TEL 306/05

Control Systems

Unit 2
System Response
Analysis
Contents
Unit overview 1

Unit outcomes 3

2.1 Inputs and disturbances 4

Introduction 4

Input force functions 5

Disturbances 8

Examples of system representation with disturbances 11

Suggested answers to activities 18

2.2 Time response analysis 21

Introduction 21

First order time response 22

Second order time response 25

Transient response analysis 38

Steady state response analysis 44

Pole placement and system response 58

Example of system design via pole placement 62

Suggested answers to activities 70

Summary of Unit 2 75

References 77
Unit Overview
It was shown in Unit 1 that the behaviour of a process can be represented
by a mathematical model that is developed by means of theoretical or
empirical approach. How do we know that the process model is valid?
This can be done by comparing the model solution with the experimental
result for a given input force function. In this unit, four forms of input force
functions that are commonly used to test the system time response will
be introduced, namely impulse, step, ramp and parabolic inputs. These
inputs represent four different scales of changes of the force function for
a system to follow. The relationship between the input and the system
can be analogised as the teaching and learning process in a classroom.
Assuming an average student (system) sitting in the class, if the teacher
teaches slowly, which resembles a step input, the student will learn very
well. As the teacher speeds up the lesson, which resembles a ramp
input, the student might find it harder to follow and may absorb only
part of the knowledge. If the teaching goes speedy, which resembles a
parabolic input, then the student may not be able to cope. He/she may
end up learning nothing and may eventually lose interest in the study!

The response of the system to an input force function is called a time

response, which consists of a transient response and a steady state
response as depicted in Figure 2.1. The transient response exhibits
the dynamics of the system that varies with respect to time, while the
steady state response happens when the response settles down at the
time approaching infinity. The time response varies for different order of
the transfer functions. The unit mainly focuses on the first and second
order transfer functions as these two are the fundamental components
constituting a higher order transfer function. It should be remarked that
all the systems discussed in this course are assumed to be linear.

UNIT 2 System response analysis 1

 Figure 2.1 Transient and steady state responses

Real life is always full of unforeseen interrupts. For example, a phone

ringing in the class will distract attention and affect the learning process.
Interrupts can be thought of as undesired input force functions, which are
denoted as disturbances in this unit. Disturbances could also be in the
form of impulse, step, ramp and parabolic functions, similar to the input
force functions described previously. It may occur at different parts of a
system and affect the time response. Comprehending the time response
of a system which includes disturbances may aid in designing a more
realistic closed loop control system, hence obtaining a better result.

2 TEL 306/05 Control Systems

Unit Outcomes
By the end of Unit 2, you should be able to:

1. Demonstrate the effect of various inputs and disturbances

to the systems.

2. Describe the time response of first order systems.

3. Evaluate the performance of second order systems.

4. Explain the relationship of pole placement and system

time response.

5. Analyse steady state errors of the systems.

UNIT 2 System response analysis 3

 2.1 Inputs and Disturbances
Introduction
In validating a model as shown in Figure 2.2, the same input signal at a
bounded range is perturbed to both model and the actual process. The
calculated model solution is then compared with the measured output.
The model is considered valid if the discrepancies between the model
solution and the actual output are within an acceptable tolerance. It
should be noted that the model is not just limited to the process, but
also applies to a complete system. The model of the system could be
derived using the block diagram reduction technique or Mason’s rule
on the signal flow graph as discussed in Unit 1. A wide range of input
signals are available for the model validation. The choice of a suitable
input signal is very much depending on the nature of the process or the
system, which will be studied in the following section.

Input signal Model solution

Model

Actual output
Process/System

Figure 2.2 Model validation

4 TEL 306/05 Control Systems

Input force functions
In general, the input signal to a process or a system can have an infinite
variety of forms. In this section, four common forms of input signal will
be described, namely an impulse, a step, a ramp and a parabolic. The
graphs and the input function in both time and Laplace domain are given
in Table 2.1, where the input signal is denoted as i(t) and (t) represents
a short impulse.

The impulse input is a large amplitude of signal applied within a small

fraction of time as illustrated in the first row of the table. An example
of the impulse response could be seen in a soccer match. To pass the
ball to the upper field, the goal keeper kicks the ball from the goal area
box. The ball flies away and drops at a new spot at the upper field. The
kick is the impulse input that causes the motion of the ball. In some of
the engineering applications, the impulse input might not be advisable
as the amplitude of the impulse input should be relatively high for the
system to response. The high amplitude, however, might have already
caused damage to the system.

Perhaps the most useful input force function is the step input which is at
the second row of the table. It represents an abrupt change in the setpoint
such as an ON-OFF state of a switch. Comprehending the response
of the system to the step input change is of considerable interest. For
instance, how long is the time needed to reach the set temperature
from the current temperature of 30°C when the air conditioner is being
switched on? By observing the output response (temperature profile), we
may justify the suitability of the air conditioner for a given size of room
and the local weather conditions.

Another form of input force function is the ramp input as depicted in

the third row of the table. The ramp input refers to the rate of change of
input signal and causes a less abrupt change as compared with the step
input. It is normally used when a step input is not possible or unsuitable
such as evaluating the velocity performance of the system. The last form
of input at the last row of the table is the parabolic input. It represents

UNIT 2 System response analysis 5

 the acceleration of an input signal in which its magnitude increases or
decreases exponentially along with time. The time and Laplace domain
functions for each form of inputs are expressed in the second and third
columns of the table. Given a transfer function, the response of the
system can be obtained by simply applying the respective input function.
For example, given the process transfer function is:

Y(S) 1
= 2
R(S) s + 3s + 2

1
⇒ Y(s) = R(s)
s2 + 3s + 2

1 1 1
For a step input, R(s) = , ⇒ Y(s) = 2 •
s (s + 3s + 2) s

The time response of the process is then calculated by inverse Laplace

of the above solution.

Activity 2.1
1
Given a process transfer function of G(s) = , compare the
ts + 1
time response when subjected to a unit impulse and a unit step
input.

6 TEL 306/05 Control Systems

 Input signal graphs Time domain Laplace domain
function function
Impulse i(t) = (t) i(s) = 1

1
Step i(t) = 1 i(s) =
s

1
Ramp i(t) = t i(s) =
s2

2
Parabolic i(t) = t 2 i(s) =
s3

Table 2.1 Common input signals for the model validation

UNIT 2 System response analysis 7

 Disturbances
Disturbances are defined as the unwanted force function that are
unavoidable in a practical system. These can be such things as changes
in the line voltage on the power supply to the house, or wind loads on the
arm of a crane. Disturbances may occur everywhere within a system and
affect the system output. Taking a position control of an antenna as an
example, the possible disturbances are shown in Figure 2.3 (Nise, 2011).

D1(s) D2(s) D3(s) D4(s)

i (s) O(s)
G1(s) G2(s) G2(s)

H2(s)

D5(s)

Figure 2.3 Disturbances in the position control of an antenna

The possible disturbances as shown in Figure 2.3 are:

θi(s) = Desired position of the antenna

H1(s) = Transfer function of the potentiometer as a signal conversion
D1(s) = Signal distortion such as electromagnetic interference
G1(s) = Transfer function of the controller
D2(s) = Distortion to the control action
G2(s) = Transfer function of the motor
D3(s) = Unexpected load to the motor such as wind blow
G3(s) = Transfer function of the antenna carriage
D4(s) = Distortion to the position of the antenna such as a bird standing
on the antenna
θo(s) = Actual position of the antenna
H2(s) = Transfer function of the potentiometer as a signal conversion
D5(s) = Distortion to the measured signal such as temperature variation

8 TEL 306/05 Control Systems

An important contribution of a closed loop control system is to minimise
the effect of disturbances on the response of the system. For an open
loop system depicted in Figure 2.4, given an input signal, R(s), applied
to the process, G(s) through an appropriate amplification ratio, K, the
output, Y(s) can be derived as:

Y(s) = KG(s)R(s) + D(s)

where D(s) is the disturbances.

It is obvious from the equation that the output, Y(s), is affected by the
disturbances, D(s) and no remedy could be proposed to reduce the effect
of disturbances. Meanwhile, for a closed loop system as illustrated in
Figure 2.5, the output, Y(s) can be derived as:

Y(s) = E(s)KG(s) + D(s)

Since the error signal, E(s), is the discrepancies between the input signal,
R(s), and the actual output, Y(s), hence

E(s) = R(s) − Y(s)

Substitute into the output signal equation,

Y(s) = [R(s) − Y(s)]KG(s) + D(s)

Y(s) = R(s)KG(s) − Y(s)KG(s) + D(s)

KG(s) 1
Y(s) = R(s) + D(s)
1 + KG(s) 1 + KG(s)

Observing the above equation, the effect of disturbances could be

minimised by increasing the amplification ratio, K.

UNIT 2 System response analysis 9

 D(s)

R(s) Y(s)
K Process, G(s)

Figure 2.4 An open loop system with disturbances

D(s)

R(s) E(s) Y(s)

K Process, G(s)

Figure 2.5 A closed loop system with disturbances

Activity 2.2

The learning experience in a classroom could be enjoyable or

annoying. It is mainly determined by the classroom’s physical
conditions, the teaching quality and the student quality. Based on
your learning experience, identify the input and the disturbances
in the learning process. Suggest solutions to minimise the effect
of disturbances if they could not be eliminated.

10 TEL 306/05 Control Systems

Examples of system representation with
disturbances
A complete system representation should include the identified
disturbances into the loop. It could be in the form of block diagrams or
signal flow graphs. Derivations of overall system transfer function which
include disturbances will be demonstrated in this section on both block
diagrams and signal flow graphs.

Example 1

A system can be represented using the block diagram as shown in Figure

2.6.

1. Derive the overall transfer function for the system with the
disturbances.

2. Construct a signal flow graph for the system and derive the
system transfer function using Mason’s rule.

3. Find the output, Y(s), of the system given that the disturbances,
D(s), is in the form of step function and the input signal, R(s),
is null. Comment on the output obtained.

D(s)
R(s) E(s) Y(s)
K1 Process, G(s)

K2

Figure 2.6 Feedback control with disturbances.

UNIT 2 System response analysis 11

 Solution

1. Y(s) = E(s)K1G(s) + D(s)G(s)

where E(s) = R(s) − K2Y(s)
therefore
Y(s) = [R(s) − K2Y(s)]K1G(s) + D(s)G(s)
Y(s) = R(s)K1G(s) − K2K1G(s)Y(s) + D(s)G(s)
[1 + K2K1G(s)]Y(s) = R(s)K1G(s) + D(s)G(s)

K1G(s) G(s)
Y(s) = R(s) + D(s)
1 + K2K1G(s) 1 + K2K1G(s)

2. The equivalent signal flow graph for the block diagram is as

follows:

D(s)

1
R(s) 1 K1 G(s) 1 Y(s)

−K2

There are two forward paths. One path originated from input
signal, R(s), another path from disturbances, D(s).

P1 = K1G(s)
P2 = G(s)

There is one loop,

L1 = −K1G(s)K2

12 TEL 306/05 Control Systems

 There are no non-touching loops
∴∆ = 1 − (L1)

Both forward paths touch L1

∴∆1 = 1, ∆2 = 1

Overall transfer function of the system is:

P1∆1R(s) + P2∆2D(s) K1G(s)R(s) + G(s)D(s)

Y(s) = =
∆ 1 + K1K2G(s)

K1G(s) G(s)
Y(s) = R(s) + D(s)
1 + K1K2G(s) 1 + K1K2G(s)

1
3. Given D(s) = and R(s) = 0,
s

K1G(s) G(s) 1
Y(s) = •0+ •
1 + K1K2G(s) 1 + K1K2G(s) s

G(s) 1
Y(s) = •
1 + K1K2G(s) s

The transfer function above indicates that by increasing the

gains K1 and K2, the effect of disturbances could be minimised.

UNIT 2 System response analysis 13

 Example 2

An active suspension system is designed to provide a more comfortable

ride especially on the bumpy roads. Figure 2.7 shows the concept of the
active suspension system where a sensor is used to provide advanced
information to the system through transfer function, H(s), before the
disturbances, D(s), act on the vehicle dynamics, G(s). The system then
adjusts the controller gain, K, actively to reject the effect of disturbances.

1. Derive the overall transfer function for the system with

disturbances.

2. Given that there is a bump as the disturbance in the form of

step function, provide the conditions so that the vehicle does
not bounce to maintain comfortability of the ride.

Bump disturbances,
D(s)

H(s)

R(s) E(s) Y(s)

K Vehicle
dynamics, G(s)

Figure 2.7 An active suspension system

14 TEL 306/05 Control Systems

Solution

1. Derive the overall transfer function for the system with the
disturbances.

Y(s) = E(s)KG(s) + D(s)G(s)

where E(s) = R(s) − Y(s) − D(s)H(s)
therefore
Y(s) = [R(s) − Y(s) − D(s)H(s)]KG(s) + D(s)G(s)
Y(s) = R(s)KG(s) − KG(s)Y(s) − D(s)H(s)KG(s) + D(s)G(s)
[1 + KG(s)]Y(s) = R(s)KG(s) − D(s)H(s)KG(s) + D(s)G(s)
[1 + KG(s)]Y(s) = R(s)KG(s) + D(s)G(s)[1 − H(s)K]

KG(s) G(s)[1 − H(s)K]

Y(s) = R(s) + D(s)
1 + KG(s) 1 + KG(s)

2. Given that there is a bump as the disturbance in the form of

step function, provide the conditions so that the vehicle does
not bounce to maintain comfortability of the ride.

To make sure the vehicle does not bounce, it is desired that

R(s) = 0. Given a step function is imposed as the disturbance,
hence

KG(s) G(s)[1 − H(s)K] 1

Y(s) = •0+ •
1 + KG(s) 1 + KG(s) s

G(s)[1 − H(s)K] 1
Y(s) = •
1 + KG(s) s

Observing the above equation, if we set

1 − H(s)K = 0

UNIT 2 System response analysis 15

 Then the effect of disturbance could be completely eliminated.
Therefore, the conditions to ensure that the vehicle does not
bounce is

H(s)K = 1

Example 3

Figure 2.8 shows the block diagram for a water heating system in the hot
water dispenser. The desired water temperature, R(s), in the dispenser is
to maintain at 90°C but fresh water will be added to refill the quantity of
hot water being dispensed. The temperature of the fresh water normally is
around room temperature at 25°C and this has introduced disturbances,
D(s), to the system. Derive the overall system transfer function so that
the effect of disturbances could be studied. Suggest the way to reduce
the effect of disturbances so that the water can be maintained at 90°C?

D(s)

R(s) E(s) Y(s)

1
K
10s + 1

Figure 2.8 Block diagram of water heating system

Solution

The overall system transfer function including disturbances is:

1 1
Y(s) = K · E(s) + D(s)
10s + 1 10s + 1

16 TEL 306/05 Control Systems

Since E(s) = R(s)

1 1
Y(s) = K · [R(s) − Y(s)] + D(s)
10s + 1 10s + 1

K K 1
Y(s) = R(s) − Y(s) + D(s)
10s + 1 10s + 1 10s + 1

K K 1
[1 + ]Y(s) = R(s) + D(s)
10s + 1 10s + 1 10s + 1

[ 10s + 1 + K
10s + 1 ]
Y(s) =
K
10s + 1
R(s) +
1
10s + 1
D(s)

K 1
Y(s) = R(s) + D(s)
10s + 1 + K 10s + 1 + K

Based on the derived transfer function, the effect of disturbances can

be reduced by increasing the gain K.

Web Reference

For more information on the effect of noises and disturbances to

a control system, you may visit this website:

https://www.controleng.com/articles/noise-and-disturbances-in-
process-control-2/

UNIT 2 System response analysis 17

 Summary

This section introduced four forms of input force functions,

namely impulse, step, ramp and parabolic that are frequently
used as the input signals for the time domain model validation
purpose. The choice of input signal is very much depending on
the characteristics of the system. The force functions are also
applicable for disturbances that intercepted the complete system.
The derivation of the transfer function for a complete system with
disturbances was demonstrated with an example to improve your
understanding of the topic.

Suggested answers to activities

Feedback

Activity 2.1

When an impulse input is applied, R(s) = 1,

Y(s) 1
G(s) = =
R(s) ts + 1

1
Y(s) = • R(s)
ts + 1

1
t
Y(s) = •1
s+ 1
t

18 TEL 306/05 Control Systems

Inverse Laplace

1 −t/t
y(t) = e
t
1
When a step input is applied, R(s) = ,
s
Y(s) 1
G(s) = =
R(s) ts + 1

1
Y(s) = • R(s)
ts + 1

1
t 1
Y(s) = •
s+ 1 s
t

Apply partial fraction expansion and inverse Laplace transform,

y(t) = 1 − e−t/t

Observing the equations of both time responses, it can be

concluded that the response of the system will be attenuated
eventually to zero when subjected to the impulse input. When a
step input is applied, the system response will converge to a unity
steady state value.

UNIT 2 System response analysis 19

 Activity 2.2

The following are some of the examples, but not limited to, inputs
and disturbances for a learning process in a classroom.

Inputs: Teaching speed and level of difficulty of content.

Disturbances: Unstable room conditions including the light and

room temperature, chatting sounds, yawning sounds, door
opening and phone-ringing sounds.

Suggestion for minimising the effect of disturbances: Maintain a

suitable temperature and room conditions, conduct interactive
teaching and learning activities, setting phone to mute mode during
classes, and avoiding late attendance so that your classmates’
attention are not distracted when you open the door.

20 TEL 306/05 Control Systems

2.2 Time Response Analysis
Introduction
In section 2.2, various forms of force functions have been introduced
to represent the input signal and disturbances that act on a system. In
this section, the time responses of the systems in both the transient
and steady state conditions will be discussed. Figure 2.9 shows a block
diagram for a position control system for a turning table, which represents
an accurate positioning system with a heavy load. The ideal conditions
for the system is that the angle of the turning table, θo(s), will accurately
follow the desired angle, θi(s), set as the input to the system, regardless
of the occurrences of disturbances, D(s).

Due to the heavy load, the ideal conditions might not be easily achieved.
It tends to suffer from instability in the response and steady state error.
The design of the turning table involves large mass with as little friction as
possible. This imposes a low damping ratio that results in an overshooting
and oscillating system response. Steady state error may also occur due
to the heavy load of the turning table. As the load approaches the desired
angle, the error signal, E(s), becomes progressively smaller to a situation
that the amplified error signal is insufficient to keep the motor driving.
Thus the load will come to rest at an angle other than the desired. The
discrepancy between the actual angle at rest and the desired angle is
defined as the steady state error.

UNIT 2 System response analysis 21

 D(s)

i(s) E(s) O(s)

Variable Power Turning
Motor Gear
resistor amplifier table

Variable
resistor

Figure 2.9 Block diagram of a turning table position control system

The details of the system response will be further discussed in the

following section. For the simplicity of calculation and the sake of
analysis, the discussion will focus mainly on the time responses of the
first order and the second order system when a step force function is
applied as the input signal.

First order time response

Consider the closed loop control system with a unity feedback as shown
in Figure 2.10. The closed loop transfer function denoted as, T(s) is:
1
Y(s) ts 1
T(s) = = =
R(s) 1 + 1 ts + 1
ts

1
where t is the time constant. Applying a unit step input R(s) = and
s
assuming zero initial conditions,

1 1
Y(s) = ∙
ts + 1 s

22 TEL 306/05 Control Systems

Perform partial franctions expansion and inverse Laplace transform

1 1
Y(s) = −
s s+ 1
t
y(t) = 1 − e−t/t

The equation above indicates that the initial output, y(0) is zero and the
output at steady state , y(∞), converges to unity, i.e.

lim y(t) = 1
t→∞

1
Yss =
s

The error, e(s), at steady state is:

lim
t→∞
e(t) = 0

Ess = 0
since E(s) = R(s) – Y(s)

R(s) E(s) 1 Y(s)

ts

Figure 2.10 First order closed loop system

The transient response of the first order system is determined by the time
constant, τ, of the system. At the time, t = τ, the system output reaches
63.2% of the final value at steady state. This can be proven as follows:

y(t) = 1 − e− t/t
y(t) = 1-e−1
y(t) = 0.632 = 63.2%

UNIT 2 System response analysis 23

 The time constant, τ, is an indicator to the speed of the system to reach
the final value. The speed of response can be quantitatively defined as
the time for the output to become a particular percentage of its final
value. A large time constant corresponds to a sluggish system and a
small time constant corresponds to a fast response as shown in Figure
2.11. The time constant, τ, can also be used to calculate the transient
response of the system as shown in Figure 2.12.

Figure 2.11 Speed of systems with respect to time constants

y(t)
63.2%
86.5%

98.2%
99.3%
95%

t
0 t 2t 3t 4t 5t

Figure 2.12 Calculation of transient response using time constant

24 TEL 306/05 Control Systems

 Activity 2.3

In an experiment, the water in the beaker is heated from

temperature of 20ºC to the boiling point of 100ºC. The time taken
for the temperature to reach 100ºC is 50 seconds. Derive the
transfer function of the boiling process.

Second order time response

Consider a closed loop control system with a unity feedback as shown
in Figure 2.13. The closed loop transfer function, T(s), is:

Y(s) w n2
T(s) = = 2
R(s) s + 2wnϚs + wn2
The transient response of the second order system could be described
by two parameters, namely the natural frequency, wn, and the damping
ratio, Ϛ. The natural frequency is the frequency of oscillation of the
system if all the damping is removed, while the damping ratio is the
ratio of the exponential decay frequency to the natural frequency (Nise,
2011). Depending on the damping ratio, Ϛ, the transient responses of
the second order system when subjected to a unit step input R(s) =

1
, can be categorised into four cases as follows:
s

R(s) E(s) Y(s)

w n2
s(s + 2Ϛwn)

Figure 2.13 Second order closed loop system

UNIT 2 System response analysis 25

 Case 1: Marginally stable when � = 0
The standard second order transfer function is given as:

Y(s) w n2
= 2
R(s) s + 2wnϚs + wn2

Since the damping ratio, Ϛ, is zero, the resultant transfer function is:

Y(s) w n2
=
R(s) s2 + wn2

Solving the characteristic equation (denominator) where s2 + w2n = 0, there

are two complex poles at:

p1, 2 = ± jwn

Figure 2.14 shows the locations of the poles on the s-plane, which are
plotted on the imaginary axis. The output response to a unit step input
can be derived from:

w n2 1 1 s
Y(s) = 2 · = −
s + w n2 s s s 2 + w n2

Inverse Laplace,

y(t) = 1 − cos wn t

The solution above reveals that the output, y(t), oscillates above a unity
and with a frequency of wn rad/s. Figure 2.15 shows a typical step response
for the case where Ϛ = 0 and wn = 2 rad/s. The oscillatory response is
denoted as a marginally stable response in the control system.

26 TEL 306/05 Control Systems

 jwn Im

Re

−jwn

Figure 2.14 Locations of poles when Ϛ = 0

y(t)

Figure 2.15 Step response for Ϛ = 0, wn = 2 rad/s

Case 2: Underdamped when 0 < Ϛ < 1

The standard second order transfer function is given as:

Y(s) w n2
= 2
R(s) s + 2wnϚs + wn2

UNIT 2 System response analysis 27

 Since the damping ratio, Ϛ, is a value between zero and unity, the
characteristic equation is:

s2 + 2wn Ϛs + wn2 = 0

Solving the characteristic equation, there are two complex poles at:

p1, 2 = −Ϛwn ± jwn√(1 − Ϛ 2) = −a± jwd

where a = Ϛwn, wd is denoted as the damped frequency. The equation

wd = wn√(1 − Ϛ 2) indicates that increasing damping ratio, Ϛ, will reduce
the damped frequency, wd. Since 0 < Ϛ < 1, wd is always less than wn.
Increasing the damping ratio, Ϛ, will shift the locations of the poles
towards real axis on the s-plane as shown in Figure 2.16.

The output response to a unit step input is given as:

y(t) = 1 −
e−a
√1 − Ϛ2 (
sin wdt + tan −1 √1 − Ϛ

Ϛ
2

)
Figure 2.17 shows a typical step response with an overshoot for the case
where � = 0.5 and wn = 2 rad/s. The transient response is denoted as an
underdamped response in the control system.

Figure 2.16 Increasing Ϛ and shifting poles

28 TEL 306/05 Control Systems

 y(t)

Figure 2.17 A typical underdamped step response

Case 3: Critically damped when Ϛ = 1

The standard second order transfer function is given as:

Y(s) w n2
= 2
R(s) s + 2wnϚs + wn2

Since the damping ratio, Ϛ, is a unity, the characteristic equation is:

s2 + 2wns + wn2 = 0

Solving the characteristic equation, there are two real poles at:
p1, 2 = −wn

The poles are plotted at the same location on the s-plane at the real
axis as shown in Figure 2.18. The output response to a unit step input
is given as:

y(t) = 1 − e−wnt (1 + wnt)

UNIT 2 System response analysis 29

 Figure 2.19 illustrates a typical step response for the case where Ϛ = 1 and
wn = 2 rad/s. The transient response is denoted as a critically damped
response in the control system where no overshoot should be exhibited.

Imaginary

Real

Figure 2.18 Both poles at the same location

y(t)

Figure 2.19 A critically damped step response

30 TEL 306/05 Control Systems

Case 4: Overdamped when Ϛ > 1

The standard second order transfer function is given as:

Y(s) w n2
= 2
R(s) s + 2wnϚs + wn2

Since the damping ratio, Ϛ, is a unity, the characteristic equation is:

s2 + 2Ϛwns + wn2 = 0

where Ϛ > 1. Solving the characteristic equation, there are two negative,
real and distinct poles at:

p1 = (Ϛ + √Ϛ2 − 1)wn and p2 = (Ϛ − √Ϛ2 − 1)wn

The poles are located at the real axis on the s-plane as depicted in Figure
2.20. The characteristics of the system response are determined by the
pole, p1, that is located near the origin of the s-plane. This is the concept
of dominant pole that will be described in the next section. The nearer
to the origin, the slower is the system response. Increasing the damping
ratio, Ϛ, will further distinct both poles apart. One of the poles, p2, will
approach the origin while another pole, p1, moves away from the origin.
Hence the system responds slower.

The output response to a unit step input is given as:

y(t) = 1 +
wn
( e−p1t
2√Ϛ 2 − 1 p1
−
e−p2t
p2 )
Figure 2.21 shows a typical step response for the case where Ϛ = 1.5
and wn = 2 rad/s. The transient response is slower as compared with the
critically damped case illustrated in Figure 2.19.

UNIT 2 System response analysis 31

 Imaginary

p1 p2
Real

Increasing Ϛ

Figure 2.20 Poles separated with increasing Ϛ

y(t)

Figure 2.21 An overdamped step response

The second order system responses at various damping ratio, Ϛ, is

illustrated in Figure 2.22. The smaller the value of the damping ratio, Ϛ,
the more oscillatory is the response. The frequency of the oscillation is
the damped frequency, wd. as described in the previous section. When

32 TEL 306/05 Control Systems

Ϛ ≥ 1, there is no overshoot and for large value of damping ratio, Ϛ, the
response becomes sluggish. The damping ratio, Ϛ ≈ 0.7 gives the system
the best performance with a small overshoot and quick system response.

Figure 2.22 The effect of damping ratio, Ϛ, on the system output, y(t)
Source: https://en.wikipedia.org/wiki/Damping_ratio

Web Reference

For more information on the effect of damping ratio to a control

system, you may visit this website:

https://en.wikipedia.org/wiki/Damping_ratio

UNIT 2 System response analysis 33

 Example 4

The transfer function for a process can be expressed as below:

Y(s) 3
= 2
U(s) 5s + 7s + 2

Find the value of the gain, K, the damping ratio, Ϛ, and the natural
frequency, wn, of the process. Sketch and comment on the system
response based on the calculations.

Solution
2
Y(s) 3 3 5
= 2 = •
U(s) 5s + 7s + 2 2 s2 + 7 s + 2
5 5

Comparing with standard second order transfer function:

2
Y(s) 3 5 w n2
= • =K· 2
U(s) 2 s2 + 7 s + 2 s + 2wnϚs + wn2
5 5

3
K = = 1.5
2

wn =
√ 2
5
= 0.6325

34 TEL 306/05 Control Systems

 7
2wnϚ = = 1.4
5

1.4
Ϛ= = 1.1067
2 · 0.6325

The figure above shows the step response of the process. The gain, K,
indicates the final value that the system converges. Since the damping
ratio is more than one, a sluggish response without overshoot is expected.

UNIT 2 System response analysis 35

 Activity 2.4

A simple closed loop control system is designed as illustrated in

Figure 2.23. Obtain:

1. The natural frequency.

2. Damped natural frequency.

3. Damping ratio for the system.

R(s) 1 Y(s)
+ 20 (s + 1)(0.2s + 1)
-

1
5

Figure 2.23 A simple closed loop control system

36 TEL 306/05 Control Systems

 Activity 2.5

A forging hammer dynamic system consists of a mass, a spring

and a damper. The free body diagram of the system with an input
force, F(t), a spring force, Fs(t) and a damper force, Fd(t) is shown
below:

F(t)
Mass, m
x(t)
dx(t)
Fs(t) = kx(t) Fd(t) = kd
dt

The system is subjected to an input force, F(t) and the output is the
displacement, x(t). The transfer function of the system is given as:
1
x(s) m
=

m ( )
F(s) s2 + kd s + k
m

Given the stiffness, k = 800N/m, mass, m = 3kg and damping

coefficient, kd = 20Ns/m, calculate:

1. Natural frequency of the system, wn.

2. Damping ratio of the system, Ϛ.

3. Sketch and label the step response of the system.

UNIT 2 System response analysis 37

 Transient response analysis
The transient response of a first order system is rather simple as the
response increases with respect to time from initial state to the final
state as shown in Figure 2.11. The speed of response can be described
by the time constant, τ, of the transfer function.

A second order system may present various transient responses

depending on the natural frequency, wn, and the damping ratio, Ϛ. The
characteristics of the transient response of a second order system could
be demonstrated in Figure 2.24 and formulated as follows:

1. Delay time, td:

Time required for the response to reach half of the final value
at very first time.

2. Rise time, tr:

Time required for the response to reach the final value at very
first time.

p−b
tr =
wd

b = tan−1
( )
wd
Ϛwn

where b is the angle of the pole as illustrated in the graph below.

38 TEL 306/05 Control Systems

 jw

jwd
wn
w n√ 1 −  2
b

−

wn

y(t)

%OS or Mp
Allowable
Final tolerance
value 5% or 2%

td tr tp ts time

Figure 2.24 Characteristics of second order transient response

3. Peak time, tp:

Time required for the response to reach the first peak of the
overshoot.

p p
tp = =
wd wn√1 − 2

UNIT 2 System response analysis 39

 4. Maximum overshoot, Mp:

Maximum peak value of the response curve measured from

unity,

Mp = 1 + exp
( −p
√1 − 2 )
Or maximum percentage overshoot,

%OS = 100exp
( −p
√1 − 2 )
5. Settling time, ts:

The time required for the system response to reach and stay
within a specific percentage tolerance band of the final steady
state value. Typical values of tolerance are 2% and 5%.

4
ts = 4t = (2% criterion)
wn

3
ts = 3t = (5% criterion)
wn

where τ is the time constant that could be expressed as:

1
t=
wn

40 TEL 306/05 Control Systems

 Example 5

Figure 2.25 depicts a feedback control for a linear system. Investigate

the transient response of the system by finding the rise time (tr), the peak
time (tp), the %OS and the settling time (ts) for a unit step input.

R(s) + + 1 1 Y(s)
25
- - (s + 2) s

0.16 s

Figure 2.25 A feedback control system

Solution

Obtaining the close loop transfer function of the system:

Y(s) 25
= 2
R(s) s + 6s + 25

Comparing the transfer function with the standard second order transfer
function:

25 wn2
⇒ 2
s2 + 6s + 25 s + 2wns + wn2

wn2 = 25 → wn = 5 rad/s

6
2n = 6 →  = = 0.6
2×5

UNIT 2 System response analysis 41

 Rise time (tr)

p−b wd
tr = wd = wn√1 − Ϛ2 = 4 rad/s, b = tan−1
wd wn

3.14 − 0.93 4
= = tan−1
4 3
= 0.55s = 0.93 rad

Peak time (tp)

p 3.14
tp = = = 0.785 s
wd 4

Percentage of overshoot %OS

%OS = 100exp ( −p

√1 − 2 )
= 9.5%

Settling time (ts)

3 3
ts = 3t = = = 1s (5% criteria)
wn 3

4 4
ts = 4t = = = 1.33s (2% criteria)
wn 3

42 TEL 306/05 Control Systems

 Activity 2.6

Figure 2.26 shows the block diagram of a unity feedback system.

Determine the:

1. Natural frequency

2. Damping ratio

3. Peak time

4. Maximum overshoot percentage

5. Settling time

6. Rise time of the system.

+ E(s)
1
R(s) s(s + 15) Y(s)
-

Figure 2.26 A unity feedback system

UNIT 2 System response analysis 43

 Steady state response analysis
The steady state response refers to the output response when the
system settles down at an infinite time. Both first order and second order
systems at steady state exhibit similar response and the performance
of the system is determined by the discrepancy between the desired
and the actual output, namely the steady state error. In this section,
the classification of the control system will be first introduced and the
steady state response of the system when subjected to different input
force functions will be investigated. Three forms of input force functions
elaborated in the previous section will be adopted, i.e. a step input, u(t),
a ramp input, tu(t), and a parabolic input, t2u(t).

Classification of control systems

A control system can be classified based on the ability of the system to

follow the input force functions. Mathematically, the classification can
be identified by referring to the denominator of an open loop transfer
function:

K(s + z1)(s + z2) ......... (s + zm)

G(s) =
sN(s + p1)(s + p2) ......... (s + pq)

where the value of the N variable defines the type of the system, K is the
process gain, z is the zero and p is the pole.

Steady state error for a unity feedback system

Consider a system shown in Figure 2.27(a), the equivalent closed loop

transfer function is depicted in Figure 2.27(b) where:

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

44 TEL 306/05 Control Systems

 + E(s)
R(s) G(s) Y(s)
R(s) G(s) Y(s)
- 1 + G(s)
Equivalent
1

(a) A unity feedback system (b) Equivalent block diagram

Figure 2.27

The error signal, E(s), is the difference between the setpoint, R(s), and
actual output, Y(s):

E(s) = R(s) – Y(s)

The error transfer function could be obtained as:

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

Applying the final value theorem to find the steady state performance
of the system:

sR(s)
ess = lim e(t) = lim sE(s) = lim
t→ s→0 s→0
1 + G(s)

The steady state error, ess, equation above provides an estimation of the
steady state performance of the system for a given transfer function of
the process, G(s), and the input, R(s). The steady state performance can
now be evaluated when the step, the ramp, and the parabolic inputs are
applied.

UNIT 2 System response analysis 45

 Step input:

1
When a step input, , is applied to the system, the steady state error is:
s

s 1
ess = lim ∙
s→0
1 + G(s) s

1
=
1 + limG(s)
s→0

The term, lim G(s), is a dc gain of the open loop transfer function, since
s→0
s is the frequency variable that approaches zero. In order to have a zero
steady state error,

lim G(s) = 
s→0

To obtain this, the denominator of the process transfer function, G(s),

must be equal to zero as s goes to zero. Effectively, this means that the
system must be at least type 1 to obtain a zero steady state error. If the
system is type 0, then a steady state error exists. Based on the process
transfer function, G(s), the magnitude of the steady state error, ess, is

z1z2 ... zm
lim G(s) =
s→0 p1p2 ... pq

46 TEL 306/05 Control Systems

Ramp input:

1
When a ramp input, , is applied to the system, the steady state error is:
s2

s 1
ess = lim ∙ 2
s→0
1 + G(s) s

1
= lim
s→0
s + sG(s)

1
=
lim sG(s)
s→0

In order to have a zero steady state error,

lim sG(s) = 
s→0

To obtain this, the denominator of the process transfer function, G(s),

must be equal to zero as s goes to zero. Effectively, this means that the
system must be at least type 2 to obtain a zero steady state error. If the
system is type 1, then a steady state error exists. Based on the process
transfer function, G(s), the magnitude of the steady state error, ess, is

z1z2 ... zm
lim sG(s) =
s→0 p1p2 ... pq

UNIT 2 System response analysis 47

 Parabolic input:

1
When a parabolic input, , is applied to the system, the steady state
s3
error is:

s 1
ess = lim ∙ 3
s→0
1 + G(s) s

1
= lim
s→0
s2 + s2G(s)

1
=
lim s2G(s)
s→0

In order to have a zero steady state error,

lim s2G(s) = 
s→0

To obtain this, the denominator of the process transfer function, G(s),

must be equal to zero as s goes to zero. Effectively, this means that the
system must be at least type 3 to obtain a zero steady state error. If the
system is type 2, then a steady state error exists. Based on the process
transfer function, G(s), the magnitude of the steady state error, ess, is

z1z2 ... zm
lim s2G(s) =
s→0 p1p2 ... pq

If the system is type 1 or type 0, then

lim s2G(s) = 0
s→0

The steady state error is infinite. The error is growing!

48 TEL 306/05 Control Systems

 Example 6

Find the steady state errors for the inputs of step u(t), ramp tu(t) and
parabolic t2u(t) to the system as shown in Figure 2.28. Comment on your
findings.

R(s) + E(s) 100(s + 2)(s + 5)(s + 6) Y(s)

-
s(s + 8)(s + 10)(s + 12)

Figure 2.28 Type 1 system

Solution

For a type 1 system as shown in Figure 2.28, the steady state error, ess
is formulated as follows:

E(s) = R(s) − Y(s)

1
E(s) = R(s)
1 + G(s)
since,

100(s + 2)(s + 5)(s + 6)

G(s) =
s(s + 8)(s + 10)(s + 12)

s(s + 8)(s + 10)(s + 12)

E(s) = R(s)
s(s + 8)(s + 10)(s + 12) + 100(s + 2)(s + 5)(s + 6)

ess = lim sE(s)

s→0

UNIT 2 System response analysis 49

 s(s + 8)(s + 10)(s + 12)
ess = lim s R(s)
s→0 s(s + 8)(s + 10)(s + 12) + 100(s + 2)(s + 5)(s + 6)

1
Applying a unit step input, R(s) =
s

s(s + 8)(s + 10)(s + 12) 1

ess = lim s
s→0 s(s + 8)(s + 10)(s + 12) + 100(s + 2)(s + 5)(s + 6) s

When s → 0
ess = 0

1
Applying a unit ramp input, R(s) =
s2

s(s + 8)(s + 10)(s + 12) 1

ess = lim s
s→0 s(s + 8)(s + 10)(s + 12) + 100(s + 2)(s + 5)(s + 6) s2

When s → 0

(8)(10)(12)
ess = = 0.16
100(2)(5)(6)

1
Applying a unit parabolic input, R(s) =
s3

s(s + 8)(s + 10)(s + 12) 1

ess = lim s
s→0 s(s + 8)(s + 10)(s + 12) + 100(s + 2)(s + 5)(s + 6) s3

When s → 0

(8)(10)(12) 1
ess = •
0 + 100(2)(5)(6) 0
ess = 

50 TEL 306/05 Control Systems

The result indicates that type 1 system follows the step input well
without any steady state error. When a ramp input is applied, the system
is slightly sluggish and exhibits a steady state error. When a parabolic
input is applied, the system becomes unstable with the steady state error
increases to infinity. If you can recall the description in the unit overview,
the type 1 system can best be analogised to the average student sitting
in the class! On the same token, can you estimate the steady state
performance for type 0 and type 2 systems?

Static error constants

In the previous section, the steady state errors are estimated by using
the final value theorem. An alternative method is by using the static error
constant. Recall that in the previous section, we derived the following
steady state functions,

For a step input, u(t)

1
ess =
1 + lim G(s)
s→0

For a ramp input, tu(t)

1
ess =
lim sG(s)
s→0

For a parabolic input, t2u(t)

1
ess =
lim s2G(s)
s→0

UNIT 2 System response analysis 51

 The static error constants are defined as:

Position constant,

Kp = lim G(s)
s→0

Velocity constant,

K = lim sG(s)
s→0

Acceleration constant,

Ka = lim s2G(s)
s→0

The static error constants appear as the denominators of the steady

state error functions. Therefore, the steady state errors increase as the
values of the static error constants decrease.

Example 7

Re-evaluate the steady state errors of the system in the previous example
using the static error constants.

52 TEL 306/05 Control Systems

 Solution

To re-evaluate the steady state error for type 1 system described in the
previous example, the open loop transfer function is:

100(s + 2)(s + 5)(s + 6)

G(s) =
s(s + 8)(s + 10)(s + 12)

1
When a unit step input, R(s) = , is applied, the static constant is:
s
Kp = lim G(s)
s→0

The steady state error is:

1 1
ess = = =0
1 + Kp 1 + 

1
When a unit ramp input, R(s) = , is applied, the static constant is:
s2
K = lim sG(s)
s→0

The steady state error is:

1 1
ess = = = 0.16
K 100(2)(5)(6)
(8)(10)(12)

UNIT 2 System response analysis 53

 1
When a unit parabolic input, R(s) = , is applied, the static constant is:
s3
Ka = lim s2G(s)
s→0

The steady state error is:

1 1
ess = = =
Ka 0

Using the static error constants to evaluate the steady state error is
comparatively easier but you should be aware that it is applicable only
to a closed loop system with unity feedback loop.

Table 2.2 summarises the steady state responses of various types of

closed loop control systems with a unity feedback loop when subjected
to different input force functions. It could be very helpful when a quick
evaluation of the system is required.

System u(t) = 1 u(t) = t u(t) = t 2

1  
Type 0
1 + Kp

Type 1 0 
1
K

Type 2 0 0
1
Ka

Table 2.2 Static error constants for various types of systems subjected to
different inputs

54 TEL 306/05 Control Systems

 Activity 2.7

Figure 2.29 shows a closed loop control system. Determine

the type of the system. With reference to Table 2.2, suggest an
appropriate input force function to obtain zero steady state error.

R(s) E(s) 100 (s)

+ s(s + 10)

1
s+5

Figure 2.29 A closed loop control system

Steady state error for a non-unity feedback system

To calculate the steady state error for a non-unity feedback system, the
same approach of derivation could be applied. For a simple closed loop
system as shown in Figure 2.30 (a), the equivalent closed loop transfer
function can be depicted as in Figure 2.30 (b). The output of the closed
loop system is:

G(s)
Y(s) = R(s)
1 + G(s)H(s)

Since the error signal equation is:

E(s) = R(s) − Y(s)

UNIT 2 System response analysis 55

 By substituting the output, Y(s), of the closed loop system into the error
signal equation, we have:

E(s) = 1 −
( G(s)
1 + G(s)H(s)
R(s)
)
Applying the final value theorem to find the steady state performance of
the system:

ess = lim e(t) = lim sE(s)

t→ s→0

ess = lim sR(s) 1 −

s→0 ( G(s)
1 + G(s)H(s) )
ess = lim sR(s)
s→0 ( 1 + G(s)H(s) − G(s)
1 + G(s)H(s) )

It should be remarked that the above steady state error, ess, is general and
applicable to both systems with a unity and a non-unity feedback loop.

+ E(s) Y(s)
R(s) G(s)
- R(s) G(s) Y(s)
Equivalent
1 + G(s)H(s)
H(s)

(a) Non-unity feedback (b) Equivalent feedback

Figure 2.30

56 TEL 306/05 Control Systems

 Example 8

Find the steady state error for the system shown below for a unit step
input.

R(s) E(s) 100 Y(s)

+- s(s + 10)

1
s + 0.5

Solution

The overall closed loop transfer function is given as:

G(s)
Y(s) = R(s)
1 + G(s)H(s)

100 1
where G(s) = and H(s) =
s(s + 10) s + 0.5

Hence the closed loop transfer function is:

Y(s) 100(s + 0.5)

=
R(s) s(s + 10)(s + 0.5) + 100

The output, Y(s), is:

100(s + 0.5)
Y(s) = • R(s)
s(s + 10)(s + 0.5) + 100

UNIT 2 System response analysis 57

 Substitute into error signal equation:

E(s) = 1 −
( 100(s + 0.5)
s(s + 10)(s + 5) + 100 ) • R(s)

E(s) =
( s(s + 10)(s + 0.5) + 100 − 100(s + 0.5)
s(s + 10)(s + 0.5) + 100 )
• R(s)

The steady state error:

ess = lim e(t) = lim sE(s)

t→ s→0

ess = lim s
s→0 ( s(s + 10)(s + 0.5) + 100 )
s(s + 10)(s + 0.5) + 100 − 100(s + 0.5)
• R(s)

ess = lim s
s→0 ( s(s + 10)(s + 0.5) + 100 )
s(s + 10)(s + 0.5) + 100 − 100(s + 0.5)
•
1
s

ess =
( 0 + 100 − 100(0.5)
100
= 0.5
)
Pole placement and system response
It should be realised that the location of poles on the s-plane influences
the response of the system. For a first order system, the closer the pole
to the origin of the s-plane, the slower is the system response. For a
second order system, when the poles are located on the imaginary axis,
the system exhibits marginally stable condition. When the poles are
located on the left hand side of the s-plane, the system is stable but it
may have overshoots depending on the damping ratio. If the damping
ratio is greater than one, there will be two real negative poles which
exhibit sluggish response without overshoot. Alternatively, when the
damping ratio is less than one, the poles will be complex conjugates,
which exhibit fast response but with overshoots.

58 TEL 306/05 Control Systems

Figure 2.31 illustrates different locations of poles on the s-plane and the
respective time responses. Generally, if a stable system is to be designed,
all the poles must be placed on the left hand side of the s-plane. Any poles
located at the right hand side of the s-plane will cause instability to the
system. Complex conjugates poles will result in oscillatory response. It
is to be noted that the gain, K, in the closed loop system will also affect
the location of poles. The details will be discussed in the next unit on
the root locus section.

Imaginary, jw

Real, 

Figure 2.31 Locations of poles and the effect on time response

Source: Gradeup
https://gradeup.co/transfer-function-concept-of-pole-zero-study-notes-for-
electrical-engineering-i-6f209990-93c4-11e7-851c-36d9cb5f5fa5

An example of poles and zeros on the s-plane is illustrated in Figure 2.32.

The positions of these poles are directly related to the damping ratio, Ϛ,
and the natural frequency, wn. The distance from the pole to the origin
is the natural frequency, wn. The angle between the real axis and the line
between the origin and the pole is:

b = cos−1 Ϛ

UNIT 2 System response analysis 59

 Thus, if specific requirements are placed upon the damping ratio Ϛ,
and the natural frequency, wn, the required position of the poles can be
identified. The concept of dominant pole is very important when a third
order system is investigated. If the third pole as shown in Figure 2.32 lies
far away from the complex conjugates poles that:

| |
1

≥ 10|Ϛwn|

then the time response of the third order system can be approximated
by the time response of the second order system with the complex
conjugates poles. It should be reminded that the poles that are located
closer to the origin exhibit slower response, hence are the dominant
poles.

jw

jwd
wn
wn√1 − Ϛ2
b

1 − Ϛw
− n


Figure 2.32 The poles on s-plane

60 TEL 306/05 Control Systems

 Reading

For information about the effects of third pole and zero on the
second order system response, please refer to:

Richard, C. D. & Robert, H. B. (2011). Modern control systems

(12th ed.). Boston: Pearson.

Activity 2.8

Figure 2.33 shows the locations of the closed-loop poles of a

negative feedback system. Determine the:

1. Natural frequency

2. Damping ratio

3. Peak time

4. Maximum percentage overshoot

5. Settling time of the system.

UNIT 2 System response analysis 61

 Imaginary,
jw

j7


Real, 
−3

Figure 2.33 The poles of a negative feedback system on s-plane

Example of system design via pole placement

In a position control of a DC motor, a tachometer with the gain function
of Kf is included to improve the stability of the system. It is observed that
the desired time responses could be obtained by varying the gain of the
tachometer, Kf. Determine the gain, Kf, and the corresponding poles when:

1. The system is critically damped, Ϛ = 1 so that no overshoot is

allowed.

2. Damping ratio, Ϛ = 0.7 to obtain a fast response when minimal

overshoots are allowed.

62 TEL 306/05 Control Systems

 i(s) E(s) 10 1 o(s)
+-
s+2 s

Kf

Figure 2.34 The negative feedback tachometer system

Solution:

The overall closed loop transfer function could be derived as follows:

i(s) E(s) 10 1 o(s)

+−
s + 2 + 10Kf s

i(s) E(s) 10 o(s)

+-
s2 + (2 + 10Kf)s

i(s) 10 o(s)
s2 + (2 + 10Kf)s + 10

UNIT 2 System response analysis 63

 Comparing the closed loop transfer function to the standard second
order transfer function:

o(s) 10 C(s) wn2

= 2 = 2
i(s) s + (2 + 10Kf)s + 10 R(s) s + 2Ϛwns + wn2

The natural frequency, n, could be obtained as:

n = √10 = 3.1623 rad/s

1. At critically damped, Ϛ = 1 so that no overshoot is allowed

2Ϛn = 2 + 10 Kf

2Ϛwn − 2
∴ Kf = = 0.4325
10
Substitute the natural frequency, n, gain, Kf, into characteristics
equation (CE):

CE = s2 + (2 + 10 * 0.4325)s + 10
Poles will be at p1 = −3.2, p2 = −3.1

2. When a minimal overshoot is allowed at damping ratio, Ϛ = 0.7

2Ϛn = 2 + 10 Kf

2Ϛwn − 2
∴ Kf = = 0.2427
10
Substitute the natural frequency, n, gain, Kf, into characteristics
equation (CE):

CE = s2 + (2 + 10 * 0.2427)s + 10
Poles will be at p1, 2 = −2.2135 + j2.2584

The example above indicates that by applying appropriate gain

on the tachometer, Kf, the location of the closed loop poles will be
altered. This will change the damping ratio and the time response.

64 TEL 306/05 Control Systems

 Example 9

A simplified tacho-generator system can be illustrated in the block

diagram as shown in Figure 2.35. Calculate the following:

1. Closed loop transfer function of the system.

2. Characteristic equation of the system.

3. Natural frequency and damping ratio in terms of KA.

4. Roots of the closed loop poles in terms of KA.

5. Steady state error of the system due to a ramp input

1
R(s) = in terms of KA.
s2

6. Damping ratio, natural frequency, damped natural frequency,

closed-loop poles, steady state error, rise time, peak time and
maximum overshoot for KA = 14.5.

7. If KA is decreased to 7.247, determine the new damping ratio

and compare the system performance between KA = 14.5 and
KA = 7.247.

θi(s) E(s) 4500 θo(s)

+ KA
-
s(s + 361.2)

Figure 2.35 Simplified block diagram of a tacho-generator

UNIT 2 System response analysis 65

 Solution

1. The open loop transfer function is given as:

o(s) 4500KA
= 2
i(s) s + 361.2s

Using the closed loop transfer function,

Y(s) 1
=
R(s) 1 + G(s)H(s)

Since H(s) = 1 for a unity feedback, hence the closed loop

transfer function can be derived as:

o(s) 4500KA
= 2
i(s) s + 361.2s + 4500KA

2. Characteristic equation (C.E.) can be obtained with reference

to the denominator of the closed loop transfer function, i.e.
C.E. = s2+ 361.2s + 4500KA

3. Comparing to the standard second order equation, s2+ 2Ϛwns

+ w2n, the natural frequency, wn and damping ratio, Ϛ, could
be obtained as follows:

wn = √4500KA

361.2 2.692
Ϛ= =
2√4500KA √KA

66 TEL 306/05 Control Systems

4. The roots of the closed loop poles can be calculated using the
formula,

−361.2 ± √361.22 − 4(1)(4500KA)

P1,2 =
2(1)
= −180.6 ± √32616.36 − 4500KA

5. When the input is a ramp, the steady state error of the system
is,

1
ess = lim sE(s) = lim
s→0 s + sG(s)H(s)

361.2 0.0803
ess = lim sE(s) = =
s→0 4500KA KA

6. Damping ratio, natural frequency, damped natural frequency,

closed-loop poles, steady state error, rise time, peak time and
maximum overshoot for KA = 14.5.

2.692
a. Damping ratio = = 0.707
√ 14.7

b. Natural frequency, wn = √4500*14.5 = 255.44

c. Damped natural frequency, wd = wn √1 − Ϛ2

= 255.44*√1 − 0.707
= 180.6 rad/sec

d. Closed loop poles = −180.6 ± √32616.36 − 4500*14.5

= −180.6 ± j 180.

UNIT 2 System response analysis 67

 0.0803
e. Steady state error = = 0.0055 rad
14.5

f. Rise time, tr =
p − cos−1Ϛ
=
p − 45 ( )
p
180
wd 180.6

= 0.013 seconds

p
g. Peak time, tp = = 0.0174 seconds
wd

h. Maximum overshoot, Mp = e−pϚ/√ 1 − Ϛ * 100 = 4.3%

2

7. New damping ratio when increasing of KA value from 7.247

to 14.5.

2.692
Damping ratio = =1
√7.247

Increase of KA value from 7.247 to 14.5 has resulted in a

decrease in damping ratio. Moderate value of Ϛ results in better
system performance compared to critical value Ϛ = 1.

68 TEL 306/05 Control Systems

 Summary

This section elaborated the time response of the first and second
order systems, which can be described as the transient response
and the steady state response. The transient response of the first
order system follows the setpoint at a speed stated by the time
constant, t. When the time approaches the time constant, t, the
system reaches 63% of the final output and gradually converges
to the final output without overshooting. The transient response
of the second order system may vary depending on the damping
ratio, Ϛ, and the natural frequency, wn. The smaller is the value of
the damping ratio, Ϛ < 1, the more oscillatory is the response. When
Ϛ ≥ 1, there is no overshoot and the response becomes sluggish
when Ϛ increases. Analyses of the second order transient response
also include calculating the rise time, peak time, overshoot, etc.

The analysis of steady state response reveals that the occurrence

of steady state error depends on the form of input signal and
the type of the system. This is akin to the teaching speed of the
teacher (input) and the learning speed of the student (system).
For example, when a type 1 system is subjected to a step input
function, the output is without any steady state error. For the same
system, when a ramp input is applied, then steady state error is
produced.

The time response is defined by the pole placement. It has been

demonstrated that by adjusting the gain of the feedback loop,
different time responses could be obtained.

UNIT 2 System response analysis 69

 Suggested answers to activities

Feedback

Activity 2.3

In an experiment of the boiling process, the step input applied is

from 20ºC to 100ºC, which is a magnitude of (100 − 20)ºC = 80ºC.
According to Figure 2.12, the total time to reach 99.3% is around
5t, which is around 50 seconds. Hence the time constant, t, of
the process is 10 seconds, which happens when the temperature
reached 63% of the 80ºC, i.e.:

Temperature at 10 seconds, y(10) = [20 + (80*0.63)]ºC = 70.4ºC

The boiling process is a typical first order system, hence the

transfer function could be obtained as:

Y(s) 1
=
R(s) 10s + 1

Activity 2.4

20
G(s) =
(s + 1)(0.2s + 1)

1
H(s) =
5

Y(s) 1 100
= = 2
R(s) 1 + G(s)H(s) s + 6s + 25

70 TEL 306/05 Control Systems

wn2 = 25

1. wn = 5rad/sec
2Ϛwn = 6

6
2. Ϛ= = 0.6
2*5

3. wd = wn(√1 − Ϛ2)
= 5(√1 − 0.62) = 4 rad/sec

Activity 2.5

1. wn =
√( )
k
m

=
√( )
800
3
= 16.322 rad/s

kd
2. 2 Ϛwn =
m
kd 20
Ϛ= =
(2wnm) (2*16.322*3)
= 0.2042

UNIT 2 System response analysis 71

 3.

Activity 2.6

The closed loop transfer function is:

Y(s) 100
= 2
R(s) s + 15s + 100

100 wn2
⇒
s2 + 15s + 100 s2 + 2wns + wn2
wn2 = 100

1. wn = 10
2wn = 15

15
2. = = 0.75
20

72 TEL 306/05 Control Systems

 p p
3. tp = = = 0.475s
wn√1 −  10√1 − (0.75)
2 2

4 4
4. ts = = = 0.533s
wn (0.75)(10)

2.16 + 0.6 2.16(0.75) + 0.6

5. tr = = = 0.222s
wn 10

6. %OS = 100e−(Ϛp/√ 1 − Ϛ ) = 2.84%

2

Activity 2.7

The block diagram could be reduced to:

R(s) E(s) Y(s)

+- 100(s + 5)
s(s + 10)(s + 5) + 100

100(s + 5)
Since G(s) =
s(s + 10)(s + 5) + 100

The system is type 1.

Referring to Table 2.2, a unit step input is suggested to obtain a

zero steady state error.

UNIT 2 System response analysis 73

 Activity 2.8

cos  = 

wd 7
tan  = =
d 3

7
 = tan−1 = 66.80º
3

1.  = cos 66.80º = 0.393

2. wn = √32 + 72 = 7.616 rad/s

p p
3. tp = = = 0.449s
wd 7
4. %OS = 100e−(p/√1 −  ) = 26.11%
2

3
5. tS = 4t = = 1.33s (3% criteria)
wn

74 TEL 306/05 Control Systems

Summary of Unit 2

Summary

1. System response analysis is conducted to validate the system

models and understand the system behaviours.

2. Various input force functions could be applied to the system

for the analysis but the unit focused mainly on four forms
of input force functions which are commonly adopted, namely
impulse, step, ramp and parabolic inputs.

3. Actual systems are unavoidably subjected to disturbances at

various levels. A more realistic system model could be
obtained when the effects of disturbances are considered.

4. The time response can be analysed in terms of transient

response and steady state response.

5. For a unit step input, a first order system exhibits a non-

overshooting exponential time response. The response speed
is determined by the time constant, t. There is one pole

1
located at − on the s-plane for a stable first order system.
t
The further away this pole is from the origin, the faster is the
response time.

UNIT 2 System response analysis 75

 6. For a second order system, the time response is depending
on the damping ratio, Ϛ. The system with the damping ratio
0 < Ϛ < 1 exhibits oscillatory response and with two complex
conjugates poles. When the damping ratio Ϛ ≥ 1, the system
is with two real poles and shows non-overshooting response
as in the first order system. The higher the damping ratio,
the more sluggish is the system response.

7. The steady state response of both first order and second

order systems are similar and could be accessed by referring
to the steady state error.

8. There are two factors that define the steady state error,
namely the input force functions and the type of the system.

9. The concept of dominant poles is important when analysing

higher order system. For example, if the third pole of a third

order system is located at

| |
1

≥10|Ϛwn|, then the system

can be approximated as a second order system.

10. The system response is reliant on the pole placement at the

s-plane. A desired system response could be achieved by
proper adjustment of the closed loop gain that alters the pole
placement.

76 TEL 306/05 Control Systems

References
Nise, N. S. (2011). Control systems engineering (6th ed.). Hoboken, N.J.:
Wiley.

Richard, C. D. & Robert, H. B. (2011). Modern control systems (12th ed.).

Boston: Pearson.

UNIT 2 System response analysis 77

COURSE TEAM
Course Team Coordinator: Dr. Magdalene Goh Wan Ching
Content Writer: Dr. Robin Tan Leong Ping
Instructional Designer: Ms. Jeanne Chow
Academic Member: Dr. Mohd Hezri Bin Marzaki

COURSE COORDINATOR
Dr. Magdalene Goh Wan Ching

EXTERNAL COURSE ASSESSOR

Associate Professor Dr. Yeap Kim Ho, Universiti Tunku Abdul Rahman

PRODUCTION
In-house Editor: Ms. Jeanne Chow
Graphic Designer: Ms. Valerie Ooi

Wawasan Open University is Malaysia’s first private not-for-profit tertiary institution dedicated to
adult learners. It is funded by the Wawasan Education Foundation, a tax-exempt entity established
by the Malaysian People’s Movement Party (Gerakan) and supported by the Yeap Chor Ee Charitable
and Endowment Trusts, other charities, corporations, members of the public and occasional grants
from the Government of Malaysia.

The course material development of the university is funded by Yeap Chor Ee Charitable and
Endowment Trusts.

© 2019 Wawasan Open University

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without prior written permission from WOU.

Wawasan Open University [DU013(P)]

Wholly owned by Wawasan Open University Sdn. Bhd. (700364-W)
54, Jalan Sultan Ahmad Shah, 10050 Penang.
Tel: (604) 2180333 Fax: (604) 2279214
Email: enquiry@wou.edu.my
Website: www.wou.edu.my