Chapter4

Time Domain Analysis of Control Systems

4.1. Introduction
 The first step in analysing a control system was to derive a
mathematical mode of the system. Once such a model is obtained,
various methods are available for the analysis of system performance.

 Performance of a system is its response to an input signal. In

designing a control system we have to choose a system with better
performance, i.e., systems are to be compared by their response to
particular test signal.
By Zewde T. 1
 Cont. …
 The commonly used test input signals are those of step functions,
ramp functions, impulse functions, sinusoidal functions and the like.

 Transient response

 The transient response of a system is a particular part of the response

of the system which tends to zero as time increases. It goes from
initial state to the final state.

 Steady state response

 The steady state response of a system is a particular part of the

response of the system which remains after the transient part has
reached zero.
By Zewde T. 2
 Cont. …
 The total solution of a system y(t) is the sum of the transient response
(𝑦𝑡𝑟 ) and the steady state response (𝑦𝑠𝑠 ).
𝑦 𝑡 = 𝑦𝑡𝑟 𝑡 + 𝑦𝑠𝑠 (𝑡)
lim 𝑦𝑡𝑟 𝑡 = 0
𝑡→∞
lim 𝑦 𝑡 = 𝑦𝑠𝑠 (𝑡)
𝑡→∞

4.2. Time Domain Transient Response Specifications

 In many practical cases, the desired performance characteristics of
control systems are specified in terms of time domain quantities.
Systems with energy storage can not respond instantaneously and will
exhibit responses whenever they are subjected to inputs or
disturbances. By Zewde T. 3
 Cont. …
 Frequently, the performance characteristics of a control system are
specified in terms of the transient response of a unit step input since it
is easy to generate and is sufficiently drastic.

 In specifying the transient response characteristics of a control system

to a unit step input, It is common to specify the following.

i. Delay time (𝒕𝒅 ): it is the time required for the response to reach half
of its final value.

𝑦 𝑡𝑑 = 50%. 𝑟 𝑡 = 1 2 . 1 = 0.5

i. Peak time (𝒕𝒑 ): it is the time required for the response to reach the

𝑑𝑦(𝑡)
first peak of the shoot. 𝑑𝑡
⃒𝑡 = 𝑡𝑝 = 0
By Zewde T. 4
 Cont. …
iii. Rise time (𝒕𝒓 ): it is the time require for the response to rise from
10% to 90%, 5% to 95% or 0% to 100% of its final value. For under
damped second order systems, the 0% to 100% rise time is normally
used. For over damped systems, the 10% to 90% rise time is
commonly used.

For 0 to 100% rising

𝑦 𝑡𝑟 = 100%. 𝑟 𝑡 = 1.1 = 1 since 𝑟 𝑡 = 1

iv. Settling time (𝒕𝒔 ): this is the time required by the response y(t) to
reach and remain within a certain range of its final value. This range
is usually from 2 to 5% of amplitude of the final value.
By Zewde T. 5
 Cont. …

v. Maximum overshoot (𝐌𝐩): it is the maximum peak value of the

response curve measured from unity. It is the difference between the

maximum value of the response and the final value (the steady state

response). The amount of maximum par cent overshoot directly

indicates the relative stability of the system.

y t p − y(∞)
Mp = ∗ 100%
y(∞)

By Zewde T. 6
 Cont. …

Fig.4.1: Unit-step response curve showing t d , t r , t p , t s , and Mp .

By Zewde T. 7
 Cont. …
4.2.1. Analysis of first order systems
 Consider a first order and its equivalent circuit diagram as shown
below

The input output relationship is given by

Y(s) 1
=
R(s) Ts + 1
By Zewde T. 8
 Cont. …
i. Unit-Step Response of First-Order Systems.

Y(s) 1 1
= →Y s = R s
R(s) Ts+1 Ts+1

1 1
But R s = then Y s =
s Ts+1 s

Expanding Y(s) into partial fractions gives

1 1 1 1
Y s = − = −
s Ts + 1 s s + 1
T
Taking the inverse Laplace transform
1
𝑦 𝑡 = 1 − 𝑒− 𝑇 𝑡
for 𝑡 ≥ 0
 From this equation, initially the output y(t) is zero and finally it becomes
unity.
By Zewde T. 9
 Cont. …
 One important characteristic of such an exponential response curve
y(t) is that at t = T the value of y(t) is 0.632, or the response y(t) has
reached 63.2% of its total change. This may be easily seen by
substituting t = T in y(t). That is,
−(T T)
At t = T, y T =1−e = 1 − e−1 = 0.632

 For 2% of final value, the settling time can be derived as

y t s = 0.98 ∗ r t = 0.98 ∗ 1 = 0.98
t
−( s T)
y t s = 0.98 → 1 − e = 0.98
→ t s ≈ 4T

 For t ≥ 4T, the response reaches to its steady state value.

By Zewde T. 10
 Cont. …
− 𝑡
Steady state response: 𝑦𝑠𝑠 𝑡 = lim 𝑦(𝑡) = lim 1−𝑒 𝑇 =1
𝑡→∞ 𝑡→∞
−𝑡 −𝑡
Transient response: 𝑦𝑡𝑟 𝑡 = 𝑦 𝑡 − 𝑦𝑠𝑠 𝑡 = 1 − 𝑒 𝑇 −1 =𝑒 𝑇

By Zewde T. 11
Fig.4.2: unit step response of first order system
 Cont. …
ii. Unit-Ramp Response of First-Order Systems.
1 1
Ys = R s = 2
TS + 1 s (Ts + 1)
1
Since the Laplace transform of the unit-ramp function is 𝑅 𝑠 = 𝑠2
Expanding Y(s) into partial fraction gives

1 T T2
Y s = 2− +
s s Ts + 1
Taking the inverse Laplace transform gives
−t
y t = t − T + Te T , for t ≥ 0

The transient response of y(t) can be obtained as

−t
ytr t = lim y(t) = lim (t − T + Te T) =0
t→∞ Byt→∞
Zewde T. 12
 Cont. …
−t
 From this limit only Te T will converges to zero as time tends to
−t
infinity. → 𝑦𝑡𝑟 𝑡 = Te T

 Steady state response:

−t −t
𝑦𝑠𝑠 𝑡 = 𝑦 𝑡 − 𝑦𝑡𝑟 𝑡 = t − T + Te T − Te T =t−T

 The error signal e(t) is then

−t −t
e t = r t − y t = t − t − T + Te T = T(1 − e T )
 At steady state the error signal will converges to T
−t
lim e(t) = lim T(1 − e T) =T
t→∞ t→∞

 The smaller the time constant T, the smaller the steady state error in
ramp input. By Zewde T. 13
 Cont. …

Fig.4.3: Ramp response of first order system.

By Zewde T. 14
 Cont. …
iii. Unit-Impulse Response of First-Order Systems.
 For the unit-impulse input, R(s) = 1 and the output of first order
system can be obtained as

1 1 1
Ys = R s = = 𝑇
TS + 1 (Ts + 1) 𝑠 + 1
𝑇
1 −𝑡
→ 𝑦 𝑡 = 𝑇𝑒 𝑇 for t ≥ 0

Fig.4.4:impulse response of first order system

Exercise 4.1: find steady state and transient response.

By Zewde T. 15
 Cont. …
 An Important Property of LTI Systems.
 The response to the derivative or integral of the input signal can be
obtained by differentiating or integrating the response of the signal
respectively.
 For unit ramp response of first order systems
−t
If u(t) 1−e T

t t −τ −t
0
u t dt = r(t) 0
1−e T dτ = t − T + Te T

 For unit impulse response of first order systems

−𝑡
If 𝑢(𝑡) 1−𝑒 𝑇

𝑑𝑢(𝑡) 𝑑 −𝑡 1 −𝑡
= 𝛿(𝑡) 1 −By Zewde
𝑒 T.𝑇 = 𝑒 𝑇
𝑑𝑡 𝑑𝑡 𝑇 16
 Cont. …
4.2.2. Analysis of second order systems
 Consider a second order system with transfer function

𝜔𝑛 2
𝐺 𝑠 = 2
𝑠 + 2ζ𝜔𝑛 𝑠 + 𝜔𝑛 2

Where the constants 𝜔𝑛 and ζ are called the natural undamped frequency
and the damping ratio of the system respectively

 G(s) can be also written in the form of the following

𝜔𝑛 2
𝐺 𝑠 =
(𝑠 + ζ𝜔𝑛 + 𝜔𝑛 ζ2 − 1)(𝑠 + ζ𝜔𝑛 − 𝜔𝑛 ζ2 − 1)
By Zewde T. 17
 Cont. …
 The poles of G(s) are

𝑠1, 2 = −ζ𝜔𝑛 ± 𝜔𝑛 ζ2 − 1 = −ζ𝜔𝑛 ± 𝜔𝑛 (−1)(1 − ζ2 )

𝑠1,2 = −ζ𝜔𝑛 ± 𝜔𝑛 −1 1 − ζ2 = −ζ𝜔𝑛 ± 𝑗𝜔𝑛 1 − ζ2

𝑠1,2 = −𝜎 ± 𝑗𝜔𝑑

where 𝜎 = ζ𝜔𝑛 is damping constant or attenuation

𝜔𝑑 = 𝜔𝑛 1 − ζ2 is damped natural frequency of the system

By Zewde T. 18
 Cont. …
Example4.1: RLC circuit

di(t)
vi t = Ri t + L dt
+ vo (t) (4.1)

1
vo t = C i t dt

dvo (t)
i t =C (4.2)
dt

Taking Laplace transform gives

vi s = RI s + LsI s + vo (s) (4.3)

I s = CSvo (s) By Zewde T. (4.4) 19

 Cont. …
 Substituting eqn.(4.4) into eqn.(4.3) and rearranging will result the
following transfer function.

𝐯𝐨 (𝐬) 𝟏

𝐯𝐢 (𝐬)
= 𝐑
𝐋𝐂
𝐬 𝟐+ 𝐬+𝟏 𝐋𝐂
𝐋

 When we comparing this result to the general second order system, we

will got the following results.

2 1 1 1
𝜔𝑛 = → 𝜔𝑛 = =
𝐿𝐶 𝐿𝐶 𝐿𝐶

𝑅 𝑅 𝑅 𝑅 𝐶
2ζ𝜔𝑛 = →ζ= = ∗ 𝐿𝐶 =
𝐿 2𝐿𝜔𝑛 2𝐿 2 𝐿
By Zewde T. 20
 Cont. …
 The dynamic behaviour of the second order systems can then be
described in terms of two parameters ζ and 𝜔𝑛 .

• case1: If 0 < ζ < 1, the closed loop poles are complex conjugates and
lie in the left half s-plane. The system is called under damped and the
transient response is oscillatory.

• Case2: If 𝜁 = 0, the transient response does not die out.

• Case3: If ζ = 1, the two poles are equal and the system is called
critically damped.

• Case4: If ζ > 1, the closed loop poles are negative real and unequal
and the system is called over damped.
By Zewde T. 21
 Cont. …
4.2.2.1. Unit step response of second order systems
ωn 2 1
C s = s(s2+2ζω 2) where R s =
n s+ωn s

1 s+2ζωn
C s = −
s s2 +2ζωn s+ωn 2

1 s+ζωn ζωn
C s = − 2 −
s s2 +2ζωn s+ωn s2 +2ζωn s+ωn 2

1 s+ζωn ζωn
C s = − −
s (s+ζωn )2 +(ωd )2 (s+ζωn )2 +(ωd )2

1 s+ζωn ζωn
C s = s − (s+ζω 2 +ω 2 (1−ζ2 ) − (s+ζω 2 +ω 2 (1−ζ2 )
n) n n) n

By Zewde T. 22
 Cont. …
 Case1: undamped system (𝜻 = 𝟎)
1 s
C s = −
s s2 +ωn 2

s2 +ωn 2 −s2 ωn 2
C s = s(s2+ωn 2)
= s(s2 +ω 2)
n

 In this case the poles of C(s) are 0 and imaginary 𝑠 = ±𝑗𝜔𝑛 . If we

expand C(s) in partial fractions, we have

1 s
C s = s − s2 +ω 2
n

1 s
∴ c t = L−1 C(s) = L−1 − = 1 − cosωn t
s s2 +ωn 2

By Zewde T. 23
 Cont. …
 As we can observe from the above equation, the response c(t) is a
sustained oscillation with constant frequency 𝜔𝑛 and constant
amplitude equal to 1. in this case the system is called undamped.

Fig.4.5: unit step response of undamped second order system.

By Zewde T. 24
 Cont. …
 Case2: Under damped system (𝟎 < 𝜻 < 𝟏):
 In this case the poles of G(s)=C(s)/R(s) are complex conjugate pair
since 𝑠1, 2 = −𝜎 ± 𝑗𝜔𝑑 .

 For a unit step input, the response C(s) can be written as

ωn 2
C s =
s(s2+2ζωn s+ωn 2 )

1 s+2ζωn
C s = s − (s2 +2ζω 2)
n s+ωn

1 s+2ζωn
C s = s − (s2 +2ζω 2 2 2 2 2
n s+ζ ωn −ζ ωn +ωn )

1 s+2ζωn
C s = −
s (s2 +2ζωn s+ζ2 ωn 2)+ωn
By2Zewde
(1−ζT. 2 ) 25
 Cont. …
But s 2 + 2ζωn s + ζ2ωn 2 = (s + ζωn )2 and ωn 2 1 − ζ2 = ωd 2

1 s+ζωn +ζωn
C s = −
s (s+ζωn )2 +ωd 2

1 s+ζωn ζωn ωd
C s = − 2 − ∗
s (s+ζωn )2 +ωd (s+ζωn )2 +ωd 2 ωd

1 s+ζωn ζωn ωd ωd
C s = s − (s+ζω 2 2 − (s+ζωn )2+ωd 2
∗ω
n ) +ωd ωd d

𝑠 ωd
But 𝐿−1 𝑠2 +ωd 2
= 𝑐𝑜𝑠𝜔𝑑 𝑡 𝐿−1
𝑠2 +ωd 2
= 𝑠𝑖𝑛𝜔𝑑 𝑡

𝑠+ζωn ωd
𝐿−1 = 𝑒 −ζ𝜔𝑛 𝑡 𝑐𝑜𝑠𝜔𝑑 𝑡 𝐿−1 = 𝑒 −ζ𝜔𝑛 𝑡 𝑠𝑖𝑛𝜔𝑑 𝑡
(s+ζωn )2 +ωd 2 (s+ζωn )2 +ωd 2

Then c t = L−1 C(s)

By Zewde T. 26
 Cont. …
ζωn
c t = 1 − e−ζωn t cosωd t − e−ζωn t sinωdt
ωd

−ζωn t ζωn
c t = 1−e cosωd t + sinωd t
ωn (1−ζ2 )

ζ
c t = 1 − e−ζωn t cosωd t + sinωdt
(1−ζ2 )
1
c t = 1− ζ𝑠𝑖𝑛ωd t + (1 − ζ2 )cosωd t e−ζωn t
(1−ζ2 )

(1−𝜁 2 )
Let 𝑐𝑜𝑠𝜃 = 𝜁, 𝑠𝑖𝑛𝜃 = (1 − 𝜁2 ) → 𝑡𝑎𝑛𝜃 = 𝜁

−1 (1−𝜁 2 )
Then 𝜃 = 𝑡𝑎𝑛
𝜁

e−ζωn t
c t =1− cosθsinωdt + sinθcosωdt
(1−ζ2 )

e−ζωn t
c t =1− sin(ωd t + θ)
1−ζ2 By Zewde T. 27
 Cont. …
𝐞−𝛇𝛚𝐧 𝐭 −𝟏 (𝟏−𝛇𝟐)
∴𝐜 𝐭 = 𝟏− 𝐬𝐢𝐧(𝛚𝐝𝐭 + 𝐭𝐚𝐧 𝛇
) for 𝑡 ≥ 0.
𝟏−𝛇𝟐

 Thus , when 0 < ζ < 1 we observe that the response c(t) is damped
oscillation which tends to 1 as t → ∞. In this case, we say that the
system is under damped.

lim c(t) = 1 = Css (t)

t→∞

The error signal for this system is

e−ζωn t
e t = r t − c t = 1 − (1 − sin ωd t + θ
1−ζ2

e−ζωn t
e t = sin(ωd t + θ)
1−ζ2
By Zewde T. 28
 Cont. …
 This error signal exhibits a damped sinusoidal oscillation. At steady
state (𝑡 → 𝜃), no error exists between the input and output.

Fig.4.6: Unit step input response of under damped system for different
damping ratio.
By Zewde T. 29
 Cont. …
 Case3: Critically damped system (𝜻 = 𝟏)
 In this case the poles of G(s)=C(s)/R(s) are real double pole −ωn . For
1
unit step input R s = , the response C(s) of critically damped case
s

becomes

ωn 2 ωn 2
C s = s(s2+2ω 2 =s s+ωn 2
n s+ωn )

If we expand C(s) in partial fractions we have

1 1 ωn
C s = − −
s s+ωn s s+ωn 2

c t = L−1 C(s) = 1 − e−ωn t − ωn te−ωn t

c t = 1 − e−ωn t (1 + ωn t), for

By Zewde T.
t ≥ 0. 30
 Cont. …
 Thus, when ζ = 1 we observe that the wave form of the response c(t)
involves no oscillations, asymptotically tends to 1 as t → ∞.

Fig.4.7: unit step response of second order critically damped system.

By Zewde T. 31
 Cont. …
Case4: overdamped system (𝜻 > 𝟏)
 In this case, the two poles of G(s)=C(s)/R(s) are negative real and
1
unequal s1,2 = −σ ± ωn ζ2 − 1. For a unit-step input, R s = and
s

C(s) can be written

ωn 2
C s = s(s2 +2ζω 2 proof after taking inverse Laplace
n s+ωn )

transform it gives,

ωn 𝐞−𝐬𝟏 𝐭 𝐞−𝐬𝟐 𝐭
𝐜 𝐭 = 𝟏+ ( − )
𝟐 ζ2 −1 𝐬𝟏 𝐬𝟐

Where s1 = ωn ζ + ζ2 − 1 and s2 = ωn (ζ − ζ2 − 1)
By Zewde T. 32
Cont. …

By Zewde T. 33
 Cont. …

 For a desirable transient response of a second order system, the

damping ratio must be between 0.4 and 0.8. The Small value of ζ

(ζ < 0.4) yield excessive overshoot in the transient response, and a

system with large value of ζ ( ζ > 0.8) responds sluggishly.

By Zewde T. 34
 Cont. …
4.2.2.2. Transient response specification parameters of
under-damped Second-Order Systems
i. Rise time (𝐭𝐫 ): Based up on the time required for the response to rise
from 0% to 100%, the rise time can be obtained from

c t = 100%r(t), where r t = u t = 1 for t ≥ 0

−ζωn t ζ
c t =1−e cosωd t + sinωdt for t ≥ 0
(1−ζ2 )
Then at t r , c t = 1
→c t =1
−ζωn 𝑡𝑟 ζ
→1−e cosωd 𝑡𝑟 + sinωd𝑡𝑟 = 1
1−ζ2

−ζωn 𝑡𝑟 ζ
→e cosωd 𝑡𝑟 + sinωd 𝑡𝑟 = 0
1−ζ2
By Zewde T. 35
 Cont. …
Since e−ζωn tr ≠ 0, the above equation can be simplified to
ζ
→ cosωd t r + sinωd t r = 0
1−ζ2

1−ζ2
→ sinωd t r = − cosωd t r
ζ

1−ζ2
→ tanωd t r = − but ωd = ωn 1 − ζ2 and σ = ζωn
ζ

ωn 1−ζ2 ωd
→ tanωd t r = − ζωn
=− σ
ωd
→ ωd t r = tan−1 (− )
σ
1 ωd
→ tr = tan−1(− )
ωd σ
π−β
→ tr =
ωd
For the smallest value of t r , ωd must be large.
By Zewde T. 36
 Cont. …
ii. Peak time (𝐭𝐩 ): it can be obtained by differentiating c(t) with respect
to time and letting this derivate equal to zero.

Exercise 4.2: Proof that the peak time corresponds to the first peak

overshoot for second order underdamped system is given by

π
tp =
ωd

By Zewde T. 37
 Cont. …
iii. Maximum overshoot (𝐌𝐩): the maximum overshoot occurs at the
peak time. Assuming that the final value of the response is unity, Mp
can be obtained from
c(tp )−c(∞)
Mp = but c ∞ = 1
C(∞)

c(tp )−1 𝜋
→ Mp = = c(t p ) − 1 where 𝑡𝑝 =
1 𝜔𝑑

−ζωn ∗𝜋 𝜔𝑑 ζ
→ Mp = 1 − e cosωd ∗ 𝜋 𝜔𝑑 + sinωd ∗ 𝜋 𝜔𝑑 −1
1−ζ2

−ζωn ∗𝜋 𝜔𝑑 ζ −ζωn ∗𝜋 𝜔𝑑
→ Mp = −e cosπ + sinπ = −e (−1 + 0)
1−ζ2

𝜎
∗𝜋 −( )𝜋
→ Mp = e−ζωn 𝜔𝑑
=𝑒 ωd

By Zewde T. 38
 Cont. …
 The maximum percent overshoot is
𝜎
−( )𝜋
Mp = 𝑒 ωd ∗ 100%
iv. Settling time (𝐭𝐬 ): for an underdamped second order system, the
transient responses are obtained from.

−ζωn t
ζ
c t =1−e cosωd t + sinωdt
(1 − ζ2 )
 The settling time corresponding to a 2% or 5% tolerance may be
measured from the determinant factors that can lead the response to
the steady state value. The determinant factors of c(t) for settling are
1 − e−ζωn t
By Zewde T. 39
 Cont. …

For 2% tolerance For 5% tolerance

𝟏 − 𝒆−𝜻𝝎𝒏 𝒕𝒔 = 𝟎. 𝟗𝟖 𝟏 − 𝒆−𝜻𝝎𝒏𝒕𝒔 = 𝟎. 𝟗𝟓

𝒆−𝜻𝝎𝒏 𝒕𝒔 = 𝟎. 𝟎𝟐 𝒆−𝜻𝝎𝒏 𝒕𝒔 = 𝟎. 𝟎𝟓

𝟒 𝟑
𝒕𝒔 = 𝜻𝝎 𝒕𝒔 = 𝜻𝝎
𝒏 𝒏

By Zewde T. 40
 Cont. …
Exercise 4.3
1) Find the time response of the following system for unit step input
C(s) 25 C(s) 20
a) = b) =
R(s) s2 +7s+25 R(s) s2 +7s+25
C(s) 50
2) For a system having = , find its time response
R(s) s2 +7s+25
specifications and the expression for the output for unit step input.
𝑑 2 𝑦(𝑡) 5𝑑𝑦(𝑡)
3) For a system given by + + 16𝑦 𝑡 = 9𝑟(𝑡), where y(t)
𝑑𝑡 2 𝑑𝑡
is the output and r(t) is the input. Determine,
a) Time response specifications.
b) Time response of the system for unit step input.
𝑘
4) A system has 𝐺 𝑠 = with unity feedback when k and T are
𝑠(𝑇𝑠+1)
constant. Determine the factor by which k should be multiplied to
reduce the overshoot from 85% to 35%.
By Zewde T. 41
 4.3. Stability of control systems
4.3.1. Definition of stability
 Stability implies that small change in system input or small change in
system initial condition or small change in system parameter do not
produce large change in system output.

4.3.2. Analysis of stability

4.3.2.1. By using the location of the pole in the complex plane
 The stability of a linear closed-loop system can be determined from
the location of the closed-loop poles in the s plane. If any of these
poles lie in the right-half s plane, the system is unstable.
By Zewde T. 42
 Cont. …
 If any poles lie in the right-half s plane, then with increasing time
they give rise to the dominant mode, and the transient response
increases monotonically or oscillates with increasing amplitude. For
such a system, as soon as the power is turned on, the output may
increase with time.

 If no saturation takes place in the system and no mechanical stop is

provided, then the system may eventually be subjected to damage
and fail since the response of a real physical system cannot increase
indefinitely.

 Therefore, closed-loop poles in the right-half s plane are not

permissible in the usual linearBy control
Zewde T.
system. 43
 Cont. …
 If all closed-loop poles lie to the left-half s plane (𝑗𝜔 axis), the system
is stable. For such a system the transient response eventually reaches
equilibrium.
 Whether a linear system is stable or unstable is a property of the
system itself and does not depend on the input or driving function of
the system. The poles of the input, or driving function, do not affect
the property of stability of the system, but they contribute only to
steady-state response terms in the solution.
 Thus, the problem of absolute stability can be solved readily by
choosing no closed-loop poles in the right-half s plane, including the
𝑗𝜔 axis (closed-loop poles on the 𝑗𝜔 axis will yield oscillations, the
amplitude of which is neither decaying
By Zewde T. nor growing with time). 44
 Cont. …
 Note: All closed-loop poles lie in the left-half s plane does not
guarantee satisfactory transient-response characteristics. If dominant
complex-conjugate closed-loop poles lie close to the 𝑗𝜔 axis, the
transient response may exhibit excessive oscillations or may be very
slow.

 Therefore, to guarantee fast, yet well-damped, transient response

characteristics, it is necessary that the closed-loop poles of the system
lie in a particular region in the complex plane, such as the region
bounded by the shaded area shown in fig.4.9.

By Zewde T. 45
 Cont. …

Fig.4.9: Region in the complex plane satisfying the conditions ζ > 0.4
and 𝑡𝑠 < 4 𝜎.
By Zewde T. 46
 Cont. …
4.3.2.2. By using Routh’s stability criterion
 The most important problem in linear control systems concerns
stability. Most linear closed-loop systems have closed-loop transfer
functions of the form

𝐶(𝑠) 𝑏0 𝑠 𝑚 + 𝑏1 𝑠 𝑚−1 + ⋯ + 𝑏𝑚−1 𝑠 + 𝑏𝑚

=
𝑅(𝑠) 𝑎0 𝑠 𝑛 + 𝑎1 𝑠 𝑛−1 + ⋯ + 𝑎𝑛−1 𝑠 + 𝑎𝑛

where the a's and b's are constants and 𝑚 ≤ 𝑛.

 Routh's stability criterion, enables us to determine the number of

closed-loop poles that lie in the right-half s plane without having to
factor the denominator polynomial. This stability criterion applies to
polynomials with only a finite number
By Zewde T. of terms. 47
 Cont. …
 The procedure in Routh's stability criterion is as follows:
1) Write the polynomial in s in the following form:
𝑎0 𝑠 𝑛 + 𝑎1 𝑠 𝑛−1 + ⋯ + 𝑎𝑛−1𝑠 + 𝑎𝑛 = 0
where the coefficients are real quantities. We assume that +𝑎𝑛 ≠ 0;
that is, any zero root has been removed.

2) The necessary but not sufficient condition for stability is that all the
coefficients of the denominator of transfer function should present
and all have a positive sign. (If all a's are negative, they can be made
positive by multiplying both sides of the equation by -1.)

 If any of the coefficients are zero or negative in the presence of at

least one positive coefficient, there is a root or roots that are imaginary
or that have positive real parts. ByIn this
Zewde T. case, the system is not stable.
48
 Cont. …
3) If all coefficients are positive, arrange the coefficients of the
polynomial in rows and columns according to the following pattern:

 The process of forming rows continues until we run out of elements.

(The total number of rows is 𝑛 + 1.) The coefficients 𝑏1 , 𝑏2 , 𝑏3 , and
so on, are evaluated as follows until the remaining ones are all zero.
By Zewde T. 49
 Cont. …

 The same pattern of cross-multiplying the coefficients of the two

previous rows is followed in evaluating the c's, d's, e's, and so on. That
is,
By Zewde T. 50
 Cont. …

And

 This process is continued until the nth row has been completed. The
complete array of coefficients is triangular. Routh's stability criterion
states

 The number of roots of the denominator of transfer function with

positive real parts is equal to the number of changes in sign of the
coefficients of the first column ByofZewde
theT. array. 51
 Cont. …
 The necessary and sufficient condition that all roots of the denominator of
transfer function lie in the left-half s plane is that all the coefficients of the
denominator should be positive and all terms in the first column of the array
have positive signs.

Example 4.2: Investigate the stability of a system with characteristics

polynomials

a) s4 + 2s3 + 3s2 + 4s + 5 = 0
b) P s = s4 + s3 + s2 + 2s + 4

c) P s = s3 + s2 + 2s + 1

By Zewde T. 52
 Cont. …
Solution: a) All the coefficients are positive numbers.
Forming the array
2∗3−1∗4 2∗5−1∗0
b1 = = 1, b2 = =5
2 2
𝑏1 ∗4−2∗b2 1∗4−2∗5
c1 = 𝑏1
= 1
= −6
𝑐1 ∗b2 −b1 ∗0 −6∗5−1∗0
d1 = 𝑐1
= −6
=5

By Zewde T. 53
 Cont. …
Special Cases
 There are two cases in which the Routh criterion, as it has been
presented above can not be applied. For these two cases certain
modifications are necessary so that the above procedure is applicable.

Case1: A zero element in the first column of the Routh array

 In this case the Routh array cannot be completed because the element
below the zero elements in the first column will become infinite. To
overcome this difficult,
• Method1: If a first-column term in any row is zero, but the
remaining terms are not zero or there is no remaining term, then the
zero term is replaced by a very small positive number 𝜖 and the rest
of the array is evaluated. By Zewde T. 54
 Cont. …
• Method2: we multiply the characteristics polynomial p(s) with a
factor (𝑠 + 𝑎), where 𝑎 > 0 and −𝑎 is not the root of p(s). The
conclusions regarding stability of the new polynomial 𝑝(s) =
s + a p(s) are obviously the same as those of the original
polynomial p(s).

Example 4.3: consider the following equation

𝑝 𝑠 = 𝑠 4 + 2𝑠 3 + 𝑠 2 + 2𝑠 + 4

Construct the Routh array as,

By Zewde T. 55
 Cont. …
Method1: Replace the zero term by a very small positive number 𝜖 and
then evaluate the rest of the array.

There are two sign changes of the

coefficients in the first column. It show
that p(s) has two roots in the right half
complex plane and therefore the system
is unstable.
Method2: If we multiply the polynomial p(s) by the factor (𝑠 + 1),

we have,

p s = s + 1 p s = s 5 + 3s 4 + 3s 3 + 3s 2 + 6s + 4

Next construct the Routh array

By Zewde T. 56
 Cont. …
According to the above Routh array,
one observes that the polynomial p(s)
and p(s) have two roots in the right half
complex plane and therefore the system
with characteristics polynomial P(s) is
unstable

Case2: A zero row in the Routh array

 In this case the Routh array cannot be completed, because in
computing the rest of the elements that follow zero rows,
indeterminate form 0 0 will appear. To overcome this difficulty:

By Zewde T. 57
 Cont. …
1) Form the auxiliary polynomial q(s) of the row which precedes the
zero row.

2) Take the derivative of q(s) with respect to s and replace the zero row
with the coefficients of q s .

3) Complete the construction of the Routh array in the usual manner.

Example 4.4: investigate the stability of a system with characteristic

polynomial

𝑝 𝑠 = 𝑠 4 + 3𝑠 3 + 4𝑠 2 + 3𝑠 + 3

By Zewde T. 58
 Cont. …
Solution: construct the Routh array as follow,
Since the row 𝑠 1 of the Routh array involves
only zeros, it is clear that the Routh array
cannot be completed. At this point form
auxiliary polynomial 𝑞 𝑠 = 3𝑠 2 + 3 of the
row 𝑠 2. Taking the derivative of q(s),
𝑞 𝑠 = 6𝑠. The replace the zero row (i.e. 𝑠 1)
with the coefficient of 𝑞 𝑠 and complete the
Routh array in the usual manner to yield

Since the first column of the new Routh array

doesn’t have sign change the system is stable.
By Zewde T. 59
 Cont. …
 Relative Stability Analysis: Routh's stability criterion provides the
answer to the question of absolute stability. This, in many practical
cases is not sufficient. We usually require information about the
relative stability of the system.

• Absolute stability: indicates stability or unstability of a system under

consideration.

• Relative stability: indicates how close a system is to instability.

 Routh's stability criterion is of limited usefulness in linear control

system analysis mainly because it does not suggest how to improve
relative stability or how to stabilize an unstable system.
By Zewde T. 60
 Cont. …
 It is possible, however, to determine the effects of changing one or
two parameters of a system by examining the values that cause
instability.
Example4.5: Let us determine the range of k such that the system with
characteristic polynomial p s = s 4 + s 3 + 2s 2 + s + 2k is
stable.
Solution:
For the system to be stable all elements of
the first column of Routh array must have
the same sign. Hence there must be
2𝑘 > 0 and 1 − 2𝑘 > 0
𝑘 > 0 and 𝑘 < 0.5
Hence the system is stable for 0 < 𝑘 < 0.5
By Zewde T. 61
 Cont. …
Exercise 4.4:
1) Determine the stability of a system with closed loop transfer
function:

1
a) G s = 2s5 +3s4 +2s3 +3s2 +2s+1

10
b) G s =
s5 +2s4 +3s3 +6s2 +5s+3

1
c) G s = 2s5 +7s4 +6s3 +42s2 +8s+56

1
d) G s = s3+5s2 +6s+30

By Zewde T. 62
 Cont. …
10
2) For a unity feedback system with 𝐺 𝑠 = determine the
𝑠(𝑠+1)(𝑠+2)
stability using Routh criterion.

𝑠+6
3) For a unity feedback system given below with 𝐺 𝑠 =
𝑠(𝑠+1)(𝑠+3)
determine the range of k that makes the system stable.

By Zewde T. 63
 Cont. …
5) Investigate the stability of the given closed loop control system by
using Routh criterion.

By Zewde T. 64
4.4. Effect of derivative and integral control actions
on system performance
 Consider the system shown in the figure below:

where
𝐺𝑐 𝑠 −Tf of the controller
𝐺𝑐 𝑠 −Tf of the plant
 The transfer function from E(s) into R(s) can be obtained from
E s = R s − C(s) but C s = Gc S G s E(s)
→ E s = R s − Gc S G s E(s)

1
→E s = R(s) By Zewde T. 65
1+Gc S G s
 Cont. …
 For a unit step reference input, the error of the system will becomes
1 1
E s = ∗
1 + Gc S G s 𝑠

1
 Let us consider the transfer function of the plant be 𝐺 𝑠 = , then
𝑇𝑠+1

i. The steady state error of the system, when the proportional control
action is added into the system is:

k
ess t = lim sE(s) where Gc s = k, Gc s G s = Ts+1
s→0

1 1 Ts+1 1
ess t = lim s ∗ k ∗ = lim =
s→0 1+ s s→0 Ts+k+1 k+1
Ts+1

By Zewde T. 66
 Cont. …
 Such a system with a proportional controller in the feed forward path
always has a steady state error in the step response. This steady state
error is called an offset. Such an error can be eliminated by including
an integral control action on the system.

Fig4.10: unit step response and offset.

By Zewde T. 67
 Cont. …
ii. The steady state error of the system when an integral control action
is added into the system is:

k k
ess t = lim sE(s) where Gc s = , Gc s G s =
s→0 s s(Ts+1)

1 1
ess t = lim s ∗ k ∗ =0
s→0 (1+ s
s(Ts+1)

 Note that integral control action, while removing offset or steady-state

error, may lead to oscillatory response of slowly decreasing amplitude
or even increasing amplitude, both of which are usually undesirable.

By Zewde T. 68
4.5. Steady state errors in unity feedback control systems
 Consider the unity-feedback control system with the following open-
loop transfer function G(s):

k Ta s + 1 Tb s + 1 … (Tm s + 1)
G s = N
s T1 s + 1 T2 s + 1 … (Tp s + 1)
 It involves the term s N in the denominator, representing a pole of
multiplicity N at the origin. The present classification scheme is based
on the number of integrations indicated by the open-loop transfer
function. A system is called type 0, type 1, type 2, . . . ,if N = 0, N = 1,
N = 2, . . . ,respectively.
 Note: As the type number is increased, accuracy is improved;
however, increasing the type number
By Zewde T.
aggravates the stability problem
69
 Cont. …
Steady-State Errors:
 consider the system shown in figure below

 The closed-loop transfer function is

C(s) G(s)
=
R(s) 1 + G(s)

 The transfer function between the error signal e(t) and the input signal
r(t) is
By Zewde T. 70
 Cont. …
E s = R s − C(s)

E(s) C s G s 1 1
→ R(s) = 1 − R s
= 1 − 1+G s
= 1+G(s) ∴ E s = 1+G(s) ∗ R(s)

 By using final value theorem, the steady state error is

sR(s)
ess = lim e(t) = lim SE(s) = lim
t→∞ s→0 s→0 1 + G(s)

i. The steady-state error of the system for a unit-step input is

s 1 1
ess = lim SE(s) = lim ∗ =
s→0 s→0 1+G(s) s 1+G(0)

 The static position error constant k p , is defined by

k p = lim G(s) = G(0) By Zewde T. 71

s→0
 Cont. …
 Thus, the steady-state error in terms of the static position error
constant k p , is given by

1
𝑒𝑠𝑠 =
1 + 𝑘𝑝
 For a type 0 system,

k Ta s + 1 Tb s + 1
k p = lim =k
s→0 T1 s + 1 T2 s + 1 …

 For a type 1 or higher system,

k Tas+1 Tb s+1
kp = lim N = ∞ for N ≥ 1
s→0 s T1 s+1 T2 s+1 …
By Zewde T. 72
 Cont. …
 For a unit-step input, the steady-state error 𝑒𝑠𝑠 , may be summarized as
follows:
1
ess = , for type 0 systems
1+k
ess = 0, for type 1 or higher systems

ii. The steady-state error of the system with a unit-ramp input is

given by

s 1 1 1
ess = lim ∗ 2 = lim = lim
s→0 1 + G(s) s s→0 s + sG(s) s→0 sG(s)

 The static velocity error constant k v is defined by

k v = lim sG(s)
Bys→0
Zewde T. 73
 Cont. …
 Thus, the steady-state error in terms of the static velocity error constant kv ,
is given by

1
ess =
kv
 For a type 0 system
sk Ta s+1 Tbs+1
kv = lim =0
s→0 T1 s+1 T2 s+1 …

 For a type 1 system

sk Ta s + 1 Tb s + 1
kv = lim =k
s→0 𝑠 T1 s + 1 T2 s + 1 …

 For type 2 or higher system

sk Ta s+1 Tbs+1
kv = lim = ∞, for 𝑁 ≥ 2
s→0 𝑠𝑁 T1 s+1By T 2 s+1
Zewde T. … 74
 Cont. …
 The steady-state error ess , for the unit-ramp input can be summarized
as follows:

ess = ∞, for type 0 system

1
ess = k, for type 1 system

ess = 0, for type 2 and higher system

iii. The steady-state error of the system with a unit-parabolic input

(acceleration input), is defined by

s 1 1 1
ess = lim ∗ = lim = lim
s→0 1+G(s) s3 s→0 𝑠2 +𝑠2 G(s) s→0 𝑠2 G(s)

By Zewde T. 75
 Cont. …
 The static acceleration error constant k a , is defined by
k a = lim s 2 G(s)
s→0

 The steady-state error interms of k a is then

1
ess =
ka
 For a type 0 system,

s 2 k Ta s + 1 Tb s + 1
k a = lim =0
s→0 T1 s + 1 T2 s + 1 …

 For a type 1 system,

s 2 k Ta s + 1 Tb s + 1
k a = lim =0
s→0 𝑠 T1 s + 1 T2 s + 1 …
By Zewde T. 76
 Cont. …
 For a type 2 system,

s 2k Ta s + 1 Tbs + 1
k a = lim 2 =k
s→0 𝑠 T1 s + 1 T2 s + 1 …

 For a type 3 or higher system,

s2 k Tas+1 Tb s+1
ka = lim 𝑁 = ∞, for 𝑁 ≥ 3
s→0 𝑠 T1 s+1 T2 s+1 …

 Thus, the steady-state error for the unit parabolic input is

ess = ∞, for type 0 and 1 systems

1
ess = , for type 2 systems
𝑘

ess = 0, for type 3 and higher systems

By Zewde T. 77
 Cont. …

Table 4.1: Summary (Steady-State Error in Terms of Gain k)

By Zewde T. 78
 Cont. …
Exercise 4.5
1) Identify the type of system given below
k(1+3s) 5s+1
a) G s = and H s =
s2 s2 +6s+7

k(1+3s)
b) G s = and H s =1
𝑠

4
c) G s = and H s =s+3
𝑠2 +6𝑠+7

𝑘(𝑠+4)
2) For a system having 𝐺 𝑠 𝐻 𝑠 = , find
𝑠(𝑠3 +5𝑠2 +6𝑠)

i. Type of the system

ii. Error coefficient
iii. Error due to parabolic input 𝑠By2Zewde T. 79
 Cont. …

3) Find the error coefficient for the system given by block diagram.

By Zewde T. 80
4.6. Feedback Characteristics of Control Systems
 Feedback is used in control systems to reduce error, the sensitivity of
the system due to parameter variations and unwanted internal and
external disturbances.
4.6.1. Effect of feedback on overall gain

 Consider an open-loop and closed loop control system shown below.

By Zewde T. 81
 Cont. …
 Due to introduction of negative feedback, the gain 𝐺 𝑠 is reduced by
1
the factor of 1+𝐺 𝑠 𝐻(𝑠)
. Therefore, due to negative feedback the
overall gain reduces.
4.6.2. Effect of feedback on stability
 Consider open-loop and a unity negative feedback control system
shown below.

By Zewde T. 82
 Cont. …
 Due to introduction of feedback, the time constant decrease and the
transient response decays more rapidly.

 Since the stability of a system depends upon the location of poles in

the s-plane, feedback may improve the stability of the system.

 Open-loop pole is stable; the closed-loop pole may be unstable.

 To control the stability of the system, a proper design and application

of the feedback is required.

4.6.3. Effect of feedback on system sensitivity

 The parameters of a control system change due to environmental

changes and other disturbances. This change has adverse affect on the
system performance. By Zewde T. 83
 Cont. …
 If the variable in a system be T due to the variation in parameter K of
the system, the sensitivity of the system parameter T to the parameter
K is given by
% 𝑐ℎ𝑎𝑛𝑔𝑒 𝑖𝑛 𝑇
𝑆=
% 𝑐ℎ𝑎𝑛𝑔𝑒 𝑖𝑛 𝐾

Mathematically, it can be written as

𝑇 𝜕𝑇 𝐾 𝜕𝑇
= 𝑇 =
𝐾
𝜕𝐾 𝑇 𝜕𝐾
𝐾
 Let T(s) be the overall transfer function of a control system and G(s)
is its forward path gain. The sensitivity of overall transfer function
T(s) with respect to the variations in G(s) is given by
By Zewde T. 84
 Cont. …

 It describes relative variation of overall transfer function due to

variation of G(s).

By Zewde T. 85
 Cont. …
1
 The sensitivity function gets reduced by a factor compared to
1+𝐻 𝑠 𝐺(𝑠)

open-loop sensitivity function due to the presence of negative feedback. In

𝑇
closed loop systems, 𝐺
is less sensitive to the variation of G(s).

4.6.4. Effect of feedback on steady state error

 The error signal is given by
e t = r t − y t or E s = R s − Y(s)

Where, r(t) is the reference input and y(t) is the actual output.

 The steady state error is given by

ess = lim 𝑒(𝑡) = lim 𝑠𝐸(𝑠) or 𝑒𝑠𝑠 = 𝑟 𝑡 − 𝑦𝑠𝑠 (𝑡)

𝑡→∞ 𝑠→0

where 𝑦𝑠𝑠 𝑡 - steady state output.

By Zewde T. 86
 Cont. …

 Since 𝐺 0 may be large value, the steady state error of open-loop

system will be large.
 In the case of negative unity feedback the steady state error is reduced
1
to . By Zewde T. 87
1+𝐺(0)
 Cont. …
4.6.5. Effect of feedback on disturbance
 Consider an open-loop and closed loop control system shown below.

 Therefore, in the case of open loop the effect of feedback is large but
in the case of closed loop by changing the value of G(s), the effect of
disturbance on the output can be reduced.
By Zewde T. 88
 Cont. …
 Note: Here in a negative unity feedback system the disturbance is
1
reduced by the factor of when compared to open-loop system.
1+𝐺(𝑠)

But based on the position of disturbance applied to a system and type

of feedback the reduction factor is not the same.

Exercise 4.2

1) Consider a feedback system given below.

By Zewde T. 89
 Cont. …
2) Find the sensitivity of over all transfer function of a system given
below with respect to ( 𝑠 = 𝑗𝜔, 𝜔 = 1.2 𝑟𝑎𝑑 𝑠𝑒𝑐).

T
a) Forward path transfer function. ( Ans. G
= 0.682)

T
b) Feedback path transfer function. ( Ans. H
= 0.953)

c) Determine the steady state system output and steady state error for

unit step input.

By Zewde T. 90
 Cont. …
3) Find the following if R = 10k and r = 8k.
a) The value of k for 4% system sensitivity due to variation of μ,
H = 0.3 and μ = 12.
b) The value of k for 3% system sensitivity due to variation of H,
μ = 18 and H = 0.25.

By Zewde T. 91
From your Text book refer the following:
 Root locus analysis
 Introduction.
 Root locus plots for negative feedback systems.
 Root locus plot of positive feedback systems.
 Conditionally stable systems.
 Frequency response Analysis
 Introduction.
 Presenting frequency-response characteristics in graphical forms.
 Polar plot.
 Nyquist plot.
 Bode plot.
 Nichols plot (Log-magnitude Versus phase plot).
By Zewde T. 92
 Nyquist stability criterion.
 Cont. …

 Control Systems Design by Root locus and frequency response

method

 Design considerations, Lead compensation, Lag compensation, Lead-

lag compensation, parallel Compensation.

 Introduction, Lead compensation, Lag compensation, Lead-lag

compensation.

By Zewde T. 93