Goals for today

• The operational amplifier

– input-output relationships
– feedback configuration
• More about zeros
– zero on the right-half plane: non-minimum phase response
– Zero-pole cancellation
• Next week
– Block diagram operations
– Analysis of a simple feedback system

2.004 Fall ’07 Lecture 10 – Friday, Sept. 28

 The operational amplifier (op-amp)
(a) Generally, vo = A (v2 − v1 ),
where A is the amplifier gain.
(b) When v2 is grounded, as is often
+V the case in practice, then vo = −Av1 .
+v1(t) v1(t) (Inverting amplifier.)
- vo(t) - vo(t)
+v2(t) A A (c) Often, A is large enough that
+ + we can approximate A → ∞.
Rather than connecting the input directly,
-V
the op—amp should then instead be used in the
feedback configuration of Fig. (c).
We have:
Z2(s)
V1 = 0; Ia = 0
Z1(s) I2(s) (because Vo must remain finite) therefore
Vi(s) V1(s)
- Vo(s)
Ia(s) I1 + I2 = 0;
I1(s) +
V i − V 1 = V i = I 1 Z1 ;
Vo − V1 = Vo = I2 Z2 .
Combining, we obtain
Figure by MIT OpenCourseWare.
Figure 2.10 Vo (s) Z2 (s)
(see also Lecture 04 – page 16) =− .
Vi (s) Z1 (s)

2.004 Fall ’07 Lecture 10 – Friday, Sept. 28

 Example: PID controller
R2 = 220 kΩ C2 = 0.1 µF Z2(s)
C1 = 5.6 µF
Z1(s) I2(s)
vi(t) v1(t) Vi(s) V1(s)
- -
vo(t) vo(s)
Ia(s)
R1 = 360 kΩ + I1(s) +

Figure by MIT OpenCourseWare. Figure 2.11 Figure 2.10

Equivalent impedances: ⇒ Transfer Function:

(R1 , C1 connected in parallel) Vo (s) s2 + 45.95s + 22.55

= −1.232 .
Vi (s) s
1 1 R1 360 × 103
= + C1 s ⇒ Z1 (s) = = ;
Z1 (s) R1 1 + R1 C1 s 1 + 2.016s

(R2 , C2 connected in series)

1 1
Z2 (s) = R2 + = 220 × 103 + −7 .
C2 s 10 s

2.004 Fall ’07 Lecture 10 – Friday, Sept. 28

 Example: all-pass filter Step response without zero:

1 1/10 1/10
Co (s) = − =− +
s (s + 10) s s + 10
R1 1 ¡ ¢
c0 (t) = − 1 − e−10t u(t).
10
R2
V1(s)
- Vo(s) Images removed due to copyright restrictions.
I(s) Please see Fig. 4.28 in Nise, Norman S. Control Systems Engineering.
Vi(s) 4th ed. Hoboken, NJ: John Wiley, 2004.
+
R3 V2(s)

Figure by MIT OpenCourseWare.

Transfer function: jω Step response:
R2 s − 10
s− σ C(s) = −
Vo (s) R1 R1 R3 C s (s + 10)
=− . −10 +10 1 2
Vi (s) R2 1
s+ = − ⇒
R3 C s s + 10
¡ ¢
c(t) = 1 − 2e−10t u(t).
Substituting R1 = R2 , R3 = 100kΩ, C = 1μF,
Non—minimum phase system
Vo (s) s − 10
=− zero in the r.h.p.
Vi (s) s + 10

2.004 Fall ’07 Lecture 10 – Friday, Sept. 28

Nonminimum-phase response
Consider a system without a zero, whose step response is Co (s)
and recall that the effect of the zero is C(s) = (s + a)Co (s) = sCo (s) + aCo (s).
In the time domain, c(t) = ċo (t) + aco (t). Therefore,
the system response with the zero is the sum of the derivative of the
original response plus the original response amplified by a gain equal to a
(”proportional term.”)

If a < 0 and the derivative term ċo (t = 0) is larger than the

proportional term aco (t = 0), then the response will initially follow
the derivative term in the opposite direction of the proportional term.

Images removed due to copyright restrictions.

Please see: Fig. 4.26 in Nise, Norman S. Control Systems Engineering. 4th ed. Hoboken, NJ: John Wiley, 2004.

2.004 Fall ’07 Lecture 10 – Friday, Sept. 28

Zero-pole cancellation
Compare the step responses

26.25(s + 4)
C1 (s) =
s(s + 3.5)(s + 5)(s + 6)
26.25(s + 4)
C2 (s) = .
s(s + 4.01)(s + 5)(s + 6)
non-negligible
Partial fraction expansion yields
negligible
1 3.5 3.5 1
C1 (s) = − + −
s s + 5 s + 6 s + 3.5
0.87 5.3 4.4 0.033
C2 (s) = − + − .
s s + 5 s + 6 s + 4.01

C1 (s) jω C2 (s) jω
−4 σ −4 σ
−6 −5 −3.5 input −6 −5 −4.01 input
pole pole

zero-pole cancellation does not occur zero-pole cancellation occurs

2.004 Fall ’07 Lecture 10 – Friday, Sept. 28