT±
on
Do not assert
loss of multiframe.
SYNCOUT±
Assert
on loss of
multiframe.
1
0
Reserved
Reserved.
0x0
R/W
R/W
EOF_MASK_0
Mask end of frame (EOF) flag, 0x0
based on output from QBD
Lane 0. Controls whether
SYNCOUT±
is asserted in
response to loss of frame sync.
0
1
SYNCOUT±
Do not assert
loss of frame.
on
on loss of
SYNCOUT±
signal
SYNCOUT±
Assert
frame.
0x312
SYNC
_GEN_1
[7:4] SYNC
[3:0] SYNC
0x0
0x0
0x0
R/W
R/W
R
_ERR_DUR
Duration of
low for purpose of sync error
report. 0 means half PCLK
cycle. Add an additional PCLK
= 4 octets for each increment
of the value.
SYNCOUT±
signal
_SYNCREQ_DUR
Duration of
low for purpose of sync
request. 0 means 5 frames +
9 octets. Add an additional
PCLK = 4 octets for each
increment of the value.
0x313
SYNC
_GEN_3
[7:0] LMFC_PERIOD
LMFC period in PCLK cycle.
This is to report the global
LMFC period based on PCLK.
Rev. 0 | Page 99 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x315
PHY_PRBS_TEST_EN
[7:0] PHY_TEST_EN
Enable PHY BER by ungating
the clocks.
0x0
R/W
1 PHY test enable.
0 PHY test disable.
Reserved.
0x316
PHY_PRBS_TEST_CTRL
7
Reserved
0x0
0x0
R
[6:4] PHY_SRC_ERR_CNT
R/W
000 Report Lane 0 error count.
001 Report Lane 1 error count.
010 Report Lane 2 error count.
011 Report Lane 3 error count.
100 Report Lane 4 error count.
101 Report Lane 5 error count.
110 Report Lane 6 error count.
111 Report Lane 7 error count.
[3:2] PHY_PRBS_PAT_SEL
Select PRBS pattern for PHY
BER test.
0x0
R/W
00 PRBS7.
01 PRBS15.
10 PRBS31.
11 Not used.
1
0
PHY_TEST_START
PHY_TEST_RESET
Start and stop the PHY PRBS
test.
0 Test not started.
1 Test started.
0x0
0x0
R/W
R/W
Reset PHY PRBS test state
machine and error counters.
0 Not reset.
1 Reset.
0x317
0x318
0x319
0x31A
0x31B
0x31C
0x31D
PHY_PRBS_TEST_
THRESHOLD_LOBITS
[7:0] PHY_PRBS_THRESHOLD_LOBITS
[7:0] PHY_PRBS_THRESHOLD_MIDBITS
[7:0] PHY_PRBS_THRESHOLD_HIBITS
[7:0] PHY_PRBS_ERR_CNT_LOBITS
[7:0] PHY_PRBS_ERR_CNT_MIDBITS
[7:0] PHY_PRBS_ERR_CNT_HIBITS
[7:0] PHY_PRBS_PASS
Bits[7:0] of the 24-bit
threshold value set the error
flag for PHY PRBS test.
0x0
0x0
0x0
R/W
R/W
R/W
R
PHY_PRBS_TEST_
THRESHOLD_MIDBITS
Bits[15:8] of the 24-bit
threshold value set the error
flag for PHY PRBS test.
PHY_PRBS_TEST_
THRESHOLD_HIBITS
Bits[23:16] of the 24-bit
threshold value set the error
flag for PHY PRBS test.
PHY_PRBS_TEST_ERRCNT_
LOBITS
Bits[7:0] of the 24-bit reported 0x0
PHY BER test error count from
selected lane.
PHY_PRBS_TEST_ERRCNT_
MIDBITS
Bits[15:8] of the 24-bit
reported PHY BER test error
count from selected lane.
0x0
R
PHY_PRBS_TEST_ERRCNT_
HIBITS
Bits[23:16] of the 24-bit
reported PHY BER test error
count from selected lane.
0x0
R
PHY_PRBS_TEST_STATUS
Each bit is for the
0xFF
R
corresponding lane. Report
PHY BER test pass/fail for each
lane.
Rev. 0 | Page 100 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x31E
PHY_DATA_SNAPSHOT_CTRL [7:5] Reserved
Reserved.
0x0
R
[4:2] PHY_GRAB_LANE_SEL
Select from which PHY lane to 0x0
grab data.
R/W
000 Grab data from Lane 0.
001 Grab data from Lane 1.
010 Grab data from Lane 2.
011 Grab data from Lane 3.
100 Grab data from Lane 4.
101 Grab data from Lane 5.
110 Grab data from Lane 6.
111 Grab data from Lane 7.
Use error trigger to grab data. 0x0
1
0
PHY_GRAB_MODE
PHY_GRAB_DATA
R/W
0 Grab data when
PHY_GRAB_DATA is set.
1 Grab data upon bit error.
Transition from 0 to 1 causes
logic to store current receive
data from one lane.
0x0
0x0
R/W
R
0x31F
0x320
0x321
0x322
0x323
PHY_SNAPSHOT_DATA_BYTE0 [7:0] PHY_SNAPSHOT_DATA_BYTE0
Stores a single byte,
PHY_SNAPSHOT_DATA,
Bits[7:0], of a 40-bit snapshot
(PHY_SNAPSHOT_DATA,
Bits[39:0]) as received on a
single PHY lane. The lane to be
captured and the capture
method is defined in
Register 0x31E.
PHY_SNAPSHOT_DATA_
BYTE1
[7:0] PHY_SNAPSHOT_DATA_BYTE1
[7:0] PHY_SNAPSHOT_DATA_BYTE2
[7:0] PHY_SNAPSHOT_DATA_BYTE3
Stores a single byte,
0x0
0x0
0x0
0x0
R
R
R
R
PHY_SNAPSHOT_DATA,
Bits[15:8], of a 40-bit snapshot
(PHY_SNAPSHOT_DATA,
Bits[39:0]) as received on a
single PHY lane. The lane to be
captured and the capture
method is defined in
Register 0x31E.
PHY_SNAPSHOT_DATA_
BYTE2
Stores a single byte,
PHY_SNAPSHOT_DATA,
Bits[23:16], of a 40-bit
snapshot (PHY_SNAPSHOT_
DATA, Bits[39:0]) as received
on a single PHY lane. The lane
to be captured and the
capture method is defined in
Register 0x31E.
PHY_SNAPSHOT_DATA_
BYTE3
Stores a single byte,
PHY_SNAPSHOT_DATA,
Bits[31:24], of a 40-bit
snapshot (PHY_SNAPSHOT_
DATA, Bits[39:0]) as received
on a single PHY lane. The lane
to be captured and the
capture method is defined in
Register 0x31E.
PHY_SNAPSHOT_DATA_
BYTE4
[7:0] PHY_SNAPSHOT_DATA_BYTE4
Stores a single byte,
PHY_SNAPSHOT_DATA,
Bits[39:32], of a 40-bit
snapshot (PHY_SNAPSHOT_
DATA, Bits[39:0]) as received
on a single PHY lane. The lane
to be captured and the
capture method is defined in
Register 0x31E.
Rev. 0 | Page 101 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x32C
SHORT_TPL_TEST_0
[7:4] SHORT_TPL_SP_SEL
Short transport layer sample
selection. Select which sample
to check from a specific DAC.
0x0
R/W
0000 Sample 0.
0001 Sample 1.
0010 Sample 2.
0011 Sample 3.
0100 Sample 4.
0101 Sample 5.
0110 Sample 6.
0111 Sample 7.
1000 Sample 8.
1001 Sample 9.
1010 Sample 10.
1011 Sample 11.
1100 Sample 12.
1101 Sample 13.
1110 Sample 14.
1111 Sample 15.
[3:2] SHORT_TPL_M_SEL
Short transport layer test DAC 0x0
selection. Select which DAC to
check.
R/W
00 DAC 0.
01 DAC 1.
10 DAC 2.
11 DAC 3.
1
0
SHORT_TPL_TEST_RESET
SHORT_TPL_TEST_EN
Short transport layer test
reset. Resets the result of short
transport layer test.
0x0
0x0
R/W
R/W
R/W
0 Not reset.
1 Reset.
Short transport layer test
enable. Enable short transport
layer test.
0 Disable.
1 Enable.
0x32D
0x32E
0x32F
SHORT_TPL_TEST_1
SHORT_TPL_TEST_2
SHORT_TPL_TEST_3
[7:0] SHORT_TPL_REF_SP_LSB
[7:0] SHORT_TPL_REF_SP_MSB
[7:1] Reserved
Short transport layer reference 0x0
sample LSB. This LSB is the
lower eight bits of expected
DAC sample, and is used to
compare with the received
DAC sample at the output of
JESD204B Rx.
Short transport layer test
reference sample MSB. This
LSB is the upper eight bits of
expected DAC sample, and is
used to compare with the
received sample at JESD204B
Rx output.
0x0
R/W
Reserved.
0x0
R
R
0
SHORT_TPL_FAIL
Short transport layer test fail. 0x0
This bit shows if the selected
DAC sample matches the
reference sample. If they
match, the test passes.
Otherwise, the test fails.
0 Test pass.
1 Test fail.
Rev. 0 | Page 102 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x334
JESD_BIT_INVERSE_CTRL
[7:0] JESD_BIT_INVERSE
Each bit of this byte inverses
the JESD204B deserialized
data from one specific
JESD204B Rx PHY. The bit
order matches the logical lane
order. For example, Bit 0
controls Lane 0, Bit 1 controls
Lane 1.
0x0
R/W
0x400
0x401
0x402
DID_REG
BID_REG
LID0_REG
[7:0] DID_RD
[7:0] BID_RD
Received ILAS configuration
on Lane 0. DID is the device ID
number. Link information
received on Lane 0 as
specified in Section 8.3 of
JESD204B.
0x0
0x0
R
R
Received ILAS configuration
on Lane 0. BID is the bank ID,
extension to DID. Link
information received on
Lane 0 as specified in
Section 8.3 of JESD204B.
7
6
Reserved
Reserved.
0x0
0x0
R
R
ADJDIR_RD
Received ILAS configuration
on Lane 0. ADJDIR is the
direction to adjust the DAC
LMFC. Link information received
on Lane 0 as specified in
Section 8.3 of JESD204B.
5
PHADJ_RD
Received ILAS configuration
on Lane 0. PHADJ is the phase
adjustment request to DAC.
Link information received on
Lane 0 as specified in
0x0
0x0
0x0
R
R
R
Section 8.3 of JESD204B.
[4:0] LL_LID0
Received ILAS LID
configuration on Lane 0. LID0
is the lane identification for
Lane 0. Link information
received on Lane 0 as specified
in Section 8.3 of JESD204B.
0x403
SCR_L_REG
7
SCR_RD
Received ILAS configuration
on Lane 0. SCR is the Tx
scrambling status. Link
information received on
Lane 0 as specified in
Section 8.3 of JESD204B.
0 Scrambling is disabled.
1 Scrambling is enabled.
Reserved.
[6:5] Reserved
[4:0] L_RD
0x0
0x0
R
R
Received ILAS configuration
on Lane 0. L is the number of
lanes per converter device.
Link information received on
Lane 0 as specified in
Section 8.3 of JESD204B.
00000 1 lane per converter device.
00001 2 lanes per converter device.
00011 4 lanes per converter device.
00111 8 lanes per converter device.
Rev. 0 | Page 103 of 138
AD9166
Data Sheet
Hex.
Addr.
0x404
Bit
Register Name
F_REG
No. Bit Name
[7:0] F_RD
Settings
Description
Reset Access
Received ILAS configuration
on Lane 0. F is the number of
octets per frame. Settings of 1,
2, and 4 are valid (value in
register is F − 1). Link infor-
mation received on Lane 0 as
specified in Section 8.3 of
JESD204B.
0x0
R
0 1 octet per frame.
1 2 octets per frame.
11 4 octets per frame.
Reserved.
0x405
K_REG
[7:5] Reserved
[4:0] K_RD
0x0
0x0
R
R
Received ILAS configuration
on Lane 0. K is the number of
frames per multiframe.
Settings of 16 or 32 are valid.
On this device, all modes use
K = 32 (value in register is K −
1). Link information received
on Lane 0 as specified in
Section 8.3 of JESD204B.
01111 16 frames per multiframe.
11111 32 frames per multiframe.
0x406
M_REG
[7:0] M_RD
Received ILAS configuration
on Lane 0. M is the number of
converters per device. Link
information received on
Lane 0 as specified in
0x0
R
Section 8.3 of JESD204B. M is 1
for real interface and 2 for
complex interface (value in
register is M − 1).
0x407
CS_N_REG
[7:6] CS_RD
Received ILAS configuration
on Lane 0. CS is the number of
control bits per sample. Link
information received on
Lane 0 as specified in
0x0
R
Section 8.3 of JESD204B. CS is
always 0 on this device.
5
Reserved
Reserved.
0x0
0x0
R
R
[4:0] N_RD
Received ILAS configuration
on Lane 0. N is the converter
resolution. Value in register is
N − 1 (for example, 16 bits =
0b01111).
0x408
NP_REG
[7:5] SUBCLASSV_RD
Received ILAS configuration
on Lane 0. SUBCLASSV is the
device subclass version. Link
information received on
Lane 0 as specified in
0x0
0x0
R
R
Section 8.3 of JESD204B.
000 Subclass 0.
001 Subclass 1.
[4:0] NP_RD
Received ILAS configuration
on Lane 0. NP is the total
number of bits per sample.
Link information received on
Lane 0 as specified in
Section 8.3 of JESD204B. Value
in register is NP − 1, for
example, 16 bits per sample =
0b01111.
Rev. 0 | Page 104 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x409
S_REG
[7:5] JESDV_RD
Received ILAS configuration
on Lane 0. JESDV is the
JESD204x version. Link
information received on
Lane 0 as specified in
0x0
R
Section 8.3 of JESD204B.
000 JESD204A.
001 JESD204B.
Received ILAS configuration
[4:0] S_RD
0x0
0x0
R
R
on Lane 0. S is the number of
samples per converter per
frame cycle. Link information
received on Lane 0 as
specified in Section 8.3 of
JESD204B. Value in register is
S − 1.
0x40A
HD_CF_REG
7
HD_RD
Received ILAS configuration
on Lane 0. HD is the high
density format. Refer to
Section 5.1.3 of JESD204B
standard. Link information
received on Lane 0 as
specified in Section 8.3 of
JESD204B.
0 Low density mode.
1 High density mode.
Reserved.
[6:5] Reserved
[4:0] CF_RD
0x0
0x0
R
R
Received ILAS configuration
on Lane 0. CF is the number of
control words per frame clock
period per link. Link
information received on
Lane 0 as specified in
Section 8.3 of JESD204B. CF is
always 0 on this device.
0x40B
0x40C
0x40D
RES1_REG
[7:0] RES1_RD
[7:0] RES2_RD
[7:0] LL_FCHK0
Received ILAS configuration
on Lane 0. Reserved Field 1.
Link information received on
Lane 0 as specified in
0x0
0x0
0x0
R
R
R
Section 8.3 of JESD204B.
RES2_REG
Received ILAS configuration
on Lane 0. Reserved Field 2.
Link information received on
Lane 0 as specified in
Section 8.3 of JESD204B.
CHECKSUM0_REG
Received checksum during
ILAS on Lane 0. Checksum for
Lane 0. Link information
received on Lane 0 as
specified in Section 8.3 of
JESD204B.
0x40E
COMPSUM0_REG
[7:0] LL_FCMP0
Computed checksum on
Lane 0. Computed checksum
for Lane 0. The JESD204B Rx
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
(Register 0x300, Bit 6) and
must match the likewise
calculated checksum in
0x0
R
Register 0x40D.
Rev. 0 | Page 105 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x412
LID1_REG
[7:5] Reserved
[4:0] LL_LID1
Reserved.
0x0
R
R
Received lane ID (LID) during 0x0
ILAS on Lane 1. Link information
received on Lane 0 as specified
in Section 8.3 of JESD204B.
0x415
CHECKSUM1_REG
[7:0] LL_FCHK1
[7:0] LL_FCMP1
Received checksum during
ILAS on Lane 1. Link information
received on Lane 0 as specified
in Section 8.3 of JESD204B.
0x0
R
R
0x416
0x41A
COMPSUM1_REG
LID2_REG
Computed checksum on
Lane 1 (see description for
Register 0x40E).
0x0
0x0
[7:5] Reserved
[4:0] LL_LID2
Reserved.
R
R
Received lane ID (LID) during 0x0
ILAS on Lane 2.
0x41D
0x41E
CHECKSUM2_REG
COMPSUM2_REG
[7:0] LL_FCHK2
[7:0] LL_FCMP2
Received checksum during
ILAS on Lane 2
0x0
R
R
Computed checksum on
Lane 2 (see description for
Register 0x40E).
0x0
0x422
LID3_REG
[7:5] Reserved
[4:0] LL_LID3
Reserved.
0x0
R
R
Received lane ID (LID) during 0x0
ILAS on Lane 3.
0x425
0x426
CHECKSUM3_REG
COMPSUM3_REG
[7:0] LL_FCHK3
[7:0] LL_FCMP3
Received checksum during
ILAS on Lane 3
0x0
R
R
Computed checksum on
Lane 3 (see description for
Register 0x40E).
0x0
0x42A
LID4_REG
[7:5] Reserved
[4:0] LL_LID4
Reserved.
0x0
0x0
R
R
Received LID during ILAS on
Lane 4.
0x42D
0x42E
CHECKSUM4_REG
COMPSUM4_REG
[7:0] LL_FCHK4
[7:0] LL_FCMP4
Received checksum during
ILAS on Lane 4
0x0
0x0
R
R
Computed checksum on
Lane 4 (see description for
Register 0x40E).
0x432
LID5_REG
[7:5] Reserved
[4:0] LL_LID5
Reserved.
0x0
0x0
R
R
Received LID during ILAS on
Lane 5.
0x435
0x436
CHECKSUM5_REG
COMPSUM5_REG
[7:0] LL_FCHK5
[7:0] LL_FCMP5
Received checksum during
ILAS on Lane 5
0x0
0x0
R
R
Computed checksum on
Lane 5 (see description for
Register 0x40E).
0x43A
LID6_REG
[7:5] Reserved
[4:0] LL_LID6
Reserved.
0x0
0x0
R
R
Received LID during ILAS on
Lane 6.
0x43D
0x43E
CHECKSUM6_REG
COMPSUM6_REG
[7:0] LL_FCHK6
[7:0] LL_FCMP6
Received checksum during
ILAS on Lane 6
0x0
0x0
R
R
Computed checksum on
Lane 6 (see description for
Register 0x40E).
0x442
LID7_REG
[7:5] Reserved
[4:0] LL_LID7
Reserved.
0x0
0x0
R
R
Received LID during ILAS on
Lane 7.
0x445
0x446
CHECKSUM7_REG
COMPSUM7_REG
[7:0] LL_FCHK7
[7:0] LL_FCMP7
Received checksum during
ILAS on Lane 7.
0x0
0x0
R
R
Computed checksum on
Lane 7 (see description for
Register 0x40E).
Rev. 0 | Page 106 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x450
ILS_DID
[7:0] DID
Device (link) identification
number (DID). DID is the
device ID number. Link
0x0
R/W
information received on Lane 0
as specified in Section 8.3 of
JESD204B. Must be set to the
value read in Register 0x400.
This signal must only be
programmed while the QBD
is held in soft reset
(Register 0x475, Bit 3), and
must not be changed during
normal operation.
0x451
0x452
ILS_BID
[7:0] BID
Bank ID, extension to DID. This 0x0
signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
R/W
ILS_LID0
7
6
Reserved
ADJDIR
Reserved.
0x0
R
Direction to adjust DAC LMFC 0x0
(Subclass 2 only). ADJDIR is
the direction to adjust DAC
LMFC. Link information received
on Lane 0 as specified in
R/W
Section 8.3 of JESD204B. This
signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
5
PHADJ
Phase adjustment to DAC
(Subclass 2 only). PHADJ is the
phase adjustment request to
the DAC. Link information
received on Lane 0 as specified
in Section 8.3 of JESD204B. This
signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
0x0
R/W
R/W
R/W
[4:0] LID0
Lane identification number
(within link). LID0 is the lane
identification for Lane 0. Link
information received on Lane 0
as specified in Section 8.3 of
JESD204B. This signal must
only be programmed while
the QBD is held in soft reset
(Register 0x475, Bit 3), and
must not be changed during
normal operation.
0x0
0x453
ILS_SCR_L
7
SCR
Scramble enable. SCR is the Rx 0x1
descrambling enable. This
signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
0 Descrambling is disabled.
1 Descrambling is enabled.
Rev. 0 | Page 107 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
[6:5] Reserved
[4:0] L
Reserved.
0x0
R
R
Number of lanes per converter 0x7
(minus 1). L is the number of
lanes per converter device.
Settings of 1, 2, 3, 4, 6, and 8
are valid. Refer to Table 15 and
Table 16.
0x454
0x455
ILS_F
ILS_K
[7:0] F
Number of octets per frame
(minus 1). This value of F is not
used to soft configure the
QBD. Register CTRLREG1 is
used to soft configure the QBD.
0x0
R
R
[7:5] Reserved
[4:0] K
Reserved.
0x0
Number of frames per
0x1F R/W
multiframe (minus 1). K is the
number of frames per
multiframe. On this device, all
modes use K = 32 (value in
register is K − 1). This signal
must only be programmed
while the QBD is held in soft
reset (Register 0x475, Bit 3),
and must not be changed
during normal operation.
01111 16 frames per multiframe.
11111 32 frames per multiframe.
0x456
0x457
ILS_M
[7:0] M
[7:6] CS
Number of converters per
device (minus 1). M is the
number of converters/device.
Settings of 1 and 2 are valid.
Refer to Table 15 and Table 16.
0x1
0x0
0x0
R
R
ILS_CS_N
Number of control bits per
sample. CS is the number of
control bits per sample. Must
be set to 0. Control bits are
not supported.
5
Reserved
Reserved.
R
R
[4:0] N
Converter resolution (minus 1). 0xF
N is the converter resolution.
Must be set to 16 (0x0F).
0x458
ILS_NP
[7:5] SUBCLASSV
Device subclass version.
0x0
R/W
SUBCLASSV is the device
subclass version. This signal
must only be programmed
while the QBD is held in soft
reset (Register 0x475, Bit 3),
and must not be changed
during normal operation.
000 Subclass 0.
001 Subclass 1.
010 Subclass 2 (not supported).
[4:0] NP
Total number of bits per
sample (minus 1). NP is the
total number of bits per
sample. Must be set to 16
(0x0F). Refer to Table 15 and
Table 16.
0xF
R
Rev. 0 | Page 108 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x459
ILS_S
[7:5] JESDV
JESD204x version. JESDV is the 0x0
JESD204x version. This signal
must only be programmed
while the QBD is held in soft
reset (Register 0x475, Bit 3), and
must not be changed during
normal operation.
R/W
000 JESD204A.
001 JESD204B.
Number of samples per
[4:0] S
0x1
R
R
converter per frame cycle
(minus 1). S is the number of
samples per converter per
frame cycle. Settings of 1 and
2 are valid. Refer to Table 15
and Table 16.
0x45A
ILS_HD_CF
7
HD
High density format. HD is the 0x1
high density mode. Refer to
Section 5.1.3 of JESD204B
standard.
0 Low density mode.
1 High density mode.
[6:5] Reserved
[4:0] CF
Reserved.
0x0
0x0
R
R
Number of control bits per
sample. CF is the number of
control words per frame clock
period per link. Must be set to
0. Control bits are not
supported.
0x45B
0x45C
0x45D
ILS_RES1
[7:0] RES1
[7:0] RES2
[7:0] FCHK0
Reserved. Reserved Field 1.
This signal must only be
programmed while the QBD is
held in soft reset
(Register 0x475, Bit 3), and
must not be changed during
normal operation.
0x0
0x0
R/W
R/W
R/W
ILS_RES2
Reserved. Reserved Field 2.
This signal must only be
programmed while the QBD is
held in soft reset
(Register 0x475, Bit 3), and
must not be changed during
normal operation.
ILS_CHECKSUM
Link configuration checksum. 0x0
Checksum for Lane 0. The
checksum for the values
programmed into
Register 0x450 to
Register 0x45C must be
calculated according to
Section 8.3 of the JESD204B
specification and written to this
register (SUM(Register 0x450 to
Register 0x45C) % 256). This
signal must only be
programmed while the QBD is
held in soft reset
(Register 0x475, Bit 3), and
must not be changed during
normal operation.
Rev. 0 | Page 109 of 138
AD9166
Data Sheet
Hex.
Addr.
0x46C
Bit
Register Name
LANE_DESKEW
No. Bit Name
Settings
Description
Reset Access
7
6
5
4
3
2
1
0
ILD7
ILS6
ILD5
ILD4
ILD3
ILD2
ILD1
ILD0
Interlane deskew status for
Lane 7 (ignore this output
when NO_ILAS = 1).
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
R
R
R
R
R
R
R
R
0 Deskew failed.
1 Deskew achieved.
Initial lane synchronization
status for Lane 6 (ignore this
output when NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
Interlane deskew status for
Lane 5 (ignore this output
when NO_ILAS = 1).
0 Deskew failed.
1 Deskew achieved.
Interlane deskew status for
Lane 4 (ignore this output
when NO_ILAS = 1).
0 Deskew failed.
1 Deskew achieved.
Interlane deskew status for
Lane 3 (ignore this output
when NO_ILAS = 1).
0 Deskew failed.
1 Deskew achieved.
Interlane deskew status for
Lane 2 (ignore this output
when NO_ILAS = 1).
0 Deskew failed.
1 Deskew achieved.
Interlane deskew status for
Lane 1 (ignore this output
when NO_ILAS = 1).
0 Deskew failed.
1 Deskew achieved.
Interlane deskew status for
Lane 0 (ignore this output
when NO_ILAS = 1).
0 Deskew failed.
1 Deskew achieved.
0x46D
BAD_DISPARITY
7
6
5
4
3
BDE7
BDE6
BDE5
BDE4
BDE3
BDE status for Lane 7.
0x0
0x0
0x0
0x0
0x0
R
R
R
R
R
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
BDE status for Lane 6.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
BDE status for Lane 5.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
BDE status for Lane 4.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
BDE status for Lane 3.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
Rev. 0 | Page 110 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
2
1
0
7
6
5
4
3
2
1
0
7
BDE2
BDE1
BDE0
NIT7
NIT6
NIT5
NIT4
NIT3
NIT2
NIT1
NIT0
UEK7
BDE status for Lane 2.
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
R
R
R
R
R
R
R
R
R
R
R
R
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
BDE status for Lane 1.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
BDE status for Lane 0.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT error status for Lane 7.
0x46E
NOT_IN_TABLE
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT error status for Lane 6.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT errors status for Lane 5.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT error status for Lane 4.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT error status for Lane 3.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT error status for Lane 2.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT error status for Lane 1.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT error status for Lane 0.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
0x46F
UNEXPECTED_KCHAR
UEK character error status for 0x0
Lane 7.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
6
5
4
3
UEK6
UEK5
UEK4
UEK3
UEK character error status for 0x0
Lane 6.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
R
R
R
R
UEK character error status for 0x0
Lane 5.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
UEK character error status for 0x0
Lane 4.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
UEK character error status for 0x0
Lane 3.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
Rev. 0 | Page 111 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
2
1
0
UEK2
UEK1
UEK0
UEK character error status for 0x0
Lane 2.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
R
R
R
UEK character error status for 0x0
Lane 1.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
UEK character error status for 0x0
Lane 0.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
0x470
CODE_GRP_SYNC
7
6
5
4
3
2
1
0
7
CGS7
CGS6
CGS5
CGS4
CGS3
CGS2
CGS1
CGS0
FS7
CGS status for Lane 7.
0 Synchronization lost.
1 Synchronization achieved.
CGS status for Lane 6.
0 Synchronization lost.
1 Synchronization achieved.
CGS status for Lane 5.
0 Synchronization lost.
1 Synchronization achieved.
CGS status for Lane 4.
0 Synchronization lost.
1 Synchronization achieved.
CGS status for Lane 3.
0 Synchronization lost.
1 Synchronization achieved.
CGS status for Lane 2.
0 Synchronization lost.
1 Synchronization achieved.
CGS status for Lane 1.
0 Synchronization lost.
1 Synchronization achieved.
CGS status for Lane 0.
0 Synchronization lost.
1 Synchronization achieved.
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
R
R
R
R
R
R
R
R
R
0x471
FRAME_SYNC
Frame sync status for Lane 7
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
6
5
4
FS6
FS5
FS4
Frame sync status for Lane 6
(ignore this output when
NO_ILAS = 1).
0x0
0x0
0x0
R
R
R
0 Synchronization lost.
1 Synchronization achieved.
Frame sync status for Lane 5
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
Frame sync status for Lane 4
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
Rev. 0 | Page 112 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
3
2
1
0
7
6
5
4
3
2
1
FS3
Frame sync status for Lane 3
(ignore this output when
NO_ILAS = 1).
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
R
R
R
R
R
R
R
R
R
R
R
0 Synchronization lost.
1 Synchronization achieved.
FS2
Frame sync status for Lane 2
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
FS1
Frame sync status for Lane 1
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
FS0
Frame sync status for Lane 0
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
0x472
GOOD_CHECKSUM
CKS7
CKS6
CKS5
CKS4
CKS3
CKS2
CKS1
Computed checksum status
for Lane 7 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Computed checksum status
for Lane 6 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Computed checksum status
for Lane 5 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Computed checksum status
for Lane 4 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Computed checksum status
for Lane 3 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Computed checksum status
for Lane 2 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Computed checksum status
for Lane 1 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Rev. 0 | Page 113 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0
7
6
5
4
3
2
1
0
7
CKS0
ILS7
ILS6
ILS5
ILS4
ILS3
ILS2
ILS1
ILS0
RX_DIS
Computed checksum status
for Lane 0 (ignore this output
when NO_ILAS = 1).
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
R
0 Checksum is incorrect.
1 Checksum is correct.
0x473
INIT_LANE_SYNC
Initial lane synchronization
status for Lane 7 (ignore this
output when NO_ILAS = 1).
R
0 Synchronization lost.
1 Synchronization achieved.
Initial lane synchronization
status for Lane 6 (ignore this
output when NO_ILAS = 1).
R
0 Synchronization lost.
1 Synchronization achieved.
Initial lane synchronization
status for Lane 5 (ignore this
output when NO_ILAS = 1).
R
0 Synchronization lost.
1 Synchronization achieved.
Initial lane synchronization
status for Lane 4 (ignore this
output when NO_ILAS = 1).
R
0 Synchronization lost.
1 Synchronization achieved.
Initial lane synchronization
status for Lane 3 (ignore this
output when NO_ILAS = 1).
R
0 Synchronization lost.
1 Synchronization achieved.
Initial lane synchronization
status for Lane 2 (ignore this
output when NO_ILAS = 1).
R
0 Synchronization lost.
1 Synchronization achieved.
Initial lane synchronization
status for Lane 1 (ignore this
output when NO_ILAS = 1).
R
0 Synchronization lost.
1 Synchronization achieved.
Initial lane synchronization
status for Lane 0 (ignore this
output when NO_ILAS = 1).
R
0 Synchronization lost.
1 Synchronization achieved.
0x475
CTRLREG0
Level input: disable deframer 0x0
receiver when this input = 1.
This signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
R/W
1 Disable character replacement
of /A/ and /F/ control characters
at the end of received frames
and multiframes.
0 Enables the substitution.
Rev. 0 | Page 114 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
6
CHAR_REPL_DIS
When this input = 1, character 0x0
replacement at the end of
R/W
frame/multiframe is disabled.
This signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
[5:4] Reserved
Reserved.
0x0
0x0
R
3
SOFTRST
Soft reset. Active high
R/W
synchronous reset. Resets all
hardware to power-on state.
1 Disables the deframer
reception.
0 Enable deframer logic.
2
FORCESYNCREQ
Command from application to 0x0
assert a sync request
R/W
SYNCOUT±
(
). Active high.
1
0
Reserved
Reserved.
0x0
R
REPL_FRM_ENA
When this level input is set, it 0x1
enables replacement of frames
received in error. This signal
must only be programmed
while the QBD is held in soft
reset (Register 0x475, Bit 3),
and must not be changed
during normal operation.
R/W
0x476
CTRLREG1
[7:5] Reserved
4 QUAL_RDERR
Reserved.
0x0
0x1
R
Error reporting behavior for
concurrent NIT and running
disparity (RD) errors. This signal
must only be programmed
while the QBD is held in soft
reset (Register 0x475, Bit 3),
and must not be changed
during normal operation.
R/W
0 NIT has no effect on RD error.
1 NIT error masks concurrent RD
error.
3
DEL_SCR
Alternative descrambler enable. 0x0
(see JESD204B Section 5.2.4)
This signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
R/W
1 Descrambling begins at
Octet 2 of user data.
0 Descrambling begins at
Octet 0 of user data. This is the
common usage.
2
CGS_SEL
Determines the QBD behavior 0x1
after code group sync has
been achieved. This signal
must only be programmed
while the QBD is held in soft
reset (Register 0x475, Bit 3),
and must not be changed
during normal operation.
R/W
Rev. 0 | Page 115 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0 After code group sync is
achieved, the QBD asserts
SYNCOUT±
only if there are
sufficient disparity errors as per
the JESD204B standard.
1 After code group sync is
achieved, if a /K/ is followed by
any character other than an /R/
or another /K/, QBD asserts
SYNCOUT±
.
1
NO_ILAS
This signal must only be pro- 0x0
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be
R/W
changed during normal
operation.
1 For single-lane operation, ILAS
is omitted. Code group sync is
followed by user data.
0 Code group sync is followed
by ILAS. For multilane
operation, NO_ILAS must
always be set to 0.
0
FCHK_N
Checksum calculation
0x0
R/W
method. This signal must only
be programmed while the
QBD is held in soft reset
(Register 0x475, Register 3),
and must not be changed
during normal operation.
0 Calculate checksum by
summing individual fields
(this more closely matches the
definition of the checksum
field in the JESD204B
standard.
1 Calculate checksum by
summing the registers
containing the packed fields
(this setting is provided in
case the framer of another
vendor performs the
calculation with this method).
0x477
CTRLREG2
7
ILS_MODE
Data link layer test mode. This 0x0
signal must only be pro-
R/W
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
0 Normal mode.
1 Code group sync pattern is
followed by a perpetual ILAS
sequence.
6
5
Reserved
Reserved.
0x0
0x0
R
REPDATATEST
Repetitive data test enable,
using JTSPAT pattern. To
enable the test, ILS_MODE
must = 0. This signal must
only be programmed while
the QBD is held in soft reset
(Register 0x475, Bit 3), and
must not be changed during
normal operation.
R/W
Rev. 0 | Page 116 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
4
QUETESTERR
Queue test error mode. This
signal must only be pro-
0x0
R/W
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
0 Simultaneous errors on
multiple lanes are reported as
one error.
1 Detected errors from all lanes
are trapped in a counter and
sequentially signaled on
SYNCOUT±
.
3
AR_ECNTR
Automatic reset of error
counter. The error counter
that causes assertion of
0x0
R/W
SYNCOUT±
is automatically
reset to 0 when AR_ECNTR = 1.
All other counters are
unaffected. This signal must
only be programmed while
the QBD is held in soft reset
(Register 0x475, Bit 3), and
must not be changed during
normal operation.
[2:0] Reserved
[7:0] KSYNC
Reserved.
0x0
R
0x478
KVAL
Number of 4 × K multiframes 0x1
during ILS. F is the number of
octets per frame. Settings of 1,
2, and 4 are valid. Refer to
R/W
Table 15 and Table 16. This
signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
0x47C
ERRORTHRES
[7:0] ETH
Error threshold value. Bad
0xFF R/W
disparity, NIT disparity, and
unexpected K character errors
are counted and compared to
the error threshold value. When
the count is equal, either an
SYNCOUT±
IRQ is generated or
is asserted per the mask register
settings or both. Function is
performed in all lanes. This
signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be
changed during normal
operation.
Rev. 0 | Page 117 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x47D
SYNC_ASSERT_MASK
[7:3] Reserved
Reserved.
0x0
0x7
R
[2:0] SYNC_ASSERT_MASK
SYNCOUT
R/W
assertion enable
mask for BD, NIT, and UEK
error conditions. Active high,
SYNCOUT
assertion enable
mask for BD, NIT, and UEK
error conditions, respectively.
When an error counter, in any
lane, has reached the error
threshold count, ETH[7:0], and
the corresponding SYNC_
ASSERT_MASK bit is set,
SYNCOUT
is asserted. The
mask bits are as follows. Note
that the bit sequence is
reversed with respect to the
other error count controls and
the error counters.
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
0x480
ECNT_CTRL0
[7:6] Reserved
Reserved.
0x0
0x7
R
[5:3] ECNT_ENA0
Error counter enable for
Lane 0. Counters of each lane
are addressed as follows:
R/W
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
[2:0] ECNT_RST0
Error counters enable for
Lane 0, active high. Counters
of each lane are addressed as
follows:
0x7
R/W
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
0x481
ECNT_CTRL1
[7:6] Reserved
Reserved.
0x0
0x7
R
[5:3] ECNT_ENA1
Error counters enable for
Lane 1, active high. Counters
of each lane are addressed as
follows:
R/W
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
[2:0] ECNT_RST1
Error counters enable for
Lane 1, active high. Counters
of each lane are addressed as
follows:
0x7
R/W
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
0x482
ECNT_CTRL2
[7:6] Reserved
Reserved.
0x0
0x7
R
[5:3] ECNT_ENA2
Error counters enable for
Lane 2, active high. Counters
of each lane are addressed as
follows:
R/W
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
Rev. 0 | Page 118 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
[2:0] ECNT_RST2
Error counters enable for
Lane 2, active high. Counters
of each lane are addressed as
follows:
0x7
R/W
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
0x483
ECNT_CTRL3
ECNT_CTRL4
ECNT_CTRL5
[7:6] Reserved
Reserved.
0x0
0x7
R
[5:3] ECNT_ENA3
Error counters enable for
Lane 3, active high. Counters
of each lane are addressed as
follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
R/W
Bit 0 = BDE.
[2:0] ECNT_RST3
Error counters enable for
Lane 3, active high. Counters
of each lane are addressed as
follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
0x7
R/W
Bit 0 = BDE.
0x484
[7:6] Reserved
Reserved.
0x0
0x7
R
[5:3] ECNT_ENA4
Error counters enable for
Lane 4, active high. Counters
of each lane are addressed as
follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
R/W
Bit 0 = BDE.
[2:0] ECNT_RST4
Error counters enable for
Lane 4, active high. Counters
of each lane are addressed as
follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
0x7
R/W
Bit 0 = BDE.
0x485
[7:6] Reserved
Reserved.
0x0
0x7
R
[5:3] ECNT_ENA5
Error counters enable for
Lane 5, active high. Counters
of each lane are addressed as
follows:
R/W
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
[2:0] ECNT_RST5
Error counters enable for
Lane 5, active high. Counters
of each lane are addressed as
follows:
0x7
R/W
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
Rev. 0 | Page 119 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x486
ECNT_CTRL6
ECNT_CTRL7
ECNT_TCH0
[7:6] Reserved
Reserved.
0x0
R
[5:3] ECNT_ENA6
Error counters enable for Lane 0x7
6, active high. Counters of
each lane are addressed as
follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
R/W
Bit 0 = BDE.
[2:0] ECNT_RST6
Error counters enable for Lane 0x7
6, active high. Counters of
each lane are addressed as
follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
R/W
Bit 0 = BDE.
0x487
[7:6] Reserved
Reserved.
0x0
0x7
R
[5:3] ECNT_ENA7
Error counters enable for
Lane 7, active high. Counters
of each lane are addressed as
follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
R/W
Bit 0 = BDE.
[2:0] ECNT_RST7
Reset error counters for
Lane 7, active high. Counters
of each lane are addressed as
follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
0x7
0x0
R/W
Bit 0 = BDE.
0x488
[7:3] Reserved
Reserved.
R
[2:0] ECNT_TCH0
Terminal count hold enable of 0x7
error counters for Lane 0.
R/W
When set, the designated
counter is to hold the terminal
count value of 0xFF when it is
reached until the counter is
reset by the user. Otherwise,
the designated counter rolls
over. Counters of each lane
are addressed as follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
This signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be
changed during normal
operation.
0x489
ECNT_TCH1
[7:3] Reserved
Reserved.
0x0
R
Rev. 0 | Page 120 of 138
Data Sheet
AD9166
Hex.
Bit
Addr.
0x48A
0x48B
Register Name
No. Bit Name
Settings
Description
Reset Access
[2:0] ECNT_TCH1
Terminal count hold enable of
error counters for Lane 1. When
set, the designated counter is to
hold the terminal count value of
0xFF when it is reached until the
counter is reset by the user.
Otherwise, the designated
counter rolls over. Counters of
each lane are addressed as
follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
This signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
0x7
R/W
ECNT_TCH2
[7:3] Reserved
Reserved.
0x0
0x7
R
[2:0] ECNT_TCH2
Terminal count hold enable of
error counters for Lane 2. When
set, the designated counter is to
hold the terminal count value of
0xFF when it is reached until the
counter is reset by the user.
Otherwise, the designated
counter rolls over. Counters of
each lane are addressed as
follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
This signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
R/W
ECNT_TCH3
[7:3] Reserved
Reserved.
0x0
R
[2:0] ECNT_TCH3
Terminal count hold enable of 0x7
error counters for Lane 3.
R/W
When set, the designated
counter is to hold the terminal
count value of 0xFF when it is
reached until the counter is
reset by the user. Otherwise,
the designated counter rolls
over. Counters of each lane are
addressed as follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
This signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be
changed during normal
operation.
Rev. 0 | Page 121 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x48C
ECNT_TCH4
ECNT_TCH5
ECNT_TCH6
[7:3] Reserved
Reserved.
0x0
R
[2:0] ECNT_TCH4
Terminal count hold enable of 0x7
error counters for Lane 4.
R/W
When set, the designated
counter is to hold the terminal
count value of 0xFF when it is
reached until the counter is
reset by the user. Otherwise,
the designated counter rolls
over. Counters of each lane are
addressed as follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
This signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be
changed during normal
operation.
0x48D
[7:3] Reserved
Reserved.
0x0
R
[2:0] ECNT_TCH5
Terminal count hold enable of 0x7
error counters for Lane 5.
R/W
When set, the designated
counter is to hold the terminal
count value of 0xFF when it is
reached until the counter is
reset by the user. Otherwise,
the designated counter rolls
over. Counters of each lane are
addressed as follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
This signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be
changed during normal
operation.
0x48E
[7:3] Reserved
Reserved.
0x0
R
[2:0] ECNT_TCH6
Terminal count hold enable of 0x7
error counters for Lane 6.
R/W
When set, the designated
counter is to hold the terminal
count value of 0xFF when it is
reached until the counter is
reset by the user. Otherwise,
the designated counter rolls
over. Counters of each lane
are addressed as follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
This signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
Rev. 0 | Page 122 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x48F
ECNT_TCH7
[7:3] Reserved
Reserved.
0x0
R
[2:0] ECNT_TCH7
Terminal count hold enable of 0x7
error counters for Lane 7.
R/W
When set, the designated
counter is to hold the terminal
count value of 0xFF when it is
reached until the counter is
reset by the user. Otherwise,
the designated counter rolls
over. Counters of each lane
are addressed as follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
This signal must only be pro-
grammed while the QBD is held
in soft reset (Register 0x475,
Bit 3), and must not be changed
during normal operation.
0x490
ECNT_STAT0
[7:4] Reserved
Reserved.
0x0
R
R
3
LANE_ENA0
This output indicates if Lane 0 0x0
is enabled.
0 Lane 0 is held in soft reset.
1 Lane 0 is enabled.
[2:0] ECNT_TCR0
Terminal count reached
0x0
R
indicator of error counters for
Lane 0. Set to 1 when the
corresponding counter terminal
count value of 0xFF has been
reached. Counters of each
lane are addressed as follows.
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
0x491
ECNT_STAT1
[7:4] Reserved
LANE_ENA1
Reserved.
0x0
R
R
3
This output indicates if Lane 1 0x0
is enabled.
0 Lane 1 is held in soft reset.
1 Lane 1 is enabled.
[2:0] ECNT_TCR1
Terminal count reached
indicator of error counters for
Lane 1. Set to 1 when the
corresponding counter
terminal count value of 0xFF
has been reached. Counters of
each lane are addressed as
follows.
0x0
R
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
0x492
ECNT_STAT2
[7:4] Reserved
LANE_ENA2
Reserved.
0x0
R
R
3
This output indicates if Lane 2 0x0
is enabled.
0 Lane 2 is held in soft reset.
1 Lane 2 is enabled.
Rev. 0 | Page 123 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
[2:0] ECNT_TCR2
Terminal count reached
indicator of error counters for
Lane 2. Set to 1 when the
corresponding counter
terminal count value of 0xFF
has been reached. Counters of
each lane are addressed as
follows.
0x0
R
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
0x493
ECNT_STAT3
ECNT_STAT4
ECNT_STAT5
[7:4] Reserved
LANE_ENA3
Reserved.
0x0
R
R
3
This output indicates if Lane 3 0x0
is enabled.
0 Lane 3 is held in soft reset.
1 Lane 3 is enabled.
[2:0] ECNT_TCR3
Terminal count reached
indicator of error counters for
Lane 3. Set to 1 when the
corresponding counter
terminal count value of 0xFF
has been reached. Counters of
each lane are addressed as
follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
0x0
R
0x494
[7:4] Reserved
LANE_ENA4
Reserved.
0x0
R
R
3
This output indicates if Lane 4 0x0
is enabled.
0 Lane 4 is held in soft reset.
1 Lane 4 is enabled.
[2:0] ECNT_TCR4
Terminal count reached
indicator of error counters for
Lane 4. Set to 1 when the
corresponding counter
terminal count value of 0xFF
has been reached. Counters of
each lane are addressed as
follows:
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
0x0
R
0x495
[7:4] Reserved
LANE_ENA5
Reserved.
0x0
R
R
3
This output indicates if Lane 5 0x0
is enabled.
0 Lane 5 is held in soft reset.
1 Lane 5 is enabled.
[2:0] ECNT_TCR5
Terminal count reached
indicator of error counters for
Lane 5. Set to 1 when the
corresponding counter
terminal count value of 0xFF
has been reached. Counters of
each lane are addressed as
follows:
0x0
R
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
Rev. 0 | Page 124 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
ECNT_STAT6
No. Bit Name
Settings
Description
Reset Access
0x496
[7:4] Reserved
Reserved.
0x0
R
R
3
LANE_ENA6
This output indicates if Lane 6 0x0
is enabled.
0 Lane 6 is held in soft reset.
1 Lane 6 is enabled.
[2:0] ECNT_TCR6
Terminal count reached
indicator of error counters for
Lane 6. Set to 1 when the
corresponding counter
terminal count value of 0xFF
has been reached. Counters of
each lane are addressed as
follows:
0x0
R
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
0x497
ECNT_STAT7
[7:4] Reserved
LANE_ENA7
Reserved.
0x0
R
R
3
This output indicates if Lane 7 0x0
is enabled.
0 Lane 7 is held in soft reset.
1 Lane 7 is enabled.
[2:0] ECNT_TCR7
Terminal count reached
indicator of error counters for
Lane 7. Set to 1 when the
corresponding counter
terminal count value of 0xFF
has been reached. Counters of
each lane are addressed as
follows:
0x0
R
Bit 2 = UEK character error.
Bit 1 = NIT.
Bit 0 = BDE.
0x498
0x499
0x49A
0x49B
0x49C
0x49D
0x49E
0x49F
0x4A0
0x4A1
0x4A2
0x4A3
0x4A4
0x4A5
BD_CNT0
BD_CNT1
BD_CNT2
BD_CNT3
BD_CNT4
BD_CNT5
BD_CNT6
BD_CNT7
NIT_CNT0
NIT_CNT1
NIT_CNT2
NIT_CNT3
NIT_CNT4
NIT_CNT5
[7:0] BD_CNT0
[7:0] BD_CNT1
[7:0] BD_CNT2
[7:0] BD_CNT3
[7:0] BD_CNT4
[7:0] BD_CNT5
[7:0] BD_CNT6
[7:0] BD_CNT7
[7:0] NIT_CNT0
[7:0] NIT_CNT1
[7:0] NIT_CNT2
[7:0] NIT_CNT3
[7:0] NIT_CNT4
[7:0] NIT_CNT5
Bad disparity 8-bit error
counters for Lane 0.
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
R
R
R
R
R
R
R
R
R
R
R
R
R
R
Bad disparity 8-bit error
counters for Lane 1.
Bad disparity 8-bit error
counters for Lane 2.
Bad disparity 8-bit error
counters for Lane 3.
Bad disparity 8-bit error
counters for Lane 4.
Bad disparity 8-bit error
counters for Lane 5.
Bad disparity 8-bit error
counters for Lane 6.
Bad disparity 8-bit error
counters for Lane 7.
Bad disparity 8-bit error
counters for Lane 0.
NIT 8-bit error counters for
Lane 1.
NIT 8-bit error counters for
Lane 2.
NIT 8-bit error counters for
Lane 3.
NIT 8-bit error counters for
Lane 4.
NIT 8-bit error counters for
Lane 5.
Rev. 0 | Page 125 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x4A6
NIT_CNT6
NIT_CNT7
UEK_CNT0
UEK_CNT1
UEK_CNT2
UEK_CNT3
UEK_CNT4
[7:0] NIT_CNT6
[7:0] NIT_CNT7
[7:0] UEK_CNT0
[7:0] UEK_CNT1
[7:0] UEK_CNT2
[7:0] UEK_CNT3
[7:0] UEK_CNT4
[7:0] UEK_CNT5
[7:0] UEK_CNT6
[7:0] UEK_CNT7
NIT 8-bit error counters for
Lane 6.
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
R
R
R
R
R
R
R
R
R
R
R
0x4A7
0x4A8
0x4A9
0x4AA
0x4AB
0x4AC
NIT 8-bit error counters for
Lane 7.
UEK character 8-bit error
counters for Lane 0.
UEK character 8-bit error
counters for Lane 1.
UEK character 8-bit error
counters for Lane 2.
UEK character 8-bit error
counters for Lane 3.
UEK character 8-bit error
counters for Lane 4.
0x4AD UEK_CNT5
UEK character 8-bit error
counters for Lane 5.
0x4AE
0x4AF
0x4B0
UEK_CNT6
UEK character 8-bit error
counters for Lane 6.
UEK_CNT7
UEK character 8-bit error
counters for Lane 7.
LINK_STATUS0
7
6
5
BDE0
BDE status for Lane 0.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT errors status for Lane 0.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT0
0x0
R
R
UEK0
UEK character errors status for 0x0
Lane 0.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
4
3
2
1
ILD0
ILS0
CKS0
FS0
Interlane deskew status for
Lane 0 (ignore this output
when NO_ILAS = 1).
0x0
0x0
0x0
0x0
R
R
R
R
0 Deskew failed.
1 Deskew achieved.
Initial lane synchronization
status for Lane 0 (ignore this
output when NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
Computed checksum status
for Lane 0 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Frame sync status for Lane 0
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
0
7
CGS0
BDE1
Code group sync status for
Lane 0.
0 Synchronization lost.
1 Synchronization achieved.
BDE status for Lane 1.
0x0
0x0
R
R
0x4B1
LINK_STATUS1
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
Rev. 0 | Page 126 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
6
NIT1
NIT errors status for Lane 1.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
0x0
R
5
UEK1
UEK character errors status for 0x0
Lane 1.
R
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
4
3
2
1
0
ILD1
ILS1
CKS1
FS1
Interlane deskew status for
Lane 1 (ignore this output
when NO_ILAS = 1).
0x0
0x0
0x0
0x0
0x0
R
R
R
R
R
0 Deskew failed.
1 Deskew achieved.
Initial lane synchronization
status for Lane 1 (ignore this
output when NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
Computed checksum status
for Lane 1 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Frame sync status for Lane 1
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
CGS1
Code group sync status for
Lane 1.
0 Synchronization lost.
1 Synchronization achieved.
BDE status for Lane 2.
0x4B2
LINK_STATUS2
7
6
5
BDE2
NIT2
0x0
0x0
R
R
R
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT errors status for Lane 2.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
UEK2
UEK character errors status for 0x0
Lane 2.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
4
3
2
ILD2
ILS2
Interlane deskew status for
Lane 2 (ignore this output
when NO_ILAS = 1).
0x0
0x0
0x0
R
R
R
0 Deskew failed.
1 Deskew achieved.
Initial lane synchronization
status for Lane 2 (ignore this
output when NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
CKS2
Computed checksum status
for Lane 2 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Rev. 0 | Page 127 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
1
FS2
Frame sync status for Lane 2
(ignore this output when
NO_ILAS = 1).
0x0
R
0 Synchronization lost.
1 Synchronization achieved.
0
CGS2
Code group sync status for
Lane 2.
0x0
R
0 Synchronization lost.
1 Synchronization achieved.
BDE status for Lane 3.
0x4B3
LINK_STATUS3
7
6
5
BDE3
NIT3
0x0
0x0
R
R
R
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT errors status for Lane 3.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
UEK3
UEK character errors status for 0x0
Lane 3.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
4
3
2
1
0
ILD3
ILS3
CKS3
FS3
Interlane deskew status for
Lane 3 (ignore this output
when NO_ILAS = 1).
0x0
0x0
0x0
0x0
0x0
R
R
R
R
R
0 Deskew failed.
1 Deskew achieved.
Initial lane synchronization
status for Lane 3 (ignore this
output when NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
Computed checksum status
for Lane 3 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Frame sync status for Lane 3
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
CGS3
Code group sync status for
Lane 3.
0 Synchronization lost.
1 Synchronization achieved.
BDE status for Lane 4.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
0x4B4
LINK_STATUS4
7
6
BDE4
NIT4
0x0
0x0
R
R
Not in table errors status for
Lane 4.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
5
UEK4
UEK character errors status for 0x0
Lane 4.
R
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
Rev. 0 | Page 128 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
4
3
2
1
0
ILD4
ILS4
CKS4
FS4
Interlane deskew status for
Lane 4 (ignore this output
when NO_ILAS = 1).
0x0
0x0
0x0
0x0
0x0
R
R
R
R
R
0 Deskew failed.
1 Deskew achieved.
Initial lane synchronization
status for Lane 4 (ignore this
output when NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
Computed checksum status
for Lane 4 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Frame sync status for Lane 4
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
CGS4
Code group sync status for
Lane 4.
0 Synchronization lost.
1 Synchronization achieved.
BDE status for Lane 5.
0x4B5
LINK_STATUS5
7
6
5
BDE5
NIT5
0x0
0x0
R
R
R
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT errors status for Lane 5.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
UEK5
UEK character errors status for 0x0
Lane 5.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
4
3
2
1
0
ILD5
ILS5
CKS5
FS5
Interlane deskew status for
Lane 5 (ignore this output
when NO_ILAS = 1).
0x0
0x0
0x0
0x0
0x0
R
R
R
R
R
0 Deskew failed.
1 Deskew achieved.
Initial lane synchronization
status for Lane 5 (ignore this
output when NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
Computed checksum status
for Lane 5 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Frame sync status for Lane 5
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
CGS5
Code group sync status for
Lane 5.
0 Synchronization lost.
1 Synchronization achieved.
Rev. 0 | Page 129 of 138
AD9166
Data Sheet
Hex.
Addr.
0x4B6
Bit
Register Name
LINK_STATUS6
No. Bit Name
Settings
Description
BDE status for Lane 6.
Reset Access
7
6
5
BDE6
0x0
R
R
R
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT errors status for Lane 6.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT6
0x0
UEK6
UEK character errors status for 0x0
Lane 6.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
4
3
2
1
0
ILD6
ILS6
Interlane deskew status for
Lane 6 (ignore this output
when NO_ILAS = 1).
0x0
0x0
0x0
0x0
0x0
R
R
R
R
R
0 Deskew failed.
1 Deskew achieved.
Initial lane synchronization
status for Lane 6 (ignore this
output when NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
CKS6
FS6
Computed checksum status
for Lane 6 (ignore this output
when NO_ILAS = 1).
0 Checksum is incorrect.
1 Checksum is correct.
Frame sync status for Lane 6
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
CGS6
Code group sync status for
Lane 6.
0 Synchronization lost.
1 Synchronization achieved.
BDE status for Lane 7.
0x4B7
LINK_STATUS7
7
6
5
BDE7
NIT7
0x0
0x0
R
R
R
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
NIT errors status for Lane 7.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
UEK7
UEK character errors status for 0x0
Lane 7.
0 Error count < ETH[7:0] value.
1 Error count ≥ ETH[7:0] value.
4
3
ILD7
ILS7
Interlane deskew status for
Lane 7 (ignore this output
when NO_ILAS = 1).
0x0
0x0
R
R
0 Deskew failed.
1 Deskew achieved.
Initial lane synchronization
status for Lane 7 (ignore this
output when NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
Rev. 0 | Page 130 of 138
Data Sheet
AD9166
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
2
1
0
CKS7
Computed checksum status
for Lane 7 (ignore this output
when NO_ILAS = 1).
0x0
0x0
0x0
R
R
R
0 Checksum is incorrect.
1 Checksum is correct.
FS7
Frame sync status for Lane 7
(ignore this output when
NO_ILAS = 1).
0 Synchronization lost.
1 Synchronization achieved.
CGS7
Code group sync status for
Lane 7.
0 Synchronization lost.
1 Synchronization achieved.
BDE counter.
0x4B8
JESD_IRQ_ENABLEA
7
6
5
4
3
2
EN_BDE
EN_NIT
EN_UEK
EN_ILD
EN_ILS
0x0
0x0
0x0
0x0
0x0
0x0
R/W
R/W
R/W
R/W
R/W
R/W
NIT error counter.
UEK error counter.
Interlane deskew.
Initial lane sync.
EN_CKS
Good checksum. This is an
interrupt that compares two
checksums: the checksum that
the transmitter sent over the
link during the ILAS, and the
checksum that the receiver
calculated from the ILAS data
that the transmitter sent over
the link. The checksum IRQ
never at any time looks at the
checksum that is programmed
over the SPI into Register 0x45D.
The checksum IRQ only looks
at the data sent by the
transmitter, and never looks at
any data programmed via the
SPI.
1
0
EN_FS
Frame sync.
Code group sync.
Reserved.
0x0
0x0
0x0
0x0
R/W
R/W
R
EN_CGS
0x4B9
JESD_IRQ_ENABLEB
[7:1] Reserved
0
EN_ILAS
Configuration mismatch
(checked for Lane 0 only). The
ILAS IRQ compares the two
sets of ILAS data that the
R/W
receiver has: the ILAS data sent
over the JESD204B link by the
transmitter, and the ILAS data
programmed into the receiver
via the SPI (Register 0x450 to
Register 0x45D). If the data
differs, the IRQ is triggered.
Note that all of the ILAS data
(including the checksum) is
compared.
0x4BA
JESD_IRQ_STATUSA
7
6
5
4
3
2
1
0
IRQ_BDE
IRQ_NIT
IRQ_UEK
IRQ_ILD
IRQ_ILS
IRQ_CKS
IRQ_FS
BDE counter.
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
NIT error counter.
UEK error counter.
Interlane deskew.
Initial lane sync.
Good checksum.
Frame sync.
IRQ_CGS
Code group sync.
Rev. 0 | Page 131 of 138
AD9166
Data Sheet
Hex.
Addr.
Bit
Register Name
No. Bit Name
Settings
Description
Reset Access
0x4BB
JESD_IRQ_STATUSB
[7:1] Reserved
Reserved.
0x0
0x0
R
0
IRQ_ILAS
Configuration mismatch
(checked for Lane 0 only).
R/W
0x800
HOPF_CTRL
[7:6] HOPF_MODE
Frequency switch (hop) mode. 0x0
Defines the phase relation
when a new frequency is
hopped to.
R/W
00 Phase continuous switch.
Changes frequency tuning
word, and the phase
accumulator continues to
accumulate to the new FTW.
01 Phase discontinuous switch.
Changes the frequency tuning
word and resets the phase
accumulator.
10 Phase coherent switch. FTW is
selected from one of the
32 hopping FTWs. Frequency
changes are phase
discontinuous from one
frequency to another but
changes back to a previous
frequency and retains the
phase accumulation of the
previous frequency.
5
Reserved
Reserved.
0x0
R
[4:0] HOPF_SEL
Hopping frequency selection 0x0
control. Enter the number of
the FTW to select the output
of that NCO.
R/W
0x806
0x807
0x808
0x809
0x80A
0x80B
0x80C
0x80D
0x80E
0x80F
0x810
0x811
0x812
0x813
0x814
0x815
0x816
0x817
0x818
0x819
0x81A
0x81B
0x81C
0x81D
0x81E
0x81F
0x820
HOPF_FTW1_0
HOPF_FTW1_1
HOPF_FTW1_2
HOPF_FTW1_3
HOPF_FTW2_0
HOPF_FTW2_1
HOPF_FTW2_2
HOPF_FTW2_3
HOPF_FTW3_0
HOPF_FTW3_1
HOPF_FTW3_2
HOPF_FTW3_3
HOPF_FTW4_0
HOPF_FTW4_1
HOPF_FTW4_2
HOPF_FTW4_3
HOPF_FTW5_0
HOPF_FTW5_1
HOPF_FTW5_2
HOPF_FTW5_3
HOPF_FTW6_0
HOPF_FTW6_1
HOPF_FTW6_2
HOPF_FTW6_3
HOPF_FTW7_0
HOPF_FTW7_1
HOPF_FTW7_2
[7:0] HOPF_FTW1[7:0]
[7:0] HOPF_FTW1[15:8]
[7:0] HOPF_FTW1[23:16]
[7:0] HOPF_FTW1[31:24]
[7:0] HOPF_FTW2[7:0]
[7:0] HOPF_FTW2[15:8]
[7:0] HOPF_FTW2[23:16]
[7:0] HOPF_FTW2[31:24]
[7:0] HOPF_FTW3[7:0]
[7:0] HOPF_FTW3[15:8]
[7:0] HOPF_FTW3[23:16]
[7:0] HOPF_FTW3[31:24]
[7:0] HOPF_FTW4[7:0]
[7:0] HOPF_FTW4[15:8]
[7:0] HOPF_FTW4[23:16]
[7:0] HOPF_FTW4[31:24]
[7:0] HOPF_FTW5[7:0]
[7:0] HOPF_FTW5[15:8]
[7:0] HOPF_FTW5[23:16]
[7:0] HOPF_FTW5[31:24]
[7:0] HOPF_FTW6[7:0]
[7:0] HOPF_FTW6[15:8]
[7:0] HOPF_FTW6[23:16]
[7:0] HOPF_FTW6[31:24]
[7:0] HOPF_FTW7[7:0]
[7:0] HOPF_FTW7[15:8]
[7:0] HOPF_FTW7[23:16]
Hopping frequency FTW1.
Hopping frequency FTW1.
Hopping frequency FTW1
Hopping frequency FTW1
Hopping frequency FTW2
Hopping frequency FTW2
Hopping frequency FTW2
Hopping frequency FTW2
Hopping frequency FTW3
Hopping frequency FTW3
Hopping frequency FTW3
Hopping frequency FTW3
Hopping frequency FTW4
Hopping frequency FTW4
Hopping frequency FTW4
Hopping frequency FTW4
Hopping frequency FTW5
Hopping frequency FTW5
Hopping frequency FTW5
Hopping frequency FTW5
Hopping frequency FTW6
Hopping frequency FTW6
Hopping frequency FTW6
Hopping frequency FTW6
Hopping frequency FTW7
Hopping frequency FTW7
Hopping frequency FTW7
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Rev. 0 | Page 132 of 138
Data Sheet
AD9166
Hex.
Bit
Addr.
0x821
0x822
0x823
0x824
0x825
0x826
0x827
0x828
0x829
0x82A
0x82B
0x82C
0x82D
0x82E
0x82F
0x830
0x831
0x832
0x833
0x834
0x835
0x836
0x837
0x838
0x839
0x83A
0x83B
0x83C
0x83D
0x83E
0x83F
0x840
0x841
0x842
0x843
0x844
0x845
0x846
0x847
0x848
0x849
0x84A
0x84B
0x84C
0x84D
0x84E
0x84F
0x850
0x851
0x852
0x853
0x854
0x855
Register Name
No. Bit Name
Settings
Description
Reset Access
HOPF_FTW7_3
HOPF_FTW8_0
HOPF_FTW8_1
HOPF_FTW8_2
HOPF_FTW8_3
HOPF_FTW9_0
HOPF_FTW9_1
HOPF_FTW9_2
HOPF_FTW9_3
HOPF_FTW10_0
HOPF_FTW10_1
HOPF_FTW10_2
HOPF_FTW10_3
HOPF_FTW11_0
HOPF_FTW11_1
HOPF_FTW11_2
HOPF_FTW11_3
HOPF_FTW12_0
HOPF_FTW12_1
HOPF_FTW12_2
HOPF_FTW12_3
HOPF_FTW13_0
HOPF_FTW13_1
HOPF_FTW13_2
HOPF_FTW13_3
HOPF_FTW14_0
HOPF_FTW14_1
HOPF_FTW14_2
HOPF_FTW14_3
HOPF_FTW15_0
HOPF_FTW15_1
HOPF_FTW15_2
HOPF_FTW15_3
HOPF_FTW16_0
HOPF_FTW16_1
HOPF_FTW16_2
HOPF_FTW16_3
HOPF_FTW17_0
HOPF_FTW17_1
HOPF_FTW17_2
HOPF_FTW17_3
HOPF_FTW18_0
HOPF_FTW18_1
HOPF_FTW18_2
HOPF_FTW18_3
HOPF_FTW19_0
HOPF_FTW19_1
HOPF_FTW19_2
HOPF_FTW19_3
HOPF_FTW20_0
HOPF_FTW20_1
HOPF_FTW20_2
HOPF_FTW20_3
[7:0] HOPF_FTW7[31:24]
[7:0] HOPF_FTW8[7:0]
Hopping frequency FTW7
Hopping frequency FTW8
Hopping frequency FTW8
Hopping frequency FTW8
Hopping frequency FTW8
Hopping frequency FTW9
Hopping frequency FTW9
Hopping frequency FTW9
Hopping frequency FTW9
Hopping frequency FTW10
Hopping frequency FTW10
Hopping frequency FTW10
Hopping frequency FTW10
Hopping frequency FTW11
Hopping frequency FTW11
Hopping frequency FTW11
Hopping frequency FTW11
Hopping frequency FTW12
Hopping frequency FTW12
Hopping frequency FTW12
Hopping frequency FTW12
Hopping frequency FTW13
Hopping frequency FTW13
Hopping frequency FTW13
Hopping frequency FTW13
Hopping frequency FTW14
Hopping frequency FTW14
Hopping frequency FTW14
Hopping frequency FTW14
Hopping frequency FTW15
Hopping frequency FTW15
Hopping frequency FTW15
Hopping frequency FTW15
Hopping frequency FTW16
Hopping frequency FTW16
Hopping frequency FTW16
Hopping frequency FTW16
Hopping frequency FTW17
Hopping frequency FTW17
Hopping frequency FTW17
Hopping frequency FTW17
Hopping frequency FTW18
Hopping frequency FTW18
Hopping frequency FTW18
Hopping frequency FTW18
Hopping frequency FTW19
Hopping frequency FTW19
Hopping frequency FTW19
Hopping frequency FTW19
Hopping frequency FTW20
Hopping frequency FTW20
Hopping frequency FTW20
Hopping frequency FTW20
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
[7:0] HOPF_FTW8[15:8]
[7:0] HOPF_FTW8[23:16]
[7:0] HOPF_FTW8[31:24]
[7:0] HOPF_FTW9[7:0]
[7:0] HOPF_FTW9[15:8]
[7:0] HOPF_FTW9[23:16]
[7:0] HOPF_FTW9[31:24]
[7:0] HOPF_FTW10[7:0]
[7:0] HOPF_FTW10[15:8]
[7:0] HOPF_FTW10[23:16]
[7:0] HOPF_FTW10[31:24]
[7:0] HOPF_FTW11[7:0]
[7:0] HOPF_FTW11[15:8]
[7:0] HOPF_FTW11[23:16]
[7:0] HOPF_FTW11[31:24]
[7:0] HOPF_FTW12[7:0]
[7:0] HOPF_FTW12[15:8]
[7:0] HOPF_FTW12[23:16]
[7:0] HOPF_FTW12[31:24]
[7:0] HOPF_FTW13[7:0]
[7:0] HOPF_FTW13[15:8]
[7:0] HOPF_FTW13[23:16]
[7:0] HOPF_FTW13[31:24]
[7:0] HOPF_FTW14[7:0]
[7:0] HOPF_FTW14[15:8]
[7:0] HOPF_FTW14[23:16]
[7:0] HOPF_FTW14[31:24]
[7:0] HOPF_FTW15[7:0]
[7:0] HOPF_FTW15[15:8]
[7:0] HOPF_FTW15[23:16]
[7:0] HOPF_FTW15[31:24]
[7:0] HOPF_FTW16[7:0]
[7:0] HOPF_FTW16[15:8]
[7:0] HOPF_FTW16[23:16]
[7:0] HOPF_FTW16[31:24]
[7:0] HOPF_FTW17[7:0]
[7:0] HOPF_FTW17[15:8]
[7:0] HOPF_FTW17[23:16]
[7:0] HOPF_FTW17[31:24]
[7:0] HOPF_FTW18[7:0]
[7:0] HOPF_FTW18[15:8]
[7:0] HOPF_FTW18[23:16]
[7:0] HOPF_FTW18[31:24]
[7:0] HOPF_FTW19[7:0]
[7:0] HOPF_FTW19[15:8]
[7:0] HOPF_FTW19[23:16]
[7:0] HOPF_FTW19[31:24]
[7:0] HOPF_FTW20[7:0]
[7:0] HOPF_FTW20[15:8]
[7:0] HOPF_FTW20[23:16]
[7:0] HOPF_FTW20[31:24]
Rev. 0 | Page 133 of 138
AD9166
Data Sheet
Hex.
Bit
Addr.
0x856
0x857
0x858
0x859
0x85A
0x85B
0x85C
0x85D
0x85E
0x85F
0x860
0x861
0x862
0x863
0x864
0x865
0x866
0x867
0x868
0x869
0x86A
0x86B
0x86C
0x86D
0x86E
0x86F
0x870
0x871
0x872
0x873
0x874
0x875
0x876
0x877
0x878
0x879
0x87A
0x87B
0x87C
0x87D
0x87E
0x87F
0x880
0x881
Register Name
No. Bit Name
Settings
Description
Reset Access
HOPF_FTW21_0
HOPF_FTW21_1
HOPF_FTW21_2
HOPF_FTW21_3
HOPF_FTW22_0
HOPF_FTW22_1
HOPF_FTW22_2
HOPF_FTW22_3
HOPF_FTW23_0
HOPF_FTW23_1
HOPF_FTW23_2
HOPF_FTW23_3
HOPF_FTW24_0
HOPF_FTW24_1
HOPF_FTW24_2
HOPF_FTW24_3
HOPF_FTW25_0
HOPF_FTW25_1
HOPF_FTW25_2
HOPF_FTW25_3
HOPF_FTW26_0
HOPF_FTW26_1
HOPF_FTW26_2
HOPF_FTW26_3
HOPF_FTW27_0
HOPF_FTW27_1
HOPF_FTW27_2
HOPF_FTW27_3
HOPF_FTW28_0
HOPF_FTW28_1
HOPF_FTW28_2
HOPF_FTW28_3
HOPF_FTW29_0
HOPF_FTW29_1
HOPF_FTW29_2
HOPF_FTW29_3
HOPF_FTW30_0
HOPF_FTW30_1
HOPF_FTW30_2
HOPF_FTW30_3
HOPF_FTW31_0
HOPF_FTW31_1
HOPF_FTW31_2
HOPF_FTW31_3
[7:0] HOPF_FTW21[7:0]
[7:0] HOPF_FTW21[15:8]
[7:0] HOPF_FTW21[23:16]
[7:0] HOPF_FTW21[31:24]
[7:0] HOPF_FTW22[7:0]
[7:0] HOPF_FTW22[15:8]
[7:0] HOPF_FTW22[23:16]
[7:0] HOPF_FTW22[31:24]
[7:0] HOPF_FTW23[7:0]
[7:0] HOPF_FTW23[15:8]
[7:0] HOPF_FTW23[23:16]
[7:0] HOPF_FTW23[31:24]
[7:0] HOPF_FTW24[7:0]
[7:0] HOPF_FTW24[15:8]
[7:0] HOPF_FTW24[23:16]
[7:0] HOPF_FTW24[31:24]
[7:0] HOPF_FTW25[7:0]
[7:0] HOPF_FTW25[15:8]
[7:0] HOPF_FTW25[23:16]
[7:0] HOPF_FTW25[31:24]
[7:0] HOPF_FTW26[7:0]
[7:0] HOPF_FTW26[15:8]
[7:0] HOPF_FTW26[23:16]
[7:0] HOPF_FTW26[31:24]
[7:0] HOPF_FTW27[7:0]
[7:0] HOPF_FTW27[15:8]
[7:0] HOPF_FTW27[23:16]
[7:0] HOPF_FTW27[31:24]
[7:0] HOPF_FTW28[7:0]
[7:0] HOPF_FTW28[15:8]
[7:0] HOPF_FTW28[23:16]
[7:0] HOPF_FTW28[31:24]
[7:0] HOPF_FTW29[7:0]
[7:0] HOPF_FTW29[15:8]
[7:0] HOPF_FTW29[23:16]
[7:0] HOPF_FTW29[31:24]
[7:0] HOPF_FTW30[7:0]
[7:0] HOPF_FTW30[15:8]
[7:0] HOPF_FTW30[23:16]
[7:0] HOPF_FTW30[31:24]
[7:0] HOPF_FTW31[7:0]
[7:0] HOPF_FTW31[15:8]
[7:0] HOPF_FTW31[23:16]
[7:0] HOPF_FTW31[31:24]
Hopping frequency FTW21
Hopping frequency FTW21
Hopping frequency FTW21
Hopping frequency FTW21
Hopping frequency FTW22
Hopping frequency FTW22
Hopping frequency FTW22
Hopping frequency FTW22
Hopping frequency FTW23
Hopping frequency FTW23
Hopping frequency FTW23
Hopping frequency FTW23
Hopping frequency FTW24
Hopping frequency FTW24
Hopping frequency FTW24
Hopping frequency FTW24
Hopping frequency FTW25
Hopping frequency FTW25
Hopping frequency FTW25
Hopping frequency FTW25
Hopping frequency FTW26
Hopping frequency FTW26
Hopping frequency FTW26
Hopping frequency FTW26
Hopping frequency FTW27
Hopping frequency FTW27
Hopping frequency FTW27
Hopping frequency FTW27
Hopping frequency FTW28
Hopping frequency FTW28
Hopping frequency FTW28
Hopping frequency FTW28
Hopping frequency FTW29
Hopping frequency FTW29
Hopping frequency FTW29
Hopping frequency FTW29
Hopping frequency FTW30
Hopping frequency FTW30
Hopping frequency FTW30
Hopping frequency FTW30
Hopping frequency FTW31
Hopping frequency FTW31
Hopping frequency FTW31
Hopping frequency FTW31
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
0x0
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Rev. 0 | Page 134 of 138
Data Sheet
AD9166
REGISTER SUMMARY: AMPLIFIER
Table 47. Amplifier Register Summary
Reg
Name
Bits Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Reset RW
0x00 SPI_INTFCONFA
[7:0] SOFTRESET_ LSBFIRST_ ADDRINC_ SDOACTIVE_ SDOACTIVE
ADDRINC
LSBFIRST
SOFTRESET 0x00 R/W
M
M
M
M
0x01 SPI_INTFCONFB
0x03 SPI_CHIPTYPE
[7:0] SINGLEINS CS
[7:0]
RESERVED
CHIP_TYPE
SOFTRESET1 SOFTRESET0 RESERVED 0x00 R/W
STALL
0x01
0x33
R
R
0x04 SPI_PRODVARIANT0 [7:0]
0x05 SPI_PRODVARIANT1 [7:0]
PROD_VARIANT0
PROD_VARIANT1
PROD_REV
0xD5 R
0x8C
0x06 SPI_PRODREV
0x0A SPI_SCRATCHPAD
0x10 POWERDOWN
[7:0]
[7:0]
[7:0]
R
SCRATCHPAD
0x00 R/W
0x39 R/W
RESERVED
PD_
NMIRROR PMIRROR
RESERVED
PD_
PD_
RESERVED PD_BG
AMP_ICM
PD_
ADCLOCK
CMDACCURENT
VOUT_TRIM
0x18 TRIM_CM
[7:0]
0x00 R/W
0xA0 R/W
0x00 R/W
0x19 DCOUTPUTVOLTAGE [7:0]
0x1B ADC_START
[7:0]
RESERVED
ST_ADC_
CLKF_1
ST_ADC_
CLKF_0
0x1C ADC_EOC
[7:0]
[7:0]
RESERVED
ADC_CODE
ADC_EOC 0x01
R
0x1D ADC_RESULTS
0xBD R
Rev. 0 | Page 135 of 138
AD9166
Data Sheet
REGISTER DETAILS: AMPLIFIER REGISTER MAP
Table 48. Amplifier Register Details
Addr Name
0x00 SPI_INTFCONFA
Bits Bit Name
Settings Description
Soft reset (mirror) Set this to mirror Bit 0.
LSB first (mirror) Set this to mirror Bit 1.
Reset Access
7
6
5
4
3
2
SOFTRESET_M
0x0
0x0
R
R
LSBFIRST_M
ADDRINC_M
SDOACTIVE_M
SDOACTIVE
ADDRINC
Address Increment (mirror) Set this to mirror Bit 2. 0x0
SDO active (mirror) Set this to mirror Bit 3.
SDO active. Enables 4-wire SPI bus mode.
Address Increment. When set, causes incrementing 0x0
of streaming addresses. Otherwise, descending
addresses are generated.
R
0x0
0x0
R
R/W
R/W
1 Streaming addresses are incremented.
0 Streaming addresses are decremented.
1
0
LSBFIRST
LSB first. When set, causes input and output data to 0x0
be oriented as LSB first. If this bit is clear, data is
oriented as MSB first.
R/W
R/W
1 Shift LSB in first.
0 Shift MSB in First.
SOFTRESET
Soft reset. This bit automatically clears to 0 after
performing a reset operation. Setting this bit
initiates a reset. This bit is autoclearing after the
soft reset is complete.
0x0
1 Pulse the soft reset line.
0 Reset the soft reset line.
Single instruction.
1 Perform single transfers.
0 Perform multiple transfers.
0x01 SPI_INTFCONFB
7
6
SINGLEINS
0x0
0x0
R/W
R/W
CS
STALL
CS_x
stalling.
0
1
CS_x
Disable
Enable
stalling.
stalling.
CS_x
[5:3] RESERVED
Reserved.
0x0
R
2
1
0
SOFTRESET1
SOFTRESET0
RESERVED
Soft Reset 1. This bit automatically clears to 0 after 0x0
performing a reset operation.
1 Pulse the Soft Reset 1 line.
0 Pulse the Soft Reset 1 line.
Soft Reset 0. This bit automatically clears to 0 after 0x0
performing a reset operation.
1 Pulse the Soft Reset 0 line.
0 Pulse the Soft Reset 0 line.
R/W
R/W
Reserved.
0x0
R
0x03 SPI_CHIPTYPE
[7:0] CHIP_TYPE
Chip type.
0x01
0x33
0xD5
0x8C
0x0
R
0x04 SPI_PRODVARIANT0 [7:0] PROD_VARIANT0
0x05 SPI_PRODVARIANT1 [7:0] PROD_VARIANT1
0x06 SPI_PRODREV
0x0A SPI_SCRATCHPAD
0x10 POWERDOWN
Product variant.
Product variant.
Revision of the product variant.
Scratch pad R/W register.
Reserved.
R
R
R
R/W
R/W
R/W
[7:0] PROD_REV
[7:0] SCRATCHPAD
[7:6] RESERVED
0x0
5
PD_NMIRROR
Force 1/10 nominal bias current in output stage.
1 Power down.
0x1
0 Normal functioning.
Force 1/10 nominal bias current in input stage.
1 Power down.
4
PD_PMIRROR
0x1
R/W
0 Normal functioning
Rev. 0 | Page 136 of 138
Data Sheet
AD9166
Addr Name
Bits Bit Name
Settings Description
Reset Access
3
PD_CMDACCURRENT
Force the DAC common-mode current to minimum 0x1 R/W
1 Power down.
0 Normal functioning.
2
1
RESERVED
PD_BG
Reserved.
0x0 R/W
Power down band gap and amplifier bias. Removes 0x0 R/W
bias to the ADC, and the input and output stages.
1 Power down.
0 Normal functioning.
0
PD_ADCCLOCK
Power down ADC clock.
1 Power down.
0x1 R/W
0 Normal functioning.
Reserved.
0x18 TRIM_CM
[7:4] RESERVED
[3:0] AMP_ICM
0x0
R
Sets the input common-mode current (ICM) of the 0x0 R/W
amplifier. To minimize common-mode voltage
(VCM) offset at the DAC output, set ICM to the nearest
setting (AMP_ICM, Bits[3:0]) that corresponds to
the full-scale current setting of the DAC
(ANA_FULL_SCALE_CURRENT, Bits[9:0],
Register 0x42 and Register 0x41).
ICM = (30.4 − 6.4) × AMP_ICM/15 + 6.4 mA
0x19 DCOUTPUTVOLTAGE [7:0] VOUT_TRIM
Adjusts the dc offset of the RF output (VOS_ADJ).
VOS_ADJ = 0.6 V × VOUT_TRIM/255 − 0.25
Reserved.
0xA0 R/W
0x1B ADC_START
[7:2] RESERVED
0x0
0x0
R
1
ST_ADC_CLKF_1
Select the ADC clock frequency (fS).
0 2 MHz.
1 250 kHz.
0
ST_ADC_CLKF_0
Set high to start ADC conversion, which takes
approximately 17 ADC clock cycles. End of
conversion is indicated by ADC_EOC, Bit 0.
0x0 R/W
0x1C ADC_EOC
[7:1] RESERVED
ADC_EOC
Reserved.
0x0
R
R
0
ADC end of conversion flag. A 0 indicates an ADC 0x1
conversion is in progress, if triggered earlier by
setting the ADC_START, Bit 0.
0 ADC conversion is in progress.
1 ADC conversion is finished.
0x1D ADC_RESULTS
[7:0] ADC_CODE
ADC output code (sample) at the end of an ADC
conversion cycle.
0xBD
R
An ADC conversion cycle can be initiated by setting
ADC_START, Bit 0 = high. The ADC code can be
read at the end of a conversion cycle (indicated by
ADC_EOC, Bit 0 = high).
The ADC is sampling an input voltage (VADC),
measured at the output of an analog mux, which is
connected to the junction temperature sensor.
VADC = VBGA × ADC_CODE/255
where VBGA = 1.09 V nominally.
V
BG varies between devices, leading to
measurement uncertainty. The variation is ±30 mV
over process, voltage (supply), and temperature
(PVT). VBGA can be measured at the AMP_VBG pin.
Rev. 0 | Page 137 of 138
AD9166
Data Sheet
OUTLINE DIMENSIONS
15.05
15.00 SQ
14.95
A1 BALL
CORNER
A1 BALL
CORNER
0.70
BSC
18 16 14 12 10
17 15 13 11
8
6
4
2
9
7
5
3
1
A
B
C
D
F
E
14.75
14.70 SQ
14.65
G
H
J
L
13.60 REF
SQ
K
M
P
T
11.80
11.60 SQ
11.50
N
R
U
0.80
BSC
V
TOP VIEW
BOTTOM VIEW
0.15 REF
R 1.0
R 0.5~1.5
DETAIL A
1.57
1.52
1.47
0.87 REF
DETAIL A
SIDE VIEW
0.29
0.25
0.21
0.40
0.35
0.30
SEATING
PLANE
COPLANARITY
0.10
0.50
0.45
0.40
BALL DIAMETER
Figure 96. 324-Ball Ball Grid Array, Thermally Enhanced [BGA_ED]
(BP-324-1)
Dimensions shown in millimeters
ORDERING GUIDE
Model1
Temperature Range
Package Description
Package Option
BP-324-1
BP-324-1
AD9166BBPZ
AD9166BBPZRL
AD9166-FMC-EBZ
−40°C to +85°C
−40°C to +85°C
324-Ball Ball Grid Array, Thermally Enhanced [BGA_ED]
324-Ball Ball Grid Array, Thermally Enhanced [BGA_ED]
Evaluation Board
1 Z = RoHS Compliant part.
2020 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D20810-7/20(0)
Rev. 0 | Page 138 of 138
相关型号:
©2020 ICPDF网 联系我们和版权申明