Module-1

Define control system: Control means regulate or direct or command. System means
interconnection of components. In a system when the output quantity is controlled by
varying the input quantity, then the system is called control system. The output quantity
is called controlled variable and input quantity is called command signal.
Example: Automobile driving system: In automobile driving control system, speed of
the automobile is a function of the position of its accelerator. The desired speed can be
maintained by controlling pressure on the accelerator pedal. Figure1 shows the general
diagrammatic representation of an automobile driving control system. In automobile
driving control system, force on the accelerator pedal is the input (command) signal,
which increase or decrease the fuel flow to the engine bringing the engine speed to the
desired value.

The diagrammatic representation of the figure 1 is known as block diagram

representation wherein each block represents an element, a plant, a mechanism, device
etc. Each block has an input signal and output signal.
Types of control system: Control system may close loop control system or open loop
control system.
Closed loop control system: In closed loop control system, its control action somehow
depends on the output signal. The closed loop control system is also known as feedback
control system.
Example for closed-loop control system
1. Automobile driving system
2. Refrigerators
3. Person walking on a road with eyes open
 4. Armature controlled Dc motor
5. Automatic water tank level control, etc
Automobile driving system
Let us reconsider the automobile driving system. The route, speed and acceleration of the
automobile are determined and controlled by the driver by observing the traffic and road
condition. Suppose the driver wants to maintain a speed of 50 km per hour (desired
output). The actual speed of the automobile is measured by the speedometer, if there is
deviation in desired speed, accordingly he takes the decision to increase or decrease the
speed. The decision is executed by change in pressure of his foot on the accelerator pedal.
These operations can be represented by block diagram as shown in figure 2. The
information about the instantaneous state of the output is feedback to the input and is
used to modify it in such a manner as to achieve the desired output. In this control
system, involves continuous manual control by a human operator, thus it is a manually
controlled closed loop control system or feedback control system.

In many complex and fast acting systems, the presence of human intervention in the
control system is undesirable. In such situations use of some equipment which performs
control action. A control system incorporating such equipment is known as automatic
control system. In fact in most situations automatic control systems perform the intended
function better than a human operator. An automatic control system could be made to
perform functions where the human operator would be impossible to do.
The general block diagram of an automatic control system which is characterized by a
feedback loop is as shown in figure 3. An error detector compares a signal obtained
through feedback element which is a function of the output signal with the reference
input. Any difference between these two signals error detector produces an error signal,
which actuates the control elements. The control element in turn alters the condition in
the plant in such a manner as to reduce the error.
Automatic tank level control system
To have a better understanding of automatic control system, let us discus a simple liquid
level control system as shown in figure 4. Let h be the actual height of the liquid level,
and H be the desired liquid lever. This automatic control system can maintain the liquid
level h (controlled output) of the tank within the accurate tolerance of the desired level
even though the output flow rate through the valve 𝑉1 is varied. The actual liquid level is
sensed by a float (feedback path element), which positions the slider arm B on a
potentiometer. The slider arm A of another potentiometer is positioned corresponding to
the desired liquid level H (the reference input). When liquid level rises or falls, the
potentiometer (error detector) gives an error voltage (error, or actuating signal)
proportional to the change in liquid level. The error voltage actuates the motor through a
power amplifier which increase or decrease the opening of the valve 𝑉2 in order to restore
the desired liquid level. Thus the control system automatically attempts to correct any
deviation from actual and desired liquid level.
Open-loop control system
In open-loop control system, Control action is independent of output. Or output has no
effect on the control action, the block diagram of control system as shown in figure 5.
The fundamental difference between open and closed loop is that feedback action. In
open-loop control system there is no feedback control action.
Example for open-loop control system
1. Traffic flow control system
2. Person walking on a road with closed eyes
3. Automatic washing machine
4. Field controlled DC motor

Traffic flow control system

Example for open loop control system is traffic control system for regulating the flow of
traffic at the crossing of two roads. This traffic controlled system is open loop control
system, because the light ON-OFF timer mechanism is set for predetermined fixed
intervals of time. The open-loop control system becomes closed-loop control system in
which the rates of traffic flow along both directions are measured and are compared, and
the difference is used to control the timing of the lights.
Person walking on a road with closed eyes
Another example for open-loop control is, the person walking on the road with closed
eyes. The person walking on the road with open eye is a closed loop control system,
because the control action is a function of the output. If the eyes are closed, the feedback
loop is broken and the system become open-loop.
Open-loop versus close loop
Sl.No Open-loop Closed-loop
1 No feedback is used. Feedback is there to compare the actual
output with the reference input
2 Open-loop systems are generally Closed loop system stability depends on the
stable design of the system
3 Construction and design are simple, Complicated design, and expensive
thus less expensive
4 Highly sensitive to the external Less sensitive to the external disturbance
disturbance
5 Error detector is absent Error detector is necessary

Mathematical model of physical systems

We know that control system is an interconnection of physical component to obtain the
desired output with the reference input. A physical component may electrical component,
electronic component, or hydraulic component etc. A collection of physical components
forms a physical system. An idealized physical system is called physical model. Once we
have a physical model of a physical system, the next step is to obtain a mathematical
model. Using Newton’s law and Kirchhoff’s laws, etc, will be used to build mathematical
models. . Mathematical models of most physical systems are represented by differential
equations. These differential equations are solved using Laplace transform.
Differential equation of physical systems
Mechanical systems
Systems containing mechanical elements are known as mechanical systems. These
mechanical systems can be classified as translations mechanical system and rotational
mechanical system. In translational system, body moves along the straight line and in
rotational system body moves around a fixed axis.
Three elements in translations mechanical system are mass M, spring constant K and
damper f. And three elements in rotational mechanical system are inertia J, torsional
spring constant K and damper f. Figure 6 shows the ideal elements in the mechanical
system. Note that mass M, inertia J and spring K are energy storage elements in the
mechanical system and damper consumes the energy in the mechanical system.

Problem 1: A mass-spring-friction mechanical system under equilibrium condition as

shown in below figure. Write the differential equation for the mechanical system.
Solution: The system equation during the equilibrium condition using D’Alembert’s
principal
𝑑 2 𝑥(𝑡) 𝑑𝑥(𝑡)
𝐹(𝑡) = 𝑀 + 𝐵 + 𝑘𝑥(𝑡)
𝑑𝑡 2 𝑑𝑡
The above equation is second order linear constant coefficient differential equation.
Problem 2: An inertia-spring-friction mechanical system under equilibrium condition as
shown in below figure. Write the differential equation for the mechanical system.

Solution: The system equation during the equilibrium condition using D’Alembert’s
principal
𝑑 2 𝜃(𝑡) 𝑑𝜃(𝑡)
𝑇(𝑡) = 𝐽 2
+𝐵 + 𝑘𝜃(𝑡)
𝑑𝑡 𝑑𝑡
Problem 1: For the mechanical system, write the system equation and find the transfer
𝑋1 (𝑠) 𝑋2 (𝑠)
function and .
𝐹(𝑠) 𝐹(𝑠)
Solution: Transfer function is defined as the ratio of Laplace transfer of the output
variable to the Laplace transfer of the input variable. The mechanical system and its
mechanical network as shown in below figure. Always number of variables
(displacement) is equal to number of masses.

At variable 𝑥1 (𝑡)
𝑑 2 𝑥1 (𝑡) 𝑑𝑥1 (𝑡)
𝑓(𝑡) = 𝑀1 2
+ 𝑓1 +𝑘1 𝑥1 (𝑡) + 𝑘2 [𝑥1 (𝑡) − 𝑥2 (𝑡)]
𝑑𝑡 𝑑𝑡
At variable𝑥2 (𝑡)
𝑑 2 𝑥2 (𝑡)
0 = 𝑀2 + 𝑘2 [(𝑥2 (𝑡) − 𝑥1 (𝑡)]
𝑑𝑡 2
Applying Laplace transform to the above equation we get
𝐹(𝑠) = 𝑀1 𝑠 2 𝑋1 (𝑠) + 𝑓1 𝑠𝑋1 (𝑠) + 𝑘1 𝑋1 (𝑠) + 𝑘2 [𝑋1 (𝑠) − 𝑋2 (𝑠)]
0 = 𝑀2 𝑠 2 𝑋2 (𝑠) + 𝑘2 [𝑋2 (𝑠) − 𝑋1 (𝑠)]
𝑀 𝑠 2 + 𝑓1 𝑠+(𝑘1 + 𝑘2 ) −𝑘2 𝑋1 (𝑠)
[𝐹(𝑠)] = [ 1 ][ ]
0 −𝑘2 𝑀2 𝑠 2 + 𝑘2 𝑋2 (𝑠)
 𝐹(𝑠) −𝑘2
[ ]
0 𝑀2 𝑠2 +𝑘2
Solving for 𝑋1 (𝑠), 𝑋1 (𝑠) = ∆

Where ∆= {𝑀1 𝑠 + 𝑓1 𝑠+(𝑘1 + 𝑘2 ) }{𝑀2 𝑠 2 + 𝑘2 }

2
– {−𝑘2 × −𝑘2 }
𝐹(𝑠)[𝑀2 𝑠 2 + 𝑘2 ]
𝑋1 (𝑠) =
{𝑀1 𝑠 2 + 𝑓1 𝑠+(𝑘1 + 𝑘2 ) }{𝑀2 𝑠 2 + 𝑘2 } – {−𝑘2 × −𝑘2 }
𝑋1 (𝑠) 𝑀2 𝑠 2 + 𝑘2
=
𝐹(𝑠) {𝑀1 𝑠 2 + 𝑓1 𝑠+(𝑘1 + 𝑘2 ) }{𝑀2 𝑠 2 + 𝑘2 } – {−𝑘2 × −𝑘2 }
And
𝑀1 𝑠 2 + 𝑓1 𝑠+(𝑘1 + 𝑘2 )) 𝐹(𝑠)
[ ]
−𝑘2 0 −𝐹(𝑠){−𝑘2 }
𝑋2 (𝑠) = =
∆ ∆
−𝐹(𝑠){−𝑘2 }
𝑋2 (𝑠) =
{𝑀1 𝑠 + 𝑓1 𝑠+(𝑘1 + 𝑘2 ) }{𝑀2 𝑠 2 + 𝑘2 } – {−𝑘2 × −𝑘2 }
2

𝑋2 (𝑠) 𝑘2
=
𝐹(𝑠) {𝑀1 𝑠 + 𝑓1 𝑠+(𝑘1 + 𝑘2 ) }{𝑀2 𝑠 2 + 𝑘2 }
2 – {−𝑘2 × −𝑘2 }

Problem 2: For the mechanical system, write the system equation and find the transfer
𝑋1 (𝑠)
function .
𝐹(𝑠)

Solution:
At variable 𝑥1 (𝑡)
𝑑 2 𝑥1 (𝑡) 𝑑𝑥1 (𝑡) 𝑑(𝑥1 (𝑡) − 𝑥2 (𝑡))
𝑓(𝑡) = 𝑘1 𝑥1 (𝑡) + 𝑀1 2
+ 𝐵1 + 𝐵3
𝑑𝑡 𝑑𝑡 𝑑𝑡
At variable 𝑥2 (𝑡)
𝑑2 𝑥2 (𝑡) 𝑑𝑥2 (𝑡) 𝑑(𝑥2 (𝑡) − 𝑥1 (𝑡))
0 = 𝑀2 2
+ 𝐵2 + 𝐵3 + 𝑘2 𝑥2 (𝑡)
𝑑𝑡 𝑑𝑡 𝑑𝑡
Apply the Laplace transform to the system equation we get
𝐹(𝑠) = 𝑘1 𝑋1 (𝑠) + 𝑀1 𝑠 2 𝑋1 (𝑠) + 𝐵1 𝑠𝑋1 (𝑠) + 𝐵3 𝑠(𝑋1 (𝑠) − 𝑋2 (𝑠))
0 = 𝑀2 𝑠 2 𝑋2 (𝑠) + 𝐵2 𝑠𝑋2 (𝑠) + 𝐵3 𝑠(𝑋2 (𝑠) − 𝑋1 (𝑠)) + 𝑘2 𝑋2 (𝑠)
𝑘1 + 𝑀1 𝑠 2 + 𝐵1 𝑠 ) + 𝐵3 𝑠 −𝐵3 𝑠 𝑋 (𝑠)
[𝐹(𝑠)] = [ 2 ][ 1 ]
0 −𝐵3 𝑠 𝑀2 𝑠 + 𝐵2 𝑠 + 𝐵3 𝑠 + 𝑘2 𝑋2 (𝑠)
𝐹(𝑠) −𝐵3 𝑠
[ ]
0 𝑀2 𝑠2 +𝐵2 𝑠+𝐵3 𝑠+𝑘2
𝑋1 (𝑠) = ∆

Where ∆= [𝑘1 + 𝑀1 𝑠 2 + 𝐵1 𝑠 ) + 𝐵3 𝑠 ] × [𝑀2 𝑠 2 + 𝐵2 𝑠 + 𝐵3 𝑠] − {−𝐵3 𝑠}{−𝐵3 𝑠}

𝑋1 (𝑠)
𝐹(𝑠)( 𝑀2 𝑠 2 + 𝐵2 𝑠 + 𝐵3 𝑠)
=
[𝑘1 + 𝑀1 𝑠 2 + 𝐵1 𝑠 ) + 𝐵3 𝑠 ] × [𝑀2 𝑠 2 + 𝐵2 𝑠 + 𝐵3 𝑠 + 𝑘2 ] − {−𝐵3 𝑠}{−𝐵3 𝑠}
Problem3: The rotational mechanical system as shown in figure below. Write the system
equation and find the transfer function of the system.
Solution: The mechanical system and its network as shown in below figure. Always
number of variables (displacement) is equal to number of inertia.

At variable 𝜃1 (𝑡)
𝑑 2 𝜃1 (𝑡)
𝑓(𝑡) = 𝐽1 + 𝑘1 [𝜃1 (𝑡) − 𝜃2 (𝑡)]
𝑑𝑡 2
At variable 𝜃2 (𝑡)
𝑑2 𝜃2 (𝑡)
0 = 𝐽2 + 𝑘1 [𝜃2 (𝑡) − 𝜃1 (𝑡)] + 𝑘2 𝜃2 (𝑡)
𝑑𝑡 2
Apply the Laplace transform to the set of above equations we get
𝐹(𝑠) = 𝐽1 𝑠 2 𝜃1 (𝑠) + 𝑘1 [𝜃1 (𝑠) − 𝜃2 (𝑠)]
0 = 𝐽2 𝑠 2 𝜃2 (𝑠) + 𝑘1 [𝜃2 (𝑠) − 𝜃1 (𝑠)] + 𝑘2 𝜃2 (𝑠)
𝑘1 + 𝐽1 𝑠 2 + 𝑘1 −𝑘1 𝜃 (𝑠)
[𝑇(𝑠)] = [ 2 ][ 1 ]
0 −𝑘1 𝐽2 𝑠 + 𝑘1 + 𝑘2 𝜃2 (𝑠)

𝑇(𝑠) −𝑘1
[ ]
0 𝐽2 𝑠2 +𝑘1 +𝑘2 𝑇(𝑠)[𝐽2 𝑠2 +𝑘1 +𝑘2 ]
𝜃1 (𝑠) = = And Transfer function
∆ ∆

𝜃1 (𝑠) [𝐽2 𝑠 2 + 𝑘1 + 𝑘2 ]
=
𝑇(𝑠) ∆
𝑘 +𝐽 𝑠2 +𝑘1 𝑇(𝑠)
[ 1 1 ]
−𝑘1 0 −𝑇(𝑠)[−𝑘1 ]
𝜃2 (𝑠) = = And Transfer function
∆ ∆
 𝜃2 (𝑠) 𝑘1
=
𝑇(𝑠) ∆
Where ∆= [𝑘1 + 𝐽1 𝑠 2 + 𝑘1 ] × [𝐽2 𝑠 2 + 𝑘1 + 𝑘2 ] − [−𝑘1 ] × [−𝑘1 ]
Problem4: The rotational mechanical system as shown in figure below. Write the system
𝜃(𝑠)
equation and find the transfer function .
𝑇(𝑠)

Solution:

At variable 𝜃1 (𝑡)
𝑑2 𝜃1 (𝑡) 𝑑(𝜃1 (𝑡) − 𝜃(𝑡))
𝑇(𝑡) = 𝐽1 2
+ 𝑘1 [𝜃1 (𝑡) − 𝜃(𝑡)] + 𝐵1
𝑑𝑡 𝑑𝑡
At variable 𝜃(𝑡)
𝑑2 𝜃(𝑡)
0 = 𝐽2 + 𝑘1 [𝜃(𝑡) − 𝜃1 (𝑡)] + 𝐵2 [𝜃(𝑡) − 𝜃1 (𝑡)]
𝑑𝑡 2
Apply the Laplace transform to the set of above equations we get
𝑇(𝑠) = 𝐽1 𝑠 2 𝜃1 (𝑠) + 𝑘1 [𝜃1 (𝑠) − 𝜃(𝑠)] + 𝐵1 𝑠[𝜃1 (𝑠) − 𝜃(𝑠)]
 0 = 𝐽2 𝑠 2 𝜃(𝑠) + 𝑘1 [𝜃(𝑠) − 𝜃1 (𝑠)] + 𝐵2 [𝜃(𝑠) − 𝜃1 (𝑠)
𝐽1 𝑠 2 + 𝐵1 𝑠 + 𝑘1 −𝑘1 − 𝐵1 𝑠 𝜃 (𝑠)
[𝑇(𝑠)] = [ 2 ][ 1 ]
0 −𝑘1 − 𝐵1 𝑠 𝐽2 𝑠 + 𝑘1 + 𝐵2 𝑠 𝜃(𝑠)
𝐽 𝑠2 +𝐵1 𝑠+𝑘1 𝑇(𝑠)
[ 1 ]
−𝑘1 −𝐵1 𝑠 0 −𝑇(𝑠)[−𝑘1 −𝐵1 𝑠]
𝜃(𝑠) = = And Transfer function
∆ ∆

𝜃(𝑠) 𝑘1 + 𝐵1 𝑠
=
𝑇(𝑠) ∆
Where ∆= [𝐽1 𝑠 2 + 𝐵1 𝑠 + 𝑘1 ] × [𝐽2 𝑠 2 + 𝑘1 + 𝐵2 𝑠] − [−𝑘1 − 𝐵1 𝑠] × [−𝑘1 − 𝐵1 𝑠]
Electrical systems
Systems containing electrical elements are known as electrical systems or electrical
circuit. The basic elements of the electrical circuits are resistor, inductor and capacitor.
The inductor and capacitor stores the energy and resistor consumes the energy. These
electrical circuits may be series circuit or parallel circuit. These electrical circuits are
analyzed using Kirchhoff’s voltage current law or current. The L-R-C series circuit is
shown in figure 7. Using Kirchhoff’s voltage law, the governing equation is
𝑑𝑖(𝑡) 1 𝑡
𝑒(𝑡) = 𝐿 + 𝑅𝑖(𝑡) + 𝐶 ∫∞ 𝑖(𝑡)𝑑𝑡
𝑑𝑡
𝑑𝑞(𝑡)
We know that current 𝑖(𝑡) = 𝑑𝑡

𝑑2 𝑞(𝑡) 𝑑𝑞(𝑡) 𝑞(𝑡)

∴ 𝑒(𝑡) = 𝐿 2
+𝑅 +
𝑑𝑡 𝑑𝑡 𝐶
𝑑2 𝑞 𝑑𝑞 𝑞
Or 𝑒 = 𝐿 + 𝑅 𝑑𝑡 + 𝐶
𝑑𝑡 2

Similarly, consider L-R-C parallel circuit is shown in figure 8. Using Kirchhoff’s current
law, the governing equation is
𝑑𝑒(𝑡) 𝑒(𝑡) 1 𝑡
𝑖(𝑡) = 𝐶 + + ∫ 𝑒(𝑡)𝑑𝑡
𝑑𝑡 𝑅 𝐿 ∞
 𝑑𝜑(𝑡) 𝑡
We know that emf due to magnetic flux is 𝑒(𝑡) = and 𝜑(𝑡) = ∫∞ 𝑒(𝑡)𝑑𝑡
𝑑𝑡
2
𝑑 𝜑(𝑡) 1 𝑑𝜑(𝑡) 1
𝑖(𝑡) = 𝐶 + + 𝜑(𝑡)
𝑑𝑡 2 𝑅 𝑑𝑡 𝐿
Analogous system
Consider the mechanical and electrical circuit as shown in figure 9, and the system
equations are
𝑑2 𝑥(𝑡) 𝑑𝑥(𝑡) 𝑑2𝑞 𝑑𝑞 𝑞
𝐹=𝑀 2
+𝑓 + 𝑘𝑥(𝑡) 𝑎𝑛𝑑 𝑒 =𝐿 2
+𝑅 +
𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡 𝐶
𝑑𝑞 𝑑2 𝑞 𝑑𝑞
= 𝑖(𝑡) And =
𝑑𝑡 𝑑𝑡 2 𝑑𝑡

𝑑𝑖 1
∴𝑒 =𝐿 + 𝑅𝑖 + ∫ 𝑖𝑑𝑡
𝑑𝑡 𝐶

On comparing these two equations, reveals that, the mechanical and electrical systems
differential equations are identical, such systems are called analogous systems. Here
force 𝐹 (torque) 𝑇 and voltage e are the variable. This is called as Force (Torque)-
Voltage analogy. List of analogous variables in the analogy is given in table 1.
Table 1: Analogues quantities in Force (Torque)-Voltage analogy
Mechanical translational Mechanical rotational system Electrical system
system
Force F Torque T Voltage e
Mass M Moment of inertia J Inductance L
Viscous friction Viscous friction coefficient f or B Resistance R
coefficient for B
Spring stiffness K Torsional spring stiffness K Reciprocal of capacitance 1/C
Linear displacement x Angular displacement 𝜃 Charge q
𝑑𝑥(𝑡) 𝑑𝜃(𝑡) Current i
𝑑𝑡 𝑑𝑡
Similarly, consider the mechanical and electrical circuit as shown in figure 10, and the
system equations are
𝑑2 𝑥(𝑡) 𝑑𝑥(𝑡)
𝐹=𝑀 2
+𝑓 + 𝑘𝑥(𝑡) 𝑎𝑛𝑑 𝑖
𝑑𝑡 𝑑𝑡
𝑑 2 𝜑(𝑡) 1 𝑑𝜑(𝑡) 1
=𝐶 + + 𝜑(𝑡)
𝑑𝑡 2 𝑅 𝑑𝑡 𝐿
𝑑𝜑(𝑡) 𝑡
We know that, emf due to magnetic flux is 𝑒(𝑡) = and 𝜑(𝑡) = ∫∞ 𝑒(𝑡)𝑑𝑡
𝑑𝑡

𝑑𝑒(𝑡) 𝑒(𝑡) 1
∴ 𝑖=𝐶 + + ∫ 𝑒(𝑡)𝑑𝑡
𝑑𝑡 𝑅 𝐿

On comparing these two equations, reveals that, the mechanical and electrical systems
differential equations are identical, such systems are called analogous systems. Here
force 𝐹 (torque) 𝑇 and current 𝑖 are the variable. This is called as Force (Torque)-
Current analogy. List of analogous variables in the analogy is given in table 2.
Table 2: Analogues quantities in Force (Torque)-Current analogy
Mechanical translational Mechanical rotational system Electrical system
system
Force F Torque T Current i
Mass M Moment of inertia J capacitance C
Viscous friction Viscous friction coefficient f or Reciprocal of resistance1/ R
coefficient f or B B
Spring stiffness K Torsional spring stiffness K Reciprocal of capacitance 1/L
Linear displacement x Angular displacement 𝜃 Magnetic flux linkage
𝜆
𝑑𝑥(𝑡) 𝑑𝜃(𝑡) Voltage e
𝑑𝑡 𝑑𝑡
Problem 5: Write the differential equation governing the mechanical system shown in
below figure. Draw the electrical circuit for force-voltage and force current analogy.

Solution: At variable 𝑥1 (𝑡)

𝑑2 𝑥1 (𝑡) 𝑑𝑥1 (𝑡) 𝑑(𝑥1 (𝑡) − 𝑥2 (𝑡))
𝑓(𝑡) = 𝑀1 + 𝐵2 + 𝐵1
𝑑𝑡 2 𝑑𝑡 𝑑𝑡
At variable 𝑥2 (𝑡)
𝑑(𝑥2 (𝑡) − 𝑥1 (𝑡)) 𝑑 2 𝑥2 (𝑡)
0 = 𝐵1 + 𝑀2 + 𝑘1 𝑥2 (𝑡)
𝑑𝑡 𝑑𝑡 2

F-V analogy refer table 1, from the table,

Force is replaced by voltage
Mass is replaced by inductance
Viscous friction is replaced by resistance
Spring is replaced by reciprocal of capacitance.
𝑑𝑥(𝑡)
is replaced by i
𝑑𝑖
The integro differential equations are
𝑑𝑖1 (𝑡)
𝑣(𝑡) = 𝐿1 + 𝑅2 𝑖1 (𝑡) + 𝑅1 [𝑖1 (𝑡) − 𝑖2 (𝑡)]
𝑑𝑡
𝑑𝑖2 (𝑡) 1
0 = 𝐿2 + ∫ 𝑖2 (𝑡)𝑑𝑡 + 𝑅1 [𝑖2 (𝑡) − 𝑖1 (𝑡)]
𝑑𝑡 𝐶1
F-I analogy refer table 2, from the table
Force is replaced by current
Mass is replaced by capacitance
Viscous friction is replaced by reciprocal of resistance
Spring is replaced by reciprocal of inductance.
𝑑𝑥(𝑡)
is replaced by e
𝑑𝑖

The integro differential equations are

𝑑𝑣1 (𝑡)
At 𝑣1 (𝑡) 𝑖(𝑡) = 𝐶1 + 𝐺2 𝑣1 (𝑡) + 𝐺1 (𝑣1 (𝑡) − 𝑣2 (𝑡))
𝑑𝑡
𝑑𝑣2 (𝑡) 1
At 𝑣2 (𝑡) 0 = 𝐶2 + 𝐿 ∫ 𝑣2 (𝑡)𝑑𝑡 + 𝐺1 (𝑣2 (𝑡) − 𝑣1 (𝑡))
𝑑𝑡 1

Problem 5: Write the differential equation governing the mechanical system shown in
below figure. Draw the electrical circuit for force-current analogy.
Solution:

𝑑2 𝑥1 (𝑡) 𝑑𝑥1 (𝑡) 𝑑𝑥1 (𝑡)

At node 𝑥1 (𝑡), 𝑓(𝑡) = 𝑀1 + 𝐵1 + 𝑘1 𝑥1 (𝑡) + 𝐵2 + 𝑘2 [𝑥1 (𝑡) − 𝑥2 (𝑡)]
𝑑𝑡 2 𝑑𝑡 𝑑𝑡
𝑑2 𝑥2 (𝑡) 𝑑𝑥1 (𝑡)
At node 𝑥2 (𝑡), 0 = 𝑘2 (𝑥2 (𝑡) − 𝑥1 (𝑡)) + 𝑀2 + 𝐵3
𝑑𝑡 2 𝑑𝑡

F-I analogy refer table 2, from the table, Force is replaced by current, mass is replaced
by capacitance, viscous friction is replaced by reciprocal of resistance and spring is
replaced by reciprocal of inductance.

The integro differential equations are

At node 𝑣1 (𝑡)
𝑑𝑣1 (𝑡) 1 1
𝑖(𝑡) = 𝐶1 + 𝐺1 𝑣1 (𝑡) + 𝐺2 𝑣1 (𝑡) + 𝐿 ∫ 𝑣1 (𝑡) 𝑑𝑡 + 𝐿 ∫[𝑣1 (𝑡) − 𝑣2 (𝑡)]𝑑𝑡
𝑑𝑡 1 2

1 𝑑𝑣2 (𝑡)
At node 𝑣2 (𝑡) 0 = 𝐿 ∫[𝑣2 (𝑡) − 𝑣1 (𝑡)]𝑑𝑡 + 𝐶2 + 𝐺3 𝑣2 (𝑡)
2 𝑑𝑡

Problem 6: Write the differential equation governing the mechanical system shown in
below figure. Draw the electrical circuit for force-voltage analogy.
Solution:

𝑑2 𝑥1 (𝑡) 𝑑[𝑥1 (𝑡)−𝑥2 (𝑡)]

At node 𝑥1 (𝑡), 𝑓(𝑡) = 𝑀1 + 𝑘1 𝑥1 (𝑡) + 𝐵 + 𝑘[𝑥1 (𝑡) − 𝑥2 (𝑡)]
𝑑𝑡 2 𝑑𝑡
𝑑[𝑥2 (𝑡)−𝑥1 (𝑡)] 𝑑2 𝑥2 (𝑡)
At node 𝑥2 (𝑡), 0 = 𝑘[𝑥1 (𝑡) − 𝑥2 (𝑡)] + 𝐵 +𝑀2 + 𝑘2 𝑥2 (𝑡)
𝑑𝑡 𝑑𝑡 2

F-V analogy refer table 1, from the table, Force is replaced by voltage, mass is replaced
by inductance, viscous friction is replaced by resistance and spring is replaced by
reciprocal of capacitance.
𝑑𝑖1 (𝑡) 1 1
𝑣(𝑡) = 𝐿1 + ∫ 𝑖1 (𝑡)𝑑𝑡 + 𝑅[𝑖1 (𝑡) − 𝑖2 (𝑡 )] + ∫[𝑖1 (𝑡) − 𝑖2 (𝑡)]𝑑𝑡
𝑑𝑡 𝐶1 𝐶
1 𝑑𝑖2 (𝑡) 1
0 = ∫[𝑖2 (𝑡) − 𝑖1 (𝑡)]𝑑𝑡 + 𝑅[𝑖2 (𝑡) − 𝑖1 (𝑡)]+𝐿2 + ∫ 𝑖2 (𝑡)𝑑𝑡
𝐶 𝑑𝑡 𝐶2
Problem: Find the translational mechanical system for the force voltage electrical circuit
shown in below figure.

Solution:
The loop equations are
𝑑𝑖1 (𝑡) 1 1
At first loop 𝑒(𝑡) = 𝐿1 + 𝑅1 𝑖1 (𝑡) + 𝐶 ∫ 𝑖1 (𝑡)𝑑𝑡 + 𝐶 ∫[𝑖1 (𝑡) − 𝑖2 (𝑡)]𝑑𝑡 +
𝑑𝑡 1 2

1
𝐶3
∫[𝑖1 (𝑡) − 𝑖3 (𝑡)]𝑑𝑡 … (1)
1
As the second loop 0 = 𝐶 ∫[𝑖3 (𝑡) − 𝑖1 (𝑡)]𝑑𝑡 + 𝑅3 [𝑖3 (𝑡) − 𝑖2 (𝑡)] … … . . (2)
3

𝑑𝑖2 (𝑡) 1
At the third loop 0 = 𝐿2 + 𝑅2 𝑖2 (𝑡) + 𝐶 ∫[𝑖2 (𝑡) − 𝑖1 (𝑡)]𝑑𝑡 + 𝑅3 [𝑖2 (𝑡) −
𝑑𝑡 2

𝑖3 (𝑡)] … (3)
F-V analogy refer table 1, from the table,
Force is replaced by voltage
Mass is replaced by inductance
Viscous friction is replaced by resistance
Spring is replaced by reciprocal of capacitance.
𝑑𝑥(𝑡)
Current i is replaced by 𝑑𝑡
2
𝑑 𝑥1 (𝑡) 𝑑𝑥1 (𝑡)
𝐹(𝑡) = 𝑀1 + 𝐵1 + 𝑘1 𝑥1 (𝑡) + 𝑘2 [𝑥1 (𝑡) − 𝑥2 (𝑡)] + 𝑘3 [𝑥1 (𝑡) − 𝑥3 (𝑡)]
𝑑𝑡 2 𝑑𝑡
 𝑑
0 = 𝑘3 [𝑥3 (𝑡) − 𝑥1 (𝑡) + 𝐵3 𝑑𝑡 [𝑥3 (𝑡) − 𝑥2 (𝑡)] … . (2)
𝑑2 𝑥2 (𝑡) 𝑑𝑥2 (𝑡) 𝑑
0 = 𝑀2 + 𝐵2 + 𝑘2 [𝑥2 (𝑡) − 𝑥1 (𝑡)] + 𝐵3 𝑑𝑡 [𝑥2 (𝑡) − 𝑥3 (𝑡)]
𝑑𝑡 2 𝑑𝑡

Now the mechanical network is

The mechanical system is

Problem 7: Write the differential equation governing the mechanical system shown in
below figure. Draw the electrical circuit for torque-voltage analogy.
Solution: The mechanical network for the system as shown in below figure

𝑑2 𝜃1 (𝑡) 𝑑𝜃1 (𝑡) 𝑑[𝜃1 (𝑡)−𝜃2 (𝑡)]

At node 𝜃1 (𝑡), 𝑇(𝑡) = 𝐽1 + 𝐵3 + 𝑘1 [𝜃1 (𝑡) − 𝜃2 (𝑡)] + 𝐵1
𝑑𝑡 2 𝑑𝑡 𝑑𝑡

At node 𝜃2 (𝑡),
𝑑[𝜃2 (𝑡) − 𝜃1 (𝑡)] 𝑑 2 𝜃2 (𝑡) 𝑑𝜃2 (𝑡)
0 = 𝑘1 [𝜃2 (𝑡) − 𝜃1 (𝑡)] + 𝐵1 + 𝐽2 2
+ 𝐵4 + 𝑘2 [𝜃2 (𝑡)
𝑑𝑡 𝑑𝑡 𝑑𝑡
𝑑[𝜃2 (𝑡) − 𝜃3 (𝑡)]
− 𝜃3 (𝑡)] + 𝐵2
𝑑𝑡
At node 𝜃3 (𝑡),
𝑑[𝜃3 (𝑡) − 𝜃2 (𝑡)] 𝑑 2 𝜃3 (𝑡) 𝑑𝜃3 (𝑡)
0 = 𝑘2 [𝜃3 (𝑡) − 𝜃2 (𝑡)] + 𝐵2 + 𝐽3 2
+ 𝐵5 + 𝑘3 𝜃3 (𝑡)
𝑑𝑡 𝑑𝑡 𝑑𝑡
Torque-Voltage analogy refer table 1, from the table, Torque is replaced by voltage,
inertia is replaced by inductance, viscous friction is replaced by resistance and spring is
replaced by reciprocal of capacitance.
𝑑𝑖1 (𝑡) 1
At the first loop 𝑣(𝑡) = 𝐿1 + 𝑅3 𝑖1 (𝑡) + 𝐶 ∫[𝑖1 (𝑡) − 𝑖2 (𝑡)]𝑑𝑡 + 𝑅1 [𝑖1 (𝑡) − 𝑖2 (𝑡)]
𝑑𝑡 1

1 𝑑𝑖2 (𝑡)
At the second loop 0 = 𝐶 ∫[𝑖2 (𝑡) − 𝑖1 (𝑡)]𝑑𝑡 + 𝑅1 [𝑖2 (𝑡) − 𝑖1 (𝑡)] + 𝐿2 +
1 𝑑𝑡
1
𝑅4 𝑖2 (𝑡) + 𝐶 ∫[𝑖2 (𝑡) − 𝑖3 (𝑡)]𝑑𝑡 + 𝑅2 [𝑖2 (𝑡) − 𝑖3 (𝑡)]
2

At the third loop

 1 𝑑𝑖3 (𝑡)
0 = ∫[𝑖3 (𝑡) − 𝑖2 (𝑡)]𝑑𝑡 + 𝑅2 [𝑖3 (𝑡) − 𝑖2 (𝑡)] + 𝐿3 + 𝑅5 𝑖3 (𝑡)
𝐶2 𝑑𝑡
1
+ ∫ 𝑖 (𝑡)𝑑𝑡
𝐶3 3

Torque-Current analogy refer table 2, from the table, torque is replaced by current,
mass is replaced by capacitance, viscous friction is replaced by reciprocal of resistance
and spring is replaced by reciprocal of inductance.
𝑑𝑣1 (𝑡) 1
At node 𝑣1 (𝑡), 𝑖(𝑡) = 𝐶1 + 𝐺3 𝑣1 (𝑡) + 𝐿 ∫[𝑣1 (𝑡) − 𝑣2 (𝑡)]𝑑𝑡 + 𝐺1 [𝑣1 (𝑡) − 𝑣2 (𝑡)]
𝑑𝑡 1

At node 𝑣2 (𝑡),
1 𝑑𝑣2 (𝑡)
0 = ∫[𝑣2 (𝑡) − 𝑣1 (𝑡)]𝑑𝑡 + 𝐺1 [𝑣2 (𝑡) − 𝑣1 (𝑡)] + 𝐶2 + 𝐺4 𝑣2 (𝑡)
𝐿1 𝑑𝑡
1
+ ∫[𝑣2 (𝑡) − 𝑣3 (𝑡)] 𝑑𝑡 + 𝐺2 [𝑣2 (𝑡) − 𝑣3 (𝑡)]
𝐿2
At node 𝑣3 (𝑡),
1 𝑑𝑣3 (𝑡)
0 = ∫[𝑣3 (𝑡) − 𝑣2 (𝑡)]𝑑𝑡 + 𝐺2 [𝑣3 (𝑡) − 𝑣2 (𝑡)] + 𝐶3 + 𝐺5 𝑣3 (𝑡)
𝐿2 𝑑𝑡
1
+ ∫ 𝑣3 (𝑡) 𝑑𝑡
𝐿3
Problem 8: Write the differential equation governing the mechanical system shown in
below figure. Draw the electrical circuit for force-voltage analogy.

Solution: The mechanical system with its mechanical network as shown in below figure

𝑑2 𝑥1 (𝑡) 𝑑𝑥1 (𝑡)

At node 𝑥1 (𝑡), 𝑓(𝑡) = 𝑀1 + 𝐵1 + 𝑘1 𝑥1 (𝑡) + 𝑘2 [𝑥1 (𝑡) − 𝑥2 (𝑡)] +
𝑑𝑡 2 𝑑𝑡
𝑑[𝑥1 (𝑡)−𝑥2 (𝑡)]
𝐵2 𝑑𝑡

At node 𝑥2 (𝑡),
𝑑[𝑥2 (𝑡) − 𝑥1 (𝑡)] 𝑑 2 𝑥2 (𝑡)
0 = 𝑘2 [𝑥2 (𝑡) − 𝑥1 (𝑡)] + 𝐵2 + 𝑀2
𝑑𝑡 𝑑𝑡 2
Force-Voltage analogy refer table 1, from the table, Force is replaced by voltage, mass
is replaced by inductance, viscous friction is replaced by resistance and spring is replaced
by reciprocal of capacitance.
At the first loop
𝑑𝑖1 (𝑡) 1 1
𝑣(𝑡) = 𝐿1 + 𝑅1 𝑖1 (𝑡) + ∫ 𝑖1 (𝑡) 𝑑𝑡 + ∫[𝑖1 (𝑡) − 𝑖2 (𝑡)] 𝑑𝑡 + 𝑅2 [𝑖1 (𝑡)
𝑑𝑡 𝐶1 𝐶2
− 𝑖2 (𝑡)]
At the second loop,
 1 𝑑𝑖2 (𝑡)
0 = ∫[𝑖2 (𝑡) − 𝑖1 (𝑡)] 𝑑𝑡 + 𝑅2 [𝑖2 (𝑡) − 𝑖1 (𝑡)] + 𝐿2
𝐶2 𝑑𝑡
The F-V analogy electrical network is

Force-Current analogy refer table 2, from the table, force is replaced by current, mass
is replaced by capacitance, viscous friction is replaced by reciprocal of resistance and
spring is replaced by reciprocal of inductance.
𝑑𝑣1 (𝑡) 1 1
At node 𝑣1 (𝑡), 𝑖(𝑡) = 𝐶1 + 𝐺1 𝑣1 (𝑡) + 𝐿 ∫ 𝑣1 (𝑡)𝑑𝑡 + [𝑣1 (𝑡) − 𝑣2 (𝑡)] +
𝑑𝑡 1 𝐿2

𝐺2 [𝑣1 (𝑡) − 𝑣2 (𝑡)]

1 𝑑𝑣2 (𝑡)
At node 𝑣2 (𝑡), 0 = 𝐿 [𝑣2 (𝑡) − 𝑣1 (𝑡)] + 𝐺2 [𝑣2 (𝑡) − 𝑣1 (𝑡)] + 𝐶2
2 𝑑𝑡

The F-current analogy electrical network is

Problem 9: Write the differential equation mechanical system shown in below figure.
Draw the electrical circuit for torque-voltage analogy.

Solution: For the mechanical system, its mechanical network as shown in below figure
At node 𝜃1 (𝑡), 𝑇(𝑡) = 𝑘1 [𝜃1 (𝑡) − 𝜃2 (𝑡)]
𝑑2 𝜃2 (𝑡) 𝑑(𝜃2 (𝑡)−𝜃3 (𝑡)) 𝑑𝜃2 (𝑡)
At node 𝜃2 (𝑡) 0 = 𝑘1 [𝜃2 (𝑡) − 𝜃1 (𝑡)] + 𝐽1 + 𝐵3 + 𝐵1
𝑑𝑡 2 𝑑𝑡 𝑑𝑡
𝑑(𝜃3 (𝑡)−𝜃2 (𝑡)) 𝑑𝜃3 (𝑡) 𝑑2 𝜃3 (𝑡)
At node 𝜃3 (𝑡), 0 = 𝐵3 + 𝑘2 𝜃3 (𝑡) + 𝐵2 + 𝐽2
𝑑𝑡 𝑑𝑡 𝑑𝑡 2

Torque-Voltage analogy refer table 1, from the table, Torque is replaced by voltage,
inertia is replaced by inductance, viscous friction is replaced by resistance and spring is
replaced by reciprocal of capacitance.
1
At the first loop, 𝑣(𝑡) = 𝐶 ∫[𝑖1 (𝑡) − 𝑖2 (𝑡)]𝑑𝑡
1

1 𝑑𝑖2 (𝑡)
At the second loop 0 = 𝐶 ∫[𝑖2 (𝑡) − 𝑖1 (𝑡)]𝑑𝑡 + 𝐿1 + 𝑅3 [𝑖2 (𝑡) − 𝑖3 (𝑡)] + 𝑅1 𝑖2 (𝑡)
1 𝑑𝑡
1 𝑑𝑖3 (𝑡)
At the third loop , 0 = 𝑅3 (𝑖3 (𝑡) − 𝑖2 (𝑡)) + 𝑘 ∫ 𝑖3 (𝑡) 𝑑𝑡 + 𝑅2 𝑖3 (𝑡) + 𝐿2
2 𝑑𝑡

From these equations we can draw the torque-voltage analogous electrical circuit as

Torque-Current analogy refer table 2, from the table, torque is replaced by current,
inertia is replaced by capacitance, viscous friction is replaced by reciprocal of resistance
and spring is replaced by reciprocal of inductance.
1
At node 𝑣1 (𝑡), 𝑖(𝑡) = 𝐿 ∫[𝑣1 (𝑡) − 𝑣2 (𝑡)]𝑑𝑡
1

1 𝑑𝑣2 (𝑡)
At node 𝑣2 (𝑡) 0 = 𝐿 ∫[𝑣2 (𝑡) − 𝑣1 (𝑡)]𝑑𝑡 + 𝐶1 + 𝐺3 (𝑣2 (𝑡) − 𝑣3 (𝑡)) + 𝐺1 𝑣2 (𝑡)
1 𝑑𝑡
1 𝑑𝑣3 (𝑡)
At node 𝑣3 (𝑡), 0 = 𝐺3 (𝑣3 (𝑡) − 𝑣2 (𝑡)) + 𝐿 ∫ 𝑣3 (𝑡)𝑑𝑡 + 𝐺2 𝑣3 (𝑡) + 𝐶2 𝑑𝑡
1
Problem 10: The electrical circuit consists cascaded two RC sections. Find the transfer
𝐸0 (𝑠)
function
𝐸𝑖 (𝑠)

Solution: Using KVL, the system equations at the first loop

1
𝑒1 (𝑡) = 𝑅𝑖1 (𝑡) + ∫(𝑖1 (𝑡) − 𝑖2 (𝑡)) 𝑑𝑡
𝐶
1 1
At the second loop 𝐶
∫(𝑖2 (𝑡) − 𝑖1 (𝑡)) 𝑑𝑡 + 𝑅𝑖2 (𝑡) + 𝐶 ∫ 𝑖2 (𝑡)𝑑𝑡 = 0
Apply Laplace equation for both the equations, we get
(𝐼1 (𝑠) − 𝐼2 (𝑠)) 1 𝐼2 (𝑠))
𝐸1 (𝑠) = 𝑅𝐼1 (𝑠) + = [𝑅 + ] 𝐼1 (𝑠) −
𝐶𝑠 𝐶𝑠 𝐶𝑠
(𝐼2 (𝑠) − 𝐼1 (𝑠)) 𝐼2 (𝑠) 𝐼1 (𝑠)) 2
0 = 𝑅𝐼2 (𝑠) + + =− + [𝑅 + ] 𝐼2 (𝑠)
𝐶𝑠 𝐶𝑠 𝐶𝑠 𝐶𝑠
1 1
𝑅+ −
[ 𝐶𝑠 𝐶𝑠 ] [𝐼1 (𝑠)] = [𝐸1 (𝑠)]
1 2 𝐼2 (𝑠) 0
− 𝑅+
𝐶𝑠 𝐶𝑠
1 1
𝑅+ − 𝑅 2 𝐶 2 𝑠 2 + 3𝑅𝐶𝑠 + 2 1 𝜏 2 𝑠 2 + 3𝜏𝑠 + 1
[ 𝐶𝑠 𝐶𝑠 ] = ∆= − 2 2=
1 2 𝐶 2𝑠2 𝐶 𝑠 𝑐2𝑠2
− 𝑅+
𝐶𝑠 𝐶𝑠
 1
𝑅 + 𝐶𝑠 𝐸1 (𝑠)
[ 1 ]
− 𝐶𝑠 0 𝐸1 (𝑠) 𝐸1 (𝑠)𝐶𝑠
𝐼2 (𝑠) = = 𝜏2 𝑠2 +3𝜏𝑠+1
=
∆ ( ) × 𝐶𝑠 𝜏 2𝑠2
+ 3𝜏𝑠 + 1
𝑐 2 𝑠2
1
Output of the circuit is 𝑒0 (𝑡) = 𝐶 ∫ 𝑖2 (𝑡)𝑑𝑡

Apply the Laplace transform to the above equation we get

𝐼2 (𝑠) 𝐸1 (𝑠)𝐶𝑠 1 𝐸1 (𝑠)
𝐸0 (𝑠) = = 2 2 × = 2 2
𝐶𝑠 𝜏 𝑠 + 3𝜏𝑠 + 1 𝐶𝑠 𝜏 𝑠 + 3𝜏𝑠 + 1
𝐸 (𝑠) 1
Transfer function 𝐸0 (𝑠) = 𝜏2𝑠2 +3𝜏𝑠+1
1

Problem 11: Derive the transfer function of the amplifier circuit with amplifier gain K as
shown in below figure.

Solution:

For the input side of the circuit, apply KVL

1 𝐼1 (𝑠) 1
𝑒𝑖 (𝑡) = 𝑅1 𝑖1 (𝑡) + 𝐶 ∫ 𝑖1 (𝑡)𝑑𝑡 , 𝐸𝑖 (𝑠) = 𝑅1 𝐼1 (𝑠) + , 𝐸𝑖 (𝑠) = (𝑅1 + 𝐶 𝑠)𝐼1 (𝑠) ,
1 𝐶1 𝑠 1

𝐸𝑖 (𝑠)
𝐼1 (𝑠) = 𝑅1𝐶1 𝑠 +1
( )
𝐶1 𝑠
1 𝐼1 (𝑠) 𝐸𝑖 (𝑠) 𝐸𝑖 (𝑠)
And 𝑣1 (𝑡) = 𝐶 ∫ 𝑖1 (𝑡)𝑑𝑡, 𝑉1 (𝑠) = = 𝑅1𝐶 𝑠 +1 =𝑅
1 𝐶1 𝑠 𝐶1 𝑠( 1 ) 1𝐶1 𝑠 +1
𝐶1 𝑠

𝑉1 (𝑠) 1
Therefore, transfer function of the input side is =𝑅 = 𝐺𝑎𝑖𝑛 𝐺1
𝐸𝑖 (𝑠) 1𝐶1 𝑠 +1

For the output side of the circuit, apply KVL

 1 𝐼2 (𝑠) 1
𝑒0 (𝑡) = 𝑅2 𝑖2 (𝑡) + 𝐶 ∫ 𝑖2 (𝑡)𝑑𝑡 , 𝐸0 (𝑠) = 𝑅2 𝐼2 (𝑠) + , 𝐸0 (𝑠) = (𝑅2 + 𝐶 𝑠)𝐼2 (𝑠) ,
2 𝐶2 𝑠 2

𝐸0 (𝑠)
𝐼2 (𝑠) = 𝑅2𝐶2 𝑠 +1
( )
𝐶2 𝑠
1 𝐼2 (𝑠) 𝐸0 (𝑠) 𝐸0 (𝑠)
And 𝑣2 (𝑡) = 𝐶 ∫ 𝑖2 (𝑡)𝑑𝑡, 𝑉2 (𝑠) = = 𝑅2𝐶 𝑠 +1 =𝑅
2 𝐶2 𝑠 𝐶2 𝑠( 2 ) 2 𝐶2 𝑠+1
𝐶2 𝑠

𝐸0 (𝑠)
Transfer function of the output side is = 𝑅2 𝐶2 𝑠 + 1 = 𝐺𝑎𝑖𝑛 𝐺2
𝑉2 (𝑠)

We know that gain of the amplifier 𝐺3 = 𝐾

Overall gain of the of the amplifier circuit is
1 𝐾(𝑅2 𝐶2 𝑠+1)
𝐺 = 𝐺1 𝐺2 𝐺3 = 𝑅 × 𝑅2 𝐶2 𝑠 + 1 × 𝐾 = (𝑅1𝐶1 𝑠 +1)
1𝐶1 𝑠 +1

Problem: Find the transfer function of the electrical circuit as shown in below figure.

Solution:

For the above circuit apply KCL to the circuit 𝑖3 = 𝑖1 +𝑖2

𝑣𝑜 (𝑡) [(𝑣𝑖 (𝑡) − 𝑣𝑜 (𝑡)] 𝑑[(𝑣𝑖 (𝑡) − 𝑣𝑜 (𝑡)]
= +𝐶
𝑅2 𝑅1 𝑑𝑡
Apply Laplace transform to the above equation we get
 𝑉𝑜 (𝑠) [(𝑉𝑖 (𝑠) − 𝑉𝑜 (𝑠)]
= + 𝐶𝑠 [(𝑉𝑖 (𝑠) − 𝑉𝑜 (𝑠)]
𝑅2 𝑅1
𝑉𝑜 (𝑠) 𝑉𝑜 (𝑠) 𝑉𝑖 (𝑠)
+ + 𝐶𝑠𝑉𝑜 (𝑠) = + 𝑉𝑖 (𝑠)𝐶𝑠
𝑅2 𝑅1 𝑅1
1 1 1
𝑉𝑜 (𝑠) [ + + 𝐶𝑠] = 𝑉𝑖 (𝑠) [ + 𝐶𝑠]
𝑅2 𝑅1 𝑅1
𝑅1 + 𝑅2 (1 + 𝑅1 𝐶𝑠) 1 + 𝑅1 𝐶𝑠
𝑉𝑜 (𝑠) [ ] = 𝑉𝑖 (𝑠) [ ]
𝑅1 𝑅2 𝑅1

𝑅1 + 𝑅2 (1 + 𝑅1 𝐶𝑠)
𝑉𝑜 (𝑠) [ ] = 𝑉𝑖 (𝑠)[1 + 𝑅1 𝐶𝑠 ]
𝑅2
𝑉𝑜 (𝑠) [1 + 𝑅1 𝐶𝑠 ] 𝑅2
=
𝑉𝑖 (𝑠) 𝑅1 + 𝑅2 (1 + 𝑅1 𝐶𝑠)
Problem: Find the transfer function of the electrical circuit as shown in below figure.

Solution:

Apply KVL to the circuit

1
𝑣𝑖 (𝑡) = 𝑅1 𝑖1 (𝑡) + ∫[𝑖1 (𝑡) − 𝑖2 (𝑡)]𝑑𝑡
𝐶1
 1 1
0 = 𝑅2 𝑖2 (𝑡) + ∫[𝑖2 (𝑡)𝑑𝑡 + ∫[𝑖2 (𝑡) − 𝑖1 (𝑡)]𝑑𝑡
𝐶2 𝐶1
Apply Laplace transform
[𝐼1 (𝑠) − 𝐼2 (𝑠)]
𝑉𝑖 (𝑠) = 𝑅1 𝐼1 (𝑠) +
𝐶1 𝑠
𝐼2 (𝑠) [𝐼2 (𝑠) − 𝐼1 (𝑠)]
0 = 𝑅2 𝐼2 (𝑠) + +
𝐶2 𝑠 𝐶1 𝑠
1 −1
𝑅1 +
𝑉 (𝑠) 𝐶1 𝑠 𝐶1 𝑠 𝐼 (𝑠)
[ 𝑖 ]= [1 ]
0 −1 1 1 𝐼2 (𝑠)
𝑅2 + +
[ 𝐶1 𝑠 𝐶1 𝑠 𝐶2 𝑠]
𝑅1 𝑅2 𝐶1 𝐶2 𝑠 2 + [𝑅1 𝐶1 +𝑅1 𝐶2 + 𝑅2 𝐶2 ]𝑠 + 1
∆=
𝐶2 𝐶1 𝑠 2
1
𝑅1 + 𝐶 𝑉𝑖 (𝑠)
1𝑠
[ −1 ]
0 𝑉𝑖 (𝑠) 𝑉𝑖 (𝑠)
𝐶1 𝑠
𝐼2 (𝑠) = = = 𝑅 𝑅 𝐶 𝐶 𝑠2 +[𝑅 𝐶 +𝑅 𝐶 +𝑅 𝐶 ]𝑠+1
∆ ∆𝐶1 𝑠 [ 1 2 1 2 1 1 1 2 2 2
]𝐶1 𝑠
𝐶 𝐶 𝑠22 1

𝑉𝑖 (𝑠)
𝐼2 (𝑠) = 𝑅1 𝑅2 𝐶1 𝐶2 𝑠2 +[𝑅1 𝐶1 +𝑅1 𝐶2 +𝑅2 𝐶2 ]𝑠+1
[ ]
𝐶2 𝑠

1
𝑣𝑜 (𝑡) = ∫ 𝑖2 (𝑡) 𝑑𝑡
𝐶2
𝐼2 (𝑠) 𝑉𝑖 (𝑠)
𝑉𝑜 (𝑠) = = 𝑅 𝑅 𝐶 𝐶 𝑠2 +[𝑅 𝐶 +𝑅 𝐶 +𝑅 𝐶 ]𝑠+1 × 𝐶2 𝑠
𝐶2 𝑠 [ 1 2 1 2 1 1 1 2 2 2
]
𝐶2 𝑠

𝑉𝑖 (𝑠)
=
𝑅1 𝑅2 𝐶1 𝐶2 𝑠2 + [𝑅1 𝐶1 +𝑅1 𝐶2 + 𝑅2 𝐶2 ]𝑠 + 1
𝑉𝑜 (𝑠) 1
= 2
𝑉𝑖 (𝑠) 𝑅1 𝑅2 𝐶1 𝐶2 𝑠 + [𝑅1 𝐶1 +𝑅1 𝐶2 + 𝑅2 𝐶2 ]𝑠 + 1
This is the transfer function of the system.
Problem 12: Find the transfer function 𝑋(𝑠)/𝐸(𝑠) for the electromechanical loud
speaker system shown in below figure. Take back emf 𝑒𝑏 = 𝑘1 𝑑𝑥/𝑑𝑡.
Solution: An equivalent circuit of a loud speaker as shown in figure (a) and (b)
From figure (a)
1
𝑒(𝑡) = 𝑅𝑖1 (𝑡) + ∫(𝑖1 (𝑡) − 𝑖2 (𝑡)𝑑𝑡
𝐶
𝑑𝑖2 (𝑡) 1 𝑑𝑥
0=𝐿 + ∫(𝑖2 (𝑡) − 𝑖1 (𝑡)𝑑𝑡 + 𝑘1
𝑑𝑡 𝐶 𝑑𝑡
Apply Laplace transform to above equation, we get
(𝐼1 (𝑠) − 𝐼2 (𝑠))
𝐸(𝑠) = 𝑅𝐼1 (𝑠) +
𝐶𝑠
(𝐼2 (𝑠) − 𝐼1 (𝑠))
− 𝑘1 𝑠𝑋(𝑠) = 𝐿 𝑠𝐼2 (𝑠) +
𝐶𝑠
1 1
𝐸(𝑠) 𝑅+ −
[ ]=[ 𝐶𝑠 𝐶𝑠 ] [(𝐼1 (𝑠)]
−𝑘1 𝑠𝑋(𝑠) 1 1 𝐼2 (𝑠)
− 𝐿𝑠 +
𝐶𝑠 𝐶𝑠
1 1 1 2 𝑅𝐿𝐶𝑠 2 + 𝐿𝑠 + 𝑅
∆= (𝑅 + ) (𝐿𝑠 + ) − ( ) =
𝐶𝑠 𝐶𝑠 𝐶𝑠 𝐶𝑠
1
𝑅 + 𝐶𝑠 𝐸(𝑠)
[ 1 ]
− 𝐶𝑠 −𝑘1 𝑠𝑋(𝑠) 𝐸(𝑠) (𝑅𝐶𝑠 + 1)𝑘1 𝑠𝑋(𝑠)
𝐼2 (𝑠) = = − … … . . (𝐴)
∆ ∆𝐶𝑠 ∆𝐶𝑠
For figure (b), the flow of current in the coil due to applied voltage 𝑒(𝑡) produces a
magnetic field, which interacts with the permanent magnet and a produces mechanical
force 𝐹(𝑡). This force is given by equation
𝐹(𝑡) = 𝑘2 𝑖2 (𝑡)
From figure (b)
𝑑2 𝑥(𝑡) 𝑑𝑥(𝑡)
𝐹(𝑡) = 𝑘2 𝑖2 (𝑡) = 𝑀 2
+𝑓 + 2𝑘𝑥(𝑡)
𝑑𝑡 𝑑𝑡
Apply the Laplace transform to the above equation, we get
𝑘2 𝐼2 (𝑠) = 𝑀 𝑠 2 𝑋(𝑠) + 𝑓𝑠𝑋(𝑠) + 2𝑘𝑋(𝑠)
From equation (A), we get
𝐸(𝑠) (𝑅𝐶𝑠 + 1)𝑘1 𝑠𝑋(𝑠)
𝑘2 [ − ] = [𝑀 𝑠 2 + 𝑓𝑠 + 2𝑘]𝑋(𝑠)
∆𝐶𝑠 ∆𝐶𝑠
𝐸(𝑠) (𝑅𝐶𝑠 + 1)𝑘1 𝑘2 𝑠
𝑘2 =[ + 𝑀 𝑠 2 + 𝑓𝑠 + 2𝑘]𝑋(𝑠)
∆𝐶𝑠 ∆𝐶𝑠

𝑘2 𝐸(𝑠) = [(𝑅𝐶𝑠 + 1)𝑘1 𝑘2 𝑠] + [∆𝐶𝑠( 𝑀 𝑠 2 + 𝑓𝑠 + 2𝑘)]𝑋(𝑠)

𝑋(𝑠) 𝑘2
=
𝐸(𝑠) [(𝑅𝐶𝑠 + 1)𝑘1 𝑘2 𝑠] + [∆𝐶𝑠( 𝑀 𝑠 2 + 𝑓𝑠 + 2𝑘)]
Problem 12: Find the transfer function 𝑋(𝑠)/𝐸(𝑠) for the electromechanical loud
speaker system shown in below figure. Take back emf 𝑒𝑏 = 𝑘1 𝑑𝑥/𝑑𝑡, where 𝑘1 is the
back emf constant.
Solution: An equivalent is

From figure
𝑑𝑖(𝑡)
𝑒(𝑡) = 𝑅𝑖1 (𝑡) + 𝐿 + 𝑒𝑏 (𝑡)
𝑑𝑡
𝑑𝑖(𝑡) 𝑑𝑥(𝑡)
𝑒(𝑡) = 𝑅𝑖1 (𝑡) + 𝐿 + 𝑘1
𝑑𝑡 𝑑𝑡

Apply Laplace transform to above equation, we get

𝐸(𝑠) = 𝑅𝐼(𝑠) + 𝐿𝑠𝐼(𝑠) + 𝑘1 𝑋(𝑠)𝑠
𝐸(𝑠) − 𝑘1 𝑋(𝑠)𝑠
𝐼(𝑠) = … … . (C)
R + Ls
The mechanical equivalent is
From above figure
𝑑 2 𝑥(𝑡) 𝑑𝑥(𝑡)
𝐹(𝑡) = 𝑀 2
+𝐵 + 𝑘𝑥(𝑡)
𝑑𝑡 𝑑𝑡
Apply Laplace transform, we get
𝐹(𝑠) = 𝑀𝑠 2 𝑋(𝑠) + 𝐵𝑠𝑋(𝑠) + 𝑘𝑋(𝑠) … . (𝐷)
We know that
𝐹(𝑡) = 𝑘2 𝑖2 (𝑡)
𝐹(𝑠) = 𝑘2 𝐼2 (𝑠)
From equation (D) becomes
𝑘2 𝐼2 (𝑠) = 𝑀𝑠 2 𝑋(𝑠) + 𝐵𝑠𝑋(𝑠) + 𝑘𝑋(𝑠)
𝑀𝑠 2 𝑋(𝑠) + 𝐵𝑠𝑋(𝑠) + 𝑘𝑋(𝑠)
𝐼2 (𝑠) = … (𝐸)
𝑘2
Since equation (C) and (E) are same, we can equate them, thus
𝑀𝑠 2 𝑋(𝑠) + 𝐵𝑠𝑋(𝑠) + 𝑘𝑋(𝑠) 𝐸(𝑠) − 𝑘1 𝑋(𝑠)𝑠
=
𝑘2 R + Ls
[R + Ls][𝑀𝑠 2 𝑋(𝑠) + 𝐵𝑠𝑋(𝑠) + 𝑘𝑋(𝑠)] = 𝑘2 𝐸(𝑠) − 𝑘2 𝑘1 𝑋(𝑠)𝑠
{[R + Ls][𝑀𝑠 2 + 𝐵𝑠 + 𝑘] + 𝑘2 𝑘1 𝑠}𝑋(𝑠) = 𝑘2 𝐸(𝑠)
𝑋(𝑠) 𝑘2
=
𝐸(𝑠) [R + Ls][𝑀𝑠 2 + 𝐵𝑠 + 𝑘] + 𝑘2 𝑘1 𝑠
 Module 2
Block diagrams and Signal flow graphs
Transfer function
Control system is an interconnection component to provide controlled output by varying
the controlling input. In order to show the function performed by each component in the
system we commonly use a diagram called the block diagram. It is a pictorial
representation of the function performed by each component and the flow of signals. We
know that transfer function is the ratio of Laplace transfer ourput to the Laplace transfer
of the input. For any linear system can be represented by it transfer function 𝐺(𝑠) =
𝐶(𝑠)/𝑅(𝑠), a block diagram representation of this transfer function is as shown in below
figure (a).
Where,𝑅(𝑠) is the Laplace transform of the signal in to the block, and 𝐶(𝑠) is the Laplace
transform of output signal coming out from the block. In addition to this, summing or
difference of signals is indicated by a symbol as shown in figure (b), while the take-off
point of a signal is represented by figure(c).

Block diagram representation of Closed-loop control system

Below figure shows the block diagram of a negative feedback system. This block
diagram is also known as canonical form feedback control system. With reference to this
figure, terminology used in block diagram of control system is given below.
𝑅(𝑠)= reference input
𝐶(𝑠)= output signal or controlled variable
𝐵(𝑠)=feedback signal
𝐸(𝑠)= actuating signal
𝐺(𝑠) = 𝐶(𝑠)/𝐸(𝑠)=forward path transfer function
𝐻(𝑠)=transfer function of the feedback element
𝐺(𝑠)𝐻(𝑠) = 𝐵(𝑠)/𝐸(𝑠)=loop transfer function
𝑇(𝑠) = 𝐶(𝑠)/𝑅(𝑠)= closed-loop transfer function
From the figure (a) 𝐶(𝑠) = 𝐺(𝑠)𝐸(𝑠)
𝐸(𝑠) = 𝑅(𝑠) − 𝐵(𝑠) = 𝑅(𝑠) − 𝐻(𝑠)𝐶(𝑠)
Now 𝐶(𝑠) = 𝐺(𝑠)𝐸(𝑠) = 𝐺(𝑠)[𝑅(𝑠) − 𝐻(𝑠)𝐶(𝑠)]
𝐶(𝑠) = 𝐺(𝑠)𝑅(𝑠) − 𝐺(𝑠)𝐻(𝑠)𝐶(𝑠)
𝐶(𝑠) + 𝐺(𝑠)𝐻(𝑠)𝐶(𝑠) = 𝐺(𝑠)𝑅(𝑠)
𝐶(𝑠)[1 + 𝐺(𝑠)𝐻(𝑠)] = 𝐺(𝑠)𝑅(𝑠)
𝐶(𝑠) 𝐺(𝑠)
Therefore transfer function of the system is 𝑅(𝑠) = 1+𝐺(𝑠)𝐻(𝑠), thus, the reduced block

diagram or simplified block diagram as shown in figure (b).

Similarly for the positive feedback system is as shown in below figure.

𝐶(𝑠) 𝐺(𝑠)
And the transfer function is 𝑅(𝑠) = 1−𝐺(𝑠)𝐻(𝑠)

The complex block diagrams can be simplified by using rules of block diagram algebra.
Some of the rules are listed in the table 3.
 𝐶(𝑠)
Problem: Reduce the block diagram shown below, and find the transfer function 𝑅(𝑠)

Solution:
 𝐶(𝑠) 𝐺1 𝐺3 𝐺1 𝐺2 + 𝐺3
=( ) (𝐺2 + ) =
𝑅(𝑠) 1 + 𝐻𝐺1 𝐺1 1 + 𝐻𝐺1
𝐶(𝑠)
Problem: Reduce the block diagram shown below, and find the transfer function 𝑅(𝑠)
Signal flow graphs
Block diagrams are very successful for representation of control systems, but for
complicated systems, the block diagram reduction process is tedious and time consuming.
An alternate approach is that of signal flow graphs developed by S J Mason. Signal flow
graph is a graphical representation of the system equation and serves as same purpose of
block diagram. The below figure show the simple block diagram and equivalent signal
flow graph.
Various signal flow graph terms
Node: It represents the system variable which is equal to the sum of all incoming signals
at the node.
Input node or source: It is node having only outgoing branches.
Output node or sink: It is node having only incoming branches.
Branch: A branch is a directed line segment joining two nodes. The arrow on the branch
indicates the direction of the signal flow and gain of the branch is the transmittance.

Node as a summing point: With respect to below signal flow graph, the node variable x1
is expressed as 𝑥1 = 𝑎21 𝑥2 + 𝑎31 𝑥3 + 𝑎41 𝑥4 x1 Sum of the incoming signal.
Node as a transmitting point: A node variable is transmitted through all branches
outgoing from the node. With respect to above signal flow graph, the node variable
𝑥5 = 𝑎15 𝑥1
𝑥6 = 𝑎16 𝑥1
Notation:𝑎𝑖𝑗 is a transmittance of branch directed from node 𝑥𝑖 to node 𝑥𝑗
Path: It is the traversal of connected branches in the direction of the branch arrows such
that no node is traversed more than once.
Forward path: It is the path from the input node to the output node. For example R-E-C
is the forward path in the signal flow graph below.
Loop: It is a path which originates and terminates at the same node. For example E-C-B-
E is a loop in the below signal flow graph.

Non-touching loop: Loops are said to be non-touching if they do not possess any
common node. For example L1=-GH.
Forward path gain: It is the product of the branch gain encountered in traverversing a
forward path. For example forward path gain of the path R-E-C in the above signal flow
graph is G.
Loop gain: It is the product of the branch gain encountered in traverversing the loop. For
example, loop gain of the loop E-C-B-E in the above signal flow graph is -GH.
Construction of signal flow graphs
Signal flow graph of the system is constructed from the describing equations. Let us
consider a system described by following set of equations.
𝑥2 = 𝑎12 𝑥1 + 𝑎23 𝑥3 + 𝑎24 𝑥4 + 𝑎52 𝑥5
𝑥3 = 𝑎23 𝑥2
𝑥4 = 𝑎34 𝑥3 + 𝑎44 𝑥4
𝑥5 = 𝑎35 𝑥3 + 𝑎45 𝑥4
Mason’s gain formula
Mason’s gain formula is used to find the overall transfer function 𝐶(𝑠)/𝑅(𝑠) of the
system. Mason’s gain formula for the determination of the overall transfer function is
given by
𝐶(𝑠) 1
𝑇= = ∑ 𝑃𝑘 ∆𝑘
𝑅(𝑠) ∆
𝑘

Where 𝑃𝑘 the gain of the kth is forward path between R(s) and C(s) and ∆ is the
determinant of the graph
∆
= 1 − (sum of loop gains of all individual loops)
+ (sum of loop gain product of all possible combination of two non touching loops)
− (sum of gain products of all possible combination of three non touching)
+ ⋯ . . i. e

∆= 1 − ∑ Pm1 + ∑ Pm2 − ∑ Pm3 + ⋯

m m m

Where Pmr is the gain product mth possible combination of r non-touching loops, and ∆𝑘
is the value of ∆ for the part of the graph not touching the kth forward path
Problem: Determine the overall gain (transfer function) of the system for the signal flow
graph as show below using Mason’s gain formula.

Solution: The overall transfer function of the signal flow graph is

𝐶(𝑠) 𝑥5 1
= = ∑ 𝑃𝑘 ∆𝑘
𝑅(𝑠) 𝑥1 ∆
𝑘

There are two forward paths between 𝑥1 and 𝑥5 , therefore 𝑘 = 2

The first forward path gain 𝑃1 = 𝑎12 𝑎23 𝑎34 𝑎45
The second forward path gain 𝑃2 = 𝑎12 𝑎23 𝑎35

∆
= 1 − (Sum of loop gains of all individual loops)
+ (Sum of loop gain product of all possible combination of two non touching loops)
− (Sum of gain products of all possible combination of three non touching)
+ ⋯ . . i. e

Sum of loop gains of all individual loops = ∑ pm1

m

There are five individual loop gains, they are ∑m pm1 = P11 + P21 + P31 + P41 + P51
P11 = a23 a32
 P21 = a23 a34 a42
P31 = a44
P41 = a23 a34 a45 a52
P51 = a23 a35 a52
Sum of loop gain product of all possible combination of two non touching loops

= ∑ Pm2
m

There are two possible combinations of two non-touching loops they are ∑m Pm2 = P12 +
P22
P12 = a23 a32 a44
P22 = 𝑎23 𝑎35 𝑎52 𝑎44
There are no combinations of three non-touching loops, four non-touching loops.
Therefore

∑ Pm3 = ∑ Pm4 … = 0
m m

Hence ∆= 1 − (P11 + P21 + P31 + P41 + P51 ) + (P12 + P22 )

∆= 1 − (a23 a32 + a23 a34 a42 + a44 + a23 a34 a45 a52 + a23 a35 a52 ) + (a23 a32 a44
+ 𝑎23 𝑎35 𝑎52 𝑎44 )
First forward path is in touch with all the loops. Therefore ∆1 = 1
The second forward path is not in touch with loop P31. Therefore ∆2 = 1 − (P31 ) = 1 −
𝑎44

𝑥5 𝑃1 ∆1 + 𝑃2 ∆2
=
𝑥1 ∆
(𝑎12 𝑎23 𝑎34 𝑎45 ) + (𝑎12 𝑎23 𝑎35 )(1 − 𝑎44 )
=
1 − a23 a32 − a23 a34 a42 − a44 − a23 a34 a45 a52 − a35 a52 + a23 a32 a44 + 𝑎23 𝑎35 𝑎52 𝑎44
Problem: Write the signal flow graph for the block diagram and determine the overall
transfer function of the system using Mason’s gain formula.
Solution: The signal flow graph for the block diagram as shown in below figure

The overall transfer function of the signal flow graph is

𝐶(𝑠) 1
= ∑ 𝑃𝑘 ∆𝑘
𝑅(𝑠) ∆
𝑘

There are two forward paths between the output and input, therefore 𝑘 = 2
The first forward path gain 𝑃1 = 𝐺1 𝐺2 𝐺3
The second forward path gain 𝑃2 = 𝐺4
∆
= 1 − (Sum of loop gains of all individual loops)
+ (Sum of loop gain product of all possible combination of two non touching loops)
− (Sum of gain products of all possible combination of three non touching)
+ ⋯ . . i. e

Sum of loop gains of all individual loops = ∑ pm1

m
There are three individual loop gains, they are ∑m pm1 = P11 + P21 + P31
P11 = G1 G2 H1
P21 = −G2 G3 H2
P31 = −G2 H1
Sum of loop gain product of all possible combination of two non touching loops

= ∑ Pm2
m

There are no possible combinations of two non-touching loops, thus ∑m Pm2 = 0

There are no combinations of three non-touching loops, four non-touching loops.
Therefore

∑ Pm3 = ∑ Pm4 … = 0
m m

Hence ∆= 1 − (P11 + P21 + P31 )

∆= 1 − G1 G2 H1 + G2 G3 H2 + G2 H1
First forward path is in touch with all the loops. Therefore ∆1 = 1 and second forward
path is not in touch with any loop. Therefore∆2 = 1 − (P11 + P21 + P31 ) = 1 − G1 G2 H1 +
G2 G3 H2 + G2 H1

𝐶(𝑠) 𝑃1 ∆1 + 𝑃2 ∆2 𝐺1 𝐺2 𝐺3 + 𝐺4 (1 − G1 G2 H1 + G2 G3 H2 + G2 H1 )
= =
𝑅(𝑠) ∆ 1 − 𝐺1 𝐺2 𝐻1 + 𝐺2 𝐺3 𝐻2 + 𝐺2 𝐻1
Problem: Determine the overall transfer function of the of the signal flow graph shown
below using Mason’s gain formula.
Solution: The overall transfer function of the signal flow graph is
𝐶(𝑠) 1
= ∑ 𝑃𝑘 ∆𝑘
𝑅(𝑠) ∆
𝑘

There are six forward paths between the output and input, therefore 𝑘 = 6
The first forward path gain 𝑃1 = 𝐺1 𝐺2 𝐺3
The second forward path gain 𝑃2 = 𝐺4 𝐺5 𝐺6
The third forward path gain 𝑃3 = 𝐺1 𝐺8 𝐺6
The forth forward path gain 𝑃4 = 𝐺4 𝐺7 𝐺3
The fifth forward path gain 𝑃5 = −𝐺4 𝐺7 𝐻1 𝐺8 𝐺6
The sixth forward path gain 𝑃6 = −𝐺1 𝐺8 𝐻2 𝐺7 𝐺3

∆
= 1 − (Sum of loop gains of all individual loops)
+ (Sum of loop gain product of all possible combination of two non touching loops)
− (Sum of gain products of all possible combination of three non touching)
+ ⋯ . . i. e

Sum of loop gains of all individual loops = ∑ pm1

m

There are three individual loop gains, they are ∑m pm1 = P11 + P21 + P31
P11 = −G2 H1
P21 = −G5 H2
P31 = G7 (−H1 )(−H2 )G8 = G7 H1 G8 H2
Sum of loop gain product of all possible combination of two non touching loops

= ∑ Pm2
m

There is one possible combination of two non-touching loops, thus ∑m Pm2 = P12
P12 = G2 (−H1 )G5 (−H2 ) = G2 H1 G5 H2
There are no combinations of three non-touching loops, four non-touching loops.
Therefore

∑ Pm3 = ∑ Pm4 … = 0
m m
 Hence ∆= 1 − (P11 + P21 + P31 ) + P12

∆= 1+G2 H1 + G5 H2 − G7 H1 H2 G8 + G2 H1 G5 H2
First forward path is not in touch with loops P31. Therefore
∆1 = 1 − (P31 ) = 1 + G5 H2
Second forward path is not in touch with loop P11, therefore ∆2 = 1 − (P11 ) = 1 + G2 H1
Third forward path is in touch with all the loops , therefore ∆3 = 1
Forth forward path is in touch with all the loops , therefore ∆4 = 1
Fifth forward path is in touch with all the loops , therefore∆5 = 1
Sixth forward path is in touch with all the loops, therefore∆6 = 1
𝐶(𝑠) 𝑃1 ∆1 + 𝑃2 ∆2 + 𝑃3 ∆3 + 𝑃4 ∆4 +𝑃5 ∆5 +𝑃6 ∆6
=
𝑅(𝑠) ∆
𝐺1 𝐺2 𝐺3 (1 + G5 H2 ) + 𝐺4 𝐺5 𝐺6 (1 + G2 H1 ) + 𝐺1 𝐺8 𝐺6 + 𝐺4 𝐺7 𝐺3 −𝐺4 𝐺7 𝐻1 𝐺8 𝐺6 − 𝐺8 𝐻2 𝐺7 𝐺3
=
1+G2 H1 + G5 H2 − G7 H1 H2 G8 + G2 H1 G5 H2
Problem: Determine the overall transfer function of the of the signal flow graph shown
below using Mason’s gain formula.

Solution: The overall transfer function of the signal flow graph is

𝐶(𝑠) 1
= ∑ 𝑃𝑘 ∆𝑘
𝑅(𝑠) ∆
𝑘

There are two forward paths between the output and input, therefore 𝑘 = 2
The first forward path gain 𝑃1 = 𝐺1 𝐺2 𝐺3 𝐺4
The second forward path gain 𝑃2 = 𝐺5 𝐺6 𝐺7 𝐺8

∆
= 1 − (Sum of loop gains of all individual loops)
+ (Sum of loop gain product of all possible combination of two non touching loops)
− (Sum of gain products of all possible combination of three non touching)
+ ⋯ . . i. e

Sum of loop gains of all individual loops = ∑ pm1

m

There are four individual loop gains, they are ∑m pm1 = P11 + P21 + P31 + P41
P11 = −G2 H1
P21 = −G3 H2
P31 = −G6 H3
P41 = −G7 H4
Sum of loop gain product of all possible combination of two non touching loops

= ∑ Pm2
m

There are four possible combinations of two non-touching loops, thus ∑m Pm2 = P12 +
P22 + P32 + P42
P12 = G2 (−H1 )G6 (−H3 ) = G2 H1 G6 H3
P22 = G2 (−H1 )G7 (−H4 ) = G2 H1 G7 H4
P32 = G3 (−H2 )G6 (−H3 ) = G3 H2 G6 H3
P42 = G3 (−H2 )G7 (−H4 ) = G3 H2 G7 H4
There are no combinations of three non-touching loops, four non-touching loops.
Therefore

∑ Pm3 = ∑ Pm4 … = 0
m m

Hence ∆= 1 − (P11 + P21 + P31 + P41 ) + (P12 + P22 + P32 + P42 )

∆= 1 + G2 H1 +G3 H2 +G6 H3 + G7 H4 + G2 H1 G6 H3 + G2 H1 G7 H4 + G3 H2 G6 H3
+ G3 H2 G7 H4
 First forward path is not in touch with loops P32 , P42. Therefore
∆1 = 1 − ( P31 + P41 . ) = 1 + G6 H3 + G7 H4
Second forward path is not in touch with loop P12 , P22 therefore
∆2 = 1 − (P11 + P21 ) = 1 + G2 H1 + G3 H2
𝐶(𝑠) 𝑃1 ∆1 + 𝑃2 ∆2
=
𝑅(𝑠) ∆
(𝐺1 𝐺2 𝐺3 𝐺4 )(1 + G6 H3 + G7 H4 ) + (𝐺5 𝐺6 𝐺7 𝐺8 )(1 + G2 H1 + G3 H2 )
=
1 + G2 H1 +G3 H2 +G6 H3 + G7 H4 + G2 H1 G6 H3 + G2 H1 G7 H4 + G3 H2 G6 H3 + G3 H2 G7 H4
Problem: For the electrical circuit, determine the overall transfer function of the circuit
using Mason’s gain formula.

Solution:
𝑣𝑖 (𝑡)−𝑣𝑎 (𝑡) 𝑉𝑖 (𝑠)−𝑉𝑎 (𝑆)
𝑖1 (𝑡) = , 𝐼1 (𝑠) = …… (1)
𝑅1 𝑅1
1 𝐼1 (𝑠)−𝐼2 (𝑠)
𝑣𝑎 (𝑡) = 𝐶 ∫(𝑖1 (𝑡) − 𝑖2 (𝑡)) 𝑑𝑡, 𝑉𝑎 (𝑆) = …… (2)
1 𝐶1 𝑠
𝑣𝑎 (𝑡)−𝑣0 (𝑡) 𝑉𝑎 (𝑠)−𝑉0 (𝑠)
𝑖2 (𝑡) = , 𝐼2 (𝑠) = …. (3)
𝑅2 𝑅2
1 𝐼2 (𝑠)
𝑣0 (𝑡) = 𝐶 ∫ 𝑖2 (𝑡)𝑑𝑡, 𝑉0 (𝑠) = ……. (4)
2 𝐶2 𝑠

Using these four equations we can write the signal flow graph for the electrical network
The overall transfer function of the signal flow graph is
𝐶(𝑠) 1
= ∑ 𝑃𝑘 ∆𝑘
𝑅(𝑠) ∆
𝑘

There is only one forward path between the output and input, therefore 𝑘 = 1
1 1 1 1 1
The first forward path gain 𝑃1 = (𝑅 ) (𝐶 𝑠) (𝑅 ) (𝐶 𝑠) = 𝑅 2
1 1 2 2 1 𝐶1 𝐶2 𝑅2 𝑠

∆
= 1 − (Sum of loop gains of all individual loops)
+ (Sum of loop gain product of all possible combination of two non touching loops)
− (Sum of gain products of all possible combination of three non touching)
+ ⋯ . . i. e

Sum of loop gains of all individual loops = ∑ pm1

m

There are three individual loop gains, they are ∑m pm1 = P11 + P21 + P31
1 1 1
P11 = (− )( ) = −
𝑅1 𝐶1 𝑠 𝑅1 𝐶1 𝑠
1 1 1
P21 = ( ) (− )=−
𝑅2 𝐶1 𝑠 R 2 C1 s
1 1 1
P31 = (− )( ) = −
𝑅2 𝐶2 𝑠 𝑅2 𝐶2 𝑠
Sum of loop gain product of all possible combination of two non touching loops

= ∑ Pm2
m
There is one possible combination of two non-touching loops, thus ∑m Pm2 = P12
1 1 1
P12 = (− ) (− )=
𝑅1 𝐶1 𝑠 𝑅2 𝐶2 𝑠 𝑅1 𝐶1 𝑅2 𝐶2 𝑠 2
There are no combinations of three non-touching loops, four non-touching loops.
Therefore

∑ Pm3 = ∑ Pm4 … = 0
m m
1 1 1 1
Hence ∆= 1 − (P11 + P21 + P31 ) + (P12 ) = 1 + 𝑅 +R +𝑅 +𝑅 2
1 𝐶1 𝑠 2 C1 s 2 𝐶2 𝑠 1 𝐶1 𝑅2 𝐶2 𝑠

Forward path is touch with all the loops, therefore, ∆1 = 1

𝑉0 (𝑠) 𝑃1 ∆1
=
𝑉𝑖 (𝑠) ∆
1
𝑅1 𝐶1 𝐶2 𝑅2 𝑠2
= 1 1 1 1
1+𝑅 +R +𝑅 +𝑅 2
1 𝐶1 𝑠 2 C1 s 2 𝐶2 𝑠 1 𝐶1 𝑅2 𝐶2 𝑠

Problem: For the electrical circuit, write the block diagram and signal flow graph and
determine the overall transfer function of the circuit using Mason’s gain formula.

𝑉𝑖 (𝑠)−𝑉1 (𝑠)
For 𝑅𝑎 , 𝐼1 (𝑠) = 𝑅𝑎
 𝑉𝑖 (𝑠)−𝑉2 (𝑠)
For 𝑅𝑏 𝐼2 (𝑠) = 𝑅𝑏

For 𝑅𝑐 𝑉1 (𝑠) = 𝑅𝑐 (𝐼1 (𝑠) − 𝐼0 (𝑠))

For 𝑅𝑑 𝑉2 (𝑠) = 𝑅𝑑 (𝐼2 (𝑠) + 𝐼0 (𝑠))
1
For 𝑅0 𝐼0 (𝑠) = 𝑅 (𝑉1 (𝑠) − 𝑉2 (𝑠))
0

Based on these equations, the block diagram is

The signal flow graph is

The overall transfer function of the signal flow graph is

𝐼0 (𝑠) 1
= ∑ 𝑃𝑘 ∆𝑘
𝑉𝑖 (𝑠) ∆
𝑘
There two forward path between the output and input, therefore 𝑘 = 2
1 1 𝑅𝑐
The first forward path gain 𝑃1 = (𝑅 ) (𝑅𝑐 ) (𝑅 ) =
𝑎 0 𝑅𝑎 𝑅0
1 1 𝑅𝑑
The Second forward path gain 𝑃2 = (𝑅 ) (𝑅𝑑 ) (− 𝑅 ) = −
𝑏 0 𝑅𝑏 𝑅0

∆
= 1 − (Sum of loop gains of all individual loops)
+ (Sum of loop gain product of all possible combination of two non touching loops)
− (Sum of gain products of all possible combination of three non touching)
+ ⋯ . . i. e

Sum of loop gains of all individual loops = ∑ pm1

m

There are four individual loop gains, they are ∑m pm1 = P11 + P21 + P31 + P41
1 𝑅𝑐
P11 = (− ) (𝑅𝑐 ) = −
𝑅𝑎 𝑅𝑎
1 𝑅𝑐
P21 = ( ) (−𝑅𝑐 ) = −
𝑅0 𝑅0
1 𝑅𝑑
P31 = (− ) ( 𝑅𝑑 ) = −
𝑅0 𝑅0
1 𝑅𝑑
P41 = (− ) (𝑅𝑑 ) = −
Rb Rb
Sum of loop gain product of all possible combination of two non touching loops

= ∑ Pm2
m

There are three possible combinations of two non-touching loops, thus ∑m Pm2 = P12 +
P12 + P32
1 1 𝑅𝑑 𝑅𝑐
P12 = [(− ) (𝑅𝑐 )] [(− ) ( 𝑅𝑑 )] =
𝑅𝑎 𝑅0 𝑅𝑎 𝑅0
1 1 𝑅𝑑 𝑅𝑐
P12 = [(− ) (𝑅𝑐 )] [(− ) (𝑅𝑑 )] =
𝑅𝑎 Rb 𝑅𝑎 𝑅𝑏
1 1 𝑅𝑑 𝑅𝑐
P32 = [( ) (−𝑅𝑐 )] [(− ) (𝑅𝑑 )] =
𝑅0 Rb 𝑅𝑜 𝑅𝑏
There are no combinations of three non-touching loops, four non-touching loops.
Therefore

∑ Pm3 = ∑ Pm4 … = 0
m m
𝑅 𝑅 𝑅 𝑅 𝑅 𝑅
Hence ∆= 1 − (P11 + P21 + P31 ) + (P12 + P12 + P32 ) = 1 + 𝑅 𝑐 + 𝑅𝑐 + 𝑅𝑑 + R𝑑 + 𝑅𝑑𝑅𝑐 +
𝑎 0 0 b 𝑎 0

𝑅𝑑 𝑅𝑐 𝑅 𝑅
+ 𝑅𝑑𝑅 𝑐
𝑅𝑎 𝑅𝑏 0 𝑏

The first forward path is not in touch with individual loop𝑃41 , therefore, ∆1 = 1 − 𝑃41 =
𝑅
1 + R𝑑
b

The second forward path is not in touch with individual loop𝑃11 , therefore, ∆2 = 1 −
𝑅
𝑃11 = 1 + 𝑅 𝑐
𝑎

The overall transfer function

𝐼0 (𝑠) 1 𝑃1 ∆1 + 𝑃2 ∆2
= ∑ 𝑃𝑘 ∆𝑘 =
𝑉𝑖 (𝑠) ∆ ∆
𝑘
𝑅 𝑅 𝑅 𝑅
𝐼0 (𝑠) ( 𝑅 𝑐𝑅 ) (1 + 𝑅𝑑 ) + (− 𝑅 𝑑𝑅 ) (1 + 𝑅 𝑐 )
𝑎 0 𝑏 𝑏 0 𝑎
=
𝑉𝑖 (𝑠) 1 + 𝑅𝑐 + 𝑅𝑐 + 𝑅𝑑 + 𝑅𝑑 + 𝑅𝑑 𝑅𝑐 + 𝑅𝑑 𝑅𝑐 + 𝑅𝑑 𝑅𝑐
𝑅𝑎 𝑅0 𝑅0 𝑅𝑏 𝑅𝑎 𝑅0 𝑅𝑎 𝑅𝑏 𝑅𝑜 𝑅𝑏

Problem: Construct the signal flow graph for the following describing equations and find
the overall transfer function using Mason’s gain formula.
𝑥2 = 𝐴12 𝑥1 + 𝐴32 𝑥3 + 𝐴42 𝑥4
𝑥3 = 𝐴13 𝑥1 + 𝐴23 𝑥2 + 𝐴33 𝑥3
𝑥4 = 𝐴34 𝑥3
Solution:
The overall transfer function of the signal flow graph is
𝑥4 1
= ∑ 𝑃𝑘 ∆𝑘
𝑥1 ∆
𝑘

There are two forward paths between the output and input, therefore 𝑘 = 2
The first forward path gain 𝑃1 = 𝐴12 𝐴23 𝐴34
The Second forward path gain 𝑃2 = 𝐴13 𝐴34

∆
= 1 − (Sum of loop gains of all individual loops)
+ (Sum of loop gain product of all possible combination of two non touching loops)
− (Sum of gain products of all possible combination of three non touching)
+ ⋯ . . i. e

Sum of loop gains of all individual loops = ∑ pm1

m

There are four individual loop gains, they are ∑m pm1 = P11 + P21 + P31
P11 = 𝐴23 𝐴32
P21 = 𝐴23 𝐴34 𝐴43
P31 = 𝐴33
There are no combinations of two or three, four non-touching loops. Therefore

∑ Pm2 = ∑ Pm3 = ∑ Pm4 … = 0

m m m
Hence ∆= 1 − (P11 + P21 + P31 ) = 1 − 𝐴23 𝐴32 − 𝐴23 𝐴34 𝐴43 + 𝐴33
The first forward path is in touch with all the individual loops ∆1 = 1
The second forward path is in touch with all individual loops, therefore, ∆2 = 1
The overall transfer function
𝑥4 1 𝑃1 ∆1 + 𝑃2 ∆2
= ∑ 𝑃𝑘 ∆𝑘 =
𝑥1 ∆ ∆
𝑘
𝑥4 𝐴12 𝐴23 𝐴34 + 𝐴13 𝐴34
=
𝑥1 1 − 𝐴23 𝐴32 − 𝐴23 𝐴34 𝐴43 + 𝐴33
State variable formulation
Consider the simple system described by first order differential equation
𝑥̇ = 𝑎𝑥 … … . . (𝑎)
As x defines the system state for 𝑡 ≥ 0, x is called state variable, and the equation (a) is
called state equation.
Take Laplace transform to equation (a) we get
𝑋̇(𝑠) = 𝑎𝑋(𝑠) … … (𝑏)
And we know that, the Laplace transform of first derivative is
𝑋̇(𝑠) = 𝑠𝑋(𝑠)
̇
𝑋(𝑠)
𝑋(𝑠) = … . . (𝑐)
𝑠
Using equation (b) and (c), the signal flow diagram is shown in below figure.

Let us consider the first order system with input and output
𝑥̇ = 𝑎𝑥 + 𝑏𝑢 , and 𝑦 = 𝑐𝑢
The modified signal flow graph is
In case of second order system, two first order state variable equations would be needed.
Problem: For the mechanical system shown below, draw the signal flow graph for the
mechanical system.

Solution:

The mechanical network is

 𝑑2 𝑥2 (𝑡) 𝑑
𝐹(𝑡) = 𝑀2 + 𝑘2 (𝑥2 (𝑡) − 𝑥1 (𝑡)) + 𝑓2 ( (𝑥 (𝑡) − 𝑥1 (𝑡))
𝑑𝑡 2 𝑑𝑡 2
𝐹(𝑡) = 𝑀2 𝑥̈ 2 + 𝑘2 (𝑥2 − 𝑥1 ) + 𝑓2 (𝑥̇ 2 − 𝑥̇ 1 )
1
𝑥̈ 2 = (𝐹(𝑡) − 𝑘2 (𝑥2 − 𝑥1 ) − 𝑓2 (𝑥̇ 2 − 𝑥̇ 1 ))
𝑀2
1
𝑥̈ 2 = (𝐹(𝑡) − 𝑘2 𝑥2 + 𝑘2 𝑥1 − 𝑓2 𝑥̇ 2 + 𝑓2 𝑥̇ 1 ) … … . . (𝐴)
𝑀2
𝑑𝑥1 (𝑡) 𝑑 2 𝑥1 (𝑡) 𝑑
0 = 𝑘1 𝑥1 (𝑡) + 𝑓1 + 𝑀1 2
+ 𝑘2 (𝑥1 (𝑡) − 𝑥2 (𝑡)) + 𝑓2 ( (𝑥1 (𝑡) − 𝑥2 (𝑡))
𝑑𝑡 𝑑𝑡 𝑑𝑡
0 = 𝑘1 𝑥1 + 𝑓1 𝑥̇ 1 + 𝑀1 𝑥̈ 1 + 𝑘2 (𝑥1 − 𝑥2 ) + 𝑓2 (𝑥̇ 1 − 𝑥̇ 2 )
1
𝑥̈ 1 = (−𝑘1 𝑥1 − 𝑓1 𝑥̇ 1 − 𝑘2 𝑥1 + 𝑘2 𝑥2 − 𝑓2 𝑥̇ 1 + 𝑓2 𝑥̇ 2 ) … . (𝐵)
𝑀1
Using equation (A) and (B) the signal flow graph as shown in below figure

Block diagram of armature controlled DC motor

Consider the armature controlled DC motor as shown in below figure.

In this system
𝑅𝑎 =Resistance of the armature
 𝐿𝑎 =Indutance of the armature winding
𝑖𝑎 = Armature current
𝑖𝑓 =Field current
𝑒𝑎 =Applied voltage to the armature
𝑒𝑏 =Back emf of the motor
𝑇𝑀 =Torque developed by the motor
𝜃 =Angular displacement of the motor shaft
𝐽 = Equivalent moment of inertia of the motor and load referred to motor shaft
𝑓0 =Friction coefficient of motor and load referred to motor shaft
In armature controlled DC Motor, torque developed by the motor is proportional to
product of flux and armature current.
𝑇𝑀 𝛼∅𝑖𝑎
Since flux is constant in armature controlled DC motor
𝑇𝑀 𝛼𝑖𝑎
𝑇𝑀 = 𝐾𝑇 𝑖𝑎
Where 𝐾𝑇 is proportionality constant
The motor back emf being proportionality to speed or angular velocity
𝑑𝜃
∴ 𝑒𝑏 𝛼
𝑑𝑡
𝑑𝜃
𝑒𝑏 = 𝐾𝑏
𝑑𝑡
Where 𝐾𝑏 is proportionality constant
The differentional equation of the armature circuit is
𝑑𝑖𝑎
𝐿𝑎 + 𝑅𝑎 𝑖𝑎 + 𝑒𝑏 = 𝑒𝑎
𝑑𝑡
The torque equation is
𝑑2 𝜃 𝑑𝜃
𝐽 2 + 𝑓0 = 𝑇𝑀
𝑑𝑡 𝑑𝑡
Taking the Laplace transform to the above equation we get
(𝐿𝑎 𝑠 + 𝑅𝑎 )𝐼𝑎 (𝑠) = 𝐸𝑎 (𝑠) − 𝐸𝑏 (𝑠)
𝐸𝑎 (𝑠) − 𝐸𝑏 (𝑠)
𝐼𝑎 (𝑠) = … … . . (𝐴)
(𝐿𝑎 𝑠 + 𝑅𝑎 )
 (𝐽𝑠 2 + 𝑓0 𝑠)𝜃(𝑠) = 𝑇𝑀 (𝑠) = 𝐾𝑇 𝐼𝑎 (𝑠)
𝐾𝑇 𝐼𝑎 (𝑠)
𝜃(𝑠) = … … … (𝐵)
(𝐽𝑠 2 + 𝑓0 𝑠)
𝐸𝑏 (𝑠) = 𝐾𝑏 𝑠𝜃(𝑠) … … (𝐶)
𝐸𝑎 (𝑠) − 𝐸𝑏 (𝑠)
(𝐽𝑠 2 + 𝑓0 𝑠)𝜃(𝑠) = 𝐾𝑇 ( )
(𝐿𝑎 𝑠 + 𝑅𝑎 )
𝐾𝑇 𝐸𝑎 (𝑠) 𝐾𝑇 𝐸𝑏 (𝑠)
(𝐽𝑠 2 + 𝑓0 𝑠)𝜃(𝑠) = −
𝐿𝑎 𝑠 + 𝑅𝑎 𝐿𝑎 𝑠 + 𝑅𝑎
𝐾𝑇 𝐸𝑎 (𝑠) 𝐾𝑇 𝐾𝑏 𝑠𝜃(𝑠)
(𝐽𝑠 2 + 𝑓0 𝑠)𝜃(𝑠) = −
𝐿𝑎 𝑠 + 𝑅𝑎 𝐿𝑎 𝑠 + 𝑅𝑎
𝐾𝑇 𝐾𝑏 𝑠 ) 𝐾𝑇 𝐸𝑎 (𝑠)
(𝐽𝑠 2 + 𝑓0 𝑠 + ) 𝜃(𝑠) =
𝐿𝑎 𝑠 + 𝑅𝑎 𝐿𝑎 𝑠 + 𝑅𝑎
[𝑠(𝐽𝑠 + 𝑓0 )( 𝐿𝑎 𝑠 + 𝑅𝑎 ) + 𝐾𝑇 𝐾𝑏 𝑠]𝜃(𝑠) = 𝐾𝑇 𝐸𝑎 (𝑠)
𝜃(𝑠) 𝑘𝑇
The transfer function of the system is 𝐺(𝑠) = 𝐸 = 𝑠[(𝐽𝑠 +𝑓
𝑎 (𝑠) 0 )( 𝐿𝑎 𝑠+𝑅𝑎 )+𝐾𝑇 𝐾𝑏 ]

From equation A, B and (C) the block diagram is

Block diagram of field controlled DC motor

Consider the field controlled DC motor as shown in below figure.
In this system,
𝑅𝑓 =Resistance of the field winding
𝐿𝑓 =Indutance of the field winding
𝑒𝑓 = Field voltage
𝑖𝑓 =Field current
𝑇𝑀 =Torque developed by the motor
𝐽 = Equivalent moment of inertia of the motor and load referred to motor shaft
𝑓 =Friction coefficient of motor and load referred to motor shaft
𝜃 = Angular displacement of the motor shaft
In field controlled DC motor, motor torque is proportional to field current, thus
𝑇𝑀 𝛼𝑖𝑓
𝑇𝑀 = 𝐾𝑇 𝑖𝑓
The differentional equation of the field circuit is
𝑑𝑖𝑓
𝐿𝑓 + 𝑅𝑎 𝑖𝑓 = 𝑒𝑓
𝑑𝑡
The torque equation is
𝑑2 𝜃 𝑑𝜃
𝐽 + 𝑓0 = 𝑇𝑀
𝑑𝑡 2 𝑑𝑡
Taking the Laplace transform to the above equation we get
(𝐿𝑓 𝑠 + 𝑅𝑎 )𝐼𝑓 (𝑠) = 𝐸𝑓 (𝑠)
𝐸𝑓 (𝑠)
𝐼𝑓 (𝑠) = … … . . (𝐷)
𝐿𝑓 𝑠 + 𝑅𝑎
𝑇𝑀 (𝑠) = 𝐾𝑇 𝐼𝑓 (𝑠)
 (𝐽 𝑠 2 + 𝑓0 𝑠)𝜃(𝑠) = 𝑇𝑀 (𝑠) = 𝐾𝑇 𝐼𝑓 (𝑠)
𝐾𝑇 𝐼𝑓 (𝑠)
𝜃(𝑠) = … … . (𝐸)
𝑠(𝐽 𝑠 + 𝑓0 )
𝐾𝑇 𝐼𝑓 (𝑠) 𝐾𝑇 𝐸𝑓 (𝑠)
𝜃(𝑠) = = ×
𝑠(𝐽 𝑠 + 𝑓0 ) 𝑠(𝐽 𝑠 + 𝑓0 ) 𝐿𝑓 𝑠 + 𝑅𝑎
𝜃(𝑠) 𝐾 𝑇 1
Transfer function 𝐸 = 𝑠(𝐽 𝑠+𝑓 ×𝐿
𝑓 (𝑠) 0) 𝑓𝑠 +𝑅𝑎

Using equation D and E the block diagram is

Module 3
Time response of feedback control systems:
Time response of the feedback controlled system is defined as the variation of the
controlled output of the system with respect to time when the system is subjected to
standard test signal. The time response is divided into two parts, they are
Transient response 𝒄𝒕𝒓 (𝒕)
Steady state response𝒄𝒔𝒔 (𝒕)
Typical time response of a controlled system for the specified test input as shown below.

Transient response is the part of the total response; it is defined as the part of the time
response that goes to zero as time becomes very large. Thus 𝑐𝑡 (𝑡) has a property that
lim 𝑐𝑡𝑟 (𝑡) = 0
𝑡→∞

Steady state response is the part of the total response that remains aster transient response
died out.
From the figure (b) the total response
 𝑐(𝑡) = 𝑐𝑡𝑟 (𝑡) + 𝑐𝑠𝑠 (𝑡)
Standard test signals
The different standard test signals that are used to test the control systems are
i. Step input
ii. Ramp input
iii. Parabolic input
iv. Impulse input
v. Sinusoidal input
Step input:
The Mathematical expression for step input is
𝑟(𝑡) = 𝐴𝑢(𝑡)
𝑂𝑟
𝐴 0≤𝑡≤∞
𝑟(𝑡) = {
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
The graph of 𝑟(𝑡) as shown in below figure

The Laplace transform of 𝑟(𝑡) is

𝐴
𝐿[𝑟(𝑡)] = 𝐿(𝐴𝑢(𝑡)) =
𝑠
𝐴
𝑅(𝑠) =
𝑠
Ramp input
The Mathematical expression for ramp input is
𝑟(𝑡) = 𝐴𝑡𝑢(𝑡)
𝑂𝑟
𝐴𝑡 0 ≤ 𝑡 ≤ ∞
𝑟(𝑡) = {
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
The graph of 𝑟(𝑡) as shown in below figure

The Laplace transform of 𝑟(𝑡) is

𝐴
𝐿[𝑟(𝑡)] = 𝐿(𝐴𝑡𝑢(𝑡)) =
𝑠2
𝐴
𝑅(𝑠) =
𝑠2
Parabolic input
The Mathematical expression for parabolic input is
𝑡2
𝑟(𝑡) = 𝐴 𝑢(𝑡)
2
𝑂𝑟
𝑡2
𝑟(𝑡) = { 2 0 ≤ 𝑡 ≤ ∞
𝐴
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
The graph of 𝑟(𝑡) as shown in below figure

The Laplace transform of 𝑟(𝑡) is

𝑡2 𝐴
𝐿[𝑟(𝑡)] = 𝐿 (𝐴 𝑢(𝑡)) = 3
2 𝑠
𝐴
𝑅(𝑠) =
𝑠3
Impulse input
The unit impulse function is defined as a signal which has zero value everywhere,
except𝑡 = 0. The mathematical expression for impulse input is
1 𝑡=0
𝑟(𝑡) = 𝛿(𝑡) = {
0 𝑡≠0
The graph of 𝑟(𝑡) as shown in below figure

Laplace transform of 𝛿(𝑡) is

𝐿[𝑟(𝑡)] = 𝐿(𝛿(𝑡)) = 1
𝑅(𝑠) = 1
Sinusoidal input
The mathematical expression for sinusoidal input is
𝐴𝑠𝑖𝑛𝜔𝑡 𝑡 ≥ 0
𝑟(𝑡) = = {
0 𝑡 <0
The graph of 𝑟(𝑡) as shown in below figure

Laplace transform of 𝑟(𝑡) is

𝐴𝜔
𝐿[𝑟(𝑡)] = 𝐿( 𝐴𝑠𝑖𝑛𝜔𝑡) =
𝑠2 + 𝜔2
Note
Laplace transform table
1 L{Au(t)} 𝐴
𝑠
2 𝐿(𝑒 ±𝑎𝑡 ) 1
𝑠∓𝑎
 3 L{sin(ωt) 𝜔
𝑠2
+ 𝜔2
4 L{cos (ωt) 𝑠
𝑠 + 𝜔2
2

5 L{𝑒 ±𝑎𝑡 sin(ωt) 𝜔

(𝑠 ∓ 𝑎)2 + 𝜔 2
6 L{𝑒 ±𝑎𝑡 cos(ωt) 𝑠∓𝑎
(𝑠 ∓ 𝑎)2 + 𝜔 2
7 𝐿(𝑡 𝑛 𝑢(𝑡)) 𝑛!
𝑠 𝑛+1

Time response of first order system to unit step input

Let us consider the first order system with unity feed back as shown in below figure.

The transfer function of the system is

𝐶(𝑠) 1
=
𝑅(𝑠) 𝑇𝑠 + 1
T is the time constant of the system, the highest power of the s defines the order of the
system.
The output of the system is expressed as
𝑅(𝑠)
𝐶(𝑠) =
𝑇𝑠 + 1
1
If the input is step input, then 𝑅(𝑠) = 𝑠 , now
1 𝐴 𝐵 1 𝑇
𝐶(𝑠) = = + = −
𝑠(𝑇𝑠 + 1) 𝑠 𝑇𝑠 + 1 𝑠 𝑇𝑠 + 1
Taking the inverse Laplace transform, we get
−𝑡
𝑐(𝑡) = 1 − 𝑒 𝑇 0≤𝑡≤∞
−𝑡
1 𝑒𝑇
𝑐(𝑡) = − 𝑡>0
𝑇𝑟𝑎𝑛𝑠𝑖𝑐𝑒𝑛𝑡
⏟ 𝑟𝑒𝑠𝑝𝑜𝑛𝑠 ⏟𝑆𝑡𝑒𝑎𝑑𝑦 𝑠𝑡𝑎𝑡𝑒 𝑟𝑒𝑠𝑝𝑜𝑛𝑠𝑒
 −𝑡
𝑢(𝑡) 𝑒 𝑇 𝑢(𝑡)
−𝑡
𝑐(𝑡) = − = (1 − 𝑒 𝑇 ) 𝑢(𝑡)
𝑇𝑟𝑎𝑛𝑠𝑖𝑐𝑒𝑛𝑡 𝑟𝑒𝑠𝑝𝑜𝑛𝑠 𝑆𝑡𝑒𝑎𝑑𝑦
⏟ ⏟ 𝑠𝑡𝑎𝑡𝑒 𝑟𝑒𝑠𝑝𝑜𝑛𝑠𝑒

The above equation state that, initially at 𝑡 = 0 the output is zero, as 𝑡 approaches to ∞
output is unity. At 𝑡 = 𝑇, the value of the output 𝑐(𝑡) = 0.632 or output 𝑐(𝑡) reaches to
63.2% of the total output. If the time constant is small, response is faster. The time
response of first order system is as shown in below figure. The slope of the tangent at 𝑡 =
1
0 is 𝑇. That is
𝑑𝑐(𝑡) 1 −𝑡 1
| = 𝑒𝑇 =
𝑑𝑡 𝑡=0 𝑇 𝑇

The error response of the system is given by

−𝑡
𝑒(𝑡) = 𝑟(𝑡) − 𝑐(𝑡) = 𝑒 𝑇
This is plotted in the above graph. The steady state error is given by
𝑒𝑠𝑠 (𝑡) = lim 𝑒(𝑡) = 0
𝑡→∞

Rise time𝑇𝑟 : The rise time is the time taken for the output to rise from 10% to 90% of its
final value. For a first-order system, the rise time is approximately equal to
𝑇𝑟 = 2.2𝑇
Settling time𝑇𝑠 : The settling time is the time required for the output to settle within a
certain percentage (usually 2% or 5%) of the final value. For a first-order system, the
settling time is approximately equal to
 𝑇𝑠 = 4𝑇
Time response of second order system to the unit step input
Let us consider the second order system with unity feed back as shown in below figure.

The closed loop transfer function of the system is

𝐶(𝑠) 𝐺(𝑠) 𝜔𝑛2
= = 2
𝑅(𝑠) 1 + 𝐺(𝑠)𝐻(𝑠) 𝑠 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
𝐶(𝑠)
We know that, is the closed loop transfer function. The roots of the denominator of
𝑅(𝑠)
𝐶(𝑠) 𝐶(𝑠)
are known as poles of the system and roots of numerator of are known as zeros
𝑅(𝑠) 𝑅(𝑠)

of the system. The denominator 1 + 𝐺(𝑠)𝐻(𝑠) = 0 or 𝑠 2 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2 = 0 is known

as characteristic equation of the system. The roots of this characteristic equation is given
by
𝑠 2 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2 = 0
−2𝜉𝜔𝑛 ± √4𝜉 2 𝜔𝑛2 − 4𝜔𝑛2
𝑠1 , 𝑠2 =
2
𝑠1 , 𝑠2 = −𝜉𝜔𝑛 ± √𝜉 2 𝜔𝑛2 − 𝜔𝑛2

𝑠1 , 𝑠2 = −𝜉𝜔𝑛 ± 𝜔𝑛 √𝜉 2 − 1
𝜔𝑛 =is known as the undamped natural frequency of oscillation in Rad/sec
𝜉 = is the damping ration
poles are purely imaginary
If the damping ratio 𝜉 = 0 𝑠1 , 𝑠2 = ± 𝑗𝜔𝑛 ; {
and the system is undamped
poles are real and equal
If the damping ratio 𝜉 = 1, 𝑠1 , 𝑠2 = −𝜔𝑛 ; {
and the system is critically damped
If the damping ratio 𝜉 > 1,𝑠1 , 𝑠2 = −𝜉𝜔𝑛 ± 𝜔𝑛 √𝜉 2 − 1
poles are real and unequal
;{
and the system is overdamped
If the damping ratio 0<𝜉<1 𝑠1 , 𝑠2 = −𝜉𝜔𝑛 ± 𝑗𝜔𝑛 √1 − 𝜉 2 = −𝜉𝜔𝑛 ± 𝑗𝜔𝑑
poles are complex
;{
and the system is underdamped or damped
Note that, 𝜔𝑑 = 𝜔𝑛 √1 − 𝜉 2 is called damped frequency of oscillation of system in
Rad/sec

Let us consider under damped case, 0 < 𝜉 < 1 and input is step input
We know that, overall transfer function or closed loop transfer function of the second
order system is given by
𝐶(𝑠) 𝜔𝑛2
= 2
𝑅(𝑠) 𝑠 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
𝜔𝑛2
𝐶(𝑠) = 2 × 𝑅(𝑠)
𝑠 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
1
If the input to the system is unit step input i.e. 𝑟(𝑡) = 𝑢(𝑡)and𝑅(𝑠) = . Therefore
𝑠

𝜔𝑛2 1 𝐴 𝐵𝑠 + 𝐶
𝐶(𝑠) = × = + 2
𝑠 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛 𝑠 𝑠 𝑠 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
2 2

Where A,B and C residues of the partial fraction, have to be determined.

𝜔𝑛2 = 𝐴(𝑠 2 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2 ) + 𝑠(𝐵𝑠 + 𝐶)
Put 𝑠 = 0
𝜔𝑛2 = 𝐴𝜔𝑛2 , ∴𝐴=1
By comparing the coefficients of 𝑠 2
0=𝐴+𝐵 ∴ 𝐵 = −1
By comparing the coefficients of 𝑠
0 = 𝐴2𝜉𝜔𝑛 + 𝐶 ∴ 𝐶 = −2𝜉𝜔𝑛
1 −𝑠 − 2𝜉𝜔𝑛
∴ 𝐶(𝑠) = + 2
𝑠 𝑠 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
1 𝑠 + 2𝜉𝜔𝑛
𝐶(𝑠) = − 2
𝑠 𝑠 + 2𝜉𝜔𝑛 𝑠 + ( 𝜉𝜔𝑛 )2 − ( 𝜉𝜔𝑛 )2 + 𝜔𝑛2
1 𝑠 + 𝜉𝜔𝑛 + 𝜉𝜔𝑛
𝐶(𝑠) = −
𝑠 (𝑠 + 𝜉𝜔𝑛 )2 + 𝜔𝑛2 − 𝜉 2 𝜔𝑛 2
1 𝑠 + 𝜉𝜔𝑛 + 𝜉𝜔𝑛
𝐶(𝑠) = −
𝑠 (𝑠 + 𝜉𝜔𝑛 )2 + 𝜔𝑛2 (1 − 𝜉 2 )
1 (𝑠 + 𝜉𝜔𝑛 ) + 𝜉𝜔𝑛
𝐶(𝑠) = −
𝑠 (𝑠 + 𝜉𝜔𝑛 )2 + 𝜔𝑑2
1 (𝑠 + 𝜉𝜔𝑛 ) 𝜉𝜔𝑛
𝐶(𝑠) = − −
𝑠 (𝑠 + 𝜉𝜔𝑛 ) + 𝜔𝑑 (𝑠 + 𝜉𝜔𝑛 )2 + 𝜔𝑑2
2 2

1 (𝑠 + 𝜉𝜔𝑛 ) 𝜔𝑑 𝜉𝜔𝑛
𝐶(𝑠) = − −
𝑠 (𝑠 + 𝜉𝜔𝑛 )2 + 𝜔𝑑2 𝜔𝑑 [(𝑠 + 𝜉𝜔𝑛 )2 + 𝜔𝑑2 ]
1 (𝑠 + 𝜉𝜔𝑛 ) (𝜔𝑛 𝜉) 𝜔𝑑
𝐶(𝑠) = − 2 – ×
2
𝑠 (𝑠 + 𝜉𝜔𝑛 ) + 𝜔𝑑 𝜔𝑑 (𝑠 + 𝜉𝜔𝑛 )2 + 𝜔𝑑2
Taking the inverse Laplace transform
(𝜔𝑛 𝜉) −𝜉𝜔 𝑡
𝑐(𝑡) = 1 − 𝑒 −𝜉 𝜔𝑛𝑡 𝑐𝑜𝑠𝜔𝑑 𝑡– 𝑒 𝑛 𝑠𝑖𝑛𝜔 𝑡
𝑑
𝜔𝑑
(𝜔𝑛 𝜉)
𝑐(𝑡) = 1 − 𝑒 −𝜉 𝜔𝑛𝑡 𝑐𝑜𝑠𝜔𝑑 𝑡– 𝑒 −𝜉𝜔𝑛𝑡 𝑠𝑖𝑛𝜔𝑑 𝑡
𝜔𝑛 √1 − 𝜉2
𝜉
𝑐(𝑡) = 1 − 𝑒 −𝜉 𝜔𝑛𝑡 𝑐𝑜𝑠𝜔𝑑 𝑡– 𝑒 −𝜉𝜔𝑛𝑡 𝑠𝑖𝑛𝜔𝑑 𝑡
√1 − 𝜉2
𝑒 −𝜉 𝜔𝑛𝑡
𝑐(𝑡) = 1 − [√1 − 𝜉 2 𝑐𝑜𝑠𝜔𝑑 𝑡 + 𝜉 𝑠𝑖𝑛𝜔𝑑 𝑡]
√1 − 𝜉2

Construct the right angle triangle using 𝜉, √1 − 𝜉 2 𝑎𝑛𝑑 1 as shown in below figure
 √1−𝜉 2
From the triangle √1 − 𝜉 2 = 𝑠𝑖𝑛𝜃 and 𝜉 = 𝑐𝑜𝑠𝜃, thus 𝑡𝑎𝑛𝜃 = , 𝜃=
𝜉

√1−𝜉 2
𝑡𝑎𝑛−1 ( )
𝜉

𝑒 −𝜉 𝜔𝑛𝑡
𝑐(𝑡) = 1 − [𝑠𝑖𝑛𝜃𝑐𝑜𝑠𝜔𝑑 𝑡 + 𝑐𝑜𝑠𝜃𝑒 −𝜉𝜔𝑛𝑡 𝑠𝑖𝑛𝜔𝑑 𝑡]
√1 − 𝜉2
𝑒 −𝜉 𝜔𝑛 𝑡
𝑐(𝑡) = 1 − [ sin (𝜔𝑑 𝑡 + 𝜃 )] For 𝑡 > 0
√1−𝜉 2

𝑒 −𝜉 𝜔𝑛𝑡 √1 − 𝜉 2
𝑐(𝑡) = 1 − [ sin (𝜔𝑑 𝑡 + 𝑡𝑎𝑛−1 ( ))]
√1 − 𝜉 2 𝜉

The steady state value of 𝑐(𝑡) is given as

𝑐𝑠𝑠 (𝑡) = lim 𝑐(𝑡) = 1
𝑡→∞

The time response of an underdamped second order system is plotted in below figure. It
is damped sinusoidal. The response reaches a steady state value 𝑐(𝑡) = 1 and steady
state error approaches to zero.

Time response specifications of second order systems

The first order system, the step response can be completely described by specifying the
value of the time constant T. The second order control systems are generally designed
with damping ratio less than one, i.e. oscillatory step response. Therefore the time
response of second order or higher control systems subjected to a step input is generally
of damped oscillatory nature as shown in below figure. From the figure, it is observed
that the step response has number of overshoots and undershoots with respect to its final
steady state response. These overshoot, undershoots decays exponentially.

The time response specifications are

Delay time 𝑡𝑑
Rise time 𝑡𝑟
Peak time 𝑡𝑝
Peak over shoot 𝑀𝑝
Settling time 𝑡𝑠
Delay time 𝒕𝒅 : Delay time is the time required for the response to reach 50% of the final
value at very first time.
Rise time 𝒕𝒓 : Rise time is the time required for the response to rise from 0% to 100% of
the final value.
Peak time 𝒕𝒑 : Peak time is the time required for the response to reach peak of time
response.
Peak over shoot 𝑴𝒑 : Peak over shoot is the maximum value of the response curve
measured from unity. It is also defined as the difference between peak time response and
steady state output, it is given by 𝑀𝑝 = 𝑐(𝑡𝑝 ) − 𝑐(∞).
More often represented as a percentage, Peak over shoot in percentage is
𝑐(𝑡𝑝 ) − 𝑐(∞)
%𝑀𝑝 = × 100%
𝑐(∞)
Settling time 𝒕𝒔 : Settling time is the time required for the response to reach and stay
within the specified band (usually 2% or 5%) of its final value.
Steady state error 𝒆𝒔𝒔 (𝒕): It indicates the error between the desired output and actual
output as t tends to infinity.
𝑒𝑠𝑠 (𝑡) = lim [𝑟(𝑡) − 𝑐(𝑡)]
𝑡→∞

Expression for rise time 𝒕𝒓 :

The response of the second order system subjected to unit step input is given by
𝑒 −𝜉 𝜔𝑛𝑡
𝑐(𝑡) = 1 − [ sin (𝜔𝑑 𝑡 + 𝜃 )] … … . (1)
√1 − 𝜉 2
From the time response specification figure, at rise time 𝑐(𝑡) = 𝑐(𝑡𝑟 ) = 1
By substituting in equation (1), we get
𝑒 −𝜉 𝜔𝑛𝑡𝑟
1= 1− [ sin (𝜔𝑑 𝑡𝑟 + 𝜃 )]
√1 − 𝜉 2
𝑒 −𝜉 𝜔𝑛𝑡𝑟
[sin(𝜔𝑑 𝑡𝑟 + 𝜃 )] = 0
√1 − 𝜉 2
sin(𝜔𝑑 𝑡𝑟 + 𝜃 ) = 0
𝜔𝑑 𝑡𝑟 + 𝜃 = 𝑛𝜋 𝑛 = 1,2,3
𝜔𝑑 𝑡𝑟 = 𝑛𝜋 − 𝜃
𝑛𝜋 − 𝜃
𝑡𝑟 =
𝜔𝑑
 𝜋−𝜃
For the first time 𝑛 = 1, then 𝑡𝑟 = 𝜔𝑑

√1−𝜉 2
𝜋 − 𝑡𝑎𝑛−1 ( 𝜉
𝑡𝑟 =
𝜔𝑛 √1 − 𝜉2
Expression for peak time 𝒕𝒑 : The response of the second order system subjected to unit
step input is given by
𝑒 −𝜉 𝜔𝑛𝑡
𝑐(𝑡) = 1 − [ sin (𝜔𝑑 𝑡 + 𝜃 )] … … . (1)
√1 − 𝜉 2
From the time response specification figure at time 𝑡 = 𝑡𝑝 , the slope of the 𝑐(𝑡)must be
𝑑𝑐(𝑡)
zero. Therefore, =0
𝑑𝑡

𝑑𝑐(𝑡) 𝑒 −𝜉 𝜔𝑛𝑡𝑝 𝑒 −𝜉 𝜔𝑛𝑡𝑝

| =0−[ (−𝜉 𝜔𝑛 ) sin(𝜔𝑑 𝑡𝑝 + 𝜃 ) + (𝜔𝑑 )cos (𝜔𝑑 𝑡𝑝 + 𝜃)]
𝑑𝑡 𝑡=𝑡𝑝 √1 − 𝜉 2 √1 − 𝜉 2
𝜉 𝜔𝑛 sin(𝜔𝑑 𝑡𝑝 + 𝜃 ) − 𝜔𝑑 cos(𝜔𝑑 𝑡𝑝 + 𝜃) = 0

Put 𝜔𝑑 = 𝜔𝑛 √1 − 𝜉 2

𝜉 𝜔𝑛 sin(𝜔𝑑 𝑡𝑝 + 𝜃 ) − 𝜔𝑛 √1 − 𝜉 2 cos(𝜔𝑑 𝑡𝑝 + 𝜃) = 0

𝜉 sin(𝜔𝑑 𝑡𝑝 + 𝜃 ) − √1 − 𝜉 2 cos(𝜔𝑑 𝑡𝑝 + 𝜃) = 0

We know that 𝑐𝑜𝑠𝜃 = 𝜉 and 𝑠𝑖𝑛𝜃 = √1 − 𝜉 2

cosθsin(𝜔𝑑 𝑡𝑝 + 𝜃 ) − 𝑠𝑖𝑛𝜃cos(𝜔𝑑 𝑡𝑝 + 𝜃) = 0
sin(𝜔𝑑 𝑡𝑝 + 𝜃 )cosθ − cos(𝜔𝑑 𝑡𝑝 + 𝜃) 𝑠𝑖𝑛𝜃 = 0
𝒔𝒊𝒏𝑨𝒄𝒐𝒔𝑩 + 𝒄𝒐𝒔𝑨𝒔𝒊𝒏𝑩 = 𝒔𝒊𝒏(𝑨 − 𝑩)
sin(𝜔𝑑 𝑡𝑝 + 𝜃 − 𝜃) = 0
sin(𝜔𝑑 𝑡𝑝 ) = 0
𝜔𝑑 𝑡𝑝 = 𝑛𝜋
𝑛𝜋
𝑡𝑝 =
𝜔𝑑
For the first peak over shoot 𝑛 = 1, then
𝜋 𝜋
𝑡𝑝 = =
𝜔𝑑 𝜔𝑛 √1 − 𝜉 2
For the second over shoot
 2𝜋
𝑡𝑝 =
𝜔𝑛 √1 − 𝜉 2
3𝜋
For the third over shoot 𝑡𝑝 =
𝜔𝑛 √1−𝜉 2

Expression for peak over shoot 𝑴𝒑 : Peak over shoot is the maximum value of the
response curve measured from unity. The response of the second order system subjected
to unit step input is given by
𝑒 −𝜉 𝜔𝑛𝑡
𝑐(𝑡) = 1 − [ sin (𝜔𝑑 𝑡 + 𝜃 )] … … . (1)
√1 − 𝜉 2
By substituting 𝑡 = 𝑡𝑝 in the above equation we get peak response
𝑒 −𝜉 𝜔𝑛 𝑡𝑝
𝑐(𝑡𝑝 ) = 1 − [ sin (𝜔𝑑 𝑡𝑝 + 𝜃 )]
√1 − 𝜉 2
𝜋 𝜋
Put 𝑡𝑝 = 𝜔 = then
𝑑 𝜔𝑛 √1−𝜉 2
𝜋
−𝜉 𝜔𝑛 ×
𝑒 𝜔𝑛 √1 − 𝜉 2 𝜋
𝑐(𝑡𝑝 ) = 1 − [ sin (𝜔𝑑 × + 𝜃 )]
√1 − 𝜉 2 𝜔𝑑
𝜉𝜋
−
√1−𝜉2
𝑒
𝑐(𝑡𝑝 ) = 1 − [ sin (𝜋 + 𝜃 )]
√1 − 𝜉 2
𝜉𝜋
−
√1−𝜉2 1
𝑐(𝑡𝑝 ) = 1 − 𝑒 (− sin 𝜃)
√1 − 𝜉 2
𝜉𝜋
−
√1−𝜉2 1
𝑐(𝑡𝑝 ) = 1 + 𝑒 sin 𝜃
√1 − 𝜉 2

We know that sin 𝜃 = √1 − 𝜉 2

𝜉𝜋
−
√1−𝜉2 √1 − 𝜉 2
𝑐(𝑡𝑝 ) = 1 + 𝑒
√1 − 𝜉 2
𝜉𝜋
−
√1−𝜉2
𝑐(𝑡𝑝 ) = 1 + 𝑒
We know that peak over shoot 𝑀𝑝 = 𝑐(𝑡𝑝 ) − 𝑐(∞)
𝜉𝜋
−
√1−𝜉2
𝑀𝑝 = 1 + 𝑒 −1
 𝜉𝜋
−
√1−𝜉2
𝑀𝑝 = 𝑒
𝜉𝜋
−
𝑐(𝑡𝑝 )−𝑐(∞) √1−𝜉2
Percentage over shoot is %𝑀𝑝 = %=𝑒 × 100%
𝑐(∞)
𝜉𝜋
−
√1−𝜉2
%𝑀𝑝 = 𝑒 × 100%

Expression for peak settling time 𝒕𝒔 : Settling time is the time required for the response
to reach and stay within the specified band (usually 2% or 5%) of its final value.
Consider the step response of the second order system we observed that the response is
oscillatory and decaying exponentially as shown in below figure.

𝑒 −𝜉 𝜔𝑛 𝑡
The response is within the envelop1 ± , from the response
√1−𝜉 2

If we considered 5% tolerance, upper bound is given by

𝑒 −𝜉 𝜔𝑛𝑡
1+ = 1.05
√1 − 𝜉 2
And the lower bound is given by
𝑒 −𝜉 𝜔𝑛𝑡
1− = 0.95
√1 − 𝜉 2
 𝑒 −𝜉 𝜔𝑛𝑡𝑠
In both the cases, = 0.05
√1−𝜉 2

𝑒 −𝜉 𝜔𝑛𝑡𝑠 = 0.05√1 − 𝜉 2
Take natural log on both sides we get
−𝜉 𝜔𝑛 𝑡𝑠 = ln(0.05) + ln (√1 − 𝜉 2 )

−𝜉 𝜔𝑛 𝑡𝑠 = −2.99 + ln (√1 − 𝜉 2 )

For low value of 𝜉 , ln (√1 − 𝜉 2 = 0 thus

−𝜉 𝜔𝑛 𝑡𝑠 = −2.99
3
𝑡𝑠 =
𝜉 𝜔𝑛
If we considered 2% tolerance, upper bound is given by
𝑒 −𝜉 𝜔𝑛𝑡
1+ = 1.02
√1 − 𝜉 2
And the lower bound is given by
𝑒 −𝜉 𝜔𝑛𝑡
1− = 0.98
√1 − 𝜉 2
𝑒 −𝜉 𝜔𝑛𝑡𝑠
In both the cases, = 0.02
√1−𝜉 2

𝑒 −𝜉 𝜔𝑛𝑡𝑠 = 0.02√1 − 𝜉 2
Take natural log on both sides we get
−𝜉 𝜔𝑛 𝑡𝑠 = ln(0.02) + ln (√1 − 𝜉 2 )

−𝜉 𝜔𝑛 𝑡𝑠 = −4 + ln (√1 − 𝜉 2 )

For low value of 𝜉 , ln (√1 − 𝜉 2 = 0 thus

−𝜉 𝜔𝑛 𝑡𝑠 = −4
4
𝑡𝑠 =
𝜉 𝜔𝑛
𝐶(𝑠) 1
Problem: A first order control system is represented by its transfer function𝑅(𝑠) = 𝑠+4.

Determine time constant and response of the system for the step input.
𝐶(𝑠) 1
Solution: 𝑅(𝑠) = 𝑠+4
𝑅(𝑠) 1 𝐴 𝐵 1 1
𝐶(𝑠) = = 𝑠(𝑠+4) = 𝑠 + 𝑠+4 = 4𝑠 − 4(𝑠+4)
𝑠+4
 1 1
𝑐(𝑡) = 𝑢(𝑡) − 𝑒 −4𝑡
4 4
1
The time constant is 𝑇 = 4 second

Problem: The time response of the second order control system for the step input
is𝑐(𝑡) = 1 + 0.2𝑒 −60𝑡 − 1.2𝑒 10𝑡 , determine the transfer function, undamped natural
frequency and damping ratio.
Solution: 𝑐(𝑡) = 1 + 0.2𝑒 −60𝑡 − 1.2𝑒 −10𝑡
Apply the Laplace transform to the above equation
1 0.2 1.2 600
𝐶(𝑠) = + − =
𝑠 𝑠 + 60 𝑠 + 10 𝑠(𝑠 + 60)(𝑠 + 10)
The transfer function of the system is
𝐶(𝑠) 600
= 2
𝑅(𝑠) 𝑠 + 70𝑠 + 600
The above equation is in the form of
𝐶(𝑠) 𝜔2
= 𝑠2 +2𝜉𝜔𝑛 𝑠+𝜔2 , 𝜔𝑛2 = 600
𝑅(𝑠) 𝑛 𝑛

Undamped natural frequency 𝜔𝑛 = √600 = 24.29Rad/sec

2𝜉𝜔𝑛 = 70
70 70
Damping ratio 𝜉 = 2𝜔 = 2×24.29 = 1.43
𝑛

Problem: A position control system as shown in below figure. What is the response 𝑐(𝑡)
to the unit step input, given 𝜉 = 0.5? Also calculate rise time, peak time, maximum
overshoot and settling time.

Solution: The closed loop transfer function

16
𝐶(𝑠) 𝐺(𝑠) 𝑠(𝑠+0.8) 16
= = 16 =
𝑅(𝑠) 1 + 𝐺(𝑠)𝐻(𝑠) 1 + (𝐾𝑠 + 1)( ) 𝑠(𝑠 + 0.8) + 16(𝐾𝑠 + 1)
𝑠(𝑠+0.8)
𝐶(𝑠) 16
= 𝑠2 +(0.8+16𝐾)𝑠+16 . The is the second order system, thus we can compare this with
𝑅(𝑠)

𝜔2
the standard closed loop transfer function 𝑠2 +2𝜉𝜔𝑛 𝑠+𝜔2 .
𝑛 𝑛

Now 𝜔𝑛2 = 16, thus 𝜔𝑛 = 4𝑅𝑎𝑑/𝑠𝑒𝑐

0.8 + 16𝐾 = 2𝜉𝜔𝑛
2𝜉𝜔𝑛 − 0.8 2 × 0.5 × 4 − 0.8
𝐾= = = 0.2
16 16
𝐶(𝑠) 16 16
= 2 = 2
𝑅(𝑠) 𝑠 + (0.8 + 16 × 0.2)𝑠 + 16 𝑠 + 4𝑠 + 16
16𝑅(𝑠) 16
𝐶(𝑠) = =
𝑠2 2
+ 4𝑠 + 16 𝑠(𝑠 + 4𝑠 + 16)
𝐴 𝐵𝑠 + 𝐶
𝐶(𝑠) = + 2
𝑠 𝑠 + 4𝑠 + 16
𝐴 = 1, 𝐵 = −1, 𝐶 = −4
1 −𝑠 − 4 1 𝑠+4
𝐶(𝑠) =
+ 2 = − 2
𝑠 𝑠 + 4𝑠 + 16 𝑠 𝑠 + 4𝑠 + 16
1 𝑠+4 1 𝑠+4 1 𝑠+4
𝐶(𝑠) = − 2 = − 2 = −
𝑠 𝑠 + 4𝑠 + 4 − 4 + 16 𝑠 𝑠 + 4𝑠 + 4 + 12 𝑠 (𝑠 + 2)2 + 12
1 𝑠+2+2 1 𝑠+2 2 √12
𝐶(𝑠) = − = − − ×
𝑠 (𝑠 + 2)2 + 12 𝑠 (𝑠 + 2)2 + 12 √12 (𝑠 + 2)2 + 12
1 𝑠+2 2 √12
𝐶(𝑠) = − − ×
𝑠 (𝑠 + 2)2 + 12 2√3 (𝑠 + 2)2 + (√12)2
1
𝑐(𝑡) = 1 − 𝑒 −2𝑡 𝑐𝑜𝑠√12𝑡 − 𝑒 −2𝑡 𝑠𝑖𝑛√12𝑡
√3
1
𝑐(𝑡) = 1 − 𝑒 −2𝑡 [ 𝑠𝑖𝑛√12𝑡 + 𝑐𝑜𝑠√12𝑡]
√3
From the given data 𝜉 = 0.5 and 𝜔𝑑 = 𝜔𝑛 √1 − 𝜉 2 = 4√1 − 0.52 = 3.464 𝑅𝑎𝑑/𝑠𝑒𝑐

Let us construct the triangle 𝜉 = 0.5 and √1 − 𝜉 2 = 0.866 as shown in below figure
 0.866
𝜃 = 𝑡𝑎𝑛−1 ( ) = 600 = 1.047𝑟𝑎𝑑
0.5
𝜋−𝜃 𝜋−1.047
Rise time𝑡𝑟 = = = 0.60𝑠𝑒𝑐
𝜔𝑑 3.464
𝜋 𝜋
Peak time 𝑡𝑝 = 𝜔 = 3.464 = 0.907𝑠𝑒𝑐
𝑑

−𝜉𝜋
−0.5𝜋
√1−𝜉2 =0.163
Maximum overshoot 𝑀𝑝 = 𝑒 = 𝑒 √1−0.52
3 3
Settling time for 5% tolerance 𝑡𝑠 = 𝜉 𝜔 = 0.5×4 = 0.5𝑆𝑒𝑐
𝑛

𝑘1
Problem: A servo mechanical system𝐺(𝑠) = , 𝐻(𝑠) = 1 + 𝑘2 , determine the value of
𝑠2

𝑘1 and 𝑘2 , so that the peak over shoot to unit step input is 0.25 and peak time is 2𝑠𝑒𝑐.
𝑘1
𝐶(𝑠) 𝐺(𝑠) 𝑠2 𝑘1
Solution: The overall transfer function 𝑅(𝑠) = 1+𝐺(𝑠)𝐻(𝑠) = 𝑘1 = 𝑠2 +𝑘
1+( 2 )(1+𝑘2 ) 1 𝑘2 𝑠+𝑘1
𝑠

Comparing 𝑠 2 + 𝑘1 𝑘2 𝑠 + 𝑘1 with 𝑠 2 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2 we get

𝑘1 = 𝜔𝑛2 and 𝑘1 𝑘2 = 2𝜉𝜔𝑛
−𝜉𝜋

√1−𝜉2
Given 𝑀𝑝 = 0.25 = 𝑒
−𝜉𝜋
𝑙𝑛(0,25) =
√1 − 𝜉 2
−𝜉𝜋
−1.38 =
√1 − 𝜉 2
(𝜉𝜋)2
Square on both the side we get 1.92 = , solving this we get 𝜉 = 0.4036
1−𝜉 2
𝜋
Peak time 𝑡𝑝 = 𝜔 = 2
𝑑

𝜋
𝜔𝑑 = = 1.57𝑟𝑎𝑑/𝑠𝑒𝑐
2
𝜔𝑑 = 𝜔𝑛 √1 − 𝜉 2
𝜔𝑑 1.57 1.57
𝜔𝑛 = = = = 1.71𝑅𝑎𝑑/𝑠𝑒𝑐
√1 − 𝜉 2 √1 − 0.40362 0.83

𝑘1 = 𝜔𝑛2 = 1.712 = 2.94

𝑘1 𝑘2 = 2𝜉𝜔𝑛
2𝜉𝜔𝑛 2 × 0.4036 × 1.71
𝑘2 = = = 0.4702
𝑘1 2.94
Problem: A negative feedback control system has the following transfer function,𝐺(𝑠) =
9
and 𝐻(𝑠) = 1. Determine the natural frequency, damping ratio, damped frequency
𝑠(𝑠+2)

of oscillation, rise time, % peak over shoot, peak time and approximate 5% settling time.
Assume unit step input.
Solution: The negative feedback system as shown in below figure, the overall transfer
𝐶(𝑠) 𝐺(𝑠)
function of the system is = 1+𝐺(𝑠)𝐻(𝑠)
𝑅(𝑠)
9
𝐶(𝑠) 𝑠(𝑠+2) 9 𝜔𝑛2
= = =
𝑅(𝑠) 1 + 9 × 1 𝑠 2 + 2𝑠 + 9 𝑠 2 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
𝑠(𝑠+2)

On comparing the equation

Natural frequency of oscillations 𝜔𝑛 =3 rad/Sec
2𝜉𝜔𝑛 = 2,
1
Damping ratio 𝜉 = 3 = 0.33

12
Damped frequency of oscillation𝜔𝑑 = 𝜔𝑛 √1 − 𝜉 2 = 3√1 − 9 = 2.828Rad/sec

1
𝜋−𝜃 √1−𝜉 2 (√1−( )2
9
Rise time𝑡𝑟 = , 𝜃= 𝑡𝑎𝑛−1 ( 𝜉 ) = 𝑡𝑎𝑛 −1
1 = 70.50
𝜔𝑑
9

𝜃 = 1.23 𝑟𝑎𝑑
𝜋−𝜃 𝜋−1.23
Rise time𝑡𝑟 = = = 0.67Sec
𝜔𝑑 2.828

% peak over shoot

𝜉𝜋
− 0.33×𝜋
√1−𝜉2 − ×100%
%𝑀𝑝 = 𝑒 × 100% = 𝑒 √1−0.332 = 32.93%
𝜋 𝜋
Peak time 𝑡𝑝 = 𝜔 = 2.828 = 1.11Sec
𝑑

3 3
Settling time for 5% tolerance 𝑡𝑠 = 𝜉𝜔 = 0.33×3 = 3Sec
𝑛

Problem: Second order control system is represented by 𝜃̈0 + 4𝜃̇0 + 25𝜃0 = 25𝜃𝑐 .
Determine the natural frequency, damped frequency of oscillation, damping factor, peak
over shoot, and peak time for the unit step input.
Solution: Consider 𝜃̈0 + 4𝜃̇0 + 25𝜃0 = 25𝜃𝑐 , corresponding differentional equation is
𝑑2 𝜃0 𝑑𝜃0
+4 + 25𝜃0 = 25𝜃𝑐
𝑑𝑡 2 𝑑𝑡
Apply Laplace transform to the above equation we get
𝑠 2 𝜃0 (𝑠) + 4𝑠𝜃0 (𝑠) + 25𝜃0 (𝑠) = 25𝜃𝑐 (𝑠)
𝐶(𝑠) 𝜃 (𝑠) 25 𝜔2
Transfer function of the system is 𝑅(𝑠) = 𝜃𝑐 (𝑠) = 𝑠2 +4𝑠+25 = 𝑠2 +2𝜉𝜔𝑛 𝑠+𝜔2
0 𝑛 𝑛

𝜔𝑛2 = 25 , Thus undamped natural frequency𝜔𝑛 = 5Rad/sec

For damping ration,2𝜉𝜔𝑛 = 4 2𝜉 × 5 = 4
𝜉 = 0.4
Damped frequency of oscillation𝜔𝑑 = 𝜔𝑛 √1 − 𝜉 2 = 5√1 − 0.42 = 4.582Rad/sec
𝜋 𝜋
Peak time 𝑡𝑝 = 𝜔 = 4.582 = 0.685Sec
𝑑
𝜉𝜋
− 0.4×𝜋
√1−𝜉2 −
Peak over shoot 𝑀𝑝 = 𝑒 =𝑒 √1−0.42 = 0.253
Problem: A unity feedback system as shown in below figure. Determine the value of A
and 𝐾1 such that damping ratio is 0.6 and damped frequency of oscillations 8rad/sec.
what is the peak value of the response.

𝐴
Solution: From the figure 𝐺(𝑠) = 𝑠(𝑠+𝑘 ) and 𝐻(𝑠) = 1
1

We know that for negative feedback system

𝐴
𝐶(𝑠) 𝐺(𝑠) 𝑠(𝑠+𝑘1 ) 𝐴 𝜔𝑛2
= = = 2 =
𝑅(𝑠) 1 + 𝐺(𝑠)𝐻(𝑠) 1 + 𝐴 𝑠 + 𝑘1 𝑠 + 𝐴 𝑠 2 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
𝑠(𝑠+𝑘 1)

On comparing transfer function with standard function

𝑘1 = 2𝜉𝜔𝑛

𝜔𝑑 = 𝜔𝑛 √1 − 𝜉 2
𝜔𝑑 8
𝜔𝑛 = = = 10 Rad/sec
√1−𝜉 2 √1−(0.6)2

𝑘1 = 2𝜉𝜔𝑛 = 2 × 0.6 × 10 = 12
𝜔𝑛2 = 𝐴
𝐴 = 100
Problem: The step response of the second order under damped unity feedback system
shown in below figure for the input 2𝑢(𝑡). Determine the closed loop and open loop
transfer function of the system.

Solution: From the figure 𝑀𝑝 = 0.5 and 𝑡𝑝 = 2𝑠𝑒𝑐 and input 𝑟(𝑡) = 2𝑢(𝑡)
𝐶(𝑠) 𝜔𝑛2
=
𝑅(𝑠) 𝑠 2 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
𝜉𝜋
−
√1−𝜉2
For step input peak overshoot 𝑀𝑝 = 𝑒 , if the input is 𝑟(𝑡) = 𝐴𝑢(𝑡) = 𝑟(𝑡) =
2𝑢(𝑡)
𝜉𝜋 𝜉𝜋
− −
√1−𝜉2 √1−𝜉2
Then peak overshoot 𝑀𝑝 = 𝐴𝑒 = 0.5 = 2𝑒
𝜉𝜋
−
0.5 √1−𝜉2
=𝑒
2
0.5 𝜉𝜋
ln ( ) = −
2 √1 − 𝜉 2
𝜉𝜋
ln( 0.25) = −
√1 − 𝜉 2
𝜉𝜋
−1.362 = −
√1 − 𝜉 2
𝜉𝜋
1.362 =
√1 − 𝜉 2
 1.362√1 − 𝜉 2 = 𝜉𝜋
Square on both sides, we get
1.92(1 − 𝜉 2 ) = (𝜋𝜉)2
1.92 − 1.92𝜉 2 = 𝜋 2 𝜉 2
1.92 = 𝜋 2 𝜉 2 + 1.92𝜉 2
1.92 = 𝜉 2 (𝜋 2 + 1.92)
1.92 = 𝜉 2 (9.82 + 1.92)
1.92 = 𝜉 2 ( 11.79)
1.92
= 𝜉2
11.79
𝜉 2 = 0.162
𝜉 = 0.4
𝜋 𝜋
Rise time𝑡𝑝 = 𝜔 = 2, 𝜔𝑑 = = 1.57 Rad/Sen
𝑑 2

𝜔𝑑 = 𝜔𝑛 √1 − 𝜉 2
𝜔𝑑 1.57 1.57 1.57 1.57
𝜔𝑛 = = √1−0.42 = = = 0.16 = 9.821Rad/sec
√1−𝜉 2 √1−0.16 √0.84

𝐶(𝑠) 𝜔2 9.8212 96.28

Closed loop transfer function 𝑅(𝑠) = 𝑠2 +2𝜉𝜔𝑛 𝑠+𝜔2 = 𝑠2 +2×0.4×9.821𝑠+9.8212 = 𝑠2 +7.84𝑠+96.28
𝑛 𝑛

96.28 96.28
𝐶(𝑠) 96.28 𝑠2 +7.84𝑠 𝑠(𝑠+7.84)
For open loop transfer function = 96.28 = 96.28 = 96.28
𝑅(𝑠) 𝑠2 +7.84𝑠(1+ 2 ) 1+ 2 1+
𝑠 +7.84𝑠 𝑠 +7.84𝑠 𝑠(𝑠+7.84)

96.28
Open loop transfer function 𝐺(𝑠) = 𝑠(𝑠+7.84) and 𝐻(𝑠) = 1

Problem: Figure shows the mechanical vibrating system, when a force of 8.9 Newton is
applied to the system, the mass oscillates as shown in below figure. Determine M, B and
k.
Solution: The equilibrium equation for the above mechanical system is
𝑑2 𝑥(𝑡) 𝑑𝑥(𝑡)
𝑀 2
+𝐵 + 𝑘𝑥(𝑡) = 𝐹(𝑡)
𝑑2 𝑑𝑡
𝑀𝑠 2 𝑋(𝑠) + 𝐵𝑠𝑋(𝑠) + 𝑘𝑋(𝑠) = 𝐹(𝑠)
𝑋(𝑠) 1
Closed loop transfer function𝐹(𝑠) = 𝑀𝑠2 +𝐵𝑠+𝑘

𝐹(𝑠)
𝑋(𝑠) =
𝑀𝑠 2 + 𝐵𝑠 + 𝑘
1
𝑋(𝑠) 1 𝑀 𝜔𝑛2
= = = 2
𝐹(𝑠) 𝑀𝑠 2 + 𝐵𝑠 + 𝑘 𝑠 2 + 𝐵 𝑠 + 𝑘 𝑠 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
𝑀 𝑀

𝐵 𝑘
𝑠2 + 𝑠 + = 𝑠 2 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
𝑀 𝑀
𝜉𝜋
−
√1−𝜉2
𝑀𝑝 = 𝐴𝑒
𝜉𝜋
−
√1−𝜉2
0.002 = 0.03𝑒
𝜉 = 0.596
𝜋 𝜋
𝑡𝑝 = 𝜔 = 2, 𝜔𝑑 = = 1.57 Rad/Sec
𝑑 2

𝜔𝑑 = 𝜔𝑛 √1 − 𝜉 2
𝜔𝑑 1.57
𝜔𝑛 = = √1−0.5962 = 1.957Rad/sec
√1−𝜉 2

𝐹(𝑡) = 8.9
8.9
𝐹(𝑠) =
𝑠
From final value theorem
𝑥(∞) = lim 𝑥(𝑡) = lim 𝑠𝑋(𝑠)
𝑡→∞ 𝑠→0

From the figure 𝑥(∞) = 0.03

0.03 = lim 𝑠𝑋(𝑠)
𝑠→0

𝐹(𝑠)
= lim 𝑠
𝑠→0 𝑀𝑠 2 + 𝐵𝑠 + 𝑘

1 8.9 8.9
= lim 𝑠 2
( ) = lim 2
=
𝑠→0 𝑀𝑠 + 𝐵𝑠 + 𝑘 𝑠 𝑠→0 𝑀𝑠 + 𝐵𝑠 + 𝑘
8.9
0.03 = ,
𝑘

𝑘 = 296.67N/m
𝑘
𝜔𝑛2 =
𝑀
𝑘 296.67
𝑀= 2
= = 77.42𝑁 − 𝑠 2 /𝑚
𝜔𝑛 1.957

𝐵
2𝜉𝜔𝑛 = 𝑀 , 𝐵 = 𝑀2𝜉𝜔𝑛 = 77.42 × 2 × 0.596 × 12.957 = 180.56𝑁 − 𝑠𝑒𝑐/𝑚
𝜋 𝜋
𝑡𝑝 = 𝜔 = 2, 𝜔𝑑 = = 1.57
𝑑 2

Problem: The unity feedback control system having an open loop transfer
𝐾
function𝐺(𝑠) = 𝑠(𝑠+10). Determine the gain K, so that the system will have a damping

ratio of 0.5. Determine settling time, peak over shoot, peak time for a unit step input.
𝐾
𝐶(𝑠) 𝐺(𝑠) 𝑠(𝑠+10) 𝐾
Solution: Closed loop transfer function 𝑅(𝑠) = 1+𝐺(𝑠) = 𝐾 = 𝑠2 +10𝑠+𝐾
1+
𝑠(𝑠+10)

𝐶(𝑠) 𝐾 𝜔𝑛2
= 2 = 2
𝑅(𝑠) 𝑠 + 10𝑠 + 𝐾 𝑠 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
Comparing the closed loop transfer function with standard closed loop transfer function,
we get
𝜔𝑛2 = 𝐾, 𝜔𝑛 = √𝐾
2𝜉𝜔𝑛 = 10
2 × 0.5 × 𝜔𝑛 = 10
𝜔𝑛 = 10 𝑅𝑎𝑑/𝑠𝑒𝑐
Thus 𝐾 = 100
−𝜉𝜋
−0.5𝜋
√1−𝜉2
Peak overshoot 𝑀𝑝 = 𝑒 =𝑒 √1−0.52 = 0.16
𝜋 𝜋 𝜋
Peak time 𝑡𝑝 = 𝜔 = = = 0.363
𝑑 𝜔𝑛 √1−𝜉 2 10√1−0.52
3 3 3
Settling time for 5% tolerance 𝑡𝑠 = 𝜉𝜔 = 0.5×10 = 5 = 0.6𝑠ec
𝑛

Problem: Consider a unity feedback control system whose open loop transfer
0.4𝑠+1
function𝐶(𝑠) = 𝑠(𝑠+0.6). Determine its transient response for step input, rise time,

maximum peak over shoot, peak time and settling time.

0.4𝑠+1
Solution: Give open loop transfer function 𝐺(𝑠) = 𝑠(𝑠+0.6) and 𝐻(𝑆) = 1
0.4𝑠+1
𝐶(𝑠) 𝐺(𝑠) 𝑠(𝑠+0.6) 0.4𝑠+1
The closed loop transfer function, = 1+𝐺(𝑠)𝐻(𝑠) = 0.4𝑠+1 = 𝑠(𝑠+0.6)+0.4+1 =
𝑅(𝑠) 1+( )(1)
𝑠(𝑠+0.6)

0.4𝑠+1
𝑠2 +𝑠+1

𝐶(𝑠) 0.4𝑠 + 1
= 2
𝑅(𝑠) 𝑠 + 𝑠 + 1
0.4𝑠 + 1
𝐶(𝑠) = × R(s)
𝑠2 + 𝑠 + 1
1
If the input to the system is step input, 𝑅(𝑠) = 𝑠
0.4𝑠 + 1
𝐶(𝑠) =
𝑠(𝑠 2 + 𝑠 + 1)
0.4𝑠+1 𝐴 𝐵𝑆+𝑐
By partial fraction expansion, 𝐶(𝑠) = 𝑠(𝑠2 +𝑠+1) = + 𝑠2 +𝑠+1
𝑠

0.4𝑠 + 1 = 𝐴(𝑠 2 + 𝑠 + 1) + (𝐵𝑆 + 𝐶)𝑠

To find 𝐴, put 𝑠 = 0 in the above equation
1=𝐴
For B and C, compare the like coefficient
0=𝐴+𝐵
0.4 = 𝐴 + 𝐶
∴ 𝐵 = −1 and 𝐶 = −0.6, thus
1 𝑆 + 0.6
𝐶(𝑠) = − 2
𝑠 𝑠 +𝑠+1
 1 𝑆 + 0.5 + 0,1 1 𝑆 + 0.5 + 0,1 1 𝑆 + 0.5 + 0,1
𝐶(𝑠) = − 1 1 = − 1 1 = − 1 2 3
𝑠 𝑠 2 + 𝑠 + ( )2 − ( )2 + 1 𝑠 (𝑠 + )2 − ( )2 + 1 𝑠 (𝑠 + ) +
⏟ 2 ⏟2 2 2 2 4

1 𝑆 + 0.6 1 𝑆 + 0.6
𝐶(𝑠) = − = −
𝑠 (𝑠 + 1)2 + (√3)2 𝑠 (𝑠 + 0.5)2 + 0.8662
2 2
1 𝑆 + 0.5 + 0.1 1 𝑆 + 0.5 0.1
𝐶(𝑠) = − = − −
𝑠 (𝑠 + 0.5)2 + 0.8662 𝑠 (𝑠 + 0.5)2 + 0.8662 (𝑠 + 0.5)2 + 0.8662
1 𝑆 + 0.5 0.1 0.866
𝐶(𝑠) = − 2 2
− 2 2
×
𝑠 (𝑠 + 0.5) + 0.866 (𝑠 + 0.5) + 0.866 0.866
1 𝑆 + 0.5 0.1 0.866
𝐶(𝑠) = − 2 2
−
𝑠 (𝑠 + 0.5) + 0.866 0.866 (𝑠 + 0.5)2 + 0.8662
Apply the inverse Laplace transform we get
𝑐(𝑡) = 1 − 𝑒 −0.5𝑡 cos(0.866𝑡) − 0.115𝑒 −0.5𝑡 sin (0.866𝑡)
𝑐(𝑡) = 1 − 𝑒 −0.5𝑡 [cos(0.866𝑡) + 0.115 sin(0.866𝑡)] … … (𝐴)
Construct the right angle triangle as shown below

1 0.115 1 1
Thus 𝑠𝑖𝑛𝜃 = 1.006 𝑎𝑛𝑑 𝑐𝑜𝑠𝜃 = 1.006, thus 𝑡𝑎𝑛𝜃 = 0.115 , 𝜃 = 𝑡𝑎𝑛−1 (0.115) = 83.430

In radian 𝜃 = 1.456 Radians

1 0.115
𝑐(𝑡) = 1 − 𝑒 −0.5𝑡 × 1.006 [ cos(0.866𝑡) + sin(0.866𝑡)]
1.006 1.006
𝑐(𝑡) = 1 − 𝑒 −0.5𝑡 × 1.006[𝑠𝑖𝑛𝜃 cos(0.866𝑡) + 𝑐𝑜𝑠𝜃 sin(0.866𝑡)]
𝑐(𝑡) = 1 − 1.006𝑒 −0.5𝑡 [sin (0.886𝑡 + 𝜃)]
The standard response of the second order system when it is subjected to step input is
𝑒 −𝜉 𝜔𝑛𝑡
𝑐(𝑡) = 1 − [ sin (𝜔𝑑 𝑡 + 𝜃 )]
√1 − 𝜉 2
Thus 𝜔𝑑 = 0.886Rad/sec
𝜋−𝜃 𝜋−1.456
Rise time 𝑡𝑟 = = = 1.9462 𝑠𝑒𝑐
𝜔𝑑 0.866
 𝜋 𝜋
Peak time𝑡𝑟 = 𝜔 = 0.866 = 3.62𝑠𝑒𝑐
𝑑

Settling time for5% tolerance is

1 + 1.006𝑒 −0.5𝑡𝑠 = 1.05
And the lower bound is given by
1 − 1.006𝑒 −0.5𝑡𝑠 = 0.95
In both the cases, 1.006𝑒 −0.5𝑡𝑠 = 0.05
0.05
𝑒 −0.5𝑡𝑠 =
1.006
Take natural log on both sides we get
0.05
−0.5𝑡𝑠 = ln (1.006)

−0.5𝑡𝑠 = −3.0017
3.0017
𝑡𝑠 = = 6.003 𝑆𝑒𝑐
0.5
𝜋 𝜋
Peak time 𝑡𝑝 = 𝜔 = 0.866 = 3.62𝑆𝑒𝑐
𝑑

Maximum peak over shoot 𝑀𝑝 = 𝑐(𝑡𝑝 ) − 𝑐(∞)

Maximum over shoot is at 𝑡 = 𝑡𝑝
𝑐(𝑡) = 𝑐(𝑡𝑝 ) = 1 − 1.006𝑒 −0.5×3.62 [sin (0.886 × 3.62 + 83.43)] = 1.179
𝑀𝑝 = 𝑐(𝑡𝑝 ) − 𝑐(∞) = 1.179 − 1 = 0.179
Steady state errors and error constants
Steady state errors: Consider the closed loop control system as shown in below figure.

Steady state error is the difference between actual output and the desired output.
From the diagram
𝐸(𝑠) = 𝑅(𝑠) − 𝐵(𝑠)
𝐸(𝑠) = 𝑅(𝑠) − 𝐶(𝑠)𝐻(𝑠)
 𝐶(𝑠) = 𝐸(𝑠)𝐺(𝑠)
∴ 𝐸(𝑠) = 𝑅(𝑠) − 𝐸(𝑠)𝐺(𝑠)𝐻(𝑠)
𝐸(𝑠) + 𝐸(𝑠)𝐺(𝑠)𝐻(𝑠) = 𝑅(𝑠)
𝐸(𝑠)[1 + 𝐺(𝑠)𝐻(𝑠)] = 𝑅(𝑠)
𝑅(𝑠)
𝐸(𝑠) =
1 + 𝐺(𝑠)𝐻(𝑠)
Apply inverse Laplace transform to above equation
𝑅(𝑠)
𝑒(𝑡) = 𝐿−1 ( )
1 + 𝐺(𝑠)𝐻(𝑠)
Let steady error is𝑒𝑠𝑠 (𝑡), steady state error is defined as the value of 𝑒(𝑡) at 𝑡 → ∞
∴ 𝑒𝑠𝑠 (𝑡) = lim e(t)
t→∞

Let us consider the final value theorem

lim e(t) = lim sE(s)
t→∞ s→0

sR(s)
∴ 𝑒𝑠𝑠 (𝑡) = lim sE(s) = lim
s→0 s→0 1 + G(s)H(s)

Type number of the control system

Consider the open loop transfer function of unity feedback system is𝐺(𝑠)). Open loop
transfer function can be expressed as a ratio of poles and zero form is.
(𝑠+𝑧 )(𝑠+𝑧 )(𝑠+𝑧3 )…….
1 2
𝐺(𝑠) = 𝐾 𝑠𝑁 (𝑠+𝑝 )(𝑠+𝑝
.
` 2 )(𝑠+𝑝3 )….

Where K is constant
𝑧1 , 𝑧2 , 𝑧3 , … are zeros of the transfer function
𝑝1 , 𝑝2 , 𝑝3 , … are poles of the transfer function
N is the number of poles present at the origin in the s-plane
If N=0, no poles at the origin in the s-plane, and the system is type-0 system
If N=1, one poles at the origin in the s-plane, and the system is type-1 system
If N=2, two poles at the origin in the s-plane, and the system is type-2 system
Static error constants or error coefficients: Error constants are classified into three
based on the type of the input signal. They are
Position error constant(𝑘𝑝 )
Velocity error constant(𝑘𝑣 )
Acceleration error constant(𝑘𝑎 )
Position error constant(𝒌𝒑 ): It is defined when the system is subjected to unit step
input. We know the expression for steady state error i.e.
sR(s)
𝑒𝑠𝑠 (𝑡) = lim
s→0 1 + G(s)H(s)

1
If the input is unit step input i.e. 𝑟(𝑡) = 𝑢(𝑡)𝑎𝑛𝑑 𝑅(𝑠) = 𝑠
1
s( s ) 1 1 1
𝑒𝑠𝑠 (𝑡) = lim = lim = =
s→0 1 + G(s)H(s) s→0 1 + G(s)H(s) 1 + lim G(s)H(s) 1 + k p
s→0

Where k p = lim G(s)H(s)

s→0
1 2 (𝑠+𝑧 )(𝑠+𝑧 )(𝑠+𝑧3 )…….
Consider loop transfer function 𝐺(𝑠)𝐻(𝑠) = 𝐾 𝑠𝑁 (𝑠+𝑝 )(𝑠+𝑝
.
` 2 )(𝑠+𝑝3 )….

Type-0 system,𝑵 = 𝟎 then

(𝑠+𝑧1 )(𝑠+𝑧2 )(𝑠+𝑧3 )……. 𝑧 .𝑧 .𝑧 …..
k p = lim G(s)H(s) = lim 𝐾 (𝑠+𝑝` )(𝑠+𝑝2 )(𝑠+𝑝3 )….
= 𝐾 𝑝1.𝑝2 .𝑝3 ….. =Constant
s→0 s→0 1 2 3

1
𝑒𝑠𝑠 (𝑡) = = constant
1 + kp
Hence in type 0 systems position error constant k p is constant and steady state error is
also constant.
Type-1 system,𝑵 = 𝟏 then
1 (𝑠+𝑧 )(𝑠+𝑧 )(𝑠+𝑧3 )…….
2
k p = lim G(s)H(s) = lim 𝐾 𝑠 (𝑠+𝑝 )(𝑠+𝑝
= ∞
s→0 s→0 ` 2 )(𝑠+𝑝3 )….

1 1
𝑒𝑠𝑠 (𝑡) = = =0
1 + kp 1 + ∞
Hence in type-1 systems position error constant k p is infinity and steady state error zero.
Type-2 system,𝑵 = 𝟐 then
1 (𝑠+𝑧 )(𝑠+𝑧 )(𝑠+𝑧3 )…….
2
k p = lim G(s)H(s) = lim 𝐾 𝑠2 (𝑠+𝑝 )(𝑠+𝑝
= ∞
s→0 s→0 ` 2 )(𝑠+𝑝3 )….

1 1
𝑒𝑠𝑠 (𝑡) = = =0
1 + kp 1 + ∞
Hence in type-2 systems position error constant k p is infinity and steady state error zero
Velocity error constant(𝒌𝒗 ): It is defined when the system is subjected to unit ramp
input. We know the expression for steady state error i.e.
sR(s)
𝑒𝑠𝑠 (𝑡) = lim
s→0 1 + G(s)H(s)
 1
If the input is unit ramp input i.e. 𝑟(𝑡) = 𝑡𝑢(𝑡)𝑎𝑛𝑑 𝑅(𝑠) = 𝑠2
1
s(s2)
𝑒𝑠𝑠 (𝑡) = lim
s→0 1 + G(s)H(s)
1 1 1
= lim = =
s→0 s[1 + G(s)H(s)] lim[s + sG(s)H(s)] 0 + lim sG(s)H(s)
s→0 s→0
1 1
𝑒𝑠𝑠 (𝑡) = =
lim sG(s)H(s) k v
s→0

Where k v = lim sG(s)H(s)

s→0
1 2 (𝑠+𝑧 )(𝑠+𝑧 )(𝑠+𝑧3 )…….
Consider loop transfer function 𝐺(𝑠)𝐻(𝑠) = 𝐾 𝑠𝑁 (𝑠+𝑝 )(𝑠+𝑝
.
` 2 )(𝑠+𝑝3 )….

Type-0 system,𝑵 = 𝟎 then

(𝑠+𝑧1 )(𝑠+𝑧2 )(𝑠+𝑧3 )…….
k v = lim sG(s)H(s) = lim 𝑠𝐾 (𝑠+𝑝` )(𝑠+𝑝2 )(𝑠+𝑝3 )….
=0
s→0 s→0

1 1
𝑒𝑠𝑠 (𝑡) = = = ∞
kv 0
Hence in type 0 systems velocity error constant k v = 0 and steady state error is infinity.
Type-1 system,𝑵 = 𝟏 then
1(𝑠+𝑧 )(𝑠+𝑧 )(𝑠+𝑧3 )…….
2 𝑧 .𝑧 .𝑧 …..
k v = lim sG(s)H(s) = lim 𝑠𝐾 𝑠 (𝑠+𝑝 )(𝑠+𝑝
= 𝐾 𝑝1.𝑝2 .𝑝3 ….. = 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡
s→0 s→0 ` 2 )(𝑠+𝑝3 )…. 1 2 3

1
𝑒𝑠𝑠 (𝑡) = = constant
kv
Hence in type-1 systems velocity error constant k v is constant and steady state error is
constant.
Type-2 system,𝑵 = 𝟐 then
1(𝑠+𝑧 )(𝑠+𝑧 )(𝑠+𝑧3 )…….
2
k v = lim sG(s)H(s) = lim 𝑠𝐾 𝑠2 (𝑠+𝑝 )(𝑠+𝑝
= ∞
s→0 s→0 ` 2 )(𝑠+𝑝3 )….

1 1
𝑒𝑠𝑠 (𝑡) = = = 0
kv ∞
Hence in type-2 systems velocity error constant k v is infinity and steady state error zero.
Acceleration error constant(𝒌𝒂 ): It is defined when the system is subjected to unit
parabolic input. We know the expression for steady state error i.e.
sR(s)
𝑒𝑠𝑠 (𝑡) = lim
s→0 1 + G(s)H(s)
 𝑡2 1
If the input is unit parabolic input i.e. 𝑟(𝑡) = 𝑢(𝑡)𝑎𝑛𝑑 𝑅(𝑠) = 𝑠3
2
1
s(s3)
𝑒𝑠𝑠 (𝑡) = lim =
s→0 1 + G(s)H(s)
1 1
= lim =
s→0 s 2 [1 + G(s)H(s)] lim[s2 + s2 G(s)H(s)]
s→0
1
=
0 + lim s2 G(s)H(s)
s→0
1 1
𝑒𝑠𝑠 (𝑡) = =
lim s 2 G(s)H(s) ka
s→0
2
Where k a = lim s G(s)H(s)
s→0
1 2 (𝑠+𝑧 )(𝑠+𝑧 )(𝑠+𝑧3 )…….
Consider loop transfer function 𝐺(𝑠)𝐻(𝑠) = 𝐾 𝑠𝑁 (𝑠+𝑝 )(𝑠+𝑝
.
` 2 )(𝑠+𝑝3 )….

Type-0 system,𝑵 = 𝟎 then

(𝑠+𝑧1 )(𝑠+𝑧2 )(𝑠+𝑧3 )…….
k a = lim s2 G(s)H(s) = lim 𝑠 2 𝐾 (𝑠+𝑝` )(𝑠+𝑝2 )(𝑠+𝑝3 )….
=0
s→0 s→0

1 1
𝑒𝑠𝑠 (𝑡) = = = ∞
ka 0
Hence in type 0 systems acceleration error constant k a = 0 and steady state error is
infinity.
Type-1 system,𝑵 = 𝟏 then
(𝑠+𝑧 )(𝑠+𝑧 )(𝑠+𝑧3 )……. (𝑠+𝑧 )(𝑠+𝑧 )(𝑠+𝑧3 )…….
k a = lim s2 G(s)H(s) = lim 𝑠 2 𝐾 𝑠 (𝑠+𝑝
1 2
)(𝑠+𝑝
1
= 𝑠𝐾 𝑠 (𝑠+𝑝 2
= 0
s→0 s→0 ` 2 )(𝑠+𝑝3 )…. )(𝑠+𝑝 ` 2 )(𝑠+𝑝3 )….

1 1
𝑒𝑠𝑠 (𝑡) = = = ∞
ka 0
Hence in type-1 systems acceleration error constant k a is zero and steady state error is
infinity
Type-2 system,𝑵 = 𝟐 then
(𝑠+𝑧 )(𝑠+𝑧 )(𝑠+𝑧3 )……. 𝑧 .𝑧 .𝑧 …..
k a = lim s2 G(s)H(s) = lim 𝑠 2 𝐾 𝑠2 (𝑠+𝑝
1 2
)(𝑠+𝑝
= 𝐾 𝑝1.𝑝2 .𝑝3 ….. = 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡
s→0 s→0 ` 2 )(𝑠+𝑝3 )…. 1 2 3

1
𝑒𝑠𝑠 (𝑡) = = constant
ka
Hence in type-2 systems acceleration error constant k a is constant and steady state error
is constant. The above discussion is summarized in the below table.
 Table: Steady state constant or error coefficients for various types of inputs and steady
state error
Type of 𝑘𝑝 𝑘𝑣 ka Steady error Steady state Steady
2 for step input
system = lim G(s)H(s) = lim
s→0
sG(s)H(s) = lim s G(s)H(s) error for state error
s→0 s→0
𝑒𝑠𝑠 (𝑡) ramp input for
𝑒𝑠𝑠 (𝑡) parabolic
input
𝑒𝑠𝑠 (𝑡)
0 Constant 0 0 Constant Infinity Infinity
1
1 + 𝑘𝑝
1 Infinity Constant 0 Zero Constant Infinity
1
𝑘𝑣
2 infinity Infinity Constant Zero Zero Constant
1
𝑘𝑎

Problem: Calculate the static error constants for the system whose loop transfer function
10(𝑠+2)
is𝐺(𝑠)𝐻(𝑠) = 𝑠(𝑠+3)(𝑠+4).

Solution: The static error constants are 𝑘𝑝 , 𝑘𝑣 , 𝑘𝑎

10(𝑠+2)
Position error constant 𝑘𝑝 = lim G(s)H(s) = lim =∞
s→0 s→0 𝑠(𝑠+3)(𝑠+4)
𝑠10(𝑠+2) 10(𝑠+2) 5
Velocity error constant 𝑘𝑣 = lim sG(s)H(s) = lim = (𝑠+3)(𝑠+4)
=3
s→0 s→0 𝑠(𝑠+3)(𝑠+4)
10(𝑠+2) 𝑠10(𝑠+2)
Acceleration error constant k a = lim s2 G(s)H(s) = lim s2 𝑠(𝑠+3)(𝑠+4) = lim (𝑠+3)(𝑠+4)
=
s→0 s→0 s→0

0
𝑠+10
Problem: Find 𝑘𝑝 , 𝑘𝑣 , 𝑘𝑎 for the system having 𝐺(𝑠) = 𝑠(𝑠3 +7𝑠2 +12𝑠) and𝐻(𝑠) = 1. Also

evaluate steady state error when the input 𝑟(𝑡)is given by

𝑟(𝑡) = 5𝑢(𝑡)
𝑟(𝑡) = 2𝑡𝑢(𝑡)
𝑟(𝑡) = 4𝑡 2 𝑢(𝑡)
Solution: The static error constants are 𝑘𝑝 , 𝑘𝑣 , 𝑘𝑎
𝑠 + 10
G(s)H(s) =
𝑠(𝑠 3 + 7𝑠 2 + 12𝑠)
𝑠+10
Position error constant 𝑘𝑝 = lim G(s)H(s) = lim =∞
s→0 s→0 𝑠(𝑠3 +7𝑠2 +12𝑠)
𝑠(𝑠+10) 𝑠+10
Velocity error constant 𝑘𝑣 = lim sG(s)H(s) = lim = = ∞
s→0 s→0 𝑠(𝑠3 +7𝑠2 +12𝑠) (𝑠3 +7𝑠2 +12𝑠)
(𝑠+10)
Acceleration error constant k a = lim s 2 G(s)H(s) = lim s2 𝑠(𝑠3 +7𝑠2 +12𝑠) =
s→0 s→0
(𝑠+10) 5
lim =6
s→0 (𝑠2 +7𝑠+12 )

Steady error 𝒆𝒔𝒔 (𝒕)

𝑟(𝑡) = 5𝑢(𝑡)
Input is step input with amplitude 5, if the input is step input, steady state error is
𝐴 5
𝒆𝒔𝒔 (𝒕) = = =0
1 + 𝑘𝑝 1 + ∞
𝑟(𝑡) = 2𝑡𝑢(𝑡)
Input is ramp input with slope 2, if the input is ramp input, steady state error is
𝐴 2
𝒆𝒔𝒔 (𝒕) = = =0
𝑘𝑣 ∞
𝑟(𝑡) = 4𝑡 2 𝑢(𝑡)
Input is parabolic input with 4, if the input is parabolic input, steady state error is
𝐴 4 24
𝒆𝒔𝒔 (𝒕) = = 5 =
𝑘𝑎 5
6

Problem: For the system shown in below figure.

Identity the type of the system.
Find the value of 𝑘𝑝 , 𝑘𝑣 , 𝑘𝑎
If 𝑟(𝑡) = 10𝑢(𝑡) find the steady state value of the output.
 𝐶(𝑠) 10 10
Solution: The open loop transfer function 𝐸(𝑠) = 𝐺(𝑠) = 𝑠2 (𝑠2 +𝑠+10)
= 𝑠4 +𝑠3 +10𝑠2 +10𝑠
10
(1+ 2 2 )×𝑠
𝑠 (𝑠 +𝑠+10)

10
𝐺(𝑠) =
𝑠(𝑠 3 + 𝑠2 + 10𝑠 + 10)
10
Now loop transfer function 𝐺(𝑆)𝐻(𝑠) = 𝑠(𝑠3 +𝑠2 +10𝑠+10)

Since the power of s is 1, thus the system is type-1 system.

The static error constants are 𝑘𝑝 , 𝑘𝑣 , 𝑘𝑎
10
G(s)H(s) =
𝑠(𝑠 3 + 𝑠 2 + 10𝑠 + 10)
10
Position error constant 𝑘𝑝 = lim G(s)H(s) = lim =∞
s→0 s→0 𝑠(𝑠3 +𝑠2 +10𝑠+10)
𝑠10 10
Velocity error constant 𝑘𝑣 = lim sG(s)H(s) = lim = =
s→0 s→0 𝑠(𝑠3 +𝑠2 +10𝑠+10) (𝑠3 +𝑠2 +10𝑠+10)

1
10
Acceleration error constant k a = lim s2 G(s)H(s) = lim s 2 𝑠(𝑠3 +𝑠2 +10𝑠+10) =
s→0 s→0
𝑠 (𝑠+10)
lim =0
s→0 (𝑠2 +7𝑠+12 )

Steady stare error for the input 𝑟(𝑡) = 10𝑢(𝑡)

Input is step input with amplitude 10, if the input is step input, steady state error is
𝐴 10
𝒆𝒔𝒔 (𝒕) = = =0
1 + 𝑘𝑝 1 + ∞
Problem: For the signal flow graph shown in below figure, mention the type number and
order of the system. Determine the steady state error for the step and ramp input.
Error 𝑒(𝑡) = 𝑟(𝑡) − 𝑐(𝑡).

𝐶(𝑠) 1
Solution: Using Mason’s gain formula, the overall transfer function 𝑅(𝑠) = ∆ ∑𝑘 𝑃𝑘 ∆𝑘
 𝐶(𝑠) 1
= ∑ 𝑃𝑘 ∆𝑘
𝑅(𝑠) ∆
𝑘

There are two forward paths between the output and input, therefore 𝑘 = 2
4 3 12
The first forward path gain 𝑃1 = × (𝑠+3) = (𝑠+3)(𝑠+4)
(𝑠+4)
−3
The second forward path gain 𝑃2 = (𝑠+3)

Sum of loop gains of all individual loops = ∑ pm1

m
4 3 −2
There is one individual loop gains, that is ∑m pm1 = p11 = (𝑠+3) × (𝑠+3) × (s+5) =
−24
(𝑠+3)(𝑠+4)(𝑠+5)

There are no combinations of non-touching loops Therefore

∑ pm2 = ∑ Pm3 = ∑ Pm4 … = 0

m m m
24
Hence ∆= 1 − p11 = 1 + (𝑠+3)(𝑠+4)(𝑠+5)

First forward path is in touch with all the loops. Therefore ∆1 = 1

Second forward path is in touch with all the loops , therefore ∆2 = 1
12 −3
𝐶(𝑠) 𝑃1 ∆1 + 𝑃2 ∆2 (𝑠+3)(𝑠+4)
+ (𝑠+3) (12 − 3𝑠 − 9)(𝑠 + 5)
= = 24 =
𝑅(𝑠) ∆ 1 + (𝑠+3)(𝑠+4)(𝑠+5) (𝑠 + 3)(𝑠 + 4)(𝑠 + 5) + 24

(12 − 3𝑠 − 9)(𝑠 + 5) (3 − 3𝑠)(𝑠 + 5)

= =
(𝑠 + 3)(𝑠 + 4)(𝑠 + 5) + 24 (𝑠 + 3)(𝑠 + 4)(𝑠 + 5) + 24
(3𝑠 − 3)(𝑠 + 5)
𝐶(𝑠) = − × 𝑅(𝑠)
(𝑠 + 3)(𝑠 + 4)(𝑠 + 5) + 24
The error is given by𝑒(𝑡) = 𝑟(𝑡) − 𝑐(𝑡).
𝐸(𝑠) = 𝑅(𝑠) − 𝐶(𝑠).

(3𝑠 − 3)(𝑠 + 5)
𝐸(𝑠) = 𝑅(𝑠) + × 𝑅(𝑠)
(𝑠 + 3)(𝑠 + 4)(𝑠 + 5) + 24
(3𝑠 − 3)(𝑠 + 5)
𝐸(𝑠) = 𝑅(𝑠)[1 + ]
(𝑠 + 3)(𝑠 + 4)(𝑠 + 5) + 24
 (𝑠 + 3)(𝑠 + 4)(𝑠 + 5) + 24 + (3𝑠 − 3)(𝑠 + 5)
𝐸(𝑠) = 𝑅(𝑠)[ ]
(𝑠 + 3)(𝑠 + 4)(𝑠 + 5) + 24
We know that, the steady state error is 𝑒𝑠𝑠 (𝑡) = lim sE(s)
s→0
1
For the step input 𝑅(𝑠) = 𝑠
1 (𝑠+3)(𝑠+4)(𝑠+5)+24+(3𝑠−3)(𝑠+5)
Therefore steady state error 𝐸(𝑠) = [ (𝑠+3)(𝑠+4)(𝑠+5)+24
𝑠

𝑠 (𝑠 + 3)(𝑠 + 4)(𝑠 + 5) + 24 + (3𝑠 − 3)(𝑠 + 5)

𝑒𝑠𝑠 (𝑡) = lim [
s→0 𝑠 (𝑠 + 3)(𝑠 + 4)(𝑠 + 5) + 24
(𝑠 + 3)(𝑠 + 4)(𝑠 + 5) + 24 + (3𝑠 − 3)(𝑠 + 5) 69
= lim =
s→0 (𝑠 + 3)(𝑠 + 4)(𝑠 + 5) + 24 84
1
For the ramp input 𝑅(𝑠) = 𝑠2
𝑠 (𝑠+3)(𝑠+4)(𝑠+5)+24+(3𝑠−3)(𝑠+5)
Steady state error 𝐸(𝑠) = [ =∞
𝑠2 (𝑠+3)(𝑠+4)(𝑠+5)+24

Problem: For a unity feedback control system the open loop transfer function𝐺(𝑠) =
10(𝑠 + 2)/𝑠 2 (𝑠 + 1). Find (a) the position, velocity and acceleration constants (b) the
1
steady state error when the input is 𝑟(𝑡) = 3 − 2𝑡 + 6 𝑡 2 .

Solution:
Error constants
Position error constant 𝑘𝑝 = lim G(s)H(s) = lim G(s)
s→0 s→0

10(s + 2)
𝑘𝑝 = lim =∞
s→0 s 2 (s + 1)
10(s+2)
Velocity error constant 𝑘𝑣 = lim sG(s)H(s) = lim 𝑠 𝐺(𝑠) = lim 𝑠 × =∞
s→0 s→0 s→0 s2 (s+1)
10(s+2)
Acceleration error constant k a = lim s2 G(s)H(s) = lim s2 𝐺(𝑠) = lim s 2 × s2(s+1) =
s→0 s→0 s→0

20
1
Steady state error for the input 𝑟(𝑡) = 3 − 2𝑡 + 6 𝑡 2 .
𝑅(𝑠)
We know that error 𝐸(𝑠) = 1+𝐺(𝑠)𝐻(𝑠)
3 2 1 10(s+2)
𝑅(𝑠) = 𝑠 − 𝑠2 + 3𝑠3 , 𝐺(𝑠) = s2 (s+1)
1 𝑛!
Because𝐿(𝑢(𝑡)) = 𝑠 , L (𝑡 𝑛 ) = 𝑠𝑛+1
 3
𝐿(3) =
𝑠
2
𝐿(2𝑡) =
𝑠2
1 1 2! 2 1
L (6 𝑡 2 ) = 6 × 𝑠2+1 = 6𝑠3 = 3𝑠3
3 2 1 3 2 1
𝑅(𝑠) 𝑅(𝑠) − 2 + 3𝑠3 − +
𝑠2 3𝑠3
∴ 𝐸(𝑠) = = = 𝑠 𝑠10(s+2) = 𝑠
s2 (s+1)+10(s+2)
1 + 𝐺(𝑠)𝐻(𝑠) 1 + 𝐺(𝑠) 1 +
s2 (s+1) s2 (s+1)

3 s 2 (s + 1) 2 s2 (s + 1)
𝐸(𝑠) = [ 2 ]− 2[ 2 ]
𝑠 s (s + 1) + 10(s + 2) 𝑠 s (s + 1) + 10(s + 2)
1 s 2 (s + 1)
+ 3[ 2 ]
3𝑠 s (s + 1) + 10(s + 2)
The steady state error 𝑒𝑠𝑠 (𝑡) is obtained from final value theorem.
𝑒𝑠𝑠 (𝑡) = lim e(t) = lim sE(s)
t→∞ s→0

∴ 𝑒𝑠𝑠 (𝑡) = lim sE(s)

s→0

3 s 2 (s + 1) 2 s2 (s + 1)
= lim s { [ 2 ]− 2[ 2 ]
s→0 𝑠 s (s + 1) + 10(s + 2) 𝑠 s (s + 1) + 10(s + 2)
1 s 2 (s + 1)
+ 3[ 2 ]}
3𝑠 s (s + 1) + 10(s + 2)
3s 2 (s + 1) 2s(s + 1) (s + 1)
= lim { 2
− 2 + 2
}
s→0 s (s + 1) + 10(s + 2) s (s + 1) + 10(s + 2) 3𝑠 (s + 1) + 10(s + 2)
1 1
𝑒𝑠𝑠 (𝑡) = 0 − 0 + =
60 60
Introduction to PI, PD and PID Controllers
In feedback control system controllers are used to increase the performance of the control
system. Controllers may
1. Proportional plus integral (PI) controller
2. Proportional plus derivative (PD) controller
3. Proportional plus integral plus derivative (PID) controller
Proportional plus integral (PI) controller
The PI controller produces an output signal consists of two terms, one proportional to
error signal and the other proportional to integral of the error signal. The transfer function
 1 𝑇𝑖 𝑠+1
of the PI controller is 𝐺𝑐 (𝑠) = 𝐾𝑝 (1 + 𝑇 𝑠) = 𝐾𝑝 ( ). Here 𝐾𝑝 proportional gain
𝑖 𝑇𝑖 𝑠

and 𝑇𝑖 is the integral time. The block diagram of unity feedback system with PI
controller is as shown in below figure.

The open loop transfer function G(s) be a second order system with transfer function,
𝜔𝑛2
𝐺(𝑠) =
𝑠(𝑠 + 2𝜉𝜔𝑛 )
Now the closed loop transfer function
𝑇 𝑠+1 𝑛 𝜔2
𝐶(𝑠) 𝐾𝑝 ( 𝑖𝑇 𝑠 ) (𝑠(𝑠+2𝜉𝜔 ) 𝐾𝑝 𝜔𝑛2 (𝑇𝑖 𝑠 + 1)
𝑖 𝑛)
= =
𝑅(𝑠) 1 + 𝐾 (𝑇𝑖 𝑠+1) ( 𝜔𝑛2 ) 𝑠 2 𝑇𝑖 (𝑠 + 2𝜉𝜔𝑛 ) + 𝐾𝑝 𝜔𝑛2 (𝑇𝑖 𝑠 + 1)
𝑝 𝑇𝑠 𝑠(𝑠+2𝜉𝜔 )
𝑖 𝑛

𝐾𝑝 𝜔𝑛2 (𝑇𝑖 𝑠 + 1)
= 3
𝑠 𝑇𝑖 + 2𝜉𝜔𝑛 𝑇𝑖 𝑠 2 + 𝐾𝑝 𝜔𝑛2 𝑇𝑖 𝑠 + 𝐾𝑝 𝜔𝑛2
𝐾𝑝 𝜔𝑛2 (𝑇𝑖 𝑠 + 1)
= 2
𝐾𝑝 𝜔𝑛
𝑠 3 + 2𝜉𝜔𝑛 𝑠 2 + 𝐾𝑝 𝜔𝑛2 𝑠 + 𝑇𝑖
2 (𝑇 𝑠+1)
𝐾𝑝 𝜔𝑛 𝐾𝑝
𝑖
= 2 𝑠+𝐾 𝜔 2 , Here 𝐾𝑖 =
𝑠3 ++2𝜉𝜔𝑛 𝑠2 +𝐾𝑝 𝜔𝑛 𝑖 𝑛 𝑇𝑖

From the closed loop transfer function, PI controller

Introduces one zero in the system
1. Increases the order of the system by one
2. Type number of the open loop system is increased by one
3. Reduces the steady state error.
4. Reduces the stability
Proportional plus derivative (PD) controller
The PD controller produces an output signal consists of two terms, one proportional to
error signal and the other proportional to derivative of the error signal. The transfer
function of the PD controller is 𝐺𝑐 (𝑠) = 𝐾𝑝 (1 + 𝑇𝑑 𝑠). Here 𝐾𝑝 proportional gain and
𝑇𝑑 is the derivative time. The block diagram of unity feedback system with PD
controller is as shown in below figure.

The open loop transfer function G(s) be a second order system with transfer function,
𝜔𝑛2
𝐺(𝑠) =
𝑠(𝑠 + 2𝜉𝜔𝑛 )
Now the closed loop transfer function
𝑛 𝜔2
𝐶(𝑠) 𝐾𝑝 (1 + 𝑇𝑑 𝑠)(𝑠(𝑠+2𝜉𝜔 ) 𝐾𝑝 𝜔𝑛2 (𝑇𝑑 𝑠 + 1)
𝑛)
= =
𝑅(𝑠) 1 + 𝐾 (1 + 𝑇 𝑠)( 𝜔𝑛2 ) 𝑠 (𝑠 + 2𝜉𝜔𝑛 ) + 𝐾𝑝 𝜔𝑛2 (𝑇𝑑 𝑠 + 1)
𝑝 𝑑 𝑠(𝑠+2𝜉𝜔 ) 𝑛

𝐾𝑝 𝜔𝑛2 (𝑇𝑑 𝑠
+ 1)
=
𝑠 2 + 2𝜉𝜔𝑛 𝑠 + 𝐾𝑝 𝜔𝑛2 𝑇𝑑 𝑠 + 𝐾𝑝 𝜔𝑛2
𝐾𝑝 𝜔𝑛2 (𝑇𝑑 𝑠 + 1)
=
𝑠 2 + (2𝜉𝜔𝑛 + 𝐾𝑝 𝜔𝑛2 𝑇𝑑 )𝑠 + 𝐾𝑝 𝜔𝑛2
2 (𝑇 𝑠+1)
𝐾𝑝 𝜔𝑛 𝑑
= 𝑠2 +(2𝜉𝜔 2 2 , here 𝐾𝑑 = 𝐾𝑝 𝑇𝑑
𝑛 +𝐾𝑑 𝜔𝑛 )𝑠+𝐾𝑝 𝜔𝑛

From the closed loop transfer function, PD controller

1. Introduces one zero in the system
2. Increases the peak overshoot
3. Reduces the rise time
4. Reduces the steady state error.
Proportional plus integral plus derivative (PID) controller
It is a combination of proportional, integral and derivative controller, will improves all
aspects of system performance. The block diagram of PID controller is as shown in
below figure.
 𝐾𝑖
The closed loop transfer function is Gc (s) = 𝐾𝑝 + + 𝐾𝑑 s. The important features of
s

PID controller are

1. The proportional controller stabilizes the gain but produces the steady state
error.
2. The integral controller reduces stead state error.
3. The derivate controller reduces the rate of change of error.

Module 4
Stability analysis
Concepts of stability
Stability in a system implies that small changes in the system input, initial conditions, or
in the system parameter, do not result in large changes in the system output. System may
be
Stable: The system is said to be stable, if the output is bounded for bounded input. It is
also known as bounded input bounded output system (BIBO)
Unstable: For unstable system, output is unbounded for bounded input
Absolute stable: The system output is stable for all variations of its parameter.
Conditionally stable: The system output is stable for a limited range of variation of its
parameter.
Asymptotically stable: A system is said to be asymptotically stable, if in the absence of
input the output tends towards zero.
Critically stable or marginally stable or limitedly stable: In marginally stable system,
natural response neither decays nor grows but remains constant and oscillates.
Consider the closed loop transfer function
 𝐶(𝑠) 𝑏0 𝑠 𝑚 + 𝑏1 𝑠 𝑚−1 + ⋯ +𝑏𝑚
= 𝐺(𝑠) =
𝑅(𝑠) 𝑎0 𝑠 𝑛 + 𝑎1 𝑠 𝑛−1 + ⋯ + 𝑎𝑛
Output of the system 𝐶(𝑠) = 𝑅(𝑠)𝐺(𝑠)
Apply the inverse Laplace transform, we get
∞
𝑐(𝑡) = 𝑟(𝑡) ∗ 𝑔(𝑡) = ∫ 𝑔(𝜏)𝑟(𝑡 − 𝜏)𝑑𝜏
0

Take absolute value on both sides we get

∞
|𝑐(𝑡)| = |∫ 𝑔(𝜏)𝑟(𝑡 − 𝜏)𝑑𝜏|
0
∞
|𝑐(𝑡)| ≤ ∫ |𝑔(𝜏)||𝑟(𝑡 − 𝜏)|𝑑𝜏
0

Where 𝑟(𝑡)input to the system, let us assume that the input is bounded, and then the
absolute value |𝑟(𝑡)| ≤ 𝑀1 < ∞ for all t. Thus 𝑀1 is finite.
∞
|𝑐(𝑡)| ≤ 𝑀1 ∫ |𝑔(𝜏)|𝑑𝜏 ≤ 𝑀2
0

The output 𝑐(𝑡) is bounded i.e. finite, if 𝑔(𝑡) is absolutely integrable. Thus input is
bounded, output is bounded and the system is stable.
The nature of 𝑔(𝑡)is depends on the poles of 𝐺(𝑠) which are the roots of the
characteristic equation. These roots may be real or complex.
Let us consider
𝐶(𝑠) 1
The closed loop transfer function𝑅(𝑠) = 𝐺(𝑠) = 𝑠+𝑎 , characteristic equation is 𝑠 + 𝑎 =

0, roots of this equation is 𝑠 = −𝑎 is real and, in the s-plane, pole is on the LHS-plane
as shown in the figure. Now 𝑔(𝑡) = 𝑒 −𝑎𝑡 is exponentially decaying and g (t) is absolutely
integrable, and the system is stable.
 𝐶(𝑠) 1
If the closed loop transfer function𝑅(𝑠) = 𝐺(𝑠) = 𝑠−𝑎 , characteristic equation is 𝑠 − 𝑎 =

0, roots of this equation is 𝑠 = 𝑎 is real and, in the s-plane, pole is on the RHS-plane as
shown in the figure. Now 𝑔(𝑡) = 𝑒 𝑎𝑡 is exponentially rising and g (t) is not absolutely
integrable, and the system is unstable.

𝐶(𝑠) 1
If the closed loop transfer function𝑅(𝑠) = 𝐺(𝑠) = 𝑠2 +2𝑠𝑎+𝑎2+𝑏2 , characteristic equation

is 𝑠 2 + 2𝑠𝑎 + 𝑎2 + 𝑏 2 = 0. The roots of this equation are 𝑠 = −𝑎 + 𝑗𝑏 , and 𝑠 =

−𝑎 − 𝑗𝑏 are complex and, in the s-plane pole are on the LHS-plane as shown in the
figure.
1 1 1
𝐺(𝑠) = = =
𝑠2 2
+ 2𝑠𝑎 + 𝑎 + 𝑏 2 (𝑠 + 𝑎 + 𝑗𝑏)(𝑠 + 𝑎 − 𝑗𝑏) (𝑠 + 𝑎 + 𝑗𝑏)(𝑠 + 𝑎 − 𝑗𝑏)
𝐴 𝐴
𝐺(𝑠) = +
(𝑠 + 𝑎 + 𝑗𝑏) (𝑠 + 𝑎 − 𝑗𝑏)
Now𝑔(𝑡) = 𝐴𝑒 −(𝑎+𝑗𝑏)𝑡 + 𝐴𝑒 −(𝑎−𝑗𝑏)𝑡 = 𝐴𝑒 −𝑎𝑡 −𝑗𝑏𝑡
𝑒 + 𝐴𝑒 −𝑎𝑡 𝑒 𝑗𝑏𝑡 = 2𝐴𝑒 −𝑎𝑡
𝑐𝑜𝑠𝑏𝑡
−𝑎𝑡
𝑔(𝑡) = 2𝐴𝑒 𝑐𝑜𝑠𝑏𝑡
The function 𝑔(𝑡) is damped and absolutely integrable, and the system is stable.

𝐶(𝑠) 1
If the closed loop transfer function𝑅(𝑠) = 𝐺(𝑠) = 𝑠2 −2𝑠𝑎+𝑎2+𝑏2 , characteristic equation

is 𝑠 2 − 2𝑠𝑎 + 𝑎2 + 𝑏 2 = 0. The roots of this equation are 𝑠 = 𝑎 + 𝑗𝑏 , and 𝑠 = 𝑎 −

𝑗𝑏 are complex and, in the s-plane pole are on the RHS-plane as shown in the figure.
1 1 1
𝐺(𝑠) = = =
𝑠2 2
− 2𝑠𝑎 + 𝑎 + 𝑏 2 (𝑠 − 𝑎 + 𝑗𝑏)(𝑠 − 𝑎 − 𝑗𝑏) (𝑠 − 𝑎 + 𝑗𝑏)(𝑠 − 𝑎 − 𝑗𝑏)
 𝐴 𝐴
𝐺(𝑠) = +
(𝑠 − 𝑎 + 𝑗𝑏) (𝑠 − 𝑎 − 𝑗𝑏)
Now𝑔(𝑡) = 𝐴𝑒 (𝑎+𝑗𝑏)𝑡 + 𝐴𝑒 (𝑎−𝑗𝑏)𝑡 = 𝐴𝑒 𝑎𝑡 −𝑗𝑏𝑡
𝑒 + 𝐴𝑒 𝑎𝑡 𝑒 𝑗𝑏𝑡 = 2𝐴𝑒 𝑎𝑡
𝑐𝑜𝑠𝑏𝑡
𝑎𝑡
𝑔(𝑡) = 2𝐴𝑒 𝑐𝑜𝑠𝑏𝑡
The function 𝑔(𝑡) is over damped and not an absolutely integrable, and the system is
unstable.

𝐶(𝑠) 1
If the closed loop transfers function𝑅(𝑠) = 𝐺(𝑠) = 𝑠2 +𝑏2 , characteristic equation is 𝑠 2 +

𝑏 2 = 0. The roots of this equation are 𝑠 = 𝑗𝑏 , and 𝑠 = −𝑗𝑏 are imaginary and, in the s-
plane pole are on the imaginary axis of the s-plane as shown in the figure.
1 1 1
𝐺(𝑠) = = =
𝑠2 +𝑏 2 (𝑠 + 𝑗𝑏)(𝑠 − 𝑗𝑏) (𝑠 + 𝑗𝑏)(𝑠 − 𝑗𝑏)
𝐴 𝐴
𝐺(𝑠) = +
(𝑠 + 𝑗𝑏) (𝑠 − 𝑗𝑏)
Now 𝑔(𝑡) = 𝐴𝑒 −𝑗𝑏𝑡 + 𝐴𝑒 𝑗𝑏𝑡 = 𝐴𝑒 −𝑗𝑏𝑡 + 𝐴𝑒 𝑗𝑏𝑡 = 2𝐴𝑐𝑜𝑠𝑏𝑡

The function 𝑔(𝑡) is undamped and absolutely integrable, and the system is oscillatory
are critically stable. Thus
If the roots of the characteristic equation are on left half of the s-plane then the system is
stable
If the roots of the characteristic equation are on right half of the s-plane then the system
is unstable
If the roots of the characteristic equation are on imaginary axis of the s-plane or roots at
the origin then the system is marginally stable or critically stable or limitedly stable.
By locating the roots of the characteristic equation in the s-plane, we can find the stability
of the system. The finding the poles of the characteristic equation is difficult by analytical
method. Hence simple criteria’s are used to judge the location of poles in the s-plane
without solving the equation. These criteria’s are
1. Routh stability criterion.
2. Hurwitz stability criterion.
Routh stability criterion
Routh stability criterion is based on arrangement of the coefficients of the charactestics
equation in to an array called Routh array.
Consider the closed loop transfer function of the system
𝐶(𝑠) 𝑏0 𝑠 𝑚 + 𝑏1 𝑠 𝑚−1 + ⋯ +𝑏𝑚
= 𝐺(𝑠) =
𝑅(𝑠) 𝑎0 𝑠 𝑛 + 𝑎1 𝑠 𝑛−1 + ⋯ + 𝑎𝑛
The characteristic equation of the system in polynomial form is given by
𝑞(𝑠) = 𝑎0 𝑠 𝑛 + 𝑎1 𝑠 𝑛−1 + ⋯ + 𝑎𝑛 , Usually 𝑎0 > 0
The necessary and sufficient condition for stability of the system is all the coefficients of
the characteristic such as 𝑎0 , 𝑎1 , 𝑎2 … be real and have the same sign. None of the
coefficient should be zero. For the Routh array construction, consider a characteristic
equation 𝑞(𝑠) = 𝑎0 𝑠 𝑛 + 𝑎1 𝑠 𝑛−1 + ⋯ + 𝑎𝑛−1 𝑠+𝑎𝑛
The third row elements 𝑏1 , 𝑏2 … …. are evaluated as follows
𝑎1 𝑎2 −𝑎0 𝑎3 𝑎1 𝑎4 −𝑎0 𝑎5
𝑏1 = , 𝑏2 =
𝑎1 𝑎1

This process is continued till we get zero as the last coefficient. In a similar way forth
row elements, 𝑐1 , 𝑐2 … …. are evaluated as follows
𝑏1 𝑎3 −𝑎1 𝑏2 𝑏1 𝑎5 −𝑎1 𝑏3
𝑐1 = , 𝑐2 = and so on
𝑏1 𝑏1

After completing the table, observe the elements in the first column of the table.
The necessary and sufficient condition for stability is that all of the elements in the
first column of the Routh array be positive. If all the elements in the first column of
Routh array have positive sign, then all the roots of the charactestics equation are
on the LHS-plane and the system is stable. If there are any sign changes in the first
column of the Routh array, then system is unstable. The number of sign changes in
the first column of the Routh array, indicate the number of roots in the RHS-plane.
This is the state of Routh stability criterion.
Problem: Using Routh criterion, determine the stability of the system represented by
characteristic equation𝑠 4 + 8𝑠 3 + 18𝑠 2 + 16𝑠 + 5 = 0. Comment on the location of the
roots of the characteristic equation.
Solution: The characteristic equation is 𝑠 4 + 8𝑠 3 + 18𝑠 2 + 16𝑠 + 5 = 0. Construct the
Routh array for the characteristic equation. All the coefficients of s are present and all are
having same sign. Thus the necessary condition is satisfied; we can perform the Routh
test

1×18−2×1 1×5−0×1
𝑠 2 Row calculation = 16, =5
1 1
16×2−5×1
𝑠1 Row calculation = 1.7
16
1.7×5−0×16
𝑠 0 Row calculation =5
1.7

In the Routh array, we observed that all the numbers of first column of Routh array have
same sign. Hence all the roots of the characteristic equation are on the LHS plane and the
system is stable.
Problem: Using Routh criterion, determine the stability of the system represented by
characteristic equation 𝑠 3 + 𝑠 2 + 2𝑠 + 24 = 0. Comment on the location of the roots of
the characteristic equation.
Solution: The characteristic equation is𝑠 3 + 𝑠 2 + 2𝑠 + 24 = 0. Construct the Routh
array for the characteristic equation.

1×2−24×1
𝑠1 Row calculation = −22
1
 −22×24−0×1
𝑠 0 Row calculation = 24
−22

There are two sign change in the first column of the Routh array i.e. plus to minus and
minus to plus. Thus the system is unstable.
Problem: A unity negative feedback control system has open loop transfer function
consisting of two poles, two zeros and a variable gain K. The zeros are located at -2 and -
1; and the poles at −0.1 and+1. Using Routh stability criterion, determine the range
value of K for which the closed loop system has 0, 1 or 2 poles in the RHS-plane.
Solution: The zeros are 𝑧1 = −1 and 𝑧2 = −1
Poles are𝑝1 = −0.1 and 𝑝2 = 1
𝐾(𝑠+𝑧1 )(𝑠+𝑧2 ) 𝐾(𝑠+1)(𝑠+2)
The open loop transfer function 𝐺(𝑠) = (𝑆+𝑝1 )(𝑠+𝑝2 )
= (𝑆+0.1)(𝑠−1)

The characteristic equation of the system is 1 + 𝐺(𝑠)𝐻(𝑠) = 0

1 + 𝐺(𝑠) = 0
𝐾(𝑠 + 1)(𝑠 + 2)
1+ =0
(𝑆 + 0.1)(𝑠 − 1)
(𝑆 + 0.1)(𝑠 − 1) + 𝐾(𝑠 + 1)(𝑠 + 2) = 0
(1 + 𝐾 )𝑠 2 + (3𝐾 − 0.9)𝑠 + (2𝐾 − 0.1) = 0
Construct the Routh array for the characteristic equation.

(3𝐾−0.9)×(2𝐾−0.1)−(1+𝐾)×0
𝑠 0 Row calculation = (2𝐾 − 0.1)
(3𝐾−0.9)

1) We have to determine the value of K for which the closed loop system has 0 poles
in RHS-Plane (i.e. No poles in the RHS-plane) and the system is stable. For stable
system, first column of Routh array have same sign. Therefore
1 + 𝑘 > 0 𝑜𝑟 𝑘 > −1
3𝑘 − 0.9 > 0 𝑜𝑟 𝐾 > 0.3
2𝐾 − 0.1 > 0 𝑜𝑟 𝐾 > 0.05
Hence 𝐾 > 0.3
 2) In this case one pole is on the RHS-plane and the system is unstable, for unstable
system we have to find the value of K. For one pole on the RHS-plans, there
should be one sign change in the first column of Routh array. Therefore
−1 < 𝐾 < 0.05
3) In this case two poles are on the RHS-plane and the system is unstable, for
unstable system we have to find the value of K. For two poles on the RHS-plane
there should be two sign changes in the first column of Routh array. Therefore
0.05 < 𝐾 < 0.3
Problem: Determine the range K for stability of unity feedback system whose open loop
𝐾
transfer function is 𝐺(𝑠) = 𝑠(𝑠+1)(𝑠+2)
𝐾
𝐶(𝑠) 𝐺(𝑠) 𝑠(𝑠+1)(𝑠+2) 𝐾
Solution: Closed loop transfer function 𝑅(𝑠) = 1+𝐺(𝑠) = 𝐾 = 𝑠(𝑠+1)(𝑠+2)+𝐾
1+
𝑠(𝑠+1)(𝑠+2)

The characteristic equation is 𝑠(𝑠 + 1)(𝑠 + 2) + 𝐾 = 0

𝑠 3 + 3𝑠 2 + 2𝑠 + 𝐾 = 0
Construct the Routh array for the characteristic equation.

For stable system, all the elements in the first column should have the same sign
From 𝑠 0 row, for the system to be stable, 𝐾 > 0
6−𝐾
From 𝑠1 row, for the system to be stable, > 0 thus the value of K should be less than
3

6. For stability, range of k is

∴ 0<𝐾<6
Problem: The open loop transfer function of unity feedback system is 𝐺(𝑠) =
𝐾
. Determine the range of K for which the system is stable. Also
(𝑠+2)(𝑠+4)(𝑠2 +6𝑠+25)

determine the value of K for which the system is marginally stable.

Solution: Closed loop transfer function
𝐾
𝐶(𝑠) 𝐺(𝑠) . 𝐾
(𝑠+2)(𝑠+4)(𝑠2 +6𝑠+25)
= = 𝐾 = 2
𝑅(𝑠) 1 + 𝐺(𝑠) 1 + . (𝑠 + 2)(𝑠 + 4)(𝑠 + 6𝑠 + 25) + 𝐾
(𝑠+2)(𝑠+4)(𝑠2 +6𝑠+25)

The characteristic equation is (𝑠 + 2)(𝑠 + 4)(𝑠 2 + 6𝑠 + 25) + 𝐾 = 0

𝑠 4 + 12𝑠 3 + 69𝑠 2 + 198𝑠 + 200 + 𝐾 = 0
𝑠 34 + 12𝑠 3 + 69𝑠 2 + 198𝑠 + (200 + 𝐾) = 0

666.25−𝐾
From 𝑠1 row, for the system to be stable, >0 thus the value of K should be less
52.5

than 666.25.
From 𝑠 0 row, for the system to be stable, 200 + 𝐾 > 0, thus 𝐾 > −200. For stability,
range of K is
∴ −200 < 𝐾 < 666.25
For marginal stability, if 𝐾 = 666.25, all the clement in the 𝑠1 row is zero. And the
corresponding auxiliary equation is 𝐴 = 52.5𝑠 2 + 200 + 𝐾 = 0
52.5𝑠 2 + 200 + 666.25 = 0
𝑠 = ±𝑗4.06
When 𝐾 = 666.25, the system is oscillatory, and the roots are on the imaginary axis,
therefore, the frequency of oscillation is 𝜔 = 4.06𝑅𝑎𝑑/𝑠𝑒𝑐.
Problem: The characteristic equations of a system in differential equation form is
𝑥̈ − (𝐾 + 2)𝑥̇ + (2𝐾 + 5)𝑥 = 0
 a. Find the values of K for which the system is
i. Stable
ii. Limitedly stable
iii. unstable
b. For the stable case for what values of K is the system
i. Underdamped
ii. Overdamped
Solution: (a) 𝑥̈ − (𝐾 + 2)𝑥̇ + (2𝐾 + 5)𝑥 = 0
Taking Laplace transform on both sides, we get
𝑠 2 − (𝐾 + 2)𝑠 + (2𝐾 + 5) = 0
Construct the Routh array using the characteristic equation.

(i) For the stable system −(𝐾 + 2) > 0 and (2𝐾 + 5) > 0
𝐾 < −2 And 𝐾 > −2.5 or −2 > 𝐾 > −2.5 or K is between −2.5 and −2
(ii) For the system to be limitedly stable
𝐾 = −2 And 𝐾 = −2.5
(iii) For the unstable system
𝐾 > −2 And 𝐾 < −2.5
(b)The roots the characteristic equation is
(𝐾 + 2) ± √(𝐾 + 2)2 − 4(2𝐾 + 5)
𝑠1 𝑠2 =
2
(i) Value of K for underdamped case
(𝐾+2)
For critically damped case, roots are equal, 𝑠1 𝑠2 = and (𝐾 + 2)2 − 4(2𝐾 + 5) =
2

0, thus when
𝐾 = 6.47, −2.47
As per stability condition of part (a) if 𝐾 = 6.47 system is unstable, we know that, for
underdamped case, the roots are complex. For underdamped case
𝐾 < −2 And 𝐾 > −2.47
(ii) Value of K for over damped case
We know that, for over damped case, the roots are real and unequal. For over damped
case
−2.47 > 𝑘 > −2.5, thus
For stable −𝟐 > 𝐾 > −𝟐. 𝟓
Limitedly stable 𝑲 = −𝟐 and 𝑲 = −𝟐. 𝟓
Unstable 𝑲 > −𝟐 And 𝑲 < −2.5
For under damping −𝟐 > 𝐾 > −𝟐. 𝟒𝟕
For over damping −𝟐. 𝟒𝟕 > 𝑘 > −𝟐. 𝟓
Special cases
Occasionally, in applying the RH criterion, certain difficulties arise causing the
breakdown of computation. Following are the difficulties
Difficulty1: When the first element of any row of the Routh array is zero. Example
Problem: Using Routh criterion, determine the stability of the system represented by
characteristic equation𝑠 5 + 𝑠 4 + 2𝑠 3 + 2𝑠 2 + 3𝑠 + 5 = 0. Comment on the location of
the roots of the characteristic equation.
Solution: Construct the Routh array using the characteristic equation.

Here, first element of 𝑠 3 row is zero, thus the Routh test breaks down. To avoid this
1
difficult, replace 𝑠 by 𝑧 in the characteristic equation then
1 1 1 1 1
( )5 + ( )4 + 2( )3 + 2( )2 + 3 + 5 = 0
𝑧 𝑧 𝑧 𝑧 𝑧
5𝑠 5 + 3𝑧 4 + 2𝑧 3 + 2𝑧 2 + 𝑧 + 1 = 0
Construct the Routh array for the new equation.
There are two sign changes in the first column of the Routh array, which tells us that
there are two roots in the RHS-plane, thus the system is unstable. The remaining roots are
on the LHS-plane.
Alternatively

Here, first element of 𝑠 3 row is zero, thus the Routh test breaks down. To avoid this
difficult, replace zero with small positive number and proceed to evaluate the rest of the
Routh array. Then examine the sign of the first column of Routh array by letting ∈→ 0.
That is

The first element of the fourth row is

2∈+2 2∈+2
, ∈→ 0 = ∞, it has a positive sign
∈ ∈

The first element of the fifth row is

−4 ∈ −4 − 5 ∈2 −4 ∈ −4 − 5 ∈2
, ∈→ 0 = −2
2 ∈ +2 2 ∈ +2
There are two sign changes in the first column of the Routh array, which tells us that
there are two roots in the RHS-plane, thus the system is unstable. The remaining roots are
on the LHS-plane.
Problem: Using Routh criterion, determine the stability of the system represented by
characteristic equation 𝑠 5 + 2𝑠 4 + 3𝑠 3 + 6𝑠 2 + 10𝑠 + 15 = 0. Comment on the location
of the roots of the characteristic equation.
Solution: Construct the Routh array using the characteristic equation.

Here, first element of 𝑠 3 row is zero, thus the Routh test breaks down. To avoid this
1
difficult, replace 𝑠 by 𝑧 in the characteristic equation then
1 1 1 1 1
( )5 + 2( )4 + 3( )3 + 6( )2 + 10( ) + 15 = 0
𝑧 𝑧 𝑧 𝑧 𝑧
15𝑧 5 + 10𝑧 4 + 6𝑧 3 + 3𝑧 2 + 2𝑧 + 1 = 0
Construct the Routh array for the new equation.
 10×6−3×15 10×2−15×1
𝑠 3 Row calculation = 1.5, = 0.5
10 10
1.5×3−0.5×10 1.5×1−0×10
𝑠 2 Row calculation = −0.333, =1
1.5 1.5
−0.333×0.5−1.5×1
𝑠1 Row calculation = −5
−0.333
−5×1−0×(−0.333)
𝑠 0 Row calculation =1
−5

There are two sign changes in the first column of Routh array, two roots in the RHS-
plane, system is unstable.
Alternatively

The first element of the fourth row is

6∈−5 6∈−5
, ∈→ 0 = −∞, it has a positive sign
∈ ∈

The first element of the fifth row is

(−6 ∈ −5)2.5 − 15 ∈ (−6 ∈ −5)2.5 − 15 ∈
, ∈→ 0 =0
(6 ∈ −5)/∈ (6 ∈ −5)/∈
There are two sign changes in the first column of the Routh array, which tells us that
there are two roots in the RHS-plane, thus the system is unstable. The remaining roots are
on the LHS-plane.
Difficulty2: When the all element of any one row of the Routh array is zero. This
indicates roots are on the imaginary axis and the system is marginally stable.
Problem: Construct the Routh array and determine the stability of the system represented
by characteristic equation 𝑠 6 + 2𝑠 5 + 8𝑠 4 + 12𝑠 3 + 20𝑠 2 + 16𝑠 + 16 = 0. Also
determine the number of roots lying on the RHS-plane, LHS-plane and on the imaginary
axis.
Solution: The characteristic equation is 𝑠 6 + 2𝑠 5 + 12𝑠 4 + 12𝑠 3 + 20𝑠 2 + 16𝑠 + 16 =
0. Construct the Routh array using the characteristic equation.

1×8−6×1 1×20−8×1 1×16−1×0

𝑠 4 Row calculation = 2, = 12, = 16
1 1 1

Divide it by 2, now the 𝑠 4 row elements are 1, 6, and 8

1×6−6×1 1×8−1×8
𝑠 3 Row calculation = 0, =0
1 1

Now all the 𝑠 3 row elements are zero, at this stage using 𝑠 4 row element construct the
auxiliary equation 𝐴 = 𝑠 4 + 6𝑠 2 + 8 . On differentiating the A with respect to s we get
𝑑𝐴
= 4𝑠 3 + 12𝑠
𝑑𝑠
𝑑𝐴
The coefficients of are used to form 𝑠 3 row element, that is 4 and 12. If these numbers
𝑑𝑠

are divided by 2, 𝑠 3 row elements are 1, and 3

1×6−3×1 1×8−0×1
𝑠 2 Row calculation = 3, =8
1 1
3×3−8×1
𝑠1 Row calculation = 0.33,
3
 0.333×8−3×0
𝑠 0 Row calculation =8
0.33

Observation: In the Routh array, we observed that all the numbers of first column of
Routh array have same sign. But 𝑠 3 with all zeros indicates the possibility of roots on the
imaginary axis, hence the system is marginally stable. To find the roots on the imaginary
axis, let us consider the auxiliary equation 𝐴 = 𝑠 4 + 6𝑠 2 + 8 and find the roots of this
equation.
𝑠 4 + 6𝑠 2 + 8 = 0 Put 𝑠 2 = 𝑥, then
𝑥 2 + 6𝑥 + 8 = 0
𝑥 = −3 ± 1 = −2 𝑎𝑛𝑑 − 4
𝑠 = ±√−2 𝑎𝑛𝑑 ± √−4
𝑠 = ±𝑗√2 And ±𝑗2
These are the roots of the auxiliary equation, are also the roots of the charactestics
equation. Thus four roots are on the imaginary axis and reaming two roots are on the
LHS-plane.
Problem: In unity feedback system find the range of K for stability and 𝐾𝑚𝑎𝑟 and 𝜔𝑚𝑎𝑟
𝐾
with 𝐺(𝑠) =
𝑠(1+0.4𝑠)(1+0.25𝑠)

Solution: The characteristic equation of the system is 1 + 𝐺(𝑠)𝐻(𝑠) = 0

𝐾
1+ =0
𝑠(1 + 0.4𝑠)(1 + 0.25𝑠)
𝑠(1 + 0.4𝑠)(1 + 0.25𝑠) + 𝐾 = 0
0.1𝑠 3 + 0.65𝑠 2 + 𝑠 + 𝐾 = 0
𝑠 3 + 6.5𝑠 2 + 10𝑠 + 10𝐾 = 0
Construct the Routh array using the characteristic equation.
 6.5×10−10𝐾×1 65−10𝐾
𝑠1 Row calculation =
6.5 6.5
65−10𝐾
×10𝐾−6.5×0
𝑠 0 Row calculation 6.5
65−10𝐾 = 10𝐾
6.5

For stability first column elements of Routh array should have the same. Therefore
65−10𝐾
10𝐾 > 0 And >0
6.5

65 − 10𝐾 > 0
10𝐾 < 65
𝐾 < 6.5
Range of K is 0 < 𝐾 < 6.5
65−10𝐾
For marginal stable = 0 , 𝐾𝑚𝑎𝑟 = 6.5
6.5

For 𝐾𝑚𝑎𝑟 , we have 𝑠1 row of zeros, and the corresponding auxiliary equation is
𝐴 = 6.5𝑠 2 + 10𝑘
For the roots of the auxiliary equation, consider 6.5𝑠 2 + 10𝑘 = 0
±√−4 × 6.5 × 10𝑘 ±𝑗√4 × 6.5 × 10𝑘
𝑠2 = =
2 × 6.5 2 × 6.5
If K=6.5
±𝑗√4 × 6.5 × 10 × 6.5
𝑠2 = = ±𝑗√10
2 × 6.5
𝑠 = ±𝑗3.16
We know that 𝑠 = 𝜎 + 𝑗𝜔
Therefore 𝜔𝑚𝑎𝑟 = 3.16 𝑅𝑎𝑑/𝑠𝑒𝑐
Problem: Find the range of K so that the system is with characteristic equation 𝑠 4 +
22𝑠 3 + 10𝑠 2 + 𝑠 + 𝐾 = 0
Solution: The characteristic equation of the system is 𝑠 4 + 22𝑠 3 + 10𝑠 2 + 𝑠 + 𝐾 = 0
Construct the Routh array using the characteristic equation.
 22×10−1 219 22×𝐾−1×0
𝑠 2 Row calculation = , =𝐾
22 22 22
219
×1−22×𝑘 219−484𝐾
1 22
𝑠 Row calculation 219 = 219
22
219−484𝐾 219
×𝐾− ×0
0 219 22
𝑠 Row calculation = 219−484𝐾 =𝐾
219

For stability first column elements of Routh array should have the same. Therefore
219−484𝐾
𝐾 > 0 and >0
219

219 − 484𝐾 > 0

219
𝐾<
484
𝐾 < 0.45
Therefore range of K is 0 < 𝐾 < 0.45
219−484𝐾
For marginal stable = 0 = 0 , 𝐾𝑚𝑎𝑟 = 0.45
219

For 𝐾𝑚𝑎𝑟 , we have 𝑠1 row of zeros, and the corresponding auxiliary equation is
219 2
𝐴= 𝑠 +𝑘
22
For the roots of the auxiliary equation, consider 219𝑠 2 + 22 × 0.45 = 0
𝑠 = ±𝑗0.213
We know that 𝑠 = 𝜎 + 𝑗𝜔
Therefore 𝜔𝑚𝑎𝑟 = 0.213 𝑅𝑎𝑑/𝑠𝑒𝑐
Problem: Using RH criterion determine the stability of the system, the system is type one
system with error constant of10 𝑠𝑒𝑐 −1, and poles at 𝑠 = −3 and s=-6.
 𝐾
Solution: The open loop transfer function is 𝐺(𝑠) = 𝑠(𝑠+3)(𝑠+6).Since it is a type one

system velocity constant 𝑘𝑣 exists and 𝑘𝑝 = ∞, 𝑘𝑎 = 0.

𝑠𝐾 𝐾
k v = 10 = lim sG(s)H(s) = lim sG(s) = lim =
s→0 s→0 s→0 𝑠(𝑠 + 3)(𝑠 + 6) 18
𝐾
= 10, 𝐾 = 180
18
𝐾 180
Now 𝐺(𝑠) = 𝑠(𝑠+3)(𝑠+6) = 𝑠(𝑠+3)(𝑠+6)

To find the stability of the system, consider the characteristic equation

1 + 𝐺(𝑠)𝐻(𝑠) = 0
180
1+ =0
𝑠(𝑠 + 3)(𝑠 + 6)
𝑠(𝑠 + 3)(𝑠 + 6) + 180 = 0
𝑠 3 + 6𝑠 2 + 18𝑠 + 180 = 0
Construct the Routh array

There are two sign changes, two roots are on the RHS-plane and the system is unstable.
Problem: Using RH criterion, determine the stability of the system represented by the
characteristic equation9𝑠 5 − 20𝑠 4 + 10𝑠 3 − 𝑠 2 − 9𝑠 − 10 = 0. Comment on the
location of the roots of the characteristic equation.
Solution: The characteristic equation of the system is equation 9𝑠 5 − 20𝑠 4 + 10𝑠 3 −
𝑠 2 − 9𝑠 − 10 = 0. On examining the coefficients of the polynomial, it is found that some
coefficients are negative, hence the system is unstable. The Routh array is constructed to
find the number of roots lying on the RHS-plane.
By examining the elements of the first column of the Routh array it is observed that there
are three sign changes and three poles on the RHS-plane, system is unstable.
Introduction to Root-Locus Techniques
We know that, using Routh array test we can find the stability of the system analytically.
But root locus is the graphical method to find the stability of the system. Root locus is
plotted on the s-plane
The root locus concepts
Each branch of root locus stats at poles and terminates at zeros or at infinity. To
understand concept of root locus, let us consider the second order unity feedback control
system as shown in below figure.

𝐾
The open loop transfer function of the system is𝐺(𝑠) = 𝑠(𝑠+𝑎), where K and a are

constants. The open loop Transfer function has two poles, one at origin 𝑠 = 0 and the
𝐶(𝑠) 𝐾
other at 𝑠 = −𝑎. The closed loop transfer function of this system is = 𝑠2 +𝑎𝑠+𝐾. The
𝑅(𝑠)

characteristic equation of the system is 𝑠 2 + 𝑎𝑠 + 𝐾 = 0. The roots of this equation

𝑎 𝑎
are 𝑠1 , 𝑠2 = − 2 ± √(2)2 − 𝐾. As parameter K and varies the roots of the characteristic
equation change. As K varied from zero to infinity, the two roots describes the locus in
the s-plane.
1. 0 ≤ 𝑘 ≤ 𝑎2 /4 the roots are real and distinct
When 𝐾 = 0, roots 𝑠1 = 0 and 𝑠2 = −𝑎, they are same as open loop poles.
2. 𝑘 = 𝑎2 /4, the roots are real and equal 𝑠1 = 𝑠2 = −𝑎/2
𝑎2
3. ≤ 𝑘 ≤ ∞ , the roots are complex conjugate.
4

The root locus for with varying K is plotted in the below figure.

The root locus drawn in the above figure has been drawn by direct solution of
characterstic equation. This procedure highly tedious for higher order syste. Evans
developed a graphical technique for root locus for second and higher order system.
Angle and magnitude critrion
Consider the feedback system and corresponding signal flow graph as shown in below
𝐶(𝑠)
figure. Using Mason’s gain formula, we can find the over all transfer function 𝑅(𝑠).

𝐶(𝑠) 1
= ∑ 𝑃𝑘 ∆𝑘
𝑅(𝑠) ∆
𝑘

Number of forward path in the signal flow graph is 𝑘 = 1, therefore

𝐶(𝑠) 1 𝑃1 ∆1
= ∑ 𝑃𝑘 ∆𝑘 =
𝑅(𝑠) ∆ ∆
𝑘

𝑃1 = 𝐺(𝑠) and ∆1 = 1

∆= 1 − ∑ 𝑃𝑚1 + ∑ 𝑃𝑚2 − ∑ 𝑃𝑚3 + ⋯

𝑚 𝑚 𝑚

∆= 1 − 𝑃11 = (−𝐺(𝑠)𝐻(𝑠) = 1 + 𝐺(𝑠)𝐻(𝑠)

∆= 1 + 𝐺(𝑠)𝐻(𝑠)
∆= 1 + 𝑃(𝑠) … … . (𝐴)
We know that 1 + 𝐺(𝑠)𝐻(𝑠) = 0 … … . (𝐵),
On comparing equation A and B, 𝑃(𝑠) = 𝐺(𝑠)𝐻(𝑠)
1 + 𝑃(𝑠) = 0
𝑃(𝑠) = −1
This 𝑃(𝑠) is a complex quantity, it has both magnitude and phase, thus
Magnitude |𝑃(𝑠)| = 1 and phase ∠𝑃(𝑠) = ±1800 (2𝑞 + 1); 𝑞 = 0,1,2,3,4 ….
Let us replace 𝑃(𝑠) = 𝐺(𝑠)𝐻(𝑠), then
|𝐺(𝑠)𝐻(𝑠)| = 1 is called magnitude criterion and
∠𝑃(𝑠) == ±1800 (2𝑞 + 1); 𝑞 = 0,1,2,3,4 …. is called angle criterion
Construction of root loci (Construction rules)
Rule 1: Root locus is symmetrical about the real axis (𝜎 𝑎𝑥𝑖𝑠)
𝐶(𝑠) 𝐺(𝑠)
Rule 2: Consider the closed loop transfer function 𝑅(𝑠) = 1+𝐺(𝑠)𝐻(𝑠) , and the loop transfer

function is 𝐺(𝑠)𝐻(𝑠) and open loop transfer function is 𝐺(𝑠). Using the open loop
transfer function in factor form we can find the poles and zeros and plot it on the s-plane.
Let the number of poles be N and number of zeros be M. Number of root locus branches
is equal to the number of poles or zeros whichever is greater. Root locus starts at poles
and terminates at zeros or at infinity.
Rule 3: A point on the real axis lies on the locus if the number of open loop pole plus
zeros on the real axis to the right of this point is odd.
Rule 4: The (𝑁 − 𝑀) branches of root locus terminates at infinity, do so along straight
line asymptotes whose angels are given by
1800 (2𝑞 + 1)
∅𝑞 = ± , 𝑞 = 0,1,2,3 … (𝑁 − 𝑀)
𝑁−𝑀
Rule 5: The asymptotes cross the real axis at a point known as centroid, determined by
the relationship:
(sum of real part of poles – sum of real part of zeros)
Centroid 𝜎𝐴 =
𝑁−𝑀
Rule 6: Breakaway point is the point at which two branches of root locus meet. To
determine breakaway point, consider the characteristic equation 1 + 𝐺(𝑠)𝐻(𝑠) = 0 and
𝑑𝑘
determined from the roots of the equation =0
𝑑𝑠

Rule 7: Angle of departure from the complex poles is given by

𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒
∅ 𝑑 = 1800 − (𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑝𝑜𝑙𝑒 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠)

𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒

+( )
𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑧𝑒𝑟𝑜𝑠 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠

∅ 𝑑 = 180 − ϕp + ϕz
Rule 8: Intersection of root locus branches with the imaginary axis can be determined
using Routh array or any other methods.
Problem: The open loop transfer function of a unity feedback system is given by𝐺(𝑠) =
𝐾(𝑠+9)
. Sketch the root locus of the system.
𝑠(𝑠2 +4𝑠+11)
𝐾(𝑠+9)
Solution: Let us consider the open loop transfer function 𝐺(𝑠) = 𝑠(𝑠2 +4𝑠+11)

Rule 1: Root locus is symmetrical about the real axis

Rule 2: The open loop poles are
𝑠(𝑠 2 + 4𝑠 + 11) = 0
The poles are, 𝑠 = 0, 𝑠 = −2 + j2.64, 𝑠 = −2 − 𝑗2.64 thus number of poles 𝑁 = 3
Zeros are𝑠 = −9 , thus number of zeros 𝑀 = 1
The poles are marked as × and zeros are marked as 0 in the s-plane
Root locus branches starts at poles and terminate at zeros or at infinity. Number of root
locus branches is equal to the number of poles or zeros whichever is greater. Therefore
number of root locus branches is equal to 3.
Rule 3: A point on the real axis lies on the locus, if the number of open loop pole plus
zeros on the real axis to the right of this point is odd.

Rule 4: Among three root locus, one root locus branch on real axis start at pole 𝑠 = 0 and
travel along negative real axis to meet zero at𝑧 = −9. The other two root locus branches
start at complex pole 𝑠 = −2 + 2.64𝑗, 𝑠 = −2 − j2.64 and terminate at infinity. Thus we
need two asymptotes. Asymptotes are straight line, gives the guideline to the root locus
branches. Therefore we have to compute the angle of asymptote.
1800 (2𝑞+1)
Angle of asymptote ∅𝑞 = ± , 𝑞 = 0,1,2,3 … (𝑁 − 𝑀) = 0,1
𝑁−𝑀
180
∅0 = ± = ±900
3−1
1800 (4 + 1)
∅1 = ± = ±4500 = ±(4500 − 3600 ) = ±900
3−1

Rule 5: Centroid is the meeting point of the asymptotes on the real axis. At this point the
angles of asymptote are measured and asymptotes are drawn with doted line.
Sum of poles − Sum of zeros 0 − 2 + 2.64j − 2 − j2.64 − (−9)
Centroid 𝜎𝐴 = = = 2.5
N−M 2
Rule 6: From the knowledge of pole-zero location, it can be concluded that there is no
possibility of breakaway point.
Rule 7: The angle departure is angle at which the root locus branch departure at the
complex pole. Angele of departure at the complex pole
 𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒
∅ 𝑑 = 1800 − (𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑝𝑜𝑙𝑒 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠)

𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒

+( )
𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑧𝑒𝑟𝑜𝑠 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠

∅ 𝑑 = 180 − ϕp + ϕz
Let us consider the complex pole at −2 + 𝑗2.64 and draw the vector from the other poles
and zeros. Let the angles of these vectors be 𝜃1 , 𝜃2 , 𝜃3 as shown in below figure.

2.64
Here 𝜃1 = 180 − 𝑡𝑎𝑛−1 ( ) = −127.10
2

𝜃2 = 900
2.64
𝜃3 = 𝑡𝑎𝑛−1 ( ) = 20.70
7
𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒
ϕ = 𝜃1 + 𝜃2 = 127.10 + 900 = 217.10
𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑝𝑜𝑙𝑒 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠 p
𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒
ϕ
𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑧𝑒𝑟𝑜𝑠 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠 z
= 𝜃3 = 20.70

𝐴𝑛𝑔𝑙𝑒 𝑜𝑓 𝑑𝑒𝑝𝑎𝑟𝑡𝑢𝑟𝑒
} ∅ = 1800 − ϕp + ϕz = 1800 − 217.10 + 20.760
𝑎𝑡 𝑡ℎ𝑒 𝑐𝑜𝑚𝑝𝑙𝑥 𝑝𝑜𝑙𝑒 − 2 + 𝑗2.64 𝑑
= −16.40
𝐴𝑛𝑔𝑙𝑒 𝑜𝑓 𝑑𝑒𝑝𝑎𝑟𝑡𝑢𝑟𝑒
} ∅ 𝑑 = 16.40
𝑎𝑡 𝑡ℎ𝑒 𝑐𝑜𝑚𝑝𝑙𝑥 𝑝𝑜𝑙𝑒 − 2 − 𝑗2.64
Mark the angles of departure at complex poles.
Rule 8: Intersection point of imaginary axis and the root locus.
For this we have to develop the characteristic equation of the closed loop transfer
function i.e. 1 + 𝐺(𝑠)𝐻(𝑠) = 0
𝐾(𝑠 + 9)
1+ =0
𝑠(𝑠 2 + 4𝑠 + 11)
𝑠(𝑠 2 + 4𝑠 + 11) + 𝐾(𝑠 + 9) = 0
𝑠 3 + 4𝑠 2 + 11𝑠 + 𝐾𝑠 + 9𝐾 = 0
Put 𝑠 = 𝑗𝜔 in the characteristic equation
(𝑗𝜔)3 + 4(𝑗𝜔)2 + 11(𝑗𝜔) + 𝐾𝑗𝜔 + 9𝐾 = 0
−𝑗𝜔 3 − 4𝜔2 + 11(𝑗𝜔) + 𝐾𝑗𝜔 + 9𝐾 = 0
Equating the imaginary part to zero
−𝑗𝜔 3 + 11(𝑗𝜔) + 𝐾𝑗𝜔 = 0
−𝜔 2 + 11 + 𝐾 = 0
𝜔 2 = 11 + 𝐾
Equating the real part to zero
−4𝜔2 + 9𝐾 = 0
Put 𝜔 2 = 11 + 𝑘, then
−4 (11 + 𝐾) + 9𝐾 = 0
𝐾 = 8.8
And 𝜔 2 = 11 + 𝐾 = 11 + 8.8, 𝜔 = ±4.4
The crossing point of root locus is±𝑗4.4.
Plot the root locus as shown in below figure. Root locus is symmetrical about the real
axis. From the plot, if 𝐾 = 8.8 root locus is at imaginary axis. If 𝐾 > 8.8 root locus
cross the imaginary axis and the system is unstable, if 𝐾 < 8.8 system is stable.
Problem: Sketch the root locus of the system whose characteristic equation is 𝑠 3 +
6𝑠 2 + 8𝑠 + 𝐾 = 0.
Solution: We know that the characteristic equation of the system is
1 + 𝐺(𝑠)𝐻(𝑠) = 0
1
1+ =0
𝐺(𝑠)𝐻(𝑠)
From this characteristic equation, we need to find the open loop transfer function 𝐺(𝑠)
𝑠 3 + 6𝑠 2 + 8𝑠 + 𝐾 = 0
𝐾
1+ =0
𝑠 3 + 6𝑠 2 + 8𝑠
𝐾 𝐾
Now 𝐺(𝑠)𝐻(𝑠) = 𝑠3 +6𝑠2 +8𝑠 = 𝑠(𝑠+2)(𝑠+4)

Rule 1: Root locus is symmetrical about the real axis (𝜎 𝑎𝑥𝑖𝑠)

Rule 2: Number of poles 𝑁 = 3, poles is 𝑠 = 0, −2, −4
Number of zeros 𝑀 = 0,
Numbers of root locus branches are 3
The poles are marked as × and zeros are marked as 0 in the s-plane
Root locus branches starts at poles and terminate at zeros or at infinity.
Rule 3: A point on the real axis lies on the locus, if the number of open loop pole plus
zeros on the real axis to the right of this point is odd.

Rule 4: The (𝑁 − 𝑀) = 3 branches of root locus terminates at infinity, do so along

straight line asymptotes whose angels are given by, that is angle of asymptotes
1800 (2𝑞 + 1)
∅𝑞 = ± , 𝑞 = 0,1,2,3 … (𝑁 − 𝑀)
𝑁−𝑀
1800
∅0 = ± = ±600
3
1800 × 3
∅1 = ± = ±1800
3
1800 × 5
∅2 = ± = ±3000
3
Rule 5: The asymptotes cross the real axis at a point known as centroid, determined by
the relationship:
(sum of real part of poles – sum of real part of zeros)
Centroid 𝜎𝐴 =
𝑁−𝑀
0−2−4
Centroid 𝜎𝐴 = = −2
3
Rule 6: Breakaway point is the point at which two branches of root locus meet. To
determine breakaway point, consider the characteristic equation 1 + 𝐺(𝑠)𝐻(𝑠) = 0 and
𝑑𝑘
determined from the roots of the equation 𝑑𝑠 = 0. The characteristic equation is 𝑠 3 +

6𝑠 2 + 8𝑠 + 𝐾 = 0
𝑘 = −𝑠 3 − 6𝑠 2 − 8𝑠 − 𝐾

𝑑𝑘
= −3𝑠 2 − 12𝑠 2 − 8 = 0
𝑑𝑠
 3𝑠 2 + 12𝑠 2 + 8 = 0
𝑠1 = −0.845 And 𝑠2 = −3.154
The point −3.154 is not on the root locus, thus it is not a valid break away point. The
valid break away point is−0.845.
Rule 7: Angle of departure from the complex poles is given by
𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒
∅ 𝑑 = 1800 − (𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑝𝑜𝑙𝑒 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠)

𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒

+( )
𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑧𝑒𝑟𝑜𝑠 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠

∅ 𝑑 = 180 − ϕp + ϕz
Rule 8: Intersection of root locus branches with the imaginary axis can be determined
using Routh array or any other methods. Let us consider the characteristic equation 𝑠 3 +
6𝑠 2 + 8𝑠 + 𝐾 = 0. Put𝑠 = 𝑗𝜔, we get
(𝑗𝜔)3 + 6(𝑗𝜔)2 + 8𝑗𝜔 + 𝐾 = 0
−𝑗𝜔3 − 6𝜔2 + 8𝑗𝜔 + 𝐾 = 0
Equating the real and imaginary to zero, then
−𝑗𝜔3 + 8𝑗𝜔 = 0 And −6𝜔2 + 𝐾 = 0
−𝑗𝜔3 + 8𝑗𝜔 = 0
− 𝜔2 + 8 = 0
𝜔2 = 8
𝜔 = ±√8 Rad/sec
−6𝜔2 + 𝐾 = 0
Put 𝜔2 = 8, the 𝐾 = 48 = 𝐾𝑚𝑎𝑟 , Root locus sketch is
The crossing point of root locus is±𝑗√8. The value of K at this point is 𝐾 = 48. Plot the
root locus as shown in below figure. Root locus is symmetrical about the real axis. From
the plot, if 𝐾 = 48 root locus is at imaginary axis, system is marginally stable. If 𝐾 > 48
root locus cross the imaginary axis and the system is unstable, if 𝐾 < 48 system is stable.
Problem: Prove that part of the root loci is a circle using angle criterion and find the centre as
𝐾(𝑠+2)
well as radius when 𝐺(𝑠)𝐻(𝑠) = .
𝑠(𝑠+1)

Solution: Let 𝑠 = 𝛼 + 𝑗𝛽 be a point on the root locus which is circle will satisfies the angle
criterion ∠𝐺(𝑠)𝐻(𝑠) = ±1800 (2𝑞 + 1)
𝐾(𝛼 + 𝑗𝛽 + 2)
𝐺(𝑠)𝐻(𝑠)|𝑠=𝛼+𝑗𝛽 =
𝛼 + 𝑗𝛽 (𝛼 + 𝑗𝛽 + 1)
If this point is on the root locus, it should satiety the angle condition
𝛽
𝑡𝑎𝑛−1 (𝛼+2)
∠𝐺(𝑠)𝐻(𝑠) = 𝛽 𝛽
= 1800
𝑡𝑎𝑛−1 (𝛼) + 𝑡𝑎𝑛−1 (𝛼+1)
𝛽 𝛽 𝛽
= 𝑡𝑎𝑛−1 ( ) − 𝑡𝑎𝑛−1 ( ) − 𝑡𝑎𝑛−1 ( ) = 1800
𝛼+2 𝛼 𝛼+1
 𝛽 𝛽 𝛽
= 𝑡𝑎𝑛−1 ( ) − [𝑡𝑎𝑛−1 ( ) + 𝑡𝑎𝑛−1 ( )] = 1800
𝛼+2 𝛼 𝛼+1
𝐴±𝐵
From the trigonometric identity 𝑡𝑎𝑛−1 𝐴 ± 𝑡𝑎𝑛−1 𝐵 = 𝑡𝑎𝑛−1 (1∓𝐴𝐵)
𝛽 𝛽
𝛽 𝛽 + 𝛽 + 2𝛼𝛽
∴ 𝑡𝑎𝑛−1 ( ) + 𝑡𝑎𝑛−1 ( ) = 𝑡𝑎𝑛−1 ( 𝛼 𝛽 𝛼+1𝛽 ) = 𝑡𝑎𝑛−1 ( 2 )
𝛼 𝛼+1 1− × 𝛼 + 𝛼 − 𝛽2
𝛼 𝛼+1

𝛽 𝛽 + 2𝛼𝛽
∠𝐺(𝑠)𝐻(𝑠) = 𝑡𝑎𝑛−1 ( ) − 𝑡𝑎𝑛−1 ( 2 ) = 1800
𝛼+2 𝛼 + 𝛼 − 𝛽2
𝛽 𝛽 + 2𝛼𝛽
𝑡𝑎𝑛−1 ( ) − 𝑡𝑎𝑛−1 ( 2 ) = 1800
𝛼+2 𝛼 + 𝛼 − 𝛽2
Take tan on both the sides, we get
𝛽 𝛽 + 2𝛼𝛽
− 2 =0
𝛼 + 2 𝛼 + 𝛼 − 𝛽2
𝛽(𝛼 2 + 𝛼 − 𝛽 2 ) − (𝛽 + 2𝛼𝛽)(𝛼 + 2)
=0
(𝛼 + 2)(𝛼 2 + 𝛼 − 𝛽 2 )
𝛽(𝛼 2 + 𝛼 − 𝛽 2 ) − (𝛽 + 2𝛼𝛽)(𝛼 + 2) = 0
(𝛼 2 + 𝛼 − 𝛽 2 ) − (1 + 2𝛼)(𝛼 + 2) = 0
𝛼 2 + 𝛼 − 𝛽 2 − 2𝛼 2 − 5𝛼 − 2 = 0
−𝛼 2 − 4𝛼 − 𝛽 2 − 2 = 0
𝛼 2 + 4𝛼 + 𝛽 2 + 2 = 0
𝛼 2 + 4𝛼 + 𝛽 2 + 2 + 2 − 2 = 0
𝛼 2 + 4𝛼 + 𝛽 2 + 2 + 2 = 2
(𝛼 + 2)2 + 𝛽 2 = (√2)2
It is similar to equation of circle(𝑥 − 𝑎)2 + (𝑦 − 𝑏)2 = 𝑐 2 . Thus above equation is circle with
centre (−2, 0)and radius√2. The root locus plot is as shown in below figure.
The open loop poles are at 𝑠1 = 0 𝑎𝑛𝑑 𝑠2 = −1. And zeros are 𝑠0 = −2.
Number of poles 𝑁 = 2 and number of zeros𝑀 = 1, Number of root locus branches are 2
Problem: The open loop transfer function of a unity feedback system is given by𝐺(𝑠) =
𝐾
. Sketch the root locus of the system.
𝑠(𝑠+4)(𝑠2 +4𝑠+20)
𝐾
Solution: Let us consider the open loop transfer function 𝐺(𝑠) = 𝑠(𝑠+4)(𝑠2 +4𝑠+20)

Rule 1: Root locus is symmetrical about the real axis

Rule 2: The open loop poles are
𝑠(𝑠 + 4)(𝑠 2 + 4𝑠 + 20) = 0
The poles are, 𝑠 = 0, s = −4, 𝑠 = −2 + j4, 𝑠 = −2 − 𝑗4 thus number of poles 𝑁 = 4
No zeros on the s-plane, thus 𝑀 = 0
The poles are marked as × in the s-plane
Root locus branches starts at poles and terminate at infinity. Number of root locus
branches is equal to the number of poles or zeros whichever is greater. Therefore number
of root locus branches is equal to 4.
Rule 3: A point on the real axis lies on the locus, if the number of open loop pole plus
zeros on the real axis to the right of this point is odd.
Rule 4: Among four root loci, two root locus branches on real axis, start at pole 𝑠 = 0
and 𝑠 = −4 travel on the negative real axis and two root loci meet at a break-away point
and terminate at infinity. The other two root locus branches start at complex pole 𝑠 =
−2 + j4, 𝑠 = −2 − j4 and terminate at infinity. Thus we need four asymptotes. Therefore
we have to compute the angle of asymptote.
1800 (2𝑞+1)
Angle of asymptote ∅𝑞 = ± , 𝑞 = 0,1,2,3 … (𝑁 − 𝑀) = 0,1, 2,3
𝑁−𝑀

180
∅0 = ± = ±450
4−0
1800 (4 + 1)
∅1 = ± = ± 1350
4−0
1800 (4 + 1)
∅2 = ± = ±225 = ∓1350
4−0
1800 (6 + 1)
∅3 = ± = ±315 = ∓450
4−0
Rule 5: Centroid is the meeting point of the asymptotes on the real axis. At this point the
angles of asymptote are measured and asymptotes are drawn with dotted line.
Sum of poles − Sum of zeros 0 − 4 − 2 + j4 − 2 − j4
Centroid 𝜎𝐴 = = = −2
N−M 4
Rule 6: Breakaway point is the point at which two branches of root locus meet. To
determine breakaway point, consider the characteristic equation 1 + 𝐺(𝑠)𝐻(𝑠) = 0 and
𝑑𝑘
determined from the roots of the equation =0
𝑑𝑠

1 + 𝐺(𝑠)𝐻(𝑠) = 0
𝐾
1 + 𝐺(𝑠)H(s) = 1 + =0
𝑠(𝑠 + 4)(𝑠 2+ 4𝑠 + 20)
𝑠(𝑠 + 4)(𝑠 2 + 4𝑠 + 20) + 𝑘 = 0
𝑘 = −𝑠(𝑠 + 4)(𝑠 2 + 4𝑠 + 20)
 𝑘 = −(𝑠 4 + 8𝑠 3 + 36𝑠 2 + 80𝑠)
𝑑𝑘
= −(4𝑠 3 + 24𝑠 3 + 72𝑠 + 80) = 0
𝑑𝑠
4𝑠 3 + 24𝑠 3 + 72𝑠 + 80 = 0
𝑠 3 + 6𝑠 3 + 18𝑠 + 20 = 0
𝑠 3 + 6𝑠 3 + 18𝑠 + 20 = (𝑠 + 2)(𝑠 2 + 4𝑠 + 10) = 0
The roots are 𝑠1 = −2, 𝑠2 = −2 + 𝑗2.45, 𝑠3 = −2 − 𝑗2.45
Rule 7: The angle departure is angle at which the root locus branch departure at the
complex pole. Angele of departure at the complex pole
𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒
∅ 𝑑 = 1800 − (𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑝𝑜𝑙𝑒 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠)
𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒
+( )
𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑧𝑒𝑟𝑜𝑠 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠

∅ 𝑑 = 1800 − ϕp + ϕz
Let us consider the complex pole at −2 + 𝑗64 and draw the vector from the other poles
and zeros. Let the angles of these vectors be 𝜃1 , 𝜃2 , 𝜃3 as shown in below figure.

4
Here 𝜃1 = 180 − 𝑡𝑎𝑛−1 (2) = 117.10

𝜃2 = 900
4
𝜃3 = 𝑡𝑎𝑛−1 ( ) = 630
2
𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒
ϕ = 𝜃1 + 𝜃2 + 𝜃3 = 117.10 + 900 + 630 = 270.10
𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑝𝑜𝑙𝑒 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠 p
 𝐴𝑛𝑔𝑙𝑒 𝑜𝑓 𝑑𝑒𝑝𝑎𝑟𝑡𝑢𝑟𝑒
} ∅ 𝑑 = 1800 − ϕp = 1800 − 270.10 = −900
𝑎𝑡 𝑡ℎ𝑒 𝑐𝑜𝑚𝑝𝑙𝑥 𝑝𝑜𝑙𝑒 − 2 + 𝑗4
𝐴𝑛𝑔𝑙𝑒 𝑜𝑓 𝑑𝑒𝑝𝑎𝑟𝑡𝑢𝑟𝑒
} ∅ 𝑑 = −900
𝑎𝑡 𝑡ℎ𝑒 𝑐𝑜𝑚𝑝𝑙𝑥 𝑝𝑜𝑙𝑒 − 2 − 𝑗4
Mark the angles of departure at complex poles.
Rule 8: Intersection point of imaginary axis and the root locus.
For this we have to develop the characteristic equation of the closed loop transfer
function i.e. 1 + 𝐺(𝑠)𝐻(𝑠) = 0
𝐾
1 + 𝐺(𝑠)H(s) = 1 + =0
𝑠(𝑠 + 4)(𝑠 2+ 4𝑠 + 20)
𝑠(𝑠 + 4)(𝑠 2 + 4𝑠 + 20) + 𝑘 = 0
𝑠 4 + 8𝑠 3 + 36𝑠 2 + 80𝑠 + 𝐾
Put 𝑠 = 𝑗𝜔 in the characteristic equation
(𝑗𝜔)4 + 8(𝑗𝜔)3 + 36(𝑗𝜔)2 + 80𝑗𝜔 + 𝐾 = 0
𝜔4 − 8𝑗𝜔3 − 36𝜔 2 + 80𝑗𝜔 + 𝐾 = 0

On equating the imaginary part to zero

−8𝑗𝜔 3 + 𝑗80𝜔 = 0
−8 𝜔 2 + 80 = 0
− 𝜔 2 + 10 = 0
𝜔 2 = 10
𝜔 = ±3.2
Equating the real part to zero
𝜔4 − 36𝜔 2 + 𝐾 = 0
𝐾 = −𝜔4 + 36𝜔 2
Put 𝜔 2 = 10, then
𝐾 = −100 + 36 × 10 = 260
The crossing point of root locus is±𝑗3.2. The value of K at this point is𝐾 = 260.
Plot the root locus as shown in below figure. Root locus is symmetrical about the real axis. From
the plot, if 𝐾 = 260 root locus is at imaginary axis. If 𝐾 > 260 root locus cross the imaginary
axis and the system is unstable, if 𝐾 < 260 system is stable.
Problem: The open loop transfer function of a unity feedback system is given by𝐺(𝑠) =
𝐾(𝑠+1.5)
. Sketch the root locus of the system.
𝑠(𝑠+1)(𝑠+5)
𝐾(𝑠+1.5)
Solution: Let us consider the open loop transfer function 𝐺(𝑠) = 𝑠(𝑠+1)(𝑠+5)

Rule 1: Root locus is symmetrical about the real axis

Rule 2: The open loop poles are
𝑠(𝑠 + 1)(𝑠 + 5) = 0
The poles are at, 𝑠 = 0, s = −1, 𝑠 = −5 thus number of poles 𝑁 = 3
The zeros are at 𝑠 = −1.5, thus 𝑀 = 1
The poles are marked as × in the s-plane and zeros are marked as 0.
Root locus branches starts at poles and terminate at zeros and infinity. Number of root
locus branches is equal to the number of poles or zeros whichever is greater. Therefore
number of root locus branches is equal to 3.
Rule 3: A point on the real axis lies on the locus, if the number of open loop pole plus
zeros on the real axis to the right of this point is odd.
Rule 4: Among three root loci, two root locus branches on real axis, start at pole 𝑠 = 0
and 𝑠 = −1 and axis approaches to infinity thus we need two asymptotes and remaining
one root loci start at −5 and terminate at −1.5.
1800 (2𝑞+1)
Angle of asymptote ∅𝑞 = ± , 𝑞 = 0,1,2,3 … (𝑁 − 𝑀) = 0,1, 2
𝑁−𝑀
180
∅0 = ± = ±900
3−1
1800 (4 + 1)
∅1 = ± = ± 4500 = ±900
3−1
Rule 5: Centroid is the meeting point of the asymptotes on the real axis. At this point the
angles of asymptote are measured and asymptotes are drawn with dotted line.
Sum of poles − Sum of zeros (0 − 1 − 5) − (−1,5)
Centroid 𝜎𝐴 = = = −2.25
N−M 4
Rule 6: Breakaway point is the point at which two branches of root locus meet. To
determine breakaway point, consider the characteristic equation 1 + 𝐺(𝑠)𝐻(𝑠) = 0 and
𝑑𝑘
determined from the roots of the equation =0
𝑑𝑠

1 + 𝐺(𝑠)𝐻(𝑠) = 0
𝐾(𝑠 + 1.5)
1 + 𝐺(𝑠)H(s) = 1 + =0
𝑠(𝑠 + 1)(𝑠 + 5)
𝑠(𝑠 + 1)(𝑠 + 5) + 𝐾(𝑠 + 1.5) = 0
−𝑠(𝑠 + 1)(𝑠 + 5) −(𝑠 3 + 6𝑠 2 + 5𝑠)
𝑘= =
𝑠 + 1.5 𝑠 + 1.5
𝑑𝑘 −(3𝑠 2 + 12𝑠 + 5)(𝑠 + 1.5) − [−(𝑠 3 + 6𝑠 2 + 5𝑠)]
= =0
𝑑𝑠 (𝑠 + 1.5)2
−(3𝑠 2 + 12𝑠 + 5)(𝑠 + 1.5) − [−(𝑠 3 + 6𝑠 2 + 5𝑠) = 0
 −2(𝑠 3 + 5.229𝑠 + 3.75) = 0
𝑠 3 + 5.229𝑠 + 3.75 = 0
The roots are𝑠1 = −0.6, 𝑠2 = −2.3 + 𝑗0.89, 𝑠3 = −2.3 − 𝑗0.89, and −0.6 is the valid
breakaway point
Rule 7: Since there is no complex poles or zeros we need not compute angle of
departure.
Rule 8: Intersection point of imaginary axis and the root locus.
For this we have to develop the characteristic equation of the closed loop transfer
function i.e. 1 + 𝐺(𝑠)𝐻(𝑠) = 0
1 + 𝐺(𝑠)𝐻(𝑠) = 0
𝐾(𝑠 + 1.5)
1 + 𝐺(𝑠)H(s) = 1 + =0
𝑠(𝑠 + 1)(𝑠 + 5)
𝑠(𝑠 + 1)(𝑠 + 5) + 𝐾(𝑠 + 1.5) = 0
𝑠 3 + 6𝑠 2 + 5𝑠 + 𝐾𝑠 + 1.5𝐾 = 0
Put 𝑠 = 𝑗𝜔 in the characteristic equation
(𝑗𝜔)3 + 6(𝑗𝜔)2 + 5𝑗𝜔 + 𝐾𝑗𝜔 + 1.5𝐾 = 0
−𝑗𝜔3 − 6𝜔 2 + 5𝑗𝜔 + 𝐾𝑗𝜔 + 1.5𝐾 = 0
On equating the imaginary part to zero
−𝑗𝜔3 + 5𝑗𝜔 + 𝐾𝑗𝜔 = 0
−𝜔2 + 5 + 𝐾 = 0
𝜔2 = 5 + 𝐾
Equating the real part to zero
−6𝜔 2 + 1.5𝐾 = 0
Put 𝜔 2 = 5 + 𝐾, then
−6(5 + 𝐾) + 1.5𝐾 = 0
30
𝐾=− = −6.67
4.5
Since the value of K is negative, there is no crossing point on the imaginary axis.
Plot the root locus as shown in below figure. Root locus is symmetrical about the real axis.
Problem: The open transfer function of a unity feedback control system is given
𝐾
by 𝐺(𝑠) = 𝑠(𝑠+2)(𝑠2+6𝑠+25), sketch the root locus as 𝐾 is varied from 0 to infinity.
𝐾
Solution: Let consider the open loop transfer function 𝐺(𝑠) = 𝑠(𝑠+2)(𝑠2 +6𝑠+25)

Rule 1: Root locus is symmetrical about the real axis (𝜎 𝑎𝑥𝑖𝑠)

Rule 2: Number of poles 𝑁 = 4, poles is 𝑠 = 0, −2, −3 + 𝑗4, −3 − 𝑗4
Number of zeros 𝑀 = 0,
Numbers of root locus branches are 4
The poles are marked as × and zeros are marked as 0 in the s-plane
Root locus branches starts at poles and terminate at infinity.
Rule 3: A point on the real axis lies on the locus, if the number of open loop pole plus
zeros on the real axis to the right of this point is odd.
Rule 4: The (𝑁 − 𝑀) = 4branches of root locus terminates at infinity, do so along
straight line asymptotes whose angels are given by, that is angle of asymptotes
1800 (2𝑞 + 1)
∅𝑞 = ± , 𝑞 = 0,1,2,3 … (𝑁 − 𝑀)
𝑁−𝑀
1800
∅0 = ± = ±450
4
1800 × 3
∅1 = ± = ±1350
4
1800 × 5
∅2 = ± = ±2250
4
1800 × 7
∅2 = ± = ±3150
4
Rule 5: The asymptotes cross the real axis at a point known as centroid, determined by
the relationship:
(sum of real part of poles – sum of real part of zeros)
Centroid 𝜎𝐴 =
𝑁−𝑀
0−2−3−3
Centroid 𝜎𝐴 = = −2
4
Rule 6: Breakaway point is the point at which two branches of root locus meet. To
determine breakaway point, consider the characteristic equation 1 + 𝐺(𝑠)𝐻(𝑠) = 0 and
𝑑𝑘
determined from the roots of the equation 𝑑𝑠 = 0. The characteristic equation is 1 +
𝐾
𝐺(𝑠)𝐻(𝑠) = 1 + 𝑠(𝑠+2)(𝑠2 +6𝑠+25) = 0

𝐾 = −(𝑠(𝑠 + 2)(𝑠 2 + 6𝑠 + 25)) = −(𝑠 4 + 8𝑠 3 + 37𝑠 2 + 50𝑠)

𝐾 = −(𝑠 4 + 8𝑠 3 + 37𝑠 2 + 50𝑠)
𝑑𝑘
= −4𝑠 3 − 24𝑠 2 − 74𝑠 − 50 = 0
𝑑𝑠
 = 4𝑠 3 + 24𝑠 2 + 74𝑠 + 50 = 0
= 𝑠 3 + 6𝑠 2 + 18.5𝑠 + 12.5 = 0
We know that, the breakaway point is between −2 and 0 . By trial and error method, if
𝑠 = −0.89 then
= 𝑠 3 + 6𝑠 2 + 18.5𝑠 + 12.5 = (𝑠 + 0.89)(𝑠 2 + 5.11𝑠 + 13.49) = 0
Then the roots are, 𝑠 = −0.89, 𝑠 = −2.55 + 𝑗5.277, 𝑠 = −2.55 − 𝑗5.277
Since the break away point is between −2 and 0, the valid break away point is 𝑠 =
−0.89.
Rule 7: Angle of departure from the complex poles is given by
𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒
∅ 𝑑 = 1800 − (𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑝𝑜𝑙𝑒 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠)

𝑠𝑢𝑚 𝑜𝑓 𝑎𝑛𝑔𝑙𝑒𝑠 𝑜𝑓 𝑣𝑒𝑐𝑡𝑜𝑟 𝑡𝑜 𝑡ℎ𝑒

+( )
𝑐𝑜𝑚𝑝𝑙𝑒𝑥 𝑧𝑒𝑟𝑜𝑠 𝑓𝑟𝑜𝑚 𝑡ℎ𝑒 𝑜𝑡ℎ𝑒𝑟 𝑣𝑒𝑐𝑡𝑜𝑟𝑠

∅ 𝑑 = 1800 − ϕp + ϕz

4
𝜃1 = 180 − 𝑡𝑎𝑛−1 ( ) = 126. 860 ~1270
−3
4
𝜃2 = 180 − 𝑡𝑎𝑛−1 ( ) = 1040
−1
𝜃3 = 900
∅ 𝑑 = 1800 − ϕp + ϕz
ϕp = 1270 + 1040 + 900 = 3210 ,ϕz = 0
∅ 𝑑 = 1800 − ϕp + ϕz
 ∅ 𝑑 = 1800 − (3210 ) = −1410
Rule 8: Intersection of root locus branches with the imaginary axis can be determined
using Routh array or any other methods. Let us consider the characteristic equation
𝑠 4 + 8𝑠 3 + 37𝑠 2 + 50𝑠 + 𝐾 = 0. Put𝑠 = 𝑗𝜔, we get
(𝑗𝜔)4 + 8(𝑗𝜔)3 + 37(𝑗𝜔)2 + 50(𝑗𝜔) + 𝐾 = 0
𝜔4 − 8 𝑗𝜔3 − 37𝜔2 + 50𝑗𝜔 + 𝐾 = 0
Equating the real and imaginary to zero, then
−8 𝑗𝜔3 + 50𝑗𝜔 = 0
−8 𝜔2 + 50 = 0
𝜔2 = 6.25
𝜔 = 2.5 𝑅𝑎𝑑/𝑠𝑒𝑐
And real part is 𝜔4 − 37𝜔2 + 𝐾 = 0
6.252 − 37 × 6.25 + 𝐾 = 0
𝐾 = 231.25 − 39.06 = 192.19
Put 𝜔2 = 6.25, the 𝐾 = 192.19 = 𝐾𝑚𝑎𝑟 , Root locus sketch is
The crossing point of root locus is±𝑗√6.25 = ±𝑗2.5. The value of K at this point is 𝐾 =
48. Plot the root locus as shown in below figure. Root locus is symmetrical about the real
axis. From the plot, if 𝐾 = 48 root locus is at imaginary axis, system is marginally stable.
If 𝐾 > 48 root locus cross the imaginary axis and the system is unstable, if 𝐾 < 48
system is stable.
 Module 5
Frequency domain analysis and stability:
Frequency domain analysis: Frequency domain analysis is defined as the steady state response
of the system due to sinusoidal input. The frequency response of the system is normally obtained
by varying the frequency of the sinusoidal input by keeping the magnitude of input at constant
value.

𝐶(𝑠) 𝐺(𝑠)
Consider a closed loop transfer function 𝑅(𝑠) = 1+𝐺(𝑠)𝐻(𝑠) = 𝑇(𝑠)
 Now if we replace 𝑠 = 𝑗𝜔, the resulting transfer function is known as sinusoidal transfer
function.
𝐶(𝑗𝜔) 𝐺(𝑗𝜔)
= = 𝑇(𝑗𝜔) = |𝑇(𝑗𝜔)|∠𝑇(𝑗𝜔)
𝑅(𝑗𝜔) 1 + 𝐺(𝑗𝜔)𝐻(𝑗𝜔)
The sinusoidal transfer function 𝑇(𝑗𝜔) is a complex quantity, it has both magnitude and
phase angle. Therefore for frequency response has two components, magnitude and
phase, thus the frequency response consists of magnitude plot and phase plot. The
magnitude plot is as shown in below figure.
Correlation between time and frequency response
Let us discus the correlation between time and frequency of the second order system. For
𝐶(𝑠) 𝜔2
second order system closed loop transfer function 𝑅(𝑠) = 𝑠2 +2𝜉𝜔𝑛 𝑠+𝜔2 . In frequency
𝑛 𝑛

domain,
𝐶(𝑗𝜔) 𝜔𝑛2 𝜔𝑛2 𝜔𝑛2
= = = 𝜔2
𝑅(𝑗𝜔) (𝑗𝜔)2 + 2𝜉𝜔𝑛 (𝑗𝜔) + 𝜔𝑛2 −𝜔 2 + 𝑗2𝜉𝜔𝑛 𝜔 + 𝜔𝑛2 𝜔
𝜔𝑛2 (− 𝜔2 ) + 𝑗2𝜉 𝜔 + 1)
𝑛 𝑛

1 1
𝑇(𝑗𝜔) = 𝜔 𝜔 =
1− (𝜔 )2 + 𝑗2𝜉 𝜔 (1 − 𝑢2 ) + 2𝑗𝜉𝑢
𝑛 𝑛
𝜔
Let 𝑢 = 𝜔 is the normalized frequency.
𝑛

1 1
𝑇(𝑗𝜔) = × 2𝜉𝑢
√(1 − 𝑢2 )2 + (2𝜉𝑢)2 𝑡𝑎𝑛−1 (1−𝑢2 )

Therefore magnitude of closed loop transfer function is

1
|𝑇(𝑗𝜔)| = 𝑀 = … … . (𝐴)
√(1 − 𝑢2 )2 + (2𝜉𝑢)2
And phase angle of closed loop transfer function is
2𝜉𝑢
∠𝑇(𝑗𝜔) = ∅ = −𝑡𝑎𝑛−1 (1−𝑢2) … … (𝐵).

Using |𝑇(𝑗𝜔)| and ∠𝑇(𝑗𝜔) we can plot the normalized frequency response for different value of
damping ratio. From equation (A) and (B)
𝑖𝑓 𝑢 = 0 𝑀=1 𝑎𝑛𝑑 𝜑 = 0
1
𝑢=1 𝑀= 𝑎𝑛𝑑 𝜑 = −900
2𝜉
𝑢→∞ 𝑀→0 𝜑 → −1800
The magnitude and phase characteristic for normalized frequency 𝑢 for certain value of ξ are
shown in below figure. The frequency where M has a peak value is known as the resonant
frequency. At this frequency slope of the magnitude curve is zero. Let 𝜔𝑟 be the resonant
𝜔
frequency and 𝑢𝑟 = 𝜔 𝑟 be the normalized resonant frequency. Then
𝑛

𝑑𝑀 1 [−4(1 − 𝑢𝑟2 )𝑢𝑟 + 8𝜉 2 𝑢𝑟 ]

| =− =0
𝑑𝑢 𝑢=𝑢𝑟 2 [(1 − 𝑢𝑟 )2 + (2𝜉𝑢𝑟 )2 ]3⁄4

−4(1 − 𝑢𝑟2 )𝑢𝑟 + 8𝜉 2 𝑢𝑟 = 0

4𝑢𝑟3 − 4𝑢𝑟 + 8𝜉 2 𝑢𝑟 = 0
𝑢𝑟2 − 1 + 2𝜉 2 = 0
𝑢𝑟 = √1 − 2𝜉 2

𝜔𝑟 = 𝜔𝑛 √1 − 2𝜉 2
And the maximum value of the magnitude known as the resonant peak is given by
1
𝑀𝑟 = … … . (𝐶)
2𝜉√(1−𝜉 2 )

And the phase angle 𝜑 at the resonant frequency is given by

−1
√1 − 2𝜉 2
𝜑𝑟 = 𝑡𝑎𝑛 [ ] … . . (𝐷)
𝜉
From equation (C) and (D) it is observed that as 𝜉 approaches to zero𝜔𝑟 approaches to 𝜔𝑛 and 𝑀𝑟
approaches to infinity. For 0 ≤ 𝜉 ≤ 1⁄ the resonant frequency 𝜔𝑟 always has a value less than
√2
𝜔𝑛 and resonant peak 𝑀𝑟 has a value greater than one.
The bandwidth is defined as the range of frequency over which the gain M is greater than or
1
equal to . The typical frequency response of a feedback control system as shown in below
√2

figure. We know that, magnitude

1
𝑀=
√(1 − 𝑢2 )2 + (2𝜉𝑢)2
𝜔
Let 𝑢𝑏 = 𝜔𝑏 be the normalized bandwidth, then magnitude
𝑛

1 1
𝑀|𝑢=𝑢𝑏 = =
√2 √(1 − 𝑢𝑏2 )2 + (2𝜉𝑢𝑏 )2
Square on both sides, we get 2 = (1 − 𝑢𝑏2 )2 + (2𝜉𝑢𝑏 )2
 1 + 𝑢𝑏4 − 2𝑢𝑏2 + 4𝜉 2 𝑢𝑏2 = 2
1
Solving for normalized bandwidth 𝑢𝑏 we get 𝑢𝑏 = [1 − 2𝜉 2 + √2 − 4𝜉 2 + 4𝜉 4 ]2
1
And demoralized band width 𝜔𝑏 = 𝑢𝑏 𝜔𝑛 = 𝜔𝑛 [1 − 2𝜉 2 + √2 − 4𝜉 2 + 4𝜉 4 ]2
Frequency response specifications: Frequency response specifications are
1. Resonant peak, 𝑴𝒓
2. Resonant frequency, 𝝎𝒓
3. Bandwidth
4. Gain margin, 𝑲𝒈
5. Phase margin, 𝜸
Resonant peak: The maximum value of the magnitude of the closed loop transfer
function.
Resonant frequency: It is the frequency at which the resonant peak occurs.
Bandwidth: It is range of the frequency the system gain is more than−3𝑑𝐵.
Gain margin: It is defined as the reciprocal of the magnitude of open loop transfer
function at phase crossover frequency. The frequency at which the phase of open loop
transfer function is 1800 is called the phase cross over frequency 𝜔𝑝𝑐 . The gain margin in
dB is expressed as
1
𝐾𝑔 = 20 log = −20 log|𝐺(𝑗𝜔𝑝𝑐 )|
|𝐺(𝑗𝜔𝑝𝑐 )|
 Phase margin: Phase margin is the amount of additional phase lag at the gain crossover
frequency 𝜔𝑔𝑐 and it is given by 𝛾 = 180 + 𝜙𝑔𝑐 . From phase margin and gain margin we
determine the stability of the system.

Problem: The forward path transfer function of a unity feedback control system is given
100
by 𝐺(𝑠) = 𝑠(𝑠+6.54). Determine resonant peak, resonant frequency and band width of the closed

loop system.
100
Solution: The open loop transfer function 𝐺(𝑠) = 𝑠(𝑠+6.54) and the closed loop transfer function

is
𝐶(𝑠) 𝐺(𝑠) 100 𝜔𝑛2
= = =
𝑅(𝑠) 1 + 𝐺(𝑠)𝐻(𝑠) 𝑠 2 + 6.54𝑠 + 100 𝑠 2 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
Thus 𝜔𝑛 = √100 = 10𝑅𝑎𝑑/𝑠𝑒𝑐
2𝜉𝜔𝑛 = 6.54𝑠, thus 𝜉 = 0.327
Resonant frequency 𝜔𝑟

𝜔𝑟 = 𝜔𝑛 √1 − 2𝜉 2 = 10√1 − 2 × (0.327)2 = 8.866𝑅𝑎𝑑/𝑠𝑒𝑐

Resonant peak 𝑀𝑟
1 1
𝑀𝑟 = = = 1.618
2𝜉√(1 − 𝜉 2 ) 2 × 0.327√1 − (0.327)2
Bandwidth
 1
𝜔𝑏 = 𝑢𝑏 𝜔𝑛 = 𝜔𝑛 [1 − 2𝜉 2 + √2 − 4𝜉 2 + 4𝜉 4 ]2
1
= 10[1 − 2 × (0.327)2 + √2 − 4 × (0.327)2 + 4(0.327)4 ]2 = 14.34𝑅𝑎𝑑/𝑠𝑒𝑐
Problem: The forward path transfer function of a unity feedback control system is given by
𝐾
𝐺(𝑠) = 𝑠(𝑠+𝑎). Determine the value of k, and a, if resonant peak 𝑀𝑟 = 1.25, resonant

frequency𝜔𝑟 = 12.65𝑅𝑎𝑑/𝑠𝑒𝑐, and also find band width and settling time.
𝐾
Solution: Open loop transfer function 𝐺(𝑠) = 𝑠(𝑠+𝑎) and closed loop transfer function

𝐶(𝑠) 𝐺(𝑠) 𝑘 𝜔𝑛2

= = 2 = 2
𝑅(𝑠) 1 + 𝐺(𝑠)𝐻(𝑠) 𝑠 + 𝑎𝑠 + 𝑘 𝑠 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
𝜔𝑛2 = 𝑘, and 2𝜉𝜔𝑛 = 𝑎
𝜔𝑛 = √𝑘
𝑎 𝑎
𝜉= =
2𝜔𝑛 2√𝑘
Resonant peak 𝑀𝑟
1
𝑀𝑟 = = 1.25
2𝜉√(1−𝜉 2 )

1 1
1.25 = =
2𝑎 𝑎 𝑎 𝑎
2√𝑘
√(1 − (2√𝑘)2 ) √𝑘
√(1 − (2√𝑘)2 )

𝑎 1
1.25 × = … … . . (1)
√𝑘 𝑎
)2 )
√(1 − (2√𝑘

Resonant frequency 𝜔𝑟

𝜔𝑟 = 𝜔𝑛 √1 − 2𝜉 2 = 12.65

𝑎
√𝑘 × √1 − 2( )2 = 12.65 … … (2)
2√𝑘

By solving equation (1) and (2) we get 𝑘 = 266.67 and 𝑎 = 14.60

𝑎 14.60
𝜉 = 2√𝑘 = 2√266.67 = 0.447, and 𝜔𝑛 = √𝑘 = √266.67 = 16.33𝑅𝑎𝑑/𝑆𝑒𝑐
4 4 8 8
Settling time 𝑡𝑠 = 𝜉𝜔 = 𝑎 = 𝑎 = 14.60 = 0.54𝑆𝑒𝑐
𝑛 ×√𝑘
2√𝑘
1
Band width 𝜔𝑏 = 𝜔𝑛 [1 − 2𝜉 2 + √2 − 4𝜉 2 + 4𝜉 4 ]2
 1
𝜔𝑏 = 16.33[1 − 2(0.447)2 + √2 − 4(0.447)2 + 4(0.447)4 ]2 = 21.70𝑅𝑎𝑑/𝑠𝑒𝑐
Problem: Find the phase margin for the system having the open loop transfer function 𝐺(𝑠) =
10
.
𝑠
10 10
Solution:𝐺(𝑠) = put 𝑠 = 𝑗𝜔 then 𝐺(𝑗𝜔) = 𝑗𝜔
𝑠

𝐺(𝑗𝜔) = |𝐺(𝑗𝜔)|∠𝐺(𝑗𝜔) = |𝑀|∠𝜑

10 00 00
Magnitude |𝐺(𝑗𝜔)| = And phase ∠𝜑 = 𝜔 = 900 = −900
𝜔 𝑡𝑎𝑛−1 ( )
0

Phase margin is calculated by equating gain to unity at gain crossover frequency. i.e.
10
=1 ∴ 𝜔 = 10𝑟𝑎𝑑/𝑠𝑒𝑐 = 𝜔𝑔𝑐
𝜔
We know that, phase margin is
Phase margin= 1800 + 𝜑|𝜔𝑔𝑐 = 1800 − 900 = 900
Problem: Find the gain margin for the system having the open loop transfer function
10
𝐺(𝑠) = 𝑠(𝑠+10)2.
10 10
Solution:𝐺(𝑠) = 𝑠(𝑠+10)2 put 𝑠 = 𝑗𝜔 then 𝐺(𝑗𝜔) = 𝑗𝜔(𝑗𝜔+10)2

𝐺(𝑗𝜔) = |𝐺(𝑗𝜔)|∠𝐺(𝑗𝜔) = |𝑀|∠𝜑

10 00 𝜔
Magnitude |𝐺(𝑗𝜔)| = and phase ∠𝜑 = 𝜔 𝜔 = −900 − 2𝑡𝑎𝑛−1 (10)
𝜔√(𝜔 2 +102 )2 𝑡𝑎𝑛−1 ( )+2𝑡𝑎𝑛−1 ( )
0 10

Gain margin is calculated by equating phase to −1800 at phase crossover frequency. i.e.
𝜔
−900 − 2𝑡𝑎𝑛−1 ( ) = −1800
10
𝜔
−2𝑡𝑎𝑛−1 ( ) = −1800 + 900 = −900
10
𝜔 900
𝑡𝑎𝑛−1 ( )= = 450
10 2
𝜔
= 1, thus 𝜔 = 10 = 𝜔𝑝𝑐 = 10𝑅𝑎𝑑/𝑠𝑒𝑐
10

Now gain margin at 𝜔 = 𝜔𝑝𝑐 is

10 10 1
Gain margin |𝐺(𝑗𝜔) |𝜔= 𝜔𝑝𝑐 = = = 200
𝜔√(𝜔 2 +102 )2 10√(102 +102 )2
1 1
Gain margin in decibel 20 log10 𝑔𝑎𝑖𝑛 = 20 log10 1 = 46𝑑𝐵
200

Bode Plots
The Bode plot is a frequency response plot of the transfer function of a system. Frequency 𝜔 is
varied from 0 to ∞. Generally we are not considering all points of 𝜔 form 0 to ∞, we are
considering only some values from 0 to ∞. Therefore Bode plot is also known as asymptotic or
corner plot. Bode plot can be plotted for both open loop and closed loop transfer function. The
Bode plot consists of magnitude plot and phase plot. In magnitude plot, the plot is logarithmic
magnitude of 𝐺(𝑗𝜔) that is 20log10 |𝐺(𝑗𝜔)| in dB and frequency 𝜔. The both magnitude and
phase plot is on the semilog paper, using logarithmic scale for frequency and linear scale for
magnitude in dB or phase angle in degree. The Bode plot is the mathematical tool to find the
stability of the system.
The logarithmic scale is a nonlinear scale used for frequency. It has some interesting properties,
that is, the distance between 1 and 2 is greater than 2 and 3 and so on. As a result, use of this
scale enables to cover a greater range of frequencies. It is interesting to note that on the log scale
that, the distance between 1 and 10 is equal to the distance between 10 and 100. This distance is
called decade. Since 𝑙𝑜𝑔 10 0 = ∞, frequency 𝜔 starts at 0.1 instead of 0.
Basic factors of 𝑮(𝒋𝝎)
The basic factors that very frequently occur in a typical transfer function 𝐺(𝑗𝜔) are,
1. Constant gain K
𝐾 𝐾
2. Integral factor 𝑗𝜔 or (𝑗𝜔)𝑛
3. Derivative factor 𝐾(𝑗𝜔) or 𝐾(𝑗𝜔)𝑛
1 1
4. First order factor in the denominator (1+𝑗𝜔𝑇) or (1+𝑗𝜔𝑇)𝑚
5. First order factor in the numerator(1 + 𝑗𝜔𝑇) or (1 + 𝑗𝜔𝑇)𝑚
1
6. Quadratic factor in the denominator 𝑗𝜔 𝑗𝜔 2
1+2𝜉( )+( )
𝜔𝑛 𝜔𝑛
𝑗𝜔 𝑗𝜔
7. Quadratic factor in the numerator1 + 2𝜉 (𝜔 ) + (𝜔 )2
𝑛 𝑛

Bode plot for basic factors of 𝑮(𝒋𝝎)

1. Constant gain K
Let 𝐺(𝑠) = 𝐾
𝐺(𝑗𝜔) = 𝐾
𝐺(𝑗𝜔) = |𝐺(𝑗𝜔)|∠𝐺(𝑗𝜔) = 𝐾∠00
Therefore Magnitude of 𝐺(𝑗𝜔) in dB 𝐴 = 20 log10 𝐾
Phase angle of 𝐺(𝑗𝜔) in degree 𝜑 = 00
The Bode plot for 𝐺(𝑗𝜔) as shown in below figure
When 𝐾 > 0, 20 log10 𝐾 , is positive
When 𝐾 < 0, 20 log10 𝐾 , is negative
When 𝐾 = 0, 20 log10 𝐾 , is zero

2. Integral factor
𝐾
Let 𝐺(𝑠) = 𝑠
𝐾
𝐺(𝑗𝜔) =
𝑗𝜔
𝐾 00 0
𝐺(𝑗𝜔) = |𝐺(𝑗𝜔)|∠𝐺(𝑗𝜔) = ∠ 𝜔 = 𝐾∠−90
𝜔 𝑡𝑎𝑛 ( )
−1
0
𝐾
Therefore Magnitude of 𝐺(𝑗𝜔) in dB 𝐴 = 20 log10 𝜔

Phase angle of 𝐺(𝑗𝜔) in degree 𝜑 = −900

The Bode plot for 𝐺(𝑗𝜔) as shown in below figure
𝐾 𝐾
When 𝜔 = 0.1 𝐾, 𝐴 = 20 log10 𝜔 = 20 log10 0.1𝐾 = 20𝑑𝐵
𝐾 𝐾
When 𝜔 = 𝐾, 𝐴 = 20 log10 = 20 log10 = 0𝑑𝐵
𝜔 𝐾
𝐾 𝐾
When 𝜔 = 10𝐾, 𝐴 = 20 log10 𝜔 = 20 log10 10𝐾 = −20𝑑𝐵
3. Derivative factor
Let 𝐺(𝑠) = 𝐾𝑠
𝐺(𝑗𝜔) = 𝐾(𝑗𝜔)
𝜔
𝐺(𝑗𝜔) = |𝐺(𝑗𝜔)|∠𝐺(𝑗𝜔) = 𝐾𝜔∠ 𝑡𝑎𝑛−1 ( ) = 𝐾𝜔∠900
0
Therefore Magnitude of 𝐺(𝑗𝜔) in dB 𝐴 = 20 log10 𝐾𝜔
Phase angle of 𝐺(𝑗𝜔) in degree 𝜑 = 900
The Bode plot for 𝐺(𝑗𝜔) as shown in below figure
When 𝜔 = 0.1/ 𝐾, 𝐴 = 20 log10 𝐾(𝜔) = 20 log10 𝐾(0.1/𝐾) = −20𝑑𝐵
When 𝜔 = 1/𝐾, 𝐴 = 20 log10 𝐾𝜔 = 20 log10 1 = 0𝑑𝐵
When 𝜔 = 10/𝐾, 𝐴 = 20 log10 𝐾𝜔 = 20 log10 10 = 20𝑑𝐵
 4. First order factor in denominator
1
Let 𝐺(𝑠) = 1+𝑠𝑇
1
𝐺(𝑗𝜔) =
1 + 𝑗𝜔𝑇
1 1 1 1
𝐺(𝑗𝜔) = |𝐺(𝑗𝜔)|∠𝐺(𝑗𝜔) = ∠ 𝜔𝑇 = ∠
√1 + 𝜔 2 𝑇 2 𝑡𝑎𝑛−1 ( 1 ) √1 + 𝜔 2 𝑇 2 𝑡𝑎𝑛−1 (𝜔𝑇)
1
Therefore Magnitude of 𝐺(𝑗𝜔) in dB 𝐴 = 20 log10 √1+𝜔2 𝑇 2

Phase angle of 𝐺(𝑗𝜔) in degree 𝜑 = ∠ − 𝑡𝑎𝑛−1 𝜔𝑇

At low frequency 𝜔𝑇 ≪ 1
𝐴 = 20 log10 1 = 0𝑑𝐵
At high frequency 𝜔𝑇 ≫ 1
1
𝐴 = 20 log10 = −20log10 √𝜔 2 𝑇 2 = −20 log10 𝜔𝑇
√1 + 𝜔 2 𝑇 2
1
At 𝜔 = 𝑇, 𝐴 = −20 log10 1 = 0
10
At 𝜔 = , 𝐴 = −20 log10 10 = −20𝑑𝐵
𝑇

Therefore magnitude plot has approximated to two straight segments, one straight line at 0𝑑𝐵 for
the frequency range 0 < 𝜔 < 1/𝑇 and other is a straight line with slope −20𝑑𝐵/𝑑𝑒𝑐for the
 1
frequency range 𝑇 < 𝜔 < ∞. The two straight are asymptotes of the exact curve. The frequency

at which the two asymptotes meet is called corner frequency or break frequency. For the
1 1
factor 𝐺(𝑗𝜔) = 1+𝑗𝜔𝑇 , the corner frequency is 𝜔𝑐 = 𝑇

The phase plot is obtained by calculating the phase angle for various values of 𝜔
Phase angle of 𝐺(𝑗𝜔) in degree 𝜑 = ∠ − 𝑡𝑎𝑛−1 𝜔𝑇
1
At the corner frequency 𝜔 = 𝜔𝑐 = 𝑇, 𝜑 = −𝑡𝑎𝑛−1 1 = −450

At the frequency 𝜔 = 0, 𝜑 = −𝑡𝑎𝑛−1 0 = 00

At the frequency 𝜔 =→ ∞, 𝜑 = −𝑡𝑎𝑛−1 ∞ = −900
The Bode plot for 𝐺(𝑗𝜔) as shown in below figure

5. First order factor in numerator

Let 𝐺(𝑠) = 1 + 𝑠𝑇
𝐺(𝑗𝜔) = 1 + 𝑗𝜔𝑇
𝜔𝑇 1
𝐺(𝑗𝜔) = |𝐺(𝑗𝜔)|∠𝐺(𝑗𝜔) = √1 + 𝜔 2 𝑇 2 ∠𝑡𝑎𝑛−1 ( )= ∠𝑡𝑎𝑛−1 (𝜔𝑇)
1 2
√1 + 𝜔 𝑇 2
 1
Therefore Magnitude of 𝐺(𝑗𝜔) in dB 𝐴 = 20 log10 √1+𝜔2 𝑇 2

Phase angle of 𝐺(𝑗𝜔) in degree 𝜑 = 𝑡𝑎𝑛−1 𝜔𝑇

At low frequency 𝜔𝑇 ≪ 1
𝐴 = 20 log10 1 = 0𝑑𝐵
At high frequency 𝜔𝑇 ≫ 1

𝐴 = 20 log10 √1 + 𝜔 2 𝑇 2 = 20log10 √𝜔 2 𝑇 2 = 20 log10 𝜔𝑇

1
At 𝜔 = 𝑇, 𝐴 = 20 log10 1 = 0
10
At 𝜔 = , 𝐴 = 20 log10 10 = 20𝑑𝐵
𝑇

Therefore magnitude plot has approximated to two straight segments, one straight line at 0𝑑𝐵 for
the frequency range 0 < 𝜔 < 1/𝑇 and other is a straight line with slope 20𝑑𝐵/𝑑𝑒𝑐for the
1
frequency range 𝑇 < 𝜔 < ∞. The two straight are asymptotes of the exact curve. The frequency

at which the two asymptotes meet is called corner frequency or break frequency. For the
1
factor 𝐺(𝑗𝜔) = 1 + 𝑗𝜔𝑇 , the corner frequency is 𝜔𝑐 = 𝑇

The phase plot is obtained by calculating the phase angle for various values of 𝜔
Phase angle of 𝐺(𝑗𝜔) in degree 𝜑 = 𝑡𝑎𝑛−1 𝜔𝑇
1
At the corner frequency 𝜔 = 𝜔𝑐 = 𝑇, 𝜑 = 𝑡𝑎𝑛−1 1 = 450

At the frequency 𝜔 = 0, 𝜑 = 𝑡𝑎𝑛−1 0 = 00

At the frequency 𝜔 =→ ∞, 𝜑 = 𝑡𝑎𝑛−1 ∞ = 900
The Bode plot for 𝐺(𝑗𝜔) as shown in below figure
 6. Quadratic factor in the denominator
1
Let 𝐺(𝑠) = 2
𝑠2 +2𝜉𝜔𝑛 𝑠+𝜔𝑛

1
𝐺(𝑠) = 𝑠 𝑠
1 + 2𝜉 + (𝜔 )2
𝜔𝑛 𝑛

1 1
𝐺(𝑗𝜔) = 𝑗𝜔 𝑗𝜔
= 𝜔 𝑗𝜔
1 + 2𝜉 + (𝜔 )2 1 − (𝜔 )2 + 2𝜉
𝜔𝑛 𝑛 𝑛 𝜔𝑛
𝜔
1 2𝜉 𝜔
𝐺(𝑗𝜔) = |𝐺(𝑗𝜔)|∠𝐺(𝑗𝜔) = ∠−𝑡𝑎𝑛−1 ( 𝜔2
𝑛
)
𝜔2 2 𝜔2 1 − 𝜔2
√(1 − 2) + 4𝜉 2 𝜔2 𝑛
𝜔𝑛 𝑛

1
Therefore Magnitude of 𝐺(𝑗𝜔) in dB 𝐴 = 20 log10 =
𝜔2 𝜔2
√(1− 2 )2 +4𝜉 2 2
𝜔𝑛 𝜔𝑛

𝜔2 𝜔2
−20 log10 √(1 − 𝜔2 )2 + 4𝜉 2 𝜔2
𝑛 𝑛

𝜔4 𝜔2 2
𝜔2 𝜔2 2) +
𝜔4
𝐴 = −20 log10 √1 + − 2 + 4𝜉 = − 20 log 10 √ 1 − (2 − 4𝜉
𝜔𝑛4 𝜔𝑛2 𝜔𝑛2 𝜔𝑛2 𝜔𝑛4

At low frequency 𝜔 ≪ 𝜔𝑛 , the magnitude is

 𝜔2 2) +
𝜔4
𝐴 = −20 log10 √1 − (2 − 4𝜉 ≈ −20 log10 1 = 0
𝜔𝑛2 𝜔𝑛4

At high frequency 𝜔 ≫ 𝜔𝑛 , the magnitude is

𝜔2 2) +
𝜔4
𝐴 = −20 log10 √1 − (2 − 4𝜉
𝜔𝑛2 𝜔𝑛4

𝜔4 𝜔2 𝜔
≈ −20 log10 √ 4
= −20 log 10 2
= −40 log10
𝜔𝑛 𝜔𝑛 𝜔𝑛
𝜔
At 𝜔 = 𝜔𝑛 , 𝐴 = −40 log10 𝜔 = 0𝑑𝐵
𝑛

At 𝜔 = 10𝜔𝑛 , 𝐴 = 20 log10 10 = −40𝑑𝐵

Therefore magnitude plot has approximated to two straight segments, one straight line at 0𝑑𝐵 for
the frequency range 0 < 𝜔 < 𝜔𝑛 and other is a straight line with slope −40𝑑𝐵/𝑑𝑒𝑐 for the
frequency range 𝜔𝑛 < 𝜔 < ∞. The two straight are asymptotes of the exact curve. The
frequency at which the two asymptotes meet is called corner frequency or break frequency. For
the quadratic factor of 𝐺(𝑗𝜔) , the corner frequency is 𝜔𝑐 = 𝜔𝑛 . Thus, the magnitude plot is
independent of the damping ratio 𝜉.
The phase plot is obtained by calculating the phase angle for various values of 𝜔
𝜔
2𝜉
−1 𝜔𝑛
Phase angle of 𝐺(𝑗𝜔) in degree 𝜑 = −𝑡𝑎𝑛 ( 𝜔2
)
1− 2
𝜔𝑛

2𝜉
At frequency 𝜔 = 𝜔𝑛 , 𝜑 = −𝑡𝑎𝑛−1 ( 0 ) = −𝑡𝑎𝑛−1 (∞) = −900

At the frequency 𝜔 = 0, 𝜑 = 00
At the frequency 𝜔 =→ ∞, 𝜑 = −𝑡𝑎𝑛−1 0 = −00 = −1800
The Bode plot for 𝐺(𝑗𝜔) as shown in below figure.
 7. Quadratic factor in the numerator
Let 𝐺(𝑠) = 𝑠 2 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2
𝑠 𝑠
𝐺(𝑠) = 1 + 2𝜉 + ( )2
𝜔𝑛 𝜔𝑛
𝑗𝜔 𝑗𝜔 𝜔 𝑗𝜔
𝐺(𝑗𝜔) = 1 + 2𝜉 + ( )2 = 1 − ( )2 + 2𝜉
𝜔𝑛 𝜔𝑛 𝜔𝑛 𝜔𝑛
𝜔
𝜔2 2 𝜔2 2𝜉 𝜔
|𝐺(𝑗𝜔)|∠𝐺(𝑗𝜔) √ 2 −1 𝑛
𝐺(𝑗𝜔) = = (1 − 2 ) + 4𝜉 2 ∠−𝑡𝑎𝑛 ( 𝜔 2 )
𝜔𝑛 𝜔𝑛 1− 2
𝜔𝑛

Therefore magnitude plot has straight segments, one straight line at 0 𝑑𝐵 for the frequency range
0 < 𝜔 < 𝜔𝑛 and other is a straight line with slope 40𝑑𝐵/𝑑𝑒𝑐 for the frequency range 𝜔𝑛 <
𝜔 < ∞. The phase plot is obtained by calculating the phase angle for various values of 𝜔. The
phase angle varies from 0 to 1800 as 𝜔 varies from 0 to ∞. The corner frequency 𝜔 = 𝜔𝑛 the
phase angle is 900 . The bode plot as shown in below figure.
Problem: Sketch the Bode plot for the following transfer function and determine phase
10
cross-over frequency and gain cross-over frequency 𝐺(𝑠) = 𝑠(1+0.4𝑠)(1+0.1𝑠).
10
Solution: Given 𝐺(𝑠) = 𝑠(1+0.4𝑠)(1+0.1𝑠)

The sinusoidal transfer function is obtained by replacing 𝑠 by 𝑗𝜔, i.e

10
𝐺(𝑗𝜔) =
𝑗𝜔(1 + 0.4𝑗𝜔)(1 + 0.1𝑗𝜔)
1 1
Corner frequencies are 𝜔𝑐1 = 0.4 = 2.5𝑅𝑎𝑑/𝑠𝑒𝑐 and 𝜔𝑐2 = 0.1 = 10𝑅𝑎𝑑/𝑠𝑒𝑐

Now the ascending order corner frequencies and corresponding term are listed in the table

Term Corner frequencies in Slope dB/dec Change in slope

Rad/sec dB/dec
10 - -20
𝑗𝜔
1 2.5 -20 −20 − 20 = −40
1 + 0.4𝑗𝜔
1 10 -20 -20 − 40 = −60
1 + 0.1𝑗𝜔
The starting frequency of the Bode plot is 𝜔 = 0.1𝑅𝑎𝑑/𝑠𝑒𝑐
Therefore starting magnitude at 𝜔 = 0.1𝑅𝑎𝑑/𝑆𝑒𝑐 is
10 10
𝐴1 = 20 log10 | = 20 log10 = 40𝑑𝐵
𝜔 𝜔=0.1 0.1
With 40dB magnitude draw a straight line with slope −20𝑑𝐵/𝑑𝑒𝑐 from 𝜔 = 0.1𝑅𝑎𝑑/𝑆𝑒𝑐 to
at 𝜔 = 2.5𝑅𝑎𝑑/𝑆𝑒𝑐 on the semilog sheet.
From 𝜔 = 2.5𝑅𝑎𝑑/𝑆𝑒𝑐 to 𝜔 = 10𝑅𝑎𝑑/𝑆𝑒𝑐 draw a another straight with slope −40𝑑𝐵/𝑑𝑒𝑐.
From 𝜔 = 10𝑅𝑎𝑑/𝑆𝑒𝑐 to 𝜔 = ∞𝑅𝑎𝑑/𝑆𝑒𝑐 , draw a straight with slope−60𝑑𝐵/𝑑𝑒𝑐. This
completes the magnitude plot.
On the same graph sheet with proper scale of 1𝑢𝑛𝑖𝑡 = 200 , the phase plot is plotted for
phase angle ϕ versus frequency 0.1 ≤ 𝜔 ≤ ∞. The phase angle ϕ for various values of
frequency ω is calculated as follows. Let us consider the sinusoidal transfer function
10
𝐺(𝑗𝜔) =
𝑗𝜔(1 + 0.4𝑗𝜔)(1 + 0.1𝑗𝜔)
0.4𝜔 0.1𝜔
The phase angle 𝜙 = ∠ 𝐺(𝑗𝜔) = −900 − 𝑡𝑎𝑛−1 ( ) − 𝑡𝑎𝑛−1 ( )
1 1

𝜙 = −900 − 𝑡𝑎𝑛−1 0.4𝜔 − 𝑡𝑎𝑛−1 0.1𝜔

𝜔 𝑡𝑎𝑛−1 0.4𝜔 𝑡𝑎𝑛−1 0.1𝜔 𝜙 = ∠ 𝐺(𝑗𝜔)
Rad/Sec degree degree degree
0.1 2.29 0.57 -92.86
1 21.80 5.71 -117.5
2.5 45 14.0 -149
4 57.99 21.8 -169.79
10 75.96 45 -210.96
20 82.87 63.43 -235.3
Problem: Sketch the Bode plot for the following transfer function and determine phase
75(1+0.2𝑠)
margin and gain margin 𝐺(𝑠) = 𝑠 (𝑠2 +16𝑠+100).
75(1+0.2𝑠) 75(1+0.2𝑠) 75(1+0.2𝑠)
Solution: Given 𝐺(𝑠) = 𝑠 (𝑠2+16𝑠+100) = 𝑠2 16𝑠
= 𝑠 (0.01𝑠2+ 0.16𝑠+1)
𝑠 ( + +1)
100 100

𝑠 2 + 16𝑠 + 100 = 𝑠 2 + 2𝜉𝜔𝑛 𝑠 + 𝜔𝑛2

∴ 𝜔𝑛 = 10𝑅𝑎𝑑/𝑠𝑒𝑐
75(1 + 0.2𝑠) 75(1 + 0.2𝑠) 0.75(1 + 0.2𝑠)
𝐺(𝑠) = = 2 =
𝑠 (𝑠 2 + 16𝑠 + 100) 𝑠 × 100 ( 𝑠 + 16𝑠 + 1) 𝑠 (0.01𝑠 2 + 0.16𝑠 + 1)
100 100

The sinusoidal transfer function is obtained by replacing 𝑠 by 𝑗𝜔, i.e

0.75(1 + 0.2𝑗𝜔) 0.75(1 + 0.2𝑗𝜔)

𝐺(𝑗𝜔) = =
𝑗𝜔 (0.01(𝑗𝜔) + 0.16𝑗𝜔 + 1) 𝑗𝜔 (1 − 0.01𝜔 2 + 0.16𝑗𝜔)
2
1
Corner frequencies are 𝜔𝑐1 = 0.2 = 5𝑅𝑎𝑑/𝑠𝑒𝑐 and 𝜔𝑐2 = 𝜔𝑛 = 10𝑅𝑎𝑑/𝑠𝑒𝑐
Now the ascending order corner frequencies and corresponding term are listed in the table
 Term Corner frequencies in Slope dB/dec Change in slope
Rad/sec dB/dec
0.75 - -20
𝑗𝜔
1 + 0.2𝑗𝜔 5 20 −20 + 20 = 0
1 𝜔𝑐2 = 𝜔𝑛 =10 -20 0 − 40 = −40
1 − 0.01𝜔 2 + 0.16𝑗𝜔

The starting frequency of the Bode plot is 𝜔 = 0.1𝑅𝑎𝑑/𝑠𝑒𝑐

Therefore starting magnitude at 𝜔 = 0.1𝑅𝑎𝑑/𝑆𝑒𝑐 is
0.75 0.75
𝐴1 = 20 log10 | = 20 log10 = 17.71𝑑𝐵
𝜔 𝜔=0.1 0.1
With 17.7dB magnitude draw a straight line with slope −20𝑑𝐵/𝑑𝑒𝑐 from 𝜔 = 0.1𝑅𝑎𝑑/𝑆𝑒𝑐
to at 𝜔 = 5𝑅𝑎𝑑/𝑆𝑒𝑐 on the semilog sheet.
From 𝜔 = 5𝑅𝑎𝑑/𝑆𝑒𝑐 to 𝜔 = 10𝑅𝑎𝑑/𝑆𝑒𝑐 draw another straight with slope0𝑑𝐵/𝑑𝑒𝑐
From 𝜔 = 10𝑅𝑎𝑑/𝑆𝑒𝑐 to 𝜔 = ∞𝑅𝑎𝑑/𝑆𝑒𝑐 , draw a straight with slope−40𝑑𝐵/𝑑𝑒𝑐. This
completes the magnitude plot.
On the same graph sheet with proper scale of 1𝑢𝑛𝑖𝑡 = 200 , the phase plot is plotted for
phase angle ϕ versus frequency 0.1 ≤ 𝜔 ≤ ∞. The phase angle ϕ for various values of
frequency ω is calculated as follows.
0.75(1+0.2𝑗𝜔)
Let us consider the sinusoidal transfer function 𝐺(𝑗𝜔) = 𝑗𝜔 (1−0.01𝜔 2+0.16𝑗𝜔)
0.2𝜔 0.16𝜔
The phase angle 𝜙 = ∠ 𝐺(𝑗𝜔) = 𝑡𝑎𝑛−1 ( ) − 900 − 𝑡𝑎𝑛−1 (1−0.01𝜔2) 𝐹𝑜𝑟 𝜔 ≤ 𝜔𝑛
1

0.2𝜔 0.16𝜔
𝜙 = ∠ 𝐺(𝑗𝜔) = 𝑡𝑎𝑛−1 ( ) − 900 − 𝑡𝑎𝑛−1 ( + 1800 ) 𝐹𝑜𝑟 𝜔 > 𝜔𝑛
1 1 − 0.01𝜔 2

𝜔 𝑡𝑎𝑛−1 0.2𝜔 0.16𝜔 𝜙 = ∠ 𝐺(𝑗𝜔)

𝑡𝑎𝑛−1 ( )
Rad/Sec degree 1 − 0.01𝜔 2 degree
degree
0.1 1. 47 0.91 -89.44
1 11.3 9.2 -87.9
5 45 48.8 -91.8
10 63.4 90 -116.6
20 75.9 -46.8+180=133.2 -147.3
50 84.3 -18.4+180 -167.3
100 87.1 -9.2+180 -173.7