a FEATURES Member of Pin-Compatible TxDAC Product Family 125 MSPS Update Rate 14-Bit Resolution Excellent SFDR and IMD Differential Current Outputs: 2 mA to 20 mA Power Dissipation: 190 mW @ 5 V to 45 mW @ 3 V Power-Down Mode: 25 mW @ 5 V On-Chip 1.20 V Reference Single +5 V or +3 V Supply Operation Packages: 28-Lead SOIC and TSSOP Edge-Triggered Latches APPLICATIONS Communication Transmit Channel: Basestations ADSL/HFC Modems Instrumentation PRODUCT DESCRIPTION The AD9764 is the 14-bit resolution member of the TxDAC series of high performance, low power CMOS digital-to-analog converters (DACs). The TxDAC family, which consists of pin compatible 8-, 10-, 12-, and 14-bit DACs, is specifically optimized for the transmit signal path of communication systems. All of the devices share the same interface options, small outline package and pinout, providing an upward or downward component selection path based on performance, resolution and cost. The AD9764 offers exceptional ac and dc performance while supporting update rates up to 125 MSPS. The AD9764's flexible single-supply operating range of 2.7 V to 5.5 V and low power dissipation are well suited for portable and low power applications. Its power dissipation can be further reduced to a mere 45 mW with a slight degradation in performance by lowering the full-scale current output. Also, a power-down mode reduces the standby power dissipation to approximately 25 mW. The AD9764 is manufactured on an advanced CMOS process. A segmented current source architecture is combined with a proprietary switching technique to reduce spurious components and enhance dynamic performance. Edge-triggered input latches and a 1.2 V temperature compensated bandgap reference have been integrated to provide a complete monolithic DAC solution. Flexible supply options support +3 V and +5 V CMOS logic families. The AD9764 is a current-output DAC with a nominal full-scale output current of 20 mA and > 100 k output impedance. TxDAC is a registered trademark of Analog Devices, Inc. REV. C 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 which may result from its use. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. 14-Bit, 125 MSPS TxDAC(R) D/A Converter AD9764 FUNCTIONAL BLOCK DIAGRAM +5V 0.1mF REFLO COMP1 AVDD +1.20V REF 0.1mF CURRENT SOURCE ARRAY FS ADJ RSET +5V DVDD DCOM CLOCK AD9764 50pF REFIO ACOM COMP2 0.1mF IOUTA SEGMENTED SWITCHES CLOCK LSB SWITCHES IOUTB LATCHES SLEEP DIGITAL DATA INPUTS (DB13-DB0) Differential current outputs are provided to support singleended or differential applications. Matching between the two current outputs ensures enhanced dynamic performance in a differential output configuration. The current outputs may be tied directly to an output resistor to provide two complementary, single-ended voltage outputs or fed directly into a transformer. The output voltage compliance range is 1.25 V. The on-chip reference and control amplifier are configured for maximum accuracy and flexibility. The AD9764 can be driven by the on-chip reference or by a variety of external reference voltages. The internal control amplifier, which provides a wide (>10:1) adjustment span, allows the AD9764 full-scale current to be adjusted over a 2 mA to 20 mA range while maintaining excellent dynamic performance. Thus, the AD9764 may operate at reduced power levels or be adjusted over a 20 dB range to provide additional gain ranging capabilities. The AD9764 is available in 28-lead SOIC and TSSOP packages. It is specified for operation over the industrial temperature range. PRODUCT HIGHLIGHTS 1. The AD9764 is a member of the TxDAC product family that provides an upward or downward component selection path based on resolution (8 to 14 bits), performance and cost. 2. Manufactured on a CMOS process, the AD9764 uses a proprietary switching technique that enhances dynamic performance beyond that previously attainable by higher power/cost bipolar or BiCMOS devices. 3. On-chip, edge-triggered input CMOS latches readily interface to +3 V and +5 V CMOS logic families. The AD9764 can support update rates up to 125 MSPS. 4. A flexible single-supply operating range of 2.7 V to 5.5 V, and a wide full-scale current adjustment span of 2 mA to 20 mA, allows the AD9764 to operate at reduced power levels. 5. The current output(s) of the AD9764 can be easily configured for various single-ended or differential circuit topologies. One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel:781/329-4700 World Wide Web Site: http://www.analog.com Fax:781/326-8703 (c) Analog Devices, Inc., 1999-2016 AD9764-SPECIFICATIONS DC SPECIFICATIONS (T MIN to TMAX , AVDD = +5 V, DVDD = +5 V, IOUTFS = 20 mA, unless otherwise noted) Parameter Min RESOLUTION Typ Max 14 Units Bits 1 DC ACCURACY Integral Linearity Error (INL) TA = +25C TMIN to TMAX Differential Nonlinearity (DNL) TA = +25C TMIN to TMAX 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 Small Signal Bandwidth (w/o CCOMP1)4 -4.5 -6.5 2.5 +4.5 +6.5 LSB LSB -2.5 -4.5 1.5 +2.5 +4.5 LSB LSB +0.025 +2 +7 20.0 1.25 % of FSR % of FSR % of FSR mA V k pF 1.32 V nA 1.25 1 1.4 V M MHz 0 50 100 50 ppm of FSR/C ppm of FSR/C ppm of FSR/C ppm/C -0.025 -2 -7 2.0 -1.0 100 5 1.08 1.20 100 0.1 TEMPERATURE COEFFICIENTS Offset Drift Gain Drift (Without Internal Reference) Gain Drift (With Internal Reference) Reference Voltage Drift POWER SUPPLY Supply Voltages AVDD5 DVDD Analog Supply Current (IAVDD ) Digital Supply Current (IDVDD)6 Supply Current Sleep Mode (IAVDD) Power Dissipation6 (5 V, IOUTFS = 20 mA) Power Dissipation7 (5 V, IOUTFS = 20 mA) Power Dissipation7 (3 V, IOUTFS = 2 mA) Power Supply Rejection Ratio8--AVDD Power Supply Rejection Ratio8--DVDD OPERATING RANGE 1 1 2.7 2.7 5.0 5.0 25 1.5 5.0 133 190 45 5.5 5.5 30 4 8.5 170 -0.4 -0.025 +0.4 +0.025 V V mA mA mA mW mW mW % of FSR/V % of FSR/V -40 +85 C NOTES 1 Measured at IOUTA , driving a virtual ground. 2 Nominal full-scale current, I OUTFS, is 32 x the I REF current. 3 Use an external buffer amplifier to drive any external load. 4 Reference bandwidth is a function of external cap at COMP1 pin and signal level. 5 For operation below 3 V, it is recommended that the output current be reduced to 12 mA or less to maintain optimum performance. 6 Measured at fCLOCK = 25 MSPS and fOUT = 1.0 MHz. 7 Measured as unbuffered voltage output with I OUTFS = 20 mA and 50 R LOAD at IOUTA and IOUTB, fCLOCK = 100 MSPS and fOUT = 40 MHz. 8 5% Power supply variation. Specifications subject to change without notice. -2- REV. C AD9764 (TMIN to TMAX , AVDD = +5 V, DVDD = +5 V, IOUTFS = 20 mA, Differential Transformer Coupled Output, DYNAMIC SPECIFICATIONS 50 Doubly Terminated, unless otherwise noted) Parameter Min DYNAMIC PERFORMANCE Maximum Output Update Rate (fCLOCK) Output Settling Time (tST) (to 0.1%)1 Output Propagation Delay (tPD) Glitch Impulse Output Rise Time (10% to 90%)1 Output Fall Time (10% to 90%)1 Output Noise (IOUTFS = 20 mA) Output Noise (IOUTFS = 2 mA) Max 125 75 73 82 dBc dBc dBc dBc dBc dBc dBc dBc dBc 85 77 70 80 77 70 58 78 76 89 dBc dBc dBc dBc 84 84 -78 NOTES 1 Measured single-ended into 50 load. Specifications subject to change without notice. -3- Units MSPS ns ns pV-s ns ns pA/Hz pA/Hz 35 1 5 2.5 2.5 50 30 AC LINEARITY Spurious-Free Dynamic Range to Nyquist fCLOCK = 25 MSPS; fOUT = 1.00 MHz 0 dBFS Output TA = +25C TMIN to TMAX -6 dBFS Output -12 dBFS Output -18 dBFS Output fCLOCK = 50 MSPS; fOUT = 1.00 MHz fCLOCK = 50 MSPS; fOUT = 2.51 MHz fCLOCK = 50 MSPS; fOUT = 5.02 MHz fCLOCK = 50 MSPS; fOUT = 20.2 MHz Spurious-Free Dynamic Range within a Window fCLOCK = 25 MSPS; fOUT = 1.00 MHz; 2 MHz Span TA = +25C TMIN to TMAX fCLOCK = 50 MSPS; fOUT = 5.02 MHz; 2 MHz Span fCLOCK = 100 MSPS; fOUT = 5.04 MHz; 4 MHz Span Total Harmonic Distortion fCLOCK = 25 MSPS; fOUT = 1.00 MHz TA = +25C TMIN to TMAX fCLOCK = 50 MHz; fOUT = 2.00 MHz fCLOCK = 100 MHz; fOUT = 2.00 MHz Multitone Power Ratio (Eight Tones at 110 kHz Spacing) fCLOCK = 20 MSPS; fOUT = 2.00 MHz to 2.99 MHz 0 dBFS Output -6 dBFS Output -12 dBFS Output -18 dBFS Output REV. C Typ -74 -72 -75 -75 dBc dBc dBc dBc 73 76 73 64 dBc dBc dBc dBc AD9764 DIGITAL SPECIFICATIONS Parameter DIGITAL INPUTS Logic "1" Voltage @ DVDD = +5 V Logic "1" Voltage @ DVDD = +3 V Logic "0" Voltage @ DVDD = +5 V Logic "0" Voltage @ DVDD = +3 V Logic "1" Current Logic "0" Current Input Capacitance Input Setup Time (tS) Input Hold Time (tH) Latch Pulsewidth (tLPW) Min Typ 3.5 2.1 5 3 0 0 -10 -10 Max Units V V V V A A pF ns ns ns 1.3 0.9 +10 +10 5 2.0 1.5 3.5 Specifications subject to change without notice. DB0-DB13 tS tH CLOCK tLPW tPD tST IOUTA OR IOUTB 0.1% 0.1% Figure 1. Timing Diagram ORDERING GUIDE ABSOLUTE MAXIMUM RATINGS* Parameter AVDD DVDD ACOM AVDD CLOCK, SLEEP Digital Inputs IOUTA , IOUTB COMP1, COMP2 REFIO, FSADJ REFLO Junction Temperature Storage Temperature Lead Temperature (10 sec) With Respect to Min Max Units ACOM DCOM DCOM DVDD DCOM DCOM ACOM ACOM ACOM ACOM -0.3 -0.3 -0.3 -6.5 -0.3 -0.3 -1.0 -0.3 -0.3 -0.3 +6.5 +6.5 +0.3 +6.5 DVDD + 0.3 DVDD + 0.3 AVDD + 0.3 AVDD + 0.3 AVDD + 0.3 +0.3 +150 +150 V V V V V V V V V V C C +300 C -65 *Stresses above those listed under Absolute Maximum Ratings may cause permanent damage to the device. This is a stress rating only; functional operation of the device at these or any other conditions above those indicated in the operational sections of this specification is not implied. Exposure to absolute maximum ratings for extended periods may affect device reliability. 1 2 RW = Small Outline IC, RU = TSSOP. Z = RoHS Compliant Part. THERMAL CHARACTERISTICS Thermal Resistance 28-Lead 300 mil SOIC JA = 71.4 C/W JC = 23 C/W 28-Lead TSSOP JA = 97.9 C/W JC = 14.0 C/W 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 the AD9764 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. -4- WARNING! ESD SENSITIVE DEVICE REV. C AD9764 PIN CONFIGURATION (MSB) DB13 1 28 CLOCK DB12 2 27 DVDD DB11 3 26 DCOM DB10 4 25 NC DB9 5 AD9764 24 AVDD DB8 6 TOP VIEW 23 COMP2 DB7 7 (Not to Scale) 22 IOUTA DB6 8 21 IOUTB DB5 9 20 ACOM DB4 10 19 COMP1 DB3 11 18 FS ADJ DB2 12 17 REFIO DB1 13 16 REFLO DB0 14 15 SLEEP NC = NO CONNECT PIN FUNCTION DESCRIPTIONS Pin No. Name Description 1 2-13 14 15 DB13 DB12-DB1 DB0 SLEEP 16 17 REFLO REFIO 18 19 20 21 22 23 24 25 26 27 28 FS ADJ COMP1 ACOM IOUTB IOUTA COMP2 AVDD NC DCOM DVDD CLOCK Most Significant Data Bit (MSB). Data Bits 1-12. Least Significant Data Bit (LSB). Power-Down Control Input. Active High. Contains active pull-down circuit; it may be left unterminated if not used. Reference Ground when Internal 1.2 V Reference Used. Connect to AVDD to disable internal reference. Reference Input/Output. Serves as reference input when internal reference disabled (i.e., Tie REFLO to AVDD). Serves as 1.2 V reference output when internal reference activated (i.e., Tie REFLO to ACOM). Requires 0.1 F capacitor to ACOM when internal reference activated. Full-Scale Current Output Adjust. Bandwidth/Noise Reduction Node. Add 0.1 F to AVDD for optimum performance. Analog Common. Complementary DAC Current Output. Full-scale current when all data bits are 0s. DAC Current Output. Full-scale current when all data bits are 1s. Internal Bias Node for Switch Driver Circuitry. Decouple to ACOM with 0.1 F capacitor. Analog Supply Voltage (+2.7 V to +5.5 V). No Internal Connection. Digital Common. Digital Supply Voltage (+2.7 V to +5.5 V). Clock Input. Data latched on positive edge of clock. REV. C -5- AD9764 DEFINITIONS OF SPECIFICATIONS Linearity Error (Also Called Integral Nonlinearity or INL) Power Supply Rejection The maximum change in the full-scale output as the supplies are varied over a specified range. 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. Settling Time 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. Differential Nonlinearity (or DNL) DNL is the measure of the variation in analog value, normalized to full scale, associated with a 1 LSB change in digital input code. Glitch Impulse 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. 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. Spurious-Free Dynamic Range The difference, in dB, between the rms amplitude of the output signal and the peak spurious signal over the specified bandwidth. Gain Error Total Harmonic Distortion 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. THD is the ratio of the sum of the rms value of the first six harmonic components to the rms value of the measured output signal. It is expressed as a percentage or in decibels (dB). Output Compliance Range Multitone Power Ratio 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. The spurious-free dynamic range for an output containing multiple carrier tones of equal amplitude. It is measured as the difference between the rms amplitude of a carrier tone to the peak spurious signal in the region of a removed tone. Temperature Drift Temperature drift is specified as the maximum change from the ambient (+25C) 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. +5V 0.1mF REFLO +1.20V REF 0.1mF REFIO +5V RETIMED CLOCK OUTPUT* LECROY 9210 PULSE GENERATOR 50pF COMP2 0.1mF MINI-CIRCUITS T1-1T IOUTA SEGMENTED SWITCHES FOR DB13-DB5 CLOCK 50V ACOM AD9764 DVDD DCOM DVDD DCOM AVDD PMOS CURRENT SOURCE ARRAY FS ADJ RSET 2kV COMP1 LSB SWITCHES IOUTB 100V TO HP3589A SPECTRUM/ NETWORK ANALYZER 50V INPUT LATCHES 50V SLEEP 50V CLOCK OUTPUT DIGITAL DATA 20pF 20pF * AWG2021 CLOCK RETIMED SUCH THAT DIGITAL DATA TRANSITIONS ON FALLING EDGE OF 50% DUTY CYCLE CLOCK. TEKTRONIX AWG-2021 Figure 2. Basic AC Characterization Test Setup -6- REV. C AD9764 Typical AC Characterization Curves (AVDD = +5 V, DVDD = +3 V, IOUTFS = 20 mA, 50 Doubly Terminated Load, Differential Output, TA = +25C, SFDR up to Nyquist, unless otherwise noted) 90 85 85 5 MSPS 80 25 MSPS SFDR - dBc 100 MSPS 60 75 -12dBFS 70 65 60 70 65 55 55 50 50 50 45 45 45 40 0.1 40 100 90 85 -6dBFS 2.5 0 SFDR - dBc 0dBFS Figure 5. SFDR vs. fOUT @ 25 MSPS 85 -6dBFS 60 50 50 45 45 40 -12dBFS 65 55 0dBFS 10mA @ 5V 20mA @ 5V 5mA @ 5V 25 Figure 6. SFDR vs. fOUT @ 50 MSPS 70 65 20mA @ 3V 5mA @ 3V 10mA @ 3V 55 0 10 20 30 40 FREQUENCY - MHz 50 0.0 50 Figure 7. SFDR vs. fOUT @100 MSPS 90 75 60 40 10 15 20 FREQUENCY - MHz 12 10 90 70 55 5 4 6 8 FREQUENCY - MHz 75 65 0 2 80 70 60 1.0 1.5 2.0 FREQUENCY - MHz 80 -12dBFS 75 0.5 Figure 4. SFDR vs. fOUT @ 5 MSPS 85 80 40 0 90 0dBFS 60 55 10 1 FREQUENCY - MHz -12dBFS 80 SFDR - dBc SFDR - dBc 65 Figure 3. SFDR vs. f OUT @ 0 dBFS SFDR - dBc 0dBFS 75 50 MSPS 70 -6dBFS 85 80 75 90 -6dBFS SFDR - dBc 90 2.0 4.0 6.0 8.0 FREQUENCY - MHz 10.0 Figure 8. SFDR vs. f OUT and IOUTFS @ 25 MSPS and 0 dBFS 90 90 455kHz @ 5 MSPS 80 9.09MHz @ 100 MSPS SFDR - dBc 80 70 60 3.38/3.63MHz @ 25 MSPS 1MHz @ 5 MSPS 2.27MHz @ 25 MSPS SFDR - dBc SFDR - dBc 80 5MHz @ 25 MSPS 70 0.675/0.725MHz @ 5 MSPS 70 60 60 13.5/14.5MHz @ 100 MSPS 20MHz @ 100 MSPS 4.55MHz @ 50 MSPS 10MHz @ 50 MSPS 50 -30 -25 -20 -15 -10 AOUT - dBFS -5 0 Figure 9. Single-Tone SFDR vs. AOUT @ fOUT = fCLOCK/11 REV. C 50 -30 -25 -20 -15 -10 AOUT - dBFS 6.75/7.25 @ 50 MSPS -5 0 Figure 10. Single-Tone SFDR vs. AOUT @ fOUT = fCLOCK/5 -7- 50 -30 -25 -20 -15 -10 AOUT - dBFS -5 0 Figure 11. Dual-Tone SFDR vs. AOUT @ fOUT = fCLOCK/7 AD9764 -70 80 85 80 2ND HARMONIC SNR - dB dBc -80 3RD HARMONIC -85 4TH HARMONIC -90 -95 000.0E+0 40.0E+6 80.0E+6 IOUTFS = 10mA, DVDD = +3V IOUTFS = 10mA, DVDD = +5V 70 Figure 13. SNR vs. fCLOCK @ fOUT = 2.0 MHz 2.0 1.5 70 IDIFF @ 0dBFS IA @ 0dBFS 50 1 10 OUTPUT FREQUENCY - MHz 2.0 80 1.5 75 -0.5 1.0 SFDR - dBc ERROR - LSB ERROR - LSB 0.0 0.5 0.0 100 Figure 14. Differential vs. SingleEnded SFDR vs. fOUT @ 50 MSPS 2.5MHz 1.0 0.5 IA @ -6dBFS 60 65 IOUTFS = 5mA, DVDD = +3V IOUTFS = 5mA, DVDD = +5V 60 0 10 20 30 40 50 60 70 80 90 100 fCLOCK - MSPS 120.0E+6 Figure 12. THD vs. f CLOCK @ fOUT = 2 MHz 75 SFDR - dBc -75 IDIFF @ -6dBFS IOUTFS = 20mA, DVDD = +3V IOUTFS = 20mA, DVDD = +5V 70 65 10MHz 60 -1.0 -0.5 55 -1.0 50 -40 -1.5 40MHz 4000 8000 CODE 12000 0 16000 Figure 15. Typical INL 8000 CODE 12000 16000 Figure 16. Typical DNL 0 -20 -20 -30 10dB - Div -40 -50 -40 -50 -60 -60 -70 -70 -80 -80 -90 000.0E+0 -90 0E+0 7.5E+6 15.0E+6 22.5E+6 Figure 18. Single-Tone SFDR 0 20 40 60 TEMPERATURE - 8C 80 0 fCLK = 50MSPS fOUT1 = 6.75MHz fOUT2 = 7.25MHz SFDR = 69dBc AMPLITUDE = 0dBFS -10 -30 -20 Figure 17. SFDR vs. Temperature @ 100 MSPS, 0 dBFS 0 fCLK = 50MSPS fOUT = 1.25MHz SFDR = 78dBc AMPLITUDE = 0dBFS -10 10dB - Div 4000 fCLK = 50MSPS fOUT1 = 6.25MHz fOUT2 = 6.75MHz fOUT3 = 7.25MHz fOUT4 = 7.75MHz SFDR = 66dBc AMPLITUDE = 0dBFS -10 -20 -30 10dB - Div -2.0 0 -40 -50 -60 -70 -80 -90 5E+6 10E+6 15E+6 20E+6 25E+6 Figure 19. Dual-Tone SFDR -8- -100 000.0E+0 7.5E+6 15.0E+6 22.5E+6 Figure 20. Four-Tone SFDR REV. C AD9764 +5V 0.1mF REFLO +1.20V REF VREFIO 0.1mF RSET 2kV COMP1 FS ADJ +5V DVDD DCOM CLOCK CLOCK ACOM AD9764 50pF REFIO IREF AVDD PMOS CURRENT SOURCE ARRAY COMP2 0.1mF VDIFF = VOUTA - VOUTB IOUTA LSB SWITCHES SEGMENTED SWITCHES FOR DB13-DB5 IOUTB IOUTA IOUTB VOUTA VOUTB RLOAD 50V LATCHES SLEEP RLOAD 50V DIGITAL DATA INPUTS (DB13-DB0) Figure 21. Functional Block Diagram IOUTB = (16383 - DAC CODE)/16384 x IOUTFS FUNCTIONAL DESCRIPTION Figure 21 shows a simplified block diagram of the AD9764. The AD9764 consists of a large PMOS current source array that is capable of providing up to 20 mA of total current. The array is divided into 31 equal currents that make up the five most significant bits (MSBs). The next four bits or middle bits consist of 15 equal current sources whose value is 1/16th of an MSB current source. The remaining LSBs are binary weighted fractions of the middle bits current sources. Implementing the middle and lower bits with current sources, instead of an R-2R ladder, enhances its dynamic performance for multitone or low amplitude signals and helps maintain the DAC's high output impedance (i.e., >100 k). All of these current sources are switched to one or the other of the two output nodes (i.e., IOUTA or IOUTB) via PMOS differential current switches. The switches are based on a new architecture that drastically improves distortion performance. This new switch architecture reduces various timing errors and provides matching complementary drive signals to the inputs of the differential current switches. As mentioned previously, IOUTFS is a function of the reference current IREF, which is nominally set by a reference voltage VREFIO and external resistor RSET. It can be expressed as: IOUTFS = 32 x IREF where IREF = VREFIO /R SET (3) (4) The two current outputs will typically drive a resistive load directly or via a transformer. If dc coupling is required, IOUTA and IOUTB should be directly connected to matching resistive loads, RLOAD, that are tied to analog common, ACOM. Note that RLOAD may represent the equivalent load resistance seen by IOUTA or IOUTB as would be the case in a doubly terminated 50 or 75 cable. The single-ended voltage output appearing at the IOUTA and IOUTB nodes is simply: VOUTA = IOUTA x RLOAD (5) VOUTB = IOUTB x R LOAD (6) The differential voltage, VDIFF , appearing across IOUTA and IOUTB is: VDIFF = (IOUTA - IOUTB) x R LOAD (7) Substituting the values of IOUTA , IOUTB and IREF; VDIFF can be expressed as: VDIFF = {(2 DAC CODE - 16383)/16384} x VDIFF = {(32 R LOAD/R SET) x VREFIO The full-scale output current is regulated by the reference control amplifier and can be set from 2 mA to 20 mA via an external resistor, RSET. The external resistor, in combination with both the reference control amplifier and voltage reference VREFIO, sets the reference current I REF, which is mirrored over to the segmented current sources with the proper scaling factor. The full-scale current, IOUTFS, is 32 times the value of IREF. (8) These last two equations highlight some of the advantages of operating the AD9764 differentially. First, the differential operation will help cancel common-mode error sources associated with IOUTA and I OUTB such as noise, distortion and dc offsets. Second, the differential code-dependent current and subsequent voltage, VDIFF , is twice the value of the single-ended voltage output (i.e., VOUTA or VOUTB), thus providing twice the signal power to the load. DAC TRANSFER FUNCTION The AD9764 provides complementary current outputs, IOUTA and IOUTB. IOUTA will provide a near full-scale current output, IOUTFS, when all bits are high (i.e., DAC CODE = 16383) while IOUTB, the complementary output, provides no current. The current output appearing at IOUTA and IOUTB is a function of both the input code and IOUTFS and can be expressed as: REV. C where DAC CODE = 0 to 16383 (i.e., Decimal Representation). Note that the full-scale value of VOUTA and VOUTB should not exceed the specified output compliance range to maintain specified distortion and linearity performance. The analog and digital sections of the AD9764 have separate power supply inputs (i.e., AVDD and DVDD) that can operate independently over a 2.7 volt to 5.5 volt range. The digital section, which is capable of operating up to a 125 MSPS clock rate, consists of edge-triggered latches and segment decoding logic circuitry. The analog section includes the PMOS current sources, the associated differential switches, a 1.20 V bandgap voltage reference and a reference control amplifier. IOUTA = (DAC CODE/16384) x IOUTFS (2) (1) -9- Note that the gain drift temperature performance for a singleended (VOUTA and VOUTB) or differential output (VDIFF) of the AD9764 can be enhanced by selecting temperature tracking resistors for RLOAD and RSET due to their ratiometric relationship as shown in Equation 8. AD9764 provides several application benefits. The first benefit relates directly to the power dissipation of the AD9764, which is proportional to IOUTFS (refer to the Power Dissipation section). The second benefit relates to the 20 dB adjustment, which is useful for system gain control purposes. REFERENCE OPERATION The AD9764 contains an internal 1.20 V bandgap reference that can be easily disabled and overridden by an external reference. REFIO serves as either an input or output, depending on whether the internal or external reference is selected. If REFLO is tied to ACOM, as shown in Figure 22, the internal reference is activated, and REFIO provides a 1.20 V output. In this case, the internal reference must be compensated externally with a ceramic chip capacitor of 0.1 F or greater from REFIO to REFLO. Also, REFIO should be buffered with an external amplifier having an input bias current less than 100 nA if any additional loading is required. The small signal bandwidth of the reference control amplifier is approximately 1.4 MHz and can be reduced by connecting an external capacitor between COMP1 and AVDD. The output of the control amplifier, COMP1, is internally compensated via a 50 pF capacitor that limits the control amplifier small-signal bandwidth and reduces its output impedance. Any additional external capacitance further limits the bandwidth and acts as a filter to reduce the noise contribution from the reference amplifier. Figure 24 shows the relationship between the external capacitor and the small signal -3 dB bandwidth of the reference amplifier. Since the -3 dB bandwidth corresponds to the dominant pole, and hence the time constant, the settling time of the control amplifier to a stepped reference input response can be approximated. +5V 0.1mF OPTIONAL EXTERNAL REF BUFFER REFLO +1.2V REF COMP1 50pF REFIO ADDITIONAL LOAD 0.1mF 2kV AVDD CURRENT SOURCE ARRAY FS ADJ AD9764 1000 Figure 22. Internal Reference Configuration BANDWIDTH - kHz The internal reference can be disabled by connecting REFLO to AVDD. In this case, an external reference may then be applied to REFIO as shown in Figure 23. The external reference may provide either a fixed reference voltage to enhance accuracy and drift performance or a varying reference voltage for gain control. Note that the 0.1 F compensation capacitor is not required since the internal reference is disabled, and the high input impedance (i.e., 1 M) of REFIO minimizes any loading of the external reference. 0.1 0.1 AVDD 0.1mF REFLO AVDD COMP1 EXTERNAL REF FS ADJ RSET IREF = VREFIO/RSET AD9764 10 100 COMP1 CAPACITOR - nF 1000 The optimum distortion performance for any reconstructed waveform is obtained with a 0.1 F external capacitor installed. Thus, if IREF is fixed for an application, a 0.1 F ceramic chip capacitor is recommended. Also, since the control amplifier is optimized for low power operation, multiplying applications requiring large signal swings should consider using an external control amplifier to enhance the application's overall large signal multiplying bandwidth and/or distortion performance. AVDD 50pF REFIO 1 Figure 24. External COMP1 Capacitor vs. -3 dB Bandwidth +1.2V REF VREFIO 10 CURRENT SOURCE ARRAY REFERENCE CONTROL AMPLIFIER Figure 23. External Reference Configuration REFERENCE CONTROL AMPLIFIER The AD9764 also contains an internal control amplifier that is used to regulate the DAC's full-scale output current, IOUTFS. The control amplifier is configured as a V-I converter, as shown in Figure 23, such that its current output, IREF, is determined by the ratio of the VREFIO and an external resistor, RSET, as stated in Equation 4. IREF is copied over to the segmented current sources with the proper scaling factor to set IOUTFS as stated in Equation 3. There are two methods in which IREF can be varied for a fixed RSET. The first method is suitable for a single-supply system in which the internal reference is disabled, and the common-mode voltage of REFIO is varied over its compliance range of 1.25 V to 0.10 V. REFIO can be driven by a single-supply amplifier or DAC, thus allowing IREF to be varied for a fixed RSET. Since the input impedance of REFIO is approximately 1 M, a simple, low cost R-2R ladder DAC configured in the voltage mode topology may be used to control the gain. This circuit is shown in Figure 25 using the AD7524 and an external 1.2 V reference, the AD1580. The control amplifier allows a wide (10:1) adjustment span of IOUTFS over a 2 mA to 20 mA range by setting IREF between 62.5 A and 625 A. The wide adjustment span of IOUTFS -10- REV. C AD9764 AVDD OPTIONAL BANDLIMITING CAPACITOR AVDD REFLO RFB 1.2V 0.1V TO 1.2V VREF OUT2 AD1580 REFIO CURRENT SOURCE ARRAY FS ADJ AGND AVDD 50pF OUT1 AD7524 COMP1 +1.2V REF VDD RSET IREF = VREF/RSET AD9764 DB7-DB0 Figure 25. Single-Supply Gain Control Circuit The second method may be used in a dual-supply system in which the common-mode voltage of REFIO is fixed, and IREF is varied by an external voltage, VGC, applied to RSET via an amplifier. An example of this method is shown in Figure 26 in which the internal reference is used to set the common-mode voltage of the control amplifier to 1.20 V. The external voltage, VGC, is referenced to ACOM and should not exceed 1.2 V. The value of RSET is such that IREFMAX and IREFMIN do not exceed 62.5 A and 625 A, respectively. The associated equations in Figure 26 can be used to determine the value of RSET. OPTIONAL BANDLIMITING CAPACITOR REFLO COMP1 AVDD 50pF CURRENT SOURCE ARRAY FS ADJ 1mF RSET IREF AD9764 IREF = (1.2-VGC)/RSET WITH VGC < VREFIO AND 62.5mA # IREF # 625A VGC The AD9764 produces two complementary current outputs, IOUTA and I OUTB, which may be configured for single-end or differential operation. IOUTA and I OUTB can be converted into complementary single-ended voltage outputs, VOUTA and VOUTB, via a load resistor, R LOAD, as described in the DAC Transfer Function section by Equations 5 through 8. The differential voltage, VDIFF, existing between VOUTA and V OUTB can also be converted to a single-ended voltage via a transformer or differential amplifier configuration. Figure 28 shows the equivalent analog output circuit of the AD9764 consisting of a parallel combination of PMOS differential current switches associated with each segmented current source. The output impedance of IOUTA and IOUTB is determined by the equivalent parallel combination of the PMOS switches and is typically 100 k in parallel with 5 pF. Due to the nature of a PMOS device, the output impedance is also slightly dependent on the output voltage (i.e., VOUTA and VOUTB) and, to a lesser extent, the analog supply voltage, AVDD, and full-scale current, IOUTFS. Although the output impedance's signal dependency can be a source of dc nonlinearity and ac linearity (i.e., distortion), its effects can be limited if certain precautions are noted. AVDD +1.2V REF REFIO ANALOG OUTPUTS Figure 26. Dual-Supply Gain Control Circuit In some applications, the user may elect to use an external control amplifier to enhance the multiplying bandwidth, distortion performance and/or settling time. External amplifiers capable of driving a 50 pF load such as the AD817 are suitable for this purpose. It is configured in such a way that it is in parallel with the weaker internal reference amplifier as shown in Figure 27. In this case, the external amplifier simply overdrives the weaker reference control amplifier. Also, since the internal control amplifier has a limited current output, it will sustain no damage if overdriven. EXTERNAL CONTROL AMPLIFIER AD9764 AVDD AVDD IOUTA IOUTB RLOAD RLOAD VREF INPUT REFIO FS ADJ RSET 50pF CURRENT SOURCE ARRAY AD9764 Figure 27. Configuring an External Reference Control Amplifier REV. C Figure 28. Equivalent Analog Output Circuit COMP1 AVDD REFLO +1.2V REF IOUTA and IOUTB also have a negative and positive voltage compliance range. The negative output compliance range of -1.0 V is set by the breakdown limits of the CMOS process. Operation beyond this maximum limit may result in a breakdown of the output stage and affect the reliability of the AD9764. The positive output compliance range is slightly dependent on the fullscale output current, IOUTFS. It degrades slightly from its nominal -11- AD9764 1.25 V for an IOUTFS = 20 mA to 1.00 V for an IOUTFS = 2 mA. Operation beyond the positive compliance range will induce clipping of the output signal which severely degrades the AD9764's linearity and distortion performance. In summary, the AD9764 achieves the optimum distortion and noise performance under the following conditions: (1) Differential Operation. (2) Positive voltage swing at IOUTA and IOUTB limited to +0.5 V. For applications requiring the optimum dc linearity, IOUTA and/ or IOUTB should be maintained at a virtual ground via an I-V op amp configuration. Maintaining IOUTA and/or IOUTB at a virtual ground keeps the output impedance of the AD9764 fixed, significantly reducing its effect on linearity. However, it does not necessarily lead to the optimum distortion performance due to limitations of the I-V op amp. Note that the INL/DNL specifications for the AD9764 are measured in this manner using IOUTA. In addition, these dc linearity specifications remain virtually unaffected over the specified power supply range of 2.7 V to 5.5 V. (3) IOUTFS set to 20 mA. (4) Analog Supply (AVDD) set at 5.0 V. (5) Digital Supply (DVDD) set at 3.0 V to 3.3 V with appropriate logic levels. Note that the ac performance of the AD9764 is characterized under the above mentioned operating conditions. DIGITAL INPUTS Operating the AD9764 with reduced voltage output swings at IOUTA and IOUTB in a differential or single-ended output configuration reduces the signal dependency of its output impedance thus enhancing distortion performance. Although the voltage compliance range of IOUTA and I OUTB extends from -1.0 V to +1.25 V, optimum distortion performance is achieved when the maximum full-scale signal at IOUTA and IOUTB does not exceed approximately 0.5 V. A properly selected transformer with a grounded center-tap will allow the AD9764 to provide the required power and voltage levels to different loads while maintaining reduced voltage swings at IOUTA and IOUTB. DC-coupled applications requiring a differential or single-ended output configuration should size RLOAD accordingly. Refer to Applying the AD9764 section for examples of various output configurations. The most significant improvement in the AD9764's distortion and noise performance is realized using a differential output configuration. The common-mode error sources of both IOUTA and IOUTB can be substantially reduced by the common-mode rejection of a transformer or differential amplifier. These common-mode error sources include even-order distortion products and noise. The enhancement in distortion performance becomes more significant as the reconstructed waveform's frequency content increases and/or its amplitude decreases. This is evident in Figure 14, which compares the differential vs. single-ended performance of the AD9764 at 50 MSPS for 0.0 and -6.0 dBFS single tone waveforms over frequency. The distortion and noise performance of the AD9764 is also slightly dependent on the analog and digital supply as well as the full-scale current setting, IOUTFS. Operating the analog supply at 5.0 V ensures maximum headroom for its internal PMOS current sources and differential switches leading to improved distortion performance as shown in Figure 8. Although IOUTFS can be set between 2 mA and 20 mA, selecting an IOUTFS of 20 mA will provide the best distortion and noise performance also shown in Figure 8. The noise performance of the AD9764 is affected by the digital supply (DVDD), output frequency, and increases with increasing clock rate as shown in Figure 13. Operating the AD9764 with low voltage logic levels between 3 V and 3.3 V will slightly reduce the amount of on-chip digital noise. The AD9764's digital input consists of 14 data input pins and a clock input pin. The 14-bit parallel data inputs follow standard positive binary coding where DB13 is the most significant bit (MSB), and DB0 is the least significant bit (LSB). IOUTA produces a full-scale output current when all data bits are at Logic 1. IOUTB produces a complementary output with the full-scale current split between the two outputs as a function of the input code. The digital interface is implemented using an edge-triggered master slave latch. The DAC output is updated following the rising edge of the clock as shown in Figure 1 and is designed to support a clock rate as high as 125 MSPS. The clock can be operated at any duty cycle that meets the specified latch pulsewidth. The setup and hold times can also be varied within the clock cycle as long as the specified minimum times are met, although the location of these transition edges may affect digital feedthrough and distortion performance. Best performance is typically achieved when the input data transitions on the falling edge of a 50% duty cycle clock. The digital inputs are CMOS-compatible with logic thresholds, VTHRESHOLD, set to approximately half the digital positive supply (DVDD) or VTHRESHOLD = DVDD/2 ( 20%) The internal digital circuitry of the AD9764 is capable of operating over a digital supply range of 2.7 V to 5.5 V. As a result, the digital inputs can also accommodate TTL levels when DVDD is set to accommodate the maximum high level voltage of the TTL drivers VOH(MAX). A DVDD of 3 V to 3.3 V will typically ensure proper compatibility with most TTL logic families. Figure 29 shows the equivalent digital input circuit for the data and clock inputs. The sleep mode input is similar with the exception that it contains an active pull-down circuit, thus ensuring that the AD9764 remains enabled if this input is left disconnected. DVDD DIGITAL INPUT Figure 29. Equivalent Digital Input -12- REV. C AD9764 Since the AD9764 is capable of being updated up to 125 MSPS, the quality of the clock and data input signals are important in achieving the optimum performance. Operating the AD9764 with reduced logic swings and a corresponding digital supply (DVDD) will result in the lowest data feedthrough and on-chip digital noise. The drivers of the digital data interface circuitry should be specified to meet the minimum setup and hold times of the AD9764 as well as its required min/max input logic level thresholds. Digital signal paths should be kept short and run lengths matched to avoid propagation delay mismatch. The insertion of a low value resistor network (i.e., 20 to 100 ) between the AD9764 digital inputs and driver outputs may be helpful in reducing any overshooting and ringing at the digital inputs that contribute to data feedthrough. For longer run lengths and high data update rates, strip line techniques with proper termination resistors should be considered to maintain "clean" digital inputs. 30 25 IAVDD - mA 20 15 10 5 0 2 4 6 8 10 12 IOUTFS - mA 14 16 18 20 Figure 30. IAVDD vs. IOUTFS Conversely, IDVDD is dependent on both the digital input waveform, fCLOCK, and digital supply DVDD. Figures 31 and 32 show IDVDD as a function of full-scale sine wave output ratios (f OUT/fCLOCK) for various update rates with DVDD = 5 V and DVDD = 3 V, respectively. Note, how IDVDD is reduced by more than a factor of 2 when DVDD is reduced from 5 V to 3 V. The external clock driver circuitry should provide the AD9764 with a low jitter clock input meeting the min/max logic levels while providing fast edges. Fast clock edges will help minimize any jitter that will manifest itself as phase noise on a reconstructed waveform. Thus, the clock input should be driven by the fastest logic family suitable for the application. 18 125MSPS 16 14 100MSPS 12 IDVDD - mA Note, that the clock input could also be driven via a sine wave, which is centered around the digital threshold (i.e., DVDD/2) and meets the min/max logic threshold. This will typically result in a slight degradation in the phase noise, which becomes more noticeable at higher sampling rates and output frequencies. Also, at higher sampling rates, the 20% tolerance of the digital logic threshold should be considered since it will affect the effective clock duty cycle and, subsequently, cut into the required data setup and hold times. 10 8 50MSPS 6 4 SLEEP MODE OPERATION 25MSPS 2 The AD9764 has a power-down function that turns off the output current and reduces the supply current to less than 8.5 mA over the specified supply range of 2.7 V to 5.5 V and temperature range. This mode can be activated by applying a logic level "1" to the SLEEP pin. This digital input also contains an active pull-down circuit that ensures the AD9764 remains enabled if this input is left disconnected. The SLEEP input with active pull-down requires <40 A of drive current. 5MSPS 0 0.01 0.1 RATIO - fOUT/fCLK 1 Figure 31. IDVDD vs. Ratio @ DVDD = 5 V 8 125MSPS 6 IDVDD - mA The power-up and power-down characteristics of the AD9764 are dependent upon the value of the compensation capacitor connected to COMP1. With a nominal value of 0.1 F, the AD9764 takes less than 5 s to power down and approximately 3.25 ms to power back up. Note, the SLEEP MODE should not be used when the external control amplifier is used as shown in Figure 27. 100MSPS 4 50MSPS 2 25MSPS POWER DISSIPATION The power dissipation, PD, of the AD9764 is dependent on several factors, including: (1) AVDD and DVDD, the power supply voltages; (2) IOUTFS, the full-scale current output; (3) fCLOCK, the update rate; and (4) the reconstructed digital input waveform. The power dissipation is directly proportional to the analog supply current, IAVDD , and the digital supply current, IDVDD. IAVDD is directly proportional to IOUTFS, as shown in Figure 30, and is insensitive to fCLOCK. REV. C 0 0.01 5MSPS 0.1 RATIO - fOUT/fCLK Figure 32. IDVDD vs. Ratio @ DVDD = 3 V -13- 1 AD9764 APPLYING THE AD9764 DIFFERENTIAL USING AN OP AMP OUTPUT CONFIGURATIONS An op amp can also be used to perform a differential-to-singleended conversion as shown in Figure 34. The AD9764 is configured with two equal load resistors, RLOAD , of 25 . The differential voltage developed across IOUTA and IOUTB is converted to a single-ended signal via the differential op amp configuration. An optional capacitor can be installed across IOUTA and IOUTB, forming a real pole in a low-pass filter. The addition of this capacitor also enhances the op amp's distortion performance by preventing the DAC's high slewing output from overloading the op amp's input. The following sections illustrate some typical output configurations for the AD9764. Unless otherwise noted, it is assumed that IOUTFS is set to a nominal 20 mA. For applications requiring the optimum dynamic performance, a differential output configuration is suggested. A differential output configuration may consist of either an RF transformer or a differential op amp configuration. The transformer configuration provides the optimum high frequency performance and is recommended for any application allowing for ac coupling. The differential op amp configuration is suitable for applications requiring dc coupling, a bipolar output, signal gain and/or level shifting. A single-ended output is suitable for applications requiring a unipolar voltage output. A positive unipolar output voltage will result if IOUTA and/or IOUTB is connected to an appropriately sized load resistor, RLOAD, referred to ACOM. This configuration may be more suitable for a single-supply system requiring a dc coupled, ground referred output voltage. Alternatively, an amplifier could be configured as an I-V converter, thus converting IOUTA or IOUTB into a negative unipolar voltage. This configuration provides the best dc linearity since IOUTA or IOUTB is maintained at a virtual ground. Note, IOUTA provides slightly better performance than IOUTB. 225V IOUTA 22 AD8047 225V IOUTB 21 COPT 500V 25V 25V Figure 34. DC Differential Coupling Using an Op Amp DIFFERENTIAL COUPLING USING A TRANSFORMER An RF transformer can be used to perform a differential-tosingle-ended signal conversion as shown in Figure 33. A differentially coupled transformer output provides the optimum distortion performance for output signals whose spectral content lies within the transformer's passband. An RF transformer such as the Mini-Circuits T1-1T provides excellent rejection of common-mode distortion (i.e., even-order harmonics) and noise over a wide frequency range. It also provides electrical isolation and the ability to deliver twice the power to the load. Transformers with different impedance ratios may also be used for impedance matching purposes. Note that the transformer provides ac coupling only. IOUTA 22 500V AD9764 The common-mode rejection of this configuration is typically determined by the resistor matching. In this circuit, the differential op amp circuit using the AD8047 is configured to provide some additional signal gain. The op amp must operate from a dual supply since its output is approximately 1.0 V. A high speed amplifier capable of preserving the differential performance of the AD9764 while meeting other system level objectives (i.e., cost, power) should be selected. The op amps differential gain, its gain setting resistor values and full-scale output swing capabilities should all be considered when optimizing this circuit. The differential circuit shown in Figure 35 provides the necessary level-shifting required in a single supply system. In this case, AVDD, which is the positive analog supply for both the AD9764 and the op amp, is also used to level-shift the differential output of the AD9764 to midsupply (i.e., AVDD/2). The AD8041 is a suitable op amp for this application. MINI-CIRCUITS T1-1T 500V AD9764 AD9764 RLOAD 225V IOUTA 22 IOUTB 21 AD8041 225V IOUTB 21 OPTIONAL RDIFF COPT 1kV AVDD Figure 33. Differential Output Using a Transformer 25V The center tap on the primary side of the transformer must be connected to ACOM to provide the necessary dc current path for both IOUTA and I OUTB. The complementary voltages appearing at I OUTA and IOUTB (i.e., VOUTA and VOUTB) swing symmetrically around ACOM and should be maintained with the specified output compliance range of the AD9764. A differential resistor, RDIFF, may be inserted in applications in which the output of the transformer is connected to the load, RLOAD, via a passive reconstruction filter or cable. RDIFF is determined by the transformer's impedance ratio and provides the proper source termination that results in a low VSWR. Note that approximately half the signal power will be dissipated across RDIFF. 25V 1kV Figure 35. Single-Supply DC Differential Coupled Circuit SINGLE-ENDED UNBUFFERED VOLTAGE OUTPUT Figure 36 shows the AD9764 configured to provide a unipolar output range of approximately 0 V to +0.5 V for a doubly terminated 50 cable since the nominal full-scale current, IOUTFS, of 20 mA flows through the equivalent RLOAD of 25 . In this case, RLOAD represents the equivalent load resistance seen by IOUTA or IOUTB. The unused output (IOUTA or IOUTB) can be connected to ACOM directly or via a matching RLOAD. Different values of -14- REV. C AD9764 IOUTFS and RLOAD can be selected as long as the positive compliance range is adhered to. One additional consideration in this mode is the integral nonlinearity (INL) as discussed in the Analog Output section of this data sheet. For optimum INL performance, the single-ended, buffered voltage output configuration is suggested. AD9764 IOUTFS = 20mA VOUTA = 0 TO +0.5V IOUTA 22 50V 50V system. In general, AVDD, the analog supply, should be decoupled to ACOM, the analog common, as close to the chip as physically possible. Similarly, DVDD, the digital supply, should be decoupled to DCOM as close as physically as possible. For those applications requiring a single +5 V or +3 V supply for both the analog and digital supply, a clean analog supply may be generated using the circuit shown in Figure 38. The circuit consists of a differential LC filter with separate power supply and return lines. Lower noise can be attained using low ESR type electrolytic and tantalum capacitors. IOUTB 21 FERRITE BEADS 25V TTL/CMOS LOGIC CIRCUITS Figure 36. 0 V to +0.5 V Unbuffered Voltage Output AVDD 100mF ELECT. 10-22mF TANT. 0.1mF CER. ACOM SINGLE-ENDED BUFFERED VOLTAGE OUTPUT CONFIGURATION Figure 37 shows a buffered single-ended output configuration in which the op amp U1 performs an I-V conversion on the AD9764 output current. U1 maintains IOUTA (or IOUTB) at a virtual ground, thus minimizing the nonlinear output impedance effect on the DAC's INL performance as discussed in the Analog Output section. Although this single-ended configuration typically provides the best dc linearity performance, its ac distortion performance at higher DAC update rates may be limited by U1's slewing capabilities. U1 provides a negative unipolar output voltage and its full-scale output voltage is simply the product of RFB and I OUTFS. The full-scale output should be set within U1's voltage output swing capabilities by scaling IOUTFS and/or RFB . An improvement in ac distortion performance may result with a reduced IOUTFS since the signal current U1 will be required to sink will be subsequently reduced. COPT RFB 200V AD9764 IOUTFS = 10mA IOUTA 22 U1 VOUT = IOUTFS 3 RFB IOUTB 21 200V Figure 37. Unipolar Buffered Voltage Output POWER AND GROUNDING CONSIDERATIONS In systems seeking to simultaneously achieve high speed and high performance, the implementation and construction of the printed circuit board design is often as important as the circuit design. Proper RF techniques must be used in device selection, placement and routing and supply bypassing and grounding. Figures 42-47 illustrate the recommended printed circuit board ground, power and signal plane layouts that are implemented on the AD9764 evaluation board. Proper grounding and decoupling should be a primary objective in any high speed, high resolution system. The AD9764 features separate analog and digital supply and ground pins to optimize the management of analog and digital ground currents in a REV. C +5V OR +3V POWER SUPPLY Figure 38. Differential LC Filter for Single +5 V or +3 V Applications Maintaining low noise on power supplies and ground is critical to obtain optimum results from the AD9764. If properly implemented, ground planes can perform a host of functions on high speed circuit boards: bypassing, shielding current transport, etc. In mixed signal design, the analog and digital portions of the board should be distinct from each other, with the analog ground plane confined to the areas covering the analog signal traces, and the digital ground plane confined to areas covering the digital interconnects. All analog ground pins of the DAC, reference and other analog components should be tied directly to the analog ground plane. The two ground planes should be connected by a path 1/8 to 1/4 inch wide underneath or within 1/2 inch of the DAC to maintain optimum performance. Care should be taken to ensure that the ground plane is uninterrupted over crucial signal paths. On the digital side, this includes the digital input lines running to the DAC as well as any clock signals. On the analog side, this includes the DAC output signal, reference signal and the supply feeders. The use of wide runs or planes in the routing of power lines is also recommended. This serves the dual role of providing a low series impedance power supply to the part, as well as providing some "free" capacitive decoupling to the appropriate ground plane. It is essential that care be taken in the layout of signal and power ground interconnects to avoid inducing extraneous voltage drops in the signal ground paths. It is recommended that all connections be short, direct and as physically close to the package as possible in order to minimize the sharing of conduction paths between different currents. When runs exceed an inch in length, strip line techniques with proper termination resistors should be considered. The necessity and value of this resistor will be dependent upon the logic family used. For a more detailed discussion of the implementation and construction of high speed, mixed signal printed circuit boards, refer to Analog Devices' application notes AN-280 and AN-333. -15- AD9764 MULTITONE PERFORMANCE CONSIDERATIONS AND CHARACTERIZATION The frequency domain performance of high speed DACs has traditionally been characterized by analyzing the spectral output of a reconstructed full-scale (i.e., 0 dBFS), single-tone sine wave at a particular output frequency and update rate. Although this characterization data is useful, it is often insufficient to reflect a DAC's performance for a reconstructed multitone or spreadspectrum waveform. In fact, evaluating a DAC's spectral performance using a full-scale, single tone at the highest specified frequency (i.e., fH) of a bandlimited waveform is typically indicative of a DAC's "worst-case" performance for that given waveform. In the time domain, this full-scale sine wave represents the lowest peak-to-rms ratio or crest factor (i.e., VPEAK/V rms) that this bandlimited signal will encounter. -10 -20 MAGNITUDE - dBm -30 -40 -50 -60 -70 -80 -90 -100 -110 2.19 2.25 2.31 2.38 2.44 2.50 2.56 2.63 2.69 2.75 2.81 FREQUENCY - MHz Figure 39a. Multitone Spectral Plot 1.0000 0.8000 0.6000 0.4000 VOLTS 0.2000 0.0000 -0.2000 -0.4000 -0.6000 -0.8000 -1.0000 TIME Figure 39b. Time Domain "Snapshot" of the Multitone Waveform However, the inherent nature of a multitone, spread spectrum, or QAM waveform, in which the spectral energy of the waveform is spread over a designated bandwidth, will result in a higher peak-to-rms ratio when compared to the case of a simple sine wave. As the reconstructed waveform's peak-to-average ratio increases, an increasing amount of the signal energy is concentrated around the DAC's midscale value. Figure 39a is just one example of a bandlimited multitone vector (i.e., eight tones) centered around one-half the Nyquist bandwidth (i.e., fCLOCK/4). This particular multitone vector, has a peak-to-rms ratio of 13.5 dB compared to a sine waves peak-to-rms ratio of 3 dB. A "snapshot" of this reconstructed multitone vector in the time domain as shown in Figure 39b reveals the higher signal content around the midscale value. As a result, a DAC's "small-scale" dynamic and static linearity becomes increasingly critical in obtaining low intermodulation distortion and maintaining sufficient carrier-to-noise ratios for a given modulation scheme. A DAC's small-scale linearity performance is also an important consideration in applications where additive dynamic range is required for gain control purposes or "predistortion" signal conditioning. For instance, a DAC with sufficient dynamic range can be used to provide additional gain control of its reconstructed signal. In fact, the gain can be controlled in 6 dB increments by simply performing a shift left or right on the DAC's digital input word. Other applications may intentionally predistort a DAC's digital input signal to compensate for nonlinearities associated with the subsequent analog components in the signal chain. For example, the signal compression associated with a power amplifier can be compensated for by predistorting the DAC's digital input with the inverse nonlinear transfer function of the power amplifier. In either case, the DAC's performance at reduced signal levels should be carefully evaluated. A full-scale single tone will induce all of the dynamic and static nonlinearities present in a DAC that contribute to its distortion and hence SFDR performance. Referring to Figure 3, as the frequency of this reconstructed full-scale, single-tone waveform increases, the dynamic nonlinearities of any DAC (i.e., AD9764) tend to dominate thus contributing to the rolloff in its SFDR performance. However, unlike most DACs, which employ an R-2R ladder for the lower bit current segmentation, the AD9764 (as well as other TxDAC members) exhibits an improvement in distortion performance as the amplitude of a single tone is reduced from its full-scale level. This improvement in distortion performance at reduced signal levels is evident if one compares the SFDR performance vs. frequency at different amplitudes (i.e., 0 dBFS, -6 dBFS and -12 dBFS) and sample rates as shown in Figures 4 through 7. Maintaining decent "small-scale" linearity across the full span of a DAC transfer function is also critical in maintaining excellent multitone performance. Although characterizing a DAC's multitone performance tends to be application-specific, much insight into the potential performance of a DAC can also be gained by evaluating the DAC's swept power (i.e., amplitude) performance for single, dual and multitone test vectors at different clock rates and carrier frequencies. The DAC is evaluated at different clock rates when reconstructing a specific waveform whose amplitude is decreased in 3 dB increments from full-scale (i.e., 0 dBFS). For each specific waveform, a graph showing the SFDR (over Nyquist) performance vs. amplitude can be generated at the different tested clock rates as shown in Figures 9-11. Note that the carrier(s)to-clock ratio remains constant in each figure. In each case, an improvement in SFDR performance is seen as the amplitude is reduced from 0 dBFS to approximately -9.0 dBFS. A multitone test vector may consist of several equal amplitude, spaced carriers each representative of a channel within a defined bandwidth as shown in Figure 39a. In many cases, one or more tones are removed so the intermodulation distortion performance -16- REV. C AD9764 of the DAC can be evaluated. Nonlinearities associated with the DAC will create spurious tones of which some may fall back into the "empty" channel thus limiting a channel's carrier-to-noise ratio. Other spurious components falling outside the band of interest may also be important, depending on the system's spectral mask and filtering requirements. AD9764 EVALUATION BOARD General Description This particular test vector was centered around one-half the Nyquist bandwidth (i.e., fCLOCK/4) with a passband of fCLOCK/16. Centering the tones at a much lower region (i.e., fCLOCK/10) would lead to an improvement in performance while centering the tones at a higher region (i.e., fCLOCK/2.5) would result in a degradation in performance. Figure 40a shows the SFDR vs. amplitude at different sample rates up to the Nyquist frequency while Figure 40b shows the SFDR vs. amplitude within the passband of the test vector. In assessing a DAC's multitone performance, it is also recommended that several units be tested under exactly the same conditions to determine any performance variability. 80 The AD9764-EB is an evaluation board for the AD9764 14-bit DAC converter. Careful attention to layout and circuit design, combined with a prototyping area, allows the user to easily and effectively evaluate the AD9764 in any application where high resolution, high speed conversion is required. This board allows the user the flexibility to operate the AD9764 in various configurations. Possible output configurations include transformer coupled, resistor terminated, inverting/noninverting and differential amplifier outputs. The digital inputs are designed to be driven directly from various word generators with the onboard option to add a resistor network for proper load termination. Provisions are also made to operate the AD9764 with either the internal or external reference or to exercise the powerdown feature. Refer to the application note AN-420, Using the AD9760/AD9764/ AD9764-EB Evaluation Board for a thorough description and operating instructions for the AD9764 evaluation board. 75 70 10 MSPS SFDR - dBc 65 60 20 MSPS 55 50 50 MSPS 45 40 100 MSPS 35 30 -20 -15 -10 AOUT - dBFS -5 0 Figure 40a. Multitone SFDR vs. AOUT (Up to Nyquist) 80 10 MSPS 20 MSPS 75 SFDR - dBc 70 65 60 50 MSPS 55 100 MSPS 50 45 40 -20 -15 -10 AOUT - dBFS -5 0 Figure 40b. Multitone SFDR vs. AOUT (Within Multitone Passband) REV. C -17- 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 -18- A J4 A J3 P1 A R38 49.9V A OUT2 R20 49.9V OUT1 R2 C13 22pF R14 0 C12 22pF A 6 5 4 1 2 3 4 5 6 7 8 9 10 10 9 8 7 6 5 4 3 2 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 1 R1 R6 1 T1 3 C4 10mF TP4 B3 A J7 JP6B C20 0 JP6A A C30 C31 C32 C33 C34 C35 C36 C19 C1 C2 C25 C26 C27 C28 C29 A AGND R13 OPEN A R12 OPEN DVDD 1 2 3 4 5 6 7 8 9 10 AVDD 10 9 8 7 6 5 4 3 2 1 R5 DVDD TP2 TP3 C3 10mF B2 DGND B1 DVDD 16 15 14 13 12 11 10 9 A B JP7B 1 2 3 4 5 6 7 R9 1kV 16 15 14 13 12 11 10 16 PINDIP RES PK 1 2 3 4 5 6 7 8 JP9 A C6 10mF TP7 B6 R4 B A A B R10 1kV A R35 1kV A B JP8 R18 1kV 2 3 U4 7 AVCC R8 AVEE 4 A C23 0.1mF R36 1kV 6 A C21 0.1mF C24 1mF R37 49.9V C22 1mF A J6 DVDD 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 J1 2 3 4 5 6 7 8 9 10 EXTCLK 10 9 8 7 6 5 4 3 2 1 R7 DVDD A AD8047 1 2 3 4 5 6 7 8 9 10 A AVCC 10 9 8 7 6 5 4 3 2 1 R3 C5 10mF TP6 B5 JP7A AVEE 16 PINDIP RES PK TP5 TP18 TP19 B4 A C18 0.1mF TP12 A 28 27 26 25 24 23 22 21 20 19 18 17 16 15 2 CLK JP1 A C16 1mF TP11 U7 A A 4 GND R44 50V VOUT REF43 A 3 2 1 AVDD JP2 VIN A C7 1mF AVDD EXTREFIN J5 2 PDIN J2 A 3 B AVCC R17 49.9V CLOCK DVDD DCOM NC AVDD COMP2 IOUTA IOUTB ACOM COMP1 FS ADJ REFIO REFLO SLEEP CT1 DB13 DB12 DB11 DB10 DB9 DB8 DB7 DB6 DB5 DB4 DB3 DB2 DB1 DB0 AD9764 U1 1 R15 49.9V TP1 C14 1mF 6 A R45 1kV R43 5kV R42 1kV A A JP4 A R46 1kV 2 3 7 A 3 4 A AD8047 U6 JP3 1 AVEE A C15 0.1mF 2 JP5 6 C17 0.1mF AVCC C10 0.1mF TP9 OUT 2 TP8 OUT 1 C9 0.1mF AVDD CW TP14 R16 2kV TP10 AVDD C11 0.1mF C8 0.1mF 1 2 3 A B TP13 AD9764 Figure 41. Evaluation Board Schematic REV. C AD9764 Figure 42. Silkscreen Layer--Top Figure 43. Component Side PCB Layout (Layer 1) REV. C -19- AD9764 Figure 44. Ground Plane PCB Layout (Layer 2) Figure 45. Power Plane PCB Layout (Layer 3) -20- REV. C AD9764 Figure 46. Solder Side PCB Layout (Layer 4) Figure 47. Silkscreen Layer--Bottom REV. C -21- AD9764 OUTLINE DIMENSIONS 18.10 (0.7126) 17.70 (0.6969) 15 28 1 10.65 (0.4193) 10.00 (0.3937) 14 2.65 (0.1043) 2.35 (0.0925) 0.30 (0.0118) 0.10 (0.0039) COPLANARITY 0.10 0.51 (0.0201) 0.31 (0.0122) 1.27 (0.0500) BSC D13783-0-1/16(C) 7.60 (0.2992) 7.40 (0.2913) SEATING PLANE 0.75 (0.0295) 45 0.25 (0.0098) 8 0 1.27 (0.0500) 0.40 (0.0157) 0.33 (0.0130) 0.20 (0.0079) 06-07-2006-A COMPLIANT TO JEDEC STANDARDS MS-013-AE CONTROLLING DIMENSIONS ARE IN MILLIMETERS; INCH DIMENSIONS (IN PARENTHESES) ARE ROUNDED-OFF MILLIMETER EQUIVALENTS FOR REFERENCE ONLY AND ARE NOT APPROPRIATE FOR USE IN DESIGN. 28-Lead Standard Small Outline Package [SOIC_W] Wide Body (RW-28) Dimensions shown in millimeters and (inches) 9.80 9.70 9.60 28 15 4.50 4.40 4.30 1 6.40 BSC 14 PIN 1 0.65 BSC 0.15 0.05 COPLANARITY 0.10 0.30 0.19 1.20 MAX SEATING PLANE 0.20 0.09 8 0 0.75 0.60 0.45 COMPLIANT TO JEDEC STANDARDS MO-153-AE PRINTED IN U.S.A. 28-Lead Thin Shrink Small Outline Package [TSSOP] (RU-28) Dimensions shown in millimeters -22- REV. C