PDF Data Sheet Rev. A

Data Sheet
JESD204B Octal Ultrasound AFE with
Digital Demodulator
AD9671
FEATURES
GENERAL DESCRIPTION
8 channels of LNA, VGA, AAF, ADC, and digital demodulator/
decimator
Low power: 150 mW per channel, time gain compensation
(TGC) mode, 40 MSPS
62.5 mW per channel, continuous wave (CW) mode;
<30 mW in power-down mode
10 mm × 10 mm, 144-ball CSP_BGA
TGC channel input referred noise: 0.82 nV/√Hz,
maximum gain
Flexible power-down modes
Fast recovery from low power standby mode: <2 μs
Low noise preamplifier (LNA)
Input referred noise: 0.78 nV/√Hz, gain = 21.6 dB
Programmable gain: 15.6 dB/17.9 dB/21.6 dB
0.1 dB compression: 1000 mV p-p/750 mV p-p/450 mV p-p
Flexible active input impedance matching
Variable gain amplifier (VGA)
Attenuator range: 45 dB, linear-in-dB gain control
Postamplifier (PGA) gain: 21 dB/24 dB/27 dB/30 dB
Antialiasing filter (AAF)
Programmable, second-order low-pass filter (LPF) from
8 MHz to 18 MHz or 13.5 MHz to 30 MHz and high-pass
filter (HPF)
Analog-to-digital converter (ADC)
Signal-to-noise ratio (SNR): 75 dB, 14 bits up to 125 MSPS
JESD204B Subclass 0 coded serial digital outputs
CW Doppler (CWD) mode harmonic rejection I/Q demodulator
Individual programmable phase rotation
Dynamic range per channel: 160 dBFS/√Hz
Close-in SNR: 156 dBc/√Hz, 1 kHz offset, −3 dBFS input
Digital demodulator/decimator
I/Q demodulator with programmable oscillator
The AD9671 is designed for low cost, low power, small size, and
ease of use for medical ultrasound applications. It contains eight
channels of a VGA with an LNA, a CW harmonic rejection I/Q
demodulator with programmable phase rotation, an AAF, an
ADC, and a digital demodulator and decimator for data
processing and bandwidth reduction.
Each channel features a maximum gain of up to 52 dB, a fully
differential signal path, and an active input preamplifier termination.
The channel is optimized for high dynamic performance and
low power in applications where a small package size is critical.
The LNA has a single-ended to differential gain that is selectable
through the serial port interface (SPI). Assuming a 15 MHz noise
bandwidth (NBW) and a 21.6 dB LNA gain, the LNA input SNR
is 94 dB. In CW Doppler mode, each LNA output drives an I/Q
demodulator that has independently programmable phase
rotation with 16 phase settings.
Power-down of individual channels is supported to increase
battery life for portable applications. Standby mode allows quick
power-up for power cycling. In CW Doppler operation, the
VGA, AAF, and ADC are powered down. The ADC contains
several features designed to maximize flexibility and minimize
system cost, such as a programmable clock, data alignment, and
programmable digital test pattern generation. The digital test
patterns include built-in fixed patterns, built-in pseudorandom
patterns, and custom user defined test patterns entered via the SPI.
APPLICATIONS
Medical imaging/ultrasound
Nondestructive testing (NDT)
Rev. A
Document Feedback
Information furnished by Analog Devices is believed to be accurate and reliable. However, no
responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other
rights of third parties that may result from its use. Specifications subject to change without notice. No
license is granted by implication or otherwise under any patent or patent rights of Analog Devices.
Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781.329.4700 ©2013–2016 Analog Devices, Inc. All rights reserved.
Technical Support
www.analog.com
AD9671
Data Sheet
TABLE OF CONTENTS
Features .............................................................................................. 1 CW Doppler Operation............................................................. 39 Applications ....................................................................................... 1 Digital Demodulator/Decimator .................................................. 40 General Description ......................................................................... 1 Vector Profile .............................................................................. 40 Revision History ............................................................................... 2 RF Decimator .............................................................................. 41 Functional Block Diagram .............................................................. 3 Baseband Demodulator and Decimator.................................. 42 Specifications..................................................................................... 4 Digital Test Waveforms.............................................................. 43 AC Specifications.......................................................................... 4 Digital Block Power Saving Scheme ........................................ 43 Digital Specifications ................................................................... 8 Serial Port Interface (SPI) .............................................................. 44 Switching Specifications .............................................................. 9 Hardware Interface..................................................................... 44 Absolute Maximum Ratings.......................................................... 12 Memory Map .................................................................................. 46 Thermal Impedance ................................................................... 12 Reading the Memory Map Table .............................................. 46 ESD Caution ................................................................................ 12 Reserved Locations .................................................................... 46 Pin Configuration and Function Descriptions ........................... 13 Default Values ............................................................................. 46 Typical Performance Characteristics ........................................... 16 Logic Levels ................................................................................. 46 TGC Mode ................................................................................... 16 Recommended Start-Up Sequence .......................................... 46 CW Doppler Mode ..................................................................... 20 Memory Map Register Descriptions ........................................ 59 Theory of Operation ...................................................................... 21 Outline Dimensions ....................................................................... 60 TGC Operation ........................................................................... 21 Ordering Guide .......................................................................... 60 Digital Outputs and Timing ...................................................... 29 Analog Test Tone Generation ................................................... 38 REVISION HISTORY
1/16—Revision A: Initial Version
Rev. A | Page 2 of 60
Data Sheet
AD9671
FUNCTIONAL BLOCK DIAGRAM
PDWN STBY
DVDD
DRVDD
AD9671
LO-A TO LO-H
CWD I/Q
DEMODULATOR
LOSW-A TO LOSW-H
LI-A TO LI-H
LNA
LG-A TO LG-H
VGA
14-BIT
ADC
AAF
DEMODULATOR/
DECIMATOR
CML
SERIALIZER
Figure 1.
Rev. A| Page 3 of 60
SERDOUT1– TO SERDOUT4–
CLK–
DATA
RATE
MULTIPLIER
CLK+
TX_TRIG–
TX_TRIG+
NCO
SDIO
CSB
SCLK
GPO0 TO
GPO3
ADDR0 TO
ADDR4
VREF
RBIAS
GAIN+
GAIN–
MLO–
MLO+
RESET–
RESET+
SERIAL
PORT
INTERFACE
REFERENCE
SERDOUT1+ TO SERDOUT4+
SYSREF+
SYSREF–
SYNCINB+
SYNCINB–
8 CHANNELS
LO
GENERATION
CWQ+
CWQ–
CWI+
CWI–
11134-001
AVDD1 AVDD2
AD9671
Data Sheet
SPECIFICATIONS
AC SPECIFICATIONS
AVDD1 = 1.8 V, AVDD2 = 3.0 V, DVDD = 1.4 V, DRVDD = 1.8 V, 1.0 V internal ADC reference, full temperature range (0°C to 85°C),
fIN = 5 MHz, low bandwidth mode, RS = 50 Ω, RFB = ∞ (unterminated), LNA gain = 21.6 dB, LNA bias = midhigh, PGA gain = 27 dB,
analog gain control, VGAIN (V) = (GAIN+) − (GAIN−) = 1.6 V, AAF LPF cutoff = fSAMPLE/3 (Mode I/Mode II) = fSAMPLE/4.5 (Mode III/
Mode IV), HPF cutoff = LPF cutoff/12.00, Mode I = fSAMPLE = 40 MSPS, Mode II = fSAMPLE = 65 MSPS, Mode III = fSAMPLE = 80 MSPS,
Mode IV = 125 MSPS, RF decimator bypassed (Mode I/Mode II), RF decimator enabled (Mode III/Mode IV), digital high-pass filter
bypassed, demodulator bypassed, baseband decimator bypassed, JESD204B link parameters: M = 8 and L = 2, unless otherwise noted.
All gain setting options are listed, which can be configured via SPI registers, and all power supply currents and power dissipations are listed for
the four mode settings (Mode I, Mode II, Mode III, and Mode IV), respectively, via slashes in Table 1.
Table 1.
Parameter1
LNA CHARACTERISTICS
Gain
Test Conditions/Comments
Min
Single-ended input to differential
output
Single-ended input to single-ended
output
Typ
Max
Unit
15.6/17.9/21.6
dB
9.6/11.9/15.6
dB
LNA gain = 15.6 dB
LNA gain = 17.9 dB
LNA gain = 21.6 dB
1000
750
450
mV p-p
mV p-p
mV p-p
LNA gain = 15.6 dB
LNA gain = 17.9 dB
LNA gain = 21.6 dB
1200
900
600
2.2
mV p-p
mV p-p
mV p-p
V
Switch off
Switch on
Switch off
Switch on
RFB = 300 Ω, LNA gain = 21.6 dB
RFB = 1350 Ω, LNA gain = 21.6 dB
High-Z
1.5
High-Z
1.5
50
200
6
22
Ω
V
Ω
V
Ω
Ω
kΩ
pF
0.83
0.82
0.78
94
2.6
nV/√Hz
nV/√Hz
nV/√Hz
dB
pA/√Hz
0.1 dB Input Compression Point
1 dB Input Compression Point
Input Common Mode (LI-x, LG-x)
Output Common Mode
LO-x
LOSW-x
Input Resistance (LI-x)
Input Capacitance (LI-x)
Input Referred Noise Voltage
Input Signal-to-Noise Ratio
Input Noise Current
FULL CHANNEL CHARACTERISTICS
AAF Low-Pass Cutoff
In Range AAF Bandwidth Tolerance
Group Delay Variation
Input Referred Noise Voltage
RS = 0 Ω
LNA gain = 15.6 dB
LNA gain = 17.9 dB
LNA gain = 21.6 dB
Noise bandwidth = 15 MHz
Time gain control (TGC)
−3 dB, programmable, low bandwidth
mode
−3 dB, programmable, high bandwidth
mode
8
18
MHz
13.5
30
MHz
f = 1 MHz to 18 MHz, VGAIN = −1.6 V to
+1.6 V
LNA gain = 15.6 dB
LNA gain = 17.9 dB
LNA gain = 21.6 dB
Rev. A| Page 4 of 60
±10
±350
%
ps
0.96
0.90
0.82
nV/√Hz
nV/√Hz
nV/√Hz
Data Sheet
Parameter1
Noise Figure
Active Termination Matched
Unterminated
Correlated Noise Ratio
Output Offset
Signal-to-Noise Ratio (SNR)
Close-In SNR
Second Harmonic
Third Harmonic
Two-Tone Intermodulation
Distortion (IMD3)
Channel-to-Channel Crosstalk
GAIN ACCURACY
Gain Law Conformance Error
Channel-to-Channel Matching
PGA Gain
GAIN CONTROL INTERFACE
Control Range
Control Common Mode
Input Impedance
Gain Range
Gain Sensitivity
Response Time
CW DOPPLER MODE
LO Frequency
Phase Resolution
Output DC Bias (Single-Ended)
Output AC Current Range
Transconductance (Differential)
Input Referred Noise Voltage
AD9671
Test Conditions/Comments
Min
LNA gain = 15.6 dB, RFB = 150 Ω
LNA gain = 17.9 dB, RFB = 200 Ω
LNA gain = 21.6 dB, RFB = 300 Ω
LNA gain = 15.6 dB, RFB = ∞
LNA gain = 17.9 dB, RFB = ∞
LNA gain = 21.6 dB, RFB = ∞
No signal, correlated/uncorrelated
Typ
5.6
4.8
3.8
3.2
2.9
2.6
−30
Differential
GAIN+, GAIN−
GAIN+, GAIN−
69
59
−130
−70
−62
−61
−55
−54
dBc
dBc
dBc
dBc
dBc
−60
−55
dB
dB
0.4
+1.3
−0.5
−0.9
+0.9
0.1
21/24/27/30
Analog
Digital step size
Analog 45 dB change
fLO = fMLO/M
Per channel, 4LO3 mode
Per channel, 8LO3 mode, 16LO3 mode
CWI+, CWI−, CWQ+, CWQ−
Per CWI+, CWI−, CWQ+, CWQ−, each
channel enabled (2 fLO and baseband
signal)
Demodulated IOUT/VIN, per CWI+,
CWI−, CWQ+, CWQ−
LNA gain = 15.6 dB
LNA gain = 17.9 dB
LNA gain = 21.6 dB
RS = 0 Ω, RFB = ∞
LNA gain = 15.6 dB
LNA gain = 17.9 dB
LNA gain = 21.6 dB
+125
−1.3
−1.6
0.7
0.8
10
45
14
3.5
750
1
Rev. A| Page 5 of 60
Unit
dB
dB
dB
dB
dB
dB
dB
LSB
dBFS
dBFS
dBc/√Hz
−125
fIN = 5 MHz at −12 dBFS, VGAIN = −1.6 V
fIN = 5 MHz at −1 dBFS, VGAIN = 1.6 V
fIN = 3.5 MHz at −0.5 dBFS, VGAIN = 0 V,
1 kHz offset
fIN = 5 MHz at −12 dBFS, VGAIN = −1.6 V
fIN = 5 MHz at −1 dBFS, VGAIN = 1.6 V
fIN = 5 MHz at −12 dBFS, VGAIN = −1.6 V
fIN = 5 MHz at −1 dBFS, VGAIN = 1.6 V
fRF1 = 5.015 MHz, fRF2 = 5.020 MHz,
ARF1 = −1 dBFS, ARF2 = −21 dBFS, VGAIN =
1.6 V, IMD3 relative to ARF2
fIN1 = 5.0 MHz at −1 dBFS
Overrange condition2
TA = 25°C
−1.6 < VGAIN < −1.28 V
−1.28 V < VGAIN < +1.28 V
1.28 V < VGAIN < 1.6 V
VGAIN = 0 V, normalized for ideal AAF
loss
−1.28 V < VGAIN < +1.28 V, 1 σ
Max
45
22.5
AVDD2/2
±2.2
dB
dB
dB
dB
dB
dB
+1.6
0.9
V
V
MΩ
dB
dB/V
dB
ns
10
MHz
Degrees
Degrees
V
mA
±2.5
3.3
4.3
6.6
mA/V
mA/V
mA/V
1.6
1.3
1.0
nV/√Hz
nV/√Hz
nV/√Hz
AD9671
Parameter1
Noise Figure
Input Referred Dynamic Range
Close-In SNR
Two-Tone Intermodulation
Distortion (IMD3)
LO Harmonic Rejection
Quadrature Phase Error
I/Q Amplitude Imbalance
Channel to Channel Matching
POWER SUPPLY
AVDD1
AVDD2
DVDD
DRVDD
IAVDD1
IAVDD2
IDVDD
IDRVDD
Total Power Dissipation
(Including Output Drivers)
Data Sheet
Test Conditions/Comments
RS = 50 Ω, RFB = ∞
LNA gain = 15.6 dB
LNA gain = 17.9 dB
LNA gain = 21.6 dB
RS = 0 Ω, RFB = ∞
LNA gain = 15.6 dB
LNA gain = 17.9 dB
LNA gain = 21.6 dB
−3 dBFS input, fRF = 2.5 MHz, fLO =
40 MHz, 1 kHz offset, 16LO mode,
one channel enabled
−3 dBFS input, fRF = 2.5 MHz, fLO =
40 MHz, 1 kHz offset, 16LO mode,
eight channels enabled
fRF1 = 5.015 MHz, fRF2 = 5.020 MHz, fLO =
80 MHz, ARF1 = −1 dBFS, ARF2 =
−21 dBFS, IMD3 relative to ARF2
Min
I to Q, all phases, 1 σ
I to Q, all phases, 1 σ
Phase I to I, Q to Q, 1 σ
Amplitude I to I, Q to Q, 1 σ
Mode I/Mode II/Mode III/Mode IV
Demodulator/decimator enabled
Typ
Max
5.7
4.5
3.4
dB
dB
dB
164
162
160
156
dBFS/√Hz
dBFS/√Hz
dBFS/√Hz
dBc/√Hz
161
dBc/√Hz
−58
dBc
−20
dBc
Degrees
dB
Degrees
dB
1.9
3.6
1.9
1.9
V
V
V
V
mA
mA
mA
0.15
0.015
0.5
0.25
1.7
2.85
1.3
1.7
TGC mode, low bandwidth mode
CW Doppler mode
TGC mode, no signal, low bandwidth
mode
TGC mode, no signal, high
bandwidth mode
CW Doppler mode
Demodulator/decimator enabled
Demodulator/decimator disabled
Four-lane mode, JESD204B lane rates =
1.6 Gbps/2.6 Gbps/1.6 Gbps/2.5 Gbps
Two-lane mode, JESD204B lane rates =
3.2 Gbps/5.0 Gbps/3.2 Gbps/5.0 Gbps
One-lane mode, demodulator/
decimator enabled, JESD204B lane
rates = 3.2 Gbps/5.0 Gbps/3.2 Gbps/
5.0 Gbps 4
TGC mode, no signal, two-lane mode,
demodulator/decimator disabled
TGC mode, no signal, two-lane mode,
demodulator/decimator enabled
CW Doppler mode, eight channels
enabled
Power-Down Dissipation
Standby Power Dissipation
1.8
3.0
1.4
1.8
148/187/223/291
4
230
239
mA
140
156/247/166/255
29/46/40/61
121/168/122/166
mA
mA
mA
mA
127/186/129/184
mA
73/105/76/105
mA
1200/1415/1365/1615
1390/1710/1550/1895
1445/1680/1635/
1910
1645/2025/1835/
2215
500
5
725
Rev. A| Page 6 of 60
Unit
mW
mW
mW
30
mW
mW
Data Sheet
Parameter1
ADC
Resolution
SNR
ADC REFERENCE
Output Voltage Error
Load Regulation at 1.0 mA
Input Resistance
AD9671
Test Conditions/Comments
Min
Typ
Max
14
75
VREF = 1 V
VREF = 1 V
Bits
dB
±50
2
7.5
Unit
mV
mV
kΩ
1
See the AN-835 Application Note, Understanding High Speed ADC Testing and Evaluation, for a complete set of definitions and information about how these tests were
completed.
2
The overrange condition is specified as 6 dB more than the full-scale input range.
3
The internal LO frequency, fLO, is generated from the supplied multiplier local oscillator frequency, fMLO, by dividing it up by a configurable divider value (M) that can be
4, 8, or 16; the MLO signal is named 4LO, 8LO, or 16LO, accordingly.
4
Baseband decimation rate = 4. M = 16.
Rev. A| Page 7 of 60
AD9671
Data Sheet
DIGITAL SPECIFICATIONS
AVDD1 = 1.8 V, AVDD2 = 3.0 V, DVDD = 1.4 V, DRVDD = 1.8 V, 1.0 V internal ADC reference, full temperature range (0°C to 85°C), unless
otherwise noted.
Table 2.
Parameter1
INPUTS
CLK+, CLK−, TX_TRIG+, TX_TRIG−
Logic Compliance
Differential Input Voltage2
Input Voltage Range
Input Common-Mode Voltage
Input Resistance (Differential)
Input Capacitance
MLO+, MLO−, RESET+, RESET−
Logic Compliance
Differential Input Voltage2
Input Voltage Range
Input Common-Mode Voltage
Input Resistance (Single-Ended)
Input Capacitance
LOGIC INPUTS
PDWN, STBY, SCLK, SDIO, ADDRx
Logic 1 Voltage
Logic 0 Voltage
Input Resistance
Input Capacitance
CSB
Logic 1 Voltage
Logic 0 Voltage
Input Resistance
Input Capacitance
LOGIC OUTPUTS
SDIO3
Logic 1 Voltage (IOH = 800 μA)
Logic 0 Voltage (IOL = 50 μA)
GPO0, GPO1, GPO2, GPO3
Logic 0 Voltage (IOL = 50 μA)
DIGITAL OUTPUTS (SERDOUTx+, SERDOUTx−)
Logic Compliance
Differential Output Voltage (VOD)
Output Offset Voltage (VOS)
DIGITAL INPUTS
SYNCINB+, SYNCINB−
Logic Compliance
Internal Bias
Differential Input Voltage Range
Input Voltage Range
Input Common-Mode Range
High Level Input Current
Low Level Input Current
Input Capacitance
Input Resistance
Temperature
Min
Typ
Max
Unit
3.6
AVDD1 + 0.2
V p-p
V
V
kΩ
pF
AVDD2 × 2
AVDD2 + 0.2
V p-p
V
V
kΩ
pF
DRVDD + 0.3
0.3
V
V
kΩ
pF
DRVDD + 0.3
0.3
V
V
kΩ
pF
CMOS/LVDS/LVPECL
0.2
GND – 0.2
0.9
15
4
25°C
25°C
LVDS/LVPECL
0.250
GND – 0.2
AVDD2/2
20
1.5
25°C
25°C
Full
Full
25°C
25°C
1.2
Full
Full
25°C
25°C
1.2
30 (SDIO = 26)
2 (SDIO = 5)
26
2
Full
Full
1.79
Full
Full
Full
400
0.75
CML
600
0.05
V
V
0.05
V
750
1.05
mV
V
LVDS
Full
Full
Full
Full
Full
Full
Full
Full
0.9
0.3
GND
0.9
−5
−5
12
Rev. A| Page 8 of 60
3.6
DRVDD
1.4
+5
+5
1
16
20
V
V
V
V
μA
μA
pF
kΩ
Data Sheet
Parameter1
SYSREF+, SYSREF−
Logic Compliance
Internal Common-Mode Bias
Differential Input Voltage
Input Voltage Range
Input Common-Mode Range
High Level Input Current
Low Level Input Current
Input Capacitance
Input Resistance
AD9671
Temperature
Min
Typ
Max
Unit
LVDS
Full
Full
Full
Full
Full
Full
Full
Full
0.9
0.3
GND
0.9
−5
−5
V
V p-p
V
V
μA
μA
pF
kΩ
3.6
DRVDD
1.4
+5
+5
4
10
8
12
1
See the AN-835 Application Note, Understanding High Speed ADC Testing and Evaluation, for a complete set of definitions and information about how these tests were
completed.
2
Specified for LVDS and LVPECL only.
3
Specified for 13 SDIO pins sharing the same connection.
SWITCHING SPECIFICATIONS
AVDD1 = 1.8 V, AVDD2 = 3.0 V, DVDD = 1.4 V, DRVDD = 1.8 V, 1.0 V internal ADC reference, L = 2, M = 8, fSAMPLE = 40 MHz, lane
data rate = 3.2 Gbps, full temperature range (0°C to 85°C), unless otherwise noted.
Table 3.
Parameter1
CLOCK2
Clock Rate (fSAMPLE)
40 MSPS (Mode I)
65 MSPS (Mode II)
80 MSPS (Mode III)3
125 MSPS (Mode IV)4
Clock Pulse Width High (tEH)
Clock Pulse Width Low (tEL)
CLOCK INPUT PARAMETERS
TX_TRIG± to CLK± Setup Time (tSETUP)
TX_TRIG± to CLK± Hold Time (tHOLD)
DATA OUTPUT PARAMETERS
Data Output Period or Unit Interval (UI)
Data Output Duty Cycle
Data Valid Time
PLL Lock Time5
Wake-Up Time
Standby
Power-Down6
Device
JESD204B Link
SYNCINB± Falling Edge to First K.28 Characters
Code Group Synchronization (CGS) Phase K.28
Characters Duration
Delay (Latency)
ADC Pipeline
RF Decimator
Digital High-Pass Filter
Baseband Decimator
TX_TRIG± to Start Code (Mode I/Mode II/Mode III/
Mode IV)
Four-Lane Mode
Two-Lane Mode
Temperature
Min
Full
Full
Full
Full
Full
Full
20.5
20.5
20.5
20.5
25°C
25°C
1
1
Typ
3.75
3.75
Max
Unit
40
65
80
125
MHz
MHz
MHz
MHz
ns
ns
ns
ns
Full
25°C
25°C
25°C
L/(20 × M × fSAMPLE)
50
0.76
26
sec
%
UI
μs
25°C
2
μs
25°C
25°C
Full
Full
375
250
μs
μs
Multiframes
Multiframe
16
11
100
16 × decimation
factor
Cycles
Cycles
Cycles
Cycles
31/42/30/36
31/33/30/30
Cycles
Cycles
4
1
Full
Full
Full
Full
Full
Full
Full
Rev. A| Page 9 of 60
AD9671
Data Sheet
Parameter1
Data Rate per Lane
Uncorrelated Bounded High Probability (UBHP) Jitter
Random Jitter at 2.5 Gbps Data Rate
Random Jitter at 5 Gbps Data Rate
Output Rise/Fall Time
TERMINATION CHARACTERISTICS
Differential Termination Resistance
APERTURE
Aperture Uncertainty (Jitter)
LO GENERATION
MLO± Frequency
4LO Mode
8LO Mode
16LO Mode
RESET± to MLO± Setup Time (tSETUP)
RESET± to MLO± Hold Time (tHOLD)
Temperature
25°C
25°C
25°C
25°C
25°C
11
80
46
64
Unit
Gbps
ps
ps rms
ps rms
ps
Full
100
Ω
25°C
<1
ps rms
Full
Full
Full
Full
Full
Min
4
8
16
1
1
Typ
Max
5.0
40
80
160
tMLO7/2
tMLO7/2
MHz
MHz
MHz
ns
ns
1
See the AN-835 Application Note, Understanding High Speed ADC Testing and Evaluation, for a complete set of definitions and information about how these tests were
completed.
Can be adjusted via the SPI.
3
Mode III must have the RF decimator enabled.
4
Mode IV must have the RF decimator enabled.
5
PLL lock time from 0 Hz to 40 MHz frequency change.
6
Wake-up time is defined as the time required to return to normal operation from power-down mode.
7
The period of the MLO clock signal is represented by tMLO.
2
CLK±, TX_TRIG± Synchronization Timing Diagram
tSETUP
tHOLD
TX_TRIG+
TX_TRIG–
tEH
tEL
11134-002
CLK–
CLK+
Figure 2. TX_TRIG± to CLK± Input Timing
CW Timing Diagram
tMLO
MLO–
MLO+
tSETUP
tHOLD
11134-003
RESET–
RESET+
Figure 3. CW Doppler Mode Input MLO±, Continuous Synchronous RESET± Timing, Sampled on the Falling MLO± Edge, 4LO Mode
Rev. A| Page 10 of 60
Data Sheet
AD9671
tMLO
MLO–
MLO+
tSETUP
tHOLD
11134-005
RESET–
RESET+
Figure 4. CW Doppler Mode Input MLO±, Continuous Synchronous RESET± Timing, Sampled on the Falling MLO± Edge, 8LO Mode
tMLO
MLO–
MLO+
11134-004
RESET–
tHOLD
tSETUP
RESET+
Figure 5. CW Doppler Mode Input MLO±, Pulse Synchronous RESET± Timing, 4LO/8LO/16LO Mode
tMLO
MLO–
MLO+
tSETUP
11134-006
RESET–
tHOLD
RESET+
Figure 6. CW Doppler Mode Input MLO±, Pulse Asynchronous RESET± Timing, 4LO/8LO/16LO Mode
Rev. A| Page 11 of 60
AD9671
Data Sheet
ABSOLUTE MAXIMUM RATINGS
THERMAL IMPEDANCE
Table 4.
Parameter
AVDD1 to GND
AVDD2 to GND
DVDD to GND
DRVDD to GND
GND to GND
AVDD2 to AVDD1
AVDD1 to DRVDD
AVDD2 to DRVDD
SERDOUTx+, SERDOUTx−, SDIO, PDWN,
STBY, SCLK, CSB, ADDRx to GND
LI-x, LO-x, LOSW-x, CWI−, CWI+, CWQ−,
CWQ+, GAIN+, GAIN−, RESET+,
RESET−, MLO+, MLO−, GPO0, GPO1,
GPO2, GPO3 to GND
CLK+, CLK−, TX_TRIG+, TX_TRIG−, VREF
to GND
Operating Temperature Range (Ambient)
Storage Temperature Range (Ambient)
Maximum Junction Temperature
Lead Temperature (Soldering, 10 sec)
Rating
−0.3 V to +2.0 V
−0.3 V to +3.9 V
−0.3 V to +2.0 V
−0.3 V to +2.0 V
−0.3 V to +0.3 V
−2.0 V to +3.9 V
−2.0 V to +2.0 V
−2.0 V to +3.9 V
−0.3 V to DRVDD + 0.3 V
−0.3 V to AVDD2 + 0.3 V
−0.3 V to AVDD1 + 0.3 V
Table 5.Thermal Impedance
Symbol
JA
JB
JT
1
Description
Junction to ambient thermal
resistance, 0.0 m/sec air flow per
JEDEC JESD51-2 (still air)
Junction to board thermal
characterization parameter, 0 m/sec
air flow per JEDEC JESD51-8 (still air)
Junction to top of package
characterization parameter, 0 m/sec
air flow per JEDEC JESD51-2 (still air)
Value1
22.0
Unit
°C/W
9.2
°C/W
0.12
°C/W
Results are from simulations. Printed circuit board (PCB) is JEDEC multilayer.
Thermal performance for actual applications requires careful inspection of
the conditions in the application to determine if they are similar to those
assumed in these calculations.
ESD CAUTION
0°C to 85°C
−65°C to +150°C
150°C
300°C
Stresses at or above those listed under Absolute Maximum
Ratings may cause permanent damage to the product. This is a
stress rating only; functional operation of the product at these
or any other conditions above those indicated in the operational
section of this specification is not implied. Operation beyond
the maximum operating conditions for extended periods may
affect product reliability.
Rev. A| Page 12 of 60
Data Sheet
AD9671
PIN CONFIGURATION AND FUNCTION DESCRIPTIONS
1
2
3
4
5
6
7
8
9
10
11
12
A
LI-E
LI-F
LI-G
LI-H
VREF
RBIAS
GAIN+
GAIN–
LI-A
LI-B
LI-C
LI-D
B
LG-E
LG-F
LG-G
LG-H
GND
GND
CLNA
GND
LG-A
LG-B
LG-C
LG-D
C
LO-E
LO-F
LO-G
LO-H
GND
GND
GND
GND
LO-A
LO-B
LO-C
LO-D
D LOSW-E LOSW-F LOSW-G LOSW-H
GND
GND
GND
GND
E
GND
AVDD2
AVDD2
AVDD2
GND
GND
GND
GND
AVDD2
AVDD2
AVDD2
GND
F
AVDD1
GND
AVDD1
GND
AVDD1
GND
GND
AVDD1
GND
AVDD1
GND
AVDD1
G
GND
AVDD1
GND
DVDD
GND
GND
GND
GND
AVDD1
GND
DVDD
GND
H
CLK–
TX_TRIG–
GND
GND
GND
GND
ADDR4
ADDR3
ADDR2
ADDR1
ADDR0
CSB
J
CLK+
TX_TRIG+
CWQ+
GND
CWI+
AVDD2
MLO+
RESET–
GPO3
GPO1
PDWN
SDIO
K
GND
GND
CWQ–
GND
CWI–
AVDD2
MLO–
RESET+
GPO2
GPO0
STBY
SCLK
L
DRVDD
NIC
NIC
SYNCINB+
SERDOUT4+ SERDOUT3+ SERDOUT2+ SERDOUT1+
SYSREF+
NIC
NIC
DRVDD
M
GND
NIC
NIC
SYNCINB–
SERDOUT4– SERDOUT3– SERDOUT2– SERDOUT1–
SYSREF–
NIC
NIC
GND
11134-007
LOSW-A LOSW-B LOSW-C LOSW-D
NIC = NOT INTERNALLY CONNECTED.
Figure 7. Pin Configuration
1
2
4
3
6
5
7
10
8
9
12
11
A
B
C
D
E
F
G
H
J
K
L
TOP VIEW
(Not to Scale)
Figure 8. CSP_BGA Pin Location
Rev. A| Page 13 of 60
11134-008
M
AD9671
Data Sheet
Table 6. Pin Function Descriptions
Pin No.
B5, B6, B8, C5 to C8, D5 to D8, E1, E5 to E8,
E12, F2, F4, F6, F7, F9, F11, G1, G3, G5 to G8,
G10, G12, H3 to H6, J4, K1, K2, K4, M1, M12
F1, F3, F5, F8, F10, F12, G2, G9
G4, G11
E2, E3, E4, E9, E10, E11, J6, K6
B7
L1, L12
C1
D1
A1
B1
C2
D2
A2
B2
C3
D3
A3
B3
C4
D4
A4
B4
H1
J1
H2
J2
H11
H10
H9
H8
H7
L2, M2, L3, M3, L10, M10, L11, M11
Mnemonic
GND
Description
Ground. These pins are tied to a quiet analog ground.
AVDD1
DVDD
AVDD2
CLNA
DRVDD
LO-E
LOSW-E
LI-E
LG-E
LO-F
LOSW-F
LI-F
LG-F
LO-G
LOSW-G
LI-G
LG-G
LO-H
LOSW-H
LI-H
LG-H
CLK−
CLK+
TX_TRIG−
TX_TRIG+
ADDR0
ADDR1
ADDR2
ADDR3
ADDR4
NIC
L4
M4
M5
L5
M6
L6
M7
L7
M8
L8
M9
L9
K11
J11
K12
J12
H12
SYNCINB+
SYNCINB−
SERDOUT4−
SERDOUT4+
SERDOUT3−
SERDOUT3+
SERDOUT2−
SERDOUT2+
SERDOUT1−
SERDOUT1+
SYSREF−
SYSREF+
STBY
PDWN
SCLK
SDIO
CSB
1.8 V Analog Supply.
1.4 V Digital Supply.
3.0 V Analog Supply.
LNA External Capacitor.
1.8 V Digital Output Driver Supply.
LNA Analog Inverted Output for Channel E.
LNA Analog Switched Output for Channel E.
LNA Analog Input for Channel E.
LNA Ground for Channel E.
LNA Analog Inverted Output for Channel F.
LNA Analog Switched Output for Channel F.
LNA Analog Input for Channel F.
LNA Ground for Channel F.
LNA Analog Inverted Output for Channel G.
LNA Analog Switched Output for Channel G.
LNA Analog Input for Channel G.
LNA Ground for Channel G.
LNA Analog Inverted Output for Channel H.
LNA Analog Switched Output for Channel H.
LNA Analog Input for Channel H.
LNA Ground for Channel H.
Clock Input Complement.
Clock Input True.
Transmit Trigger Complement.
Transmit Trigger True.
Chip Address Bit 0.
Chip Address Bit 1.
Chip Address Bit 2.
Chip Address Bit 3.
Chip Address Bit 4.
Not Internally Connected. These pins are not connected internally. Allow the
NIC pins to float, or connect them to ground. Avoid routing high speed
signals through these pins because noise coupling may result.
Active Low JESD204B LVDS SYNC Input—True.
Active Low JESD204B LVDS SYNC Input—Complement.
Serial Lane 4 CML Output Data—Complement.
Serial Lane 4 CML Output Data—True.
Serial Lane 3 CML Output Data—Complement.
Serial Lane 3 CML Output Data—True.
Serial Lane 2 CML Output Data—Complement.
Serial Lane 2 CML Output Data—True.
Serial Lane 1 CML Output Data—Complement.
Serial Lane 1 CML Output Data—True.
Active Low JESD204B LVDS System Reference (SYSREF) Input—Complement.
Active Low JESD204B LVDS SYSREF Input—True.
Standby Power-Down.
Full Power-Down.
Serial Clock.
Serial Data Input/Output.
Chip Select Bar.
Rev. A| Page 14 of 60
Data Sheet
Pin No.
B9
A9
D9
C9
B10
A10
D10
C10
B11
A11
D11
C11
B12
A12
D12
C12
K10
J10
K9
J9
J8
K8
K7
J7
A8
A7
A6
A5
K5
J5
K3
J3
AD9671
Mnemonic
LG-A
LI-A
LOSW-A
LO-A
LG-B
LI-B
LOSW-B
LO-B
LG-C
LI-C
LOSW-C
LO-C
LG-D
LI-D
LOSW-D
LO-D
GPO0
GPO1
GPO2
GPO3
RESET−
RESET+
MLO−
MLO+
GAIN−
GAIN+
RBIAS
VREF
CWI−
CWI+
CWQ−
CWQ+
Description
LNA Ground for Channel A.
LNA Analog Input for Channel A.
LNA Analog Switched Output for Channel A.
LNA Analog Inverted Output for Channel A.
LNA Ground for Channel B.
LNA Analog Input for Channel B.
LNA Analog Switched Output for Channel B.
LNA Analog Inverted Output for Channel B.
LNA Ground for Channel C.
LNA Analog Input for Channel C.
LNA Analog Switched Output for Channel C.
LNA Analog Inverted Output for Channel C.
LNA Ground for Channel D.
LNA Analog Input for Channel D.
LNA Analog Switched Output for Channel D.
LNA Analog Inverted Output for Channel D.
General-Purpose Open-Drain Output 0.
General-Purpose Open-Drain Output 1.
General-Purpose Open-Drain Output 2.
General-Purpose Open-Drain Output 3.
Synchronizing Input for LO Divide by M Counter Complement.
Synchronizing Input for LO Divide by M Counter True.
CW Doppler Multiple Local Oscillator Input Complement.
CW Doppler Multiple Local Oscillator Input True.
Gain Control Voltage Input Complement.
Gain Control Voltage Input True.
External Resistor to Set the Internal ADC Core Bias Current.
Voltage Reference Input/Output.
CW Doppler I Output Complement.
CW Doppler I Output True.
CW Doppler Q Output Complement.
CW Doppler Q Output True.
Rev. A| Page 15 of 60
AD9671
Data Sheet
TYPICAL PERFORMANCE CHARACTERISTICS
TGC MODE
Mode I = fSAMPLE = 40 MSPS, fIN = 5 MHz, low bandwidth mode, RS = 50 Ω, RFB = ∞ (unterminated), LNA gain = 21.6 dB, LNA bias =
midhigh, PGA gain = 27 dB, VGAIN (V) = (GAIN+) − (GAIN−) = 1.6 V, AAF LPF cutoff = fSAMPLE/3, HPF cutoff = LPF cutoff/12.00
(default), RF decimator bypassed, digital demodulator and baseband decimator bypassed, unless otherwise noted.
25
2.0
PERCENTAGE OF UNITS (%)
1.5
GAIN ERROR (dB)
1.0
0°C
0.5
0
25°C
–0.5
85°C
–1.0
20
15
10
5
–1.5
–0.4
0
0.4
0.8
1.2
1.6
VGAIN (V)
0
GAIN ERROR (dB)
Figure 9. Gain Error vs. VGAIN
Figure 12. Gain Error Histogram, VGAIN = 1.28 V
25
20
20
PERCENTAGE OF UNITS (%)
15
10
5
GAIN ERROR (dB)
10
5
0
–1.0
–0.9
–0.8
–0.7
–0.6
–0.5
–0.4
–0.3
–0.2
–0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
11134-010
–1.0
–0.9
–0.8
–0.7
–0.6
–0.5
–0.4
–0.3
–0.2
–0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0
15
CHANNEL TO CHANNEL GAIN MATCHING (dB)
Figure 10. Gain Error Histogram, VGAIN = −1.28 V
11134-013
PERCENTAGE OF UNITS (%)
11134-012
–0.8
–1.0
–0.9
–0.8
–0.7
–0.6
–0.5
–0.4
–0.3
–0.2
–0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
–1.2
11134-009
–2.0
–1.6
Figure 13. Gain Matching Histogram, VGAIN = −1.2 V
35
20
PERCENTAGE OF UNITS (%)
25
20
15
10
15
10
5
5
CHANNEL TO CHANNEL GAIN MATCHING (dB)
Figure 14. Gain Matching Histogram, VGAIN = 1.2 V
Figure 11. Gain Error Histogram, VGAIN = 0 V
Rev. A| Page 16 of 60
11134-014
11134-011
GAIN ERROR (dB)
–1.0
–0.9
–0.8
–0.7
–0.6
–0.5
–0.4
–0.3
–0.2
–0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0
0
–1.0
–0.9
–0.8
–0.7
–0.6
–0.5
–0.4
–0.3
–0.2
–0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
PERCENTAGE OF UNITS (%)
30
Data Sheet
AD9671
1.4
70
LNAGAIN = 17.9dB
66
64
1.0
SNR (dBFS)
INPUT REFERRED NOISE (nV/√Hz)
68
1.2
0.8
LNAGAIN = 15.6dB
62
60
LNAGAIN = 21.6dB
58
56
0.6
54
0.4
50
10
3
4
5
6
7
8
9
10
FREQUENCY (MHz)
30
35
40
50
55
PGAGAIN = 21dB
72
70
–136
PGAGAIN = 24dB
SNR (dBFS)
68
–138
–140
66
64
PGAGAIN = 27dB
62
60
–142
PGAGAIN = 30dB
58
–144
0
5
10
15
20
25
30
35
40
45
54
–5
AMPLITUDE (dBFS)
62
PGAGAIN = 27dB
58
PGAGAIN = 30dB
45
50
55
CHANNEL GAIN (dB)
11134-017
40
45
50
55
SPEED MODE = I (40MSPS)
LOW BANDWIDTH MODE
–6
–7
–9
35
40
–5
52
30
35
–4
–8
25
30
–3
54
20
25
–2
PGAGAIN = 24dB
15
20
–1
64
56
15
0
PGAGAIN = 21dB
60
10
Figure 19. SNR vs. Channel Gain and PGA Gain, AIN = −45 dBm
70
66
5
CHANNEL GAIN (dB)
Figure 16. Short-Circuit, Output Referred Noise vs. Channel Gain,
LNA Gain = 21.6 dB, PGA Gain = 21 dB, VGAIN = 1.6 V
68
0
11134-019
56
CHANNEL GAIN (dB)
50
10
45
74
PGA GAIN = 21dB
11134-016
OUTPUT REFERRED NOISE (dBc/√Hz)
25
Figure 18. SNR vs. Channel Gain and LNA Gain, AOUT = −1.0 dBFS
–134
–146
–5
SNR (dBFS)
20
CHANNEL GAIN (dB)
Figure 15. Short-Circuit, Input Referred Noise vs. Frequency
–132
15
11134-018
2
Figure 17. SNR vs. Channel Gain and PGA Gain, AOUT = −1.0 dBFS
–10
0
5
10
15
INPUT FREQUENCY (MHz)
Figure 20. Antialiasing Filter (AAF) Pass-Band Response,
LPF Cutoff = 1 × (1/3) × fSAMPLE, HPF = 1/12 × LPF Cutoff
Rev. A| Page 17 of 60
20
11134-020
1
11134-015
52
MIN VGAIN, AOUT = –12.0dBFS
MAX VGAIN, AOUT = –1.0dBFS
–20
–30
–40
THIRD-ORDER, MIN VGAIN
–50
THIRD-ORDER, MAX VGAIN
–60
–70
SECOND-ORDER, MIN VGAIN
–80
–90
SECOND-ORDER, MAX VGAIN
2
3
4
5
6
7
8
9
10
11
INPUT FREQUENCY (MHz)
–30
–40
–50
–60
LNAGAIN = 17.9dB
–70
LNAGAIN = 21.6dB
–80
LNAGAIN = 15.6dB
–90
–100
10
15
20
25
30
35
40
45
50
CHANNEL GAIN (dB)
VGAIN = –1.2V
–60
VGAIN = 0V
–70
–80
–90
VGAIN = +1.6V
–100
–110
–120
–40
–35
–30
–25
–20
–15
–10
–5
0
–20
–30
–40
VGAIN = –1.2V
–50
–60
VGAIN = 0V
–70
–80
–90
VGAIN = +1.6V
–100
–110
–120
–40
–35
–30
–25
–20
–15
–10
–5
0
ADC OUTPUT LEVEL (dBFS)
–100
PGAGAIN = 24dB
–110
PHASE NOISE (dBc/√Hz)
–20
–30
–40
LNAGAIN = 17.9dB
–50
–60
–10
Figure 25. Third-Order Harmonic Distortion vs. ADC Output Level (AOUT)
Figure 22. Second-Order Harmonic Distortion vs. Channel Gain,
AOUT = −1.0 dBFS
LNAGAIN = 21.6dB
LNAGAIN = 15.6dB
–70
–80
–120
–130
–140
–150
–100
10
15
20
25
30
35
40
45
CHANNEL GAIN (dB)
Figure 23. Third-Order Harmonic Distortion vs. Channel Gain,
AOUT = −1.0 dBFS
–160
100
1k
10k
OFFSET FREQUENCY FROM CARRIER (Hz)
Figure 26. TGC Path Phase Noise,
LNA Gain = 21.6 dB, PGA Gain = 27 dB, VGAIN = 0 V
Rev. A| Page 18 of 60
100k
11134-026
–90
11134-023
THIRD-ORDER HARMONIC DISTORTION (dBFS)
–50
Figure 24. Second-Order Harmonic Distortion vs. ADC Output Level (AOUT)
THIRD-ORDER HARMONIC DISTORTION (dBFS)
PGAGAIN = 24dB
–20
–10
–40
0
0
–10
0
–30
ADC OUTPUT LEVEL (dBFS)
11134-022
SECOND-ORDER HARMONIC DISTORTION (dBFS)
Figure 21. Second-Order and Third-Order Harmonic Distortion vs. Input
Frequency
–20
11134-025
–100
11134-021
HARMONIC DISTORTION (dBFS)
–10
0
–10
11134-024
0
Data Sheet
SECOND-ORDER HARMONIC DISTORTION (dBFS)
AD9671
AD9671
0
8
7
6
5
4
3
2
1
0
100k
–10
–20
fIN1 = 5.0MHz
fIN2 = 5.01MHz
FUND1 LEVEL = –1dBFS
FUND2 LEVEL = –21dBFS
–30
1M
10M
FREQUENCY (Hz)
0
–10
–20
–30
–40
–50
–60
–70
–80
–90
100k
IMD3 (dBFS)
–40
100M
–50
–60
VGAIN = –1.2V
–70
–80
–90
–100
VGAIN = +1.6V
–110
10M
FREQUENCY (Hz)
100M
VGAIN = 0V
–120
–40
–35
–20
–10
–5
0
7
FUND1 LEVEL = –1dBFS
FUND2 LEVEL = –21dBFS
RS = 50Ω
6
NOISE FIGURE (dB)
–30
–40
–50
–60
–70
RIN = 1000Ω
–80
5
4
3
2
–90
RIN = 50Ω
20
25
30
35
RIN = 300Ω
40
CHANNEL GAIN (dB)
45
50
11134-028
IMD3 (dBFS)
–15
Figure 29. IMD3 vs. ADC Output Level
fIN1 = 2.3MHz
fIN2 = 2.31MHz
–100
15
–20
Figure 28. IMD3 vs. Channel Gain
1
0
2
4
6
8
10
12
14
16
18
20
FREQUENCY (MHz)
Figure 30. Noise Figure vs. Frequency
RS = RIN = 100 Ω, LNA Gain = 17.9 dB, PGA Gain = 30 dB, VGAIN = 1.6 V
Rev. A| Page 19 of 60
11134-031
0
–25
ADC OUTPUT LEVEL (dBFS)
Figure 27. LNA Input Impedance Magnitude and Phase, Unterminated
–10
–30
11134-029
1M
11134-027
PHASE (Degrees)
MAGNITUDE (kΩ)
Data Sheet
AD9671
Data Sheet
CW DOPPLER MODE
fIN = 5 MHz, fLO = 20 MHz, 4LO mode, RS = 50 Ω, LNA gain = 21.6 dB, LNA bias = midhigh, all CW channels enabled, phase rotation = 0°.
10
165
9
160
155
SNR (dBc/√Hz)
7
6
5
4
3
150
145
140
2
0
0
1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
BASEBAND FREQUENCY (Hz)
Figure 31. Noise Figure vs. Baseband Frequency
130
0
1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
BASEBAND FREQUENCY (Hz)
Figure 32. Output Referred SNR vs. Baseband Frequency
Rev. A| Page 20 of 60
11134-033
135
1
11134-032
NOISE FIGURE (dB)
8
Data Sheet
AD9671
THEORY OF OPERATION
MLO–
RESET+
RESET–
RFB1
RFB2
T/R
SWITCH C
S
LO
GENERATION
LO-x
CWI+
CWI–
LOSW-x
LI-x
LG-x
CSH
CLG
TRANSDUCER
CWQ+
CWQ–
ATTENUATOR
–45dB TO 0dB
LNA
15.6dB,
17.9dB,
21.6dB
POST
AMP
PIPELINE
ADC
21dB,
24dB,
27dB,
30dB
GAIN
INTERPOLATOR
gm
FILTER
DEMOD/
DEC
SERIAL
CML
SERDOUT1+
TO SERDOUT4+
SERDOUT1–
TO SERDOUT4–
SYSREF+
NCO
SYSREF–
SYNCINB+
SYNCINB–
GAIN+
TX_TRIG+ TX_TRIG–
GAIN–
11134-034
MLO+
Figure 33. Simplified Block Diagram of a Single Channel
Each channel in the AD9671 contains both a TGC signal path and
a CW Doppler signal path. Common to both signal paths, the
LNA provides four user adjustable input impedance termination
options for matching different probe impedances. The CW
Doppler path includes an I/Q demodulator with programmable
phase rotation needed for analog beamforming. The TGC path
includes a differential X-AMP® VGA, an AAF, an ADC, and a
digital demodulator and decimator. Figure 33 shows a simplified
block diagram with external components.
TGC OPERATION
The system gain for TGC operation is distributed as listed in
Table 7.
In addition to the analog VGA attenuation described in Equation 2,
the attenuation level can be digitally controlled in 3.5 dB increments.
Equation 3 is still valid, and the value of VGAATT is equal to the
attenuation level set in Address 0x011, Bits[7:4].
Table 7. Channel Analog Gain Distribution
Section
LNA
Attenuator
VGA Amplifier
Filter
ADC
Nominal Gain (dB)
15.6/17.9/21.6 (LNAGAIN)
−45 to 0 (VGAATT)
21/24/27/30 (PGAGAIN)
0
0
Low Noise Amplifier (LNA)
Each LNA output is dc-coupled to a VGA input. The VGA
consists of an attenuator with a range of −45 dB to 0 dB followed
by an amplifier with 21 dB, 24 dB, 27 dB, or 30 dB of gain. The
X-AMP gain interpolation technique results in low gain error
and uniform bandwidth, and differential signal paths minimize
distortion.
The linear in dB gain (law conformance) range of the TGC path is
45 dB. The slope of the gain control interface is 14 dB/V, and the
gain control range is −1.6 V to +1.6 V. Equation 1 is the expression
for the differential voltage, VGAIN, at the gain control interface.
Equation 2 is the expression for the VGA attenuation, VGAATT,
as a function of VGAIN.
VGAIN (V) = (GAIN+) − (GAIN−)
(1)
VGAATT (dB) = −14 (dB/V) × (1.6 − VGAIN)
(2)
Then calculate the total channel gain as in Equation 3.
ChannelGain (dB) = LNAGAIN + VGAATT + PGAGAIN
In its default condition, the LNA has a gain of 21.6 dB (12×),
and the VGA postamplifier gain is 24 dB. If the voltage on the
GAIN+ pin is 0 V and the voltage on the GAIN− pin is 1.6 V
(44.8 dB attenuation), the total gain of the channel is 0.8 dB if
the LNA input is unmatched. The channel gain is −5.2 dB if the
LNA is matched to 50 Ω (RFB = 300 Ω). However, if the voltage on
the GAIN+ pin is 1.6 V and the voltage on the GAIN− pin is 0 V
(0 dB attenuation), VGAATT is 0 dB. This results in a total gain of
45.6 dB through the TGC path if the LNA input is unmatched, or
in a total gain of 39.6 dB if the LNA input is matched. Similarly,
if the LNA input is unmatched and has a gain of 21.6 dB (12×),
and the VGA postamp gain is 30 dB, the channel gain is
approximately 52 dB with 0 dB VGAATT.
(3)
Good system sensitivity relies on a proprietary ultralow noise
LNA at the beginning of the signal chain, which minimizes the
noise contribution in the following VGA. Active impedance
control optimizes noise performance for applications that
benefit from input impedance matching.
The LNA input, LI-x, is capacitively coupled to the source.
An on-chip bias generator establishes dc input bias voltages of
approximately 2.2 V and centers the output common-mode
levels at 1.5 V (AVDD2 divided by 2). A capacitor, CLG, of the
same value as the input coupling capacitor, CS, is connected
from the LG-x pin to ground.
The LNA supports three gains, 21.6 dB, 17.9 dB, or 15.6 dB, set
through the SPI. Overload protection ensures quick recovery
time from large input voltages.
Low value feedback resistors and the current driving capability
of the output stage allow the LNA to achieve a low input
referred noise voltage of 0.78 nV/√Hz (at a gain of 21.6 dB).
On-chip resistor matching results in precise single-ended gains,
Rev. A| Page 21 of 60
AD9671
Data Sheet
The LNA consists of a single-ended voltage gain amplifier with
differential outputs. The negative output is externally available
on two output pins, LO-x and LOSW-x, that are controlled via
internal switches. This configuration allows the active input
impedance synthesis of three different impedance values (and
an unterminated value) by connecting up to two external
resistances in parallel and controlling the internal switch states
via the SPI. For example, with a fixed gain of 8× (17.9 dB), an
active input termination is synthesized by connecting a feedback resistor between the negative output pin, LO-x, and the
positive input pin, LI-x. This well known technique is used for
interfacing multiple probe impedances to a single system. The
input resistance (RIN) calculation is shown in Equation 4.
RIN 
(RFB1  20 ) || (RFB2  20 )  30 
A
(1  )
2
(4)
where:
RFB1 and RFB2 are the external feedback resistors.
20 Ω is the internal switch on resistance.
30 Ω is an internal series resistance common to the two internal
switches.
A/2 is the single-ended gain or the gain from the LI-x inputs to
the LO-x outputs.
RFB can be equal to RFB1, RFB2, or (RFB1 + 20 Ω)||(RFB2 + 20 Ω)
depending on the connection status of the internal switches.
Because the amplifier has a gain of 8× from its input to its
differential output, it is important to note that the gain, A/2,
is the gain from Pin LI-x to Pin LO-x and that it is 6 dB less
than the gain of the amplifier, or 12.1 dB (4×). The input
resistance is reduced by an internal bias resistor of 6 kΩ in
parallel with the source resistance connected to Pin LI-x, with
Pin LG-x ac grounded. Use the more accurate Equation 5 to
calculate the required RFB for a desired RIN, even for higher
values of RIN.
R IN
(R  20 ) || (RFB 2  20 )  30 
 FB1
|| 6 k 
A
(1  )
2
(5)
Table 8. Active Termination Example for LNA Gain = 21.6 dB,
RFB1 = 650 Ω, RFB2 = 1350 Ω
Addr. 0x02C
Value
00 (default)
01
10
11
1
RS (Ω)
100
50
200
N/A1
LO-x
Switch
On
On
Off
Off
LOSW-x
Switch
Off
On
On
Off
RFB (Ω)
RFB1
RFB1||RFB2
RFB2
∞
RIN (Ω)
(Eq. 4)
100
66
200
∞
N/A means not applicable.
The bandwidth (BW) of the LNA is greater than 80 MHz.
Ultimately, the BW of the LNA limits the accuracy of the
synthesized RIN. For RIN = RS up to about 200 Ω, the best match
is between 100 kHz and 10 MHz, where the lower frequency
limit is determined by the size of the ac coupling capacitors, and
the upper limit is determined by the LNA BW. Furthermore, the
input capacitance and RS limit the BW at higher frequencies.
Figure 34 shows RIN vs. frequency for various values of RFB.
1k
RS = 500Ω, RFB = 2kΩ
RS = 200Ω, RFB = 800Ω
100
RS = 100Ω, RFB = 400Ω, CSH = 20pF
RS = 50Ω, RFB = 200Ω, CSH = 70pF
10
100k
1M
10M
FREQUENCY (Hz)
100M
11134-035
Active Impedance Matching
RFB is the resulting impedance of the RFB1 and RFB2 combination
(see Figure 33). Use Register 0x02C in the SPI memory to
program the AD9671 for four impedance matching options:
three active terminations and unterminated. Table 8 shows an
example of how to select RFB1 and RFB2 for 66 Ω, 100 Ω, and 200 Ω
input impedance for LNA gain = 21.6 dB (12×).
INPUT RESISTANCE (Ω)
which are critical for accurate impedance control. The use of a
fully differential topology and negative feedback minimizes
distortion. Low second-order harmonic distortion is particularly
important in harmonic ultrasound imaging applications.
Figure 34. RIN vs. Frequency for Various Values of RFB
(Effects of RSH and CSH Are Also Shown)
However, for larger RIN values, parasitic capacitance starts
rolling off the signal BW before the LNA can produce peaking.
CSH further degrades the match; therefore, do not use CSH for
values of RIN that are greater than 100 Ω.
For example, to set RIN to 200 Ω with a single-ended LNA gain of
12.1 dB (4×), the value of RFB1 from Equation 1 must be 950 Ω
while the switch for RFB2 is open. If the more accurate equation
(Equation 5) is used to calculate RIN, the value is then 194 Ω
instead of 200 Ω, resulting in a gain error of less than 0.27 dB.
Some factors, such as the presence of a dynamic source resistance,
may influence the absolute gain accuracy more significantly. At
higher frequencies, the input capacitance of the LNA must be
considered. The user must determine the level of matching
accuracy and adjust RFB accordingly.
Rev. A| Page 22 of 60
Data Sheet
AD9671
Table 9 lists the recommended values for RFB and CSH in terms
of RIN. CFB is needed in series with RFB because the dc levels at
Pin LO-x and Pin LI-x are unequal.
Figure 36 shows the noise figure as it relates to RS for various
values of RIN, which is helpful for design purposes.
8
Table 9. Active Termination External Component Values
7
LNA Gain (dB)
15.6
17.9
21.6
15.6
17.9
21.6
15.6
17.9
21.6
6
Minimum CSH (pF)
90
70
50
30
20
10
Not applicable
Not applicable
Not applicable
4
3
2
0
The short-circuit noise voltage (input referred noise) is an important
limit on system performance. The short-circuit noise voltage for
the LNA is 0.78 nV/√Hz at a gain of 21.6 dB, including the VGA
noise at a VGA postamp gain of 27 dB. These measurements,
which were taken without a feedback resistor, provide the basis for
calculating the input noise and noise figure (NF) performance.
Figure 35 and Figure 36 are simulations of noise figure vs. RS results
with different input configurations and an input referred noise
voltage of 2.5 nV/√Hz for the VGA. Unterminated (RFB = ∞)
operation exhibits the lowest equivalent input noise and noise
figure. Figure 36 shows the noise figure vs. source resistance rising
at low RS, where the LNA voltage noise is large compared with the
source noise, and at high RS due to the noise contribution from RFB.
The lowest NF is achieved when RS matches RIN.
10
100
RS (Ω)
1k
11134-037
1
LNA Noise
Figure 36. Noise Figure vs. RS for Various Fixed Values of RIN,
Active Termination Matched Inputs, VGAIN = 1.6 V
CLNA Connection
CLNA (Pin B7) must have a 1 nF capacitor attached to AVDD2.
DC Offset Correction/High-Pass Filter
The AD9671 LNA architecture is designed to correct for dc offset
voltages that can develop on the external CS capacitor due to
leakage of the Tx/Rx switch during ultrasound transmit cycles.
The dc offset correction, as shown in Figure 37, provides a
feedback mechanism to the LG-x input of the LNA to correct
for this dc voltage.
Figure 35 shows the relative noise figure performance. With an LNA
gain of 21.6 dB, the input impedance is swept with RS to preserve
the match at each point. The noise figures for a source impedance
of 50 Ω are 7 dB, 4 dB, and 2.5 dB for the shunt termination,
active termination, and unterminated configurations, respectively.
The noise figures for 200 Ω are 4.5 dB, 1.7 dB, and 1 dB,
respectively.
AD9671
RFB1
LO-x
RFB2
LOSW-x
Tx/Rx
SWITCH C
S
LI-x
LNA
LG-x
CSH
15.6dB,
17.9dB,
21.6dB
CLG
TRANSDUCER
gm
DC OFFSET
CORRECTION
12.0
10.5
Figure 37. Simplified LNA Input Configuration
9.0
The feedback acts as high-pass filter providing dynamic
correction of the dc offset. The cutoff frequency of the highpass filter response is dependent on the value of the CLG capacitor,
the gain of the LNA (LNAGAIN) and the trandsconductance ( gm) of
the feedback transconductance amplifier. The gm value is
programmed in Register 0x120, Bits[4:3]. Ensure that CS is
equal to CLG for proper operation.
7.5
SHUNT TERMINATION
6.0
4.5
3.0
1.5
0
ACTIVE TERMINATION
UNTERMINATED
10
100
RS (Ω)
1k
11134-036
NOISE FIGURE (dB)
5
11134-038
RFB (Ω)
150
200
300
350
450
650
750
950
1350
NOISE FIGURE (dB)
RIN (Ω)
50
50
50
100
100
100
200
200
200
RIN = 50Ω
RIN = 75Ω
RIN = 100Ω
RIN = 200Ω
UNTERMINATED
Figure 35. Noise Figure vs. RS for Shunt Termination, Active
Termination Matched and Unterminated Inputs, VGAIN = 1.6 V
Rev. A| Page 23 of 60
AD9671
Data Sheet
249Ω
Address
0x120[4:3]
00 (default)
01
10
11
gm
(mS)
0.5
1.0
1.5
2.0
LNAGAIN =
15.6 dB
41 kHz
83 kHz
133 kHz
167 kHz
LNAGAIN =
17.9 dB
55 kHz
110 kHz
178 kHz
220 kHz
LNAGAIN =
21.6 dB
83 kHz
167 kHz
267 kHz
330 kHz
AD9671
GAIN+
±0.8V DC
100Ω AT 0.8V CM
0.01µF
249Ω
ADA4938-x1
100Ω
GAIN–
0.01µF
31.3kΩ
±1.6V
0.8V CM
249Ω
10kΩ
±0.8V DC
AT 0.8V CM
249Ω
1ADA4938-x REFERS TO THE ADA4938-1 AND ADA4938-2 DEVICES
11134-039
Table 10. High-Pass Filter Cutoff Frequency, fHP, for CLG = 10 nF
Figure 38. Differential GAIN± Pin Configuration
For other values of CLG, determine the high-pass filter cutoff
frequency by scaling the values from Table 10 or calculating
based on CLG, LNAGAIN, and gm, as shown in Equation 6.
f HP (C LG ) 
10 nF
1
g
 LNA GAIN  m  f HP 
2 
C LG
C LG
Use Address 0x011, Bits[7:4], to disable the analog gain control
and to control the attenuator digitally. The control range is
45 dB and the step size is 3.5 dB.
(6)
where fHP is the high-pass filter cutoff frequency (see Table 10).
Variable Gain Amplifier (VGA)
The differential X-AMP VGA provides precise input attenuation and interpolation. It has a low input referred noise of
2.5 nV/√Hz and excellent gain linearity. The VGA is driven by
a fully differential input signal from the LNA. The X-AMP architecture produces a linear in dB gain law conformance and low
distortion levels—deviating only ±0.5 dB or less from the ideal.
The gain slope is monotonic with respect to the control voltage
and is stable with variations in process, temperature, and supply.
The resulting total gain range is 45 dB, which allows for range
loss at the endpoints.
The X-AMP inputs are part of a PGA that completes the VGA.
The PGA in the VGA can be programmed to a gain of 21 dB, 24
dB, 27 dB, or 30 dB, allowing optimization of channel gain for
different imaging modes in the ultrasound system. The VGA
bandwidth is greater than 100 MHz. The input stage is designed
to ensure excellent frequency response uniformity across the gain
setting. For TGC mode, this input stage minimizes time delay
variation across the gain range.
VGA Noise
In a typical application, a VGA compresses a wide dynamic
range input signal to within the input span of an ADC. The
input referred noise of the LNA limits the minimum resolvable
input signal, whereas the output referred noise, which depends
primarily on the VGA, limits the maximum instantaneous dynamic
range that can be processed at any one particular gain control
voltage. This latter limit is set in accordance with the total noise
floor of the ADC.
The output referred noise is a flat 40 nV/√Hz (postamp gain =
24 dB) over most of the gain range because it is dominated by
the fixed output referred noise of the VGA. At the high end of the
gain control range, the noise of the LNA and the source prevail.
The input referred noise reaches its minimum value near the
maximum gain control voltage, where the input referred
contribution of the VGA is miniscule.
Gain Control
At lower gains, the input referred noise and, therefore, the noise
figure increase as the gain decreases. The instantaneous dynamic
range of the system is not lost, however, because the input capacity
increases as the input referred noise increases. The contribution of
the ADC noise floor has the same dependence. The important
relationship is the magnitude of the VGA output noise floor
relative to that of the ADC.
The analog gain control interface, GAIN±, is a differential
input. VGAIN varies the gain of all VGAs through the interpolator
by selecting the appropriate input stages connected to the input
attenuator. The nominal VGAIN range is 14 dB/V from −1.6 V to
+1.6 V, with the best gain linearity from approximately −1.44 V
to +1.44 V, where the error is typically less than ±0.5 dB. For
VGAIN voltages of greater than 1.44 V and less than −1.44 V, the
error increases. The value of GAIN± can exceed the supply
voltage by 1 V without gain foldover.
Gain control noise is a concern in very low noise applications.
Thermal noise in the gain control interface can modulate the
channel gain. The resulting noise is proportional to the output
signal level and is usually evident only when a large signal is
present. Take care to minimize noise impinging at the GAIN±
inputs. Use an external RC filter to remove VGAIN source noise.
Ensure that the filter bandwidth is sufficient to accommodate the
desired control bandwidth and attenuate unwanted switching noise
from the external DACs used to drive the gain control.
Gain control response time is typically 750 ns to settle within 10%
of the final value for a change from minimum to maximum gain.
The AD9671 can bypass the GAIN± inputs and control the gain
of the attenuator digitally (see the Gain Control section). This
mode removes any external noise contributions when active gain
control is not needed.
The differential input pins, GAIN+ and GAIN−, can interface
to an amplifier, as shown in Figure 38. Decouple and drive the
GAIN+ and GAIN− pins to accommodate a 3.2 V full-scale input.
Rev. A| Page 24 of 60
Data Sheet
AD9671
Antialiasing Filter (AAF)
The filter that the signal reaches prior to the ADC is used to
reject dc signals and to band limit the signal for antialiasing.
The antialiasing filter is a combination of a single-pole, high-pass
filter and a second-order, low-pass filter. Configure the highpass filter as a ratio of the low-pass filter cutoff frequency using
Address 0x02B, Bits[1:0].
The filter uses on-chip tuning to trim the capacitors and, in turn, to
set the desired low-pass cutoff frequency and reduce variations.
The default −3 dB low-pass filter cutoff is 1/3, 1/4.5, or 1/6 of the
ADC sample clock rate. The cutoff can be scaled to 0.75, 0.8, 0.9,
1.0, 1.13, 1.25, or 1.45 times this frequency using Address 0x00F.
The cutoff tolerance (±10%) is maintained from 8 MHz to 18 MHz
for low bandwidth mode or 13.5 MHz to 30 MHz for high
bandwidth mode.
Table 11 and Table 12 calculate the valid SPI-selectable low-pass
filter settings and expected cutoff frequencies for the low bandwidth and high bandwidth modes at the minimum sample
frequency and the maximum sample frequency in each speed
mode.
Tuning is normally off to avoid changing the capacitor settings
during critical times. The tuning circuit is enabled through the
SPI. It is disabled automatically after 512 cycles of the ADC sample
clock. Initialize the tuning of the filter after initial power-up and
after reprogramming of the filter cutoff scaling or the ADC
sample rate. The tuning is initiated using Address 0x02B, Bit 6.
Four SPI-programmable settings allow users to vary the highpass filter cutoff frequency as a function of the low-pass cutoff
frequency. Two examples are shown in Table 13: an 8 MHz lowpass cutoff frequency and an 18 MHz low-pass cutoff frequency. In
both cases, as the ratio decreases, the amount of rejection on the
low end frequencies increases. Therefore, making the entire AAF
frequency pass band narrow can reduce low frequency noise or
maximize dynamic range for harmonic processing.
Table 11. SPI-Selectable Low-Pass Filter Cutoff Options for Low Bandwidth Mode at Example Sampling Frequencies
Address
0x00F,
Bits[7:3]
0 0000
LPF Cutoff
Frequency (MHz)
1.45 × (1/3) × fSAMPLE
Sampling Frequency (MHz)
20.5
9.91
0 0001
1.25 × (1/3) × fSAMPLE
8.54
0 0010
1.13 × (1/3) × fSAMPLE
0 0011
1.0 × (1/3) × fSAMPLE
0 0100
0.9 × (1/3) × fSAMPLE
0 0101
0.8 × (1/3) × fSAMPLE
0 0110
0.75 × (1/3) × fSAMPLE
0 1000
1.45 × (1/4.5) × fSAMPLE
0 1001
1.25 × (1/4.5) × fSAMPLE
0 1010
1.13 × (1/4.5) × fSAMPLE
0 1011
1.0 × (1/4.5) × fSAMPLE
0 1100
0.9 × (1/4.5) × fSAMPLE
0 1101
0.8 × (1/4.5) × fSAMPLE
0 1110
0.75 × (1/4.5) × fSAMPLE
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
40
Out of tunable filter
range
16.67
10.67
65
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
17.33
10.00
16.25
12.89
20.94
11.11
18.06
10.00
16.25
8.89
14.44
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
17.78
8.00
13.00
16.00
Out of tunable filter 11.56
range
Out of tunable filter 10.83
range
14.22
15.00
13.33
12.00
Rev. A| Page 25 of 60
80
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
16.82
13.33
125
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
17.50
AD9671
Address
0x00F,
Bits[7:3]
1 0000
LPF Cutoff
Frequency (MHz)
1.45 × (1/6) × fSAMPLE
1 0001
1.25 × (1/6) × fSAMPLE
1 0010
1.13 × (1/6) × fSAMPLE
1 0011
1.0 × (1/6) × fSAMPLE
1 0100
0.9 × (1/6) × fSAMPLE
1 0101
0.8 × (1/6) × fSAMPLE
1 0110
0.75 × (1/6) × fSAMPLE
Data Sheet
Sampling Frequency (MHz)
20.5
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
40
9.67
65
15.71
8.33
13.54
80
Out of tunable filter
range
16.67
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
12.19
15.00
10.83
13.33
9.75
12.00
8.67
10.67
125
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
16.67
8.13
10.00
15.63
Table 12. SPI-Selectable Low-Pass Filter Cutoff Options for High Bandwidth Mode at Example Sampling Frequencies
Address
0x00F,
Bits[7:3]
0 0000
LPF Cutoff
Frequency (MHz)
1.45 × (1/3) × fSAMPLE
Sampling Frequency (MHz)
0 0001
1.25 × (1/3) × fSAMPLE
0 0010
1.13 × (1/3) × fSAMPLE
0 0011
1.0 × (1/3) × fSAMPLE
0 0100
0.9 × (1/3) × fSAMPLE
0 0101
0.8 × (1/3) × fSAMPLE
0 0110
0.75 × (1/3) × fSAMPLE
0 1000
1.45 × (1/4.5) × fSAMPLE
0 1001
1.25 × (1/4.5) × fSAMPLE
0 1010
1.13 × (1/4.5) × fSAMPLE
0 1011
1.0 × (1/4.5) × fSAMPLE
0 1100
0.9 × (1/4.5) × fSAMPLE
0 1101
0.8 × (1/4.5) × fSAMPLE
0 1110
0.75 × (1/4.5) × fSAMPLE
20.5
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
40
19.33
16.67
65
Out of tunable filter
range
27.08
15.00
24.38
80
Out of tunable filter
range
Out of tunable filter
range
30.00
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
21.67
26.67
19.50
24.00
17.33
21.33
16.25
20.00
20.94
25.78
18.06
22.22
16.25
20.00
14.44
17.78
125
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
27.78
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
16.00
25.00
14.22
22.22
Out of tunable filter
range
20.83
Rev. A| Page 26 of 60
Data Sheet
Address
0x00F,
Bits[7:3]
1 0000
LPF Cutoff
Frequency (MHz)
1.45 × (1/6) × fSAMPLE
1 0001
1.25 × (1/6) × fSAMPLE
1 0010
1.13 × (1/6) × fSAMPLE
1 0011
1.0 × (1/6) × fSAMPLE
1 0100
0.9 × (1/6) × fSAMPLE
1 0101
0.8 × (1/6) × fSAMPLE
1 0110
0.75 × (1/6) × fSAMPLE
AD9671
Sampling Frequency (MHz)
20.5
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
40
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable
filter range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
65
15.71
80
19.33
13.54
16.67
125
Out of tunable filter
range
26.04
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
15.00
23.44
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
Out of tunable filter
range
20.83
18.75
16.67
15.63
Table 13. High-Pass Filter Cutoff Options
Address 0x02B[1:0] High-Pass Filter Cutoff
00 (default)
01
10
11
Ratio = low-pass filter cutoff frequency/high-pass filter cutoff frequency.
For debug and testing, there is a bypass switch to view the AAF
output on the GPO2 and GPO3 pins. Enable this mode via SPI
Address 0x109, Bit 4. The differential AAF output of only one
channel can be accessed at a time. The dc output voltage is 1.5 V
(or AVDD2/2) and the maximum ac output voltage is 2 V p-p.
into the AD9671 to approximately 0.8 V p-p differential. This
limit prevents the large voltage swings of the clock from feeding
through to other portions of the AD9671, and it preserves the
fast rise and fall times of the signal, which are critical to low
jitter performance.
3.3V
0.1µF
ADC
OUT
The AD9671 uses a pipelined ADC architecture. The quantized
output from each stage is combined into a 14-bit result in the
digital correction logic. The pipelined architecture permits the
first stage to operate on a new input sample and the remaining
stages to operate on preceding samples. Sampling occurs on the
rising edge of the clock.
The output staging block aligns the data, corrects errors, and
passes the data to the output buffers. The data is then serialized
and aligned to the frame and output clocks.
Clock Input Considerations
For optimum performance, clock the AD9671 sample clock
inputs (CLK+ and CLK−) with a differential signal. This signal
is typically ac-coupled into the CLK+ and CLK− pins via a
transformer or capacitors. These pins are biased internally and
require no additional bias.
Figure 39 shows the preferred method for clocking the AD9671.
A low jitter clock source, such as the Valpey Fisher oscillator,
VFAC3AHL-1 80.000, is converted from single-ended to
differential using an RF transformer. The back to back Schottky
diodes across the secondary transformer limit clock excursions
VFAC3
50Ω
MINI-CIRCUITS®
ADT1-1WT, 1:1Z
0.1µF
XFMR
CLK+
100Ω
ADC
0.1µF
CLK–
0.1µF
SCHOTTKY
DIODES:
HSM2812
11134-040
AAF/VGA Test Mode
Figure 39. Transformer-Coupled Differential Clock
If a low jitter clock is available, another option is to ac couple a
differential positive emitter-coupled logic (PECL) signal to the
sample clock input pins, as shown in Figure 40. Analog Devices,
Inc., offers a family of clock drivers with excellent jitter performance, including the AD9516-0, AD9516-1, AD9516-2, AD9516-3,
and AD9516-5 (these five devices are represented by AD9516-x in
Figure 40, Figure 41, and Figure 42), as well as the AD9524.
3.3V
AD9516-x OR AD9524
VFAC3
0.1µF
0.1µF
CLK+
CLK
OUT
50Ω*
0.1µF
100Ω
PECL DRIVER
0.1µF
CLK
240Ω
240Ω
*50Ω RESISTOR IS OPTIONAL.
Rev. A| Page 27 of 60
ADC
CLK–
Figure 40. Differential PECL Sample Clock
11134-041
1
Ratio
12.00
9.00
6.00
3.00
High-Pass Cutoff Frequency
Low-Pass Cutoff = 8 MHz
Low-Pass Cutoff = 18 MHz
670 kHz
1.5 MHz
890 kHz
2.0 MHz
1.33 MHz
3.0 MHz
2.67 MHz
6.0 MHz
1
AD9671
Data Sheet
Treat the clock input as an analog signal in cases where aperture
jitter may affect the dynamic range of the AD9671. Separate power
supplies for clock drivers from the ADC output driver supplies
to avoid modulating the clock signal with digital noise. Low jitter,
crystal controlled oscillators make the best clock sources, such
as the Valpey Fisher VFAC3 series. If the clock is generated from
another type of source (by gating, dividing, or other methods), it is
retimed by the original clock during the last step.
A third option is to ac couple a differential LVDS signal to the
sample clock input pins, as shown in Figure 41.
3.3V
AD9516-x OR AD9524
0.1µF
0.1µF
CLK+
CLK
OUT
50Ω*
0.1µF
ADC
100Ω
0.1µF
LVDS DRIVER
CLK–
11134-042
CLK
*50Ω RESISTOR IS OPTIONAL.
For more information on how jitter performance relates to
ADCs, refer to the AN-501 Application Note and the AN-756
Application Note.
Figure 41. Differential LVDS Sample Clock
130
In some applications, it is acceptable to drive the sample clock
inputs with a single-ended CMOS signal. In such applications,
drive CLK+ directly from a CMOS gate, and bypass the CLK−
pin to ground with a 0.1 μF capacitor (see Figure 42).
110
AD9516-x OR AD9524
0.1µF
CLK
50Ω*
SNR (dB)
3.3V
VFAC3
OUT
CMOS DRIVER
OPTIONAL
0.1µF
100Ω
CLK+
100
16 BITS
90
14 BITS
80
50
0.1µF
12 BITS
70
60
ADC
CLK
RMS CLOCK JITTER REQUIREMENT
120
10 BITS
8 BITS
CLK–
40
11134-043
0.1µF
*50Ω RESISTOR IS OPTIONAL.
30
Figure 42. Single-Ended 1.8 V CMOS Sample Clock
Typical high speed ADCs use both clock edges to generate a
variety of internal timing signals. As a result, these ADCs can be
sensitive to the clock duty cycle. Commonly, a 5% tolerance is
required on the clock duty cycle to maintain dynamic performance
characteristics. The AD9671 contains a duty cycle stabilizer (DCS)
that retimes the nonsampling edge, providing an internal clock
signal with a nominal 50% duty cycle. This DCS allows a wide
range of clock input duty cycles without affecting the performance
of the AD9671. When the DCS is on, noise and distortion performance are nearly flat for a wide range of duty cycles. However,
some applications may require the DCS function to be off. When
the DCS function is off, the dynamic range performance can be
affected.
The duty cycle stabilizer uses a delay-locked loop (DLL) to create
the nonsampling edge. As a result, any changes to the sampling
frequency require approximately eight clock cycles to allow the
DLL to acquire and lock to the new rate.
Clock Jitter Considerations
High speed, high resolution ADCs are sensitive to the quality of the
clock input. Calculate the degradation in SNR at a given input
frequency (fA) due only to aperture jitter (tJ) as follows:
In this equation, the rms aperture jitter represents the root
mean square of all jitter sources, including the clock input,
analog input signal, and ADC aperture jitter (see Figure 43).
10
100
ANALOG INPUT FREQUENCY (MHz)
1000
Figure 43. Ideal SNR vs. Input Frequency and Jitter
Clock Duty Cycle Considerations
SNR Degradation = 20 × log 10(1/2 × π × fA × tJ)
1
0.125ps
0.25ps
0.5ps
1.0ps
2.0ps
11134-044
VFAC3
(7)
Power Dissipation and Power-Down Mode
The power dissipated by the AD9671 is proportional to its
sample rate. The digital power dissipation does not vary
significantly because it is determined primarily by the DRVDD
supply and the bias current of the LVDS output drivers. The
AD9671 features scalable LNA bias currents (see Table 33,
Address 0x012). The default LNA bias current settings are
midhigh.
By asserting the PDWN pin high, the AD9671 is placed into
power-down mode. In this state, the device typically dissipates
5 mW. During power-down, the LVDS output drivers are placed
into a high impedance state. The AD9671 returns to normal
operating mode when the PDWN pin is pulled low. This pin is
only 1.8 V tolerant. To drive the PDWN pin from a 3.3 V logic
level, insert a 1 kΩ resistor in series with this pin to limit the
current.
By asserting the STBY pin high, the AD9671 is placed in
standby mode. In this state, the device typically dissipates
725 mW. During standby, the entire device is powered down
except the internal references. The LVDS output drivers are
placed into a high impedance state. This mode is well suited for
applications that require power savings because it allows the
device to be powered down when not in use and then quickly
powers up. The time to power up the device is also greatly
reduced. The AD9671 returns to normal operating mode when
the STBY pin is pulled low. This pin is only 1.8 V tolerant. To
Rev. A| Page 28 of 60
Data Sheet
AD9671
In power-down mode, low power dissipation is achieved by
shutting down the reference, reference buffer, PLL, and biasing
networks. The decoupling capacitors on VREF are discharged
when entering power-down mode and must be recharged when
returning to normal operation. As a result, the wake-up time is
related to the time spent in power-down mode: shorter cycles
result in proportionally shorter wake-up times. To restore the
device to full operation, approximately 375 μs is required when
using the recommended 1 μF and 0.1 μF decoupling capacitors
on the VREF pin and the 0.01 μF decoupling capacitors on the
GAIN± pins. Most of this time is dependent on gain decoupling;
higher value decoupling capacitors on the GAIN± pins result in
longer wake-up times.
A number of other power-down options are available when using
the SPI port interface. The user can individually power down
each channel or place the entire device into standby mode. When
fast wake-up times are required, standby mode allows the user
to keep the internal PLL powered up. The wake-up time is slightly
dependent on gain. To achieve a 2 μs wake-up time when the
device is in standby mode, apply 0.8 V to the GAIN± pins.
Power and Ground Connection Recommendations
When connecting power to the AD9671, use two separate 1.8 V
supplies: one for analog (AVDD1) and one for digital (DRVDD).
If only one 1.8 V supply is available, route it to the AVDD1 pin
first and then tap it off and isolate it with a ferrite bead or a
filter choke preceded by decoupling capacitors for the DRVDD
pin.
If the user does not use the digital demodulator/decimator
functions for post ADC processing, the DVDD pin can be tied
to the 1.8 V DRVDD supply. When this is done, route the DVDD
supply first, tap it off, and isolate it with a ferrite bead or filter
choke preceded by decoupling capacitors for the DRVDD pin. It
is not recommended to use the same supply for AVDD1, DVDD,
and DRVDD.
For both high and low frequencies, use several decoupling
capacitors on all supplies. Place these capacitors near the point
of entry at the PCB level and near the device, with minimal trace
lengths.
When using the AD9671, a single PCB ground plane is sufficient.
With proper decoupling and smart partitioning of the analog,
digital, and clock sections of the PCB, optimum performance
can be easily achieved.
Advanced Power Control
For an ultrasound system, not all channels are needed during all
scanning periods. The POWER_START and POWER_STOP
values in the vector profile can be used to delay the channel
startup and turn the channel off after a certain number of samples.
These counters are relative to TX_TRIG±. The analog circuitry
needs to power up before the digital one and the advance time
(POWER_SETUP) for powering up the analog circuitry, before
POWER_START, is set up in Address 0x112 (see Table 33).
POWER_STOP
(PROFILE SPECIFIC)
TX_TRIG±
DIGITAL
POWER
ANALOG
POWER
POWER_START
(PROFILE SPECIFIC)
POWER_SETUP
(SPI SET)
11134-045
drive the STBY pin from a 3.3 V logic level, insert a 1 kΩ
resistor in series with this pin to limit the current.
Figure 44. Power Sequencing
DIGITAL OUTPUTS AND TIMING
JESD204B Transmit Top Level Description
The AD9671 digital output complies with the JEDEC Standard
JESD204B, Serial Interface for Data Converters. JESD204B is a
protocol to link the AD9671 to a digital processing device over
a serial interface up to 5 Gbps link speeds. The benefits of the
JESD204B interface include a reduction in required board area
for data interface routing, and enables smaller packages for
converter and logic devices. The AD9671 supports single, dual,
or quad lane interfaces.
JESD204B Overview
The JESD204B data transmit block, as shown in Figure 45,
assembles the parallel channel data from the ADC or digital
processing block into frames and uses 8B/10B encoding as
well as optional scrambling to form serial output data. Lane
synchronization is supported through the use of special
characters during the initial establishment of the link, and
additional synchronization is embedded in the data stream
thereafter. A matching external receiver is required to lock onto
the serial data stream and recover the data and clock. For
additional details on the JESD204B interface, users are
encouraged to refer to the JESD204B standard.
The AD9671 JESD204B transmit block maps the eight channel
outputs over a link. A link can be configured to use either
single, dual, or quad serial differential outputs, which are called
lanes. The JESD204B specification refers to a number of
parameters to define the link, and these parameters must match
between the JESD204B transmitter (AD9671 output) and
receiver.
The JESD204B link is described according to the parameters
listed in Table 14.
Rev. A| Page 29 of 60
AD9671
Data Sheet
Table 14. JESD204B Parameters
Description
Samples transmitted per single converter per frame cycle
Number of converters per converter device
L
N
N’
CF
Number of lanes per converter device
Converter resolution
Total number of bits per sample
Number of control words per frame clock cycle per
converter device
Number of control bits per conversion sample
Number of frames per multiframe
High density mode
Octets per frame
Control bit
Tail bit
Scrambler enable/disable
Checksum for the JESD204B parameters
CS
K
HD
F
C
T
SCR
FCHK
TRANSPORT
LAYER
SAMPLES
FROM CHANNEL
SAMPLE
CONSTRUCTION
AD9671 Value
1
8 with demodulator/decimator disabled, 16 with
demodulator/decimator enabled
1, 2, or 4
12, 14, or 16
16
0
0
Configurable on the AD9671
0
4, 8, 16, or 32 (dependent on L = 4, 2, or 1, respectively)
0
Available on the AD9671
Configurable on the AD9671
Automatically calculated and stored in the register map
DATA LINK
LAYER
FRAME
CONSTRUCTION
SCRAMBLER
LANE
ALIGNMENT
CHARACTER
GENERATION
PHYSICAL
LAYER
8B/10B
ENCODER
SERIALIZER
OUTPUT
11134-046
Parameter
S
M
Figure 45. AD9671 Transmit Link Simplified Block Diagram
Figure 45 shows a simplified block diagram of the AD9671
JESD204B link. By default, the AD9671 is configured to use
eight channels and four lanes. Channel A and Channel B data
is output to SERDOUT1±, Channel C and Channel D data is
output to SERDOUT2±, Channel E and Channel F data is
output to SERDOUT3±, and Channel G and Channel H data
is output to SERDOUT4±. The AD9671 allows other configurations such as combining the outputs of the eight channels onto a
single lane.
By default in the AD9671, the 14-bit converter word from each
converter is broken into two octets (eight bits of data). Bit 0
(MSB) through Bit 7 are in the first octet. The second octet
contains Bit 8 through Bit 13 (LSB) and two tail bits. The tail
bits can be configured as zeros or a pseudorandom number
sequence.
The two resulting octets can be scrambled. Scrambling is optional
but is available to avoid spectral peaks when transmitting similar
digital data patterns. The scrambler uses a self synchronizing
polynomial-based algorithm defined by the equation: 1 + x14 +
x15. The descrambler in the receiver must be a self synchronizing
version of the scrambler polynomial.
The two octets are then encoded with an 8B/10B encoder. The
8B/10B encoder works by taking eight bits of data (an octet) and
encoding them into a 10-bit symbol. Figure 46 shows how the
14-bit data is taken from the ADC, the tail bits are added, the two
octets are scrambled, and how the octets are encoded into two
10-bit symbols. Figure 46 illustrates the default data format.
At the data link layer, in addition to the 8B/10B encoding, the
character replacement allows the receiver to monitor frame
alignment. The character replacement process occurs on the
frame and multiframe boundaries, and implementation depends
on which boundary is occurring and if scrambling is enabled.
If scrambling is disabled, the following applies. If the last
scrambled octet of the last frame of the multiframe equals the
last octet of the previous frame, the transmitter replaces the last
octet with the control character /A/ = /K28.3/. On other frames
within the multiframe, if the last octet in the frame equals the
last octet of the previous frame, the transmitter replaces the last
octet with the control character /F/ = /K28.7/.
If scrambling is enabled, the following applies. If the last octet of
the last frame of the multiframe equals 0x7C, the transmitter
replaces the last octet with the control character /A/ = /K28.3/.
On other frames within the multiframe, if the last octet equals
0xFC, the transmitter replaces the last octet with the control
character /F/ = /K28.7/.
Refer to JEDEC Standard JESD204B (July 2011) for additional
information about the JESD204B interface. Section 5.1
describes the transport layer and data format details, and
Section 5.2 describes scrambling and descrambling.
Rev. A| Page 30 of 60
Data Sheet
AD9671
JESD204B Synchronization Details
The four multiframes have the following properties:
The AD9671 is a JESD204B Subclass 0 device and establishes
synchronization of the link through three control signals, TX_
TRIG, SYSREF, and SYNCINB, and typically a common device
clock. SYSREF, TX_TRIG, and SYNCINB are assumed to be
common to all converter devices for alignment purposes at the
system level.

The synchronization process is accomplished over three phases:
code group synchronization (CGS) phase, initial lane alignment
sequence (ILAS) phase, and data transmission phase. Note that
if scrambling is enabled, the bits are not actually scrambled until
the data transmission phase. The CGS and ILAS phases do not
use scrambling.
CGS Phase
In this phase, the JESD204B transmit block transmits /K28.5/
characters in response to a synchronization request from the
receiver (SYNCINB asserted). The receiver (external logic
device) must locate K28.5 characters in its input data stream
using clock and data recovery (CDR) techniques.
After a certain number of consecutive K28.5 characters are
detected on all link lanes, the receiver can optionally initiate a
SYS_REF edge so that the AD9671 transmit data establishes a
local multiframe clock (LMFC) internally. The AD9671 is a
Subclass 0 device that does not mandate SYS_REF for multidevice synchronization. The use of SYS_REF reduces the latency
variation between devices and reduces the absolute latency of
each device to some extent. However, SYS_REF does not meet
the full requirements of a JESD204B Subclass 1 device, and the
primary synchronization tool on the AD9671 is to use the global
TX_TRIG signal to embed a START_CODE simultaneously into
the data stream for all devices.
After synchronizing all lanes, the receiver or logic device
deasserts the SYNCINB signal (SYNCINB± goes high), and the
transmitter block begins the ILAS phase, if enabled, on the next
internal LMFC boundary.
ILAS Phase
In the ILAS phase, the transmitter sends out a known pattern
and the receiver aligns all lanes of the link and verifies the
parameters of the link.
The ILAS phase begins after SYNCINB± is deasserted (goes
high). The transmit block begins to transmit four multiframes.
Dummy samples are inserted between the required characters
so that full multiframes are transmitted.



Multiframe 1 begins with an /R/ character (K28.0) and
ends with an /A/ character (K28.3).
Multiframe 2 begins with an /R/ character, followed by a /Q/
(K28.4) character and link configuration parameters over
14 configuration octets (see Table 15), and ends with an
/A/ character. Many of the parameter values are of the
notation of the value − 1.
Multiframe 3 is the same as Multiframe 1.
Multiframe 4 is the same as Multiframe 1.
Data Transmission Phase
By the end of the ILAS phase, data transmission starts. Initiating a
global TX_TRIG signal resets any sampling edges within the
ADC and replaces a sample with the START_CODE (see
Address 0x18B and Address 0x18C in Table 33). Aligning the
data on all lanes based on the START_CODE guarantees the
synchronization across multiple lanes and across multiple devices.
In the data transmission phase, frame alignment is monitored
with control characters. Character replacement is used at the
end of frames. Character replacement in the transmitter occurs
in the following instances:


If scrambling is disabled and the last octet of the frame or
multiframe equals the octet value of the previous frame.
If scrambling is enabled and the last octet of the multiframe is
equal to 0x7C, or the last octet of a frame is equal to 0xFC.
Table 15. 14 Configuration Octets of the ILAS Phase
No.
0
1
2
3
4
5
6
7
8
9
10
11
12
13
Rev. A| Page 31 of 60
Bit 7
(MSB)
Bit 6
Bit 5
0
0
SCR
0
0
0
0
0
0
0
0
0
Bit 4 Bit 3
DID[7:0]
0
Bit 2
BID[3:0]
LID[4:0]
L[4:0]
F[7:0]
K[4:0]
M[7:0]
CS[1:0]
0
0
0
0
HD
0
0
0
0
0
Bit 1
N[4:0]
N’[4:0]
S[4:0]
CF[4:0]
Reserved, don’t care
Reserved, don’t care
FCHK[7:0]
Bit 0
(LSB)
AD9671
Data Sheet
Link Setup Parameters
Table 17. JESD204B Configurable Identification Values
The following steps demonstrate how to configure the AD9671
JESD204B interface and the outputs.
DID Value
LID (SERDOUT1±)
LID (SERDOUT2±)
LID (SERDOUT3±)
LID (SERDOUT4±)
DID
BID
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Disable lanes before changing the configuration.
Select the converter and lane configuration.
Configure the tail bits and control bits.
Set the lane identification values.
Set the number of frame per multiframe, K.
Enable scramble, SCR.
Set the lane synchronization options.
Verify FCHK, checksum of JESD204B interface parameters.
Set additional digital output configuration options.
Reenable lane(s) after configuration.
Disable Lanes
Converter and Lane Configuration
If the digital demodulator/decimator is disabled, the JESD204B
M parameter (number of converters) is set to 8 (Address 0x153 =
0x07). Otherwise M = 16 when the channel output is complex data.
The lane configuration is set in Address 0x150, Bits[1:0] such
that 00 = one lane per link, 01 = two lanes per link, or 11 = four
lanes per link. The channel data (A to H) is placed on the
JESD204B lanes according Table 16.
2
4
Set Number of Frames per Multiframe, K
Per the JESD204B specification, a multiframe is defined as a group
of K successive frames, where K is between 1 and 32, and it
requires that the number of octets be between 17 and 1024. The
K value is set to 32 by default in Register 0x152, Bits[4:0]. Note
that Register 0x152 represents a value of K − 1.
32 ≥ K ≥ Ceil (17/F)
The JESD204B specification also requires that the number of
octets per multiframe (K × F) be between 17 and 1024. The F
value is fixed based on the value of M and L. F can be read from
Address 0x151.
F
M 2
L
Enable Scramble, SCR
Table 16. Channel to JESD204B Lane Mapping
SERDOUT1±
A, B, C, D, E, F,
G, H
A, B, C, D
A, B
Value Range
0 to 31
0 to 31
0 to 31
0 to 31
0 to 255
0 to 15
The K value can be changed; however, it must comply with a
few conditions. The AD9671 uses a fixed value for octets per
frame, F. K must also be a multiple of 4 and conform to the
following equation:
Before modifying the JESD204B link parameters, disable the
link and hold it in reset. This is accomplished by writing a
Logic 1 to Address 0x142, Bit 0.
L
1
Register, Bits
0x148, [4:0]
0x149, [4:0]
0x14A, [4:0]
0x14B, [4:0]
0x146, [7:0]
0x147, [3:0]
SERDOUT2±
Power-down
SERDOUT3±
Power-down
SERDOUT4±
Power-down
Power-down
C, D
E, F, G, H
E, F
Power-down
G, H
Configure the Tail Bits and Control Bits
With N’ = 16 and N = 14, two tail bits are available per sample
for transmitting additional information over the JESD204B link.
Tail bits are dummy bits sent over the link to complete the two
octets and do not convey any information about the input signal.
Tail bits can be fixed zeros (default) or pseudorandom numbers
(Address 0x142, Bit 6).
Set Lane Identification Values
JESD204B allows parameters to identify the device and lane.
These parameters are transmitted during the ILAS phase, and they
are accessible in the internal registers.
There are three identification values: device identification
(DID), bank identification (BID), and lane identification (LID).
DID and BID are device specific; therefore, they can be used for
link identification.
Scrambling can be enabled or disabled by setting Address 0x150,
Bit 7. By default, scrambling is enabled. Per the JESD204B protocol,
scrambling is only functional after the lane synchronization is
complete.
Set Lane Synchronization Options
Most of the synchronization features of the JESD204B interface
are enabled by default for typical applications. In some cases,
these features can be disabled or modified as follows.
ILAS enabling is controlled in Address 0x142, Bits[3:2] and is
enabled by default. Optionally, to support some unique instances of
the interfaces (such as NMCDA-SL), the JESD204B interface
can be programmed to either disable the ILAS sequence or
continually repeat the ILAS sequence. Additionally, the ILAS can
be repeated for a fixed count, as programmed in Address 0x145,
Bits[7:0].
The AD9671 has fixed values of some of the JESD204B interface
parameters, and they are as follows:


Rev. A| Page 32 of 60
N’ = 16: number of bits per sample is 16. Read only value
from Address 0x155, Bits[3:0] = 15 (N’ − 1).
CF = 0: number of control words per frame clock cycle per
converter is 0, in Address 0x157, Bits[4:0].
Data Sheet
AD9671
Table 18. JESD204B Configuration Table Used in ILAS and
Checksum Calculation
The AD9671 calculates values for some JESD204B parameters
based on other settings, particularly the quick configuration
register selection. The following read only values are available in
the register map for verification:

No.
0
1
2
3
4
5
6
7
8
9
10
F: octets per frame can be 32, 16, 8, or 4; read the value
(F − 1) from Address 0x151, Bits[4:0]
M: number of converters per link can be 8 or 16; read the
value (M − 1) from Address 0x153, Bits[3:0]
S: samples per converter per frame is 1 by default; read the
value (S − 1) from Address 0x156, Bit 0.


Verify FCHK, Checksum of JESD204B Interface
Parameters
The JESD204B parameters can be verified through a checksum
value (FCHK) of the JESD204B interface parameters. Each lane
has a FCHK value associated with it. The FCHK value is
transmitted during the ILAS second multiframe and can be
read from the internal registers.
Bit 7
(MSB)
Bit 6
Bit 5
Bit 4 Bit 3
DID[7:0]
Bit 2
Bit 1
Bit 0
(LSB)
BID[3:0]
LID[4:0]
L[4:0]
SCR
F[7:0]
K[4:0]
M[7:0]
CS[1:0]
N[4:0]
N’[4:0]
S[4:0]
CF[4:0]
Set Additional Digital Output Configuration Options
The JESD204B outputs are configured by default to produce a
peak differential voltage of 262 mV. This voltage satisfies the
JESD204B specification for a transmit eye mask for an LV-OIF11G-SR-based operation target of between 180 mV and 385 mV
peak differential voltage, but other peak differential voltages can
be accommodated. Address 0x015, Bits[6:4] settings allow output
peak voltages. Additional options include the following:
Checksum value is the modulo 256 sum of the parameters listed
as Octet 0 to Octet 10 in Table 18. Checksum is calculated by
adding the parameter fields before they are packed into the
octets.
The FCHK value for the lane configuration for data coming out
of SERDOUT1± can be read from Address 0x15A. Similarly,
FCHK for the lane defined for SERDOUT2± can be read from
Address 0x15B.




Invert polarity of the serial output data: Address 0x014, Bit 2
Flip (mirror) 10-bit word before output: Address 0x143, Bit 0
Channel data format (offset binary, twos complement, gray
code): Address 0x014, Bits[1:0]
Options for interpreting the signal on the SYNCINB± pin:
Address 0x156, Bit 5
Reenable Lanes After Configuration
After modifying the JESD204B link parameters, enable the link
and then the synchronization process can begin. This enable is
accomplished by writing a Logic 0 to Address 0x142, Bit 0.
JESD204B
TEST PATTERN
8-BIT
JESD204B
TEST PATTERN
10-BIT
OCTET1
A8
A9
A10
A11
A12
A13
T0
T1
A0
A1
A2
A3
A4
A5
A6
A7
S8
S9
S10
S11
S12
S13
S14
S15
S0
S1
S2
S3
S4
S5
S6
S7
SERIALIZER
E10
E11
E12
E13
E14
E15
E16
E17
E18
E19
E0
E1
E2
E3
E4
E5
E6
E7
E8
E9
Figure 46. AD9671 Digital Processing of JESD204B Lanes
Rev. A| Page 33 of 60
SERDOUTx±
E0 E1 E2 E3 E4 E5 E6 E7 E8 E9 . . . E19
SYNCINB
t
SYSREF
11134-047
8B/10B
ENCODER/
CHARACTER
REPLACMENT
OPTIONAL
SCRAMBLER
1 + x14 + x15
OCTET0
A
A0
A1
A2
A3
A4
A5
CHANNEL A6
A7
A8
TX_TRIG A9
A10
A11
PATH
A12
A13
ADC
TEST PATTERN
16-BIT
AD9671
Data Sheet
Table 19. AD9671 JESD204B Frame Alignment Monitoring and Correction Replacement Characters
Character to be Replaced
Last octet in frame repeated from previous frame
Last octet in frame repeated from previous frame
Last octet in frame repeated from previous frame
Last octet in frame equals D28.7
Last octet in frame equals D28.3
Last octet in frame equals D28.7
Frame and Lane Alignment Monitoring and Correction
Frame alignment monitoring and correction is part of the
JESD204B specification. The 14-bit word requires two octets to
transmit all the data. The two octets (MSB and LSB), where
F = 2, make up a frame. During normal operating conditions
frame alignment is monitored via alignment characters that are
inserted under certain conditions at the end of a frame. Table 19
summarizes the conditions for character insertion along with
the expected characters under the various operation modes. If
lane synchronization is enabled, the replacement character value
depends on whether the octet is at the end of a frame or at the
end of a multiframe.
Based on the operating mode, the receiver can ensure that it is
still synchronized to the frame boundary by correctly receiving
the replacement characters.
Super Frame and Output Zero Stuffing
To handle the various decimation rates and to handle complex
(IQ) vs. real samples, a wrapper around the JESD204B transmitter
was created. Each word in the standard JESD204B frame represents
a word in the super frame. However, in most cases, the frame
boundary for the super-frame does not occur at the same time
as the JESD204B frame boundary.
As the decimation rates increase, relatively large amounts of
zero stuffing can occur. The zero stuffer can be configured to
add additional codes into the data stream to facilitate super
frame synchronization.
It is highly recommended to configure the device to autocalculate the size of the JESD204B and the super frames.
Last Octet in
Multiframe
No
Yes
Not applicable
No
Yes
Not applicable
Replacement Character
K28.7
K28.3
K28.7
K28.7
K28.3
K28.7
connection as shown in Figure 47. Place a 0.1 μF series
capacitor on each output pin and use a 100 Ω differential
termination close to the receiver side. The 100 Ω differential
termination results in a nominal 600 mV p-p differential swing
at the receiver. In the case where the receiver inputs do not
provide their own common-mode bias, single-ended 50 Ω
terminations can be used. When single-ended terminations are
used, the termination voltage (VRXCM) must be chosen to match
the input requirements of the receiver.
For receivers whose input common-mode voltage requirements
match the output common-mode voltage (DRVDD/2) of the
AD9671, a dc-coupled connection can be used. The common
mode of the digital output automatically biases itself to half of
DRVDD (0.9 V for DRVDD = 1.8 V) (see Figure 48).
If there is no far end receiver termination or if there is poor
differential trace routing, timing errors may result. To avoid
such timing errors, it is recommended that the trace length be
less than six inches and that the differential output traces be
adjacent and at equal lengths.
Figure 49 through Figure 54 show examples of the digital output
(default) data eyes, time interval error (TIE) jitter histograms, and
bathtub curves.
SINGLE-ENDED
TERMINATION
VRXCM
DRVDD
100Ω
DIFFERENTIAL
0.1µF TRACE PAIR
50Ω
50Ω
SERDOUTx+
100Ω
SERDOUTx–
Digital Outputs and Timing
The AD9671 has differential digital outputs that power up
by default. The driver current is derived on chip and sets the
output current at each output equal to a nominal 3 mA. Each
output presents a 100 Ω dynamic internal termination to reduce
unwanted reflections.
0.1µF
VCM = Rx VCM
OUTPUT SWING = 600mV p-p
The AD9671 digital outputs can interface with custom ASICs and
FPGA receivers, providing superior switching performance in
noisy environments. Single point-to-point network topologies are
recommended with a single differential 100 Ω termination resistor
placed as close to the receiver logic as possible.
For receiver inputs that provide their own common-mode bias,
or whose input common-mode requirements are not within the
bounds of the AD9671 DRVDD supply, use an ac-coupled
Rev. A| Page 34 of 60
RECEIVER
11134-048
Lane Synchronization
On
On
Off
On
On
Off
Figure 47. AC-Coupled Digital Output Termination Example
DRVDD
SERDOUTx+
100Ω
DIFFERENTIAL
TRACE PAIR
100Ω
RECEIVER
SERDOUTx–
OUTPUT SWING = 600mV p-p
VCM = DRVDD/2
Figure 48. DC-Coupled Digital Output Termination Example
11134-049
Scrambling
Off
Off
Off
On
On
On
Data Sheet
AD9671
MASK HITS1: EYE DIAGRAM
400
–
200
200
100
100
0
–100
–200
–100
–300
–200
0
TIME (ps)
200
400
EYE: ALL BITS
OFFSET: –0.0018
MASK: TEMP_MSK
ULS: 6000; 493327, TOTAL: 6000; 493327
–400
–200
PERIOD1: HISTOGRAM
3500
200
PERIOD1: HISTOGRAM
4
6000
–
–
5000
2500
4000
1500
3000
1000
2000
500
1000
–7.5
0
TIME (ps)
7.5
15.0
22.5
0
Figure 50. Digital Outputs Histogram, External 100 Ω Terminations at 2.5 Gbps
–10
–5
0
TIME (ps)
5
10
15
Figure 53. Digital Outputs Histogram, External 100 Ω Terminations at 5.0 Gbps
TJ AT BER1: BATHTUB
TJ AT BER1: BATHTUB
1
3
–
1–2
–15
3
–
1–2
1–4
1–6
1–6
BER
1–4
1–8
1–8
1–10
1–10
1–12
1–12
1–14
1–14
0.81
0.5
11134-349
0
UIs
0.75
Figure 51. Digital Outputs Bathtub, External 100 Ω Terminations
at 2.5 Gbps
1–16
–0.5
0
UIs
0.5
Figure 54. Digital Outputs Bathtub, External 100 Ω Terminations
at 5.0 Gbps
Rev. A| Page 35 of 60
11134-350
–15.0
11134-249
–22.5
11134-250
HITS
2000
1–16
–0.5
100
4
3000
1
0
TIME (ps)
Figure 52. Digital Outputs Data Eye, External 100 Ω Terminations
at 5.0 Gbps
Figure 49. Digital Outputs Data Eye, External 100 Ω Terminations
at 2.5 Gbps
0
–100
11134-150
–400
11134-149
EYE: ALL BITS
OFFSET: 0.0018
MASK: TEMP_MSK
ULS: 8000; 993330, TOTAL: 8000; 993330
–400
HITS
0
–200
–300
BER
1
–
300
VOLTAGE (mV)
VOLTAGE (mV)
300
MASK HITS1: EYE DIAGRAM
400
1
AD9671
Data Sheet
Additional SPI options allow the user to further increase the
output driver voltage swing of all four outputs to drive longer
trace lengths (see Address 0x015 in Table 33). Even though this
produces sharper rise and fall times on the data edges and is less
prone to bit errors, the power dissipation of the DRVDD supply
increases when this option is used. See the Memory Map section
for more details.
Preemphasis
Preemphasis enables the receiver eye diagram mask to be met in
conditions where the interconnect insertion loss is not in accordance with the JESD204B specification. In conditions where preemphasis is not needed to achieve sufficient signal integrity for
the link, it is best to disable the preemphasis to conserve power.
Enabling preemphasis on a short link and increasing the deemphasis value too high may cause the receiver eye diagram to fail
in cases where it passes with no de-emphasis. The transmitter eye
diagram does not necessarily pass when preemphasis is enabled.
Furthermore, using more preemphasis than necessary may increase
EMI; therefore, consider EMI when choosing an insertion loss
compensation strategy. To enable preemphasis, write a Logic 1
to Address 0x015, Bit 1.
There are several methods to select test data patterns on the
JESD204B link, as shown in Figure 55. These methods serve
different purposes in the testing process of establishing the link.
The processed samples from the ADC can be replaced by nine
digital output test pattern options. The replacement is initiated
through the SPI using Address 0x00D, Bits[3:0]. These options
are useful when validating receiver capture and timing. See
Table 21 for the output test mode bit sequencing options. Some
test patterns have two serial sequential words, which the user
can alternate in various ways, depending on the test pattern
chosen. Note that some patterns may not adhere to the data
format select option. In addition, custom user defined test
patterns are assigned in the user pattern registers (Address 0x019
through Address 0x020). All test mode options except PN
sequence short and PN sequence long can support 8-bit to 14-bit
word lengths to verify data capture to the receiver.
The PN sequence short pattern produces a pseudorandom bit
sequence that repeats itself every 29 − 1 bits, or 511 bits. For a
description of the PN sequence short pattern and how it is
generated, see Section 5.1 of the ITU-T O.150 (05/96) standard.
The only difference from the standard is that the starting value
is a specific value instead of all 1s (see Table 20 for the initial
values).
The PN sequence long pattern produces a pseudorandom bit
sequence that repeats itself every 223 − 1 bits, or 8,388,607 bits.
For a description of the PN sequence long pattern and how it is
generated, see Section 5.6 of the ITU-T O.150 (05/96) standard.
The only differences from the standard are that the starting
value is a specific value instead of all 1s and that the AD9671
inverts the bit stream (see Table 20 for the initial values). The
output sample size depends on the selected bit length.
Table 20. PN Sequence Initial Values
Sequence
PN Sequence Short
PN Sequence Long
Initial
Value
0x092
0x003
First Three Output Samples
(MSB First, 16-Bit)
0x496F, 0xC9A9, 0x980C
0xFF5C, 0x0029, 0xB80A
See the Memory Map section for information on how to change
these additional digital output timing features through the SPI.
Test patterns are initiated at the input of the scrambler block by
setting Address 0x144, Bits[5:4] = 10 or at the output of the
8B/10B encoder by setting Address 0x144, Bits[5:4] = 01. The
test pattern generated is selected in Address 0x144, Bits[3:0],
and is specified in Table 22.
Rev. A| Page 36 of 60
Data Sheet
AD9671
Digital Output Test Patterns
TEST
PATTERNS
TEST
PATTERNS
SERIALIZER
FRAME
CONSTRUCTION
SCRAMBLER
PROCESSED
SAMPLE FROM
ADC
SAMPLE
CONSTRUCTION
FRAME/LANE
ALIGNMENT
CHARACTER
GENERATION
OUTPUT
8B/10B
ENCODER
11134-052
TEST
PATTERNS
Figure 55. Example of Data Flow Block Diagram
Table 21. Flexible Output Test Modes—Address 0x00D
Output Test
Mode Bit
Sequence
0000
0001
0010
0011
0100
1000
Pattern Name
Off (default)
Midscale short
+Full-scale short
−Full-scale short
Checkerboard
output
PN sequence long
PN sequence short
One-/zero-word
toggle
User input
1001 to 1110
1111
Reserved
Ramp output
0101
0110
0111
Digital Output
Word 1
Not applicable
10 0000 0000 0000
11 1111 1111 1111
00 0000 0000 0000
10 1010 1010 1010
Digital Output
Word 2
Not applicable
Same
Same
Same
01 0101 0101 0101
Digital Output
Word 3
Not applicable
Same
Same
Same
10 1010 1010 1010
Digital Output
Word 4
Not applicable
Same
Same
Same
01 0101 0101 0101
Subject to
Resolution
Select
Not applicable
Yes
Yes
Yes
No
Not applicable
Not applicable
11 1111 1111 1111
Not applicable
Not applicable
00 0000 0000 0000
Not applicable
Not applicable
11 1111 1111 1111
Not applicable
Not applicable
00 0000 0000 0000
Yes
Yes
No
Address 0x019 and
Address 0x01A
Not applicable
00 0000 0000 0000
Address 0x01B and
Address 0x01C
Not applicable
00 0000 0000 0001
Address 0x01D and
Address 0x01E
Not applicable
00 0000 0000 0000
Address 0x01F and
Address 0x020
Not applicable
00 0000 0000 0001
No
Digital Output
Word 1
Not applicable
10 1010 1010 1010
Digital Output
Word 2
Not applicable
01 0101 0101 0101
Digital Output
Word 3
Not applicable
10 1010 1010 1010
Digital Output
Word 4
Not applicable
01 0101 0101 0101
Subject to
Resolution
Select
Not applicable
No
11 1111 1111 1111
00 0000 0000 0000
11 1111 1111 1111
00 0000 0000 0000
No
Not applicable
Not applicable
Address 0x019 and
Address 0x01A
Address 0x019 and
Address 0x01A
00 0000 0000 0000
See JESD204B
specification
Not applicable
Not applicable
Not applicable
Address 0x01B and
Address 0x01C
Address 0x01B and
Address 0x01C
00 0000 0000 0001
See JESD204B
specification
Not applicable
Not applicable
Not applicable
Address 0x01D and
Address 0x01E
Address 0x01D and
Address 0x01E
00 0000 0000 0000
See JESD204B
specification
Not applicable
Not applicable
Address 0x01F and
Address 0x020
Address 0x01F and
Address 0x020
00 0000 0000 0001
See JESD204B
specification
Yes
Yes
No
No
Yes
Table 22. Flexible Output Test Modes—Address 0x144
Output Test
Mode Bit
Sequence
0000
0001
0010
0011
0100
0101
0110
0111
1000
1001 to 1111
Pattern Name
Off (default)
Alternating
checkerboard
One-/zero-word
toggle
PN sequence long
PN sequence short
Continuous/repeat
user test pattern
Single user test
pattern
Ramp output
Modified RPAT
sequence
Reserved
Rev. A| Page 37 of 60
No
Yes
Not applicable
No
AD9671
Data Sheet
SDIO Pin
TX_TRIG± Pins
The SDIO pin is required to operate the SPI. The SDIO pin has
an internal 30 kΩ pull-down resistor that pulls it low and is only
1.8 V tolerant. To drive the SDIO pin from a 3.3 V logic level, insert
a 1 kΩ resistor in series with this pin to limit the current.
The TX_TRIG± function has several uses within the AD9671
and is initiated with an external hardware trigger either on the TX_
TRIG± pins or by a software trigger by setting Address 0x10C, Bit 5
to 1. The hardware trigger has the advantage of guaranteed
synchronous triggering of multiple AD9671 devices in a system.
The setup and hold time for each TX_TRIG± hardware input is
given in Table 3 as 1 ns. Due to the asynchronous SPI function,
the software trigger cannot guarantee synchronization of multiple
AD9671 devices. If the TX_TRIG± hardware trigger is not used,
tie the TX_TRIG± pins in a low logic state.
SCLK Pin
The SCLK pin is required to operate the SPI. The SCLK pin has
an internal 30 kΩ pull-down resistor that pulls it low and is only
1.8 V tolerant. To drive the SCLK pin from a 3.3 V logic level,
insert a 1 kΩ resistor in series with this pin to limit the current.
CSB Pin
The CSB pin is required to operate the SPI. The CSB pin has an
internal 70 kΩ pull-up resistor that pulls it high and is only 1.8 V
tolerant. To drive the CSB pin from a 3.3 V logic level, insert a
1 kΩ resistor in series with this pin to limit the current.
RBIAS Pin
To set the internal core bias current of the ADC, place a resistor
nominally equal to 10.0 kΩ to ground at the RBIAS pin. Using a
resistor other than the recommended 10.0 kΩ resistor for RBIAS
degrades the performance of the device. Therefore, use at least a
1% tolerance on this resistor to achieve consistent performance.
VREF Pin
A stable and accurate 0.5 V voltage reference is built into the
AD9671. This voltage reference is amplified internally by a factor of
2, setting VREF to 1.0 V, which results in a full-scale differential
input span of 2.0 V p-p for the ADC. VREF is set internally by
default, but the user can drive the VREF pin externally with a
1.0 V reference to achieve more accuracy. However, the AD9671
does not support ADC full-scale ranges less than 2.0 V p-p.
When applying the decoupling capacitors to the VREF pin, use
ceramic, low equivalent series resistance (ESR) capacitors. Ensure
that these capacitors are near the reference pin and on the same
layer of the PCB as the AD9671. The VREF pin must have both
a 0.1 μF capacitor and a 1 μF capacitor that are connected in
parallel to analog ground. These capacitor values are recommended
for the ADC to properly settle and acquire the next valid sample.
GPOx Pins
Use the general-purpose output pins, GPO0, GPO1, GPO2, and
GPO3, in a system to provide programmable inputs to other chips
in the system. The value of each pin is set via Address 0x00E to
either Logic 0 or Logic 1 (see Table 33).
ADDRx Pins
Use the chip address pins to address individual AD9671 devices in
a system. Chip address mode is enabled using Address 0x115,
Bit 5 (see Table 33). If the value written to Bits[4:0] matches the
value on the chip address bit pins (ADDR4 to ADDR0), the device
is selected and any subsequent SPI writes or reads to addresses
indicated as chip registers are written only to that device. If chip
address mode is disabled, write all addresses regardless of the value
on the address pins.
The TX_TRIG± function is used to reset circuits in the digital
demodulator and decimator (see the Baseband Demodulator
and Decimator section), initiate the advanced power mode (see
the Advanced Power Control section), and synchronize the data
serialization in the JESD204B block (see the JESD204B
Overview section).
ANALOG TEST TONE GENERATION
The AD9671 can generate analog test tones that the user can
then switch to the input of the LNA of each channel for channel
gain calibration. The test tone amplitude at the LNA output is
dependent on LNA gain, as shown in Table 23.
Table 23. Test Signal Fundamental Amplitude at LNA Output
Address 0x116[3:2],
Analog Test Tones
00 (default)
01
10
11
LNA Gain
15.6 dB
80 mV p-p
160 mV p-p
320 mV p-p
Reserved
LNA Gain
17.9 dB
98 mV p-p
196 mV p-p
391 mV p-p
Reserved
LNA Gain
21.6 dB
119 mV p-p
238 mV p-p
476 mV p-p
Reserved
Calculate the test signal amplitude at the input to the ADC
given the LNA gain, attenuator control voltage, and the PGA
gain. Table 24 and Table 25 list example calculations.
Table 24. Test Signal Fundamental Amplitude at ADC Input,
VGAIN = 0 V, PGA Gain = 21 dB
Address 0x116[3:2],
Analog Test Tones
00 (default)
01
10
11
LNA Gain
15.6 dB
−29 dBFS
−23 dBFS
−17 dBFS
Reserved
LNA Gain
17.9 dB
−28 dBFS
−22 dBFS
−16 dBFS
Reserved
LNA Gain
21.6 dB
−26 dBFS
−20 dBFS
−14 dBFS
Reserved
Table 25. Test Signal Fundamental Amplitude at ADC Input,
VGAIN = 0 V, PGA Gain = 30 dB
Address 0x116[3:2],
Analog Test Tones
00 (default)
01
10
11
Rev. A| Page 38 of 60
LNA Gain
15.6 dB
−20 dBFS
−14 dBFS
−8 dBFS
Reserved
LNA Gain
17.9 dB
−19 dBFS
−13 dBFS
−7 dBFS
Reserved
LNA Gain
21.6 dB
−17 dBFS
−11 dBFS
−5 dBFS
Reserved
Data Sheet
AD9671
CW DOPPLER OPERATION
Each channel of the AD9671 includes an I/Q demodulator. Each
demodulator has an individual programmable phase shifter.
The I/Q demodulator is ideal for phased array beamforming
applications in medical ultrasound. Each channel can be
programmed for 16 phase settings/360° (or 22.5°/step), selectable
via the SPI port. The device has a RESET± input that is used to
synchronize the LO dividers of each channel. If multiple AD9671
devices are used, a common reset across the array ensures a
synchronized phase for all channels. If the RESET± input is not
used, tie each input pin to ground. Internal to the AD9671, the
individual Channel I and Channel Q outputs are current summed.
If multiple AD9671 devices are used, current sum and convert
the I and Q outputs from each AD9671 to a voltage using an
external transimpedance amplifier.
Quadrature Generation
The internal 0° and 90° LO phases are digitally generated by a
divide-by-M logic circuit, where M is 4, 8, or 16. The internal
divider is selected via Address 0x02E, Bits[2:0] (see Table 33). The
divider is dc-coupled and inherently broadband; the maximum
LO frequency is limited only by its switching speed. Ensure that
the duty cycle of the quadrature LO signals is as near 50% as
possible for the 4LO and 8LO modes. The 16LO mode does not
require a 50% duty cycle. Furthermore, the divider is implemented
such that the MLO signal reclocks the final flip-flops that generate
the internal LO signals and thereby minimizes noise introduced
by the divide circuitry.
For optimum performance, the MLO signal input is driven
differentially, as on the AD9671 evaluation board. The commonmode voltage on each pin is approximately 1.2 V with the
nominal 3 V supply. It is important to ensure that the MLO source
have very low phase noise (jitter), a fast slew rate, and an
adequate input level to obtain optimum performance of the CW
signal chain.
Beamforming applications require a precise channel-to-channel
phase relationship for coherence among multiple channels. The
RESET± input is provided to synchronize the LO divider circuits in
different AD9671 devices when they are used in arrays. The
RESET± input is a synchronous edge-triggered input that resets the
dividers to a known state after power is applied to multiple
AD9671 devices. The RESET± signal can be either a continuous
signal or a single pulse, and it can be either synchronized with the
MLO± clock edge (recommended) or it can be asynchronous. If a
continuous signal is used for the RESET± then it has to be at the
LO rate. For synchronous RESET±, the device can be configured to
sample the RESET± signal with either the falling or rising edge of
the MLO± clock, which makes it easier to align the RESET± signal
with the opposite MLO± clock edge. Register 0x02E is used to
configure the RESET signal behavior. Synchronize the RESET±
input to the MLO signal input. Achieve accurate channel-tochannel phase matching via a common clock on the RESET± input
when using more than one AD9671.
I/Q Demodulator and Phase Shifter
The I/Q demodulators consist of double-balanced, harmonic
rejection, passive mixers. The RF input signals are converted
into currents by transconductance stages that have a maximum
differential input signal capability of matching the LNA output
full scale. These currents are then presented to the mixers that
convert them to baseband (RF − LO) and 2× RF (RF + LO).
The signals are phase shifted according to the codes that are
programmed into the SPI latch (see Table 26). The phase shift
function is an integral part of the overall circuit. The phase shift
listed in Table 26 is defined as being between the baseband I or
Q channel outputs. As an example, for a common signal applied
to a pair of RF inputs to an AD9671, the baseband outputs are
in phase for matching phase codes. However, if the phase code
for Channel 1 is 0000 and the phase code for Channel 2 is 0001,
Channel 2 leads Channel 1 by 22.5°.
Table 26. Phase Select Code for Channel-to-Channel Phase Shift
Phase Shift
0°
22.5°
45°
67.5°
90°
112.5°
135°
157.5°
180°
202.5°
225°
247.5°
270°
292.5°
315°
337.5°
Rev. A| Page 39 of 60
I/Q Demodulator Phase (Address 0x02D[3:0])
0000
0001 (not valid for 4LO mode)
0010
0011 (not valid for 4LO mode)
0100
0101 (not valid for 4LO mode)
0110
0111 (not valid for 4LO mode)
1000
1001 (not valid for 4LO mode)
1010
1011 (not valid for 4LO mode)
1100
1101 (not valid for 4LO mode)
1110
1111 (not valid for 4LO mode)
AD9671
Data Sheet
DIGITAL DEMODULATOR/DECIMATOR
The AD9671 contains digital processing capability. Each
channel has three stages of processing that are available: RF
decimator, baseband (BB) demodulator, and baseband
decimator. For test purposes, the input to the demodulator/
decimator can serve as a test waveform. Normally, the input is
the output of the ADC. The output of the demodulator/decimator
is sent to the framer/serializer for output formatting.
The maximum data rate of the BB demodulator and decimator
is 65 MSPS. Therefore, if the sample of the ADC is greater than
65 MSPS, enable the RF decimator (fixed rate of 2). The ADC
resolution is 14 bits. The maximum resolution at the output of
the digital processing is 16 bits. Saturation of the ADC is
determined after the dc offset calibration to ensure maximum
dynamic range. Depending on decimation rate, the loss in
output SNR due to truncation to 16 bits is negligible.
VECTOR PROFILE
To minimize the time needed to reconfigure device settings
during operation, the device supports configuration profiles.
The user can store up to 32 profiles in the device. A profile is
selected by a 5-bit index. A profile consists of a 64-bit vector, as
described in Table 27. Each parameter is concatenated to form the
64-bit profile vector. The profile memory starts at Register 0xF00
and ends at Register 0xFFF. Write the memory in either stream
or address selected data mode. However, the user must read the
memory using stream mode. When writing or reading in
stream mode while the SPI configuration is set to MSB first
mode (default setting for Register 0x000), the write/read address
needs to refer to the last register address, not the first one. For
example, when writing or reading the first profile that spans the
address space between Register 0xF00 and Register 0xF07, with the
SPI port configured as MSB first, the referenced address must
be Register 0xF07 to allow reading or writing the 64 profile bits
in MSB mode. For more information about stream mode, see
the AN-877 Application Note, Interfacing to High Speed ADCs
via SPI.
There is a buffer used to store the current profile data. When
the profile index is written in Register 0x10C, the selected profile is
read from memory and stored in the current profile buffer. The
profile memory is read/written in the SPI clock domain. After
the SPI writes the profile index value, it takes four SPI clock
cycles to read the profile from RAM and store it in the current
profile buffer. If the SPI is in LSB mode, these additional SPI
clock cycles are provided when the profile index register is
written. If the SPI is in MSB mode, an additional byte must be
read or written to update the profile buffer.
Updating profile memory does not affect the data in the profile
buffer. The profile index register must be written to cause a
refresh of the current profile data, even if the profile index
register is written with the same value.
NUMERICALLY
CONTROLLED
OSCILLATOR
ADC OUTPUT OR
TEST WAVEFORM
MULTIBAND AAF
DECIMATE BY 2
DC OFFSET
CALIBRATION
Cos
HIGH-PASS
FILTER
BB DECIMATOR
I
Q
RF DECIMATOR
LOW-PASS
FILTER
DECIMATOR
LOW-PASS
FILTER
DECIMATOR
BB DEMODULATOR
Figure 56. Simplified Block Diagram of a Single Channel of Demodulator/Decimator
Table 27. Profile Definition
Field
f
No. of
Bits
16
Description
Demodulation frequency (fD)
fD = f × fSAMPLE/216, where (f ) = [0,(216 − 1)] and fSAMPLE is the effective sample rate
0x0000: fD = 0 (dc, I = cos(0) = 1, Q = sin(0) = 0)
0x0001: fD = fSAMPLE/216
…
0x8000: fD = fSAMPLE/2
…
0xFFFF (216 − 1): fD = fSAMPLE (216 − 1)/216 = −fSAMPLE/216
Rev. A| Page 40 of 60
FRAMER
SERIALIZER
11134-053
–Sin
Data Sheet
Field
AD9671
P
No. of
Bits
8
M
5
g
3
HPF Bypass
1
POWER_START
15
Reserved
POWER_STOP
1
15
Description
Pointer to coefficient block. The coefficients used begin at Coefficient P × 8 and continues for M × 8 coefficients,
for example,
0000 0000: points to Coefficient 0 and continues M × 8 coefficients
0000 0001: points to Coefficient 8 and continues M × 8 coefficients
Decimation factor
M = N – 1, where N = decimation factor
0x00: decimate by 1 (no decimation, just filtering)
0x01: decimate by 2
0x02: decimate by 3
…
0x1F: decimate by 32
Digital gain compensation
Gain = 2
000: gain = 1 (no shift)
001: gain = 2 (shift by 1)
010: gain = 4 (shift by 2)
…
111: gain = 128 (shift by 7)
Digital high-pass filter bypass
0 = disable (filter enabled)
1 = enable (filter bypassed)
ADC clock cycles counted from the TX_TRIG signal assertion when the active channels are powered up
0x0000 = 0 clock cycles
0x0001 = 1 clock cycle
…
0x7FFF = 32,767 clock cycles
Reserved
ADC clock cycles counted from the TX_TRIG signal assertion when the active channels are powered down
0x0000 = 0 clock cycles
0x0001 = 1 clock cycle
…
0x7FFF = continuous run mode
RF DECIMATOR
The input to the RF decimator is either the ADC output data or
a test waveform, as described in the Digital Test Waveforms section.
The test waveforms are enabled per channel using Address 0x11A
(see Table 33).
fSAMPLE/2. Figure 57 and Figure 58 show the frequency response
of the filter, depending on the mode. Figure 57 shows the
attenuation amplitude over the Nyquist frequency range.
Figure 58 shows the pass band response as nearly flat.
10
DC Offset Calibration
The user can reduce dc offset through a manual system
calibration process. Measure the dc offset of every channel in
the system and then set a calibration value using Address 0x110
and Address 0x111. Note that these registers are both chip and
local addresses, meaning that they are accessed using the chip
address and device index. Bypass the dc offset calibration using
Address 0x10F, Bits[2:0].
AMPLITUDE (dBFS)
0
LOW BAND FILTER
–10
HIGH BAND FILTER
–20
–30
–40
–50
The multiband filter is a finite impulse response (FIR) filter. It is
programmable with low or high bandwidth filtering. The filter
requires 11 input samples to populate the filter. The decimation
rate is fixed at 2×. Therefore, the decimation frequency is fDEC =
Rev. A| Page 41 of 60
–60
0
2
4
6
8
10
12
14
16
18
20
FREQUENCY (MHz)
Figure 57. AAF Frequency Response (Frequency Scale Assumes
fADC = 2 × fDEC = 40 MHz)
11134-054
Multiband AAF and Decimate by 2
AD9671
Data Sheet
2
Coefficient Memory
1
The coefficient memory stores the eight coefficients per
decimation, with a maximum decimation of 32, in a coefficient
memory block. At a maximum decimation of 32, 32 × 8 = 256
coefficients is needed. The coefficient memory is available at SPI
Address 0x1000 to Address 0x1FFF. This memory is sufficient
space to store up to 2048 coefficients. Each vector profile has a
pointer, P, to the coefficient block within coefficient memory.
AMPLITUDE (dBFS)
0
–1
LOW BAND FILTER
HIGH BAND FILTER
–2
–3
–4
–5
–6
–8
0
2
4
6
8
10
12
14
16
18
20
FREQUENCY (MHz)
11134-055
–7
Figure 58. AAF Frequency Response, Zoomed In (Frequency Scale Assumes
fADC = 2 × fDEC = 40 MHz)
High-Pass Filter
The user can apply a second-order Butterworth, high-pass
infinite impulse response (IIR) filter after the RF decimator. The
filter has a cutoff of 700 kHz for an encode clock of 50 MHz. The
filter has a settling time of 2.5 μs. Therefore, if the ADC clock is
50 MHz, ignore the first 125 samples (2.5 μs/0.02 μs). Bypass or
enable the filter in the vector profile if the filter is enabled in
Register 0x113, Bit 5. If the filter is bypassed by setting
Register 0x113, Bit 5 = 1, the filter cannot be enabled from
the vector profile.
BASEBAND DEMODULATOR AND DECIMATOR
The demodulator downconverts the RF signal to a baseband
quadrature signal. The excess oversampling is reduced by the
decimator.
Numerically Controlled Oscillator
The numerically controlled oscillator (NCO) generates I and Q
signals (cos and –sin) for the demodulator. A division of the
effective sample clock generates the oscillator frequency. If the
RF decimator is bypassed, the effective sample clock is the same
as the ADC clock. If the RF decimator is enabled, the effective
clock rate is ½ the ADC sample clock frequency. The divider is
set in the vector profile. The oscillator has a frequency resolution
of 1 kHz. To synchronize different devices, the NCO is reset
upon assertion of TX_TRIG±.
Decimation Filter
The purpose of the decimation filter is to band limit the
demodulated signal prior to decimation. The filter is a polyphase
FIR filter and uses 16 taps per decimation with symmetrical
coefficients. Therefore, there are eight unique 14-bit coefficients
per decimation. The decimation rate and a pointer to the coefficients used by the filter are set in the vector profile. Digital gain
from 1 to 128 is applied to the filter response. The digital gain
compensation is set in the vector profile.
The filter is reset upon assertion of TX_TRIG±. The decimation
filter takes 32× the decimation input samples or 32 output
samples to populate.
Coefficients are written using the SPI in stream mode during
startup. Coefficients are written in 14-bit × 8-word = 112-bit
blocks. There are 256 coefficient blocks. The 14 bits × 8-word
coefficients are packed into 14 bytes × 8 bits, as shown in Table 29.
Writes and reads from a coefficient block must begin on a
coefficient block boundary and an entire coefficient block must
be written or read. After a coefficient block is written, the
coefficient block address automatically increments/decrements
(depending on the LSB/MSB SPI setting in Register 0x000) to
the next coefficient block.
Having a direct map between SPI memory address and coefficient block address requires a divide by 7, which is not simple to
accomplish in hardware (the address must be mapped within a
single cycle). Therefore, each block is padded to a 16-byte
boundary, but the SPI does not need to shift in these extra two
bytes when loading coefficient memory sequentially. If the SPI
is configured LSB first, SPIADDR[3:0] is all 0s. If the SPI is configured MSB first, SPIADDR[3:0] is all 1s. In other words, in
LSB mode, the referenced addresses for the coefficient memory
blocks are 0x1000, 0x2000, and so on, whereas in MSB SPI mode,
the referenced block addresses are 0x100F, 0x200F, and so on.
Coefficient block order and how words/bytes are split across
each other are shown in Table 29. When the SPI is configured
LSB first, C0[0] = B0[0] is written first, and C7[13] = B13[7] is
written last. When the SPI is configured MSB first, C7[13] =
B13[7] is written first, and C0[0] = B0[0] is written last.
The position of a coefficient, Cn, in memory is determined
from its index (i, j) by
n = M(1 + i) − (1 + j), if i is even
(8)
n = M × i + j, if i is odd
(9)
where
M is the decimation factor.
j is the decimation phase from 0 to M − 1.
i is the index within the coefficient block, from 0 to 7.
Due to symmetry, Coefficient C0 is multiplied by the newest
and oldest samples.
As an example, the coefficient memory for a decimation factor
of M = 4 is shown in Table 28.
The upper 16 bits of the filter output are used as the data output
of the channel. The filter output may have gain applied according
to g, from the vector profile. Additionally, a gain of 4 can be
applied using the filter output gain in Register 0x113, Bit 4.
Rev. A| Page 42 of 60
Data Sheet
AD9671
Table 28. Coefficient Memory for M = 4
Index (i)
Decimation Phase (j)
0
1
2
3
7
28
29
30
31
6
27
26
25
24
5
20
21
22
23
4
19
18
17
16
3
12
13
14
15
2
11
10
9
8
1
4
5
6
7
0
3
2
1
0
C1[13:0]
27:14
C0[13:0]
13:0
B1[7:0]
15:8
B0[7:0]
7:0
Table 29. Coefficient Block Mapping into SPI Memory Location
C7[13:0]
111:98
B13[7:0]
111:104
C6[13:0]
97:84
B12[7:0]
103:96
B11[7:0]
95:88
C5[13:0]
83:70
B10[7:0]
87:80
B9[7:0]
79:72
Coefficients (Eight Words × 14 Bits)
C4[13:0]
C3[13:0]
69:56
55:42
SPI Memory (14 Bytes)
B8[7:0] B7[7:0] B6[7:0] B5[7:0] B4[7:0]
71:64
63:56
55:48
47:40
39:32
C2[13:0]
41:28
B3[7:0]
31:24
B2[7:0]
23:16
DIGITAL TEST WAVEFORMS
DIGITAL BLOCK POWER SAVING SCHEME
Digital test waveforms can be used in the digital processing block
instead of the ADC output. To enable digital test waveforms,
use Address 0x11B. Enable each channel individually in
Address 0x11A.
To reduce power consumption in the digital block, the
demodulator and decimation filter start in an idle state after
running the chip (Register 0x008, Bits[2:0] = 000). In the digital
idle state, the chip JESD204B block outputs zeroes and there is
no unnecessary digital processing of the ADC output data. The
digital block only switches to a running state when the negative
edge of the TX_TRIG± pulse is detected, or with a software
TX_TRIG± write (Register 0x10C, Bit 5 = 1).
For testing and debugging, use a programmable waveform
generator in place of ADC data. The waveform generator can
vary offset, amplitude, and frequency. The generator uses the ADC
sample frequency, fSAMPLE, and ADC full-scale amplitude, AFULL-SCALE,
as references. The values are set in Address 0x117, Address 0x118,
and Address 0x119 (see Table 33).
x = C + A × sin(2 × π × N)
N
f SAMPLE  n
, see Address 0x117
64
(10)
(11)
A
A  FULL xSCALE , see Address 0x118
2
(12)
C = AFULL-SCALE × a × 2−(13 − b), see Address 0x119
(13)
To put the digital block back into the idle state (while the rest of
the chip is still running) and to save power, enact one of the
following three events: raise the TX_TRIG± signal high, write to
the profile index (Register 0x10C, Bits[0:4]), or allow the power
stop to expire by using the advanced power control feature.
Figure 59 illustrates the digital block power saving scheme.
CHIP IN POWER-DOWN,
STANDBY,
OR CW MODE
RUN CHIP
DIGITAL
DECIMATOR/FILTER
IDLE
Channel ID and Ramp Generator
In Channel ID test mode, the output is a concatenated value.
Bits[6:0] are a ramp. Bit 7 is 0 in real data mode or I channel
and 1 for Q channel in complex data mode. Bits[10:8] are the
channel ID such that Channel A is coded as 000 and Channel B
is 001. Bits[15:11] are the chip address.
TX_TRIG± IS HIGH, PROFILE
INDEX WRITE, OR POWER
STOP EXPIRES
Filter Coefficients
NEGATIVE EDGE TX_TRIG±
OR S/W TX_TRIG±
DIGITAL
DECIMATOR/FILTER
RUNNING
Figure 59. Digital Block Power Saving Scheme
To check the filter coefficients, use a sequence of 1 followed by
0s for the input to the decimating FIR filter. The number of 0s is
the decimation rate times the number of taps (16). The output
shifter outputs the LSBs of the filter.
Rev. A| Page 43 of 60
11134-358
Waveform Generator
AD9671
Data Sheet
SERIAL PORT INTERFACE (SPI)
Table 30. Serial Port Pins
Pin
SCLK
SDIO
CSB
Function
Serial clock. Serial shift clock input. SCLK
synchronizes serial interface reads and writes.
Serial data input/output. Dual-purpose pin that
typically serves as an input or an output, depending
on the instruction sent and the relative position in
the timing frame.
Chip select bar (active low). This control gates the
read and write cycles.
The falling edge of CSB, in conjunction with the rising edge of
SCLK, determines the start of the framing sequence. During the
instruction phase, a 16-bit instruction is transmitted, followed
by one or more data bytes, which is determined by Bit Field W0
and Bit Field W1. An example of the serial timing and its
definitions are shown in Figure 61 and Table 31.
During normal operation, CSB signals to the device that SPI
commands are to be received and processed. When CSB is
brought low, the device processes SCLK and SDIO to execute
instructions. Normally, CSB remains low until the communication
cycle is complete. However, if connected to a slow device, CSB
can be brought high between bytes, allowing older microcontrollers
enough time to transfer data into shift registers. CSB can be
stalled when transferring one, two, or three bytes of data. When
W0 and W1 are set to 11, the device enters streaming mode and
continues to process data, either reading or writing, until CSB is
taken high to end the communication cycle. CSB being high allows
complete memory transfers without the need for additional
instructions. Regardless of the mode, if CSB is taken high in the
middle of a byte transfer, the SPI state machine is reset, and the
device waits for a new instruction.
The SPI port can be configured to operate in different manners.
CSB can also be tied low to enable 2-wire mode. When CSB is
tied low, SCLK and SDIO are the only pins required for
In addition to word length, the instruction phase determines
whether the serial frame is a read or write operation, allowing
the serial port to be used both to program the chip and to read
the contents of the on-chip memory. If the instruction is a readback operation, performing a readback causes the serial data
input/output (SDIO) pin to change direction from an input to
an output at the appropriate point in the serial frame.
The user can send data in MSB first mode or LSB first mode.
MSB first mode is the default at power-up and is changed by
adjusting the configuration register (Address 0x000). For more
information about this and other features, see the AN-877
Application Note, Interfacing to High Speed ADCs via SPI.
HARDWARE INTERFACE
The pins described in Table 30 constitute the physical interface
between the programming device and the serial port of the
AD9671. The SCLK and CSB pins function as inputs when
using the SPI. The SDIO pin is bidirectional, functioning as an
input during write phases and as an output during readback.
If multiple SDIO pins share a common connection, ensure that
proper VOH levels are met. Figure 60 shows the number of SDIO
pins that can be connected together and the resulting VOH level,
assuming the same load for each AD9671.
1.800
1.795
1.790
1.785
1.780
1.775
1.770
1.765
1.760
1.755
1.750
1.745
1.740
1.735
1.730
1.725
1.720
1.715
0
10
20
30
40
50
60
70
80
90
NUMBER OF SDIO PINS CONNECTED TOGETHER
100
11134-056
Three pins define the serial port interface: SCLK, SDIO, and CSB
(see Table 30). The SCLK (serial clock) pin synchronizes the read
and write data presented to the device. The SDIO (serial data
input/output) pin is a dual-purpose pin that allows data to be
sent to and read from the internal memory map registers of the
device. The CSB (chip select bar) pin is an active low control that
enables or disables the read and write cycles.
communication. Although the device is synchronized during
power-up, exercise caution when using 2-wire mode to ensure
that the serial port remains synchronized with the CSB line.
When operating in 2-wire mode, use a 1-, 2-, or 3-byte transfer
exclusively. Without an active CSB line, streaming mode can be
entered but not exited.
VOH (V)
The AD9671 SPI allows the user to configure the signal chain for
specific functions or operations through the structured register
space provided inside the chip. The SPI offers the user added
flexibility and customization, depending on the application.
Addresses are accessed via the serial port and can be written to
or read from via the port. Memory is organized into bytes that
can be further divided into fields, as documented in the Memory
Map section. For detailed operational information, see the AN-877
Application Note, Interfacing to High Speed ADCs via SPI.
Figure 60. SDIO Pin Loading
This interface is flexible enough to be controlled either by serial
programmable read only memories (PROMs) or by PIC
microcontrollers, which provide the user with an alternative to a
full SPI controller for programming the device (see the AN-812
Application Note, Microcontroller-Based Serial Port Interface
(SPI®) Boot Circuit).
Rev. A| Page 44 of 60
Data Sheet
AD9671
tDS
tS
tHIGH
tCLK
tH
tDH
tLOW
CSB
DON’T
CARE
SDIO
DON’T
CARE
DON’T
CARE
R/W
W1
W0
A12
A11
A10
A9
A8
A7
D5
D4
D3
D2
D1
D0
DON’T
CARE
Figure 61. Serial Timing Details
Table 31. Serial Timing Definitions
Parameter
tDS
tDH
tCLK
tS
tH
tHIGH
tLOW
tEN_SDIO
Timing (ns min)
12.5
5
40
5
2
16
16
15
tDIS_SDIO
15
Description
Setup time between the data and the rising edge of SCLK
Hold time between the data and the rising edge of SCLK
Period of the clock
Setup time between CSB and SCLK
Hold time between CSB and SCLK
Minimum period that SCLK must be in a logic high state
Minimum period that SCLK must be in a logic low state
Minimum time for the SDIO pin to switch from an input to an output relative to the SCLK falling
edge (not shown in Figure 61)
Minimum time for the SDIO pin to switch from an output to an input relative to the SCLK rising
edge (not shown in Figure 61)
Rev. A| Page 45 of 60
11134-057
SCLK
AD9671
Data Sheet
MEMORY MAP
READING THE MEMORY MAP TABLE
RESERVED LOCATIONS
Each row in the memory map register table has eight bit locations.
The memory map is roughly divided into three sections: the chip
configuration register map (Address 0x000 to Address 0x19C), the
profile register map (Address 0xF00 to Address 0xFFF), and the
coefficient register map (Address 0x1000 to Address 0x1FFF).
Registers that are designated as local registers use the device
index in Address 0x004 and Address 0x005 to determine to
which channels of a device the command is applied. Registers
that are designated as chip registers use the chip address mode
in Address 0x115 to determine whether the device is to be
updated by writing to the chip register.
Do not write to undefined memory locations except when
writing the default values suggested in this data sheet. Addresses
that have values marked as 0 must be considered reserved and
have a 0 written into their registers during power-up.
The first column of the memory map indicates the register address,
and the default value is shown in the second rightmost column.
The Bit 7 (MSB) column is the start of the default hexadecimal
value given. For example, Address 0x011, the LNA and VGA gain
adjustment register, has a default value of 0x06, meaning that
Bit 7 = 0, Bit 6 = 0, Bit 5 = 0, Bit 4 = 0, Bit 3 = 0, Bit 2 = 1, Bit 1 =
1, and Bit 0 = 0, or 0000 0110 in binary. This setting is the default
for GAIN± pins enabled, PGA gain = 24 dB and LNA gain =
21.6 dB.
For more information about the SPI memory map and other
functions, see the AN-877 Application Note, Interfacing to
High Speed ADCs via SPI.
DEFAULT VALUES
After a reset, critical registers are automatically loaded with default
values. These values are indicated in Table 33, where an X refers
to an undefined feature (don’t care).
LOGIC LEVELS
An explanation of various registers follows: “bit is set” is
synonymous with “bit is set to Logic 1” or “writing Logic 1 for
the bit.” Similarly, “bit is cleared” is synonymous with “bit is set
to Logic 0” or “writing Logic 0 for the bit.”
RECOMMENDED START-UP SEQUENCE
To save system power during programming, the AD9671 powers
up in power-down mode. To start the device up and initialize
the data interface, the SPI commands listed in Table 32 are
recommended. At a minimum, write the profile memory for an
index of 0 (Address 0xF00 to Address 0xF07; see Table 27). If
additional profiles and coefficient memory are required, write these
after Profile File Memory 0.
Table 32. AD9671 SPI Write Start-Up Sequence Example
Address
0x000
0x002
0x0FF
0x004
0x005
0x113
0x011
0xF00
0xF01
0xF02
0xF03
0xF04
0xF05
0xF06
0xF07
0x10C1
0x014
0x008
0x021
0x199
0x142
0x188
0x18B
0x18C
Value
0x3C
0x0X (default)
0x01
0x0F
0x3F
0x03
0x06 (default)
0xFF
0x7F
0x00
0x80
0x0C
0x00
0x00
0x20
0x00 (default)
0x00
0x00
0x12
0x80
0x04
0x01
0x27
0x72
Description
Initiate SPI reset
Set speed mode to 40 MSPS
Enable speed mode change
Set local registers to all channels
Set local registers to all channels
Bypass demodulator and decimator, bypass RF decimator, enable high pass filter
Set LNA gain= 21.6 dB, GAIN± pins enabled, and PGA gain = 24 dB
Continuous run mode enable; do not power down channels (POWER_STOP LSB)
Continuous run mode enable; do not power down channels (POWER_STOP MSB)
Power up all channels 0 clock cycles after TX_TRIG± signal assertion (POWER_START LSB)
Digital high-pass bypassed (POWER_START MSB)
Decimate by 2 (M = 00001); digital gain = 16 (g = 100)
Point to Coefficient Block 00
demodulation frequency = fSAMPLE/8
demodulation frequency = fSAMPLE/8
Set index profile (required after profile memory writes)
Set output data format
Chip run (TGC mode)2
16-bit, four-lane mode
Enables automatic serializer/deserializer (SERDES) sample clock counter
ILAS enabled
Enable start code identifier
Set START_CODE MSB
Set START_CODE LSB
Rev. A| Page 46 of 60
Data Sheet
AD9671
Address
0x150
0x182
0x181
0x186
Value
0x03
0x82
0x02
0xAA
0x10C3
0x00F
0x02B
0x20
0x18
0x40
Description
JESD204B scrambler disabled and four-lane configuration (L = 4)
Autoconfigures PLL
PLL N divider = ÷20
Disable continuous data resync (continuous data resync is not recommended during real-time
scanning; one-time data resync is sufficient)
Set SPI TX_TRIG± and index profile
Set low-pass filter cutoff frequency, bandwidth mode
Set analog LPF and HPF to defaults, tune filters4
1
Setting the profile index requires an additional SPI write in SPI MSB mode before the chip is run to complete the current profile buffer update.
Running the chip from full power-down mode requires 375 μs wake-up time as listed in Table 3.
3
The software TX_TRIG trigger switches the demodulator/decimator digital block to a running state. It may not be needed if hardware the TX_TRIX signal is used to run
the digital block.
4
Tuning the filters requires 512 ADC clock cycles.
2
Rev. A| Page 47 of 60
AD9671
Data Sheet
Table 33. AD9671 Memory Map Registers
Addr. Register
(Hex) Name
Bit 7 (MSB)
Chip Configuration Registers
0
0x000 CHIP_
PORT_
CONFIG
0x001
CHIP_ID
0x002
CHIP_
GRADE
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0 (LSB)
Default
Value
LSB first
0 = off
(default)
1 = on
SPI reset
0 = off
(default)
1 = on
1
1
SPI reset
0 = off
(default)
1 = on
LSB first
0 = off
(default)
1 = on
0
0x18
Chip ID Bits[7:0]
AD9671 = 0xA7 (default)
X
X
Speed mode
(identify device
variants of chip ID)
00: Mode I
(40 MSPS) (default)
01: Mode II (65 MSPS)
10: Mode III (80 MSPS)
11: Mode III (125 MSPS)
0xA7
X
X
X
X
0x0X
Device Index and Update Registers
0x004 DEVICE_
X
X
INDEX_2
X
X
Data
Channel H
0 = off
1 = on
(default)
Data
Channel G
0 = off
1 = on
(default)
Data
Channel F
0 = off
1 = on
(default)
Data
Channel E
0 = off
1 = on
(default)
0x0F
0x005
1
1
Data
Channel D
0 = off
1 = on
(default)
Data
Channel C
0 = off
1 = on
(default)
Data
Channel B
0 = off
1 = on
(default)
Data
Channel A
0 = off
1 = on
(default)
0x3F
DEVICE_
INDEX_1
X
X
Rev. A| Page 48 of 60
Comments
Nibbles
mirrored so
that LSB or
MSB first mode
is set correctly,
regardless of
shift mode. SPI
reset reverts
all registers
(including the
JESD ones),
except Reg.
0x000 to their
default values
and Reg. 0x000,
Bit 2 and Bit 5
are automatically
cleared.
Default is
unique chip
ID, different
for each
device; read
only register.
Speed mode
used to
differentiate
ADC speed
power modes
(must update
Reg. 0x0FF to
initiate mode
setting).
Bits are
set to determine which
on-chip device
receives the
next write
command.
Bits are
set to determine which
on-chip device
receives the
next write
command.
Data Sheet
Addr.
(Hex)
0x0FF
Register
Name
DEVICE_
UPDATE
AD9671
Bit 7 (MSB)
X
Program Function Registers
0x008 GLOBAL_
X
MODES
0x009
GLOBAL_
CLOCK
X
Bit 5
X
Bit 4
X
Bit 3
X
Bit 2
X
LNA input
impedance
0 = 6 kΩ
(default)
1= 3 kΩ
X
X
0
0
0x01
Determines
generic modes
of chip
operation
(global).
X
X
X
Internal power-down mode
000 = chip run (TGC mode)
001 = full power-down (default)
010 = standby
011 = reset all JESD registers
100 = CW mode (TGC power-down)
X
X
DCS
0 = off
1 = on
(default)
0x01
X
X
0x00
Turns the
internal duty
cycle stabilizer
(DCS) on and
off (global).
Monitor PLL
lock and link
ready status
(read only,
global).
0x00A PLL_
STATUS
PLL lock
status
0 = not
locked
1 = locked
X
X
X
0x00D TEST_IO
User test
mode
0 = continuous,
repeat user
patterns
(1, 2, 3, 4,
1, 2, 3, 4,
…)
(default)
1 = single
clock cycle
user
patterns,
then zeros
(1, 2, 3, 4,
0, 0, …)
X
X
Reset PN
long gen
0 = on, PN
long
running
(default)
1 = off, PN
long held
in reset
Reset PN
short gen
0 = on,
PN short
running
(default)
1 = off,
PN short
held in
reset
X
X
X
0x00E
GPO
Bit 1
X
X
Bit 0 (LSB)
X
Default
Value
0x00
Bit 6
X
JESD204B
link ready
status
0 = link not
ready
(default)
1 = link
ready, PLL
locked
Comments
A write to
Reg. 0x0FF
(the value
does not
matter) resets
all default
register values
(analog and
ADC registers
only not,
JESD204B
registers and
not Reg. 0x000
or Reg. 0x002,
Bits[5:4]) if
Reg 0x02 has
been previously written
since the last
reset/load of
defaults.
Output test mode
0000 = off (default)
0001 = midscale short
0010 = +FS short
0011 = −FS short
0100 = checkerboard output
0101 = PN sequence long
0110 = PN sequence short
0111 = one-/zero-word toggle
1000 = user input
1001 to 1110 = reserved
1111 = ramp output
0x00
When this
register is set,
the test data is
placed on the
output pins in
place of
normal data
(local).
General-purpose digital outputs
0x00
Values placed
on GPO0 to
GPO3 pins
(global).
Rev. A| Page 49 of 60
AD9671
Addr.
(Hex)
0x00F
Register
Name
FLEX_
CHANNEL_
INPUT
0x010
FLEX_
OFFSET
FLEX_
GAIN
0x011
Data Sheet
Bit 7 (MSB)
X
Bit 6
Bit 5
Bit 4
Filter cutoff frequency control
0 0000 = 1.45 × (1/3) × fSAMPLE
0 0001 = 1.25 × (1/3) × fSAMPLE
0 0010 = 1.13 × (1/3) × fSAMPLE
0 0011 = 1.0 × (1/3) × fSAMPLE (default)
0 0100 = 0.9 × (1/3) × fSAMPLE
0 0101 = 0.8 × (1/3) × fSAMPLE
0 0110 = 0.75 × (1/3) × fSAMPLE
0 0111 = not applicable
0 1000 = 1.45 × (1/4.5) × fSAMPLE
0 1001 = 1.25 × (1/4.5) × fSAMPLE
0 1010 = 1.13 × (1/4.5) × fSAMPLE
0 1011 = 1.0 × (1/4.5) × fSAMPLE
0 1100 = 0.9 × (1/4.5) × fSAMPLE
0 1101 = 0.8 × (1/4.5) × fSAMPLE
0 1110 = 0.75 × (1/4.5) × fSAMPLE
0 1111 = not applicable
1 0000 = 1.45 × (1/6) × fSAMPLE
1 0001 = 1.25 × (1/6) × fSAMPLE
1 0010 = 1.13 × (1/6) × fSAMPLE
1 0011 = 1.0 × (1/6) × fSAMPLE
1 0100 = 0.9 × (1/6) × fSAMPLE
1 0101 = 0.8 × (1/6) × fSAMPLE
1 0110 = 0.75 × (1/6) × fSAMPLE
1 0111 = not applicable
X
1
0
Bit 3
Bit 2
BW mode
0 = low
(default,
8 MHz to
18 MHz)
1 = high
(13.5 MHz
to 30 MHz)
Bit 1
X
Bit 0 (LSB)
X
Default
Value
0x18
0
0
0
0
0x20
Reserved.
0x06
LNA and PGA
gain
adjustment
(global).
LNA bias
00 = high
01 = midhigh (default)
10 = midlow
11 = low
0
0
0x09
LNA bias
current
adjustment
(global).
0x00
Reserved.
Output data format
00 = offset binary
01 = twos complement
(default)
10 = gray code
11 = reserved
0x01
Data output
modes (local).
LNA gain
00 = 15.6 dB
01 = 17.9 dB
10 = 21.6 dB
(default)
11 = reserved
PGA gain
00 = 21 dB
01 = 24 dB (default)
10 = 27 dB
11 = 30 dB
Digital VGA gain control
0000 = GAIN± pins enabled (default)
0001 = 0.0 dB (maximum gain, GAIN± pins disabled)
0010 = −3.5 dB
0011 = −7.0 dB
…
1110 = −45 dB
1111 = reserved (do not use)
X
X
X
X
1
PGA bias
0 =100%
(default)
1 = 60%
0
0
X
Output data
invert
0 = disable
(default)
1 = enable
X
X
Output
preemphasis
0 = off
(default)
1 = on
1
0x61
Data output
levels (global).
X
X
X
X
0x00
Reserved
(global).
Reserved
(global).
Reserved
(global).
User-Defined
Pattern 1, LSB
(global).
0x012
BIAS_
CURRENT
0x013
RESERVED_
13
OUTPUT_
MODE
0
0
0
X
X
X
0x015
OUTPUT_
ADJUST
X
0x016
RESERVED_
16
RESERVED_
17
FLEX_VREF
X
Output
data
enable
0=
enable
(default)
1=
disable
CML output drive level adjustment
000 = reserved
001 = reserved
010 = 368 mV
011 = reserved
100 = 293 mV
101 = 286 mV
110 = 262 mV (default)
111 = 238 mV
X
X
X
X
X
X
X
X
X
X
X
0x00
X
X
X
X
X
1
0
0
0x04
USER_
PATT1_LSB
B7
B6
B5
B4
B3
B2
B1
B0
0x00
0x014
0x017
0x018
0x019
Comments
Antialiasing
filter cutoff
(global).
0
Rev. A| Page 50 of 60
Data Sheet
Addr. Register
(Hex) Name
0x01A USER_
PATT1_
MSB
0x01B USER_
PATT2_LSB
AD9671
Bit 7 (MSB)
B15
Bit 6
B14
Bit 5
B13
Bit 4
B12
Bit 3
B11
Bit 2
B10
Bit 1
B9
Bit 0 (LSB)
B8
Default
Value
0x00
B7
B6
B5
B4
B3
B2
B1
B0
0x00
0x01C USER_
PATT2_
MSB
0x01D USER_
PATT3_LSB
B15
B14
B13
B12
B11
B10
B9
B8
0x00
B7
B6
B5
B4
B3
B2
B1
B0
0x00
0x01E
USER_
PATT3_
MSB
USER_
PATT4_LSB
B15
B14
B13
B12
B11
B10
B9
B8
0x00
B7
B6
B5
B4
B3
B2
B1
B0
0x00
USER_
PATT4_
MSB
FLEX_
SERIAL_
CTRL
B15
B14
B13
B12
B11
B10
B9
B8
0x00
0
X
X
X
Lane low
rate:
0 = normal
(default)
1 = low
output
rate
(<1 Gbps)
X
X
SERIAL_
CH_STAT
Lane mode
00 = reserved (default)
01 = 2 channels/lane
(4 lanes)
10 = 4 channels/lane
(2 lanes)
11 = 8 channels/lane
(1 lane)
X
X
0x02B FLEX_
FILTER
X
X
X
Bypass
analog
HPF
0 = off
(default)
1 = on
X
0x02C LNA_
TERM
X
Enable
automatic
low-pass
tuning
1 = on
(self
clearing)
X
X
X
X
X
0x02D CW_
ENABLE_
PHASE
X
X
X
CW
Doppler
channel
enable
1 = on
0 = off
0x01F
0x020
0x021
0x022
X
X
Output word length
00 = 12 bits (default)
01 = 14 bits
10 = 16 bits
11 = reserved
0x00
Channel
powerdown
1 = on
0 = off
(default)
Analog high-pass filter
cutoff
00 = fLP/12.00 (default)
01 = fLP/9.00
10 = fLP/6.00
11 = fLP/3.00
0x00
Used to power
down
individual
channels
(local).
0x00
Filter cutoff
(global)
(fLP = low-pass
filter cutoff
frequency).
0x00
LNA active
termination/
input
impedance
(global).
Phase of
demodulators
(local, chip).
LO-x, LOSW-x connection
00 = RFB1 + 50 Ω (default)
01 = (RFB1||RFB2) + 50 Ω
10 = RFB2 + 50 Ω
11 = ∞
I/Q demodulator phase
0000 = 0° (default)
0001 = 22.5° (not valid for 4LO mode)
0010 = 45°
0011 = 67.5° (not valid for 4LO mode)
0100 = 90°
0101 = 112.5° (not valid for 4LO mode)
0110 = 135°
0111 = 157.5° (not valid for 4LO mode)
1000 = 180°
1001 = 202.5° (not valid for 4LO mode)
1010 = 225°
1011 = 247.5° (not valid for 4LO mode)
1100 = 270°
1101 = 292.5° (not valid for 4LO mode)
1110 = 315°
1111 = 337.5° (not valid for 4LO mode)
Rev. A| Page 51 of 60
Comments
User-Defined
Pattern 1, MSB
(global).
User-Defined
Pattern 2, LSB
(global).
User-Defined
Pattern 2, MSB
(global).
User-Defined
Pattern 3, LSB
(global).
User-Defined
Pattern 3, MSB
(global).
User-Defined
Pattern 4, LSB
(global).
User-Defined
Pattern 4, MSB
(global).
Lane setting
control (global).
0x00
AD9671
Addr.
(Hex)
0x02E
Register
Name
CW_LO_
MODE
0x02F
CW_
OUTPUT
0x102
RESERVED_
102
RESERVED_
103
RESERVED_
104
RESERVED_
105
RESERVED_
106
RESERVED_
107
RESERVED_
108
VGA_TEST
Data Sheet
Bit 6
RESET±
with
MLO±
clock edge
0=
synchronous
(default)
1=
asynchronous
Bit 5
Synchronous
RESET±
sampling
MLO±
clock edge
0 = falling
(default)
1 = rising
Bit 4
RESET±
polarity
0=
active
high
(default)
1=
active
low
Bit 3
MLO± and
RESET±
buffer
enable (in
all modes
except CW
mode)
0 = powerdown
(default)
1 = enable
Bit 2
0
0
0
0
0
0
0
0x80
Global.
0
0
0
0
0
0
0
0X00
Reserved.
0
0
0
0
0
0
0
0
0X00
Reserved.
0
0
1
1
1
1
1
1
0x3F
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
0
0
0
0
0
0
X
X
Reserved.
0
0
0
0
0
0
0
0
Read
only
0x00
X
X
X
VGA/
AAF test
enable
0 = off
(default)
1 = on
X
0x00
VGA/AAF test
mode enables
AAF output to
the GPO2/
GPO3 pins
(global).
0x10C PROFILE_
INDEX
X
X
0x00
Index for
profile
memory
selects active
profile
(global).
0x10D RESERVED_
10D
0x10E RESERVED_
10E
1
1
Manual
TX_TRIG
signal
0 = off,
use pin
(default)
1 = on,
autogenerate
TX_TRIG
(self clears)
1
1
1
1
1
1
0xFF
Reserved.
1
1
1
1
1
1
1
1
0xFF
Reserved.
0x103
0x104
0x105
0x106
0x107
0x108
0x109
Bit 1
Bit 0 (LSB)
LO mode
00X = 4LO, 3rd to 5th odd harmonic
rejection (default)
010 = 8LO, 3rd to 5th odd harmonic
rejection
011 = 8LO, 3rd to 13th odd harmonic
rejection
100 = 16LO, 3rd to 5th odd harmonic
rejection
101 = 16LO, 3rd to 13th odd harmonic
rejection
11X = reserved
Default
Value
0x00
Bit 7 (MSB)
Enable
JESD during CW
0: JESD
link
disabled
during CW
(default)
1: JESD
link
enabled
during CW
(switching
activity
can degrade CW
performance)
CW output
dc bias
voltage
0 = bypass
1 = enable
(default)
0
VGA/AAF output test mode
000 = Channel A (default)
001 = Channel B
010 = Channel C
011 = Channel D
100 = Channel E
101 = Channel F
110 = Channel G
111 = Channel H
Profile index[4:0]
Rev. A| Page 52 of 60
Comments
CW mode
functions
(global).
Reserved.
Data Sheet
Addr.
(Hex)
0x10F
Register
Name
DIG_
OFFSET_
CAL
0x110
AD9671
Bit 7 (MSB)
0
Bit 6
0
Bit 5
0
Bit 4
0
DIG_
OFFSET_
CORR1
DIG_
OFFSET_
CORR2
D7
D6
D5
D4
0x112
POWER_
MASK_
CONFIG
X
0x113
DIG_
DEMOD_
CONFIG
X
0x115
CHIP_
ADDR_EN
X
0x116
ANALOG_
TEST_
TONE
X
X
X
0x117
DIG_SINE_
TEST_FREQ
X
X
X
0x111
D15
Bit 3
Digital
offset
calibration
status
0 = not
complete
(default)
1=
complete
D3
Bit 2
Bit 1
Bit 0 (LSB)
Digital offset calibration
000 = disable correction, reset correction
value (default)
001 = average 210 samples
010 = average 211 samples
…
111 = average 216 samples
D2
D1
D0
D14
D13
D12
D11
D10
D9
D8
Digital offset calibration (read back if autocalibration enabled with Register 0x10F; otherwise, force
correction value)
Offset correction = [D15:D0] × full scale/216
0111 1111 1111 1111 (215 − 1) = +1/2 full scale − 1/216 full scale
0111 1111 1111 1110 (215 – 2) = +1/2 full scale − 2/216 full scale
…
0000 0000 0000 0001 (+1) = +1/216 full scale
0000 0000 0000 0000 = no correction (default)
1111 1111 1111 1111 (−1) = −1/216 full scale
…
1000 0000 0000 0000 (−215) = −1/2 full scale
X
X
Power-up setup time (POWER_SETUP)
0 0000 = 0
0 0001 = 1 × 40/fSAMPLE
0 0010 = 2 × 40/fSAMPLE (default)
0 0011 = 3 × 40/fSAMPLE
…
1 1111 = 31 × 40/fSAMPLE
DemodBaseband
Decimator and filter
DeciX
Digital
decimator ulator
enable
mator
high-pass
0 = enable 0 = enable
00 = RF 2× decimator
gain
filter
(default)
(default)
bypassed (default)
0 = enable scale
1 = bypass 1 = bypass
01 = RF 2× decimator
0 = no
(default)
enabled and low
1 = bypass gain
bandwidth filter
(default)
1X = RF 2× decimator
1 = 4×
enabled and high
gain
bandwidth filter
(shift
decimator
output
by 2)
Chip address qualifier
X
Chip
0 0000 (default)
address
(If read, returns the state of ADDR0 to ADRR4 pins)
mode
0 = disable
(default)
1 = enable
X
Analog test tone
amplitude
(see Table 23 to Table 25)
Analog test tone
frequency
00 = fSAMPLE/4 (default)
01 = fSAMPLE/8
10 = fSAMPLE/16
11 = fSAMPLE/32
Digital test tone frequency
0 0000 = 1 × fSAMPLE/64
0 0001 = 2 × fSAMPLE/64
…
1 1111 = 32 × fSAMPLE/64
Rev. A| Page 53 of 60
Default
Value
0x00
0x00
0x00
Comments
Control digital
offset
calibration
enable and
number of
samples used
(global).
Offset
correction LSB
(local, chip).
Offset
correction
MSB (local,
chip).
0x02
POWER_SETUP
time is used to
set the powerup time
(global).
0x00
Enable stages
of the digital
processing
(global).
0x00
Chip address
mode enables
the addressing
of devices if
the value of
chip address
qualifier equals
the state on
the address
pins, ADDRx
(global).
Analog test
tone amplitude and
frequency
(global).
0x00
0x00
Digital sine
test tone
frequency
(global).
AD9671
Addr.
(Hex)
0x118
Register
Name
DIG_SINE_
TEST_AMP
0x119
DIG_SINE_
TEST_
OFFSET
0x11A
TEST_
MODE_
CH_
ENABLE
Data Sheet
Bit 5
X
Bit 4
X
Bit 3
Bit 2
Bit 1
Bit 0 (LSB)
Digital test tone amplitude
0000 = AFULL-SCALE (default)
0001 = AFULL-SCALE/2
0010 = AFULL-SCALE/22
…
1111 = AFULL-SCALE/215
Offset exponent (b)
000 = 0 (default)
001 = 1
…
111 = 7
Default
Value
0x00
Bit 7 (MSB)
X
Bit 6
X
Comments
Digital sine
test tone
amplitude
(global).
0x00
Digital sine
test tone
offset (global).
Ch. H
enable
0 = off
(default)
1 = on
X
0x00
Enable
channels for
test mode
(global).
0x00
Enable digital
test modes
(local).
0x00
Reserved.
0x11C RESERVED_
11C
0x11D RESERVED_
11D
0x11E RESERVED_
11E
0x11F RESERVED_
11F
0x120 CW_TEST_
TONE
0
Offset multiplier (a)
0 1111 = 15
0 1110 = 14
…
0 0000 = 0 (default)
1 1111 = −1
…
1 0000 = −16
Offset = AFULL-SCALE × a × 2−(13 − b)
Offset range is ~0.5 dB
Maximum positive offset = 15 × 2−(13 − 7) = 0.25 × AFULL-SCALE
Maximum negative offset = −16 × 2−(13 − 7) ≈ −0.25 × AFULL-SCALE
Ch. A
Ch. B
Ch. C
Ch. D
Ch. E
Ch. F
Ch. G
enable
enable
enable
enable
enable
enable
enable
0 = off
0 = off
0 = off
0 = off
0 = off
0 = off
0 = off
(default)
(default)
(default)
(default)
(default)
(default)
(default)
1 = on
1 = on
1 = on
1 = on
1 = on
1 = on
1 = on
X
X
X
X
Datapath test mode selection
000 = disable test modes (default)
001 = enable digital sine test mode
010 = enable decimator filter test
(output of decimator is the sequence of
filter coefficients)
011 = enable channel ID test mode
(16-bit data = digital ramp (7 bits) + I/Q bit
+ Channel ID (3 bits) + Chip Address (5 bits)
100 = enable analog test tone
101 = reserved
…
111 = reserved
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0x00
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
0
CW I/Q
output
swap
0 = disable
(default)
1 = enable
LNA offset
cancellation
0 = enable
(default)
1 = disable
LNA offset cancellation
transconductance
00 = 0.5 mS (default)
01 = 1.0 mS
10 = 1.5 mS
11 = 2.0 mS
0
0x00
0x142
JESD204B
power
during
standby
0 = remain
powered
up
(default)
1 = powerdown
JESD204B
tail bit
value
0 = zeros
(default)
1: PN
sequence
JESD204B
test mode
enable
0 = disable
(default)
1 = enable
JESD204B
lane sync
enable
0=
disable
(default)
1=
enable
Power down
JESD204B
link
0 = link
enabled
(default)
1 = link
powered
down
0x00
Sets the
frequency of
the analog test
tone to fLO in
CW Doppler
mode; enables
I/Q output
swap; LNA
offset cancellation control
(global).
JESD204B
configuration
(global).
0x11B TEST_
MODE_
CONFIG
JTX_LINK_
CTRL1
CW analog test tone
override for Reg. 0x116,
Bits[1:0]
00 = disable override
(default)
01 = set analog test tone
frequency to fLO
1X = set analog test tone
frequency to dc
JESD204B ILAS enable
00 = disable (default)
01 = enable
10 = always on, test
mode
11 = reserved
Rev. A| Page 54 of 60
JESD204B
serial
frame
alignment
character
insertion
(FACI)
disable
0: FACI
enabled
1: FACI
disabled
Data Sheet
Addr.
(Hex)
0x143
Register
Name
JTX_LINK_
CTRL2
0x144
JTX_LINK_
CTRL3
0x145
JTX_LINK_
CTRL4
0x146
JTX_DID_
CFG
JTX_BID_
CFG
JTX_LID0_
CFG
JTX_LID1_
CFG
JTX_LID2_
CFG
JTX_LID3_
CFG
RESERVED_
14C
RESERVED_
14D
RESERVED_
14E
RESERVED_
14F
JTX_SCR_
L_CFG
0x147
0x148
0x149
0x14A
0x14B
0x14C
0x14D
0x14E
0x14F
0x150
AD9671
Bit 7 (MSB)
Bit 6
Checksum
enable
0 = enable
(default)
1 = disable
Checksum
algorithm
0 = add
parameter
(default)
1 = add
packed
octets
Bit 5
SYNCINB
signal
polarity
0 = not
inverted
(default)
1=
inverted
Bit 1
Bit 0 (LSB)
10B
10B
transmit bit
transmit
mirror
bit invert
0 = not
0 = not
mirrored
inverted
(default)
(default)
1=
1=
mirrored
inverted
SERDOUTx±
JESD204B test mode selection
JESD204B test pattern
0000 = off (default)
input selection
0001 = alternating checkerboard
00 = reserved (default)
0010 = 1-/0-word toggle
01 = 10-bit test data
0011 = PN sequence long
injected at output of
0100 = PN sequence short
8B/10B encoder
0101 = continuous/repeat user test pattern
10 = 8-bit test data
0110 = single user test pattern
injected at input of
0111 = ramp output
scrambler
1000 = modified RPAT sequence
11 = reserved
1001 = reserved
…
1111 = reserved
Initial lane alignment sequence repeat count
0000 0000 = 4 × K + 1 (default)
0000 0001 = 4 × K + 2
…
1111 1111 = 4 × K + 128
JESD204B serial device identification (DID) number
X
X
X
X
X
X
Bit 4
0
X
Bit 3
0
Bit 2
8B/10B
encoder
0 = enable
(default)
1 = bypass
(test mode
only)
Default
Value
0x00
Comments
JESD204B
configuration
(global).
0x00
JESD204B test
mode and
checksum
controls
(global).
0x00
JESD204B ILAS
repeat count
(global).
0x00
Global.
0x00
Global.
X
JESD204B serial bank identification (BID) number
(extension to DID)
Serial lane identification (LID) number for Lane 1
0x00
Global.
X
X
Serial lane identification (LID) number for Lane 2
0x01
Global.
X
X
X
Serial lane identification (LID) number for Lane 3
0x02
Global.
X
X
X
Serial lane identification (LID) number for Lane 4
0x03
Global.
0
0
0
0
0
0
0
0
0x00
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
JESD204B
serial
scrambler
mode
0=
disabled
1=
enabled
(default)
X
X
X
X
X
0x83
JESD204B
scrambler and
lane
configuration
(global).
Rev. A| Page 55 of 60
Lanes per link
00 = one lane (L = 1)
01 = two lanes (L = 2)
10 = reserved
11 = four lanes (L = 4)
(default)
AD9671
Data Sheet
Addr.
(Hex)
0x151
Register
Name
JTX_F_CFG
Bit 7 (MSB)
X
Bit 6
X
Bit 5
X
0x152
JTX_K_CFG
X
X
X
0x153
JTX_M_
CFG
X
X
X
X
0x154
JTX_CS_N_
CFG
X
Control
bits per
sample
0 = none
(CS = 0,
default,
read only)
X
0
0x155
JTX_SCV_
NP_CFG
0
0
0
0
0x156
JTX_JV_S_
CFG
X
X
0
0x157
JTX_HD_
CF_CFG
0
0
Number of
clocks
SYNCINB
signal
must be
low for
synchronization to
begin
0=2
frame
clock
cycles
1=4
frame
clock
cycles
(default)
0
0x158
JTX_RES1_
CFG
JTX_RES2_
CFG
0
0
0
0
0
0
0x159
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0 (LSB)
Number of octets per frame (F)
F = (M × 2)/(L)
0 0000 = reserved
…
0 0011 = 4 octets ( M = 8, L = 4, default)
0 0100 = reserved
…
0 0111 = 8 octets (M = 8, L = 2) or (M = 16, L = 4)
0 1000 = reserved
…
0 1111 = 16 octets (M = 8, L = 1) or (M = 16, L = 2)
1 0000 = reserved
…
1 1111 = 32 octets (M = 16, L = 1)
Number of frames per multiframe (K)
0 0000 = 1
0 0001 = 2
…
1 1111 = 32 (default)
Number of converters per link
0000 = reserved
…
0111 = 8 channels, real data (M = 8)
1000 = reserved
…
1111 = 8 channels, quadrature data (M = 16)
Output resolution (N)
0000 = reserved
…
1011 = 12 bits
1100 = reserved
1101 = 14 bits
1110 = reserved
1111 = 16 bits (default)
Bits per output sample (N’)
0000 = reserved
…
1110 = reserved
1111 = 16 (default)
0
0
0
Samples per
channel per
frame (S)
0=1
sample
(default,
read only)
1=2
samples
Default
Value
0x03
Comments
JESD204B
number of
octets per
frame (read
only, global).
0x0F
JESD204B
frames per
multiframe
(global).
0x07
JESD204B
number of
converter per
link (read only,
global).
0x0F
JESD204B
serializer
number of bits
per channel
(global).
0x0F
JESD204B
number of bits
per samples
(global, read
only).
Number of
clocks SYNCINB
signal must be
low for synchronization to
begin (global).
0x20
0x00
0
Control words per frame clock per link
0 0000 = 0 (default)
0 0001 = reserved
…
1 1111 = reserved
0
0
0
0
0x00
JESD204B
control words
per frame
(global, read
only).
Reserved.
0
0
0x00
Reserved.
0
Rev. A| Page 56 of 60
0
0
Data Sheet
Addr. Register
(Hex) Name
0x15A JTX_
CHKSUM0_
CFG
AD9671
Bit 7 (MSB)
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Checksum value for Lane 1 (FCHK)
Bit 1
Bit 0 (LSB)
Default
Value
0x3C
RESERVED_
15E
RESERVED_
15F
RESERVED_
160
RESERVED_
161
RESERVED_
170
RESERVED_
171
RESERVED_
172
RESERVED_
173
RESERVED_
174
JTX_CLK_
CNTL_1
JTX_CLK_
CNTL_2
0
1
1
0
1
1
0
0
0x3C
Comments
JESD204B
checksum
value Lane 1
(global, read
only).
JESD204B
checksum
value Lane 2
(global, read
only).
JESD204B
checksum
value Lane 3
(global, read
only).
JESD204B
checksum
value Lane 4
(global, read
only).
Reserved.
0
1
1
0
1
1
0
0
0x3C
Reserved.
0
1
1
0
1
1
0
0
0x3C
Reserved.
0
1
1
0
1
1
0
0
0x3C
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
1
1
1
1
1
1
1
1
0xFF
Reserved.
1
1
1
1
1
1
1
1
0xFF
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
0
0
0
0
1
1
1
1
0x0F
Reserved.
1
0
0
0
0
1
1
1
0x87
Reserved.
0
0
0
0
0
0x00
PLL N divider
setting (Z)
(global).
0x182
PLL_
STARTUP
0
0
0
0
0x02
PLL control
(global).
0x183
RESERVED_
183
RESERVED_
184
PLL autoconfigure
0 = disable
(default)
1 = enable
0
PLL N divider setting (in powers of 2)
000 = divide by 1 (Z = ÷5, default)
001 = divide by 2 (Z = ÷10)
010 = divide by 4 (Z = ÷20)
011 = divide by 8 (Z = ÷40)
100 = divide by 16 (Z = ÷80)
101 = reserved
110 = reserved
111 = reserved
0
1
0
0
0
0
0
1
1
1
0x07
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
0x15B JTX_
CHKSUM1_
CFG
Checksum value for Lane 2 (FCHK)
0x3D
0x15C JTX_
CHKSUM2_
CFG
Checksum value for Lane 3 (FCHK)
0x3E
0x15D JTX_
CHKSUM3_
CFG
Checksum value for Lane 4 (FCHK)
0x3F
0x15E
0x15F
0x160
0x161
0x170
0x171
0x172
0x173
0x174
0x180
0x181
0x184
Rev. A| Page 57 of 60
AD9671
Addr.
(Hex)
0x186
Register
Name
DATA_
VALID_
RESYNC
0x188
START_
CODE_EN
Data Sheet
Bit 7 (MSB)
1
Bit 6
0
Bit 5
1
Bit 4
0
Bit 3
One time
data resync
with
JESD204B
clock after
TX_TRIG
0: disable
resync
1: enable
resync
(default)
Bit 2
Continuous
data resync
with
JESD204B
clock
0: disable
resync
1: enable
resync
(default)
0
0
0
0
0
0
Bit 1
One time
SYSREF
resync
with
JESD204B
clock after
TX_TRIG
0: disable
resync
1: enable
resync
(default)
0
RESERVED
_189
0x18A RESERVED
_18A
0x18B START_
CODE_
MSB
0x18C START_
CODE_LSB
0x190 FRAME_
SIZE_MSB
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
1
X
X
X
0x191
0
0
0
0x189
0x192
0x193
0x194
0x195
0x196
0x197
0x198
0x199
RESERVED_
191
RESERVED_
192
RESERVED_
193
RESERVED_
194
RESERVED_
195
RESERVED_
196
RESERVED_
197
RESERVED_
198
SAMPLE_
CLOCK_
COUNTER
0x19A RESERVED_
19A
0x19B RESERVED_
19B
Bit 0 (LSB)
Continuous
SYSREF
resync with
JESD204B
clock
0: disable
resync
(default)
1: enable
resync
Default
Value
0xAE
Comments
Data and
SYSREF resync.
0x01
Enable start
code identifier
(global).
0
Start code
identifier
0 = disable
1 = enable
(default)
0
0x00
Reserved.
0
0
0
0x00
Reserved.
0
1
1
1
0x27
Start code
MSB (global).
1
0
0
1
0
0x72
X
X
X
X
0x10
0
Automatically
set frame
size
0=
disable
1=
enable
(default)
0
Start code LSB
(global).
Automatically
set frame size
(global).
0
0
0
0
0x00
Reserved.
0
0
1
1
0
0
0
0x18
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
0
0
0
1
1
1
0
0
0x1C
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
0
0
0
1
1
0
0
0
0x18
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
0
0
0
0
0
0
0
0
0x00
Reserved.
SERDES
clock
counter
0 = disable
(default)
1 = enable
0
0
0
0
0
0
0
0
0x00
Enables
automatic
SERDES
sample clock
counter.
0
0
0
0
0
0
0
0x00
Reserved.
0
1
1
1
0
0
0
0
0x70
Reserved.
Rev. A| Page 58 of 60
Data Sheet
Addr. Register
(Hex) Name
0x19C JTX_
FRAME_
SIZE
AD9671
Bit 7 (MSB)
X
0x19D RESERVED_ 0
19D
0x19E RESERVED_ 0
19E
Coefficient Registers
0x1000 Coefficient
memory
to
0x1FFF
Profile Memory Registers
0xF00 Profile
to
memory
0xFFF
Bit 3
X
Bit 2
X
Bit 1
0
Bit 0 (LSB)
0
Default
Value
0x10
0
Bit 4
Set frame
size
automatically
0=
disable
1=
enable
(default)
0
0
0
0
0
0x00
Reserved.
0
1
0
0
0
0
0x10
Reserved.
256 × 112 bits
0x00
Global.
32 × 64 bits
0x00
Global.
Bit 6
X
Bit 5
X
0
0
Comments
Automatically
set JESD204B
frame size
(global).
MEMORY MAP REGISTER DESCRIPTIONS
Profile Index and Software TX_TRIG (Register 0x10C)
For more information about the SPI memory map and other
functions, consult the AN-877 Application Note, Interfacing
to High Speed ADCs via SPI.
The vector profile is selected using the profile index in
Register 0x10C, Bits[4:0]. The software TX_TRIG control in Bit 5
generates a TX_TRIG signal internal to the device. This signal
is asynchronous to the ADC sample clock. Therefore, do not
use this signal to align the data output, reset the digital
demodulator and decimator, or initiate advanced power mode
across multiple devices in the system. The external pin-driven
TX_TRIG± control is recommended for systems that require
synchronization of these features across multiple AD9671
devices.
Update (Register 0x0FF)
All registers except Register 0x002 are updated as soon as they
are written. Writing to Register 0x0FF (the value written is don’t
care) initializes and updates the speed mode (Address 0x002)
and resets all other registers to their default values (analog and
ADC registers only; not the JESD204B registers, Register 0x000,
or Register 0x002). Set the speed mode in Register 0x002 and
write to Register 0x0FF at the beginning of the setup of the SPI
writes after the device is powered up to avoid rewriting other
registers after Register 0x0FF is written.
Rev. A| Page 59 of 60
AD9671
Data Sheet
OUTLINE DIMENSIONS
A1 BALL
CORNER
10.10
10.00 SQ
9.90
A1 BALL
CORNER
12 11 10 9 8
7 6 5
4
3
2
1
A
B
C
D
8.80
BSC SQ
E
F
G
H
0.80
J
K
L
M
TOP VIEW
0.60
REF
BOTTOM VIEW
DETAIL A
*1.40 MAX
DETAIL A
0.65 MIN
0.25 MIN
0.50
COPLANARITY
0.45
0.20
0.40
BALL DIAMETER
*COMPLIANT WITH JEDEC STANDARDS MO-275-EEAB-1
WITH EXCEPTION TO PACKAGE HEIGHT.
10-21-2010-B
SEATING
PLANE
Figure 62. 144-Ball Chip Scale Package, Ball Grid Array [CSP_BGA]
(BC-144-1)
Dimensions shown in millimeters
ORDERING GUIDE
Model1
AD9671KBCZ
AD9671EBZ
1
Temperature Range
0°C to 85°C
Package Description
144-Ball Chip Scale Package, Ball Grid Array [CSP_BGA]
Evaluation Board
Z = RoHs Compliant Part.
©2013–2016 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D11134-0-1/16(A)
Rev. A| Page 60 of 60
Package Option
BC-144-1