a +3 V, Dual, Serial Input Complete 12-Bit DAC AD8303 FEATURES Complete Dual 12-Bit DAC Pretrimmed Internal Voltage Reference Single +3 V Operation 0.5 mV/Bit with 2.0475 V Full Scale Low Power: 9.6 mW 3-Wire Serial SPI Compatible Interface Power Shutdown IDD < 1 mA Compact SO-14, 1.75 mm Height Package APPLICATIONS Portable Communications Digitally Controlled Calibration Servo Controls PC Peripherals FUNCTIONAL BLOCK DIAGRAM VDD CS En CLK SDI (DATA) S H I F T R E G I S T E R R E D G A I C S T A E D R P D LDA LDB B A N D G A P R R E D G A I C S T B E R P OP AMP A DAC A R E F E R E N C E VOUTA REF BUF VREF REF BUF DAC B R OP AMP B VOUTB AD8303 DGND GENERAL DESCRIPTION RS MSB AGND SHDN The AD8303 is a complete (includes internal reference) dual, 12-bit, voltage output digital-to-analog converter designed to operate from a single +3 volt supply. Built using a CBCMOS process, this monolithic DAC offers the user low cost and easeof-use in single-supply +3 volt systems. Operation is guaranteed over the supply voltage range of +2.7 V to +5.5 V making this device ideal for battery operated applications. A double buffered serial data interface offers high speed, threewire, DSP and SPI microcontroller compatible inputs using data in (SDI), clock (CLK) and load strobe (LDA + LDB) pins. A chip-select (CS) pin simplifies connection of multiple DAC packages by enabling the clock input when active low. Additionally, an RS input sets the output to zero scale or to 1/2 scale based on the level applied to the MSB pin. A power shutdown feature reduces power dissipation to less than 3 µW. The 2.0475 V full-scale voltage output is laser-trimmed to maintain accuracy over the operating temperature range of the device. The binary input data format provides an easy-to-use one-half millivolt-per-bit software programmability. The voltage outputs are capable of sourcing 3 mA. The AD8303 is specified over the extended industrial (–40°C to +85°C) temperature range. AD8303s are available in plastic DIP and low profile 1.75 mm height SO-14 surface mount packages. For single-channel DAC applications, see the AD8300 which is offered in the 8-lead DIP and SO-8 packages. 1.0 2 VDD = +5V VDD = +5V TA = –40°C, +25°C, +85°C 1.5 INL LINEARITY ERROR – LSB 0.8 0.6 DNL – LSB 0.4 0.2 0 –0.2 –0.4 –0.6 +25°C 0.5 0 –0.5 –1 +85°C –1.5 –0.8 –1.0 –40°C 1 0 1024 2048 3072 DIGITAL INPUT CODE – Decimal 4096 Figure 1. Differential Nonlinearity Error vs. Code –2 0 1024 2048 3072 DIGITAL INPUT CODE – Decimal 4096 Figure 2. Linearity Error vs. Digital Code and Temperature REV. 0 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. © Analog Devices, Inc., 1996 One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 617/329-4700 Fax: 617/326-8703 AD8303–SPECIFICATIONS +3 V OPERATION (@ VDD = +2.7 V to +3.6 V, –408C ≤ TA ≤ +858C, unless otherwise noted) Condition Symbol STATIC PERFORMANCE Resolution2 Relative Accuracy2 Differential Nonlinearity2 Differential Nonlinearity2 Zero-Scale Error Full-Scale Voltage3 Full-Scale Tempco3, 4 N INL DNL DNL VZSE VFS TCVFS ANALOG OUTPUTS Output Current Output Resistance to GND Capacitive Load 4 IOUT ROUT CL Data = 800H, ∆VOUT < 3 mV Data = 000H No Oscillation3 30 500 mA Ω pF VREF Load > 1 MΩ 1 V Monotonic, TA = +25°C Monotonic Data = 000H Data = FFF H2 Min Typ1 Parameter 12 –2 –3/4 –1 ± 1/2 ± 1/4 ± 1/2 1.25 2.039 2.0475 16 Max Units +2 +3/4 +1 +4.5 2.056 ±3 Bits LSB LSB LSB mV Volts ppm/°C REFERENCE OUTPUT Output Voltage LOGIC INPUTS Logic Input Low Voltage Logic Input High Voltage Input Leakage Current Input Capacitance4 VIL VIH IIL CIL INTERFACE TIMING SPECIFICATIONS 4, 5 Clock Width High Clock Width Low Load Pulse Width Data Setup Data Hold Reset Pulse Width Load Setup Load Hold Select Deselect tCH tCL tLDW tDS tDH tRS tLD1 tLD2 tCSS tCSH AC CHARACTERISTICS4 Voltage Output Settling Time6 Voltage Output Settling Time6 Shutdown Recovery Time Output Slew Rate DAC Glitch Digital Feedthrough tS tS tDSR SR Q Q To ± 0.1% of Full Scale To ± 1 LSB of Final Value To ± 0.1% of Full Scale Data = 000H to FFFH to 000H SUPPLY CHARACTERISTICS Power Supply Range Shutdown Current Supply Current7 Power Dissipation Power Supply Sensitivity VDD RANGE IDD_SD IDD PDISS PSS DNL < ± 1 LSB SHDN = 0, No Load, VIL = 0 V, TA = +25°C VDD = 3 V, VIL = 0 V, No Load VDD = 3 V, VIL = 0 V, No Load ∆VDD = ± 5% 0.6 2.1 10 10 40 40 40 15 15 40 15 40 40 40 V V µA pF ns ns ns ns ns ns ns ns ns ns µs µs µs V/µs nV/s nV/s 4 14 10 2.0 15 15 2.7 0.02 2 6 0.001 5.5 1 3.2 9.6 0.004 V µA mA mW %/% NOTES 1 Typical readings represent the average value of room temperature operation. 2 1 LSB = 0.5 mV for 0 V to +2.0475 V output range. The first two codes (000 H, 001H) are excluded from the linearity error measurement. 3 Includes internal voltage reference error. 4 These parameters are guaranteed by design and not subject to production testing. 5 All input control signals are specified with t R = tF = 2 ns (10% to 90% of +3 V) and timed from a voltage level of 1.6 V. 6 The settling time specification does not apply for negative going transitions within the last 6 LSBs of ground. 7 See Figure 6 for a plot of incremental supply current consumption as a function of the digital input voltage levels. Specifications subject to change without notice. –2– REV. 0 SPECIFICATIONS +5 V OPERATION (@ V DD AD8303 = +5 V 6 10%, –408C ≤ TA ≤ +858C, unless otherwise noted) Condition Symbol STATIC PERFORMANCE Resolution2 Relative Accuracy2 Differential Nonlinearity2 Differential Nonlinearity2 Zero-Scale Error Full-Scale Voltage3 Full-Scale Tempco3, 4 N INL DNL DNL VZSE VFS TCVFS ANALOG OUTPUTS Output Current Output Resistance to GND Capacitive Load 4 IOUT ROUT CL Data = 800 H, ∆VOUT < 3 mV Data = 000 H No Oscillation 30 500 mA Ω pF REFERENCE OUTPUT Output Voltage VREF Load > 1 MΩ 1 V LOGIC INPUTS Logic Input Low Voltage Logic Input High Voltage Input Leakage Current Input Capacitance4 VIL VIH IIL CIL INTERFACE TIMING SPECIFICATIONS 4, 5 Clock Width High Clock Width Low Load Pulse Width Data Setup Data Hold Reset Pulse Width Load Setup Load Hold Select Deselect tCH tCL tLDW tDS tDH tRS tLD1 tLD2 tCSS tCSH AC CHARACTERISTICS4 Voltage Output Settling Time6 Voltage Output Settling Time6 Shutdown Recovery Time Output Slew Rate DAC Glitch Digital Feedthrough tS tS tSDR SR Q Q To ± 0.1% of Full Scale To ± 1 LSB of Final Value 5 To ± 0.1% of Full Scale Data = 000H to FFFH to 000H SUPPLY CHARACTERISTICS Power Supply Range Shutdown Supply Current Positive Supply Current7 Power Dissipation Power Supply Sensitivity VDD RANGE IDD_SD IDD PDISS PSS DNL < ± 1 LSB SHDN = 0, No Load, VIL = 0 V, TA = +25°C VDD = 5 V, VIL = 0 V, No Load VDD = 5 V, VIL = 0 V, No Load ∆VDD = ± 10% Monotonic, TA = +25°C Monotonic Data = 000 H Data = FFF H Min Typ1 Parameter 12 –2 –3/4 –1 ± 1/2 ± 1/4 ± 1/2 1.25 2.039 2.0475 16 Units +2 +3/4 +1 +4.5 2.056 Bits LSB LSB LSB mV Volts ppm/°C ±3 0.8 2.4 10 10 30 30 30 15 15 30 15 30 30 30 –3– µs µs µs V/µs nV s nV s 4 12 10 2 15 15 2.7 V V µA pF ns ns ns ns ns ns ns ns ns ns 3.0 0.02 2.1 10.5 0.001 NOTES 1 Typical readings represent the average value of room temperature operation. 2 1 LSB = 0.5 mV for 0 V to +2.0475 V output range. The first two codes (000 H, 001H) are excluded from the linearity error measurement. 3 Includes internal voltage reference error. 4 These parameters are guaranteed by design and not subject to production testing. 5 All input control signals are specified with t R = tF = 2 ns (10% to 90% of +5 V) and timed from a voltage level of 1.6 V. 6 The settling time specification does not apply for negative going transitions within the last 6 LSBs of ground. 7 See Figure 6 for a plot of incremental supply current consumption as a function of the digital input voltage levels. Specifications subject to change without notice. REV. 0 Max 5.5 1 3.4 17 0.004 V µA mA mW %/% AD8303 ABSOLUTE MAXIMUM RATINGS* ORDERING GUIDE VDD to GND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . –0.3 V, +8 V Logic Inputs to GND . . . . . . . . . . . . . . . . . . . . . . –0.3 V, +8 V VOUT to GND . . . . . . . . . . . . . . . . . . . . . . –0.3 V, VDD + 0.3 V IOUT Short Circuit to GND . . . . . . . . . . . . . . . . . . . . . . 50 mA Package Power Dissipation . . . . . . . . . . . . . . . (TJ MAX –TA)/θJA Thermal Resistance θJA 14-Pin Plastic DIP Package (N-14) . . . . . . . . . . . . 103°C/W 14-Lead SOIC Package (R-14) . . . . . . . . . . . . . . . . 158°C/W Maximum Junction Temperature (TJ MAX) . . . . . . . . . . . 150°C Operating Temperature Range . . . . . . . . . . . . –40°C to +85°C Storage Temperature Range . . . . . . . . . . . . . –65°C to +150°C Lead Temperature (Soldering, 10 secs) . . . . . . . . . . . . .+300°C Temperature DNL Range Model Package Description Package Option AD8303AN ± 0.75 –40°C to +85°C 14-Pin P-DIP N-14 AD8303AR ± 0.75 –40°C to +85°C 14-Lead SOIC R-14 The AD8303 contains 700 transistors. The die size measures 70 mil × 99 mil. *Stress above those listed under “Absolute Maximum Ratings” may cause permanent damage to the device. This is a stress rating only and 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 rating conditions for extended periods may affect device reliability. SDI D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 CLK tCSS tCSH CS tLD2 LDA, B tLD1 SDI tDS CLK tDH tCL tCH tLDW LDA, B tRS RS tS FS ±1 LSB ERROR BAND VOUT ZS tS a. SHDN tSDR IDD b. Figure 3. Timing Diagrams 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 AD8303 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. 0 AD8303 Table I. Control-Logic Truth Table CS CLK RS MSB SHDN LDA/B Serial Shift Register Function DAC Register Function H L L L ↑+ H H X X X X X X L H ↑+ L X X X X X X X X X X X X X X H H L X X Latched Latched Latched Latched Latched Updated with Current Shift Register Contents Transparent Loaded with 800H Latched with 800H Loaded with All Zeros Latched All Zeros No Effect H H H H H H H L ↑+ L ↑+ X H H H H H H H H H H H L H H H H H ↓– L X H X H X No Effect No Effect No Effect Shift-Register-Data Advanced One Bit No Effect No Effect No Effect No Effect No Effect No Effect No Effect No Effect NOTES 1↑+ positive logic transition; ↓– negative logic transition; X Don’t Care. 2Do not clock in serial data while LDA or LDB is LOW. PIN DESCRIPTIONS Pin No. Name Function 1 2 AGND VOUTA 3 VREF 4 5 DGND CS 6 7 8 CLK SDI LDA 9 10 RS LDB 11 MSB 12 13 14 SHDN VDD VOUTB Analog Ground. DAC voltage output, 2.0475 V full scale with 0.5 mV per bit. An internal temperature stabilized reference maintains a fixed full-scale voltage independent of time, temperature and power supply variations. Reference Voltage Output Terminal. Very high output resistance must be buffered if used as a virtual ground. Digital Ground Chip Select, Active Low Input. Disables shift register loading when high. Does not effect LDA or LDB operation. Clock Input, positive edge clocks data into shift register. Serial Data Input, input data loads directly into the shift register. Load DAC register strobes, active low. Transfers shift register data to DAC A register. Asynchronous active low input. See Control Logic Truth Table for operation. Resets DAC register to zero condition or half-scale depending on MSB pin. Asynchronous active low input. Load DAC register strobes, active low. Transfers shift register data to DAC B register. Asynchronous active low input. See Control Logic Truth Table for operation. Digital Input: Logic High presets DAC registers to half-scale 800 H (sets MSB bit to one) when the RS pin is strobed; Logic Low clears all DAC registers to zero (000H ) when the RS pin is strobed. Active low shutdown control input. Does not affect register contents as long as power is present on VDD . Positive power supply input. Specified range of operation +2.7 V to +5.5 V DAC voltage output, 2.0475 V full scale with 0.5 mV per bit. An internal temperature stabilized reference maintains a fixed full-scale voltage independent of time, temperature and power supply variations. PIN CONFIGURATION 14-Pin P-DIP (N-14) 14-Lead SOIC (R-14) AGND 1 14 VOUTB VOUTA 2 VREF 3 13 VDD AD8303 12 SHDN TOP VIEW 11 MSB (Not to Scale) 10 LDB CS 5 DGND 4 9 RS CLK 6 8 LDA SDI 7 REV. 0 –5– AD8303–Typical Performance Characteristics 7 120 VDD = +5V BROADBAND NOISE – 200µV/DIV OUTPUT CURRENT – mA 80 40 DATA = 800H RL TIED TO +1.024V 0 –40 NEGATIVE CURRENT LIMIT –80 –120 0 100 90 10 0% 5 4 VDD = +5V 3 VDD = +3V 2 1 0 TIME = 100µs/DIV 2 1 OUTPUT VOLTAGE – Volts TA = +25°C DATA = 000H 6 TA = +25°C NBW = 635kHz SUPPLY CURRENT – mA POSITIVE CURRENT LIMIT Figure 5. Broadband Noise Figure 4. IOUT vs. VOUT 0 1 2 3 4 LOGIC VOLTAGE – Volts 5 Figure 6. Supply Current vs. Logic Input Voltage POWER SUPPLY REJECTION – dB 75 VDD = +5V ± 10% 50mV 1V 5µs 60 100 90 TA = +25°C DATA = 800H 45 100 90 VOUT VOUT VDD = +3V ± 10% 30 LD LD 15 10 10 0% 0% 5V 0 10 100 1k 10k 100k FREQUENCY – Hz 5V CODE 800H TO 7FFH 1M Figure 7. Power Supply Rejection vs. Frequency 200ns Figure 8. Midscale Transition Performance Figure 9. Large Signal Settling Time 120 10mV 100 2.5 TUE = ∑ (INL+ZS+FS) SS = 200 UNITS VDD = +2.7V NO LOAD SS = 200 UNITS NORMALIZED TO +25°C 2.0 FREQUENCY 90 VOUT VOUT DRIFT – mV 100 80 60 40 1.5 1.0 VDD = +5.5V VDD = +2.7V 0.5 10 0% CLK 20 2V 0 1µs 0 Figure 10. Clock Feedthrough vs. Time –5 –3 –1 1 3 5 7 9 11 13 15 TOTAL UNADJUSTED ERROR – LSB Figure 11. Total Unadjusted Error Histogram –6– –0.5 –55 –35 –15 5 25 45 65 85 105 125 TEMPERATURE – °C Figure 12. Full-Scale Voltage Drift vs. Temperature REV. 0 AD8303 0.5 VDD = +4.5V 0.0 –0.5 –1.0 –55 –35 –15 5 25 45 65 85 105 125 TEMPERATURE – °C Figure 13. Zero-Scale Voltage Drift vs. Temperature VDD = +5V DATA = FFFH 1 10 100 1k FREQUENCY – Hz 10k χ 20 15 χ –2σ 10 VDD = +5V SS = 212 UNITS 5 0 100 200 300 400 500 HOURS OF OPERATION AT +150°C IDD SUPPLY CURRENT – mA 25 –0.5 ZERO SCALE (DATA = 000H) –1 100k 100 200 300 400 500 HOURS OF OPERATION AT +150°C 60 VDD = +3.6V, VLOGIC = 2.1V, DATA = FFFH 3 VDD = +2.7V SS = 200 UNITS TA = –40 TO +85°C 40 30 20 2 VDD = +3.0V OR +5.0V, VLOGIC = 0V, DATA = 000H 10 0 Figure 17. Supply Current vs. Temperature 8 16 24 –40 –32 –24 –16 –8 0 TEMPERATURE COEFFICIENT – ppm/ °C Figure 18. Full-Scale Output Tempco Histogram 1000 IDD SHUTDOWN CURRENT – nA 500mV 500mV 100 90 100 90 VOUT 100 VOUT VDD = +5.5V 10 10 SHDN 0% 0% 1µs 5V 10 –55 –35 –15 5 25 45 65 95 TEMPERATURE – °C REV. 0 SHDN 5V 1µs 105 125 Figure 19. Shutdown Current vs. Temperature 600 70 VDD = +5.5V, VLOGIC = 2.4V, DATA = FFFH 50 4 0 Figure 15. Long-Term Drift Accelerated by Burn-In 1 –60 –40 –20 0 20 40 60 80 100 120 140 TEMPERATURE – °C 600 Figure 16. Shutdown Current vs. Time Accelerated by Burn-In 5 FULL SCALE (DATA = FFFH) 0 6 χ +2σ SHUTDOWN CURRENT – nA 0.5 Figure 14. Output Voltage Noise Density vs. Frequency 30 0 1 0.1 1 VDD = +2.7V SS = 212 UNITS 1.5 FREQUENCY VOUT DRIFT – mV VDD = +2.7V 2 10 NOMINAL VOLTAGE CHANGE – mV NO LOAD SS = 200 UNITS NORMALIZED TO +25°C 1.5 1.0 OUTPUT VOLTAGE NOISE DENSITY – µV/Hz 2.0 Figure 20. Shutdown Recovery Time –7– Figure 21. Shutdown Time AD8303 THEORY OF OPERATION OUTPUT SECTION The AD8303 is a complete, ready-to-use, dual, 12-bit digital-toanalog converter. Only one +2.7 V to +5.5 V power supply is necessary for operation. It contains two voltage-switched, 12-bit, laser-trimmed digital-to-analog converters, a curvaturecorrected bandgap reference, rail-to-rail output op amps, input shift register, and two DAC registers. The serial data interface consists of a serial data input (SDI), clock (CLK), chip select (CS) and two DAC load strobe pins (LDA and LDB). The rail-to-rail output stage of this amplifier has been designed to provide precision performance while operating near either power supply. Figure 23 shows an equivalent output schematic of the rail-to-rail amplifier with its N-channel pull-down FETs that will pull an output load directly to GND. The output sourcing current is provided by a P-channel pull-up device that can source current to GND terminated loads. The rail-to-rail output stage permits operation at supply voltages down to +2.7 V. The N-channel output pull-down MOSFET shown in Figure 23 has a 35 Ω ON resistance which sets the sink current capability near ground. In addition to resistive load driving capability, the amplifier has also been carefully designed and characterized for up to 500 pF capacitive load driving capability. For battery operation and similar low power applications, a shutdown feature (SHDN) is available to reduce power supply current to less than 1 µA. In addition an asynchronous reset pin (RS) will set both DAC outputs to either zero volts or to midscale, depending on the logic value applied to the MSB pin. This function is useful for power-on reset or system failure recovery to a known state. VDD P-CH D/A CONVERTER SECTION Each of the two DACs is a 12-bit device with an output that swings from GND potential to 0.4 V generated from the internal bandgap voltage (Figure 22). Each DAC uses a laser-trimmed segmented R-2R ladder that is switched by n-channel MOSFETs. The output voltage of the DAC has a constant resistance independent of digital input code. The DAC output is internally connected to the rail-to-rail output op amp. N-CH AGND Figure 23. Equivalent Analog Output Circuit VREF 1.0V BANDGAP REF 2kΩ 1.0V 12-BIT DAC 0.4V REFERENCE SECTION The internal curvature-corrected bandgap voltage reference is laser trimmed for both initial accuracy and low temperature coefficient. Figure 18 provides a histogram of total output performance of full-scale versus temperature, which is dominated by the reference performance. 0.4V FS VOUT 2.047V FS 10kΩ 2.5kΩ VOUT 10kΩ VREF Output The internal reference drives two resistor-divider networks. One divider provides a 0.4 V reference for the DAC. The second divider is trimmed to 1.0 V and is available at the VREF pin. The VREF output is useful for ratiometric applications, and also for generating a “false ground” or bipolar offset. See Figures 30 and Figure 31 for typical applications. Since VREF has a high output impedance, it must be buffered if it is required to deliver current to an external load. Figure 22. AD8303 Equivalent Schematic of Analog Section AMPLIFIER SECTION The internal DAC’s output is buffered by a low power consumption, precision amplifier. This low power amplifier contains a differential PNP pair input stage that provides low offset voltage and low noise, as well as the ability to amplify the zero-scale DAC output voltages, The rail-to-rail amplifier is configured with a gain of approximately five in order to set the 2.0475 volt full-scale output (0.5 mV/LSB). An equivalent circuit schematic for the amplifier section is shown in Figure 22. The op amp has a 4 µs typical settling time to 0.1% of full scale. There are slight differences in settling time for negative slewing signals versus positive. Also, negative transition settling time to within the last 6 LSBs of zero volts has an extended settling time. See the oscilloscope photos in the typical performances section of this data sheet. –8– REV. 0 AD8303 Whether or not a separate power supply trace is available, however, generous supply bypassing will reduce supply-line induced errors. Local supply bypassing consisting of a 10 µF tantalum electrolytic in parallel with a 0.1 µF ceramic capacitor is recommended in all applications (Figure 25). POWER SUPPLY The very low power consumption of the AD8303 is a direct result of a circuit design optimizing the use of a CBCMOS process. By using the low power characteristics of CMOS for the logic, and the low noise, tight matching of the complementary bipolar transistors, excellent analog accuracy is achieved. +2.7V TO +5.5V One advantage of the rail-to-rail output amplifiers used in the AD8303 is the wide range of usable supply voltage. The part is fully specified and tested for operation from +2.7 V to +5.5 V. If reduced linearity and source current capability near full scale can be tolerated, operation of the AD8303 is possible down to +2.7 V. SDI CLK CS LDA LDB RS POWER SUPPLY BYPASSING AND GROUNDING MSB Precision analog products, such as the AD8303, require a well filtered power source. Since the AD8303 operates from a single +3 V to +5 V supply, it seems convenient to simply tap into the digital logic power supply. Unfortunately, the logic supply is often a switch-mode design, which generates noise in the 20 kHz to 1 MHz range. In addition, fast logic gates can generate glitches hundred of millivolts in amplitude due to wiring resistances and inductances. The power supply noise generated thereby means that special care must be taken to insure that the inherent precision of the DAC is maintained. Good engineering judgment should be exercised when addressing the power supply grounding and bypassing of the AD8303. SHDN 5 8 AD8303 10 14 10µF VOUTA VOUTB 9 11 12 AGND DGND 4 Figure 25. Recommended Supply Bypassing for the AD8303 INPUT LOGIC LEVELS All digital inputs are protected with a Zener-type ESD protection structure (Figure 26) that allows logic input voltages to exceed the VDD supply voltage. This feature can be useful if the user is driving one or more of the digital inputs with a 5 V CMOS logic input voltage level while operating the AD8303 on a +3 V power supply. If this mode of interface is used, make sure that the VOL of the 5 V CMOS meets the VIL input requirement of the AD8303 operating at 3 V. See Figure 6 for a graph for digital logic input threshold versus operating VDD supply voltage. VDD LOGIC IN +5V 10-22µF TANT. GND 0.1µF CER. +5V RETURN Figure 26. Equivalent Digital Input ESD Protection For power consumption-sensitive applications, it is important to note that the internal power consumption of the AD8303 is strongly dependent on the actual logic input voltage levels present in the SDI, CLK, CS, LDA, LDB, SHDN, RS and MSB pins. Since these inputs are standard CMOS logic structures, they contribute static power dissipation which depends on the actual driving logic VOH and VOL voltage levels. Consequently, using CMOS logic versus TTL will provide minimal dissipation in the static state. +5V POWER SUPPLY Figure 24. Use Separate Traces to Reduce Power Supply Noise REV. 0 2 TO ANALOG GROUND FERRITE BEAD: 2 TURNS, FAIR-RITE #2677006301 100µF ELECT. 0.1µF 6 1 The AD8303 should be powered directly from the system power supply. This arrangement, shown in Figure 24, employs an LC filter and separate power and ground connections to isolate the analog section from the logic switching transients. Analog and digital ground pins of the AD8303 should be connected together directly at the IC package. TTL/CMOS LOGIC CIRCUITS 13 VDD 7 –9– AD8303 MSB DAC REGISTER A SDI CLK 12 DAC A VOUTA DAC B VOUTB RESET LOAD D 12-BIT SHIFT REGISTER 12 CLK Q11–Q0 en MSB CS DAC REGISTER B AD8303 12 RESET LOAD MSB RS LDA LDB SHDN Figure 27. AD8303 Digital Section Functional Block Diagram DIGITAL INTERFACE The AD8303 has a double-buffered serial data input. The serial-input register is separate from the two DAC registers, which allows preloading of a new data value into the serial register without disturbing the present DAC values. A functional block diagram of the digital section is shown in Figure 27, while Table I contains the truth table for the control logic inputs. Three pins control the serial data input. Data at the Serial Data Input (SDI) is clocked into the shift register on the rising edge of CLK. Data is entered in MSB-first format. Twelve clock pulses are required to load the 12-bit DAC value. If additional bits are clocked into the shift register, for example when a µC sends two 8-bit bytes, the MSBs are ignored (Figure 28). The CLK pin is only enabled when Chip Select (CS) is low. If only one AD8303 is connected to a serial data bus, then CS can be tied (hardwired) to ground. BYTE 1 BYTE 2 MSB LSB B15 B14 B13 B12 B11 B10 B9 X X X X D11 D10 D9 MSB LSB B8 B7 B6 B5 B4 B3 B2 B1 B0 D8 D7 D6 D5 D4 D3 D2 D1 D0 D11–D0: 12-BIT DAC VALUE X = DON'T CARE THE MSB OF BYTE 1 IS THE FIRST BIT THAT IS LOADED INTO THE DAC Figure 28. Typical AD8303-Microprocessor Serial Data Input Format Separate Load pins (LDA and LDB) are provided to control the flow of data from the shift register to the DAC registers. After the new value is loaded in the serial-input register, it can be asynchronously transferred to either DAC register by strobing the appropriate Load pin (LDA or LDB). The Load pins are level sensitive, so they should be returned high before any new data is loaded into the serial-input register. RESET (RS) AND MSB PINS The RS pin forces both of the DAC registers to a known state, based on the logic level on the MSB pin. If MSB is a logic zero, then forcing RS low will set the DAC latches to all zeros and the DAC output voltage will be zero volts. If MSB is a logic one, then RS will force the DAC latches to one-half scale (800H) and the DAC outputs will be 1.024 V. The half-scale reset is useful for systems where the DAC output is referenced to a “false ground” (see the Generating Bipolar Outputs with a Single Supply section of this data sheet for more information). The reset function is useful for setting the DAC outputs to zero at power-up or after a power supply interruption. Test systems and motor controllers are two of many applications which benefit from powering up to a known state. The reset pulse can be generated by the microprocessor’s power-on RESET signal, by an output from the microprocessor (Figure 33), or by an external resistor and capacitor (Figure 34). RS and MSB have level-sensitive thresholds. The RS input overrides other logic inputs (specifically, LDA and LDB). However, LDA and LDB should be set high before RS goes high. If LDA or LDB are kept low, then the contents of the shift register will be transferred to the DAC register as soon as RS goes high. –10– REV. 0 AD8303 SHUTDOWN (SHDN) The shutdown feature is activated when SHDN is pulled low. While the AD8303 is in shutdown mode, the voltage reference, DACs, and output amplifiers are all turned off. Supply current is less than 1 µA. The DAC output voltage goes to 0 V, pulled to GND by the 12.5 kΩ feedback resistors (Figure 22). If power (i.e., VDD ) is maintained to the AD8303 during shutdown, the value stored in the DAC input latches will not change. When the SHDN pin is driven high, the DACs will return to the same voltages as before shutdown. The CMOS logic section of the AD8303 remains active while SHDN is low. Thus, new data can be loaded while the DACs are shut down and, when SHDN goes high, the DACs will assume the new output voltage. The AD8303 recovers from shutdown very quickly. The voltage output settling time after shutdown is typically only a few microseconds longer than the normal settling time (Figure 20). GENERATING “BIPOLAR” OUTPUTS WITH A SINGLE SUPPLY To maximize output signal swings in single supply operation, many circuit designs employ a “false-ground” configuration. This method defines a voltage, usually at one half of full scale or at one half of the power supply, as the “ground” reference. Signals are then measured differentially from the false ground, which produces a “quasi-bipolar” output swing. The AD8303’s voltage reference output, combined with an op amp, can provide a temperature compensated false-ground reference, as shown in Figure 30. The op amp amplifies the AD8303’s 1.0 V reference by 1.024 to provide an analog common (false ground) at one-half scale (1.024 V). With this method, the DAC output is ± 1.024 V (referenced to the false ground). The “Quasi-Bipolar” code table is given in Table III. +3V 13 VDD +3V TO +5V VOUTA 13 SDI CLK CS LDA LDB RS MSB SHDN 0.1µF 10µF VREF 6 3 100Ω SIGNAL GROUND (FALSE GROUND, +1.024V) OP193 5 8 VOUT = ±1.024V (REFERENCED TO SIGNAL GROUND) +3V AD8303 VDD 7 2 AGND DGND AD8303 10 VOUTA 9 11 VOUTB 14 1 0V ≤ VOUT ≤ 2.0475V 2, 14 2kΩ 0.022µF 500pF 12 R2A 97.6kΩ AGND DGND 1 4 VOUTA, VOUTB R1 2.4kΩ 1µF R2B* 2kΩ 4 *ZERO-SCALE TRIM Figure 29. Unipolar Output Operation Figure 30. A False-Ground Generator UNIPOLAR OUTPUT OPERATION This is the basic mode of operation for the AD8303. As shown in Figure 29, the AD8303 has been designed to drive loads as low as 2 kΩ in parallel with 500 pF. The code table for this operation is shown in Table II. Table III. Quasi-Bipolar Code Table DAC Analog Hexadecimal Decimal Output Common “Bipolar” Number Number In Voltage (False-Ground) Analog in DAC Register DAC Register (V) Voltage (V) Voltage (V) Table II. Unipolar Code Table Hexadecimal Number Decimal Number in DAC Register in DAC Register Analog Output Voltage (V) FFF 801 800 7FF 000 2.0475 1.0245 1.024 1.0235 0 REV. 0 4095 2049 2048 2047 0 FFF 801 800 7FF 000 4095 2049 2048 2047 0 2.0475 1.0245 1.024 1.0235 0 1.024 1.024 1.024 1.024 1.024 +1.2035 0.0005 0 –0.0005 –1.024 Since the AD8303’s reference voltage output limits are typical, a trim potentiometer is included so that the “false-ground” output can be adjusted to exactly 1.024 V. To maintain accuracy, resistors R1 and R2A must be of the same type (preferably metal film) to insure temperature coefficient matching. The circuit includes compensation to allow for a 1 µF bypass capacitor at the false-ground output. The benefit of a large capacitor is that not only does the false ground present a very low dc resistance to the load, but its ac impedance is low as well. –11– AD8303 important to maintain accuracy. Resistor pairs R1-R2 and R3-R4 should be selected to match within 0.01%. In addition, these resistors must be of the same type (preferably metal film) to insure temperature coefficient matching. Mismatching between R1 and R2 causes offset and gain errors while an R3 to R4 mismatch yields gain errors. BIPOLAR OUTPUT OPERATION Although the AD8303 has been designed for single-supply operation, the output can also be configured for bipolar operation. A typical circuit is shown in Figure 31. This circuit uses the AD8303’s internal voltage reference to generate a bipolar offset. Since VREF must source current in this application, one half of an OP293 dual op amp is used as a buffer. The other op amp then amplifies the DAC output voltage to produce a bipolar output swing. The output voltage is coded in offset binary and is given by: GENERATING A NEGATIVE SUPPLY VOLTAGE R4 R2 R2 VO = 0.5 mV × Digital Code × × 1 + – 1.0 V × R3 + R4 R1 R1 where 0.5 mV represents the pretrimmed value for one LSB of the AD8303, Digital Code is the digital code sent to the DAC, and 1.0 V is the AD8303 reference voltage. +3V 13 7 6 5 8 10 9 11 12 R1 10kΩ VDD SDI VREF CLK 3 1/2 OP293 AD8303 CS LDA LDB VOUTA RS MSB VOUTB 2 14 +3V +3V R3 10kΩ R4 19.08kΩ 8 1/2 OP293 2 VOUT = ±2.048V 10µF –3V OPTIONAL FULL-SCALE TRIM SHDN AGND DGND 1 R2 OPTIONAL 20.48kΩ ZERO TRIM Some applications may require a bipolar output configuration, as shown in Figure 31, but only have a single power supply rail available. This is very common in data acquisition systems using microprocessor-based systems. In these systems, +12 V, +15 V, and/or +5 V only are available. Single supply rails are, of course, common in battery-powered systems. Shown in Figure 32 is a method of generating a negative supply using a single IC and two capacitors. The ADM8660 employs a charge pump technique to invert supply voltages as low as 1.5 V. A shutdown feature on the ADM8660 complements the shutdown of the AD8303. Note, however, that the ADM8660 requires about 500 µs to turn on after exiting the shutdown state. V+ CAP+ OSC 5 ADM8660 4 CAP– SHUTDOWN LV FC 6 1 –3V 10µF 7 GND 3 4 SHDN FROM AD8303 1/6 74HC04 Figure 31. Bipolar Output Operation For a ± 2.048 V full scale using the circuit values shown, the transfer function becomes: VO = 1 mV × Digital Code – 2.048 V Note that the full-scale span has increased from 2.048 V to 4.096 V (± 2.048 V). Therefore, although each AD8303 LSB represents 0.5 mV, each output LSB of the bipolar circuit has been scaled to 1 mV. The code table for this circuit is shown in Table IV. Table IV. Bipolar Code Table Hexadecimal Number Decimal Number in DAC Register in DAC Register Analog Output Voltage (V) FFF 801 800 7FF 000 2.047 0.001 0 –0.001 –2.048 4095 2049 2048 2047 0 Figure 32. Generating a Negative Supply Voltage MICROCOMPUTER INTERFACES The AD8303 serial data input provides an easy interface to a variety of single-chip microcomputers (µCs). Many µCs have a built-in serial data capability which can be used for communicating with the DAC. In cases where no serial port is provided, or it is being used for some other purpose (such as an RS-232 communications interface), the AD8303 can easily be addressed in software. Twelve data bits are required to load a value into the AD8303. If more than 12 bits are transmitted before the Chip Select input goes high, the extra (i.e., the most significant) bits are ignored. This feature is valuable because most µCs only transmit data in 8-bit increments. Thus, the µC sends 16 bits to the DAC instead of 12 bits. The AD8303 will only respond to the last 12 bits clocked into the SDI input, however, so the serial data interface is not affected. As with the false-ground generator circuit, resistor matching is –12– REV. 0 AD8303 AD8303-MC68HC11 INTERFACE The circuit illustrated in Figure 33 shows a serial interface between the AD8303 and the MC68HC11 8-bit microprocessor. The MOSI output drives the AD8303’s serial data input, SDI, while SCK drives the clock (CLK). The DAC’s CS, LDA, LDB, MSB and RS inputs are driven by lines PD5 and PC0–PC3, respectively. (PD3) MOSI SDI (PD4) SCK CLK (PD5) SS MC68HC11 CS PC0 LDA PC1 LDB PC2 MSB PC3 RS AD8303 NOTE: ADDITIONAL PINS OMITTED FOR CLARITY Figure 33. AD8303-MC68HC11 Serial Interface To load data into the AD8303, the 68HC11’s CPOL and CPHA bits are set high. This action configures the µC to transfer data on the rising edge of the serial clock. After CS is set low, two bytes of data are sent to the AD8303 using the format shown in Figure 28. Then LDA or LDB are strobed low, transferring the serial-input register contents to the appropriate DAC. The RS and MSB inputs allow the DAC to be reset to either zero volts or half scale at any time. AN 8051 µC INTERFACE A typical interface between the AD8303 and an 8051 µC is shown in Figure 34. This interface also uses the µC’s internal serial port. The serial port is programmed for Mode 0 operation, which functions as a simple 8-bit shift register. The 8051’s Port 3.0 pin functions as the serial data output, while Port 3.1 serves as the serial clock. The LDA and LDB pins are controlled by the 8051’s Port 1.0 and Port 1.1 lines, respectively. 7 (P3.0) RxD 6 (P3.1) TxD 8 P1.0 P1.1 10 SDI CLK LDA AD8303 LDB 80CL51 10k VDD 1µF RS CS 9 5 MSB 11 SHDN 12 + VDD NOTE: ADDITIONAL PINS OMITTED FOR CLARITY The 8051’s serial data transmission is straightforward. When data is written to the serial buffer register (SBUF, at Special Function Register location 99H), the data is automatically converted to serial format and clocked out via Port 3.0 and Port 3.1 After 8 bits have been transmitted, the Transmit Interrupt flag (SCON.1) is set and the next 8 bits can be transmitted. The circuit of Figure 34 demonstrates “hardwiring” many of the AD8303 features which may not have to be changed within a given design. For example, the reset feature is controlled by a resistor and capacitor. This produces a power-on reset pulse without requiring a µC I/O pin. The MSB pin can be hardwired to V DD or ground, depending on whether a reset to 0 V or half scale is required. If the AD8303 is the only device on the serial interface, CS can also be tied to ground. Finally, SHDN can be tied to VDD if the shutdown feature will not be used. Software for the interface of Figure 34 is shown in Figure 35. This routine sends the 12-bit value placed in registers DAC_VAL0 and DAC_VAL1 to the DAC addressed by the two LSBs of DAC_ADDR. The subroutine begins by setting appropriate bits in the Serial Control register to configure the serial port for Mode 0 operation. The MSBs of the DAC value are obtained from memory location DAC_VAL1, adjusted to compensate for the 8051’s serial data format, and moved to the serial buffer register. At this point, serial data transmission begins automatically. When all 8 bits have been sent, the Transmit Interrupt bit is set, and the subroutine then proceeds to send the LSBs of the DAC value, stored at location DAC_VAL0. Next the LDA and LDB bits from DAC_ADDR are logically ANDed with Port1. This action sets the appropriate AD8303 DAC select input low and transfers the DAC value from the serialinput register to the DAC register, causing the DAC output voltage to change. Finally the LDA and LDB inputs are driven high to await the next DAC update. The 8051 sends data out of its shift register LSB first, while the AD8303 requires data MSB first. The subroutine therefore includes a BYTESWAP subroutine to reformat the data. This routine transfers the MSB-first byte at location SHIFTREG to an LSB-first byte at location SENDBYTE. The routine rotates the MSB of the first byte into the carry with a Rotate Left Carry instruction, then rotates the carry into the MSB of the second byte with a Rotate Right Carry instruction. After 8 loops, SENDBYTE contains the data in the proper format. The BYTESWAP routine in Listing C is convenient because the DAC data can be calculated in normal LSB form. Figure 34. AD8303-80CL51 Serial Interface REV. 0 –13– AD8303 ;AD8303.ASM ; ; This subroutine loads an AD8303 shift register with a 12-bit ; DAC value, and transfers the value to DAC A or DAC B. ; The DAC value is stored at location DAC-VAL1 (MSB) and DAC_VAL0 (LSB) ; The DAC address (A or B) is stored at DAC_ADDR, (b0=0 for A, b1=0 for B) ; ; Primary controls $MOD51 $TITLE(AD8303 Interface, Using the Serial Port in Mode 0) ; ; Variable declarations ; PORT1 DATA 90H ;SFR register for port 1 DAC_VAL0 DATA 40H ;LSBs of 12-bit DAC Value DAC_VAL1 DATA 41H ; MSBs of DAC Value DAC_ADDR DATA 42H ;DAC address, format is: ; 1,1,1,1,1,1,LDB,LDA ; Set bit low to select DAC LOOPCOUNT DATA 43H ;Count loops for byte swap SHIFTREG DATA 44H ;Shift reg. for byte swap SENDBYTE DATA 45H ; Destination reg. for SR ; ORG 100H ;arbitrary starting address DO_8303: CLR SCON.7 ;set serial CLR SCON.6 ; data mode 0 CLR SCON.5 ;Clr SM2 for mode 0 CLR SCON.1 ;Clr the transmit flag MOV SHIFTREG,DAC_VAL1 ;Get Most Significant Byte ACALL SEND_IT ; send to AD8303 MOV SHIFTREG,DAC_VAL0 ;Get Least Significant Byte ACALL SEND_IT ; send it to the AD8303 MOV A,PORT1 ;Get I/O port contents ANL A,DAC_ADDR ;Clr LDA/LDB, other bits unchanged MOV PORT1,A ;Send to I/O port ORL A,#00000011B ;Set LDA and LDB high MOV PORT1,A ;Send to I/O port RET ;Done ; ;Convert the byte to LSB-first format and send it to the AD8303 SEND_IT: MOV LOOPCOUNT,#8 ;Shift 8 bits BYTESWAP: MOV A,SHIFTREG ;Get source byte RLC A ;rotate MSB to carry MOV SHIFTREG,A ;Save new source byte MOV A,SENDBYTE ;get destination byte RRC A ;Move carry into MSB MOV SENDBYTE,A ;Save DJNZ LOOPCOUNT,BYTESWAP ;Done? MOV SBUF,SENDBYTE ;Send the byte SEND_WAIT: JNB SCON.1,SEND_WAIT ;Wait until 8 bits are send CLR SCON.1 ;Clear the serial flag RET ;Done END Figure 35. Software Listing for the AD8303-80CL51 Interface –14– REV. 0 AD8303 OUTLINE DIMENSIONS Dimensions shown in inches and (mm). 14-Lead Epoxy DIP (N-14) 0.795 (20.19) 0.725 (18.42) 14 8 1 7 0.280 (7.11) 0.240 (6.10) 0.060 (1.52) 0.015 (0.38) PIN 1 0.210 (5.33) MAX 0.325 (8.25) 0.300 (7.62) 0.195 (4.95) 0.115 (2.93) 0.150 (3.81) MIN 0.200 (5.05) 0.125 (3.18) 0.022 (0.558) 0.014 (0.356) 0.100 0.070 (1.77) (2.54) 0.045 (1.15) BSC 0.015 (0.38) 0.008 (0.20) SEATING PLANE 14-Lead Narrow Body SOIC (R-14) 0.3444 (8.75) 0.3367 (8.55) 0.1574 (4.00) 0.1497 (3.80) 14 8 1 7 PIN 1 0.0098 (0.25) 0.0040 (0.10) SEATING PLANE REV. 0 0.0500 (1.27) BSC 0.2440 (6.20) 0.2284 (5.80) 0.0688 (1.75) 0.0532 (1.35) 0.0192 (0.49) 0.0138 (0.35) –15– 0.0099 (0.25) 0.0075 (0.19) 0.0196 (0.50) x 45° 0.0099 (0.25) 8° 0° 0.0500 (1.27) 0.0160 (0.41) PRINTED IN U.S.A. C2098–18–1/96 AD8303 –16– REV. 0