UNIT 3 CONTROL SYSTEMS

Structure
3.1 Introduction
Objectives
3.2 I3asic'~onceptsof Control
3.3 Open Loop and Feedback Control
3.3.1 Open Loop Contry1 Systems
3.3.2 Closed Loop (Feedback) Control Systems
3.3.3 Definitions of Terms
3.4 Manual and Automatic Control
3.5 Regulators and Servomechanisms
3.6 Examples of Autorr~ticControl Systems (ACS)
3.6.1 ACS at Home
3.6.2 Hydraulic and Pneumatic Control Systems
3.6.3 Mechanical and Thermal Control Systems
3.6.4 Electrical and Electromechanical Control Systems
3.6.5 Biological Control Systems
3.6.6 Other Control Systems
3.7 Modelling of Feedback Control Systems (FBCS)
3.7.1 Derivation of Equations for the System
3.7.2 Transfer Function
3.8 Analysis of Feedback Control Systems
3.8.1 Standard Inputs
3.8.2 Response of First Order Systems
3.8.3 Response of Second Order Systems
3.8.4 Time Response Specifications
3.8.5 Steady State Response Analysis
3.8.6 Response of Higher Order Systems
3.9 Stability of Control Systems
3.9.1 Definition of Stability
3.9.2 Conditions for Stability
3.9.3 Tests for Stability
3.9.4 Routh-Hurwitz Criterion for Stability
3.10 Basic Concepts of Control System Synthesis
3.10.1 Statement of Synthesis Problem
3.10.2 Series Compensation : PID Controllers
3.10.3 Synthesis of PID Controller for Chemical Processes
3.1 1 Control System Software
3.12 Summary
3.13 Further Reading
3.14 Answers to SAQs

3.1 INTRODUCTION
You have been introduced to the concepts of systems and system theory in Unit 1. Then you
studied how to model physical systems such as electrical, electro-mechanical, mechanical,
fluidic, thennal, and structural in Unit 2. You have learnt that, although there is diversity in
physical systems and the components employed therein, there is unity in their mathematical
modelling. Simple physical systems are represented by first and second order differential
equations. Consequently, the responses of these systems are basically of two types:
(a) Exponentially falling or rising hctions, such as A e "'or ~ e + ~ ' ;
(b) Sinusoidal oscillations with exponentially rising or falling amplitude like
~ e + ~ ' s $r
i nor Ce-'' sin at.
With this background, you are now ready to learn two more fundamenla1concepts of system
theory :
Systems and Control (i) control, and
(ii) feedback
These concepts are developed in this unit.
Objectives
The main objective of this unit is to introduce to you the fundamental concepts of control
and feedback. Accordingly, you learn what is meant by control, as well as open loop and
feedback control. Examples from several fields are given to illustrate these basic concepts.
You will then study mathematical modelling of control systems. This will enable you to
detennine the responses of control systems.
Another fundamental concept in feedback control is the possibility of instability, which
vitiates the very purpose of control. Hence, you will study the concept of stability as well as
stability criteria.
Lastly, you will be introduced to the basic concepts of design of a control system.
The computer has become so pervasive that it is but natural that computers be employed for
simulating control systems. You will get acquainted with simulation of control systems.
Thus, after reading this unit, for simple control systems you will be able to :
derive mathematical model,
analyse responses,
determine stability,
know design methodology, and
simulate them on a PC.

3.2 BASIC CONCEPTS OF CONTROL

The concept of control is intuitive. Control is essential for our safety and progress. For,
example, traffic on roads must be well regulated and controlled; else there would be
accidents, resulting in injuries, deaths and damage. Sluice gates in dams across rivers must
be controlled, especially in the rainy season; otherwise there would be floods downstream,
causing damage to crops, houses and even deaths. The rate of nuclear fission in a nuclear
power plant has to be well controlled; otherwise uncontrolled fission will lead to nuclear
bomb explosion, as in Chernobyl nuclear power station in Russia. In industry, the size of
nuts and bolts have to be precisely controlled; else they will not fit and the production will
go waste.
Examples such as these are sufficient to bring home the importance of the concept of
control. This unit is meant to explain the basic concept of control. Let us take a few
examples to illustrate this concept.
Example 3.1
In an automobile, the driver controls its speed by increasing or decreasing the
pressure on the accelerator pedal. We, therefore, conclude that pressure on
accelerator pedal controls the speed of the car. We illustrate this concept.as shown in
Figure 3.l(a).
Example 3.2
In a water tank, a control valve is provided to control the inflow of water into the
tank. By regulating the opening of the control valve, the level of water in the tank is
controlled. Hence we say, opening of control valve, controls the water level in the
tank. This relationship is shown in Figure 3.l(b).
Example 3.3
In the domestic fan, we know that the setting of the fan regulator controls its speed.
Figure 3.1 (c) illustrates this relationship.
From these examples, we observe that there exists cause-andeffect relationships between
variables like pressure, opening and setting on the one hand and speed, level and speed on
the other. Such cause-and-effect relationships constitute the basis of control. We assert that
the variable 'cause' (i.e. input) controls the variable 'effect' (i.e. output). We illustrate
this concept as shown in Figure 3.2. Note that Figure 3.2 is applicable to all types of control
systems such as those shown in Figure 3.1. It is independent of (a) the physical nature of the
 t

-
Control Systems
Accelerator
Pressure Automobile Speod
-
-r
Pedal I

(a) Automobile Speed Control System

--
L

Opening Control Valve 5 Water Tank -water Level

*

(b) W a t e r Level C o n t r o l S y s t e m

Setting I- Speed Regulator D Fan L, Fan Speed

2 J

(c) Fan Speed C o n t r o l S y s t e m

Figure 3.1 : Examptes of Control Systems

input and output variables, as well as (b) nature of elements used in the system. It
concentrates on the functional relationship between input and output variables. Hence
Figure 3.2 is called functional block diagram.

nManipulated n

Figure 3.2 : Functional Block Diagram of a Control System

At this point, we are in a position to define a control system. To recapitulate, we have

defined, in Unit 1, a system as a collection of interconnected elements so configured as to
achieve a certain purpose or aim. In a control system, our aim is to control its output. Hence
we define a control system as follows :
Definition 1
A control system is a system which is meant to control or regulate its outputs in
I' accordance with its input variables.
Furthermore, we observe from Figures 3.1 and 3.2, that a typical control system has two
essential components
(i) controller to which input signal is applied.
(ii) plant or process which receives signals, called/manipulated variables, from
controller and produces desired outputs.

SAQ 1
Identify the controller and the plantlprocess in each of the three systems shown in
Figure 3.1.
 3.3 OPEN LOOP AND FEEDBACK CONTROL
Having understood the basic concept of control, we now proceed to learn the next basic
concept : open loop and closed loop (feedback) control systems.
3.3.1 Open Loop Control System
Figure 3.2 shows a typical open loop control system. In this system, the controller is
connected in series with the plant/process. Input fit), is applied to the controller; the
controller generates a signal called manipulated variable, designated as u(t). This
manipulated variable acts upon the plantlprocess and causes it to produce the desired
output, fit).
In Figure 3.2, success of control over output, y(t), depends upon how well we know the
functional relation between input r(t) and output y(t), inasmuch as
fit) = F [r(t), tl ...(3.1)
The more exact our knowledge of this functional relationship, the finer will be the
regulation of output, y(t). Equation (3.1) is defined as the control law or calibration of the
system.
Another factor also affects the success of control, viz. how well the calibration hold good
during the operation of the systems. Why we bring in this practical point is that, during
opzration of the system, several phenomena such as the following occur :
(i) temperature of system components rises.
(ii) wear and tear of mechanical components take place.
(iii) there are fluctuations in AC mains voltage and frequency.
(ivj variations in ambient temperature, humidity and pressure occur.
All such phenomena disturb the calibration of the system; as a consequence, the regulation
of the output suffers.
To sum up, we observe that the success of the open loop control system of Figure 3.2
depends on two factors :
(i) exactness of control law or calibration, F [r(t), t]
(ii) presence of factors, such as mentioned above, which cause actual system output
to differ from the theoretical value calculated from equation (3.1).
In case the actual output differs from theoretical value, nothing can be done, because the
controller does not know what the actual output is.
3.3.2 Closed Loop (Feedbackj Control Systems
In order to remedy this situation, we must let the controller know what the actual output of
the system is. This requires that we measure the actual output of the system and feed it back
to the controller as shown in Figure 3.3. The element which measures the actual output is
-
.
Input
r(t)
- Controller Plant / Process
L
J
y(t)
I

Feedback
Signal
b(t)
Feedback Element - Output
Y (t)
A

Figure 3.3 : Closed Loop or Feedback Control System

named in Figure 3.3 as the 'feedback element'. The feedback signal b(t), produced is a
function of output, i.e.

The controller now compares the input signal,r(t) with feedback signal, b(t), and produces
an output
 where e(t) = r(t) - b(t) = error/deviation/discrepancy signal. ... (3.4) Control Systems

The system of Figure 3.3 is called Feedback Control System or Closed Loop Control
System.
An important characteristic'of the feedback control system is that it operates in such a way
as to cause the error signal, e(t), defined by equation (3.4),tend to zero. When, ideally,
e(t) = 0, perfect regulation of y(t) is achieved. Thus the feedback system is an error driven
system which causes the error 4 0.This aspect is brought out clearly in Figure 3.4.

Manipulated
Variable , p l a n t or Output
~ ( t ) - ' Process y(t) 1
-
Feed bac k
Signal Feedbac k
b(t) Element

Figure 3.4 : Block Diagram of Feedback Control System

3.3.3 Definitions of Terms

All the terms associated with Figure 3.4 have been introduced in Section 3.3.2. For sake of
clarity, we define below the terms associated with Figure 3.4 :
Reference Input Signal (IQ)) is the signal which is required to be applied to the system to
cause it produce the desired output y(t).
Feedback Signal (b(t))is the output of the feedback device. It is a function of the output
signal y(t).
Error Signal (e(t))is the difference between r(t) and, b(t) i.e. e(t) = r(t) - b(t).
Error Detector is the element to which r(t) and b(t) signals are applied, and which
produces the error signal e(t).
Manipulated Variable (u(t))isproduced by the controller, and is a function of e(t).
Controller is the element which, upon receipt of error signal e(t),produces manipulated
variable (u(t))which causes the plant/process to give desired output y(t).
Plant or Process is the subsystem which is to be controlled; it receives input signal u(t) and
produces output signal y(t).
Output Signal Cy(t))is the regulated or controlled output of the control system.
Note that the objective of the control system is to regulate its output y(t) in accordance with
reference input r(t).
Feedback Element measures system output signal y(t) and produces feedback signal b(t)
which is a function of y(t).
Feedback Signal (b(t))is the output of the feedback element. It is a function of y(t), such as
the whole or a fraction of y(t), or involve derivative or integral of y(t).
The feedback signal is compared with the reference input r(t) by the error detector to
generate the error signal e(t), which actuates the control system.
Having studied the basic concepts of open loop and closed loop (i.e. feedback) control, it is
the time now to apply these concepts to a practical example. Example 3.4, which follows,
serves this purpose. It is also meant to bring out the advantages and disadvantages of closed
loop control systems vis-a-vis open loop control systems.

SAQ 2
We use electric iron at home to iron our clothes. Cotton and linen clothes require a
high temperature, while polyester, terelyne and silk require ironing at much lower
temperature. With a view to catering to these needs, a temperature controlling
mechanism is incorporated into the iron.
Describe the automatic temperature control system incorporated into the electric iron,
and draw a functional block diagram for it. Is it an open loop or closed loop system?
Systems and Control Example 3.4
Motor Speed Control System
Figure 3.5 shows a simple speed control system for a DC motor. Herein the desired
speed is set by input potentiometer, whose output voltage VRcorresponds to desired
speed. The actual speed 'hr (in rpm) of the motor is measured by a tachometer (TG),
which is mounted on the motor shaft. The tachometer output voltage VBis given by
V B =H . N ... (3.5)
where H = tachometer sensitivity (in voltslrpm). The tachometer voltage VB is
compared with reference voltage VR to generate error signal
l?= V R - V B = V R - H .N ...(3.6)
at the input of the differential amplifier. The amplifier output is in the form of current
If given by
I F = K A .E Z K - ~ ( V ~ - H . N ) ... (3.7)
where KA = amplifier gain (in Amps/volt). If flows through the field winding of the
DC motor, which is separately excited. The armature current ZA is kept constant by
connecting it to a separate DC source. The output torque of the motor.
TM=KT.If=K&A(VR-H.N) ...(3.8)
where KT = torque constant (in N- &Amp)
Ref. Potentiometer = constant
+ -
Differential
E
Ampli ter
-
%
N
V,=H.N.

(a) Schematic Diagram

Differential Amplifier D C Motor +Load

" '
K~ '-, KT
rn

A
.TM
I

--, "B
w

J
N -

l a c home ter
V, = H.N -
H IN

(b) Functional Block Diagram

Figure 3.5 : Motor Speed Control System

The motor torque T, is utilized to (i) overcome the load torque TL(e.g. cutting force of a
machine tool cutting a metal surface in a lathe), and (ii) to move the load having moment of
inertia J (~g-m2) and coefficient of friction B (N-drpm). Under steady state conditions of
operation, which we are considering now, there is no acceleration; hence no torque is
required to accelerate the load. However, torque T, is required to overcome friction. Now
TF= B . N. Hence we have the torque balance equation
 or KT.KA(VR-H.N)=B.N+TL ...(3.9) Control Systems

combining these equations and simplifying, we get

KA . KT
where G = -- - Forward gain and H = feedback gain.
i B
Speed Control System Operation
I
~o'~oad Etfect o t €ff&ct of
Load TI Parameter Change

(ii) (iii) (iv) in^ in H

I I
OL
r-l CL CL
I
(v) (vil (vii)
Figure 3.6 : Conditions of Operation of Speed Control System of Example 3.4

We now study the operation of this system under various conditions as shown in Figure 3.6.
(i) No Load, Open Loop (i.e. TL= 0, H = 0. Switch S in position A in Figure 3.4(a))
From equation (3.10), we get
N=G.VR ... (3.1 la)
Hence, reference voltage VROrequired to obtain desired speed No is given by

(ii) No Load, Closed Loop (TL= 0, H # 0. Switch S in position B)

Hence, reference VRcrequired to run the system at desired speed No is given by

Note that VBO= H . NO= voltage output of tachometer at desired speed No .

(iii) Open Loop, Effect of Load TL(TL# 0, H = 0, VR= VRO. Switch S in position A)

Thus, the drop in speed resulting from application of load is given by

TL
(An),, = (rpm) ... (3.14)
Systems and Control (iv) Closed Loop, Effect of Load (H f 0, TLf 0, VR= VRC . Switch Sin position B)

Hence, drop in speed on application of load torque TLis given by

dN
DG . From
Effects of changes in C with TL= 0 can be studied by determining -
equation (3.10) for N, we get

The sensitivity to changes in G is defined as

(v) Open Loop, Effect of Changes in G

(H = 0, VR = VRo,TL= 0. Switch S in position A)

(s:), = 1 (i.e. 100% )

(vi) Closed Loop, Effect of Changes in G
(H a, VR = VRC. TL= 0. Switch S in position B)

(vii) Closed Loop, Effect of Changes in H

(H # 0, VR= VRc,TL= 0. Switch Sin position B)

dN -
- G2 ... (3.21)
~H--(I+GH)~
Sensitivity to changes in His

i.e. almost 100%.

These results are summarized for convenience in Table 3.1.
Table 3.1 :Performance of Open Loop and Closed Loop Systems

Condition of System
Condition of Operation
Open Loop Closed Loop
No Load (TL= 0) No VRc=VRo+H.N,
R
'O =

Effect of Load T~ (Anlo

(A40 = jj- (An) --
C-~+GH

Sensitivity to Changes
in
0) G (a,=1 (@CL =
(sC)O

(ii) H GH
( G c L = z = l
 To make these matters concrete, Control Systems

let KA= 100 AmpNolt, KT= 1.5 N-dA,

B = 0.15 N-drpm, H = 0.01 Vfrpm
Let no load desired speed be 1200 rpm, and load torque TL= 30 N-m.

Then G=-- KA. KT- 1000 (rpm/V)

B
1
1 and GH = (1000) (0.01) = 10
Substituting these values in Table 3.1, we get resulp given in Table 3.2.
I
Table 3.2 : Numerical Results for Performance of Open and Closed Loop Systems
Operating Condition Open Loop System Closed Loop System
NOLoad vRO=1.2v VRc= 1.2+ 1 2 = 13.2V
\

Effect of Load (An), = 200 rpm ------

(An) - 200 -18.2rpm
- (1 + lo)
Sensitivity to Changes in
(a) G (s;), = 1 = 100%

(b) H (s:)~=---- GH 10 - 91%

i + ~ ~ - i i
This example has been worked out in great detail with a view to bringing out the advantages
of closed loop systems vis-a-vis open loop systems. These are :
1
(i) Drop in speed on application of load is reduced by the factor
(1 + GH) '
(ii) Sensitivity to changes in forward gain G is also reduced by the same factor

Disadvantages of closed loop system are :

(i) Considerably large reference input is required to obtain given desired output
since reference input must overcome feedback voltage VB = H . N.
(ii) A feedback &vice, such as a tachometer, is required to be connected to the
motor.
(iii) Factors (i) and (ii) make closed loop systems more bulky, expensive and power
hoggers than open loop systems.
(iv) As we shall discover later, closed loop systems are liable to become unstable in
operation.
-
3.4 MANUAL AND AUTOMATIC CONTROL
Both open loop and feedback control systems could be either manually operated or be
automatic. We illustrate this point with the help of examples.
Example 3.5
Filling a Tank
Suppose you are asked by your mother to fill up a stainless steel or plastic tank in
your kitchen for storing drinking water. What do you do? You connect a
rubberlplastic hose pipe to the kitchen tap and open the tap as shown in Figure 3.7(a).
The tank begins to fill up. When the tank gets full, you promptly close the water tap.
You cannot afford to let the tank overflow, mess up your kitchen and earn the wrath
of your mother.
Have you realised that you have just now operated a closed loop water level control system?
You were observing the actual water level, comparing it with desired level and regulating
the tap opening. You were thus acting as both feedback element and controller of the system
as shown in Figure 3.8 (a). Since you operated the system, it is manual.
Systems and Control Plastic Hose

Inlet
Pipe

Water Tank
Iq) Manual Control (b) Automatic Ccntrol

Figure 3.7 : Water Level Control System

Desired
Lev o l
+
Your Hand
and Tap
- Water
Tank
Actual
-
(a) Water Level Control System ( M a n u a l )

Desired Actual

+-
Level , Float Mechanism
Inlet Valve

Float.
Mechan~sm

(b) Water Level Control System (Automatic)

Figure 3.8 :Block Diagrams of Water Level Control Systems of Figure 3.7

Example 3.6
Suppose now that you have to fill up the overhead water tank of your housefbuilding.
It is obvious that you cannot adopt, the same method now as you used for filling your
kitchen tank. The main reason is that you cannot see the level of water in the
overhead tank. That is why a simple float mechanism is used in which the float rises
as the overhead tank gets filled. When the water has reached a predetermined level,
the float mechanism closes the inlet valve. Block diagram for this system is shown in
Figure 3.8(b). Note that the float mechanism acts both as feedback element and
controller. Since there is no need for human intervention, this system is automatic.
Example 3.7
Vehicle Driving
Your aim, when driving a vehicle (car, scooter, motor cycle, etc.) is to reach your
destination safely and as quickly as condition of the traffic and the condition of road
permit. The essence of driving is to steer your vehicle clear of obstacles like buses,
tfucks, cars, scooters, pedestrians, cows, buffalows, donkeys and so on (this list, like
the proverbial tail of Lord Hanuman, is unending). As if this is not enough, there are
innumerable pot holes on the road. In order to steer clear of such obstacles, your
vehicle has a steering mechanism consisting of steering wheel or handle, and a
number of mechanical'linkages. As you do turn the steering wheel, the vehicle also
turns in the desired direction to the extent desired. This is a feedback control system
in which the driver acts a$ both feedback element and controller. Figure 3.9 shows
 this manual control system.Such a system is known, in modem terminology, as C ~ n i mSystems
l
man-in-the-loopsystem

Desired
Direction
+ - Driveis hands
Steorrng
Mechanism
Vehicle
Actual

3.9 : Vehicle Steering System

Example 3.8
Piloting an Aircraft
Every aircraft (such as Indian Airlines Airbus A-302 or Jet Airways Boeing 737-400)
I
must have a pilot to fly the passengers to their destination. The functions of the pilot
are :
(i) to manage the critical take off and landing operations,
(ii) to set the correct flight course to destination, and
(iii) to ensure that the aircraft remains on course inspite of disturbances such as
rough weather, rains and storms, pockets of high or low air pressure, etc.
The pilot performs the first two tasks manually. The third task, which is a long d r a m
out and tedious task, is now-a-days performed by an autopilot. This is an excellent
example of a combined manual and automatic control system. You could call ~tthe
symbiosis of manual and automatic control.

SAQ 3
Cite examples of (i) manual, (ii) automatic, and (iii) combined manual-auto control
systems. (At least one example of each type is required.)

SAQ 4
Consider the domestic temperature contzolled electric iron. It is an example of a
- temperature control system. Draw a block diagram for it similar to Figure 3.4.
Identify the reference input, error and output signals, as well as error detector,
controller, and plant.

SAQ 5
Classify the following as open loops or feedback control systems :
(i) concrete mixer
(ii) domestic refrigerator
(iii) multi-speed domestic mixer
(iv) dumper as a land levelling system
(v) elevator or 'lift' in a multi-storeyed building.
In each case, identify the purpose of the control system and draw block diagrams,
similar to Figure 3.2 or Figure 3.4, as the case may be.

81
Systems and Control
3.5 REGULATORS AND SERVOMECHANISMS
Having understood control systems : open loop and closed loop, you are now ready for a
frner study of these systems. Towards this end, let us go back to Example 3.5 and
Example 3.7. Both of them are closed loop control systems. However, there exists a subtle
difference between them. The aim of this Section is to explain this difference.
In Figure 3.7, the reference input, viz. desired water level, is constant with respect to time;
consequently the system output is also a constant.
On the other hand, in Figure 3.9, the reference input, viz. the desired direction, goes on
changing with time, depending on traffic and road conditions. As a result, the output
variable also changes correspondingly.
Thus the purpose of Figure 3.7 is to regulate the output in accordance with a fixed reference
input. However, in Figure 3.9, the purpose of the system is to cause the output to track or
follow the time variations of reference input.
To appreciate this difference further, consider the following example.
Example 3.9 -
Anti-Aircraft Gun
Figure 3.10 illustrates the operation of this system which is meant for shooting down
enemy war-planes which have invaded your territory. Here the anti-aircraft (AA) gun
is first trained on the enemy aircraft. As the aircraft is flying, its position is changing
with respect to time. Furthermore, the pilot of the enemy plane is bound to take
diversionary measures to avoid getting hit by shells fired from your AA gun. Hence,
you have to not only track the movements of the efiemy plane but also estimate its
future position. Then you fire shells in the direction of that future position so that
they have the maximum probability of hitting the target plane. This is a game of 'one
upmanship' between you, the gunner, and the pilot of the target plane. Obviously the
loser perishes.
Alrptane Projected position of
- - -- ~ i r p l a n ewhen the
1 \ shell arrives
/ \ /
/
/
/
/
/
/ Lead
1 angle

~ c t u a t i n gsignal

Figem 3.10 : Anti-Aircraft Radar Tracking Control System

From examples such as these you will appreciate that there are two basic types of control
systems :
(1) Systems which regulate their output in accordance with fixed ref. input signal. Such
systems are called Regulators.
 (2) Systems in which the ref. input signal goes on changing with reference to time; the Control Systems
system causes the output to track the variation, with time, of the reference input. Such
systems are called Servomechanismsor Servo Systems or Tracking Systems.

3.6 EXAMPLES OF AUTOMATIC CONTROL SYSTEMS

(ACS)
In this Section, you will study examples of automatic control systems taken from various
fields. This will enable you to appreciate that the principle of automatic control are universal
and are applicable to all sorts of applications under the sun; may be even beyond the sun.
3.6.1 ACS at Home
I You have already come across several ACSs in our homes:
(i) speed regulator for the domestic fan (Figure 3.1 (c)).

r (11) water level control system for the overhead water tank (Figure 3.7 (b)).
(ili) temperature controlled electric iron (SAQ 2).
(iv) domestic refngerator (SAQ 3 (ii)).
(v) multispeed mixer (SAQ 3 (ui)).
Such examples are enough to convince you that appl~cationsof ACSs are quite common at
home. Notice that many of these applications are open loop control system (OLCSs),
because
(a) OLCSs are the simplest and least expensive ACSs
1 (b) the limited regulation of output they offer is quite sufficient for our needs.
'
For example, there are two speed settings available in the domestic mixer : low speed
(300 rprn) and high speed (600 rpm). We are not particularly hothered whether. at high
! speed setting, the mixer actually runs at 600 rpm or 540 rpm. The ~ n l y difference it makes
is that instead of running the mixer for 30 sec, we may have to run it for 35 sec which we
I may not even notice. Hence, the simple open loop ACS IS quite sufficient to meet our needs.
3.6.2 Hydraulic and Pneumatic Control Systems
Level and flow rate control systems are the two basic types of hydraulic control systems.
We, however, describe below the hydraulic braking system employed in motor cars. This
system is shown in Figure 3.1 1.

Broke pedal
\

a;m shoe.

Cylinder
waster CJ linda
\

V
0Lm x,eas Drum Shon
REAR BRAKES FRONT BRAKES

Figure 3.11 : Hydraulic Break System

Systems and Control When the car driver presses the brake paddle, it actuates the piston of the master cylinder.
Brake 011, under pressure, is thereby forced into wheel cylinders, and actuates pistons in the
wheel cylinders They. in turn, press the corresponding brake drums hard against the wheels
and thereby brake the car.
TOa first order of approximatton, there is a linear relationship between the force with which
the car drtver presses the brake paddle and the force with which the brake drums press
against thc wheels of the car. That is why the car driver can slow the car down to the desired
extent by varying the force exerted on the brake paddle.

3.6.3 Mechanical and Thermal Control Systems

Heat exchangers are used tn process industt~esboth for preheating fluids or for
condensinglcooling flutds. Varlous types of heat exchangers are m use such as
sheli-and-tube exchangers, double pipe exchangers, condensers, tubular furnaces, etc. as
shown m F~gure3.12

Stearn

(b) Doubla Pipe

Figure 3.12 : Heat Exchanger

Let us consider the shell and tube type heat exchanger of Figure 3.12 (a). Herem the water
to be heated flows through the pipe embedded m a steam jacket or shell. The steam
corldenses on the surface of the tubes; m this process, heat flows from the hot steam into the
colder water in the tube. 'Ihe mathematical model of this heat exchange process is quite
complex. However, a reasonable simple model may be derived as shown in Figure 3 13.

Ther mo ma t ~ r

Rgure 3.13 : Mathematical Model of Shell- and-Tube Heat Exchanger

Control of outlet water temperature is exercised by (i) controlling the rate of flow of water
through the tube, or (ii) controlling the steam flow. The former method is employed in the
application described above.

3.6.4 Electrical and Electromechanical Control Systems

The electric power we receive from the mains is nominally 230V, 50 Hz. As per Indian
Electricity Act, electricity boards and electric supply companies, such as Tata Electric
Power Supply Co., are expected to mamtatn the voltage at 230V k15% and supply
+
frequency at 50 Hz 2Hz. For this purpose, they have to install (a) automattc voltage
regulators (AVRs) and (b) load frequency controllers (LFCs) in their generating stations as
well as receiving stations. However, due to the gross over loading, the mains voltage is
known to have gone down from the nominal 230V to a mirumurn of 165V. Similarly, the
mains frequency goes down to 46 - 47 Hz On the other hand, during light load period
(about 2 AM), the mains voltage and frequency may rise upto 270V and 52.5 Hz.
 Control Systems
Such wide fluctuations are dangerous to the electrical equipment connected to them.,In
order to protect sensitive electrical equipment, consumers have to use voltage stabilizers or
regulators. They maintain the load voltage at 230V f 2% for AC mains voltage fluctuation
of 230~+:~~$, i.e. 180V-270V.
A typical AC voltage stabilizer is shown in Figure 3.14. This is the so called servo voltage
stabilizer. In this a servo system is used for regulating the load voltage. This system works
as follows: The actual load voltage is sensed by voltage feedback element. It is compared
with the reference voltage. The error voltage actuates the servo motor which adjusts the
position of the wiper on the variac. The variac output is fed to the p r i m 1 winding of the
buck-boost transformer. The voltage induced in the secondary winding either adds to or
subtracts from AC mains voltage such that load voltage remains constant.

Figure 3.14 : Servo Voltage Stabilizer

3.6.5 Biological Control Systems

A number of control systems are to be found in the plant and animal kingdom. The human
being is replete with such control systems. Simple actions we perform, like picking up an
object and placing it in some other place, standing erect, walking, sitting down on a chair
etc. would not be possible without in-built automatic control systems. These are examples
of voluntary controls. In addition there are a number of involuntary controls which are in
charge of vital body functions such as digestive system for digesting the food we eat,
circulatory systems for ensuring circulation of blood throughout our body, respiratory
system for breathing, etc. Suffice it to say that, without such sophisticated, delicately
balanced and well coordinated control systems, we would not be able to stay alive.
3.6.6 Other Control Systems
Cited below are examples of control systems encountered in other fields :
(a) Pollution Control Systems

pollutants coming out of chimneys of petrochemical, cement, fertilizer,

metallurgical, etc. plants.
(ii) Sewage treatment plants for treating sewage from cities.
(iii) Emission control systems in petrol and diesel vehicles for controlling the
emission of toxic chemicals.
(b) Economic Control Systems for regulating the economy of the country, region or firm
such as accounting, auditing, taxation, imports-exports, etc.
(c) Political Control Systems such as parliament and central govt; legislative assembly
and state Govt; local self governments; etc.
(d) Judidal, educational, religions, etc. systems.
Systems and Control --
SA- xO (i-
Cite at least one example each of hydraulic, mechanical, thermal and
electro-mechanical control systems. Explain their operation with the help of block
diagrams.

3.7 MODELLING OF FEEDBACK CONTROL SYSTEMS

Having studied the fundamental conceDts and tv~icala~~lications.vou are now readv to
embark upon analysis and design of F ~ C SThe
. krst st& in this diiection is to derivk
mathematical model for the FBCS under study. In Unit 2, you have already studied how to
derive mathematical and simulation models of physical systems. That study will help you
derive mathematical model for FBCS, because FBCS are also physical systems.
In this section you will study mathematical modelling of simple FBCS. This will lead you to
derivation of transfer function models of FBCS.
Example 3.9
Dynamic Analysis of Speed Control System
In Example 3.4, we studied a simple motor speed control system in its steady state
operation. We continue with the same example and study its dynamic performance
i.e. performance when motor speed, motor torque, motor field current and feedback
signal are varying with time. Such dynamic conditions arise when
(a) with motor at rest, we suddenly apply reference input VRat t = 0.
(b) with motor running at a speed No, load torque TLis suddenly applied at
time t = 0.
3.7.1 Derivation of Equations for the System
Equations (3.5) - (3.8) are applicable to dynamic operations of the system also. As a result,
we have
V,(t) = H . N(t) ... (3.5)
E(t) = Vp(t) - Vn(t)= Vp(t) - H . N(t) (7 6)

and
The motor torque is utilized for
(i) counter acting load torque TL(t) J
(ii) accelerating the motor and the load, the accelerating torque being

(iii) overcoming friction

... TF(t) = B . N(t)
.JAl/.\
c"v(ll
Thus T,(t) = J + B . N(t) + Tl.(t) (
dt ... (3.24)
I
Substituting for T,(t) fromequation (3.8) and simplifying, we get I

Note that dynamic operation of the system is governed by first order differential
equation (3.25). This system is thus a first order system.
3.7.2 Transfer Function Control Systems

Recollect that you learnt Laplace transform, in Section 2.6.3 of Unit 2, as an aid to the
solution of differential equations. A very important concept associated with Laplace
transforms is transfer function. We define transfer function and illustrate its application in
analysis of dynamic systems.
Definition of Transfer Function
The transfer function may be defined for a dynamic system or any element thereof. It is
defined as
Y(s) Laplace transform of Output ... (3.26)
G(s) = -
X(s) Laplace transform of Input
It is defined under two conditions :
(i) the system or element is at rest; thus all initial conditions are zero, i.e.
y(o+) = j,(O+) = .... = 0.
(ii) no loading of the system or element.
Derivation of Transfer Function (TF)
The procedure for derivation of TF is as follows
(i) Derive equations governing dynamic performance of the element or system.
(ii) Take Laplace transform of the dynamic equation, assuming zero initial
conditions.
(iii) Take the ratio of Laplace transform of the output to Laplace transform of the
input. This ratio is the desired TF.
Example 3.10
Derive TFs of the elements of the speed control system
- of Example 3.9. Hence derive
transfer function of the entire system.
Solution

Element Dynamic Equation Transfer Function 1

(Tachometer)

Differential Amplifier Equation (3.7)

Motor and (motor + load)

impedance

I assuming no load i.e.

TL(t)= 0 1
We substitute these transfer functions in the block diagram of Figure 3.5 to obtain the
block diagram containing transfer functions, as shown in Figure 3.15.

Figure 3.15 : Block Diagram of Speed Control System with Transfer Function

Block Diagram Reduction

Block diagram, such as shown in Figure 3.15, can be reduced to its simplest form wlth the
help of rules given in Table 3.3.
Control Systems

Table 3.3 :Block Diagram Simplification Rules

Standard Transfer Functions for a Closed Loop Control System

Consider the reduced block of the closed loop system shown in Figure 3.16. We define the
Following terms and transfer functions :
(1) Forward Path is a path going frorn a given input to a given output of a closed
loop system.

Forward P o t h ,

- Feed back Poth

Figure 3.16 : Standard Block Diagram of a Closed Loop System

(2) Forward Transfer Function (or Open Loop Transfer Function)

(3) Feedback Path comes back from a given output y(t) to error detector.
(4) Feedback Transfer Function

(5) Loop is a closed path going from e(t) to y(t] and back from y(t) to b(t).
(6) Loop Transfer Function

(7) Closed Loop Transfer Function

(8) Characteristic Function

F(s) = 1 + G(s) H(s)
 C d r o l systems
Example 3.11
Derive the open loop and closed loop transfer functions of the speed control system
of Example 3.9.
Solution
Note, from Figure 3.15, that there are two forward paths in this system as shown in
Figure 3.17. The forward, feedback and closed loop transfer functions associated
with these paths are given below.
(a) Forward Path No. 1 from VRto N

Forward TF. m
G1(s)= E(s)

Vds)
Feedback TF, HI(s) = --
N(s) -

Closed Loop TF, TI(s) = m

R( )

1
Error TF, !@l=
VR(S) [1 + G1(s)Hl(s:~1

J J
where , = - and,'T
2 =
B B + KAK+
(b) Forward Path No. 2 from - TLto Y

Forward TF, G2(s) = 3?l

TE($)

1 Feedback TF, H2(s)= KAK+

Clqed Loop TF, T2 ( s ) = Y(s)

(- T L ( ~ ) )

Numerical Values
KA = 100 AmpNolt B = 0.15 N-mlrpm
KT = 1.5 N-dAmp J = 0.03 kg-m2
H = 0.01 VIrpm TL= 30 N-m
Systems and Control

(a) For ward P a t h e l from VR to N (b) Forward Path W 2 tmrn -I1. to N

Inn.

T i m (~sec )

Time ( suc )
Figure 3.17 : Simplified Block Diagram of Speed Control System of Example 3.11
 (100)(1.5j Control System
TI ( s ) = -
0.03s + 0.15 + (100)(1.5)(0.01)
- 150 - 5000 - 90.91
0.03s + 1.65 - s + 55 - 1 + 0.018s

1 - - - 6.67
- -33.33
*Iso
G2(s)= 0.03s + 0.15 - s + 5 1 + 0.2s
H2(s)= (100)(1.5)(0.01)= 1.5
0.61
T2(s) = ---
1 -
-
1 - --33.33 -
- -
0.03s+0.15+1.5 0.03s+1.65 s+55 1+0.018s

3.8 ANALYSIS OF FEEDBACK CONTROL SYSTEMS

Having derived the transfer functions of a feedback control system, you are now ready to
analyse the behaviour of these systems. Following types of analyses are common in control
system, as illustrated in Figure 3.18. Now you will study time response analysis under
(a) open loop and (b) closed loop conditions.
System Analysis

Locus Contour
Plots Plots

Figure 3.18 :Bird's Eyeview of System Analysis

3.8.1 Standard Inputs

In Unit 2, you have already studied the standard inputs applied for studying time response of
dynamic systems. They are shown in Figure 3.19 below for your ready reference. You will
study response of the control systems, when the above inputs are applied. Numerical
examples would be the best way to illustrate the procedure.

Figure 3.19 :Standard Inputs

Systems and Control
 Control Systems

200

(a) Step Response (b) Ramp Respons

Figure 3.20 : Open Loop Responses of Speed Control System

(a) Step Response (b) Ramp Response

Figure 3.21 : Closed Loop Responses of Speed Control System

In(t)
0- t lSec)
- 0.61 -
Closed Loop

Figure 3.22 :Load Responses of Open Loop and Closed Loop System

3.8.3 Response of Second Order Systems

Consider the elementary servo system shown in Figure 3.23 (a). In this system, the
reference input signal, OXt), varies with time. The objective of this closed loop system is to
cause the output position O,(t), track the time variations of O,.(t).
In this system, the reference input position, 0,.(t), is converted by the input potentiometer
into reference input voltage VR(t).The output position, 0,(t), is similarly measured by output
potentiometer and converted to feedback voltage, V,(t). These two signals are applied to a
differential servo amplifier. The servoamplifier output is the current, Iht), flowing through
the field winding of the field controlled DC servo motor. The armature current, IA, is kept
constant so that the motor torque, Tdt), is proportional to field current, Ibt). The
servomotor positions the load through a gear train which reduces the motor speed by the
factor 'a' < 1. The output potentiometer is mounted on the output shaft.
 Systems d Input Pot Constant
J + r n

Shaft Motor -
I I mhTr
(a) Schematic Diagram

E If

Differential DC Servo Load Gear Train

Servo Amplifier

L m
Output Pot
*Y

(b) Functional Block Diagram

(c) Simplified Block Diaararn

Figure 3.23 : Elementary Servo System

Derivation of Equations
The input and output potentiometers are circular, precision transducers whose output
voltage is directly proportional to angular position of its shaft. Thus,

v,<t> = Kp . 0,(t)
where K, = potentiometer sensitivity (Vlrad)
The differential servoamplifier consists of (i) operational amplifier and (ii) power amplifier.
Its output
q t ) = KA (VR - V,) ... (3.39)
where KA = amplifier gain (A I V)
The servomotor used here is the field controlled DC motor employed in the speed control
system studied before. Hence motor torque is, as before, given by

Tdt) = KT . 9 (t) ... (3.40)

I
where K T = motor torque constant (N-m I A)
The motor torque is utilized for
(a) driving the load to &sired position, through the gear train, so that Oy(t)= 0Xt);
(b) overcoming constant load torque, TL.
 Control Systems
A detailed discussion of the gear train is given in Section 4.4 of the Unit 4. The equations of
motion for the motor + gear train + load combination are :

where J = J, + a 2 ~=, equivalent moment of inertia

B = B, + a2BL = equivalent damping coefficient, both referred to motor shaft,
and om= motor speed (radsec)
q ( t ) = am, = load speed (radsec)
t
and B,(r) = I
0
oy(t)dt = load position (rad).

Derivation of Transfer Functions

It can be done in two ways :
(i) from system equations, by taking Laplace transforms and simplifying, and
(ii) by block diagram simplification.
We take the latter approach. The simplified block diagram of the system is shown in
Figure 3.23(c). From this block diagram, we get the following system transfer functions :
Forward 'IT

K ~ K A K aT
where K= = system gain (sec-')
J
B
P = - = open loop pole
J

r, = = motor + load time constant (sec)

K
K = - = static gain (sec ')
P
Closed Loop TF

where 9
on= - = undamped natural frequency
J

and amping ratio

C = x =
Systems and Control Load TF

Error TF

Note that the elementary servo system is a second order system. You have already studied
such second order systems in Unit 2. Recall that the second order system could have the
following types of responses, depending on the value of damping ratio (zeta).<
Damping Ratio Nature of Roots Nature of Response

O<(<l Complex conjugate Underdamped, oscillatory

Real and Equal Critically damped non-oscillatory

I <>l -
Real and Distinct Overdamped I
Under damped oscillatory response is very common. We shall therefore consider this case
in detail. Table 3.5 gives responses of the second order servo system.
Table 3.5 :Responses of the Second Order Servo System

Step Response Ramp Response Parabolic Response Load Response

u(t) t-~(t) 0

TL(~) 0 0 - TL . U(t)
exs) -
1 -
1 1
- 0
S s2 s3
TL(s) 0 0 0 2
S

ey(s) 4 4 4 - a2rL
-.a,"
s(s2 + 2<mns+ a;) s2(s2+ 2561,s + o;) s3(s2+ 2<wns + 0;) K
s(s2 + 2<o,p + 0;)
ey(t) t - - 25
+ - e-w
1- T e - m
1-
1
c2 an ad K
sin(@d+ $1 sh(wd+ v)

eys 1 -a 2 ~ L
K
ess = o r - Qyss o %=B t-
1-4c2 '
-3
a
- 2~L
on K On 4 K
 B Control Systems
where a = Ton= -= damping factor
2J
o,= w, -= damped frequency

These responses are shown in Figure 3.24. Note that the responses are governed by damping
ratio, 5, and undamped natural frequzncy, on.

Figure 3.24 : Responses of Second Order Systems

3.8.4 Time Response Specifications

The response to step input, shown in Figure 3.25 is used for defining time domain
specificationsfor control system. These specificationsare defined in Table 3.6.
Table 3.6 :Time Domain Specificationsfor Second Order System
Y(s) /r(s) = o: / (s2+ 2504,s + 4)

Specification Definition 1 Expression

Peak over-shoot (PO)
Peak Time, tp Time at which PO occurs
Rise nme, t,, Time for y(t) to rise from 0.1 Y,,to 0.9
yss
Rise Time, tn Time at which Y(t) = Y,, for the first
time
Delay Time, td Time at which Y(t) = 0.5YS, (1 + 0.70 /on
Settling Time, t, Time at and after which Y(t) 5 x% of Y,, ts5 = 3 / < o n for x = f 5
,t = 4 / < o n for x = + 2
ts, = 5 / < o n for x = f l
Equivalent Time Constant t, = [on(-- 2p)V21-
No. of Oscillations, N No. of oscillations of Y(t) about Y,, N = 2-/1tc
before settling down to f 2% of Yss
Steady state errors, e,, eps= 0 (step input)
e,, = 25 /.on (ramp input)
e, = a2r,/ K (Load torque. Td
 Figure 3.25 : Typical Underdamped Response to a Step Function

Damping ratio, 5
Figure 3.26 : Variation of PO and with 6

SAQ 7
Derive expressions for (i) peak time, fp, (ii) peak overshoot, PO, and (iii) rise time,
m.

Table 3.7 :Transient Response Characteristics of Second Order Systems

for Several Values of Damping Ratio, 5

Damphg Ratio, 5 0.8

-
Peak overshoot PO 1.5
Peak time, ant, 3.245 3.430 3.628 3.927 4.444 5.236
Rise time, wn t, 1.28 1.66 1.96 , 2.28 2.63 2.94
Settling time, antss 12 7.5 6.0 5.O 4.24 3.75
No. of oscillations, N 2.47 1.46 1.W 0.85 0.64 0.48

Damped Freq. ad/a,, 0.968 0.916 0.866 0.8 0.707 0.6

Note that peak overshoot is determined by damping ratio anly. P.O. = 100% for 5 = 0 and
decreases to 0% as 5 -+ 1. Figure 3.26 shows PO and ad/ anas a function of 5.
 c
Settling time t, depends on both and a,,.For a given an,t, decreases as< increases.
1
Similarly, for a given 5, tsdecreases as co,,increases. Optimum value of = ;
r c
2 = 0.707.
Example 3.13
Until now we have derived expressions for responses and time domain specifications
for the elementary servo system However, an example is completely solved only
when you have obtained numerical answers. Towards this end, we solve the
example with the following numerical data :
Potentiometer Sensitivity, Kp = 1Vtdeg.
Amplifier Gain, KA=lOAN
Motor Torque Constant, KT = 1.12 x 10- N-m/A
Total Moment of Inertia, J = 1.6 x 10- kg-m2
4 Nm
Total Friction Coefficient, B = 1.337 x 10- -
radsec

Gear Ratio, a=- 1

30
System Gain, K=Kp.KA.KT.a
= 2.135 x N-m/rad)
Hence,

This yields an==

- 115.5 radsec

o, = a,,-= 107.5 sec

'
$ = ms- ( = 68.8".
The response of the system to step input

zz 1 --I F".' sin (107.5t+ 68.8')

107.5
Response to ramp input is

= r - 0.00627 + 0.0093 e- 41.fusin (107.51+ 137.6")

Response to load torque TLis

= - 0.052 T, [I - 1.07&-"." sin (107.51+ 68.8")]

-

These responses are shown in Figure 3.27.The time domain performance of the
system is characterized by
Systcmspod Control

Time ( sec 1

(a) Step Response

Output response

o.oL-4
0.0
__
0.02
t
0.04
----
I 7
-- r --
a06 0.08 0.X) 0.12
- -1-- T --1
0.14 0.16 0.18 0.20
Time (sec 1 Time (sec 1

(b) Ramp Rap- (c) Load Response

Figure 3.27
 Coatrol Systems
Peak overshoot, PO = e-xc'm= 29.5% = 30%

I
Peak time, tp = -= 0.029 sec
IL

ad

Settling time, ts = = 0.096 sec

can

No. of oscillations,N = r=
2 v
nc
1.64

Steady state error for ramp input, e, = -=

a n 0.00627 rad = u.30 aeg.

2
Steady state error for load torque, e m = -
a TL - 0.052TL rad = 2.98TL deg.
K
1

SAQ 8
j The block diagram of a servo system employing arrmture controlled DC motor is
shnwn in Figure 3.28. Show that its closed loop transfer function
m= 42.3 . Hence determine :
R(s) (s2 + 7.7s+ 42.3)
(i) its step and ramp responses;
(ii) peak over shoot, settling time; and
(iii) steadv state errors for step and ramp inputs.

ein - 1 -3
I
r(t)
t -E+
v
10
a -
Ra
0.2
1
- (ohm)

Figure 3.28 : Armature Controlled DC Motor Servo System

A two tank water level control system is shown in Figure 3.29 (a). Its block diagram
is shown in Figure 3.29 (b). The parameters of the system are
c(t) = system output variable = water level in lower tank
RlR2 = hydraulic resistance of lower and upper tanks
C,, C2= hydraulic capacities of lower and upper tanks
T I ,T2 = time constants of lower and upper tanks
Kc = Controller gain
 controlkr
Q

(a) Schematic Diagram

Ib) Block Diagram

Figure 3.29 :Two Capacity Water Level Control System

The system parameters have numerical values as follows : TI = T, = 60 sec, RIKc= 8.

Determine
(a) response y(t) when ref. input v(t) = u(t),
(b) system error eL, when disturbance u(t) = 1/ 8, and
(c) system error ep when v(t) is a step input.

3.8.5 Steady State Response Analysis

This part of analysis concentrates its attention on steady conditions which will be reached
after the transients have settled down. We are particularly interested in the steady state
errors.
The starting point for the study of steady state errors is the error transfer function,
R(5) '
Refer the definition of error transfer function.

The expression for error E(s) in response to a given input R(s) is given by
r 1 1

The steady state error is obtained by applying the final value theorem of Laplace transform
Thus, Control Systems

e,, = lirn e(t) = lim s . E(s)

I+- s+o
Typical inputs used for study of steady state errors are again step, ramp and parabolic
inputs. The steady state errors in response to these inputs are given by the following
expressions
1
(a) for step input, r(t) = u(t), R(s) =.;
1 -- 1
e~~= [ I + lim G(s)H(s)]- (1 + Kp]
s+o
1
(b) for ramp input, r(t) = t . u(t), R(s) = 7
S
1 --1
evs = [lim
sG(s)H(s)]- K,,

(c) for parabolic input, r(t) =

e, =
if1
-

1
u(t) so that R(s) = l/s3. Hence steady state error

- 1
--
[lim s2 G(s)H(s)] K,
s+O

The constants, Kp, Kv and Ka which appear in the above expressions for steady state error
are &fined as steady state error constants. Thus
Kp = lim - G(s)H(s)= Step Error Constant ... (3.54 a)
s+o
Kv = lim s . G(s)H(s)= Ramp Error Constant ... (3.54 b)
s+o
and Ka = lim s2 G(s)H(s)= Parabolic Error Constant
s+o . ,

Whether these error constants will be zero, finite value or infinite depends on the nature of
the loop transfer function G(s)H(s).In general, G(s)H(s)is given by

where

In this expression for G(s)H(s),the expression si in the denominator represents the number
of integrations in the loop. We may classify control systems depending on the value of 'j'.
This classification and its interpretation of steady state errors is illustrated in Table 3.8.
Table 3.8 :Control Systems Classification Based on Steady State Errors

i S stem G(s) H(s) K, K, K, ep em e~

{YP
0 0 K(l+sz&l+sTb)... KO 0 0 1 00 oo

( 1 + ST,)(^ + ST,). .. 1 + Kp

1 1 K(l + sTJ ... m K, 0 0 1

- w
s ( l + ST,) ... Kv
2 2 K(l +ST,,... 00 K2 0 0 1
-
s2(1 + sTl). .. 'a
Systems and Cmtml SAQ 10
Evaluate the steady state errors to step, ramp and parabolic inputs of systems of
Examples 3.11,3.12and 3.13.

3.8.6 Response of Higher Order Systems

We studied first and second order systems first because
(a) they are simple to analyse
(b) they still illustrate typical features of practical control systems, such as
oscillatory and non-oscillatory responses, steady state errors, etc.
However, most of the practical control systems are higher order systems. We, therefore, turn
our attention to such higher order systems. Another important motivation to study higher
order systems is that they exhibit one vital characteristic which low order system do not,
viz. instability. Again we take an example to illustrate the principle.
Example 3.14
CNC System
The application we consider here is computer numerical control (CNC) system for a
machine tool, such as a lathe. Suppose we have a block of metal, such as brass; we
wish to machine a flower pot out of it. The shape of the cylindrical block and the
shape of the flower pot are shown in Figure 3.30 (a). In order to cut this shape from
the raw cylindrical block, we have to cause the cutting tool to follow the contour (or
shape) A-B-C-D.

A
Desired Shape
to Be Cut

Figure 3.30 (a) : The Desired Shape to be Cut

This task is performed by the DC servo system shown in Figure 3.30 (b). In this system the
particular contour A-B-C-D is generated on a computer and is supplied as reference input
signal r(t). The corresponding electrical signal is V,(t). The load to be positioned is the
cutting tool. Its actual position y(t) is measured by the output potentiometer whose output

Figure 3.30 (b) :Schematic Diagram of Servo System

 ----- two sienals.
These ----~~
VJt),~
r. - and
'
VJt) --
"., are a ~ d i e to
d thei differential servo amplifier, whose Control Systems

output

I
The armature voltage Va(t) drives annature current, Ia(t) which is given by

where R,, La are armature resistance (in Ohms) and inductance (in Henry) respectively
and
is the back emf of the armature controlled DC servomotor,
where
K, = back emf constant (Volts/(rad/sec)) ,and
o,(t) = motor speed (rad/sec)
The torque generated by the armatuPe controlled DC servomotor is given by
Tdt) = Kt . Ia(t) ...(3.61)
where Kt = torque constant (N-m/Amp), assuming field current If=constant.
The motor is utilized for (i) accelerating the motor-gear train- load, (ii) overcoming friction
in motor-gear train-load, and (iii) counteracting load torque TLwhich is due to cutting force.

Thus,

where J, B are moment of inertia (kg-m2)and coefficient of friction (N-m per radjsec) of
motor-gear train-load reflected to motor shaft and TL = Load torque due to cutting
force (N-m).
The servomotor is coupled to the load through gear train, so that load speed

o,(tj = a . a,(f) ...(3.63)

where a = gear ratio ( c 1).
and ... (3.64)
I Taking Laplace transforms of the above equations which govern system operation, we get
the block diagram shown in Figure 3.30 (c).

I -
%
'j,--J
p

I Figure 3.30 (c) :Block Diagram of Servo System

II Note that there are two inputs acting on the system, viz. (i) reference input, r(t), and (ii) load
torque TLwhich acts as a disturbance signal. Hence it is possible to simplify the block
o nshown in
diagram with respect to each input signal. The result of the ~ i m p ~ c a t i is
Figure 3.30 (d) and Figure 3.30 (e). From these simplified block diagrams, we get
I
Systems and Control

where %=J.La, al=Ra-B+Kb.KI

and

Figure 3.30 (d) : Simplified Block Diagram for TI,= 0 and r(t) Applied

Figure 3.30 (e) : Simplified Block Diagram for r(l) = 0 and f ' ~ ( l )Applied

Numerical Data

Let

R, = 5 ohms, La = 0.1 Henry

J=2.7116 x kg-m2
B = 1.3558 x N-m/(rad/sec)
K, = 0.6779 N-rn/Amp
Kb= 0.6779 VI(rad1sec)
a=0.1
and KA = amplifier gm = variable 1
Substituting these numerical values we get :
a, = L a . J = (0.1) (2.7116 x = 2.7116 x

a,= La B + Ra J = (0.1) (1.3558 x + (5) (2.7116 x

= 13.6936 x

a, = Ra B + KbK, = (5) (1.3558 x + 0.6779~

= 0.46633

1
% = K = K, KAKp = -(K,) (0.6779) (0.1) 3
57.3 I

= 0.01183 KA
with these numerical values,we get
The transient response of the system is governed by the roots of the characteristicequation

The nature of the roots is dependent upon the value of KA.The values of KA and
corresponding rmts are as follows :

K~ 0 68.2 135.4 343.4 877.0

Real Root 0 - 15.50 30.50 - 50.5 - 70.50

Complex - 25.50 +j32.7 - 20 +j26.45 - 10 +j31.78 +

-j41.4 +j55.5
10 -
Roots

Observe that for KA = 0, the roots are at the origin and at - 25.50 +j32.7. As the value of KA
increases, the complex roots move towards the imaginary axis of the s-plane, whereas the
third real root moves away from the origin. For KA = 343.4, we have two imaginary roots at
+ j41.4, while the real root is at - 50.50. As KA increases still further, the complex roots
-
cross over to the right cf 91ejo-axis.
The question we must ponder over is : what is the significance of the complex roots moving
towards the imaginary axis and then crossing over to the right of the imaginary axis? To
answer this question let us consider step response of the system :
250 KA
(') = s [s3+ 50.5 s2+ 1720 s + 250 KAJ ... 3:' 5' )

-
Since the forward gain is 0, the system response will be zero.
For KA= 68.2.

B,s + C,
--A1 +
-
S
+- Dl
(s + 2 0 ) ~+ (26.451~ S + 30.5
:. y(t) = A, + B,' e-20 sin (26.45 t t $,) + D, e-30.5'
This shows that the transient response terms decay to zero with time, so that the system
outputy(t) settles to steady state value = A, = 1 (see Figure 3.31 (a)).

0.0 I I I I I I I I I
0.0 0.1 0.2 0.3 0.1 0.5 0.6 0.7 0.8 0.9 1.0
Time ( see )

Figom 3.31 (a) r Response of System for K - 68.2

 ;. y (t) = A, + B,' sin(41.4t + $J + D2 e- 50.5 ' ... (3.72)
This shows that although the third term tends to zero with time, the second term shows
continuous oscillations about the steady state value A, . y (t) thus, never settles dowh to a
steady state value (see Figure 3.31 (b)).

I K volue is 135.4 I

0.0 0.1 0.2 0.3 0.L 0.5 0.6 0.7 0.8 0.9 1.0
Time ( s e c )

Figure 3.31 (b) : Response of System for K = 135.4

250 x 877
For KA= 877, y(s) =
+ 55S21[s + 70.51
s[(s - 1 0 ) ~

A3 B3s + C3 +- D3
Y(') = + (s - l o ) + 55.52 s + 70.5

Note that in the second term, the factor e+lmgrows exponentially with time. As a result, the
amplitude of sinusoidal os/ciIlations also goes on exponentially increasing (see Figure 3.31
(c)). Theoretically the amplitude of oscillations would tend to -. In practice, the amplitude
of oscillations would increase until the motor or gear train or couplings breaks down.
Hence, you must promptly switch off the system, the moment it starts oscillating wildly;
otherwise the system is liable to breakdown.

Figure 3.31 (c) : Response o t System tor K = 347.4

Control Systems
 Stable '
& ' Unstable

' A '
Region 1 ' Region

Figure 3-32: Stable Regions

3.9.3 Tests for Stability

I
From the above discussion it is clear that test for stability is essentially a test for all roots of
the characteristicequation (3.75)to have negative real parts. l

There are certain very simple tests for all roots of equation (3.67)to have negative real parts; 1
( 1 ) All coefficients, ai (i = 0 , 1 , ..., n), must be positive.
(2) None of these ai s should be zero. Let us take examples to illustrate these
conditions.
Example 3.15
Let F(s)= s3 + 5s2 + 2s -%= 0. Here uo is negative. Condition 1 above is not
. -. . . ..
satisfied. Hence at least one root must have positive real pan. rnis is mrne out oy me
A .~
fact that F(s)can be factorized into-
F(s)= ( S - l ) ( s- 2 ) ( +
~ 4) = 0
The roots of the system are located at s = +I,- 2,- 4. The root at s = +1 indicating
that the system is unstable.
Example 3.16
Let F (s)= s3 + 3 s2 + 2 s = 0. Here = 0. Hence, the root s = 0 is on jw-axis,
implying instability.
Example 3.17

F(s)= s3 + 2s2 - 7s + 1 = 0 has a, = - 7 .

This shows that at least one root must be having positive real part. Factorising
F(s)= ( s - 112(s+ 4 ) = 0 which shows two roots at s = + l .
The above conditions are necessary, but not sufficient. This is illustrated by the
following example.
Example 3.18
Let F (s)= s3 + 2 s2 + 2 s + 40 = 0. Here all coefficients are positive; however,
factorization yields

F (s) = (s"- 2 s + 10) ( S + 4 ) = [ ( S - 1)" + 3"] ( S + 4)= u

It has two complex roots at s = 1 +_ j3
We must therefore, look for tests which lead to necessary and sufficient conditions for all
roots of the characteristicequation (3.67)to have negative real parts. Such
developed by the British mathematician Routh and Gennan rnathemauclan nurwitz
..
.. . a test wa!
,.
1

independently. The conditions they derived were in different forms. .

However, they were
..L--.
shown to be equivalent to each other. Hence, this test bears their comolnea n:lme.
 -
3.9.4 Routh Hurwitz Criterion for Stability Control Systems

F(s)=a$"+a,-,s"-'+ ...+ a,s+a,=O ... (3.67)

be the characteristic equation of a control system. Then form the routh array of coefficients
as shown in Table 3.9.
Table 3.9 :Routh's Array of Coefficients

Row Coefficients
pP

S" an an-2 an-4 ...

Sn -1
an - 1 an-3 an-5 ...
bn - 1 $n-2
bn-2 bn - 3 ...
-- -
$n -3
Cn-1 Cn - 2 Cn - 3

s2 P2 P1 Po
s1 41 40
I
so '-0

bn-l = ( a n - l . a n - 2 - a n . a n - 3 ) / a n - l ... (3.76 a)

bn-2 = ( a n - l ~ a n - 4 - a n - a n - 5 ) / a n - l ... (3.76 b)
c , - ~= ( b n - 1 . a n - 3 - a n - l - b n - 2 ) / b n - l ... (3.76 c)
c , , - ~= (bn-l a n - 5 - a n - 1 . b n - 3 ) / b n - 1 ... (3.76 d)
(41P1 - P2 40) ... (3.77)
ro =
41
The Routh's array of coefficients has the following properties :
It has n + 1 rows, which have been numbered above as f ,..., s0.
(2) The f i t row starts with coefficient a, of the highest power f and has alternate
coefficients.
(3) The second row starts with coefficient a, - 1 and has again alternate
coefficients. The last coefficient is taken to be zero if n is even.
(4) The table of coefficients goes on shrinking, in the sense that
(a) 3rd and 4th rows have one coefficient less than 1st and 2nd rows;
(b) 5th and 6th rows have one coefficient than 3rd and 4th rows; etc. until
(c) row (n + 1) has only one coefficient.
(5) Coefficients in the third row onwards are derived from those of the previous
rows as shown in equation (3.77).
Routh Hurwitz condition for stability states that, for stability (i.e. roots to have negative real
parts) the necessary and sufficient condition is that all coefficients in the first column of
Routh's array be positive.
If some coefficients in the first column are negative, the system is unstable and the number
of roots with positive real parts is equal to number of changes, of sign of the coefficients in
the first column.
1,c.t US take a few examples to illustrate the concepts.
Systems and Control Example 3.19

Since all coefficients in the first column are positive, the system is stable.
Example 3.20

Since there is one negative coefficient in the first column, the system is unstable.
Furthermore, there are two changes in sign, viz +2 to - 18 and - 18 to + 40. Hence
there are two roots with positive real parts.
Example 3.21
As our last example, we consider the third order system of Example 3.14, which led
us to all this discussion on stability.

The condition for stability is 86860 - 250KA > 0 or KA < 347.44, the value
KA = 347.4 is called marginal value or critical value of gain.
For K > 347.4 the first coefficient of s' row becomes negative, indicating instability
of the system. There are two changes of sign, viz. +50.5 to 86860 - 250 KA and from
86860 - 250KA to 250KA,indicating two complex roots with positive real parts.
+
KA = 877 does lead to complex roots at +10 j55.5 corroborating this conclusion.
Example 3.22
In the above example, we employed Routh - Hurwitz stability criterion to determine
the critical value of one unknown parameter, viz. the system gain. In this example,
we see how we can determine the values of two unknown parameters for stable
operation of the system. Consider the system shown in Figure 3.33.
The characteristic equation of the system is given by
F(s) = 1 + G(s)H(s)

Applying Routh - Hurwitz criterion,

s3 2zi l+K
s2 2+zm K
s1 [(2+ zm)(l+ K ) - 2 2 A / ( 2 + 7,)
so K
The conditions for stability are :
( i ) z, > 0, (ii) K >+Oand (iii) ( 2 + z,)(l + K) - 2z,K > 0 or K < ( z , + 2) / ( z , - 2)
 Control Systems
The relation between K and 7, is shown in Figure 3.33(b). It is a hyperbola in which
(i) K + m as 2, -+2, and (ii) K + 1 as 2, + m.The shaded region in the first quadrant,
bound by (i) K = 1, (ii) 2, = 2, and (iii) K = (2, + 2) /(T, - 2), represents the stable region
.
of operation of the system.

(a) System Block Diagram

(b) Relation Between K and T,,,

Figure 3.33 for Example 3.22

3.10 BASIC CONCEPTS OF CONTROL SYSTEM

SYNTHESIS
We are given a control system. We evaluate its performance by
(i) computing its responses to various types of inputs,
(ii) determining the effect of loading and disturbances, and
(iii) investigating the stability of the system.
It is quite common to discover, as a result of analysis,-thatthe performance of the system is
unsatisfactory in the sense that
(a) the system is either unstable or highly oscillatory, or
(b) steady state errors are too large or
(c) both.
Then what do we do ?The answer to this question constitutes the synthesis problem.
We define the synthesis problem and learn the basic concepts of solution of the synthesis
problem.
 3.10.1 Statement of Synthesis Problem
As an introduction to this problem, consider the following example.
Example 3.23
The block diagram of an instrument servo system is shown in Figure 3.34 (a). Its
characteristicequation is given by
0.04s2+s+~=0=s2+25s+25~ ...(3.80)
Determine K so as to satisfy the following ratio specifications :
(i) damping ratio ,C 2 0.5
(ii) steady state error for ramp input I0.01 (i.e. 1%)

(a) Basic System

-
Deriva tiva Channel

1 Y
.
'
h ( O . O L h+1)

Proportional Channel

(b) System with PD Controller

1.6

1.1 - lp\, Original s y s t e m s

I
I
\
----- Cornpensatad systems

- - - H /
c--
---

1 I I 1
0.0 0.1 0.2 0.3 0.4 0.5 0.6
Time ( w c )

Figure 3.34 (c) : Response of the System With nnd Without Conpernation
Solution Control Sptems

Undamped natural frequency, on=

12'5
While damping ratio 6 = --- and steady state error for ramp input
mq

12.5 . .
The value of K which satisfies the first specification viz 6 = 0.5 = ~ 2 5 IS~glven
p by
i
Kp = 25. For this value of K , steady state error for ramp input of e , = - =I
Kp 25 = 0.04
whereas we require e,, I 0.01.

On the other hand, let us set K so as to satisfy e , = < 0.01.

K~ -
This yields Kp = 100 for which the damping ratio

whereas we require 6 2 0.5.

This shows that it is impossible to set Kp SO as to satisfy both 6 and e , specifications
simultaneously. This is typical of stability-accuracydilemma in control system?,
vi/. if we satisfy stability requirements, accuracy suffers and vice-versa.
In order to resolve this dilemma, it is necessary to introduce new elements in the
system. Addition of such new elements with a view to satisfying both stability and
accuracy requirements simultaneously is called compensation. The rationale for
compensation is the essence of synthesis.
There are three standard methods of compensation : (a) series, (b) feedback, and
(c) load cn~ill~cnsation.
We study the basics of these methods of compensation.
3.10.2 Series Compensation : BID Controllers
Scrics compensation consists of connecting the compensator in series with the plant in the
forward path as shown in Figure 3.35. Common types of series compensation are :
(a) PID controllers employed especially in chemical process control;
(b) phase lead, phase lag and phase lead-lag compensators used in electro-
mechanical and electric control system.

E - Series u
-Controller Plant
Compensator

B FQQ ~ ,
bock
E I Q ~t Q ~

Figure 3.35

We study the simple PID controller in this section.

Synthesis of PD Controller : In this controller, a derivative channel is provided in parallel
with the existing proportional channel as shown in Figure 3.34 (b). As a result, the forward
transfer function of the system

Kp and K,, the derivative channel gain, can be independently selected so as to satisfy
specifications on stability and accuracy.
-

and Example 3.24

Determine Kp and Kd in Figure 3.34 (b) so as to satisfy the specifications given in the
previous example.
Solution
For the PD compensated system of Figure 3.34 (b),

Hence, the characteristic equation of the system is given by

The undamped natural frequency, o,= m a s before, whereas damping ratio

U+KJ (1+Kd)

The error transfer function of the system :

The error for ramp input,

~ ( 0 . 0+4 1~)
E(s) =
S2[o.~S2+ ( 1 + kd)S+ Kp]
The steady state error for ramp input,
s2[0.04s+ 11
evs= lim sE(s)=
S+O s2[0.04s2+ ( 1 + Kd)s+ Kp]

--
-
1
KP
We have to determine Kp and Kd such that

(i) e, = = 0.01, yielding Kp = 100

KP

Substituting Kp = 100, we get Kd = 1 . To recheck let us substitute Kp = 100 and

Kd = 1 in the characteristic equation.

This yields on m= 50 - 0.5. Finally

50 radlsec and 5 = 2x50 -

All the specificationshave thus been satisfied. Consequently the synthesis of the
compensator is complete.
 Synthesis of PI Controller: Consider the second order system of Section 3.8.3, particularly Control Systems
Figure 3.23 (c). We obtain the response of the system to application of load torque, with
e,= 0, as
K a2~L(s)
ey2 (s) = - ...
[F)
Js2 + BS + K
For a constant load torque, of magnitude TL,we have

K a2~L
eY2(s) = -
s(Js2+ Bs + K) [
TI
Hence, the steady state output

The steady state error

Now, suppose our specification requires that there be no steady state error when a constant
load torque TLis applied, i.e. e,, = - eyk = 0
How do we achieve this objective? The answer to this question is to have a PI controller as
shown in Figure 3.36 of the PI controller has two channels :
(a) proportional channel with gain K,, and
1 (b) integral channel having gain KI.

Proportional Channel
I

Integral Channel
BY

( a ) Block Diagram

d
-
b) Simplified Block Diagram

(c) Response to Load Torque TI

Figure 3.36 : Servo Control System with PI Controller

The response of the system to load torque TLis given by

where, K = K$*KT a, QI = K$&T a

TL
For a constant load torque T,,TL(s)= 7
Hence

and 8 = lims . g y p )= 0 ... (3.86a)

Yz,s s-0

Consequently, eeSs= - 0Y7.w = 0 ... (3.86b)

We thus see that the error Qe has been eliminated by applying PI control Ql= amount of
integral control, applied controls, the time $at which e,($) becomes zero.

SAQ 13
Show that application of PI control also reduces the steady state error in response to
ramp input to zero (Hint :Refer to'Figure 3.30 (d)).

3.10.3 Synthesis of PID Controller for Chemical Processes

PID controllers find widest application in chemical process control. As our fmal example,
we consider a simple chemical process control system and learn how to synthesise a PID
controller for it.
There are number of rules for synthesizing or "tuning" the controller for a given process.
The best known rules are the Ziegler-Nichols tuning rules which were enunciated in 1942.
Since then, they have become the industry standard. They have stood the test of time for
more than 50 years by virtue of their simplicity and universality. It is, therefore, appropriate
that we study these rules for "tuning" the controller.
Process Model :The first step in Ziegler-Nichols rules is to derive the model of the process
being controlled. We study here the simplest FOPDT (first order plant with dead - time lag)
model which is based on the "process reaction curve" (i.e. transient response of the
process). Towards obtaining the process reaction curve, we proceed as follows :
(1) Open the process control immediately after the electronic controller, leaving the
final control element such as control value connected, as shown in Figure.
(2) Maintain the system at nonnal operating values of all variables.
(3) Introduce a small step input into system.
(4) Record the response of the system at point 'B' .
A typical process reaction curve is shown in Figure 3.37 (b). From this we get the process
model as

where Kp = process gain, L = dead-time lag, and z = process time constant.

 ControI Systems

GI = Electronic Controller
G2 = Final Con trot Element such as Control valve
G3 = Process
f4 = Feedback Element

(a) Set-up for Obtaining Process Reaetion Curve

(b) Typical Process Reaction Curve

Figure 3.37 : Deriving Process Model for Ziegler-Nichols Method

Ziegler-Nichols Tuning Rules :On the basis of the process reaction curve, Ziegler-Nichols
have recommended the following controller tuning formulae.

I P type controller
I

I
I
1 ... (3.88)
b , ( s ) = Kc E(s), where Kc = ;
;;
I V L

PI Controller
09 L
, where Kc = -, Ti = -= 3.33L
NL 0.3
PID controller
f 1 \ 1 a

i s I L
.

7 ; : L= , TD=0.5L
0.5
Example 3.25
Open loop transient response test on a temperature controller system yielded the
i following results :Dead-time lag L= 0.8 min.
Reaction rate N = O.S°C/min
' Temperature change of 10°C for 10% change in control valve setting.
1 From these data we get NL = 0.5 x 0.8 = 0.4. Based on these process parameters, we
obtain P, PI, PID controller settings as indicate below :
Systems and Control
~ontro~ier Kc Ti Td

P . 1 1
- 2.5
NL - 0.4
PI Oe9 O" - 2.25 - 2.67
NL - 0.4 0.3 - 0.3
PID ----
1.2 1.2-3 0'8-1.6 0.5L = 0.5 x 0.8 = 0.4
NL - 0.4 0.5 - 0.5

The process responses with these controller settings are shown in Figure 3.37.
SAQ 14
Compute PI and PID controller settings for a process having dead time lag = 0.45
min, time constant t = 1.0 min. and process reaction rate, N = 20" Clmin.

3.11 CONTROL SYSTEM SOFTWARE

Computer software for control system analysis and design on PC/AT is available. This
software enables the user to perform a variety of tasks such as
(i) computation and displayfplot of transient response of a given system to inputs
like step/ramp/parabolic and load changes,
(ii) evaluation and displaylplot of frequency response and root loci,
(iii) stability analysis,
(iv) controller synthesis, and
(v) state space analysis and state feedback controller synthesis.
Two levels of control system softwares are available :
(a) For student use
(i) program listings given at the end of text book by N. K. Sinha, and
(ii) program listisngs on floppy disks available from well known authors like
B. C. Kuo, R. C. Dorf, etc.
(b) For professional use
(i) MATLAB plus control system design tool box,
(ii) CACSD, and
(iii) ORACLE, etc.
Such software are available in
(a) IITs and IISc,
(b) Regional Engineering Colleges,
(c) University departments such as Roorkee University, M.S. University- Baroda,
Banaras Hindu University - Institute of Technology etc.

3.12 SUMMARY
Having come to the end, it is now time for us to look back and to take a stock of what we
have learnt, when we do this exercise, we find that we have picked up the following basic
concepts of control.
(a) It is really the cause-and-effect relationships in systems which form the basis of
control. We assert that the variable 'cause', controls the variable 'effect'.
(b) Any control system has two essential elements :the controller and the plant or
process to be controlled.
 Control Systems
There are two basic configurations of control systems : open loop and closed
loop control systems. The latter, viz. closed loop control systems have an
additional basic element, viz feedback element.
Closed loop systems are superior to open loop systems in as much as
(i) they yield much better accuracy of control,
(ii) they reduce sensitivity of the'system to parameter variations,
(iii) they attenuate the effects of external factors like load variations; set point
variations, etc, and
(iv) they improve the transient response of the system.
However, we pay for these good effects in terms of
(i) increased system complexity, bulk and power consumption, and
(ii) possibility of instability.
There are two basic types of closed loop systems :
(i) regulators, and
(ii) servo systems or tracking systems.
There are two basic problems to be studied in control systems :
(i) analysis, and
(ii) synthesis.
The analysis problem consists of determining system responses to
(1) step, ramp and parabolic inputs,
(ii) sinusoidal inputs, and
(iii) random inputs.
Accordingly we have the following methods of andysis :
(i) transient response,
(ii) steady state response,
(iii) frequency response,
(v) root locus, and
(vi) stochastic analysis.
We have studied (i) and (ii) for simplicity.
Stability is an important aspect of closed loop system performance. This
problem arises in third ~t*.d higher order systems. We have two basic stability
criteria :
(i) Routh-Hurwitz criterion, and
(ii) Nyquist criterion.
We have studied Routh-Hurwitz criterion for stability.
With the basic configuration, it is often impossible to meet stability and
accuracy requirements. It is then that the need for synthesis of compensators
arises.
The problem of synthesis consists of synthesizing compensators to enable the
system to satisfy the stability and accuracy requirements simultaneously.
There are two methods of compensation
(i) series, and (ii) feedback compensation.
PD, PI and PID controllers are the basic methods of series compensation. We
have studied the well-known Ziegler-Nichols rules for tuning PID controllers
for process control systems. ,

3.13 FURTHER READING

Kuo, B . C., Automatic Control Systems (6th Edition), Prentice Hall of India, 1992.
Dorf, R. C., Modern Control Systems (6th Edition).
Campbell, D. P., Process Dynamics, John Wiley & Sons, New York, 1958.
- - - .- . . ---3 TO SAQs
ANSWERS
_--
1
SAQ 1

System Controller Plant I

I
-- -
(a)
(b)
(c) 1
Accelerator Pedal
Control Valve

Regulator 1
Automobile
Water Tank

1
1
2
Electric Iron

terrp of
on-time
- Heating toast ,
t ~ m p .- 1 switch 1 I elements 1

Bimetallic
strip
Main switch
-
"
2
.
S 0
230V A
Cold
element
main
1 0

(a 1 Schematic d i a g r a m , I

setting st r i p
-4 H e a t in g
element -1 Actual temp

( b ) Block d i a g r a m

Figure 3.38 : Automatic Electric Iron

When the mains switch is turned 'on', current flows through heating element and
bimetallic strip. As bimetallic strip heats up, it begins to bend as shown in
Figure 3.38 (a). When the upper limit of temperature is reached, bimetallic strip
bends so much as to break the circuit. Thereupon both heating element and bimetal
strip cool down. The bimetal strip unbends. When its temperature has come down to
the lower limit, the bimetal strip recloses the electric circuit, and the cycle is
repeated. From this is derived the block diagram shown in Figure 3.38 (b).
Settines For Automatlc Electric Iron I
Fabric Wash Rayon Silk Wool Cotton Linen Steam
& Power
Wear

Temp(OF) 200 245 360 395 460 510 350

Range (OF) +20 k50 + 50 + 50 f 50 + 50 +40
This type of control is called 'on-off control'. Being least expensive, it is very
common in refrigerators, air conditioners, etc.
Temperature of the electric iron can be set by adjusting the gap between poles 1 and 1
2 of the bimetallic strip shown in Figure 3.38 (a). The smaller this gap, the longer (

will the iron remain connected to AC mains and the higher will be the temperature.
Table shows the typical temperature settings of the iron.
 Control Sptems

(i) Manual control system :Lathe operated by a skilled turner

(ii) Automatic control system :CNC (i.e. computer numerically controlled) lathe
which automatically turns a workpiece m accordance with the program stored in
the computer.
(iii) Man-machine system :Power steering mechanism for a heavy vehicle like
truck or bus. Herein, as the vehicle driver turns the steering wheel to the right or
left as the case may be, a hydraulieQowerservo turns the heavy front wheels of
the vehicle accordingly. -

Automatic pop-up toaster is an on-off temperature control system operated by a

mechanical timer. When you insert two slices of bread in the toaster and press the
lever down, the heating elements in the toaster are connected to AC mains. Hence,
the slices get heated from both sides simultaneously. After the preset time, the lever
is released, electric supply to the heating elements is broken and the toasts pop-up.

On - Temperature of Toast
Desued
TIMER TIme HEATING ____c
I
Temperature
SWITCH ELEMENTS

0- 0 A. C. Mans
Figure 3.39 : Block Diagram of Pop-up Toaster

Temperature setting knob is provided. As you move this knob towards 'dark' side,
the longer is on-time of the heating elements, the higher is the temperature to which
the slices are heated, and hence the browner will be the toasts.

System OLlCL Purpose

(i) Concrete Mixer OL To control the time for which
cement, sand, mettle, etc. are
mixed with water to form concrete.
(ii) Refrigeration CL To regulate the temperature of
food stored by using a thermostat.
(iii) Kitchen Mixer OL To control speed of mixer motor.
No sensing of speed is done. Time
for which the motor runs is
controlled.
(iv) Mechanical Dumper CL Man-machine system for levelling
of land.
(v) Elevator CL To transport passengers to desired
floor. It is a position control
system with floor switch acting as
sensor + controller.

(i) Hydraulic system : Electro-hydraulic servo for positioning heavy loads.

(ii) Mechanical :Governor for controlling the speed of a steam engine.
(iii) Thennal :Temperature controller for a baking oven.
(iv) Electro-mechanical:Variable speed drives used in industry.
Systems and Control

on - a t .
Start from y(t) = 1 - -e sm(odt + g).
ad
For finding peak overshoot, set dy(t) /dt = 0.
This will yield tP = n /adas the time at which y = y, = 1 + e- OUt/Od

= 1 + e - ,LC /m
~m,- 1
Then peak overshoot = x 100 = e(- @/-1 lmA
1
To find tr2, set y(ta) = 1. This yields sin(odta + $) = 0 or + g = IG. Hence

Substituting numerical values, we get

Y(s) / R(s) = 42.3 / (s2+ 7.7s + 42.3)
Hence, o,=a=6.50, < = 7.2 / 2 m = 0.592 = 0.6

a=<on=3.85
On - a t .
and y(t) = 1- -e sin(odt + $)
ad

= 1 - 1.25e 3.85'sin(5.2t + 53.1")

n
t, = :=0.6sec, tn=
Ed!
,. = 0.42, ts = &= 0.78
"Jd "Jd 5%

(a) Compute on, 5, a, $, a d and hence

o n -at
y(t) = 1 - -e sin(odt + $)
ad

= 1 - 1.34e-0.0333'sin(0.0373t+ 48.2")
 1 R1 1 Rl
--.-
(b) eu=-.
8 1+KCRlc8 9

KAKt/ B
Example 3.11 :G(s) = , H(s) = H.
1 + STm
This is a type 0 system with Kp = KAKt/B.
eps= l / ( l + Kp)=0.15/1.65=0.09, evs=e,=m
Example 3.13 :G(s) = K/s(Js + B) = Kv/s(l + 7,s)
Thi's is a type 1system with K, = ( K / B ) = 2.135 x 1.337 x = 159.7
Hence eps = 0,evs= 1/ Kv= 0.00627, e, =

(a) Stable,
(b) Unstable, two roots in RHF.
(c) Unstable for all ,z > 0 and K, > 0.

(a) K 128.
(b) K 11.55.

eYl(s) = K / ( J s ~ +BS + K + B/S) - R(s).

Substitute R(s) = 1 /s2 for ramp input. Apply final value theorem The desired result
will follow.

N = 2QC/min,L = 0.5 min, NL = 1. From these data you may calculate PI and PID
controller settings.