Lecture 28 Operational Amplifiers Reading: Jaeger 11.1-11.5 and Notes
Georgia Tech
ECE 3040 - Dr. Alan Doolittle
Operational Amplifier •Operational Amplifier or “Op-Amp” is a multistage amplifier that is used for general electrical signal manipulation. •The numbers of applications possible with Op-amps are two numerous to list. •Most everyone agrees: “Op-Amp analysis is significantly easier than transistor analysis.” •Though they are often internally complex, their use in circuits most often simplifies the overall design. •The circuit is modeled by an ideal voltage amplifier. Model
Circuit Symbol
+ Vin -
+ + Vout -
Georgia Tech
Vin -
+ Vout -
ECE 3040 - Dr. Alan Doolittle
Ideal Operational Amplifier •Rin = Infinity, •Voltage Gain, Av=Infinity at all frequencies •Rout=0
Model
+ Vin -
Georgia Tech
+ Vout -
ECE 3040 - Dr. Alan Doolittle
Ideal Operational Amplifier •Infinite gain means that the device is useless without adding “Feedback” to control the overall gain to a finite value.
i+ v+ iin •Consider the circuit to the right with vin=0 vout = Av (− v − )
vin
R1 v − = vout R1 + R 2 R1 v − = − Av v − R1 + R 2 If A v → ∞, the above equation is only satisfied for v - = 0
iout
vi-
vout i2
•Feedback forces the two input voltages to be equal! This is known as a “virtual ground”. •R1 and R2 form a “Feedback Network” Georgia Tech
ECE 3040 - Dr. Alan Doolittle
Inverting Amplifier •Finite voltage gain results from an infinite voltage gain amplifier with “negative feedback” (feedback that takes a fraction of the output voltage and mixes it back into the negative summation node). i+
1) vin − iin R1 − i2 R2 − vo = 0
v+
2) iin = i− + i2 = i2 due to infinite input resistance iin
v -v i in = in R1 but v - = 0 due to the virtual ground v 3) i in = in R1 Combining 1, 2 and 3, v v vin − in R1 − in R2 − vo = 0 R1 R1 vo R =− 2 v in R1
Georgia Tech
iout
vi-
vin
vout i2
Overall circuit gain is finite, negative (for this feedback configuration) and set by the feedback resistor network. ECE 3040 - Dr. Alan Doolittle
Inverting Amplifier •Input Resistance:
vin = R1 Rin = iin
i+ v+ iin
iout
vi-
vin
Georgia Tech
vout i2
ECE 3040 - Dr. Alan Doolittle
Inverting Amplifier •Output Resistance:
v t = i1 R 3 + i 2 R 1 but, i1 = 0 since v - = 0 and
v+
it
i1 v-
i2
vt
i1 = i 2 thus, vt = 0 + 0 Rout
vt = =0 it
Georgia Tech
ECE 3040 - Dr. Alan Doolittle
Non-Inverting Amplifier iin
Virtual Ground or “short”
v+ vin
vi-
v − = vout
vout
R1 R1 + R2
The virtual ground requires that v + = v - so, v in = v -
i2
so, vout vout R1 + R2 = = vin v− R1 Av =
Georgia Tech
vout R = 1+ 2 vin R1
Rin =
vin vin = =∞ iin 0
Rout = 0 (Same circuit as for Non - inverting case)
ECE 3040 - Dr. Alan Doolittle
Unity Gain Buffer or “Voltage Follower”
Same as Non − inverting amplifier except R2 = 0 and R1 = ∞ Av =
vout R = 1 + 2 ⇒ Av = 1 v− R1
Rin =
vin vin = =∞ iin 0
Rout = 0 (Same circuit as for Non - inverting case)
•Can be used to isolate a high impedance circuit from a low impedance circuit Georgia Tech
ECE 3040 - Dr. Alan Doolittle
Summing Amplifier v+ iin,a
v-
i2 = iin ,a + iin ,b + iin ,c vin ,a vin ,b vin ,c v − out = + + R 2 R1a R1b R1c R2 R2 R2 − vin ,b − vin ,c vout = −vin ,a R1a R1b R1c R2 R2 R2 + vin ,b + vin ,c vout = − vin ,a R c R b 1 1 1 R a
vin,a
Georgia Tech
vout
iin,b vin,b iin,c
•Output is a scaled sum of inputs. •Scaling can be controlled by ratios of resistors
i2
vin,c ECE 3040 - Dr. Alan Doolittle
Difference Amplifier v+
v+
vin,a
vth
v-
v-
vout
vout vin,b
vin,b vth = vin ,a
R2 R1 + R2
and
Rth = R1 || R2
Using Superposition we can combine the results of the Inverting and Non-inverting solutions: vin,b=0 vin,a=0 v out
Georgia Tech
R = vth 1 + 2 R1
v out = vin ,a
R2 R1 + R2
v out = vin ,a
R2 R1
vout = − v in,b R2 1 + R 1
R2 R1
vout = v in,a
R2 R − v in,b 2 R1 R1
vout = (v in,a − v in,b )
R2 R1
This circuit amplifies the difference of two signals ECE 3040 - Dr. Alan Doolittle
Non-Ideal (Real World) Operational Amplifiers Finite Open-Loop Gain
•Real op-amps do not have “infinite” “open loop (without feedback)” gain. •Voltage gains are typically large but finite: ~104-106 V/V •Finite gain causes a deviation from ideal amplifier behavior v − = v out
R1 = βv out R1 + R 2
where β =
v+
R1 is known as the feedback factor R1 + R 2
v out = A openloop (v + − v − ) = A openloop (v + − βv out )
vin
v-
vout
so, A v, closed loop = If
A openloop v out = , where β A openloop is the loop gain v in 1 + β A openloop
β A openloop >> 1
A v, closed loop =
Georgia Tech
1
β
= 1+
R2 R1
⇒ approaches the infinite gain result
ECE 3040 - Dr. Alan Doolittle
Non-Ideal (Real World) Operational Amplifiers Finite Open-Loop Gain
•Finite open-loop gain means the Virtual Ground is not perfect! v + − v − = v + − β vout
A openloop v+ v + = = v+ − 1+ β A 1 + β A openloop openloop
Small but finite offset between + and - terminals
v+ vin
v-
vout
•The Gain Error (GE) that results from the Non-infinite open-loop gain can be quantified as: 1 A openloop v+ GE = − v+ = β (1 + β A openloop ) β 1 + β A openloop Georgia Tech
ECE 3040 - Dr. Alan Doolittle
Non-Ideal (Real World) Operational Amplifiers Finite Output Impedance •Real Op-Amps have a small but finite output impedance, Ro . •We want to find the Output impedance of the various circuits we have examined . •All the configurations have a common circuit for calculating the output impedance .
Georgia Tech
ECE 3040 - Dr. Alan Doolittle
Non-Ideal (Real World) Operational Amplifiers Finite Output Impedance Rout =
vx ix
i x = io + i 2 io =
v x − Av ,openloop (v + − v − ) Ro
v x = i2 (R1 + R2 ) ⇒ i2 = v− =
vx (R1 + R2 )
R1 v = β vx (R1 + R2 ) x
(Rout )
−1
i x 1 + Av ,openloop β 1 = = + (R1 + R2 ) vx Ro
Ro Rout = 1+ A v ,openloop β
(R1 + R2 )
Ro is very small so this term is EXTREMELY small! Georgia Tech
ECE 3040 - Dr. Alan Doolittle
Non-Ideal (Real World) Operational Amplifiers Finite Input Impedance: Non-Inverting Case
•Real Op-Amps have a large but finite input resistance, RID ix =
v x − v− R ID
Neglecting the current i x compared to i1 and i 2 (due to R ID >> R 1 or R 2 ) v − = i1 R1 ≈ i2 R1 v− = v− =
R1 vout = β vout = Av ,openloop β (v x − v − ) R1 + R2 Av ,openloop β 1 + Av ,openloop β
vx
Av ,openloop β vx − vx 1+ A vx v ,openloop β = ix = (1 + Av,openloop β )RID R ID Rin = (1 + Av ,openloop β )R ID
Georgia Tech
RID is very large so Rin is EXTREMELY large! ECE 3040 - Dr. Alan Doolittle
Non-Ideal (Real World) Operational Amplifiers Finite Input Impedance: Inverting Case
•Real Op-Amps have a large but finite input resistance, RID
Rin R’in
•Rin=R1 +R’in •Find R’in by forming a new test circuit i1 = i− + i2 =
(R'in )−1 =
v1 + Av ,openloop v1 v −v v1 v + 1 out = 1 + R ID R2 R ID R2
i1 1 1 + Av ,openloop = + v1 R ID R2
R2 R 'in = (RID ) 1+ A v , openloop Thus,
R’in
R2 Rin = R1 + (RID ) 1+ A v , openloop
Since RID>>R2/(1+Av,openloop) and Av,openloop is very large, R2 Rin = R1 + (R ID ) 1+ A v ,openloop Georgia Tech
R2 ⇒ Rin = R1 + 1+ A v , openloop
≈ R1 ECE 3040 - Dr. Alan Doolittle