14-Bit, 200 MSPS/500 MSPS TxDAC+® with 2×/4×/8× Interpolation and Signal Processing AD9784 Preliminary Technical Data FEATURES PRODUCT DESCRIPTION 14-bit resolution, 200 MSPS input data rate Selectable 2×/4×/8× interpolation filters Selectable fDAC/2, fDAC/4, fDAC/8 modulation modes Single or dual-channel signal processing Selectable image rejection Hilbert transform Flexible calibration engine Direct IF transmission features Serial control interface Versatile clock and data interface SFDR: 90 dBc @10 MHz WCDMA ACLR = 80 dBc @ 40 MHz IF DNL = ±0.75 LSB INL = ±1.5 LSB 3.3 V compatible digital Interface On-chip 1.2 V reference 80-lead thermally enhanced TQFP package The AD9784 is a 14-bit, high speed, CMOS DAC with 2×/4×/8× interpolation and signal processing features tuned for communications applications. It offers state of the art distortion and noise performance. The AD9784 was developed to meet the demanding performance requirements of multicarrier and third generation base stations. The selectable interpolation filters simplify interfacing to a variety of input data rates while also taking advantage of oversampling performance gains. The modulation modes allow convenient bandwidth placement and selectable sideband suppression. The flexible clock interface accepts a variety of input types such as 1 V p-p sine wave, CMOS, and LVPECL in single ended or differential mode. Internal dividers generate the required data rate interface clocks. The AD9784 provides a differential current output, supporting single-ended or differential applications; it provides a nominal full-scale current from 10 mA to 20 mA. The AD9784 is manufactured on an advanced low cost 0.25 µm CMOS process. APPLICATIONS Digital quadrature modulation architectures Multicarrier WCDMA, GSM, TDMA, DCS, PCS, CDMA Systems FUNCTIONAL BLOCK DIAGRAM 2× fDAC/2 fDAC/4 fDAC/8 0 90 CLK– 2× ×2 2× ×4 ×8 CLOCK DISTRIBUTION AND CONTROL LPF CSB SCLK ×2/×4/×8/×16 DATA PORT SYNCHRONIZER 2× IOUTA IOUTB SDO ×1 LATCH 16-BIT DAC HILBERT Q CLOCK MULTIPLIER CLK+ ZERO STUFF REFIO SDIO 0 90 DATACLK/ PLL_LOCK ∆t FSADJ RESET 03152-PrD-001 P2B[15:0] 90 SPI P1B[15:0] DATA ASSEMBLER 0 Re()/Im() I REFERENCE CIRCUITS 2× ×1/×2/×4/×8/×16 2× CALIBRATION LATCH Figure 1. Rev. PrC 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 www.analog.com Fax: 781.326.8703 © 2004 Analog Devices, Inc. All rights reserved. AD9784 Preliminary Technical Data TABLE OF CONTENTS Product Highlights ........................................................................... 3 Digital Filter Specifications ........................................................... 23 AD9784–Specifications.................................................................... 4 Digital Interpolation Filter Coefficients.................................. 23 DC Specifications ......................................................................... 4 AD9784 Clock/Data Timing..................................................... 24 Dynamic Specifications ............................................................... 5 Interpolation Modes .................................................................. 27 Digital Specifications ................................................................... 6 Real and Complex Signals......................................................... 28 Pin Configuration and Function Descriptions............................. 7 Modulation Modes..................................................................... 29 Clock .............................................................................................. 7 Power Dissipation ...................................................................... 34 Analog............................................................................................ 8 Dual Channel Complex Modulation with Hilbert ................ 35 Data ................................................................................................ 8 Hilbert Transform Implementation......................................... 36 Serial Interface .............................................................................. 9 Operating the AD9784 Rev E Evaluation Board........................ 40 Definitions of Specifications ......................................................... 10 Power Supplies............................................................................ 40 Typical Performance Charatceristics ........................................... 12 PECL Clock Driver .................................................................... 40 Serial Control Interface.................................................................. 17 Data Inputs.................................................................................. 41 General Operation of the Serial Interface ............................... 17 SPI Port ........................................................................................ 41 Instruction Byte .......................................................................... 17 Operating with PLL Disabled ................................................... 41 Serial Interface Port Pin Descriptions ..................................... 17 Operating with PLL Enabled .................................................... 42 MSB/LSB Transfers..................................................................... 18 Analog Output ............................................................................ 42 Notes on Serial Port Operation ................................................ 18 Outline Dimensions ....................................................................... 52 Mode Control (via SPI Port) ......................................................... 19 ESD Caution................................................................................ 52 REVISION HISTORY Revision PrC: Preliminary Version Rev. PrC | Page 2 of 52 Preliminary Technical Data AD9784 PRODUCT HIGHLIGHTS 1. The AD9784 is a member of a high speed interpolating TxDAC+ family with 16-/14-/12-bit resolutions. 6. Flexible clock with single-ended or differential input: CMOS, 1 V p-p sine wave and LVPECL capability. 2. 2×/4×/8× user selectable interpolating filter eases data rate and output signal reconstruction filter requirements. 7. 3. 200 MSPS input data rate. 4. Ultrahigh speed 500 MSPS DAC conversion rate. Complete CMOS DAC function operates from a 2.7 V to 3.6 V single analog (AVDD) supply and a 2.5 V (DVDD) digital supply. The DAC full-scale current can be reduced for lower power operation, and a sleep mode is provided for low-power idle periods. 5. Internal PLL/clock divider provides data rate clock for easy interfacing. 8. On-chip voltage reference: The AD9784 includes a 1.20 V temperature-compensated band gap voltage reference. Rev. PrC | Page 3 of 52 AD9784 Preliminary Technical Data AD9784–SPECIFICATIONS DC SPECIFICATIONS Table 1. TMIN to TMAX, AVDD1, AVDD2 = 3.3 V, ACVDD, ADVDD, CLKVDD, DVDD, DRVDD = 2.5 V, IOUTFS = 20 mA, unless otherwise noted Parameter RESOLUTION DC Accuracy1 Integral Nonlinearity Differential Nonlinearity ANALOG OUTPUT Offset Error Gain Error (Without Internal Reference) Gain Error (With Internal Reference) Full-Scale Output Current2 Output Compliance Range Output Resistance Output Capacitance REFERENCE OUTPUT Reference Voltage Reference Output Current3 REFERENCE INPUT Input Compliance Range Reference Input Resistance (Ext Reference Mode) Small Signal Bandwith TEMPERATURE COEFFICIENTS Unipolar Offset Drift Gain Drift (Without Internal Reference) Gain Drift (With Internal Reference) Reference Voltage Drift POWER SUPPLY AVDD1, AVDD2 Voltage Range Analog Supply Current (IAVDD1) Analog Supply Current (IAVDD2) IAVDD1 in SLEEP Mode ACVDD, ADVDD Voltage Range Analog Supply Current (IACVDD) Analog Supply Current (IADVDD) CLKVDD Voltage Range Clock Supply Current (ICLKVDD) DVDD Voltage Range Digital Supply Current (IDVDD) DRVDD Voltage Range Digital Supply Current (IDRVDD) Nominal Power Dissipation4 OPERATING RANGE Min Typ 14 Unit Bits 1.5 0.75 LSB LSB TBD 3 % of FSR % of FSR % of FSR mA V kΩ pF 10 –1.0 1.14 Max 20 +1.0 1.20 1 0.1 1.26 V µA 1.25 V MΩ MHz 10 0.5 ppm of FSR/°C ppm of FSR/°C ppm of FSR/°C ppm /°C 3.1 3.3 3.5 V mA mA mA 2.35 2.5 2.65 V mA mA 2.35 2.5 2.65 V mA 2.35 2.5 2.65 V mA 2.35 2.5/3.3 3.5 V mA W °C 1.25 –40 1 Measured at IOUTA driving a virtual ground. Nominal full-scale current, IOUTFS, is 32× the IREF current. 3 Use an external amplifier to drive any external load. 4 Measured under the following conditions: fDATA = 125 MSPS, fDAC = 500 MSPS, 4× Interpolation, fDAC/4 Modulation, Hilbert Off. 2 Rev. PrC | Page 4 of 52 +85 Preliminary Technical Data AD9784 DYNAMIC SPECIFICATIONS Table 2. TMIN to TMAX, AVDD1, AVDD2 = 3.3 V, ACVDD, ADVDD, CLKVDD, DVDD, DRVDD = 2.5 V, IOUTFS = 20 mA, Differential Transformer Coupled Output, 50 Ω Doubly Terminated, unless otherwise noted Parameter DYNAMIC PERFORMANCE Maximum DAC Output Update Rate (fDAC) Output Settling Time (tST) (to 0.025%) Output Propogation Delay5 (tPD) Output Rise Time (10%–90%)6 Output Fall Time (90%–10%)6 Output Noise (IOUTFS = 20 mA) AC LINEARITY—BASEBAND MODE Spurious-Free Dynamic Range (SFDR) to Nyquist (fOUT = 0 dBFS) fDATA = 160 MSPS; fOUT= 1 MHz fDATA = MSPS; fOUT = MHz fDATA = MSPS; fOUT = MHz fDATA = MSPS; fOUT = MHz fDATA = MSPS; fOUT = MHz fDATA = MSPS; fOUT = MHz Two-Tone Intermodulation (IMD) to Nyquist (fOUT1 = fOUT2 = –6 dBFS) fDATA = 160 MSPS; fOUT1 = 25 MHz; fOUT2 = 31 MHz fDATA = MSPS; fOUT1 = MHz; fOUT2 = MHz fDATA = MSPS; fOUT1 = MHz; fOUT2 = MHz fDATA = MSPS; fOUT1 = MHz; fOUT2 = MHz fDATA = MSPS; fOUT1 = MHz; fOUT2 = MHz fDATA = MSPS; fOUT1 = MHz; fOUT2 = MHz Total Harmonic Distortion (THD) fDATA = MSPS; fOUT = MHz; 0 dBFS fDATA = MSPS; fOUT = MHz; 0 dBFS Signal-to-Noise Ratio (SNR) fDATA = MSPS; fOUT = MHz; 0 dBFS fDATA = MSPS; fOUT = MHz; 0 dBFS Adjacent Channel Power Ratio (ACPR) WCDMA with MHz BW, MHz Channel Spacing IF = 16 MHz, fDATA = 65.536 MSPS IF = 32 MHz, fDATA = 131.072 MSPS Four-Tone Intermodulation MHz, MHz, MHz and MHz at –12 dBFS (fDATA = MSPS, Missing Center) AC LINEARITY—IF MODE Four-Tone Intermodulation at IF = MHz MHz, MHz, MHz and MHz at dBFS fDATA = MSPS, fDAC = MHz 5 6 Propagation delay is delay from CLK input to DAC update. Measured single-ended into 50 Ω load. Rev. PrC | Page 5 of 52 Min Typ 500 Max Unit MSPS ns ns ns ns pA√Hz 95 dBc dBc dBc dBc dBc 80 dBc dBc dBc dBc dBc dBc dB dB dBFS dBFS dBc dBc dBFS dBFS AD9784 Preliminary Technical Data DIGITAL SPECIFICATIONS Table 3. TMIN to TMAX, AVDD1, AVDD2 = 3.3 V, ACVDD, ADVDD, CLKVDD, DVDD = 2.5 V, IOUTFS = 20 mA, unless otherwise noted Parameter DIGITAL INPUTS Logic 1 Voltage Logic 0 Voltage Logic 1 Current Logic 0 Current Input Capacitance LOCK INPUTS Input Voltage Range Common-Mode Voltage Differential Voltage PLL CLOCK ENABLED Input Setup Time (ts) Input Hold Time (tH) Latch Pulse Width (tLPW) PLL CLOCK DISABLED Input Setup Time (ts) Input Hold Time (tH) Latch Pulse Width (tLPW) CLK to PLLLOCK Delay (tOD) Min Typ DRVDD – 0.9 DRVDD 0 –10 –10 Max Unit 0.9 +10 +10 V V µA µA pF 5 0 0.75 0.5 1.5 1.5 2.65 2.25 V V V ns ns ns ns ns ns ns Rev. PrC | Page 6 of 52 Preliminary Technical Data AD9784 DNC ADVDD ADGND ACVDD ACGND AVDD2 AVDD1 AGND2 AGND1 IOUTA IOUTB AGND1 AVDD1 AGND2 AVDD2 ACGND ADGND ACVDD ADVDD NC PIN CONFIGURATION AND FUNCTION DESCRIPTIONS 80 79 78 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 62 61 60 FSADJ 59 REFIO CLKVDD 3 58 RESET CLKGND 4 57 CSB CLK+ 5 56 SCLK CLK– 6 55 SDIO CLKGND 7 54 SDO DGND 8 53 DGND DVDD 9 52 DVDD 51 P2B0 50 P2B1 P1B13 12 49 P2B2 P1B12 13 48 P2B3 P1B11 14 47 P2B4 P1B10 15 46 P2B5 DGND 16 45 DGND DVDD 17 44 DVDD P1B9 18 P1B8 19 43 P2B6 42 P2B7 P1B7 20 41 P2B8 CLKVDD 1 LPF 2 PIN 1 IDENTIFIER AD9784 P1B15 10 TOP VIEW (Not to Scale) P1B14 11 03151-PrD-001 P2B10 P2B9 P2B11 P2B12 DGND DVDD IQSEL/P2B15 ONEPORTCLOCK/P2B14 P2B13 DRVDD DATACLK/PLL_LOCK P1B0 P1B2 P1B1 DVDD DGND P1B3 P1B4 P1B5 NC = NO CONNECT P1B6 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Figure 2. Pin Configuration CLOCK Table 4. Clock Pin Function Descriptions Pin No. 5, 6 2 31 Mnemonic CLK+, CLK– LPF DATACLK/PLL_LOCK Direction I I/O I/O Description Differential Clock Input. PLL Loop Filter. PLOCKEXT DCLKEXT 04h[0] 02h[3] 0 0 0 1 1, 3 4, 7 CLKVDD CLKGND 1 X Mode Pin configured for input of channel data rate or synchronizer clock. Internal clock synchronizer may be turned on or off with DCLKCRC (02h[2]). Pin configured for output of channel data rate or synchronizer clock Internal Clock PLL Status Output: 0: Internal clock PLL is not locked. 1: Internal clock PLL is locked. Clock Domain 2.5 V. Clock Domain 0 V. Rev. PrC | Page 7 of 52 AD9784 Preliminary Technical Data ANALOG Table 5. Analog Pin Function Descriptions Pin No. 59 60 70, 71 61 62, 79 63, 78 64, 77 65, 76 66, 75 67, 74 68, 73 69, 72 Mnemonic REFIO FSADJ IOUTB, IOUTA DNC ADVDD ADGND ACVDD ACGND AVDD2 AGND2 AVDD1 AGND1 Direction A A A Description Reference. Full-Scale Adjust. Differential DAC Output Currents. Do not connect. Analog Domain Digital Content 2.5 V. Analog Domain Digital Content 0 V. Analog Domain Clock Content 2.5 V. Analog Domain Clock Content 0 V. Analog Domain Clock Switching 3.3 V. Analog Domain Switching 0 V. Analog Domain Quiet 3.3 V. Analog Domain Quiet 0 V. DATA Table 6. Data Pin Function Descriptions Pin No. 10–15, 18–24, 27–29 Mnemonic P1B15–P1B0 Direction I 32 IQSEL/P2B15 I 33 ONEPORTCLK/P2B14 I/O 34, 37–43, 46–51 30 9, 17, 26, 36, 44, 52 8, 16, 25, 35, 45, 53 P2B13–P2B0 I Description Input Data Port One. ONEPORT 02h[6] Mode 0 Latched Data Routed for 1 Channel Processing. 1 Latched Data Demultiplexed by IQSEL and Routed for Interleaved I/Q Processing. ONEPORT IQPOL IQSEL/ 02h[6] 02h[1] P2B15 Mode (IQPOL == 0) 0 X X Latched data routed to Q channel bit 15(MSB) processing. 1 0 0 Latched data on data port one routed to Q channel processing. 1 0 1 Latched data on data port one routed to I channel processing. 1 1 0 Latched data on data port one routed to I channel processing. 1 1 1 Latched data on data port one routed to Q channel processing. ONEPORT 02h[6] 0 Latched data routed for Q channel Bit 14 processing. 1 Pin configured for output of clock at twice the channel data route. Input Data Port Two Bits 13–0. DRVDD DVDD Digital Output Pin Supply, 2.5 V or 3.3 V. Digital Domain 2.5 V. DGND Digital Domain 0 V. Rev. PrC | Page 8 of 52 Preliminary Technical Data AD9784 SERIAL INTERFACE Table 7. Serial Interface Pin Function Descriptions Pin No. 54 Mnemonic SDO Direction O 55 SDIO I/O 56 57 58 SCLK CSB RESET I I I Description SDIODIR CSB 00h[7] Mode 1 X High Impedance. 0 0 Serial Data Output. 0 1 High Impedance. SDIODIR CSB 00h[7] Mode 1 X High Impedance. 0 0 Serial Data Output. 0 1 Serial Data Input/Output Depending on Bit 7 of the Serial Instruction Byte. Serial interface clock. Serial interface chip select. Resets entire chip to default state. Rev. PrC | Page 9 of 52 AD9784 Preliminary Technical Data DEFINITIONS OF SPECIFICATIONS Linearity Error (Integral Nonlinearity or INL) Glitch Impulse Linearity error is defined as the maximum deviation of the actual analog output from the ideal output, determined by a straight line drawn from zero to full scale. Asymmetrical switching times in a DAC give rise to undesired output transients that are quantified by a glitch impulse. It is specified as the net area of the glitch in pV-s. Differential Nonlinearity (or DNL) Spurious-Free Dynamic Range DNL is the measure of the variation in analog value, normalized to full scale, associated with a 1 LSB change in digital input code. The difference, in dB, between the rms amplitude of the output signal and the peak spurious signal over the specified bandwidth. Monotonicity Total Harmonic Distortion A D/A converter is monotonic if the output either increases or remains constant as the digital input increases. THD is the ratio of the rms sum of the first six harmonic components to the rms value of the measured fundamental. It is expressed as a percentage or in decibels (dB). Offset Error The deviation of the output current from the ideal of zero is called offset error. For IOUTA, 0 mA output is expected when the inputs are all 0s. For IOUTB, 0 mA output is expected when all inputs are set to 1s. Signal-to-Noise Ratio (SNR) S/N is the ratio of the rms value of the measured output signal to the rms sum of all other spectral components below the Nyquist frequency, excluding the first six harmonics and dc. The value for SNR is expressed in decibels. Gain Error The difference between the actual and ideal output span. The actual span is determined by the output when all inputs are set to 1s, minus the output when all inputs are set to 0s. Output Compliance Range The range of allowable voltage at the output of a current-output DAC. Operation beyond the maximum compliance limits may cause either output stage saturation or breakdown, resulting in nonlinear performance. Interpolation Filter If the digital inputs to the DAC are sampled at a multiple rate of fDATA (interpolation rate), a digital filter can be constructed which has a sharp transition band near fDATA/2. Images which would typically appear around fDAC (output data rate) can be greatly suppressed. Pass-Band Frequency band in which any input applied therein passes unattenuated to the DAC output. Temperature Drift Temperature drift is specified as the maximum change from the ambient (+25°C) value to the value at either TMIN or TMAX. For offset and gain drift, the drift is reported in ppm of full-scale range (FSR) per degree C. For reference drift, the drift is reported in ppm per degree C. Stop-Band Rejection The amount of attenuation of a frequency outside the passband applied to the DAC, relative to a full-scale signal applied at the DAC input within the pass-band. Group Delay Power Supply Rejection The maximum change in the full-scale output as the supplies are varied from minimum to maximum specified voltages. Number of input clocks between an impulse applied at the device input and peak DAC output current. A half-band FIR filter has constant group delay over its entire frequency range Settling Time Impulse Response The time required for the output to reach and remain within a specified error band about its final value, measured from the start of the output transition. Response of the device to an impulse applied to the input. Rev. PrC | Page 10 of 52 Preliminary Technical Data AD9784 Adjacent Channel Power Ratio (or ACPR) Complex Image Rejection A ratio in dBc between the measured power within a channel relative to its adjacent channel. In a traditional two part upconversion, two images are created around the second IF frequency. These images are redundant and have the effect of wasting transmitter power and system bandwidth. By placing the real part of a second complex modulator in series with the first complex modulator, either the upper or lower frequency image near the second IF can be rejected. Complex Modulation The process of passing the real and imaginary components of a signal through a complex modulator (transfer function = ejwt = coswt + jsinwt) and realizing real and imaginary components on the modulator output. Rev. PrC | Page 11 of 52 AD9784 Preliminary Technical Data TYPICAL PERFORMANCE CHARATCERISTICS –000 –000 –000 –000 ALL CAPS (Initial caps) ALL CAPS (Initial caps) (TMIN to TMAX, AVDD1, AVDD2 = 3.3 V, ACVDD, ADVDD, CLKVDD, DVDD, DRVDD = 2.5 V, IOUTFS = 20 mA, Differential Transformer Coupled Output, 50 Ω Doubly Terminated, unless otherwise noted) TBD –000 –000 –000 –000 –000 –000 –000 –000 ALL CAPS (Initial caps) –000 –000 –000 –000 TBD –000 –000 –000 –000 ALL CAPS (Initial caps) –000 Figure 6. Single-Tone Spectrum @ FDATA = 78 MSPS with FOUT = FDATA/3 ALL CAPS (Initial caps) ALL CAPS (Initial caps) –000 –000 –000 Figure 3 Single-Tone Spectrum@ FDATA = 65 MSPS With FOUT = FDATA/3 –000 TBD –000 –000 –000 –000 –000 –000 –000 ALL CAPS (Initial caps) –000 –000 –000 Figure 4. In-Band SFDR vs. FOUT @ FDATA = 65 MSPS –000 –000 –000 –000 TBD –000 –000 –000 –000 –000 –000 ALL CAPS (Initial caps) –000 Figure 7. In-Band SFDR Vs. FOUT @ FDATA = 78 MSPS –000 ALL CAPS (Initial caps) ALL CAPS (Initial caps) TBD –000 –000 TBD –000 –000 –000 –000 ALL CAPS (Initial caps) –000 Figure 5. Out-of-Band SFDR vs. FOUT @ FDATA = 65 MSPS –000 –000 –000 –000 –000 ALL CAPS (Initial caps) Figure 8. Out-of-Band SFDR vs. FOUT @ FDATA = 78 MSPS Rev. PrC | Page 12 of 52 –000 AD9784 –000 –000 –000 –000 ALL CAPS (Initial caps) ALL CAPS (Initial caps) Preliminary Technical Data TBD –000 –000 –000 –000 –000 ALL CAPS (Initial caps) –000 –000 –000 Figure 9. Single-Tone Spectrum @ FDATA = 160 MSPS with FOUT = FDATA/3 –000 –000 –000 TBD –000 –000 –000 –000 ALL CAPS (Initial caps) –000 Figure 12. Third Order IMD Products vs. FOUT @ FDATA = 65 MSPS –000 ALL CAPS (Initial caps) ALL CAPS (Initial caps) TBD –000 –000 –000 –000 –000 TBD –000 –000 –000 –000 –000 –000 ALL CAPS (Initial caps) –000 –000 –000 Figure 10. In-Band SFDR vs. FOUT @ FDATA = 160 MSPS –000 –000 –000 –000 TBD –000 –000 –000 –000 –000 –000 ALL CAPS (Initial caps) –000 Figure 13. Third Order IMD Products vs. FOUT @ FDATA = 78 MSPS –000 ALL CAPS (Initial caps) ALL CAPS (Initial caps) –000 –000 TBD –000 –000 –000 –000 ALL CAPS (Initial caps) –000 Figure 11. Out-of-Band SFDR vs. FOUT @ FDATA = 160 MSPS –000 –000 –000 –000 –000 ALL CAPS (Initial caps) –000 Figure 14. Third Order IMD Products vs. FOUT @ FDATA = 160 MSPS Rev. PrC | Page 13 of 52 Preliminary Technical Data –000 –000 –000 –000 –000 ALL CAPS (Initial caps) ALL CAPS (Initial caps) AD9784 TBD –000 –000 –000 –000 TBD –000 –000 –000 –000 ALL CAPS (Initial caps) –000 –000 –000 Figure 15. TPC 13. Third Order IMD Products vs. FOUT and Interpolation Rate 1× – FDATA = 160 MSPS –000 –000 –000 ALL CAPS (Initial caps) –000 Figure 18. 3rd Order IMD Products vs. AVDD @ FOUT = 10 MHz, FDAC = 320 MSPS, FDATA = 160 MSPS 2× – FDATA = 160 MSPS 4× – FDATA = 80 MSPS 8× – FDATA = 50 MSPS –000 ALL CAPS (Initial caps) ALL CAPS (Initial caps) –000 –000 –000 TBD –000 TBD –000 –000 –000 –000 –000 –000 –000 –000 –000 –000 ALL CAPS (Initial caps) –000 –000 –000 ALL CAPS (Initial caps) –000 –000 Figure 19. SNR vs. Data Rate for fOUT = 5 MHz Figure 16. Third Order IMD Products vs. AOUT and Interpolation Rate FDATA = 50 MSPS for All Cases 1× – FDAC = 50 MSPS 2× – FDAC = 100 MSPS 4× – FDAC = 200 MSPS –000 –000 –000 –000 –000 ALL CAPS (Initial caps) ALL CAPS (Initial caps) 8× – FDAC = 400 MSPS TBD –000 –000 –000 –000 TBD –000 –000 –000 –000 ALL CAPS (Initial caps) –000 –000 –000 Figure 17. SFDR vs. AVDD @ FOUT = 10 MHz; FDAC = 320 MSPS FDATA = 160 MSPS Rev. PrC | Page 14 of 52 –000 –000 –000 ALL CAPS (Initial caps) Figure 20. SFDR vs. Temperature @ fOUT = fDATA/11 –000 AD9784 –000 –000 –000 –000 –000 ALL CAPS (Initial caps) ALL CAPS (Initial caps) Preliminary Technical Data TBD –000 –000 –000 –000 –000 –000 ALL CAPS (Initial caps) –000 –000 –000 TBD –000 –000 –000 ALL CAPS (Initial caps) –000 –000 –000 –000 –000 –000 –000 ALL CAPS (Initial caps) –000 TBD TBD –000 –000 –000 ALL CAPS (Initial caps) –000 –000 TBD –000 –000 –000 –000 –000 Figure 25. Single Tone Spurious Performance, FOUT = 10 MHz, FDATA = 80 MSPS, Interpolation = 4× Figure 22. Two Tone IMD Performance, FDATA = 150 MSPS, No Interpolation –000 –000 –000 ALL CAPS (Initial caps) –000 –000 –000 –000 –000 Figure 24. Two Tone IMD Performance, FDATA = 90 MSPS, Interpolation = 4× ALL CAPS (Initial caps) ALL CAPS (Initial caps) –000 –000 –000 –000 –000 TBD –000 Figure 21. Single Tone Spurious Performance, fOUT = 10 MHz, fDATA = 150 MSPS, No Interpolation ALL CAPS (Initial caps) –000 –000 –000 –000 ALL CAPS (Initial caps) –000 –000 –000 Figure 23. Single Tone Spurious Performance, FOUT = 10 MHz, FDATA = 150 MSPS, Interpolation = 2× –000 –000 –000 ALL CAPS (Initial caps) –000 Figure 26. Two Tone IMD Performance, FOUT = 10 MHz, FDATA = 50 MSPS, Interpolation = 8× Rev. PrC | Page 15 of 52 Preliminary Technical Data –000 –000 –000 –000 –000 ALL CAPS (Initial caps) ALL CAPS (Initial caps) AD9784 TBD –000 –000 –000 –000 TBD –000 –000 –000 –000 ALL CAPS (Initial caps) –000 –000 –000 Figure 27. Single Tone Spurious Performance, FOUT = 10 MHz, FDATA = 50 MSPS, Interpolation = 8× –000 –000 –000 ALL CAPS (Initial caps) –000 Figure 28. Eight Tone IMD Performance, FDATA = 160 MSPS, Interpolation = 8× Rev. PrC | Page 16 of 52 Preliminary Technical Data AD9784 SERIAL CONTROL INTERFACE INSTRUCTION BYTE SDO (PIN 54) SCLK (PIN 56) AD9784 SPI PORT INTERFACE CSB (PIN 57) The instruction byte contains the following information: 03151-PrD-002 SDIO (PIN 55) Table 8. Figure 29. AD9784 SPI Port Interface The AD9784 serial port is a flexible, synchronous serial communications port allowing easy interface to many industrystandard microcontrollers and microprocessors. The serial I/O is compatible with most synchronous transfer formats, including both the Motorola SPI® and Intel® SSR protocols. The interface allows read/write access to all registers that configure the AD9784. Single or multiple byte transfers are supported as well as MSB first or LSB first transfer formats. The AD9784’s serial interface port can be configured as a single pin I/O (SDIO) or two unidirectional pins for in/out (SDIO/SDO). GENERAL OPERATION OF THE SERIAL INTERFACE There are two phases to a communication cycle with the AD9784. Phase 1 is the instruction cycle, which is the writing of an instruction byte into the AD9784, coincident with the first eight SCLK rising edges. The instruction byte provides the AD9784 serial port controller with information regarding the data transfer cycle, which is Phase 2 of the communication cycle. The Phase 1 instruction byte defines whether the upcoming data transfer is read or write, the number of bytes in the data transfer, and the starting register address for the first byte of the data transfer. The first eight SCLK rising edges of each communication cycle are used to write the instruction byte into the AD9784. A logic high on the CS pin, followed by a logic low, will reset the SPI port timing to the initial state of the instruction cycle. This is true regardless of the present state of the internal registers or the other signal levels present at the inputs to the SPI port. If the SPI port is in the midst of an instruction cycle or a data transfer cycle, none of the present data will be written. The remaining SCLK edges are for Phase 2 of the communication cycle. Phase 2 is the actual data transfer between the AD9784 and the system controller. Phase 2 of the communication cycle is a transfer of 1, 2, 3, or 4 data bytes as determined by the instruction byte. Normally, using one multibyte transfer is the preferred method. However, single byte data transfers are useful to reduce CPU overhead when register access requires one byte only. Registers change immediately upon writing to the last bit of each transfer byte. N1 0 0 1 1 N2 0 1 0 1 Description Transfer 1 Byte Transfer 2 Bytes Transfer 3 Bytes Transfer 4 Bytes R/W, Bit 7 of the instruction byte, determines whether a read or a write data transfer will occur after the instruction byte write. Logic high indicates read operation. Logic 0 indicates a write operation. N1, N0, Bits 6 and 5 of the instruction byte, determine the number of bytes to be transferred during the data transfer cycle. The bit decodes are shown in the following table: Table 9. MSB 17 R/W 16 N1 15 N0 14 A4 13 A3 12 A2 11 A1 LSB 10 A0 A4, A3, A2, A1, A0, Bits 4, 3, 2, 1, 0 of the instruction byte, determine which register is accessed during the data transfer portion of the communications cycle. For multibyte transfers, this address is the starting byte address. The remaining register addresses are generated by the AD9784. SERIAL INTERFACE PORT PIN DESCRIPTIONS SCLK—Serial Clock. The serial clock pin is used to synchronize data to and from the AD9784 and to run the internal state machines. SCLK’s maximum frequency is 15 MHz. All data input to the AD9784 is registered on the rising edge of SCLK. All data is driven out of the AD9784 on the falling edge of SCLK. CSB—Chip Select. Active low input starts and gates a communication cycle. It allows more than one device to be used on the same serial communications lines. The SDO and SDIO pins will go to a high impedance state when this input is high. Chip select should stay low during the entire communication cycle. SDIO—Serial Data I/O. Data is always written into the AD9784 on this pin. However, this pin can be used as a bidirectional data line. The configuration of this pin is controlled by Bit 7 of register address 00h. The default is Logic 0, which configures the SDIO pin as unidirectional. SDO—Serial Data Out. Data is read from this pin for protocols that use separate lines for transmitting and receiving data. In the case where the AD9784 operates in a single bidirectional I/O mode, this pin does not output data and is set to a high impedance state. Rev. PrC | Page 17 of 52 AD9784 Preliminary Technical Data MSB/LSB TRANSFERS INSTRUCTION CYCLE R/W N1 N0 A4A 3A 2A 1A SDO D30 D20 D10 D00 D7 D6N D5N D30 D20 D10 D00 INSTRUCTION CYCLE DATA TRANSFER CYCLE CSB SDIO A0 A1 A2 A3 A4 N0 N1 R/W D00 D10 D20 D4N D5N D6N D7N D00 D10 D20 D4N D5N D6N D7N SDO 03152-PrD-005 SCLK Figure 31. Serial Register Interface Timing LSB First tDS tSCLK CSB tPWH The same considerations apply to setting the software reset, SWRST (00h[5]) bit. All other registers are set to their default values but the software reset doesn’t affect the bits in register address 00h and 04h. D7 D6N D5N Figure 30. Serial Register Interface Timing MSB First NOTES ON SERIAL PORT OPERATION The AD9784 serial port configuration bits reside in Bits 6 and 7 of register address 00h. It is important to note that the configuration changes immediately upon writing to the last bit of the register. For multibyte transfers, writing to this register may occur during the middle of communication cycle. Care must be taken to compensate for this new configuration for the remaining bytes of the current communication cycle. 0 03152-PrD-004 SDIO tPWL SCLK tDS SDIO tDH INSTRUCTION BIT 7 INSTRUCTION BIT 6 03152-PrD-006 The AD9784 serial port controller address will increment from 1Fh to 00h for multibyte I/O operations if the MSB first mode is active. The serial port controller address will decrement from 00h to 1Fh for multibyte I/O operations if the LSB first mode is active. SCLK Figure 32. Timing Diagram for Register Write CSB SCLK tDV SDIO SDO It is recommended to use only single byte transfers when changing serial port configurations or initiating a software reset. DATA BIT n DATA BIT n–1 Figure 33. Timing Diagram for Register Read Rev. PrC | Page 18 of 52 03152-PrD-007 The AD9784 serial port can support both most significant bit (MSB) first or least significant bit (LSB) first data formats. This functionality is controlled by register address DATADIR (00h[6]). The default is MSB first. When this bit is set active high, the AD9784 serial port is in LSB first format. That is, if the AD9784 is in LSB first mode, the instruction byte must be written from least significant bit to most significant bit. Multibyte data transfers in MSB format can be completed by writing an instruction byte that includes the register address of the most significant byte. In MSB first mode, the serial port internal byte address generator decrements for each byte required of the multibyte communication cycle. Multibyte data transfers in LSB first format can be completed by writing an instruction byte that includes the register address of the least significant byte. In LSB first mode, the serial port internal byte address generator increments for each byte required of the multibyte communication cycle. DATA TRANSFER CYCLE CSB Preliminary Technical Data AD9784 MODE CONTROL (VIA SPI PORT) Table 10. Address COMMS FILTER DATA MODULATE PLL DCLKCRC VERSION CALMEMCK MEMRDWR MEMADDR MEMDATA DCRSTAT 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 Bit 7 SDIODIR INTERP[1] DATAFMT CHANNEL PLLON DATADJ[3] Bit 6 DATADIR INTERP[0] ONEPORT HILBERT PLLMULT[1] DATADJ[2] RESERVED CALSTAT MEMADDR[7] RESERVED CALEN MEMADDR[6] Bit 5 SWRST Bit 4 SLEEP DCLKSTR MODDUAL PLLMULT[0] DATADJ[1] CALMEM[1] XFERSTAT MEMADDR[5] MEMDATA[5] Bit 3 PDN ZSTUFF DCLKPOL DCLKEXT SIDEBAND MOD[1] PLLDIV[1] PLLDIV[0] DATADJ[0] MODSYNC Reserved Reserved Reserved Reserved Reserved Reserved Reserved VERSION[3] CALMEN[0] XFEREN SMEMWR MEMADDR[4] MEMADDR[3] MEMDATA[4] MEMDATA[3] Bit 2 HPFX8 DCLKCRC MOD[0] PLLAZ[1] MODADJ[2] VERSION[3] CALCKDIV[2] SMEMRD MEMADDR[2] MEMDATA[2] DCRSTAT[2] Bit 1 PLLLOCK HPFX4 IQPOL Bit 0 EXREF HPFX2 CRAYDIN PLLAZ[0] MODADJ[1] PLOCKEXT MODADJ[0] VERSION[3] CALCKDIV[2] FMEMRD MEMADDR[1] MEMDATA[1] DCRSTAT[1] VERSION[3] CALCKDIV[2] UNCAL MEMADDR[0] MEMDATA[0] DCRSTAT[0] Table 11. COMMS(00) SDIODIR Bit 7 Direction I Default 0 DATADIR 6 I 0 SWRST SLEEP PDN PLLOCK 5 4 3 1 I I I O 0 0 0 0 EXREF 0 I 0 Description 0: SDIO pin configured for input only during data transfer 1: SDIO configured for input or output during data transfer 0: Serial data uses MSB first format 1: Serial data uses LSB first format 1: Default all serial register bits, except addresses 00h and 04h 1: DAC output current off 1: All analog and digital circuitry, except serial interface, off 0: With PLL on, indicates that PLL is not locked 1: With PLL on, indicates that PLL is locked 0: Internal band gap reference 1: External reference Table 12. FILTER(01) INTERP[1:0] Bit [7:6] Direction I Default 00 ZSTUFF HPFX8 3 2 I I 0 0 HPFX4 1 I 0 HPFX2 0 I 0 Description 00: No interpolation 01: Interpolation 2× 10: Interpolation 4× 11: Interpolation 8× 1: Zero Stuffing on 0: ×8 interpolation filter configured for low pass 1: ×8 interpolation filter configured for high pass 0: ×4 interpolation filter configured for low pass 1: ×4 interpolation filter configured for high pass 0: ×2 interpolation filter configured for low pass 1: ×2 interpolation filter configured for high pass Rev. PrC | Page 19 of 52 AD9784 Preliminary Technical Data Table 13. DATA(02) DATAFMT Bit 7 Direction I Default 0 ONEPORT 6 I 0 DCLKSTR 5 I 0 DCLKPOL 4 I 0 DCLKEXT 3 I 0 DCLKCRC 2 I 0 IQPOL 1 I 0 GRAYDIN 0 I 0 Description 0: Twos complement data format 1: Unsigned binary input data format 0: I and Q input data onto ports one and two respectively 1: I and Q input data interleaved onto port one 0: DATACLK pin 12 mA drive strength 1: DATACLK pin 24 mA drive strength 0: Input data latched on DATACLK rising edge 1: Input data latched on DATACLK falling edge 0: With PLOCKEXT off, DATACLK pin inputs channel data rate or modulator synchronizer clock 1: With PLOCKEXT off, DATACLK pin outputs channel data rate or modulator synchronizer clock 0: With PLOCKEXT off, and DATACLK pin as input, DATACLK clock recovery off 1: With PLOCKEXT off, and DATACLK pin as input, DATACLK clock recovery on 0: In one port mode, IQSEL = 1 latches data into I channel, IQSEL = 0 latches data into Q channel 1: In one port mode, IQSEL = 0 latches data into I channel, IQSEL = 1 latches data into Q channel 0: Gray decoder off 1: Gray decoder on Table 14. MODULATE(03) CHANNEL Bit 7 Direction I Default 0 HILBERT MODDUAL 6 5 I I 0 0 SIDEBAND 4 I 0 MOD[1:0] [3:2] I 00 Description MODDUAL CHANNEL 03h [5] 03h[7] 0 0 I channel processing routed to DAC 0 1 Q channel processing routed to DAC 1 0 Modulator real output routed to DAC 1 1 Modulator imaginary output routed to DAC 1: With MODDUAL on, Hilbert transform on 0: Modulator uses a single channel 1: Modulator uses both I and Q channels 0: With MODDUAL on, lower sideband rejected 1: With MODDUAL on, upper sideband rejected 00: No modulation 01: fS/2 modulation 10: fS /4 modulation 11: fS /8 modulation Rev. PrC | Page 20 of 52 Preliminary Technical Data AD9784 Table 15. PLL(04) PLLON Bit 7 Direction I Default 0 PLLMULTI[1:0] [6:5] I 00 PLLDIV[1:0] [4:3] I 00 PLLAZBW[1:0] [2:1] I 00 PLOCKEXT 0 I 0 Description 0: PLL off 1: PLL on PLL MULTIPLY FACTOR 00: ×2 00: ×4 00: ×8 00: ×16 PLLMULT rate divide factor 00:/1 00:/2 00:/4 00:/8 PLL Autozero settling bandwidth as fraction of CLK ±rate 00: /8 (lowest) 01: /4 10: /2 (highest) 0: With PLL on, DATACLK/PLL_LOCK pin configured for DATACLK input/output 1: With PLL on, DATACLK/PLL_LOCK pin configured for output of PLLLOCK Table 16. DCLKCRC(05) DATADJ[3:0] Bit [7:4] Direction I Default 0000 MODSYNC 3 I 00 MODADJ[2:0] [2:0] I 000 Description DATACLK offset. Twos complement respresentation 0111: +7 : 0000: 0 : 1000: -8 0: With PLOCKEXT off, channel data rate clock synchronizer mode 1: With PLOCKEXT off, state machine clock synchronizer mode Modulator coefficient offset fS/8 fS/4 fS/2 000 1 1 1 001 1/√2 0 –1 010 0 –1 1 011 –1/√2 0 –1 100 –1 1 1 101 –1/√2 0 –1 110 0 –1 1 111 1/√2 0 –1 Table 17. VERSION(0D) VERSION[3:0] Bit [3:0] Direction O Default – Rev. PrC | Page 21 of 52 Description Hardware version identifier AD9784 Preliminary Technical Data Table 18. CALMEMCK(OE) CALMEM Bit [5:4] Direction O Default 00 CALCKDIV[2:0] [2:0] I 00 Description Calibration memory 00: Uncalibrated 01: Self Calibration 10: Factory calibration 11: User input Calibration clock divide ratio from channel data rate 000: /32 001: /64 : 110: /2048 111: /4096 Table 19. MEMRDWR(OF) CALSTAT Bit 7 Direction O Default 0 CALEN XFERSTAT 6 5 I O 0 0 XFEREN SMEMWR SMEMRD FMEMRD UNCAL 4 3 2 1 0 I I I I I 0 0 0 0 0 Description 0: Self Calibration cycle not complete 1: Self Calibration cycle complete 1: Self Calibration in progress 0: Factory memory transfer not complete 1: Factory memory transfer complete 1: Factory memory transfer in progress 1: Write static memory data from external port 1: Read static memory to external port 1: Read factory memory data to external port 1: Use uncalibrated Table 20. MEMADDR(10) MEMADDR [7:0] Bit [7:0] Direction I/O Default 00000000 Description Address of factory or static memory to be accessed Table 21. MEMDATA(11) MEMDATA [5:0] Bit [5:0] Direction I/O Default 000000 Description Data or factory or static memory access Table 22. DCRCSTAT(12) DCRCSTAT (2) Bit 2 Direction O Default 0 DCRCSTAT(1) 1 O 0 DCRCSTAT(0) 0 O 0 Description 0: With DATACLK CRC on, lock has never been achieved 1: With DATACLK CRC on, lock has been achieved at least once 0: With DATACLK CRC on, system is currently not locked 1: With DATACLK CRC on, system is currently locked 0: With DATACLK CRC on, system is currently locked 1: With DATACLK CRC on, system lost lock due to jitter Rev. PrC | Page 22 of 52 Preliminary Technical Data AD9784 DIGITAL FILTER SPECIFICATIONS DIGITAL INTERPOLATION FILTER COEFFICIENTS –20 Upper Coefficient H(43) H(42) H(41) H(40) H(39) H(38) H(37) H(36) H(35) H(34) H(33) H(32) H(31) H(30) H(29) H(28) H(27) H(26) H(25) H(24) H(23) Integer Value 9 0 –27 0 65 0 –131 0 239 0 –407 0 665 0 –1070 0 1764 0 –3273 0 10358 16384 –40 –60 –80 –100 –120 –140 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.1 0.2 0.3 0.4 0.5 Figure 34. ×2 Interpolation Filter Response 0 –20 –40 –60 –80 –100 0.5 03152-PrD-009 0.5 03152-PrD-010 –120 Table 24. Stage 2 Interpolation Filter Coefficients Lower Coefficient H(1) H(2) H(3) H(4) H(5) H(6) H(7) H(8) H(9) H(10) Upper Coefficient H(19) H(18) H(17) H(16) H(15) H(14) H(13) H(12) H(11) Integer Value 19 0 –120 0 436 0 –1284 0 5045 8192 03152-PrD-008 Table 23. Stage 1 Interpolation Filter Coefficients Lower Coefficient H(1) H(2) H(3) H(4) H(5) H(6) H(7) H(8) H(9) H(10) H(11) H(12) H(13) H(14) H(15) H(16) H(17) H(18) H(19) H(20) H(21) H(22) 0 –140 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.1 0.2 0.3 0.4 Figure 35. ×4 Interpolation Filter Response 0 –20 –40 –60 –80 Table 25. Stage 3 Interpolation Filter Coefficients Lower Coefficient H(1) H(2) H(3) H(4) H(5) H(6) Upper Coefficient H(11) H(10) H(9) H(8) H(7) Integer Value 7 0 –53 0 302 512 –100 –120 –140 –0.5 Rev. PrC | Page 23 of 52 –0.4 –0.3 –0.2 –0.1 0 0.1 0.2 0.3 0.4 Figure 36. ×8 Interpolation Filter Response AD9784 Preliminary Technical Data AD9784 CLOCK/DATA TIMING DLL Disabled, Two-Port Data Mode, DATACLK as Output With the interpolation set to 1×, the DATACLK output is a delayed and inverted version of DACCLK at the same frequency. Note that DACCLK refers to the differential clock inputs applied at Pins 5 and 6. As Figure 37 shows, there is a constant delay between the rising edge of DACCLK and the falling edge of DATACLK. The DCLKPOL bit (Reg 02 Bit 4) allows the data to be latched into the AD9784 on either the rising or falling edge of DACCLK. With DCLKPOL = 1, the data is latched in on the rising edge of Diff Clk, as shown in Figure 37. With DCLKPOL = 0, as shown in Figure 38, data is latched in on the falling edge of DACCLK. The setup and hold times are always with respect to the latched edge of DACCLK. DACCLKIN DATACLKOUT tD = 5ns TYP t12 tH = 2.9ns TYP DATA 03152-PrD-066 tS = –0.5ns TYP Figure 37. Data Timing, DLL Off, 1× Interpolation, DCLKPOL = 1 DACCLKIN DATACLKOUT tD = 6ns TYP tH = 2.9ns TYP DATA Figure 38. Data Timing, DLL Off, 1× Interpolation, DCLKPOL = 0 Rev. PrC | Page 24 of 52 03152-PrD-067 tS = –0.5ns TYP Preliminary Technical Data AD9784 With the interpolation set to 2×, the DACCLK input runs at twice the speed of the DATACLK. Data is latched into the AD9784’s inputs on every other rising edge of DACCLK, as shown in Figure 40 and Figure 41. With DCLKPOL = 1, as shown in Figure 40, the latching edge of DACCLK is the rising edge that occurs just before the falling edge of DATACLK. With DCLKPOL = 0, as in Figure 41, the latching edge of DACCLK is the rising edge of DACCLK that occurs just before the rising edge of DATACLK. The setup and hold time values are identical to those in Figure 37 and Figure 38. interpolation mode. Again, similar to operation in the 2× interpolation mode, with DCLKPOL = 1, the latching edge of DACCLK is the rising edge that occurs just before the falling edge of DATACLK. With DCLKPOL = 0, the latching edge of DACCLK is the rising edge that occurs just before the rising edge of DATACLK. The setup and hold time values are identical to those in 1× and 2× interpolation 03152-PrD-068 Note that there is a slight difference in the delay from the rising edge of DACCLK to the falling edge of DATACLK, and the delay from the rising edge of DACCLK to the rising edge of DATACLK. As Figure 39 shows, the DATACLK duty cycle is slightly less than 50%. This is true in all modes. With the interpolation set to 4× or 8×, the DACCLK input runs at 4× or 8× the speed of the DATACLK output. The data is latched in on a rising edge of DACCLK, similar to the 2× interpolation mode. However, the latching edge is every fourth edge in 4× interpolation mode and every eighth edge in the 8× Figure 39 DACCLKIN DATACLKOUT tD = 5ns TYP tH = 2.9ns TYP DATA 03152-PrD-069 tS = –0.5ns TYP Figure 40. Data Timing, DLL Off, 2× Interpolation, DCLKPOL = 1 DACCLKIN DATACLKOUT tD = 6ns TYP tH = 2.9ns TYP DATA Figure 41. Data Timing, DLL Off, 2× Interpolation, DCLKPOL = 0 Rev. PrC | Page 25 of 52 03152-PrD-070 tS = –0.5ns TYP AD9784 Preliminary Technical Data DATAADJUST Synchronization When designing the digital interface for high speed DACs, care must be taken to ensure that the DAC input data meets setupand-hold requirements. Often, compensation must be used in the clock delay path to the digital engine driving the DAC. The AD9784 has the on chip capability to vary the DACCLK’s latching edge. With the interpolation function enabled, this allows the user the choice of multiple edges upon which to latch the data. For instance, if the AD9784 is using 8× interpolation, the user may latch from one of eight edges before the rising edge of DATACLK, or seven edges after this rising edge. The specific edge upon which data is latched is controlled by SPI Register 05h, Bits 7:4. Table 26 shows the relationship of the latching edge of DACCLK and DATACLK with the various settings of the DATAADJ bits. Figure 42, Figure 43, and Figure 44 show the alignment for the latching edge of DACCLK with 4× interpolation and different settings for DATAADJ. In Figure 42, DATAADJ is set to 0000, with DCLKPOL set to 0 so that the latching edge of DACCLK is immediately before the rising edge of DATACLK. The data transitions shown in Figure 42 are synchronous with the DACCLK, so that DACCLK and data are constant with respect to each other. The only visible change when DATAADJ is altered is that DATACLK moves, indicating the latching edge has moved as well. Note that when DATAADJ is altered, the latching edge with respect to DATACLK remains the same, but the latching edge of DACCLK follows the edge of DATACLK. RISING EDGE OF DATACLK CONCURRENT WITH LATCHING EDGE OF DACCLK Table 26. Bit 4 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 Latching Edge wrt DATACLK 0 +1 +2 +3 +4 +5 +6 +7 –8 –7 –6 –5 –4 –3 –2 –1 03152-PrD-071 DACCLK LATCHING EDGE DATA TRANSITION Figure 42. DATAADJ = 0000 Figure 43 shows the same conditions, but now DATAADJ is set to 1111. This moves DATACLK to the left in the plot, indicating that it occurs one DACCLK cycle before it did in Figure 42. As explained previously, the latching edge of DACCLK also moves one cycle back in time. Note that the data in Figure 40 and Figure 41 was taken with the DATAADJ default of 0000. With DCLKPOL = 0, the latching edge of DACCLK is just previous to the rising edge of DATACLK; with DCLKPOL = 1, the latching edge of DACCLK is just previous to the falling edge of DATACLK. With 8× interpolation, the user has the capability of using one of 16 edges to latch the data. This is due to the fact that there are eight DAC clock edges before and after the DATACLK until the next DATACLK latching edge. With 4× interpolation, there are only four latching edges of DACCLK available before and after each DATACLK edge. Therefore, in 4× interpolation, only the even numbered values for DATAADJ are available, and the options are changed from +3 cycles to –4 cycles. With 2× interpolation, there are only two edges available before and after DATACLK, so the choices for DATAADJ are diminished to +1 cycle to –2 cycles. Rev. PrC | Page 26 of 52 RISING EDGE OF DATACLK CONCURRENT WITH LATCHING EDGE OF DACCLK DACCLK LATCHING EDGE DATA TRANSITION Figure 43. DATAADJ = 1111 03152-PrD-072 Bit 7 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 SPI Reg 05h Bit 6 Bit 5 0 0 0 0 0 1 0 1 1 0 1 0 1 1 1 1 0 0 0 0 0 1 0 1 1 0 1 0 1 1 1 1 Preliminary Technical Data AD9784 Figure 44 shows the same conditions, with DATAADJ now set to 0001, thus moving DATACLK to the right in the plot. This indicates that it occurs one DACCLK cycle after it did in Figure 42. Now the latching edge of DACCLK moves forward in time one cycle. RISING EDGE OF DATACLK CONCURRENT WITH LATCHING EDGE OF DACCLK 03152-PrD-073 DACCLK LATCHING EDGE DATA TRANSITION Figure 44. DATAADJ = 0001 INTERPOLATION MODES Table 27. INTERP[1] 0 0 1 1 INTERP[0] 0 1 0 1 Mode No Interpolation ×2 Interpolation ×4 Interpolation ×8 Interpolation The digital filter’s frequency domain response exhibits symmetry about half the output data rate and dc. It will cause images of the input data to be shaped by the interpolation filter’s frequency response. This has the advantage of causing input data images, which fall in the stop band of the digital filter to be rejected by the stop-band attenuation of the interpolation filter; input data images falling in the interpolation filter’s passband will be passed. In band-limited applications, the images at the output of the DAC must be limited by an analog reconstruction filter. The complexity of the analog reconstruction filter is determined by the proximity of the closest image to the required signal band. Higher interpolation rates yield larger stop-band regions, suppressing more input images and resulting in a much relaxed analog reconstruction filter. A DAC shapes its output with a sinc function, having a null at the sampling frequency of the DAC. The higher the DAC sampling rate compared to the input signal bandwidth, the less the DAC sinc function will shape the output. Figure 45 shows the interpolation filters of the AD9784 under different interpolation rates, normalized to the input data rate, fSIN. The higher the interpolation rate the more input data images fall in the interpolation filter stop band and are rejected; the band-width between passed images is larger with higher interpolation factors. The sinc function shaping is also reduced with a higher interpolation factor. Table 28. Interpolation is the process of increasing the number of points in a time domain waveform by approximating points between the input data points; on a uniform time grid, this produces a higher output data rate. Applied to an interpolation DAC, a digital interpolation filter is used to approximate the interpolated points, having an output data rate increased by the interpolation factor. Interpolation filter responses are achieved by cascading individual digital filter banks, whose filter coefficients are given in Table 1; filter responses are shown in Figure 34. Mode No Interpolation ×2 Interpolation ×4 Interpolation ×8 Interpolation Rev. PrC | Page 27 of 52 Sinc Shaping at 0.43fSIN (dB) –2.8241 –0.6708 –0.1657 –0.0413 Bandwidth to First Image fSIN 2fSIN 4fSIN 8fSIN AD9784 Preliminary Technical Data SINC RESPONSE NO INTERPOLATION 0 –50 INTERP[1] = 0 INTERP[0] = 0 –100 –150 –8 –6 –4 –2 –0 2 4 6 8 fSIN ×2 INTERPOLATION 0 –50 INTERP[1] = 0 INTERP[0] = 1 –100 –150 –8 –6 –4 –2 0 2 4 6 8 fSIN ×4 INTERPOLATION 0 –50 INTERP[1] = 1 INTERP[0] = 0 –100 –150 –8 –6 –4 –2 0 2 4 6 8 fSIN ×8 INTERPOLATION 0 –50 –100 –150 –8 –6 –4 –2 0 2 4 6 8 fSIN 03152-PrD-011 INTERP[1] = 1 INTERP[0] = 1 Figure 45. Interpolation Modes REAL AND COMPLEX SIGNALS A complex signal contains both magnitude and phase information. Given two signals at the same frequency, if a point in time can be taken such that the signal leading in phase is cosinusoidal and the lagging signal is sinusoidal, then information pertaining to the magnitude and phase of a combination of the two signals can be derived; the combination of the two signals can be considered a complex signal. The cosine and sine can be represented as a series of exponentials; recalling that a multiplication by j is a counter clockwise rotation about the Re/Im plane, the phasor representation of a complex signal, with frequency f, can be shown Figure 46. Im Im Re C A/2 A/2 2πft Re A/2 A –f 0 +f FREQUENCY A/2 Acos(2πft) = A Asin(2πft) = A e+j2πft + e–j2πft 2 e+j2πft + e–j2πft 2j = = A 2 A 2 [e+j2πft + e–j2πft] [ je+j2πft + e–j2πft] 03152-PrD-012 C = Ae2πft = Acos(2πft) + jAsin(2πft) The cosine term represents a signal on the real plane with mirror symmetry about dc; this is referred to as the real, inphase or I component of a complex signal. The sine term represents a signal on the imaginary plane with mirror asymmetry about dc; this term is referred to as the imaginary, quadrature or Q complex signal component. The AD9784 has two channels of interpolation filters, allowing both I and Q components to be shaped by the same filter transfer function. The interpolation filters’ frequency response is a real transfer function. Two DACs are required to represent a complex signal. A single DAC can only synthesize a real signal. When a DAC synthesizes a real signal, negative frequency components fold onto the positive frequency axis. If the input to the DAC is mirror symmetrical about dc, the folded negative frequency components fold directly onto the positive frequency components in phase producing constructive signal summation. If the input to the DAC is not mirror symmetric about dc, negative frequency components may not be in phase with positive frequency components and will cause destructive signal summation. Different applications may or may not benefit from either type of signal summation. Figure 46. Complex Phasor Representation Rev. PrC | Page 28 of 52 Preliminary Technical Data AD9784 MODULATION MODES Table 29. Single Channel Modulation MODDUAL 0 0 0 0 0 0 0 0 CHANNEL 0 0 0 0 1 1 1 1 MOD[1] 0 0 1 1 0 0 1 1 MOD[0] 0 1 0 1 0 1 0 1 Either channel of the AD9784’s interpolation filter channels can be routed to the DAC and modulated. In single channel operation the input data may be modulated by a real sinusoid; the input data and the modulating sinusoid will contain both positive and negative frequency components. A double sideband output results when modulating two real signals. At the DAC output the positive and negative frequency components will add in phase resulting in constructive signal summation. As the modulating sinusoidal frequency becomes a larger fraction of the DAC update rate, fDAC, the more the sinc function of the DAC shapes the modulated signal bandwidth, and the closer the first image moves. As the AD9784 interpolation filter’s pass band represents a large portion of the input data’s Nyquist band, under certain modulation and interpolation modes it is possible for modulated signal bands to touch or overlap images if sufficient interpolation is not used. Figure 48 shows the effect of real modulation under all interpolation modes. The sinc shaping at the corners of the modulated signal band and the bandwidth to the first image for those cases whose pass bands do not touch or overlap are tabulated. Mode I Channel, no modulation I Channel, modulation by fDAC/2 I Channel, modulation by fDAC/4 I Channel, modulation by fDAC/8 Q Channel, no modulation Q Channel, modulation by fDAC/2 Q Channel, modulation by fDAC/4 Q Channel, modulation by fDAC/8 Table 30. Modulation none fDAC/2 fDAC/4 fDAC/8 None fSIN fSIN Overlap Overlap Interpolation ×2 ×4 2 fSIN 4 fSIN 2 fSIN 4 fSIN Touching 2 fSIN Overlap Touching ×8 8 fSIN 8 fSIN 4 fSIN 6 fSIN Table 31. Modulation None fDAC/4 None 0 –2.8241 –0.0701 –22.5378 Overlap fDAC/8 Overlap fDAC/2 Interpolation ×2 ×4 0 0 –0.6708 –0.1657 –1.1932 –2.3248 –9.1824 –6.1190 Touching –0.2921 –1.9096 Overlap Touching Modulated pass band edges sinc shaping(lower/upper). Rev. PrC | Page 29 of 52 ×8 0 –0.0413 –3.0590 –4.9337 –0.5974 –1.3607 –0.0727 –0.4614 AD9784 Preliminary Technical Data fDAC 7fDAC/8 3fDAC/4 fDAC/2 fDAC/2 3fDAC/4 3fDAC/8 3fDAC/8 5fDAC/8 fDAC/4 fDAC/4 5fDAC/8 fDAC/8 fDAC/8 0 –fDAC/8 –fDAC/4 –3fDAC/8 –fDAC/2 –5fDAC/8 –3fDAC/4 –7fDAC/8 –fDAC FILTERED INTERPOLATION IMAGES fDAC 03152-PrD-013 7fDAC/8 –fDAC/8 –fDAC/4 –3fDAC/8 –fDAC/2 –5fDAC/8 –3fDAC/4 –7fDAC/8 –fDAC fS/8 MODULATION Figure 47. Double Sideband Modulation NO INTERPOLATION 0 INTERP[1] = 0 INTERP[0] = 0 MOD[1] = 0 MOD[0] = 1 –50 –100 –6 –4 –2 0 2 4 6 8 fSIN ×2 INTERPOLATION 0 INTERP[1] = 0 INTERP[0] = 1 MOD[1] = 0 MOD[0] = 1 –50 –100 –150 –8 –6 –4 –2 0 2 4 6 8 fSIN ×4 INTERPOLATION 0 INTERP[1] = 1 INTERP[0] = 0 MOD[1] = 0 MOD[0] = 1 –50 –100 –150 –8 –6 –4 –2 0 2 4 6 INTERP[1] = 1 INTERP[0] = 1 MOD[1] = 0 MOD[0] = 1 –50 –100 –150 –8 8 fSIN ×8 INTERPOLATION 0 –6 –4 –2 0 2 4 Figure 48. Real Modulation by fDAC/2 under all Interpolation Modes Rev. PrC | Page 30 of 52 6 8 fSIN 03152-PrD-014 –150 –8 Preliminary Technical Data AD9784 NO INTERPOLATION 0 INTERP[1] = 0 INTERP[0] = 0 –50 MOD[1] = 1 –100 –150 –8 MOD[0] = 0 –6 –4 –2 0 2 4 6 8 fSIN ×2 INTERPOLATION 0 INTERP[1] = 0 –50 INTERP[0] = 1 MOD[1] = 1 –100 –150 –8 MOD[0] = 0 –6 –4 –2 0 2 4 6 8 fSIN ×4 INTERPOLATION 0 INTERP[1] = 1 –50 INTERP[0] = 0 MOD[1] = 1 –100 –150 –8 MOD[0] = 0 –6 –4 –2 0 2 4 6 8 fSIN ×8 INTERPOLATION 0 INTERP[1] = 1 MOD[1] = 1 –100 –150 –8 MOD[0] = 0 –6 –4 –2 0 2 4 6 8 fSIN 03215-PrD-015 INTERP[0] = 1 –50 Figure 49. Real Modulation by fDAC/4 under all Interpolation Modes NO INTERPOLATION 0 INTERP[1] = 0 INTERP[0] = 0 –50 MOD[1] = 1 –100 –150 –8 MOD[0] = 0 –6 –4 –2 0 2 4 6 8 fSIN ×2 INTERPOLATION 0 INTERP[1] = 0 INTERP[0] = 1 –50 MOD[1] = 1 –100 –150 –8 MOD[0] = 0 –6 –4 –2 0 2 4 0 8 fSIN 6 ×4 INTERPOLATION INTERP[1] = 1 INTERP[0] = 0 –50 MOD[1] = 1 –100 MOD[0] = 0 –6 –4 –2 0 2 4 0 8 fSIN 6 ×8 INTERPOLATION INTERP[1] = 1 INTERP[0] = 1 –50 MOD[1] = 1 –100 –150 8– MOD[0] = 0 –6 –4 –2 0 2 4 Figure 50. Real Modulation by fDAC/8 under all Interpolation Modes Rev. PrC | Page 31 of 52 6 8 fSIN 03152-PrD-017 –150 –8 AD9784 Preliminary Technical Data Table 32. Dual Channel Complex Modulation MODSING 0 0 0 0 0 0 0 0 REALIMAG 0 0 0 0 1 1 1 1 MOD[1] 0 0 1 1 0 0 1 1 MOD[0] 0 1 0 1 0 1 0 1 Mode Real output, no modulation Real output, modulation by fDAC/2 Real output, modulation fDAC/4 Real output, modulation fDAC/8 Image output, no modulation Imag output, modulation by fDAC/2 Imag output, modulation by fDAC/4 Imag output, modulation by fDAC/8 Table 33. In dual channel mode, the two channels may be modulated by a complex signal, with either the real or imaginary modulation result directed to the DAC. Assume initially that the complex modulating signal is defined for a positive frequency only; this causes the output spectrum to be translated in frequency by the modulation factor only. No additional sidebands are created as a result of the modulation process, and therefore the bandwidth to the first image from the baseband bandwidth is the same as the output of the interpolation filters. Furthermore, pass bands will not overlap or touch. The sinc shaping at the corners of the modulated signal band are tabulated. Figure 52 shows the complex modulations. Modulation None None 0 –2.8241 –0.0701 –22.5378 –0.4680 –6.0630 –1.3723 –4.9592 fDAC/2 fDAC/4 fDAC/8 Interpolation ×2 ×4 0 0 –0.6708 –0.1657 –1.1932 –2.3248 –9.1824 –6.1190 –0.0175 –0.2921 –3.3447 –1.9096 –0.1160 –0.0044 –1.7195 –0.7866 ×8 0 –0.0413 –3.0590 –4.9337 –0.5974 –1.3607 –0.0727 –0.4614 Modulated passband edges sinc shaping(lower/upper). Figure 51. Complex Modulation Rev. PrC | Page 32 of 52 5fDAC/8 3fDAC/4 7fDAC/8 5fDAC/8 3fDAC/4 7fDAC/8 fDAC fDAC/2 fDAC/2 3fDAC/8 fDAC 03152-PrD-018 3fDAC/8 fDAC/4 NO NEGATIVE SIDEBAND fDAC/8 0 –fDAC/8 –fDAC/4 –3fDAC/8 –fDAC/2 –5fDAC/8 –3fDAC/4 –7fDAC/8 –fDAC fS/8 MODULATION fDAC/4 fDAC/8 0 –fDAC/8 –fDAC/4 –3fDAC/8 –fDAC/2 –5fDAC/8 –3fDAC/4 –7fDAC/8 –fDAC FILTERED INTERPOLATION IMAGES Preliminary Technical Data AD9784 ×2 INTERPOLATION 0 INTERP[1] = 0 INTERP[0] = 1 –50 MOD[1] = 0 –100 –150 –8 MOD[0] = 1 –6 –4 –2 0 2 4 0 6 8f SIN ×4 INTERPOLATION INTERP[1] = 1 INTERP[0] = 0 –50 MOD[1] = 0 –100 –150 –8 MOD[0] = 1 –6 –4 –2 0 2 4 0 6 8f SIN ×8 INTERPOLATION INTERP[1] = 1 INTERP[0] = 1 –50 –150 –8 MOD[0] = 1 –6 –4 –2 0 2 4 6 8f SIN 03152-PrD-019 MOD[1] = 0 –100 Figure 52. Complex Modulation by fDAC/2 under all Interpolation Modes ×2 INTERPOLATION 0 INTERP[1] = 0 INTERP[0] = 1 –50 MOD[1] = 1 –100 –150 –8 MOD[0] = 0 –6 –4 –2 0 2 4 0 6 8 fSIN ×4 INTERPOLATION INTERP[1] = 1 INTERP[0] = 0 –50 MOD[1] = 1 –100 –150 –8 MOD[0] = 0 –6 –4 –2 0 2 4 0 6 8 fSIN ×8 INTERPOLATION INTERP[1] = 1 INTERP[0] = 1 –50 –150 –8 MOD[0] = 0 –6 –4 –2 0 2 4 6 8 fSIN 03152-PrD-020 MOD[1] = 1 –100 Figure 53. Complex Modulation by fDAC/4 under all Interpolation Modes ×2 INTERPOLATION 0 INTERP[1] = 0 INTERP[0] = 1 –50 MOD[1] = 1 –100 –150 –8 MOD[0] = 1 –6 –4 –2 0 2 4 0 6 8 fSIN ×4 INTERPOLATION INTERP[1] = 1 INTERP[0] = 0 –50 MOD[1] = 1 –100 –150 –8 MOD[0] = 1 –6 –4 –2 0 2 4 0 6 8 fSIN ×8 INTERPOLATION INTERP[1] = 1 INTERP[0] = 1 –50 –150 –8 MOD[0] = 1 –6 –4 –2 0 2 4 Figure 54. Complex Modulation by fDAC/8 under all Interpolation Modes Rev. PrC | Page 33 of 52 6 8 fSIN 03152-PrD-021 MOD[1] = 1 –100 AD9784 Preliminary Technical Data POWER DISSIPATION 60 The AD9784 has seven power supply domains: two 3.3 V analog domains (AVDD1 and AVDD2), two 2.5 V analog domains (ADVDD and ACVDD), one 2.5 V clock domain (CLKVDD), and two digital domains (DVDD, which runs from 2.5 V, and DRVDD, which can run from 2.5 V or 3.3 V). 4× 2× 40 IDVDD (mA) The current needed for the 3.3 V analog supplies, AVDD1 and AVDD2, is consistent across speed and varying modes of the AD9784. Nominally, the current for AVDD1 is 29 mA across all speeds and modes, while the current for AVDD2 is 20 mA. 8× 50 30 1× 20 0 25 50 75 100 125 150 FDATA (MSPS) 175 200 225 250 Figure 56. CLKVDD Supply Current vs. Clock Speed and Interpolation Rates 30 425 400 375 350 325 300 275 250 225 200 175 150 125 100 75 50 25 0 2× fs/8 4× fs/8 8× fs/8 4× fs/4 25 8× fs/4 2× fs/4 8× 2× 4× 20 4× 8× IDVDD (mA) 2× 15 1× 10 1× 0 25 50 75 100 125 150 FDATA (MSPS) 175 200 225 250 Figure 55. DVDD Supply Current vs. Clock Speed, Interpolation, and Modulation Rates 0 0 25 50 75 100 125 150 FDATA (MSPS) 175 200 225 250 Figure 57. ADVDD and ACVDD Supply Current vs. Clock Speed and Interpolation Rates Rev. PrC | Page 34 of 52 03152-PrD-079 5 03152-PrD-077 IDVDD (mA) 0 03152-PrD-078 10 The current for the 2.5 V analog supplies and the digital supplies varies depending on speed and mode of operation. Figure 55, Figure 56, and Figure 57 show this variation. Note that CLKVDD, ADVDD, and ACVDD vary with clock speed and interpolation rate, but not with modulation rate. Preliminary Technical Data AD9784 fDAC 7fDAC/8 3fDAC/4 5fDAC/8 fDAC/2 3fDAC/8 fDAC/4 fDAC/8 0 –fDAC/8 –fDAC/4 –3fDAC/8 –fDAC/2 –5fDAC/8 –3fDAC/4 –fDAC –7fDAC/8 FILTERED INTERPOLATION IMAGES fDAC 7fDAC/8 3fDAC/4 5fDAC/8 fDAC/2 3fDAC/8 fDAC/4 fDAC/8 0 –fDAC/8 –fDAC/4 –3fDAC/8 –fDAC/2 –5fDAC/8 –3fDAC/4 –fDAC –7fDAC/8 fS/8 MODULATION fDAC 03152-PrD-022 7fDAC/8 3fDAC/4 5fDAC/8 fDAC/2 3fDAC/8 fDAC/4 fDAC/8 0 –fDAC/8 –fDAC/4 –3fDAC/8 –fDAC/2 –5fDAC/8 –3fDAC/4 –fDAC –7fDAC/8 fS/4 MODULATION Figure 58. Complex Modulation with Negative Frequency Aliasing DUAL CHANNEL COMPLEX MODULATION WITH HILBERT Table 34. HILBERT 0 1 Mode Hilbert transform off Hilbert transform on When complex modulation is performed, the entire spectrum is translated by the modulation factor. If the resulting modulated spectrum is not mirror symmetric about dc, when the DAC synthesizes the modulated signal, negative frequency components will fall on the positive frequency axis and can cause destructive summation of the signals. For some applications, this can be seen as distorting the modulated output signal. By performing a second complex modulation with a modulating signal having a fixed π/2 phase difference, Figure 59 (Y), relative to the original complex modulation signal, Figure 59 (X), taking the Hilbert transform of the new resulting complex modulation, and subtracting it from the original complex modulation output all negative frequency components can be folded in phase to the positive frequency axis before being synthesized by the DAC. When the DAC synthesizes the modulated output there are no negative frequency components to fold onto the positive frequency axis out of phase; consequently no distortion is produced as a result of the modulation process. 0 ALIASED NEGATIVE FREQUENCY INTERPOLATION IMAGES X = Ae j2π(f + fm)t Y = Ae j2π(f + fm)t – π/2 Im Im C=X–Z Re Im Re –50 A/2 A/2 A/2 f A/2 A/2 A A/2 00 00 A/2 f A/2 A/2 f A/2 f A 03152-PrD-023 A/2 dBFS A/2 Re Figure 59. Negative Frequency Image Rejection –100 –150 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.1 0.2 0.3 0.4 Figure 60. Negative Frequency Aliasing Distortion Rev. PrC | Page 35 of 52 0.5 03152-PrD-024 Re Z = HILBERT(Y) Im AD9784 Preliminary Technical Data Figure 60 shows this effect at the DAC output for a mirror asymmetic signal about dc produced by complex modulation without a Hilbert transform. The signal bandwidth was narrowed to show the aliased negative frequency interpolation images. The transfer function of an ideal Hilbert transform has a +90° phase shift for negative frequencies, and a –90° phase shift for positive frequencies. Because of the discontinuities that occur at 0 Hz and at 0.5 × Sample Rate, any real implementation of the Hilbert Transform trades off bandwidth versus ripple. In contrast, Figure 61 shows the same waveform with the Hilbert transform applied. Clearly, the aliased interpolation images are not present. Figure 62 and Figure 63 show the gain of the Hilbert transform versus frequency. Gain is essentially flat, with a pass-band ripple of 0.1dB over the frequency range 0.07 × Sample Rate to 0.43 × Sample Rate. 0 dBFS –50 –150 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.1 0.2 0.3 0.4 0.5 03152-PrD-025 –100 Figure 64 shows the phase response of the Hilbert transform implemented in the AD9784. The phase response for positive frequencies begins at –90° at 0 Hz, followed by a linear phase response (pure time delay) equal to nine filter taps (nine clock cycles). For negative frequencies, the phase response at 0 Hz is +90°, again followed by a linear phase delay of nine filter taps. To compensate for the unwanted 9-cycle delay, an equal delay of nine taps is used in the AD9784 digital signal path opposite to the Hilbert transform. This delay block is noted as t on the data sheet. Figure 61. Effects of Hilbert Transform 10 0 –10 –20 If the output of the AD9784 is to be used with a quadrature modulator, negative frequency images are cancelled without the need of a Hilbert transform. –30 –40 –50 HILBERT TRANSFORM IMPLEMENTATION –60 The Hilbert transform on the AD9784 is implemented as a 19coefficient FIR. The coefficients are given in Table 35 –70 900 1000 03152-PrD-074 900 1000 03152-PrD-075 –90 Table 35. Coefficient H(1) H(2) H(3) H(4) H(5) H(6) H(7) H(8) H(9) H(10) H(11) H(12) H(13) H(14) H(15) H(16) H(17) H(18) H(19) –80 –100 Integer Value –6 0 –17 0 –40 0 –91 0 –318 0 318 0 91 0 40 0 17 0 6 100 200 300 400 500 600 700 800 Figure 62. Hilbert Transform Gain 1.0 0.8 0.6 0.4 0.2 0 –0.2 –0.4 –0.6 –0.8 –1.0 100 200 300 400 500 600 700 800 Figure 63. Hilbert Transform Ripple Rev. PrC | Page 36 of 52 Preliminary Technical Data AD9784 A baseband double sideband signal modulated to IF increases IF filter complexity and reduces power efficiency. If the baseband signal is complex, a single sideband IF modulation can be used, relaxing IF filter complexity and increasing power efficiency. 4 3 2 1 The AD9784 has the ability to place the baseband single sideband complex signal either above the IF frequency or below it. Figure 66 illustrates the baseband selection. 0 –1 –2 –4 100 200 400 600 800 1000 1200 0 03152-PrD-076 –3 –50 dBFS Figure 64. Phase Response of Hilbert Transform Table 36. Dual Channel Complex Modulation Sideband Selection –100 –150 –0.5 AD9784 LO Q AD9784 Im() –0.3 –0.2 –0.1 0 0.1 0.2 0.3 0.4 0.5 0.4 0.5 Figure 66. Upper IF Sideband Rejected Re() 0 90 03151-PrD-003 I –0.4 03152-PrD-027 Mode Lower IF sideband rejected Upper IF sideband rejected 03152-PrD-028 Sideband 0 1 0 Figure 65. AD9784 Driving Quadrature Modulator The AD9784 can be configured to drive a quadrature modulator, representatively as in Figure 65. Where two AD9784s are used with one AD9784 producing the real output, the second AD9784 produces the imaginary output. By configuring the AD9784 as a complex modulator coupled to a quadrature modulator, IF image rejection is possible. The quadrature modulator acts as the real part of a complex modulation producing a double sideband spectrum at the local oscillator (LO) frequency, with mirror symmetry about dc. dBFS –50 Rev. PrC | Page 37 of 52 –100 –150 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.1 0.2 0.3 Figure 67. Lower IF Sideband Rejected AD9784 Preliminary Technical Data 0 0 IF fIF –fIF –fIF BASEBAND fIF SIDEBAND = 0 03152-PrD-029 fIF 0 –fIF SIDEBAND = 1 Figure 68. IF Quadrature Modulation of Real and Complex Baseband Signals Table 37. Data Port Synchronization PLOCKEXT 1 0 0 0 0 0 0 DCLKEXT X 0 0 1 1 1 1 MODSYNC X 0 1 0 0 1 1 DCLKCRC X X X 0 1 0 1 Mode PLL output Dataclk Master Modulator Master Dataclk Slave Dataclk Slave Modulator Slave Modulator Slave In applications where two or more AD9784s are used to synthesize several digital data paths, it may be necessary to ensure that the digital inputs to each device are latched synchronously. In complex data processing applications, digital modulator phase alignment may be required between two AD9784s. In order to allow data synchronization and phase alignment, only one AD9784 should be configured as a master device, providing a reference clock for another slave-configured AD9784. With synchronization enabled, a reference clock signal is generated on the DATACLK/PLL_LOCK pin of the master. The DATACLK/PLL_LOCK pins on the slave devices act as inputs for the reference clock generated by the master. The DATACLK/ PLL_LOCK pin on the master and all slaves must be directly connected. All master and slave devices must have the same clock source connected to their respective CLK+/CLK– pins. When configured as a master, the reference clock generated may take one of two forms. In modulator master mode, the reference clock will be a square wave with a period equal to 16 cycles of the DAC update clock. Internal to the AD9784 is a 16-state finite state machine, running at the DAC update rate. This state machine generates all internal and external synchronization clocks and modulator phasings. The rising edge of the master reference clock is time aligned to the internal state machine’s state zero. Slave devices use the master’s reference clock to synchronize their data latching and align their modulator’s phase by aligning their local state machine state zero to the master. Function PLL locked flag output, synchronizer disabled Channel data rate clock output Modulator synchronization clock output Input channel data rate clock, DLL off Input channel data rate clock, DLL on Input modulator synchronizer clock, DLL off Input modulator synchronizer clock, DLL on The second master mode, DATACLK master mode, generates a reference clock that is at the channel data rate. In this mode, the slave devices align their internal channel data rate clock to the master. If modulator phase alignment is needed, a concurrent serial write to all slave devices is necessary. To achieve this, the CSB pin on all slaves must be connected together and a group serial write to the MODADJ register bits must be performed; the modulator coefficient alignment is updated on the next rising edge of the internal state machine following a successful serial write, Figure 69. Modulator master mode does not need a concurrent serial write as slaves lock to the master phase automatically. In a slave device, the local channel data rate clock and the digital modulator clock are created from the internal state machine. The local channel data rate clock is used by the slave to latch digital input data. At high data rates, the delay inherent in the signal path from master to slave may cause the slave to lag the master when acquiring synchronization. To account for this, an integer number of the DAC update clock cycles may be programmed into the slave device as an offset. The value in DATADJ allows the local channel data rate clock in the slave device to advance by up to eight cycles of the DAC clock or delayed by up to seven cycles, Figure 70. The digital modulator coefficients are updated at the DAC clock rate and decoded in sequential order from the state machine according to Figure 71. The MODADJ bits can be used to align a different coefficient to the finite state machine’s zero state as shown in Figure 72. Rev. PrC | Page 38 of 52 Preliminary Technical Data AD9784 DAC CLOCK STATE MACHINE 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 MODULATOR COEFFICIENT 1 0 –1 0 1 0 –1 0 1 0 –1 0 –1 0 1 0 –1 0 1 0 –1 0 1 0 MODADJ 1 0 –1 0 000 –1 0 1 0 000 03152-PrD-030 STATE MACHINE CYCLE CLOCK CHANNEL DATA RATE CLOCK Figure 69. Synchronous Serial Modulator Phase Alignment DATADJ[3:0] 0000 1111 0001 DAC CLOCK 03152-PrD-031 RECEIVED CHANNEL DATA RATE CLOCK LOCAL CHANNEL DATA RATE CLOCK –1 +1 Figure 70. Local Channel Data Rate Clock Synchronized with Offset STATE 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 DECODE 1 0 1/ 2 0 0 0 –1/ 2 0 –1 0 –1/ 2 0 0 0 –1/ 2 0 fs/8 0 0 0 2 3 4 1 5 6 2 7 03152-PrD-032 fs/4 fs/2 1 3 1 Figure 71. Digital Modulator State Machine Decode MODADJ[2:0] 000 010 101 DAC CLOCK 14 15 0 1 2 3 15 0 1 15 0 1 2 MODULATOR COEFFICIENT –1 0 1 0 –1 0 0 –1 0 1 0 –1 0 03152-PrD-033 STATE MACHINE STATE MACHINE CYCLE CLOCK Figure 72. Local Modulator Coefficient Synchronized with Offset Rev. PrC | Page 39 of 52 AD9784 Preliminary Technical Data OPERATING THE AD9784 REV E EVALUATION BOARD This section helps the user get started with the AD9784 evaluation board. Because it is intended to provide starter information to power up the board and verify correct operation, a description of some of the more advanced modes of operation has been omitted. For a description of the various SPI registers and the effect they have on the operating modes of the AD9784, see the Mode Control (via SPI Port) section. POWER SUPPLIES The AD9784 Rev E Evaluation Board has five power supply connectors, labeled VDDIN, CVDIN, VDD2IN, VDD3IN, and AVDIN. The AD9784 itself actually has seven power supply domains. To reconcile the power supply domains on the chip with the power supply connectors on the evaluation board, use Table 38. Additionally, the DRVDD power supply on the AD9784 is used to supply power for the digital input bus. DRVDD can be run from 2.5 V or 3.3 V. On the evaluation board, DRVDD is jumper selectable by JP1, just to the left of the chip on the evaluation board. With the jumper set to the 3.3 V position, DRVDD chip receives its power from VDD3IN. With the jumper set to the 2.5 V position, DRVDD receives its power from AVDIN. CLKVDDS PECL CLOCK DRIVER The AD9784 system clock is driven from an external source via connector S1. The AD9784 Evaluation Board includes an OnSemiconductor MC100EPT22 PECL clock driver. In the factory, the evaluation board is set to use this PECL driver as a single-ended-to-differential clock receiver. The PECL driver can be set to run from 2.5 V from the CVDIN power connector, or 3.3 V from the VDD3IN power connector. This setting is done via jumper, JP2, situated next to the CVDIN power connector, and by setting input bias resistors R23 and R4 on the evaluation board. The factory default is for the PECL driver to be powered from CVDIN at 2.5 V (R23 = 90.9 Ω, R4 = 115 Ω). To operate the PECL driver with a 3.3 V supply, R23 must be replaced with a 115 Ω resistor and R4 must be replaced with a 115 Ω resistor, as well as changing the position of JP2. The schematic of the PECL driver section of the evaluation board is shown below in Figure 73. A low jitter sine wave can be used as the clock source. Care must be taken to make sure the clock amplitude does not exceed the power supply rails for the PECL driver. CLKVDDS CLK+ ACLKX R23 115Ω 7 R5 50Ω MC100EPT22 1 COND;5 U2 CLKVDDS;8 2 R4 90.9Ω R6 50Ω R7 50Ω CLK– 03152-PrD-080 C32 0.1µF Figure 73. PECL Driver on AD9784 Rev E Evaluation Board Table 38. Evaluation Board Label VDDIN CVDIN VDD2IN VDD3IN AVDIN PS Domain on Chip DVDD CLKVDD ACVDD and ADVDD AVDD2 AVDD1 Nominal Power Supply Voltage (V) 2.5 2.5 2.5 3.3 3.3 Description SPI port Clock circuitry Analog circuitry containing clock and digital interface circuitry Switching analog circuitry Analog output circuitry Rev. PrC | Page 40 of 52 Preliminary Technical Data AD9784 DATA INPUTS SPI PORT Digital data inputs to the AD9784 are accessed on the evaluation board through connectors J1 and J2. These are 40 pin right angle connectors that are intended to be used with standard ribbon cable connectors. The input levels should be either 3.3 V or 2.5 V CMOS, depending on the setting of the DRVDD jumper JP1. The data format is selectable through Register 02h, Bit 7 (DATAFMT). With this bit set to a default 0, the AD9784 assumes that the input data is in twos complement format. With this bit set to 1, data should be input in offset binary format. SW1 is a hard reset switch that sets the AD9784 to its default state. It should be used every time the AD9784 power supply is cycled or the clock is interrupted, or if new data is to be written via the SPI port. For a description of the various SPI registers and the effect they have on the operating modes of the AD9784, see the Mode Control (via SPI Port) section. Set the SPI software to read back data from the AD9784 and verify that when the software is run, the expected values are read back. When the evaluation board is first powered up and the clock and data are running, it is recommended that the proper operating current is verified. Depress reset switch SW1 to ensure that the AD9784 is in the default mode. The default mode for the AD9784 is for the internal PLL to be disabled, and the interpolation set to 1×. The modulator is turned off in the default mode. The nominal operating currents for the evaluation board in the power-up default mode are shown in Table 39. Additionally, the DRVDD power supply on the AD9784 is used to supply power for the digital input bus. DRVDD can be run from 2.5 V or 3.3 V. On the evaluation board, DRVDD is jumper selectable by JP1, just to the left of the chip on the evaluation board. With the jumper set to the 3.3 V position, DRVDD chip receives its power from VDD3IN. With the jumper set to the 2.5 V position, DRVDD receives its power from AVDIN. OPERATING WITH PLL DISABLED The SPI registers referenced in this section are shown in Table 40. With the PLL disabled, the evaluation board clock input must be run at the intended DAC sample rate, up to the specified limit of 500 MSPS. At the same time, the interpolation rate should be set so the input data rate does not exceed the 200 MSPS limit. In the default mode with the PLL disabled, the DATACLK signal from the AD9784 is available at connector S2. The rate of this clock is the system clock applied at S1, divided by the interpolation rate. DATACLK can be used to synchronize the external data into the AD9784. Table 39. Nominal Operating Currents in Power-Up Default Mode Evaluation Board Power Supply VDDIN CVDIN VDD2IN VDD3IN AVDIN 50 MSPS 24 79 1 30 27 Nominal Current @ Speed (mA) 100 MSPS 150 MSPS 49 74 83 87 4 6 30 30 27 27 200 MSPS 99 92 8 30 27 Table 40. SPI Registers Register 01h 04h Bit 7 0 0 1 1 Bit 7 INTERP[1] PLLON Bit 6 INTERP[0] PLLMULT[1] Interpolation Rate Bit 6 Rate 0 1× 1 2× 0 4× 1 8× Bit 6 0 0 1 1 Bit 5 Bit 4 Bit 3 Bit 0 PLLMULT[0] PLLDIV[1] PLLDIV[0] PLOCKEXT PLL Multiplier Bit 5 0 1 0 1 Rev. PrC | Page 41 of 52 Mult 2× 4× 8× 16× Bit 4 0 0 1 1 PLL Divider Bit 3 0 1 0 1 Div ÷1 ÷2 ÷4 ÷8 AD9784 Preliminary Technical Data OPERATING WITH PLL ENABLED ANALOG OUTPUT Note that a specific revision of the AD9784 on the Rev E Evaluation Board has a nonfunctioning PLL. This revision can be identified by the xxx. The analog output of the AD9784 is accessed via connector S3. Once all settings are selected and current levels, PLL lock state, and SPI port functionality are verified, the analog signal at S3 can be viewed. For most of the AD9784’s applications, a spectrum analyzer is the instrument of choice to verify proper performance. A typical spectral plot is shown in Figure 74, with the AD9784 synthesizing a two-tone signal in the default mode with a 200 MSPS sample rate. A single tone CW signal should provide output power of approximately +0.5 dBm to the spectrum analyzer. With the AD9784 PLL enabled, the evaluation board clock input must be run at the data input rate, up to the specified 200 MSPS limit. The PLL controls the internal clock multiplication and drives the interpolation filters and digital modulator. The internal PLL has a VCO in the control loop that is designed to operate optimally over the 200 MHz to 500 MHz range. The VCO speed can be calculated as follows: VCO Speed = Input Data Rate × PLLMULT[1,0] The interpolation rate is set by Bits 6 and 7. With the PLL enabled, the settings for the interpolation rate, the PLL multiplier, and the PLL divide are interrelated. The interpolation rate must meet the following criteria: Interpolation Rate = [Settings of Bits 6, 7] = [PLLMULT ÷ PLLDIVIDER] If the spectrum does not look correct at this point, the data input may be violating setup and hold times with respect to the input clock. To correct this, the user should vary the input data timing. If this is not possible, SPI Register 02h, Bit 4 can be inverted. This bit controls the clock edge upon which the data is latched. If these methods do not correct the spectrum, it is unlikely that the issue is timing related. This note should then be reread to verify that all instructions have been followed. Therefore, assuming the input data rate is constant and the VCO is at optimal speed, if the interpolation rate is increased by a factor of M, the PLLMULT setting must be decreased by the same factor M. 10 0 –10 –20 –30 –40 –50 –60 –60 –70 –80 –90 –100 START 100 kHz 19.9 MHz/ STOP 200 MHz Figure 74. Typical Spectral Plot Rev. PrC | Page 42 of 52 03152-PrD-081 With the PLL enabled, DATACLK connector S2 indicates the lock state of the PLL. A Logic 1 from S2 indicates lock; a Logic 0 indicates the PLL is not currently locked. Rev. PrC | Page 43 of 52 Figure 75. Power Supply Distribution 2.5VQ CGND;3,4,5 03152-PrD-082 S11 SMAEDGE 3.3VQ AGND; 3,4,5 CLKVDD_IN 2 SMAEDGE 1 S10 3.3V AGND2; 3,4,5 AVDD_IN S9 SMAEDGE 2.5VN DGND; 3,4,5 ADVDD3_IN S5 SMAEDGE 2.5V AGND2; 3,4,5 DVDD_IN S7 SMAEDGE ADVDD2_IN TP6 RED TP4 RED TP2 RED TP18 BLK TP13 RED TP1 RED C69 0.1µF C68 0.1µF AVD1 C67 0.1µF C48 0.1µF C47 0.1µF POWER INPUT FILTERS FERRITE C63 + 22µF 16V L1 FERRITE + C64 22µF 16V L2 FERRITE C65 + 22µF 16V L3 FERRITE + C46 22µF 16V TP17 BLK L9 FERRITE + C45 22µF 16V L8 FERRITE L12 L11 TP7 BLK JP5 CVD TP5 BLK JP10 TP3 BLK AVD2 JP9 TP16 BLK VDD JP34 1 3 A B JP33 JP30 C76 0.1µF C34 0.1µF JP6 JP8 JP7 CLKVDDS DRVDD AVDD2 ACVDD ADVDD BLK BLK BLK BLK ACLKX BLK TP30 TP31 TP32 TP33 TP34 FERRITE L6 JP1 2 1 3 A B C29 22µF 16V JP36 C75 0.1µF JP1 2 CLKVDD AVDD AVDD2 DVDDS DVDD DVDD TP12 FERRITE BLK AVD3 C32 0.1µF L7 VAL L10 VAL L13 VAL L14 VAL R4 90.9Ω 7 R23 115Ω CLKVDDS BLK TP36 BLK TP35 C35 0.1µF CLKVDDS MC100EPT22 1 CGND; 5 U2 CLKVDDS; 8 2 + C28 4.7µF 6.3V 6 4 CLKVDDS; 8 CGND; 5 U2 MC100EPT22 3 CLK– AUX CLOCK 50Ω R6 50Ω R5 50Ω R7 CLK+ Preliminary Technical Data AD9784 Figure 76. Local Circuitry Rev. PrC | Page 44 of 52 TP14 WHT IQ B A S6 1 2 3 03152-PrD-083 DGND; 3,4,5 OPCLK_3 JP28 BD15 C33 0.1µF OPCLK JP27 BD14 + C7 10µF 6.3V DVDD + C8 10µF 6.3V DVDD + C9 10µF 6.3V DVDD OPCLK S4 DATACLK S2 C54 0.001µF DGND; 3,4,5 + C31 10µF 6.3V DRVDD + C10 10µF 6.3V C26 0.001µF C23 0.001µF C24 0.001µF C25 0.001µF C36 0.1µF C39 0.1µF C41 0.1µF C40 0.1µF RESET 58 SPI_CSB 57 22 P1B5 23 P1B4 24 P1B3 AD05 AD04 AD03 DVDD6 52 29 P1B0LSB AD00 P2B2 49 P2B3 48 32 P2B15MSB-IQSEL 33 P2B14-OPCLK P2B7 42 39 P2B10 AD9786BTSP P2B8 41 P2B6 43 38 P2B11 BD11 BD10 40 P2B9 DVDD5 44 37 P2B12 BD12 BD09 P2B5 46 DCOM5 45 36 DVDD4 BD08 BD07 BD06 BD05 BD03 BD04 P2B4 47 34 P2B13 U1 BD01 P2B1 50 31 DCLK-PLLL BD02 BD00 SPSDO SPSDI SPCLK SPCSB RESET TP11 WHT C37 0.1µF + C30 10V 10µF C17 0.1µF C19 0.1µF C15 0.1µF C66 10µF 6.3V C2 10µF 6.3V + C3 10µF 6.3V 4 C22 0.001µF DVDD + C6 10µF 6.3V C38 0.1µF 3 6 4 DVDD R42 49.9Ω 2 1 DRVDD AGND; 3,4,5 S3 TP29 BLK SW1 FLOAT; 5 4 3 AVDD2 AGND; 3,4,5 S3 OUT1 C61 0.001µF S P 1 TTWB-1-B T2B NC = 5 S P TC1-1T + C5 10µF 6.3V 6 5 4 1 2 3 T2A NC = 5 R9 49.9Ω R10 49.9Ω C18 0.001µF RESET C21 0.001µF 6 T3 S 1 C4 0.1µF 3 P ADVDD AVDD C62 0.1µF R8 C16 2.000kΩ 0.1µF 0.01% TP8 WHT TP10 WHT + + C55 0.001µF C14 0.1µF C20 0.001µF ACVDD C49 0.1µF P2B0LSB 51 35 DCOM4 BD13 DCOM6 53 28 P1B1 AD01 30 DRVDD1 SP-SDO 54 27 P1B2 AD02 SP-SDI 55 REFIO 59 21 P1B6 AD06 SP-CLK 56 DNC1 61 FSADJ 60 20 P1B7 AD07 26 DVDD3 ADVDDP2 62 19 P1B8 AD08 25 DCOM3 ADCOMP2 63 18 P1B9 AD10 AD09 AVDD2P2 66 15 P1B10 AD11 ACVDDP2 64 ACOM2P2 67 14 P1B11 AD12 17 DVDD2 AVDD1P1 68 13 P1B12 ACCOM2P2 65 ACOM1P21 69 12 P1B13 AD13 16 DCOM2 IOUTB 70 11 P1B14 IOUTA 71 ACOM1P11 72 AVDD1P2 73 ACOM2P12 74 AD14 10 P1B15MSB 9 DVDD1 8 DCOM1 7 CLKCOM2 S1 CGND; 3,4,5 JP23 AVDD2P1 75 ACCOMP1 78 3 CLKVDD2 DNC2 80 ADVDDP1 79 2 LPF 1 CLKVDD1 C49 1pF 6 CLK– AD15 C42 0.1µF ACLKX CLK+ C11 0.1µF ACVDDP1 77 CLK– DVDD C12 0.1µF ACOM2P1 76 R1 50Ω +C1 10µF 6.3V CLKVDD 5 CLK+ 3 4 JP22 CLKVDD 4 CLKCOM1 2 5 T1 T1-1T 1 T2A 6 TP15 WHT C13 0.1µF R3 10kΩ ADTL1-12 R2 10kΩ AD9784 Preliminary Technical Data Preliminary Technical Data R29 100Ω 1 4 3 6 5 8 7 10 9 12 11 14 13 16 15 18 17 20 19 22 21 24 23 26 25 28 27 30 29 32 31 34 33 36 35 38 37 40 39 3 4 5 6 7 8 9 10 2 AX13 3 AX12 4 AX11 5 AX10 6 AX09 7 AX08 8 AX07 1 AX06 2 AX05 3 AX04 4 AX03 5 AX02 6 AX01 7 AX00 8 1 2 3 4 5 6 7 8 9 10 RP1 22 RP1 22 RP1 22 RP1 22 RP1 22 RP1 22 RP1 22 RP1 22 RP2 22 RP2 22 RP2 22 RP2 22 RP2 22 RP2 22 RP2 22 RP2 22 RP6 DNP R1 R2 R3 R4 R5 R6 R7 R8 R9 AX00 R38 100Ω AX04 2 RP5 DNP AX14 AX07 AX05 R1 R2 R3 R4 R5 R6 R7 R8 R9 1 RIBBON J1 AX06 JP12 AX11 AX15 RCOM 2 JP3 AX10 R33 100Ω 1 DATA-A AX09 R32 100Ω RCOM AX12 R28 100Ω AX08 R31 100Ω R39 100Ω R40 100Ω R34 100Ω R41 100Ω 2 3 4 5 6 7 8 9 10 RP7 DNP 16 AD15 15 AD14 14 AD13 13 AD12 12 AD11 11 AD10 10 AD09 9 AD08 16 AD07 15 AD06 14 AD05 13 AD04 12 AD03 11 AD02 10 AD01 9 AD00 1 2 3 4 5 6 7 8 9 10 RP8 DNP R1 R2 R3 R4 R5 R6 R7 R8 R9 JP21 R44 100Ω R43 100Ω 1 R1 R2 R3 R4 R5 R6 R7 R8 R9 AX01 JP19 AX02 AX03 03152-PrD-084 AX13 R27 100Ω R30 100Ω RCOM AX14 R26 100Ω RCOM AX15 AD9784 R46 100Ω Figure 77. Digital Data Port A Input Terminations Rev. PrC | Page 45 of 52 AD9784 Preliminary Technical Data R60 100Ω BX13 R64 100Ω 3 4 5 6 7 8 9 BX14 2 BX13 3 BX12 4 BX11 5 BX10 6 BX09 7 BX08 8 BX07 1 BX06 2 BX05 3 BX04 4 BX03 5 BX02 6 BX01 7 BX00 8 6 5 8 7 10 9 12 11 14 13 16 15 18 17 20 19 22 21 24 23 26 25 28 27 30 29 32 31 34 33 SDO 36 35 CLK 38 37 SDI 40 39 CSB 1 RIBBON J2 2 3 4 5 6 7 8 9 10 RP3 22 RP3 22 RP3 22 RP3 22 RP3 22 RP3 22 RP3 22 RP3 22 RP4 22 RP4 22 RP4 22 RP4 22 RP4 22 RP4 22 RP4 22 RP4 22 RP11 DNP R1 R2 R3 R4 R5 R6 R7 R8 R9 BX00 BX07 R55 100Ω BX04 2 RP12 10 DNP 1 3 BX05 R1 R2 R3 R4 R5 R6 R7 R8 R9 BX15 4 BX06 JP31 BX11 R54 100Ω R53 100Ω R56 100Ω R47 100Ω 2 3 4 5 6 7 8 9 10 RP9 DNP 16 BD15 15 BD14 14 BD13 13 BD12 12 BD11 11 BD10 10 BD09 9 BD08 16 BD07 15 BD06 14 BD05 13 BD04 12 BD03 11 BD02 10 BD01 9 BD00 1 2 3 4 5 6 7 8 9 10 RP10 DNP R1 R2 R3 R4 R5 R6 R7 R8 R9 JP25 R51 100Ω R49 100Ω 1 R1 R2 R3 R4 R5 R6 R7 R8 R9 BX01 JP24 BX02 BX03 03152-PrD-085 1 RCOM 2 JP26 BX10 R63 100Ω 1 DATA-B BX09 R59 100Ω RCOM BX12 BX08 R58 100Ω RCOM R61 100Ω BX14 R57 100Ω RCOM R62 100Ω BX15 R52 100Ω Figure 78. Digital Data Port B Input Terminations Rev. PrC | Page 46 of 52 Preliminary Technical Data AD9784 DVDDS Q 5 OPCLK_3 + C52 4.7µF 6.3V C53 0.1µF 6 Q_ CLR 15 DGND;8 74LCX112 DVDDS;16 U7 SDO SDI CLK 10 PRE 11 9 J Q 13 CLK 12 7 K Q_ CLR 14 74LCX112 DGND;8 U7 DVDDS;16 CSB SW2 SW3 SW4 A A A 3 2 2 B 1 2 SPCSB U5 12 1 4 U5 10 3 SPSDI U5 8 5 1 U6 2 13 74AC14 R21 10kΩ R20 10kΩ 3 U5 U5 9 R45 9kΩ U6 11 4 U6 6 B 1 B 1 SPI PORT P1 1 2 3 4 5 6 12 U6 DVDDS 10 74AC14 9 74AC14 U6 74AC14 74AC14 5 R48 9kΩ B 1 3 2 74AC14 74AC14 SPSDO 11 2 74AC14 74AC14 6 13 A 3 74AC14 74AC14 SPCLK U5 R50 9kΩ 3 SW5 U6 8 + C43 4.7µF 6.3V 74AC14 Figure 79. SPI and One-Port Clock Circuitry Rev. PrC | Page 47 of 52 C50 0.1µF + C44 4.7µF 6.3V C51 0.1µF 03152-PrD-086 OPCLK 3 J 1 CLK 2 K 4 PRE Preliminary Technical Data 03152-PrD-087 AD9784 03152-PrD-088 Figure 80. PCB Assembly, Primary Side Figure 81. PCB Assembly, Secondary Side Rev. PrC | Page 48 of 52 AD9784 03152-PrD-089 Preliminary Technical Data 03152-PrD-090 Figure 82. PCB Assembly, Layer 1 Metal Figure 83. PCB Assembly, Layer 6 Metal Rev. PrC | Page 49 of 52 Preliminary Technical Data 03152-PrD-091 AD9784 03152-PrD-092 Figure 84. PCB Assembly, Layer 2 Metal (Ground Plane) Figure 85. PCB Assembly, Layer 3 Metal (Power Plane) Rev. PrC | Page 50 of 52 AD9784 03152-PrD-093 Preliminary Technical Data 03152-PrD-094 Figure 86. PCB Assembly, Layer 4 Metal (Power Plane) Figure 87. PCB Assembly, Layer 5 Metal (Ground Plane) Rev. PrC | Page 51 of 52 AD9784 Preliminary Technical Data OUTLINE DIMENSIONS 14.00 SQ 1.20 MAX 0.75 0.60 0.45 12.00 SQ 80 60 60 SEATING PLANE 80 61 61 1 1 PIN 1 TOP VIEW (PINS DOWN) BOTTOM VIEW 20 41 21 6.00 SQ 20 41 40 40 21 0.15 0.05 1.05 1.00 0.95 7° 3.5° 0° 0.20 0.09 COPLANARITY 0.08 0.50 BSC 0.27 0.22 0.17 GAGE PLANE 0.25 COMPLIANT TO JEDEC STANDARDS MS-026-ADD-HD Figure 88. 80-Lead Thermally Enhanced TQFP (SV-80) Dimensions shown in millimeters) ESD CAUTION ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily accumulate on the human body and test equipment and can discharge without detection. Although this product features proprietary ESD protection circuitry, permanent damage may occur on devices subjected to high energy electrostatic discharges. Therefore, proper ESD precautions are recommended to avoid performance degradation or loss of functionality. © 2004 Analog Devices, Inc. All rights reserved. Trademarks and registered trademarks are the property of their respective owners. PR03151–0–3/04(PrC) Rev. PrC | Page 52 of 52