Chapter III: Monolithic Instrumentation Amplifiers

Chapter III
MONOLITHIC INSTRUMENTATION AMPLIFIERS
Advantages Over Op Amp In-Amps
Monolithic IC instrumentation amplifiers were developed to satisfy the demand for in-amps that would be
easier to apply. These circuits incorporate variations in
the 3-op amp and 2-op amp in-amp circuits previously
described, while providing laser-trimmed resistors
and other benefits of monolithic IC technology. Since
both active and passive components are now within
the same die, they can be closely matched—this
will ensure that the device provides a high CMR. In
addition, these components will stay matched over
temperature, ensuring excellent performance over a
wide temperature range. IC technologies such as laser
wafer trimming allow monolithic integrated circuits
to be tuned up to very high accuracy and provide low
cost, high volume manufacturing. An additional advantage of monolithic devices is that they are available
in very small, very low cost SOIC, MSOP, or LFCSP
(chip scale) packages designed for use in high volume
production. Table 3-1 provides a quick performance
summary of Analog Devices in-amps.
Which to Use—an In-Amp or a Diff Amp?
Although instrumentation amplifiers and difference
amplifiers share many properties, the first step in the
design process should be which type of amplifier
to use.
A difference amplifier is basically an op amp subtractor,
typically using high value input resistors. The resistors
provide protection by limiting the amplifier’s input
current. They also reduce the input common-mode
and differential voltage to a range that can be handled
by the internal subtractor amplifier. In general, difference amplifiers should be used in applications where the
common-mode voltage or voltage transients may exceed
the supply voltage.
Table 3-1. Latest Generation Analog Devices In-Amps Summarized1
Product Features
Power
Supply
Current
Typ
AD8221 Precision, high BW
0.9 mA
AD620 General-purpose
0. 9 mA
AD8225 Precision gain = 5
1.1 mA
AD8220R-R, JFET input
750 mA
AD8222 Dual, precision, high BW
1.8 mA
AD8230R-R, zero drift
2.7 mA
AD8250 High BW, programmable gain 3.5 mA
AD8251 High BW, programmable gain 3.5 mA
AD8553 Auto-zero with shutdown
1.1 mA
AD8555 Zero drift dig prog
2.0 mA
AD8556 Dig prog IA with filters
2.0 mA
AD622Low cost
0.9 mA
AD621 Precise gain
0.9 mA
AD623Low cost, S.S.
375 A
AD627 Micropower, S.S.
60 A
–3 dB
BW
Typ
(G = 10)
CMR
G = 10
(dB)
Min
Input
Offset
Voltage
Max
VOS
RTI
Input
Drift Noise2 Bias
(V/C) (nV/√Hz) Current
Max
(G = 10) (nA) Max
560 kHz
800 kHz
900 kHz4
1500 kHz
750 kHz
2 kHz
3.5 MHz
3.5 MHz
1 kHz
700 kHz6
700 kHz6
800 kHz
800 kHz
800 kHz
80 kHz
1003
953
834, 5
100
1003
110
100
100
100
806
806
863
933
903
100
60 V
125 V
150 V
250 mV
120 V
10 V
100 V
100 V
20 V
10 V
10 V
125 V
250 V7
200 V
250 V
0.4
1
0.3
5
0.4
10
1
1
0.1
0.07
0.07
1
2.57
2
3
NOTES
S.S. = single supply.
1
Refer to ADI website at www.analog.com for latest products and specifications.
2
At 1 kHz. RTI noise = √((eni)2 + (eno/G)2).
3
For dc to 60 Hz, 1 k source imbalance.
4
Operating at a gain of 5.
5
For 10 kHz, 1 k source imbalance.
6
Operating at a gain of 70.
Referred to input (RTI).
7
3-1
11 max
16 max
45 typ4
17 typ
11 max
240 typ
13 typ
13 typ
150 typ
32 typ
32 typ
14 typ
17 max7
35 typ
42 typ
1.5
2
1.2
10 pA
2
1
15
15
1
22
54
5
2
25
10
VB
I
A1
IB COMPENSATION
I
A2
IB COMPENSATION
C1
10k�
C2
+VS
10k�
+VS
–IN
R1
24.7k�
400�
Q1
+VS
R2
24.7k�
+VS
+VS
+VS
–VS
400�
Q2
+IN
RG
–VS
OUTPUT
A3
10k�
REF
10k�
–VS
–VS
–VS
–VS
Figure 3-1. AD8221 simplified schematic.
In contrast, an instrumentation amplifier is most commonly
an op amp subtractor with two input buffer amplifiers
(these increase the input Z and thus reduce loading of the
input source). An in-amp should be used when the total
input common-mode voltage plus the input differential
voltage, including transients, is less than the supply voltage. In-amps are also needed in applications where the
highest accuracy, best signal-to-noise ratio, and lowest
input bias current are essential.
Monolithic In-Amp Design—The
Inside Story
High Performance In-Amps
Analog Devices introduced the first high performance
monolithic instrumentation amplifier, the AD520,
in 1971.
In 2003, the AD8221 was introduced. This in-amp is
in a tiny MSOP package and offers increased CMR
at higher bandwidths than other competing in-amps.
It also has improved ac and dc specifications over the
industry-standard AD620 series in-amps.
The AD8221 is a monolithic instrumentation amplifier
based on the classic 3-op amp topology (Figure 3-1).
Input transistors Q1 and Q2 are biased at a constant
current so that any differential input signal will force the
output voltages of A1 and A2 to be equal. A signal applied
to the input creates a current through RG, R1, and R2
such that the outputs of A1 and A2 deliver the correct
voltage. Topologically, Q1, A1, R1 and Q2, A2, R2 can
be viewed as precision current feedback amplifiers. The
amplified differential and common-mode signals are
applied to a difference amplifier, A3, which rejects the
common-mode voltage, but processes the differential
voltage. The difference amplifier has a low output
offset voltage as well as low output offset voltage drift.
Laser-trimmed resistors allow for a highly accurate
in-amp with gain error typically less than 20 ppm and
CMRR that exceeds 90 dB (G = 1).
Using superbeta input transistors and an I B compensation scheme, the AD8221 offers extremely high input
impedance, low I B, low IOS, low I B drift, low input
bias current noise, and extremely low voltage noise of
8 nV/√Hz.
The transfer function of the AD8221 is
G=
49.4 kΩ
+1
RG
49.4 kΩ
G −1
Care was taken to ensure that a user could easily and
accurately set the gain using a single external standard
value resistor.
RG =
Since the input amplifiers employ a current feedback
architecture, the AD8221’s gain bandwidth product
increases with gain, resulting in a system that does not
suffer from the expected bandwidth loss of voltage feedback architectures at higher gains.
In order to maintain precision even at low input levels,
special care was taken with the AD8221’s design and
layout, resulting in an in-amp whose performance
satisfies even the most demanding applications (see
Figures 3-3 and 3-4).
3-2
–IN 1
8
+VS
RG 2
7
VOUT
RG 3
6
+IN 4
5
AD8221
VREF
–VS
OUT1
OUT2
VEE
The AD8222 (Figure 3-5) is a dual version of the
AD8221 in-amp, with similar performance and
specifications. Its small size allows more amplifiers
per PC board. In addition, the AD8222 is the first
in-amp to be specified for differential output performance. It is available in a 4 mm 3 4 mm, 16-lead
LFCSP package.
VCC
A unique pinout enables the AD8221 to meet an
unparalleled CMRR specification of 80 dB at 10 kHz
(G = 1) and 110 dB at 1 kHz (G = 1000). The balanced
pinout, shown in Figure 3-2, reduces the parasitics that
had, in the past, adversely affected CMR performance.
In addition, the new pinout simplifies board layout
because associated traces are grouped. For example,
the gain setting resistor pins are adjacent to the inputs,
and the reference pin is next to the output.
16
15
14
13
–IN1
1
12
–IN2
RG1
2
11
RG2
RG1
3
10
RG2
+IN1
4
9
+IN2
140
CMRR (dB)
100
8
Figure 3-5. AD8222 connection diagram.
For many years, the AD620 has been the industrystandard, high performance, low cost in-amp. The
AD620 is a complete monolithic instrumentation
amplifier offered in both 8-lead DIP and SOIC packages.
The user can program any desired gain from 1 to
1000 using a single external resistor. By design, the
required resistor values for gains of 10 and 100 are
standard 1% metal film resistor values.
GAIN = 100
120
7
VEE
GAIN = 1000
6
REF2
160
5
VCC
Figure 3-2. AD8221 pinout.
REF1
TOP VIEW
GAIN = 10
GAIN = 1000
GAIN = 1
GAIN = 10
80
GAIN = 100
60
40
0.1
1
10
100
1k
10k
100k
1M
FREQUENCY (Hz)
Figure 3-3. CMRR vs. frequency (RTI) of the AD8221.
GAIN = 1000
GAIN (dB)
GAIN = 10
GAIN = 1
–10
–20
–30
100
1k
63
). /54054
2%&
!$
The AD620 (see Figure 3-7) is a second-generation
version of the classic AD524 in-amp and embodies a
modification of its 3-op amp circuit. Laser trimming
of on-chip thin film resistors, R1 and R2, allows the
user to accurately set the gain to 100 within 0.3%
max error, using only one external resistor. Monolithic
construction and laser wafer trimming allow the tight
matching and tracking of circuit components.
10
0
n). Figure 3-6. AD620 pin configuration.
GAIN = 100
30
20
2'
4/06)%7
.OTTO3CALE
50
40
n63 70
60
2' 10k
100k
FREQUENCY (Hz)
1M
Figure 3-4. AD8221 closed-loop gain vs.
frequency.
10M
A preamp section comprised of Q1 and Q2 provides
additional gain up front. Feedback through the Q1-A1R1 loop and the Q2-A2-R2 loop maintains a constant
collector current through the input devices Q1 and Q2,
3-3
+VS
I1
20�A
VB
I2
20�A
IB
COMPENSATION
IB COMPENSATION
A1
A2
10k�
C2
C1
10k�
OUTPUT
A3
– IN
R3
400�
R1
10k�
R2
Q1
Q2
R4
400�
RG
GAIN
SENSE
10k�
REF
+IN
GAIN
SENSE
–VS
Figure 3-7. A simplified schematic of the AD620.
The AD620 also has superior CMR over a wide frequency range, as shown in Figure 3-9.
The value of RG also determines the transconductance
of the preamp stage. As RG is reduced for larger gains,
the transconductance increases asymptotically to that
of the input transistors.This has important advantages:
First, the open-loop gain is boosted for increasing
programmed gain, thus reducing gain related errors.
Second, the gain bandwidth product (determined by
C1, C2, and the preamp transconductance) increases
with programmed gain, thus optimizing the amplifier’s
frequency response. Figure 3-8 shows the AD620’s
closed-loop gain vs. frequency.
1000
GAIN (V/V)
100
10
1
0.1
100
1k
10k
100k
FREQUENCY (Hz)
1M
10M
Figure 3-8. AD620 closed-loop gain vs. frequency.
3-4
160
140
120
100
CMR (dB)
thereby impressing the input voltage across the external
gain setting resistor, RG. This creates a differential
gain from the inputs to the A1/A2 outputs given by
G = (R1 + R2)/RG + 1. The unity-gain subtractor, A3,
removes any common-mode signal, yielding a singleended output referred to the REF pin potential.
80
G = 1000
G = 100
G = 10
G=1
60
40
20
0
0.1
1
10
100
1k
FREQUENCY (Hz)
10k
100k
Figure 3-9. AD620 CMR vs. frequency.
1M
Figures 3-10 and 3-11 show the AD620’s gain nonlinearity and small signal pulse response.
100�V
The value of 24.7 k was chosen so that standard
1% resistor values could be used to set the most
popular gains.
Low Cost In-Amps
2V
The AD622 is a low cost version of the AD620 (see
Figure 3-6). The AD622 uses streamlined production
methods to provide most of the performance of the
AD620 at lower cost.
100 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
90
Figures 3-12, 3-13, and 3-14 show the AD622’s CMR
vs. frequency, gain nonlinearity, and closed-loop gain
vs. frequency.
10
0%
.... .... .... ........ .... .... .... ........
160
CMR (dB)
Figure 3-10. AD620 gain nonlinearity
(G = 100, RL = 10 k, vertical scale: 100 V =
10 ppm, horizontal scale: 2 V/div).
20V
140
G = 1000
120
G = 100
100
G = 10
80
G=1
60
40
.... .... .... ........ ........ .... ........
100
90
20
0
0.1
1
10
100
1k
FREQUENCY (Hz)
10k
100k
1M
Figure 3-12. AD622 CMR vs. frequency
((RTI) 0 to 1 k source imbalance).
10
.... .... .... ........ ........ .... ........
0%
10�s
10�s
2V
100
Figure 3-11. Small signal pulse response of
the AD620 (G = 10, RL= 2 k, CL = 100 pF).
90
Finally, the input voltage noise is reduced to a value of
9 nV/√Hz, determined mainly by the collector current
and base resistance of the input devices.
The internal gain resistors, R1 and R2, are trimmed to
an absolute value of 24.7 k, allowing the gain to be
programmed accurately with a single external resistor.
The gain equation is then
G=
Figure 3-13. AD622 Gain nonlinearity
(G = 1, RL = 10 k, vertical scale: 20 V = 2 ppm).
49.4 kΩ
+1
RG
So that
RG =
10
0%
49.4 kΩ
G −1
Where resistor RG is in k.
3-5
Pin-Programmable, Precise Gain In-Amps
1000
The AD621 is similar to the AD620, except that for
gains of 10 and 100 the gain setting resistors are on the
die—no external resistors are used. A single external
jumper (between Pins 1 and 8) is all that is needed to
select a gain of 100. For a gain of 10, leave Pin 1 and
Pin 8 open. This provides excellent gain stability over
temperature, as the on-chip gain resistor tracks the TC
of the feedback resistor. Figure 3-15 is a simplified
schematic of the AD621. With a max total gain error of
0.15% and 5 ppm/C gain drift, the AD621 has much
greater built-in accuracy than the AD620.
GAIN (V/V)
100
10
1
0
100
1k
10k
100k
FREQUENCY (Hz)
1M
The AD621 may also be operated at gains between 10
and 100 by using an external gain resistor, although gain
error and gain drift over temperature will be degraded.
Using external resistors, device gain is equal to
10M
Figure 3-14. AD622 closed-loop
gain vs. frequency.
G = (R1 + R2)/RG + 1
+VS
7
I1
20�A
VB
20�A
IB
COMPENSATION
I2
IB COMPENSATION
A1
A2
C1
10k�
C2
10k�
A3
– IN
R3
400�
2
R1
Q1
25k� R2
R5
5555.6�
R6
555.6�
1
G = 100
25k�
Q2
10k�
R4
400�
6
10k�
3
+IN
8
G = 100
4
–VS
Figure 3-15. A simplified schematic of the AD621.
3-6
OUTPUT
5
REF
Figures 3-16 and 3-17 show the AD621’s CMR vs.
frequency and closed-loop gain vs. frequency.
Figures 3-18 and 3-19 show the AD621’s gain nonlinearity and small signal pulse response.
160
100�s
140
GAIN = 100
100
90
120
GAIN = 10
100
CMR (dB)
2V
80
60
10
40
0%
20
0
0.1
1
10
100
1k
10k
100k
1M
Figure 3-18. AD621 gain nonlinearity
(G = 10, RL = 10 k, vertical scale: 100 V/div
= 100 ppm/div, horizontal scale 2 V/div).
FREQUENCY (Hz)
Figure 3-16. AD621 CMR vs. frequency.
1000
20mV
10�s
CLOSED-LOOP GAIN (V/V)
100
90
100
10
10
1
0%
0.1
100
1k
10k
100k
FREQUENCY (Hz)
1M
10M
Figure 3-19. Small signal pulse response of
the AD621 (G = 10, RL = 2 k, CL = 100 pF).
Figure 3-17. AD621 closed-loop gain vs.
frequency.
3-7
n). 63
2' 6/54
2' 62%&
n63
). !$
A single resistor sets the gain from 1 to 1000. The AD8220
operates on both single and dual supplies and is wellsuited for applications where input voltages close to those
of the supply are encountered. In addition, its rail-to-rail
output stage allows for maximum dynamic range, when
constrained by low single-supply voltages.
Auto-Zeroing Instrumentation Amplifiers
4/06)%7
Figure 3-20. AD8220 connection diagram.
The AD8220 is a FET input, gain-programmable, high
performance instrumentation amplifier with a max input
bias current of 10 pA. It also features excellent high
frequency common-mode rejection (see Figure 3-20).
The AD8220 maintains a minimum CMRR of 70 dB
up to 20 kHz, at G = 1. The combination of extremely
high input impedance and high CMRR over frequency
makes the AD8220 useful in applications such as patient
monitoring. In these applications, input impedance is high
and high frequency interference must be rejected.
Auto-zeroing is a dynamic offset and drift cancellation
technique that reduces input referred voltage offset to
the V level, and voltage offset drift to the nV/C level.
The AD8230 (Figure 3-22) is an instrumentation amplifier that utilizes an auto-zeroing topology and combines
it with high common-mode signal rejection.
CMRR (dB)
GAIN = 1000
GAIN = 100
GAIN = 10
BANDWIDTH
LIMITED
100
GAIN = 1
80
100
1k
10k
+VS 2
7
RG
VREF1 3
6
VREF2
5
–IN
AD8230
TOP VIEW
Figure 3-22. AD8230 connection diagram.
60
40
10
VOUT
The internal signal path consists of an active differential sample-and-hold stage (preamp), followed by
a differential amplifier (gain amp). Both amplifiers
implement auto-zeroing to minimize offset and drift.
A fully differential topology increases the immunity
of the signals to parasitic noise and temperature effects.
Amplifier gain is set by two external resistors for
convenient TC matching. The AD8230 can accept
input common-mode voltages within and including
the supply voltages (5 V).
160
120
8
+IN 4
The rail-to-rail output, low power consumption
and small MSOP/CSP package make this precision
instrumentation amplifier attractive for use in multichannel applications.
140
–VS 1
100k
FREQUENCY (Hz)
Figure 3-21. Typical AD8220 CMRR vs. frequency.
3-8
The signal sampling rate is controlled by an on-chip,
10 kHz oscillator and logic to derive the required nonoverlapping clock phases. For simplification of the functional
description, two sequential clock phases, A and B, will
be used to distinguish the order of internal operation as
depicted in Figures 3-23 and 3-24, respectively.
During Phase A, the sampling capacitors are connected
to the input signals at the common-mode potential. The
input signal’s difference voltage, VDIFF, is stored across
the sampling capacitors, CSAMPLE. The common-mode
potential of the input affects CSAMPLE insofar as the
sampling capacitors are at a different common-mode
potential than the preamp. During this period, the gain
amp is disconnected from the preamp so that its output
remains at the level set by the previously sampled input
signal, held on CHOLD in Figure 3-23.
In Phase B, upon sampling the analog input signals,
the input common-mode component is removed.
The common-mode output of the preamp is held at the
reference potential, VREF. When the bottom plates of the
sampling capacitors connect to the output of the preamp,
the input signal common-mode voltage is pulled to the
amplifier’s common-mode voltage, VREF. In this manner,
the sampling capacitors are brought to the same commonmode voltage as the preamp. The remaining differential
signal is presented to the gain amp, refreshing the hold
capacitors’ signal potentials, as shown in Figure 3-24.
Figures 3-25 through 3-28 show the internal workings
of the AD8230 in depth. As noted, both the preamp and
gain amp auto-zero. The preamp auto-zeroes during
phase A, shown in Figure 3-25, while the sampling caps
are connected to the signal source. By connecting the
PREAMP
GAIN AMP
V+IN
VDIFF
+VCM
CHOLD
CSAMPLE
+
–
–
+
VOUT
CHOLD
V–IN
VREF
RG
RF
Figure 3-23. The AD8230 in Phase A sampling phase. The differential component of the input signal
is stored on sampling capacitors, CSAMPLE. The gain amp conditions the signal stored on the hold
capacitors, CHOLD. Gain is set with the RG and RF resistors.
PREAMP
GAIN AMP
V+IN
VDIFF
+VCM
CHOLD
CSAMPLE
+
–
–
+
VOUT
CHOLD
V–IN
RG
VREF
RF
Figure 3-24. In Phase B, the differential signal is transferred to the hold capacitors, refreshing the value
stored on CHOLD. The gain amp continues to condition the signal stored on the hold capacitors, CHOLD.
3-9
preamp differential inputs together, the resulting output
referred offset is connected to an auxiliary input port to the
preamp. Negative feedback operation forces a canceling
potential at the auxiliary port, which is subsequently
held on a storage capacitor, CP_HOLD.
While in Phase A, the gain amp shown in Figure 3-26
reads the previously sampled signal held on the holding
capacitors, CHOLD.The gain amp implements feedforward
offset compensation to allow for transparent nulling of
the main amp and a continuous output signal. A differential
signal regimen is maintained throughout the main amp and
feedforward nulling amp by utilizing a double differential
input topology. The nulling amp compares the input of
the two differential signals. As a result, the offset error
is fed into the null port of the main amp, VNULL, and
stored on CM_HOLD. This operation effectively forces
the differential input potentials at both the signal and
feedback ports of the main amp to be equal. This is the
requirement for zero offset.
PREAMP
V+IN
A
GAIN AMP
B
B
VDIFF
+VCM
CSAMPLE
A
+
–
–
+
B
VOUT
A
B
A
CHOLD
B
A
CHOLD
A
B
V–IN
CP_HOLD
VREF
RG
RF
Figure 3-25. Detailed schematic of the preamp during Phase A. The differential signal is
stored on the sampling capacitors. Concurrently, the preamp nulls its own offset and stores
the correction voltage on its hold capacitors, CP_HOLD.
PREAMP
GAIN AMP
Sn
NULLING AMP
B
A
A
B
fn
B
A
CN_HOLD
CHOLD
+
–
–
CM_HOLD
B
+
B
s
MAIN
AMP
CHOLD
VNULL
VOUT
f
VREF
RG
RF
Figure 3-26. Detailed schematic of the gain amp during Phase A. The main amp conditions
the signal held on the hold capacitors, CHOLD. The nulling amplifier forces the inputs of the
main amp to be equal by injecting a correction voltage into the VNULL port, removing the
offset of the main amp. The correction voltage is stored on CM_HOLD.
3-10
During Phase B, the inputs of the preamp are no longer
shorted, and the sampling capacitors are connected to the
input and output of the preamp as shown in Figure 3-27.
The preamp, having been auto-zeroed in Phase A, has
minimal offset. When the sampling capacitors are connected to the preamp, the common mode of the sampling
capacitors is brought to VREF. The preamp outputs the
difference signal onto the hold capacitors, CHOLD.
The main amp continues to output the gained difference signal, shown in Figure 3-28. Its offset is kept to a
minimum by using the nulling amp’s correction potential
stored on CM_HOLD from the previous phase. During this
phase, the nulling amp compares its two differential inputs
and corrects its own offset by driving a correction voltage
into its nulling port and, ultimately, onto CN_HOLD. In
this fashion, the nulling amp reduces its own offset in
Phase B before it corrects for the main amp’s offset in
the next phase, Phase A.
PREAMP
V+IN
A
GAIN AMP
B
B
VDIFF
+VCM
CSAMPLE
A
+
–
–
+
B
VOUT
A
B
A
CHOLD
B
A
CHOLD
A
B
V–IN
CP_HOLD
VREF
RG
RF
Figure 3-27. Detailed schematic of the preamp during Phase B. The preamp’s offset remains low because
it was corrected in the previous phase. The sampling capacitors connect to the input and output of the
preamp, and the difference voltage is passed onto the holding capacitors, CHOLD.
PREAMP
GAIN AMP
Sn
NULLING AMP
B
A
A
B
fn
B
A
CN_HOLD
CHOLD
+
–
–
CM_HOLD
B
+
B
s
MAIN
AMP
CHOLD
VNULL
VOUT
f
VREF
RG
RF
Figure 3-28. Detailed schematic of the gain amp during Phase B. The nulling amplifier nulls its own
offset by injecting a correction voltage into its own auxiliary port and storing it on CN_HOLD. The main
amplifier continues to condition the differential signal held on CHOLD, yet maintains minimal offset
because its offset was corrected in the previous phase.
3-11
Two external resistors set the gain of the AD8230. The
gain is expressed in the following function:

R 
Gain = 2 1 + F 
RG 

Figure 3-30 shows the AD8230’s common-mode rejection vs. frequency. Figure 3-31 is a plot of AD8230’s
gain flatness vs. frequency at a gain of 10.
+VS
0.1MF
#-27)4(./3/52#%)-"!,!.#%
10MF
0.1MF
#-2D"
–VS
10MF
2
4
AD8230
VREF2
5
VREF1
3
6
1
RG
8
#-27)4(K63/52#%)-"!,!.#%
VOUT
7
RF
RG
K
K
&2%15%.#9(Z
Figure 3-30. Common-mode rejection vs. frequency.
Figure 3-29. Gain setting.
Table 3-2. Gains Using Standard 1% Resistors
RF
2
10
50
100
200
500
1000
0 V (short)None
2 kV
8.06 kV
499 V
12.1 kV
200 V
9.76 kV
100 V
10 kV
200 V
49.9 kV
200 V
100 kV
RG
Actual Gain
2
10
50.5
99.6
202
501
1002
'!).D"
Gain
Figure 3-29 and Table 3-2 provide an example of some
gain settings. As Table 3-2 shows, the AD8230 accepts a
wide range of resistor values. Since the instrumentation
apmplifier has finite driving capability, ensure that the
output load in parallel with the sum of the gain setting
resistors is greater than 2 kV.
RL ( RF + RG ) > 2 kΩ
Offset voltage drift at high temperature can be minimized
by keeping the value of the feedback resistor, RF, small.
This is due to the junction leakage current on the RG
pin, Pin 7.
n
K
K
K
&2%15%.#9(Z
Figure 3-31. Gain vs. frequency, G = 10.
The AD8553 is a precision current-mode auto-zero
instrumentation amplifier capable of single-supply
operation. The current-mode correction topology
results in excellent accuracy, without the need for
trimmed resistors on the die.
3-12
2'
2'
n).
'.$
).
6##
6/54
62%&
6&"
%.!",%
AD8553
The pinout of the AD8553 allows the user to access the signal
current from the output of the voltage-to-current converter
(Pin 5). The user can choose to use the AD8553 as a
current-output device instead of a voltage-output device.
The AD8555 is a zero-drift, sensor signal amplifier with
digitally programmable gain and output offset. Designed
to easily and accurately convert variable pressure sensor
and strain bridge outputs to a well-defined output voltage
range, the AD8555 also accurately amplifies many other
differential or single-ended sensor outputs.
Figure 3-32. AD8553 connection diagram.
Figure 3-32 is the AD8553 connection diagram while
Figure 3-33 shows a simplified schematic illustrating
the basic operation of the AD8553 (without correction).
The circuit consists of a voltage-to-current amplifier
(M1 to M6), followed by a current-to-voltage amplifier
(R2 and A1). Application of a differential input voltage
forces a current through external resistor R1, resulting
in conversion of the input voltage to a signal current.
Transistors M3 to M6 transfer twice this signal current
to the inverting input of the op amp A1. Amplifier A1 and
external resistor R2 form a current-to-voltage converter
to produce a rail-to-rail output voltage at VOUT.
Figure 3-34 shows the pinout and Figure 3-35 the
simplified schematic.
6$$ &),4$)'/54
$)'). I
M5
IR1 =
VIN+
VIN+ – VIN–
R1
M1
6/54
6#,!-0
60/3
Figure 3-34. AD8555 connection diagram.
The AD8555 (and AD8556) use both auto-zeroing
and “chopping” techniques to maintin zero drift. A1,
A2, R1, R2, R3, P1, and P2 form the first gain stage of
the differential amplifier. A1 and A2 are auto-zeroed op
amps that minimize input offset errors. P1 and P2 are
digital potentiometers, guaranteed to be monotonic.
Programming P1 and P2 allows the first stage gain to
be varied from 4.0 to 6.4 with 7-bit resolution, giving a
fine gain adjustment resolution of 0.37%. R1, R2, R3,
P1, and P2 each have a similar temperature coefficient,
so the first stage gain temperature coefficient is lower
than 100 ppm/8C.
C2
I – IR1
R2
2IR1
VIN–
A1
I + IR1
M3
2I
633
M6
I – IR1
M2
4/06)%7
.OTTO3CALE
An external reference voltage is applied to the noninverting input of A1 for output-offset adjustment. Because
the AD8553 is essentially a chopper in-amp, some type of
low-pass filtering of the ouput is usually required. External
capacitor C2 is used to filter out high frequency noise.
R1
!$
6.%' Op amp A1 is a high precision auto-zero amplifier. This
amplifier preserves the performance of the autocorrection
current-mode amplifier topology while offering the user
a true voltage-in, voltage-out instrumentation amplifier.
Offset errors are corrected internally.
I
M4
VBIAS
VOUT = VREF +
VREF
2I
Figure 3-33. AD8553 simplified schematic.
3-13
2R2
R1
VIN+ – VIN–
VDD
VCLAMP
VDD
VNEG
A1
R4
P3
VDD
A2
VPOS
VDD
P1
RF
A3
R3
VDD
R6
VSS
R1
VSS
A5
A4
VOUT
P2
VSS
R2
R5
VDD
FILT/
DIGOUT
R7
VSS
P4
VSS
DAC
VSS
Figure 3-35. AD8555 simplified schematic.
A3, R4, R5, R6, R7, P3, and P4 form the second gain stage
of the differential amplifier. A3 is also an auto-zeroed op
amp that minimizes input offset errors. P3 and P4 are
digital potentiometers, allowing the second stage gain
to be varied from 17.5 to 200 in eight steps; they allow
the gain to be varied over a wide range. R4, R5, R6, R7,
P3, and P4 each have a similar temperature coefficient,
so the second stage gain temperature coefficient is lower
than 100 ppm/8C.
A5 implements a voltage buffer, which provides the
positive supply to the amplifier output buffer A4. Its
function is to limit VOUT to a maximum value, useful for
driving analog-to-digital converters (ADC) operating on
supply voltages lower thanVDD.The input to A5,VCLAMP,
has a very high input resistance. It should be connected
to a known voltage and not left floating. However, the
high input impedance allows the clamp voltage to be
set using a high impedance source (e.g., a potential
divider). If the maximum value of VOUT does not need
to be limited, VCLAMP should be connected to VDD.
A4 implements a rail-to-rail input and output unity-gain
voltage buffer. The output stage of A4 is supplied from
a buffered version of VCLAMP instead of VDD, allowing
the positive swing to be limited. The maximum output
current is limited between 5 to 10 mA.
An 8-bit digital-to-analog converter (DAC) is used to
generate a variable offset for the amplifier output. This
DAC is guaranteed to be monotonic. To preserve the
ratiometric nature of the input signal, the DAC references
are driven from VSS and VDD, and the DAC output can
swing from VSS (Code 0) to VDD (Code 255). The 8-bit
resolution is equivalent to 0.39% of the difference between
VDD and VSS (e.g., 19.5 mV with a 5 V supply). The DAC
output voltage (VDAC) is given approximately by
 Code + 0.5 
VDAC ≈ 
(V − VSS ) + VSS
 256  DD
The temperature coefficient of VDAC is lower than
200 ppm/8C.
3-14
The amplifier output voltage (VOUT) is given by
VOUT = GAIN (VPOS − VNEG ) + VDAC
60
CLOSED-LOOP GAIN (dB)
where GAIN is the product of the first and second
stage gains.
VS = 2.5V
GAIN = +70
120
CMRR (dB)
VS = 2.5V
GAIN = +1280
40
GAIN = +70
20
0
80
1k
100k
FREQUENCY (Hz)
1M
Figure 3-37. AD8555 closed-loop gain vs.
frequency measured at output pin.
40
0
100
10k
1k
10k
FREQUENCY (Hz)
100k
The AD8556 is essentially the same product as the
AD8555, except that the former includes internal RFI
filtering. The block diagram for the AD8556 is shown in
Figure 3-38. For theory of operation, refer to the previous
section that covers the AD8555.
1M
Figure 3-36. AD8555 CMRR vs. frequency.
Figures 3-36 and 3-37 show the AD8555’s CMRR vs.
frequency and its closed-loop gain vs. frequency.
See the AD8555 product data sheet for more details.
DIGIN
VDD
VCLAMP
VDD
1 +IN
EMI
FILTER
1 +IN
EMI
FILTER
2
OUT
3
–IN
VSS
VDD
VPOS
2
DAC
LOGIC
A5
VSS
A1
OUT
R5
3
–IN
P4
R7
R2
VSS
P2
VDD
EMI
FILTER
R3
VNEG
2
EMI
FILTER
OUT
VSS
A3
OUT
–IN
VDD
3
RF
EMI
FILTER
1 +IN
2
A4
OUT
–IN
VSS
A2
–IN
2
P1
VDD
1 +IN
1 +IN
R1
3
R4
VSS
R6
P3
AD8556
VSS
FILT/DIGOUT
Figure 3-38. AD8556 block diagram showing EMI/RFI built-in filters.
3-15
3
VOUT
+VS
+VS
+VS
VB
–IN
4006
Q1
A1
–VS
4006
Q2
R2
C2
UNITYGAIN
BUFFERS
A2
+IN
–VS
C1
R1
15k6
+VS
3k6
A3
3k6
GAIN-OF-5
DIFFERENCE AMPLIFIER
–VS
15k6
Figure 3-39. AD8225 simplified schematic.
Fixed Gain (Low Drift) In-Amps
VOUT
+VS
VREF
–VS
130
The AD8225 has a wide gain bandwidth product,
resulting from its being compensated for a fixed gain of
5, as opposed to the usual unity-gain compensation of
variable gain in-amps. High frequency performance is
also enhanced by the innovative pinout of the AD8225.
Since Pin 1 and Pin 8 are uncommitted, Pin 1 may be
connected to Pin 4. Since Pin 4 is also ac common, the
stray capacitance at Pins 2 and 3 is balanced.
Figure 3-40 shows the AD8225’s CMR vs. frequency
while Figure 3-41 shows its gain nonlinearity.
120
110
CMR (dB)
100
90
80
70
60
50
40
30
1
10
100
1k
FREQUENCY (Hz)
10k
100k
Figure 3-40. AD8225 CMR vs. frequency.
4
100mV
2V
3
100
2 90
NONLINEARITY (ppm)
The AD8225 is a precision, gain-of-5, monolithic
in-amp. Figure 3-39 shows that it is a 3-op amp instrumentation amplifier. The unity-gain input buffers
consist of superbeta NPN transistors Q1 and Q2 and
op amps A1 and A2. These transistors are compensated
so that their input bias currents are extremely low,
typically 100 pA or less. As a result, current noise is
also low, only 50 fA/√Hz. The input buffers drive a
gain-of-5 difference amplifier. Because the 3 k and
15 k resistors are ratio matched, gain stability is better
than 5 ppm/ C over the rated temperature range.
1
0
–1
–2 10
0
–3
–4
–10
0
OUTPUT VOLTAGE (V)
Figure 3-41. AD8225 gain nonlinearity.
3-16
10
Monolithic In-Amps Optimized for
Single-Supply Operation
Single-supply in-amps have special design problems
that need to be addressed. The input stage must be able
to amplify signals that are at ground potential (or very
close to ground), and the output stage needs to be able
to swing to within a few millivolts of ground or
the supply rail. Low power supply current is also
important. And, when operating from low power
supply voltages, the in-amp needs to have an adequate
gain bandwidth product, low offset voltage drift, and
good CMR vs. gain and frequency.
The AD623 is an instrumentation amplifier based
on the 3-op amp in-amp circuit, modified to ensure
operation on either single- or dual-power supplies, even
at common-mode voltages at, or even below, the negative
supply rail (or below ground in single-supply operation).
Other features include rail-to-rail output voltage swing,
low supply current, MSOP packaging, low input and
output voltage offset, microvolt/dc offset level drift,
high common-mode rejection, and only one external
resistor to set the gain.
As shown in Figure 3-42, the input signal is applied to
PNP transistors acting as voltage buffers and dc level
shifters. A resistor trimmed to within 0.1% of 50 k
in each amplifier’s (A1 and A2) feedback path ensures
accurate gain programmability.
The output voltage at Pin 6 is measured with respect
to the reference potential at Pin 5. The impedance of the
reference pin is 100 k. Internal ESD clamping diodes
allow the input, reference, output, and gain terminals
of the AD623 to safely withstand overvoltages of 0.3 V
above or below the supplies. This is true for all gains,
and with power on or off. This last case is particularly
important, since the signal source and the in-amp may
be powered separately. If the overvoltage is expected to
exceed this value, the current through these diodes should
be limited to 10 mA, using external current limiting
resistors (see Input Protection Basics for ADI In-Amps
section in Chapter 5). The value of these resistors is
defined by the in-amp’s noise level, the supply voltage,
and the required overvoltage protection needed.
The bandwidth of the AD623 is reduced as the gain
is increased since A1 and A2 are voltage feedback op
amps. However, even at higher gains, the AD623 still
has enough bandwidth for many applications.
+VS
7
1.5�A
+
–A1
Q1
+IN
3
50k�
4
–VS
The differential output is
+VS
7

100 kΩ 
 VO = 1 + R  VC


G
1.5�A
–IN
where RG is in k.
2
The differential voltage is then converted to a single-ended
voltage using the output difference amplifier, which also
rejects any common-mode signal at the output of the
input amplifiers.
Since all the amplifiers can swing to either supply rail,
as well as have their common-mode range extended to
below the negative supply rail, the range over which the
AD623 can operate is further enhanced.
Note that the base currents of Q1 and Q2 flow directly
out of the input terminals, unlike dual-supply, inputcurrent-compensated in-amps such as the AD620.
Since the inputs (i.e., the bases of Q1 and Q2) can
operate at ground (i.e., 0 V or, more correctly, 200 mV
below ground), it is not possible to provide input
current compensation for the AD623. However, the
input bias current of the AD623 is still very small: only
25 nA max.
50k�
1
GAIN
RESISTOR
8
50k�
50k�
50k�
–
A3
+
50k�
OUTPUT
6
REF
5
–
A2
+
Q2
4
–VS
Figure 3-42. AD623 simplified schematic.
The AD623’s gain is resistor-programmed by R G
or more precisely by whatever impedance appears
between Pins 1 and 8. Figure 3-43 shows the gain vs.
frequency of the AD623. The AD623 is laser-trimmed
to achieve accurate gains using 0.1% to 1% tolerance
resistors.
3-17
70
120
VREF = 2.5V
60
110
50
40
x10
90
30
CMR (dB)
GAIN (dB)
x1000
100
20
10
x1
80
x100
70
60
0
–10
50
–20
40
–30
100
1k
10k
FREQUENCY (Hz)
100k
30
1M
VREF = 2.5V
1
10
100
1k
FREQUENCY (Hz)
10k
100k
Figure 3-44. AD623 CMR vs. frequency (VS = 5 V).
Figure 3-43. AD623 closed-loop gain
vs. frequency.
Figure 3-45 shows the gain nonlinearity of the AD623.
Table 3-3. Required Value of Gain Resistor
Desired Gain 2
5
10 20 33 40 50 65 100 200 500 1000 1% Std.
Value of RG () 100 k 24.9 k 11 k 5.23 k 3.09 k 2.55 k 2.05 k 1.58 k 1.02 k 499 200 100 Calculated Gain
Using 1% Resistors
2
5.02
10.09
20.12
33.36
40.21
49.78
64.29
99.04
201.4
501
1001
Figure 3-45. AD623 gain nonlinearity
(G = –10, 50 ppm/div).
Table 3-3 shows required values of RG for various gains.
Note that for G = 1, the RG terminals are unconnected
(RG = ). For any arbitrary gain, RG can be calculated
using the formula
Figure 3-46 shows the small signal pulse response of
the AD623.
RG = 100 k/(G – 1)
Figure 3-44 shows the AD623’s CMR vs. frequency.
Note that the CMR increases with gain up to a gain of
100 and that CMR also remains high over frequency, up
to 200 Hz. This ensures the attenuation of power line
common-mode signals (and their harmonics).
Figure 3-46. AD623 small signal pulse response
(G = 10, RL = 10 k, CL = 100 pF).
3-18
The AD627 is a single-supply, micropower instrumentation amplifier that can be configured for gains between
5 and 1000 using just a single external resistor. It provides
a rail-to-rail output voltage swing using a single 3 V to
30 V power supply. With a quiescent supply current of
only 60 A (typical), its total power consumption is less
than 180 W, operating from a 3 V supply.
Figure 3-47 is a simplified schematic of the AD627.
The AD627 is a true instrumentation amplifier built
using two feedback loops. Its general properties are
similar to those of the classic 2-op amp instrumentation
amplifier configuration and can be regarded as such,
but internally the details are somewhat different. The
AD627 uses a modified current feedback scheme,
which, coupled with interstage feedforward frequency
compensation, results in a much better CMRR at
frequencies above dc (notably the line frequency of
50 Hz to 60 Hz) than might otherwise be expected of
a low power instrumentation amplifier.
As shown in Figure 3-47, A1 completes a feedback loop,
which, in conjunction with V1 and R5, forces a constant
collector current in Q1. Assume for the moment that
the gain-setting resistor (RG) is not present. Resistors
R2 and R1 complete the loop and force the output of
A1 to be equal to the voltage on the inverting terminal
with a gain of (almost exactly) 1.25. A nearly identical
feedback loop completed by A2 forces a current in Q2,
which is substantially identical to that in Q1, and A2 also
provides the output voltage. When both loops are balanced, the gain from the noninverting terminal to VOUT
is equal to 5, whereas the gain from the output of A1 to
VOUT is equal to –4. The inverting terminal gain of A1
(1.25), times the gain of A2 (–4), makes the gain from
the inverting and noninverting terminals equal.
The differential mode gain is equal to 1 + R4/R3,
nominally 5, and is factory trimmed to 0.01% final
accuracy (AD627B typ). Adding an external gain setting
resistor (RG) increases the gain by an amount equal to
(R4 + R1)/RG. The gain of the AD627 is given by the
following equation:
 200 kΩ 
G = 5+

 RG 
Laser trims are performed on resistors R1 through R4
to ensure that their values are as close as possible to
the absolute values in the gain equation. This ensures
low gain error and high common-mode rejection at all
practical gains.
Figure 3-48 shows the AD627’s CMR vs. frequency.
120
110
100
90
G = 1000
80
CMR (dB)
Low Power, Single-Supply In-Amps
70
G = 100
60
50
G=5
40
30
20
10
0
1
10
100
1k
10k
FREQUENCY (Hz)
Figure 3-48. AD627 CMR vs. frequency.
EXTERNAL GAIN RESISTOR
REF
R1
100k�
R2
25k�
+VS
–IN
2k�
R4
100k�
RG
R3
25k�
2k�
Q2
Q1
–VS
+VS
+IN
–VS
A1
A2
R5
200k�
OUTPUT
R6
200k�
V1
–VS
Figure 3-47. AD627 simplified schematic.
3-19
100k
Gain-Programmable In-Amps
Figures 3-49 and 3-50 show the AD627’s gain vs.
frequency and gain nonlinearity.
The AD8250 and AD8251 (Figure 3-52) are digitally gainprogrammable instrumentation amplifiers that have high
(G) input impedances and low distortion, making them
suitable for sensor interfacing and driving high sample rate
analog-to-digital converters. The two products are nearly
identical, except for their gain ranges. The AD8250 has
programmable gains of 1, 2, 5, and 10, while the AD8251
has a range of 1, 2, 4, and 8 (for binary applications). Both
products have high bandwidths of 10 MHz, low distortion,
and a settling time of 0.5 s to 0.01%. Input offset drift and
gain drift are only 1 V/C and 10 ppm/C, respectively. In
addition to their wide input common-voltage range, they
boast a high common-mode rejection of 80 dB at
G = 1 from dc to 100 kHz. The combination of precision
dc performance coupled with high speed capabilities makes
the AD8250 and AD8251 excellent candidates for data
acquisition and medical applications. Furthermore, these
monolithic solutions simplify design and manufacturing,
while boosting their performance, by maintaining a tight
match of internal resistors and amplifiers.
70
60
G = 1000
CLOSED-LOOP GAIN (dB)
50
40
G = 100
30
20
10
G = 10
G=5
0
–10
–20
–30
100
1k
10k
FREQUENCY (Hz)
100k
Figure 3-49. AD627 closed-loop gain
vs. frequency.
M6$)6)3)/.
+VS
–IN
A1
PPM6%24
$)6)3)/.
DGND
WR
A1
A2
6/54
6$)6)3)/.
GAIN
LOGIC
A3
OUT
A2
+IN
Figure 3-50. AD627 gain nonlinearity
(VS = 2.5 V, G = 5, 4 ppm/vertical division).
–VS
The AD627 also has excellent dynamic response, as
shown in Figure 3-51.
REF
Figure 3-52. AD8250 and AD8251 simplified
schematic.
The AD8250 and AD8251 user interfaces are comprised
of a parallel port that allows users to set the gain in one
of three different ways (Figure 3-52). A 2-bit word sent to
A1 and A2 via a bus may be latched using the CLK input.
An alternative is to set the gain within 1 s by using the
gain port in transparent mode. The last method is to strap
A1 and A2 to a high or low voltage potential, permanently
setting the gain.
Figure 3-51. Small signal pulse response of the
AD627 (VS = 5 V, G = +10, RL = 20 k, CL = 50 pF,
20 s/horizontal division, 20 mV/vertical division).
The AD8250 and AD8251 are available in a 10-lead
MSOP package and are specified over the –40C to
+125C temperature range, making them an excellent
solution for applications where size and packing density
are important considerations. To simplify matters, their
pinout was chosen to optimize layout and increase ac
performance.
3-20