3320100XZ8L - Control System Engineering - Solution Manual

3 Transfer Function and Impulse Response
Solutions of Examples for Practice
Example 3.3.7
Solution : Taking Laplace transform of the network,
R1 R2
1 1
Ei(t) sC1 sC2 Eo(s)
Fig. 3.1 (a)
Divide the network into two parts,

Part 1 : R1
1 é 1 ù
E i (s) = I(s) R 1 + I(s) = I(s) êR 1 + ú
sC 1 ë sC 1û
Ei(s) 1
I(s) sC1 V1(s)
... (1)
1
but, V1 (s) = I(s)
sC 1 Fig. 3.1 (b)
I(s) = V1(s) (sC1)
é 1 ù
\ E i (s) = V1 (s) (s C 1 ) êR 1 + ú
ë sC 1û
Port 2 R2
V1 ( s) 1
\ =
E i ( s) (R 1 C 1 s + 1)
1
Part 2 : V2(s) I(s) sC2 Eo(s)
æ 1 ö
V2(s) = ç R2 + ÷ I(s)
è sC 2 ø
Fig. 3.1 (c)
(3 - 1)
TM
TECHNICAL PUBLICATIONS - An up thrust for knowledge

Control System Engineering 3-2 Transfer Function and Impulse Response
1
E o (s) = I(s)
sC 2
1
E o (s) sC 2 1 1
= = =
V2 (s) æ 1 ö æ 1 ö R 2 C2 s + 1
ç R2 + ÷ sC 2 ç R 2 + ÷
è sC 2 ø è sC 2 ø
Amplifier gain = 10 hence V2(s) = 10 V1(s)
10E i (s) E o (s) 10

Eo(s) (1+R2C2s) = i.e. =
1 + R 1C1s E i (s) (1 + sR 1 C1 )(1 + sR 2 C2 )
Example 3.3.8
Solution : The s - domain network of the 1
sC
given circuit is shown in the Fig. 3.2 (a). + –
+
Applying KVL to the loop,
1 +
- I(s) - I(s) R + Vi (s) = 0 Vi(s) R Vo(s)
sC
I(s) –
Vi (s) = I(s) é + Rù
1
\ –
êë sC úû
Fig. 3.2 (a)
Vi (s) 1 + sCR
= é + Rù =
1
\
I (s) ê
ë sC úû sC
And, Vo (s) = I(s) R i.e. Vo (s)

= R
I(s)
Example 3.3.9
Solution : Applying we get the equations as,
di 1
dt C ò
E i = iR + L + idt ... (1)
Input = E i ; Output = E o
F(s)
Laplace transform of ò f(t) dt = , … Neglecting initial conditions
s
df(t)
and Laplace transform of = sF(s) … Neglecting initial conditions
dt
Take Laplace transform,
I(s)
E i (s) = I(s) éR + sL +
1 ù 1
\ i.e. = ... (2)
êë sC úû E i ( s) é 1 ù
R + sL +
êë sC úû
TM

1 1
Cò
Now Eo = idt i.e. E o (s) = I(s) i.e. I(s) = sCE o (s) ... (3)
sC
Substituting value of I(s) in equation (2),
sCE o (s) 1
\ =
E i (s) éR + sL + 1 ù
êë sC úû
E o (s) 1 1
= =
\ E i (s) é 1 ù RsC + s 2 LC + 1
sC R + sL +
êë sC úû
So we can represent the system as in the 1

Fig. 3.3 (a). Ei(s) Eo(s)
2
s LC + sRC + 1
Fig. 3.3 (a) Transfer function model
Example 3.3.10
Solution : The s-domian network is shown in the Fig. 3.4 (a).
1
R1 || –––
sC1
R1
= –––––––
1+sR1C1
R1 1 Z1(s)
–––
sC1
Ei(s) Ei(s)
1
R2 ––– Eo(s) Z2(s) Eo(s)
sC2
1
R2 || –––
sC2
R2
= –––––––
1+sR2C2
(a) (b)
Fig. 3.4
R2
Eo(s) Z2 1 + s R 2 C2
\ T.F. = = =
E1(s) Z1 + Z 2 R1 R2
+
1 + sR1C1 1 + sR 2 C2
R 2 (1 + s R 1 C 1 )
\ T.F. =
(R 1 + R 2 ) + sR 1 R 2 (C 1 + C 2 )
TM

Example 3.4.2 Kept this unsolved example for student's practice.
Example 3.5.2
Solution : From pole-zero plot given, the transfer function has 3 poles at s = – 1, – 2+j
and – 2– j. And it has one zero at s = – 3.
K(s + 3) K(s + 3) K(s + 3)
\ T(s) = = =
(s + 1) (s + 2 + j) (s + 2 - j) 2 2
(s + 1) [(s + 2) - (j) ] (s + 1) [s 2 + 4s + 5]
Now d.c. gain is value of T(s) at s = 0 which is given as 10.
K´ 3 50
\ d.c. gain = T(s)|at s= 0 i.e. 10 = i.e. K = = 16.667
1´5 3
16.667 (s + 3)
\ T(s) = ... Transfer function
( s + 1) ( s 2 + 4s + 5)
Example 3.5.3
Solution : From the pole-zero plot shown in the Fig. 1, there is 1 pole at origin, s= 0 ,
1 pole at s = –1, 1 pole at s = –2, 1 zero at s = –3.
Hence the overall transfer function with gain 2.5 is,
2.5( s + 3)
\ T(s) =
s( s + 1)( s + 2)
qqq
TM

Mathematical Models
4 of Control Systems
Example 4.6.1
Solution : As discussed earlier work done by the gears remains same.
T1 N1 q2
\ = =
T2 N2 q1
Now T2 is supplying torque to load.
d2 q2 dq 2
\ T2 (t) = J 2 + B2 ... (1)
dt 2 dt
N2
\ As T2 = T1 , Substituting in equation (1)
N1
N1 d2 q2 N1 dq 2
\ T1 = J2 + B2 ... (2)
N2 dt 2 N2 dt
N1
Now q2 = q 1 , Substituting in equation (2)
N2
2
N1 æ N ö d q1 N1 æ N ö dq 1
T1 = J2 ç 1 ÷ + B2 ç 1 ÷ ... (3)
N2 è N 2 ø dt 2 N2 è N 2 ø dt
Hence 2
é N1 ù
J2 ê N ú = Equivalent M.I. referred to primary = J eq
ë 2û
2
and é N1 ù
B2 ê N ú = Equivalent friction referred to primary = B eq
ë 2û
\ Taking Laplace of equation (3) and assuming initial conditions zero.

T1 (s) = J eq s 2 q 1 (s) + B eq s q 1 (s) = q 1 (s) [s 2 J eq + s B eq ]
q 1 (s) 1
\ =
T1 (s) s [s J eq + B eq ]
where J eq = Equivalent M.I., B eq = Equivalent friction.
(4 - 1)
TM

Control System Engineering 4-2 Mathematical Models of Control Systems
Example 4.9.10
x1 K1 x2
Solution : Two displacements : No element
under x 1 (t) alone as force is directly applied to a
spring K1 . So it will store energy and hence is the
cause to change the force applied to M2 . Hence M2
F K2
displacement of M2 is x 2 and as B 2 and K2 are
B2
connected to fixed supports both are under x 2 (t)
only as shown in the equivalent system.
At node 1, f(t) = K1 (x 1 - x 2 ) ... (1) Equivalent system
d2 x 2 dx 2
At node 2, 0 = K1 (x 2 - x 1 ) + K2 x 2 + M2 + B2 ... (2)
dt 2 dt
dx d2 x di
i) F-V analogy : M ® L, B ® R, K ® 1/C, x ® ò i dt, ® i, ®
dt dt 2 dt
1
v(t) =
C1 ò (i 1 - i 2 ) dt ... (3) Loop (1)
1 di 2 1
0 =
C1 ò (i 2 - i 1 ) dt + L 2 dt
+
C2 ò i 2 dt + R 2 i 2 ... (4) Loop (2)
The F-V analogous network is shown in the Fig. 4.1.

L2
C2
(1/K2)
V(s) C1
I1 I2 R2 Same displacement same current.
(1/K1)
Fig. 4.1
dx d2 x dv
ii) F - I analogy : M ® C, B ® 1/R, K ® 1/L, x ® ò v dt, ® v, ®
dt dt 2 dt
1
\ i(t) =
L1 ò (v 1 - v 2 ) dt ... (5) Node (1)
1 dv 2 v 2 1
0 =
L1 ò (v 2 - v 1 ) dt + C2 dt
+ +
R2 L2 ò v 2 dt ... (6) Node (2)
TM

The F-I analogous system is shown in the Fig. 4.1 (a).
v1 L1(1/K1) v2
2
1
Same displacement same voltage.

I(s) L2 R2
C2
(1/K2) (1/B2)
Fig. 4.1 (a)

Example 4.9.11
Solution : There are three displacements x i , x o and x y . The input is x i and output is x o .
So transfer function of the mechanical system is X o (s)/X i (s). The equilibrium equations
are,
d(x o - x i ) d(x o - x y )
B1 + K1 (x o - x i ) + B 2 = 0. ... (1)
dt dt
d(x y - x o )
B2 + K2 x y = 0 ... (2)
dt
Taking Laplace transform of both the equations, neglecting initial conditions,
B 1 sX o (s) - B 1 sX i (s) + K1 X o (s) - K1 X i (s) + B 2 sX o (s) - B 2 sX y (s) = 0
\ X o (s)[sB 1 + K1 + sB 2 ] - X i (s) [sB 1 + K1 ] - B 2 sX y (s) = 0 ... (3)
B2 s
B 2 sX y (s) - B 2 sX o (s) + K2 X y (s) = 0 i.e. X y (s) = X (s) ... (4)
sB 2 + K2 o
Substituting in equation (3),
sB 2 sB 2 X o (s)
X o (s)[sB 1 + K1 + sB 2 ] - X i (s)[sB 1 + K1 ] - =0
sB 2 + K2
X o (s) (sB 1 + K1 ) (sB 2 + K2 )
Solving, =
X i (s) s 2 B 1 B 2 + sB 1 K2 + sK1 B 2 + K1 K2 + s 2 B 22 + sK2 B 2 - s 2 B 22
æ B öæ B ö
ç1 + s 1 ÷ ç1 + s 2 ÷
X o (s) è K 1ø è K 2 ø
\ = ... Required transfer function
X i (s) æ B öæ B ö B
ç1 + s 1 ÷ ç1 + s 2 ÷ + s 2
è K 1ø è K 2 ø K1
TM

The s domain network for the given electric network is,

R2
Z(s)
R1 1
Ei(s) 1 Eo(s) Ei(s) R1 + Eo(s)
sC2 sC1
1 I(s)
sC1
(a) (b)
Fig. 4.2
E i (s)
Now I(s) =
1
Z(s) + R 1 +
sC 1
é 1 ù 1
and E o (s) = I(s) êR 1 + ú with Z(s) = R 2||
ë sC 1û sC 2
é 1 ù
E i (s) êR 1 +
ë sC 1 úû E i (s)(1 + sR 1 C 1 )
\ E o (s) = =
1 sC 1 Z(s) + sC 1 R 1 + 1
Z(s) + R 1 +
sC 1
1
R2 ´
sC 2 R2
Substituting Z(s) = =
1 1 + sR 2 C 2
R2 +
sC 2
E o (s) (1 + sR 1 C 1 ) (1 + sR 1 C 1 ) (1 + sR 2 C 2 )
= =
\ E i (s) sC 1 R 2 (1 + sR 1 C 1 ) (1 + sR 2 C 2 ) + sC 1 R 2
+ (1 + sR 1 C 1 )
(1 + sR 2 C 2 )
1
As K ® , B ® R the two transfer functions are identical hence the two systems are
C
analogous in nature.
Example 4.9.12
Solution : The displacements are q 1 and q as shown in the Fig. 4.3 (a). The J 1 is under q 1 .
The displacement changes to q due to K. While J2 is under q.
Thus the equivalent system is as shown in the Fig. 4.3 (a). (See Fig. 4.3 (a) on next page)
The equilibrium equations are,
d2 q1
T = J1 + K( q 1 - q) …(1)
dt 2
TM

fv3
x1(t)
K
J1 J2
K2
f(t)
T q1 q1 q q K1 M1 fv1 K3 M2 fv2
(Same) (Same)
(a) (b)
Fig. 4.3
2
d q
0 = K( q - q 1 ) + J 2 …(2)
dt 2
Taking Laplace transform of equation (1) and equation (2),
q1 K q
\ T(s) = J 1 s 2 q 1 (s) + Kq 1 ( s) - Kq( s) …(3)
\ 0 = Kq(s) - Kq 1 (s) + J 2 s 2 q(s) …(4) T J1 J2

é K+ J 2 s2 ù
From equation (4), q 1 (s) = ê ú q(s) …(5)
êë K úû
Using equation (5) in equation (1), Fig. 4.3 (c)
é K+ J 2 s 2 ù
T(s) = q 1 (s) [ J 1 s 2 + K] - Kq(s) = ê ú ( J 1 s 2 + K) q(s) - Kq(s)
êë K úû
ìï (K + J 2 s 2 ) (J 1 s 2 + K) üï
\ T(s) = í - Ký q(s)
ïî K ïþ
q(s) K K
\ = =
T(s) 2 2
(K + s J 2 ) (K + s J 1 ) - K 2 s J 1 J 2 + Ks 2 J 2 + Ks 2 J 1
4
\ q(s) K
=
T(s) 2 2
s [s J 1 J 2 + K( J 1 + J 2 )]
Example 4.9.13
Solution : The mass M1, K1 and fv3
friction fv1 are under the
x1(t)
displacement x1(t) while mass M2, K3
and friction fv2 are under the
K2
displacement x2(t). The spring K2 and
f(t)
the friction fv3 are between x1(t) and K1 M1 fv1 K3 M2 fv2
x2(t). The equivalent mechanical
system is shown in the Fig. 4.4 (a).
The equilibrium equations are, Fig. 4.4 (a)
TM

d2 x 1 dx 1 d (x 1 - x 2 )
f(t) = M1 + K1 x1 + fv1 + K2 (x1 – x2) + fv3 …(1)
dt 2 dt dt
d (x 2 - x 1 ) d2 x 2 dx 2
0 = K2 (x2 – x1) + fv3 + M2 + fv2 + K3 x2 …(2)
dt dt 2 dt
Taking Laplace transform of both the equations,

2
F(s) = X1(s) [M1s + s(fv1 + fv3) + (K1 + K2)] – X2(s) [sfv3 + K2] …(3)
0 = – X1(s) [K2 + sfv3] + X2(s) [M2s2 + s (fv2 + fv3) + (K2 + K3)] …(4)
é M s 2 + s(fv2 + fv3 ) + (K2 + K 3 ) ù
From equation (4), X1(s) = ê 2 ú X2(s)
êë K2 + sfv3 úû
Using in equation (3),

ì [M s 2 + s(fv1 + fv3 ) + (K1 + K2 )][M2 s 2 + s(fv2 + fv3 ) + (K2 + K3 )] ü
F(s) = X 2 (s) í 1 - (sfv3 + K2 ) ý
(sfv3 + K2 )
î þ
\ X 2 (s) (sfv3 + K2 )
=
F(s) [M1 s + s(fv1 + fv3 ) + (K1 + K2 )] [M2 s 2 + s(fv2 + fv3 ) + (K2 + K 3 )] - [sfv3 + K2 ] 2
2
Example 4.9.14
Solution : Write the equilibrium equation
At 1, F(t) = K(x 1 - x 2 ) ... (1)
d2 x 2 d x2
At 2, 0 = K(x 2 - x 1 ) + M + B ... (2)
dt 2 dt
Taking Laplace of both, neglecting initial conditions
F(s) = KX 1 (s) - KX 2 (s) ... (3)
0 = KX 2 (s) - KX 1 (s) + Ms 2 X 2 (s) + BsX 2 (s) ... (4)
1
from equation (3) X 1 (s) = [F(s) + KX 2 (s)]
K
0 = KX 2 (s) - Kìí [F(s) + KX 2 (s)]üý Ms 2 X 2 (s) + Bs X 2 (s)

1
Substituting in (4)
îK þ
\ 0 = KX 2 (s) - F(s) KX 2 (s) + Ms 2 X 2 (s) + Bs X 2 (s)
\ F(s) = sX 2 (s) [Ms + B]
\ X 2 (s) 1
=
F(s) s(Ms + B)
TM

Example 4.9.15
Solution : The equilibrium equation at node X o is,
d (x o - x i ) dx o
B1 + K1 (x o - x i ) + B 2 + K2 x o = 0
dt dt
Taking Laplace transform and neglecting initial conditions,
B 1 s X o (s) - B 1 s X i (s) + K1 X o (s) - K1 X i (s) + B 2 sX o (s) + K2 X o (s) = 0
\ X o (s) [s B 1 + K1 + s B 2 + K2 ] = [K1 + s B 1 ] X i (s)
\ X o (s) s B 1 + K1
=
X i (s) s (B 1 + B 2 ) + (K1 + K2 )
1
For force-current analogy, B ® , 1
R R1( )
· B1
1
K ® , x ® f, x = V, x = ò V dt Vi Vo
L
1 1
L1 ò
\ [V - Vi ] + (Vo - Vi ) dt 1
R1 o L1(
K1
)
1 1 Vi R2 L2 (1/K2) Vo
+ V +
R2 o L2 ò Vo dt = 0
(1/B2)
The force-current equivalent analogous electrical

network is as shown in the Fig. 4.5 (a).
Fig. 4.5 (a)
Example 4.9.16
Solution : q(t) = Deflection and e(t) is the applied voltage which is to be measured.
Torque developed by coil is proportional to current passing through coil
T µ i(t)
T = KT i(t)
dq(t)
e b = Kb
dt
Now torque produced has to overcome spring torque and M. I. J of system.
d 2 q(t)
\ T = Ks q (t) + J
dt 2
Now analysing coil circuit,
dq(t)
e(t) = i(t) R + e b (t) i.e. e(t) = i(t) R + Kb
dt
Taking Laplace of all the equations,
T(s) = KT I(s) ... (1)
T(s) = Ks q (s) + s 2 J q (s) ... (2)
e(s) = I (s) R + Kb s q (s) ... (3)

TM

\ Equating (1) and (2), KT I(s) = Ks q(s) + s 2 J q(s)

é K + s2 Jù
\ I(s) = q (s) ê s ú
êë KT úû
\ Substituting in (3),
q (s) R ( Ks + s 2 J ) é R (Ks + s 2 J) + Kb s KT ù
e(s) = + Kb s q (s) = q (s) ê ú
KT êë KT úû
q (s) KT
\ = is the required transfer function
e(s) 2
J R s + Kb KT s + Ks R
Example 4.9.17
Solution : Due to force applied, K2
mass M1 will be displaced by
x 1 (t). Now K 3 will consume same
x1 x2 x3
energy hence one end of B 1 will
get displaced by displacement K3
B1
x 2 (t) which is less than x 3 (t).
Now due to B 1 and K2 , the force M2
F(t) M1 K1
on M2 will be different and M2
will be displaced by x 1 (t) which
is other than x 1 (t) and x 2 (t).
So mass M1 and K1 are under
influence of x 1 (t).
Spring K 3 under influence of x 1 - x 2
Spring K2 under influence of x 1 - x 3
Friction B 1 under influence of x 2 - x 3 .
Hence equivalent mechanical system will be as shown in above figure.
The equations of equilibrium are,
d2 x 1
At x 1 , F(t) = M1 + K1 x 1 + K 3 (x 1 - x 2 ) + K2 (x 1 - x 3 ) ... (1)
dt 2
d(x 2 - x 3 )
At x 2 , 0 = K 3 (x 2 - x 1 ) + B 1 ... (2)
dt
d(x 3 - x 2 ) d2 x 3
At x 3 , 0 = B1 + K2 (x 3 - x 1 ) + M2 ... (3)
dt dt 2
Use F-V analogy and the replacements,
1 dx d2 x di
M ® L, B ® R, K® , x ® q, ® i, ®
C dt dt 2 dt
and x ® q ® ò i dt
TM

di 1 1 1 1
V(t) = L 1 +
dt C 1 ò i 1 dt + C 3 ò (i 1 - i 2 )dt + C2 ò (i 1 - i 3 )dt ... (1)
1
0 =
C3 ò (i 2 - i 1 )dt + R 1 (i 2 - i 3 ) ... (2)
1 di 3
0 = R 1 (i 3 - i 2 ) +
C2 ò (i 3 - i 1 )dt + L 2 dt
... (3)
Simulating using loop basis,
L1 C1 (1/K1)
C3 (1/K3) R1
i2
V +
_ L2
i3
C2 (1/K2)
i1
Currents through various elements,
L1 , G1 i1 alone
C3 i1 - i2
R1 i2 - i3
C2 i1 - i3
L2 i3 alone
Note The elements in parallel in equivalent mechanical system drawn based on nodal analysis
appears in series in F-V analogous system based on loop analysis. Similarly elements in series
in mechanical system appears in parallel in F-V analogous system. Remember this, while
drawing F-V analagous system.
In this problem it can be observed that
M1 , K1 in parallel ® L 1 , C 1 in series
K 3 , B 1 in series ® C 3 , R 1 in parallel
(K 3 , B 1 series) parallel with (K2 ) ® (C 3 , R 1 in parallel) series with (C 2 )
TM

Control System Engineering 4 - 10 Mathematical Models of Control Systems
Example 4.9.18
Solution : Due to applied force, M1 will be displaced by x 1 . Spring K1 will be under
same displacement.
Due to friction B 1 , K 3 will get displaced by x 2 as friction causes change in displacement.
Due to spring K2 , friction B 2 will be under influence of x 3 which is different than x 2 .
This is due to storing of energy by K2 .
Due to B 2 and K 3 , mass M2 will again under different displacement x 4 .
Due to friction B 3 , again there will be change in force applied to K4 hence K4 will be
under the influence of displacement x5 .
So mass M1 and K1 are under x 1 .
B 1 between x 1 and x 2 .
K2 between x 1 and x 3 .
K 3 between x 2 and x 4 .
B 2 between x 3 and x 4 .
B 3 between x 4 and x5 .
K4 under x5 alone.
Hence equivalent system is as shown,
K2
B2 B3
x1 x2 x3
x4 x5
B1
F(t) M1 K1 K3 M2 K4
The equations of equilibrium are,

d2 x 1 d(x 1 - x 2 )
At x 1 , F(t) = M1 + K1 x 1 + B 1 + K2 (x 1 - x 3 ) ... (1)
dt 2 dt
d(x 2 - x 1 )
At x 2 , 0 = B1 + K 3 (x 2 - x 4 ) ... (2)
dt
d(x 3 - x 4 )
At x 3 , 0 = K2 (x 3 - x 1 ) + B 2 ... (3)
dt
TM

d(x 4 - x 3 ) d(x 4 - x5 ) d2 x 4
At x 4 , 0 = K 3 (x 4 - x 2 ) + B 2 + B3 + M2 ... (4)
dt dt dt 2
d(x5 - x 4 )
At x5 , 0 = K 4 x5 + B 3 ... (5)
dt
Using F-V analogy, the replacements are,
1 dx d2 x di
M ® L, B ®R K® , x ® q, ® i, ®
C dt dt 2 dt
dq
and x ® q ® ò i dt as i =
dt
di 1 1 1
V(t) = L 1 +
dt C 1 ò i 1 dt + R 1 (i 1 - i 2 ) + C2 ò (i 1 - i 3 )dt ... (6)
1
0 = R 1 (i 2 - i 1 ) +
C3 ò (i 2 - i 4 )dt ... (7)
1
0 =
C2 ò (i 3 - i 1 )dt + R 2 (i 3 - i 4 ) ... (8)
1 di 4
0 =
C3 ò (i 1 - i 2 )dt + R 2 (i 4 - i 3 ) + R 3 (i 4 - i5 ) + L 2 dt
... (9)
1
0 =
C4 ò i5 dt + R 3 (i5 - i 4 ) ... (10)
Elements Currents
L1 , C1 i1
R1 i1 - i2
C2 i1 - i3
C3 i2 - i4
R2 i3 - i4
R3 i4 - i5
C4 i5
L2 i4
Remember parallel elements in series and series elemets in parallel.
Mechanical F-V
M1, K1 parallel L1 , C1 series
TM

B1, K 3 series R1, C3 parallel
K 2 , B2 series C2 , R2 parallel
B1 , K 3 series in R1 C3 parallel in series

parallel with K 2 , B2 with C2 ,R2 parallel
Thus the F-V analogous network based on loop basis is,
L1 C1 (1/K1) L3
R1 C3 (1/K3)
i2
+ R3 C4 (1/K4)
V (t) _
i5
C2 (1/K2) i4
R2
i1
i3
Check the elements and currents as per the table made above.
Example 4.9.19
Solution : For the mass M 1 , the displacement is x 1 due to force f(t), against the friction
force,
d 2 x1 dx
\ f (t) = M1 + B1 1 K (1)
dt 2 dt
Let force transmitted to mass M 2 be f2 (t) which will cause the displacement of x 2 of
mass M 2 against spring K 2 and friction B2 .
d 2 x2 dx2
\ f2 (t) = M 2 + K1 x2 +B2 K (2)
dt 2 dt
Now taking moments about fulcrum,
f (t) ´ L1 = f 2 (t) ´ L 2
L1
\ f 2 (t) = ´ f (t) ... (3)
L2
TM

The equivalent mechanical system is,
x1 x2
B1 M1 f(t) L 1/ L 2 f2(t) M2 B2 K2
For F-V analogy, M ® L , B ® R, K ® 1/C, x ® q, dx/dt ® I

Hence equations (1) and (2) get modified as,
dI
V(t) = L1 1 + I1 R1 K (4)
dt
dI 2 1
V2 (t) = L 2
dt 2
+ I2 R 2 +
C2 ò I 2 dt K (5)
L1
and V2 (t) = . V(t) K (6)
L2
L1
The equation (6) indicates that it acts as a transformer with turns ratio . Hence F-V
L2
analogous network is,
L1 R1 L2 R2
I1
v(t) V2 I2 C2 (1/K2)
(L1 / L2 )
For obtaining the transfer function X 2 (s) /X1 (s), take laplace transform of the equations (1)
and (2),
F (s) = M1 s 2 ´ X1 (s) + B1 s X1 (s) K (7)
F2 (s) = M2 s 2 X 2 (s) + B 2 s X 2 (s) + K2 X 2 (s) ... (8)
L1
and F2 (s) = F (s) ... (9)
L2
TM

Substituting (7) and (8) in (9) we get,
L1
M 2 s 2 X 2 (s) + B2 s X 2 (s) + K 2 X 2 (s) =
L2 [
M1s 2 X1 (s) + B1s X1 (s) ]
L1
\ [
X 2 (s) M 2 s 2 + B2 s + K 2 ]= [
X ( s) M1s 2 + B1s
L2 1 ]
X2 ( s) L1s (M1s + B1 )
=
X1 ( s)
(
L 2 M 2 s 2 + B2 s + K 2 )
This is the required transfer function.
qqq
TM

Block Diagram
5 Models of Control Systems
Solutions for Examples for Practice
Example 5.3.11
Solution :
The blocks G 1 and G 2 are in series.
Minor loop
G3 3 blocks in
parallel
+
R(s) + G1 G2 G4 G6 C(s)
+
– +
H1 G5
H2
Series
R(s) + G1 G2
G3 + G4 + G5 G6 C(s)
1 + G1 G2 H1
–
H2
Minor loop
R(s) + G1 G2 (G3 + G4 + G5)

G6 C(s)
1 + G1 G2 H1
–
H2
(5 - 1)
TM

Control System Engineering 5-2 Block Diagram Models of Control Systems
G1 G2 (G3 + G4 + G5)
R(s) 1 + G1 G2 H1
G6 C(s)
1 + G1 G2 (G3 + G4 + G5)H2
1 + G1 G2 H1
R(s) G1 G2 G6 (G3 + G4 + G5) C(s)

1 + G1 G2 H1 + G1 G2 H2 G3 + G1 G2 H2 G4 + G1 G2 G5 H2
Fig. 5.1
Example 5.3.12
Solution : Shifting the take off point to the right as shown we get,
Parallel
G3
G2
R(s) + +
G1 G2 1 C(s)
– –
H1
The dotted portion shown has a parallel combination of block of 1 and G 3 / G 2 , so they
add up. Then shift summing point to the left of G 1 .
R(s) + G3
G1 G2 1+ C(s)
– G2
–
H1
G2+G3
R(s) G1 G2 C(s)
G2
– –
1
G1
H1
TM

Interchanging the positions of summing points using Associative law, ... Rule 1
Minor feedback loop
R(s) G2 + G3 C(s)
G1 G2 ––––––
G2
– –
H1
G1G2
–––––––––
1+G1G2H1
1
––
G1
Combining two blocks in forward path we get canonical form from which we can write,
G 1 G 2 (G 2 + G 3 )
C(s) G 2 (1 + G 1 G 2 H 1 ) G 1 (G 2 + G 3 )
= = ... Ans
R(s) G G (G + G 3 ) 1 1 + G 1G 2 H 1 + G 2 + G 3
1+ 1 2 2 ×
G 2 (1 + G 1 G 2 H 1 ) G 1
Example 5.3.13
Solution : Combine the parallel blocks G 2 and G 3 to give G 2 + G 3 and minor loop of
G 1 and H 1 .
G4
+
R(s) G1 C(s)
–––––––––– G2 + G3
– 1 + G1H1
Minor feedback
loop
H2
G1
Reduce the minor feedback loop of G 1 and H 1 to give .
1+ G 1 H 1
G4
G1(G2 + G3) +
R(s) C(s)
1 + G1H1 + G1H2(G2+G3)
TM

Reducing inner minor loop we get,

The two blocks are in parallel, so adding them we get the single equivalent block as
C(s) G4+ G4G1H1+ G4G1G2H2+G4G1G3H2+ G1G2 + G1G3

= … Ans.
R(s) 1 + G1H1 + G1G2H2 + G1G3H2
R(s) G1G2 + G1G3 C(s)

G4 +
1 + G1H1 + G1G2H2 + G1G3H2
Example 5.3.14
Solution : The blocks G 2 and G 3 in parallel, use rule 3.
Minor loop of
G1 1 and G6
– –
+ +
R(s) G2 + G3 G5 G6 C(s)
– +
G4
G7
G1 G2 + G3 Minor loop with G = 1, H = G1(G2 + G3)
–
+ + G6
R(s) G2 + G3 G5 C(s)
1 + G6
+
–
Series
G4
G7
TM

G4 + G5(G2 + G3)
Parallel blocks,
Rule 3
+ 1 G6
R(s) 1 + G1(G2 + G3) G5(G2 + G3) C(s)
+ 1 + G6
– +
G4
G7
+
1 G6
R(s) C(s)
1 + G1(G2 + G3) G4 + G5(G2 + G3) 1 + G6
–
G7
G 2 G5 G 6 + G 3 G5 G 6 + G 4 G 6
\ G eq =
(1 + G 1 G 2 + G 1 G 3 ) (1 + G 6 )
\
H eq = G 7
Geq
R(s) C(s)
1 + GeqHeq
Example 5.3.15
Solution : Rearranging the block diagram we get,
R + + C
G3 G8 G5 G2 G1
– –
G7 G4
+ Separate
+ paths
G6
TM

Separate the feedback paths through G 4 and G 7 , combined at summing point before G 6 .
So G 6 will appear separately in each path of G 4 and G 7 , as shown below.
Minor loop
R
G3 G8 G2G5 G1 C
– –
–
G6 G7
G6 G4
Reducing inner minor feedback loop,
R G8
G3 G2G5 G1 C
1 + G6G7G8
– –
G4G6
Hint : Combine inner blocks, reduce minor feedback loop. Again combine blocks in
forward path to get simple form.
G
Using we get,
1 + GH
C(s) G 1 G 2 G 3 G5 G 8
\ = ... Ans.
R(s) 1 + G 6 G7 G 8 + G 2 G 4 G5 G 6 G 8 + G 1 G 2 G 3 G5 G 8
TM

Example 5.3.16
Solution : Shifting take off point before G 2 . ... Rule 6
Critical
rule G3
+
R(s) + C(s)
G1 G2
– –
H2
H1 G2
Shifting take off point before summing point using critical rule No. 10,
G3
+
R(s) C(s)
G1 G2
– +
– H2
H1G2
For this negative sign,
include block of '–1'
while separating the path.
Separating the paths in the feedback path shown dotted,
G3
+
R(s) C(s)
G1 G2
– – –
H1G2 H2
H1G2 –1 H2
Key Point Remember that though the paths through summing point are separated, signs at the
summing points to those paths must be carried as it is. Hence after H 2 , carry the negative
sign and then H 1 G 2 to get – H 1 H 2 G 2 .
TM

Shifting summing point as shown and then interchanging the two summing points using
associative Law we get, ... Rule 4 and 1
Parallel
G3
Minor loop
G2
+
R(s) G1 C(s)
1 G2
1 + H1G1G2
– –
H2
– H1H2G2
Rule 3 Rule 8
R(s) G1 G3 G2 C(s)
1+
1 + H1G1G2 G2 1 + G2H2
–
– H1H2G2
Minor loop Geq
R(s) G1 (G2 + G3) C(s)

(1 + G2H2) (1 + H1G1G2)
–
– H1H2G2
Heq
C(s) G eq G 1 (G 2 + G 3 )
Rule 8 gives, = =
R(s) 1 + G eq H eq 1 + G 2 H2 + H1 G 1 G 2 - G 1 G 2 G 3 H1 H2
TM

Example 5.3.17
Solution : Shift summing point before G1 and take off point of H2 after the block G4.
1/G1 H2 1/G4
–
R Y
G1 G2 G3 G4
– –
H3
H1
Minor G3G4
––––––––
loop 1+G3G4H3
H2
––––
G1G4
–
R G3G4 Y
G1G2 ––––––––
1+G3G4H3
–
H1
G1G2G3G4
––––––––––
Minor 1+G3G4H3
loop = –––––––––––––––––
G1G2G3G4 H2
1+ –––––––––– ´ ––––
1+G3G4H3 G1G4
R G1G2G3G4
–––––––––––––––– Y
1+G3G4H3+G2G3H2
–
H1
G1G2G3G4
–––––––––––––––– G1G2G3G4
R 1+G3G4H3+G2G3H2 Y
–––––––––––––––––– = –––––––––––––––––––––––––––––
G1G2G3G4 1+G3G4H3+G2G3H2+G1G2G3G4H1
1 + –––––––––––––––––– ´ H1
1+G3G4H3+G2G3H2
TM

Control System Engineering 5 - 10 Block Diagram Models of Control Systems
Example 5.3.18
Solution : Shift the take off point of V3(s) after G2.
1/G2
+
R(s) + + C(s)
G1 G2 G3
– – –
H2 H3
H1
Parallel
1 1+G2
Minor 1+ –– = –––– Minor
loop G2 G2 loop
G2 G3
–––––– ––––––
1+G2H2 1+G3H3
R(s) G2 1+G2 G3 C(s)

G1 –––––– ––––– ––––––
1+G2H2 G2 1+G3H3
–
H1 G1G2
Minor loop –––––––
1+G2H2 G1G2
–––––––––– = ––––––––––––––
G1G2H1 1+G 2H2+G1G2H1
1+ –––––––
1+G2H2
R(s) G1G2 1+G2 G3 C(s)

–––––––––––––– ––––– ––––––
1+G2H2+G1G2H1 G2 1+G3H3
Series
R(s) G1G3(1+G2) C(s)

––––––––––––––––––––––––––––––––––––––––
1+G2H2+G1G2H1+G3H3+G2G3H2H3+G1G2G3H1H3
Fig. 5.2
TM

Example 5.3.19
Solution : Shifting take off point before G 2 .
G3
+
R(s) + C(s)
G1 G2
– –
H2
H1 G2
Shifting take off point before summing point using critical rule No. 10,
G3
+
R(s) C(s)
G1 G2
– –
+
H2
H1G2 –
Separating the paths in the feedback path shown dotted,
Minor G3
feedback loop
+
R(s) C(s)
G1 G2
– – –
H1G2 H2
– H1H2G2
Key Point Remember that though the paths through summing point are separated, signs at
the summing points to those paths must be carried as it is.
Hence after H 2 , carry the negative sign and then H 1 G 2 to get – H 1 H 2 G 2 .
Shifting summing point as shown and then interchanging the two summing points using
Associative Law we get,
TM

Blocks in parallel
G3
G2
Minor feedback loop
+
R(s) G1 C(s)
G2
1 + H1G1G2
– –
H2
– H1H2G2
G2 + G3
G2
R(s) G1 G3 G2 C(s)
1+
1 + H1G1G2 G2 1 + G2H2
–
– H1H2G2
R(s) G1 (G2 + G3) C(s)

(1 + G2H2) (1 + H1G1G2)
–
– H1H2G2
G 1 (G 2 + G 3 )
C(s) (1 + G 2 H 2 ) (1 + H 1 G 1 G 2 )
=
R(s) G (G + G 3 ) ( - H 1 H 2 G 2 )
1+ 1 2
(1 + G 2 H 2 ) (1 + H 1 G 1 G 2 )
\ C(s) G 1 (G 2 + G 3 )
=
R(s) 1 + G 2 H2 + H1 G 1 G 2 - G 1 G 2 G 3 H1 H2
Example 5.3.20
Solution : Separating the feedback paths,
TM

R(s) C(s)
G1 G2 G3
– – –
+ H1
S –
H2
Separating the dotted paths from the summing points 'S'.

S1 S2
R(s) T1 T2 C(s)
G1 G2 G3
– – –
4
1 H1 3
+ H1
–
2
H2
Separating the paths from T1 to S 1 , T1 to S 2 , T2 to S 1 and T2 to S 2 .

4
H1 –H2
Parallel
–
R(s) C(s)
G1 G2 G3
–
– – 1 –
H1
H1 H1 Minor feedback
loop
2
– H2
– H1H2
–
R(s) G3 C(s)
G1 – H1 G2
1 + G3(–H1H2) Shift the
– – summing point
H1
TM

–H1H2
–
R(s) G3 C(s)
G1–H1 G2
1 – G3H1H2 Interchange
– –
Interchange the
summing points
1
H1
G2
–H1H2
– C(s)
R(s) + G2G3
G1–H1
1 – G3H1H2
– –
H1
G2
Parallel Minor loop
G2G3
1 – G3H1H2–G2G3H1H2
G2G3
R(s) H1 1 – G3H1H2 C(s)
G1 – H1 –
G2 G2G3
–
1+ × –H1H2
1 – G3H1H2
G1G2–H1G2–H1
G2
R(s) G1G2 – H1(1 + G2) G2G3 C(s)

×
G2 1 – G3H1H2 – G2G3H1H2
–
[G 1 G 2 - H 1 (1 + G 2 ) ] G 3
C(s) 1 - G 3 H 1H 2 - G 2 G 3 H 1H 2
\ =
R(s) [G 1 G 2 - H 1 (1 + G 2 ) ] G 3
1+
1 - G 3 H 1H 2 - G 2 G 3 H 1H 2
TM

\ C(s) G 1G 2 G 3 - H 1G 3 - H 1G 2 G 3
=
R(s) 1 - G 3 H 1H 2 - G 2 G 3 H 1H 2 + G 1G 2 G 3 - H 1G 3 - H 1G 2 G 3
Example 5.3.21
Solution : Separating two feedback from second takeoff point which is after block having
transfer function G 2 as shown, we get,
Minor feedback
loop H2
–
R(s) C(s)
G1 G2 G3
+ – +
H1
H1
G4
H2
–
R(s) G2 C(s)
G1 G3
1 + G2H1
+ +
H1
G4
Shifting summing point behind the block having transfer function ‘G 1 ’ as shown we get,
1 / G1 H2
–
R(s) G2 C(s)
G1 G3
1 + G2H1
+ +
H1
G4
TM

Use Associative law for the two summing points and interchange their positions, we get,
H2
G1
–
R(s) G2 C(s)
G1 G3
1 + G2H1
+ +
Series Minor loop
H1
G4
G1G2
1 + G2H1 G1G2
=
G1G2 1 + G2H1 – G1G2H1
1– × H1
1 + G2H1
Minor loop G1G2G3

1 + G2H1 – G1G2H1 + G2G3H2
H2
G1
G1G2G3
– R(s) 1 + G2H1 – G1G2H1 C(s)
R(s) G1G2G3 C(s)
1 + G2H1 – G1G2H1 G1G2G3 H2
+ 1+ +
1 + G2H1 – G1G2H1 G1
G4 G4
G4 and other block are in parallel hence,
R(s) G1G2G3 C(s)

G4 +
1 + G2H1 – G1G2H1 + G2G3H2
\ C(s) G 4 + G 4 G 2 H1 – G 4 G 1 G 2 H1 + G 2 G 3 G 4 H2 + G 1 G 2 G 3
=
R(s) 1 + G 2 H1 – G 1 G 2 H1 + G 2 G 3 H2
Example 5.3.22
Solution : Shift summing point to the left and take off point to the right as shown in the
Fig. 5.3 (a).
TM

sC1 R2
–
R(s) + 1 1 + 1 1
sC1 R1 sC2 R2 C(s)
– –
Fig. 5.3 (a)
Interchange the summing points and solve minor feedback loops.
sC1R2
–
R(s) + 1 1
sC1R1 sC2R2 C(s)
– –
Loop 1 Loop 2
Fig. 5.3 (b)

1 1
Loop 1 = = sC1R2
sC 1 R 1 1 + sC 1 R 1
1 –
1+ R(s) 1 1
sC 1 R 1 1 + sC1R1 1 + sC2R2 C(s)
1 1 Fig. 5.3 (c)

Loop 2 = =
sC 2 R 2 1 + sC 2 R 2
1
1+
sC 2 R 2
æ 1 ö æ 1 ö
ç ÷ ç ÷
C( s) è 1 + sC R
1 1ø è 1 + sC R
2 2 ø
\ =
R( s) 1 1
1+ ´ sC 1 R 2
(1 + sC 1 1) (
R 1 + sC 2 R2 )
C( s) 1
=
R( s) s C 1 C 2 R 1 R 2 + s (R 1 C 1 + R 2 C 2 + R 2 C 1 ) + 1
2
Example 5.3.23
Solution : No blocks are in series or parallel and no minor feedback loop is existing so
shifting summing point towards left i.e. behind block with transfer function G 1 as shown,
we get,
TM

G3 1 / G1
+
R(s) C(s)
G1 G2
– –
H1
H2
Use Associative law for the summing points, we get,
Parallel G3
G3
1 + —— G1 Minor
G1 loop
+
R(s) C(s)
G1 G2
– –
H1
H2
G1 + G3
G1 Series
R(s) G3 G1 C(s)
1 + —— G2
G1 1 + G1H1
–
H2
R(s) G1 + G3 G1 G2 C(s) R(s) (G1 + G3) G1 G2 C(s)

G1 1 + G1 H1 G1 1 + G1 H1
– –
H2 H2
G2 (G1 + G3)
R(s) G2 (G1 + G3) C(s) R(s) 1 + G1H1 C(s)
1 + G1H1 G2 (G1 + G3) H2
– 1+
1 + G1H1
H2
TM

\ C(s) G1 G2 + G2 G 3
=
R(s) 1 + G 1 H1 + G 1 G 2 H2 + G 2 H2 G 3
Example 5.3.24
Solution : Separate all the paths linked by take-off point and summing point in the
feedback path.
Parallel of ‘1’ and G3
Minor loop
+
R(s) G1 + G2 G3 C(s)
– +
–
+
H1
+
Separate the dotted paths
Minor loop
+ G2
R(s) G1 ––––––– 1 + G3 C(s)
– 1 + G2
–
H1
H1
G2
––––––––––––
1 + G2 + G2H1
G2
–––––––
1 + G2
R(s) G1 + 1 + G3 C(s)
G2H1
1 + –––––––
– 1 + G2
H1
TM

G2 (1 + G3)
R(s) G1 + ––––––––––– C(s)
1 + G2+ G2H1 G2 (1 + G3)
R(s) G1 –––––––––––
– 1 + G2+ G2H1 C(s)
G2 (1 + G3)H1
1 + –––––––––––
1 + G2+ G2H1
H1
Minor loop
\ C(s) G 1 G 2 (1 + G 3 )
=
R(s) 1 + G 2 + 2G 2 H 1 + G 2 G 3 H 1
Example 5.4.3
Solution : In this case there are two inputs and two outputs. Consider one input at a time
assuming other zero and one output at a time. Consider R 1 acting, R 2 = 0 and C 2 not
considered R 2 = 0 and C 2 is suppressed (not considered). C 2 suppressed does not mean
that C 2 = 0. Only it is not the focus of interest while C 1 is considered. As R 2 = 0,
summing point at R 2 can be removed but block of '–1' must be introduced in series with
the signal which is shown negative at that summing point.
Forward path
R1 G1 C1 Minor
– loop
G3 R1 G1 C1
» –
G4 – G2G3G4
Feedback
–1 G2 path
C1 G1 G1
= =
R1 1 + [G 1 ] [- G 2 G 3 G 4 ] 1 - G 1 G 2 G 3 G 4
C2
For , assume C 1 suppressed.
R1
R1 G1
–
Forward G3 R1 – G1G3G2 C2
path
» –
G4 Feedback G4
path
–1 G2 C2 Minor loop
TM

\ C2 - G1 G 3 G2 - G1 G2 G 3
= =
R1 1 + [- G 1 G 3 G 2 ] [G 4 ] 1 - G 1 G 2 G 3 G 4
C1
For , R 1 = 0 and C 2 is suppressed.
R2
–1 G1 C1
Feedback
Forward path
G3 R2 – G1G2G4 C1
path
–
»
G4 G3
–
R2 G2 Minor loop
\ C1 - G1 G2 G4 -G1 G2 G4
= =
R2 1 + [- G 1 G 2 G 4 ] [G 3 ] 1- G1 G 2 G 3 G4
C2
For , R 1 = 0 and C 1 is suppressed.
R2
–1 G1
G3 R2 G2 C2
Feedback » –
path
G4 – G1G3G4
–
R2 G2 C2 Minor loop
Forward path
C2 G2 G2
= =
R2 1 + [G 2 ] [- G 1 G 3 G 4 ] 1 - G 1 G 2 G 3 G 4
Example 5.4.4
Solution : i) With N(s) = 0 block diagram becomes
TM

Minor loop
3
+
R(s) E(s) 10 C(s)
s+4
s(s + 1)
– –
0.5 s
10
s (s + 1) 10 10
Minor feedback loop = = =
10 2 2
1+ 0.5s s + s + 5s s + 6s
s (s + 1)
+
R(s) E(s) 10 C(s)
s+4 2
s + 6s
–
3
s+4
+
R(s) E(s) 10 C(s)
s+4 2
X(s) s + 6s
–
Assume output of second summing points as X(s),

10 (s + 4)
Hence E(s) = R(s) – C(s) …(i) and C(s) = X(s) ... (ii)
s 2 + 6s
3
While X(s) = E(s) + R(s) ... (iii)
s+ 4
Substituting value of X(s) and R(s) from (i) and (ii) in (iii) we get,
s 2 + 6s 3 3
C(s) = E(s) + E(s) + C(s)
10 (s + 4) s+ 4 s+ 4
Solving, C(s) 10 (s + 7)
= when N(s) = 0
E(s) 2
s + 6s - 30
C(s)
ii) To find , we have to reduce block diagram solving minor feedback loop and
R(s)
shifting summing point to the left as shown earlier in (i).
TM

So referring to block diagram after these two steps i.e.
3
s+4
+
R(s) E(s) 10 C(s)
s+4 2
s + 6s
–
Exchanging two summing points using associative law,
3
s+4
+
R(s) 10 (s + 4) C(s)
2
s + 6s
–
3
Parallel of '1' and ––––
s+4
3 10 (s + 4)
= 1 + –––– ––––––––
s+4 G s2 + 6s 10(s+4)
Minor loop –––––– = = –––––––––––
1 + GH 2
s+7 10 (s + 4) s +16s+40
= –––– 1 + ––––––––
s+4 s2 + 6s
C(s) æ s + 7 ö æç 10 (s + 4) ö÷ 10(s + 7)
\ = ç ÷ ´ =
R(s) è s + 4 ç 2 ÷ 2
ø è s + 16s + 40 ø s + 16s + 40
iii) With R(s) = 0 block diagram becomes,
N(s)
+
+ 10 + C(s)
s+4
s(s + 1)
– –
0.5 s
The block of ‘3’ will not exist as R(s) = 0. Similarly first summing point will also vanish
but student should note that negative sign of feedback must be considered as it is, though
summing point gets deleted.
TM

In general while deleting

N(s)
summing point, it is +
necessary to consider the +
10 + C(s)
signs of the different signals –1 s+4
s(s + 1)
at that summing points and –
should not be disturbed. So 0.5 s
introducing block of ‘–1’ to
consider negative sign.
N(s)
+ N(s)
+
10 + C(s)
10 + C(s)
s(s + 1)
+ – s(s + 1)
+
0.5 s
– (s + 4) – 0.5 s
– (s + 4) After simplification ( – 1.5 s – 4)

Parallel
Removing summing point, as sign is positive no need of adding a block.

N(s)
+
+
10 + C(s) N(s) C(s)
s(s + 1)
+
– 10 (1.5 s + 4)
– (1.5 s + 4)
s(s + 1)
Minor loop with G = 1
C(s) 1 1 s(s + 1)
\ = = =
N(s) é -10 (1.5s + 4) ù 15s + 40
s2
1-ê ú 1+ + 16s + 40
ë s(s + 1) û s(s + 1)
Example 5.4.5
Solution : Consider R 1 alone R 2 , R 3 , R 4 are zero.
Note Whenever R 1 is zero, as the sign of feedback at R 1 is negative, while removing

summing point at R 1 , do not forget to insert a block of '–1' to consider effect of
negative sign.
TM

Minor loop
R1 C R1 G2 C
G1 G2 G1
1 + G2H2
– – –
H2
H1
H1
G 1G 2
C 1 + G 2 H2 G 1G 2 G 1G 2 R 1
\ = = i.e. C =
R1 G 1G 2 H 1 1 + G 2 H 2 + G 1G 2 H 1 1 + G 2 H 2 + G 1G 2 H 1
1+
1 + G 2 H2
Consider R 2 alone, with R 3 = R 1 = R 4 = 0

Note
this R2 R2
Minor loop
+ +
+ C + G2 C
–1 G1 G2
1 + G2H2
–
H2
–G1H1
H1
G2
C 1 + G 2 H2
G 2 R2
\ = i.e. C=
1 + G 2 H 2 + G 1G 2 H 1
R2 æ G2 ö
1 -ç ÷ ( -G 1 H 1 )
è 1 + G H
2 2 ø
Consider R 3 alone, R 1 = R 2 = R 4 = 0
Note R3
this
–
+ + C
–1 G1 G2
–
H2
H1
TM

Combining two summing points we get,

Minor loop
R3 – C R3 – G2 C
G2
1 + G2H2
+ – +
H2
–G1H1
–G1H1
G2
C 1 + G 2 H2 G2
\ = =
-R3 æ G 2 ö 1 + G 2 2 + G 1G 2 H 1
H
1 -ç ÷ ( -G 1 H 1 )
è 1 + G 2 H2 ø
\ - R3G2
C =
1 + G 2 H 2 + G 1G 2 H 1
Consider R 1 alone, with R 1 = R 2 = R 3 = 0

Note
this
C
–1 G1 G2
+
–
R4 G2 C
H2 –G1H1
1 + G2H2
+
+
H1
+
R4
-G 1 G 2 H 1
C 1 + G 2 H2 -G 1 G 2 H 1
\ = =
R4 æ G 2 ö 1 + G 2 H 2 + G 1G 2 H 1
1 - ( -G 1 H 1 )ç ÷
è 1 + G 2 H2 ø
\ - G 1G 2 H 1R 4
C =
1 + G 2 H 2 + G 1G 2 H 1
Combining all the values of C, we get
G 1 G 2 R 1 + G 2 (R 2 - R 3 ) - G 1 G 2 H 1 R 4
C =
1 + G 2 H 2 + G 1G 2 H 1
TM

Example 5.4.6
Solution : i) Input R at station I
R(s) H3
+ –
+ + C(s)
G1 G2 G3 Shift the
takeoff point
– –
H2
H1
R(s) 1
H3
G3
+ –
+ C(s)
G1 G2 G3
+
– –
Minor loop
H2
H1
R(s) H3 Minor loop

G3
+ –
+ G3
– G1 G2 C(s)
1 + G3H2
H1
G2G3
1 + G3H2 + G2H3
R(s)
G2G3
+ 1 + G3H2 C(s)
G1
G2G3 H3
– 1+ ×
1 + G3H2 G3
H1
TM

G 1G 2 G 3
C(s) 1 + G 3 H2 + G 2 H 3 G1 G2G 3
\ = =
R(s) G 1G 2 G 3 1 + G 3 H 2 + G 2 H 3 + G 1G 2 G 3 H 1
1+ ´ H1
1 + G 3 H2 + G 2 H 3
ii) Input R at station II

Note that at station I, through input is zero while removing the summing point, the
negative sign of H 1 as it is. Thus add a block of '– 1' as shown in the system.
Note
H3 R(s)
this
– +
C(s)
G1 + G2 + G3
–1
–
H2
H1
Shift the summing point to the left of G 2 and interchange the two summing points.
Minor loop H3 R(s)
– +
+ C(s)
–1 G1 G2 G3
– +
H2 1
G2
H1
R(s)
+
G2 C(s)
G3
1+G2H3 +
+ –
H2
G2
– G1H1
Blocks in parallel
TM

R(s)
+
G2 + C(s)
G3
1+G2H3
H
– G1H1 – 2
G2
– G1G2H1– H2
G2
C(s) G3 R(s)
\ =
R(s) G 3 (- G 1G 2 H 1 - H 2 )
1– +
1+ G2H3 C(s)
G3
+
\ C(s) G 3 (1 + G 2 H 3 )
=
R(s) 1 + G 2 H 3 + G 3 H 2 + G 1G 2 G 3 H 1
– G1G2H1– H2
Example 5.4.7 1 + G2H3
Solution : Assume R( s) = 0 but to consider negative

sign at the summing point introduce block of '–1' as shown in the Fig. 5.5 (a). Rearranged
block diagram is shown in the Fig. 5.5 (b).
4 W(s)
Note
this
1 Y(s) W(s)
–1 2 4 Y(s)
s+1 +
+
– 14
7 s+1
(a) (b)
Fig. 5.5
é ù
Y( s) ê 1 ú 4( s + 1)
= 4´ ê ú =
W( s) ê1 - 1 ´ é -14 ù ú s + 15
êë êë s + 1 úû úû
Example 5.4.8
Solution : Solving the minor feedback loop of ‘G 4 ’ and ‘1’.
TM

–
G1 G2 G3 C1
+
H2
H1
+
G4
R1 G5
+ 1 + G4
As C 2 is not the focus of interest, G 6 becomes meaningless in the block diagram. Shifting
take off point to the right of G 2 .
Minor
–
loop
G1 G2 G3 C1
+
Shift
1
H2
G2
H1
+ G4G5
R1
1 + G4
G1G2
G3 C1
1 + G1G2
1 1
H2
G2 G3
H1
+ G4G5
R1
+ 1 + G4
TM

Combining all the blocks in series.
G1G2G3G4G5H1
R1 C1
(1 + G1G2) (1 + G4)
+
H2
G2G3
G 1 G 2 G 3 G 4 G5 H1
C1 (1 + G 1 G 2 ) (1 + G 4 )
\ =
R1 G G G G G H H2
1- 1 2 3 4 5 1 ×
(1 + G 1 G 2 ) (1 + G 4 ) G 2 G 3
\ C1 G 1 G 2 G 3 G 4 G5 H1
=
R1 1 + G 1 G 2 + G 4 + G 1 G 2 G 4 - G 1 G 4 G5 H1 H2
1 1
Example 5.5.2 sC1 sC2
V1(s)
Solution : The s-domain network, currents
and voltages are shown in the Fig. 5.6
V (s) - V1 (s) Vi(s) R1 R2 Vo(s)
I 1 (s) = i
1
I1(s) I2(s)
sC 1
Fig. 5.6
= sC 1 [Vi (s) - V1 (s)] …(1)

V1 (s) = R 1 [I 1 (s) - I 2 (s)] …(2)
I 2 (s) = sC 2 [V1 (s) - Vo (s)] …(3)
Vo (s) = R 2 I 2 (s) …(4)
The block diagrams for equations are,
I1(s) Vi(s) I2(s)

Vi(s) sC1 I1(s) R1 V1(s) sC2
– – – I2(s) Vo(s)
R2
V1(s) I2(s) Vo(s)
sC1 R1 sC2
Equation (1) Equation (2) Equation (3) Equation (4)
TM

The complete block diagram is shown in the Fig. 5.7.
Vi(s) I1 V1 I2 Vo(s)
sC1 R1 sC2 R2
Shift R1 Shift
sC1 sC2
Fig. 5.7
Interchange Interchange
Vi(s)
sC1 R1 sC2 R2 Vo(s)
– – –
1 1
––
––
R1 R1 R2
sC1 sC2
Minor
loops
Vi(s) Vo(s)
sC1 R1 sC2 R2
– –
sC1 sC2
1 1 1
–– x R1 x –– = ––
R1 R2 R 2
R1 R2
–––––– –––––––––
1+s R1C1 1+s R2 C2
Series
R1 R2
Vi(s) sC1 ––––––– sC2 ––––––– Vo(s)
1+sC1R1 1+s R2 C 2
–
1
––
R2 Minor
loop
s R1R2C2
Vi(s) Vo(s)
(1 + sC1R1)(1 + sC2R2)
sC1
sR1R2C2
1+ x1
(1 + sC1R1)(1 + sC2R2) R1
TM

Vo (s) sR 1 R 2 C 2
\ = sC 1 ´
Vi (s) (1 + sC 1 R 1 ) (1 + sC 2 R 2 ) + sR 1 C 2
s 2 R 1R 2 C1C2
=
s 2 R 1 R 2 C 1 C 2 + s[C 1 R 1 + C 2 R 2 + R 1 C 2 ]+ 1
qqq
TM

Signal Flow Graph
6 Models of Control Systems
Example 6.5.14
Solution : T1 = G 1 G 2 G 3 G 4 ... Only one forward path
No combination of nontouching loops.
G2 G3 G1 G2 G3 G4
–H1 – H2
L1 = –G2H1 L2 = –G3H2
– H3
L3 = –G1G2G3G4H3
\ D = 1– [L 1 + L 2 + L 3 ] = 1 + G 2 H 1 + G 3 H 2 + G 1 G 2 G 3 G 4 H 3
D1 = 1 ... All loops touching to T1
\ C(s) TD G 1G 2 G 3 G 4
= 1 1 =
R(s) D 1 + G 2 H 1 + G 3 H 2 + G 1G 2 G 3 G 4 H 3
Example 6.5.15
Solution : Name all the summing and take off points as shown below and representing
each separately as a node draw the signal flow graph.
G4
s1 s2 s3 s4
t1 t2 + t3
R(s) C(s)
G1 G2 G3
– – –
H2
H1
(6 - 1)
TM

Control System Engineering 6-2 Signal Flow Graph Models of Control Systems
G4
1 s1 1 s2 1 t1 G1 s3 G2 t2 G3 s4 1 t3 1
R(s) C(s)
–H1 –H2
–1
Number of forward paths are K = 2
Forward path gains are
T1 = G 1 G 2 G 3
T2 = G 4
Individual feedback loops are
G4
1 G1 G2 G2 G3 1 1 G2 1
–H2 –H1
–H1 –H2
L1 = –G1 G2 H1 L2 = –G2 G3 H2 L3 = +G2 G4 H1 H2
G4
1 G1 G2 G3 1 1 1 1
–1 –1
L5 = –G4 L4 = –G1 G1 G3
There are no combination of nontouching loops

\ D = 1 - [L 1 + L 2 + L 3 + L 4 + L5 ]
All the loops are touching to both the forward paths
\ D1 = D2 = 1
\ Using Mason’s gain formula
C(s) T D + T2 D 2
= 1 1
R(s) D
TM

C(s) G1 G 2 G 3 + G 4
\ =
R(s) 1 + G1 G 2 H1 + G 2 G 3 H 2 + G1 G 2 G 3 + G 4 - G 2 G 4 H1 H 2
Example 6.5.16
Solution : Name the various summing and take-off points to draw the signal flow graph
as shown,
G3
+
H1
–
s3
s1 +
R(s) G1 G2 C(s)
+ t1 t2
Input s2 Output
–
H2
The corresponding signal flow graph is,

s3
+G3 –H1
1
1 1 G1 G2 1
Input Output
s1 t1 s2 t2
–H2
Forward path gains are,

T1 = G 1 G 2 T2 = G 3 G 2
The various loops are,
+G3
1
G1 G2 1 1 G2 –H1
1
G2
–H2 –H2
L1 = – G 1 G 2 H2 L2 = - G 2 G 3 H2 L 3 = – G 2 H1
No combinations of non toucting loops.
\D = 1 – [L 1 + L 2 + L 3 ] = 1 + G 1 G 2 H 2 + G 2 G 3 H 2 + G 2 H 1
For T1 , All loops are touching, \ D 1 = 1
For T2 , All loops are touching, \ D 2 = 1
TM

According to Mason's gain formula,

C(s) T D + T2 D 2
= 1 1
R(s) D
C(s) G1 G2 + G2 G 3
=
R(s) 1 + G 1 G 2 H2 + G 2 G 3 H2 + G 2 H1
Example 6.5.17
Solution : Representing each take off point and summing point by separate node, the
signal flow graph is as shown in the Fig. 6.1.
G6 G5
1 1 G1 G2 G3 G4 1 1
R(s) C(s)
–H1
–H2
Fig. 6.1
Forward paths : T1 = G1G2G3G4 , T2 = G 6G 2G 3G4,
T3 = G1G2G5 , T 4 = G 6 G2 G5
Individual feedback loops : L1 = – G2G3H1, L2 = – G1G2G3G4H2, L3 = – G1G2G5H2
No combination of two non touching loops.
\ D = 1– [L 1 + L 2 + L 3 ] = 1 + G2G3H1 + G1G2G3G4H2 + G1G2G5H2
All loops are touching to T1, T2, T3, T4 hence D 1 = D 2 = D 3 = D 4 = 1
C( s) T1 D 1 + T2 D 2 + T3 D 3 + T4 D 4 G 1G 2 G 3 G 4 + G 6 G 2 G 3 G 4 + G 1G 2 G5 + G 6 G 2 G5
\ = =
R( s) D 1 + G 2 G 3 H 1 + G 1G 2 G 3 G 4 H 2 + G 1G 2 G5 H 2
Example 6.5.18
Solution : The various forward paths are,
G3 G4 G3 G4
G1 G2 G5 G6
T1= G1G2 T2 = G3G4 G2 G1

T4 = G3G5G2 T5 = G1G6G4
G7
T3 = G7
Fig. 6.2
The various individual loop gains are,
H2 H1 G5 G6
The combinations of two non-touching loops are,
L 1 L2 = H 1 H 2
L 1 = H2 L 2 = H1 L3 = G5G6
No combination of three non-touching loops. Fig. 6.2 (a)
TM

H2
H2 All loops
are non-touching
G5 G6 to T3 G3 G4
L1 non-touching to T1
G1 G2 H1
H1
L2 non-touching to T2
G7
Fig. 6.2 (b)

\ D = 1 – [L1 + L2 + L3] + [L1 L2]
All loops are touching to T4 and T5.
\ D 1 = 1 – L1 , D 2 = 1 – L2 , D 3 = 1 – [L1 + L2 + L3] + [L1 L2] , D 4 = D5 = 1
C(s) T D + T2 D 2 + T3 D 3 + T4 D 4 + T5 D5
\ = 1 1
R(s) D
G 1 G 2 (1 - H 2 ) + G 3 G 4 (1 - H 1 ) + G 7 [1 - H 1 -H 2 - G 5 G 6 + H 1 H 2 ]
\ C(s) + G 3 G5 G 2 + G 1 G 6 G 4
=
R(s) 1 - H 1 - H 2 - G5 G 6 + H 1H 2
Example 6.5.19
Solution : System node variables are Y1 , Y2 , Y 3 , Y4 .
Consider equation 1 : This indicates Y2 depends on Y1 and Y 3
Consider equation 2 : This indicates Y 3 depends on Y1 , Y 2 and Y 3
G4
G1 Y2 Y3
G5
G3 Y1 Y2 Y3
G2
S.F.G. for equation (1)
Consider equation 3 : This indicates Y4 depends on, Y 3 and Y2

G4
G6
G5
G1 G2 G7
Y1 Y2 Y3 Y4
G3
Y2 Y3 Y4
G7
G6
Fig. 6.3
TM

Combining all three we get, complete S.F.G. as shown in Fig. 6.3,

No. of forward paths = K = 4
4 T D T D + T2 D 2 + T3 D 3 + T4 D 4
\ T.F. = å K K = 1 1 ... Mason's gain formula
D D
K= 1
T1 = G 1 G 2 G 7 ,T2 = G 4 G 7 , T3 = G 1 G 6 , T4 = G 4 G 3 G 6
Individual loops are,
G2 G5
L1 L2
G3 SELF LOOP
L1 = G2 G3 L2 = G5
\ D = 1 - [L 1 + L 2 ] = 1 - G 2 G 3 - G 5
No nontouching loop combinations.
For T1 , T2 and T4 , all loops are touching. \ D1 = D2 = D4 = 1
L2
And for T3 , `G 5 ' self loop is nontouching, \ D3 = 1 -G 5
Y4 T D + T2 D 2 + T3 D 3 + T4 D 4 G5
= 1 1
Y1 D
G1
G 1 G 2 G 7 × 1 + G 4 G 7 × 1 + G 1 G 6 (1 - G 5 ) + G 4 G 3 G 6 × 1
= T3
D G6
L2 nontouching to T3
Y4 G G G + G 4 G 7 + G 1 G 6 (1 - G 5 ) + G 4 G 3 G 6
\ = 1 2 7
Y1 1 - G 2 G 3 - G5
Example 6.5.20
1 G1 1
Solution : There are two inputs and two R1 C1
outputs hence there are four gains possible as
G3 H1
C1 C2 C2 C C
, , and 2 . Let us find the gain 1 G4
R1 R2 R1 R2 R1 H2
by Mason's gain formula. The input R2 = 0 and

G2
C2 is suppressed as shown in the Fig. 6.4 (a).
The forward paths are, T1 = G1 (only one Fig. 6.4 (a)
forward path).
The individual feedback loops are,
The loops L1 and L2 are nontouching to each other, hence L1 L2 = G3G4H1H2
\ D = 1 – [L1 + L2 + L3] + [L1 L2]
TM

All loops are touching to T1 hence D1 = 1.

G1
H2
G4
G3 H2
H1 H1
G2
L1 = G3 H2 L2 = G4 H1 L3 = G1 G2 H1 H2
Fig. 6.4 (b)
C1 T D G1
= 1 1 =
R1 D 1 - G 3 H 2 - G 4 H 1 - G 1G 2 H 1H 2 + G 3 G 4 H 1H 2
Similarly other gains can be obtained.
Example 6.5.21
Solution : Forward paths : T1 = G1 G2 G3 G4 G5 G7, T2 = G1 G2 G3 G4 G6 G7
Individual feedback loops are,
G6
G2 G4 G4 G5 G4
H1 H2
H3 H3
L1 = G2H1 L2 = G4H2 L3 = G4G5H3 L4 = G4G6H3
Fig. 6.5
Two nontouching loop combinations : L1L2, L1L3, L1L4

\ D = 1 – [L1 + L2 + L3 + L4] + [L1L2 + L1L3 + L1L4]
All the loops are touching to T1 and T2 hence D1 = D2 = 1.
ì ü
C(s) T1 D 1 + T2 D 2 ï + ï
ï G G G G
1 2 3 4 5 7 G G G G G G
1 2 3 4 6 7 G G ï
\ =í
1 - G 2 H 1 - G 4 H 2 - G 4 G5 H 3 - G 4 G 6 H 3 + G 2 G 4 H 1H 2 ý
=
R(s) D
ï ï
ïî + G 2 G 4 G5 H 1H 3 + G 2 G 4 G 6 H 1H 3 ïþ
Example 6.5.22
Solution : Only one forward path, T1 = G 1 G 2 G 3 G 4 G 5
TM

The individual feedback loops are,

G1 G3 G1 G2 G3 G4 G5
–H1 –H3 –H4 –H5

–H2
L1 = – G1H1 L2 = – G3H3 L3 = – G1G2G3H2 L4 = – G4H4 L5 = – G5H5

Fig. 6.6
The nontouching loops : L 1 L 2 , L 1 L 4 , L 1 L5 , L 2 L5 , L 3 L5

Three nonotouching loops : L 1 L 2 L5
\D = 1 - [L 1 + L 2 + L 3 + L 4 + L5 ] + [L 1 L 2 + L 1 L 4 + L 1 L5 + L 2 L5 + L 3 L5 ] - [L 1 L 2 L5 ]
D 1 = 1 as all the loops are touching to T1
C(s) TD G 1G 2 G 3 G 4 G5
\ = 1 1 =
R(s) D {1 + G 1 H 1 + G 3 H 3 + G 1 G 2 G 3 H 2 + G 4 H 4 + G 5 H5 + G 1 G 3 H 1 H 3
+ G 1 G 4 H 1 H 4 + G 1 G 5 H 1 H5 + G 3 G 5 H 3 H5 + G 1 G 2 G 3 G 5 H 2 H5
+ G 1 G 3 G 5 H 1 H 3 H5 }
Example 6.5.23
Solution : Number of forward paths K = 2
2
S TK D K
C(s) K= 1 T1 D 1 + T2 D 2
\ T.F. = = = ... Mason's gain formula
R(s) D D
T1 = 5 × 4 × 2 × 10 = 400 and T2 = 1 × 10 × 2 × 10 = 200

Key Point Self loops cannot be part of forward path or another individual loop as node
containing self loop gets traced twice which is not allowed.
Individual loops are,
4 2 –1
–1 –2 SELF LOOP
L1 = – 4 L2 = – 4 L 3 = –1
Combinations of two non-touching loops are,
i) L 1 and L 3 and ii) L 2 and L 3
No combination of three non-touching loops :
\ D = 1 -[L 1 + L 2 + L 3 ] + [L 1 L 3 + L 2 L 3 ]
= 1 – [– 4 – 4 – 1] + [4 + 4] = 1 + 9 + 8 = 18
Consider T1 , L3 loop is non-touching.
TM

T1
–1
1 5 4 2 10
L3
\ D 1 = 1 – L 3 = 1 – [ –1] = 2
Consider T2 , L 1 is non-touching,
–1
T2
10 2
1 5 4 2 10
L1
–1
D 2 = 1 – L 1 = 1 – [ – 4] = 5
C(s) T1 D 1 + T2 D 2 400 ´ 2 + 200 ´ 5

\ = = = 100
R(s) D 18
Example 6.5.24
Solution : The various forward path gains are,
T1 = G 1 G 2 G 3 G 6 G 7 G 8 , T2 = G 1 G 4 G 6 G 7 G 8
T3 = G 1 G 4 G 5 G 8 , T4 = G 1 G 2 G 3 G 5 G 8
The individual feedback loop gains are,
G3 G6 G7 G2 G3 G6 G7 G8
– H2
– H1
L1 = – G3G6G7H2 L2 = – G2G3G6G7G8H1
G4 G4 G5 G5
G6 G7 G8 G8 G2 G3 G8
– H1
– H1 – H1
L3 = – G4G6G7G8H1 L4 = – G4G5G8H1 L5 = – G2G3G5G8H1

Fig. 6.7
TM

Control System Engineering 6 - 10 Signal Flow Graph Models of Control Systems
All loops are touching to each other, no combination of non touching loops.
\ D = 1 – [L 1 + L 2 + L 3 + L 4 + L5 ]
All loops are touching to all the forward paths hence,
D1 = D2 = D 3 = D4 = 1
x2 T D + T2 D 2 + T3 D 3 + T4 D 4
= 1 1
x1 1 – [L 1 + L 2 + L 3 + L 4 + L5 ]
G1G 2 G 3 G 6G 7G 8 + G1G 4 G 6G 7G 8 + G1G 4 G5 G 8 + G1G 2 G 3 G5 G 8

=
1 + G 3 G 6G 7H2 + G 2 G 3 G 6G 7G 8 H1 + G 4 G 6G 7G 8 H1 + G 4 G5 G 8 H1 + G 2 G 3 G5 G 8 H1
Example 6.5.25
Solution : The forward path gains are,
T1 = G 1 G 2 G 3 G 4 , T2 = G 1 G 6 G 4 , T3 = G 1 G 7
The individual feedback loop gains are,
L 1 = G 2H1 L2 = G5 (self loop)
Both loops are nontouching to each other hence L1L2 = G 2 G5 H 1
\ D = 1 – [L 1 + L 2 ] + [L 1 L 2 ]
For T1 , both loops touching hence D 1 = 1
For T2 , both loops touching hence D 2 = 1
For T3 , L2 is nontouching hence D 3 = 1 – L2 = 1 – G5
x5 T D + T2 D 2 + T3 D 3 G 1 G 2 G 3 G 4 + G 1 G 6 G 4 + G 1 G 7 (1 – G 5 )
\ = 1 1 =
x1 D 1 – G 2 H1 – G5 + G 2 G5 H1
Example 6.7.4
Solution : Laplace transform of the given network is as shown in following figure.
R1 V1(s) R2
1
Vi(s) sL Vo(s)
sC
I1(s) I2(s)
Key Point Assume the network variables alternately as the loop current and node voltage and
then write the equations by analysing the horizontal and vertical branches alternately.
(Vi - V1 ) 1
I 1 (s) = ... (I) V1 = (I 1 - I 2 ) ... (II)
R1 sC
TM

( V1 - Vo )
I2 = ... (III) Vo = I 2 sL ... (IV)
R2
S.F.G. for equation (I) S.F.G. for equation (II)
1 1
+ V1
R1 I1 I1 sC V1 I2
Vi
– 1 –1
R1 sC
S.F.G. for equation (III) S.F.G. for equation (IV)

1
+
V1 R2 I2 Vo
I2 sL Vo
–1
R2
1 1 +1
Total Signal Flow Graph for the network is as R1 I1 sC V1 R2 I2 sL
follows. Vi Vo
V
Use Mason's gain formula to find o . –1 –1 –1
Vi
R1 sC R2
Vo S TK D K
= ; Number of forward paths = 1
Vi D
Vo T D1
\ = 1
Vi D
L
T1 =
R1 R2 C
1 1 sL
L1 = – , L2 = – , L3 = –
sR 1 C sR 2 C R2
L 1 and L 3 are non-touching.
\ D = 1 – [L 1 + L 2 + L 3 ] + [L 1 L 3 ]
1 1 sL L
D = 1+ + + +
sR 1 C sR 2 C R2 R1 R2 C
As all loops are touching to T1 , D 1 = 1,
L
Vo R1 R2 C
\ =
Vi 1 1 sL L
1+ + + +
sR 1 C sR 2 C R 2 R 1 R 2 C
TM

Vo sL
\ =
Vi sR 1 R 2 C + R 2 + R 1 + s 2 L R 1 C + sL
Vo (s) sL
\ =
Vi (s) 2
s LR 1 C + s[L + R 1 R 2 C] + (R 1 + R 2 )
Example 6.7.5
Solution : Laplace transform of the given network is,
R1 R3
V1
Vi(s) R2 R4 Vo(s)
I1(s) I2(s)
Equations for different currents and voltages are
1
R1 I1 V1 1
S.F.G. (I) Vi I1 = ( Vi - V1) ´ ... (1)
R1
–1
R1
S.F.G. (II) I1 R2 V1 I2 V1 = (I1 - I 2 ) R 2 ... (2)
–R2
1
V1 R3 I2 Vo
1
S.F.G. (III) I 2 = (V1 - Vo ) ´ ... (3)
–1 R3
R3
S.F.G. (IV) I2 R4 Vo Vo = I 2 R 4 ... (4)
Total S.F.G. is as shown in the following figure.
1 1
R1 I1 R 2 V1 R 3 I2 R4
Vi Vo
–1 –R2 –1
R1 R3
TM

Vo
Use Mason's gain formula to find
Vi
Vo S TK D K
=
Vi D
T1 D 1
Number of forward paths = K = 1 =
D
R2 R4
\ T1 =
R1 R 3
R2 R2 R4
L1 = - , L2 = – , L3 = –
R1 R3 R3
L 1 and L 3 are non-touching
D = 1 – [L 1 + L 2 + L 3 ]+ [L 1 L 3 ]
R R R R R
= 1+ 2 + 2 + 4 + 2 4 …All loops are touching to T1
R1 R 3 R 3 R1 R 3
\ D1 = 1
R1 R 3 + R2 R 3 + R1 R2 + R1 R4 + R2 R4
D =
R1 R 3
Vo T D
= 1 1
Vi D
Vo R2 R4
\ =
Vi R2 R4 + R1 R4 + R1 R2 + R2 R 3 + R1 R 3
Example 6.7.6
Solution : The Laplace domain representation of the given network is shown below.
The various branch currents are shown,
I2(s) R1 I2(s)
(I1+I2) V1(s)
I1(s) I2(s)
1 1
Vi(s) sC sC Vo(s)
Vi(s) Vo(s)
R
(I1+I2)
(I1+I2)
Vi (s) - V1 (s)
\ I 1 (s) = = sC Vi (s) - sC V1 (s) ... (1)
æ 1 ö
ç sC ÷
è ø
Vi (s) – Vo (s) 1 1
Then, I 2 (s) = = V (s) – Vo (s) ... (2)
R R i R
TM

V1 (s) = [I 1 (s) + I 2 (s)]R = R I 1 (s) + R I 2 (s) ... (3)

Vo (s) - V1 (s)
and also, I 2 (s) = = sC Vo (s) - sC V1 (s)
1
sC
From this obtain the equation for Vo (s) as I 2 (s) equation is already obtained.
Note Write the separate equation for separate branch and each element must be
considered at least once.
1
\ Vo (s) = Vi (s) +I (s) ... (4)
sC 2
Hence the signal flow graph is,
1 –1
R R R
1
sC R sC
Vi(s) V1(s) Vo(s)
I1(s) I2(s)
–sC
+1
The forward path gains are,

1 1
T1 = sCR T2 = T3 = ´ R´ 1 = 1
sCR R
The various loop gains are,
1 1
L1 = – sCR L2 = - L3 = - ´ R´ 1 = – 1
sCR R
The loops L 1 and L 2 are non touching.
\ L1L2 = 1
Hence system determinant is,
D = 1 - [L 1 + L 2 + L 3 ] + [L 1 L 2 ]
1 3sCR + s 2 C 2 R 2 + 1
= 1 + sCR + +1 +1 =
sCR sCR
For T1 , D1 = 1 … All loops touching to T1
For T2 , D2 = 1 – L1 … As L 1 is non touching to T2
= (1 + sCR)
For T3 , D3 = 1 … All loops touching to T3
According to Mason’s gain formula,
Vo (s) T D + T2 D 2 + T3 D 3
= 1 1
Vi (s) D
TM

æ s 2 C 2 R 2 + 1 + sCR + sCR ö
sCR + æç
1 ö ç ÷
÷(1 + sCR) + 1 ç sCR ÷
è sCR ø è ø
= =
æ 3sCR + s 2 C 2 R 2 + 1 ö æ 3sCR + s 2 C 2 R 2 + 1 ö
ç ÷ ç ÷
ç sCR ÷ ç sCR ÷
è ø è ø
Vo (s) s 2 C 2 R 2 + 2sCR + 1
\ Vi (s)
=
s 2 C 2 R 2 + 3sCR + 1
Example 6.7.7
Solution : For the given current distributions,
V - V2 1 1
For branch r1 , i1 = 1 = V - V … (1)
r1 r1 1 r1 2
For branch r 3 , V2 = (i 1 - i 2 ) r 3 = i 1 r 3 - i 2 r 3 … (2)
Now current through r2 is (i 2 - a i 1 ) . Hence according to Ohm's law,
V2 - V 3
i2 - a i1 =
r2
1 1
\ i2 = V - V + a i1 ...(3)
r2 2 r2 3
And V 3 = i 2 r4 … (4)
The simulations of all the equations are,
a
1
V1 r1 i1 V2 i1 r3 V2 i2 V2 i2 V3 i2 r4 V3
i1
1
r2
1 – r3 1
– –
r1 r2
Equation (1) Equation (2) Equation (3) Equation (4)
Thus the overall signal flow graph is as shown, a
There are two forward paths, 1 1

V1 r1 i1 r3 V2 r2 i2 r4 V3
r r
T1 = 3 4
r1 r2
a r4 1 – r3 1
T2 = – –
r1 r2
r1
The various loops and loop gains are,
a
1
r3 r2 r4
1 – r3 1 1 – r3
– – –
r1 r2 r1
TM

- r3 - r3 - r4 a r3
L1 = L2 = L3 = L4 =
r1 r2 r2 r1
One combination of two non-touching loops is L 1 L3 .
\ D = 1 - [L 1 + L 2 + L 3 + L 4 ] + [L 1 L 3 ]
All loop are touching to all the forward paths,
\ D1 = D2 = 1
r 3 r4 a r4
+
V3 T1 D 1 + T2 D 2 r1 r2 r1
\ = =
V1 D 1 - [L 1 + L 2 + L 3 + L 4 ]+ [L 1 L 3 ]
r 3 r4 a r4
+
r1 r2 r1
=
r r r ar r r
1+ 3 + 3 + 4 - 3 + 3 4
r1 r2 r2 r1 r1 r2
V3 r 3 r4 + a r4 r2
\ =
V1 r1 r2 + r 3 r2 + r 3 r1 + r1 r4 - a r2 r 3 + r 3 r4
qqq
TM

7 State Variable Models
Example 7.4.4
Solution : Applying KVL to the two loops,
di di 1 R 1 1
– L1 1 – i1 R1 – eT + VC = 0 i.e. = – 1 i1 + V - e …(1)
dt dt L1 L1 C L1 T
di 2 di 2 R 1
– L2 – i 2 R2 + V C = 0 i.e. = – 2 i2 + V …(2)
dt dt L2 L2 C
dVC
And (i1 + i2) = – C … Polarities of VC are opposite
dt
dVC 1 1
\ = – i1 - i2 …(3)
dt C C
Selecting state variables as X1 = i1, X2 = i2, X3 = VC, U = eT
· R 1 1
\ X 1 = – 1 X1 + X - U
L1 L1 3 L1
· R2 1
\ X2 = – X + X
L2 2 L2 3
· 1 1
\ X3 = – X - X
C 1 C 2
é R1 1 ù
ê- L 0
L1 ú é- 1 ù
ê 1 ú ê L1 ú
ê R2 1 ú ê ú
\ A=ê 0 - ú,B=ê 0 ú
ê L2 L2 ú
ê ú
ê ú ê 0 ú
ê-1 -
1
0 ú ë û
êë C C úû
·
where state mode is X = AX + BU
(7 - 1)
TM

Control System Engineering 7-2 State Variable Models
Example 7.4.5
Solution : i) For transfer function model,
obtain its s-domain network as shown in the
R 1/sC
Fig. 7.1. Applying KVL
+ – + –
V(s) = I( s) éR + + sLù
1
+
ëê sC ûú V(s) + sL Y(s)
–
I(s) –
and Y(s) = I( s) ´ sL
V( s)
= ´ sL
éR + 1 + sLù Fig. 7.1
ëê sC ûú
V( s) s 2 LC
=
s 2 LC + sRC + 1
Y( s) s2 … T.F. model
=
\ V( s) R 1
s2 + s +
L LC
ii) For state variable model, apply KVL in time domain

L di( t ) di( t )
v( t ) = i( t )R + e( t ) + and y(t) = L
dt dt
Let i(t) = X1 and e(t) = X2 , v(t) = U

di( t ) R 1 1 ·
\ = - X1 - X2 + U = X1
dt L L L
de( t ) de( t ) 1 ·
and C = i( t ) i.e. = X1 = X2
dt dt C
·
While y(t) = L X 1 = -RX 1 - X 2 + U
·
Hence the state variable model is X = AX + BU and
Y = CX + DU
é- R - ù
1
é1 ù
ê Lú , B = êL ú , C = [-R -1] , D = [1]
where, A=ê L ú
ê
1
0 ú ê0 ú
ë C û ë û
TM

Example 7.4.6
Solution :
R1
i) For transfer function model refer
+ –
example 3.3.1 (Chapter - 3). +
vC C
–
E o ( s) 1 + s CR 2
= ei(t) + eo(t)
\ E i ( s) 1 + sC [R 1 + R 2 ] – +
i(t) R2
–
ii) For state variable model consider network
shown in the Fig. 7.2. Fig. 7.2
Applying KVL,
– i( t )R 1 - v C ( t ) - i( t )R 2 + e i ( t ) = 0
ei ( t) - v C ( t)
\ i( t ) = … (1)
R1 + R2
dv C ( t ) dv C ( t ) i( t ) -v C ( t ) ei ( t)
C = i( t ) i.e. = = + … (2)
dt dt C C (R 1 + R 2 ) C (R 1 + R 2 )
Let X1 = v C ( t ) and e i ( t ) = U
· –X 1 U ·
X1 = + i.e. X = AX+BU … (3)
C(R 1 + R 2 ) C(R 1 + R 2 )
é e (t) - v C (t)ù
\ e o ( t ) = v C ( t ) + i( t )R 2 = v C ( t ) + ê i ú R2
ë R1 + R2 û
R1 R2 R1 R2
\ eo ( t) = v C (t) + ei ( t) = X1 + U … (4)
R1 + R2 R1 + R2 R1 + R2 R1 + R2
\ Y = CX + DU
The equations (3) and (4) give the state variable model.
Example 7.4.7
Solution : Assume the loop currents as
shown in following figure R = 1 M W V2(t) R = 1 MW V1(t)
Applying KVL to first loop, +
+ – + –
- I 1 R - V2 + u( t ) = 0 –
+ +
u( t ) - V2 u(t)
+
C = 1 mF C = 1 mF
i.e. I1 = … (1) – –
R I1(t) I2(t)
Applying KVL to second loop,
–
TM

V2 - V1
- I 2 R - V1 + V2 = 0 i.e. I2 = … (2)
R
Now current through capacitors are I1 – I2 and I2,
dV2 dV1
\ I1 – I 2 = C … (3) and I2 = C … (4)
dt dt
Substituting I1 and I2,
dV2 u( t ) - V2 V2 - V1 u( t ) - 2V2 + V1
C = - = … (5)
dt R R R
dV1 V2 - V1
C = … (6)
dt R
Now RC = 1 ´ 10 6 ´ 1 ´ 10 -6 = 1
· dV1 · dV2
Using u(t) = U, X1(t) = V1, X2(t) = V2, X1 = , X2 =
dt dt
· U + X 1 - 2X 2 · U + X 1 - 2X 2
\ CX 2 = i.e. X2 = = U + X1 – 2X2 … (7)
R RC
· X2 - X1 · -X 1 + X 2
and CX 1 = i.e. X1 = = – X 1 + X2 … (8)
R RC
Ans Y(t) = V1(t) = X1(t)
The equations (7), (8) and (9) gives the required state model.
é· ù -1 1 ù é X 1 ù é0ù
êX 1 ú = é éX ù
\ + U and Y(t) = [1 0] ê 1 ú
ê ú ë 1 -2úû êëX 2 úû êë1 úû
· ê
ëX 2 û
ëX 2 û
Example 7.4.8
Solution : Applying KVL to the first loop,
di (t) di 1 (t) 1 1
– L1 1 - VC (t) + e i (t) = 0 i.e. =– V (t) + e (t) …(1)
dt dt L1 C L1 i
Applying KVL to the second loop,

di 2 (t) di 2 (t) R 1
– L2 – i2(t) R2 + VC(t) = 0 i.e. = – 2 i2(t) + V (t) …(2)
dt dt L2 L2 C
dVC (t) dVC (t) 1 1
i1(t) – i2(t) = C i.e. = i (t) – i (t) …(3)
dt dt C 1 C 2
eo(t) = i2(t) R2 ... (4)
Use the state variables as, X1 = i1(t), X2 = i2(t), X3 = VC(t)
Using in the equations (1) to (4), we get
TM

é 1 ù é ù
é· ù ê0 -
L1 ú
0 ê1 ú
êX 1ú ê ú éX1 ù êL ú
·
êX 2 ú = ê 0 - R 2 1 ú êX ú + ê 1 ú U … U = ei(t)
ê· ú ê L2 L2 ú ê 2ú ê 0 ú
êX 3 ú ê1 ú êëX 3 úû ê ú
1
ë û ê - 0 ú ê 0 ú
êë C C úû ë û
éX1 ù
Y = [0 1 0] ê X 2 ú … Y = eo(t)
ê ú
êëX 3 úû
·
This is the required state model in the form, X = AX + BU and Y = CX.
Example 7.7.4
Solution : System is 3 rd order, n = 3
3 integrators and variables are required. Select y = X 1 and then successive differentiation
of y as next variable.
·
\ X 1 = X 2 = dy/dt ... (1)
· d2 y
X2 = X3 = ... (2)
dt 2
· ·
Now as 3 variables are defined, X 3 ¹ X 4 but X 3 must be obtained by substituting all
selected variables in original differential equation.
· · d3y
\ X 3 + 6X 3 + 11X 2 + 10 X 1 = 3U as X3 =
dt 3
·
\ X 3 = 3U – 10X 1 – 11X 2 – 6X 3 ... (3)
y = X1 which is output equation.

\ State model can be written as,
·
X = AX + BU
and Y = CX
TM

é 0 1 0ù é0 ù
where A=ê 0 0 1 ú , B = ê0 ú , C = [ 1 0 0 ]
ê ú ê ú
êë– 10 –11 –6úû êë 3úû
State diagram :
X3 X3 X2 X1
U(s) 3 Y(s)
– –
–
11
10
Example 7.7.5
Solution : Dividing both N and D by s 2 ,
3 4
+
Y (s) s s2
=
U (s) 5 6
1+ +
s s2
é 5 6 ù 5 6
D = 1 - ê- - \ L1 = - , L2 = -
s 2ú s
ë s û s2
The numerator is T1 D 1 + T2 D 2 , i.e. K = 2 = Forward paths
3 4
T1 = , T2 = , D1 = D2 = 1
s s2
3
1
__ 1
__
1 X2 s X2 s 4 1
U(s) Y(s)
X1
X1
–5
–6
Fig. 7.3
Complete signal flow graph is,

·
\ X 1 = X2
·
X 2 = U (s) - 6 X 1 - 5 X 2
Y = 4 X1 + 3 X2
TM

·
\Model is, X = AX + BU and Y = CX + DU
é0 1ù é0ù
where A=ê ú , B = ê1 ú , C = [4 3], D = [0]
ë- 6 - 5 û ë û
Example 7.7.6
Solution : The closed loop transfer function of the system is,
K (s + z)
Y(s) s (s + a)(s + p) Ks + Kz
= =
U(s) K(s + z) s + (a + p) s 2 + (ap + K) s + Kz
3
1+
s (s + a)(s + p)
Decompose the denominator as,

{([s + a + p] s + (ap + K)) s + Kz}
For numerator shift take off point once for Ks. The entire state diagram is shown in the
Fig. 7.4.
K
+
U(s) X3 X2 X1 X1 +
1/s 1/s 1/s Kz Y(s)
X3 X2
– – –
a+p
ap + K
Kz
Fig. 7.4
· · ·
\ X 1 = X 2, X 2 = X3, X 3 = U – Kz X1 – (ap + K) X2 – (a + p) X3
while Y = Kz X1 + KX2
·
Hence, X = AX + BU and Y = CX + DU where,
é 0 1 0 ù é0ù
ê
A= 0 0 1 ú , B = ê0ú , C = [Kz K 0], D = [0]
ê ú ê ú
êë-Kz -(ap + K) -(a + p) úû êë1 ûú
Example 7.7.7
Solution : By direct decomposition,
y( s) 5
=
u( s ) {[( ) + 6]s + 7}
s + 0 s
TM

X3 X3 = X 2 X2 = X 1
Hence the state diagram is
u(s) X1
as shown in the Fig. 7.5. 1/s 1/s 1/s 5 y(s)
Assigning output of each – –
integrator as the state
variable, 6
7
Fig. 7.5
· · ·
X 1 = X 2 , X 2 = X 3 , X 3 = – 6 X 2 – 7 X 1 + U, Y = 5X 1
·
\ X = AX + BU
and Y = CX + DU where
é0 1 0ù é0ù
A = ê0 0 1 ú , B = ê0ú , C = [5 0 0], D = [0]
ê ú ê ú
êë–7 –6 0úû êë1 úû
Example 7.8.3
Solution : Finding factors of denominator
Y ( s) 6 A B C
= = + +
U ( s) ( s + 1)( s + 2)( s + 3) s + 1 s + 2 s + 3
Taking partial fractions,
Y ( s) 3 6 3
= - + U(s) X1 X1
U ( s) s + 1 s + 2 s + 3
–
ò 3
Hence the state diagram is

as shown in the Fig. 7.6.
This is Foster's form of X2 X2 +
representation. From –
ò 6
–
Y(s)
+
Fig. 7.6, we get,
2
·
X 1 = U ( s) - X 1 , X3 X3
· –
ò 3
X 2 = U ( s) - 2 X 2
· 3
X 3 = U ( s) - 3 X 3
Fig. 7.6
TM

and Y (s) = 3X 1 - 6X 2 + 3X 3
Hence the state space model is,
·
X = AX + BU and Y = CX
é-1 0 0 ù é1ù
where A = 0 -2 0 , B = ê1ú ,
ê ú C = [3 -6 3]
ê ú ê ú
êë 0 0 -3úû êë1ûú
Example 7.10.3
Solution : T.F. = C [sI – A] –1 B
Adj [sI – A]
[sI – A ]–1 = |sI – A|
é1 0ù é-2 -3ù é s 0ù é -2 -3ù
[sI – A] = s ê ú -ê ú=ê ú -ê ú
ë0 1 û ë 4 2 û ë0 s û ë+ 4 2 û
és + 2 3 ù
[sI – A] = ê
ë–4 s – 2úû
éC C 12 ù T és – 2 4 ùT és – 2 –3ù
Adj = ê 11 ú =ê– 3 ú =ê
ëC 21 C 22 û ë s + 2û ë 4 s + 2úû
|sI – A| = (s + 2) (s – 2) + 12 = s 2 – 4 + 12 = s 2 + 8
és – 2 –3 ù
ê 4 s + 2úû
[sI – A] –1 = ë
s2 + 8
és – 2 –3 ù é 3ù é 3s – 21ù
[1 1] ê ú ê ú ê5s + 22 ú
ë 4 s + 2û ë5 û [8s + 1]
\ T.F. = =[1 1] ë û =
s2 + 8 s2 + 8 s2 + 8
Example 7.10.4
Solution : The transfer function is given by,
T(s) = C [sI - A] - 1 B
é1 0 0ù é- 1 0 0 ù és + 1 0 0 ù
ê ú ê
[sI – A] = s 0 1 0 - 0 - 1 0 = 0 ú ê s+ 1 0 ú
ê ú ê ú ê ú
êë0 0 1 úû êë 0 0 - 3úû êë 0 0 s + 3úû
T
és 2 + 4s + 3 0 0 ù
ê ú
Adj [sI – A] = [co factor [sI - A]] T =ê 0 2
s + 4s + 3 0 ú
ê 0 0 s 2 +
2s + 1ú
ë û
TM

Control System Engineering 7 - 10 State Variable Models
és 2 + 4s + 3 0 0 ù
ê 2 ú
= ê 0 s + 4s + 3 0 ú
ê 0 0 2
s + 2s + 1ú
ë û
Adj [sI - A]
|sI – A| = (s + 1) 2 (s + 3) and [sI - A] - 1 =
|sI - A|
és 2 + 4s + 3 0 0 ù é 3ù
ê 2 ú ê- 6ú
[1 1 1] ê 0 s + 4s + 3 0 ú ê ú
ê 0 0 2
s + 2s + 1ú êë 3 úû
\ T(s) = ë û
(s + 1) 2 (s + 3)
é 3 (s 2 + 4s + 3) ù
ê ú
[1 1 1] ê- 6 (s 2 + 4s + 3) ú
ê 3 (s 2 + 2s + 1) ú 2 2 2
= ë û = [3s + 12s + 9 - 6s - 24s - 18 + 3s + 6s + 3]
(s + 1) 2 (s + 3) (s + 1) 2 (s + 3)
- 6s - 6 - 6 (s + 1)
\ T(s) = =
(s + 1) 2 (s + 3) (s + 1) 2 (s + 3)
Example 7.10.5
Solution : The T.F. is given by,
és + 5 6ù
T.F. = C[sI - A ] -1 B + D, [sI - A ] = ê
ë -1 s úû
éC C 12 ù T és 1 ùT és -6 ù
Adj [sI - A ] = ê 11 ú = ê ú =ê ú
ëC 21 C 22 û ë-6 s + 5û ë1 s + 5û
|sI - A| = s (s + 5) + 6 = s 2 + 5s + 6 = ( s + 2)( s + 3)
és -6 ù é1 ù ésù
[1 1 ]ê ú ê0ú [1 1]ê1ú
C Adj [sI - A ]B ë1 s + 5 ûë û = ë û
\ T.F. = =
|sI - A| ( s + 2)( s + 3) ( s + 2)( s + 3)
( s + 1)
\ T.F. =
( 2)( s + 3)
s +
Example 7.16.6
Solution : Use Laplace transform method,
é1 0ù é 0 - 1ù é s + 1ù
[sI – A] = s ê ú -ê ú= ê-2 s + 3ú
ë0 1 û ë+ 2 -3 û ë û
TM

T
és + 3 2ù és + 3 -1ù
\ Adj [sI – A] = ê ú =ê
ë -1 s û ë 2 s úû
é s + 1ù 2
|sI – A| = ê ú = s + 3s + 2 = (s + 1) (s + 2)
ë-2 s + 3 û
és + 3 -1ù
Adj [sI - A] ê 2 s úû
= ë
–1
\ [sI – A] =
|sI - A| ( s + 1) ( s + 2)
é s+ 3 -1 ù
ê(s + 1) (s + 2) (s + 1) (s + 2) ú
=ê ú
–1
\ f (s) = [sI – A]
ê 2 s ú
ê (s + 1)(s + 2) (s + 1)(s + 2) úû
ë
é s+ 3 -1 ù
ê(s + 1) (s + 2) (s + 1) (s + 2) ú
\ e
At
= L
–1
[sI – A]
–1
=L
–1 ê ú
ê 2 s ú
ê(s + 1) (s + 2) (s + 1) (s + 2) úû
ë
Using partial fraction expansion for all the elements.
é 2 - 1 -1 1 ù
+
At –1 ês + 1 s + 2 s +1 s + 2ú é 2e - t - e -2t -e - t + e -2t ù
e = L ê ú= ê -t -2t ú
ê 2 - 2 -1 2 ú êë2e - 2e -e - t + 2e -2t úû
êë s + 1 s + 2 +
s + 1 s + 2 úû
Example 7.16.7
é- 2 3ù
Solution : From the given state equation, A = ê ú
ë 0 - 3û
eAt = L- 1 {(sI - A) - 1} = State transition matrix
é1 0ù é- 2 3ù és + 2 - 3 ù
[sI – A] = s ê ú -ê =
- 3 úû êë 0 s + 3úû
ë0 1 û ë 0
éC 11 C 12 ù T és + 3 0 ùT
Adj [sI - A] êC C 22 úû ê 3 s + 2úû
[sI - A] - 1 = = ë 21 = ë
sI - A sI - A s+2 - 3
0 s+ 3
és + 3 3 ù
é 1 3 ù
ê 0 + 2úû ê s + 2
ë s (s + 2) (s + 3) ú
= =ê ú
(s + 2) (s + 3) 1
ê 0 ú
ë (s + 3) û
TM

ìé 1 3 ùü
At -1 -1 -1
ïï ê(s + 2) (s + 2) (s + 3) ú ïï
\ e = L [(sI - A) ]= L íê úý
1
ïê 0 úï
ïî ë (s + 3) û ïþ
é 1 æ 3 3 öù
ê ç - ÷ú
= L- 1 ê(s + 2) è (s + 2) (s + 3) øú
ê 0 1 ú
ë s+ 3 û
ée - 2t 3e - 2t - 3e - 3t ù
\ eAt = ê ú
êë 0 e - 3t úû
Example 7.16.8
Solution : From the given state model
é- 1 1 0ù
ê
A = 0 -1 1 ú
ê ú
ëê 0 0 - 1úû
ì Adj [sI - A ] ü
eAt = L- 1 { [sI - A - 1 ] } = L- 1 í ý
î sI - A þ
é1 0 0ù é- 1 1 0 ù és + 1 - 1 0 ù
ê ú ê
[sI – A] = s 0 1 0 - 0 - 1 1 = 0 ú ê s +1 - 1 ú
ê ú ê ú ê ú
êë0 0 1 úû êë 0 0 - 1úû êë 0 0 s + 1úû
é(s + 1) 2 0 0 ù
T ê ú
Adj [sI – A] = [Cofactor matrix] = ê (s + 1) (s + 1) 2 0 ú
ê 1 (s + 1) (s + 1) 2 ú
ë û
sI - A = (s + 1) 3
é 1 1 1 ù
ê(s + 1) ú
ê (s + 1) 2 (s + 1) 3ú
1 1 ú
\ [sI - A ] -1
= ê 0
ê (s + 1) (s + 1) 2 ú
ê ú
ê 0 1 ú
0
êë (s + 1) úû
é 1 ù é 1 ù 1
L- 1 æç
1 ö -t
÷ = e , L- 1 ê ú = t e- t , L- 1 ê ú = t 2 e- t
è +1 ø
s êë( s + 1) 2 úû êë( s + 1) 3 úû 2
TM

ée - t t e- t
1 2 -tù
t e
ê 2 ú
\eAt = L- 1 [sI - A ] - 1 =ê 0 e- t t e- t ú ... State transition matrix
ê ú
ê 0 0 e- t ú
ë û
Example 7.16.9
Solution : The matrix A is,
é-3 1 ù
A = ê ú
ë 0 -1û
The state transition matrix is,
eAt = L-1 [sI - A ] -1
é1 0ù é-3 1 ù és + 3 -1 ù
sI – A = s ê ú -ê ú=ê s + 1úû
ë0 1 û ë 0 -1û ë 0
és + 1 0 ù T és + 1 1 ù
Adj [sI – A] = ê ú =ê
ë 1 s + 3û ë 0 s + 3úû
|sI – A| = (s + 1) (s + 3)
é ( s + 1) 1 ù é 1 1 ù
-1 Adj [sI - A ] ê( s + 1)( s + 3) ( s + 1)( s + 3) ú ê s + 3 ( s + 1)( s + 3) ú
[sI - A ] = =ê ú=
s I -A ê ( s + 3) ú ê 1 ú
0 ê 0 ú
êë ( s + 1)( s + 3) úû ë s +1 û
Finding the partial fractions,

é 1 0.5
-
0.5 ù
ê( s + 3) ( s + 1) ( s + 3) ú
[sI - A ] -1 = ê
1 ú
ê 0 ú
êë ( s + 1) úû
ée -3t 0.5 e - t - 0.5 e -3t ù
eAt = L-1 [sI - A ] -1 = ê ú
êë 0 e- t úû
Example 7.16.10
Solution : System is homogeneous so ZSR = 0
\ X(t) = eAt X(0)
éA A2ù
Let A = ê 1
ëA 3 A 4 úû
é e – 2t ù · é – 2e – 2 t ù
Now X(t) = ê ú hence X(t) = ê ú
êë– 2 e –2 t úû êë 4e – 2 t úû
TM

é– 2ù
and at t = 0 X(t) = ê ú
ë 4 û
· é– 2ù é 1ù
So X(0) = ê ú when X(0) = ê ú
ë 4 û ë– 2û
ée–t ù · é– e – t ù · é– 1ù
Similarly X(t) = ê ú, X(t) = ê ú at t = 0 X(t) = ê ú
êë– e – t úû êë e – t úû ë+ 1û
· é– 1ù é+ 1ù
So X(0) = ê ú when X(0) = ê ú
ë+ 1û ë– 1û
·
Substituting in X(t) = A X(t)
é–2ù éA 1 A 2 ù é 1ù é–1ù éA A 2 ù é 1ù
\ ê 4ú = êA and ê ú = ê 1
ë û ë 3 A 4 úû êë–2úû 1
ë û ëA 3 A 4 úû êë–1úû
\ A 1 – 2A 2 = – 2, A 3 – 2A 4 = + 4 ... (1 and 2)
+ A 1 – A 2 = – 1, A 3 – A4 = 1 ... (3 and 4)
Solving these 4 equations simultaneously,
A 2 = 1, A 1 = 0, A 4 = – 3, A3 = – 2
é 0 1 ù
\ A = ê ú
ë– 2 – 3û
Example 7.16.11
é s 0ù é 0 1 ù é s –1 ù
Solution : [sI – A] = ê ú -ê ú=ê ú
ë0 s û ë-2 0û ë2 s + 3û
éC C 12 ù T és + 3 –2ù T és + 3 1ù
Adj [sI – A] = ê 11 ú =ê ú =ê
ëC 21 C 22 û ë 1 sû ë –2 súû
|sI – A| = s 2 + 3s + 2 = (s +1) (s + 2)
\ eAt = L–1 [(sI – A) –1 ]
é s+ 3 1 ù
ê(s + 1) (s + 2) (s + 1) (s + 2) ú
= ê ú = f (s)
ê –2 s ú
ëê(s + 1) (s + 2) (s + 1) (s + 2) úû
Finding partial fractions,
é 2 – 1 1
–
1 ù
ê s+1 s+ 2 s+1 s+ 2 ú
eAt = L–1 ( f(s)) = ê ú
ê –2 + 2 –1
+
2 ú
êë s + 1 s+ 2 s+1 s + 2 úû
TM

é 2e – t – e – 2 t e – t – e – 2t ù
= ê ú
êë– 2e – t + 2e – 2 t – e – t + 2e – 2t úû
é1 ù é 2e – e
–t – 2t ù
ZIR = eAt X(0) = eAt ê ú = ê ú
ë0û êë – 2e - t + 2e – 2 t úû
To find ZSR = L–1 { f(s) B U(s) }
U(t) = Unit step \U(s) = 1/s
ìé s+ 3 1 ù ü
ï ê(s + 1) (s + 2) (s + 1) (s + 2) ú é0ù ï
ï ï
\ ZSR = L–1 í ê ú ê ú [1 / s]ý
ïê –2 s 5
ú ë û ï
ïî êë(s + 1) (s + 2) (s + 1) (s + 2) úû ïþ
ì 5 ü ì 2.5 – 5 + 2.5 ü
ïï s (s + 1) (s + 2) ïï ï
–1 ï s s + 1 s + 2 ïï é2.5 – 5e – t + 2.5 e – 2 t ù
= L–1 í ý = L í ý =ê ú
ï 5 ï ï 5 5 ï ê 5 e –t –
5e – 2t úû
– ë
ïî (s + 1) (s + 2) ïþ ïî s + 1 s + 2 ïþ
é2.5 – 3e –t + 1.5 e – 2 t ù
\ X(t) = ZIR +ZSR = ê ú
êë 3 e – t – 3e – 2 t úû
é 0 1ù
\ Y(t) = ê ú X(t)
ë–2 –3û
é 0 1ù é2.5 – 3e + 1.5 e
–t – 2t ù é 3e – t – 3e – 2t ù
é Y1 (t) ù
êY (t) ú = ê–2 –3ú ê ú =ê ú
ë 2 û ë û êë 3 e – t – 3e – 2 t úû êë – 5 + 6e – 2t – 3e – t úû
\ Y1 (t) = 3 (e – t – e – 2t ) and Y2 (t) = – 5 + 3 (2e –2t – e –t )
These are the outputs for unit step input applied.

Example 7.16.12
Solution : Given system is homogeneous whose solution is X(t) = eAt X(0)
Now eAt = L–1 {(sI – A) – 1}
é1 0ù é 0 1 ù é s –1ù
[sI – A] = s ê ú – ê– 2 0ú = ê2 s ú
ë0 1 û ë û ë û
Adj (sI – A)
(sI – A) –1 =
|sI – A|
éC C 12 ù T és –2ù
T
é s 1ù
Adj (sI – A) = ê 11 = ê1 s ú = ê–2 sú and |sI – A| = s 2 + 2
ëC 21 C 22 úû ë û ë û
TM

é s 1 ù é s 1ù
ês 2 + 2 s + 2ú
2 ês 2 + 2 s + 2ú
2
(sI – A) –1 = ê ú i.e. eAt = L–1 ê ú
ê –2 s ú ê –2 s ú
ê 2 ú ê 2 ú
ës + 2 s2 + 2û ës + 2 s2 + 2û
ìï s üï ì
–1 ï s üï
L–1 í 2 ý = L í 2 2ý
= cos 2t
îï s + 2 þï îï s + ( 2) ïþ
ìï 1 üï ì
–1 ï 1 2 üï 1
L–1 í 2 ý = L í × ý = sin 2 t
2 2
ïî s + 2 ïþ ïî 2 s + ( 2) ïþ 2
ìï –2 üï –1 ï
ì 2 üï
L–1 í 2 ý = L í – 2 × 2 ý =– 2 sin 2t
îï s + 2 þï îï s + ( 2) 2 þï
é 1
sin 2 tù
At ê cos 2 t ú
\ e = 2
ê – 2 sin 2 t ú
êë cos 2 t úû
é 1
sin 2 t ù
At At é1ù ê cos 2 t + ú
\ X(t) = e X(0) = e ê1ú = ê 2
ë û ú
êë cos 2 t – 2 sin 2 t úû
\Output response
écos 2 t + 1 sin 2 t ù
Y(t) = [1 – 1] X(t) = [1 – 1] ê 2 ú
ê ú
êë cos 2 t – 2 sin 2 t ûú
3
Y(t) = sin 2 t is the required response.
2
qqq
TM

The Performance of
9 Feedback Control Systems
Solutions of Examples For Practice
Example 9.5.7
K
Solution : G( s)H( s) =
s ( s + 1)( s + 4)
2
Lim Lim K
Kp = s ® 0 G( s)H( s) = s®0 =¥
s ( s + 1)( s + 4)
2
Lim Lim sK
Kv = s ® 0 sG( s)H( s) = s®0 =¥
s ( s + 1)( s + 4)
2
Lim Lim s2K K

Ka = s ® 0 s 2 G( s)H( s) = s®0 =
s2 ( s + 1)( s + 4) 4
18 2 A
r( t ) = 1 + 8t + t = A 1 + A 2t + 3 t2
2 2
A1 A A 1 8 18 72
\ e ss = e ss1 + e ss2 + e ss3 = + 2 + 3 = + + = 0+0 +
1 + Kp Kv Ka 1+¥ ¥ K K
4
But e ss is to be limited to 0.8
72
\ 0.8 = i.e. K = 90
K
Example 9.5.8
Solution : The static error coefficients are,
Lim Lim 10K 10K
Kp = G(s)H(s) = = = ¥
s®0 s®0 2
s(s + 4)(s + 2s + 5) 0´ 4´5
Lim Lim s ´ 10K 10K

Kv = sG(s)H(s) = = = 0.5 K
s®0 s®0 2
s(s + 4)(s + 2s + 5) 4´5
Lim s 2 10K 0 ´ 10K

Ka = s 2 G(s)H(s) = Lim = = 0
s®0 s®0
s(s + 4)(s 2 + 2s + 5) 4´5
(9 - 1)
TM

Control System Engineering 9-2 The Performance of Feedback Control Systems
Example 9.5.9
Solution : Using superposition principle, consider inputs separately.
a) R acting, TL = 0
R + C(s)
5 100
1 + 0.1s (s + 1) (1 + 0.2s)
–
500
\ G(s)H(s) =
(1 + 0.1s) (s + 1) (1 + 0.2s)
For step input
Kp = Lim G(s)H(s) = 500
s®0
A 10 10
\ e ss1 = where A = magnitude of step = =
1 + Kp 1 + 500 501
b) TL acting, R = 0 hence E(s) = – C(s)
E C(s)
5 100
1 + 0.1s (s + 1) (1 + 0.2s)
– –
TL
As system is not in standard form, error coefficient method cannot be used.
5 + 100 C(s)
–1
1 + 0.1s (s + 1) (1 + 0.2s)
–
TL
– C(s)
TL 100
(s + 1) (1 + 0.2s)
+
–5
1 + 0.1s
100
C(s) (1 + s) (1 + 0.2s)
=
T(s) -5 ö
´æ
100
1-
(1 + s) (1 + 0.2s) çè 1 + 0.1s ÷ø
TM

C(s) 100(1 + 0.1s)

=
T(s) (1 + s) (1 + 0.2s) ´ (1 + 0.1s) + 500
-4
Now T(s) = –ve sign as TL applied with –ve sign.
s
– 400(1 + 0.1s)
\ C(s) =
s[(1 + s) (1 + 0.2s) ´ (1 + 0.1s) + 500]
+400(1 + 0.1s)
but E(s) = – C(s) =
s[(1 + s) (1 + 0.2s) ´ (1 + 0.1s) + 500]
400(1 + 0.1s)
\ e ss2 = Lim sE(s) = Lim
s®0 s®0 [(1 + s) (1 + 0.2s) ´ (1 + 0.1s) + 500]
= é
400 ù + 400
=
êë1 + 500 úû 501
10 400 410
\Total error e ss = e ss1 + e ss2 = + = = 0.8183
501 501 501
Example 9.5.10
Solution : i) This is Type 0 system. It will produce finite steady state error for step input.
20
\ Kp = Lim G( s)H( s) = Lim = 3.3333
s®0 s ® 0 ( s + 2 )( s + 3 )
A 1
\ e ss = = = 0.2307 … Assuming unit step
1 + Kp 1 + 3. 333
ii) This is Type 2 system. It will produce finite steady state error for parabolic input.
s 2 20( s + 1)
\ Ka = Lim s 2 G(s)H(s) = Lim = 2.5
s®0 s®0 s 2 ( s + 2)( s + 4)
A 1
\ e ss = = = 0.4 … Assuming unit parabolic input
Ka 2.5
iii) This is Type 1 system. It will produce finite steady state error for ramp input.
\ Kv = Lim sG(s)H(s) = Lim

(
s( 2.5) s 2 + 2s + 1) = 0.1
s®0
(
s ® 0 s( s + 1 ) s + 5s + 25
2
)
A 1
\ e ss = = = 10 … Assuming unit ramp input
Kv 0.1
Example 9.5.11
K
Solution : G(s)H(s) = … H(s) =1
s( s + 1)( 0.5s + 1)
TM

sK
Kv = Lim s G( s)H( s) = Lim =K
s®0 s®0 s( s + 1)( 0.5s + 1)
i) K = 1.5 , r(t) = 5 t i.e. A=5

A 5
\ ess = = = 3.333
Kv 1.5
ii) ess £ 0.1 , unit ramp i.e. A = 1

A 1 1
\ ess = = i.e. 0.1 £ i.e. K ³ 10
Kv K K
Example 9.5.12
Solution :
20
G(s)H(s) = … H(s) =1
(
s( s + 2) s 2 + 2s + 20 )
i) r(t) = 5, A = 5
20
Kp = Lim G( s )H( s) = Lim =¥
s®0 s®0
(
s( s + 2) s 2 + 2s + 20 )
A 5
\ ess = = =0
1 + Kp 1 + ¥
ii) r(t) = 5 t , A = 5
s ´ 20
Kv = Lim sG( s)H( s) = Lim = 0.5
s®0 s®0
(
s( s + 2) s 2 + 2s + 20 )
A 5
\ ess = = = 10
Kv 0.5
3 2
iii) r(t) = t ,A=3
2
s 2 ´ 20
Ka = Lim s 2 G( s)H( s) = Lim =0
s®0 s®0
(
s( s + 2) s 2 + 2s + 20 )
A 3
\ ess = = =¥
Ka 0
Example 9.5.13
Solution : Unity feedback system \ H(s) = 1
To determine type of the system it is required to bring G(s) H(s) into its time constant
form.
K(s + 2)
G(s) H(s) =
s(s 3 + 7s 2 + 12s)
TM

2K(1 + 0.5s)
=
s2 (s 2 + 7s + 12)
2K(1 + 0.5s)
=
s2 (s + 4) (s + 3)
2K(1 + 0.5s)
=
s 2 × 4 (1 + 0.25 s) 3 × (1 + 0.33s)
2K
(1 + 0.5s)
G(s) H(s) = 12
s 2 (1 + 0.25s) (1 + 0.33s)
Comparing this with,
A(1 + T1 s) (1 + T2 s) .........
G(s) H(s) = Where j = Type of the system
s j (1 + Ta s) (1 + Tb s) .......
\ j = 2, So system is Type 2 system.

Error coefficients calculation :
Lim
Kp = G(s) H(s)
s®0
K
(1 + 0.5s)
= Lim 6 =¥
s ® 0 s 2 (1 + 0.25s) (1 + 0.33s)
Lim
Kv = sG(s) H(s)
s®0
K
(1 + 0.5s)
= Lim 6 =¥
s ® 0 s (1 + 0.25s) (1 + 0.33s)
Lim 2
Ka = s G(s) H(s)
s®0
K
(1 + 0.5s) K
= Lim 6 =
s®0 (1 + 0.25s) (1 + 0.33s) 6
To find error use,
s R(s)
e ss = Lim
s ® 0 1 + G(s) H(s)
where R(s) = Laplace Transform of input r(t) to the system

R 2 R
r(t) = t , R(s) =
2 s3
TM

R
s×
\ e ss = Lim s3
s ® 0 K
(1 + 0.5 s)
1+ 6
s 2 (1 + 0.25s) (1 + 0.33s)
R
= Lim
s ® 0 é K
(1 + 0.5s) ù
s2 ê1 + 6 ú
ê s 2 (1 + 0.25s) (1 + 0.33s) ú
êë ûú
R
= Lim
s ® 0 K
(1 + 0.5s)
s2 + 6
(1 + 0.25s) (1 + 0.33s)
6R
e ss =
K
R
Steady state error can also be determined as e ss = as system is type 2 system, where
Ka
R is magnitude of the input.
R 6R
e ss = =
K K
6
Example 9.5.14
Solution : To determine type of the system arrange G(s) H(s) its time constant form as,
10(1 + s)
G(s) H(s) =
2
s × 2(1 + 0.5s) (10) (1 + 0.1s)
0.5 (1 + s)
=
s2 (1 + 0.5s) (1 + 0.1s)
Comparing this with standard form
K(1 + T1 s) (1 + T2 s) .......
G(s) H(s) = where j = Type
s j (1 + Ta s) (1 + Tb s) ........
Type of system is 2.
Error coefficients are,
Lim
Kp = G(s) H(s)
s®0
Lim 0.5 (1 + s)
=
s ® 0 s (1 + 0.5s) (1 + 0.1s)
2
\ Kp = ¥
TM

Lim
Kv = s G(s) H(s)
s®0
Lim 0.5 (1 + s)
=
s ® 0 s (1 + 0.5s) (1 + 0.1s)
\ Kv = ¥
Lim
Ka = s 2 G(s) H(s)
s®0
Lim 0.5 (1 + s)
=
s®0 (1 + 0.5s) (1 + 0.1s)
\ Ka = 0.5
The steady state error for step input is,
A1
e ss1 =
1+ Kp
where A 1 = magnitude of step input part of total input

Here A1 = 1 and Kp = ¥
1
\ e ss1 =
1+¥
= 0
The steady state error for ramp input is,
A2
e ss2 =
Kv
where A 2 = magnitude of ramp input = 4

A
\ e ss2 =
¥
= 0
The steady stade error for parabolic input is,
A3
e ss 3 =
Ka
where A 3 = magnitude of parabolic input

1
The parabolic input is expressed as t 2 in R(t)
2
\ A3 = 1
A 2
Remember that for parabolic input, A is the magnitude when input is expressed as t .
2
1
\ e ss 3 = = 2
0.5
TM

As the input is combination of all the three inputs, the total error is also the algebraic sum
of the steady state errors due to the individual type of inputs.
\ e ss = e ss1 + e ss2 + e ss3
= 0+0+2
\ e ss = 2
Example 9.5.15
Solution : Simplifying the given system by eliminating minor feedback loop in between
we get,
R(s) G1 C(s)
25 1/2
1 + G1H1
–
R(s) 250 C(s)

2
(s +214s + 40)
–
20
G1 (s + 4) (s + 10) 20 20
Now = = =
1 + G 1 H1 20 (s + 4) (s + 10) + 200s s 2 + 214s + 40
1+ 10s
(s + 4) (s + 10)
Combining all blocks in series and simplifying.
250
So G(s) = , H(s) = 1
2
s + 214s + 40
Now comparing G(s) H(s) with standard time constant form
K(1 + s T1 ) (1 + s T2 ) ......
i.e. it is clear that quadratic denominator can be expressed as
s j (1 + s Ta ) (1 + s Tb ) .......
( 1 + sTa ) (1 + sTb ) hence value of `j' for given system is zero. Hence it is Type Zero
system.
Error coefficients :
Lim Lim 250 25
Kp = G(s) H(s) = =
s®0 s ® 0 s + 214s + 40
2 4
Lim Lim æ 250 ö

Kv = sG(s) H(s) = sç ÷=0
s®0 s ® 0 è s + 214s + 40 ø
2
Lim Lim 2 æ 250 ö

Ka = s 2 G(s) H(s) = s ç ÷=0
s®0 s®0 2
è s + 214s + 40 ø
TM

Now for first input,

10
i) r(t) = 10 \ R(s) = , so A = 10
s
A
\ e ss =
1 + Kp
10 40
= = = 1.379
25 29
1+
4
5
ii) r(t) = 5t, \ R(s) = , so A = 5 ramp
s2
A 5
\ e ss = = =¥
Kv 0
6 2
iii) r(t) = 10 + 5t + t
2
i.e. A 1 = 10 step, A 2 = 5 ramp, A 3 = 6 parabolic
A1 A2 A3 10 5 6
\ e ss = + + = + + =¥
1 + Kp Kv Ka 25 0 0
1+
4
Thus as system is type 0 system, error is finite only for step input.
Example 9.5.16
Solution : From the system shown we can write,
K
G(s) = H(s) = 1
s(s + 1)
The input is r(t) = 0.1 t i.e. ramp of magnitude 0.1. For ramp input K V controls the error.
Lim Lim s. K
\ Kv = sG(s)H(s) = =K
s®0 s ® 0 s(s + 1)
A 0.1
\ e ss = =
Kv K
Maximum e ss allowed is 0.005
0.1
\ 0.005 =
K
0.1
\ K = = 20
0.005
For any value of K greater than 20, e ss will be less than 0.005. Hence the range of value of
K for e ss £ 0.005 is,
20 £ K < ¥
Example 9.5.17
Solution : The input is 2t i.e. ramp of magnitude 2, so Kv will control the error.
Lim Lim 200
\ Kv = sG(s) H(s) = s× ×1 = 25
s®0 s ® 0 s(s + 8)
TM

Control System Engineering 9 - 10 The Performance of Feedback Control Systems
A 2
\ e ss = = = 0.08
Kv 25
This error is to be reduced by 5% of existing value, with new gain of G(s) as K2 instead of
200.
e ss1 = e ss - æç ´ e ss ö÷
5
\ ... reduce by 5%
è 100 ø
5 ´ 0.08
= 0.08 – = 0.076
100
New error is 0.076.
K2
New G(s) = and H(s) = 1 with same input
s(s + 8)
Lim
\ Kv = sG(s) H(s)
s®0
s × K2 K
= = 2
s(s + 8) 8
A
\ e ss1 =
Kv
2 16
= =
æ K2 ö K2
ç 8 ÷
è ø
16
\ 0.076 =
K2
\ K2 = 210.52
So new gain is 210.52
Example 9.5.18
Solution : The Kp, Kv, Ka are given by,
Kp = Lim G(s)H(s), Kv = Lim s G(s)H(s), Ka = Lim s 2 G(s)H(s)

s®0 s®0 s®0
10
i) Kp = ¥, Kv = ¥, Ka = Lim s 2 ´ ´ 0.7 = 7
s®0 s2
ii) Kp = 0.7, Kv = 0 Ka = 0
10
iii) Kp = ¥, Kv = ¥, Ka = ´ 0.8 = 1.6
5
Example 9.7.2
Solution : According to generalized error coefficient method,
K
ess(t) = K0 r(t) + K1r ¢(t) + 2 r ¢¢(t) + …
2!
TM

2
Now r(t) = 1 + 2t + t , r ¢(t) = 2 + 2t, r ¢¢(t) = 2
Lim Lim d F1 (s) Lim d 2 F1 (s)

\ K0 = s®0
F1(s), K1 = s®0
, K2 = s®0
ds ds 2
1 1 ( 0 . 1s + 1) s
F1(s) = = =
1 + G(s) 50 s (1 + 0 . 1s) + 50
1+
s (1 + 0 . 1s)
0 .1 s 2 + s s 2 + 10 s
= =
0 . 1 s 2 + s + 50 s 2 + 10 s + 500
Lim
\ K0 = s®0
F1(s) = 0
dF1 (s) (s 2 + 10s + 500) (2s + 10) - (s 2 + 10s) (2s + 10)

=
ds (s 2 + 10s + 500) 2
(2s + 10) [s 2 + 10 s + 500 - s 2 - 10s] 1000 s + 5000

= =
(s 2 + 10 s + 500) 2 ( s 2 + 10s + 500) 2
Lim d F1 (s) 5000

\ K1 = s®0
= = 0.02
ds (500) 2
d 2 F1 (s) (s 2 + 10s + 500) 2 [1000] - (1000 s + 5000) (2) (s 2 + 10 s + 500) (2 s + 10)

=
ds 2 (s 2 + 10 s + 500) 4
d 2 F1 (s) (500) 2 (1000) - (5000)( 2)(500)(10)

\ K2 = Lim
s®0
= = 3.2 ´ 10– 3
2
ds (500) 4
3 . 2 ´ 10 -3 ( 2)
\ ess(t) = 0.02 (2 + 2t) + = 0.04t + 0.0032
2!
Example 9.9.4
Solution : C(s) 1
Let the first order system has transfer function =
R(s) ( s + T)
1
Input r(t) = Unit ramp input = t hence R(s) =
s2
1 A B C
\ C(s) = = + +
s2 (s + T) s2 s s+T
2 2
\ A (s + T) + Bs (s + T) + C s = 1 i.e. (B + C) s + (A + BT) s + AT = 1
1 1
\ AT = 1 \ A= and A + BT = 0 \ B= -
T T2
TM

1
B+C =0 \ C= c(t), r(t)
T2 r(t) = t
1 T 1 T2 1 T2
\ C(s) = - + 6 S.S. error
s2 s s+T
c(t) for T = 1
4
Taking Laplace inverse,
1 1 1 -T t 2
\ c(t) = æç ö÷ t - + e
T
è ø T 2
T2 Time
0 2 4 6 t
The sketch of c(t) and r(t) is shown in
Fig. 9.1. Take T = 1 for plotting c(t). Fig. 9.1
Example 9.9.5
Solution : Let the transfer function of a first order system is,
C(s) 1 1 2 1
= and r(t) = t i.e. R(s) =
R(s) (s + T) 2 s3
R (s) 1 A B C D
\ C(s) = = = + + +
(s + T) 3
s (s + T) s 3 s 2 s s+T
\ A (s + T) + B s (s + T) + C s 2 (s + T) + D s 3 = 1
C + D = 0, B + C T = 0, A + B T = 0, A T = 1
\ A = 1 , -1 1 -1
B= , C= , D=
T T 2 T 3 T3
(1 / T) (1 / T 2 ) (1 / T 3 ) (1 / T 3 )
\ C (s) = - + -
s3 s2 s s+T
1 2 1 1 1 - Tt
\ C(t) = L-1 [C(s)] = 2T
t -
2
t+
3
- e
T T T3
Example 9.9.6
1
Solution : R(s) =
s
C(s) K K
T(s) = = hence C(s) =
R(s) s+ a s(s + a)
Finding partial fractions, we get
A1 A2 K K
C(s) = + and A 1 = , A2 = -
s s+ a a a
æ Kö æ Kö
ça÷ ça÷
C(s) = è ø - è ø
K K - at
\ i.e. c(t) = - e
s s+ a a a
TM

Lim
The steady state of c(t) is 2 hence c(t) = 2
t®¥
K
\ = 2
a
2.0 dc(t)
While slope at t = 0 is = 1 i.e. =1
2.0 dt t= 0
K
\ – ´ ( -a) e - at = 1
a t= 0
\ K = 1 and a = 1/2
Example 9.14.15
Solution : w2n
Comparing the T.F. with the standard form
s 2 + 2x wn s + w2n
w2n = 25 and 2x wn = 6
wn = 5 \ x = 0.6
é 1 - x2 ù
q = tan -1 ê ú = 0.9272 radians
ê x ú
ë û
wd = wn 1 - x 2 = 5 1 - (0.6) 2 = 4 rad/sec.
p - q p - 0.9272 p p
Tr = = = 0.5535 sec. , Tp = = = 0.785 sec.
wd 4 wd 4
-px 1 - x2 4
% Mp = e ´ 100 = 9.48 % , Ts = = 1.33 sec.
x wn
e- x wn t e - 3t
And c(t) = 1 – sin ( wd t + q) = 1 – sin(4t + 0.9272)
1 - x2 1 - (0.6) 2
\ c(t) = 1 – 1.5625 e - 3t sin (4t + 0.9272)
Example 9.14.16
20
C(s) (s + 1) (s + 4) 20
Solution : = =
R(s) 20 2
s + 5s + 24
1+
(s + 1) (s + 4)
Key Point Now though T.F. is not in standard form, denominator always reflect 2x wn and
wn2 from middle term and the last term respectively.
\ Comparing, s 2 + 5s + 24 with s 2 + 2x wn s + wn2
TM

\ w2n = 24 \ wn = 4.8989 rad/sec.
2 x wn = 5 \ x = 0.51031
wd = wn 1 - x 2 = 4.2129 rad/sec.
C(s)
Now, for c(t) we can use standard expression for in standard form. So writing
R(s)
C(s) 20 ïì 24 ïü
= ×
R(s) 24 íï s 2 + 5s + 24 ýï
î þ
For the bracket term use standard expression, and then c(t) can be obtained by multiplying
20
this expression by constant .
24
é ù
20 e - x wn t
\ c(t) = ê1 - sin ( wd t + q) ú
24 ê ú
ë 1 - x2 û
1 - x2
q = tan - 1 radians = 1.03 radians
x
\ c(t) =
20
24 [
1 - 1.1628 e - 2.5 t sin (4.2129 t + 1.03) ]
Example 9.14.17
Solution : G(s) = K1 , H(s) = 1 + K s
2
s2
K1
G(s) s2 K1
T.F. = = =
1 + G(s)H(s) K1 s 2 + K1 K2 s + K1
1+ (1 + K2 s)
s2
\ Comparing with standard form,
1
wn = K1 , 2x wn = K1 × K2 , \ x= K1 × K2
2
Now, Mp is function of x alone,
1 - x2 1 - x2
\ % Mp = e - p x ´ 100 i.e. 25 = e - p x ´ 100
1 - x2 -px
\ 0.25 = e - p x i.e. ln (0.25) =
1 - x2
-px
\ – 1.3862 = and solving x = 0.4037
1 - x2
p p
Now, Tp = = 4 sec. i.e. =4
wd
wn 1 - x2
TM

p
\ =4 i.e. wn = 0.8584 rad/sec.
wn 1 - (0.4037) 2
Now, wn = K1 i.e. K 1 = w2n = 0.7369

1
and, x = K1 × K2 = 0.4037 i.e. K 2 = 0.9405
2
Example 9.14.18
Solution : System differential equation is,
d2 y dy
+ 8y = 8x
+ 4
dt 2 dt
Y(s)
To find T.F. , take Laplace transform from above equation and neglect initial
X(s)
conditions.
s 2 Y(s) + 4s Y(s) + 8 Y(s) = 8 X(s) i.e. Y(s) [s 2 + 4s + 8] = 8 X(s)
Y(s) 8 w2n
\ T.F. = =
X(s) s 2 + 4s + 8 s 2 + 2x w s + w2
n n
\ w2n = 8 i.e. wn = 2.83 rad/sec.
2 x wn = 4 \ x = 0.7067
\ wd = wn 1 - x 2 = 2.83 1 - (0.7067) 2 = 2.002 rad/sec.
p p
\ Tp = Time for peak overshoot = = = 1.57 sec.
wd 2.002
1 - x2 1 - (0.706) 2 ´ 100
% Mp = e - p x ´ 100 = e - p ´ 0.706 = 4.33 %
4 4
Ts = Settling time = = = 2 sec.
x wn 0.7067 ´ 2.83
Example 9.14.19
Solution : The characteristic equation is 1 + G(s)H(s) = 0
16
\ 1+ =0 i.e. s2 + 6s + 16 = 0
s (s + 6)
2
Comparing with s + 2 x wn s + w2n = 0,
wn = 4 rad /s, 2 xwn = 6 i.e. x = 0.75
wd = wn 1 - x 2 = 2.6457 rad/s
p-q –1 1 - x2
Tr = where q = tan = 0.7227 rad
wd x
p - 0.7227
\ Tr = = 0.9142 sec
2.6457
TM

p p
TP = = = 1.1874 sec
wd 2.6457
% MP = 100 e - px / 1- x 2 = 2.8375 %
4 4
Ts = = = 1.333 sec
x wn 0.75 ´ 4
Example 9.14.20
Solution : The applied force f(t) gets consumed in moving a mass through x(t) against
spring and friction force.
d 2x dx
\ f(t) = M + B + K x( t ) K (1)
dt 2 dt
Taking Laplace and neglecting initial conditions,

F(s) = M s 2 X(s ) + B s X (s ) + K X(s ) = X(s) [M s 2 + B s + K]
X(s ) 1 1/M
\ = = K (2)
F (s ) 2
M s + B s +K s + B s + K
2
M M
Comparing denominator with s 2 + 2 x wn s + w2n ,
K K B B
w2n = i.e. wn = and 2 x wn = i.e. x = K (3)
M M M 2 MK
8.9
Now f(t) = 8.9 N i.e. F(s) =
s
8.9 (1 M)
\ X(s) = .
s B K
s2 +
s+
M M
The final value of x(t) is 0.03 m.
8.9 (1 M)
\ Final value = Lim s . X(s) i.e. Lim s . . = 0.03
s® 0 s® 0 s B K
s2 + s+
M M
8.9
\ = 0.03 i.e. K = 296.67 N/m
K
The overshoot is 0.0029. But this is for the final value of 0.03. Hence the % M p can be
calculated as,
0.0029 2
% Mp = ´ 100 = 9.67 % = e - p x 1 - x ´ 100
0.03
Solving, x = 0.5966
p p
Now Tp = 2 sec = i.e. wn =
wn 1 - x2 2 1 - x2
TM

p
\ wn = = 1.957 rad/sec
2 1 -( 0 . 5966) 2
Equating to equation (3),

K 296.67
wn = i.e. 1.957 = i.e. M = 77.44 kg
M M
B B
And, x = i.e. 0.5966 = i.e. B = 180.855 N/M/sec
2 MK 2 77.44 ´ 296.67
Example 9.14.21
Solution : From the T.F. the poles are at s = – 3 ± 7j and there are no zeros.
\ Pole-zero plot is as shown.
The transfer function is, X j7
C ( s) 1 1
= =
R ( s) (s + 3 + j7)(s + 3 - j7) ( s + 3) - (j7) 2
2
–3 0
1
=
s2 + 6s + 58 X – j7
Comparing denominator with s 2 + 2 xwn s + w2n ,

Fig. 9.2
w2n = 58 hence wn = 7.6157 rad/sec
6
2 xwn = 6 hence x = = 0.3939
2 ´ 7.6157
wd = wn 1 - x 2 = 7.6157 × 1 - ( 0. 3939) 2 = 7 rad/sec
p p
\ Tp = Peak time = = = 0.4487 sec
wd 7
% Mp = e - px 1- x 2
´ 100 = 23.11%
Example 9.14.22
Solution : The closed loop T.F. is,
(s + 2)
C(s) G(s) s(s + 1) s+2
= = =
R(s) 1 + G(s) (s + 2) 2
s + 2s + 2
1+
s(s + 1)
Comparing denominator with s 2 + 2 xwn s + w2n ,
w2n = 2 i.e. wn = 2 = 1.414 rad/sec
2
2 xwn = 2 i.e. x = = 0.7071
2´ 2
TM

wd = wn 1 - x 2 = 1.414 1 - (0.7071) 2 = 1 rad/s
é 1 - x2 ù p
q = tan -1 ê ú = 45° = rad.
êë x úû 4
p-q p
\ Tr = = 2.3561 sec, Tp = = 3.1416 sec
wd wd
1 + 0.7 x 4
\ Td = = 1.0572 sec, Ts = = 4 sec
wn xwn
-p x 1 - x2
% Mp = 100 e = 4.32 %
Key Point As [C(s) / R(s)] is not in the standard form and numerator contains s term, the
standard expression of c(t) cannot be used.
s+2 (s + 2) 1
\ C(s) = R(s) ´ = …R(s) =
2
s + 2s + 2 s(s 2 + 2s + 2) s
Find the partial fractions,

A Bs + C A(s 2 + 2s + 2) + s(Bs + C)
C(s) = + =
s s 2 + 2s + 2 s(s 2 + 2s + 2)
\ A + B = 0, 2A + C = 1, 2A = 2
i.e. A = 1, B = –1, C = –1
1 (s + 1) 1 ì s +1 ü
\ C(s) = - = -í
s s + 2s + 2 s î s + 2s + 1 + 1 ýþ
2 2
1 ìï s +1 üï
= -í
s ï (s + 1) 2 + (1) 2 ýï
î þ
\ c(t) = L-1 [C(s)] = 1 - e - t cost
Example 9.14.23
Solution : The closed loop transfer function is,
K (s + a)
C(s) G(s) (s + b) 2 K (s + a)
= = =
R(s) 1 + G(s)H(s) K (s + a) s 2 +
s (2 b + K) + (ka + b 2 )
1+
2
(s + b)
Comparing denominater with s 2 + 2xwn s + w2n
w2n = K a b 2 i.e. wn = K a + b2 … (1)
2b + K
2 x wn = 2 b + K i.e. x = … (2)
2´ K a +b 2
TM

For unit step input, A = 1.

K( s + a ) Ka
Kp = Lim G( s) H( s ) = Lim =
s®0 s®0
(s + b )
2 b2
A 1 b2
\ e ss = = = … (3)
Kp Ka Ka
b2
But e ss is 10 % hence,
b2
0.1 = i.e. K a = 10 b 2 … (4)
Ka
10
From (1), 10 = 10 b 2 + b 2 i.e. b 2 = i.e. b = 0.9534
11
2 ´ (0.9534) + K
From (2), 0.5 = i.e. K = 1.2552.
2´ 10 ´ (0.9534) 2 + (0.9534) 2
From (4), 1.2552 ´ a = 10 ´ (0.9534) 2 i.e. a = 7.2416
Example 9.14.24
Solution : Substituting value of e,
d2 c dc
+ 6.4 = 160 (r - 0.4 c)
dt 2 dt
d2 c dc
\ + 6.4 + 64 c = 160 r
d t2 dt
Taking Laplace transform of both the sides, and neglecting initial conditions,
s 2 C (s) + 64 s C(s) + 64 C (s) = 160 R(s)
C(s) 160
\ = … Transfer function
R (s) 2
s + 6.4 s + 64
Comparing denominater with s 2 + 2 x wn s + w2n ,
wn 2 = 64 and 2 x wn = 6.4
6.4
i.e. wn = 8 rad/sec. = 0.4
and x=
2´ 8
As the damping ratio x is less than unity, the
c(t)
system is underdamped and will produce the
time response with damped oscillatory transients
for the step input as shown in the figure
TM

Example 9.14.25
Solution: The closed loop transfer function is,
C( s) G( s) 1 + 0.4s 1 + 0.4s
= = =
R( s) 1 + G( s)H( s) s( s + 0.6) s2 + s +1
(1 + 0.4s)
1+
s( s + 0.6)
It is not standard second order system hence standard expression of c(t) cannot be used.
Use partial fractions.
é 1 + 0.4s ù 1 + 0.4s A Bs + C
C( s) = R ( s) ê = = +
ë s 2 + + ú
s 1 û s s + s +1
2 s s (
2 + +
s 1 )
\ ( )
A s 2 + s + 1 + ( B s + C) s = 1 + 0.4s
\ A + B = 0, A + C = 0.4, A = 1 i.e. B = –1, C = – 0.6

1 ( s + 0.6) 1 s + 0.6
\ C(s) = – = –
s s2 + s +1 s 1 3
s2 + s + +
4 4
1 s + 0.6 1 s + 0.5 + 0.1
= – = –
s 2 2 s ( s + 0.5) 2 + ( 0.866) 2
æ s + 1 ö + æç 3 ö÷
ç 2 ÷ø è 2 ø
è
1 ( s + 0.5) 0.1 0.866

= – – ´
s ( s + 0.5) 2 + ( 0.866) 2 0.866 ( s + 0.5) 2 + ( 0.866) 2
\ c(t) = 1 – e –0.5t cos 0.866t – 0.1154 e –0.5t sin ( 0.866t )

From denominator of C(s)/R(s), w2n = 1 , 2xwn = 1
\ wn = 1, x = 0.5 , wd = wn 1 – x 2 = 0.866
1– x 2 p
\ Mp = 100 e – px / = 16.3 %, Tp = = 3.627 sec
wd
Example 9.14.26
K
Solution : For the given system, G(s) = , H(s) = 1 + bs
s(s + 2)
K
C( s) G(s) s(s + 2) K
\ = = =
R( s) 1+ G(s)H(s) K(1 + bs) 2
s + s( 2 + Kb) + K
1+
s( s + 2)
Comparing denominator with s 2 + 2xwn s + wn2 ,
w2n = K i.e. wn = K ... (1)
TM

(2 + Kb)
2xwn = 2 + Kb i.e. x= ...(2)
2 K
Given specifications are Mp = 20 % and Tp = 1 sec
1- x 2 – px
\ 20 = e - px / ´ 100 i. e. ln(0.2) =
1 – x2
Solving, x = 0.4559
p p
and Tp = i.e. 1=
2 wn 1 - ( 0.4559) 2
wn 1 - x
\ wn = 3.5297 rad/sec.
From equation (1), 3.5297 = K i.e. K = 12.4587
(2 + 12.4587 ´ b)
From equation (2), 0.4559 = i.e. b = 0.0978.
2 ´ 3.5297
wd = wn 1 - x 2 = 3.1415 rad / sec
æ 1 - x2 ö
q = tan -1 ç ÷ = 62.8771º = 1.0974 rad
ç x ÷
è ø
p - q p - 1.0974
\ Tr = = = 0.651 sec.
wd 3.1415
4 4
\ Ts = = = 2.4857 sec.
xwn 0.4559 ´ 3.5297
Example 9.14.27
Solution : i) All the three elements M, K and B are under same displacement x(t). Thus
the differential equation is,
d 2 x( t ) dx( t )
f(t) = M +B + K x( t )
dt 2 dt
Taking Laplace of both sides,
F(s) = Ms 2 X( s) + Bs X( s) + K X( s) ... Neglect initial conditions
\ X( s) 1 1M
= =
F ( s) Ms + Bs + K s 2 + B s + K
2
M M
ii) For K = 33, B = 15, M = 3

X( s) 1 3 0.333
= =
F( s) 2 2
s + 5s + 11 s + 5s + 11
TM

w2n = 11 hence wn = 3.3166 rad/sec.

2xwn = 5 hence x = 0.7537
% Mp = e - px 1- x 2
´ 100 = 2.7234 %
wd = wn 1 - x 2 = 2.1797 rad/sec
p
\ Tp = = 1.4412 sec
wd
4
and Ts = = 1.6 sec
xwn
Example 9.14.28
di(t) 1 1
+ ò i(t) dt
Cò
Solution : ei(t) = R i(t) + L and eo(t) = i(t) dt
dt C
I(s)
Ei(s) = I(s) éR + sL + ù and Eo(s) =
1
\
ëê sC ûú sC
E o (s) 1 1 / LC
\ = =
E i (s) s LC + sRC + 1 s + R s + 1
2 2
L LC
2 2
Comparing denominator with s + 2x wn s + wn ,
1 R
w2n = and 2x wn =
LC L
1 R
\ wn = and x=
LC 2 L/ C
qqq
TM

The Stability of Linear
10 Feedback Systems
Example 10.6.2
Solution : The Routh's array is,
s4 1 26 80
s3 6 56 0
No sign change in the first column of the
s2 16.67 80 0 Routh's array hence the system is stable in
nature.
s1 27.2 0
s0 80
Example 10.6.3
Solution :
The Routh's array is,
s4 1 23 50
There are no sign changes in the first column
s3 6 40 0
of the Routh's array, hence the system is
s2 16.333 50 0 stable in nature.
s1 21.633 0
s0 50
Example 10.6.4
Solution : a 0 = 1, a 1 = 6, a 2 = 11, a 3 = 6, n = 3
s3 1 11
s2 6 6
s1 11 ´ 6 - 6 0
= 10
6
s0 6
As there is no sign change in first column, system is stable.

(10 - 1)
TM

Control System Engineering 10 - 2 The Stability of Linear Feedback Systems
Example 10.6.5
Solution : a 0 = 1, a 1 = 4, a 2 = 1, a 3 = 16
s3 1 1
2 +4 16
s
s1 4 - 16 0
= -3
4
s0 + 16
As there are two sign changes, system is unstable.

Number of roots located in the right half of s-plane = Number of sign changes = 2.
Example 10.6.6
Solution : a) The characteristic equation is, 1 + G( s)H( s) = 0.
1
1+ = 0 i.e. s 2 + 6s + 9 = 0
( s + 2 )( s + 4)
Routh's array is,
2
s 1 9 No sign change in the
first column, hence
1
s 6 0 system is stable.
0
s 9
9
b) 1+ = 0 i.e. s3 + 2s2 + 9 = 0
2
s ( s + 2)
Routh's array is,
3
s 1 0 There are two sign changes in the
first column, hence system is
2
s 2 9 unstable.
1
s – 4.5 0
0
s 9
Example 10.7.6
Solution :
s5 1 2 3
s4 1 2 15
s3 0 – 12 0
Replace 0 by small positive number e.
TM

s5 1 2 3
s4 1 2 15
s3 e –12 0
s2 2e + 12 15 0
e
æ 2e + 12 ö
ç ÷( -12) - 15 e
è e ø
s1 2e + 12 0 0
e
s0 15
Lim 2e + 12 12
= = 2+ = +¥
e®0 e e
( 2 e+ 12 )
Lim e
( -12) - 15 e Lim -24e - 144 - 15e 2
=
e®0 2 e+ 12 e®0 2e + 12
e
0 - 144 - 0
= = – 12
0 + 12
s5 1 2 3
s4 1 2 15
s 3 e – 12 0
s2 +¥ 15 0
There are two sign
s1 –12 0 changes, so system
is unstable.
0 15
s
Example 10.7.8
Solution : s 4 + 0s 3 + 2s 2 + 0s + 1 = 0
Routh’s array is,
s4 1 2 1
s3 0 0
\ Row of zeros, special case 2,

\ Auxiliary equation is,
A(s) = s 4 + 2s 2 + 1 = 0
dA(s)
= 4s 3 + 4s
ds
TM

\ Routh’s array is,
s4 1 2 1
s3 4 4 0
s2 1 1 0
s1 0 0
\ Routh’s array
s4 1 2 1
s3 4 4 0
s2 1 1 0
s1 2 0
s0 1
No sign change so no root in right half.

Solving A(s) = 0 i.e. s 4 + 2s 2 + 1 = 0
-2 ± 4 - 4
\ s2 = = –1, –1
2
\ s = ± j, ±j
\ Number of roots on imaginary axis = 4
Number of roots in left half = 0 = Number of roots in right half.
Example 10.7.9
Solution :
s6 1 3 – 16 – 48
s5 4 0 – 64 0
s4 3 0 – 48 0
s3 0 0 0 Ü Special case 2
A(s) = 3s4 - 48 = 0
dA
= 12s3
ds
s6 1 3 – 16 – 48
s5 4 0 – 64 0
s4 3 0 – 48 0
TM

s3 12 0 0 0
s2 [ e] 0 – 48 0 0 Ü Special case 2
s1 576 0 0 Lim 576

=+¥
e e® 0 e
s0 – 48
\ One sign change and system is unstable. With one root in right half of s-plane.
4
Solve, A(s) = 3s 4 - 48 = 0 i.e. s = 16
s2 = 4 i.e. s = ± 2 and s2 = – 4 i.e s = ± 2j
So two roots (± j2) located on jw axis and remaining in left half of s-plane.
Roots with positive real part ® One

Roots with negative real part ® Three
Roots with zero real part ® Two.
Example 10.7.10
Solution : The characteristic equation is , s 6 + s5 + 3s 4 + 3s 3 + 2s 2 + s 1 + 1 = 0,
The Routh array is ,
3 e -1 s
6
1 3 2 1
For s 3 , Lim ® Negative
e® 0 e 5
s 1 3 1 0
æ 3 e -1 ö s
4
0 +e 1 1 Special case 1
ç e ÷ - ( e - 1)
For s , Lim è
2 ø
s
3 3e – 1 e–1 0
e® 0 æ 3 e -1 ö e e
ç e ÷
è ø 3e – 1
– ( e – 1)
2 e
® Positive i.e. X ® positive s 1
(3e – 1)
e -1 ö æ 3 e -1 ö
Xæç ÷ -ç ÷
e X
1 è e ø è e ø
For s , Lim e–1 3e – 1
e® 0 X 1 X –
s e e 0
® Negative X
0
Thus there are four sign changes in the s 1
first column of the Routh's array.
Hence the system is unstable with four
roots lying in the right hand side of s -
plane.
Example 10.8.7
Solution : The characteristic equation is 1 + G(s)H(s) = 0
TM

K
\ 1+ = 0 ...H(s) = 1
s (s + 2) (s +3) (s + 4)
\ s 4 + 9s 3 + 26s 2 + 24s + K = 0
s4 1 26 K From the row of s 0 , K > 0

s3 9 24 0 From the row of s 1 , 560 – 9K > 0
s2 23.333 K \ 560 > 9K
560 – 9 K \ K < 62.222

s1 0
23. 333 Hence the range of K for stability is,
s0 K 0 < K < 62.222
Example 10.8.8
Solution : F(s) = s [s 3 + 5s 2 + 5s + 4] + K = 0 i.e. s 4 + 5s 3 + 5s 2 + 4s + K = 0
s4 1 5 K For system to be stable there

should not be sign change in the
s3 5 4 0
first coumn.
s2 4.2 K 0
\ K>0 from s0
s1 16.8 - 5K 0 0
\ 16.8 – 5 K > 0 from s
4.2
\ 16.8 > 5 K
s0 K
\ 3.36 > K \ K < 3.36
\ Range of K is 0 < K < 3.36
Example 10.9.5
Solution : The characteristic equation is 1 + G(s) H(s) = 0
K
\ 1+ = 0 i.e. s 4 + 4s 3 + 4s 2 + 3s + K = 0
2
s(s + 3)(s + s + 1)
s4 1 4 K The Rouths array is,

For marginal value of K, row of s 1 must be row of zeros.
s3 4 3 0
\ 9.75 – 4 Kmar = 0 i.e. Kmar = 2.4375
s2 3.25 K 0
For this value of K, the auxiliary equation is,
s1 9.75 - 4K 0
3.25 - Kmar -2.4375
A(s) = 3.25 s 2 + Kmar = 0 i.e. s 2 =
3.25 3.25
s0 K
\s = ± j 0.866
TM

Thus the frequency of oscillations in 0.866 rad/sec.
Example 10.9.6
Solution : The characteristic equation is,
s(s 2 + 25 +2) (s 2 + 65 +10) + K = 0 i.e. s5 + 8s 4 + 24s 3 + 32s 2 + 20s + K = 0
Routh's array is,

For marginal K1 row of s 1 must be row of zeros.
s5 1 24 20 480 + K ö æ 160 - K ö
\æç ÷ç ÷ - 20K = 0
s4 8 32 K è 20 ø è 8 ø
s3 20 160 - K 0 \76800 + 160K - 480 K - K2 - 3200 k = 0

8
i.e. - K2 - 3520K + 76800 = 0
480 + K
s2 K
20 \( K + 3541.6845) (K - 21.6845) = 0
\ Kmar = + 21 .6845 … Selecting + ve value

+
A(s) = æç mar ö 2
480 K
For this value,
20 ÷ s + Kmar = 0
è ø
21.6845
\ s2 = - = 0.8644
(480 + 21.6845)
20
\ s = ± j 0.9297 = ± jw
Hence the frequency of oscillations is 0.9297 rad/sec.
Example 10.9.7
Solution : The characteristic equation of the system is
1 + G(s) H(s) = 0
K(s + 5)
1+ = 0
s(1 + Ts) (1 + 2s)
\ s(1 + Ts) (1 + 2s) + K(s + 5) = 0
\ s(2Ts 2 + (T + 2)s + 1) + K(s + 5) = 0
\ 2Ts 3 + (T+2) s 2 + s(K + 1) + 5K = 0
TM

\ Routh’s array is
s3 2T K+1
s2 (T + 2) 5K
s1 (T + 2) (K + 1) - 2T ´ 5K 0
(T + 2)
s0 5K
From last row 5K > 0

K must be positive.
From row of s 1
(T + 2) (K + 1) – 10KT > 0
KT + 2K + T + 2 – 10KT > 0
\ 2K + T + 2 – 9KT > 0
Limiting value is
2K + T + 2 – 9KT > 0
2+ T
i.e. K <
9T - 2
\ Region in which a closed loop system is stable is

T
Considering positive
Unstable 2+T
K= values of K and T
9T – 2
Stable Unstable
K
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Unstable
Example 10.9.8
Solution : (i) Loop T.F. has pole at s = 0, s = –1, s = –3 and zero at s = –5, gain K = 10
10(s + 5)
G(s) H(s) =
s(s + 1) (s + 3)
Characteristic equation.
1 + G(s) H(s) = 0
10(s + 5)
1+ = 0
s(s + 1) (s + 3)
TM

s(s 2 + 4s + 3) + 10s + 50 = 0
s 3 + 4s 2 + 13s + 50 = 0
s3 1 13
s2 4 50
s1 0.5 0
s0 50
No sign change in the first column hence system is absolutely stable in nature.
ii) Type 1, Kv = 10 sec -1 , Poles at s = –3, s = – 6
K(1 + T1 s)...
Type 1 G(s) H(s) =
s(1 + T2 s) (1 + T3 s)...
K/ 18
G(s) H(s) =
s æç 1 + ö÷ æç 1 + ö÷
s s
è 3 øè 6 ø
lim lim 5 ´ K/ 18 K
Now Kv = sG(s) H(s) = =
s®0 s®0 18
s æç 1 + ö÷ æç 1 + ö÷
s s
è 3 øè 6ø
K
Now = 10 \ K = 180
18
180
\ G(s) H(s) =
s(s + 3) (s + 6)
Charcteristic equation : 1 + G(s) H(s) = 0
s(s 2 + 9s + 18) + 180 = 0
s 3 + 9s 2 + 18s + 180 = 0
s3 1 18
s2 9 180
s1 –2 0
s0 180
There are two sign changes so two roots are located in R.H.S. of s-plane
Hence system is Unstable.
Example 10.9.9
Solution : The open loop T.F.,
K
G(s)H(s) =
s (1 + sT1 ) (1 + sT2 )
TM

The characteristic equation is,
1 + G(s)H(s) = 0
K
\ 1+ = 0
s (1 + sT1 ) (1 + sT2 )
\ T1 T2 s 3 + (T1 + T2 ) s 2 + s + K = 0

s3 T1T2 1
s2 (T1 + T2 ) K
s1 (T1 + T2 ) - T1T2 K 0
(T1 + T2 )
s0 K
K>0
For stability,
and (T1 + T2 ) - T1 T2 K > 0
\ T1 T2 K < T1 + T2
æ 1 1 ö
\ K<ç + ÷
T
è 1 T2 ø
The range of values of K for stability,

æ 1 1 ö
0<K<ç + ÷
T
è 1 T2 ø
Example 10.9.10
Solution : The open loop T.F. is,
K
G(s) =
(s + 4) 3
The characteristic equation is,
1 + G(s) = 0
K
\ 1+ =0
(s + 4) 3
\ (s + 4)3 + K = 0
3 3
\ s + 12s 48s + (64 + K) = 0
TM


3
s 1 48
2
s 12 64 + K
1 576 - 64 - K
s 0
12
0
s 64+K
0
From s , 64 + K > 0
\ K > – 64
and from s1, 576 – 64 – K > 0
\ 512 – K > 0
\ K < 512
So range of values of K for system stability is,
– 64 < K < 512
0 < K < 512
i.e. effectively
Example 10.9.11
Solution : The characteristic equation is,
1+G(s) = 0 H(s) = 1
K
\ 1+ = 0
s(s + 10) (s 2 + 4s + 5)
\ s 4 + 14s 3 + 45s 2 + 50s + K = 0
s4 1 45 K
s3 14 50 0
s2 41.428 K
s1 2071.4 - 14K 0
41.428
s0 K
From s 0 , K > 0
From s 1 , 2071.4 – 14 K > 0
2071.4
\ K < < 147.957
14
TM

So range of values of K is, 0 < K < 147.957.
Example 10.9.12
Solution : i) All roots must have negative real parts i.e. all roots lying in L.H.S.
s3 1 50 0
s2 4+K K
1 12+ 6K 0
s
4+ K
s0 12
\ 4+K > 0
\ K > –4
and 12 + 6K > 0
\ 12 > – 6 K
\ – 12 < 6K
\ K > –2
So effectively for all values of K greater than – 2 to ¥, the polynomial roots will have
negative real part.
ii) To have roots with zero real parts, there must exist row of zeros in Routh's array, other
than row of s 0 .
s4 1 24 K
3 8 32 0
s
2 20 K
s
s1 640 - 8K 0
20
s0 K
So for s 1 as row of zeros,

640 – 8 K = 0
\ K = 80
Thus for K = 80, A(s) = auxiliary equation
= 20s 2 + K = 0
\ 20s 2 + 80 = 0
\ s2 = – 4
\ s = ± j2
Thus we get the roots with zero real parts.
qqq
TM

11 The Root Locus Method

Pole under consideration
Example 11.5.8
Solution : P = 4 , Z = 1 j
Poles are at s = 0 , – 4 , – 1 ± j Zero at

s = –2.
fP3 fZ1 fP1
Draw pole-zero plot.
–4 –2 –1 0
Let us calculate fd at the pole s = –1 + j.
Join all other poles to this pole and
fP2
measure or calculate the angles –j
fP1, fP2 , fP3 as shown in the
Fig. 11.1 (a). Fig. 11.1 (a)
Join all zeros to this pole and calculate
fZ1 .
Then , å fP = fP1 + fP2 + fP3 while
å fZ = fZ1
From geometry of the Fig. 11.1 (a) we can calculate,
fP1 = 135° , fP2 = 90° , fP3 = 18.43°
\ å fP = 135° + 90° + 18.43° = 243.43°
å fZ = fZ1 = 45°
Tangential to j o
\ f = å fP – å fZ = 243.43°– 45° fd line –18.43
= 198.43°
–4 –2 –1 0
o
fd = 180° – f = 180° – 198.43° = – 18.43° +18.43
\ Root locus branch leaving this pole will depart –j
tangentially to the line whose angle is given by

fd = – 18.43° as shown in the Fig. 11.1 (b)
Fig. 11.1 (b)
(11 - 1)
TM

Control System Engineering 11 - 2 The Root Locus Method
For second complex conjugate pole, sign of fd will be just opposite as root locus is always
symmetrical about real axis. So root locus branch departing from s = – 1– j will depart
tangentially to the line whose angle is given by fd = +18.43°.
Example 11.7.13
Solution :
Step 1 : P = 4, Z = 0, N = P = 4, P – Z = 4 branches to ¥
Starting points : s = 0, – 2, – 4, – 8.
Terminating points : s = ¥, ¥, ¥, ¥
Step 2 : Sections of real axis as shown in jw
the Fig. 11.2 (a).
NRL NRL NRL
Step 3 : Angles of asymptotes X X X X s
–8 –4 –2 0
(2q + 1)180º
q = , q = 0, 1, 2, 3
P– Z Two breakaway
q 1 = 45º, q 2 = 135º, q 3 = 225º, q 4 = 315º points
Fig. 11.2 (a)
Step 4 : Centroid =
å R. P. of O. L. poles – å R. P. of O. L. zeros = –2 – 4 – 8 – 0 = – 3.5
P– Z 4
Step 5 : Breakawary points

1 + G( s)H( s) = 0 i.e. s 4 + 14s 3 + 56s 2 + 64s + K = 0 … (1)
K = –s 4 - 14s 3 - 56s 2 - 64s …(2)

dK
\ = 0 gives 4s 3 + 42s 2 + 112s + 64 = 0
ds
Solving, s = – 0.785, – 6.652, – 3.061
Valid breakaway points are s = – 0.785 , – 6.652
Step 6 : Intersection with imaginary axis
s4 1 56 K
from equation (1), Routh's array is,
s3 14 64 0 From row of s1 for Kmar,
s2 51.42 K 0 3291.42 – 14K = 0
s1 3291.42 – 14K 0 \ Kmar = 235.1
51.42
A(s) = 51.42 s2 + K = 0
s0 K 2 -235.1
s =
51.42
\s = ± j2.138
Step 7 : Angle of departure not required.
TM

Step 8 : The root locus is shown in the Fig. 11.2 (b). To find dominant poles for x =
–1
0.707, draw a line making angle q = cos x = 45º, with respect to negative real axis. It
intersects root locus at P = – 0.7 + j 0.7, from graph.
Scale :
On both axes : 1unit = 1unit
Imj
q2 = 135º q = 45º q1 = 45º

for
¥ x = 0.707 ¥
+j2.138
q = 45º
P = –0.75 + j0.75
for x = 0.707
P
X X X X Real
–4 2 0
–8
Breakaway Centroid
point –3.5
–6.652
Breakaway
point –j2.138
–0.785
¥ ¥
q3 = 225º q4 = 315º
Fig. 11.2 (b)
Thus for x = 0.707, the dominant poles are located at – 0.75 ± j0.75.
Example 11.7.14
Solution : Step 1 : P = 4, Z = 0, N = P = 4, P – Z = 4 branches to ¥
Starting : 0, – 3, -1.5 ± j1.5, Terminating : ¥, ¥, ¥, ¥
TM

Step 2 : Sections of real axis as shown in the Fig. 11.3 (a).
Note that when the open loop poles are distributed

One
symmetrically about a point on the real axis then breakaway imj
dK point X
point os symmetry is one of the roots of = 0. NRL
ds NRL s
X X
dK –3 –1.5 0
Also check the validity of all the roots of = 0 as
ds X
breakaway points as root locus of such system has
special behavior. Fig. 11.3 (a)
(2q + 1)180º
Step 3 : q= , q = 0, 1, 2, 3
P-Z
\q 1 = 45º, q 2 = 135º, q 3 = 225º, q 4 = 315º
Step 4 : Centroid = å R. P. of O. L. poles - å R. P. of O. L. zeros = 0 - 3 - 1.5 - 1.5

= -1.5
P-Z 4
Step 5 : Breakaway point

1 + G( s)H( s) = 0, i.e. s 4 + 6s 3 + 13.5s 2 + 13.5s + K = 0 … (1)
\ K = -s 4 - 6s 3 - 13.5s 2 - 13.5s … (2)

dK
= 0 gives 4s 3 + 18s 2 + 27s + 13.5 = 0
ds
Solving , s = – 1.5, – 1.5, – 1.5
From equation (2), K = + 5.0625 for s = – 1.5.
As there is symmetry in both directions (x and y) about a point s = – 1.5, all three
breakaway points are situated at point of symmetry s = – 1.5. This gives special nature of
root locus.
Step 6 : Intersection with imaginary axis.

From equation (1), Routh's array is :
s4 1 13.5 K From row of s1 for Kmar,
s3 6 13.5 0 151.875 – 8K = 0
\ Kmar = 25.3125
s2 11.25 K 0
A(s) = 11.25s2+ K = 0
s1 151.875 - 8K 0
2 –25.3125
11.25 \ s =
11.25
s0 K
\ s = ± j1.5
TM

Step 7 : Angle of departure at complex poles consider

pole at – 1.5 + j 1.5.
X j1.5
fP = 135º, fP2 = 90º , fP3 = 45º
1
fP3 fP1
\ å fP = 270º,å fZ = 0º X X
0
fP2
\ f = å fP - å fZ = 270º X – j1.5
\ fd = 180º- f = - 90º at – 1.5 + j 1.5

fd = + 90º at – 1.5 – j 1.5 Fig. 11.3 (b)
Step 8 : The nature of this root locus is typical as
shown in the Fig. 11.3 (c). All branches meet at s = – 1.5 and then breakaway along
asymptotes approaching to ¥.
Scale
on both axes
2 units = 1unit
Imj q1 = 45º
q2 = 135º
¥
¥ X j 1.5
fd = – 90º
X X Real
–3 0
Centroid
and
three breakaway
points s = –1.5
fd = +90º
¥ X –j 1.5
q3 = 225º ¥
q4 = 315º
Fig.11.3 (c)
Example 11.7.15
Solution : The open loop poles are located at,
- 4 ± 16 - 80
s = 0, – 4 and = – 2 ±j4
2
TM

Step 1 : P = 4, Z = 0, N = P = 4, P – Z = 4 branches
approaching to ¥. One X j4
breakaway
Starting points, s = 0, – 4, –2 + j4 , –2 – j4 point
NRL NRL
Terminating points : s = ¥, ¥, ¥, ¥ X X
–4 –2 0
Step 2 : Sections of real axis shown in the Fig. 11.4(a).
Step 3 : Angles of asymptotes X – j4
(2q + 1) 180°
q = , q = 0, 1, 2, 3
P-Z Fig. 11.4 (a)
\ q1 = 45º, q 2 = 135º, q 3 = 225º, q 4 = 315º
Step 4 : Centroid =
å R. P. of O. L. poles - å R. P. of O. L. zeros = 0 - 4 - 2 - 2 = – 2
P-Z 4
The characteristic equation is 1 + G(s)H(s) = 0 i.e. s 4 + 8s 3 + 36s 2 + 80s + K = 0 … (1)
\ K = - s 4 - 8 s 3 - 36s 2 - 80s ... (2)
dK
\ = – 4s 3 - 24s 2 - 72s - 80 = 0 i.e. s 3 + 6s 2 +18s + 20 = 0
ds
Solving, s = – 2, –2 ± j 2. 45
Due to over all symmetry about s = –2, the root locus has special nature. In such case all
the breakaway points may be valid. So confirm the validity of complex conjugate
breakaway points using angle condition.
K
ÐG( s)H( s) s = - 2 + j2.45 = Ð
s( s + 4)(s + 2 + j4)(s + 2 - j4)
s = - 2 + j2.45
0º
=
Ð - 2 + j2.45º Ð2 + j2.45 Ðj6.45 Ð - j1.55
0º
=
129.22º , 50.78º , 90º , -90º
= -180º … Angle condition is satisfied.
Hence both -2 ± j2.45 are valid breakaway points.
While the value of K at - 2 ± j 2.45 can be obtained using magnitude condition.
K
G(s)H(s) s = -2 + j 2.45 = 1 i.e. =1
s(s + 4) (5 + 2 + j4) (5 + 2 - j4) s = - 2 + j2.45
K
\ =1
-2 + j2.45 -2 + j2.45 + 4 -2 + j2.45 + 2 + j4 -2 + j2.45 +2 - j4
\ K = 3.1626 ´ 3.1626 ´ 6.45 ´ 1.55 = 100
As – 2 – j 2.45 is complex conjugate, K = 100 also for s = – 2 – j 2.45.
For s = – 2, K = - ( -2) 4 - 8( -2) 3 - 36( -2) 2 - 80( -2) = + 64 ... from (2)
TM

s 4 + 8s 3 + 36s 2 + 80s + K = 0 ... Characteristic equation
From the row of s 1 , 2080 – 8 K = 0

s4 1 36 K 2080
\ Kmar = = 260
8
s3 8 80 0
For this Kmar, A(s) = 0 is,
s2 26 K 0
26s 2 + Kmar = 0
s1 2080 - 8K 0 -Kmar
26 \ s2 = = – 10
26
s0 K
\ s = ± j 3.162
These are intersection points with imaginary axis.
Step 7 : Angle of departure

Consider complex pole at – 2+ j4.
X j4
From the geometry,
4 fP3 fP1
fP1 = 180º – tan -1 = + 116.56º
2 X X
–4 –2 0
fP2 = 90º fP2
4 X – j4
fP3 = tan -1 = + 63.43º
2
å fP = 270º, åfZ = 0º Fig. 11.4 (b)

\ f = å fP - å f Z = 270
\ fd = 180º – f = 180º – 270º = – 90º ... at –2 + j4
fd = + 90º ... at – 2 – j4
4 branches
s=0 s=–4 s = – 2 + j4 s = –2– j4
Breakaway
point 1
K = 64 s = –2
s = –2 + j2.45 s = –2 – j2.45 Breakaway

point 3
Breakaway
point 2
¥ ¥ ¥ ¥
Both breakaway points 2 and 3 occur simultaneously at K = +100

Fig. 11.4 (c)
TM

Step 8 : Draw the root locus to the scale on the graph paper as shown Fig. 11.4(d).
Imj
Scale fd = –90º
2 units = 1 unit
on both axes +j4
q 2= 135º ¥ ¥
q= 45º
Breakaway q1= 45º
point + j 3.162
–2 + j2-45
q= 30º + j 2.45
Q
–4 0 Real
Breakaway –2
point 1 and
centroid
s = –2
– j 2.45
Breakaway
point – j 3.162
q3= 255º q4= 315º
–2 – j2-45
+¥ fd = +90º ¥
–j4
Fig. 11.4 (d)
TM

i) For x = 0.707, draw a line at q = cos -1 0.707 = 45º with respect to negative real axis ,
meeting root locus at Q. From graph Q is (– 2 + j 1.925). Using the magnitude condition,
K
= 1 i.e. K = 94.7357 at Q
s(s + 4) (s + 2 + j4) (s + 2 - j4) At Q
\ K = + 94.7357 ... for x = 0.707
ii) For x = 0.866, draw a line at q = cos -1 0.866 = 30º with respect to negative real axis,
meeting root locus at R. From graph R is (–2 + j 1.125). Using the magnitude condition,
K
= 1 i.e. K = 77.58 at R
s(s + 4) (s + 2 + j4) (s + 2 - j4) At R
\ K = +77.58 ... for x = 0.866

iii) For x = 1, K = +64 which is K at breakaway point s = – 2.
Example 11.7.16
Solution : The given G(s)H(s) is,
K K
G(s)H(s) = =
s(1 + 0.02s)(1 + 0.1s) s ´ 0.02(s + 50) ´ 0.1 ´ (5 + 10)
500K K1
= = with K1 = 500 K
s(s + 50)(s + 10) s(s + 50) (s + 10)
Step 1 : P = 3, Z = 0, N = P = 3, P – Z = 3 branches towards ¥ Starting points :
s = 0, – 10, – 50.
One
Terminating points : s = ¥, ¥, ¥ breakaway
point exists
Step 2 : Sections of real axis shown in the NRL
Fig. 11.5 (a) ´ ´ ´0 NRL
– 50 – 10
180° (2q + 1)
Step 3 : Angles of asymptotes =
P-Z
q = 0, 1, 2 Fig. 11.5 (a)
q1 = 60º , q 2 = 180º, q 3 = 300º
Step 4 : Centroid =
å R. P. of open loop poles - å R. P. of open loop zeros
P-Z
0 - 10 - 50
= = – 20
3
K1
1 + G(s)H(s) = 1 + =0
s(s + 50)(s + 10)
\ s 3 + 60s 2 + 500s + K1 = 0 i.e. K1 = – s 3 - 60s 2 - 500s ... (1)
TM

dK1
\ = - 3s 2 - 120s - 500 = 0 i.e. s 2 + 40s +166.667 = 0
ds
Solving, s = – 4.7247, 35.2752.
Thus s = – 4.7247 is valid breakaway point at which the value of K1 = + 1128.4510, using
equation (1). But K1 = 500 K hence,
K = +2.2569 at breakaway point s = – 4.7247.
s3 1 500 Rouths array from characteristic equation is,
s2 60 K1 Thus marginal value of K1 is 30000 making row of s 1
as row of zeros.
s1 30000 - K1 0 A(s) = 60 s 2 + K1 mar = 0
60
30000
s0 K1 \ s2 = -
= – 500
60
\ s = ± j 22.3606 … Intersection with imaginary axis
Step 7 : No angle of departure as no complex poles.
Step 8 : Sketch the root locus, to the scale on graph paper. (Refer Fig. 11.5 on next page)
For x = 0.4, draw a line at q = cos -1 x = 66.4218º with respect to negative real axis. Let it
meets root locus at point Q(- 4 + j 9.25) . Find K1 at this point Q using the magnitude
condition.
The co-ordinates of Q (– 4 + j 9.25). Using magnitude condition,
K1 K1
=1 i.e. =1
s s + 10 s + 50 At Q - 4 + j 9.25 6 + j 9.25 46 + j9.25
\ K1 = 10.0778 ´ 11.0255 ´ 46.9208 = 5213.5019

K1
\ K = = 10.427 for x = 0.4
500
Example 11.7.17
K Imj.
Solution : G(s)H(s) =
s(s + 3)(s + 5)
Step 1 : P = 3, Z = 0, N = P = 3,
NRL NRL
P – Z = 3 branches approach to ¥. Real
–5 –3 0
Starting points : s = 0, – 3, – 5
Terminating points : s = ¥, ¥, ¥
Step 2 : Sections of real axis Fig. 11.6 (a)
One breakaway point exists between s = 0 and s = – 3
TM

Imj
q1 = 60º
Scale
1 unit = 5 units ¥
on both axes
–1
q = cos x
= 66.42º
+ j 22.3606
Q = –4 + j 9.25
q2 = 180º Real
¥ – 40 – 30 –20 –10 0
– 50
Centroid
K1mar = 30000
For critical damping Breakaway
K = 2.2569 i.e. K at Point – 4.7247 \Kmar = 60
breakaway point
– j 22.3606
q3 = 300º
Fig. 11.5 (b)
Step 3 : Angles of asymptotes

(2q + 1) 180 °
q= , q = 0, 1, 2
P- Z
\q 1 = 60º, q 2 = 180º, q 3 = 300º
Step 4 : Centroid =
å R. P. of O. L. poles - å R. P. of O. L. zeros = 0 - 3 - 5 - 0 = – 2.667
P- Z 3
K
1 + G(s)H(s) = 1 + =0
s(s + 3)(s + 5)
TM

\ s 3 + 8s 2 + 15s + K = 0 i.e. K = - s 3 - 8s 2 - 15s … (1)

dK
\ = - 3s 2 - 16s - 15 = 0 i.e. 3s 2 + 16s + 15 = 0
ds
\ s = – 1.2137, – 4.119 thus s = – 1.213 is valid breakaway point
Using equation (1), K = + 10.2837 at breakaway point
1 + G(s)H(s) = s 3 + 8s 2 + 15s + K = 0
Routh's array,
s3 1 9 For Kmar ,
s2 6 K 120 – K = 0
s1 120 - K 0 \ Kmar = 120

8
A(s) = 8s 2 + K = 0
0
s K
- 120
\ s2 = = - 15
8
\ s = ± j 3.873 … Intersection with imaginary axis
Step 7 : No complex poles or zeros hence not required.
Step 8 : Draw the root locus as shown in the Fig. 11.6 (b).
For x = 0.6, draw the line at angle q = cos - 1 x = 53.13º, measured with respect to negative
real axis.
It intersects root locus at point P (– 1 + j1.25).
i) Closed loop dominant poles for x = 0.6 are s = – 1 ± j1.25.
ii) The characteristic equation for x = 0.6 is,
(s + 1 + j1.25) (s + 1 – j1.25) = 0 i.e. s 2 + 2s + 2.5625 = 0
Comparing with s 2 + 2 x wn s + w2n , w2n = 2.5625
\ wn = 1.6 rad/sec.
iii) K for x = 0.6 is to be obtained from magnitude condition.
K
|G(s)H(s)|P = 1 i.e. = 1
s(s + 3)(s + 5) s = - 1 + j1.25
K
\ =1
| - 1 + j1.25| |3 - 1 + j1.25| |5 - 1 + j1.25|
\ K = 1.6 ´ 2.3585 ´ 4.1907 = 15.8142 … Gain at x = 0.6
TM

q1 = 60º
Scale Imj.
on both axes : 1 unit = 1 ¥
+ j3.87
–1
q = cos x
–1
= 53.13º The cos x line
intersects root
P locus at point
P (–1 + j1.25)
q2 = 180º
¥ –5 –3 0 Real
Centroid
s = – 2.667
Breakaway point
s =– 1.2137
– j3.87
q3 = 300º
Fig. 11.6 (b)
Example 11.7.18
K K
Solution : Step 1 : G(s)H(s) = =
s (s 2 + 6s + 25) s (s + 3 + j4)(s + 3 - j4)
P = 3, Z = 0, N = Number of branches = P = 3
P – Z = 3 branches approaching to ¥.
TM

Starting points Þ s = 0, – 3 + j4, – 3 – j4

Terminating points Þ s = ¥, ¥, ¥
Step 2 : Sections of real axis are as shown
No breakaway point.
X j4
Step 3 : Angles of asymptotes :
(2q + 1)180º
q = ,
P-Z RL NRL
X
–3 0
q = 0, 1, 2
\ q 1 = 60º, q 2 = 180º,
X – j4
q 3 = 300º
Step 4 : Centroid :
s =
å R × P × of poles -å R × P × of zeros = 0-3-3
=–2
P-Z 3
Step 5 : Breakaway point :
K
1 + G(s)H(s) = 1 + =0
s ( s2 + 6s + 2s)
\ s 3 + 6s 2 + 25s + K = 0 i.e. K = -s 3 - 6s 2 - 25s = 0 … (1)
\ K = -s 3 - 6s 2 - 25s = 0
dK
\ = -3s 2 - 12s - 25 = 0 i.e. s 2 + 4s + 8. 3333 = 0
ds
- 4 ± 16 - 4 ´ 8. 333
\ s = = -2 ± j2.0816
2
The complex breakaway point is not possible hence there is no breakaway point.
Validity of breakaway point can be confirmed using angle condition, if required.

From characteristic equation (1),
Routh’s array is
3
s 1 25
2 From s1, 150 – K = 0 i.e. Kmar = 150
s 6 K
150 - K A(s) = 6s 2 + K = 0
s1 0
6 At Kmar, 6s 2 + 150 = 0 i.e. s2 = – 25
0
s K
\s = ± j5 … Intersection with imaginary axis
TM

Step 7 : Angle of departure at complex pole :

– 14
fP1 = 180º – x = 180º – tan = 126.869º X j4
3
fP2 = 90º fP1
x
X
å fP = fP1 + fP2 = 216.869º –3 0
å fZ = 0 … No zeros fP2
f = å fP - å fZ = 216.869º X – j4
\ fd1 = 180º – f = 180º – 216.869º

= – 36.869º at (– 3 + j4)
\ fd2 = + 36.869º at (– 3 – j4)
Step 8 : The root locus is shown in the Fig. 11.7.
jw
¥
q1 = 60º
Scale :
On both axes
1 unit = 1 unit + j5
– j4
fd = –36.86º
Kmar = +150
Centroid
s
q2 = 180º ¥ –3 –2 0
fd = +36.86º
– j4
– j5
q3 = 300º
¥
Fig. 11.7
TM

Example 11.7.19
j
Solution : Step 1 : P = 3, Z = 1,
N=P=3 NRL RL NRL
–4 –1 0
Branches will originate at the open loop
poles –j
s = 0, s = – 1 + j, s = – 1 – j
One branch will terminate at finite zero located at s = – 4. Remaining P – Z = 2 branches
will approach to ¥.
Step 2 : Pole-zero plot , sections of real axis.

Section between s = 0 and s = – 4 is the part of root locus. This is nothing but the branch
terminating at s = – 4. No breakaway point according to general predictions. Two
separate branches will start from the remaining complex poles and will approach to ¥
along the asymptotes.
Step 3 : Angles of asymptotes

Number of branches approaching to ¥ = 2.
Number of asymptotes required = 2
(2 q + 1) 180º
\ q = , q = 0, 1
P– Z
180º (2 + 1) 180°
\ q1 = = + 90° , q 2 = = + 270º
2 2
Step 4 : Centroid :
jw q1
270º
90º Centroid may or
may not be
–1 0 +1 s on the root
–4
Centroid locus.
q2
s =
å R . P. of poles – å R. P. of zeros
=
0 – 1 – 1 – (– 4)
=+1
P– Z 2
TM

Step 5 : No breakaway point exists as per general predictions.

K(s + 4)
Characteristic equation is, 1 + G(s)H(s) = 0 i.e. 1+ = 0
s (s 2 + 2s + 2)
\ s 3 + 2s 2 + 2s + Ks + 4K = 0
Routh’s array :
s3 1 K+2
4 – 2K = 0
s2 2 4K \ Kmar = + 2 making row of s 1 as row of zeros.
s1 4 – 2K 0 A(s) = 2s 2 + 4K =0
2
At K = +2, A(s) = 2s 2 + 8 = 0
s0 4K \s 2 = – 4 \s = ± j 2 … Intersection with j w axis
Step 7 : Angle of dapartures. Consider complex

pole – 1 + j and join all poles and zeros to it.
j
fP1 = +135° , fP2 = 90° fP1
1 fZ1
fZ1 = + tan -1 = 18.43°
3 –4 –1 0
fP2
\ S fP = fP1 + fP2 = 225°
–j
S fZ = fZ1 = 18.43°
\ f = S fP - S fZ = 206.56°
\ fd = 180° - f = – 26.56°, at s = – 1 + j
fd = + 26.56°, at s = – 1 – j
Step 8 : Complete root locus is as shown in the Fig. 11.8.
Step 9 : Prediction about stability.

For 0 < K < 2 : All roots are in left half of s-plane system is absolutely stable.
At Kmar = + 2 : Dominant roots are on imaginary axis. System is marginally
stable which oscillates at 2 rad/sec.
For K > 2 : Dominant roots are in right half of s-plane. System is unstable in
nature.
TM

¥
Imj q1 = 90º
Scale :
on both axes
2 units = 1 unit
+j2
+j
fd = –26.56º
Centroid
Real
–4 –1 0 +1
fd = +26.56º
–j
–j2
¥ q2 = 270º
Fig. 11.8
Example 11.7.20
Solution : Step 1 : P = 4, Z = 0, N = 4, P – Z = 4 branches to ¥
Starting points : s = 0, – 1, – 3, – 4
Terminating points : ¥, ¥, ¥, ¥
TM

jw
Step 2 : Sections of real axis. Two breakaway points
possible
NRL NRL NRL
Step 3 : Angles of asymptotes, q 1 = 45°, s
–4 –3 –1 0
q 2 = 135°, q 3 = 225°, q 4 = 315°
0 -1 - 3 - 4
Step 4 : Centroid = =–2 Two breakaway
4 points
Step 5 : Breakaway points
1 + G(s)H(s) = 0
i.e. s4 + 8s3 + 19s2 + 12s + K = 0
4 3 2
\ K = – s – 8s – 19s – 12s
dK
\ = – 4s3 – 24s2 – 38s – 12 = 0 i.e.
3 2
s + 6s + 9.5 s + 3 = 0
ds
\ s = – 0.4188, – 3.5811 … Breakaway points
Step 6 : Intersection with jw axis.
4
s 1 19 K \ 210 – 8 K = 0
210
s
3
8 12 0 \ Kmar = = 26.25
8
2 2
s 17.5 K A(s) = 17.5 s + Kmar = 0
26.25
1 210 - 8 K \ s2 = – i.e. s = ± j 1.224
s 0 17.5
17.5
0
s K
Step 7 : The nature of root locus is shown in the Fig. 11.9.
q2 = 135º Scale jw q1 = 45º

on both axes
1 unit = 1 unit
¥
j 1.22
Centroid
s
–4 –3 –2 –1 0
– j 1.22
¥
Breakaway Breakaway
q3 = 225º point point q4 = 315º
– 3.5811 – 0.4188
Fig. 11.9
TM

Example 11.7.21
Solution : Step 1 : P = 2, Z = 1, N = P = 2, P – Z = 1 branch to ¥
Starting points : s = 0, + 1 2
Breakaway
Terminating points : s = – 1, ¥ points
Step 2 : Sections of real axis
NRL
NRL
Step 3 and 4 : Not required as there
–1 0 +1
is only one asymptote which is
negative real axis. So centroid not
necessary.
Fig. 11.10 (a)
K (s + 1)
1 + G(s)H(s) = 1 + =0 i.e. s2 + s (K – 1) + K = 0 …(1)
s (s - 1)
-s2 + s
\ K = …(2)
s +1
dK (s + 1) ( -2s + 1) - ( - s 2 + s) (1)
\ = =0
ds (s + 1) 2
i.e. – 2s2 – 2s + s + 1 + s2 – s = 0 i.e. + s2 + 2s – 1 = 0

\ s = + 0.4142, – 2.4142 …Both valid breakaway points
From equation (2), K = + 0.1715, + 5.8284 …At breakaway points
From equation (1), the Routh's array is
1
\ From row of s , K–1=0
s2 1 K
\ Kmar = + 1
s1 (K – 1) 0 A(s) = s2 + K = 0 2
i.e. s = – 1
s0 K \ s=±j …Intersection points
Step 7 : Not required as no complex open loop poles.

Step 8 : Proof of circular nature of root locus :
For a complex point, s = s + jw
K[s + 1 + jw]
\ G(s)H(s) =
[s + jw] [s - 1 + jw]
w
tan -1 w ù ì w w ü
\ Ð G(s)H(s) = s+1 = tan -1 é - í tan -1 + tan -1
étan -1 w + tan -1 w ù ê
ë s + 1 ú
û î s s - 1 ýþ
êë s s - 1 úû
TM

é w+ w ù
é w ù ê ú A +B
= tan -1 - tan -1 ê s s - 1 ú
–1 –1 –1
…tan A + tan B = tan
êë s + 1 úû w w 1 - AB
êë1 - s ´ s - 1 úû
w ù é w ( 2s - 1) ù
= tan -1 é - tan -1 ê ú
êë s + 1 úû
êë s(s -1) - w2 úû
ì w w( 2s - 1) ü
ï s+1 - 2 ï
ï s(s -1) - w ï A -B
tan -1 í
–1 –1 –1
w( 2s - 1) ý
= …tan A – tan B = tan
w 1 + AB
ï1 + ´ ï
ïî s + 1 [s(s -1) - w2 ] ï
þ
For a points to be on the root locus, Ð G(s)H(s) = 180 º
ì w w( 2s - 1) ü
ï s+1 - 2 ï
ï s(s -1) - w ï
\ 180 º = tan -1 í ý
w s -
ï1 + w ´
( 2 1)
ï
ïî s + 1 [s(s -1) - w2 ] ï
þ
Taking tan of both sides and using tan 180 º = 0
w w( 2s - 1)
- = 0
s + 1 s(s -1) - w2
\ s (s – 1) – w2 – (2 s – 1) (s + 1) = 0 …w ¹ 0
2 2
\ s – s – w2 – 2 s – s + 1 = 0
Imj
Real
–1 0 +1
Breakaway
Breakaway point
point
s = + 0.4142
s = – 2.4142
K = + 0.1715
K = + 5.8284
Fig. 11.10 (b)

TM

\ – s 2 – 2 s – w2 + 1 = 0 i.e.
2
s + 2 s + w2 – 1 = 0
2 2 2 2
\ s + 2 s + 1 + w2 – 1 – 1 = 0 i.e. (s + 1) + (w – 0) = ( 2)
This is an equation of a circle with centre (– 1, 0) and radius 2.
Step 9 : The complete root locus is shown in the Fig. 11.10 (b) (See Fig. 11.10(b) on
previous page).
Exaple 11.7.22
Solution : Step 1 : P = 4, Z = 0, N = 4, P – Z = 4 branches to ¥
Starting = 0, – 3, – 1.5 ± j3 Terminating = ¥, ¥, ¥, ¥
Step 2 : Sections of real axis as shown in the
Fig. 11.11(a).
NRL
(2q + 1) 180º NRL
Step 3 : q =
P-Z
One
q = 0, 1, 2, 3 breakaway
q 1 = 45º, q 2 = 135º, q 3 = 225º, q 4 = 315º point possible
Fig. 11.11 (a)
å R.P
O.L poles å O.L zeros
of – R.P of
0 - 3 -1 .5 -1 .5
Step 4 : Centroid = = = – 1.5
P-Z 4
K
1 + G(s)H(s) = 0 i.e. 1+ =0
(
s ( s + 3) s 2 + 3s + 11.25 )
\ s 4 + 6s 3 + 20.25s 2 + 33.75s + K = 0 ... Characteristic equation
4 3 2
\ K = - s - 6s - 20.25s - 33.75s ... (1)
dK
= - 4s 3 - 18s 2 - 40.5s - 33.75 = 0
ds
Solving, s = – 1.5, – 1.5 ± j 1.8371
At s = – 1.5, K = +20.25 from equation (1) so it is valid.
Test s = – 1.5 + j 1.8371 by angle test.
ÐG(s)H(s) = ± (2q + 1) 180º ... At s = – 1.5 + j 1.8371
ÐK + j0
\
Ðs Ðs + 3 Ð s + 1.5 + j3 Ð s + 1.5 – j3 s = –1.5 + j1.8371
ÐK + j0
=
Ð - 1.5 + j1.8371 Ð1.5 + j1.8371 Ðj4.8371 Ð - j1.1629
0º
= = – 180º
129.23º 50 . 77 º 90º - 90º
TM

Thus both – 1.5 ± j 1.8371 are valid breakaway points.

To find K at – 1.5 ± j 1.8371, use magnitude condition.
G(s)H(s) s = -1.5 + j1.8371 = 1

K
\ = 1 i.e. K = + 31.6393
2. 3716 ´ 2. 3716 ´ 4.8371 ´ 1.1629
Thus both s = – 1.5 ± j 1.8371, occur simultaneously as breakaway points at K = + 31.6393.

From characteristic equation obtained in step 5,
s4 1 20.25 K From row of s 1 ,
s3 6 33.75 0 493 . 59
Kmar = = 82.265
6
s2 14.625 K 0
s1 493.59 - 6K 0 A(s) = 14.625 s 2 + Kmar = 0

14.625
\ s 2 = – 5.625
Step 7 : Angles of departure at – 1.5 ± j3

3
fP1 = 180º – tan -1 = 116.56º j3
1 .5
fP1
3 fP3
fP2 = 90º, fP3 = tan -1 = 63.43º
1 .5 –3 0
fP2
å fP = 270º, å fZ = 0º – j3
\ f = å fP – å fZ = 270º Fig. 11.11 (b)
\ fd = 180º – f = – 90º at s = – 1.5 + j3
and fd = + 90º at s
= – 1.5 – j3
0 –3 –1.5+j3 –1.5–j3
Step 8 : Nature of root
locus branches is shown in
the Fig. 11.11 (c). k = 20.25 –1.5 Both breakaway
points occur
The overall root locus is 1
j1.837 simultaneously
shown in the Fig. 11.11 (d) k = 31.6393 –1.5+
on graph paper. –1.5+j1.8371
¥ ¥ ¥ ¥
q1 = 45º q2 q3 q4
= 135º = 225º = 315º
Fig. 11.11 (c)
TM

+ j3 jw
q2 = 135º
q1 = 45º
fd = –90º
+ j 2.37
–1.5 + j1.83
Breakaway point
Kmar = 82.26
Breakaway point = –1.5
and centroid
s
–3 0
Scale on both axes
2 Units = 1.5 Units
–1.5 + j1.83
Breakaway point
–j 2.37
fd = +90º
q3 = 225º q4 = 315º
– j3
Fig. 11.11 (d)
Step 9 : Comment on stability
For 0 < K < 82.26 system is stable.

At K = 82.26 system is marginally stable.
For K > 82.26 system is unstable.
Example 11.7.23
Solution : Step 1 : P = 4, Z = 0, N = 4, All branches approaching to ¥. Starting points
–3 ± 9 – 12
s = 0, –3 and i.e. 0, –3, –1.5 ± j 0.866.
2
Step 2 : Pole-zero plot is as shown.
Minimum one breakaway point exists between 0 j 0.866
and – 3.
NRL NRL
Step 3 and 4 : Real parts of poles are not –3 –1.5 0
changed.
–j 0.866
Centroid –1.5 and angles q 1 = 45°, q 2 = 135°,
q 3 = 225°, q 4 = 315°.
TM


1 + G(s)H(s) = 0
K
1+ = 0
s (s + 3) (s 2 + 3 s + 3)
s 4 + 6s 3 + 12s 2 + 9s + K = 0
\ K = – s 4 – 6s 3 – 12s 2 – 9s ... (1)

dK
\ = – 4s 3 – 18s 2 – 24s – 9 = 0 i.e. 4s 3 + 18s 2 + 24s + 9 = 0
ds
Solving, s = – 1.5, – 0.633, – 2.366
\ Breakaway points are = –1.5, – 0.633, –2.366 and all are valid.
For s = –1.5, K = 1.6875 ... Using equation (1)
s = - 0.633 , K= 2.25 ü
ý These two occur simultaneously. ... Using equation (1)
s = - 2. 366 , K= 2.25 þ

Characteristic equation : s 4 + 6s 3 + 12s 2 + 9s + K = 0
Routh’s array,
s4 1 12 K \ 94.5 – 6K = 0
Kmar = 15.75
s3 6 9 0
A(s) = 10.5s 2 + K = 0
s2 10.5 K 0
94.5 – 6 K \10.5s 2 + 15.75 = 0
10.5
s1 0 s2 = – 1.5
s0 K \ s = ± j 1.224
Step 7 : Angle of departure.

fd at –1.5 + j 0.866 is – 90º. While fd at –1.5 – j 0.866 is + 90º.
Step 8 : In this problem s = –1.5 will occur as a breakaway point first as K = 1.6875 for
this breakaway point. But branches from complex poles will reach at s = –1.5 rather than
branches from real open loop poles. This is because the complex poles are more closer to
s = – 1.5 than the real poles. The remaining breakaway points will occur later
simultaneously. These breakaway points will exist after the complex poles branches break
into real branches at s = –1.5. The root locus is shown in the Fig. 11.12.
TM

jw
q2 K = 2.25
q1
¥ Breakaway
point = –0.633 ¥
j 1.224
K = 2.25
Breakaway j0.866
point = –2.366
Kmar = 15.75
s
–3 0
–j 0.866
–j 1.224
¥ Centroid & B.P. ¥

q3 at s = –1.5
q4
K = 1.6875
Fig. 11.12
Step 9 : For 0 < K < 15.75 system is stable.
At K = 15.75 system is marginally stable.
For K > 15.75 system is unstable.
Example 11.7.24
K(s + 4) 2 breakaway
Solution : G(s)H(s) = points exist
(s + 2) (s + 3)
Step 1 : P = 2, Z = 1, N = P = 2 NRL NR L
X X
P – Z = 1 branch approach to ¥ –4 –3 –2
Starting points : s = –2, –3
Terminating points : s = – 4, ¥
Step 2 : Sections of real axis.
Step 3 : Angle of asymptote

(2q + 1)180°
q = ,q=0
P-Z
\ q1 = 180°
Step 4 : Centroid not required as only one asymptote exists.
K(s + 4) - s 2 - 5s - 6
1+ = 0 i.e. s 2 + 5s + 6 + K(s + 4) = 0 i.e. K = …(1)
(s + 2)(s + 3) (s + 4)
dK (s + 4)( -2s - 5) - ( - s 2 - 5s - 6)(1)
\ = =0
ds (s + 4) 2
\ -2s 2 - 8s - 5s - 20 + s 2 + 5s + 6 = 0 i.e. s 2 + 8s + 14 = 0
\ s = – 2.5857, – 5.4142
TM

Both are valid breakaway points.

At s = – 2.5857, K = + 0.1715 …From equation (1)
At s = – 5.4142, K = + 5.8284 …From equation (1)
1 + G(s)H(s) = 0 i.e. s 2 + 5s + 6 + K(s + 4) = 0
\ s 2 + s(5 + K) + (6 + 4K) = 0
s2 1 6 + 4K From row of s 1 , for Kmar,

s1 5+K 0 5+K=0
s0 6 + 4K i.e. Kmar = – 5
As it is negative, root locus does not
intersect imaginary axis and lies
completely in left half of s-plane.
Step 7 : Angle of departure not required as no complex poles.

Step 8 : The root locus is as shown in the Fig. 11.13.
Imj
Scale :
On both axes
2 Units = 1 Unit
Breakin
point
Breakaway
–5.41
point
–2.58
q1 = 180º
X X
¥ 0 Real
–4 –3 –2
Fig. 11.13
The system is absolutely stable as for any value of K > 0, all the roots lie in the left half
of s-plane.
TM

Example 11.13.4
Solution : Without derivative rate feedback controller, compare denominator with
s 2 + 2xwn s + wn2 ,
w2n = 16 i.e. wn = 4 and 2xwn = 1.6 i.e. x = 0.2
æ 1 - x2 ö
wd = wn 1 - x 2 = 3.9192 rad/sec, q = tan -1 çç ÷ = 1.3694 rad
÷÷
ç x
è ø
p-q p
\ Tr = = 0.4521 sec, Tp = = 0.8016 sec
wd wd
1- x 2
% Mp = e - px / ´ 100 = 52.662 %
C(s) 16 16
For = , G(s) = , H(s) = 1
R(s) 2
s + 1.6s + 16 s(s + 1.6)
s ´ 16 A 1
\ Kv = Lim sG(s)H(s) = Lim = 10, e ss = = = 0.1
s ®0 s ®0 s(s + 1.6) K v 10
Now use rate feedback controller,
C(s) G(s) 16
\ = =
R(s) 1 + G(s)H(s) s 2 + s(1.6 + 16 K ) + 16
t
\ w2n = 16 i.e. wn = 4 … Remains same
1.6 + 16 Kt
2xwn = 1.6 + 16 Kt i.e. x= = 0.8 (given)
2´ 4
\ Kt = 0.3 … For x = 0.8
\ wd = wn 1 - x 2 = 4 ´ 1 -( 0.8) 2 = 2.4 rad/sec
16
R 1 s(s + 1.6) C
– –
s Kt
Rate feedback controller

Fig. 11.14 (a)
16
16 R C
G(s) = s(s + 1.6 + 16 K ) – s(s + 1.6 + 16 K t)
t
H(s) = 1
Fig. 11.14 (b)
TM

é 1 - x2 ù
\ q = tan -1 ê ú = 0.6435 rad
ê x ú
ë û
p-q p
\ Tr = = 1.04 sec, Tp = = 1.308 sec.
wd wd
% Mp = e - px / 1- x 2
´ 100 = 1.516 %
16
Kv = Lim sG(s)H(s) = Lim s ´ = 2.5
s®0 s®0 s(s + 1.6+ 16 ´ 0.3)
A 1
\ e ss = = = 0.4
Kv 2.5
The steady state error increases but overshoot decreases drastically.
Example 11.13.5
Solution : i) With K0 = 0, the system is,
KA
C(s) s(s + 2) R(s) 1 C(s)
= KA
R(s) KA s(s + 2)
1+ –
s(s +2)
KA
=
S2 + 2s + KA
Comparing denominater with s 2 + 2x wn s + w2n ,
w2n = KA i.e. wn = KA = 10 = 3.16 rad / sec
2
2xw2n = 2 i.e. x = = 0.3162
2 ´ 10
KA 10
G(s)H(s) = =
s(s + 2) s(s + 2)
10
\ Kv = Lim sG(s)H(s) = =5
s®0 2
A 1
\ e ss = = = 0.2 ... For unit ramp A = 1
Kv 5
ii) With K0 , the system becomes ,

10
2
C(s) s + s(2 + K0 ) 10
\ = =
R(s) 10 2
s + s (2 + K0 ) + 10
1+
s 2 + s (2 + K0 )
\ w2n = 10 i.e. wn = 10 rad/sec

TM

1
s2 + s(2 +K0)
Minor loop
1
C(s) R(s) s(s + 2) C(s)
R(s) 1 KA = 10
KA = 10 sK0
s(s + 2) 1+
– – » – s(s + 2)
sK0
2 + K0
and 2xwn = 2 + K0 i.e x=
2 10
2 + K0
But given x = 0.6 hence, 0.6 =
2 10
\ K0 = 1.7947
10 10
\ G(s)H(s) = =
s2 + s(2 + 1.7949) s(s + 3.7949)
10
\ Kv = Lim sG(s)H(s) = = 2.6351
s®0 3.7949
A 1
\ e ss = = = 0.3795
Kv 2.6351
Example 11.13.6
Solution : Without controller,
100
G(s) = H(s) = 1
s (s + 12)
C(s) 100
\ =
R(s) 2
s + 12s + 100
\ w2n = 100 \ wn = 1 0
2x wn = 12 \ x = 0.6
\ wd = wn 1 - x 2 = 10 ´ 0.8 = 8 rad/sec
- p x 1 – x2
\ % Mp = e = 9.47 %
4
Ts = = 0.666 sec
x wn
TM

With controller we get,
R(s) 100 C(s)

s (s + 12)
– +
1
s
30
æ 1 + s ö 100
ç 30 ÷ø (s + 30) ´ 3.33
G(s) = è = , H(s) = 1
s (s + 12) s(s + 12)
(s + 30) 3.33
C(s) s(s + 12) 3.33(s + 30)
= =
R(s) (s + 30) 3.33 2
s + 12s + 3.33s + 100
1+
s (s + 12)
3.33 (s + 30)
=
s2 + 15.33s + 100
\ w2n = 100 \ wn = 10
2x wn = 15.33
15.33
\ x = = 0.7665
2 ´ 10
\ x is improved, wd = wn 1 - x 2 = 10 1 - (0.7665) 2
= 6.4224 rad/sec
1 - x2
% Mp = e - p x ´ 100 = 2.353 %
Overshoot decreased to 2.3 % from 9.47 %.
4
Ts = = 0.5218 sec
x wn
TM

Comparison : Following figure shows comparison between system with controller and
system without controller.
9.47%
c(t) c(t)
2.353%
1 1
0.66 sec time time

0.5218 sec
Without controller
With controller
qqq
TM

Frequency Response Methods
12 (Bode Plot Method)
Solutions of Selected Examples for Practice
Example 12.6.4
Solution : The open loop transfer function is
K
G(s) = and H(s) = 1
s(1 + sT)
K
C(s) G(s) s(1 + sT) K K/ T
\ = = = =
R(s) 1 + G(s)H(s) K Ts + s + K s + 1 s + K
2 2
1+
s(1 + sT) T T
K
Comparing this with standard form, wn = ... (1)
T
1
x = ... (2)
2 KT
Now 0.2 = e - p x / 1- x 2
Mp = 20 % i.e.
Solving, x = 0.455 and wr = 6 rad/sec ... Given
\ wr = wn 1 - 2 x 2
\ 6 = wn 1 - 2 ´ (0.455) 2
\ wn = 7.8382 rad/sec
K K
Using equation (1), 7.8382 = i.e. 61.437 =
T T
1 1
2 KT KT
Substituting from equation (1) into equation (2),
1
61.437 = i.e. T 2 = 0.0196 i.e. T = 0.14
0.8281T 2
\ K = 8.6133
(12 - 1)
TM

Control System Engineering 12 - 2 Frequency Response Mehtods (Bode Plot Method)
1 1
Mr = = = 1.234
2 x 1 - x2 2 ´ 0.455 1 - (0.455) 2
Example 12.6.5
Solution : For the second order system,
1- x2
Mp = 10 % i.e. 10 = e - p x ´ 100
-px px
= ln ( 0.1) i.e. = 2.3025
1 - x2 1 - x2
Solving, x = 0.5911
p-q 1 - x2
Tr = where q = tan -1 = 0.9383 rad
wd x
p - 0.9383 2.2032
\ 0.1 = i.e. wn =
wn 1 - x 2 0.1 ´ 1 - ( 0.5911) 2
\ wn = 27.3158 rad/sec
From these values, Mr and B.W. can be calculated as,
1 1
Mr = = = 1.0487
2
2 x 1 -x 2 ´ 0.5911 ´ 1 - ( 0.5911) 2
B.W. = wn ´ 1 - 2 x 2 + 2 - 4 x 2 + 4 x 4
= 27.3158 ´ 1 - 2 ´ ( 0.5911) 2 + 2 - 4 ´ ( 0.5911) 2 + 4 ( 0.5911) 4
= 31.686
Example 12.6.6
Solution : G(s) = K
, H(s) = 1
s(1 + st)
K K
C(s) s(1 + st) K t
\ = = =
R(s) K ts 2 +s+K 2 +1 + K
1+ s s
s(1 + st) t t
Comparing denominator with s 2 + 2 x wn s + w2n
K K
\ w2n = i.e. wn = ...(1)
t t
1 1
and 2xwn = i.e. x = ...(2)
t 2 Kt
TM

1
Mr = = 1.06
2x 1 – x2
\ x 1- x 2 = 0.4716 i.e. x 2 (1 - x 2 ) = 0.2225
\ x 4 - x 2 + 0.2225 = 0 i.e. x 2 = 0.6658, 0. 3341

\ x = 0.8159, 0.578 but x cannot be more than 0.707.
\ x = 0.578
K K
Using equation (1), wn = i.e. 12 = i.e. K = 144 t
t t
1 1
2 144t ´ t 2 ´ 12 ´ t
\ t = 0.072, K = 10.3806
B.W. = wn 1 - 2 x 2 + 2 - 4 x 2 + 4 x 4
= 12 1 - 2 ´ ( 0 . 578) 2 + 2 - 4 ´ ( 0 . 578) 2 + 4 ´ ( 0 . 578) 4
= 14.1246 rad/sec
Example 12.6.7
Solution : From the table Mp = 0.12 i.e. 12 % and Tp = 0.2 sec.
2 1- x 2
But, Mp = 100 e - p x / 1- x i.e. 0.12 = e - p x /
Solving, x = 0.5594
p p p
Tp = = i.e. 0.2 =
wd
wn 1 - x 2 wn 1 - ( 0.5594) 2
wn = 18.9504 rad/sec.
1 1
\ Mr = = = 1.078
2
2x 1-x 2 ´ 0.5594 ´ 1 - (0.5594) 2
wr = wn 1 - 2 x 2 = 11.5914 rad/sec
wb = wn 1 - 2 x2 + 2 - 4 x 2 + 4 x 4 = 22.1162
Example 12.6.8
Solution : For the given pole-zero diagram,
K 104
T(s) = =
(s + 2 - j10) (s + 2 + j10) s 2 + 4 s + 104
TM

Comparing denominater with s 2 + 2 x wn s + w2n ,

w2n = 104 i.e. 10.198 rad/sec.
2 x wn = 4 i.e. x = 0.1961
1 1
Mr = = = 2.6
2x 1 - x2 2 ´ 0.1961 ´ 1 - (0.1961) 2
wr = wn 1 – 2 x 2 = 10.198 ´ 1 - 2 ´ (0.1961) 2 = 9.4136 rad/sec.
wb = B.W. = wn 1 - 2 x 2 + 2 - 4 x 2 + 4 x 4 = 15.4066
Example 12.6.9
Solution : Mp = 50 %, Td = Time period = 0.2 sec.
1 - x2 1 - x2
Now, Mp = 100 e - p x i.e. 0.5 = e - p x
-px x
ln (0.5) = i.e. 0.2206 =
1 - x2 1 - x2
Solving, x = 0.2154
1
And Td = 0.2 sec. i.e. fd = = 5 Hz
Td
\ wd = 2 p fd = 10 p rad / sec = wn 1 - x 2
10 p
\ wn = = 32.1711 rad/sec
1 - (0. 2154) 2
1
i) Mr = = 2.377
2x 1 - x2
ii) wr = wn 1 - 2 x 2 = 30.6421 rad/sec.
iii) B.W. = wn 1 - 2 x2 + 2 - 4 x2 4 x4 = 48.336
Example 12.15.17
Solution :
4 (1 + 0.5 s)
Step 1 : G(s) = … Time constant form
æ s 2 ö÷
s (1 + 2s) ç 1 + 0.05 s +
ç 64 ÷
è ø
Step 2 : Factors
i) K = 4 i.e. 20 Log 4 = 12 dB
1
ii) i.e. one pole at the origin. So magnitude plot is straight line of slope
s
– 20 dB/dec, passing through intersection point of w = 1 and 0 dB.
TM

1 1
iii) i.e. simple pole i.e. T1 = 2 i.e. wC1 = = 0.5
1 + 2s T1
So straight line of 0 dB till w = 0.5 and then slope of – 20 dB/dec.

1
iv) 1 + 0.5 s i.e. simple zero i.e. T2 = 0.5 i.e. wC2 = =2
T2
So straight line of 0 dB till w = 2 and then slope of + 20 dB/dec.

1 1
v) , quadratic pole, comparing with
æ s2 ö 2x s2
ç 1 + 0.05 s + ÷ 1+ s+
ç 64 ÷ wn w2
è ø n
w2n = 64 i.e. wC = wn = 8 rad / s
3
Straight line of slope – 40 dB/dec for w ³ 8.
The resultant slope table is,
Range of w 0 < w< 0.5 0.5 £ w < 2 2 £w < 8 8 £w < ¥
Resultant slope –20 – 20 –20 = – 40 –40 + 20 = – 20 –20 – 40 = – 60

in dB/dec
4(1 + 0.5 jw)

Step 3 : Phase angle plot G (jw) =
éæ w2 ö÷ ù
jw(1 + 2 jw)êç 1 – + 0.05 jwú
ç 64 ÷
êëè ø úû
w 1 – tan –1 2 w + tan –1 0.5 w é ù fR

jw ê 0.05w ú
– tan –1 ê ú
ê w2 ú
êë1 – 64 úû
0.2 –90º – 21.8º + 5.71º – 0.573º – 106.67 º
4 –90º – 82.87º + 63.43º – 14.93º – 124.37 º
7 –90º – 85.91º + 74.05º – 56.19º – 158.05 º
8 –90º – 86.42º + 75.96º – 90º – 190.45 º
¥ –90º – 90º + 90º – 180º – 270 º
Step 4 : The Bode plot is shown in the Fig. 12.1.
TM

Semi-log paper (5 cycles ´ 1/10)
1 2 3 4 5 6 7 891 2 3 4 5 6 7 891 2 3 4 5 6 7 891 2 3 4 5 6 7 891 2 3 4 5 6 7 891
MR
in dB
Scale
On Y axis : 1 unit = 30º, 20 dB
+60
– 60
dB/dec
+40 – 20 dB/dec – 40 dB/dec
– 40
Control System Engineering
dB/dec –20 dB/dec

+20
– 20 K=4
dB/dec
0 dB
G.M. = +22 dB
–20
fR
in deg – 60 dB/dec
12 - 6
(1/s)
TM
–90º
Fig. 12.1
–120º
P.M.
–150º +64º
System stable

–180º
–210º
wgc
–240º 1.3
–270º
wpc
= 7.4
0.1 1 10 100 Log w

Frequency Response Mehtods (Bode Plot Method)
Example 12.15.18
Solution : Step 1 : G(s)H(s) in time constant form.
Step 2 : In this example, K is unknown. So draw the magnitude plot without K. The
effect of K is to shift the magnitude plot upwards or downwards by 20 Log K dB. Thus
shift the magnitude plot so as to match the given specifications. Equating this shift
required to 20 Log K, the required value of K can be determined.
The various factors other than K are,
1
1) , pole at the origin, straight line of slope –20 dB/dec passing through intersection
s
point of w = 1 and 0 dB.
1 1
2) , simple pole, T1 = 0.5, wC = =2
1 + 0.5 s 1 T1
Straight line of slope –20 dB/dec for w ³ 2.

1 1
3) , simple pole, T2 = 0.2, wC = =5
1 + 0.2 s 2 T2
Straight line of slope –20 dB/dec for w ³ 5.
Range of w 0<w <2 2 £w < 5 5 £w < ¥
Resultant slope in – 20 – 20 – 20 = – 40 – 40 – 20 = – 60
dB/dec
K
Step 3 : Phase angle table : G(jw)H(jw) =
jw(1 + 0.5 jw)(1 + 0.2 jw)
w 1 – tan –1 0.5 w – tan –1 0.2 w fR

jw
0.2 – 90º – 5.71º – 2.29º – 98º
2 – 90º – 45º – 21.8º – 156.8º
4 – 90º – 63.43º – 38.65º – 192.08º
5 – 90º – 68.19º – 45º – 203.19º
¥ – 90º – 90º – 90º – 270º
Step 4 : Sketch the Bode plot as shown in the Fig. 12.2 from which wpc = 3.3 rad/sec,
which is not affected by value of K. (See Fig. 12.2 on next page)
TM

1 2 3 4 5 6 7 891 2 3 4 5 6 7 891 2 3 4 5 6 7 891 2 3 4 5 6 7 891 2 3 4 5 6 7 891
MR
in dB
Scale
On Y axis : 1 unit = 30º, 20 dB
+60
– 60
dB/dec
– 20 dB/dec
+40
– 40
dB/dec
+20 – 40 dB/dec Line of w

pc
– 20
dB/dec
C'
0 dB
A' 6 dB Line for G.M. = + 6 dB Shift A – A'
C = 8 dB upwards
–20 A for G.M. = 6 dB
– 60 dB/dec
fR 1
in deg s
12 - 8
TM
–90º
Plot without K
Fig. 12.2
–120º
B
Line for P.M. = +60º
P.M.
–150º Shift C'– C = 4 dB
+ 60º
downwards
for P.M. = 60º

–180º
–210º
New wgc
required for
P.M. = 60º
–240º
wpc
–270º 3.3
0.1 1 10 100 Log w

Steps to find K for given G.M. :
1) Draw the horizontal line below 0 dB at a distance of given G.M., till it intersects
vertical line of wpc . This is point A ¢.
2) Draw the vertical line from wpc , till it intersects magnitude plot without K. This is
point A.
3) The point A must be at A ¢ to match given G.M. Hence A to A ¢ is the shift required
i.e. contribution by K which is 20 Log K dB.
4) Upward shift must be taken positive and downward shift negative.
20 Log K = Shift (AA ¢)
In this example, 20 Log K = + 8 dB (Upwards) … For G.M.=6 dB

Shift is positive as upwards hence K = 2.511 for G.M. = 6 dB.
Steps to find K for given P.M. :
1) Draw the horizontal line above –180º line at a distance of given P.M., till it
intersects phase angle plot. This is point B.
2) Draw the vertical line from B, till it intersects the magnitude plot without K. This is
point C¢.
3) This C¢ must be on 0 dB line at C, as the vertical line through point B must be new
wgc line for given P.M.
4) Thus distance of C¢ from 0 dB line is the shift required to satisfy given P.M. This
must be 20 Log K dB.
5) Upward shift (C¢ C) must be taken positive while downward shift (C¢ C) must be
taken negative.
20 Log K = Shift (C¢ C)
In this example, 20 Log K = – 4 dB (Downwards) …for P.M. = + 60º

Shift is negative as downwards hence K = 0.631 for P.M. = + 60º.
Note that two different values of K are required to satisfy two different specifications.
Example 12.15.19
8 (1 + 0.2s)
Solution : Step 1 : G( s)H( s) = … Time constant form
s 2 (1 + 0.02s)
Step 2 : Analysis of factors
i) K= 8, 20 Log K = 18.06 dB, straight line parallel to Log w
TM

1
ii) , two poles at origin , straight line of slope – 40 dB/dec passing through
s2
intersection point of w = 1 and 0 dB.
1
iii) (1+0.2 s), simple zero, T1 = 0.2, wC1 = =5
T1
Straight line of slope + 20 dB/dec for w ³ 5.

1 1
iv) , simple pole, T2 = 0.02, wC = = 50
1 + 0.02s 2 T2
Straight line of slope –20 dB/dec for w ³ 50
Range of w 0<w <5 5 £ w < 50 50 £ w < ¥
Resultant slope in – 40 – 40+20 = – 20 – 20 – 20 = – 40

dB/dec
8 (1 + 0.2 jw)
(jw) 2 (1 + 0.02 jw)
w 1 + tan –1 0.2 w –tan –1 0.02 w fR

( jw) 2
0.5 –180º + 5.71º – 0.57º –174.86º
5 –180º + 45º – 5.71º –140.71º
10 –180º + 63.43º – 11.31º –127.88º
50 –180º + 84.28º – 45º –140.72º
¥ –180º + 90º – 90º –180º
Step 4 : The Bode plot is shown in the Fig. 12.3 (See Fig. 12.3 on next page) which
includes the values of wgc , wpc , G.M. and P.M.
Example 12.15.20
4 (1 + s)
Solution : Step 1 : G( s)H( s) = … Time constant form
s 2 (1 + 0.25 s)æç 1 + ö÷
s
è 12 ø
TM

1 2 3 4 5 6 7 891 2 3 4 5 6 7 891 2 3 4 5 6 7 891 2 3 4 5 6 7 891 2 3 4 5 6 7 891
MR
in dB
Scale
On Y axis : 1 unit = 20 dB,15º
60
– 40 dB/dec
40
wgc = 2.7
20 K = 8 (18 dB)
0 dB
– 20 – 20 dB/dec
G.M. +¥
– 40
– 40 dB/dec
12 - 11
TM
fR
in deg –90º
Fig. 12.3
–105º
1
2
–120º s
wgc = 2.7 rad/s
wpc 0

–135º
G.M.= + ¥ dB
P.M.= + 24º
–150º System stable
–165º
wpc 0 P.M. = + 24º
–180º
0.1 1 10 100 Log w


1) K = 4, 20 Log K = 12.04 dB, straight line parallel to Log w axis
1
2) , two poles of origin, staight line of slope – 40 dB/dec passing through
s2
intersection of w = 1 and 0 dB.
1
3) (1+s), simple zero, T1=1, wC1 = = 1, straight line of slope + 20 dB/dec for w ³ 1.
T1
1 1
4) , simple pole, T2 = 0.25, wC2 = = 4 , straight line of slope –20 dB/dec
(1 + 0.25 s) T2
for w ³ 4.
1 1 1
5) , simple pole , T3 = , wC3 = = 12 , straight line of slope –20 dB/dec for
æ1 + s ö 12 T3
ç 12 ÷
è ø
w ³ 12.
Range of w 0<w <1 1 £w < 4 4 £ w < 12 12 £ w < ¥
Resultant slope – 40 – 40 + 20 = – 20 – 20 – 20 = – 40 – 40 – 20 = – 60
in dB/dec
4(1 + jw)
w
(jw) 2 (1 + 0.25 jw)æç 1 + j 12 ö÷
è ø
w 1 + tan –1 w – tan –1 0.25 w w fR

– tan –1
( jw) 2 12
0.4 –180º + 21.8º – 5.71º – 1.91º – 165.82º

4 –180º + 75.96º – 45º – 18.43º – 167.47º
8 –180º + 82.87º – 63.43 – 33.7º –194.25º
12 –180º + 85.23º – 71.56º – 45º – 211.33º
¥ –180º +90º – 90º – 90º – 270º
Step 4 : The Bode plot and the values of G.M. snd P.M. are shown in the Fig. 12.4.
TM

1 2 3 4 5 6 7 891 2 3 4 5 6 7 891 2 3 4 5 6 7 891 2 3 4 5 6 7 891 2 3 4 5 6 7 891
MR
in dB
Scale
On Y axis : 1 unit = 20 dB , 30º
60
– 60 – 40 dB/dec wgc = 4
dB/dec
40
– 40
dB/dec – 20 dB/dec wpc = 5.6

20
K = 4 (12 dB)
– 20
dB/dec
0 dB
G.M. = 6 dB
– 20 – 40 dB/dec
fR
– 60 dB/dec
in deg wgc = 4 rad/s
12 - 13
TM
wpc = 5.6 rad/s
–120º G.M. = +6 dB
Fig. 12.4
1 P.M. = +13º
2 System stable
–150º s
P.M. = 13º
–180º

–210º
–240º
–270º
0.1 1 10 100 Log w

Example 12.15.21
Solution :
2
æ1 + s ö
2
æ1 + j w ö
ç ÷ ç ÷
è Aø
G(s)H(s) = è
Aø
i.e. G(jw )H(jw ) =
s3 (jw) 3
2
æ 1 + w2 ö æ 2 ö
ç ÷ ç1 + w ÷
ç A2 ÷ ç ÷
è ø è A2 ø
Now |G(jw )H(jw )| = M = =
w3 w3
For phase margin, w = wgc , and at wgc M = 1
2
wgc
1+ 2
wgc
\ 1 = A2 i.e. 3 =
wgc 1+ i.e. 3 A2 =
wgc A 2 + wgc
2
3
wgc A 2
Now P.M. = 180°+Ð G(jw) H(jw)|w = w gc ... (1)

wgc
\ 50° = 180° + 2 tan -1 - 270°
A
wgc wgc wgc
\ 2 tan -1 = 140° i.e. tan -1 = 70° i.e. = 2.7474 ... (2)
A A A
wgc = 2.7474 A
Substituting in equation (1), we get
(2.7474 A) 3 A 2 = A 2 + (2.7474 A) 2 i.e. 20.74 A 5 = 8.5482 A 2
\ A3 = 0.4121 i.e. A = 0.745 for P.M. = + 50°
Example 12.15.22
Solution : Step 1 : Time constant form of G(s).
s s
50 ´ 3 ´ (1 + ) 3(1 + )
G(s) = 3 = 3
s ´ 50 ´ (1 + 0.1s + 0.02 s 2 ) 5 [1 + 0.1s + 0.02 s 2 ]
i) K = 3, 20 Log K = 20 Log 3 = 9.54 dB
Straight line parallel to Log w axis, above 0 dB line.
1
ii) , one pole at the origin.
s
Straight line of slope – 20 dB/dec, passing through intersection of w = 1 and 0 dB.
s 1 1
iii) (1 + ) , simple zero, T1 = , wC1 = =3
3 3 T1
Straight line of slope + 20 dB/dec for w ³ wC1 i.e. w³ 3.
TM

1
iv) , quadratic pole
(1 + 0.1 s + 0.02 s 2 )
The original quadratic is s 2 + 5 s + 50. Comparing with s 2 + 2 xwn s + wn2
wC2 = wn = 50 = 7.071
Straight line of slope – 40 dB/dec for w ³ 7.071.
5
2 x wn = 5 i.e. x = = 0.3535
2 ´ 50
\ correction at wC2 = –20 Log 2 x = + 3 dB at w = 7.071.
Resultant slope table :
Range of w Resultant slope
0 < w < wC1(3) – 20 dB/dec
3 < w < wC2 (7.07) - 20 + 20 = 0 dB/dec
7.07 < w < ¥ 0 – 40 = – 40 dB/dec
Step 3 : Phase angle table

w w
3(1 + j ) 3 (1 + j )
G(jw) H(jw) = 3 = 3
j w [ 1 + 0.1jw + 0.02(jw) 2 ] jw [(1 - 0.02w2 ) + j0.1w]
w 1 w 0.1 w fR
jw + tan-1 - tan-1 [ ]
3 1- 0.02 w2
0.5 – 90º + 9.46º – 2.87º – 83.4º
5 – 90º + 59.03º – 45º – 75.9º
7.07 – 90º + 67. 01º – 90º – 113º
20 – 90º +81.46º – 164.05º – 172.59º
¥ – 90º + 90º – 180º – 180º
Step 4 : Sketch the Bode plot. From the Bode plot,

G. M. = + ¥ dB, P.M. = + 62º The system is absolutely stable.
(See Fig. 12.5 on next page)
Example 12.15.23
Solution : Step 1 : G(s)H(s) in the time constant form.
Step 2 : The factors are,
i) K = 20, 20 Log K = 26.02 dB, straight line parallel to Log w axis.
ii) s, a zero at origin, straight line of slope + 20 dB/dec passing through intersection
point of w = 1 and 0 dB.
TM

SEMI-LOG PAPER (5 CYCLES X 1/10)
MR 1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
dB +60
–20 dB/dec 0 dB/dec
Scale
+40 On Y axis, 1 unit = 20 dB, 15º
– 40 With correction
dB/dec wC1 of +3 dB at w = 7.07
+20
– 20
K = 3 (9.54 dB)
dB/dec
0 dB
– 40 dB/dec
–20 wC2
(1/s)
f R in deg Without
correction
–75º
12 - 16
TM
–90º wgc = 7.6
Fig. 12.5
–105º
–120º
–135º

P.M.
–150º +62º
–165º
wpc ® ¥
–180º
–195º
0.1 1 wgc = 3 10 100 Log w
1 1 1
iii) , simple pole, , T1 = 0.1, wC1 = = 10 rad/sec. Straight line of
1 + 0.1 s 1 + T1 s T1
slope – 20 dB/dec for w ³ 10 rad sec .
The resultant slope table is,
Range of frequency Resultant slope
0 < w < wC1 (10) + 20 dB/dec
10 < w < ¥ + 20 – 20 = 0 dB/dec

20 jw
G( j w)H( j w) =
(1 + 0.1 jw)
w jw - tan - 1 0.1 w fR
0.1 + 90º – 0.57º + 89.43º

1 + 90º – 5.71º + 84.28º
10 + 90º – 45º + 45º
100 + 90º – 84.28º + 5.71º
¥ + 90º – 90º 0º
Step 4 : The Bode plot is shown in the Fig. 12.6. (Refer Fig. 12.6 on next page)
Example 12.15.24
Solution : Step 1 : Obtain time constant form of G(s)H(s).
40 ´ 5 æç 1 + ö÷
s
è 5 ø 10(1 + 0.2s)
G(s)H(s) = =
s(1 + 0.1s)(1 + 0.5s)
s ´ 10 ´ 2 ´ æç 1 + ö÷ æç 1 + ö÷
s s
è 10 ø è 2 ø
Step 2 : Factors
1) K = 10, 20 Log K = 20 Log 10 = 20 dB
1
2) , one pole at origin.
s
Straight line of slope –20 dB/dec passing through intersection of w = 1 and 0 dB.
1 1
3) , simple pole, T1 = 0.5, wC 1 = = 2 rad/sec
(1 + 0.5s) T1
Straight line of slope –20 dB/dec for w > 2.

1
4) (1 + 0.2s), simple zero, T2 = 0.2, wC 2 = = 5 rad/sec.
T2
TM

MR in dB 1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
+80
Scale
+60 wC1 = 10 0 dB/dec On Y-axis :
1 unit = 20 dB, 15º
s(+20 dB/dec)
+20 dB/dec
+40
K = 20, 26.02 dB
+20 dB
0 dB
– 20
fR in
degrees
12 - 18
TM
+90º
Fig. 12.6
+75º
+60º

+45º
+30º
+15º
0º
0.1 1 10 100 Log w

Straight line of slope +20 dB/dec for w > 5.

1 1
5) , simple pole, T3 = 0.1, wC 3 = = 10 rad/sec.
(1 + 0.1s) T3
Straight line of slope –20 dB/dec for w > 10.
Resultant slope table :
0 < w < wC1 (2) – 20 dB/dec (Pole at origin)
2 < w < wC 2 (5) – 20 – 20 = – 40 dB/dec
5 < w < wC 3 (10) – 40 + 20 = – 20 dB/dec
10 < w < ¥ – 20 – 20 = – 40 dB/dec

10(1 + 0.2 jw)
G( jw) H( jw) =
jw(1+ 0.1 jw)(1 + 0.5 jw)
w 1 -tan -1 0.5 w + tan -1 0.2 w -tan -1 0.1 w fR

jw
0.2 - 90° - 5.71° + 2.29 ° - 1.14 ° - 94.56 °
5 - 90° - 68.19 ° + 45 ° - 26.56 ° - 139.75 °
10 - 90° - 78.69 ° + 63.43 ° - 45 ° - 150.25 °
50 - 90° - 87.71° + 84.29 ° - 78.69 ° - 172.11°
¥ - 90° - 90 ° + 90 ° - 90 ° - 180 °
Step 4 : The Bode plot is shown in the Fig. 12.7. (See Fig. 12.7 on next page)
From the plot, wgc = 4.4 rad/sec, wpc = ¥, G.M. = + ¥ dB P.M. = + 42 °
As G.M. = + ¥ dB, the system is absolutely stable in nature.
TM

1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
+ 80
Scale :
+ 60 – 20 dB/dec on Y-axis : 1 unit = 20 dB, 30°
+ 40 wC1 – 40 dB/dec
wC2 K = 10,20 dB
+ 20
– 20 dB/dec
0 dB wC3
– 40 dB/dec
– 20
wgc
4.4
– 40 1/s
12 - 20
Solution :
TM
wgc = 4.4 rad/sec
Fig. 12.7
wpc ® ¥
– 90°
G.M. = + ¥ dB
P.M. = +42°
– 120° System is absolutely
stable

– 150°
P.M.
+42° wpc ® ¥
– 180°
– 210°
0.1 1 10 100 Log w

Example 12.15.25
Solution : Step 1 : G(s)H(s) in time constant form.
20 ´ 10 æç 1 + ö÷
s
è 10 ø 14.2857 (1 + 0.1 s)
G(s)H(s) = =
s 2 ´ 2 ´ æç 1 + ö÷ ´ 7 ´ æç 1 + ö÷ s (1 + 0.5 s) (1 + 0.1428 s)
s s 2
è 2ø è 7ø
Step 2 : Factors of G(s)H(s)
1) K = 14.2875, 20 Log K = 23.098 dB
1
2) , 2 poles at origin, straight line of slope – 40 dB/dec passing through the
s2
intersection of 0 dB and w = 1 lines.
1 1
3) wC1 = = = 2, simple pole, slope – 20 dB/dec
T1 0.5
1 1
4) wC2 = = = 7, simple pole, slope – 20 dB/dec
T2 0.1428
1 1
5) wC3 = = = 10, simple zero, slope + 20 dB/dec
T3 0.1
Frequency range Resultant slope

0<w<2 – 40 dB/dec
2< w<7 – 40 – 20 = – 60 dB/dec
7 < w < 10 – 60 – 20 = – 80 dB/dec
10 < w < ¥ – 80 + 20 = – 60 dB/dec

14.2857 (1 + 0.1 jw)
G(jw)H(jw) =
(jw) 2 (1 + 0.5 jw) (1 + 0.1428 jw)
w 1 - tan - 1 0.5 w - tan - 1 0.1428 w + tan - 1 0.1 w fR

(jw) 2
0.2 - 180° - 5.71° - 1.6359° + 1.1457° - 186.2°

2 - 180° - 45 ° - 15.93° + 11.3° - 229.63°
10 - 180° - 78.69° - 55 ° + 45 ° - 268.69°
¥ - 180° - 90° - 90° + 90° - 270°
Step 4 : The Bode plot and the specifications are shown in the Fig. 12.8.
TM

MR in dB 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
80
Scale
– 40 dB/dec On Y-axis
60 wgc 1 Unit = 20 dB, 30º
40
20 K, 23.09 dB
– 60 dB/dec
0dB
wC2
– 20
wC1 wC3
– 40
– 80 dB/dec – 60 dB/dec
fR in deg
Ans. :
– 120º
wgc = 2.9 rad/sec
– 150º wpc 0 rad/sec

P.M.= – 66º
– 180º
G.M.= – ¥ dB
Unstable system
– 210º P.M.
– 240º
– 270º
0.1 1 10 100 1000 Log w
Fig. 12.8
Example 12.15.26
Solution : Step 1 : G(s) in time constant form
K 40
G(s) =
s æç 1 + ö÷ æç 1 + ö÷
s s
è 2øè 20 ø
Step 2 : Sketch magnitude plot without K/40. For 1/s, a line of slope – 20 dB/dec,
passing through intersection of w = 1 and 0 dB.
æ ö
ç 1 ÷ 1 1
ç ÷ gives T1 = , wC 1 = = 2. So it is a straight line of slope – 20 dB/dec for
s
ç1+ ÷ 2 T1
è 2ø
w > wC 1 i.e. w > 2.
So resultant line of slope after wC2 = 2 is – 20 – 20 = – 40 dB/dec.
æ ö
ç 1 ÷ 1 1
ç ÷ gives T2 = , wC2 = = 20. So it is a straight line of slope – 20 dB/dec for
ç1+ s ÷ 20 T 2
è 20 ø
w > wC2 . So resultant slope after wC2 = 20 is – 40 – 20 = – 60 dB/dec.
Step 3 : Phase angle plot
TM

K 40
G(jw) =
æ jw ö æ jw ö
jw ç 1 + ÷ ç 1 + ÷
è 2 øè 20 ø
w 1 w w fR
- tan -1 - tan -1
jw 2 20
0.2 – 90º – 5.71º – 0.572º – 96.28º
2 – 90º – 45º – 5.71º – 140.71º
6 – 90º – 71.56º – 8.53º – 170.1º
20 – 90º – 84.28º – 45º – 219.28º
¥ – 90º – 90º – 90º – 360º
Step 4 : Sketching Bode plot as shown in the Fig. 12.9. (See Fig. 12.9 on next page)
wpc = 8 rad/sec for all values of K
K
i) For K = 40, 20 Log = 20 Log 1 = 0 dB
40
Hence the plot obtained is for K = 40 for which
G.M. = + 30 dB, P.M. = + 54º, System stable
K
ii) For K = 400, 20 Log = 20 Log 10 = 20 dB
40
Shift the magnitude plot up by 20 dB with wpc constant. For this case,
G.M. = + 10 dB, P.M. = + 20º, System stable
K
iii) For K = 4000, 20 Log = 20 Log 100 = 40 dB
40
Shift the magnitude plot up by 40 dB with wpc constant. For this case,
G.M. = – 10 dB, P.M. = – 24º, System unstable
Example 12.15.27
Solution :
K
Let G(s)H(s) =
s (1 + 0.1s) (1 + 0.05s)
Step 1 : G(s)H(s) is in time constant form.

Step 2 : Find Bode plot without effect of K.
i) One pole at origin, straight line of slope – 20 dB/dec passing through intersection of 0
dB and w = 1.
TM

1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
+ 60
Scale on
+ 40 wpc constant Y-axis
1 unit = 20 dB, 30º
+ 20
A"
wgc2 wgc3
0 dB wgc1 O
A'
– 20
A Shift of 40 dB
– 40
For 1/s
For K = 4000
fR
Shift of 20 dB
12 - 24
TM
For K = 400
Fig. 12.9
Without K/40
– 90º i.e. for K = 40
– 120º
P.M.
for

– 150º K = 40
P.M.
for K = 400
– 180º
P.M. (– ve)
for
– 210º K = 4000
– 240º
wpc = 8
Log w
0.1
1 10 100
1 1
ii) Simple pole, , T1 = 0.1, wC1 = = 10
1 + 0.1 s T1

1 1
iii) Simple pole, , T = 0.05, wC2 = = 20
1 + 0.05 s 2 T2

The resultant slope table is
0 < w < wC1 (10) – 20 dB/dec
10 £ w < wC2 (20) – 20 – 20 = – 40 dB/dec
20 £ w < ¥ – 40 – 20 = – 60 dB/dec
Step 3 :
–1 –1
w 1 – tan 0.1 w – tan 0.05 w fR
jw
0.2 – 90º – 1.14º – 0.57º – 91.71º
10 – 90º – 45º – 26.56º – 161.56º
20 – 90º – 63.43º – 45º – 198.43º
40 – 90º – 75.96º – 63.43º – 229.34º
¥ – 90º – 90º – 90º – 270º
Step 4 : The Bode plot without K is shown in the Fig. 12.10. (See Fig. 12.10 on next page).
i) For K = 10, the magnitude plot is to be shifted upwards by 20 log K = 20 dB
\ wgc = 10 rad/s, wpc = 14 rad/s, G.M. = + 6 dB, P.M. = + 18º
ii) For G.M. = 20 dB, draw horizontal line from – 20 dB till it intersects vertical line of wpc
at A. Then the shift A ¢A is contribution by K.
\ Shift A ¢A = 20 Log K = + 6 dB (upwards)
\ K = 1.99 » 2 … for G.M. = 20 dB
iii) For P.M. = 24º, draw horizontal line at 24º above – 180º line i.e. from – 156º till it meets
phase angle plot at B. Draw vertical line from B till it meets plot without K at C¢. Then
C¢C is the shift to be contributed by K.
\ Shift + CC¢ = 20 Log K = + 18 dB (upwards)
\ K = 7.94 … for P.M. = 24º
TM

MR 1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
in dB
80
Scale :
60 On Y axis : 1 unit = 20 dB, 30º
wgc = 10 for K = 10
40
– 20 dB/dec wpc = 14
20
G.M. = + 6 dB
0 dB C Shift A' – A = + 6 dB
G.M. for G.M. = + 20 dB
+ 20 C' A
–20
Line for G.M. = 20 dB
A' Plot with K = 10
12 - 26
fR in deg
TM
–90º – 40 dB/dec
Fig. 12.10
–120º Plot without K 1/s
(K = 1)
– 60 dB/dec
–150º B
P.M. P.M.
24º Shift C' – C = + 18 dB
+ 18º for P.M. = + 24º
–180º

–210º
–240º
–270º
Log w
0.1 1 10 100
Example 12.15.28
Solution : Step 1 :
10(1 + 0.05s)
G(s)H(s) = ... Time constant form
(1 + s) (1 + 0.5s) (1 + 0.1s)
Step 2 : Factors : K = 1, 20 Log K = 0 dB
1
Simple pole, , wC1 = 1, – 20 dB/dec
1+s
1
1 + 0.5s
K = 10 hence 20 log 10 = 20 dB, line parallel to log w axis.
1
Simple pole, , wC 3 = 10, – 20 dB/dec
1 + 0.1s
Simple zero, 1 + 0.05s, wC4 = 20, + 20 dB/dec
Thus resultant magnitude plot is 20 dB line parallel to Log w axis upto wC1 = 1, – 20 dB
line from w = 1 to wC2 = 2, – 40 dB line from w = 2 to wC 3 = 10, – 60 dB line from w = 10
to wC4 = 20 and finally – 40 dB line from w = 20 to w = ¥.
10(1 + 0.05 jw)
Step 3 : Phase table : G(jw)H(jw) =
(1 + jw) (1 + 0.5 jw) (1 + 0.1 jw)
–1 –1 –1 –1
w – tan w – tan (0.5w) – tan (0.1w) + tan (0.5w) fR
0.5 – 26.56º – 14.03º – 2.86º + 1.43º – 42.01º
5 – 78.69 º – 68.19º – 26.56º + 14.03º – 159.4º
8 – 82.87º – 75.96º – 38.66º + 21.8º – 175.68º
20 – 87.13º – 84.28º – 63.43º + 45º – 189.84º
¥ – 90º – 90º – 90º + 90º – 180º
The Bode plot is shown in the Fig. 12.11. (See Fig. 12.11 on next page.)
From the Bode plot, wgc = 4.6 rad/s, wpc = 9 rad/s, G.M. = + 12 dB, P.M. = + 24º, System
is stable.
TM

MR in dB
1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
Scale :
+ 60 On Y-axis : 1 Unit = 20 dB, 30º
–60 dB/dec
wgc = 4.6 wpc = 9
+ 40
–40 dB/dec
–20 dB/dec
+ 20 20 dB, (K = 10)
–20 dB/dec
–40 dB/dec
0 dB 0 dB
G.M. = +12 dB
0 dB
– 20 –60 dB/dec
– 40 –40 dB/dec
12 - 28
fR in deg
TM
– 30º
Fig. 12.11
– 60º
– 90º
– 120º

– 150º
P.M.
= + 24º
– 180º
– 210º
0.1 1 wpc = 4.6 wpc = 9 10 100 Log w

Example 12.17.4
Solution : Ke –0.2s (K 16)e –0.2s
G( s) = =
s( s + 2)( s + 8)
s æç 1 + ö÷ æç 1 + ö÷
s s
è 2 øè 8 ø
K
Let A = be the new constant term. Now obtain the Bode plot without the effect
16
of A.
The various factors are,
1
1) , one pole at the origin. So magnitude plot is straight line of slope – 20 dB dec
s
passing through intersection of w = 1 and 0 dB.
1 1 1
2) , simple pole, T1 = , wC1 = =2
s 2 T1
1+
2
Straight line of slope – 20 dB dec for w ³ 2.
1 1 1
3) , simple pole, T2 = , wC2 = =8
s 8 T2
1+
8
Straight line of slope – 20 dB dec for w ³ 8.
4) e –0.2s = e –0.2jw = cos 0.2w – j sin 0.2w
\ e –0.2s = cos 2 0.2w + sin 2 0.2w = 1 i.e. 20 Log 1 = 0 dB
Hence the term e –0.2s does not contribute to magnitude plot.

é – sin 0.2wù
Ð e –0.2s = + tan –1 ê ú = –tan [tan 0.2w]
–1
ë cos 0.2 w û
– 0.2´ w´ 180º
= – 0.2w rad = degrees.
p
0 < w < w C1( 2) – 20 dB dec
2 < w < w C2 ( 8) – 20 – 20 = – 40 dB dec
8<w<¥ – 40 – 20 = – 60 dB dec
The phase angle table can be obtained as,

Ae –0.2jw
G(jw) =
w w
jw æç 1 + j ö÷ æç 1 + j ö÷
è 2 øè 8ø
TM

1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
MR in
dB
Scale
On Y axis ; 1 unit = 30º, 20 dB
60
40
–20 dB/dec
20
wpc = 2.3
C,D
0 dB
B +6 dB G.M.
A Shift AB ® 4 dB for G.M. = +6 dB
–20
fR in degree –40 dB/dec
–90º (1/s)
12 - 30
TM
–120º Magnitude plot without
F effect of A
Fig. 12.12
–60 dB/dec
–150º P.M.
= +45º No shift required for P.M. = +45º
as point C is on 0 dB line
–180º
–210º

–240º
–270º
0.1 1 10 100 Log w

w 1 0.2w 180º –1 w –1 w fR
– – tan – tan
jw p 2 8
0.1 – 90º – 1.14º – 2.86º – 0.716º – 94.7º
1 – 90º – 11.45º – 26.56º – 7.125º – 135.13º
2 – 90º – 22.91º – 45º – 14.03º – 171.94º
4 – 90º – 45.83º – 63.43º – 26.56º – 225.82º
Obtain the Bode plot as shown in the Fig. 12.12 (See Fig. 12.12 on previous page)
From the Bode plot, wpc = 2.3 rad sec
Refer Ex. 12.15.3 for obtaining K for given G. M. and P. M.
i) For G. M . = + 6 dB, shift AB = 4 dB
\ 20 Log A = 4
\ A = 1.5848 i.e. K = 16 A = 25.36
ii) For P. M . = 45º, shift CD = 0 dB as point C is on 0 dB line hence points C and
D represent single point.
\ 20 Log A =0 dB
\ A = 1 i.e. K = 16 A = 16.
qqq
TM

13 Polar and Nyquist Plots
Example 13.2.9
Solution : G(jw) = 1
1 + j wT
(1 - j w T)
\ G(jw) = … Rationalizing
(1 + j w T) (1 - j w T)
1 - j wT 1 wT
= = -j = X + jY
1 + w2 T2 1+w T 2 2 1 + w2 T 2
1 - wT
Thus X = ,Y=
1 + w2 T2 1 + w2 T 2
2 2
é ö 1ù
æ X - 1 ö + Y 2 = êæç é ù
2
Now ç ÷
1 ÷ - ú + ê - wT ú
è 2ø êëçè 1 + w T ÷ø 2 úû
2 2 2 2
êë1 + w T úû
2 2
é 1 - w2 T 2 ù é - wT ù (1 - w2 T 2 ) 2 + 4 w2 T 2
= ê ú +ê ú =
êë 2 (1 + w2 T 2 ) úû 2 2
ëê1 + w T úû 2 (1 + w2 T 2 ) 2
1 - 2 w2 T 2 + w4 T 4 + 4 w2 T 2 1 + 2 w2 T 2 + w4 T 4
= =
2 (1 + w2 T 2 ) 2 2 (1 + w2 T 2 ) 2
(1 + w2 T 2 ) 2 1
= =
2 (1 + w2 T2 ) 2 2 ( 1 ,0)
2
w=0
2 2
æ X - 1 ö + Y2 = æ 1 ö w=¥
ç ç ÷ 0º
2 ÷ø
Hence 0 1
è è 2ø
The X and Y co-ordinates of G(jw) satisfies above
equation which is the equation of the circle with centre R= 1
æ 1 , 0 ö and radius R = 1 . Ö2
ç2 ÷ – 90º
è ø 2
Fig. 13.1
(13 - 1)
TM

Control System Engineering 13 - 2 Polar and Nyquist Plots
1
This proves that polar plot of G(s) = is a semicricle as shown in the Fig. 13.1.
1 + Ts
1
G(jw) =
1 + j wT
1
M = , f = - tan -1 wT
2 2
1+w T
w= 0 M=1 f = 0°
w= ¥ M=0 f = –90°
Example 13.2.10
Solution : The s-domain network is shown in the R
Fig. 13.2 (a). Using potential divider rule
1
E o ( s) 1
sC 1 Ei(s) Eo(s)
= = = T( s) sC
E i ( s) 1 1 + sRC
R+
sC
1
\ T(jw) = Fig. 13.2 (a)
1 + jwRC
1
\ MR = , f = -tan -1 wRC
2 2 2 0 Ð – 90º
1+w R C 1 Ð 0º
At w = 0, MR = 1, f = 0º and at w = ¥, MR = 0, f = -90º 0º
\ Rotation = -90º -0º = -90º clockwise in direction hence w=¥

w=0
the polar plot is as shown in the Fig. 13.2 (b).
– 90º
Example 13.5.3
1 Fig. 13.2 (b)
Solution : G (j w) =
j w(1 + j w)(1 + 2 j w)
To sketch the polar plot, vary w from 0 to ¥.
0°
At w = 0, G (j w) = ¥ , Ð G (j w) = = - 90°
+ 90° , 0° , 0°
0°
At w = ¥, G (j w) = 0, Ð G (j w) = = - 270°
90° , 90° , 90°
Rotation of plot = - 270°- (- 90°) = - 180° clockwise.
To find intersection with negative real axis, rationalize G (j w).
G (j w) =
- j w(1 - j w)(1 - 2 j w)
=
[ ]
- j w 1 - 3 j w - 2 w2
(j w)(- j w)(1 + j w)(1 - j w)(1 + 2 j w) (1 - 2 j w) (w2 )(1 + w2 )(1 + 4 w2 )
=
- 3 w2
-
(
j w 1 - 2 w2 )
(
w2 1 + w2 )(1 + 4 w2 ) (w2 )(1 + w2 )(1 + 4 w2 )
TM

For imaginary part to be zero,

1
1 - 2 w2 = 0 i.e. w2 = = 0.5 i.e. wpc = 0.7071 rad/sec
2
So intersection with negative real axis is,
-3
Point Q = = – 0.666
(1 + 0.5) (1 + 4 ´ 0.5)
Plot polar plot on graph paper to the scale to obtain G.M. and P.M. To plot it let us obtain
G (j w) = MR and Ð G (j w) = fR for two more frequencies, as given in the Fig. 13.3.
–270º
Scale
–261.41º 1 cm = 0.5 units
–198.43º Unit radius

circle
w=¥
–1+j0
P.M. 0
Q(–0.666)
wpc P
w MR fR
wgc –4
10 4.969 ´ 10 – 261.41º
1 0.3162 – 198.43º
Polar
plot
–125º
w=0
– 90º
Fig. 13.3 [For Ex. 13.5.3]
From the graph, unit radius circle intersects the polar plot at P.
\ P.M. = 180°+ Ð G (j w) = 180°-125° = + 55°
wgc
1 1
and G.M. = 20 log = 20 log = + 3.53 dB.
l ( OQ ) 0.666
Example 13.5.4
Solution : Convert transfer function to the frequency domain.
K
G(jw)H(jw) =
(jw)(1 + 2jw)(1 + 0.1jw)
TM

K
\ M = , f = - 90°- tan -1 2w - tan –1 - 0.1w
2 2
w 1 + 4w 1 + 0.01w
Start w = 0 M=¥ f = -90° -270°-( -90°) = - 180° clockwise
End w = ¥ M=0 f = -270°
Rationalize G(jw)H(jw) to find intersection with negative real axis.

K( -jw)(1 - 2jw)(1 - 0.1jw)
G(jw)H(jw) =
(jw)( -jw)(1 + 2jw)(1 - 2jw)(1 + 0.1jw)(1 - 0.1jw)
-jKw [1 - 2.1jw - 0.2w2 ] -2.1K jKw(1 - 0.2w2 )
= = -
w2 (1 + 4w2 ) (1 + 0.01w2 ) (1 + 4w2 )(1 + 0.01w2 ) w2 (1 + 4w2 )(1 + 0.01w2 )
Equating imaginary part to zero, 1 - 0.2w2 = 0 i.e. wPC = 5 rad/sec
Using in real part,

- 2.1K
Q = = – 0.09523 K
[1 + 4 ´ 5] [1 + 0.01 ´ 5]
The polar plot is shown in the Fig. 13.4. – 270º
For stability, the critical point must not be

enclosed i.e. it must lie to the left of Q.
\ l (OQ) < 1
w=¥
– 1 + j0
\ 0.09523 K < 1 Q
For O
\ K < 10.5 stability
\ 0 < K < 10.5

... Range for stability
Routh's method : The characteristic equation
is,
K
1 + G(s)H(s) = 0 i.e. 1+ =0 w=0
s(1 + 2s) (1 +0.1s) – 90º
3 2 Fig. 13.4
\ 0.2 s + 2.1s + s + K = 0
s3 0.2 1
s2 2.1 K From row of s 1 , 2.1 – 0.2 K > 0 i.e. 2.1 > 0.2 K
2.1 - 0.2K 2.1

s1 \K < < 10.5
2.1 0.2
\0 < K < 10.5 ... Range of stability
TM

Example 13.5.5
Solution : The frequency domain transfer function is,
K(1 + jw)
G(jw)H(jw) =
(jw)(jw) 2 + jw)(4 + jw)
K 1 + w2 w w
M = , f = + tan -1 w - 180°-tan -1 - tan -1
2 4
w2 4 + w2 16 + w2
Start w = 0 M=¥ f = - 180° -270° - ( -180°) = - 90° Clockwise
End w = ¥ M=0 f = - 270°
Rationalize G(jw)H(jw) to obtain intersection with negative real axis

K(1 + jw)(2 - jw)(4 - jw)
G(jw)H(jw) =
( -w2 ) (2 + jw) (2 - jw) (4 + jw) (4 - jw)
K(1 + jw)[8 - 6jw - w2 ] K [8 - 6jw - w2 + 8jw + 6w2 - jw 3 ]

= =
( -w2 )(4 + w2 )(16 + w2 ) ( -w2 )(4 + w2 )(16 + w2 )
K(8 + 5w2 ) Kjw(2 - w2 )

= +
( -w2 )(4 + w2 )(16 + w2 ) ( -w2 )(4 + w2 ) (16 + w2 )
Equating imaginary part to zero, 2 - w2 = 0
\ w2 = 2 i.e. wPC = 2
Using in real part,
K(8 + 5 ´ 2)
Q = = - 0.08333 K
( -2) (4 + 2)(16 + 2)
The polar plot is shown in the Fig. 13.5.
\ l(OQ) < 1 i.e. 1 – 0.08333 K|< 1 i.e. K < 12 … For stability
–270º
Critical point must

be to the left of Q
for stability
Note that while going
through 90º clockwise
w= ¥
–1+j0 rotation, the plot
–180º Q intersects negative
O real axis at point Q
w=0 –0.0833 K
Fig. 13.5
TM

Routh's criterion : The characteristic equation is 1 + G(s)H(s) = 0

K(s +1)
1+ = 0 i.e. s 4 + 6s 3 + 8s 2 + Ks + K = 0 s 2 (s + 2) (s + 4)
s 2 (s + 2)(s + 4)
s4 1 8 K For stability
s3 6 K 0 48 - K > 0 ... from s2
\K < 48 ... (1)
s2 48 - K K 0 From row of s1
6
s1 æ 48 - K ö 0 æ 48 - K ö
ç ÷(K) - 6K ç ÷ (K) - 6K > 0
è 6 ø è 6 ø
æ 48 - K ö
ç ÷
è 6 ø
s0 K i.e. 48 - K - 36 > 0
i.e. 12 - K > 0
i.e. K < 12 …(2)
Hence the ultimate condition for stability is 0 < K < 12 as obtained by the polar plot.
Example 13.5.6
–270º
Solution : Refer Example
13.5.5 for the procedure and
verify the polar plot as shown For stability,
in the Fig. 13.6. –1 + j0
must be
to the left –0.025 K
For K = 45, system is unstable. w= ¥
of Q Q
Range of K is –180º
O
0 < K < 40. w=0
For K = 45
Q = –1.125
Fig. 13.6
Example 13.5.7
e - jwT
Solution : G(jw) =
1 + jwT
Now e - jwT = cos( wT) - j sin( wT) i.e. |e - jwT|= 1
1
M = |G(jw)| =
1 + w2 T 2
TM

é - sin wT ù
and f = ÐG(jw) = tan -1 ê -1 -1
ú - tan wT = - wT - tan wT
ë cos wT û
For w = 0, M = 1 and f = 0º
w = ¥, M = 0 and f = not necessary as M = 0.
To find intersection with negative real axis rationalize G(jw).
(1 - jwT)[cos( wT) - j sin( wt)]
G(jw) =
(1 - jwt)(1+ jwT)
(cos wT - wT sin wT) j(sin wT + wT cos wT)
= -
1+ w2 T 2 1+ w2 T 2
\ sin wT+ wT cos wT = 0
\ sin wT = - wT cos wT i.e. tan wT = - wT i.e. wT = tan -1 ( - wT)
( -wT) 3
Now tan -1 ( -wT) = ( - wT) - …
3
Considering first two terms we get,
( -wT) 3 ( wT) 3
wT = – ( - wT) - i.e. 2wT = i.e. ( wT) 2 = 6
3 3
6 6
\ w2 = i.e. wpc =
T2 T
(cos 6 - 6 sin 6 )
\ Intersection point = = – 0.333.
(1 + 6)
For higher values of w , the plot spirals
round the origin.
Hence polar plot begins at 1 Ð 0° and then 1Ð0º
revolves repeatedly about the origin

6
between 0° and – 360° as w increases. So wpc = T
polar plot is as shown in the Fig. 13.7.
Fig. 13.7
TM

Example 13.13.10 Nyquist path

j¥
Solution : Step 1 : P = 0 Section
Section I IV
Step 2 : N = – P = 0 for stability
j0 R®¥
Step 3 : Two poles at origin.
\ Nyquist path is as shown
(1 + 0.5 jw) Section II
– j0
Step 4 : G( jw) H( jw) = R®0
( jw) ( jw) (1 + 0.1 jw) (1 + 0.02 jw)
Section III
Section I : s = + j¥ to s = +j0 s - plane
– j¥
Fig.13.8
Starting point w® + ¥ 90° – 180 ° – (–270)° = + 90°

0Ð = 0 Ð – 270 °
90°×90°×90°×90° Anticlockwise rotation
Terminating w =+ 0 0°
¥Ð = ¥ Ð – 180°
point 90°×90°×0°×0°
Section II : s = + j0 to s = – j 0
Starting point w= +0 ¥ Ð – 180 ° + 180 (–180) = + 360°

Anticlockwise rotation
Terminating w® -0 0°
¥Ð = ¥ Ð + 180 º
point - 90°× - 90°×0°×0°
Section III : Is mirror image of section I.

Section IV : Is not required.
(1 + 0.5 jw)
Step 5 : G( jw) H( jw) =
(– w2 ) (1 + 0.1 jw) (1 + 0.02 jw)
Rationalizing G( jw) H( jw) and separating real and imaginary part we get,
(1 + 0.5 jw) [1 – 0.12 jw – 0.002 w2 ]
G( jw) H( jw) =
(– w2 ) (1 + 0.01 w2 ) (1 + 0.0004 w2 )
(1 + 0.058 w2 ) jw [0.38 – 0.001 w2 ]

= +
D D
Equating imaginary part to zero,
0.38
w ( 0. 38 - 0.001 w2 ) = 0 \ w2 = = 380 \wpc = 19.4935 rad/sec
0.001
TM

Substituting in real part,

(1 + 0.058 ´ 380)
Point Q = = – 0.0109
(– 380) (1 + 0.01 ´ 380) (1 + 0.0004 ´ 380)
Step 6 : The section I starts from origin tangential to – 270° and crossing negative real
axis at point Q = – 0.0109 and is terminating at ¥ Ð - 180° i.e. mapping of s = + j0. So
rotation of plot in section I is 90° anticlockwise but it is crossing negative real axis while
doing so.
\ Nyquist plot is,
–270º
s = – j0 Section III
–180º s =+ j¥
–1+j0 O
¥
+180º s = –j¥ ¥
Section I Q
s = + j0 –0.0109
¥Ð –180º Section II
and ¥Ð +180º N=0 360º anticlockwise rotation
is same point
Fig. 13.9
Step 7 : Critical point –1 + j0 is getting encircled once in clockwise and once in

anticlockwise.
\ N = 0 This satisfies the stability criterion
\ System is stable.
1 1
G.M. = 20 Log = 20 Log dB = + 39.19 dB
|OQ | | 0.0109|
Example 13.13.11
Solution : Step 1 : No open loop pole in right half of s-plane so P = 0.
Step 2 : For closed loop stability N = – P = 0 + j¥

i.e. critical point –1 + j 0 should not get encircled Section Section
by Nyquist plot. I III
Step 3 : Nyquist path is as shown in the 0

Section
Fig. 13.10 (a).
II
3(2 - jw)
Step 4 : G(jw)H(jw) = – j¥
(1 + jw)(5 + jw) Fig. 13.10 (a)
Secton I : s = + jw to s = 0
TM

Starting w ® +¥ 0Ð
-90º
= 0 Ð - 270º 0º -( – 270º) = + 270º
90º 90º Anticlockwise rotation
Terminating w = 0 1.2 Ð
0º
= 1.2 Ð 0º
0º 0º
Section II : Mirror image of section I about real axis
Section III : Not required for closed loop stability.
Step 5 : Intersection with negative real axis

3(2 - jw)(1 - jw)(5 - jw)
G(jw)H(jw) = … Rationalize
(1 + jw)(1 - jw)(5 + jw)(5 - jw)
=
( ) ( )
3 10 - 8w2 + 3 jw w2 - 17
(1 + w )(25 + w )
2 2
For imaginary part zero, w = 0 or w2 = 17 – 270º

i.e. wpc = 17 . At w = 0, plot intersects positive
real axis at 1.2Ð0º. At w = 17 , plot intersects
negative real axis at point Q whose co-ordinates
are,
Q=
(
3 10 - 8 w2 ) =
3( -126)
= - 0.5
– 1 + j0 Q
– 0.5
O
1.2Ð 0º
(1 + w2 )(25 + w2 ) w2 = 17 18 ´ 42
N=0
Step 6 : Nyquist plot is shown in the

Fig. 13.10 (b).
Fig. 13.10 (b)
Step 7 : As N = 0 for – 1 + j0, it matches
criterion of step 2, hence system is stable.
1 1
G.M. = 20 Log = 20 Log = + 6.02 dB
l ( OQ ) 0.5
To find P.M., find wgc for which |G(jw)H(jw)|= 1

3´ 4 + w2
\
1+ w2 25 + w2
=1 i.e. ( ) (
9 4 + w2 = 1 + w2 )(25 + w2 )
\ w4 + 17 w2 - 11 = 0
\ w2 = 0.6241, w2 = – 17.6241 i.e. wgc = 0.79 rad/s.
\ P.M. = 180º + ÐG(jw)H(jw)

w = wgc
Ð2 - j0.79
= 180º + = 180º - 21.55º - 38.308º - 8.97º
Ð 1 + j0.79 Ð5 + j0.79
\ P.M. = +111.17º
TM

Example 13.13.12
(2 + jw)(8 + jw)
Solution : G(jw)H(jw) =
(jw) 3 +j¥
IV
Step 1 : No pole in r.h.s. of s-plane, P = 0
I
Step 2 : For stability, N = – P = 0
+j0 II
Step 3 : Nyquist path is, shown in the ¥
R
Fig. 13.11 (a). – j0 ®
¥
Step 4 : Section I : III
– j¥
Fig. 13.11 (a)
s = + j¥ 90º 90º – 270º – (–90º) = – 180º

0Ð = 0 Ð - 90º
270º
Clockwise
s = + j0 0º 0º
¥Ð = ¥ Ð - 270º
270º
Section II :
s = + j0 i.e. w = + 0 ¥ Ð - 270º 270º – (–270º) = 540º

Anticlockwise
s = – j0 i.e. w = – 0 ¥ Ð + 270º
Section III is mirror image of section - I, about real axis.
Section IV not required.

–270º
Step 5 : Calculate point Q which is
+j0
intersection of Nyquist plot with negative
real axis.
(2 + jw)(8 + jw)
G(jw)H(jw) =
(-jw) 3 –0.625
1 –1+j0 –j¥
... = –j –180º ¥ +j¥
j
Q
=
[
j 16 + 10jw - w2 ]
3
w
= =
-10
w2
+
w3
j
(16 - w2 ) –j0
N=2 – 90º
Fig. 13.11 (b)
TM

\ 16 - w2 = 0 ... for making imaginary part zero.

\ wpc = 4
-10
\ Point Q = = – 0.625
16
Step 6 : The Nyquist plot is shown in the Fig. 13.11 (b) (See Fig. 13.11 (b) on previous
page).
Step 7 : N = 2 from the Nyquist plot and must be zero for stability.
Thus there are 2 closed loop poles located in right half of s-plane, making system
unstable.
Example 13.12.13
Solution : Step 1 : No open loop pole in right half of
s-plane hence P = 0. Nyquist path
j¥
Section
Step 2 : According to Nyquist criteria, N = –P = 0
Section I IV
for stability.
j0 R®¥
Step 3 : Due to pole at origin, the Nyquist path is
as shown in the Fig. 13.12
Section II
R®0 – j0
Step 4 : Analysis of sections
1
G( jw) H( jw) = Section III
jw(1 + jw) s - plane
Section I : s = + j ¥ to + j0 – j¥
Fig. 13.12
Starting point w® ¥ 0Ð
0º
= 0 Ð – 180º –90º–(–180º)
90º 90º
= + 90º
Terminating point w ® +0 ¥Ð
0º
= ¥ Ð – 90º
90º 0º Anticlockwise
Section II : s = + j ¥ to s = – j0
Starting point w ® +0 ¥ = Ð– 90º 90º–(–90º) = + 180º

Terminating point w ® -0 ¥ = Ð+ 90º Anticlockwise
Section III : Mirror image of section I about real axis.

Section IV : Not required for closed loop stability.
( -jw) (1 - jw)
G( jw) H( jw) = ... Rationalize
( jw)( -jw)(1 + jw)(1 - jw)
TM

- w2 - jw 1 jw + 90º
= =- -
w2 (1 + w2 ) 1 + w2 w2 (1 + w2 ) – j0
Thus imaginary part is zero for

w = 0 but w = 0 is not on the Nyquist
path hence the Nyquist plot does not
intersect the negative real axis at finite
– 1 + j0 – j¥ ¥
point. – 180º 0º
+ j¥
Step 6 : The Nyquist plot is shown N=0
in the Fig. 13.13.
Step 7 : The Nyquist plot does not

encircle –1+j0 hence N = 0. This + j0
matches with the criteria in step 2.
Thus the given system is stable in – 90º
nature. Fig. 13.13
Example 13.12.14
Solution : Step 1 : No pole in right half of s-plane Nyquist path
hence P = 0. j¥
Section
Step 2 : According to Nyquist criterion, N = – P = 0 IV
Section I
So Nyquist plot should not encircle point – 1 + j0, for
j0 R®¥
stability.
Step 3 : Four poles at origin, so Nyquist path is as Section II

R®0 – j0
shown in Fig. 13.14.
Section III
Step 4 : Analysis of sections : s - plane
– j¥
Fig. 13.14
Section I : s = + j¥ to s = + j0
Starting point w® + ¥ 90° – 360° – (–360° ) = 0°

0Ð = 0 Ð – 360°
360°× 90° No rotation of plot
Terminating point w® + 0 0°
¥ Ð = ¥ Ð – 360°
360°× 0°
5w
Now angle contribution by 1 + 5s is, tan -1 æç ö÷ while the angle contribution by
è 1 ø
w
is - tan -1 æç ö÷
1
1+s è1ø
TM

5w w
Now tan -1 æç ö÷ > tan -1 æç ö÷ ... (1)
è ø
1 è1ø
5w w
So net angle, fR = - 360°+ tan -1 æç ö÷ - tan -1 æç ö÷ ... (2)
è 1 ø è1ø
And from equation (1), fR > - 360° i.e. - 359° , - 356° , - 350°
Hence section I will be always in first quadrant.
Section II : s = + j0 to s = – j0
Starting point w® + 0 ¥ Ð – 360° 360° – (–360° ) = + 720°

Terminating point w® – 0 ¥ Ð + 360°
Section III : Mirror image of section I, i.e. in third quadrant.

Section IV : Not required.
Step 5 : No finite intersection with negative real axis.
Step 6 : Nyquist plot is shown below.
Section II Section I
720º rotation in first quadrant
+j¥ A +j0 360º

–1+j0
¥ ¥ 0 B
–j¥ –j0 –360º
Both points A and B

are actually same
N = +2 located at ¥
Fig. 13.15
Step 7 : N = +2 from the plot

But N should be zero for stability, as per step 1.
There are two roots of characteristic equation in right half of s-plane.
Z = 2
Hence the given system is Unstable.
TM

Example 13.12.15 s = + j¥
Solution : I
+
s = + j2
Step 1 : No pole in right half of s-plane i.e. + j2 II
–
P = 0. s = + j2
III R®¥
Step 2 : For stability, N = –P = 0.
0 ¥
Step 3 : Nyquist path is so as to exclude the IV
+
poles of G(s)H(s). Now G(s)H(s) has s = – j2 VII
– j2 V
poles at s 2 = -4 i.e. s = ± j2. So Nyquist –
s = – j2
path should not encircle s = ± j2. VI
Hence Nyquist path is as shown in the s = – j¥
Fig. 13.16. It is divided in 7 section.
Fig. 13.16
For ease of analysis, Section III is
considered upto s = 0.
Step 4 : Analysis of sections.

Section I : s = + j ¥ to s = + j2 +
Starting point 0°
0Ð = 0 Ð - 270°
w= + ¥ 90° 180°
Terminating point M1 Ð f 1
w= + 0
K
Now G(jw) H(jw) = … s = jw
(2+ jw) ( - w2 + 4)
w
Ð - tan -1 æç ö÷ - ( Ð - w2 + 4)
K
\ G(jw) H(jw) = MÐf =
( w2 + 4 )( 4 - w2 ) è2ø
Now s = + j2 + means w = 2.0001.

K
\ M1 = Þ¥
( 4 + ( 2.0001) ) (4 - 2.0001 2 )
2
0° 0°
f1 = Þ Þ – 225°
-1 æ 2.0001 ö 45 ° .180°
tan ç ÷ 180°
è 2 ø
Note Ð 4 - w2 = Ð - 4.0001 ´ 10 -4 + j0 = 180°

w = 2.0001
Thus section I becomes,
w= + ¥ 0 Ð - 270° - 225 ° - ( - 270°)

= + 45 ° Anticlockwise
w ® + 2+ ¥ Ð - 225 °
TM

Section II : s = + j2 + to s = + j2 -
Thus w = + 1.9999 i.e. close to j2 but slightly less than +j2.

+
w = +2 ¥ Ð - 225 ° - 45 ° - ( - 225 °)
0° = + 180° Anticlockwise
w ® + 2- ¥Ð = ¥ Ð- 45 °
45 ° 0°
Ð 4 - w2 = Ð + 3.9999 ´ 10 -5 + j0 = 0°
w = 1.9999
Section III : s = + j2 - to s = 0
w = + j2- ¥ Ð - 45 ° 0° - ( - 45 °)
= + 45 ° Anticlockwise
w=0 K
Ð 0°
8
Section IV is mirror image of section III about real axis.

Section V is mirror image of section II about real axis.
Section VI is mirror image of section I about real axis.
Section VII is not required as encirclements are about origin.
K(2 - jw)
Rationalize, G(jw) H(jw) =
( 2 + jw) ( 2 - jw)( - w2 + 4)
K
Thus at w = 0, it intesects real axis but at w = 0 it intersects positive real axis at .
8
Thus there is no finite intersection with negative real axis.
Step 6 : The Nyquist plot is shown in the Fig. 13.17.
– 270°
¥ Ð–225º ¥
– 180° – 1 + j0 s = + j¥
0°
K s=0
s = – j¥
8
– 90° R®¥ ¥ Ð – 45°

N = +2
Fig. 13.17
TM

Step 7 : For any value of K, N = +2 and system is absolutely unstable.

For no value of K, system is stable.
Example 13.12.16
Solution : Step 1 : No open loop pole in right half of
+ j¥
s-plane hence P = 0
I R®¥
Step 2 : N = – P = 0 i.e. Nyquist plot should not III
0
encircle – 1 + j0 for stability. ¥
Step 3 : Nyquist path is shown in the Fig. 13.18. II
For ease of analysis, it is divided into three sections. – j¥
2 (1 - jw)
Step 4 : G(jw)H(jw) =
(2 + jw) (3 + jw) Fig. 13.18
Section I : s = + j ¥ to s = 0
Start w ® + ¥ - 90º 0º – (– 270º) = + 270º

0Ð = 0 Ð – 270º
90º 90º Anticlockwise
End w = 0 1
Ð
0º 1
= Ð 0º
3 0º 0º 3
Note that Ð 1 – jw |w ® + ¥ = – 90º

Section II : Mirror image of section I about real axis.
Section III : Not required for closed loop stability.
Step 5 : Intersection with negative real axis.

2 (1 - jw) (2 - jw) (3 - jw)
G(jw)H(jw) = …Rationalize
(2 + jw) (2 - jw) (3 + jw) (3 - jw)
2 [(6 - 6w2 ) - jw (11 - w2 )

=
(4 + w2 ) (9 + w2 )
For imaginary part to be zero, w = 0, w = 11

1
At w = 0, plot intersects with positive real axis at .
3
TM

At w = 11, Plot intersects with negative real axis at Q.

2 ´ [6 - 6 ´ 11]
Q = = – 0.4
(4 + 11) (9 + 11)
– 270º
Step 6 : Nyquist plot is
shown in the Fig. 13.19.
Step 7 : Nyquist plot

does not encircle – 1 + j0.
–1+j0 +j¥
Hence P = 0. This matches w=0
–180º 0º
with the criterion in step 2. –j¥ 1
Ð 0º
3
Hence the system is stable Q
N=0 – 0.4
in nature.
– 90º
Fig. 13.19
Example 13.12.17
Solution : Step 1 : No pole in right half of s-plane hence P = 0.
Step 2 : N = – P = 0, for stability Nyquist plot should not encircle – 1 + j 0 point.
+j¥
Section- I
+j0 ¥
®
R Section-IV
–j0
Section II
Section-III
s - plane
–j¥
Fig. 13.20
Step 3 : As 2 poles at origin the Nyquist path is as shown in the Fig. 13.20.
Step 4 : Analysis of sections, replace s by j w
10
\ G(jw ) H (jw) =
2
(j w) (1 + 0.25 j w) (1 + 0.5 j w)
Section I : s = + j ¥ to s = + j 0
Starting point w® + ¥ 0 Ð 0° – 180° – (–360° )

= 0 Ð – 360°
180°× 90°× 90° = + 180°
Terminating point w® + 0 ¥ Ð 0° Anticlockwise rotation
= ¥ Ð – 180°
180°× 0°× 0°
TM

Section II : s = + j 0 to s = – j 0
Starting point w® + 0 ¥ Ð – 180° 180° – (–180° )

0° = + 360°
Terminating w® – 0 ¥Ð = ¥ Ð + 180°
- 180°× 0°× 0° Anticlockwise rotation
point
Section III : Mirror image of section I

Rationalizing G (j w) H (j w),
10(1 – 0.25jw) (1 – 0.5jw)
G(j w) H(j w) =
- w2 × (1 + 0.0625w2 ) (1 + 0.25w2
10 [1 - 0.75 jw - 0.125 w2 ]
=
- w2 (1 + 0.0625 w2 ) (1 + 0.25 w2 )
10 (1 - 0.125 w2 ) j 0.75 w
= –
- w2 (1 + 0.0625 w2 ) (1 + 0.25 w2 ) - w2 (1 + 0.0625 w2 ) (1 + 0.25 w2 )
\ w = 0 rad/sec
But w = 0 is not the part of Nyquist path hence there is no finite intersection of Nyquist
plot with negative real axis.
Step 6 : Nyquist plot is shown below.
I
tion
Sec
+j0 +j¥
–1+j0
–180° ¥ 0° or –360°
–j0 –j¥
R
All points are at ¥ Sect
®
ion I o
II
¥
360 anticlockwise
Section II
N = +2
Fig. 13.21
Step 7 : N=+2 from plot
TM

But for stability N = 0 as obtained in step 2.

So there are two roots of characteristic equation in right half of s-plane.
Hence the given system is Unstable.
Example 13.12.18
K
Solution : G(s)H(s) =
(s + 1) (s 2 + 6s + 10)
Step 1 : No pole in right half so P = 0
Step 2 : As N = – P = 0, for stability Nyquist plot should net encircle – 1 + j 0 point.
Step 3 : Nyquist path is
j¥
Section I
Section III
0
R
¥
Section II
–j¥
Fig. 13.22
Though it is actually divided into two sections + j ¥ to - j¥ and -j ¥ to + j¥, for case of
analysis, let us divide it into three sections as shown in Fig. 13.22.
Step 4 : Analysis of sections.
K
G(jw)H(jw) =
(1 + jw) ( - w2 + 6jw+ 10)
Section I : s = + j ¥ to s = 0 i.e. w ® ¥ to w = 0
Starting point w ® +¥ 0° 0° – (–270° )

0Ð = 0 Ð –270°
90°× 18 0° = + 270°
0° Anticlockwise rotation
Terminating w= 0 K
Ð =
K
Ð 0°
point 10 - 0°× 0° 10
Section II : Mirror image of section I.

Section III : Not required.
K(1 - jw) [(10 - w2 ) - 6jw]
Rationalise, G(jw) H(jw) =
(1 + jw) (1 - jw) [(10 - w2 ) + 6jw] [(10 - w2 ) - 6jw]
TM

K[10 - w2 - 6jw-10jw+ jw 3 - 6w2 ]

=
(1 + w2 ) [(10 - w2 ) 2 + 36w2 ]
K(10 - 7 w2 ) jw( w2 - 16)

= +
D D
w2 – 16 = 0
\ w2 = 16
\ w = 4 i.e. wpc = 4
Substituting in the real part,
K(10 - 7 ´ 16) -102K
Point Q = =
2
(1 + 16)[(10 - 16) + 36 ´ 16] 10404
Step 6 : Nyquist plot
–270°
Section II
K
––
– 1 + j0 Q 10
0°
N=0
–102K
–––––
10404
Section I
Fig. 13.23
Step 7 : For stability, encirclements of – 1 + j0 must be zero.

\ l(OQ) < 1
-102K
\ < 1
10404
10404
\ K < < 102
102
Thus the range of values of K is 0 < K < 102. At K = 102, Q point will be –1 +j0 and
system will become critically stable.
\ Kcritical = 102
TM


j¥
Solution : Step 1 : P = 1 as 1 pole in right half of Section
s-plane. IV
Section I
Step 2 : N = – P = – 1 i.e. Nyquist plot must j0 R®¥
encircle –1 + j0 point once in clockwise for
stability. Section II
R®0 – j0
Step 3 : One pole at origin so Nyquist path is as
shown in Fig. 13.24. Section III
s - plane
Step 4 : 10 ( 3 + jw)
G( jw) H( jw) = – j¥
jw ( + jw - 1)
Fig. 13.24
Section I : s = + j ¥ ® s = + j0
Starting point w® + ¥ 0°×90° – 270° – (– 90°) = – 180°
0Ð = 0 Ð- 90°
90°×90° Clockwise rotation
Terminating point w®+ 0 0°×0°
¥Ð = ¥ Ð- 270°
90°+180°
Section II : s = + j0 ® s = – j0
Starting point w® + 0 ¥ Ð- 270° - 90° – (– 270°) = + 180°

Terminating point w® -0 0°×0°
¥Ð = ¥ Ð- 90°
- 90° + 180° *
* Contribution by (s –1) at w ® + 0 is + 180° as discussed earlier and remains same for

w = – 0.
Section III : Mirror image of section I.
Step 5 : Rationalizing G(jw)H(jw)

10 (3 + j w) (– jw) ( jw + 1) 10 jw [ 3 + 4jw – w2 ]
G( jw)H( jw) = =
(jw) (– jw ) (jw – 1) (jw + 1) w2 ´ ( - w2 – 1 )
10 jw [ 3 – w2 + 4 jw]
= ... Taking negative sign outside from denominator
w2 ´ (1 + w2 )
– 40 w2 10 w (3 – w 2 )
= +j where D = w2 (1 + w2 )
D D
Equating imaginary part to zero
w ( 3 - w2 ) = 0 i.e. w2 = 3 i.e. wpc = 3
Substituting in real part we get,
– 40 ´ 3
Q = = –10
3 ´ (1 + 3)
TM

Step 6 : Nyquist plot is shown in the Fig. 13.25.

–270º , + 90º
s = + j0
Section I
Section II
–1+j0
s = – j¥
Q
¥
–10 s = + j¥
Section III mirror image

of section I
s = –j0
–90º
N = –1
Fig 13.25
Step 7 : Net encirclements of –1 + j0 is N = –1.

This matches with the stability criterion.
Hence given system is stable in nature. This is the rare example of open loop unstable
system behaves as stable closed loop system.
Example 13.14.4
Solution : The O.L.T.F. is
(s + 1)
G(s) H(s) =
2
s (s - 4)
Step 1 : There is one pole in the right half at s = 4

\ P = 1
\ N = –P=–1
j¥
For stability there must be one
clockwise encirclement. Section I
Section II
+ j0 Section IV
Step 2 : Nyquist path as shown in
Fig. 13.26. O
– j0 R
Step 3 : Analysis of sections : ® s - plane
Section III ¥
(1 + jw)
G(jw) H(jw) =
– j¥
( jw) 2 ( jw- 4)
Fig. 13.26
TM

Section I : + j¥ to + j0
w=+ ¥ 90° – 360 – (–180)

0Ð = 0 Ð- 180°
180°. 90° = –180°
w=+ 0 0° Clockwise
¥Ð = ¥ Ð- 360°
180°.180°
Note that the pole in right half contributes + 90° for w ® ¥ and contributes +180° for both
w ® +0 and – 0.
Section II : + j0 to – j0
w=+ 0 0° 0 – (–360)
¥Ð = ¥ Ð- 360°
180°.180° = + 360° Anticlockwise
w=–0 0°
¥Ð = ¥ Ð 0°
-180°.180°
Section III : Mirror image of section I.


Rationalize the given G( j w)H( j w),
(1+ jw)( jw+ 4) jw- w2 + 4 + 4 jw ( 4 - w2 ) + 5 jw
G( j w)H( j w) = = =
- w2 ( jw+ 4)( jw - 4) ( - w2 ) ( - w2 - 16) w2 ( w2 + 16)
The imaginary part is zero only when w = 0. But w = 0 is not existing on path.
Hence there is no finite intersection of Nyquist plot with imaginary axis.
Step 5 : The plot is,
90º
– 180º
-1 + j0 + j¥ + j0 0º
¥
¥ - j¥ - j0
+ 180º – 360º
N=+1
Fig. 13.27
TM

Step 6 : As – 1 + j0 is encircled once in anticlockwise direction,

N = +1 but N must be –1 for stability.
As N = 1, P = 1 and N=Z–P
\ 1 = Z–1 i.e. Z=2
There are two roots in right half of s-plane. The closed loop system is unstable in nature.
s+2 (s + 2) j¥
Solution : G(s)H(s) = = Section
s 2 (s 2 - 9) s 2 (s + 3) (s - 3)
Section I IV
Step 1 : P = 1, right half open loop pole
j0 R®¥
Step 2 : For stability, N = – P = – 1 clockwise
Section II
Step 3 : The Nyquist path is as shown in Fig. 13.28. R®0 – j0
Step 4 : Analysis of sections Section III

(2 + jw) s - plane
G(jw) H(jw) =
2
(jw) (3 + jw) ( - 3 + jw)
– j¥
Section I : + j¥ to + j0 Fig. 13.28
w= + ¥ 90°
0Ð = 0 Ð - 270°
180° 90° 90° - 360° - ( - 270°) = - 90°
0° Clockwise
w= + 0 ¥Ð = ¥ Ð - 360°
180° 0°180°
Section II : + j0 to – j0 [Note right half pole contribution to angle]
w= + 0 ¥ Ð - 360°
0° - ( - 360°) = + 360°
w= – 0 0°
¥Ð = ¥ Ð 0° Anticlockwise
- 180° 0°180°
Section III : Mirror image of section - I.

Step 5 : Intersection with real axis

(2 + jw) 2 + jw
G(jw) H(jw) = =
2 2
( - w ) [(jw) - 9] ( - w ) ( - w2 - 9)
2
2 w
= +j
w2 ( w2 + 9) w2 ( w2 + 9)
No finite intersection with real axis, as w = 0 for getting zero imaginary part.
TM

Step 6 : The Nyquist plot is as shown in the Fig. 13.29.

– 270º
+ j0
– 1 + j0 + j¥ ¥ 0º
¥ – j¥ – 360º
– j0
N=1
– 90º
Fig. 13.29
Step 7 : As – 1 + j0 is encircled once in anticlockwise direction, N = + 1
\ N = Z – P i.e. 1 = Z – 1 i.e. Z = 2
So two roots in R.H.S of s-plane given hence, system is unstable.
Example 13.14.6
Solution : Step 1 : P = 1 as one open loop pole in right hand side of s-plane.
Step 2 : According to Nyquist criterion, N = – P = – 1 i.e. once in clockwise direction.
Step 3 : The Nyquist path is shown in the
Fig. 13.30 j¥ Section
Section I III
Step 4 : Analysis of sections
¥
K(1 + 0.4 jw)(1 + jw) R®
G(jw)H(jw) = o
(1 + 8 jw)(-1 + jw)
K 1 + (0.4w) 2 1 + w2 Section II
|M| = –j ¥
1 + 64w2 1 + w2
K´ 0.4 Fig. 13.30
\ Lim |M| = = 0.05 K
w®¥ 8
Section I : s = j ¥ to s = 0
Starting w® ¥ 0.05 K Ð
90º 90º
= 0.05 K Ð 0º -180º -0º
90º 90º = -180º
Clockwise rotation
Terminating w = 0 KÐ
0º 0º
= K Ð - 180º
0º 180º
Section II : Mirror image of section I about real axis.

TM

Section III : Not required for closed loop analysis.

K(1 + 0.4 jw)(1 + jw)(1 - 8jw)(-1 - jw)
G(jw)H(jw) =
(1 + 8jw)(1 - 8jw)(-1 + jw)(-1 - jw)
=
K {(3.2w 4
) (
- 17.4w2 - 1 + jw 5.6 - 14w2 )}
(
1 + 64w2 )(1 + w2 )
For imaginary part zero, 5.6 - 14 w2 = 0 i.e. wPC = 0.4 = 0.63245
Substituting in real part, Q=

[ 3.2 ´ 0.4 2 - 17.4 ´ 0.4 - 1]K
= – 0.2 K
(1 + 64 ´ 0.4)(1 + 0.4)
Step 6 : The Nyquist plot is shown in the Fig. 13.31
+ 180º
– Q –1 0.05 K
0º
K
–0.2 K
N = –1
Fig. 13.31
Step 7 : For stability, N = – 1. This is possible when point – 1 + j0 is to the right of point
Q.
Hence for stability, O Q > 1 i.e. - 0.2 K > 1
\ 0.2 K > 1 i.e. K > 5
Hence range of K for stability is 5 < K < ¥.

Ks j¥
Solution : G(s) = Section
(s - 1) 2 (s + 5)
Section I IV
Step 1 : Two poles in R.H.S. of s-plane.
j0 R®¥
\ P = 2
Step 2 : According to Nyquist stability
criterion, Section II
R®0 – j0
N = –P=–2
So Nyquist plot must encircle – 1 + j0 Section III
s - plane
twice in clockwise direction.
Step 3 : The Nyquist path is, as shown in – j¥
Fig. 13.32. Fig. 13.32
TM

Kjw
Step 4 : G(jw) =
( -1+ jw) ( -1+ jw) (5+ jw)
Section I : s = + j¥ to s = + j0
Starting point w ® ¥ 1 90°

Ð = 0 Ð – 180°
¥ 90° 90° 90° – 270º – (– 180º) = – 90º
Terminating point w ® + 0 90°
0Ð = 0 Ð - 270° Clockwise
180°180° 0°
Section - II : s = + j0 to s = – j0
w® + 0 0 Ð – 270°
– 450º – (– 270º) = – 180º
w® – 0 - 90°
0Ð = 0 Ð - 450° Clockwise
180°180° 0°
Key Point Pole in right half contributes + 180º for both w ® + 0 and w ® – 0.
Section III : Mirror image of Section I about real axis.

Kjw(5 - jw) ( -1 - jw) 2
G(jw) = … Rationalizing
( -1 + jw) 2 ( -1 - jw) 2 (5 - jw) (5 + jw)
Kjw[5 + 9jw - 3w2 + jw 3 ] -Kw2 ( 9 + w2 ) + Kjw(5 - 3w2 )
= =
(1 + w2 ) 2 ( 25 + w2 ) (1 + w2 ) 2 ( 25 + w2 )
2
Equating imaginary part to zero, 5 – 3 w = 0
2 5
\ w = = 1.667 i.e. wpc = 1.2909 rad/sec
3
Substituting in real part intersection with negative real axis is obtained,
- K´ 1.667 ´ (9 + 1.667)
\ Q = = – 0.09374 K
(1 + 1.667) 2 (25 + 1.667)
Step 6 : The Nyquist plot is as – 270º
shown in the Fig. 13.33. Section I
Note that while approaching +j0

from 0 Ð – 180º to Ð – 270º in
–1 + j0 Q – j¥
clockwise direction, Section I of – 180º
–1 + j0 +j¥
plot has to intersect negative real
axis at point Q. So from origin, it – j0
N=0
will travel through third unstable Section I
N=–2
quadrant and through point Q stable
will reach to 0 Ð – 270º.
Fig. 13.33
TM

Step 7 : For stability, N = – 2 which is possible only if,

|OQ| > 1 i.e. | – 0.09374 K | > 1
1
\ K > > 10.667
0 . 09374
So range of K for stability is 10.667 < K < ¥.
Thus for K = 10, the system is unstable.
qqq
TM

13
14 TheCompensation
Design of Feedback Control
of Control Systems
Systems
Example 14.4.3
Solution : The transfer function of a lead compensator is,
a (1 + Ts)
T(s) = and comparing with given transfer function, T = 0.12, aT = 0.04
(1 + aTs)
1 1 1
The maximum phase shift occurs at wm = = ×
T a T aT
1 1
\ wm = ´ = 14.4337
0.12 0.04
–1 –1
fm = tan wm T – tan wm a T … Maximum phase shift
–1 –1
= tan (14.4337 ´ 0.12) – tan (14.4337 ´ 0.04) = 30º.
Example 14.4.4
Solution : Step 1 : Assume a lead compensator as,
(1 + Ts) K (1 + Ts)
G c ( s) = Kc a =
(1 + a Ts) (1 + a Ts)
K (1 + Ts) 1 (1 + Ts)
\ G c ( s)G( s) = ´ = G ( s)
(1 + a Ts) s ( s + 2)( s + 30) (1 + a Ts) 1
K
\ G 1 ( s) =
s ( s + 2)( s + 30)
1
For unit ramp input, e ss = = 0.04
Kv
\ Kv = 25
s´ K
But Kv = lim s G 1 ( s) i.e. 25 = lim
s® 0 s® 0 s ( s + 2)( s + 30)
\ K = 25 ´ 2 ´ 30 = 1500
(14 - 1)
TM

Control System Engineering 14 - 2 The Design of Feedback Control Systems
1500
Step 2 : Sketch Bode plot of G 1 ( s) =
s ( s + 2)( s + 30)
25
\ G 1 ( s) =
s (1 + 0.5 s) (1 + 0.033 s)
Factors : K = 25, 20 log K = 27.95 » 28 dB
One pole at origin, straight line of slope – 20 dB/dec.
1 1
Simple pole, , T1 = 0.5, w C1 = = 2, slope – 20 dB/dec
1 + 0.5 s T1
1 1
Simple pole, , T2 = 0.0333, wC2 = = 30, slope – 20 dB/dec
1 + 0.033 s T2
Frequency range Resulting slope

0 < w < w C1 ( 2) – 20 dB/dec
2 < w < 30 – 20 – 20 = – 40 dB/dec
30 < w < ¥ – 40 – 20 = – 60 dB/dec
25
Now G 1 (j w) =
j w(1 + 0.5 j w)(1 + 0.033 j w)
The phase angle table is,
w 1 - tan - 1
0.5 w - tan - 1
0.033 w fR
5w
0.2 - 90° - 5.71° - 0.38° - 96.09°

2 - 90° - 45° - 3.81° - 138.81°
10 - 90° - 78.69° - 18.41° - 187.1°
30 - 90° - 86.18° - 45° - 221.18°
¥ - 90° - 90° - 90° - 270°
The Bode plot is shown in the Fig. 14.1 (a) from which wgc = 7.2, wpc = 8.6,
G.M. = + 4 dB and P.M. = f1 = + 8°.
Step 3 : fs = required P.M. = 35°
\ fm = fs - f1 + e ... e = + 5°
= 35°- 8°+ 5° = 32°
1 -a 1 -a
Step 4 : sin fm = i.e. sin 32° = = 0.5299
1+a 1+a
\ 1 - a = 0.5299 (1 + a ) i.e. a = 0.3071
Choose a = 0.31
Step 5 : - 10 log æç ö÷ = – 5.08 dB for a = 0.31.

1
èaø
TM

Find frequency at which gain of uncompensated G 1 ( s) is – 5.08 dB from the

Fig. 14.1.
5 6 7 8 91
28 dB, (K = 25)
Uncompensated system
Log w
4
3
2
– 60 dB/dec
5 6 7 8 91
4
3
2
(1/s)
wC2
100
5 6 7 8 91
– 40 dB/dec
4
= +4 dB
wpc = 8.6
3
G.M.
wm = 9.4
2
wgc 10
5 6 7 8 91
7.2
4
P.M.= +80
wC1
3
2
1
5 6 7 8 91
– 20 dB/dec
4
3
– 5dB
2
0.1
MR in 1
–90º
–120º
–150º
–180º
–210º
–240º
–270º
0dB
60
40
20
–20
–40
dB 80
Fig. 14.1 (a) Bode plot of uncompensated system in example 14.4.4

TM

\ wm = 9.4 rad/sec
1 1 1
Now wm = i.e. 9.4 = i.e. = 5.2336
T a T ´ 0.31 T
Step 6 : Two corner frequencies of lead compensater are,

1 1
wC1 = = 5.2336 and wC2 = = 16.88
T aT
Step 7 : K = Kc a
K 25
\ Kc = = = 80.64
a 0.31
Use K = 25 of G 1 ( s) adjusted in time constant form.
80.64 ´ 0.31 ´ æç 1 +
s ö
è 5.2336 ÷ø
\ G c ( s) =
æ1 + s ö
ç 16.88 ÷
è ø
25 (1 + 0.191 s)
\ G c ( s)G ( s) =
s (1 + 0.5 s)(1 + 0.059 s)(1 + 0.033 s)
Draw the Bode plot of this compensated system and check G.M. and P.M.
Factors : K = 25, 20 log K = 28 dB
One pole at origin, – 20 dB/dec
1
1 + 0.5 s
1
Simple zero, (1 + 0.191 s), wC2 = = 5.23, + 20 dB/dec
0.191
1 1
Simple pole, , wC3 = = 16.8, – 20 dB/dec
(1 + 0.059 s) 0.059
1
(1 + 0.033 s)
Thus resultant slope table is,
0 < w < wC1 (2) – 20 dB/dec
2 < w < wC 2 (5.23) – 20 – 20 = – 40 dB/dec
5.23 < w < wC 3 (16.8) – 40 + 20 = – 20 dB/dec
16.8 < w < wC 4 (30) – 20 – 20 = – 40 dB/dec
30 < w < ¥ – 40 – 20 = – 60 dB/dec
TM

MR
2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
in dB 1
80
Compensated system
60 – 20 wC1 – 40
wC2
40
28 dB, (K = 25)
20 – 20
wC3
B
0dB
– 40
A
–20 wC4
–40 (1/s)
fR
– 60
TM
–90º
14 - 5
–120º
C
–150º
P.M. wgc = 8.6 rad/sec
–180º wpc = 24 rad/sec
D

G.M. = + 16 dB = AB
–210º P.M. = + 36º = CD
–240º
Fig. 14.1 (b) Bode plot of compensated system in example 14.4.4

–270º
0.1 1 10 100 Log w

wgc = 8.6 wpc = 24
The Design of Feedback Control Systems
w 1 - tan - 1
0.5 w + tan - 1
0.191 w - tan - 1
0.059 w - tan - 1 0.033 w fR
jw
0.2 - 90° - 5.71° +2.18° - 0.67° - 0.38° - 94.58°
2 - 90° - 45° +20.9° - 6.73° - 3.81° - 124.64°
10 - 90° - 78.69° +62.36° - 30.54° - 18.41° - 155.28°
30 - 90° - 86.18° +80.1° - 60.53° - 45° - 201.61°
¥ - 90° - 90° + 90° - 90° - 90° - 270°
The Bode plot is shown in the Fig. 14.1 (b) from which, (See Fig. 14.1 (b) on previous page)
wgc = 8.6 rad/sec, wpc = 24 rad/sec, G.M. = + 16 dB, P.M. = + 36°
Thus compensated system satisfies required specifications.
Example 14.4.5
1
Solution : Step 1 : Let G(s) =
s 2 (0.2s +
1)
Kc a (1 + Ts) K(1 + Ts)
where G c (s) = = ... Lead compensator
(1 + aTs) (1 + aTs)
s 2 K (1 + Ts)
Ka = Lim G c (s)G(s) s 2 = Lim = K = 10 (Given)
s® 0 s® 0 s 2 (1 + 0.2s) (1 + aTs)
10
\ G 1 (s) =
s 2 (1 + 0.2s)
Step 2 : Sketch the Bode plot of G 1 (s) as shown in the Fig. 14.2.
(See Fig. 14.2 on next page)
Factors : K = 10 i.e. 20 Log K = 20 dB
Two poles at the origin, straight line of slope – 40 dB/dec.
1 1
Simple Pole, , T1 = 0.2 , wC1 = = 5, slope - 20 dB / dec
1 + 0.2s T1
0 < w < w C1(5) – 40 dB/dec
5 £w< ¥ – 40 – 20 = – 60 dB/dec
10
Phase angle table : G 1 (jw) =
2
(jw) (1 + 0.2 jw)
TM

MR 1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
in dB
Scale
60 On Y axis : 1 unit = 20 dB,30º
– 40 dB/dec wgc = 3 rad/sec
– 60 dB/dec
40
– 40 dB/dec
20 20 Log K = 20 dB
wm = 6.6 rad/sec
0 dB
– 60 dB/dec
–20
– 15.08 dB
–40
fR 2
(1/s )
in deg
TM
–90º
14 - 7
Fig. 14.2 (a)

–120º
–150º
–180º

P.M.
= – 30º
–210º
–240º
–270º
0.1 1 10 100 Log w

w 1 -tan -1 0.2 w fR
2
( jw)
0.5 – 180º – 5.71º – 185.71º
2 – 180º – 21.8º – 201.8º
5 – 180º – 45º – 225º
¥ – 180º – 90º – 270º
From the Bode plot,

f1 = P.M. = - 30º, wgc = 3 rad sec, G.M. = – ¥ dB
Step 3 : Required P.M. fs = + 35º
\ fm = fs - f1 + e = 35 - ( - 30º ) + 5º = 70º
1 -a 1 -a
Step 4 : sin fm = i.e. sin 70º = = 0.9397
1+a 1+a
\ a = 0.031
-10 Log æç ö÷ = – 15.08 dB for a = 0.031

1
Step 5 :
èaø
Find the frequency from the Fig. 14.2 (a) at which gain of the uncompensated system is
– 15.08 dB. This is wm .
Thus wm = 6.6 rad sec
1 1
\ But wm = i.e. = 6.6 ´ 0.031 = 1.162
T a T
Step 6 : The two corner frequencies of lead compensator are,
1 1
wC1 = = 1.162 and wC2 = = 37.48
T aT
K 10
Step 7 : K = Kc a i.e. Kc = = = 322.58
a 0.031
322.58 ´ 0.031 (1 + 0.8606 s) 10 (1 + 0.8606 s)
Step 8 : G c (s) = =
(1 + 0.02667s) (1 + 0.02667 s)
Thus the compensated systems is,
10 (1 + 0.8606 s)
G c (s)G(s) =
s 2 (1 + 0.02667s)(1 + 0.2 s)
Draw the bode plot of compensated system.

Factors : K = 10, 20 Log K = 20 dB
Two poles at the origin, – 40 dB/dec straight line
Simple zero, (1 + 0.8606 s), T1 = 0.8606, wC1 = 1.16, + 20 dB/dec.
TM

MR SEMI-LOG PAPER (5 CYCLES X 1/10)
in dB 1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
80
Scale
60 On Y axis : 1 unit = 20 dB,30º
– 40 dB/dec
– 60 dB/dec
40
– 40 dB/dec wgc = 6 rad/sec
wC1
– 20 dB/dec 20 Log K = 20 dB
20
– 20 dB/dec wC2
0 dB
G.M. = + 10 dB
– 40 dB/dec
–20 wC3
– 60 dB/dec
–40
fR 2
(1/s )
in deg
TM
14 - 9
–90º
Fig. 14.2 (b)

–120º
–150º
P.M. = 36º
–180º

–210º
–240º
–270º
wpc = 11 rad/sec
0.1 1 10 100 Log w

1
Simple pole, , T = 0.2, wC2 = 5, – 20 dB/dec.
(1 + 0.2 s) 2
1
Simple pole, , T3 = 0.02667, wC3 = 37.48, – 20 dB/dec.
1 + 0.0266 s
0 < w < wC1 (1.16) – 40 dB/dec.

1.16 £ w < wC2 (5) – 40 + 20 = – 20 dB/dec.
5 £ w £ wC3 (37.48) – 20 – 20 = – 40 dB/dec.
37.48 £ w < ¥ – 40 – 20 = – 60 dB/dec.
10 (1 + 0.8606 jw)
Phase angle table : G c (jw) G(jw) =
2
(jw) (1 + 0.2 jw) (1 + 0.02667 jw)
w 1 + tan - 1 0.8606 w - tan - 1 0.2 w - tan - 1 0.02667 w fR

(jw) 2
0.5 – 180º + 23.28º – 5.71º – 0.76º – 163.19º
2 – 180º + 59.84º – 21.8º – 3.05º – 145º
5 – 180º + 76.91º – 45º – 7.59º – 155.68º
10 – 180º + 83.37º – 63.43º – 14.93º – 175º
40 – 180º + 88.33º – 82.87º – 46.85º – 221.4º
¥ – 180º + 90º – 90º – 90º – 270º
Sketch the bode plot as shown in the Fig. 14.2 (b). (Fig. 14.2 (b) see on previous page).
From the plot, wgc = 6 rad/sec., wpc = 11 rad/sec.
G.M. = + 10 dB P.M. = 36º
Thus the compensated system satisfies all the specifications.
Example 14.5.3
100 K
Solution : Step 1 : Assume G1(s) = KG( s) =
s( s + 1)
Lim Lim
Kv = s G1(s)Gc(s) and originally Kv = sG(s) = 100
s®0 s®0
(1 + sT)
For lag compensator, Gc(s) =
(1 + sbT)
Lim 100 K (1 + sT)
\ Kv = s× ´ = 100 K
s®0 s( s + 1) (1 + sbT)
For not to sacrifice value of Kv , 100 K = 100 hence K = 1
TM

100 100
\ G 1 ( s) = i.e. G 1 (j w) =
s( s + 1) jw(1 + jw)
Step 2 : Draw the Bode plot of G 1 ( s)

K = 100, 20 Log K = 40 dB
One pole at the origin hence slope – 20 dB/dec.
1
Simple pole of (1/1+s) with T1 = 1, wC1 = = 1 hence straight line of slope –20 dB/dec
T1
for w > 1.
0 < w < 1 ( wC1) – 20 dB/dec
1 < w< ¥ – 20 –20 = – 40 dB/dec
w 1 –tan–1w fR
jw
0.1 – 90º – 5.71º – 95.71º
1 – 90º – 45º – 135º
2 – 90º – 63.43º – 153.43º
10 – 90º – 84.28º – 174.28º
¥ – 90º – 90º – 180º
The bode plot is shown in the Fig. 14.3. (See Fig. 14.3 on next page)
From the plot, wgc = 10 rad/sec, wpc = ¥, G.M. = +¥ dB, P.M. = + 6º
Step 3 : fs = 15º specified P.M., e = 5º
\ f2 = fs + e = 15+5 = 20º
Step 4 : Find w2 which gives 20º P.M. from the Bode plot.
\ w2 = 3 rad/sec ...From Bode plot
Step 5 : So this w2 must be new wgc hence magnitude plot must be brought down by
21 dB, from the Bode plot.
\ –20 Log b = –21 ...Negative as down shift
\ b = 11.22
w 3
Step 6 : Choose wC2 = 2 = = 0.3 rad/sec
10 10
1 1
But wC2 = i.e. T= = 3.333
T 0. 3
1 1
and wC1 = = = 0.026, b T = 37.4
bT 11.22 ´ 3. 333
TM

MR in dB SEMI-LOG PAPER (5 CYCLES X 1/10)
1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
80
Scale on Y axis
1 unit = 20 dB, 15º
60 –20 dB/dec
–40 dB/dec
40 K = 100
wgc = 10
20
21 dB
down
0 dB
– 20
fR
(1/s)
TM
– 90º
14 - 12
Fig. 14.3
– 105º
– 120º
– 135º

Point for
– 150º P.M. 20º
– 165º New
P.M. 20º
P.M. = 6º
– 180º
0.1 1 w2 10 100 1000 Log w

3
Step 7 : Hence the lag compensator is,
(1 + 3. 333 s)
G c ( s) =
(1 + 37.4 s)
Thus the total transfer function is,
100(1 + 3. 333 s) 100 (1 + 3. 333 jw)
G 1 ( s) G c ( s) = , G 1 (jw) G c (jw) =
s (1 + 37.4 s) (1 + s) jw(1 + 37.4 jw)(1 + jw)
Draw the Bode plot to check new P.M.

K = 100, 20 Log K = 40 dB
One pole at origin, straight line of slope –20 dB/dec
1 1
Simple pole, , T = 37.4, wC1 = = 0.026, Slope –20 dB/dec for w > wC1
1 + 37.4 s 1 T1
1
Simple zero, (1+3.333 s), T2 = 3.333 , wC2 = = 0.3 , Slope +20 dB/dec for w > wC2
T2
1 1
Simple pole, , T3 = 1, wC 3 = = 1, Slope – 20 dB/dec for w > wC 3
1+s T3

0 < w < 0.026 (wC1) – 20 dB/dec
0.026 < w < 0.3 (wC2 ) – 20 – 20 = – 40 dB/dec
0.3 < w < 1(wC3 ) – 40 + 20 = – 20 dB/dec
1 < w< ¥ – 20 – 20 = – 40 dB/dec
w 1 -tan -1 37. 4 w t tan -1 3.333w -tan -1 w fR

jw
0.026 – 90º – 45º + 5.02º – 1.51º – 131.49º

0.3 – 90º – 84.9º + 45º – 16.7º – 146.6º
0.5 – 90º – 86.9º + 59.04º – 26.56º – 144.42º
1 – 90º – 88.46º + 73.3º – 45º – 150.16º
5 – 90º – 89.69º + 86.56º – 78.69º – 171.82º
The modified Bode plot is shown in the Fig. 14.3 (a). (See Fig. 14.3 (a) on next page)
From the Bode plot, wgc = 2.4 , wpc = ¥
G.M. = + ¥ dB , P.M. = +20º
Thus the required specification of P.M. = 15º is matched with Kv unchanged.
TM

–20 dB/dec
1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
80
Scale on Y axis
1 unit = 20 dB, 15º
60
–40 dB/dec
–20 dB/dec
40 K = 100
wgc = 2.4
20
0 dB
–40 dB/dec
– 20
(1/s)
fR
TM
14 - 14
– 90º
Fig. 14.3 (a)

– 105º
– 120º

– 135º
– 150º
– 165º
P.M.
+ 20º wpc ® ¥
– 180º
0.01 0.1 1 10 100 Log w

Example 14.5.4
Solution :
K
Step 1 : G1(s) = G(s) =
s (1 + s)(1 + 0.2s)
\ Kv = 5 = Lim
s ®0
s G 1 ( s) = K i.e. K=5
5 5
\ G 1 ( s) = , G 1 (jw) =
s(1 + s) (1 + 0.2 s) jw (1 + jw)(1 + 0. 2 jw)
Step 2 : Draw the Bode plot of G1(s)

K = 5, 20 Log K = 13.98 dB
One pole at origin, straight line of slope –20 dB/dec
1 1
Simple pole, , T1 = 1, wC1 = = 1, Straight line of slope –20 dB/dec for w > 1
1+s T1
1 1
Simple pole, , T2 = 0.2, wC2 = = 5, Straight line of slope –20 dB/dec for w >5.
1 + 0.2s T2
0 < w < 1 (w C1) – 20 dB/dec
1 < w < 5 (w C2) – 20 – 20 = – 40 dB/dec
5 < w< ¥ – 40 – 20 = – 60 dB/dec
Phase angle table
w 1 -tan -1 w -tan -1 0. 2 w fR
jw
0.1 – 90º – 5.71º – 1.145º – 96.85º
1 – 90º – 45º – 11.3º – 146.31º
3 – 90º – 71.56º – 30.96º – 192.52º
5 – 90º – 78.56º – 45º – 213.69º
¥ – 90º – 90º – 90º – 270º
The Bode plot is shown in the Fig. 14.4 (See Fig. 14.4 on next page). From the Bode plot,
wpc = 2.4 rad/sec, wgc = 2.4 rad/sec, G.M. = 0 dB, P.M. = 0º
Uncompensated system is marginally stable.
TM

1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
80
Scale on Y axis
1 unit = 20 dB, 30º
60
–20 dB/dec
–40 dB/dec
40
20
K=5
wgc = 2.4
0 dB
20 dB
down
– 20
fR
TM
Point
– 90º
for
14 - 16
(1/s)
Fig. 14.4
P.M. = 50º
– 120º
–60 dB/dec
– 150º
New P.M.
+50º wpc = 2.4 G.M. = P.M. = 0

– 180º
– 210º
– 240º
w2 = 0.5 line
– 270º
0.1 1 10 100 1000 Log w

Step 3 : fs = 45º given P.M. , e = 5º

New P.M. f2 = fs + e = 45 + 5 = 50º
Step 4 : From the Bode plot in the Fig. 14.4, find w2 which gives 50º P.M. Thus
w2 = 0.5 rad/sec from the Bode plot.
Step 5 : Let w2 = 0.5 is the new wgc . For this, the magnitude plot must be brought
down by 20 dB.
\ –20 Log b = –20 ...–ve as downward shift.
\ b = 10
w2 0 .5
Step 6 : Choose wC2 = = = 0.05 rad/sec
10 10
1 1
Now, wC2 = i.e. T= = 20
T 0.05
1 1
\ wC1 = = = 0.005, b T = 200
bT 20 ´ 10
Step 7 : The log compensator is,
(1 + 20 s)
G c ( s) =
(1 + 200 s)
Thus the transfer function of the compensated system is,
5(1 + 20 s)
G 1 ( s)G c ( s) =
s (1 + 200 s) (1 + s) (1 + 0.2 s)
5(1 + 20 jw)
G 1 (jw)G c (jw) =
jw (1 + 200 jw)(1 + jw)(1 + 0.2 jw)
For checking the specification, draw the Bode plot of compensated system.
K = 5, 20 Log K = 13.97 dB
One pole at the origin, straight line of slope –20 dB/sec.
1 1
Simple pole, , T1 = 200, wC1 = = 0.005, Straight line of slope –20 dB/dec for
1 + 200 s T1
w > wC1
1
Simple zero, 1+20s, T2 = 20, wC2 = = 0.05, Straight line of slope +20 dB/dec for
T2
w > wC2
1 1
Simple pole , , T3 = 1, wC 3 = = 1, Straight line of slope –20 dB/dec for
1+s T3
w > wC 3
TM

1 1
Simple pole, , T4 = 0.2, wC4 = = 5, Straight line of slope –20 dB/dec for
1 + 0.2s T4
w >wC4
0 < w < 0.005 –20 dB/dec
0.005 < w < 0.05 – 20 – 20 = – 40 dB/dec
0.05 < w < 1 – 40 + 20 = –20 dB/dec
1 < w< 5 – 20 – 20 = – 40 dB/dec
5 < w< ¥ – 40 – 20 = – 60 dB/dec
w 1 -tan -1 200 w t tan -1 20w -tan -1 w -tan -1 0. 2 w fR

jw
0.005 – 90º – 45º + 5.71º – 0.28º – 0.05º – 129.62º
0.05 – 90º – 84.28º + 45º – 2.86º – 0.57º –132.71º
1 – 90º – 89.71º + 87.13º – 45º – 11.3º –148.88º
2 – 90º – 89.85º + 88.56º – 63.43º – 21.8º –174.23º
5 – 90º – 89.94º + 89.42º – 78.69º – 45º –214.21º
¥ – 90º – 90º + 90º – 90º – 90º – 270º
Draw the Bode plot as shown in the Fig. 14.4 (a). (See Fig. 14.4 (a) on next page)
From the Bode plot,
wgc = 0.8 rad/sec, wpc = 2.4 rad/sec, G.M. = +16 dB, P.M. = +45º
The compensated system is stable with all the secpifications are matched, due to
compensation.
TM

–20 dB/dec

1 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91 2 3 4 5 6 7 8 91
80
Scale on Y axis
– 40 dB/dec 1 unit = 20 dB, 30º
60
–20 dB/dec
40
wgc = 0.8
20
K=5
0 dB
–40 dB/dec
– 20
G.M.
+16 dB
– 60 dB/dec 1
––
s
((
TM
fR
14 - 19
– 90º
Fig. 14.4 (a)

– 120º
– 150º
P.M.
45º

– 180º
– 210º
– 240º
wpc = 2.4
– 270º
0.001 0.01 0.1 1 10 Log w

Example 14.8.2
K
Solution : G(s) =
s (s + 7)
0.15 = e - px / 1- x 2
MP = 15 % i.e. i.e. x = 0.5169
K
i) 1 + G(s) = 1 + =0
s (s + 7)
2
\ s + 7s + K = 0 … Characteristic equation
\ wn = K, 2xwn = 7
7
\ K = = 6.7711 i.e. K = 45.8482 and wn = 6.7711
2 ´ 0.5169
4
\ Ts = = 1.1428 sec
x wn
1
ii) Ts(new) = T (old) = 0.3809 sec
3 s
4
Ts(new) =
x w¢n
4
\ w¢n = = 20.3162 rad/sec.
0.5169 ´ 0.3809
The desired closed loop poles are
= – xw¢n ± jw¢n 1 - x 2 = – 10.5 ± j 17.3916
K a (1 + Ts )
Let Gc(s) =
(1 + a Ts )
Ð G(s)H(s) at dominant pole
K
= Ð
s( s + 7) s = -10 . 5 + j 17.3916
0° 0°
= = = – 222.49°
Ð -10.5 + j 17.3916 Ð - 3.5 + j 17.3916 121 . 12° 101 . 38°
Angle to be contributed by lead compensator
= – 180 – (– 222.49°) = + 42.49° … (Given)
Compensator's zero given to be at – 10.
\ s = – 10 … Zero
1
\ = 10 \ T = 0.1
T
To find the pole,
Approximately location of pole can be obtained from the Fig. 14.5 (a).
TM

Let DC = x, DP » 10
CD P
\ tan f = +j17.3916
DP
CD f
\ tan 42.49° = –
10 f 2
–
\ CD = 9.16 2
\ Point C = – 19.16 42.49º

C D
1 X
\ = 19.16 –10.5 –10 0
aT
\ aT = 0.052 Location
of compensator pole
\ a = 0.52 < 1
(1 + 0.1 s) Fig. 14.5 (a)
\ Gc(s) =
(1 + 0.052 s)
K (1 + 0.1 s)
\ G(s)Gc(s) =
s ( s + 7) (1 + 0.052 s)
Now at s = – 10.5 + j 17.3916, |G(s)H(s)| = 1
|K | | -0.05 + j 1.7391 |
\ =1
| -10.5 + j 17.3916 | | -3.5 + j 17.3916 | | 0.454 + j 0.9043 |
\ K = 209.607
Hence compensated system is,
209.607 (1 + 0.1s) 403.1 (s + 10)
G(s)Gc(s) = =
s (s + 7) (1 + 0.052 s) s (s + 7) (s + 19.23)
The corresponding rough root locus is shown in the Fig. 14.5 (b).
NRL NRL
–19.23 –7 0
–10
Fig. 14.5 (b)
qqq
TM

3320100XZ8L - Control System Engineering - Solution Manual

Uploaded by

Copyright:

Available Formats

3320100XZ8L - Control System Engineering - Solution Manual

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

3320100XZ8L - Control System Engineering - Solution Manual

Uploaded by

Copyright:

Available Formats

3 Transfer Function and Impulse Response

Solutions of Examples for Practice

Fig. 3.1 (a)

Divide the network into two parts,

TECHNICAL PUBLICATIONS - An up thrust for knowledge

Amplifier gain = 10 hence V2(s) = 10 V1(s)

10E i (s) E o (s) 10

And, Vo (s) = I(s) R i.e. Vo (s)

TECHNICAL PUBLICATIONS - An up thrust for knowledge

So we can represent the system as in the 1

TECHNICAL PUBLICATIONS - An up thrust for knowledge

Example 3.4.2 Kept this unsolved example for student's practice.

TECHNICAL PUBLICATIONS - An up thrust for knowledge

Solutions of Examples for Practice

\ Taking Laplace of equation (3) and assuming initial conditions zero.

where J eq = Equivalent M.I., B eq = Equivalent friction.

TECHNICAL PUBLICATIONS - An up thrust for knowledge

The F-V analogous network is shown in the Fig. 4.1.

TECHNICAL PUBLICATIONS - An up thrust for knowledge

The F-I analogous system is shown in the Fig. 4.1 (a).

Same displacement same voltage.

Fig. 4.1 (a)

TECHNICAL PUBLICATIONS - An up thrust for knowledge

The s domain network for the given electric network is,

TECHNICAL PUBLICATIONS - An up thrust for knowledge

\ 0 = Kq(s) - Kq 1 (s) + J 2 s 2 q(s) …(4) T J1 J2

Using equation (5) in equation (1), Fig. 4.3 (c)

TECHNICAL PUBLICATIONS - An up thrust for knowledge

Taking Laplace transform of both the equations,

Using in equation (3),

0 = KX 2 (s) - Kìí [F(s) + KX 2 (s)]üý Ms 2 X 2 (s) + Bs X 2 (s)

\ F(s) = sX 2 (s) [Ms + B]

TECHNICAL PUBLICATIONS - An up thrust for knowledge

The force-current equivalent analogous electrical

e(s) = I (s) R + Kb s q (s) ... (3)

TECHNICAL PUBLICATIONS - An up thrust for knowledge

\ Equating (1) and (2), KT I(s) = Ks q(s) + s 2 J q(s)

TECHNICAL PUBLICATIONS - An up thrust for knowledge

Simulating using loop basis,

Currents through various elements,

TECHNICAL PUBLICATIONS - An up thrust for knowledge

The equations of equilibrium are,

TECHNICAL PUBLICATIONS - An up thrust for knowledge

Remember parallel elements in series and series elemets in parallel.

M1, K1 parallel L1 , C1 series

TECHNICAL PUBLICATIONS - An up thrust for knowledge

B1, K 3 series R1, C3 parallel

B1 , K 3 series in R1 C3 parallel in series

Thus the F-V analogous network based on loop basis is,

TECHNICAL PUBLICATIONS - An up thrust for knowledge

The equivalent mechanical system is,

For F-V analogy, M ® L , B ® R, K ® 1/C, x ® q, dx/dt ® I

F2 (s) = M2 s 2 X 2 (s) + B 2 s X 2 (s) + K2 X 2 (s) ... (8)

TECHNICAL PUBLICATIONS - An up thrust for knowledge

Substituting (7) and (8) in (9) we get,

TECHNICAL PUBLICATIONS - An up thrust for knowledge

Solutions for Examples for Practice

R(s) + G1 G2 (G3 + G4 + G5)