You are on page 1of 117

EE101: Op Amp circuits (Part 2)

M. B. Patil
mbpatil@ee.iitb.ac.in
www.ee.iitb.ac.in/~sequel
Department of Electrical Engineering
Indian Institute of Technology Bombay

M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

Consider a bridge circuit for sensing temperature, pressure, etc., with Ra = Rb = Rc = R .


Rd = R + R varies with the quantity to be measured. Typically, R is a small fraction of R.

M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

Consider a bridge circuit for sensing temperature, pressure, etc., with Ra = Rb = Rc = R .


Rd = R + R varies with the quantity to be measured. Typically, R is a small fraction of R.
The bridge converts R to a signal voltage which can then be suitably amplified and used for
display or control.

M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

Consider a bridge circuit for sensing temperature, pressure, etc., with Ra = Rb = Rc = R .


Rd = R + R varies with the quantity to be measured. Typically, R is a small fraction of R.
The bridge converts R to a signal voltage which can then be suitably amplified and used for
display or control.
Assuming that the amplifier has a large input resistance,

M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

Consider a bridge circuit for sensing temperature, pressure, etc., with Ra = Rb = Rc = R .


Rd = R + R varies with the quantity to be measured. Typically, R is a small fraction of R.
The bridge converts R to a signal voltage which can then be suitably amplified and used for
display or control.
Assuming that the amplifier has a large input resistance,
v1 =

R
1
VCC = VCC .
R +R
2

M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

Consider a bridge circuit for sensing temperature, pressure, etc., with Ra = Rb = Rc = R .


Rd = R + R varies with the quantity to be measured. Typically, R is a small fraction of R.
The bridge converts R to a signal voltage which can then be suitably amplified and used for
display or control.
Assuming that the amplifier has a large input resistance,
v1 =

R
1
VCC = VCC .
R +R
2

v2 =

(R + R)
1 1+x
1
1
VCC =
VCC (1 + x) (1 x/2) VCC = (1 + x/2) VCC ,
R + (R + R)
2 1 + x/2
2
2

where x = R/R .

M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

Consider a bridge circuit for sensing temperature, pressure, etc., with Ra = Rb = Rc = R .


Rd = R + R varies with the quantity to be measured. Typically, R is a small fraction of R.
The bridge converts R to a signal voltage which can then be suitably amplified and used for
display or control.
Assuming that the amplifier has a large input resistance,
v1 =

R
1
VCC = VCC .
R +R
2

v2 =

(R + R)
1 1+x
1
1
VCC =
VCC (1 + x) (1 x/2) VCC = (1 + x/2) VCC ,
R + (R + R)
2 1 + x/2
2
2

where x = R/R .
For example, with VCC = 15 V , R = 1 k, R = 0.01 k ,
v1 = 7.5 V ,
v2 = 7.5 + 0.0375 V .
M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

v1 = 7.5 V , v2 = 7.5 + 0.0375 V .

M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

v1 = 7.5 V , v2 = 7.5 + 0.0375 V .


The amplifier should only amplify v2 v1 = 0.0375 V (since that is the signal arising from R).

M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

v1 = 7.5 V , v2 = 7.5 + 0.0375 V .


The amplifier should only amplify v2 v1 = 0.0375 V (since that is the signal arising from R).
Definitions:
Given v1 and v2 ,
1
vc = (v1 + v2 ) = common-mode voltage,
2
vd = (v2 v1 ) = differential-mode voltage.

M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

v1 = 7.5 V , v2 = 7.5 + 0.0375 V .


The amplifier should only amplify v2 v1 = 0.0375 V (since that is the signal arising from R).
Definitions:
Given v1 and v2 ,
1
vc = (v1 + v2 ) = common-mode voltage,
2
vd = (v2 v1 ) = differential-mode voltage.
v1 and v2 can be rewritten as,
v1 = vc vd /2 , v2 = vc + vd /2 .

M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

v1 = 7.5 V , v2 = 7.5 + 0.0375 V .


The amplifier should only amplify v2 v1 = 0.0375 V (since that is the signal arising from R).
Definitions:
Given v1 and v2 ,
1
vc = (v1 + v2 ) = common-mode voltage,
2
vd = (v2 v1 ) = differential-mode voltage.
v1 and v2 can be rewritten as,
v1 = vc vd /2 , v2 = vc + vd /2 .
In the above example, vc 7.5 V , vd = 37.5 mV .

M. B. Patil, IIT Bombay

Common-mode and differential-mode voltages


VCC
Ra

Rc

v1

Amplifier

v2
Rb

Vo

Rd

v1 = 7.5 V , v2 = 7.5 + 0.0375 V .


The amplifier should only amplify v2 v1 = 0.0375 V (since that is the signal arising from R).
Definitions:
Given v1 and v2 ,
1
vc = (v1 + v2 ) = common-mode voltage,
2
vd = (v2 v1 ) = differential-mode voltage.
v1 and v2 can be rewritten as,
v1 = vc vd /2 , v2 = vc + vd /2 .
In the above example, vc 7.5 V , vd = 37.5 mV .
Note that the common-mode voltage is quite large compared to the differential-mode voltage.
This is a common situation in transducer circuits.
M. B. Patil, IIT Bombay

Common-Mode Rejection Ratio

v+
Amplifier

vo

v+ = vc + vd /2
v = vc vd /2

An ideal amplifier would only amplify the difference (v+ v ) , giving


vo = Ad (v+ v ) = Ad vd ,
where Ad is called the differential gain or simply the gain (AV ).

M. B. Patil, IIT Bombay

Common-Mode Rejection Ratio

v+
Amplifier

vo

v+ = vc + vd /2
v = vc vd /2

An ideal amplifier would only amplify the difference (v+ v ) , giving


vo = Ad (v+ v ) = Ad vd ,
where Ad is called the differential gain or simply the gain (AV ).
In practice, the output can also have a common-mode component:
vo = Ad vd + Ac vc ,
where Ac is called the common-mode gain.

M. B. Patil, IIT Bombay

Common-Mode Rejection Ratio

v+
Amplifier

vo

v+ = vc + vd /2
v = vc vd /2

An ideal amplifier would only amplify the difference (v+ v ) , giving


vo = Ad (v+ v ) = Ad vd ,
where Ad is called the differential gain or simply the gain (AV ).
In practice, the output can also have a common-mode component:
vo = Ad vd + Ac vc ,
where Ac is called the common-mode gain.
The ability of an amplifier to reject the common-mode signal is given by the
Common-Mode Rejection Ratio (CMRR):
CMRR =

Ad
.
Ac

M. B. Patil, IIT Bombay

Common-Mode Rejection Ratio

v+
Amplifier

vo

v+ = vc + vd /2
v = vc vd /2

An ideal amplifier would only amplify the difference (v+ v ) , giving


vo = Ad (v+ v ) = Ad vd ,
where Ad is called the differential gain or simply the gain (AV ).
In practice, the output can also have a common-mode component:
vo = Ad vd + Ac vc ,
where Ac is called the common-mode gain.
The ability of an amplifier to reject the common-mode signal is given by the
Common-Mode Rejection Ratio (CMRR):
Ad
.
Ac
For the 741 Op Amp, the CMRR is 90 dB (' 30,000), which may be considered to be infinite in
CMRR =

many applications. In such cases, mismatch between circuit components will determine the overall
common-mode rejection performance of the circuit.

M. B. Patil, IIT Bombay

Op Amp circuits (linear region)

R2

i1
Vi1
Vi2

R1

i
i+

Vo

R3

RL
R4

M. B. Patil, IIT Bombay

Op Amp circuits (linear region)

R2

i1
Vi1
Vi2

R1

i
i+

Vo

R3

RL
R4

Method 1:
Large input resistance of Op Amp i+ = 0, V+ =

R4
Vi2 .
R3 + R4

M. B. Patil, IIT Bombay

Op Amp circuits (linear region)

R2

i1
Vi1
Vi2

R1

i
i+

Vo

R3

RL
R4

Method 1:
Large input resistance of Op Amp i+ = 0, V+ =
Since V+ V 0, i1 =

R4
Vi2 .
R3 + R4

1
1
(Vi1 V )
(Vi1 V+ ) .
R1
R1

M. B. Patil, IIT Bombay

Op Amp circuits (linear region)

R2

i1
Vi1
Vi2

R1

i
i+

Vo

R3

RL
R4

Method 1:
Large input resistance of Op Amp i+ = 0, V+ =

R4
Vi2 .
R3 + R4

1
1
(Vi1 V )
(Vi1 V+ ) .
R1
R1
R2
i1 R2 V+
(Vi1 V+ ) .
R1

Since V+ V 0, i1 =
i 0 V o = V

M. B. Patil, IIT Bombay

Op Amp circuits (linear region)

R2

i1
Vi1
Vi2

R1

i
i+

Vo

R3

RL
R4

Method 1:
Large input resistance of Op Amp i+ = 0, V+ =

R4
Vi2 .
R3 + R4

1
1
(Vi1 V )
(Vi1 V+ ) .
R1
R1
R2
i1 R2 V+
(Vi1 V+ ) .
R1

Since V+ V 0, i1 =
i 0 V o = V

Substituting for V+ and selecting R3 /R4 = R1 /R2 , we get (show this),


Vo =

R2
(Vi2 Vi1 ) .
R1

M. B. Patil, IIT Bombay

Op Amp circuits (linear region)

R2

i1
Vi1
Vi2

R1

i
i+

Vo

R3

RL
R4

Method 1:
Large input resistance of Op Amp i+ = 0, V+ =

R4
Vi2 .
R3 + R4

1
1
(Vi1 V )
(Vi1 V+ ) .
R1
R1
R2
i1 R2 V+
(Vi1 V+ ) .
R1

Since V+ V 0, i1 =
i 0 V o = V

Substituting for V+ and selecting R3 /R4 = R1 /R2 , we get (show this),


R2
(Vi2 Vi1 ) .
R1
The circuit is a difference amplifier.
Vo =

M. B. Patil, IIT Bombay

Difference amplifier

R2
Vi1
Vi2

R2
Vi1

R1

R1

Vo
R3

RL

R2

Vo1
R3

R4

RL

R1

AND
Vi2

R4

Case 1

Vo2
R3

RL
R4

Case 2

Method 2:
Since the Op Amp is operating in the linear region, we can use superposition:

M. B. Patil, IIT Bombay

Difference amplifier

R2
Vi1
Vi2

R2
Vi1

R1

R1

Vo
R3

RL

R2

Vo1
R3

R4

RL

R1

AND
Vi2

R4

Case 1

Vo2
R3

RL
R4

Case 2

Method 2:
Since the Op Amp is operating in the linear region, we can use superposition:
Case 1:

Inverting amplifier (note that V+ = 0 V ).


R2
Vi1 .
Vo1 =
R1

M. B. Patil, IIT Bombay

Difference amplifier

R2
Vi1
Vi2

R2
Vi1

R1

R2

R1

Vo
R3

RL

Vo1
R3

R4

RL

R1

AND
Vi2

R4

Case 1

Vo2
R3

RL
R4

Case 2

Method 2:
Since the Op Amp is operating in the linear region, we can use superposition:
Case 1:

Case 2:

Inverting amplifier (note that V+ = 0 V ).


R2
Vi1 .
Vo1 =
R1
R4
Non-inverting amplifier, with Vi =
Vi2 .

R3 + R4
R2
R4
Vo2 = 1 +
Vi2 .
R1
R3 + R4

M. B. Patil, IIT Bombay

Difference amplifier

R2
Vi1
Vi2

R2
Vi1

R1

R2

R1

Vo
R3

Vo1
R3

RL
R4

RL

R1

AND
Vi2

R4

Case 1

Vo2
R3

RL
R4

Case 2

Method 2:
Since the Op Amp is operating in the linear region, we can use superposition:
Case 1:

Case 2:

Inverting amplifier (note that V+ = 0 V ).


R2
Vi1 .
Vo1 =
R1
R4
Non-inverting amplifier, with Vi =
Vi2 .

R3 + R4
R2
R4
Vo2 = 1 +
Vi2 .
R1
R3 + R4

The net result is,


Vo = Vo1 + Vo2 =

1+

R2
R1

R4
R3 + R4

Vi2

R2
R2
Vi1 =
(Vi2 Vi1 ) , if R3 /R4 = R1 /R2 .
R1
R1
M. B. Patil, IIT Bombay

Difference amplifier

1k

1k

Ra

Rc

v1

10 k

R2
1 k R1

v2

Vo

Rb
1k

VCC

Rd
1k

1 k R3

RL
R4
10 k

Bridge

Difference amplifier

M. B. Patil, IIT Bombay

Difference amplifier

1k

1k

Ra

Rc

v1

10 k

R2
1 k R1

v2

Vo

Rb
1k

VCC

Rd
1k

1 k R3

RL
R4
10 k

Bridge

Difference amplifier

The resistance seen from v2 is (R3 + R4 ) which is small enough to cause v2 to change.
This is not desirable.

M. B. Patil, IIT Bombay

Difference amplifier

1k

1k

Ra

Rc

v1

10 k

R2
1 k R1

v2

Vo

Rb
1k

VCC

Rd
1k

1 k R3

RL
R4
10 k

Bridge

Difference amplifier

The resistance seen from v2 is (R3 + R4 ) which is small enough to cause v2 to change.
This is not desirable.
need to improve the input resistance of the difference amplifier.

M. B. Patil, IIT Bombay

Difference amplifier

1k

1k

Ra

Rc

v1

10 k

R2
1 k R1

v2

Vo

Rb
1k

VCC

Rd
1k

1 k R3

RL
R4
10 k

Bridge

Difference amplifier

The resistance seen from v2 is (R3 + R4 ) which is small enough to cause v2 to change.
This is not desirable.
need to improve the input resistance of the difference amplifier.
We will discuss an improved difference amplifier later. Before we do that, let us
discuss another problem with the above difference amplifier which can be important
for some applications (next slide).

M. B. Patil, IIT Bombay

Difference amplifier
R2
vi1
vi2

R1

vo

vi1 = vc vd /2
vi2 = vc + vd /2

RL

R3 = R1
R4 = R2

M. B. Patil, IIT Bombay

Difference amplifier
R2
vi1
vi2

R1

vo

vi1 = vc vd /2
vi2 = vc + vd /2

RL

R3 = R1
R4 = R2

Consider the difference amplifier with R3 = R1 , R4 = R2 Vo =

R2
(vi2 vi1 ) .
R1

The output voltage depends only on the differential-mode signal (vi2 vi1 ),
i.e., Ac (common-mode gain) = 0.

M. B. Patil, IIT Bombay

Difference amplifier
R2
vi1
vi2

R1

vo

vi1 = vc vd /2
vi2 = vc + vd /2

RL

R3 = R1
R4 = R2

Consider the difference amplifier with R3 = R1 , R4 = R2 Vo =

R2
(vi2 vi1 ) .
R1

The output voltage depends only on the differential-mode signal (vi2 vi1 ),
i.e., Ac (common-mode gain) = 0.
In practice, R3 and R1 may not be exactly equal. Let R3 = R1 + R .

M. B. Patil, IIT Bombay

Difference amplifier
R2
vi1
vi2

R1

vo

vi1 = vc vd /2
vi2 = vc + vd /2

RL

R3 = R1
R4 = R2

Consider the difference amplifier with R3 = R1 , R4 = R2 Vo =

R2
(vi2 vi1 ) .
R1

The output voltage depends only on the differential-mode signal (vi2 vi1 ),
i.e., Ac (common-mode gain) = 0.
In practice, R3 and R1 may not be exactly equal. Let R3 = R1 + R .

R2
R2
R2
vo =
1+
vi2
vi1
R1 + R + R2
R1
R1
R2
R
'
(vd x vc ) , with x =
(show this)
R1
R1 + R2

M. B. Patil, IIT Bombay

Difference amplifier
R2
vi1
vi2

R1

vo

vi1 = vc vd /2
vi2 = vc + vd /2

RL

R3 = R1
R4 = R2

Consider the difference amplifier with R3 = R1 , R4 = R2 Vo =

R2
(vi2 vi1 ) .
R1

The output voltage depends only on the differential-mode signal (vi2 vi1 ),
i.e., Ac (common-mode gain) = 0.
In practice, R3 and R1 may not be exactly equal. Let R3 = R1 + R .

R2
R2
R2
vo =
1+
vi2
vi1
R1 + R + R2
R1
R1
R2
R
'
(vd x vc ) , with x =
(show this)
R1
R1 + R2
R2
R2
|Ac | = x
 |Ad | =
, since x 0.01 (with 1% tolerance resistors).
R1
R1

M. B. Patil, IIT Bombay

Difference amplifier
R2
vi1
vi2

R1

vo

vi1 = vc vd /2
vi2 = vc + vd /2

RL

R3 = R1
R4 = R2

Consider the difference amplifier with R3 = R1 , R4 = R2 Vo =

R2
(vi2 vi1 ) .
R1

The output voltage depends only on the differential-mode signal (vi2 vi1 ),
i.e., Ac (common-mode gain) = 0.
In practice, R3 and R1 may not be exactly equal. Let R3 = R1 + R .

R2
R2
R2
vo =
1+
vi2
vi1
R1 + R + R2
R1
R1
R2
R
'
(vd x vc ) , with x =
(show this)
R1
R1 + R2
R2
R2
|Ac | = x
 |Ad | =
, since x 0.01 (with 1% tolerance resistors).
R1
R1
However, since vc can be large compared to vd , the effect of Ac cannot be ignored.
M. B. Patil, IIT Bombay

Difference amplifier

R2
vi1
vi2

R1

vo

vi1 = vc vd /2
vi2 = vc + vd /2

RL

R3 = R1
R4 = R2

|Ac | = x

R2
R2
R
, |Ad | =
, where x =
.
R1
R1
R1 + R2

M. B. Patil, IIT Bombay

Difference amplifier

R2
vi1
vi2

R1

vo

vi1 = vc vd /2
vi2 = vc + vd /2

RL

R3 = R1
R4 = R2

R2
R2
R
, |Ad | =
, where x =
.
R1
R1
R1 + R2
In our earlier example, vc = 7.5 V vd = 0.0375 V .
|Ac | = x

M. B. Patil, IIT Bombay

Difference amplifier

R2
vi1
vi2

R1

vo

vi1 = vc vd /2
vi2 = vc + vd /2

RL

R3 = R1
R4 = R2

R2
R2
R
, |Ad | =
, where x =
.
R1
R1
R1 + R2
In our earlier example, vc = 7.5 V vd = 0.0375 V .
|Ac | = x

With R1 = 1 k, R2 = 10 k, x = 0.01 ,
|Ac vc | = x

R2
vc = 0.01 10 7.5 = 0.75 V .
R1

M. B. Patil, IIT Bombay

Difference amplifier

R2
vi1
vi2

R1

vo

vi1 = vc vd /2
vi2 = vc + vd /2

RL

R3 = R1
R4 = R2

R2
R2
R
, |Ad | =
, where x =
.
R1
R1
R1 + R2
In our earlier example, vc = 7.5 V vd = 0.0375 V .
|Ac | = x

With R1 = 1 k, R2 = 10 k, x = 0.01 ,
R2
vc = 0.01 10 7.5 = 0.75 V .
R1
R2
|Ad vd | =
vd = 10 0.0375 = 0.375 V .
R1
|Ac vc | = x

M. B. Patil, IIT Bombay

Difference amplifier

R2
vi1
vi2

R1

vo

vi1 = vc vd /2
vi2 = vc + vd /2

RL

R3 = R1
R4 = R2

R2
R2
R
, |Ad | =
, where x =
.
R1
R1
R1 + R2
In our earlier example, vc = 7.5 V vd = 0.0375 V .
|Ac | = x

With R1 = 1 k, R2 = 10 k, x = 0.01 ,
R2
vc = 0.01 10 7.5 = 0.75 V .
R1
R2
|Ad vd | =
vd = 10 0.0375 = 0.375 V .
R1
|Ac vc | = x

The (spurious) common-mode contribution is substantial.


need a circuit which will reduce the common-mode component at the output.

M. B. Patil, IIT Bombay

Improved difference amplifier

Vi1

A1

i1

R2

R4
R3
A3

R1

Vo
B

Vi2

Vo1

A2

RL

R3

R2

R4
Vo2

Improved difference amplifier

Vi1

A1

i1

R2

R4
R3
A3

R1

Vo
B

Vi2

Vo1

A2

RL

R3

R2

R4
Vo2

V+ V VA = Vi1 , VB = Vi2 , i1 =

1
(Vi1 Vi2 ) .
R1

Improved difference amplifier

Vi1

A1

i1

R2

R4
R3
A3

R1

Vo
B

Vi2

Vo1

A2

RL

R3

R2

R4
Vo2

1
(Vi1 Vi2 ) .
R1
Large input resistance of A1 and A2 the current through the two resistors marked R2 is also
equal to i1 .
V+ V VA = Vi1 , VB = Vi2 , i1 =

Improved difference amplifier

Vi1

A1

i1

R2

R4
R3
A3

R1

Vo
B

Vi2

Vo1

A2

RL

R3

R2

R4
Vo2

1
(Vi1 Vi2 ) .
R1
Large input resistance of A1 and A2 the current through the two resistors marked R2 is also
equal to i1 .
V+ V VA = Vi1 , VB = Vi2 , i1 =

Improved difference amplifier

Vi1

A1

i1

R2

R4
R3
A3

R1

Vo
B

Vi2

Vo1

A2

RL

R3

R2

R4
Vo2

1
(Vi1 Vi2 ) .
R1
Large input resistance of A1 and A2 the current through the two resistors marked R2 is also
equal to i1 .

1
2 R2
Vo1 Vo2 = i1 (R1 + 2 R2 ) =
(Vi1 Vi2 ) (R1 + 2 R2 ) = (Vi1 Vi2 ) 1 +
.
R1
R1
V+ V VA = Vi1 , VB = Vi2 , i1 =

Improved difference amplifier

Vi1

A1

i1

R2

R4
R3
A3

R1

Vo
B

Vi2

Vo1

A2

RL

R3

R2

R4
Vo2

1
(Vi1 Vi2 ) .
R1
Large input resistance of A1 and A2 the current through the two resistors marked R2 is also
equal to i1 .

1
2 R2
Vo1 Vo2 = i1 (R1 + 2 R2 ) =
(Vi1 Vi2 ) (R1 + 2 R2 ) = (Vi1 Vi2 ) 1 +
.
R1
R1

R4
R4
2 R2
Finally, Vo =
(Vo2 Vo1 ) =
1+
(Vi2 Vi1 ) .
R3
R3
R1
V+ V VA = Vi1 , VB = Vi2 , i1 =

Improved difference amplifier

Vi1

A1

i1

R2

R4
R3
A3

R1

Vo
B

Vi2

Vo1

A2

RL

R3

R2

R4
Vo2

1
(Vi1 Vi2 ) .
R1
Large input resistance of A1 and A2 the current through the two resistors marked R2 is also
equal to i1 .

1
2 R2
Vo1 Vo2 = i1 (R1 + 2 R2 ) =
(Vi1 Vi2 ) (R1 + 2 R2 ) = (Vi1 Vi2 ) 1 +
.
R1
R1

R4
R4
2 R2
Finally, Vo =
(Vo2 Vo1 ) =
1+
(Vi2 Vi1 ) .
R3
R3
R1
V+ V VA = Vi1 , VB = Vi2 , i1 =

This circuit is known as the instrumentation amplifier.

M. B. Patil, IIT Bombay

Instrumentation amplifier

Vi1

A1

Vo1
R4

R1

i1

R2

A3

R3

Vo

R2
R3

RL
R4

Vi2

A2

Vo2

Instrumentation amplifier

Vi1

A1

Vo1
R4

R1

i1

R2

A3

R3

Vo

R2
R3

RL
R4

Vi2

A2

Vo2

The input resistance seen from Vi1 or Vi2 is large (since an Op Amp has a large input
resistance).

Instrumentation amplifier

Vi1

A1

Vo1
R4

R1

i1

R2

A3

R3

Vo

R2
R3

RL
R4

Vi2

A2

Vo2

The input resistance seen from Vi1 or Vi2 is large (since an Op Amp has a large input
resistance).
the amplifier will not load the preceding stage, a desirable feature.

Instrumentation amplifier

Vi1

A1

VCC
Ra

R1

i1

v2
Rb

R4

Rc

v1

Vo1

R2

Vo

R2
R3

Rd

A3

R3

RL
R4

Vi2

A2

Vo2

The input resistance seen from Vi1 or Vi2 is large (since an Op Amp has a large input
resistance).
the amplifier will not load the preceding stage, a desirable feature.

Instrumentation amplifier

Vi1

A1

VCC
Ra

R1

i1

v2
Rb

R4

Rc

v1

Vo1

R2

Vo

R2
R3

Rd

A3

R3

RL
R4

Vi2

A2

Vo2

The input resistance seen from Vi1 or Vi2 is large (since an Op Amp has a large input
resistance).
the amplifier will not load the preceding stage, a desirable feature.
As a result, the voltages v1 and v2 in the bridge circuit will remain essentially the same
when the bridge circuit is connected to the instrumentation amplifier.

M. B. Patil, IIT Bombay

Instrumentation amplifier
vi1

A1 v
o1

R2

R4
R3

R1

A3

i1
B

vi2

vi1 = vc vd /2
vo

R3

R2
A2

R4

vi2 = vc + vd /2

RL

vo2

As we have seen earlier, vi1 and vi2 can have a large common-mode component (vc ).
What is the effect of vc on the amplifier output vo ?

M. B. Patil, IIT Bombay

Instrumentation amplifier
vi1

A1 v
o1

R2

R4
R3

R1

A3

i1
B

vi2

vi1 = vc vd /2
vo

R3

R2
A2

R4

vi2 = vc + vd /2

RL

vo2

As we have seen earlier, vi1 and vi2 can have a large common-mode component (vc ).
What is the effect of vc on the amplifier output vo ?
v+ v vA = vc vd /2 , vB = vc + vd /2 .

M. B. Patil, IIT Bombay

Instrumentation amplifier
vi1

A1 v
o1

R2

R4
R3

R1

A3

i1
B

vi2

vi1 = vc vd /2
vo

R3

R2
A2

R4

vi2 = vc + vd /2

RL

vo2

As we have seen earlier, vi1 and vi2 can have a large common-mode component (vc ).
What is the effect of vc on the amplifier output vo ?
v+ v vA = vc vd /2 , vB = vc + vd /2 .
i1 =

1
1
1
(vA vB ) =
((vc vd /2) (vc + vd /2)) = vd .
R1
R1
R1

M. B. Patil, IIT Bombay

Instrumentation amplifier
vi1

A1 v
o1

R2

R4
R3

R1

A3

i1
B

vi2

vi1 = vc vd /2
vo

R3

R2
A2

R4

vi2 = vc + vd /2

RL

vo2

As we have seen earlier, vi1 and vi2 can have a large common-mode component (vc ).
What is the effect of vc on the amplifier output vo ?
v+ v vA = vc vd /2 , vB = vc + vd /2 .
i1 =

1
1
1
(vA vB ) =
((vc vd /2) (vc + vd /2)) = vd .
R1
R1
R1

vc has simply got cancelled! (And this holds even if R2 and R20 are not exactly matched.)

M. B. Patil, IIT Bombay

Instrumentation amplifier
vi1

A1 v
o1

R2

R4
R3

R1

A3

i1
B

vi2

vi1 = vc vd /2
vo

R3

R2
A2

R4

vi2 = vc + vd /2

RL

vo2

As we have seen earlier, vi1 and vi2 can have a large common-mode component (vc ).
What is the effect of vc on the amplifier output vo ?
v+ v vA = vc vd /2 , vB = vc + vd /2 .
i1 =

1
1
1
(vA vB ) =
((vc vd /2) (vc + vd /2)) = vd .
R1
R1
R1

vc has simply got cancelled! (And this holds even if R2 and R20 are not exactly matched.)
The instrumentation amplifier is very effective in minimising the effect of the common-mode
signal. (Note that component mismatch in the second stage will cause a finite CMRR, but the first
stage has effectively amplified only vd while leaving vc unchanged; so the overall CMRR has
improved.)
M. B. Patil, IIT Bombay

Current-to-voltage conversion

Some circuits produce an output in the form of a current. It is convenient to convert


this current into a voltage for further processing.

Current-to-voltage conversion

Some circuits produce an output in the form of a current. It is convenient to convert


this current into a voltage for further processing.
Current-to-voltage conversion can be achieved by simply passing the current through a
resistor: Vo1 = Is R .

Is

Vo1

Current-to-voltage conversion

Some circuits produce an output in the form of a current. It is convenient to convert


this current into a voltage for further processing.
Current-to-voltage conversion can be achieved by simply passing the current through a
resistor: Vo1 = Is R .

Ro
Is

Vo1

Vi

Ri

AV Vi

Vo2

amplifier

However, this simple approach will not work if the next stage in the circuit (such as an
amplifier) has a finite Ri , since it will modify Vo1 to Vo1 = Is (Ri k R) , which is not
desirable.

M. B. Patil, IIT Bombay

Current-to-voltage conversion

i
Is

Vo
RL

Current-to-voltage conversion

i
Is

Vo
RL

V V+ , and i 0 Vo = V Is R = Is R .

Current-to-voltage conversion

i
Is

Vo
RL

V V+ , and i 0 Vo = V Is R = Is R .
The output voltage is proportional to the source current, irrespective of the value
of RL , i.e., irrespective of the next stage.

Current-to-voltage conversion

i
Is

Vo
RL

V V+ , and i 0 Vo = V Is R = Is R .
The output voltage is proportional to the source current, irrespective of the value
of RL , i.e., irrespective of the next stage.
Example: a photocurrent detector.

Current-to-voltage conversion

i
Is

Vo

Is

RL

Vo
RL

Vbias (negative)

V V+ , and i 0 Vo = V Is R = Is R .
The output voltage is proportional to the source current, irrespective of the value
of RL , i.e., irrespective of the next stage.
Example: a photocurrent detector.

M. B. Patil, IIT Bombay

Current-to-voltage conversion

i
Is

Vo

Is

RL

Vo
RL

Vbias (negative)

V V+ , and i 0 Vo = V Is R = Is R .
The output voltage is proportional to the source current, irrespective of the value
of RL , i.e., irrespective of the next stage.
Example: a photocurrent detector.
Vo = Is R . The diode is under a reverse bias, with Vn = 0 V and Vp = Vbias .

M. B. Patil, IIT Bombay

Op Amp circuits (linear region)


C

Vi

Vc

i1
R

i
Vo
RL

M. B. Patil, IIT Bombay

Op Amp circuits (linear region)


C

Vi

Vc

i1
R

i
Vo
RL

V V+ = 0 V i1 = Vi /R .

M. B. Patil, IIT Bombay

Op Amp circuits (linear region)


C

Vi

Vc

i1
R

i
Vo
RL

V V+ = 0 V i1 = Vi /R .
Since i 0 , the current through the capacitor is i1 .
C

dVc
Vi
= i1 =
.
dt
R

M. B. Patil, IIT Bombay

Op Amp circuits (linear region)


C

Vi

Vc

i1
R

i
Vo
RL

V V+ = 0 V i1 = Vi /R .
Since i 0 , the current through the capacitor is i1 .
C

dVc
Vi
= i1 =
.
dt
R

Vc = V Vo = 0 Vo = Vo C

dVo
dt

Vi
R

M. B. Patil, IIT Bombay

Op Amp circuits (linear region)


C
Vc

i1

Vi

i
Vo
RL

V V+ = 0 V i1 = Vi /R .
Since i 0 , the current through the capacitor is i1 .
C

dVc
Vi
= i1 =
.
dt
R

Vc = V Vo = 0 Vo = Vo C
Vo =

1
RC

dVo
dt

Vi
R

Z
Vi dt

The circuit works as an integrator.


M. B. Patil, IIT Bombay

Integrator

C
Vi

R = 1 k , C = 0.2 F

R
Vo
RL

Vo =

1
RC

Vi dt

Integrator

C
Vi

R = 1 k , C = 0.2 F

R
Vo
RL

Vo =

Vi
3

Vo
3
0

0.5

1
1.5
t (msec)

2.5

1
RC

Vi dt

Integrator

C
Vi

R = 1 k , C = 0.2 F

R
Vo
RL

Vo =

1
RC

Vi dt

Vi

Vi

3
0

Vo

0
3

Vo
3

6
0

0.5

1
1.5
t (msec)

2.5

0.5

1
1.5
t (msec)

2.5

Integrator

C
Vi

R = 1 k , C = 0.2 F

R
Vo
RL

Vo =

1
RC

Vi dt

Vi

Vi

3
0

Vo

0
3

Vo
3

6
0

0.5

1
1.5
t (msec)

2.5

0.5

1
1.5
t (msec)

2.5

SEQUEL files: ee101 integrator 1.sqproj, ee101 integrator 2.sqproj


M. B. Patil, IIT Bombay

Offset voltage

Vi

Vo

0
1
Vi (mV)

Offset voltage
Real Op Amp
Ideal Op Amp

Vi

VOS

Vo

Vo

Vsat

Vsat

0
1
Vi (mV)

0
1
Vi (mV)

Offset voltage
Real Op Amp
Ideal Op Amp

Vi

VOS

Vo

Vo

Vsat

Vsat

0
1
Vi (mV)

0
1
Vi (mV)

For the real Op Amp, Vo = AV ((V+ + VOS ) V ) .

Offset voltage
Real Op Amp
Ideal Op Amp

Vi

VOS

Vo

Vo

Vsat

Vsat

0
1
Vi (mV)

0
1
Vi (mV)

For the real Op Amp, Vo = AV ((V+ + VOS ) V ) .


For Vo = 0 V , V+ + VOS V = 0 V+ V = VOS .

Offset voltage
Real Op Amp
Ideal Op Amp

Vi

VOS

Vo

Vo

Vsat

Vsat

0
1
Vi (mV)

0
1
Vi (mV)

For the real Op Amp, Vo = AV ((V+ + VOS ) V ) .


For Vo = 0 V , V+ + VOS V = 0 V+ V = VOS .
Vo versus Vi curve gets shifted.

Offset voltage
Real Op Amp
Ideal Op Amp

Vi

VOS

Vo

Vo

Vsat

Vsat

0
1
Vi (mV)

0
1
Vi (mV)

For the real Op Amp, Vo = AV ((V+ + VOS ) V ) .


For Vo = 0 V , V+ + VOS V = 0 V+ V = VOS .
Vo versus Vi curve gets shifted.
741: 6 mV < VOS < 6 mV .
OP-77: 50 V < VOS < 50 V .
M. B. Patil, IIT Bombay

Effect of VOS

10 k R2

10 k
1k

Vi

R2

1k

Vi

R1
Vo
RL

R1

Real

VOS

Ideal

Vo
RL

M. B. Patil, IIT Bombay

Effect of VOS

10 k R2

10 k
1k

Vi

R2

Real

1k

Vi

R1

R1

VOS

Vo

R2
Vi + VOS
R1

Vo
RL

RL

By superposition, Vo =

Ideal

1+

R2
R1

M. B. Patil, IIT Bombay

Effect of VOS

10 k R2

10 k
1k

Vi

R2

Real

1k

Vi

R1

R1

VOS

Vo

R2
Vi + VOS
R1

Vo
RL

RL

By superposition, Vo =

Ideal

1+

R2
R1

For VOS = 2 mV , the contribution from VOS to Vo is 22 mV ,

M. B. Patil, IIT Bombay

Effect of VOS

10 k R2

10 k
1k

Vi

R2

Real

1k

Vi

R1

R1

VOS

Vo

R2
Vi + VOS
R1

Vo
RL

RL

By superposition, Vo =

Ideal

1+

R2
R1

For VOS = 2 mV , the contribution from VOS to Vo is 22 mV ,


i.e., a DC shift of 22 mV .

M. B. Patil, IIT Bombay

Effect of VOS
Vc

Vc
C

C
i1
Vi

i1
Vi

R
Vo
RL

Real

VOS

Ideal

Vo
RL

M. B. Patil, IIT Bombay

Effect of VOS
Vc

Vc
C

C
i1
Vi

i1
Vi

Real

Vo
RL

V V+ = VOS i1 =

VOS

Ideal

Vo
RL

1
dVc
(Vi VOS ) = C
.
R
dt

M. B. Patil, IIT Bombay

Effect of VOS
Vc

Vc
C

C
i1
Vi

i1
Vi

Real

Vo
RL

VOS

Ideal

Vo
RL

1
dVc
V V+ = VOS i1 = (Vi VOS ) = C
.
R
dt
Z
1
i.e., Vc =
(Vi VOS ) dt .
RC

M. B. Patil, IIT Bombay

Effect of VOS
Vc

Vc
C

C
i1
Vi

i1
Vi

Real

Vo
RL

VOS

Ideal

Vo
RL

1
dVc
V V+ = VOS i1 = (Vi VOS ) = C
.
R
dt
Z
1
i.e., Vc =
(Vi VOS ) dt .
RC
Even with Vi = 0 V , Vc will keep rising or falling (depending on the sign of VOS ).
Eventually, the Op Amp will be driven into saturation.

M. B. Patil, IIT Bombay

Effect of VOS
Vc

Vc
C

C
i1
Vi

i1
Vi

Real

Vo
RL

VOS

Ideal

Vo
RL

1
dVc
V V+ = VOS i1 = (Vi VOS ) = C
.
R
dt
Z
1
i.e., Vc =
(Vi VOS ) dt .
RC
Even with Vi = 0 V , Vc will keep rising or falling (depending on the sign of VOS ).
Eventually, the Op Amp will be driven into saturation.
need to address this issue!

M. B. Patil, IIT Bombay

Effect of VOS
Integrator with Vi = 0 V :
R

C
i1

i1

Vc
R

Vc
R

Vo
VOS

(a)

Vo

RL

VOS

RL

(b)

M. B. Patil, IIT Bombay

Effect of VOS
Integrator with Vi = 0 V :
R

C
i1

i1

Vc
R

Vc
R

Vo
VOS

(a)
(a)

Vo

RL

VOS

RL

(b)

VOS
dVc
= C
R
dt
Z
1
Vc =
VOS dt Op Amp saturates.
RC

i1 =

M. B. Patil, IIT Bombay

Effect of VOS
Integrator with Vi = 0 V :
R

C
i1

i1

Vc
R

C
Vc

R
Vo
VOS

Vo

RL

(a)

VOS

RL

(b)

VOS
dVc
= C
R
dt
Z
1
Vc =
VOS dt Op Amp saturates.
RC

(a)

i1 =

(b)

There is a DC path for the current.

R0
Vo = 1 +
VOS .
R

M. B. Patil, IIT Bombay

Effect of VOS
Integrator with Vi = 0 V :
R

C
i1

i1

Vc
R

C
Vc

R
Vo
VOS

Vo

RL

(a)

VOS

RL

(b)

VOS
dVc
= C
R
dt
Z
1
Vc =
VOS dt Op Amp saturates.
RC

(a)

i1 =

(b)

There is a DC path for the current.

R0
Vo = 1 +
VOS .
R

R 0 should be small enough to have a negligible effect on Vo .

M. B. Patil, IIT Bombay

Effect of VOS
Integrator with Vi = 0 V :
R

C
i1

i1

Vc
R

C
Vc

R
Vo
VOS

Vo

RL

(a)

VOS

RL

(b)

VOS
dVc
= C
R
dt
Z
1
Vc =
VOS dt Op Amp saturates.
RC

(a)

i1 =

(b)

There is a DC path for the current.

R0
Vo = 1 +
VOS .
R

R 0 should be small enough to have a negligible effect on Vo .


However, R 0 must be large enough to ensure that the circuit still functions as an integrator.

M. B. Patil, IIT Bombay

Effect of VOS
Integrator with Vi = 0 V :
R

C
i1

i1

Vc
R

C
Vc

R
Vo
VOS

Vo

RL

(a)

VOS

RL

(b)

VOS
dVc
= C
R
dt
Z
1
Vc =
VOS dt Op Amp saturates.
RC

(a)

i1 =

(b)

There is a DC path for the current.

R0
Vo = 1 +
VOS .
R

R 0 should be small enough to have a negligible effect on Vo .


However, R 0 must be large enough to ensure that the circuit still functions as an integrator.
R 0  1/C at the frequency of interest.
M. B. Patil, IIT Bombay

Input bias currents

Q8

741

Q2

Q1

I+
B

I
B
Q3

Q4

Q7
Q5

R1

Q6

R3

R2

Input bias currents

Q8

741

Real Op Amp

I+
B

Ideal Op Amp

Vo
I
B
+

Q2

Q1

I+
B

I
B
Q3

Q4

Q7
Q5

R1

Q6

R3

R2

I+
B
I
B

Input bias currents

Q8

741

Real Op Amp

I+
B

Ideal Op Amp

Vo
I
B
+

Q2

Q1

I+
B

I
B
Q3

Q4

Q7
Q5

R1

Q6

R3

R2

IB+ and IB are generally not exactly equal.


|IB+ IB | : offset current (IOS )
(IB+ + IB )/2 : bias current (IB ).

I+
B
I
B

Input bias currents

Q8

741

Real Op Amp

I+
B

Ideal Op Amp

Vo
I
B
+

Q2

Q1

I+
B

I+
B

I
B
Q3

Q4

Q7
Q5

R1

I
B

Q6

R3

R2

Op Amp

IB

IOS

VOS

741

80 nA

20 nA

OP77

1.2 nA

0.3 nA

10 V

BJT input

411

50 pA

25 pA

0.8 mV

FET input

1 mV

BJT input

IB+ and IB are generally not exactly equal.


|IB+ IB | : offset current (IOS )
(IB+ + IB )/2 : bias current (IB ).
M. B. Patil, IIT Bombay

Effect of bias currents


Inverting amplifier:
i2
R2
Real

i1
Vi

Ideal

R1

Vo
RL

I
B
I+
B

M. B. Patil, IIT Bombay

Effect of bias currents


Inverting amplifier:
i2
R2
Real

i1
Vi

Ideal

R1

Vo
RL

I
B
I+
B

Assume that the Op Amp is ideal in other respects (i.e., VOS = 0 V , etc.).

M. B. Patil, IIT Bombay

Effect of bias currents


Inverting amplifier:
i2
R2
Real

i1
Vi

Ideal

R1

Vo
RL

I
B
I+
B

Assume that the Op Amp is ideal in other respects (i.e., VOS = 0 V , etc.).
V V+ = 0 V i1 = Vi /R1 .

M. B. Patil, IIT Bombay

Effect of bias currents


Inverting amplifier:
i2
R2
Real

i1
Vi

Ideal

R1

Vo
RL

I
B
I+
B

Assume that the Op Amp is ideal in other respects (i.e., VOS = 0 V , etc.).
V V+ = 0 V i1 = Vi /R1 .
i2 = i1 IB Vo = V i2 R2 = 0

Vi

IB
R1

R2 =

R2

Vi + IB R2 ,
R1

M. B. Patil, IIT Bombay

Effect of bias currents


Inverting amplifier:
i2
R2
Real

i1
Vi

Ideal

R1

Vo
RL

I
B
I+
B

Assume that the Op Amp is ideal in other respects (i.e., VOS = 0 V , etc.).
V V+ = 0 V i1 = Vi /R1 .

Vi

IB
R1
i.e., the bias current causes a DC shift in Vo .
i2 = i1 IB Vo = V i2 R2 = 0

R2 =

R2

Vi + IB R2 ,
R1

For IB = 80 nA, R2 = 10 k, Vo = 0.8 mV .

M. B. Patil, IIT Bombay

Effect of bias currents


Non-nverting amplifier:
i2
R2
Real

i1

Ideal

R1

Vo
Vi
RL

I
B
I+
B

M. B. Patil, IIT Bombay

Effect of bias currents


Non-nverting amplifier:
i2
R2
Real

i1

Ideal

R1

Vo
Vi
RL

I
B
I+
B

Assume that the Op Amp is ideal in other respects (i.e., VOS = 0 V , etc.).

M. B. Patil, IIT Bombay

Effect of bias currents


Non-nverting amplifier:
i2
R2
Real

i1

Ideal

R1

Vo
Vi
RL

I
B
I+
B

Assume that the Op Amp is ideal in other respects (i.e., VOS = 0 V , etc.).
V V+ = Vi i1 = Vi /R1 .

M. B. Patil, IIT Bombay

Effect of bias currents


Non-nverting amplifier:
i2
R2
Real

i1

Ideal

R1

Vo
Vi
RL

I
B
I+
B

Assume that the Op Amp is ideal in other respects (i.e., VOS = 0 V , etc.).
V V+ = Vi i1 = Vi /R1 .
i2 = i1 IB =

Vi

IB .
R1

M. B. Patil, IIT Bombay

Effect of bias currents


Non-nverting amplifier:
i2
R2
Real

i1

Ideal

R1

Vo
Vi
RL

I
B
I+
B

Assume that the Op Amp is ideal in other respects (i.e., VOS = 0 V , etc.).
V V+ = Vi i1 = Vi /R1 .
Vi

IB .
R1

Vi
R2

Vo = Vi i2 R2 = Vi
IB
R2 = Vi 1 +
+ IB R2 .
R1
R1
Again, a DC shift Vo .
i2 = i1 IB =

M. B. Patil, IIT Bombay

Effect of bias currents


Integrator:
i2

C
Vc

Real

i1
Vi

Ideal

Vo
RL

I
B
I+
B

Effect of bias currents


Integrator:
i2

C
Vc

Real

i1
Vi

Ideal

Vo
RL

I
B
I+
B

Even with Vi = 0 V , Vc =

1
C

IB dt will drive the Op Amp into saturation.

Effect of bias currents


Integrator:
R
i2

C
Vc

Real

i1
Vi

Ideal

Vo
RL

I
B
I+
B

Z
1

IB dt will drive the Op Amp into saturation.


C
Connecting R 0 across C provides a DC path for the current, and results in a DC shift
Vo = IB R 0 at the output.
Even with Vi = 0 V , Vc =

M. B. Patil, IIT Bombay

Effect of bias currents


Integrator:
R
i2

C
Vc

Real

i1
Vi

Ideal

Vo
RL

I
B
I+
B

Z
1

IB dt will drive the Op Amp into saturation.


C
Connecting R 0 across C provides a DC path for the current, and results in a DC shift
Vo = IB R 0 at the output.
Even with Vi = 0 V , Vc =

As we have discussed earlier, R 0 should be small enough to have a negligible effect on Vo .


However, R 0 must be large enough to ensure that the circuit still functions as an integrator.

M. B. Patil, IIT Bombay