VTU Edusat Programme 5

th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
1
Dr. D.V. GIRISH
Professor & Head
Department of Mechanical Engineering
Malnad College of Engineering
Hassan 573201
Email dv_girish@rediffmail.com
Fax : 08172 -245683
Phone (O) 08172-245319 (R) 08172-266102
Mobile : 9448639079
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
2
Transfer Functions:
It is defined as the ratio of the Laplace transform of output (response) to the
Laplace transform of input (excitation) assuming all the initial conditions to
be zero.
Fig (a) System in time domain Fig (b) System in Laplace domain
Fig: Transfer Functions of a system
If G(S) be the transfer function of the system, we can write mathematically
This Transfer function is a property of the system itself, independent of the
input or driving function. The T.F includes the units necessary to relate the
input to the output, However, it does not provide any information concerning
the physical structure of the system. i.e., the T.F of many physically different
systems can be identical
STEPS TO OBTAIN TRANSFER FUNCTION
a) Write the appropriate equation which defines the behaviour of the
element.
b) Transform this equation assuming all initial conditions to be zero.
c) Form the ratio of output C(S) to input R(S)
g (t)
r(t)
C(t)
G (S)
R(S) C(S)
L.T. of output
L.T. of Input
G(S) =
all initial conditions are Zero
C(S)
R(S)
=
C(S)
R(S)
=
G(S)
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
3
Consider a spring-mass-damper (k-m-c) system on which the force F acts
and displacement x of the mass is the output.
Draw the free body diagram as shown
Equation of Motion
= F cx kx = mx
Taking laplace transform of each term of this equation (assuming Zero initial
condition), we can write,
F(s) = ms
2
X(s) + cs X(s) + kX(s)
Now, taking the ratio of X(s) to F(S) we can write the transfer function of the
system
Here the highest power of the complex variable S, in the denominator of the
transfer function determines the order of the system. Thus the k-m-c
system under consideration is a second order system.
Similarly we can write for k-m system
i.e., c = 0
and if m=0, we can write
x
F
M
0
Kx Cx
x
F
M
0
mx + cx + kx = F
.
=
..
X (S)
F (S)
=
=
G(S) =
1
ms
2
+CS+K
X (S)
F (S)
=
1
ms
2
+K
X (S)
F (S)
=
1
CS+K
K
C
.
. ..
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
4
Block diagram Reduction:
We know that,
Input-Output behavior of a Linear System or Element of a Linear System is
given by Transfer Function,
G(s)= C(s)/R(s)
Where, R(s) = Laplace transformation of the input Variable
C(s) = Laplace transformation of the output Variable
A Convenient graphical representation of this behaviour, i.e., short hand
pictorial representation of the cause and effect relationship between the
input and out put of a physical system is known as BLOCK DIAGRAM
This is shown in Fig (i)
(Input) (Output)
Fig (i)
Here, the signal into the black represents the input R(S) and signal out of the
black represents the output C(S), while the block itself stands for transfer
function G(S).
The flow of information (Signal) is unidirectional from input to the output,
with the out put being equal to the input multiplied by the transfer function
of the block.
A complex system comprising of several elements is represented by the
interconnection of the blocks for Individual elements.
The blocks are connected by lines with arrows indicating the unidirectional,
flow of information from the out put of one block to the input of the other. In
addition to this summing or differencing of signals is indicated by the
symbols shown in the fig (ii) (a) (b) & (c), while take off point of a signal is
represented by fig(iii)
i.e., in summing point two or more signals can be added or subtracted.
G (S)
R(S)
C(S)
Fig (ii) (a)
A
A + B
+
+
B
Fig (ii) (b)
A
A - B
+
-
B
AB+C
C
Fig (ii) (c)
A +
+
B
-
Fig (iii)
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
5
The points at which the output signal of any block can be applied to two or
more points is known as Take off Point.
(This output is analogous to voltage but not to current)
Forward Path:
The direction of flow of signal from Input to output is Known as Forward
Path.
Feedback Path:
The direction of flow of signal from output to Input is Known as Feedback
Path.
The following Equations refer to a Tension Regulating Apparatus such as
used in the Paper Industry.
(a) Main Input x =Fr A.,
(b) Lever Measurement. e = (x-y)/2
(c) The change in torque provided by motor tm = [Km/(1+tp) e
(d) Roll tension Fc = tm/R
(e) Tension of Control Spring y = 2 Fc/K
Draw individual Block Diagram and Determine overall Transfer function

Output
Input
=
C(S)
R(S)
G
1
(S) G
2
(S) G
3
(S)
=
R(s)
G (S)
H (S)
C(s)
Input Output
G
1
(S)
R (S) R(S) G
1
(S)
G
2
(S)
R (S) G
1
(S) G
2
(S)
G
3
(S)
C (S)
= R (S) G
1
(S) G
2
(S) G
3
(S)
Output
Input
=
C(S)
R(S)
= ?
Feed Forward
Feed back
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
6
Solution: Individual Block Diagrams:
(a) Main Input x =Fr A.,
(b) Lever Measurement. e = (x-y)/2
(c) The change in torque provided by motor tm= (Km/(1+ tp))e
(d) Roll Tension, Fc = tm/R,
(e) Tension of Control Spring, y = 2 Fc/K,
Complete Block diagram and overall transfer function:
Overall Transfer Function
CANNONICAL FORM OF CLOSED LOOP SYSTEM:
If a Block diagram which consists of a forward path having one block, a feed
back path having one block, a take off point and a summing point, it
represents a Cannonical form of a closed loop system
K
m
(1+t
p
)
t
m
1/2
A
Fr
e
1/R
2/K
Fc
x-y
A
Fr
x
e
X +
1/2
Y -
X - Y
K
m
/(1+t
p
)
2/K
F
c
y
1/R
t
m
F
c
F
c
F
r
=
=
?
e
tm
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
7
R (s) Laplace Transform of Reference Input (t)
C (s) Laplace Transform of Controlled Output c(t)
E (s) Laplace Transform of Error signal e(t)
B (s) Laplace Transform of Feed back signal b(t)
G (S) Equivalent forward path Transform function = (c(s)/E(s))
H (S) Equivalent feedback path Transform function = (c(s)/B(s))
BLOCK DIAGRAM ALGEBRA
Block diagrams of some of the control systems turn out to be very complex
such that the evaluation of their performance requires simplification (or
reduction) of block diagrams which is carried out by block diagram
rearrangements, using the rules of BLOCK DIAGRAM ALGEBRA
Some of the important Block diagram rearrangements are discussed here
Block diagram Reduction:
(i) Block diagram of a closed loop system
In the above figure, we can write
E (S) = R (S) B (S) &
B (S) = C (S) . H (S)
R(s)
G (S)
C(s)
E(s)
H (S)
+
+
B(S)
G (S)
R (S) +
C (S)
B(S)
H(s)
E(S)
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
8
C (S) = G (S) . E (S)
C (S) = G (S) . [R (S) B (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) / R (S)
= G (S) / 1 + G (S) H (S)
Similarly it can be shown for a positive feed back
C (S) / R (S) = G (S) / 1 G(S) H (S)
In General we can write
C (S) / R (S) = G (S) / 1 G (S) H (S)
In Block diagram representation we can write
C (S)
R (S)

= Transfer Function
=
G(S)
1+G(S) H(S)
G (S)
R (S) +
C (S)
B(S)
H(s)
E(S)
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
9
RULES OF BLOCK DIAGRAM ALGEBRA (Fourteen)
Original Block Diagram Equivalent Block Diagram Transformation
1)
Rearranging Summing
Points
2) Splitting the Summing
Points
3)
4)
5)
G (S)
1 G (S) H (S)
C (S)
R (S)
A + A -B
A B+C
B C
-
+
+
A +
A+C A B+C
B
C
-
+
-
A-B+C
-
A +
B
A +
A - B A B+C
B
-
+
-
Interchanging the
Blocks
G
1
G
2
A G
1
G
2
A
G1G
2
A G
1
G
2
A
G
1
G
2
A G
1
G
2
A
Combining the
Blocks in Cascade
G
1
G
2
A G
1
G
2
A
A G
1
G
1
G
1
A
A(G
1
+G
2
)
A G
2
Combining the
Blocks in Parallel
OR Eliminating a
Forward Loop
G1+G
2
A G
1
+G
2
A
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
10
6)
7)
8)
9)
10)
11)
G
1
A
A G
1
- B
B
A G
1
-
+ Moving a
Summing Point
ahead of a Block
G
1
+
A
A G
1
- B
A-B/ G
1
1/ G
1
B
B / G
1
-
G
1
A - B
A G
1
B G
1
+
B
-
A
Moving a
Summing Point
Beyond a Block
-
B G
1
G
1
+
G
1
A G
1
A G
1
B G
1
B B
G
1
A G
1 A
A G
1
G
1
G
1
A
A
A G
1
A G
1
Moving a take off
point ahead of a
Block
G
1
A G
1
A
A G
1
1/G
1
A G
1
A G
1
A
A
Moving a take off
point beyond a
Block
A -B
A
+
-
B
A -B
A
+
A -B
B
A -B
B
-
+
Moving a take off
point before (ahead
of) a summing Block
Moving a take off
point after (beyond)
a Summing block
A
+
- (A B)
A
(A B)
-
+
+
A
B
A -B +
A -B
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
11
12)
13)
(No. 10, 11, 12 & 13 to be used very carefully or better to be avoided)
14)
(No. 6, 7, 8, 9 & 14 are very important / to be used frequently)
Procedure for reduction of
Block Diagram
Step. 1 : Reduce the Cascade Blocks
Step. 2 : Reduce the parallel Blocks
Step. 3 : Reduce the internal Feed back loops
Step. 4 : It is advisable to shift take off points towards right
and summing points towards left. It is always better
to avoid rules 10 and 11
[i.e. shafting take off point before a summing block and
shifting of take off point after summing block]
Step. 5 : Repeat steps 1 to step 4 until the simple form is
obtained
A G
1
+ A G
2
G
1
G
2
A G
1
A
A
+
Removing a Block
from a Forward Path
B = (A-B G
2
) G
1
B = (AG
1
- B G
1
G
2
)
G
1
G
2
A
-
+ B
Removing a Block
from a Feed Back
Loop
B
G
2
1/G
2
G
1
-
+
B = (A / G
2
B) G
1
G
2
B = (AG
1
- B G
1
G
2
)
B = (A-B H
1
) G
1
B = (AG
1
- B G
1
H
1
)
G
1
H
1
A
-
+ B
+
A Eliminating a
Feedback Loop
A G
1
A G
2
G
2
1/G
2
G
1
A G
1
+ A G
2
A G
2
+
A
A
A
G
1
1+ G
1
H
1
B
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
12
Step. 6 : Find transfer function of the over all system using the
formula C(S) / R (S)
Any complicated system can be brought into simple form by reduction of
block diagram using block diagram Algebra, discussed earlier.
EXERCISES
Ex - 01
Ex - 02
G
1
H
1
G
2
R
H
1
G1+G2
+
R +
R
G
1
+ G
2
1-(G
1
+ G
2
) H
1
C
=
G1+G2
1-(G
1
+ G
2
) H
1
R
C
G
1
G
2
H
1
R
C
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
13
G
2
G
1
G
1
H
1
R
C
R
C
H
1
G
1
G
1
+ G
2
R
C
H
1
G
1
G
1
+ G
2
1 (G
1
+G
2
)
G
1
+ G
2
1-(G
1
+ G
2
) H
1
G
1
1
G
1
+ G
2
C
R
.
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
14
Ex - 03
Example 04:
(Jan/Feb-2003)
G
2
R
G
1
1- G
1
H
1
C
G
1
+ G
2
1-G
1
H
1
R C
G
1
+ G
2
1-G
1
H
1
C
R
=
G
2
G
1
H
1
R
C
C
G
1
1- G
1
H
1
G
1
R
G
1
1- G
1
H
1
G
2
C
R
=
H
2
H
1
G
1 S
2
G
2
G
3
G
4
S
1
C(S)
R(S)
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
15
Example: 05
July/Aug-2003
H
1
G
1 S
2
G
2
G
4
C(s)
H
2
S
1
1/G
4
G
4
R(s)
G
2
G
3
G
4
1- G
2
G
3
G
4
H
1
S
1
H
2
/ G
4
G
1
C(s)
R(s)
G
1
G
2
G
3
G
4
1-G
2
G
3
G
4
H
1
+ G
1
G
2
G
3
H
2
C(s) R(s)
H
3
G
2
G
1
G
3
H
1
H
2
C(s)
R(s)
R(s)
G
1
G
2
G
3
G
4
1-G
2
G
3
G
4
H
1
G
1
G
2
G
3
G
4
H
2
1-G
2
G
3
G
4
H
1
G
4
C(s)
1 +
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
16
Simplify further . . . . . .
Example: 06
R
(s)
G
2
G
1
H
1
H
3
/G
3
G
3
1+G
3
H
2
C
(s)
G
1
H
1
G
2
G
3
G
2
G
3
H
3
1+G
3
H
2
1+G
3
H
2
G
3
R(s) C(s)
1 +
G
1 G
2
G
4
G
5
G
6
H
1
H
2
-
- -
+
+
C(s)
+
G
3
R(s)
+
1/G
3
H
3
G
2
G
1
C(s)
H
1
H
2
R(s)
G
3
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
17
G
1
G
2
1+ G
1
G
2
H
1
G
3
G
4
+G
5 G
6
H
2
+
R(s)
C(s)
G
1
G
2
(G
3
G
4
+G
5
)
1 + G
1
G
2
H
1
G
1
G
2
(G
3
G
4
+G
5
)
1+ G
1
G
2
H
1
G
6
H
2
1 +
C (S)
R (S)
G
1
G
2
(G
3
G
4
+G
5
)
1+G
1
G
2
H
1
+ G
1
G
2
(G
3
G
4
+G
5
) H
2
x G
6
C (S)
R (S)
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
18
SYSTEM STABILITY CRITERION
A control system, if it is to be of any practical value, must be STABLE.
This means that, in response to some input, the system will not oscillate
violently or drive itself to some limiting value of the controlled variable but
rather will attain some useful response.
More specifically a stable system is one in which the transients decay, that
is, the transient response disappears with increasing values of time.
This latter statement constitutes the basic concept of stability. There are so
many techniques for determining the stability, all are based on the previous
definition.
For example, consider the following equations:
(i) My + Ky = 0 (Spring mass system)
= Mp2 + K = 0

= jen
(ii) cy + Ky = 0
= cp + K = 0
) (root
c
k
P

.
|

\
|
=
y = C1ept = c1 e
(iii) My + Cy +ky = 0
= Mp2 + Cp + K = 0
,
m
mk c
m
c
P
2
4
2
2
1
2

=
= -a jed Solution = y = e
-at
{C1 Sined + C2 cos edt}
2
1

.
|

\
|
=
m
k
j P
m
k
n
=
t
c
k

..

.
..
.
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
19
Equation 2, and 3, are stable (but not eq.1), since transients will die as
t , because of exponentially decreasing terms.
Now consider
Here characteristic equation is p-1 = 0
p=1
and transient solution is, Yt = c,e
1.t
Here it can be seen that the response does not vanish, but instead grows
rapidly with increasing values of time (i.e, yt as t )
The response of the actual system would not of course approach infinity but
rather would proceed to some extreme limiting value determined by the
physical nature of the system. Such a response is of limited practical value
Again consider the equation
Ch Eq. P
2
- 2 P + 5 = 0
= 1 j2
and the transient solution is Yt= e
1t
(C1 Sin 2t+ C2 cos 2t)
Here again an oscillatory response is indicated with an amplitude which
increases with time rather than decreases. The physical system would tend
to oscillate violently, perhaps destroying itself
These above examples illustrates the response of systems when the roots of
the characteristic equation are -ve or +ve real numbers or complex numbers
with +ve real parts, indicating STABLE and UNSTABLE systems
respectively.
Here, it should be noted that the system input has no effect on stability.
Therefore a linear system which is STABLE to one input is STABLE to all
inputs.
So far we have discussed the possibility of having a complex root with +ve or
-ve real parts.
x y
dt
dy
=
0 5 2
2
2
= + y
dt
dy
dt
y d
2
5 4 4 2
= p
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
20
But what about the possibility of having complex roots in which the real
part is a zero?
Now again consider My + Ky = 0
Ch Eq., =
Mp2 + K = 0
= jen
Solution, : C1 sin ent + C2 cos ent
The response turned out to be a persistent oscillation as shown.
The amplitude of oscillation neither decaying nor growing with time. A
system with this response is said to possess LIMITED STABILITY
Hence border line case between absolute stability and instability is called as
LIMITED STABILITY
Consider a 100th degree characteristic equation, suppose that 99 of its
roots are either -ve real number or pairs of complex numbers with -ve real
parts, but that remaining single root is a positive number is the system is
STABLE ?
No the system is UNSTABLE
Why?
Because there will be one exponential term in the transient solution which
will grow with time and so the transient solution will approach infinity as
time approaches infinity.
This above example will emphasise the basic concept of stability and states
that
For absolute stability, all roots must be negative real numbers or
complex numbers with -ve real parts.
Since solving a higher degree characteristic equation for its roots is time
consuming, the Routh / Hurwitz criterion can be used to determine whether
or not there are +ve roots with out actually solving it.
We Know,
Every system has to pass through a Transient Stage for a small period before
reaching Steady State.
So, Naturally the question comes in to playWill the system reach its
steady state after passing through transients
2
1

.
|

\
|
=
m
k
j P
..

VTU Edusat Programme 5

th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
21
To find an answer to the above question study of STABILITY is of utmost
important
STABILITY?
Even after excitation by a bounded input, output must be bounded.
In the absence of Input, Output must be zero irrespective of Initial
Conditions.
If its impulse response approaches zero as time approaches infinity.
A Control System, if it is to be of any practical value must be stable This
means
In response to some input the system will not oscillate violently or
drive itself to some limiting value of the controlled value, but rather
will attain some useful response
i.e. A System is Stable if its impulse response approaches zero as time
approaches infinity. The Stability of a System is determined by its response
to inputs or disturbances.
In general,
A Stable System is one that will remain at rest unless excited by an external
source and will return to rest if all excitations are removed.
Such condition requires that the co-efficient of t in the exponential terms of
the Transient Solution be negative real numbers or Complex numbers with
negative real parts.
Ex: e
-5t
decays while e
+5t
grows as time advances.
These co-efficient are of course the roots of the characteristic equation.
Therefore,
From the Physical view point, a stable system is one in which the transient
die out and the system Settles down to some useful response.
Mathematically, a system is said to be stable if the roots of the characteristic
equation are negative real number or Complex numbers with negative real
parts.
In a vast majority of practical systems, the following statements on stability
are quite useful.
(i) If all the roots of the characteristic equation have ve real parts the
system is STABLE.
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
22
(ii) If any root of the characteristic equation has a +ve real part or if there
is a repeated root on the je-axis, the system is unstable
(iii) If condition (i) is satisfied except for the presence of one or more non
repeated roots on the je-axis the system is limitedly STABLE
In further subdivision of the concept of stability, a linear system is
characterized as:
(i) Absolutely stable with respect to a parameter of the system if it is
stable for all values of this parameter.
(ii) Conditionally stable with respect to a parameter, if the system is
stable for only certain bounded ranges of values of this parameter.
It follows from the above discussion that stability can be established by
determining the roots of characteristic equation. Unfortunately, no general
formula in algebraic form is available to determine the roots of characteristic
equation of higher than second order. Though the various numerical
methods exist for root determination of characteristic equation, these are
quite cumber some even for third and fourth order systems.
However, simple graphical and algebraic criteria have been developed which
permit the study of stability of a system with in the need of actually
determining the roots of its characteristic equation. These criteria answer
the question, whether a system be stable or not, in YES or NO form
The roots of characteristic equations of several systems are given below.
Determine in each case if the set of roots represents stable, marginally stable
or unstable systems.
(a) -1, -4, -6, -8, (b) -3, +3, -2, -6,
(c) -5, -4, 0, -6, (d) -2+j, -2-j, -3+j, -4+j,
(e) -2+j, -2-j, 2j, -2j (f) 4, -3, -2, 6,
(g) -4, -6, 8, 5, (h) -3+2j, -3-2j, -2, -4,
(i) j, j, -2, 2, (j) -2, -2+j, -2-j, -4,
Comment on the STABILITY of the following:
(a) (d) (h) & (j) represents stable system since all the roots
have negative real parts.
(c.) (e.) represents marginally stable systems since
all the roots have non +ve real parts, that is
zero or ve.
(b) (f) (g) and (i) represents unstable systems since each has at least one
root with a +ve real part
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
23
ii. A system has poles at -2, -4, 6, and zeros 2, 4, -3, is the system is stable?
The system is STABLE since the poles which are the roots of the system
characteristic equation have negative real parts. The fact that the system
has a zero with a +ve real part does not affect its stability.
iii. (S+2).(S+3)2.(S-4) = 0, is the system is stable The characteristic
equation has the roots -2, -3, -3 and +4 and therefore represents an
UNSTABLE system since there is a +ve real root.
(a) An integrator (it may be written as dy/dt=x)
(b) A step input
(c) Abounded input that produces an unbounded output.
(d) x(t)= Cos wt
(e) x(t) = e-t sin4t
iv. (a) The characteristic equation of an integrator is S=0. Since the root
does not have a ve real part an integrator is NOT STABLE. Since
there is no root with a +ve real part an integrator is MARGINALLY
STABLE.
(b) Since a step function gives the system is MARGINALLY STABLE.
(c) The system is UNSTABLE
(d) The system is UNSTABLE since there is no decay
(e) The system is STABLE because there is decay exponentially
Rouths Creterion
E.J. Routh (1877) developed a method for determining whether or not an
equation has roots with + ve real parts with out actually solving for the
roots.
A necessary condition for the system to be STABLE is that the real parts of
the roots of the characteristic equation have -ve real parts. This insures that
the impulse response will decay exponentially with time.
If the system has some roots with real parts equal to zero, but none with +ve
real parts the system is said to be MARGINALLY STABLE.
It determines the poles of a characteristic equation with respect to the left
and the right half of the S-plane with out solving the equation.
The roots of this characteristic equation represent the closed loop poles. The
stability of the system depends on these poles. The necessary, but not
S
S F ) (
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
24
sufficient conditions for the system having no roots in the right half S-Plane
are listed below.
(i) All the co-efficients of the polynomial must have the same sign.
(ii) All powers of S, must present in descending order.
(iii) The above conditions are not sufficient
In a vast majority of practical systems. The following statements on stability
are quite useful.
(i) If all the roots of the characteristic equation have ve real parts the
system is STABLE.
(ii) If any root of the characteristic equation has a +ve real part or if there
is a repeated root on the je-axis, the system is unstable
(iii) (iii) If condition (i) is satisfied except for the presence of one or more
non repeated roots on the je-axis the system is limitedly STABLE
In this instance the impulse response does not decay to zero although it is
bounded. Additionally certain inputs will produce outputs. Therefore
marginally stable systems are UNSTABLE
The Routh Stability criterion is a method for determining system stability
that can be applied to an nth order characteristic equation of the form
an s
n
+ an-1 s
n-1
+ an-2 s
n-2
+ an-3 s
n-3
+. a1 s
1
+ a0 = ZERO
The criterion is applied through the use of a Routh Array (Routh table)
Defined as follows:
S
n
an an-2 an-4 .
S
n-1
an-1 an-3 an-5 .
S
n-2
b1 b2 b3
S
n-3
c1 c2 c3
S
n-4
d1 d2
.
.
S
2
e1 a0
S
1
f1
S
0
a0
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
25
Where an an-1,. a0 are the co-efficients of characteristic equation and
b1 b2C1 C2 evaluated as follows
1
3 2 1
1

=
n
n n n n
a
a a a a
b
1
5 4 1
2

=
n
n n n n
a
a a a a
b
This Process is continued till we get a ZERO as the last co-efficient in the
third row.
In a similar way the co-efficient of 4
th
, 5
th
, 6
th
..n
th
and (n+1)
th
rows are
evaluated.
1
2 1 3 1
1
b
b a a b
c
n n

=
1
3 1 5 1
2
b
b a a b
c
n n

=
. . . . . . .et.al
d1 =
d2 =
This table is continued horizontally and vertically until only zeros are
obtained. Any row can be multiplied by a constant before the next row is
computed with out disturbing the properties of the table.
This process is continued until s
0
is obtained, which is equal to a0.
The ROUTH STABILITY CRITERION is stated as follows,
All the terms in the first column of Rouths Array should have same sign,
and there should not be any change of sign.
This is a necessary and sufficient condition for the system to be stable
On the other hand any change of sign in the first column of Rouths Array
indicates,
(i) The System is Unstable, and
(i) The Number of changes of sign gives the number of roots lying in the
right half of S-Plane
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
26
Examples:
Ex. 01:
S
3
+6S
2
+12S+8=0 using Rouths method
S3 1 12 0
S2 6 8 0
S1
6
64
6
1 8 12 6
=
1
]
1

0 0
S0
8
6
64
0 8
6
64
=

0 0
There are no sign changes in the first column of the array and so there are
no roots with +ve real parts and hence, the system in question is stable.
Ex. 02:
S
5
+3S
4
+7S
3
+20S
2
+6S+15=0
S5 1 7 6
S4 3 20 15
S3 1/3 1 0
S2 11 15 0
S1 6/11 0 0
S0 15
System is stable
Ex. 03:
S
4
+2S
3
+3S2+8S+2=0
S4 1 3 2
S3 2 8 0
S3 1 4 0
S2 -1 2 0
S1 6 0 0
S0 2
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
27
There are two sign changes (plus to minus & minus to plus) in the first
column, showing that there are two roots with +ve real parts. Therefore the
system is unstable.
Ex. 04:
3S
4
+10S
3
+5S
2
+5S+2=0
S4 3 5 2
S3 10 5 0
S3 (2) (1) 0
S2 7/2 2 0
S1 -1/7 0
S0 2
Here two roots are +ve (2 changes of sign) and hence the system is unstable.
Ex. 05:
Examine the stability of
S
5
+2S
4
+4S
3
+8S
2
+3S+1=0
S5 1 4 3
S4 2 8 1
S3 0 2.5 0
S1 - - - - - -
How to proceed? What to do?
Ex. 06:
S
5
+2S
4
+2S
3
+4S
2
+4S+8=0
S5 1 2 4
S4 2 4 8
(S4) (1) (2) (4)
S3 0 0 0
How to proceed? What to do?
Rouths array failed
Rouths array failed
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
28
Special cases:
Occasionally, in applying the Routh stability criterion certain difficulties
arise causing the break down of the Rouths test.
The difficulties encountered are generally of the following types.
Difficulty 01:
When the first term in any row of the Routh array is zero while rest of the
row has at least one non zero term.
[Because of this zero term, the terms in the next row become infinite and
Rouths test breaks down]
Ex 05, refers to above difficulty. Now how to solve?
Difficulty 02:
When all the elements in any one row of the Routh array are zero.
Because of a zero row in the array the Rouths test breaks down.
Ex. 06 refers to this difficulty.
Now, How to solve?
Difficulty 01:
Let me repeat the problem No.5
i.e., S
5
+2S
4
+4S
3
+8S
2
+3S+1=0
We have already written the Rouths array and it is given as
S5 1 4 3
S4 2 8 1
S3 0 2.5 0
S1 - - - - - -
Here it indicates that first term in horizontal 3rd row of Routh's array is zero
while rest of the row has atleast one non zero term
To remove this difficulty, the following methods can be used,
METHOD 1
(i) Replace zero by a small +ve number c and complete the Rouths array
(ii) Examine the signs of the first column of Rouths array by letting c 0
Rouths array failed
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
29
Following the above method, continue the Rouths array
S
5
1 4 3
S
4
2 8 1
S
3
e 2.5 0
S
2
e
e 5 8
1
0 1 .
=
e
e
0
S
1
e
e
e
.
|

\
|
e
e
5 8
5 . 2
5 8
0 0
S
0
1
Now, lt 0 e
e
e 5 8
= t 1 8 e =
e
8
5
=
Therefore, the sign is ve.
Again,
It
e
e
e

.
|

\
|
e
e
5 8
5 . 2
5 8
It
,
5 8
5 . 2 5 8
2
e
e e
0 e 0 e
= It
5 . 2
5
5 . 12
=

0 e
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
30
Therefore, the array is given by
S
5
1 4 3
S
4
2 8 1
S
3
c 2.5 0
S
2
- 1 0
S
1
2.5 0 0
S
0
1 0 0
There are two changes of sign in the first column, and hence the system is
UNSTABLE (i.e. having two poles in the right half of S-plane)
METHOD 2
(i) Modify the original characteristic equations by replacing S by 1/Z
(ii) Apply the Rouths test on the modified equation in terms of Z
(iii) The number of Z-roots with +ve real parts are the same as the number of
S-roots with +ve real parts
(iv) Use same Routh criterion to determine status of the system with given
polynomial for the same above
For the problem discussed above, (i.e, example 05) let us try this method 2
Example
S
5
+2S
4
+4S
3
+8S
2
+3S+1=Zero
Replace S, by 1/Z, ( i.e. put S=1/z) The polynomial becomes
0 1
3 8 4 2 1
2 3 4 5
= + + + + +
Z Z Z Z Z
OR
0 1 2 4 8 3
2 3 4 5
= + + + + + Z Z Z Z Z
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
31
Rouths array will be
Z
5
1 8 2
Z
4
3 4 1
Z
3
6.67 1.67 0
Z
2
3.25 1 0
Z
1
-0.382 0 0
Z
0
1 0 0
There are two changes of sign in the first column of the Rouths array which
tells us that there are two Z-roots in the right half Z - plane. Therefore the
number of S-roots in the right half S - plane is also two
Hence the system is UNSTABLE
Note: This method works in most but not all cases.
Difficulty 02:
Let me repeat the problem No.6
i.e., S
5
+2S
4
+2S
3
+4S
2
+4S+8=0
We know that Rouths array will be
S
5
1 2 4
S
4
2 4 8
S
3
0 0 0
Here it indicates that all the elements in the 3rd row of the routh array is
zero
This condition indicates that there are symmetrically located roots in the S-
plane (pair of real roots with opposite signs and/or pair (pairs) of conjugate
roots on the imaginary axis and / or complex conjugate roots forming
quadrates in the S-plane)
Rouths array failed
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
32
To eliminate the above difficulty
(i) Form an equation by using the co-efficients of the array which is just
above the row of zeros. This equation is known as auxiliary equation.
[This polynomial gives the number and location of root pairs of
characteristic equation which are symmetrically located in the
S-plane. The order of the auxiliary polynomial is always even]
(ii) Replace the row of zeros in the Rouths array by a row of co-efficients
of the polynomial generated by taking the first derivative of the
auxiliary polynomial.
(iii) Continue the Rouths array and status of the polynomial may be
determined by using the same Rouths criterion discussed earlier.
Now, the auxiliary equation is formed from the coefficient of the S
4
row, which is given by
A(S) = 2S
4
+4S
2
+8
The derivative of the above polynomial with respect to S, is
The zeros in the S3 row are now replaced by the co-efficients 8 and 8 and
continue the Rouths array
S
5
1 2 4
or
S
5
1 2 4
S
4
2 4 8 S
4
1 2 4
S3 8 8 0 S3 4 4 0
S
2
2 8 0 S
2
1 4 0
S
1
- 24 0 S
1
-12 0
S
0
8 S
0
4
Here, there are two changes of sign and hence, the system is UNSTABLE
Solve: (Home Work)
(i) S
6
+S
5
-2S
4
-3S
3
-7S
2
-4S-4=0
(ii) S
6
+2S
5
+8S
4
+12S
3
+20S
2
+16S+16=0
S S
dS
S dA
8 8
) (
3
+ = =
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
33
Ex: . . . . . ?
The characterstic equation of a given system is S
4
+6S
3
+11S
2
+6S+K=0.
What restrictions must be placed upon the parameter K, in order to insure
that the system is stable. The routh table for the system is
S
4
1 11 K
S
3
6 6 0
S
2
10 K 0
S
1
10
6 60 k
0 0
S
0
K
For the system to be stable, the following the following restrictions must be
placed upon the parameter K:
i.e., 60-6K >0, or K<10, and K>0
Thus K must be greater than zero and less than 10
K>0
K<10
Determine the range of values of K so that the system having the following
characteristic equation will be stable.
S(S
2
+2S+3) (S+2) + K=0
(S
3
+2S
2
+3S) (S+2) + K=0
= S
4
+2S
3
+3S
2
+2S
3
+ 4S
2
+6S+K=0
on simplification,
We get
S
4
+4S
3
+7S
2
+6S+K=0
Rouths array becomes
S
4
1 7 K
S
3
4 6 0
S
2
5.5 K
i.e,
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
34
S
1
5 . 5
4 33 k
0
S
0
K
The system will stable if no change of sign occurs in the first column.
K > 0 and
5 . 5
4 33 k
> 0 = 33-4K > 0 or 4K < 33 K< 33/4
or K < 8.25 Range of K, is 0 < K < 8.25
Example:
(i) S
2
+KS+(2K-1) = 0
S
2
1 (2K-1)
S
1
K 0
S
0
2K-1
2K-1 > 0, K> 0
2K > 1
K >
If K > system is stable
(ii) S
4
+6S
3
+ 11S
2
+ 6S = K = 0
S
4
1 11 K
S
3
6 6 0
S
2
10 K 0
S
1
10
6 60 k
0 0
S
0
K
For system to be stable
60 60 K > 0 or K < 10 and K > 0
Range of K, V 0 < K < 10
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
35
HURWITZ STABILITY CRITERION
The Hurwitz Stability Criterion is another method for determining whether or
not all the roots of a characteristics equation have ve real parts. This
criterion is applied through the use of determinants formed from the co-
efficients of characteristic equation.
It is assumed that the first co-efficient, an is positive. The determinants i
for i = 1, 2, 3, .n-1 are formed as the Principal minor determinants of the
following arrangement (Called Hurtwitz determinant)
W.K.T. Characteristic equation
=an s
n
+ an-1 s
n-1
+ an-2 s
n-2
+. a1 s
1
+ a0 =0
Then n, can be written as,
The determinates are thus formed as below
A
1
= a
n-1
A
2
=
a
n-1
a
0
a
n
a
n-2
For Ex:
If n=3, then Hurwitz determinant can be written as
WKT Characteristic Eq: a3S
3
+a2S
2
+a1S
1
+a0 = 0
H =
a
2
a
0
0
a
3
a
1
0
0 a
2
a
0
then A1 = a2,

n
=
H =
an-1 a n-3 o . o
a n a n-2 o . o
o a n-1 a n-3 ----------- o
o a n a n-2 ----------- o
-----------------------------------
o ----------------------------------an
ao if n, odd
a1 if n, even
a1 if n, odd
a0 if n, even
H = A
n
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
36
A
2
=
a
2
a
0
a
3
a
1
A
3
=
a
2
a
0
0
a
3
a
1
0
0 a
2
a
0
Thus all the roots of characteristic equation have ve real part if
A
1
= a
2
> 0
A
2
= a
2
a
1
a
3
a
0
> 0
A
3
= a
1
a
2
a
0
a
0
2
a
3
> 0
Then the system is STABLE
Again for Ex: if n = 4, The Characteristic Equation:
= a4S4 + a3S
3
+ a2S
2
+ a1S
1
+a0 = 0
H =
a
3
a
1
0 0
a
4
a
2
a
0
0
0 a
3
a
1
0
0 a
4
a
2
a
0
The disadvantages of Hurwitz Criterion are as follows
(i) It is very complicated and time consuming for solving higher order
system.
(ii) This method is unable to find the exact number of poles located in
the right half of S-plane
(iii) It is very tough to predict marginal stability
EX: 01
s
4
+ 3 s
3
+ 6 s
2
+ 9 s+ 12 =0
The Hurwitz arrangement is given below
3 9 0 0
1 6 12 0
0 3 9 0
0 1 6 12
for stability A1, A2, A3
& A4 all must be greater
than zero
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
37
1 = 3>0
2 = 18 - 9 = 9 > 0
3 = 3(54-36) 9(9-0) + 0( ) = 54 - 81 = -27 <0
4 = 3(648 432) 972 > 0
System is UNSTABLE
Ex- 02
S
2
+ KS + (2K-1) = 0
H =
K 0
1 (2K-1)
in order for these determinants be +ve, it is necessary that
K > 0 and 2K 1> 0
i.e., K >
If K > 1/2 , then the system is stable
EXERCISES CONDITION?
4 s
3
+ 3s
2
+ 2s + 5 = 0
(March 2001)
S
6
+ 2s
5
+ 8s
4
+ 12s
3
+20S
2
+ 16S +16=0
(Aug- 2003) (Feb- 2004)
s
4
+ 5s
3
+5S
2
+ 4S +K = 0
(Feb- 2005)
s
5
+ 4s
4
+ 12s
3
+20S
2
+ 30S +100 = ZERO
(Aug- 2005)
A
1
= K, A
2
= K (2K-1)-0
= 2K
2
- K
VTU Edusat Programme 5
th
Semester Mechanical - ME 55 Control Engineering
Dr.D.V.Girish, Professor & Head (Mechanical) M.C.E., Hassan
38
CHARACTERISTIC EQUATION
The denominator polynomial in terms of S of a transfer function is known
as characteristic polynomial. If this polynomial is equated to Zero,
characteristic equation will be obtained.
The characteristic polynomial of the transfer function G(S) of the equation
, is ansn + an + sn-1 + . . . + a0
and characteristic equation is given by
anS
n
+ an + S
n-1
+ . . . . + a0 = ZERO
Solving the characteristic equation of a transfer function we get poles of the
transfer function.
) (
) (
) (
S R
S C
S G