625 kSPS, 24-Bit, 109 dB Σ−Δ ADC With On-Chip Buffer AD7762 FEATURES FUNCTIONAL BLOCK DIAGRAM VIN– VIN+ MULTIBIT Σ-Δ MODULATOR DIFF AVDD1 AVDD2 AVDD3 VREF+ RECONSTRUCTION BUF AVDD4 DECAPA/B RBIAS AD7762 PROGRAMMABLE DECIMATION AGND VDRIVE MCLK DVDD CONTROL LOGIC I/O OFFSET AND GAIN REGISTERS SYNC RESET CS DGND FIR FILTER ENGINE RD/WR DRDY DB0 TO DB15 05477-001 120 dB dynamic range at 78 kHz output data rate 109 dB dynamic range at 625 kHz output data rate 112 dB SNR at 78 kHz output data rate 106 dB SNR at 625 kHz output data rate 625 kHz maximum fully filtered output word rate Programmable over-sampling rate (32× to 256×) Fully differential modulator input On-chip differential amplifier for signal buffering Low-pass finite impulse response (FIR) filter with default or user-programmable coefficients Overrange alert bit Digital offset and gain correction registers Filter bypass modes Low power and power-down modes Synchronization of multiple devices via SYNC pin Figure 1. APPLICATIONS Data acquisition systems Vibration analysis Instrumentation GENERAL DESCRIPTION The AD7762 is a high performance, 24-bit Σ-Δ analog-todigital converter (ADC). It combines wide input bandwidth and high speed with the benefits of Σ-Δ conversion with a performance of 106 dB SNR at 625 kSPS, making it ideal for high speed data acquisition. Wide dynamic range combined with significantly reduced antialiasing requirements simplify the design process. An integrated buffer to drive the reference, a differential amplifier for signal buffering and level shifting, an overrange flag, internal gain and offset registers, and a low-pass digital FIR filter make the AD7762 a compact, highly integrated data acquisition device requiring minimal peripheral component selection. In addition, the device offers programmable decimation rates, and the digital FIR filter can be adjusted if the default characteristics are not appropriate to the application. The AD7762 is ideal for applications demanding high SNR without a complex front end signal processing design. The differential input is sampled at up to 40 MSPS by an analog modulator. The modulator output is processed by a series of lowpass filters, the final filter having default or user-programmable coefficients. The sample rate, filter corner frequencies, and output word rate are set by a combination of the external clock frequency and the configuration registers of the AD7762. The reference voltage supplied to the AD7762 determines the analog input range. With a 4 V reference, the analog input range is ±3.2 V differential biased around a common mode of 2 V. This common-mode biasing can be achieved using the on-chip differential amplifier, further reducing the external signal conditioning requirements. The AD7762 is available in an exposed paddle, 64-lead TQFP and is specified over the industrial temperature range from −40°C to +85°C. Table 1. Related Devices Part No. AD7760 AD7763 Description 24-bit, 2.5 MSPS, 100 dB Σ-Δ, parallel interface 24-bit, 625 kSPS, 109 dB Σ-Δ, serial interface 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 that may result from its use. Specifications subject to change without notice. No license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Trademarks and registered trademarks are the property of their respective owners. One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A. Tel: 781.329.4700 www.analog.com Fax: 781.461.3113 © 2005 Analog Devices, Inc. All rights reserved. AD7762 TABLE OF CONTENTS General Description ......................................................................... 1 Bias Resistor Selection ............................................................... 17 Specifications..................................................................................... 3 Decoupling and Layout Recommendations................................ 18 Timing Specifications....................................................................... 5 Supply Decoupling ..................................................................... 19 Timing Diagrams.......................................................................... 5 Additional Decoupling .............................................................. 19 Absolute Maximum Ratings............................................................ 6 Reference Voltage Filtering ....................................................... 19 ESD Caution.................................................................................. 6 Differential Amplifier Components ........................................ 19 Pin Configuration and Function Descriptions............................. 7 Layout Considerations............................................................... 19 Terminology ...................................................................................... 9 Programmable FIR Filter............................................................... 20 Typical Performance Characteristics ........................................... 10 Downloading a User-Defined Filter ............................................ 21 Theory of Operation ...................................................................... 13 Example Filter Download ......................................................... 21 AD7762 Interface............................................................................ 14 AD7762 Registers ........................................................................... 23 Reading Data............................................................................... 14 Control Register 1—Reg 0x0001 .............................................. 23 Sharing the Parallel Bus ............................................................. 14 Control Register 2—Address 0x0002 ...................................... 23 Writing to the AD7762 .............................................................. 14 Status Register (Read Only) ...................................................... 24 Reading Status and Other Registers......................................... 14 Offset Register—Address 0x0003............................................. 24 Clocking the AD7762 ................................................................ 15 Gain Register—Address 0x0004............................................... 24 Example 1 .................................................................................... 15 Overrange Register—Address 0x0005..................................... 24 Example 2 .................................................................................... 15 Outline Dimensions ....................................................................... 25 Driving the AD7762....................................................................... 16 Ordering Guide .......................................................................... 25 Using the AD7762 ...................................................................... 17 REVISION HISTORY 8/05—Revision 0: Initial Version Rev. 0 | Page 2 of 28 AD7762 SPECIFICATIONS AVDD1 = DVDD = VDRIVE = 2.5 V, AVDD2 = AVDD3 = AVDD4 = 5 V, VREF = 4.096 V, MCLK amplitude = 5 V, TA = 25°C, normal mode, using on-chip amplifier with components as shown in Table 8, unless otherwise noted. 1 Table 2. Parameter DYNAMIC PERFORMANCE Decimate by 256 Dynamic Range Signal-to-Noise Ratio (SNR) 2 Test Conditions/Comments MCLK = 40 MHz, ODR = 78 kHz, FIN = 1 kHz Modulator inputs shorted Decimate by 64 Dynamic Range Input amplitude = −0.5 dBFS Input amplitude = −60 dBFS Nonharmonic, input amplitude = −6 dBFS Input amplitude = −60 dBFS Input amplitude = −0.5 dBFS Input amplitude = −6 dBFS Input amplitude = −60 dBFS MCLK = 40 MHz, ODR = 312.5 kHz, FIN = 1 kHz Modulator inputs shorted Signal-to-Noise Ratio (SNR)2 Spurious-Free Dynamic Range (SFDR) Decimate by 32 Dynamic Range Input amplitude = −0.5 dBFS Nonharmonic, input amplitude = −6 dBFS MCLK = 40 MHz, ODR = 625 kHz, FIN =100 kHz Modulator inputs shorted Spurious-Free Dynamic Range (SFDR) Total Harmonic Distortion (THD) Signal-to-Noise Ratio (SNR)2 Spurious-Free Dynamic Range (SFDR) Total Harmonic Distortion (THD) DC ACCURACY Resolution Differential Nonlinearity Integral Nonlinearity Zero Error Gain Error Zero Error Drift Gain Error Drift DIGITAL FILTER RESPONSE Decimate by 32 Group Delay Decimate by 64 Group Delay Decimate by 256 Group Delay ANALOG INPUT Differential Input Voltage Input Capacitance Specification Unit 119 120.5 112 59 126 77 −105 −106 −75 dB min dB typ 112 114 109.5 126 dB typ dB typ dBc typ dBc typ dB typ dB typ dB typ dB min dB typ dB typ dBc typ 108 109.5 107 120 −108 −106 dB typ dBc typ dB typ dB typ 24 Bits 0.00076 0.014 0.02 0.015 0.019 0.0002 % typ % typ % max % typ %/°C typ %/°C typ MCLK = 40 MHz 47 μs typ MCLK = 40 MHz 91.5 μs typ MCLK = 40 MHz 358 μs typ VIN(+) – VIN(−), VREF = 2.5 V VIN(+) – VIN(−), VREF = 4.096 V At internal buffer inputs At modulator inputs ±2 ±3.25 5 55 V p-p V p-p pF typ pF typ Input amplitude = −0.5 dBFS Nonharmonic, input amplitude = −6 dBFS Input amplitude = −0.5 dBFS Input amplitude = −6 dBFS dB min dB typ Guaranteed monotonic to 24 bits Rev. 0 | Page 3 of 28 AD7762 Parameter REFERENCE INPUT/OUTPUT VREF Input Voltage VREF Input DC Leakage Current VREF Input Capacitance POWER DISSIPATION Total Power Dissipation Standby Mode POWER REQUIREMENTS AVDD1 (Modulator Supply) AVDD2 (General Supply) AVDD3 (Diff Amp Supply) AVDD4 (Ref Buffer Supply) DVDD VDRIVE Normal Mode AIDD1 (Modulator) AIDD2 (General) AIDD4 (Reference Buffer) Low Power Mode AIDD1 (Modulator) AIDD2 (General) AIDD4 (Reference Buffer) AIDD3 (Diff Amp) DIDD DIGITAL I/O MCLK Input Amplitude 3 Input Capacitance Input Leakage Current Three-State Leakage Current (D15:D0) VINH VINL VOH 4 VOL4 Test Conditions/Comments Specification Unit VDD3 = 3.3 V ± 5% VDD3 = 5 V ± 5% +2.5 +4.096 ±6 5 V max V max μA max pF max Normal mode Low power mode Clock stopped 958 661 6.35 mW max mW max mW max ±5% ±5% +2.5 +5 +3.15/+5.25 +3.15/+5.25 +2.5 +1.65/+2.7 V V V min/max V min/max V V min/max AVDD4 = 5 V 49/51 40/42 34/36 mA typ/max mA typ/max mA typ/max AVDD4 = 5 V 26/28 20/23 9/10 mA typ/max mA typ/max mA typ/max AVDD3 = 5 V, both modes Both modes 41/44 63/70 mA typ/max mA typ/max 5 7.3 ±5 ±5 0.7 × VDRIVE 0.3 × VDRIVE 1.5 0.1 V typ pF typ μA max μA max V min V max V min V max ±5% 1 See the Terminology section. SNR specifications in dBs are referred to a full-scale input, FS. Tested with an input signal at 0.5 dB below full scale, unless otherwise specified. 3 While the AD7762 can function with an MCLK amplitude of less than 5 V, this is the recommended amplitude to achieve the performance as stated. 4 Tested with a 400 μA load current. 2 Rev. 0 | Page 4 of 28 AD7762 TIMING SPECIFICATIONS AVDD1 = DVDD = VDRIVE = 2.5 V, AVDD2 = AVDD3 = AVDD4 = 5 V, TA = 25°C, normal mode, unless otherwise noted. Table 3. Parameter fMCLK fICLK t1 1, 2 t2 t3 t4 t5 t6 t7 t8 t9 t10 t11 t12 1 2 Limit at TMIN, TMAX 1 40 500 20 0.5 × tICLK 10 3 (0.5 × tICLK) + 16 ns tICLK tICLK 3 11 4 × tICLK 4 × tICLK 5 0 Unit MHz min MHz max kHz min MHz max typ ns min ns min max min min ns min ns max min min ns min ns min Description Applied master clock frequency Internal modulator clock derived from MCLK DRDY pulse width DRDY falling edge to CS falling edge RD/WR setup time to CS falling edge Data access time CS low read pulse width CS high pulse width between reads RD/WR hold time to CS rising edge Bus relinquish time CS low write pulse width CS high period between address and data Data setup time Data hold time tICLK = 1/fICLK. When ICLK = MCLK, DRDY pulse width depends on the mark/space ratio of applied MCLK. TIMING DIAGRAMS DRDY t5 t1 CS t6 t2 t7 t3 RD/WR t8 t4 DATA MSW LSW + STATUS 05477-002 D[0:15] Figure 2. Parallel Interface Timing Diagram CS t9 t10 t11 D[0:15] t12 REGISTER ADDRESS Figure 3. AD7762 Register Write Rev. 0 | Page 5 of 28 REGISTER DATA 05477-004 RD/WR AD7762 ABSOLUTE MAXIMUM RATINGS TA = 25°C, unless otherwise noted. Table 4. Parameters AVDD1 to GND AVDD2–AVDD4 to GND DVDD to GND VDRIVE to GND VIN+, VIN– to GND Digital input voltage to GND1 MCLK to MCLKGND VREF to GND2 AGND to DGND Input Current to Any Pin Except Supplies3 Operating Temperature Range Commercial Storage Temperature Range Junction Temperature TQFP Exposed Paddle Package θJA Thermal Impedance θJC Thermal Impedance Lead Temperature, Soldering Vapor Phase (60 sec) Infrared (15 sec) ESD Rating −0.3 V to +3 V −0.3 V to +6 V −0.3 V to +3 V −0.3 V to +3 V −0.3 V to +6 V −0.3 V to DVDD + 0.3 V −0.3 V to +6 V −0.3 V to AVDD4 + 0.3 V −0.3 V to +0.3 V ±10 mA Stresses above those listed under Absolute Maximum Ratings may cause permanent damage to the device. This is a stress rating only; functional operation of the device at these or any other conditions above those 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 150°C 92.7°C/W 5.1°C/W 215°C 220°C 600 V 1 Absolute maximum voltage on digital inputs is 3.0 V or DVDD + 0.3 V, whichever is lower. 2 Absolute maximum voltage on VREF input is 6.0 V or AVDD4 + 0.3 V, whichever is lower. 3 Transient currents of up to 200 mA do not cause SCR latch-up. 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. 0 | Page 6 of 28 AD7762 64 63 62 61 60 59 58 DB11 DB10 DB9 DB8 DGND DB7 DB6 DB5 DB4 DB3 DB2 DB1 DB0 DGND VDRIVE DGND PIN CONFIGURATION AND FUNCTION DESCRIPTIONS 57 56 55 54 53 52 51 50 49 DGND 1 48 DB12 47 DB13 MCLK 3 46 DB14 AVDD2 4 45 DB15 AGND2 5 44 VDRIVE AVDD1 6 AD7762 43 DGND AGND1 7 TOP VIEW (Not to Scale) 42 DGND 41 DVDD 40 CS VREF+ 10 39 RD/WR AGND4 11 38 DRDY AVDD4 12 37 RESET AGND2 13 36 SYNC AVDD2 14 35 DGND AVDD2 15 34 AGND1 AGND2 16 33 AVDD1 PIN 1 MCLKGND 2 DECAPA 8 REFGND 9 05477-005 AGND3 AGND3 DECAPB AGND3 AGND2 AVDD2 VIN– VIN+ AVDD3 AGND3 VOUTA+ VOUTA– VINA– VINA+ RBIAS AGND2 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Figure 4. 64-Lead TQFP Pin Configuration Table 5. Pin Function Descriptions Pin No. 6, 33 Mnemonic AVDD1 4, 14, 15, 27 AVDD2 24 AVDD3 12 AVDD4 7, 34 5, 13, 16, 18, 28 23, 29, 31, 32 11 9 41 AGND1 AGND2 AGND3 AGND4 REFGND DVDD 44, 63 VDRIVE 1, 35, 42, 43, 53, 62, 64 19 20 21 22 25 26 10 DGND VINA+ VINA− VOUTA− VOUTA+ VIN+ VIN− VREF+ 8 DECAPA Description 2.5 V Power Supply for Modulator. These pins should be decoupled to AGND1 with 100 nF and 10 μF capacitors on each pin. 5 V Power Supply. These pins should be decoupled to AGND2 with 100 nF capacitors on each of Pin 4, Pin 14, and Pin 15. Pin 27 should be connected to Pin 14 via a 15 nH inductor. 3.3 V to 5 V Power Supply for Differential Amplifier. These pins should be decoupled to AGND3 with a 100 nF capacitor. 3.3 V to 5 V Power Supply for Reference Buffer. This pin should be decoupled to AGND4 with a 10 nF capacitor in series with a 10 Ω resistor. Power Supply Ground for Analog Circuitry Powered by AVDD1. Power Supply Ground for Analog Circuitry Powered by AVDD2. Power Supply Ground for Analog Circuitry Powered by AVDD3. Power Supply Ground for Analog Circuitry Powered by AVDD4. Reference Ground. Ground connection for the reference voltage. 2.5 V Power Supply for Digital Circuitry and FIR Filter. This pin should be decoupled to DGND with a 100 nF capacitor. Logic Power Supply Input, 1.8 V to 2.5 V. The voltage supplied at these pins determines the operating voltage of the logic interface. Both these pins must be connected together and tied to the same supply. Each pin should also be decoupled to DGND with a100 nF capacitor. Ground Reference for Digital Circuitry. Positive Input to Differential Amplifier. Negative Input to Differential Amplifier. Negative Output from Differential Amplifier. Positive Output from Differential Amplifier. Positive Input to the Modulator. Negative Input to the Modulator. Reference Input. The input range of this pin is determined by the reference buffer supply voltage (AVDD4). See the Reference Voltage Filtering section for more details. Decoupling Pin. A 100 nF capacitor must be inserted between this pin and AGND1. Rev. 0 | Page 7 of 28 AD7762 Pin No. 30 17 Mnemonic DECAPB RBIAS 45 to 52, 54 to 61 DB15 to DB8 DB7 to DB0 37 RESET 3 MCLK 2 36 MCLKGND SYNC 39 RD/WR 38 DRDY 40 CS Description Decoupling Pin. A 33 pF capacitor must be inserted between this pin and AGND3. Bias Current Setting Pin. A resistor must be inserted between this pin and AGND1. For more details, see the Bias Resistor Selection section. 16-Bit Bidirectional Data Bus. These are three-state pins that are controlled by the CS pin and the RD/WR pin. The operating voltage for these pins is determined by the VDRIVE voltage. See the AD7762 Interface section for more details. A falling edge on this pin resets all internal digital circuitry and powers down the part. Holding this pin low keeps the AD7762 in a reset state. Master Clock Input. A low jitter digital clock must be applied to this pin. The output data rate depends on the frequency of this clock. See the section Clocking the AD7762 for more details. Master Clock Ground Sensing Pin. Synchronization Input. A falling edge on this pin resets the internal filter. This can be used to synchronize multiple devices in a system. Read/Write Input. This pin, in conjunction with the chip select pin, is used to read and write data to and from the AD7762. If this pin is low when CS is low, a read takes place. If this pin is high and CS is low, a write occurs. See the AD7762 Interface section for more details. Data Ready Output. Each time that new conversion data is available, an active low pulse, ½ ICLK period wide, is produced on this pin. See the AD7762 Interface section for more details. Chip Select Input. Used in conjunction with the RD/WR pin to read and write data to and from the AD7762. See the AD7762 Interface section for more details. Rev. 0 | Page 8 of 28 AD7762 TERMINOLOGY Signal-to-Noise Ratio (SNR) SNR is the ratio of the rms value of the actual input signal to the rms sum of all other spectral components below the Nyquist frequency, excluding harmonics and dc. The value for SNR is expressed in decibels. Total Harmonic Distortion (THD) The ratio of the rms sum of harmonics to the fundamental. For the AD7762, it is defined as THD (dB ) = 20 log V22 + V32 + V42 + V52 + V62 V1 where: V1 is the rms amplitude of the fundamental. V2, V3, V4, V5,.and V6 are the rms amplitudes of the second to the sixth harmonics. Nonharmonic Spurious-Free Dynamic Range (SFDR) The ratio of the rms signal amplitude to the rms value of the peak spurious spectral component, excluding harmonics. Dynamic Range Dynamic range is the ratio of the rms value of the full scale to the rms noise measured with the inputs shorted together. The value for dynamic range is expressed in decibels. Integral Nonlinearity (INL) The maximum deviation from a straight line passing through the endpoints of the ADC transfer function. Differential Nonlinearity (DNL) The difference between the measured and the ideal 1-LSB change between any two adjacent codes in the ADC. Zero Error The zero error is the difference between the ideal midscale input voltage (0 V) and the actual voltage producing the midscale output code. Zero Error Drift The change in the actual zero error value due to a temperature change of 1°C. It is expressed as a percentage of the zero error at room temperature. Gain Error The first transition (from 100…000 to 100…001) should occur for an analog voltage 1/2 LSB above the nominal negative full scale. The last transition (from 011…110 to 011…111) should occur for an analog voltage 1 1/2 LSB below the nominal full scale. The gain error is the deviation of the difference between the actual level of the last transition and the actual level of the first transition, from the difference between the ideal levels. Gain Error Drift The change in the actual gain error value due to a temperature change of 1°C. It is expressed as a percentage of the gain error at room temperature. Rev. 0 | Page 9 of 28 AD7762 TYPICAL PERFORMANCE CHARACTERISTICS 0 –25 –25 –50 –50 –75 –100 –125 –75 –100 –125 –150 –150 –175 –175 –200 0 4000 8000 12000 16000 20000 05477-009 AMPLITUDE (dB) 0 05477-006 AMPLITUDE (dB) AVDD1 = DVDD = VDRIVE = 2.5 V, AVDD2 = AVDD3 = AVDD4 = 5 V, VREF = 4.096 V, TA = 25°C, normal mode, unless otherwise noted. All FFTs are generated from 65536 samples using a 7-term Blackman-Harris window. –200 24000 0 4000 FREQUENCY (Hz) 0 –25 –25 –50 –50 –100 –125 –75 –100 –125 –150 –150 –175 –175 –200 0 4000 8000 12000 16000 20000 –200 24000 0 4000 FREQUENCY (Hz) –25 –50 –50 –75 –100 –125 –125 –175 –175 16000 24000 –100 –150 12000 20000 –75 –150 8000 16000 20000 05477-011 AMPLITUDE (dB) –25 05477-008 AMPLITUDE (dB) 0 4000 12000 Figure 9. Low Power FFT, 1 kHz, −6 dB Input Tone, 256× Decimation 0 0 8000 FREQUENCY (Hz) Figure 6. Normal Mode FFT, 1 kHz, −6 dB Input Tone, 256× Decimation –200 24000 05477-010 AMPLITUDE (dB) 0 –75 20000 Figure 8. Low Power FFT, 1 kHz, −0.5 dB Input Tone, 256× Decimation 05477-007 AMPLITUDE (dB) Figure 5. Normal Mode FFT, 1 kHz, −0.5 dB Input Tone, 256× Decimation 8000 12000 16000 FREQUENCY (Hz) –200 24000 0 FREQUENCY (Hz) 4000 8000 12000 16000 20000 24000 FREQUENCY (Hz) Figure 7. Normal Mode FFT, 1 kHz, −60 dB Input Tone, 256× Decimation Figure 10. Low Power FFT, 1 kHz, −60 dB Input Tone, 256× Decimation Rev. 0 | Page 10 of 28 0 –25 –25 –50 –50 –75 –100 –125 –75 –100 –125 –150 –150 –175 –175 –200 0 60000 120000 180000 240000 05477-063 AMPLITUDE (dB) 0 05477-060 AMPLITUDE (dB) AD7762 –200 300000 0 60000 FREQUENCY (Hz) –25 –50 –50 –75 –100 –125 –75 –100 –125 –150 –150 –175 –175 –200 180000 300000 240000 05477-064 AMPLITUDE (dB) –25 05477-061 AMPLITUDE (dB) 0 120000 240000 Figure 14. Low Power FFT, 100 kHz, −0.5 dB Input Tone, 32× Decimation 0 60000 180000 FREQUENCY (Hz) Figure 11. Normal Mode FFT, 100 kHz, −0.5 dB Input Tone, 32× Decimation 0 120000 –200 300000 0 60000 FREQUENCY (Hz) 120000 180000 240000 300000 FREQUENCY (Hz) Figure 12. Normal Mode FFT, 100 kHz, −6 dB Input Tone, 32× Decimation Figure 15. Low Power FFT, 100 kHz, −6 dB Input Tone, 32× Decimation 120 116 –60dB 118 –60dB 116 112 108 –0.5dB 110 05477-062 108 106 0 64 128 192 –0.5dB 05477-065 114 –6dB SNR (dBFS) SNR (dBFS) 112 –6dB 104 256 0 DECIMATION RATE (x) 64 128 192 256 DECIMATION RATE (x) Figure 13. Normal Mode SNR vs. Decimation Rate, 1 kHz Input Tone Figure 16. Low Power SNR vs. Decimation Rate, 1 kHz Input Tone Rev. 0 | Page 11 of 28 AD7762 4500 3000 4000 2500 3000 OCCURRENCE OCCURRENCE 3500 2500 2000 1500 2000 1500 1000 1000 0 8385222 8385238 8385254 0 8383530 83835246 8383562 8385270 24-BIT CODE 05477-058 05477-055 500 500 8383578 8383594 8383610 24-BIT CODE Figure 17. Normal Mode, 24-Bit Histogram, 256× Decimation Figure 20. Low Power, 24-Bit Histogram, 256× Decimation 0.0010 0.0015 +85°C +85°C 0.0010 0.0005 +25°C INL (%) INL (%) 0.0005 0 +25°C 0 –40°C –40°C –0.0005 –0.0010 0 4194304 8388608 12582912 –0.0010 16777216 24-BIT CODE 0.4 0 –0.2 –0.4 05477-057 DNL (LSB) 0.2 –0.6 8388608 4194304 8388608 12582912 Figure 21. 24-Bit INL, Low Power Mode 0.6 4194304 0 24-BIT CODE Figure 18. 24-Bit INL, Normal Mode 0 05477-059 05477-056 –0.0005 12582912 16777216 24-BIT CODE Figure 19. 24-Bit DNL Rev. 0 | Page 12 of 28 16777216 AD7762 THEORY OF OPERATION The AD7762 employs a Σ-Δ conversion technique to convert the analog input into an equivalent digital word. The modulator samples the input waveform and outputs an equivalent digital word to the digital filter at a rate equal to ICLK. rate to be chosen from 4× to 32×. The third filter has a fixed decimation rate of 2×, is user programmable, and has a default configuration. It is described in detail in the Programmable FIR Filter section. This filter can be bypassed. Due to the high oversampling rate, that spreads the quantization noise from 0 to fICLK, the noise energy contained in the band of interest is reduced (Figure 22 a). To further reduce the quantization noise, a high order modulator is employed to shape the noise spectrum; so that most of the noise energy is shifted out of the band of interest (Figure 22 b). Table 6 lists some characteristics of the default filter. The group delay of the filter is defined to be the delay to the center of the impulse response and is equal to the computation + filter delays. The delay until valid data is available (the DVALID status bit is set) is equal to 2× the filter delay + the computation delay. a. The digital filtering that follows the modulator removes the large out-of-band quantization noise (Figure 22 c) while also reducing the data rate from fICLK at the input of the filter to fICLK/8 or less at the output of the filter, depending on the decimation rate used. QUANTIZATION NOISE fICLK\2 BAND OF INTEREST b. Digital filtering has certain advantages over analog filtering. It does not introduce significant noise or distortion and can be made perfectly linear phase. NOISE SHAPING The AD7762 employs three FIR filters in series. By using different combinations of decimation ratios and filter selection and bypassing, data can be obtained from the AD7762 at a large range of data rates. The first filter receives data from the modulator at ICLK MHz where it is decimated by four to output data at ICLK/4 MHz. This partially filtered data can also be output at this stage. The second filter allows the decimation fICLK\2 BAND OF INTEREST c. fICLK\2 BAND OF INTEREST 04975-037 DIGITAL FILTER CUTOFF FREQUENCY Figure 22. Σ-Δ ADC Table 6. Configuration with Default Filter ICLK Frequency 20 MHz 20 MHz 20 MHz 20 MHz 20 MHz 20 MHz 20 MHz 12.288 MHz 12.288 MHz 12.288 MHz 12.288 MHz Filter 1 4× 4× 4× 4× 4× 4× 4× 4× 4× 4× 4× Filter 2 4× 8× 8× 16× 16× 32× 32× 8× 16× 32× 32× Filter 3 2× Bypassed 2× Bypassed 2× Bypassed 2× 2× 2× Bypassed 2× Data State Fully filtered Partially filtered Fully filtered Partially filtered Fully filtered Partially filtered Fully filtered Fully filtered Fully filtered Partially filtered Fully filtered Computation Delay 1.775 μs 2.6 μs 2.25 μs 4.175 μs 3.1 μs 7.325 μs 4.65 μs 3.66 μs 5.05 μs 11.92 μs 7.57 μs Rev. 0 | Page 13 of 28 Filter Delay 44.4 μs 10.8 μs 87.6 μs 20.4 μs 174 μs 39.6 μs 346.8 μs 142.6 μs 283.2 μs 64.45 μs 564.5 μs Pass-Band Bandwidth 250 kHz 140.625 kHz 125 kHz 70.3125 kHz 62.5 kHz 35.156 kHz 31.25 kHz 76.8 kHz 38.4 kHz 21.6 kHz 19.2 kHz Output Data Rate (ODR) 625 kHz 625 kHz 312.5 kHz 312.5 kHz 156.25 kHz 156.25 kHz 78.125 kHz 192 kHz 96 kHz 96 kHz 48 kHz AD7762 AD7762 INTERFACE READING DATA WRITING TO THE AD7762 The AD7762 uses a 16-bit bidirectional parallel interface. This interface is controlled by the RD/WR and CS pins. While the AD7762 is configured to convert analog signals with the default settings on reset, there are many features and parameters on this part that the user can change by writing to the device. Because some of the programmable registers are 16 bits wide, two write operations are required to program a register. The first write contains the register address while the second write contains the register data. An exception is when a user filter is being downloaded to the AD7762. This is described in detail in the Downloading a User-Defined Filter section. The AD7762 Registers section contains the register addresses and more details. When a new conversion result is available, an active low pulse is output on the DRDY pin. To read a conversion result from the AD7762, two 16-bit read operations are performed. The DRDY pulse indicates that a new conversion result is available. Both RD/WR and CS go low to perform the first read operation. Shortly after both these lines go low, the data bus becomes active and the 16 most significant bits (MSBs) of the conversion result are output. The RD/WR and CS lines must return high for a full ICLK period before the second read is performed. This second read contains the 8 least significant bits (LSBs) of the conversion result along with 6 status bits. These status bits are shown in Table 7. Descriptions of the other status bits are in Table 15. Table 7. Status Bits During Data Read D7 DValid Ovr UFilt LPwr FiltOk DLOk 0 D0 0 Shortly after RD/WR and CS return high, the data bus returns to a high impedance state. Both read operations must be completed before a new conversion result is available because the new result overwrites the contents on the output register. If a DRDY pulse occurs during a read operation, the data read is invalid. Figure 3 shows a write operation to the AD7762. The RD/WR line is held high while the CS line is brought low for a minimum of 4 ICLK periods. The register address is latched during this period. The CS line is brought high again for a minimum of 4 ICLK periods before the register data is put onto the data bus. If a read operation occurs between the writing of the register address and the register data, the register address is cleared and the next write must be the register address again. This also provides a method to get back to a known situation if the user forgets whether the next write is an address or data. Generally, the AD7762 is written to and configured on powerup and very infrequently, if at all, after that. Following any write operation, the full group delay of the filter must pass before valid data is output from the AD7762. READING STATUS AND OTHER REGISTERS SHARING THE PARALLEL BUS By its nature, the high accuracy of the AD7762 makes it sensitive to external noise sources. These include digital activity on the parallel bus. For this reason, it is recommended that the AD7762 data lines are isolated from the system data bus by means of a latch or buffer to ensure that there is no digital activity on the D0 to D15 pins that is not controlled by the AD7762. If multiple, synchronized AD7762 parts that share a properly distributed common MCLK signal exist in a system, these parts can share a common bus without being isolated from each other. This bus can then be isolated from the system bus by a single latch or buffer. The AD7762 features a number of programmable registers. To read back the contents of these registers or the status register, the user must first write to the control register of the device, setting a bit corresponding to the register to be read. The next read operation outputs the contents of the selected register instead of a conversion result. The AD7762 Registers section provides more information on the relevant bits in the control register. Rev. 0 | Page 14 of 28 AD7762 CLOCKING THE AD7762 EXAMPLE 2 The AD7762 requires an external low jitter clock source. This signal is applied to the MCLK pin, and the MCLKGND pin is used to sense the ground from the clock source. An internal clock signal (ICLK) is derived from the MCLK input signal. The ICLK controls the internal operations of the AD7762. The maximum ICLK frequency is 20 MHz, but due to an internal clock divider, a range of MCLK frequencies can be used. There are two ways to generate the ICLK: Take a second example from Table 6, where: ICLK = MCLK (CDIV = 1) ICLK = MCLK/2 (CDIV = 0) These options are selected from the control register (see the AD7762 Registers section for more details). On power-up, the default is ICLK = MCLK/2 to ensure that the part can handle the maximum MCLK frequency of 40 MHz. For output data rates equal to those used in audio systems, a 12.288 MHz ICLK frequency can be used. As shown in Table 6, output data rates of 192 kHz, 96 kHz, and 48 kHz are achievable with this ICLK frequency. As mentioned previously, this ICLK frequency can be derived from different MCLK frequencies. ODR = 48 kHz fICLK = 12.288 MHz fIN (max) = 19.2 kHz SNR = 120 dB t j(rms ) = 256 = 133 ps 2 × π × 19.2 × 10 3 × 10 6 The input amplitude also has an effect on these jitter figures. If, for example, the input level was 3 dB below full scale, the allowable jitter would be increased by a factor of √2, increasing the first example to 2.53 ps rms. This happens when the maximum slew rate is decreased by a reduction in amplitude. Figure 23 and Figure 24 illustrate this point, showing the maximum slew rate of a sine wave of the same frequency but with different amplitudes. 1 0.5 The MCLK jitter requirements depend on a number of factors and are given by 0 OSR t j(rms ) = 2 × π × f IN × 10 SNR ( dB ) 20 OSR = Over-sampling ratio = f ICLK ODR –1.0 04975-038 –0.5 where: fIN = Maximum input frequency Figure 23. Maximum Slew Rate of Sine Wave with Amplitude of 2 V p-p SNR(dB) = Target SNR EXAMPLE 1 1 This example can be taken from Table 6, where: ODR = 625 kHz fICLK = 20 MHz fIN (max) = 250 kHz SNR = 108 dB 0.5 0 This is the maximum allowable clock jitter for a full-scale, 250 kHz input tone with the given ICLK and output data rate. –0.5 –1.0 04975-039 t j(rms ) 32 = = 3.6ps 2 × π × 250 × 103 × 106 Figure 24. Maximum Slew Rate of Same Frequency Sine Wave with Amplitude of 1 V p-p Rev. 0 | Page 15 of 28 AD7762 DRIVING THE AD7762 To achieve the specified performance in normal mode, the differential amplifier should be configured as a first-order antialias filter, as shown in Figure 25. Any additional filtering should be carried out in previous stages using low noise, high performance op amps, such as the AD8021. Suitable component values for the first-order filter are listed in Table 8. The values in Table 8 yield a 10 dB attenuation at the first alias point of 19 MHz. +2.5V +3.685V 0V +2.048V VIN+ A –2.5V +0.410V +2.5V +3.685V B 0V +2.048V –2.5V +0.410V VIN– 04975-041 The AD7762 has an on-chip differential amplifier that operates with a supply voltage (AVDD3) from 3.15 V to 5.25 V. For a 4.096 V reference, the supply voltage must be 5 V. Figure 26. Differential Amplifier Signal Conditioning CFB CFB RFB VIN– CS A1 B 2R VIN RIN RM VIN– CS RFB RM 18 Ω 04975-040 RFB CFB Table 8. Normal Mode Component Values RFB 655 Ω RM VIN+ RIN Figure 25. Differential Amplifier Configuration RIN 1 kΩ A1 R CFB VREF 4.096 V RM AD8021 VIN+ RIN RFB 2R RM 04975-042 RIN Figure 27. Single-Ended-to-Differential Conversion CS 5.6 pF CFB 33 pF VIN+ Figure 26 shows the signal conditioning that occurs using the circuit in Figure 25 with a ±2.5 V input signal biased around ground and having the component values and conditions in Table 8. The differential amplifier always biases the output signal to sit on the optimum common mode of VREF/2, in this case 2.048 V. The signal is also scaled to give the maximum allowable voltage swing with this reference value. This is calculated as 80% of VREF, that is, 0.8 × 4.096 V ≈ 3.275 V p-p on each input. To obtain maximum performance from the AD7762, it is advisable to drive the ADC with differential signals. Figure 27 shows how a bipolar, single-ended signal biased around ground can drive the AD7762 with the use of an external op amp, such as the AD8021. CS1 SS1 SH3 CPA SH1 CPB1 SS3 ANALOG MODULATOR CS2 SS2 SH4 SH2 CPB2 SS4 05477-043 A Figure 28. Equivalent Input Circuit The AD7762 employs a double sampling front end, as shown in Figure 28. For simplicity, only the equivalent input circuit for VIN+ is shown. The equivalent input circuitry for VIN− is the same. With a 4.096 V reference, a 5 V supply must be provided to the reference buffer (AVDD4). With a 2.5 V reference, a 3.3 V supply must be provided to AVDD4. Rev. 0 | Page 16 of 28 AD7762 The sampling switches SS1 and SS3 are driven by ICLK, whereas the sampling switches SS2 and SS4 are driven by ICLK. When ICLK is high, the analog input voltage is connected to CS1. On the falling edge of ICLK, the SS1 and SS3 switches open, and the analog input is sampled on CS1. Similarly, when ICLK is low, the analog input voltage is connected to CS2. On the rising edge of ICLK, the SS2 and SS4 switches open, and the analog input is sampled on CS2. Capacitors CPA, CPB1, and CPB2 represent parasitic capacitances that include the junction capacitances associated with the MOS switches. Table 9. Equivalent Component Values Mode Normal Low Power CS1 51 pF 13 pF CS2 51 pF 13 pF CPA 12 pF 12 pF CPB1/2 20 pF 5 pF USING THE AD7762 Data can then be read from the part using the default filter, offset, gain, and overrange threshold values. The conversion data read is not valid, however, until the group delay of the filter has passed. When this has occurred, the DVALID bit read with the data LSW is set, indicating that the data is indeed valid. The user can then download a different filter, if required (see Downloading a User-Defined Filter). Values for gain, offset, and overrange threshold registers can be written or read at this stage. BIAS RESISTOR SELECTION The AD7762 requires a resistor to be connected between the RBIAS pin and AGND1. The value for this resistor is dependant on the reference voltage being applied to the device. The resistor value should be selected to give a current of 25 μA through the resistor to ground. For a 2.5 V reference voltage, the correct resistor value is 100 kΩ and for a 4.096 V reference, the correct resistor value is 160 kΩ. The following is the recommended sequence for powering up and using the AD7762. 1. Apply power. 2. Start the clock oscillator, applying MCLK. 3. Take RESET low for a minimum of 1 MCLK cycle. 4. Wait a minimum of 2 MCLK cycles after RESET has been released. 5. Write to Control Register 2 to power up the ADC and the differential amplifier as required. The correct clock divider (CDIV) ratio should be programmed now. 6. Write to Control Register 1 to set the output data rate. 7. Wait a minimum of 5 MCLK cycles after CS has been released. 8. Take SYNC low for a minimum of 4 MCLK cycles, if required, to synchronize multiple parts. Rev. 0 | Page 17 of 28 AD7762 DECOUPLING AND LAYOUT RECOMMENDATIONS PIN 41 24 27 44 63 41 AVDD3 AVDD2 VDRIVE VDRIVE DVDD PIN 44 PIN 63 6 33 AVDD1 AVDD1 PIN 24 PIN 27 12 AVDD4 PIN 6 PIN 33 4 14 15 19 VINA+ 20 VINA– 21 VOUTA– 22 VOUTA+ U2 DB (0:15) 8 DECAPA 30 DECAPB C7 100nF 25 VIN+ 26 VIN– VIN+ VIN– C64 33pF AD7762BSV 10 VREF+ 9 REFGND AVDD2 DGND DGND DGND DGND DGND DGND DGND AVDD4 L1 PIN 4 (RHS) C48 100nF L3 PIN 15 (VBIAS) C50 100nF PIN 12 (VBUF) L2 PIN 14 (LHS) C62 100nF L9 PIN 27 AVDD3 RESET SYNC DRDY 3 2 MCLK MCLKGND 11 AGND4 AGND3 AGND3 AGND3 AGND3 23 29 31 32 AVDD1 CS RD/WR 37 36 38 RESET SYNC DRDY AGND2 AGND2 AGND2 AGND2 AGND2 R19 160kΩ 1 35 42 43 53 62 64 DB0 DB1 DB2 DB3 DB4 DB5 DB6 DB7 DB8 DB9 DB10 DB11 DB12 DB13 DB14 DB15 40 39 CS RD/WR RBIAS 5 13 16 18 28 17 7 AGND1 34 AGND1 VREF 61 60 59 58 57 56 55 54 52 51 50 49 48 47 46 45 DB0 DB1 DB2 DB3 DB4 DB5 DB6 DB7 DB8 DB9 DB10 DB11 DB12 DB13 DB14 DB15 MCLK VDRIVE DVDD L4 L5 R38 10Ω C59 10nF PIN 5 (VMOD1) C52 100nF L11 PIN 33 (VMOD2) L6 PIN 24 (VDIF1) C53 100nF Figure 29. Simplified Connection Diagram Rev. 0 | Page 18 of 28 C54 100nF L7 PIN 44 (VDRV1) C56 100nF L12 PIN 63 (VDRV2) C57 100nF L8 PIN 41 (DVDD) C58 100nF 05477-046 INA+ INA– OUTA– OUTA+ AVDD2 AVDD2 AVDD2 PIN 4 PIN 14 PIN 15 PIN 12 Due to the high performance nature of the AD7762, correct decoupling and layout techniques are required to obtain the performance as stated within this datasheet. Figure 29 shows a simplified connection diagram for the AD7762. AD7762 SUPPLY DECOUPLING DIFFERENTIAL AMPLIFIER COMPONENTS Every supply pin must be connected to the appropriate supply via a ferrite bead and decoupled to the correct ground pin with a 100 nF, 0603 case size, X7R dielectric capacitor. There are two exceptions to this: The correct components for use around the on-chip differential amplifier are detailed in Table 8. Matching the components on both sides of the differential amplifier is important to minimize distortion of the signal applied to the amplifier. A tolerance of 0.1% or better is required for these components. Symmetrical routing of the tracks on both sides of the differential amplifier also assists in achieving stated performance. • Pin 12 (AVDD4) must have a 10 Ω resistor inserted between the pin and a 10 nF decoupling capacitor. • Pin 27 (AVDD2) does not require a separate decoupling capacitor or a direct connection to the supply, but instead is connected to Pin 14 via a 15 nH inductor. LAYOUT CONSIDERATIONS While using the correct components is essential to achieve optimum performance, the correct layout is just as important. The Design Tools section of the AD7762 product page on the Analog Devices website contains the gerber files for the AD7762 evaluation board. These files should be used as a reference when designing any system using the AD7762. ADDITIONAL DECOUPLING There are two other decoupling pins on the AD7762—Pin 8 (DECAPA) and Pin 30 (DECAPB). Pin 8 should be decoupled with a 100 nF capacitor, and Pin 30 requires a 33 pF capacitor. The location and orientation of some of the components mentioned in previous sections is critical, and particular attention must be paid to the components which are located close to the AD7762. Locating these components further away from the devices can have a direct impact on the maximum performance achievable. REFERENCE VOLTAGE FILTERING A low noise reference source, such as the ADR431 (2.5 V) or ADR434 (4.096 V), is suitable for use with the AD7762. The reference voltage supplied to the AD7762 should be decoupled and filtered, as shown in Figure 30. The recommended scheme for the reference voltage supply is a 100 Ω series resistor connected to a 100 μF tantalum capacitor, followed by series resistor of 10 Ω, and finally a 10 nF decoupling capacitor very close to the VREF+ pin. U3 2 +12V C15 10μF + C9 100nF +VIN VOUT GND 4 6 C10 100nF R30 100Ω R17 10Ω PIN10 + C11 100μF C46 10nF 04975-047 ADR434 The use of ground planes also should be carefully considered. To ensure that the return currents through the decoupling capacitors are flowing to the correct ground pin, the ground side of the capacitors should be as close to the ground pin associated with that supply. A ground plane should not be relied on as the sole return path for decoupling capacitors because the return current path using ground planes is not easily predictable. Figure 30. Reference Connection Rev. 0 | Page 19 of 28 AD7762 PROGRAMMABLE FIR FILTER No. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Dec. Value 53656736 25142688 −4497814 −11935847 −1313841 6976334 3268059 −3794610 −3747402 1509849 3428088 80255 −2672124 −1056628 1741563 1502200 −835960 −1528400 93626 1269502 411245 −864038 −664622 434489 Hex Value 332BCA0 17FA5A0 444A196 4B62067 4140C31 6A734E 31DDDB 439E6B2 4392E4A 1709D9 344EF8 1397F 428C5FC 4101F74 1A92FB 16EBF8 40CC178 4175250 16DBA 135EFE 6466D 40D2F26 40A242E 6A139 No. 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 Dec. Value 700847 −70922 −583959 −175934 388667 294000 −183250 −302597 16034 238315 88266 −143205 −128919 51794 121875 16426 −90524 −63899 45234 114720 102357 52669 15559 1963 Hex Value AB1AF 401150A 408E917 402AF3E 5EE3B 47C70 402CBD2 4049E05 3EA2 3A2EB 158CA 4022F65 401F797 CA52 1DC13 402A 401619C 400F99B B0B2 1C020 18FD5 CDBD 3CC7 7AB The default filter should be sufficient for almost all applications. It is a standard brick wall filter with a symmetrical impulse response. The default filter has a length of 96 taps in nonaliasing with 120 dB of attenuation at Nyquist. This filter not only performs signal antialiasing, but also suppresses out-ofband quantization noise produced by the analog-to-digital conversion process. Any significant relaxation in the stop-band attenuation or transition bandwidth relative to the default filter can result in a failure to meet the SNR specifications. • The filter must be even, symmetrical FIR. • The coefficients are in sign-and-magnitude format with 26 magnitude bits and sign coded as positive = 0. • The filter length must be between 12 taps and 96 taps in steps of 12. • Because the filter is symmetrical, the number of coefficients that must be downloaded is half the filter length. The default filter coefficients exemplify this with only 48 coefficients listed for a 96-tap filter. • Coefficients are written from the center of impulse response (adjacent to the point of symmetry) outwards. • The coefficients are scaled so that the in-band gain of the filter is equal to 134217726 with the coefficients rounded to the nearest integer. For a low-pass filter, this is the equivalent of having the coefficients sum arithmetically (including sign) to a 67108863 (0x3FF FFFF) positive value over the half-impulse response coefficient set (maximum 48 coefficients). Any deviation from this introduces a gain error. 0 PASS-BAND RIPPLE = 0.05dB –0.1dB FREQUENCY = 251kHz –3dB FREQUENCY = 256kHz STOP BAND = 312.5kHz –20 –40 –60 –80 –100 –120 –140 05477-044 Table 10. Default Filter Coefficients To create a filter, note the following: AMPLITUDE (dB) As previously mentioned, the third FIR filter on the AD7762 is user programmable. The default coefficients that are loaded on reset are given in Table 10 and the frequency responses are shown in Figure 31. The frequencies quoted in Figure 31 scale directly with the output data rate. –160 0 100 200 300 400 500 600 FREQUENCY (kHz) Figure 31. Default Filter Frequency Response (625 kHz ODR) The procedure for downloading a user-defined filter is detailed in the Downloading a User-Defined Filter section. Rev. 0 | Page 20 of 28 AD7762 DOWNLOADING A USER-DEFINED FILTER EXAMPLE FILTER DOWNLOAD The following is an example of downloading a short userdefined filter with 24 taps. The frequency response is shown in Figure 32. 10 0 –10 AMPLITUDE (dB) –20 –30 –40 –50 –60 To download a user filter: –70 1. –80 Write to Control Register 1, setting the DL_Filt bit and also the correct filter length bits corresponding to the length of the filter to be downloaded (see Table 11). 0 200 300 400 500 600 Figure 32. 24-Tap FIR Frequency Response Write the first half of the current coefficient data (00000, Sign bit, Magnitude [25:16]). The first coefficient to be written must be the one adjacent to the point of filter symmetry. 3. Write the second half of the current coefficient data (Magnitude [15:0]). The coefficients for the filter are listed in Table 12 and are shown from the center of symmetry outwards. The raw coefficients were generated using a commercial filter design tool and scaled appropriately so their sum equals 67108863 (0x3FF FFFF). Table 12. 24-Tap FIR Coefficients 4. Repeat Step 2 and Step 3 for each coefficient. 5. Write the 16-bit checksum. 6. Use these methods to verify that the filter coefficients are downloaded correctly: a. Read the status register, checking the DL_OK bit. b. Read data and observe the status of the DL_OK bit. Note that because the user coefficients are stored in RAM, they are cleared after a RESET operation or a loss of power. Table 11. Filter Length Values Number of Coefficients Default 6 12 18 24 30 36 42 48 100 FREQUENCY (kHz) 2. FLEN[3:0] 0000 0001 0011 0101 0111 1001 1011 1101 1111 04975-045 As previously mentioned, the filter coefficients are 27 bits in length; 1 sign and 26 magnitude bits. Because the AD7762 has a 16-bit parallel bus, the coefficients are padded with 5 MSB 0s to generate a 32-bit word and split into two 16-bit words for downloading. The first 16-bit word for each coefficient becomes (00000, Sign bit, Magnitude [25:16]), while the second word becomes (Magnitude [15:0]). To ensure that a filter is down-loaded correctly, a checksum must also be generated and then downloaded following the final coefficient. The checksum is a 16-bit word generated by splitting each 32-bit word into 4 bytes and summing all bytes from all coefficients up to a maximum of 192 bytes (48 coefficients × 4 bytes). The same checksum is generated internally in the AD7762 and compared with the checksum downloaded. The DL_OK bit in the status register is set if these two checksums agree. Coefficient 1 2 3 4 5 6 7 8 9 10 11 12 Filter Length Default 12 24 36 48 60 72 84 96 Rev. 0 | Page 21 of 28 Raw 0.365481974 0.201339905 0.009636604 −0.075708848 −0.042856209 0.019944246 0.036437914 0.007592007 −0.021556583 −0.024888355 −0.012379538 −0.001905756 Scaled 53188232 29300796 1402406 −11017834 −6236822 2902466 5302774 1104856 −3137108 −3621978 −1801582 −277343 AD7762 Table 13 shows the hex values (in sign and magnitude format) that are downloaded to the AD7762 to realize this filter. The table is also split into the bytes that are all summed to produce the checksum. The checksum generated from these coefficients is 0x0E6B. Table 13. Filter Hex Values Coefficient 1 2 3 4 5 6 7 8 9 10 11 12 Word 1 Byte 1 Byte 2 03 2B 01 BF 00 15 04 A8 04 5F 00 2C 00 50 00 10 04 2F 04 37 04 1B 04 04 Word 2 Byte 3 Byte 4 96 88 18 3C 66 26 1E 6A 2A 96 49 C2 E9 F6 DB D8 DE 54 44 5A 7D 6E 3B 5F Table 14 lists the 16-bit words the user would write to the AD7762 to set up the ADC and download this filter, assuming an output data rate of 625 kHz has already been selected. Table 14. Word 0x0001 0x8079 0x032B 0x9688 0x01BF 0x183C … 0x0404 0x3B5F 0x0E6B 0x0001 0x0879 Rev. 0 | Page 22 of 28 Description Address of Control Register 1. Control register data. DL filter, set filter length = 24, set output data rate = 625 kHz. First coefficient, Word 1. First coefficient, Word 2. Second coefficient, Word 1. Second coefficient, Word 2. Other coefficients. Twelfth (final) coefficient, Word 1. Final coefficient, Word 2. Checksum. Wait (0.5 × tICLK × Number of Unused Coefficients) for AD7762 to fill remaining unused coefficients with 0s. Address of control register. Control register data. Set read status and maintain filter length and decimation settings. Read contents of status register. Check Bit 7 (DL_OK) to determine that the filter was downloaded correctly. AD7762 AD7762 REGISTERS The AD7762 has a number of user-programmable registers. The control registers are used to set the decimation rate, the filter configuration, the clock divider, and so on. There are also digital gain, offset, and overrange threshold registers. Writing to these registers involves writing the register address first, then a 16-bit data-word. Register addresses, details of individual bits, and default values are given here. CONTROL REGISTER 1—REG 0X0001 Default Value 0x001A MSB DL_ Filt RD Ovr RD Gain RD Off RD Stat 0 SYNC FLEN3 FLEN2 FLEN1 FLEN0 BYP F3 1 DEC2 DEC1 LSB DEC0 Table 15. Bit 15 Mnemonic DL_Filt 1 14 RD Ovr1, 2 13 12 11 10 9 RD Gain1, 2 RD Off1, 2 RD Stat1, 2 0 SYNC1 8-5 4 3 2-0 FLEN3:0 BYP F3 1 DEC2:0 1 2 Description Download Filter. Before downloading a user-defined filter, this bit must be set. The Filter Length bits must also be set at this time. The write operations that follow are interpreted as the user coefficients for the FIR filter until all the coefficients and the checksum have been written. Read Overrange. If this bit has been set, the next read operation outputs the contents of the Overrange Threshold Register instead of a conversion result. Read Gain. If this bit has been set, the next read operation outputs the contents of the digital gain register. Read Offset. If this bit has been set, the next read operation outputs the contents of the digital offset register. Read Status. If this bit has been set, the next read operation outputs the contents of the status register. 0 must be written to this bit. Synchronize. Setting this bit initiates an internal synchronization routine. Setting this bit simultaneously on multiple devices synchronizes all filters. Filter Length Bits. These bits must be set when the DL Filt bit is set and before a user-defined filter is downloaded. Bypass Filter 3. If this bit is 0, Filter 3 (programmable FIR) is bypassed. 1 must be written to this bit. Decimation Rate. These bits set the decimation rate of Filter 2. All 0s implies that the filter is bypassed. A value of 1 corresponds to 2× decimation, a value of 2 corresponds to 4× decimation, and so on up to the maximum value of 5, corresponding to 32× decimation. Bit 15 to Bit 9 are all self clearing bits. Only one of the bits from Bit 14 to Bit 11 can be set in any write operation because it determines the contents of the next read operation. CONTROL REGISTER 2—ADDRESS 0X0002 Default Value 0x009B MSB 0 0 0 0 0 0 0 0 0 0 CDIV 0 PD LPWR 1 LSB D1PD Table 16. Bit 5 Mnemonic CDIV 3 2 PD LPWR 1 0 1 D1PD Description Clock Divider Bit. This sets the divide ratio of the MCLK signal to produce the internal ICLK. Setting CDIV = 0 divides the MCLK by 2. If CDIV = 1, then the ICLK frequency is equal to the MCLK. Power Down. Setting this bit powers down the AD7762, reducing the power consumption to 6.35 mW. Low Power. If this bit is set, the AD7762 is operating in a low power mode. The power consumption is reduced for a 6 dB reduction in noise performance. Write 1 to this bit. Differential Amplifier Power Down. Setting this bit powers down the on-chip differential amplifier. Rev. 0 | Page 23 of 28 AD7762 STATUS REGISTER (READ ONLY) MSB PART 1 PART 0 DIE 2 DIE 1 DIE 0 DVALID LPWR OVR DL OK Filter OK U Filter BYP F3 1 DEC2 DEC1 LSB DEC0 Table 17. Bit 15, 14 13 to 11 10 9 8 7 Mnemonic PART1:0 DIE2:0 DVALID LPWR OVR DL OK 6 Filter OK 5 4 3 2-0 U Filter BYP F3 1 DEC2:0 Comment Part Number. These bits are constant for the AD7762. Die Number. These bits reflect the current AD7762 die number for identification purposes within a system. Data Valid. This bit corresponds to the DVALID bit in the status word output in the second 16-bit read operation. Low Power. If the AD7762 is operating in low power mode, this bit is set to 1. If the current analog input exceeds the current overrange threshold, this bit is set. When downloading a user filter to the AD7762, a checksum is generated. This checksum is compared to the one downloaded following the coefficients. If these checksums agree, this bit is set. When a user-defined filter is in use, a checksum is generated when the filter coefficients pass through the filter. This generated checksum is compared to the one downloaded. If they match, this bit is set. If a user-defined filter is in use, this bit is set. Bypass Filter 3. If Filter 3 is bypassed by setting the relevant bit in Control Register 1, this bit is also set. This bit is always set. Decimation Rate. These correspond to the bits set in Control Register 1. OFFSET REGISTER—ADDRESS 0X0003 Non-bitmapped, Default Value 0x0000 The offset register uses twos complement notation and is scaled such that 0x7FFF (maximum positive value) and 0x8000 (maximum negative value) correspond to an offset of +0.78125% and −0.78125%, respectively. Offset correction is applied after any gain correction. Using the default gain value of 1.25 and assuming a reference voltage of 4.096V, the offset correction range is approximately ±25 mV. GAIN REGISTER—ADDRESS 0X0004 Non-bitmapped, Default Value 0xA000 The gain register is scaled such that 0x8000 corresponds to a gain of 1.0. The default value of this register is 1.25 (0xA000). This gives a full-scale digital output when the input is at 80% of VREF. This ties in with the maximum analog input range of ±80% of VREF p-p. OVERRANGE REGISTER—ADDRESS 0X0005 Non-bitmapped, Default Value 0xCCCC The overrange register value is compared with the output of the first decimation filter to obtain an overload indication with minimum propagation delay. This is prior to any gain scaling or offset adjustment. The default value is 0xCCCC which corresponds to 80% of VREF (the maximum permitted analog input voltage). Assuming VREF = 4.096 V, the bit is then set when the input voltage exceeds approximately 6.55 V p-p differential. Note that the overrange bit is also set immediately if the analog input voltage exceeds 100% of VREF for more than four consecutive samples at the modulator rate. Rev. 0 | Page 24 of 28 AD7762 OUTLINE DIMENSIONS 0.75 0.60 0.45 12.20 12.00 SQ 11.80 1.20 MAX 64 49 1 49 64 1 48 48 PIN 1 10.20 10.00 SQ 9.80 TOP VIEW (PINS DOWN) 7.50 BSC SQ EXPOSED PAD 0° MIN 1.05 1.00 0.95 0.15 0.05 SEATING PLANE 0.20 0.09 7° 3.5° 0° 0.08 MAX COPLANARITY BOTTOM VIEW (PINS UP) 16 33 17 32 VIEW A 16 33 32 0.50 BSC LEAD PITCH VIEW A 17 0.38 0.32 0.22 ROTATED 90° CCW COMPLIANT TO JEDEC STANDARDS MS-026-ACD-HD Figure 33. 64-Lead Thin Quad Flat Package, Exposed Pad [TQFP_EP] (SV-64-4) Dimensions shown in millimeters ORDERING GUIDE Model AD7762BSVZ 1 AD7762BSVZ-REEL1 EVAL-AD7762EB 1 Temperature Range −40°C to +85°C −40°C to +85°C Package Description 64-Lead Thin Quad Flat Package, Exposed Pad (TQFP_EP) 64-Lead Thin Quad Flat Package, Exposed Pad (TQFP_EP) Evaluation Board Z = Pb-free part. Rev. 0 | Page 25 of 28 Package Option SV-64-4 SV-64-4 AD7762 NOTES Rev. 0 | Page 26 of 28 AD7762 Rev. 0 | Page 27 of 28 AD7762 NOTES © 2005 Analog Devices, Inc. All rights reserved. Trademarks and registered trademarks are the property of their respective owners. D05477–0–8/05(0) Rev. 0 | Page 28 of 28