AD AD7762BSVZ-REEL

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