2.

5 MSPS, 24-Bit, 100 dB

Sigma-Delta ADC with On-Chip Buffer
AD7760
FEATURES FUNCTIONAL BLOCK DIAGRAM
120 dB dynamic range at 78 kHz output data rate VIN– VIN+

100 dB dynamic range at 2.5 MHz output data rate

112 dB SNR at 78 kHz output data rate DIFF
MULTIBIT AVDD1
Ȉ-ǻ
100 dB SNR at 2.5 MHz output data rate MODULATOR AVDD2

AVDD3
2.5 MHz maximum fully filtered output word rate VREF+
BUF RECONSTRUCTION AVDD4
Programmable oversampling rate (8× to 256×)
DECAPA/B
Fully differential modulator input RBIAS
On-chip differential amplifier for signal buffering AD7760 PROGRAMMABLE AGND
DECIMATION
Low-pass finite impulse response (FIR) filter with default or VDRIVE
user-programmable coefficients MCLK
CONTROL LOGIC DVDD
I/O
Modulator output mode SYNC OFFSET AND GAIN DGND
REGISTERS
Overrange alert bit RESET FIR FILTER
ENGINE
Digital offset and gain correction registers

04975-001
Filter bypass modes
CS RD/WR DRDY DB0 TO DB15
Low power and power-down modes
Figure 1.
Synchronization of multiple devices via SYNC pin

APPLICATIONS
Data acquisition systems
Vibration analysis
Instrumentation

GENERAL DESCRIPTION
The AD7760 is a high performance, 24-bit Σ-Δ analog-to-digital coefficients. The sample rate, filter corner frequencies, and output
converter (ADC). It combines wide input bandwidth and high word rate are set by a combination of the external clock frequency
speed with the benefits of Σ-Δ conversion to achieve a perfor- and the configuration registers of the AD7760.
mance of 100 dB SNR at 2.5 MSPS, making it ideal for high
speed data acquisition. Wide dynamic range combined with The reference voltage supplied to the AD7760 determines the
significantly reduced antialiasing requirements simplify the analog input range. With a 4 V reference, the analog input range
design process. An integrated buffer to drive the reference, a is ±3.2 V differential biased around a common mode of 2 V.
differential amplifier for signal buffering and level shifting, an This common-mode biasing can be achieved using the on-chip
overrange flag, internal gain and offset registers, and a low-pass differential amplifier, further reducing the external signal
digital FIR filter make the AD7760 a compact, highly integrated conditioning requirements.
data acquisition device requiring minimal peripheral component The AD7760 is available in an exposed paddle, 64-lead TQFP
selection. In addition, the device offers programmable decimation and is specified over the industrial temperature range from
rates, and the digital FIR filter can be adjusted if the default −40°C to +85°C.
characteristics are not appropriate for the application. The
AD7760 is ideal for applications demanding high SNR Table 1. Related Devices
without a complex front-end signal processing design. Part No. Description
AD7762 24-bit, 625 kSPS, 109 dB, Σ-∆ parallel interface
The differential input is sampled at up to 40 MSPS by an analog AD7763 24-bit, 625 kSPS, 109 dB, Σ-∆ serial interface
modulator. The modulator output is processed by a series of low-
pass filters, with the final filter having default or user-programmable

Rev. A
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 One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
license is granted by implication or otherwise under any patent or patent rights of Analog Devices. Tel: 781.329.4700 www.analog.com
Trademarks and registered trademarks are the property of their respective owners. Fax: 781.461.3113 ©2006 Analog Devices, Inc. All rights reserved.
AD7760

TABLE OF CONTENTS
Features .............................................................................................. 1 Writing to the AD7760 .............................................................. 23

Applications....................................................................................... 1 Clocking the AD7760 .................................................................... 24

Functional Block Diagram .............................................................. 1 Buffering the MCLK signal....................................................... 24

General Description ......................................................................... 1 MCLK Jitter Requirements ....................................................... 24

Revision History ............................................................................... 3 Driving the AD7760....................................................................... 26

Specifications..................................................................................... 4 Using the AD7760 ...................................................................... 27

Timing Specifications .................................................................. 6 Decoupling and Layout Recommendations................................ 28

Timing Diagrams.......................................................................... 7 Supply Decoupling ..................................................................... 29

Absolute Maximum Ratings............................................................ 8 Additional Decoupling .............................................................. 29

ESD Caution.................................................................................. 8 Reference Voltage Filtering ....................................................... 29

Pin Configuration and Function Descriptions............................. 9 Differential Amplifier Components ........................................ 29

Terminology .................................................................................... 11 Bias Resistor Selection ............................................................... 29

Typical Performance Characteristics ........................................... 12 Layout Considerations............................................................... 29

Theory of Operation ...................................................................... 18 Exposed Paddle........................................................................... 29

Modulator Data Output Mode...................................................... 19 Programmable FIR Filter............................................................... 30

Modulator Inputs........................................................................ 19 Downloading a User-Defined Filter ............................................ 31

Modulator Data Output Scaling ............................................... 19 Example Filter Download ......................................................... 31

Modulator Data Output Mode Interface ..................................... 20 AD7760 Registers ........................................................................... 33

Clock Divide-by-1 Mode (CDIV = 1) ..................................... 20 Control Register 1—Address 0x0001 ...................................... 33

Clock Divide-by-2 Mode (CDIV = 0) ..................................... 20 Control Register 2—Address 0x0002 ...................................... 33

Using the AD7760 in Modulator Output Mode..................... 21 Status Register (Read Only) ...................................................... 34

AD7760 Interface............................................................................ 22 Offset Register—Address 0x0003............................................. 34

Reading Data............................................................................... 22 Gain Register—Address 0x0004............................................... 34

Reading Status and Other Registers......................................... 22 Overrange Register—Address 0x0005..................................... 34

Sharing the Parallel Bus ............................................................. 22 Outline Dimensions ....................................................................... 35

Synchronization.......................................................................... 22 Ordering Guide .......................................................................... 35

Rev. A | Page 2 of 36
 AD7760
REVISION HISTORY
8/06—Rev. 0 to Rev. A
Updated Package Option................................................... Universal Added MCLK Jitter Requirements Heading ...............................24
Change to Features............................................................................1 Changes to Driving the AD7760 Section.....................................26
Changes to Specifications.................................................................4 Changes to Figure 51 ......................................................................26
Changes to Absolute Maximum Ratings........................................8 Added Figure 52 ..............................................................................26
Changes to Terminology Section ..................................................11 Changes to Figure 55 ......................................................................28
Added Figure 36 Through Figure 39 ............................................17 Changes to Figure 56 ......................................................................29
Added Modulator Data Output Mode Section ...........................19 Added Exposed Paddle Section.....................................................29
Added Figure 41 Through Figure 47 ............................................19 Change to Control Register 2—Address 0x0002 Section ..........33
Added Modulator Data Output Mode Interface Section...........20 Changes to Status Register (Read Only) Section ........................34
Changes to Reading Data Section.................................................22
Added Synchronization Section....................................................22 7/05—Revision 0: Initial Version
Changes to Clocking the AD7760 Section...................................24
Added Buffering the MCLK Signal Section.................................24

Rev. A | Page 3 of 36
AD7760

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
the on-chip amplifier with components as shown in Table 8, unless otherwise noted. 1
Table 2.
Parameter Test Conditions/Comments Specification Unit
DYNAMIC PERFORMANCE
Decimate by 256 MCLK = 40 MHz, ODR = 78 kHz, fIN = 1 kHz
Dynamic Range Modulator inputs shorted 119 dB min
120.5 dB typ
Signal-to-Noise Ratio (SNR) 2 Input amplitude = −0.5 dBFS 112 dB typ
Input amplitude = −60 dBFS 59 dB typ
Spurious-Free Dynamic Range (SFDR) Nonharmonic, input amplitude = −6 dBFS 126 dBc typ
Input amplitude = −60 dBFS 77 dBc typ
Total Harmonic Distortion (THD) Input amplitude = −0.5 dBFS −105 dB typ
Input amplitude = −6 dBFS −106 dB typ
Input amplitude = −60 dBFS −75 dB typ
Decimate by 32 MCLK = 40 MHz, ODR = 625 kHz, fIN =100 kHz
Dynamic Range Modulator inputs shorted 108 dB min
109.5 dB typ
Signal-to-Noise Ratio (SNR)2 Input amplitude = −0.5 dBFS 107 dB typ
Spurious-Free Dynamic Range (SFDR) Nonharmonic, input amplitude = −6 dBFS 120 dBc typ
Total Harmonic Distortion (THD) Input amplitude = −0.5 dBFS −105 dB typ
Input amplitude = −6 dBFS −106 dB typ
Decimate by 8 MCLK = 40 MHz, ODR = 2.5 MHz
Dynamic Range Modulator inputs shorted 99 dB min
100.5 dB typ
Signal-to-Noise Ratio (SNR)2 fIN = 1 kHz, input amplitude = −0.5 dBFS 100 dB typ
fIN = 100 kHz, input amplitude = −0.5 dBFS 99 dB typ
fIN = 1 MHz, input amplitude = −0.5 dBFS 98 dB typ
Spurious-Free Dynamic Range (SFDR) Nonharmonic, fIN = 100 kHz, input amplitude = −6 dBFS 120 dBc typ
Nonharmonic, fIN = 1 MHz, input amplitude = −6 dBFS 114 dBc typ
Total Harmonic Distortion (THD) Input amplitude = −0.5 dBFS, fIN = 100 kHz −103 dB typ
Input amplitude = −6 dBFS, fIN = 100 kHz −102 dB typ
IMD Second Order fIN A = 989.95 kHz, fIN B = 999.95 kHz −115 dB typ
IMD Third Order fIN A = 989.95 kHz, fIN B = 999.95 kHz −89 dB typ
DC ACCURACY
Resolution 24 Bits
Differential Nonlinearity Guaranteed monotonic to 24 bits
Integral Nonlinearity 0.00076 % typ
Zero Error 0.014 % typ
0.02 % max
Gain Error 0.016 % typ
Zero Error Drift 0.00001 % FS/°C typ
Gain Error Drift 0.0002 % FS/°C typ
DIGITAL FILTER RESPONSE
Decimate by 8
Group Delay MCLK = 40 MHz 12 µs typ
Decimate by 32
Group Delay MCLK = 40 MHz 47 µs typ
Decimate by 256
Group Delay MCLK = 40 MHz 358 µs typ

Rev. A | Page 4 of 36
 AD7760
Parameter Test Conditions/Comments Specification Unit
ANALOG INPUT
Differential Input Voltage VIN(+) – VIN(−), VREF = 2.5 V ±2 V p-p
VIN(+) – VIN(−), VREF = 4.096 V ±3.25 V p-p
Input Capacitance At internal buffer inputs 5 pF typ
At modulator inputs 55 pF typ
REFERENCE INPUT/OUTPUT
VREF Input Voltage VDD3 = 3.3 V ± 5% +2.5 V max
VDD3 = 5 V ± 5% +4.096 V max
VREF Input DC Leakage Current ±6 µA max
VREF Input Capacitance 5 pF max
POWER DISSIPATION
Total Power Dissipation Normal mode 958 mW max
Low power mode 661 mW max
Standby Mode Clock stopped 6.35 mW max
POWER REQUIREMENTS
AVDD1 (Modulator Supply) ±5% +2.5 V
AVDD2 (General Supply) ±5% +5 V
AVDD3 (Differential Amplifier Supply) +3.15/+5.25 V min/max
AVDD4 (Reference Buffer Supply) +3.15/+5.25 V min/max
DVDD ±5% +2.5 V
VDRIVE +1.65/+2.7 V min/max
Normal Mode
AIDD1 (Modulator) 49/51 mA typ/max
AIDD2 (General) 3 40/42 mA typ/max
AIDD4 (Reference Buffer) AVDD4 = 5 V 34/36 mA typ/max
Low Power Mode
AIDD1 (Modulator) 26/28 mA typ/max
AIDD2 (General)3 20/23 mA typ/max
AIDD4 (Reference Buffer) AVDD4 = 5 V 9/10 mA typ/max
AIDD3 (Differential Amplifier) AVDD3 = 5 V, both modes 41/44 mA typ/max
DIDD Both modes 63/70 mA typ/max
DIGITAL I/O
MCLK Input Amplitude 4 5 V typ
Input Capacitance 7.3 pF typ
Input Leakage Current ±5 μA max
Three-State Leakage Current (D15:D0) ±5 μA max
VINH 0.7 × VDRIVE V min
VINL 0.3 × VDRIVE V max
VOH 5 1.5 V min
VOH 6 2.4 V typ
VOL4 0.1 V max
1
See the Terminology section.
2
SNR specifications in decibels are referred to a full-scale input, FS. Tested with an input signal at 0.5 dB below full scale, unless otherwise specified.
3
Current scales with ICLK frequency. See the Typical Performance Characteristics section.
4
Although the AD7760 can function with an MCLK amplitude of less than 5 V, this is the recommended amplitude to achieve the performance as stated.
5
Tested using the minimum VDRIVE voltage of 1.65 V with a 400 µA load current.
6
Tested using VDRIVE = 2.5 V with a 400 μA load current.

Rev. A | Page 5 of 36
AD7760
TIMING SPECIFICATIONS
AVDD1 = DVDD = VDRIVE = 2.5 V, AVDD2 = AVDD3 = AVDD4 = 5 V, TA = 25°C, normal mode, unless otherwise noted.
Table 3.
Parameter Limit at TMIN, TMAX Unit Description
fMCLK 1 MHz min Applied master clock frequency
40 MHz max
fICLK 500 kHz min Internal modulator clock derived from MCLK
20 MHz max
t1 1, 2 0.5 × tICLK typ DRDY pulse width
t2 10 ns min DRDY falling edge to CS falling edge
t3 3 ns min RD/WR setup time to CS falling edge
t4 (0.5 × tICLK) + 16 ns max Data access time
t5 tICLK min CS low read pulse width
t6 tICLK min CS high pulse width between reads
t7 3 ns min RD/WR hold time to CS rising edge
t8 11 ns max Bus relinquish time
t92 0.5 × tICLK typ DRDY high period
t102 0.5 × tICLK typ DRDY low period
t11 (0.5 × tICLK) + 16 ns max Data access time
t12 3, 4 23 ns min Data valid prior to DRDY rising edge
t133, 4 19 ns min Data valid after DRDY rising edge
t14 11 ns max Bus relinquish time
t15 4 × tICLK min CS low write pulse width
t16 4 × tICLK min CS high period between address and data
t17 5 ns min Data setup time
t18 0 ns min Data hold time
t194, 5 23 ns min Data valid prior to MCLK falling edge while DRDY is logic low
t204, 5 19 ns min Data valid after MCLK falling edge while DRDY is logic low
1
tICLK = 1/fICLK.
2
When ICLK = MCLK, DRDY pulse width depends on the mark-space ratio of applied MCLK.
3
Valid when using the modulator output mode with CDIV = 1.
4
See the Modulator Data Output Mode section for timing diagrams.
5
Valid when using the modulator output mode with CDIV = 0.

Rev. A | Page 6 of 36
 AD7760
TIMING DIAGRAMS

DRDY
t1 t5 t6

CS
t2
t3 t7

RD/WR

t4 t8

04975-002
D[0:15] DATA MSW LSW + STATUS

Figure 2. Filtered Output—Parallel Interface Timing Diagram

CS

t15 t16
RD/WR

t17 t18

04975-004
D[0:15] REGISTER ADDRESS REGISTER DATA

Figure 3. AD7760 Register Write

Rev. A | Page 7 of 36
AD7760

ABSOLUTE MAXIMUM RATINGS

TA = 25°C, unless otherwise noted.
Table 4.
Parameters Rating
Stresses above those listed under Absolute Maximum Ratings
may cause permanent damage to the device. This is a stress
AVDD1 to GND −0.3 V to +3 V
rating only; functional operation of the device at these or any
AVDD2:AVDD4 to GND −0.3 V to +6 V
other conditions above those listed in the operational sections
DVDD to GND −0.3 V to +3 V
of this specification is not implied. Exposure to absolute
VDRIVE to GND −0.3 V to +3 V
maximum rating conditions for extended periods may affect
VIN+, VIN– to GND1 −0.3 V to +6 V
device reliability.
VINA+, VINA− to GND1 −0.3 V to +6 V
Digital Input Voltage to GND2 −0.3 V to DVDD + 0.3 V
MCLK to MCLKGND −0.3 V to +6 V
VREF+ to GND3 −0.3 V to AVDD4 + 0.3 V ESD CAUTION
AGND to DGND −0.3 V to +0.3 V
Input Current to Any Pin Except ±10 mA
Supplies4
Operating Temperature Range
Commercial −40°C to +85°C
Storage Temperature Range −65°C to +150°C
Junction Temperature 150°C
TQFP Exposed Paddle Package
θJA Thermal Impedance 92.7°C/W
θJC Thermal Impedance 5.1°C/W
Lead Temperature, Soldering
Vapor Phase (60 sec) 215°C
Infrared (15 sec) 220°C
ESD 600 V
1
Absolute maximum voltage for VIN−, VIN+ and VINA−, VINA+ is 6.0 V or AVDD3 + 0.3 V,
whichever is lower.
2
Absolute maximum voltage on digital inputs is 3.0 V or DVDD + 0.3 V,
whichever is lower.
3
Absolute maximum voltage on VREF+ input is 6.0 V or AVDD4 + 0.3 V,
whichever is lower.
4
Transient currents of up to 200 mA do not cause SCR latch-up.

Rev. A | Page 8 of 36
 AD7760

PIN CONFIGURATION AND FUNCTION DESCRIPTIONS

VDRIVE
DGND

DGND

DGND

DB10
DB11
DB0
DB1
DB2
DB3
DB4
DB5
DB6
DB7

DB8
DB9
64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49

DGND 1 48 DB12
PIN 1
MCLKGND 2 47 DB13
MCLK 3 46 DB14
AVDD2 4 45 DB15
AGND2 5 44 VDRIVE
AVDD1 6 43 DGND
AGND1 7 42 DGND
AD7760
DECAPA 8 TOP VIEW 41 DVDD
REFGND 9 (Not to Scale) 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

17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
AVDD3

VIN–
AVDD2
VIN+
VINA+
VINA–
VOUTA–
VOUTA+
RBIAS
AGND2

AGND3

AGND2
AGND3

AGND3
AGND3
DECAPB

04975-005
Figure 4. 64-Lead TQFP Pin Configuration

Table 5. Pin Function Descriptions

Pin No. Mnemonic Description
6, 33 AVDD1 2.5 V Power Supply for Modulator. These pins should be decoupled to AGND1 (Pin 7 and Pin 34, respectively)
with 100 nF and 10 µF capacitors on each pin. See the Decoupling and Layout Recommendations section
for details.
4, 14, 15, 27 AVDD2 5 V Power Supply. These pins should be decoupled to AGND2 (Pin 5 and Pin 13, 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. See the
Decoupling and Layout Recommendations section for details.
24 AVDD3 3.3 V to 5 V Power Supply for Differential Amplifier. This pin should be decoupled to AGND3 (Pin 23) with a
100 nF capacitor. See the Decoupling and Layout Recommendations section for details.
12 AVDD4 3.3 V to 5 V Power Supply for Reference Buffer. This pin should be decoupled to Pin 9 with a 10 nF capacitor
in series with a 10 Ω resistor.
7, 34 AGND1 Power Supply Ground for Analog Circuitry Powered by AVDD1.
5, 13, 16, 18, 28 AGND2 Power Supply Ground for Analog Circuitry Powered by AVDD2.
23, 29, 31, 32 AGND3 Power Supply Ground for Analog Circuitry Powered by AVDD3.
11 AGND4 Power Supply Ground for Analog Circuitry Powered by AVDD4.
9 REFGND Reference Ground. Ground connection for the reference voltage.
41 DVDD 2.5 V Power Supply for Digital Circuitry and FIR Filter. This pin should be decoupled to DGND with a 100 nF
capacitor.
44, 63 VDRIVE 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 of these pins must be connected together and tied to the same supply.
Each pin should also be decoupled to DGND with a 100 nF capacitor.
1, 35, 42, 43, DGND Ground Reference for Digital Circuitry.
53, 62, 64
19 VINA+ Positive Input to Differential Amplifier.
20 VINA− Negative Input to Differential Amplifier.
21 VOUTA− Negative Output from Differential Amplifier.
22 VOUTA+ Positive Output from Differential Amplifier.
25 VIN+ Positive Input to the Modulator.
26 VIN− Negative Input to the Modulator.
10 VREF+ 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.
8 DECAPA Decoupling Pin. A 100 nF capacitor must be inserted between this pin and AGND.
Rev. A | Page 9 of 36
AD7760
Pin No. Mnemonic Description
30 DECAPB Decoupling Pin. A 33 pF capacitor must be inserted between this pin and AGND3.
17 RBIAS Bias Current Setting Pin. A resistor must be inserted between this pin and AGND. For more details, see the
Bias Resistor Selection section.
45 to 52, DB15:DB8, 16-Bit Bidirectional Data Bus. These are three-state pins that are controlled by the CS pin and the RD/WR
54 to 61 DB7:DB0 pin. The operating voltage for these pins is determined by the VDRIVE voltage. See the Modulator Data
Output Mode and AD7760 Interface sections for more details.
37 RESET A falling edge on this pin resets all internal digital circuitry and powers down the part. Holding this pin low
keeps the AD7760 in a reset state.
3 MCLK Master Clock Input. A low jitter, buffered digital clock must be applied to this pin. The output data rate
depends on the frequency of this clock. See the Clocking the AD7760 section for more details.
2 MCLKGND Master Clock Ground Sensing Pin.
36 SYNC Synchronization Input. A falling edge on this pin resets the internal filter. This can be used to synchronize
multiple devices in a system. See the Synchronization section for more details.
39 RD/WR Read/Write Input. This pin, in conjunction with the chip select pin, is used to read and write data to and
from the AD7760. If this pin is low when CS is low, a read takes place. If this pin is high when CS is low, a
write occurs. See the Modulator Data Output Mode and AD7760 Interface sections for more details.
38 DRDY Data Ready Output. Each time new conversion data is available, an active low pulse, ½ ICLK period wide, is
produced on this pin. See the Modulator Data Output Mode and AD7760 Interface sections for more details.
40 CS Chip Select Input. Used in conjunction with the RD/WR pin to read and write data from and to the AD7760.
See the Modulator Data Output Mode and AD7760 Interface sections for more details.

Rev. A | Page 10 of 36
 AD7760

TERMINOLOGY
Signal-to-Noise Ratio (SNR) Integral Nonlinearity (INL)
SNR is the ratio of the rms value of the actual input signal to the INL is the maximum deviation from a straight line passing
rms sum of all other spectral components below the Nyquist through the endpoints of the ADC transfer function.
frequency, excluding harmonics and dc. The value for SNR is
expressed in decibels. Differential Nonlinearity (DNL)
DNL is the difference between the measured and the ideal
Total Harmonic Distortion (THD) 1 LSB change between any two adjacent codes in the ADC.
THD is the ratio of the rms sum of harmonics to the fundamental.
For the AD7760, it is defined as Zero Error
Zero error is the difference between the ideal midscale input
V 22  V 32  V 42  V52  V62
THD dB  20 log
voltage (when both inputs are shorted together) and the actual
V1 voltage producing the midscale output code.

where: Zero Error Drift

V1 is the rms amplitude of the fundamental. Zero error drift is the change in the actual zero error value due
V2, V3, V4, V5, and V6 are the rms amplitudes of the second to to a temperature change of 1°C. It is expressed as a percentage
the sixth harmonics. of full scale at room temperature.

Nonharmonic Spurious-Free Dynamic Range (SFDR) Gain Error

SFDR is the ratio of the rms signal amplitude to the rms value The first transition (from 100 … 000 to 100 … 001) should
of the peak spurious spectral component, excluding harmonics. occur for an analog voltage ½ LSB above the nominal negative
full scale. The last transition (from 011 … 110 to 011 … 111)
Dynamic Range should occur for an analog voltage 1½ LSB below the nominal
Dynamic range is the ratio of the rms value of the full scale to full scale. The gain error is the deviation of the difference
the rms noise measured with the inputs shorted together. The between the actual level of the last transition and the actual
value for the dynamic range is expressed in decibels. level of the first transition, from the difference between the
ideal levels.
Intermodulation Distortion
With inputs consisting of sine waves at two frequencies, fa and Gain Error Drift
fb, any active device with nonlinearities creates distortion products Gain error drift is the change in the actual gain error value due
at sum and difference frequencies of mfa ± nfb, where m, n = 0, to a temperature change of 1°C. It is expressed as a percentage
1, 2, 3, and so on. Intermodulation distortion terms are those of full scale at room temperature.
for which neither m nor n are equal to 0. For example, the second-
order terms include (fa + fb) and (fa − fb), and the third-order
terms include (2fa + fb), (2fa − fb), (fa + 2fb), and (fa − 2fb).
The AD7760 is tested using the CCIF standard, where two input
frequencies near the top end of the input bandwidth are used.
In this case, the second-order terms are usually distanced in
frequency from the original sine waves, and the third-order
terms are usually at a frequency close to the input frequencies.
As a result, the second- and third-order terms are specified
separately. The calculation of the intermodulation distortion is
as per the THD specification, where it is the ratio of the rms
sum of the individual distortion products to the rms amplitude
of the sum of the fundamentals expressed in decibels.

Rev. A | Page 11 of 36
AD7760

TYPICAL PERFORMANCE CHARACTERISTICS

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 65,536 samples using a 7-term Blackman-Harris window.
0 0

–25 –25

–50 –50

AMPLITUDE (dB)
AMPLITUDE (dB)

–75 –75

–100 –100

–125 –125

–150 –150

–175 –175

04975-009
04975-006

–200 –200
0 4000 8000 12000 16000 20000 24000 0 4000 8000 12000 16000 20000 24000
FREQUENCY (Hz) FREQUENCY (Hz)

Figure 5. Normal Mode FFT, 1 kHz, −0.5 dB Input Tone, 256× Decimation Figure 8. Low Power FFT, 1 kHz, −0.5 dB Input Tone, 256× Decimation

0
0

–25
–25

–50
–50
AMPLITUDE (dB)
AMPLITUDE (dB)

–75
–75

–100
–100

–125
–125

–150
–150

–175

04975-010
–175
04975-007

–200
–200
0 4000 8000 12000 16000 20000 24000 0 4000 8000 12000 16000 20000 24000
FREQUENCY (Hz) FREQUENCY (Hz)

Figure 6. Normal Mode FFT, 1 kHz, −6 dB Input Tone, 256× Decimation Figure 9. Low Power FFT, 1 kHz, −6 dB Input Tone, 256× Decimation

0 0

–25 –25

–50 –50
AMPLITUDE (dB)
AMPLITUDE (dB)

–75 –75

–100 –100

–125 –125

–150 –150
04975-011

–175 –175
04975-008

–200 –200
0 4000 8000 12000 16000 20000 24000 0 4000 8000 12000 16000 20000 24000
FREQUENCY (Hz) 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. A | Page 12 of 36
 AD7760
0 0

–25 –25

–50 –50
AMPLITUDE (dB)

AMPLITUDE (dB)
–75 –75

–100 –100

–125 –125

–150 –150

04975-015
–175

04975-012
–175

–200 –200
0 250 500 750 1000 1250 0 250 500 750 1000 1250
FREQUENCY (kHz) FREQUENCY (kHz)

Figure 11. Normal Mode FFT, 100 kHz, −0.5 dB Input Tone, 8× Decimation Figure 14. Low Power FFT, 100 kHz, −0.5 dB Input Tone, 8× Decimation

0 0

–25 –25

–50 –50
AMPLITUDE (dB)

AMPLITUDE (dB)

–75 –75

–100 –100

–125 –125

–150 –150

04975-016
04975-013

–175 –175

–200 –200
0 250 500 750 1000 1250 0 250 500 750 1000 1250
FREQUENCY (kHz) FREQUENCY (kHz)
Figure 12. Normal Mode FFT, 100 kHz, −6 dB Input Tone, 8× Decimation Figure 15. Low Power FFT, 100 kHz, −6 dB Input Tone, 8× Decimation

0 0

–25 –25

–50 –50
AMPLITUDE (dB)

AMPLITUDE (dB)

–75 –75

–100 –100

–125 –125

–150 –150
04975-014

04975-017

–175 –175

–200 –200
0 250 500 750 1000 1250 0 250 500 750 1000 1250
FREQUENCY (kHz) FREQUENCY (kHz)

Figure 13. Normal Mode FFT, 1 MHz, −0.5 dB Input Tone, 8× Decimation Figure 16. Low Power FFT, 1 MHz, −0.5 dB Input Tone, 8× Decimation

Rev. A | Page 13 of 36
AD7760
0 0

–25 –25

–50 –50
AMPLITUDE (dB)

AMPLITUDE (dB)
–75 –75

–100 –100

–125 –125

–150 –150

–175 –175

04975-018

04975-021
–200 –200
0 250 500 750 1000 1250 0 250 500 750 1000 1250
FREQUENCY (kHz) FREQUENCY (kHz)

Figure 17. Normal Mode FFT, 1 MHz, −6 dB Input Tone, 8× Decimation Figure 20. Low Power FFT, 1 MHz, −6 dB Input Tone, 8× Decimation

0 0
TONE A: 999.75kHz TONE A: 999.75kHz
–25 TONE B: 1.00025MHz –25 TONE B: 1.00025MHz

–50 –50
AMPLITUDE (dB)

AMPLITUDE (dB)

–75 –75

–100 –100

–125 –125

–150 –150

–175 –175
04975-019

04975-022
–200 –200
0 250 500 750 1000 1250 0 250 500 750 1000 1250
FREQUENCY (kHz) FREQUENCY (kHz)

Figure 18. Normal Mode IMD, 1 MHz Center Frequency, 8× Decimation Figure 21. Low Power IMD, 1 MHz Center Frequency, 8× Decimation

0 0
TONE A: 999.75kHz TONE A: 999.75kHz
–20 TONE B: 1.00025MHz –20 TONE B: 1.00025MHz
SECOND-ORDER IMD: –105.6dB SECOND-ORDER IMD: –115.7dB
–40 –40
AMPLITUDE (dB)

AMPLITUDE (dB)

–60 –60

–80 –80

–100 –100

–120 –120

–140 –140
04975-020

04975-023

–160 –160
0 2000 4000 6000 8000 10000 0 2000 4000 6000 8000 10000
FREQUENCY (kHz) FREQUENCY (kHz)

Figure 19. Normal Mode IMD, 1 MHz Center Frequency, 8× Decimation Figure 22. Low Power IMD, 1 MHz Center Frequency, 8× Decimation

Rev. A | Page 14 of 36
 AD7760
0 0
TONE A: 999.75kHz TONE A: 999.75kHz
TONE B: 1.00025MHz TONE B: 1.00025MHz
–20 –20
THIRD-ORDER THIRD-ORDER
IMD: –89.15dB IMD: –87.67dB
–40 –40
AMPLITUDE (dB)

AMPLITUDE (dB)
–60 –60

–80 –80

–100 –100

–120 –120

–140 –140

04975-024

04975-027
–160 –160
995 997 999 1001 1003 1005 995 997 999 1001 1003 1005
FREQUENCY (kHz) FREQUENCY (kHz)

Figure 23. Normal Mode IMD, 1 MHz Center Frequency, 8× Decimation Figure 26. Normal Mode IMD, 1 MHz Center Frequency, 8× Decimation

100.5 –105
NORMAL MODE
100.0
–107

99.5
NORMAL MODE
–109
99.0
SNR (dBFS)

THD (dBc)

98.5 –111

98.0
–113
LOW POWER MODE LOW POWER MODE
97.5
–115
97.0
04975-025

04975-028
96.5 –117
0 10 20 30 40 0 10 20 30 40
MCLK FREQUENCY (MHz) MCLK FREQUENCY (MHz)

Figure 24. SNR vs. MCLK Frequency, 8× Decimation, −6 dB, 1 kHz Input Tone Figure 27. THD vs. MCLK Frequency, 8× Decimation, −6 dB, 1 kHz Input Tone

120 116

–60dB
–60dB
–6dB
115 112 –6dB

–0.5dB
–0.5dB
SNR (dBFS)

SNR (dBFS)

110 108

105 104

100 100
04975-026

04975-029

95 96
0 64 128 192 256 0 64 128 192 256
DECIMATION RATE DECIMATION RATE

Figure 25. Normal Mode SNR vs. Decimation Rate, 1 kHz Input Tone Figure 28. Low Power SNR vs. Decimation Rate, 1 kHz Input Tone

Rev. A | Page 15 of 36
AD7760
4500 3000

4000
2500
3500

3000 2000
OCCURRENCE

OCCURRENCE
2500
1500
2000

1500 1000

1000
500

04975-030

04975-032
500

0 0
8385222 8385238 8385254 8385270 8383530 83835246 8383562 8383578 8383594 8383610
24-BIT CODE 24-BIT CODE

Figure 29. Normal Mode, 24-Bit Histogram, 256× Decimation Figure 32. Low Power, 24-Bit Histogram, 256× Decimation

600 450

400
500
350

400 300
OCCURRENCE

OCCURRENCE

250
300
200

200 150

100
100
04975-031

04975-033
50

0 0
8385016 8385116 8385216 8385316 8385416 8385516 8383236 8383386 8383536 8383686 8383836
24-BIT CODE 24-BIT CODE

Figure 30. Normal Mode, 24-Bit Histogram, 8× Decimation Figure 33. Low Power, 24-Bit Histogram, 8× Decimation

0.0010 0.0015

+85°C
+85°C 0.0010
0.0005

0.0005
+25°C
INL (%)

INL (%)

+25°C
0

–40°C –40°C
–0.0005
–0.0005
04975-034

04975-036

–0.0010 –0.0010
0 4194304 8388608 12582912 16777216 0 4194304 8388608 12582912 16777216
24-BIT CODE 24-BIT CODE

Figure 31. 24-Bit INL, Normal Mode Figure 34. 24-Bit INL, Low Power Mode

Rev. A | Page 16 of 36
 AD7760
0.6 50

45
0.4
40

35
0.2
30
DNL (LSB)

DIDD (mA)
0 25

20
–0.2
15

10
–0.4

04975-035

04975-059
5

–0.6 0
0 4194304 8388608 12582912 16777216 0 2 4 6 8 10 12 14 16 18 20
24-BIT CODE ICLK FREQUENCY (MHz)

Figure 35. 24-Bit DNL Figure 38. Decimate × 32, DIDD vs. ICLK Frequency (DVDD = 2.5 V)

40 45

40
35

35
30
30
25
DIDD (mA)
AIDD2 (mA)

25
20
20
15
15

10 10

04975-060
5 5
04975-057

0 0
0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20
ICLK FREQUENCY (MHz) ICLK FREQUENCY (MHz)

Figure 36. AIDD2 vs. ICLK Frequency (AVDD2 = 5 V) Figure 39. Decimate × 256, DIDD vs. ICLK Frequency (DVDD = 2.5 V)

60

50

40
DIDD (mA)

30

20

10
04975-058

0
0 2 4 6 8 10 12 14 16 18 20
ICLK FREQUENCY (MHz)

Figure 37. Decimate × 8, DIDD vs. ICLK Frequency (DVDD = 2.5 V)

Rev. A | Page 17 of 36
AD7760

THEORY OF OPERATION
The AD7760 employs a Σ-Δ conversion technique to convert The third filter has a fixed decimation rate of 2×, is user
the analog input into an equivalent digital word. The modulator programmable, and has a default configuration. It is described
samples the input waveform and outputs an equivalent digital in detail in the Programmable FIR Filter section. This filter can
word to the digital filter at a rate equal to ICLK. also be bypassed.

By employing oversampling, the quantization noise is spread Table 6 shows some characteristics of the default filter. The
across a wide bandwidth from 0 to fICLK. This means that the group delay of the filter is defined to be the delay to the center
noise energy contained in the signal band of interest is reduced of the impulse response and is equal to the computation plus
(see Figure 40a). To further reduce the quantization noise in the the filter delays. The delay until valid data is available (the
signal band of interest, a high order modulator is employed to DVALID status bit is set) is equal to twice the filter delay plus
shape the noise spectrum so that most of the noise energy is the computation delay.
shifted out of the signal band (see Figure 40b).
a.
The digital filtering that follows the modulator removes the
large out-of-band quantization noise (see Figure 40c) while also QUANTIZATION NOISE
reducing the data rate from fICLK at the input of the filter to
fICLK\2
fICLK/8 or less at the output of the filter, depending on the BAND OF INTEREST

decimation rate used.

b.
Digital filtering has certain advantages over analog filtering: It
does not introduce significant noise or distortion and can be
NOISE SHAPING
made perfectly linear in terms of phase.
fICLK\2
The AD7760 employs three FIR filters in series. By using BAND OF INTEREST
different combinations of decimation ratios, filter selection,
and bypassing, data can be obtained from the AD7760 at a large c.
range of data rates. Multibit data from the modulator can be
obtained at the ICLK rate (see Modulator Data Output Mode DIGITAL FILTER CUTOFF FREQUENCY

section). The first filter receives the data from the modulator at

04975-037
a maximum frequency of 20 MHz and decimates it by 4 to output BAND OF INTEREST
fICLK\2

the data at 5 MHz. The partially filtered data can be output at Figure 40. Σ-∆ ADC
this stage. The second filter allows the decimation rate to be
chosen from 2× to 32× or to be completely bypassed.

Table 6. Configuration with Default Filter

ICLK Computation Pass-Band Output Data
Frequency Filter 1 Filter 2 Filter 3 Data State Delay Filter Delay Bandwidth Rate (ODR)
20 MHz Bypassed Bypassed Bypassed Unfiltered 0 0 10 MHz 20 MHz
20 MHz 4× Bypassed Bypassed Partially filtered 0.325 µs 1.2 µs 1.35 MHz 5 MHz
20 MHz 4× Bypassed 2× Fully filtered 1.075 µs 10.8 µs 1 MHz 2.5 MHz
20 MHz 4× 2× Bypassed Partially filtered 1.35 µs 3.6 µs 562.5 kHz 2.5 MHz
20 MHz 4× 2× 2× Fully filtered 1.625 µs 22.8 µs 500 kHz 1.25 MHz
20 MHz 4× 4× Bypassed Partially filtered 1.725 µs 6 µs 281.25 kHz 1.25 MHz
20 MHz 4× 4× 2× Fully filtered 1.775 µs 44.4 µs 250 kHz 625 kHz
20 MHz 4× 8x Bypassed Partially filtered 2.6 µs 10.8 µs 140.625 kHz 625 kHz
20 MHz 4× 8× 2× Fully filtered 2.25 µs 87.6 µs 125 kHz 312.5 kHz
20 MHz 4× 16× Bypassed Partially filtered 4.175 µs 20.4 µs 70.3125 kHz 312.5 kHz
20 MHz 4× 16× 2× Fully filtered 3.1 µs 174 µs 62.5 kHz 156.25 kHz
20 MHz 4× 32× Bypassed Partially filtered 7.325 µs 39.6 µs 35.156 kHz 156.25 kHz
20 MHz 4× 32× 2× Fully filtered 4.65 µs 346.8 µs 31.25 kHz 78.125 kHz
12.288 MHz 4× 8× 2× Fully filtered 3.66 µs 142.6 µs 76.8 kHz 192 kHz
12.288 MHz 4× 16× 2× Fully filtered 5.05 µs 283.2 µs 38.4 kHz 96 kHz
12.288 MHz 4× 32× Bypassed Partially filtered 11.92 µs 64.45 µs 21.6 kHz 96 kHz
12.288 MHz 4× 32× 2× Fully filtered 7.57 µs 564.5 µs 19.2 kHz 48 kHz

Rev. A | Page 18 of 36
 AD7760

MODULATOR DATA OUTPUT MODE

Operating the AD7760 in modulator output mode enables the scaled to 15 bits. The transfer function in Figure 42 shows the
output of data directly from the Σ-Δ modulator. This mode of scaling involved for the 16 data bits output from Modulator
operation bypasses the AD7760 on-board digital filtering Pins D[15:0] vs. the maximum differential voltage input
capabilities, outputting data in its unfiltered form. allowed for the modulator inputs (VIN+ and VIN−).
D[15:0]
As discussed in the Theory of Operation section, the AD7760 +4.096V
0011 1111 1111 1111
operates using oversampling, which spreads quantization noise
+3.275V = MODULATOR FULL SCALE = 80% OF +4.096V VIN+ = 3.6855V
over a wide bandwidth. The decrease in the quantization noise 0011 0011 0011 0010
VIN– = 0.4105V

energy in the resulting signal band is illustrated in Figure 40a.

By coupling the use of oversampling with the use of a high
order, multibit Σ-Δ modulator, the AD7760 further reduces the
quantization noise in the signal band. Figure 41 is an FFT of 0000 0000 0000 0000
VIN+ = 2.048V
VIN– = 2.048V
unfiltered data output from the AD7760 when it is used in
modulator output mode. This clearly demonstrates the shaping
of the quantization noise performed by the AD7760’s Σ-Δ
modulator. VIN+ = 0.4105V
1100 1100 1100 1100
80% OF +4.096V = MODULATOR FULL SCALE = –3.275V VIN– = 3.6855V
MODULATOR INPUTS

04975-049
1100 0000 0000 0000
The maximum voltage input to each differential modulator –4.096V

input pin is 0.8 × 4.096 V ≈ 3.275 V (80% of VREF), which must Figure 42. Modulator Output Data Scaling
sit on a common mode of VREF/2. This maximum differential As the nature of the modulator output is coarse relative to
input voltage is shown as the conditioned output of the AD7760’s the fully filtered output of the AD7760 (due to the associated
on-board differential amplifier in Figure 52 in the Driving the quantization noise of the modulator output), Bits D[3:0] of the
AD7760 section. modulator output are zero when operating in modulator data
output mode. Thus, the data outputs for the calculations listed
Further details on the signal conditioning implemented by the
in Example 1 and Example 2 for inputs to the modulator pins
AD7760’s on-board differential amplifier and the recommended
VIN+ and VIN− show Bits D[3:0] of the modulator output as zero.
external circuitry that accompanies it is described in the
Driving the AD7760 section. Example 1
0
VIN+ = 3.5 V
–10 VIN− = 0.595 V
–20 Modulator Output Code = ([VIN(+) − VIN(−)]/4.096 V) × 16384
–30
–40
= [(3.5 V − 0.595 V)/4.096 V] × 16384
–50 = +11620
AMPLITUDE (dB)

–60
Direct Scaling: [0010 1101 0110 0100]
–70
–80 Value Output on Data Output Pins D[15:0]:
–90 D [15:0] = [0010 1101 0110 0000].
–100
–110 Example 2
–120 VIN+ = 0.595 V
–130
VIN− = 3.5 V
–140
–150 Modulator Output Code = ([VIN(+) − VIN(−)]/4.096 V) × 16384
–160 = [(0.595 V − 3.5 V)/4.096 V] × 16384
04975-048

0 1 2 3 4 5 6 7 8 9 10
FREQUENCY (MHz)
= −11620
Figure 41. FFT of Data Output by the AD7760 in Modulator Output Mode
Direct Scaling: [1101 0010 1001 1100]
Value Output on Data Output Pins D[15:0]:
MODULATOR DATA OUTPUT SCALING D [15:0] = [1101 0010 1001 0000].
In modulator output mode, data is output in a 16-bit twos
complement format on Pins D [15:0]; however, this data is

Rev. A | Page 19 of 36
AD7760

MODULATOR DATA OUTPUT MODE INTERFACE

The AD7760 can be configured in modulator data output mode output mode when operating with CDIV = 1 (that is, ICLK =
(bypassing the default decimation filtering) by writing 0 to each MCLK). A DRDY pulse is generated for each word. The data on
of the bits contained in Control Register 1: BYP F1, BYP F3, each of the 16 data output pins, D [15:0], is valid on the rising
and DEC [2:0]. This will bypass all digital decimation filtering edge of the DRDY pulse. The DRDY pulse can be used to latch
offered by the AD7760. See the AD7760 Registers section for the modulator data into a FIFO or as a DMA control signal. Shortly
further details. after the RD/WR and CS lines return high, the AD7760 stops
outputting data and the data bus returns to high impedance.
When the AD7760 is operating in modulator data output mode,
a different parallel interfacing scheme than that used for config- CLOCK DIVIDE-BY-2 MODE (CDIV = 0)
urations, where the AD7760’s data output is filtered is necessary.
When operating in modulator output mode with CDIV = 0
The data output rate depends on the clock divider ratio that is (that is, ICLK = MCLK/2), the frequency of the DRDY signal
used. When the CDIV bit in Control Register 2 is set to logic created is half that of the MCLK frequency input to the device.
high, data is output at the MCLK frequency. If the CDIV bit is The timing scheme that is used when CDIV = 0 depends on the
set to logic low, data is output at a frequency of MCLK/2. See number of MCLK cycles that occur between RESET and SYNC.
the Clocking the AD7760 section. If the number of MCLK cycles (n) between the rising edge of
RESET and the rising edge of SYNC (see Figure 44) is an even
CLOCK DIVIDE-BY-1 MODE (CDIV = 1) value, use the interface timing shown in Figure 43. If n is an odd
When obtaining data from the AD7760 in modulator output value, use the interface timing shown in Figure 45.
mode, both the RD/WR and CS lines must be held low. This
brings the data bus out of its high impedance state. Figure 43
shows the timing diagram for reading data in the modulator data

t9 t10

DRDY

CS, RD/WR t13

t12 t14
t11
04975-050
D[0:15] INVALID DATA MOD DATA M MOD DATA M + 1 MOD D...

Figure 43. AD7760 Modulator Output Mode (CDIV = 1) and (CDIV = 0, n is even)

RESET

MCLK

n × tMCLK
04975-051

SYNC

Figure 44. AD7760 Relative Timing Between RESET and SYNC in Modulator Output Mode CDIV = 0

Rev. A | Page 20 of 36
 AD7760
t9 t10

DRDY

t20

D[0:15] INVALID DATA MOD DATA M MOD DATA M + 1 MOD D...

t19

MCLK

t14
t11

04975-052
CS, RD/WR

Figure 45. AD7760 Modulator Output Mode (CDIV = 0, n is odd)

In the case where n is an odd number of MCLK cycles, the 5. Write to Control Register 2 to power up the ADC and the
modulator data output on Pins D [15:0] is output on the rising differential amplifier as required. The correct clock divider
edge of DRDY. In this case, the modulator data should be read (CDIV) ratio should be programmed at this time.
on the falling edge of MCLK when DRDY is logic low. Figure 45
shows timing details to be used when reading the modulator 6. Write to Control Register 1 to set the bypass filter bits, BYP F1
output data where CDIV = 0 and there is an odd number of and BYP F3, and the decimation rate bits, DEC [2:0], to 0.
MCLK cycles between the rising edge of RESET and the rising 7. Wait a minimum of six MCLK cycles after the rising edge
edge of SYNC. The edge of MCLK that should be used under of CS has been released.
these conditions is illustrated in Figure 45 by arrows on the
MCLK falling edges in question. 8. Take SYNC low for a minimum of four MCLK cycles, if
required, to synchronize multiple parts.
USING THE AD7760
IN MODULATOR OUTPUT MODE Using this sequence results in an even number of MCLK cycles
The following is the recommended sequence for powering up between the rising edge of RESET and the rising edge of SYNC.
and using the AD7760: Therefore, when using this sequence with CDIV = 0, the interface
timing shown in Figure 43 should be implemented.
1. Apply power.
Note that whether the number of MCLK cycles between the
2. Start the clock oscillator, applying MCLK. rising edge of RESET and SYNC is odd or even is irrelevant
3. Take RESET low for a minimum of one MCLK cycle. when the AD7760 is operated with CDIV = 1.

4. Wait a minimum of two MCLK cycles after the rising edge When using the AD7760 in modulator output mode, the offset,
of RESET. gain, and overrange registers are not operational. The only
registers that can be used are Control Register 1 and Control
Register 2.

Rev. A | Page 21 of 36
AD7760

AD7760 INTERFACE
READING DATA If multiple synchronized AD7760 parts that share a properly
When the AD7760 is outputting data at a 5 MHz output data distributed common MCLK signal exist in a system, these parts
rate or less, the interface operates in a conventional mode, as can share a common bus without being isolated from each
shown in Figure 2, using a 16-bit bidirectional parallel interface. other. This bus can then be isolated from the system bus by a
This interface is controlled by the RD/WR and CS pins. The 24-bit single latch or buffer.
conversion data is output in twos complement format. When a SYNCHRONIZATION
new conversion result is available, an active low pulse is output
The SYNC input to the AD7760 provides a synchronization
on the DRDY pin.
function that allows the user to begin gathering samples of the
To read a conversion result from the AD7760, two 16-bit read analog front-end input from a known point in time.
operations are performed. The DRDY pulse indicates that a new
The SYNC function allows multiple AD7760s, operated from
conversion result is available. Both RD/WR and CS go low to
the same MCLK, RESET, and SYNC signals, to be synchronized
perform the first read operation. Shortly after both lines go low,
so that each ADC simultaneously updates its output register.
the data bus becomes active and the 16 most significant bits
The distribution of the signals that are common to each of the
(MSBs) of the conversion result are output. The RD/WR and CS
devices that are to be synchronized is extremely important in
lines must return high for a full ICLK period before the second ensuring that the timing of each of the AD7760 devices is
read is performed. This second read contains the eight least correct, that is, that each AD7760 device sees the same digital
significant bits (LSBs) of the conversion result along with six edges synchronously.
status bits. These status bits are shown in Table 7. Descriptions
of the other status bits are found in Table 17. The SYNC signal is sensed on the falling edge of MCLK. On the
Table 7. Status Bits During Data Read first falling edge of MCLK after SYNC goes logic low, the digital
MSB LSB filter sequencer is reset to 0. The filter is held in a reset state until
D7 D6 D5 D4 D3 D2 D1 D0 a falling edge of the MCLK senses SYNC logic high. The SYNC
DVALID OVR UFILT LPWR FILTOK DLOK 0 0 signal must remain logic low for a minimum of four MCLK
cycles. Figure 46 shows the recommended timing for the SYNC
Shortly after RD/WR and CS return high, the data bus returns
signal with respect to MCLK.
to a high impedance state. Both read operations must be
completed before a new conversion result is available because DEVICE SYNCHRONIZED
FROM THIS POINT IN TIME
the new result overwrites the contents of the output register.
If a DRDY pulse occurs during a read operation, the data read MCLK
is invalid.
MIN SYNC LOGIC LOW

04975-053
SYNC 4 × tMCLK
READING STATUS AND OTHER REGISTERS
The AD7760 features a number of programmable registers. To Figure 46. Recommended SYNC Timing
read back the contents of these registers or the status register,
the user must first write to the control register of the device, The rising edge of SYNC should be coincident with the rising
setting a bit that corresponds to the register to be read. The next edge of MCLK. Thus, the next falling edge of MCLK senses
read operation outputs the contents of the selected register SYNC logic high and takes the filter out of its reset state. By
instead of a conversion result. The AD7760 Registers section applying this signal scheme to multiple ADCs using the same
provides more information on the relevant bits in the control MCLK and SYNC signals, all of the devices will gather input
register. samples synchronously.

SHARING THE PARALLEL BUS Following a SYNC signal, the digital filter needs time to settle
By its nature, the high accuracy of the AD7760 makes it before valid data can be read from the AD7760. The DVALID
sensitive to external noise sources. These include digital activity status bit (D7 in Table 7) output with each conversion indicates
on the parallel bus. For this reason, it is recommended that the when valid data is being output by the converter. The time from
AD7760 data lines be isolated from the system data bus by the rising edge of SYNC until the DVALID bit is asserted is
means of a latch or buffer to ensure all digital activity on the dependent on the filter configuration used. See the Theory of
D0 to D15 pins is controlled by the AD7760. Operation section and the values listed in Table 6 for details on
calculating the time until DVALID is asserted.

Rev. A | Page 22 of 36
 AD7760
WRITING TO THE AD7760 Figure 3 shows a write operation to the AD7760. The RD/WR
line is held high while the CS line is brought low for a minimum
There are many features and parameters that the user can
of four ICLK periods. The register address is latched during this
change by writing to the AD7760 device. See the Using the
period. The CS line is brought high again for a minimum of
AD7760 section, which details the writing sequence needed to
initialize the operation of the part. four ICLK periods before the register data is put onto the data
bus. If a read operation occurs between the writing of the
The AD7760 has programmable registers that are 16 bits wide. register address and the register data, the register address is
This means that two write operations are required to program a cleared and the next write must be the register address. This
register. The first write contains the register address, and the also provides a method to revert back to a known situation if
second write contains the register data. An exception is when a the user forgets whether the next write is an address or data.
user-defined filter is being downloaded to the AD7760. This is
described in detail in the Downloading a User-Defined Filter Generally, the AD7760 is written to and configured on power-
section. The AD7760 Registers section contains the register up and very infrequently, if at all, after that. Following any write
addresses and details. operation, the full group delay of the filter must elapse before
valid data is output from the AD7760.

Rev. A | Page 23 of 36
AD7760

CLOCKING THE AD7760

The AD7760 requires an external low jitter clock source. This MCLK JITTER REQUIREMENTS
signal is applied to the MCLK pin, and the MCLKGND pin is The MCLK jitter requirements depend on a number of factors
used to sense the ground from the clock source. An internal and are given by
clock signal (ICLK) is derived from the MCLK input signal. The
ICLK controls all internal operations of the AD7760. The OSR
t j(rms )
2 u S u f IN u 10
maximum ICLK frequency is 20 MHz, but due to an internal SNR (dB)
clock divider, a range of MCLK frequencies can be used. There 20
are two ways to generate the ICLK:
where:
ICLK = MCLK (CDIV = 1) OSR = oversampling ratio = fICLK/ODR.
ICLK = MCLK/2 (CDIV = 0) fIN = maximum input frequency.
SNR(dB) = target SNR.
These options are selected from the control register (see the
AD7760 Registers section for more details). On power-up, the Example 1
default is ICLK = MCLK/2 to ensure that the part can handle This example can be taken from Table 6, where:
the maximum MCLK frequency of 40 MHz. For output data ODR = 2.5 MHz.
rates equal to those used in audio systems, a 12.288 MHz ICLK fICLK = 20 MHz.
frequency can be used. As shown in Table 6, output data rates of fIN (max) = 1 MHz.
192 kHz, 96 kHz, and 48 kHz are achievable with this ICLK SNR = 108 dB.
frequency. As mentioned previously, this ICLK frequency can
8
2 u S u 10 6 u 10 5.4
be derived from different MCLK frequencies. t j (rms) 1.79 ps

It is recommended that the MCLK signal applied to the

This is the maximum allowable clock jitter for a full-scale,
AD7760 has a 50-50 mark-space ratio. When operating in clock
1 MHz input tone with the given ICLK and output data rate.
divide-by-1 mode (that is, CDIV = 1), using higher mark-space
ratios reduces the maximum MCLK frequency that can be Example 2
applied to the AD7760 yielding maximum performance. For Take a second example from Table 6, where:
example, using a mark-space ratio of 60-40 (with CDIV = 1) ODR = 48 kHz.
reduces the maximum MCLK frequency that will yield the fICLK = 12.288 MHz.
maximum INL and THD performance to 16 MHz. fIN (max) = 19.2 kHz.
SNR = 120 dB.
BUFFERING THE MCLK SIGNAL
256
2 u S u 19.2 u 103 u 106
The MCLK signal for the AD7760 must be buffered before t j(rms ) 133 ps
being input to the MCLK pin on the AD7760 device. This can
be done simply by routing the MCLK signal to both inputs of an
The input amplitude also has an effect on these jitter figures.
AND gate (see Figure 47).
For example, if the input level was 3 dB below full-scale, the
The recommended buffer is the NC7SZ08M5, which is a two- allowable jitter would be increased by a factor of √2, increasing
input AND gate from Fairchild Semiconductor. Using the buffer the first example to 2.53 ps rms. This happens when the
with a supply voltage of 5 V is advised to achieve optimum maximum slew rate is decreased by a reduction in amplitude.
performance from the AD7760. Figure 48 and Figure 49 illustrate this point, showing the
maximum slew rate of a sine wave of the same frequency but
AD7760 with different amplitudes.
MCLK
SOURCE
3 MCLK

NC7SZ08M5
04975-054

(AND GATE)

Figure 47. Buffering the MCLK Signal Using the NC7SZ08M5 AND Gate

Rev. A | Page 24 of 36
 AD7760
1.0 1.0

0.5 0.5

0 0

–0.5 –0.5

04975-038

04975-039
–1.0 –1.0

Figure 48. Maximum Slew Rate of Sine Wave with Amplitude of 2 V p-p Figure 49. Maximum Slew Rate of Sine Wave (with the Same Frequency as in
Figure 48) with Amplitude of 1 V p-p

Rev. A | Page 25 of 36
AD7760

DRIVING THE AD7760

The AD7760 has an on-chip differential amplifier that operates signal to sit on the optimum common mode of VREF/2, in this
with a supply voltage (AVDD3) within the 3.15 V to 5.25 V range. case 2.048 V. The signal is also scaled to give the maximum
For a 4.096 V reference, the supply voltage must be 5 V. 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
To achieve the specified performance in normal mode, the on each input.
differential amplifier should be configured as a first-order
antialias filter, as shown in Figure 50. Any additional filtering With a 4.096 V reference, a 5 V supply must be provided to the
should be carried out in previous stages using low noise, high reference buffer (AVDD4). With a 2.5 V reference, a 3.3 V supply
performance op amps, such as the AD8021. must be provided to AVDD4.

Suitable component values for the first-order filter are listed in Figure 51 shows the transfer function in terms of the 24-bit
Table 8. Using the values in the table as an example yields a digital output codes (twos complement coding) of the AD7760
10 dB attenuation at the first alias point of 19 MHz. vs. the voltage signals VA and VB applied to the on-board
differential amplifier A1, as shown in Figure 52.
CFB
24 BITS

RFB 011…111

011…110
RIN RM
A VIN–
CS A1 000…010
RM
B VIN+ 000…001
AD7760
RIN 24-BIT 000…000
RFB OUTPUT
111…111
04975-040

CFB 111…110

Figure 50. Differential Amplifier Configuration 100…001

Table 8. Normal Mode Component Values 100…000

04975-056
VREF RIN RFB RM CS CFB B = +2.5V A = 0V A = +2.5V
A = –2.5V B = 0V B = –2.5V
4.096 V 1 kΩ 655 Ω 18 Ω 5.6 pF 33 pF
Figure 51. Transfer Function for the AD7760 Filtered Output Where VA and VB
Figure 52 shows the signal conditioning that occurs using the are Inputs to the On-Board Differential Amplifier A1
circuit shown in Figure 50 with a ±2.5 V input signal biased
around ground and the component values and conditions listed
in Table 8. The differential amplifier always biases the output

+2.5V +3.685V

VIN+
0V +2.048V

–2.5V +0.410V

+2.5V +3.685V

VIN–
0V +2.048V

–2.5V +0.410V
04975-055

INPUTS TO THE AD7760 DIFFERENTIAL AMPLIFIER OUTPUTS OF THE AD7760 DIFFERENTIAL AMPLIFIER

Figure 52. Differential Amplifier Signal Conditioning

Rev. A | Page 26 of 36
 AD7760
To obtain maximum performance from the AD7760, it is USING THE AD7760
advisable to drive the ADC with differential signals. Figure 53 The following is the recommended sequence for powering up
shows how a bipolar, single-ended signal biased around ground and using the AD7760:
can drive the AD7760 with the use of an external op amp, such
as the AD8021. 1. Apply power.
CFB
2. Start the clock oscillator, applying MCLK.

2R RFB 3. Take RESET low for a minimum of one MCLK cycle.

2R
VIN
RIN RM 4. Wait a minimum of two MCLK cycles after RESET has
AD8021 VIN–
CS A1
been released.
RM
R
VIN+
5. Write to Control Register 2 to power up the ADC and the
RIN RFB
differential amplifier as required. The correct clock divider
(CDIV) ratio should be programmed at this time.

04975-042
CFB

Figure 53. Single-Ended-to-Differential Conversion 6. Write to Control Register 1 to set the output data rate.

The AD7760 employs a double-sampling front end, as shown in 7. Wait a minimum of five MCLK cycles after CS has been
Figure 54. For simplicity, only the equivalent input circuit for VIN+ released.
is shown. The equivalent input circuitry for VIN− is the same.
8. Take SYNC low for a minimum of four MCLK cycles, if
VIN+
CS1
required, to synchronize multiple parts.
SS1

CPA
SH3 Data can then be read from the part using the default filter,
SH1 SS3
CPB1 offset, gain, and overrange threshold values. The conversion
ANALOG data read is not valid, however, until the group delay of the filter
MODULATOR
has elapsed. Once this has occurred, the DVALID bit read with
the data LSW is set, indicating that the data is indeed valid.
SS2 CS2

SH4 The user can then download a different filter if required (see the
SH2
CPB2
SS4 Downloading a User-Defined Filter section). Values for gain,
04975-043

offset, and overrange threshold registers can be written or read

at this stage.
Figure 54. Equivalent Input Circuit

Sampling Switches SS1 and SS3 are driven by ICLK, whereas

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 CS1 (pF) CS2 (pF) CPA (pF) CPB1/2 (pF)
Normal 51 51 12 20
Low Power 13 13 12 5

Rev. A | Page 27 of 36
AD7760

DECOUPLING AND LAYOUT RECOMMENDATIONS

Due to the high performance nature of the AD7760, correct decoupling and layout techniques are required to obtain the performance as
stated within this data sheet. Figure 55 shows a simplified connection diagram for the AD7760.

PIN 14
PIN 15

PIN 12

PIN 33

PIN 24
PIN 27

PIN 44
PIN 63

PIN 41
PIN 4

PIN 6
U2

14
15

12

33

24
27

44
63

41
4

6
DB [0:15]

AVDD2
AVDD2
AVDD2

AVDD4

AVDD1
AVDD1

AVDD3
AVDD2

DVDD
VDRIVE
VDRIVE
19 61
VINA+ VINA+ DB0 DB0
20 60
VINA– VINA– DB1 DB1
21 59
VOUTA– VOUTA– DB2 DB2
22 58
VOUTA+ VOUTA+ DB3 DB3
57
DB4 DB4
56
DB5 DB5
8 55
DECAPA DB6 DB6
30 54
DECAPB DB7 DB7
52
DB8 DB8
51
DB9 DB9
25 50
VIN+ VIN+ DB10 DB10
C7 C64 26 49
100nF 33pF VIN– VIN– DB11 DB11
48
DB12 DB12
AD7760BSV 47
DB13 DB13
10 46
VREFX VREF+ DB14 DB14
9 45
REFGND DB15 DB15

17 40
RBIAS CS CS
39
RD/WR RD/WR
1
DGND
35 37
DGND RESET RESET
42 36
DGND SYNC SYNC
43 38
DGND DRDY DRDY
53
R19 DGND
160kŸ 62 3
DGND MCLK MCLK
7 AGND1
34 AGND1

AGND2
AGND2
AGND2
AGND2
AGND2

AGND3
AGND3
AGND3
AGND3

11 AGND4

64 2
DGND MCLKGND
5
13
16
18
28

23
29
31
32

AVDD2 AVDD4 AVDD1 AVDD3 VDRIVE DVDD

PIN 12 L4
(VBUF)
L1 L3 L2 L5 L11 L6 L7 L12 L8
PIN 4 PIN 15 PIN 14 PIN 5 PIN 33 PIN 24 PIN 44 PIN 63 PIN 41
(RHS) (VBIAS) (LHS) L9 PIN 27 (VMOD1) (VMOD2) (VDIF1) (VDRV1) (VDRV2) (DVDD)
R38
10Ÿ

C48 C50 C62 C59 C52 C53 C54 C56 C57 C58
100nF 100nF 100nF 10nF 100nF 100nF 100nF 100nF 100nF 100nF

04975-046
Figure 55. Simplified Connection Diagram

Rev. A | Page 28 of 36
 AD7760
SUPPLY DECOUPLING BIAS RESISTOR SELECTION
Every supply pin must be connected to the appropriate supply The AD7760 requires a resistor to be connected between the
via a ferrite bead and decoupled to the correct ground pin with RBIAS and AGND pins. The value of this resistor is dependent
a 100 nF, 0603 case size, X7R dielectric capacitor. There are two on the reference voltage being applied to the device. The resistor
exceptions to this: value should be selected to produce a current of 25 µA through

x
the resistor to ground. For a 2.5 V reference voltage, the correct
Pin 12 (AVDD4) must have a 10 Ω resistor inserted between resistor value is 100 kΩ, and for a 4.096 V reference, the correct
the pin and a 10 nF decoupling capacitor, which is resistor value is 160 kΩ.
connected to ground at Pin 9.

x
LAYOUT CONSIDERATIONS
Pin 27 (AVDD2) does not require a separate decoupling
While using the correct components is essential to achieve
capacitor or a direct connection to the supply, but instead
optimum performance, the correct layout is just as important.
is connected to Pin 14 via a 15 nH inductor.
The AD7760 product page on the Analog Devices website
ADDITIONAL DECOUPLING contains the Gerber files for the AD7760 evaluation board.
These files should be used as a reference when designing any
There are two other decoupling pins on the AD7760—Pin 8
system using the AD7760.
(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 of this data sheet are critical,
REFERENCE VOLTAGE FILTERING
and particular attention must be paid to the components that
A low noise reference source, such as the ADR431 (2.5 V) or are located close to the AD7760. Locating these components
ADR434 (4.096 V), is suitable for use with the AD7760. The farther away from the device can have a direct impact on the
reference voltage supplied to the AD7760 should be decoupled achievable maximum performance.
and filtered as shown in Figure 56.
The use of ground planes should also be carefully considered.
The recommended scheme for the reference voltage supply is a To ensure that the return currents through the decoupling
100 Ω series resistor connected to a 100 μF tantalum capacitor, capacitors are flowing to the correct ground pin, the ground side
followed by a series resistor of 10 Ω, and finally a 10 nF capacitor of the capacitors should be as close as possible to the ground pin
placed as close as possible to the VREF+ pin, decoupling this associated with that supply. A ground plane should not be relied
capacitor to the associated ground pin, Pin 11. on as the sole return path for decoupling capacitors because the
U3 return current path using ground planes is not easily predictable.
ADR434 R30 R17
100Ÿ 10Ÿ
7.5V
2
VIN VOUT
6
PIN 10 EXPOSED PADDLE
C15
+
C9 GND C11
+
C46 The AD7760 64-lead TQFP employs a 6 mm × 6 mm exposed
10µF 100nF 4 100µF 10nF
paddle (see Figure 59). The paddle reduces the thermal
04975-047

resistance of the package by providing a path for heat energy to

Figure 56. Reference Connection flow between the package and the PCB and, in turn, increases
the heat transfer efficiency from the AD7760 package.
DIFFERENTIAL AMPLIFIER COMPONENTS
Connecting the exposed paddle to the AGND plane of the PCB
The correct components for use around the on-chip differential is essential in creating the conditions that allow the AD7760
amplifier are detailed in Table 8. Matching the components on package to perform to the highest specifications possible. The
both sides of the differential amplifier is important to minimize exposed paddle should not be connected directly to any of the
distortion of the signal applied to the amplifier. A tolerance of ground pins on the AD7760 and should only be connected to
0.1% or better is required for these components. Symmetrical the analog ground plane. Best practice is to use multiple vias
routing of the tracks on both sides of the differential amplifier connecting the exposed paddle to the AGND plane of the PCB.
also assists in achieving the stated performance.

Rev. A | Page 29 of 36
AD7760

PROGRAMMABLE FIR FILTER

As previously mentioned, the third FIR filter on the AD7760 To create a filter, note the following:

x
is user programmable. The default coefficients that are loaded
upon reset are given in Table 10, and the frequency responses The filter must be an even, symmetrical FIR.

x
are shown in Figure 57. The frequencies quoted in Figure 57
The coefficients are in sign-and-magnitude format, with
scale directly with the output data rate.
26 magnitude bits and sign coded as positive = 0.

x
Table 10. Default Filter Coefficients
Dec Hex Dec Hex The filter length must be between 12 taps and 96 taps in
No. Value Value No. Value Value steps of 12.

x
0 53656736 332BCA0 24 700847 AB1AF
1 25142688 17FA5A0 25 −70922 401150A
Because the filter is symmetrical, the number of
2 −4497814 444A196 26 −583959 408E917
coefficients that must be downloaded is half the filter
3 −11935847 4B62067 27 −175934 402AF3E
length. The default filter coefficients exemplify this with
only 48 coefficients listed for a 96-tap filter.
4 −1313841 4140C31 28 388667 5EE3B
5 6976334 6A734E 29 294000 47C70 x Coefficients are written from the center of the impulse
6 3268059 31DDDB 30 −183250 402CBD2 response (adjacent to the point of symmetry) outwards.
7 −3794610 439E6B2 31 −302597 4049E05
8 −3747402 4392E4A 32 16034 3EA2 x The coefficients are scaled so that the in-band gain of the
9 1509849 1709D9 33 238315 3A2EB filter is equal to 134,217,726, with the coefficients rounded
10 3428088 344EF8 34 88266 158CA to the nearest integer. For a low-pass filter, this is the
11 80255 1397F 35 −143205 4022F65 equivalent of having the coefficients summed arithmetically
12 −2672124 428C5FC 36 −128919 401F797 (including sign) to a +67,108,863 (0x3FF FFFF) positive
13 −1056628 4101F74 37 51794 CA52 value over the half-impulse-response coefficient set (a
14 1741563 1A92FB 38 121875 1DC13 maximum of 48 coefficients). Any deviation from this
15 1502200 16EBF8 39 16426 402A introduces a gain error.
16 −835960 40CC178 40 −90524 401619C
0 PASS-BAND RIPPLE = 0.05dB
17 −1528400 4175250 41 −63899 400F99B –0.1dB FREQUENCY = 1.004MHz
–3dB FREQUENCY = 1.06MHz
18 93626 16DBA 42 45234 B0B2 –20
STOP BAND = 1.25MHz
19 1269502 135EFE 43 114720 1C020 –40
20 411245 6466D 44 102357 18FD5
AMPLITUDE (dB)

–60
21 −864038 40D2F26 45 52669 CDBD
22 −664622 40A242E 46 15559 3CC7 –80
23 434489 6A139 47 1963 7AB
–100

The default filter should be sufficient for most applications. It is –120

a standard brick wall filter with a symmetrical impulse response. –140

The default filter has a length of 96, is nonaliasing, and provides
–160
120 dB of attenuation at Nyquist. This filter not only performs
04975-044

0 500 1000 1500 2000 2500

signal antialiasing, but also suppresses out-of-band quantization FREQUENCY (kHz)

noise produced by the analog-to-digital conversion process. Any Figure 57. Default Filter Frequency Response (2.5 MHz ODR)
significant relaxation in the stop-band attenuation or transition
bandwidth relative to the default filter can result in a failure to The procedure for downloading a user-defined filter is detailed
meet the SNR specifications. in the Downloading a User-Defined Filter section.

The default filter characteristics scale with both the MCLK

frequency applied and the decimation rate chosen by the user.

Rev. A | Page 30 of 36
 AD7760

DOWNLOADING A USER-DEFINED FILTER

As previously mentioned, the filter coefficients are 27 bits in EXAMPLE FILTER DOWNLOAD
length—one sign and 26 magnitude bits. Because the AD7760 The following is an example of downloading a short user-
has a 16-bit parallel bus, the coefficients are padded with 5 MSB defined filter with 24 taps. The frequency response is shown in
0s to generate a 32-bit word, split into two 16-bit words for Figure 58.
downloading. The first 16-bit word for each coefficient becomes
10
(00000, sign bit, Magnitude [25:16]), whereas the second word
becomes (Magnitude [15:0]). To ensure that a filter is downloaded 0

correctly, a checksum must also be generated and then downloaded –10

following the final coefficient. The checksum is a 16-bit word
–20
generated by splitting each 32-bit word into four bytes and

AMPLITUDE (dB)
summing the bytes from all coefficients up to a maximum of –30

192 bytes (48 coefficients × four bytes). The same checksum –40
is generated internally in the AD7760 and compared with the
–50
downloaded checksum. The DL_OK bit in the status register is
set if these two checksums agree. –60

–70
To download a user filter
–80

04975-045
0 100 200 300 400 500 600
1. Write to Control Register 1, setting the DL_FILT bit and
FREQUENCY (kHz)
the correct filter length bits corresponding to the length of
Figure 58. 24-Tap FIR Frequency Response
the filter to be downloaded (see Table 11).
The coefficients for the filter are listed in Table 12 and are
2. Write the first half of the current coefficient data
shown from the center of symmetry outwards. The raw
(00000, sign bit, Magnitude [25:16]). The first coefficient
coefficients were generated using a commercial filter design
to be written must be the one adjacent to the point of filter
tool and were scaled appropriately so that their sum equals
symmetry.
67,108,863 (0x3FF FFFF).
3. Write the second half of the current coefficient data Table 12. 24-Tap FIR Coefficients
(Magnitude [15:0]). Coefficient Raw Scaled
4. Repeat Step 2 and Step 3 for each coefficient. 1 0.365481974 53188232
2 0.201339905 29300796
5. Write the 16-bit checksum. 3 0.009636604 1402406
4 −0.075708848 −11017834
6. Use the following methods to verify that the filter coefficients 5 −0.042856209 −6236822
are downloaded correctly: 6 0.019944246 2902466
a. Read the status register, checking the DL_OK bit. 7 0.036437914 5302774
8 0.007592007 1104856
b. Read data and observe the status of the DL_OK bit. 9 −0.021556583 −3137108
10 −0.024888355 −3621978
Note that because the user coefficients are stored in RAM, they
11 −0.012379538 −1801582
are cleared after a RESET operation or a loss of power.
12 −0.001905756 −277343
Table 11. Filter Length Values
FLEN [3:0] Number of Coefficients Filter Length
0000 Default Default
0001 6 12
0011 12 24
0101 18 36
0111 24 48
1001 30 60
1011 36 72
1101 42 84
1111 48 96

Rev. A | Page 31 of 36
AD7760
Table 13 shows the hexadecimal values (in sign-and-magnitude Table 14 lists the 16-bit words the user would write to the
format) that are downloaded to the AD7760 to realize this filter. AD7760 to set up the ADC and download this filter, assuming
The table is also split into the bytes that are summed to produce an output data rate of 1.25 MHz has been selected.
the checksum. The checksum generated from these coefficients Table 14. Sequence of Write Instructions to Set Up Device
is 0x0E6B. and Download the Filter Example
Table 13. Filter Hexadecimal Values Word Description
Word 1 Word 2 0x0001 Address of Control Register 1.
Coefficient Byte 1 Byte 2 Byte 3 Byte 4 0x8079 Control register data. DL filter: set filter length = 24,
1 03 2B 96 88 set output data rate = 1.25 MHz.
2 01 BF 18 3C 0x032B First coefficient, Word 1.
3 00 15 66 26 0x9688 First coefficient, Word 2.
4 04 A8 1E 6A 0x01BF Second coefficient, Word 1.
5 04 5F 2A 96 0x183C Second coefficient, Word 2.
6 00 2C 49 C2 … Other coefficients.
7 00 50 E9 F6 0x0404 Twelfth (final) coefficient, Word 1.
8 00 10 DB D8 0x3B5F Final coefficient, Word 2.
9 04 2F DE 54 0x0E6B Checksum. Wait (0.5 × tICLK × number of unused
10 04 37 44 5A coefficients) for AD7760 to write 0s to the remaining
unused coefficients.
11 04 1B 7D 6E
0x0001 Address of control register.
12 04 04 3B 5F
0x0879 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
if the filter was downloaded correctly.

Rev. A | Page 32 of 36
 AD7760

AD7760 REGISTERS
The AD7760 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 in this section.

CONTROL REGISTER 1—ADDRESS 0x0001

Default Value 0x001A
MSB LSB
D15 D14 D13 D12 D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0
DL_ RD RD RD RD 0 SYNC FLEN3 FLEN2 FLEN1 FLEN0 BYP F3 BYP F1 DEC2 DEC1 DEC0
FILT OVR GAIN OFF STAT

Table 15. Bit Descriptions of Control Register 1

Bit Mnemonic Description
15 DL_FILT 1 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.
14 RD OVR1, 2 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.
13 RD GAIN1, 2 Read Gain. If this bit has been set, the next read operation outputs the contents of the digital gain register.
12 RD OFF1, 2 Read Offset. If this bit has been set, the next read operation outputs the contents of the digital offset register.
11 RD STAT1, 2 Read Status. If this bit has been set, the next read operation outputs the contents of the status register.
10 0 0 must be written to this bit.
9 SYNC1 Synchronize. Setting this bit initiates an internal synchronization routine. Setting this bit simultaneously on
multiple devices synchronizes all filters.
8 to 5 FLEN [3:0] Filter Length Bits. These bits must be set when the DL_FILT bit is set before a user-defined filter is downloaded.
4 BYP F3 Bypass Filter 3. If this bit is 0, Filter 3 (programmable FIR) is bypassed.
3 BYP F1 Bypass Filter 1. If this bit is 0, Filter 1 is bypassed. This should only occur when the user requires unfiltered
modulator data to be output.
2 to 0 DEC [2:0] 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.
1
Bit 15 to Bit 9 are self-clearing bits.
2
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 After RESET: 0x009B
Recommended register setting for power-up and normal operation using clock divide-by-2 (CDIV = 0) mode: 0x0002
MSB LSB
D15 D14 D13 D12 D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0
0 0 0 0 0 0 0 0 0 0 CDIV 0 PD LPWR 1 D1PD

Table 16. Bit Descriptions of Control Register 2

Bit Mnemonic Description
5 CDIV 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, the ICLK frequency is equal to the MCLK.
3 PD Power Down. Setting this bit powers down the AD7760, reducing the power consumption to 6.35 mW.
2 LPWR Low Power. If this bit is set, the AD7760 is operating in a low power mode. The power consumption is reduced for a 6 dB
reduction in noise performance.
1 1 Write 1 to this bit.
0 D1PD Differential Amplifier Power Down. Setting this bit powers down the on-chip differential amplifier.

Rev. A | Page 33 of 36
AD7760
STATUS REGISTER (READ ONLY)
MSB LSB
D15 D14 D13 D12 D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0
PART 1 PART 0 DIE 2 DIE 1 DIE 0 0 LPWR OVR DL_OK FILTOK UFILT BYP F3 BYP F1 DEC2 DEC1 DEC0

Table 17. Bit Descriptions of Status Register

Bit Mnemonic Comment
15, 14 PART [1:0] Part Number. These bits are constant for the AD7760.
13 to 11 DIE [2:0] Die Number. These bits reflect the current AD7760 die number for identification purposes within a system.
10 0 This bit is set to 0.
9 LPWR Low Power. If the AD7760 is operating in low power mode, this bit is set to 1.
8 OVR If the current analog input exceeds the current overrange threshold, this bit is set.
7 DL_OK When downloading a user filter to the AD7760, a checksum is generated. This checksum is compared to the one
downloaded following the coefficients. If these checksums agree, this bit is set.
6 FILTOK 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.
5 UFILT If a user-defined filter is in use, this bit is set.
4 BYP F3 Bypass Filter 3. If Filter 3 is bypassed by setting the relevant bit in Control Register 1, this bit is also set.
3 BYP F1 Bypass Filter 1. If Filter 1 is bypassed by setting the relevant bit in Control Register 1, this bit is also set.
2 to 0 DEC [2:0] Decimation Rate. These bits correspond to the bits set in Control Register 1.

OFFSET REGISTER—ADDRESS 0x0003 OVERRANGE REGISTER—ADDRESS 0x0005

Non-bit-mapped, Default Value 0x0000 Non-bit-mapped, Default Value 0xCCCC
The offset register uses twos complement notation and is scaled The overrange register value is compared with the output of
such that 0x7FFF (maximum positive value) and 0x8000 (max- the first decimation filter to obtain an overload indication with
imum negative value) correspond to an offset of +0.78125% and minimum propagation delay. This is prior to any gain scaling
−0.78125%, respectively. Offset correction is applied after any or offset adjustment. The default value is 0xCCCC, which
gain correction. Using the default gain value of 1.25 and assuming corresponds to 80% of VREF (the maximum permitted analog
a reference voltage of 4.096 V, the offset correction range is input voltage). Assuming VREF = 4.096 V, the bit is then set when
approximately ±25 mV. the input voltage exceeds approximately 6.55 V p-p differential.
Once the overrange bit is set, the DVALID bit in the status bits
GAIN REGISTER—ADDRESS 0x0004 of the AD7760 ouptut is set to zero, providing another indication
Non-bit-mapped, Default Value 0xA000 that an input overrange has occurred. Note that the overrange
The gain register is scaled such that 0x8000 corresponds to a bit is set immediately if the analog input voltage exceeds 100% of
gain of 1.0. The default value of this register is 1.25 (0xA000). VREF for more than four consecutive samples at the modulator rate.
This results in a full-scale digital output when the input is at
80% of VREF, tying in with the maximum analog input range of
±80% of VREF p-p.

Rev. A | Page 34 of 36
 AD7760

OUTLINE DIMENSIONS
12.20
12.00 SQ
0.75 1.20
MAX 11.80
0.60
0.45 64 49 49 64
1 48 48 1
PIN 1

10.20
TOP VIEW 10.00 SQ
EXPOSED 6.00
(PINS DOWN) 9.80 PAD BSC SQ

1.05 0° MIN
0.20
1.00 0.09 BOTTOM VIEW
0.95 (PINS UP)
7° 16 33 33 16
3.5° 17 32 32 17
0.15 0°
SEATING
0.05 PLANE 0.08 MAX VIEW A
COPLANARITY 0.50 0.38
BSC 0.32
LEAD PITCH
0.22
VIEW A
ROTATED 90° CCW
COMPLIANT TO JEDEC STANDARDS MS-026-ACD-HD

Figure 59. 64-Lead Thin Quad Flat Package, Exposed Pad [TQFP_EP]
(SV-64-2)
Dimensions shown in millimeters

ORDERING GUIDE
Model Temperature Range Package Description Package Option
AD7760BSVZ 1 −40°C to +85°C 64-Lead Thin Quad Flat Package, Exposed Pad [TQFP_EP] SV-64-2
AD7760BSVZ-REEL1 −40°C to +85°C 64-Lead Thin Quad Flat Package, Exposed Pad [TQFP_EP] SV-64-2
EVAL-AD7760EB Evaluation Board
1
Z = Pb-free part.

Rev. A | Page 35 of 36
AD7760

NOTES

©2006 Analog Devices, Inc. All rights reserved. Trademarks and

registered trademarks are the property of their respective owners.
D04975-0-8/06(A)

Rev. A | Page 36 of 36