a Ultralow Drift,

Dual BiFET Op Amp

AD647
FEATURES
Low Offset Voltage Drift
Matched Offset Voltage
Matched Offset Voltage Over Temperature
Matched Bias Currents
Crosstalk: 124 dB at 1 kHz
Low Bias Current: 35 pA max Warmed Up
Low Offset Voltage: 250 mV max
Low Input Voltage Noise: 2 mV p-p
High Open Loop Gain: 108 dB
Low Quiescent Current: 2.8 mA max
Low Total Harmonic Distortion
Standard Dual Amplifier Pinout
Available in Hermetic Metal Can Package, Hermetic
Surface Mount (20-Pin LCC) and Chip Form
MIL-STD-883B Processing Also Available
Single Version Available: AD547

PRODUCT DESCRIPTION
The AD647 is an ultralow drift, dual JFET amplifier that com-
bines high performance and convenience in a single package.
The AD647 uses the most advanced ion-implantation and laser
wafer drift trimming technologies to achieve the highest perfor-
mance currently available in a dual JFET. Ion-implantation per- PRODUCT HIGHLIGHTS
mits the fabrication of matched JFETs on a monolithic bipolar 1. The AD647 is guaranteed and tested to tight matching speci-
chip. Laser wafer drift trimming trims both the initial offset volt- fications to ensure high performance and to eliminate the se-
age and its drift with temperature to provide offsets as low as lection and matching of single devices.
100 V (250 V max) and drifts of 2.5 V/C max.
2. Laser wafer drift trimming reduces offset voltage and offset
In addition to outstanding individual amplifier performance, the voltage drifts to 250 V and 2.5 V/C max.
AD647 offers guaranteed and tested matching performance on
3. Voltage noise is guaranteed at 4 V p-p max (0.1 Hz to
critical parameters such as offset voltage, offset voltage drift and
10 Hz) on K, L and S grades.
bias currents.
4. Bias current (35 pA K, L, S; 75 pA J) is specified after five
The high level of performance makes the AD647 especially well
minutes of operation.
suited for high precision instrumentation amplifier applications
that previously would have required the costly selection and 5. Total supply current is a low 2.8 mA max.
matching of space wasting single amplifiers. 6. High open loop gain ensures high linearity in precision instru-
The AD647 is offered in four performance grades, three com- mentation amplifier applications.
mercial (the J, K and L) and one extended (the S). All are sup- 7. The standard dual amplifier pinout permits the direct substi-
plied in hermetically sealed 8-pin TO-99 packages and are tution of the AD647 for lower performance devices.
available processed to MIL-STD-883B. The LCC version is
8. The AD647 is available in chip form.
also available processed to MIL-STD-883B.

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
which may result from its use. No license is granted by implication or One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
otherwise under any patent or patent rights of Analog Devices. Tel: 617/329-4700 Fax: 617/326-8703
AD647SPECIFICATIONS (@ +258C and V = 615 V dc) S

Model AD647J AD647K AD647L AD647S

Min Typ Max Min Typ Max Min Typ Max Min Typ Max Units
OPEN LOOP GAIN
VO = 10 V, RL 2 k 100,000 250,000 250,000 250,000 V/V
TMIN to TMAX, RL = 2 k 100,000 250,000 250,000 100,000 V/V
OUTPUT CHARACTERISTICS
Voltage @ R L = 2 k, TMIN to TMAX 610 12 610 12 610 12 610 12 V
Voltage @ RL = 10 k, TMIN to TMAX 612 13 612 13 612 13 612 13 V
Short Circuit Current 25 25 25 25 mA
FREQUENCY RESPONSE
Unity Gain Small Signal 1.0 1.0 1.0 1.0 MHz
Full Power Response 50 50 50 50 kHz
Slew Rate, Unity Gain 2.0 3.0 2.0 3.0 2.0 3.0 2.0 3.0 V/s
INPUT OFFSET VOLTAGE 1
Initial Offset 1.0 0.5 0.25 0.5 mV
Input Offset Voltage vs. Temperature 10 5 2.5 5.0 V/C
Input Offset Voltage vs. Supply,
TMIN to TMAX 200 100 100 100 V/V
INPUT BIAS CURRENT 2
Either Input 10 75 10 35 10 35 10 35 pA
Offset Current 5 2 2 2 pA
MATCHING CHARACTERISTICS 3
Input Offset Voltage 1.0 0.5 0.25 0.5 mV
Input Offset Voltage T MIN to TMAX 10 5 2.5 10.0 V/C
Input Bias Current 35 25 25 25 pA
Crosstalk 124 124 124 124 dB
INPUT IMPEDANCE
Differential 1012i6 1012i6 1012i6 1012i6 MipF
Common Mode 1012i6 1012i6 1012i6 1012i6 MipF
INPUT VOLTAGE RANGE
Differential 4 20 20 20 20 V
Common Mode 610 12 610 12 610 12 610 12 V
Common-Mode Rejection 76 80 80 80 dB
INPUT NOISE
Voltage 0.1 Hz to 10 Hz 2 4 4 4 V p-p
f = 10 Hz 70 70 70 70 nV/Hz
f = 100 Hz 45 45 45 45 nV/Hz
f = 1 kHz 30 30 30 30 nV/Hz
f = 10 kHz 25 25 25 25 nV/Hz
POWER SUPPLY
Rated Performance 15 15 15 15 V
Operating 5 18 5 18 5 18 5 18 V
Quiescent Current 2.8 2.8 2.8 2.8 mA
TEMPERATURE RANGE
Operating, Rated Performance 0 +70 0 +70 0 +70 55 +125 C
Storage 65 +150 65 +150 65 +150 65 +150 C
PACKAGE OPTION
TO-99 Style (H-08B) AD647JH AD647KH AD647LH AD647SH
LCC (E-20A) AD647SE
AD647SE/883BH
NOTES
1
Input Offset Voltage specifications are guaranteed after 5 minutes of operation at TA = +25C.
2
Bias Current specifications are guaranteed at maximum at either input after 5 minutes of operation at TA = +25C. For higher temperatures, the current doubles every 10C.
3
Matching is defined as the difference between parameters of the two amplifiers.
4
Defined as the maximum safe voltage between inputs, such that neither exceeds 10 V from ground.
Specifications shown in boldface are tested on all production units at final electrical test. Results from those tests are used to calculate outgoing quality levels. All min and max specifications
are guaranteed, although only those shown in boldface are tested on all production units.
Specifications subject to change without notice.
METALIZATION PHOTOGRAPH
Dimensions shown in inches and (mm).
Contact factory for latest dimensions.

2 REV. A
 Typical CharacteristicsAD647

REV. A 3
AD647

4 REV. A
 AD647
APPLICATION NOTES THE AD647 USED WITH THE AD7546
The AD647 is fully specified under actual operating conditions Figure 24 shows the AD647 used with the AD7546 16-bit
to insure high performance in any application, but there are some segment DAC. In this application, amplifier performance is
steps that will improve on even this high level of performance. critical to the overall performance of the AD7546. A1 is used as
The bias current of a JFET amplifier doubles with every 10C a dual precision buffer. Here the offset voltage match, low offset
increase in junction temperature. Any heat source that can be voltage and high open loop gain of the AD647 ensure monoto-
eliminated or minimized will significantly improve bias current nicity and high linearity over the entire operating temperature
performance. To account for normal power dissipation, the larg- range. A2 serves a dual function amplifier A is a Track and
est contributor to chip self-heating, the bias currents of the Hold circuit that deglitches the DAC output and amplifier B
AD647 are guaranteed fully warmed up with 15 V supplies. A acts as an output amplifier. The performance of the amplifiers
decrease in supply voltage will decrease power consumption, of A2 is crucial to the accuracy of the system. The errors of
resulting in a corresponding drop in bias currents. these amplifiers are added to the errors due strictly to DAC im-
perfections. For this reason great care should be used in the se-
Open loop gain and bias currents, to some extent, are affected lection of these amplifiers. The matching characteristics, low
by output loading. In applications where high linearity is essen- bias current and low temperature coefficients of the AD647
tial, load impedance should be kept as high as possible to mini- make it ideal for this application.
mize degradation of open loop gain.
The outstanding ac and dc performance of the AD647 make it
an ideal choice for critical instrumentation applications. In such
applications, leakage paths, line losses and external noise
sources should be considered in the layout of printed circuit
boards. A guard ring surrounding the inputs and connected to a
low impedance potential (at the same level as the inputs) should
be placed on both sides of the circuit board. This will eliminate
leakage paths that could degrade bias current performance. All
signal paths should be shielded to minimize noise pickup.

Figure 24. AD647 Used with AD7546 16-Bit DAC

USING THE AD647 IN LOG AMPLIFIER APPLICATIONS

Log amplifiers or log ratio amplifiers are useful in a wide range
of analog computational applications, ranging from the simple
linearization of exponential transducer outputs to the use of
logarithms in computations involving multi-term products or ar-
bitrary exponents. Log amps also facilitate the compression of
wide ranging analog input signals into a range that can be easily
handled using standard circuit techniques.
Figure 23. AD647 Used as DAC Output Amplifier

A CMOS DAC AMPLIFIER

The output impedance of a CMOS DAC, such as the AD7541,
varies with digital input code. This causes a corresponding
variation in the noise gain of the DAC-amplifier combination.
This noise gain modulation introduces a nonlinearity whose
magnitude is dependent on the amount of offset voltage present.
Laser wafer drift trimming lowers the initial offset voltage and
the offset voltage drift of the AD647, therefore minimizing the
effect of this nonlinearity and its drift with temperature. This, in
conjunction with the low bias current and high open loop gain,
makes the AD647 ideal for DAC output amplifier applications.

Figure 25. Log-Ratio Amplifier

REV. A 5
AD647
The picoamp level input current and low offset voltage of the cuits. DC error sources such as offset voltage and bias currents
AD647 make it suitable for wide dynamic range log amplifiers. represent the largest individual contributors to output error.
Figure 25 is a schematic of a log ratio circuit employing the Offset voltages will be passed by the filtering network and may,
AD647 that can achieve less than 1% conformance error over depending on the design of the filter circuit, be amplified and
5 decades of current input, 1 nA to 100 A. For voltage inputs, generate unacceptable output offset voltages. In filter circuits for
the dynamic range is typically 50 mV to 10 V for 1% error, low frequency ranges large value resistors are used to generate
limited on the low end by the amplifiers input offset voltage. the low-pass filter function. Input bias currents passing through

C683a32/84
The conversion between current (or voltage) input and log out- these resistors will generate an additional offset voltage that will
put is accomplished by the base-emitter junctions of the dual also be passed to the output of the filter.
transistor Q1. Assuming Q1 has > 100, which is the case for The use of the AD647 will minimize these error sources and,
the specified transistor, the base-emitter voltage on side 1 is to a therefore, maximize filter accuracy. The wide variety of perfor-
close approximation mance levels of the AD647 allows for just the amount of accu-
VBE A = kT/q ln I1/IS1 racy required for any given application.
This circuit is arranged to take the difference of the VBEs of Q1A AD647 AS AN INSTRUMENTATION AMPLIFIER
and Q1B, thus producing an output voltage proportional to the The circuit shown in Figure 26 uses the AD647 to construct an
log of the ratio of the inputs ultra high precision instrumentation amplifier. In this type of
KkT application the matching characteristics of a monolithic dual
VOUT = K (VBE A VBE B ) = q (ln I1/IS1 ln I2/IS2) amplifier are crucial to ensure high performance.
VOUT = K kT/q ln Il /I2
The scaling constant, K is set by R1 and RTC to about 16, to
produce a 1 V change in output voltage per decade difference in
input signals. RTC is a special resistor with a +3500 ppm/C
temperature coefficient, which makes K inversely proportional
to temperature, compensating for the T in kT/q. The log ratio
transfer characteristic is therefore independent of temperature.
This particular log ratio circuit is free from the dynamic prob-
lems that plague many other log circuits. The 3 dB bandwidth
is 50 kHz over the top 3 decades, 100 nA to 100 A, and de-
creases smoothly at lower input levels. This circuit needs no
additional frequency compensation for stable operation from
input current sources, such as photodiodes, which may have
100 pF of shunt capacitance. For larger input capacitances a
20 pF integration capacitor around each amplifier will provide a Figure 26. Precision FET Input Instrumentation Amplifier
smoother frequency response. The use of an AD647L as the input amplifier A1, guarantees
This log ratio amplifier can be readily adjusted for optimum maximum offset voltage of 250 V, drift of 2.5 V/C and bias
accuracy by following this simple procedure. First, apply V1 = currents of 35 pA. A2 serves two less critical functions in the
V2 = 10.00 V and adjust Balance for VOUT = 0.00 V. Next amplifier and, therefore can be an AD647J. Amplifier A is an ac-
apply V1 = 10.00 V, V2 = 100 V and adjust gain for VOUT = tive data guard which increases ac CMRR and minimizes extra-
+1.00 V. Repeat this procedure until gain and balance readings neous signal pickup and leakage. Amplifier B is the output
are within 2 mV of ideal values. amplifier of the instrumentation amplifier. To attain the preci-
sion available from this configuration, a great deal of care
ACTIVE FILTERS should be taken when selecting the external components.
In active low-pass filtering applications the dc accuracy of the CMRR will depend on the matching of resistors R1, R2, R3,
and R4. The gain drift performance of this circuit will be af-
PRINTED IN U.S.A.
amplifiers used is critical to the performance of the filter cir-
fected by the matching TC of the resistors used.
OUTLINE DIMENSIONS
Dimensions shown in inches and (mm).
TO-99 E-20A

6 REV. A