Non-ideal Operational

Amplifier
• Various error terms arise in practical operational
amplifiers due to non-ideal behavior.
• Some of the non-ideal characteristics include:
1. Finite open-loop gain that causes gain error
2. Nonzero output resistance
3. Finite input resistance
4. Finite CMRR
5. Common-mode input resistance
6. DC error sources
7. Output voltage and current limits

Zulfiqar Ali 1
 vo = Av = A(vs − v ) = A(vs − βvo )
id 1
v A
Av = o =
v s 1 + Aβ
Aβ is called loop gain.
For Aβ >>1,

1 R
A = = 1+ 2
ideal β R
1
v = vs − v = vs − βvo
R id 1
v = 1 v = βv Aβ v
1 R +R o o = vs − vs = s
1 2 1 + Aβ 1 + Aβ
R is called
β= 1 feedback
No longer zero, vid is small for
R +R factor.
1 2 large Aβ.
Gain Error is given by
GE= (ideal gain)-(actual gain)
For non-inverting amplifier,

1 A 1
GE = − =
β 1+ Aβ β (1+ Aβ )

Gain error is also expressed as a fractional or percentage error.

A 1
−
FGE = β 1+ Aβ =
1 1
≅
1 1 + Aβ Aβ
β
Problem: Find ideal and actual gain and gain error is percent
Given data: Closed-loop gain of 200 (46 dB), open-loop gain of op amp is
10,000 (80 dB).
Approach:Amplifier is designed to give ideal gain and deviations from ideal
case are determined. Hence, .

R1 and R2 aren’t designed to compensate for finite open-loop gain of

amplifier.
Analysis:

A 104
Av = = = 196
1 + Aβ 10 4
1+
200
200 −196
FGE = = 0.02
200
Output terminal is driven by test source vx
and current ix is calculated to determine
output resistance (all independent sources
are turned off).The equivalent circuit is same
For both inverting and non-inverting
amplifiers. v
Rout = x
ix
 R
Analysis: ix = io + i v = 1 v = βv
2 Also, vid= -v1 and 1 R +R x x
v x - Av vx 1 2
io = id i = 1 i 1 + Aβ 1
Ro 2 R +R ∴ = x = +
1 2 R vx Ro R +R
out 1 2
Thus, shunt feedback at output reduces Rout.

Ro ⎛ ⎞
∴Rout = ⎜R + R ⎟
⎜ ⎟
1+ Aβ ⎝ 1 2 ⎠
Ro If A is infinite, Rout=0
Since, Ro/(1+Aβ)<<(R1+R2), Rout ≅
1 + Aβ
• Problem: Design non-inverting amplifier and find
open-loop gain
• Given Data: Av=35 dB, Rout =0.2Ω, Ro = 250Ω
• Analysis:
Av = 1035dB / 20dB = 56.2 β= 1 = 1
Av 56.2
R
Rout = o ≤ 0.2Ω
1 + Aβ
⎛ ⎞
⎜ ⎟
1 Ro ⎛ 250 ⎞
∴A ≥ ⎜
⎜ −1⎟⎟ = 56.2⎜⎜ −1⎟⎟ = 7.03×104 = 96.9dB
β R
⎜⎜ ⎟⎟
⎜
⎝ 0.2
⎟
⎠
out ⎠
⎝
 Assuming i-<<i2 implies i1 = i2.

R
v ≅ 1 v = βv
1 R +R o o
1 2
= β(Av )= Aβ(v x − v )
id 1
Aβ
v = v
1 1+ Aβ x
Aβ
vx − vx
Test voltage source vx is applied to 1+ A β vx
input and current ix is calculated. ∴ix = =
R (1+ Aβ )R
vx − v id id
ix = 1 v =i R ≅i R R = R (1+ Aβ )
R 1 1 1 2 1 in id
id
A real amplifier responds to signal common to both inputs, called
common-mode input voltage. In general,
⎛
⎜
v + v ⎞⎟
vo = A(v − v ) + Acm 1 2 ⎟
⎜
1 2 ⎜ 2 ⎟⎠
⎝
= A(v ) + Acm (v )
id ic
 Ideal gain for voltage follower is unity,
gain error
A
1−
GE = 1− Av = 2CMRR
⎛ 1 ⎞
1+ A⎜⎜1− ⎟
vs + v o ⎜
⎝ 2CMRR
⎟⎟
⎠
v = vs − vo v =
id ic 2 Since, both A and CMRR are
normally >>1,
(v + v ) ⎞
vo = A⎜⎜ (vs − vo )+ s o ⎟⎟
⎛

2CMRR ⎠ 1 1
⎝
GE ≅ −
⎛ 1 ⎞ A CMRR
A⎜⎜1+ ⎟
⎟⎟
vo ⎜
2CMRR First term is due to finite amplifier
Av = = ⎝ ⎠
vs 1+ A⎛⎜1− 1 ⎞
⎟ gain, second term shows that
⎜⎜ ⎟⎟ CMRR may introduce an even larger
⎝ 2CMRR ⎠
error.

Zulfiqar Ali 10
With inputs being zero, the amplifier To include effect of offset voltage
output rests at some dc voltage level ⎛ v ⎞

instead of zero. The equivalent dc

⎜
vo = A v +
⎜
ic +V ⎟⎟
⎜
⎝
id CMRR OS ⎟⎠
input offset voltage is
V If vid =0,
V = o
OS A v ⎛ ⎞
The amplifier is connected as vo = A ⎜
ic
⎜ +V ⎟⎟ = A(v )
voltage-follower to give output CMRR OS ⎟⎠
⎜ OS
⎝
voltage equal to offset voltage. v
∴CMRR = v ic μV/V
OS
Thus, CMRR is a measure of
how total offset voltage changes
from its dc value when common-
mode voltage is applied.