32-Channel, 16/14, Serial Input, Voltage-Output DACs AD5372/AD5373 Preliminary Technical Data Power-on reset Digital reset (RESET) Clear function to user-defined SIGGND (CLR pin) Simultaneous update of DAC outputs (LDAC pin) FEATURES 32-channel DAC in 56-LFCSP and 64-LQFP AD5372 Guaranteed monotonic to 16 bits AD5373 Guaranteed monotonic to 14 bits Maximum output voltage span of 4 × VREF (20 V) Nominal output voltage range of -4 V to +8 V Multiple, independent output spans available System calibration function allowing user-programmable offset and gain Channel grouping and addressing features Thermal Monitoring Function DSP/microcontroller-compatible serial interface 2.5 V to 5.5 V JEDEC-compliant digital levels APPLICATIONS Level setting in automatic test equipment (ATE) Variable optical attenuators (VOA) Optical switches Industrial control systems Instrumentation FUNCTIONAL BLOCK DIAGRAM DVCC VDD VSS 8 n n n A/B SELECT REGISTER SERIAL INTERFACE SDI n n n SCLK VREF0 14 8 TO MUX 2's n n X1 REGISTER n M REGISTER A/B MUX n C REGISTER ·· · ·· · SYNC LDAC n = 16 FOR AD5372 n = 14 FOR AD5373 n CONTROL REGISTER AGND DNGD n ·· · · · · · · ·· · · n X1 REGISTER n M REGISTER ·· · ·· · A/B MUX n OFS0 REGISTER GROUP 0 BUFFER OFFSET DAC 0 BUFFER X2A REGISTER X2B REGISTER · · · · · · MUX n 2 · · · · · · X2A REGISTER X2B REGISTER MUX n 2 n DAC 0 REGISTER · · · · · · DAC 7 REGISTER OUTPUT BUFFER AND POWER DOWN CONTROL VOUT0 DAC 0 · · · · · · · · · · · · VOUT2 VOUT7 DAC 7 OUTPUT BUFFER AND POWER DOWN CONTROL n BUSY 8 n CLR n STATE MACHINE n n n n 8 TO MUX 2's n A/B MUX n X1 REGISTER n M REGISTER n C REGISTER · ·· ·· · n POWER-ON RESET A/B SELECT REGISTER n · ·· · ·· · ·· · ·· n X1 REGISTER n M REGISTER VOUT4 VOUT5 VOUT6 GROUP 1 14 RESET VOUT3 n C REGISTER SDO · ·· ·· · A/B MUX n OFS1 REGISTER VOUT1 SIGGND0 VREF1 OFFSET DAC 1 BUFFER X2A REGISTER X2B REGISTER · · · · · · MUX n 2 · · · · · · X2A REGISTER X2B REGISTER MUX n 2 n DAC 0 REGISTER · · · · · · DAC 7 REGISTER OUTPUT BUFFER AND POWER DOWN CONTROL VOUT8 · · · · · · · · · · · · VOUT10 DAC 7 OUTPUT BUFFER AND POWER DOWN CONTROL DAC 0 n VOUT9 VOUT11 VOUT12 VOUT13 VOUT14 VOUT15 SIGGND1 n C REGISTER AD5372/ AD5373 VREF1 SUPPLIES GROUP 1 TO 3 GROUP 2 TO GROUP 3 ARE IDENTICAL TO GROUP 1 VOUT16 TO VOUT31 5372-0001B SIGGND2 SIGGND3 Figure 1. AD5372/AD5373—Protected by U.S. Patent No. 5,969,657; other patents pending Rev. PrF Information furnished by Analog Devices is believed to be accurate and reliable. However, no responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other rights of third parties that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective companies. One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. www.analog.com Tel: 781.329.4700 Fax: 781.326.8703 © 2006 Analog Devices, Inc. All rights reserved. AD5372/AD5373 Preliminary Technical Data TABLE OF CONTENTS Specifications......................................................................................4 Clear Function ............................................................................ 16 AC Characteristics.........................................................................5 BUSY and LDAC Functions...................................................... 16 Timing Characteristics .................................................................6 Power-Down Mode .................................................................... 16 Absolute Maximum Ratings.............................................................8 Thermal Monitor Function....................................................... 16 ESD Caution...................................................................................8 Toggle Mode................................................................................ 17 Terminology .................................................................................... 11 Serial Interface ................................................................................ 18 Functional Description .................................................................. 12 SPI Write Mode........................................................................... 18 DAC Architecture—General..................................................... 12 SPI Readback Mode ................................................................... 19 Channel Groups.......................................................................... 12 Register Update Rates ................................................................ 19 A/ B Reigsters And Gain/Offset Adjustment.......................... 13 Channel Addressing And Special Modes ................................ 19 Load DAC.................................................................................... 13 Special Function Mode .............................................................. 20 Offset DACs ................................................................................ 13 Power Supply Decoupling ......................................................... 22 Output Amplifier........................................................................ 14 Power Supply Sequencing ......................................................... 22 Transfer Function ....................................................................... 14 Interfacing Examples...................................................................... 23 Reference Selection .................................................................... 14 Outline Dimensions ....................................................................... 24 Calibration................................................................................... 15 Ordering Guide........................................................................... 24 AD5372 Calibration Example................................................... 15 Reset Function ............................................................................ 16 REVISION HISTORY Pr B1 Modified SPI timing diagrams Added Reference Selection and Calibration text Pr. B2 Added Reset Function text Pr. B3 Added Power Down Mode text Pr. B4 Added Terminology and Power Supply Sequencing sections Pr D Rewrote calibration section Changed SPI read diagram Pr F. Changed LFCSP Vout8 and Vout9 positions Rev. PrF| Page 2 of 25 Preliminary Technical Data AD5372/AD5373 General Description The AD5372 and AD5373 contain 32, 16-bit or 14-bit DACs in a single, 56-lead, LFCSP or 64-lead LQFP package. The AD5372/AD5373 provides buffered voltage outputs with a span 4 times the reference voltage. The gain and offset of each DAC can be independently trimmed to remove errors. For even greater flexibility, the device is divided into 4 groups of 8 DACs. Two offset DACs allow the output range of the groups to be altered. Group 0 can be adjusted by Offset DAC 0, and group 1 to group 3 can be adjusted by Offset DAC 2. The ADAD5372/AD5373 offers guaranteed operation over a wide supply range with VSS from -4.5 V to -16.5 V and VDD from+8 V to +16.5 V. The output amplifier headroom requirement is 1.4 V operating with a load current of 1 mA. The ADAD5372/AD5373 has a high-speed serial interface, which is compatible with SPI®, QSPI™, MICROWIRE™, and DSP interface standards and can handle clock speeds of up to 50 MHz. The DAC outputs are updated on reception of new data into the DAC registers. All the outputs can be updated simultaneously by taking the LDAC input low. Each channel has a programmable gain and an offset adjust register. Each DAC output is gained and buffered on-chip with respect to an external SIGGND input. The DAC outputs can also be switched to SIGGND via the CLR pin. Table 1. High Channel Count Bipolar DACs Model Resolution AD5360BCPZ AD5360BSTZ AD5361BCPZ AD5361BSTZ AD5362BCPZ AD5362BSTZ AD5363BCPZ AD5363BSTZ AD5370BCPZ AD5370BSTZ AD5371BCPZ AD5371BSTZ AD5372BCPZ AD5372BSTZ AD5373BCPZ AD5373BSTZ 16 Bits 16 Bits 14 Bits 14 Bits 16 Bits 16 Bits 14 Bits 14 Bits 16 Bits 16 Bits 14 Bits 14 Bits 16 Bits 16 Bits 14 Bits 14 Bits Nominal Output Span 4 × V REF (20 V) 4 × V REF (20 V) 4 × V REF (20 V) 4 × V REF (20 V) 4 × V REF (20 V) 4 × V REF (20 V) 4 × V REF (20 V) 4 × V REF (20 V) 4 × V REF (12 V) 4 × V REF (12 V) 4 × V REF (12 V) 4 × V REF (12 V) 4 × V REF (12 V) 4 × V REF (12 V) 4 × V REF (12 V) 4 × V REF (12 V) Output Channels 16 16 16 16 8 8 8 8 40 40 40 40 32 32 32 32 Linearity Error (LSB) ±4 ±4 ±1 ±1 ±4 ±4 ±1 ±1 ±4 ±4 ±1 ±1 ±4 ±4 ±1 ±1 Rev. PrF | Page 3 of 25 Package Description 56-Lead LFCSP 52-Lead LQFP 56-Lead LFCSP 52-Lead LQFP 56-Lead LFCSP 52-Lead LQFP 56-Lead LFCSP 52-Lead LQFP 64-Lead LFCSP 64-Lead LQFP 100-Ball CSPBGA 80-Lead LQFP 56-Lead LFCSP 64-Lead LQFP 56-Lead LFCSP 64-Lead LQFP Package Option CP-56 ST-52 CP-56 ST-52 CP-56 ST-52 CP-56 ST-52 CP-64 ST-64 BC-100-2 ST-80 CP-56 ST-64 CP-56 ST-64 AD5372/AD5373 Preliminary Technical Data SPECIFICATIONS DVCC = 2.3 V to 5.5 V; VDD = 8 V to 16.5 V; VSS = −4.5 V to −16.5 V; VREF = 3 V; AGND = DGND = SIGGND = 0 V; RL = Open Circuit; Gain (m), Offset(c) and DAC Offset registers at default value; all specifications TMIN to TMAX, unless otherwise noted. Table 2. Performance Specifications AD53721 B Version AD53731 B Version Unit 16 ±4 ±1 ±20 ±20 100 100 ±35 14 ±1 ±1 ±20 ±20 100 100 ±35 Bits LSB max LSB max mV max mV max µV max µV max mV VOUT Temperature Coefficient 5 5 DC Crosstalk2 0.5 0.5 ppm FSR/°C typ mV max 1 60 2/5 1 60 2/5 MΩ min nA max V min/max Typically 100 MΩ. Per input. Typically ±30 nA. ±2% for specified operation. 55 ±0.5 55 ±0.5 kΩ min V min/max Typically 60 kΩ. VSS + 1.4 VDD − 1.4 5 ±1 2200 1 VSS + 1.4 VDD − `.4 5 ±1 2200 1 V min V max mA max mA max pF max Ω max ILOAD = 1 mA. ILOAD = 1 mA. 1.7 2.0 0.8 ±8 1.7 2.0 0.8 ±8 V min V min V max µA max ±1 10 ±1 10 µA max pF max All other digital input pins. 0.5 DVCC − 0.5 −5 10 0.5 DVCC − 0.5 −5 10 V max V min µA max pF typ Sinking 200 µA. Sourcing 200 µA. SDO only. Parameter ACCURACY Resolution Relative Accuracy Differential Nonlinearity Offset Error Gain Error Offset Error2 Gain Error2 Gain Error of Offset DAC REFERENCE INPUTS (VREF1, VREF2)2 VREF DC Input Impedance VREF Input Current VREF Range SIGGND INPUT (SIGGND0 TO SIGGND4)2 DC Input Impedance Input Range OUTPUT CHARACTERISTICS2 Output Voltage Range Short Circuit Current Load Current Capacitive Load DC Output Impedance DIGITAL INPUTS Input High Voltage Input Low Voltage Input Current (with pull-up/pulldown) Input Current (no pull-up/pull-down) Input Capacitance2 DIGITAL OUTPUTS (SDO) Output Low Voltage Output High Voltage (SDO) High Impedance Leakage Current High Impedance Output Capacitance2 Rev. PrF| Page 4 of 25 Test Conditions/Comments2 Guaranteed monotonic by design over temperature. Before Calibration Before Calibration After Calibration After Calibration Positive or Negative Full Scale. See Offset DACs section for details Includes linearity, offset, and gain drift. Typically 100 µV. Measured channel at mid-scale, fullscale change on any other channel JEDEC compliant. IOVCC = 2.5 V to 3.6 V. IOVCC = 3.6 V to 5.5 V. IOVCC = 2.5 V to 5.5 V. CLR and RESET pin only. Preliminary Technical Data Parameter POWER REQUIREMENTS DVCC VDD VSS Power Supply Sensitivity2 ∆ Full Scale/∆ VDD ∆ Full Scale/∆ VSS ∆ Full Scale/∆ VCC DICC IDD ISS Power Dissipation Power Dissipation Unloaded (P) Junction Temperature3 1 2 3 AD5372/AD5373 AD53721 B Version AD53731 B Version Unit Test Conditions/Comments2 2.3/5.5 8/16.5 −4.5/−16.5 2.3/5.5 8/16.5 −4.5/−16.5 V min/max V min/max V min/max −75 −75 −90 2 14 14 −75 −75 −90 2 14 14 dB typ dB typ dB typ mA max mA max mA max VCC = 5.5 V, VIH = VCC, VIL = GND. Outputs unloaded. Outputs unloaded. 350 130 350 130 mW °C max TJ = TA + PTOTAL × θJ. Temperature range for B Version: −40°C to +85°C. Typical specifications are at 25°C. Guaranteed by design and characterization, not production tested. Where θJ represents the package thermal impedance. AC CHARACTERISTICS DVCC = 2.5 V; VDD = 15 V; VSS = −15 V; VREF = 3 V; AGND = DGND = SIGGND = 0 V; CL = 200pF; RL = 10 kΩ; Gain (m), Offset(c) and DAC Offset registers at default value; all specifications TMIN to TMAX, unless otherwise noted. Table 3. AC Characteristics Parameter DYNAMIC PERFORMANCE1 Output Voltage Settling Time Slew Rate Digital-to-Analog Glitch Energy Glitch Impulse Peak Amplitude Channel-to-Channel Isolation DAC-to-DAC Crosstalk Digital Crosstalk Digital Feedthrough Output Noise Spectral Density @ 10 kHz AD5372/ AD5373 Unit Test Conditions/Comments 20 30 1 20 10 100 40 10 0.1 1 250 µs typ µs max V/µs typ nV-s typ mV max dB typ nV-s typ nV-s typ nV-s typ nV-s typ nV/(Hz)1/2 typ Full-scale change DAC latch contents alternately loaded with all 0s and all 1s. 1 Guaranteed by design and characterization. Not production tested Rev. PrF | Page 5 of 25 VREF(+) = 2 V p-p, 1 kHz. Between DACs in the same group. Between DACs from different groups. Effect of input bus activity on DAC output under test. VREF = 0 V. AD5372/AD5373 Preliminary Technical Data TIMING CHARACTERISTICS DVCC = 2.3 V to 5.5 V; VDD = 8 V to 16.5 V; VSS = −4.5 V to −16.5 V; VREF = 3 V; AGND = DGND = SIGGND = 0 V; RL = Open Circuit; Gain (m), Offset(c) and DAC Offset registers at default value; all specifications TMIN to TMAX, unless otherwise noted. SPI INTERFACE (Figure 4 and Figure 5) Parameter1, 2, 3 t1 t2 t3 t4 t5 t6 t7 t8 t93 t10 Limit at TMIN, TMAX 20 8 8 11 20 10 5 5 42 1.25 Unit ns min ns min ns min ns min ns min ns min ns min ns min ns max µs max Description SCLK Cycle Time. SCLK High Time. SCLK Low Time. SYNC Falling Edge to SCLK Falling Edge Setup Time. Minimum SYNC High Time. 24th SCLK Falling Edge to SYNC Rising Edge. Data Setup Time. Data Hold Time. SYNC Rising Edge to BUSY Falling Edge. t11 t12 t13 t14 500 20 10 3 ns max ns min ns min Single-Channel Update Cycle Time 24th SCLK Falling Edge to LDAC Falling Edge. LDAC Pulse Width Low. BUSY Rising Edge to DAC Output Response Time. t15 t16 0 3 ns min µs max t17 t18 t19 t20 20/30 125 330 400 µs typ/max ns max ns min µs max t21 t225 270 25 ns min ns max µs max BUSY Pulse Width Low (Single-Channel Update.) See Table 7. BUSY Rising Edge to LDAC Falling Edge. LDAC Falling Edge to DAC Output Response Time. DAC Output Settling Time. CLR/RESET Pulse Activation Time. RESET Pulse Width Low. RESET Time Indicated by BUSY Low. Minimum SYNC High Time in Readback Mode. SCLK Rising Edge to SDO Valid. 1 Guaranteed by design and characterization, not production tested. All input signals are specified with tr = tf = 2 ns (10% to 90% of VCC) and timed from a voltage level of 1.2 V. 3 See Figure 4 and Figure 5. 4 This is measured with the load circuit of Figure 2. 5 This is measured with the load circuit of Figure 3. 2 V CC 200µA RL TO OUTPUT PIN 2.2kΩ TO OUTPUT PIN CL IOL V OH (min) - V OL (max) 2 50pF V OL CL 50pF 200µA Figure 2. Load Circuit for BUSY Timing Diagram IOL Figure 3. Load Circuit for SDO Timing Diagram Rev. PrF| Page 6 of 25 Preliminary Technical Data AD5372/AD5373 t1 SCLK 1 24 2 t3 24 t11 t4 SYNC 1 t2 t6 t5 t7 t8 SDI DB0 DB23 t9 t10 BUSY t12 t13 LDAC1 t17 t14 VOUT1 t15 t13 LDAC2 t17 VOUT2 t 16 CLR t18 VOUT t19 RESET VOUT t18 t20 05814-004A BUSY 1LDAC ACTIVE DURING BUSY. 2LDAC ACTIVE AFTER BUSY. Figure 4.SPI Write Timing t22 SCLK 48 24 t21 SYNC SDI DB23 DB0 INPUT WORD SPECIFIES REGISTER TO BE READ NOP CONDITION DB0 SDO DB0 DB23 DB23 DB0 5371-0005D LSB FROM PREVIOUS WRITE Figure 5.SPI Read Timing Rev. PrF | Page 7 of 25 SELECTED REGISTER DATA CLOCKED OUT AD5372/AD5373 Preliminary Technical Data ABSOLUTE MAXIMUM RATINGS TA = 25°C, unless otherwise noted. Transient currents of up to 100 mA do not cause SCR latch-up. Table 4. Absolute Maximum Ratings Parameter VDD to AGND VSS to AGND DVCC to DGND Digital Inputs to DGND Digital Outputs to DGND VREF1, VREF2 to AGND VOUT0–VOUT39 to AGND SIGGND to AGND AGND to DGND Operating Temperature Range (TA) Industrial (B Version) Storage Temperature Range Junction Temperature (TJ max) θJA Thermal Impedance 56-LFCSP 64-LQFP Reflow Soldering Peak Temperature Time at Peak Temperature Rating −0.3 V to +17 V −17 V to +0.3 V −0.3 V to +7 V −0.3 V to VCC + 0.3 V −0.3 V to VCC + 0.3 V −0.3 V to +7 V VSS − 0.3 V to VDD + 0.3 V VSS − 0.3 V to VDD + 0.3 V −0.3 V to +0.3 V Stresses 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 listed in the operational sections of this specification is not implied. Exposure to absolute maximum rating conditions for extended periods may affect device reliability. −40°C to +85°C −65°C to +150°C 130°C 24°C/w 45.5°C/w 230°C 10 s to 40 s ESD CAUTION ESD (electrostatic discharge) sensitive device. Electrostatic charges as high as 4000 V readily accumulate on the human body and test equipment and can discharge without detection. Although this product features proprietary ESD protection circuitry, permanent damage may occur on devices subjected to high energy electrostatic discharges. Therefore, proper ESD precautions are recommended to avoid performance degradation or loss of functionality. Rev. PrF| Page 8 of 25 VOUT26 VOUT25 VOUT24 AGND DGND DVCC SDO SDI SCLK SYNC DVCC DGND VOUT7 VOUT6 VOUT6 DVCC AD5372/AD5373 DGND VOUT7 SYNC SCLK SDO SDI DGND DVCC VOUT24 AGND CLR LDAC VOUT26 VOUT25 Preliminary Technical Data 48 VOUT5 47 VOUT4 VOUT27 3 46 SIGGND0 SIGGND3 4 45 VOUT28 5 44 VOUT3 VOUT2 VOUT29 6 43 PIN 1 IDENTIFIER VOUT30 7 AD5372 AD5373 TOP VIEW (Not to Scale) VOUT31 8 NC 9 NC 10 LDAC CLR RESET BUSY VOUT27 SIGGND3 VOUT28 VOUT29 VOUT30 VOUT31 NC VDD VSS VREF1 VOUT1 42 VOUT0 41 VREF0 VOUT23 39 VOUT22 40 NC 11 38 NC 12 37 NC 13 36 NC 14 35 NC 15 VDD 16 34 33 VOUT21 VOUT20 VSS VDD SIGGND2 VOUT19 VOUT18 VOUT17 VOUT16 VOUT14 VOUT15 VOUT13 VOUT12 VOUT11 SIGGND1 VOUT8 VOUT9 VOUT10 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 VSS PIN 1 INDICATOR AD5372/ AD5373 TOP VIEW (Not to scale) 42 41 40 39 38 37 36 35 34 33 32 31 30 29 VOUT5 VOUT4 SIGGND0 VOUT3 VOUT2 VOUT1 VOUT0 VREF0 VOUT23 VOUT22 VOUT21 VOUT20 VSS VDD NC = NO CONNECT 100605 VREF1 NC NC 1 2 3 4 5 6 7 8 9 10 11 12 13 14 VOUT9 15 VOUT8 16 VOUT10 17 VOUT11 18 SIGGND1 19 VOUT12 20 VOUT13 21 VOUT14 22 VOUT15 23 VOUT16 24 VOUT17 25 VOUT18 26 VOU1T9 27 SIGGND2 28 BUSY 2 56 55 54 53 52 51 50 49 48 47 46 45 44 43 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 RESET 1 5372-0060 Figure 7. 56-Lead LFCSP Pin Configuration Figure 6.64-Lead LQFP Pin Configuration Table 5. Pin Function Descriptions Pin DVCC VSS VDD AGND DGND VREF0 VREF1 VOUT0 to VOUT31 SYNC1 SCLK1 SDI1 SDO1 CLR LDAC RESET Function Logic Power Supply; 2.5 V to 5.5 V. These pins should be decoupled with 0.1 µF ceramic capacitors and 10 µF capacitors. Negative Analog Power Supply; −11.4 V to −16.5 V for specified performance. These pins should be decoupled with 0.1 µF ceramic capacitors and 10 µF capacitors. Positive Analog Power Supply; +11.4 V to +16.5 V for specified performance. These pins should be decoupled with 0.1 µF ceramic capacitors and 10 µF capacitors. Ground for All Analog Circuitry. All AGND pins should be connected to the AGND plane. Ground for All Digital Circuitry. All DGND pins should be connected to the DGND plane. Reference Input for DACs 0 to 7. This reference voltage is referred to AGND. Reference Input for DACs 8 to 31. This reference voltage is referred to AGND. DAC Outputs. Buffered analog outputs for each of the 40 DAC channels. Each analog output is capable of driving an output load of 10 kΩ to ground. Typical output impedance of these amplifiers is 1 Ω. Active Low Input. This is the frame synchronization signal for the serial interface. Serial Clock Input. Data is clocked into the shift register on the falling edge of SCLK. This pin operates at clock speeds up to 50 MHz. Serial Data Input. Data must be valid on the falling edge of SCLK. Serial Data Output. CMOS output. SDO can be used for readback. Data is clocked out on SDO on the rising edge of SCLK and is valid on the falling edge of SCLK. Asynchronous Clear Input (level sensitive, active low). See the Clear Function section for more information Load DAC Logic Input (Active Low).See the BUSY AND LDAC FUNCTIONS section for more information. Asynchronous Digital Reset Input. Rev. PrF | Page 9 of 25 AD5372/AD5373 Pin BUSY SIGGND0 SIGGND1 SIGGND1 SIGGND3 EXPOSED PADDLE Preliminary Technical Data Function Digital Input/Open-Drain Output. BUSY is open-drain when an output. See the BUSY AND LDAC FUNCTIONS section for more information. Reference Ground for DACs 0 to 7. VOUT0 to VOUT7 are referenced to this voltage. Reference Ground for DACs 8 to 15. VOUT7 to VOUT15 are referenced to this voltage. Reference Ground for DACs 16 to 23. VOUT16 to VOUT23 are referenced to this voltage. Reference Ground for DACs 24 and 31. VOUT24 to VOUT31 are referenced to this voltage. The Lead Free Chip Scale Package (LFCSP) has an exposed paddle on the underside. This should be connected to VSS Rev. PrF| Page 10 of 25 Preliminary Technical Data AD5372/AD5373 TERMINOLOGY Relative Accuracy Relative accuracy, or endpoint linearity, is a measure of the maximum deviation from a straight line passing through the endpoints of the DAC transfer function. It is measured after adjusting for zero-scale error and full-scale error and is expressed in least significant bits (LSB). Differential Nonlinearity Differential nonlinearity is the difference between the measured change and the ideal 1 LSB change between any two adjacent codes. A specified differential nonlinearity of 1 LSB maximum ensures monotonicity. Zero-Scale Error Zero-scale error is the error in the DAC output voltage when all 0s are loaded into the DAC register. Zero-scale error is a measure of the difference between VOUT (actual) and VOUT (ideal) expressed in mV. Zero-scale error is mainly due to offsets in the output amplifier. Full-Scale Error Full-scale error is the error in DAC output voltage when all 1s are loaded into the DAC register. Full-scale error is a measure of the difference between VOUT (actual) and VOUT (ideal) expressed in mV. It does not include zero-scale error. Gain Error Gain error is the difference between full-scale error and zero-scale error. It is expressed in mV. Gain Error = Full-Scale Error − Zero-Scale Error VOUT Temperature Coefficient This includes output error contributions from linearity, offset, and gain drift. DC Output Impedance DC output impedance is the effective output source resistance. It is dominated by package lead resistance. DC Crosstalk The DAC outputs are buffered by op amps that share common VDD and VSS power supplies. If the dc load current changes in one channel (due to an update), this can result in a further dc change in one or more channel outputs. This effect is more significant at high load currents and reduces as the load currents are reduced. With high impedance loads, the effect is virtually immeasurable. Multiple VDD and VSS terminals are provided to minimize dc crosstalk. Output Voltage Settling Time The amount of time it takes for the output of a DAC to settle to a specified level for a full-scale input change. Digital-to-Analog Glitch Energy The amount of energy injected into the analog output at the major code transition. It is specified as the area of the glitch in nV-s. It is measured by toggling the DAC register data between 0x1FFF and 0x2000. Channel-to-Channel Isolation Channel-to-channel isolation refers to the proportion of input signal from one DAC’s reference input that appears at the output of another DAC operating from another reference. It is expressed in dB and measured at midscale. DAC-to-DAC Crosstalk DAC-to-DAC crosstalk is the glitch impulse that appears at the output of one converter due to both the digital change and subsequent analog output change at another converter. It is specified in nV-s. Digital Crosstalk The glitch impulse transferred to the output of one converter due to a change in the DAC register code of another converter is defined as the digital crosstalk and is specified in nV-s. Digital Feedthrough When the device is not selected, high frequency logic activity on the device’s digital inputs can be capacitively coupled both across and through the device to show up as noise on the VOUT pins. It can also be coupled along the supply and ground lines. This noise is digital feedthrough. Output Noise Spectral Density Output noise spectral density is a measure of internally generated random noise. Random noise is characterized as a spectral density (voltage per √Hz). It is measured by loading all DACs to midscale and measuring noise at the output. It is measured in nV/(Hz) Rev. PrF | Page 11 of 25 1/2 AD5372/AD5373 Preliminary Technical Data FUNCTIONAL DESCRIPTION DAC ARCHITECTURE—GENERAL The ADAD5372/AD5373 contains 32 DAC channels and 32 output amplifiers in a single package. The architecture of a single DAC channel consists of a 16-bit (AD5372) or 14-bit (AD5373) resistor-string DAC followed by an output buffer amplifier. The resistor-string section is simply a string of resistors, each of value R, from VREF to AGND. This type of architecture guarantees DAC monotonicity. The 16-bit (AD5372) or 14-bit (AD5373) binary digital code loaded to the DAC register determines at which node on the string the voltage is tapped off before being fed into the output amplifier. The output amplifier multiplies the DAC out voltage by 4. The output span is 12 V with a 3 V reference and 20 V with a 5 V reference. CHANNEL GROUPS The 32 DAC channels of the AD5372/AD5373 are arranged into four groups of 8 channels. The eight DACs of Group 0 derive their reference voltage from VREF0. Group 1 to Group 3 derive their reference voltage from VREF1. Each group has its own signal ground pin. Table 6. AD5372(AD5373) Registers Register Name Word Length (Bits) Description X1A (group)(channel) 16(14) Input data register A, one for each DAC channel. X1B (group) (channel) M (group) (channel) C (group) (channel) X2A (group)(channel) 16(14) 16(14) 16(14) 16(14) Input data register B, one for each DAC channel. Gain trim registers, one for each DAC channel. Offset trim registers, one for each DAC channel. X2B (group) (channel) 16(14) DAC (group) (channel) OFS0 OFS1 Control 14 14 3 Output data register A, one for each DAC channel. These registers store the final, calibrated DAC data after gain and offset trimming. They are not readable, nor directly writable. Output data register B, one for each DAC channel. These registers store the final, calibrated DAC data after gain and offset trimming. They are not readable, nor directly writable. Data registers from which the DACs take their final input data. The DAC registers are updated from the X2A or X2B registers. They are not readable, nor directly writable. Offset DAC 0 data register, sets offset for Group 0. Offset DAC 1 data register, sets offset for Groups 1 to 3. Bit 2 = A/B. 0 = global selection of X1A input data registers. 1 = X1B registers. Bit 1 = Enable Temp Shutdown. 0 = disable temp shutdown. 1 = enable. Bit 0 = Soft Power Down. 0 = soft power up. 1 = soft power down. A/B Select 0 8 Each bit in this register determines if a DAC in Group 0 takes its data from register X2A or X2B (0 = X2A, 1 = X2B) A/B Select 1 8 A/B Select 2 8 A/B Select 3 8 Each bit in this register determines if a DAC in Group 1 takes its data from register X2A or X2B (0 = X2A, 1 = X2B) Each bit in this register determines if a DAC in Group 2 takes its data from register X2A or X2B (0 = X2A, 1 = X2B) Each bit in this register determines if a DAC in Group 3 takes its data from register X2A or X2B (0 = X2A, 1 = X2B) Rev. PrF| Page 12 of 25 Preliminary Technical Data AD5372/AD5373 A/ B REIGSTERS AND GAIN/OFFSET ADJUSTMENT LOAD DAC Each DAC channel has seven data registers. The actual DAC data word can be written to either the X1A or X1B input register, depending on the setting of the A/B bit in the Control Register. If the A/B bit is 0, data will be written to the X1A register. If the A/B bit is 1, data will be written to the X1B register. Note that this single bit is a global control and affects every DAC channel in the device. It is not possible to set up the device on a per-channel basis so that some writes are to X1A registers and some writes are to X1B registers. All DACs in the AD5372/AD5373 can be updated simultaneously by taking LDAC low, when each DAC register will be updated from either its X2A or X2B register, depending on the setting of the A/B select registers. The DAC register is not readable, nor directly writable by the user. MUX MUX X1B REGISTER X2B REGISTER DAC REGISTER DAC M REGISTER C REGISTER Figure 8. Data Registers Associated With Each DAC Channel Each DAC channel also has a gain (M) and offset (C) register, which allow trimming out of the gain and offset errors of the entire signal chain. Data from the X1A register is operated on by a digital multiplier and adder controlled by the contents of the M and C registers. The calibrated DAC data is then stored in the X2A register. Similarly, data from the X1B register is operated on by the multiplier and adder and stored in the X2B register. Although a multiplier and adder symbol are shown for each channel, there is only one multiplier and one adder in the device, which are shared between all channels. This has implications for the update speed when several channels are updated at once, as described later. Each time data is written to the X1A register, or to the M or C register with the A/B control bit set to 0, the X2A data is recalculated and the X2A register is automatically updated. Similarly, X2B is updated each time data is written to X1B, or to M or C with A/B set to 1. The X2A and X2B registers are not readable, nor directly writable by the user. Data output from the X2A and X2B registers is routed to the final DAC register by a multiplexer. Whether each individual DAC takes its data from the X2A or X2B register is controlled by an 8-bit A/B Select Register associated with each group of 8 DACs. If a bit in this register is 0, the DAC takes its data from the X2A register; if 1 the DAC takes its data from the X2B register (bit 0 controls DAC 0 through bit 7 controls DAC 7). Note that, since there are 32 bits in 4 registers, it is possible to set up, on a per-channel basis, whether each DAC takes its data from the X2A or X2B register. A global command is also provided that sets all bits in the A/B Select Registers to 0 or to 1. Rev. PrF | Page 13 of 25 5 RESERVED 4 3 2 1 0 0 5370-0200 X2A REGISTER In addition to the gain and offset trim for each DAC, there are two 14-bit Offset DACs, one for Group 0, and one for Group 1 to Group 3. These allow the output range of all DACs connected to them to be offset within a defined range. Thus, subject to the limitations of headroom, it is possible to set the output range of Group 0 or Group 1 to Group3 to be unipolar positive, unipolar negative, or bipolar, either symmetrical or asymmetrical about zero volts. The DACs in the AD5372/AD5373 are factory trimmed with the Offset DACs set at their default values. This gives the best offset and gain performance for the default output range and span. When the output range is adjusted by changing the value of the Offset DAC an extra offset is introduced due to the gain error of the Offset DAC. The amount of offset is dependent on the magnitude of the reference and how much the Offset DAC moves from its default value. This offset is quoted on the specification page. The worst case offset occurs when the Offset DAC is at positive or negative full-scale. This value can be added to the offset present in the main DAC of a channel to give an indication of the overall offset for that channel. In most cases the offset can be removed by programming the channels C register with an appropriate value. The extra offset cause by the Offset DACs only needs to be taken into account when the Offset DAC is changed from its default value. Figure 9 shows the allowable code range which may be loaded to the Offset DAC and this is dependant on the reference value used. Thus, for a 5V reference, the Offset DAC should not be programmed with a value greater than 8192 (0x2000). VREF (V) X1A REGISTER OFFSET DACS 4096 8192 12288 OFFSET DAC CODE Figure 9. Offset DAC Code Range 16383 AD5372/AD5373 Preliminary Technical Data OUTPUT AMPLIFIER As the output amplifiers can swing to 1.4 V below the positive supply and 1.4 V above the negative supply, this limits how much the output can be offset for a given reference voltage. For example, it is not possible to have a unipolar output range of 20V, since the maximum supply voltage is ±16.5 V. S1 DAC CHANNEL AD5373 Transfer Function OUTPUT Code applied to DAC from X1A or X1B register:S2 R6 10kΩ CLR DAC_CODE = INPUT_CODE × (m+1)/214 + c - 213 DAC output voltage:- R5 R1 CLR VOUT = 4 × VREF × (DAC_CODE – OFFSET_CODE )/214 +VSIGGND CLR S3 R4 R3 R2 SIGGND SIGGND OFFSET DAC 2049-0008 c = code in offset register - default code = 214. OFFSET_CODE is the code loaded to the offset DAC. It is multiplied by 4 in the transfer function as this DAC is a 14 bit device. On power up the default code loaded to the offset DAC is 5461 (0x1555). With a 3V reference this gives a span of -4 V to +8 V. CHECK VALUE OF R1 &R5 R1,R2,R3 = 20kΩ R4,R5 = 60kΩ R6 = 10kΩ Figure 10. Output Amplifier and Offset DAC Figure 10 shows details of a DAC output amplifier and its connections to the Offset DAC. On power up, S1 is open, disconnecting the amplifier from the output. S3 is closed, so the output is pulled to SIGGND (R1 and R2 are very much greater than R6). S2 is also closed to prevent the output amplifier being open-loop. If CLR is low at power-up, the output will remain in this condition until CLR is taken high. The DAC registers can be programmed, and the outputs will assume the programmed values when CLR is taken high. Even if CLR is high at powerup, the output will remain in the above condition until VDD > 6 V and VSS < -4 V and the initialization sequence has finished. The outputs will then go to their power-on default value. TRANSFER FUNCTION The output voltage of a DAC in the AD5372/AD5373 is dependent on the value in the input register, the value of the M and C registers, and the offset from the Offset DAC. The transfer functions for the AD5372 and AD5373 are shown below. Notes DAC_CODE should be within the range of 0 to 16383. For 12 V span VREF = 3.0 V. For 20 V span VREF = 5.0 V. X1A, X1B default code = 5461 m = code in gain register - default code = 214 – 1. c = code in offset register - default code = 213. OFFSET_CODE is the code loaded to the offset DAC. It is multiplied by 4 in the transfer function as this DAC is a 14 bit device. On power up the default code loaded to the offset DAC is 5461 (0x1555). With a 3V reference this gives a span of -4 V to +8 V. REFERENCE SELECTION The AD5372/AD5373 has two reference input pins. The voltage applied to the reference pins determines the output voltage span on VOUT0 to VOUT31. VREF0 determines the voltage span for VOUT0 to VOUT7 (Group 0) and VREF1 determines the voltage span for VOUT8 to VOUT31 (Group 1 to Group 3). The reference voltage applied to each VREF pin can be different, if required, allowing the groups to have a different voltage spans. The output voltage range can be adjusted further by programming the offset and gain registers for each channel as well as programming the offset DACs. If the offset and gain features are not used (i.e. the m and c registers are left at their default values) the required reference levels can be calculated as follows: AD5372 Transfer Function VREF = (VOUTmax – VOUTmin)/4 Code applied to DAC from X1A or X1B register:16 DAC_CODE = INPUT_CODE × (m+1)/2 + c - 2 DAC output voltage:- 15 VOUT = 4 × VREF × (DAC_CODE – OFFSET_CODE )/216 +VSIGGND Notes DAC_CODE should be within the range of 0 to 65535. For 12 V span VREF = 3.0 V. For 20 V span VREF = 5.0 V. X1A, X1B default code = 21844 m = code in gain register - default code = 216 – 1. If the offset and gain features of the AD5372/AD5373 are used, then the required output range is slightly different. The chosen output range should take into account the system offset and gain errors that need to be trimmed out. Therefore, the chosen output range should be larger than the actual, required range. The required reference levels can be calculated as follows: Rev. PrF | Page 14 of 25 1. Identify the nominal output range on VOUT. 2. Identify the maximum offset span and the maximum Preliminary Technical Data AD5372/AD5373 Reducing Zero-scale and Full-scale Error gain required on the full output signal range. 3. 4. 5. Calculate the new maximum output range on VOUT including the expected maximum offset and gain errors. Choose the new required VOUTmax and VOUTmin, keeping the VOUT limits centered on the nominal values. Note that VDD and VSS must provide sufficient headroom. Zero-scale error can be reduced as follows: 1. Set the output to the lowest possible value. 2. Measure the actual output voltage and compare it with the required value. This gives the zero-scale error. 3. Calculate the number of LSBs equivalent to the error and subtract this from the default value of the C register. Note that only negative zero-scale error can be reduced. Calculate the value of VREF as follows: VREF = (VOUTMAX – VOUTMIN)/4 Reference Selection Example Full-scale error can be reduced as follows: Nominal Output Range = 12V (-4V to +8V) Offset Error = ±70mV Gain Error = ±3% SIGGND = AGND = 0V 1. Measure the zero-scale error. 2. Set the output to the highest possible value. 3. Gain Error = ±3% => Maximum Positive Gain Error = +3% => Output Range incl. Gain Error = 12 + 0.03(12)=12.36V Measure the actual output voltage and compare it with the required value. Add this error to the zero-scale error. This is the full-scale error. 4. Offset Error = ±70mV => Maximum Offset Error Span = 2(70mV)=0.14V => Output Range including Gain Error and Offset Error = 12.36V + 0.14V = 12.5V Calculate the number of LSBs equivalent to the full-scale error and subtract it from the default value of the M register. Note that only positive full-scale error can be reduced. 5. The M and C registers should not be programmed until both zero-scale and full-scale errors have been calculated. 1) 2) 3) VREF Calculation Actual Output Range = 12.5V, that is -4.25V to +8.25V (centered); VREF = (8.25V + 4.25V)/4 = 3.125V AD5372 CALIBRATION EXAMPLE If the solution yields an inconvenient reference level, the user can adopt one of the following approaches: 1. Use a resistor divider to divide down a convenient, higher reference level to the required level. 2. Select a convenient reference level above VREF and modify the Gain and Offset registers to digitally downsize the reference. In this way the user can use almost any convenient reference level but may reduce the performance by overcompaction of the transfer function. 3. Use a combination of these two approaches CALIBRATION The user can perform a system calibration on the AD5372 and AD5373 to reduce gain and offset errors to below 1 LSB. This is achieved by calculating new values for the M and C registers and reprogramming them. This example assumes that a −4 V to +8 V output is required. The DAC output is set to −4 V but measured at −4.03 V. This gives an zero-scale error of −30 mV. 1. 1 LSB = 12 V/65536 = 183.105 µV 2. 30 mV = 164 LSB 3. 164 LSB should be added to the default C register value: (32768 + 164) = 32932 4. 32932 should be programmed to the C register The full-scale error can now be removed. The output is set to +8 V and a value of +8.02 V is measured. The full-scale error is +20 mV – (–30 mV) = +50 mV This is a full-scale error of +50 mV. 1. 50 mV = 273 LSBs 2. 273 LSB should be subtracted from the default M register value: (65535 − 273) = 65262 3. 65262 should be programmed to the M register Rev. PrF | Page 15 of 25 AD5372/AD5373 Preliminary Technical Data RESET FUNCTION When the RESET pin is taken low, the DAC buffers are disconnected and the DAC outputs VOUT0 to VOUT31 are tied to their associated SIGGND signals via a 10 kΩ resistor. On the rising edge of RESET the AD5372/AD5373 state machine initiates a reset sequence to reset the X, M and C registers to their default values. This sequence typically takes 300µs and the user should not write to the part during this time. When the reset sequence is complete, and provided that CLR is high, the DAC output will be at a potential specified by the default register settings which will be equivalent to SIGGGND. The DAC outputs will remain at SIGGND until the X, M or C registers are updated and LDAC is taken low. CLEAR FUNCTION CLR is an active low input which should be high for normal operation. The CLR pin has in internal 500kΩ pull-down resistor. When CLR is low, the input to each of the DAC output buffer stages, VOUT0 to VOUT31, is switched to the externally set potential on the relevant SIGGND pin. While CLR is low, all LDAC pulses are ignored. When CLR is taken high again, the DAC outputs remain cleared until LDAC is taken low. The contents of input registers and DAC registers 0 to 31 are not affected by taking CLR low. To prevent glitches appearing on the outputs CLR should be brought low whenever the output span is adjusted by writing to the offset DAC. As described later, the ADAD5372/AD5373 has flexible addressing that allows writing of data to a single channel, all channels in a group, the same channel in groups 0 to 3 or groups 1 to 4, or all channels in the device. This means that 1, 4, 8 or 32 DAC register values may need to be calculated and updated. As there is only one multiplier shared between 32 channels, this task must be done sequentially, so the length of the BUSY pulse will vary according to the number of channels being updated. Table 7. BUSY Pulse Widths Action Loading Input, C, or M to 1 channel Loading Input, C, or M to 4 channels Loading Input, C, or M to 8 channels Loading Input, C, or M to 32 channels BUSY Pulse Width (µs max) 1.25 2.75 4.75 16.75 BUSY Pulse Width = ((Number of Channels +1) × 500ns) +250ns The AD5372/AD5373 contains an extra feature whereby a DAC register is not updated unless its X2A or X2B register has been written to since the last time LDAC was brought low. Normally, when LDAC is brought low, the DAC registers are filled with the contents of the X2A or X2B registers, depending on the setting of the A/B Select Registers. However the AD5372/AD5373 updates the DAC register only if the X2 data has changed, thereby removing unnecessary digital crosstalk. POWER-DOWN MODE BUSY AND LDAC FUNCTIONS The value of an X2 (A or B) register is calculated each time the user writes new data to the corresponding X1, C, or M registers. During the calculation of X2, the BUSY output goes low. While BUSY is low, the user can continue writing new data to the X1, M, or C registers (see the Register Update Rates section for more details), but no DAC output updates can take place. The BUSY pin is bidirectional and has a 50 kΩ internal pullup resistor. Where multiple AD5372 or AD5373 devices may be used in one system the BUSY pins can be tied together. This is useful where it is required that no DAC in any device is updated until all other DACs are ready. When each device has finished updating the X2 (A or B) registers it will release the BUSY pin. If another device hasn’t finished updating its X2 registers it will hold BUSY low, thus delaying the effect of LDAC going low. The DAC outputs are updated by taking the LDAC input low. If LDAC goes low while BUSY is active, the LDAC event is stored and the DAC outputs update immediately after BUSY goes high. A user can also hold the LDAC input permanently low. In this case, the DAC outputs update immediately after BUSY goes high. BUSY also goes low, for approximately 500ns, whenever the A/B Select Registers are written to. The AD5372/AD5373 can be powered down by setting Bit 0 in the control register. This will turn off the DACs thus reducing the current consumption. The DAC outputs will be connected to their respective SIGGND potentials. The power-down mode doesn’t change the contents of the registers and the DACs will return to their previous voltage when the power-down bit is cleared. THERMAL MONITOR FUNCTION The AD5372/AD5373 can be programmed to power down the DACs if the temperature on the die exceeds 130°C. Setting Bit 1 in the control register (see Table 15) will enable this function. If the die temperature exceeds 130°C the AD5372/AD5373 will enter a temperature power-down mode, which is equivalent to setting the power-down bit in the control register. To indicate that the AD5372/AD5373 has entered temperature power-down mode Bit 4 of the control register is set. The AD5372/AD5373 will remain in temperature shutdown mode, even if the die temperature falls, until Bit 1 in the control register is cleared. Rev. PrF | Page 16 of 25 Preliminary Technical Data AD5372/AD5373 TOGGLE MODE The AD5372/AD5373 has two X2 registers per channel, X2A and X2B, which can be used to switch the DAC output between two levels with ease. This approach greatly reduces the overhead required by a micro-processor which would otherwise have to write to each channel individually. When the user writes to either the X1A ,X2A, M or C registers the calculation engine will take a certain amount of time to calculate the appropriate X2A or X2B values. If the application only requires that the DAC output switch between two levels, such as a data generator, any method which reduces the amount of calculation time encountered is advantageous. For the data generator example the user need only set the high and low levels for each channel once, by writing to the X1A and X1B registers. The values of X2A and X2B will be calculated and stored in their respective registers. The calculation delay therefore only happens during the setup phase, i.e. when programming the initial values. To toggle a DAC output between the two levels it is only required to write to the relevant A/B Select Register to set the MUX2 register bit. Furthermore, since there are 8 MUX2 control bits per register it is possible to update eight channels with a single write. Table 17 shows the bits that correspond to each DAC output. Rev. PrF | Page 17 of 25 AD5372/AD5373 Preliminary Technical Data SERIAL INTERFACE determine what is done with the data. Two bits are reserved on the AD5373. The AD5372/AD5373 contains a high-speed SPI serial interface operating at clock frequencies up to 50 MHz (20MHz for read operations). To minimize both the power consumption of the device and on-chip digital noise, the interface powers up fully only when the device is being written to, that is, on the falling edge of SYNC. The serial interface is 2.5 V LVTTL compatible when operating from a 2.7 V to 3.6 V DVCC supply. It is controlled by four pins, as follows. The serial interface works with both a continuous and a burst (gated) serial clock. Serial data applied to SDI is clocked into the AD5372AD5373 by clock pulses applied to SCLK. The first falling edge of SYNC starts the write cycle. At least 24 falling clock edges must be applied to SCLK to clock in 24 bits of data, before SYNC is taken high again. If SYNC is taken high before the 24th falling clock edge, the write operation will be aborted. SYNC If a continuous clock is used, SYNC must be taken high before the 25th falling clock edge. This inhibits the clock within the AD5372/AD5373. If more than 24 falling clock edges are applied before SYNC is taken high again, the input data will be corrupted. If an externally gated clock of exactly 24 pulses is used, SYNC may be taken high any time after the 24th falling clock edge. Frame synchronization input. SDI Serial data input pin. SCLK Clocks data in and out of the device. SDO The input register addressed is updated on the rising edge of SYNC. In order for another serial transfer to take place, SYNC must be taken low again Serial data output pin for data readback. SPI WRITE MODE The AD5372AD5373 allows writing of data via the serial interface to every register directly accessible to the serial interface, which is all registers except the X2A and X2B registers and the DAC registers. The X2A and X2B registers are updated when writing to the X1A, X1B, M and C registers, and the DAC registers are updated by LDAC. The serial word (see Table 8 or Table 9) is 24 bits long. 16 or 14 of these bits are data bits, six bits are address bits, and two bits are mode bits that Table 8. AD5372 Serial Word Bit Assignation I23 I22 I21 I20 I19 I18 I17 I16 I15 I14 I13 I12 I11 I10 I9 I8 I7 I6 I5 I4 I3 I2 I1 I0 M1 M0 A5 A4 A3 A2 A1 A0 D15 D14 D13 D12 D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 Table 9. AD5373 Serial Word Bit Assignation I23 I22 I21 I20 I19 I18 I17 I16 I15 I14 I13 I12 I11 I10 I9 I8 I7 I6 I5 I4 I3 I2 I1* I0* M1 M0 A5 A4 A3 A2 A1 A0 D13 D12 D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 0 0 * Reserved bits. Set to 0 when writing. Bits are read back as 0 Rev. PrF | Page 18 of 25 Preliminary Technical Data AD5372/AD5373 Table 10. Group Addressing SPI READBACK MODE The ADAD5372/AD5373 allows data readback via the serial interface from every register directly accessible to the serial interface, which is all registers except the DAC data registers. In order to read back a register, it is first necessary to tell the ADAD5372/AD5373 which register is to be read. This is achieved by writing to the device a word whose first two bits are the special function code 00. The remaining bits then determine if the operation is a readback, and the register which is to be read back, or if it is a write to of the special function registers such as the control register. After the special function write has been performed, if it is a readback command then data from the selected register will be clocked out of the SDO pin during the next SPI operation. The SDO pin is normally three-state but becomes driven as soon as a read command has been issued. The pin will remain driven until the registers data has been clocked out. See Figure 5 for the read timing diagram. Note that due to the timing requirements of t5 (25ns) the maximum speed of the SPI interface during a read operation should not exceed 20MHz. A5 0 0 0 0 1 1 1 1 A4 0 0 1 1 0 0 1 1 A3 0 1 0 1 0 1 0 1 Group Selected All groups, all DACs 0 1 2 3 4 1, 2, 3, 4, 5 2, 3, 4, 5 Table 11. Channel Addressing A2 0 0 0 0 1 1 1 1 A1 0 0 1 1 0 0 1 1 A0 0 1 0 1 0 1 0 1 REGISTER UPDATE RATES Table 12. Mode Bits As mentioned previously the value of the X2 (A or B) register is calculated each time the user writes new data to the corresponding X1, C or M registers. The calculation is performed by a three stage process. The first two stages take 500ns each and the third stage takes 250ns. When the write to one of the X1, C or M registers is complete the calculation process begins. If the write operation involves the update of a single DAC channel the user is free to write to another register provided that the write operation doesn’t finish until the first stage calculation is complete, i.e. 500ns after the completion of the first write operation. If a group of channels is being updated by a single write operation the first stage calculation will be repeated for each channel, taking 500ns per channel. In this case the user should not complete the next write operation until this time has elapsed. M1 1 1 0 0 M0 1 0 1 0 Channel Selected 0 1 2 3 4 5 6 7 Action Write DAC data (x) register Write DAC offset (m) register Write DAC gain (m) register Special function, used in combination with other bits of word The AD5372/AD5373 has very flexible addressing that allows writing of data to a single channel, all channels in a group, the same channel in groups 0 to 3 or groups 1 to 3, or all channels in the device Table 10 shows all these address modes. CHANNEL ADDRESSING AND SPECIAL MODES If the mode bits are not 00, then the data word D13 to D0 is written to the device. Address bits A5 to A0 determine which channel or channels is/are written to, while the mode bits determine to which register (X1A, X1B, C or M) the data is written, as shown in Table 8 and Table 9. If data is to be written to the X1A or X1B register, the setting of the A/B bit in the Control Register determines which (0 Æ X1A, 1 Æ X1B). Rev. PrF | Page 19 of 25 AD5372/AD5373 Preliminary Technical Data Table 13. Group and Channel Addressing This table shows which group(s) and which channel(s) is/are addressed for every combination of address bits A5 to A0. ADDRESS BITS A5 TO A3 ADDRESS BITS A2 TO A0 000 001 010 011 100 101 110 111 000 All groups, all channels Group 0, channel 0 Group 1, channel 0 Group 2, channel 0 Group 3, channel 0 Reserved Groups 0,1,2,3 channel 0 Groups 1,2,3 channel 0 001 Group 0, all channels Group 0, channel 1 Group 1, channel 1 Group 2, channel 1 Group 3, channel 1 Reserved Groups 0,1,2,3 channel 1 Groups 1,2,3 channel 1 010 Group 1, all channels Group 0, channel 2 Group 1, channel 2 Group 2, channel 2 Group 3, channel 2 Reserved Groups 0,1,2,3 channel 2 Groups 1,2,3 channel 2 011 Group 2, all channels Group 0, channel 3 Group 1, channel 3 Group 2, channel 3 Group 3, channel 3 Reserved Groups 0,1,2,3 channel 3 Groups 1,2,3 channel 3 100 Group 3, all channels Group 0, channel 4 Group 1, channel 4 Group 2, channel 4 Group 3, channel 4 Reserved Groups 0,1,2,3 channel 4 Groups 1,2,3 channel 4 101 Reserved Group 0, channel 5 Group 1, channel 5 Group 2, channel 5 Group 3, channel 5 Reserved Groups 0,1,2,3 channel 5 Groups 1,2,3 channel 5 110 Reserved Group 0, channel 6 Group 1, channel 6 Group 2, channel 6 Group 3, channel 6 Reserved Groups 0,1,2,3 channel 6 Groups 1,2,3 channel 6 111 Reserved Group 0, channel 7 Group 1, channel 7 Group 2, channel 7 Group 3, channel 7 Reserved Groups 0,1,2,3 channel 7 Groups 1,2,3 channel 7 SPECIAL FUNCTION MODE The codes for the special functions are shown in Table 15. Table 16 shows the addresses for data readback. If the mode bits are 00, then the special function mode is selected, as shown in Table 14. Bits I21 to I16 of the serial data word select the special function, while the remaining bits are data required for execution of the special function, for example the channel address for data readback. Table 14. Special Function Mode I23 I22 I21 I20 I19 I18 I17 I16 I15 I14 I13 I12 I11 I10 I9 I8 I7 I6 I5 I4 I3 I2 I1 I0 0 0 S5 S4 S3 S2 S1 S0 F15 F14 F13 F12 F11 F10 F9 F8 F7 F6 F5 F4 F3 F2 F1 F0 Rev. PrF | Page 20 of 25 Preliminary Technical Data AD5372/AD5373 Table 15. Special Function Codes ACTION SPECIAL FUNCTION CODE DATA S5 S4 S3 S2 S1 S0 F15-F0 0 0 0 0 0 0 0000 0000 0000 0000 NOP 0 0 0 0 0 1 XXXX XXXX XXXX X[F2:F0] Write control register F2 = 1 Æ Select B reg for input; F2 = 0 Æ Select A reg for input F1 = 1 Æ En temp shutdown; F1 = 0 Æ Disable temp shutdown F0 = 1 Æ Soft power down; F0 = 0 Æ soft power up 0 0 0 0 1 0 XX[F13:F0] Write data in F13:F0 to OFS0 register 0 0 0 0 1 1 XX[F13:F0] Write data in F13:F0 to OFS1 register 0 0 0 1 0 0 XX[F13:F0] Reserved 0 0 0 1 0 1 See Table 14 Select register for readback 0 0 0 1 1 0 XXXX XXXX[F7:F0] Write data in F7:F0 to A/B Select Register 0 0 0 0 1 1 1 XXXX XXXX[F7:F0] Write data in F7:F0 to A/B Select Register 1 0 0 1 0 0 0 XXXX XXXX[F7:F0] Write data in F7:F0 to A/B Select Register 2 0 0 1 0 0 1 XXXX XXXX[F7:F0] Write data in F7:F0 to A/B Select Register 3 0 0 1 0 1 0 XXXX XXXX[F7:F0] Reserved 0 0 1 0 1 1 XXXX XXXX[F7:F0] Block write A/B Select Registers F7:F0 = 0, write all 0’s (all channels use X2A register) F7:F0 = 1, wrote all 1’s (all channels use X2B register) Table 16. Address Codes for Data Readback F15 F14 F13 0 0 0 F12 F11 F10 F9 F8 F7 REGISTER READ 0 0 1 0 1 0 0 1 1 1 0 0 0 0 0 0 0 1 Control Register 1 0 0 0 0 0 0 1 0 OFS0 Data Register 1 0 0 0 0 0 0 1 1 OFS1 Data Register 1 0 0 0 0 0 1 0 0 Reserved 1 0 0 0 0 0 1 1 0 A/B Select Register 0 1 0 0 0 0 0 1 1 1 A/B Select Register 1 1 0 0 0 0 1 0 0 0 A/B Select Register 2 1 0 0 0 0 1 0 0 1 A/B Select Register 3 1 0 0 0 0 1 0 1 0 Reserved X1A Register Bits F12 to F7 select channel to be read back, from Channel 0 = 001000 to Channel 31 = 100111 X1B Register C Register M Register Note: F6 to F0 are don’t care for data readback function. Rev. PrF | Page 21 of 25 AD5372/AD5373 Preliminary Technical Data Table 17. DACs Select by A/B Select Registers A/B Select Register F7 F6 F5 F4 Bits F3 F2 F1 F0 0 VOUT7 VOUT6 VOUT5 VOUT4 VOUT3 VOUT2 VOUT1 VOUT0 1 VOUT15 VOUT14 VOUT13 VOUT12 VOUT11 VOUT10 VOUT9 VOUT8 2 VOU23 VOUT22 VOUT21 VOUT20 VOUT19 VOUT18 VOUT17 VOUT16 3 VOUT31 VOUT30 VOUT29 VOUT28 VOUT27 VOUT26 VOUT25 VOUT24 POWER SUPPLY DECOUPLING this package during the assembly process. In any circuit where accuracy is important, careful consideration of the power supply and ground return layout helps to ensure the rated performance. The printed circuit board on which the AD5372/AD5373 is mounted should be designed so that the analog and digital sections are separated and confined to certain areas of the board. If the AD5372/AD5373 is in a system where multiple devices require an AGND-to-DGND connection, the connection should be made at one point only. The star ground point should be established as close as possible to the device. For supplies with multiple pins (VSS, VDD, VCC), it is recommended to tie these pins together and to decouple each supply once. POWER SUPPLY SEQUENCING When the supplies are connected to the AD5372/AD5373 it is important that the AGND and DGND pins are connected to the relevant ground plane before the positive or negative supplies are applied. In most applications this is not an issue as the ground pins for the power supplies will be connected to the ground pins of the AD5372/AD5373 via ground planes. Where the AD5372/AD5373 is to be used in a hot-swap card care should be taken to ensure that the ground pins are connected to the supply grounds before the positive or negative supplies are connected. This is required to prevent currents flowing in directions other than towards an analog or digital ground. The AD5372/AD5373 should have ample supply decoupling of 10 µF in parallel with 0.1 µF on each supply located as close to the package as possible, ideally right up against the device. The 10 µF capacitors are the tantalum bead type. The 0.1 µF capacitor should have low effective series resistance (ESR) and effective series inductance (ESI), such as the common ceramic types that provide a low impedance path to ground at high frequencies, to handle transient currents due to internal logic switching. Digital lines running under the device should be avoided, because these couple noise onto the device. The analog ground plane should be allowed to run under the AD5372/AD5373 to avoid noise coupling. The power supply lines of the AD5372/AD5373 should use as large a trace as possible to provide low impedance paths and reduce the effects of glitches on the power supply line. Fast switching digital signals should be shielded with digital ground to avoid radiating noise to other parts of the board, and should never be run near the reference inputs. It is essential to mini mize noise on all VREF lines. Avoid crossover of digital and analog signals. Traces on opposite sides of the board should run at right angles to each other. This reduces the effects of feedthrough through the board. A microstrip technique is by far the best, but not always possible with a double-sided board. In this technique, the component side of the board is dedicated to ground plane, while signal traces are placed on the solder side. As is the case for all thin packages, care must be taken to avoid flexing the package and to avoid a point load on the surface of Rev. PrF | Page 22 of 25 Preliminary Technical Data AD5372/AD5373 INTERFACING EXAMPLES The SPI interface of the AD5372/AD5373 is designed to allow the parts to be easily connected to industry standard DSPs and micro-controllers. Figure 11 shows how the AD5372/AD5373 could be connected to the Analog Devices Blackfin® DSP. The Blackfin has an integrated SPI port which can be connected directly to the SPI pins of the AD5372/AD5373 and programmable I/O pins which can be used to set or read the state of the digital input or output pins associated with the interface. transmit and receive clocks (TCLK and RCLK) are also connected together. The user can write to the AD5372/AD5373 by writing to the transmit register. A read operation can be accomplished by first writing to the AD5372/AD5373 to tell the part that a read operation is required. A second write operation with a NOP instruction will cause the data to be read from the AD5372/AD5373. The DSPs receive interrupt can be used to indicate when the read operation is complete. AD537x SPISELx SYNC SCK SCLK MOSI MISO ADSP-BF531 PF10 SDI SDO RESET PF9 LDAC PF8 CLR PF7 BUSY ADSP-21065L AD537x TFSx RFSx SYNC TCLKx RCLKx SCLK DTxA SDI DRxA SDO FLAG0 RESET FLAG1 LDAC FLAG2 CLR FLAG3 BUSY 537x-0101 Figure 12. Interfacing to an ADSP-21065L DSP 537x-0101 Figure 11. Interfacing to a Blackfin DSP The Analog Devices ADSP-21065L is a floating point DSP with two serial ports (SPORTS). Figure 12 shows how one SPORT can be used to control the AD5372/AD5373. In this example the Transmit Frame Synchronization (TFS) pin is connected to the Receive Frame Synchronization (RFS) pin. Similarly the Rev. PrF | Page 23 of 25 AD5372/AD5373 Preliminary Technical Data OUTLINE DIMENSIONS 0.75 0.60 0.45 12.00 BSC SQ 1.60 MAX 64 49 1 48 SEATING PLANE PIN 1 10.00 BSC SQ TOP VIEW (PINS DOWN) 10° 6° 2° 1.45 1.40 1.35 0.15 0.05 SEATING PLANE 0.20 0.09 7° 3.5° 0° 0.08 MAX COPLANARITY VIEW A 16 33 32 17 0.27 0.22 0.17 0.50 BSC VIEW A ROTATED 90° CCW COMPLIANT TO JEDEC STANDARDS MS-026BCD Figure 13. 64-Lead Low Profile Quad Flat Package [LQFP] (ST-64-2) Dimensions shown in millimeters 8.00 BSC SQ 0.60 MAX 0.60 MAX 0.30 0.23 0.18 43 42 PIN 1 INDICATOR TOP VIEW PIN 1 INDICATOR 56 1 6.25 6.10 SQ 5.95 EXPOSED PAD (BOTTOM VIEW) 7.75 BSC SQ 0.50 0.40 0.30 29 28 15 14 0.25 MIN 1.00 0.85 0.80 12° MAX 6.50 REF 0.80 MAX 0.65 TYP 0.50 BSC SEATING PLANE 0.05 MAX 0.02 NOM COPLANARITY 0.08 0.20 REF COMPLIANT TO JEDEC STANDARDS MO-220-VLLD-2 Figure 14. 56-Lead Free Chip Scale Package [LFCSP] (CP-56) Dimensions shown in millimeters ORDERING GUIDE Model AD5372BSTZ AD5372BCPZ AD5373BSTZ AD5373BCPZ Temperature Range -40°C to +85°C -40°C to +85°C -40°C to +85°C -40°C to +85°C Package Description 64-Lead LQFP 56-Lead LFCSP 64-Lead LQFP 56-Lead LFCSP Rev. PrF | Page 24 of 25 Package Option ST-64 CP-56 ST-64 CP-56 Preliminary Technical Data AD5372/AD5373 NOTES ©2006 Analog Devices, Inc. All rights reserved. Trademarks and registered trademarks are the property of their respective owners. PR05815-0-10/06(PrF) Rev. PrF | Page 25 of 25