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