C H A P T E R 6

The Stability of Linear Feedback

Systems

Exercises
E6.1 The Routh array is

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
s3 1 w
K +1
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r

s2 K 6
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es

s1
of or stu e o tat

b 0
ity s w g us d S

is
te f t ss th nite

so 6
e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te

where
st f a s d s ec
de o rse de ot
ill le u vi pr

K(K + 1) − 6
w r sa co pro is

b= .
o eir is rk
th nd wo

K
a his
T

For stability, we require K > 0 and b > 0. Therefore, using the condition
that b > 0, we obtain

K2 + K − 6 > 0 ,

and solving for K yields K > 2 and K < −3. We select K > 2, since we
also have the condition that K > 0.
E6.2 The Routh array is

s3 1 2
s2 15 40
s1 -2/3 0
so 40
The system is unstable since the first column shows two sign changes.

235

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 236 CHAPTER 6 The Stability of Linear Feedback Systems

E6.3 The Routh array is

s4 1 32 10
s3 10 37
s2 28.3 10
s1 33.5
s0 10

By the Routh-Hurwitz criterion, the system is stable (i.e., all the numbers
in the first column are positive).
E6.4 The closed-loop transfer function is

−K(s − 2)
T (s) = .

)
eb
s3 + 5s2 + (4 − K)s + 2K

er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

Therefore, the characteristic equation is

.
r
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es
of or stu e o tat

s3 + 5s2 + (4 − K)s + 2K = 0 .
ity s w g us d S

is
te f t ss th nite
e rt ss fo U
gr hi in e

The corresponding Routh array is given by

th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot

s3 1 4−K
ill le u vi pr
w r sa co pro is
o eir is rk

s2 5 2K
th nd wo
a his
T

s1 b 0
so 2K

where
5(4 − K) − 2K 20 − 7K
b= = .
5 5
For stability we require K > 0 and b > 0 . Thus, the range of K for
stability is 0 < K < 20/7.
E6.5 The closed-loop transfer function is

K
T (s) = .
s3 + 10s2 + 27s + 18 + K

When K = 20, the roots of the characteristic polynomial are

s1,2 = −1.56 ± j1.76

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Exercises 237

and

s3 = −6.88 .

E6.6 When K = 252, the roots of the characteristic equation are on the imag-
inary axis. The roots are

s1,2 = ±j5.2 and s3 = −10 .

E6.7 (a) The closed-loop system characteristic equation is

K(s + 2)
1 + GH(s) = 1 + =0,
s(s − 1)
or

s2 + (K − 1)s + 2K = 0 .

)
eb
√

er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch
We have the relationships ωn = 2K and 2ζωn = K − 1, where

e
d on g. in t la

m ld a
an ing rnin tors igh

ζ = 0.707. Thus,
.
r
or ud a uc y
w cl le tr p
e in nt ns co

1 √
th k ( de f i es

 
of or stu e o tat

2 √ 2K = K − 1 ,
ity s w g us d S

2
is
te f t ss th nite
e rt ss fo U
gr hi in e

or
th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec

2 2
2 K−1
 
de o rse de ot

√ √
ill le u vi pr

= ,
w r sa co pro is

2 2K
o eir is rk
th nd wo
a his

and
T

K 2 − 6K + 1 = 0 .

Solving for K yields K = 5.83 and K = 0.17. However, for stability

we require K > 1 (from the Routh array), so we select K = 5.83.
√
(b) The two roots on the imaginary axis when K = 1 are s1,2 = ±j 2.
E6.8 The closed-loop system characteristic equation is
3
+ 20s2 + (100 + K)s + 20K = 0 .

The corresponding Routh array is

s3 1 (100 + K)
s2 20 20K
s1 b 0
so 20K

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 238 CHAPTER 6 The Stability of Linear Feedback Systems

where
20(100 + K) − 20K 20(100)
b= = = 100 .
20 20
Therefore, the system is stable for all K > 0.
E6.9 The characteristic equation is

s3 + 52 + (K + 1)s + 10 = 0 ,

and the Routh array is given by

s3 1 K +1
s2 5 10
s1 b 0

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch
so 10

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r
or ud a uc y
w cl le tr p

where
e in nt ns co

D
th k ( de f i es
of or stu e o tat

5(K + 1) − 10
ity s w g us d S

b= =K −1 .
is
te f t ss th nite

5
e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te

Setting b = 0, yields
st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is

K −1=0 ⇒ K>1.
o eir is rk
th nd wo
a his
T

E6.10 The characteristic equation is

s3 + 42 + s + 4 = 0 ,

and the Routh array is given by

s3 1 1
s2 4 4 ⇔ A(s) = 4s2 + 4
s1 0 0 ⇐ Row of 0’s
so

The auxiliary equation is

A(s) = 4s2 + 4.

Solving A(s) = 0 yields two poles on the imaginary axis at s = ±j.

Restarting the Routh array on the s1 -row with dA(s)/ds = 8s, we have

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Exercises 239

.. ..
. .
s1 8 0
so 4

Hence the system is marginally stable with two poles on the jω-axis and
one pole in the right half-plane.
E6.11 The tracking error, E(s) = Y (s) − R(s) is given by

s4 + 6s3 + s2 − 9s − 7
E(s) = .
s4 + 6s3 + s2 + s + 3

Computing the steady-state error to a unit step input yields ess = −7/3.

s4 1 1 3

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
s3 w
t p W em ch
6 1

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r

s2
or ud a uc y

0.833 3
w cl le tr p
e in nt ns co

D
th k ( de f i es
of or stu e o tat

s1 -20.6
ity s w g us d S

is
te f t ss th nite

so 3
e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te

The system is unstable. The closed-loop poles are s1 = −5.8431, s2 =

st f a s d s ec
de o rse de ot
ill le u vi pr

−0.8196 and s3,4 = 0.3314 ± j0.7188.

w r sa co pro is
o eir is rk
th nd wo

E6.12 The characteristic equation is

a his
T

s2 + as + b = 0,

so, the Routh array is

s2 1 b
s1 a 0
so b

The system is stable if and only if a > 0 and b > 0. For a second-order
system, a necessary and sufficient condition for stability is that all the
coefficients have the same sign.
E6.13 The characteristic equation is

s2 + (KD − 2)s + KP = 0.

The Routh array is

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 240 CHAPTER 6 The Stability of Linear Feedback Systems

s2 1 KP
s1 KD − 1 0
so KP
The system is stable if and only if KP > 0 and KD > 1.
E6.14 The characteristic equation associated with the system matrix is

s3 + 3s2 + 5s + 6 = 0 .

The roots of the characteristic equation are s1 = −2 and s2,3 = −5±j1.66.

The system is stable.
E6.15 The roots of q(s) are s1 = −4, s2 = −3, s3,4 = −1 ± j2 and s5,6 = ±j0.5.
The system is marginally stable. The Routh array is

)
eb
s6
er or in ing
1 31.25 67.75 15

ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

s5 9 61.25 14.75

.
r
or ud a uc y
w cl le tr p
e in nt ns co

s4 24.44 66.11 15
th k ( de f i es
of or stu e o tat
ity s w g us d S

s3 31.909 9.2273 0
is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by

s2 60 15
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot

s1 0 0
ill le u vi pr
w r sa co pro is
o eir is rk

so
th nd wo
a his
T

The auxillary equation is

60s2 + 15 = 0 .

Solving the auxillary equation yields two roots at s1,2 = ±j0.5. After
accounting for the row of zeros, the completed Routh array verifies that
the system has no poles in the right half-plane.
E6.16 The Routh array is

s4 1 45 50
s3 9 87
s2 35.33 50
s1 74.26 0
so 50

The system is stable. The roots of q(s) are s1,2 = −3 ± j4, s3 = −2 and

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Exercises 241

s4 = −1.
E6.17 The characteristic equation is

s3 + 7s2 + 36s + 24 = 0 .

The system is stable. The roots of the characteristic equation are s1 =

−0.77, s2,3 = −3.12 ± 4.64i.
E6.18 The roots of q(s) are s1 = −20 and s2,3 = ±j2.24. The system is marginally
stable. The Routh array is

s3 1 5
s2 20 100
s1 0 0

)
eb
er or in ing
ed id n
so

W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r
or ud a uc y

The auxillary equation is

w cl le tr p
e in nt ns co

D
th k ( de f i es
of or stu e o tat

20s2 + 100 = 0 .
ity s w g us d S

is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by

The roots are s = ±j2.24. So, the system has roots at s = ±j2.24.
in o e r
y y p d le d
ro n an o te
st f a s d s ec

Completing the Routh array (after accounting for the row of zeros) verifies
de o rse de ot
ill le u vi pr

that no poles lie in the right half-plane.

w r sa co pro is
o eir is rk
th nd wo

E6.19 (a) Unstable.

a his
T

(b) Unstable.
(c) Stable.
E6.20 (a) The roots are s1,2 = −2 and s3 = −1.
(b) The roots are s1,2,3 = −3.
E6.21 This system is not stable. The output response to a step input is a ramp
y(t) = kt.
E6.22 The characteristic equation is

(sn − 2)3 + 10(sn − 2)2 + 29(sn − 2) + K = 0

or

s3n + 4s2n + sn − 26 + K = 0 .

The Routh array is

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 242 CHAPTER 6 The Stability of Linear Feedback Systems

s3 1 1
s2 4 K − 26
30−K
s1 4 0
so K − 26

If K = 30, then the auxillary equation is 4s2n + 4 = 0 or sn = ±j.

Therefore, s = sn − 2 implies s = −2 ± j.
E6.23 The characteristic polynomial is

s3 + 4s2 + ks + 8 = 0 .

The Routh array is

)
eb
er or in ing
ed id n
W
s3 no the iss tea s

itt W tio
1 w k
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

s2
.
r

4 8
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es

4k−8
s1
of or stu e o tat

4
ity s w g us d S

is
te f t ss th nite

so 8
e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te

So, k > 2 for stability.

st f a s d s ec
de o rse de ot
ill le u vi pr

The transfer function is

w r sa co pro is

E6.24
o eir is rk
th nd wo
a his
T

G(s) = C(sI − A)−1 B + D

  
 s −1 0
 0 
  
= [ 1 0 0 ]
 0 s −1 
 0 
 
  
k k s+k 1
  
2
 s + ks + k s+k 1  0 
   1
= [ 1 0 0 ] −k s2 + ks s 
 0 
 
 ∆(s)

 
−ks −ks − k s2 1

where ∆(s) = s3 + ks2 + ks + k. Thus, the transfer function is

1
G(s) = .
s3 + ks2 + ks + k

The Routh array is

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Exercises 243

s3 1 k
s2 k k
s1 k−1
so k
For stability k > 1.
E6.25 The closed-loop transfer function is
Ks + 2
T (s) = .
s2 (s + p) + Ks + 2
Therefore, the characteristic equation is

s3 + ps2 + Ks + 2 = 0 .

)
eb
er or in ing
ed id n
The Routh array is

W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

s3
.
r

1 K
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es

s2 p 2
of or stu e o tat
ity s w g us d S

is
te f t ss th nite

s1 (pK − 2)/p
e rt ss fo U
gr hi in e
th a a ly by

k
in o e r

so
y y p d le d

2
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr

We see that the system is stable for any value of p > 0 and pK − 2 > 0.
w r sa co pro is
o eir is rk
th nd wo

E6.26 The closed-loop transfer function is

a his
T

4
T (s) = .
2s2 + (K − 2)s + 4 − K
Therefore, the characteristic equation is

2s2 + (K − 2)s + 4 − K = 0 .

The Routh array is

s2 2 4−K
s1 K −2
so 4−K
We see that the system is stable for any value of K > 20 and any K < 4.
Therefore, the system is stable for all 0 < K < 4..

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 244 CHAPTER 6 The Stability of Linear Feedback Systems

Problems
P6.1 (a) Given

s2 + 5s + 2 ,

we have the Routh array

s2 1 2
s1 5 0
so 2
Each element in the first column is positive, thus the system is stable.
(b) Given

s3 + 4s2 + 8s + 4 ,

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
we have the Routh array
an ing rnin tors igh

.
r
or ud a uc y
w cl le tr p

s3 1 8
e in nt ns co

D
th k ( de f i es
of or stu e o tat

s2 4 4
ity s w g us d S

is
te f t ss th nite

s1
e rt ss fo U

7 0
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te

so 4
st f a s d s ec
de o rse de ot
ill le u vi pr

Each element in the first column is positive, thus the system is stable.
w r sa co pro is
o eir is rk
th nd wo

(c) Given
a his
T

s3 + 2s2 − 6s + 20 ,

we determine by inspection that the system is unstable, since it is

necessary that all coefficients have the same sign. There are two roots
in the right half-plane.
(d) Given

s4 + s3 + 2s2 + 12s + 10 ,

we have the Routh array

s4 1 2 10
s3 1 12 0
s2 -10 10 0
s1 13 0
so 10

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Problems 245

There are two sign changes in the first column, thus the system is
unstable with two roots in the right half-plane.
(e) Given

s4 + s3 + 3s2 + 2s + K ,

we have the Routh array

s4 1 3 K
s3 1 2 0
s2 1 K
s1 2−K 0
so K
Examining the first column, we determine that the system is stable

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
for 0 < K < 2. w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

(f) Given
.
r
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es

s5 + s4 + 2s3 + s + 6 ,
of or stu e o tat
ity s w g us d S

is
te f t ss th nite

we know the system is unstable since the coefficient of the s2 term is

e rt ss fo U
gr hi in e
th a a ly by

missing. There are two roots in the right half-plane.

in o e r
y y p d le d
ro n an o te
st f a s d s ec

(g) Given
de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk

s5 + s4 + 2s3 + s2 + s + K ,
th nd wo
a his
T

we have the Routh array

s5 1 2 1
s4 1 1 K
s3 1 1−K
s2 K K
s1 −K 0
so K
Examining the first column, we determine that for stability we need
K > 0 and K < 0. Therefore the system is unstable for all K.
P6.2 (a) The closed-loop characteristic polynomial is

s4 + 27.88s3 + 366.4s2 + 1500s + 1500ka = 0 .

The Routh array is

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 246 CHAPTER 6 The Stability of Linear Feedback Systems

s4 1 366.4 1500ka
s3 27.88 1500
s2 312.6 1500ka
s1 b
so 1500ka
where

b = 1500 − 133.78ka .

Examining the first column of the Routh array, we find that b > 0
and 1500ka > 0 for stability. Thus,

0 < ka < 11.21 .

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la
(b) With
m ld a
an ing rnin tors igh

.
r
or ud a uc y
w cl le tr p

4
e in nt ns co

Ts = 1.5 = ,
th k ( de f i es

ζωn
of or stu e o tat
ity s w g us d S

is
te f t ss th nite

we determine that
e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te

ζωn = 2.67 .
st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk

So, shift the axis by s = so − 2.67, and

th nd wo
a his
T

(so − 2.67)4 + 27.88(so − 2.67)3 + 366.4(so − 2.67)2 + 1500(so − 2.67) +

1500ka = s4o + 17.2s3o + 185.85s2o + 63.55so − 1872.8 + 1500ka .

The Routh array is

s4 1 185.85 1500ka -1872.8
s3 17.2 63.55
s2 182.16 1500ka -1872.8
s1 b
so 1500ka -1872.8
where

b = 240.38 − 141.63ka .

Examining the first column of the Routh array, we find that b > 0
and 1500ka − 1872.8 > 0. Thus, 1.25 < ka < 1.69.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Problems 247

P6.3 (a) Given

K
G(s) = ,
(s + 1)(s + 2)(0.5s + 1)
and
1
H(s) = ,
0.005s + 1
the closed-loop transfer function is
K(0.005s + 1)
T (s) = .
0.0025s4 + 0.5125s3 + 2.52s2 + 4.01s + 2 + K
Therefore, the characteristic equation is

0.0025s4 + 0.5125s3 + 2.52s2 + 4.01s + (2 + K) = 0 .

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch
The Routh array is given by

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r

s4
or ud a uc y

0.0025 2.52 2+K

w cl le tr p
e in nt ns co

D
th k ( de f i es
of or stu e o tat

s3 0.5125 4.01 0
ity s w g us d S

is
te f t ss th nite

s2 2.50 2+K
e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d

s1
ro n an o te

3.6 − 0.205K 0
st f a s d s ec
de o rse de ot
ill le u vi pr

so 2+K
w r sa co pro is
o eir is rk
th nd wo

Examining the first column, we determine that for stability we require

a his
T

−2 < K < 17.6 .

(b) Using K = 9, the roots of the characteristic equation are

s1 = −200 , s2,3 = −0.33 ± 2.23j , and s4 = −4.35 .

Assuming the complex roots are dominant, we compute the damping

ratio ζ = 0.15. Therefore, we estimate the percent overshoot as
√
2
P.O. = 100e−πζ/ 1−ζ = 62% .

The actual overshoot is 27%, so we see that assuming that the complex
poles are dominant does not lead to accurate predictions of the system
response.
P6.4 (a) The closed-loop characteristic equation is
K(s + 40)
1 + GH(s) = 1 + =0,
s(s + 10)(s + 20)

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 248 CHAPTER 6 The Stability of Linear Feedback Systems

or

s3 + 30s2 + 200s + Ks + 40K = 0 .

The Routh array is

s3 1 200 + K
s2 30 40K
K
s1 200 − 3 0
so 40K
Therefore, for stability we require 200 − K/3 > 0 and 40K > 0. So,
the range of K for stability is

0 < K < 600 .

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
(b) At K = 600, the auxilary equation is
an ing rnin tors igh

.
r
or ud a uc y
w cl le tr p
e in nt ns co

30s2 + 40(600) = 0 or s2 + 800 = 0 .

th k ( de f i es
of or stu e o tat
ity s w g us d S

is
te f t ss th nite

The roots of the auxiliary equation are

e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te

s = ±j28.3 .
st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk

(c) Let K = 600/2 = 300. Then, to the shift the axis, first define so =
th nd wo
a his

s + 1. Substituting s = so − 1 into the characteristic equation yields

T

(so −1)3 +30(so −1)2 +500(so −1)+12000 = s3o +27s2o +443so +11529 .

The Routh array is

s3 1 443
s2 27 11529
s1 16 0
so 11529
All the elements of the first column are positive, therefore all the
roots lie to left of s = −1. We repeat the procedure for s = so − 2
and obtain

s3o + 24s2o + 392so + 10992 = 0 .

The Routh array is

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Problems 249

s3 1 392
s2 24 10992
s1 -66 0
so 10992
There are two sign changes in the first column indicating two roots
to right of s = −2. Combining the results, we determine that there
are two roots located between s = −1 and s = −2. The roots of the
characteristic equation are

s1 = −27.6250 and s2,3 = −1.1875 ± 20.8082j .

We see that indeed the two roots s2,3 = −1.1875±20.8082j lie between
-1 and -2.

)
eb
er or in ing
ed id n
P6.5 (a) Given the characteristic equation,

W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

s3 + 5s2 + 6s + 2 = 0 ,
.
r
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es
of or stu e o tat

we compute the roots s1 = −3.4, s2 = −1.0, and s3 = −0.59.

ity s w g us d S

is
te f t ss th nite

(b) The roots of the characteristic equation

e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te

s4 + 9s3 + 30s2 + 42s + 20 = 0

st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is

are s1 = −1, s2 = −2, and s3,4 = −3 ± j1.

o eir is rk
th nd wo
a his

(c) The roots of the characteristic equation

T

s3 + 20s2 + 100s + 200 = 0

are s1 = −13.8 and s2,3 = −3.1 ± 2.2j.

P6.6 (a) The characteristic equation is

1 + G(s) = 0 ,

or

s3 + s2 + 10s + 2 = 0 .

The roots are: s1 = −0.2033, and s2,3 = −0.3984 ± j3.1112.

(b) The characteristic equation is

s4 + 10s3 + 35s2 + 50s + 24 = 0 .

The roots are s1 = −1, s2 = −2, s3 = −3, and s4 = −4.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 250 CHAPTER 6 The Stability of Linear Feedback Systems

(c) The characteristic equation is

s3 + 11s2 + 29s + 6 = 0 .

The roots are s1 = −0.2258, s2 = −3.8206 and s3 = −6.9536.

P6.7 (a) The closed-loop characteristic equation is

s3 + 101s2 + (100 + 10KKa )s + 100KKa = 0 .

The Routh array is

s3 1 100 + 10KKa
s2 101 100KKa
s1 b
so 100KKa

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
where
an ing rnin tors igh

.
r
or ud a uc y

910
w cl le tr p
e in nt ns co

b = 100 + KKa > 0 .

th k ( de f i es

101
of or stu e o tat
ity s w g us d S

is
te f t ss th nite

Thus, examing the first column, we determine that KKa > 0 stabi-
e rt ss fo U
gr hi in e
th a a ly by

lizes the system.

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec

(b) The tracking error is

de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk

100 100
th nd wo

e(s) = lim s(1 − T (s)) = .

a his

s→0 s 2 KKa
T

We require E(s) < 1o = 0.01745. So,

100
KKa > = 5729 .
0.01745
When KKa = 5729, the roots of the characteristic polynomial are

s1 = −10.15 and s2,3 = −45.43 ± j233.25 .

P6.8 (a) The closed-loop characteristic equation is

K
1+ =0,
(0.5s + 1)(s + 1)( 14 s + 1)
or

s3 + 7s2 + 14s + 8(1 + K) = 0 .

The Routh array is

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Problems 251

s3 1 14
s2 7 8(1 + K)
s1 b
so 8(1 + K)
where
7(14) − 8(1 + K)
b= .
7
For stability, we require b > 0 and 8(1 + K) > 0. Therefore, the range
of K for stability is

−1 < K < 11.25 .

(b) Let K = 11.25/3 = 3.75. Then, the closed-loop transfer function is

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
3.37
an ing rnin tors igh

T (s) = .

.
r

s3 7s2
or ud a uc y

+ + 14s + 38
w cl le tr p
e in nt ns co

D
th k ( de f i es

The settling time to a step input is Ts = 5.63 seconds.

of or stu e o tat
ity s w g us d S

is
te f t ss th nite

(c) We want Ts = 4 sec, so

e rt ss fo U
gr hi in e
th a a ly by

4
in o e r
y y p d le d
ro n an o te

Ts = 4 = implies ζωn = 1 .
st f a s d s ec
de o rse de ot

ζωn
ill le u vi pr
w r sa co pro is
o eir is rk

Our desired characteristic polynomial is

th nd wo
a his
T

(s + b)(s2 + 2ζωn s + ωn2 ) = s3 + (2 + b)s2 + (ωn2 + 2b)s + bωn2

where we have used the fact that ζωn = 1 and ωn and b are to be
determined. Our actual characteristic polynomial is

s3 + 7s2 + 14s + 8(1 + K) = 0 .

Comparing the coefficients of the actual and desired characteristic

polynomials, we find the following relationships:

2+b=7
ωn2 + 2b = 14
bωn2 = 8(1 + K) .

Solving these three equations yields

b=5, ωn = 2 and K = 1.5 .

The actual settling time is Ts = 4.17 s when K = 1.5. Slightly adjust-

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 252 CHAPTER 6 The Stability of Linear Feedback Systems

ing K to K = 1.41 yields the desired settling time. The closed-loop

poles are

s1 = −4.96 and s2,3 = −1.02 ± 1.69j .

P6.9 (a) The closed-loop characteristic equation is

10K
1 + GH(s) = 1 + ,
(s + 100)(s + 20)2
or

s3 + 140s2 + 4400s + 40000 + 10K = 0 .

The Routh array is

s3 1 4400

)
eb
er or in ing
ed id n
s2

W
no the iss tea s
140 40000 + 10K

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

s1 b
.
r
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es

so 40000 + 10K
of or stu e o tat
ity s w g us d S

is

where
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by

140(4400) − (40000 + 10K)

in o e r
y y p d le d
ro n an o te

b= .
st f a s d s ec

140
de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk

Examining the first column and requiring all the terms to be positive,
th nd wo
a his

we determine that the system is stable if

T

−4000 < K < 57600 .

(b) The desired characteristic polynomial is

(s+b)(s2 +1.38ωn s+ωn2 ) = s3 +(1.38ωn +b)s2 +(ωn2 +1.38ωn b)s+bωn2

where we have used the fact that ζ = 0.69 to achieve a 5% over-

shoot, and ωn and b are to be determined. The actual characteristic
polynomial is

s3 + 140s2 + 4400s + 40000 + 10K = 0 .

Equating the coefficients of the actual and desired characteristic poly-

nomials, and solving for K, b, and ωn yields

b = 104.2 , ωn = 25.9 and K = 3003 .

So, a suitable gain is K = 3003.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Problems 253

P6.10 (a) The closed-loop characteristic equation is

s4 + 7s3 + 20s2 + (24 + K)s + 10K = 0 .

The Routh array is

s4 1 20 10K
s3 7 24 + K 0
116−K
s2 7 10K
s1 b
so 10K
where
 
116−K
7 (24 + K) − 70K
b= .

)
eb
 

er or in ing
116−K

ed id n
W
no the iss tea s

itt W tio
w
t p W em ch
7

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r

Setting b > 0 yields

or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es
of or stu e o tat

2784 − 398K − K 2 > 0 ,

ity s w g us d S

is
te f t ss th nite
e rt ss fo U

which holds when

gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec

−404.88 < K < 6.876 .

de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk

Examining the first column, we also find that K < 116 and K > 0
th nd wo

for stability. Combining all the stability regions, we determine that

a his
T

for stability

0 < K < 6.876 .

(b) When K = 6.876, the roots are

s1,2 = −3.5 ± 1.63j , and s3,4 = ±2.1j .

P6.11 Given

s3 + (1 + K)s2 + 3s + (1 + 7K) = 0 ,

the Routh array is

s3 1 3
s2 1+K 1 + 7K
s1 b
so 1 + 7K

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 254 CHAPTER 6 The Stability of Linear Feedback Systems

where
3(1 + K) − (1 + 7K) 2 − 4K
b= = .
1+K 1+K
Given that K > 0, we determine that the system is stable when 2−4K > 0
or

0 < K < 1/2 .

When K = 1/2, the s2 -row yields the auxillary equation

3s2 + 9 = 0 .
√
The
√ roots are s = ±j 3. So, the system frequency of oscillation is ωn =
3 rad/s.
P6.12 The characteristic equation is

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch
s3 + as2 + bs + c = 0,

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r
or ud a uc y
w cl le tr p

so, the Routh array is

e in nt ns co

D
th k ( de f i es
of or stu e o tat
ity s w g us d S

s3 1 b
is
te f t ss th nite
e rt ss fo U
gr hi in e

s2 a c
th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec

ab−c
s1
de o rse de ot

a
ill le u vi pr
w r sa co pro is

s0
o eir is rk

c
th nd wo
a his
T

For the system to be stable, we require that a > 0, ab − c > 0 and c > 0.
When a > 0 and c > 0, we know that b > 0. So, a necessary condition for
stability is that all coefficients a, b, and c be positive. The necessary and
sufficient conditions for stability also require that b > c/a, in addition to
a > 0 and c > 0.
P6.13 The characteristic equation is

s3 + (p + 2ζωn )s2 + (2ζωn p + Kωn2 )s + Kωn2 z = 0.

The conditions for stability (see P6.12) are p + 2ζωn > 0, 2ζωn p + Kωn2 >
(Kωn2 z)/(p + 2ζωn ), and Kωn2 z > 0. Since we know that K > 0, ζ > 0,
and ωn > 0, it follows that for stability z > 0, p > −2ζωn , and
Kωn2 z
2ζωn p + Kωn2 > .
p + 2ζωn
P6.14 The system has the roots

s1,2 = ±2j , s3,4 = ±2j , and s5,6 = −1 ± 2j ,

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Problems 255

Therefore, the system is not stable since there are repeated roots on the
jω-axis.
P6.15 (a) Neglecting the zeros and poles, we have the characteristic equation

s4 + 30s3 + 325s2 + 2500s + K = 0 .

The Routh array is

s4 1 325 K
s3 30 2500 0
s2 241.67 K
s1 b
so K

)
where

eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
604166.67 − 30K
an ing rnin tors igh

b= .
.
r
or ud a uc y

241.67
w cl le tr p
e in nt ns co

D
th k ( de f i es
of or stu e o tat

Therefore, the system is stable for 0 < K < 20139.

ity s w g us d S

is
te f t ss th nite

(b) Without neglecting the zeros and poles, the closed-loop characteristic
e rt ss fo U
gr hi in e
th a a ly by

equation is
k
in o e r
y y p d le d
ro n an o te
st f a s d s ec

s6 + 90s5 + 5525s4 + 12400s3 + (1255000 + K)s2

de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk

+ (8500000 + 30K)s + 1125K = 0 .

th nd wo
a his
T

This is stable for

0 < K < 61818 .

We see that the additional poles and zero makes the system stable
for a much larger gain K.
P6.16 (a) The Routh array is
s3 1 5
s2 5 6
s1 3.8
so 6
Examining the first column of the Routh array, we see no sign changes.
So, the system is stable.
(b) The roots of the system are s1 = −0.3246 and s2,3 = −2.3377 ±
3.6080j.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 256 CHAPTER 6 The Stability of Linear Feedback Systems

(c) The step response is shown in Figure P6.16.

Step Response
0.18

0.16

0.14

0.12
Amplitude

0.1

0.08

0.06

0.04

)
0.02

eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
0
an ing rnin tors igh

0 5 10 15

.
r
or ud a uc y
w cl le tr p

Time (sec )
e in nt ns co

D
th k ( de f i es
of or stu e o tat

FIGURE P6.16
ity s w g us d S

is
te f t ss th nite

Unit step response.

e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr

P6.17 The closed-loop transfer function is

w r sa co pro is
o eir is rk
th nd wo
a his

K +1
T

T (s) = .
s3 + 3s2 + 3s + K + 1

The Routh array is

s3 1 3
s2 3 K+1
8−K
s1 3

so K +1

So, for stability we require −1 < K < 8.

P6.18 The system characteristic equation is

s2 + (h − k)s + ab − kh = 0 .

For stability we require h > k and ab > kh. If k > h, the system is
unstable.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Problems 257

P6.19 (a) The characteristic equation is

s3 + 18s2 + (K − 40)s + 5K = 0 .

The Routh array is

s3 1 K − 40
s2 18 5K
13K−720
s1 18

so 5K
For stability

K > 720/13 .

)
eb
er or in ing
ed id n
W
(b) When K = 720/13, the system is marginally stable. The roots are
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r
or ud a uc y

s1,2 = ±j3.92 and s3 = −18.

w cl le tr p
e in nt ns co

D
th k ( de f i es
of or stu e o tat
ity s w g us d S

P6.20 The closed-loop characteristic equation is

is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by

q(s) = s5 + s4 + 9s3 + Ks2 + 2Ks + K .

in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr

The range of stability for the vertical-liftoff vehicle is

w r sa co pro is
o eir is rk
th nd wo
a his

5.177 < K < 7.823 .

T

Therefore, for K = 6, the system is stable. When K = 6 we have

q(s) = s5 + s4 + 9s3 + 6s2 + 12s + 6

The Routh array is

s5 1 9 12
s4 1 6 6
s3 3 6
s2 4 6
s1 1.5
so 6

All entries in the first column are positive, so the system is stable.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 258 CHAPTER 6 The Stability of Linear Feedback Systems

P6.21 The state transition matrix is

 
1 (k − p1 )e−p1 t − (k2 − p2 )e−p2 t e−p1 t − e−p2 t
Φ(t, 0) =  2 
p2 − p1 −k1 e −p 1 t + k1 e−p 2 t −p1 e−p 1 t + p2 e−p 2 t

where p1 p2 = k1 and p1 + p2 = k2 . We assume that p1 6= p2 . In the case

when p1 = p2 , the state transition matrix will change, but the factors e−p1 t
and e−p2 t will remain. The eigenvalues of A are given by the solution to

det |λI − A| = λ2 + k2 λ + k1 = 0 .
q
Therefore, the eigenvalues are λ1,2 = −k2 /2 ± k22 − 4k1 . If k2 > 0 and
k1 > 0, then the eigenvalues are in the left half-plane, and the system is
stable. The transfer function is given by
s−1
G(s) = C (sI − A)−1 B = − .

)
eb
er or in ing
s2 + k2 s + k1

ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

Therefore the characteristic equation is s2 + k2 s + k1 = 0 and the poles

.
r
or ud a uc y
w cl le tr p
e in nt ns co

q
are s1,2 = −k2 /2 ± k22 − 4k1 . If k2 > 0 and k1 > 0, then the poles are in
th k ( de f i es
of or stu e o tat

the left half-plane, and the system is stable. Notice that the values of λ1,2
ity s w g us d S

is
te f t ss th nite

and s1,2 are the same. Also, the eigenvalues are the same as the values of
e rt ss fo U
gr hi in e
th a a ly by

−p1 and −p2 . So, if the eigenvalues are negative, then the elements of the
k
in o e r
y y p d le d
ro n an o te

state transition matrix will decay exponentially.

st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
a his
T

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Advanced Problems 259

Advanced Problems
AP6.1 The Routh array is

s4 1 K1 K2
s3 20 4
20K1 −4
s2 20 K2
s1 b 0
so K2
where
20K1 − 4 − 100K2
b= .
5K1 − 1

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
For stability, we require K2 > 0, K1 > 0.2, and b > 0. Therefore, using
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

the condition that b > 0, we obtain

.
r
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es

K2 < 0.2K1 − 0.04 .

of or stu e o tat
ity s w g us d S

is
te f t ss th nite

The stability region is shown in Figure AP6.1.

e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec

0. 4
de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo

0.35
a his
T

0. 3

0.25
K2

0. 2

0.15

0. 1

0.05

STABLE REGION

0
0. 2 0. 4 0. 6 0. 8 1 1. 2 1. 4 1. 6 1. 8 2
K1

FIGURE AP6.1
Stability region.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 260 CHAPTER 6 The Stability of Linear Feedback Systems

AP6.2 The characteristic equation is

s4 + 14s3 + 85s2 + (K − 100)s + 2K = 0.

The Routh array is

s4 1 85 2K
s3 14 K − 100
1290−K
s2 14 2K
s1 b 0
so 2K
where
(1290 − K)(K − 100) − 392K

)
eb
b= .

er or in ing
ed id n
W
no the iss tea s

itt W tio
1290 − K w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r

Therefore, using the condition that b > 0, we obtain the stability range
or ud a uc y
w cl le tr p
e in nt ns co

152.59 < K < 845.41.

th k ( de f i es
of or stu e o tat
ity s w g us d S

AP6.3 (a) The steady-state tracking error to a step input is

is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by

ess = lim s(1 − T (s))R(s) = 1 − T (0) = 1 − α .

in o e r
y y p d le d
ro n an o te

s→0
st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is

We want
o eir is rk
th nd wo
a his

|1 − α| < 0.05 .
T

This yields the bounds for α

0.95 < α < 1.05 .

(b) The Routh array is

s3 1 α
s2 1+α 1
s1 b 0
so 1
where
α2 + α − 1
b= .
1+α
Therefore, using the condition that b > 0, we obtain the stability

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Advanced Problems 261

range for α:

α > 0.618 .

(c) Choosing α = 1 satisfies both the steady-state tracking requirement

and the stability requirement.
AP6.4 The closed-loop transfer function is

K
T (s) = .
s3 + (p + 1)s2 + ps + K

The Routh array is

s3 1 p
s2 1+p K

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
s1 b w 0
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

so
.
r

K
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es
of or stu e o tat

where
ity s w g us d S

is
te f t ss th nite

p2 + p − K
e rt ss fo U
gr hi in e
th a a ly by

b= .
k
in o e r
y y p d le d

1+p
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr

Therefore, using the condition that b > 0, we obtain the the relationship
w r sa co pro is
o eir is rk
th nd wo
a his

K < p2 + p .
T

The plot of K as a function of p is shown in Figure AP6.4.

AP6.5 The closed-loop transfer function is

30K1 K2
T (s) = .
(s + 1 + K1 K3 )(s − 10)(2s + K2 K3 − 4) + 30K1 K2 K4

The Routh array is

s3 2 a
s2 b c
s1 d 0
so c

where a = −9K2 K3 + 16 + K1 K2 K32 − 24K1 K3 , b = 2K1 K3 + K2 K3 − 22,

and c = −10K2 K3 + 40 − 10K1 K2 K32 + 40K1 K3 and d = (ab − 2c)/b .

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 262 CHAPTER 6 The Stability of Linear Feedback Systems

120

100

80

60
K

40
STABLE REGION

20

0
0 1 2 3 4 5 6 7 8 9 10

)
eb
p

er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

FIGURE AP6.4

.
r
or ud a uc y
w cl le tr p

Stability region.
e in nt ns co

D
th k ( de f i es
of or stu e o tat
ity s w g us d S

is
te f t ss th nite

The conditions for stability are

e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec

2K1 K3 + K2 K3 − 22 > 0
de o rse de ot
ill le u vi pr

−10K2 K3 + 40 − 10K1 K2 K32 + 40K1 K3 > 0

w r sa co pro is
o eir is rk
th nd wo

−2(−10K2 K3 + 40 − 10K1 K2 K32 + 40K1 K3 ) + (9K2 K3

a his
T

+16 + K1 K2 K32 − 24K1 K3 )(2K1 K3 + K2 K3 − 22) > 0

Valid values for the various gains are: K1 = 50, K2 = 30, K3 = 1, and
K4 = 0.3. The step response is shown in Figure AP6.5.
AP6.6 The characteristic equation is

s3 + 9s2 + (KD + 26)s + KP + 24 = 0.

For stability we require that KP > −24 and

KP − 210
KD > .
9
The relationship between KD and KP is shown in Figure AP6.6.
AP6.7 The characteristic equation is

0.1s4 + 2.05s3 + s2 + 8KP s + 8KI = 0.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Advanced Problems 263

Step Response

350

300

250

Amplitude
200

150

100

50

0
0 5 10 15 20 25 30
Time (sec)

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
FIGURE AP6.5
an ing rnin tors igh

Stability region.

.
r
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es
of or stu e o tat
ity s w g us d S

is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by

40
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
ill ale ou ovi pr

30
s c r is
or eir is p rk
th nd wo

STABLE
a his

20
T

10
Kw
D

UNSTABLE
−10

−20

0 50 100 150 200 250 300 350 400 450 500

K
P

FIGURE AP6.6
Stability region.

From the Routh array we find the conditions for stability are

0 < KI < 0.3125

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 264 CHAPTER 6 The Stability of Linear Feedback Systems

p p
1.2812 − 1.6416 − 5.2531KI < KP < 1.2812 + 1.6416 − 5.2531KI

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es
of or stu e o tat
ity s w g us d S

is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
a his
T

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Design Problems 265

Design Problems
CDP6.1 The plant model with parameters given in Table CDP2.1 in Dorf and
Bishop is given by:
θ(s) 26.035
= .
Va (s) s(s + 33.142)
In the above transfer function we have neglected the motor inductance
Lm . The closed-loop transfer function from the input to the output is
θ(s) 26.035Ka
= 2 .
R(s) s + 33.142s + 26.035Ka
The Routh array is

s2 1 26.035Ka

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
s1
d on g. in t la

m ld a
33.142 0
an ing rnin tors igh

.
r
or ud a uc y

s0
w cl le tr p

26.035Ka
e in nt ns co

D
th k ( de f i es
of or stu e o tat
ity s w g us d S

Stability is achieved for any 0 < Ka < ∞.

is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by

k
in o e r

The closed-loop characteristic polynomial is

y y p d le d

DP6.1
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr

1 1 1
s3 + s2 (5 + p + K) + s( Kp + K + 5p) + K = 0 .
w r sa co pro is
o eir is rk

5 5 5
th nd wo
a his
T

(i) When p = 2, we have

1 3
s3 + s2 (7 + K) + s(10 + K) + K = 0 .
5 5
The Routh array is
s3 1 10 + 35 K
K
s2 7+ 5 K
s1 b
so K
where
(7 + K/5)(10 + 3K/5) − K
b= .
7 + 15 K

When −32.98 < K < −17.69, we find that b > 0. Examining the
other terms in the first column of the array, we find that the system

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 266 CHAPTER 6 The Stability of Linear Feedback Systems

is stable for any K > 0.

(ii) When p = 0, we have
1 1
s3 + s2 (5 + K) + s( K) + K = 0 .
5 5
The Routh array is
1
s3 1 5K

s2 5 + 15 K K
s1 b
so K
where
(5 + 51 K) 15 K − K K 2 /25
b= = .

)
eb
er or in ing
(5 + K/5) (5 + K/5)

ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

Again, examination of the first column reveals that any K > 0 results

.
r
or ud a uc y
w cl le tr p
e in nt ns co

in a stable system. So, we just need to select any K > 0; e.g. K = 10.
th k ( de f i es
of or stu e o tat

(a) The closed-loop characteristic equation is

ity s w g us d S

DP6.2
is
te f t ss th nite
e rt ss fo U
gr hi in e

20(Ks + 1)
th a a ly by

1+ =0,
in o e r
y y p d le d
ro n an o te

s2 (s + 20)
st f a s d s ec
de o rse de ot
ill le u vi pr

or
w r sa co pro is
o eir is rk
th nd wo

s3 + 20s2 + 20Ks + 20 = 0 .
a his
T

The Routh array is

s3 1 20K
s2 20 20
s1 b
so 1
where
20K − 1
b= .
1
For stability, we require K > 0.05.
(b) The desired characteristic polynomial is

(s2 + as + b)(s + 5) = s3 + s2 (a + 5) + s(5a + b) + 5b = 0 .

Equating coefficients with the actual characteristic equation we can

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Design Problems 267

solve for a, b and K, yielding b = 4, a = 15, and

5a + b 79
K= = .
20 20
(c) The remaining two poles are s1 = −14.73 and s2 = −0.27.
(d) The step response is shown in Figure DP6.2.

0.9

0.8

0.7

0.6

)
eb
er or in ing
ed id n
W
no the iss tea s
y(t)

itt W tio
0.5 w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

0.4

.
r
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es

0.3
of or stu e o tat
ity s w g us d S

0.2
is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by

0.1
in o e r
y y p d le d
ro n an o te
st f a s d s ec

0
de o rse de ot

0 2 4 6 8 10 12 14 16 18 20
ill le u vi pr

time (sec)
w r sa co pro is
o eir is rk
th nd wo
a his

FIGURE DP6.2
T

Mars guided vehicle step response.

DP6.3 (a) The closed-loop characteristic equation is

2τ s3 + (τ + 2)s2 + (K + 1)s + 2K = 0 .

The Routh array is

s3 2τ K +1
s2 τ +2 2K
s1 b
so 2K
where
(τ + 2)(K + 1) − 4Kτ
b= .
(τ + 2)

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 268 CHAPTER 6 The Stability of Linear Feedback Systems

Examining the first column of the Routh array, we determine that for
stability τ > 0, K > 0 and setting b > 0 yields the relationships:
τ +2 2 2
(1) K < when τ > (2) K > 0 when 0 < τ ≤ .
3τ − 2 3 3
The plot of τ versus K is shown in Figure DP6.3a.

4.5

3.5

3
tau

)
eb
er or in ing
2.5

ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r
or ud a uc y
w cl le tr p
e in nt ns co

1.5
th k ( de f i es
of or stu e o tat
ity s w g us d S

1
is
te f t ss th nite

STABLE REGION
e rt ss fo U
gr hi in e

0.5
th a a ly by

0 1 2 3 4 5 6 7
in o e r
y y p d le d
ro n an o te
st f a s d s ec

K
de o rse de ot
ill le u vi pr
w r sa co pro is
o ir s rk

FIGURE DP6.3
th nd wo

(a) The plot of τ versus K.

a his
i
T

(b) The steady-state error is

A
ess = , where Kv = 2K .
Kv
So,
ess 1
= .
A 2K
We require that ess ≤ 0.25A, therefore

K≥2.

One solution is to select τ = 0.5, then we remain within the stable

region.
(c) The step response is shown in Figure DP6.3b. The percent overshoot
is P.O. = 57%.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Design Problems 269

1.6
P.O. = 56.77 %

1.4

1.2

y(t) 0.8

0.6

0.4

0.2

0
0 2 4 6 8 10 12 14 16 18 20

)
eb
time (sec)

er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

FIGURE DP6.3

.
r
or ud a uc y
w cl le tr p

CONTINUED: (b) Closed-loop system step response.

e in nt ns co

D
th k ( de f i es
of or stu e o tat
ity s w g us d S

is
te f t ss th nite

DP6.4 (a) The closed-loop characteristic polynomial is

e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec

s3 + Ks2 + [(2 + m)K − 1]s + 2mK = 0 .

de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo

The Routh array is

a his
T

s3 1 2K + mK − 1
s2 K 2mK
s1 b
so 2mK
Examining the first column of the Routh array, we see that for sta-
bility we require m > 0, K > 0, and b > 0, where

(2K + mK − 1)K − 2mK

b= = (2 + m)K − (1 + 2m) > 0 ,
K
or

1 + 2m
K> .
2+m

The plot of K vs m is shown in Figure DP6.4a.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 270 CHAPTER 6 The Stability of Linear Feedback Systems

1.6

1.4
STABLE REGION

1.2

1
K

0.8

0.6

0.4
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

)
eb
m

er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

FIGURE DP6.4

.
r
or ud a uc y
w cl le tr p

(a) The plot of K versus m.

e in nt ns co

D
th k ( de f i es
of or stu e o tat
ity s w g us d S

is
te f t ss th nite

1.8
e rt ss fo U
gr hi in e
th a a ly by

P.O. = 64.3208 %
in o e r
y y p d le d
ro n an o te

1.6
st f a s d s ec
de o rse de ot
ill le u vi pr
w r sa co pro is

1.4
o eir is rk
th nd wo
a his

1.2
T

1
y(t)

0.8

0.6

0.4

0.2

0
0 1 2 3 4 5 6 7 8 9 10
time (sec)

FIGURE DP6.4
CONTINUED: (b) Shuttle attitude control step response.

(b) The steady-state error is

ess 1 1
= = < 0.10 ,
A Kv 2mK

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Design Problems 271

or mK > 5. For example, we can select m = 0.5 and K = 2.

(c) See Figure DP6.4b for the step response where P.O. = 64.3%.
DP6.5 The closed-loop transfer function is

K
T (s) = .
s3 + 10s2 + 20s + K

The range of K for stability is 0 < K < 200. If we let K = Km /N where

Km = 200, then N = 6.25 results in a step response with P.O. = 15.7%
and Ts = 1.96 seconds.
DP6.6 The closed-loop system is given by
   
0 1 0
ẋ =  x +  r

)
2 − K1 −2 − K2 1

eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

The characteristic polynomial is s2 +(2+K2 )s+K1 −2h = 0. Soithe system

.
r
or ud a uc y
w cl le tr p
e in nt ns co

D
th k ( de f i es

is stable for K1 > 2 and K2 > −2. Selecting K = 10 2 results in

of or stu e o tat
ity s w g us d S

closed-loop eigenvalues at s = −2 ± 2j. The closed-loop step response has

is
te f t ss th nite

a settling time of 2.11 s and a percent overshoot of 4.32%.

e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot
ill le u vi pr

Im(s)
w r sa co pro is
o eir is rk
th nd wo
a his
T

sin-1 ζ = sin-1 0.69=43.63ο

Re(s)

desired
region
for eigenvalues

ζωn = -1

FIGURE DP6.6
Performance region.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 272 CHAPTER 6 The Stability of Linear Feedback Systems

DP6.7 (a) The inner loop closed-loop transfer function is

Y (s) 20s
= 3 2
.
U (s) s + 10s + 20s + 20K1
The Routh array is
s3 1 20
s2 ω 20K1
200−20K1
s1 10

so 20K1
For stability 0 < K1 < 10.
(b) The fastest response (that is, the quickest settling time) occurs when
K1 = 2.2

)
eb
(c) With K1 = 2.2, the closed-loop transfer function is

er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

Y (s) 20K2 s

.
r

= 3 .
or ud a uc y
w cl le tr p
e in nt ns co

R(s) s + 10s2 + (20 + 20K2 )s + 44

th k ( de f i es
of or stu e o tat
ity s w g us d S

The Routh array is

is
te f t ss th nite
e rt ss fo U

s3
gr hi in e

1 20(K2 + 1)
th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec

s2 10 44
de o rse de ot
ill le u vi pr
w r sa co pro is

200K2 +156
s1
o eir is rk
th nd wo

10
a his

so 44
T

For stability, we require

200K2 + 156 > 0 .

Therefore, K2 > −0.78.

DP6.8 The closed-loop characteristic equation is

s2 + 4KD s + 4(KP + 1) = 0.

So, it is possible to find KP and KD to stabilize the system. For example,

any KP > 0 and KD > 0 leads to stability. Choosing KP ≥ 9 results in a
steady-state tracking error
√ less than 0.1 due to a unit step input. Then,
the damping ratio ζ = 2/2 is achieved by selecting
√ √
2 KP + 1
KD = .
2

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Computer Problems 273

Computer Problems
CP6.1 The m-file script is shown in Figure CP6.1.

ans =

-0.7400 + 4.3224i
-0.7400 - 4.3224i
-0.5200 + 0.0000i

ans =
pa=[1 2 20 10]; roots(pa)
pb=[1 8 24 32 16]; roots(pb) -2.0004 + 0.0000i
pc=[1 0 2 1]; roots(pc) -2.0000 + 0.0004i
-2.0000 - 0.0004i
-1.9996 + 0.0000i

ans =

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la
0.2267 + 1.4677i

m ld a
an ing rnin tors igh

0.2267 - 1.4677i

.
r
or ud a uc y
w cl le tr p

-0.4534 + 0.0000i
e in nt ns co

D
th k ( de f i es
of or stu e o tat

FIGURE CP6.1
ity s w g us d S

is
te f t ss th nite

Computing the polynomial roots with the roots function.

e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec

CP6.2 The m-file script is shown in Figure CP6.2.

de o rse de ot
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
a his

K1=1;K2=2;K3=5; den=[1 4 2];

T

num1=K1*[1 -2 4];num2=K2*[1 -2 4];num3=K3*[1 -2 4];

sys1 = tf(num1,den); sys2 = tf(num2,den); sys3 = tf(num3,den);
sys1_cl=feedback(sys1,[1]);
sys2_cl=feedback(sys2,[1]);
sys3_cl=feedback(sys3,[1]);
p1 = pole(sys1_cl), p2 = pole(sys2_cl), p3 = pole(sys3_cl)

p1 = p3 =

-0.5000 + 1.6583i 0.5000 + 1.8484i

-0.5000 - 1.6583i p2 =
0.5000 - 1.8484i
0.0000 + 1.8257i
0.0000 - 1.8257i

FIGURE CP6.2
K = 1 is stable;K = 2 is marginally stable; and K = 5 is unstable.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 274 CHAPTER 6 The Stability of Linear Feedback Systems

CP6.3 The closed-loop transfer function and the roots of the characteristic equa-
tion are shown in Figure CP6.3.

Transfer function:
numg=[1 1]; deng=[1 4 6 10]; s+1
sysg = tf(numg,deng); ----------------------
sys = feedback(sysg,[1]) s^3 + 4 s^2 + 7 s + 11
r=pole(sys)
r=

-2.8946
-0.5527 + 1.8694i
-0.5527 - 1.8694i

FIGURE CP6.3
Closed-loop transfer function and roots.

)
eb
er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
CP6.4 There are no poles in the right half-plane, but the system is unstable
an ing rnin tors igh

.
r
or ud a uc y

since there are multiple poles on the jω-axis at s = ±j and s = ±j (see

w cl le tr p
e in nt ns co

D
th k ( de f i es

Figure CP6.4).
of or stu e o tat
ity s w g us d S

is
te f t ss th nite
e rt ss fo U
gr hi in e

Step Response
th a a ly by

From: U(1)
in o e r
y y p nd le d
tro an a so cte

25
es of es ed ote
ill ale ou ovi pr

20
is
d
o ei is rk
th nd wo

num=[1]; den=[1 2 2 4 1 2];

Amplitude w r s r c pr

15
s
a his

sys = tf(num,den);
T

pole(sys) 10

t = 0:0.1:100;
d

step(sys,t) 5
To: Y(1)

ans = 0

-5
-2.0000
0.0000 + 1.0000i -10
0.0000 - 1.0000i
-0.0000 + 1.0000i -15
-0.0000 - 1.0000i
-20

-25
0 10 20 30 40 50 60 70 80 90 100

Time (sec.)

FIGURE CP6.4
Unstable system step response.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Computer Problems 275

CP6.5 The closed-loop system poles for the slow/fast pilots are shown in Fig-
ure CP6.5. The maximum allowable time delay is 0.2045seconds. At the
maximum allowable time delay, the system has roots on the jω-axis at
s = ±2.6j. The slow pilot destabilizes the aircraft.

nume=[-10]; dene=[1 10]; syse = tf(nume,dene);

numg=[-1 -6]; deng=[1 3 6 0]; sysg = tf(numg,deng);
%
% Fast pilot
%
tau=0.1; tau1=2; K=1; tau2=0.5;
nump=-K*[tau1*tau tau-2*tau1 -2]; closed-loop
denp=[tau2*tau tau+2*tau2 2]; system poles
sysp = tf(nump,denp);
fast_pilot =
sysa = series(sysp,syse);
sysb = series(sysa, sysg);
-19.6267
sys = feedback(sysb,[1]);

)
eb
-10.7712

er or in ing
ed id n
W
no the iss tea s
fast_pilot=pole(sys)

itt W tio
w
t p W em ch
-3.8885

e
d on g. in t la

m ld a
%
an ing rnin tors igh

-0.1697 + 2.7880i
% Slow pilot

.
r
or ud a uc y

-0.1697 - 2.7880i
w cl le tr p
e in nt ns co

%
-0.3742
th k ( de f i es

tau=0.6; tau1=2; K=1; tau2=0.5;

of or stu e o tat
ity s w g us d S

nump=-K*[tau1*tau tau-2*tau1 -2];

is
te f t ss th nite

denp=[tau2*tau tau+2*tau2 2];

slow_pilot =
e rt ss fo U
gr hi in e

sysp = tf(nump,denp);
th a a ly by

k
in o e r
y y p d le d

sysa = series(sysp,syse);
ro n an o te

-9.4526
st f a s d s ec

sysb = series(sysa, sysg);

de o rse de ot

-4.5228 + 2.2595i
ill le u vi pr

sys = feedback(sysb,[1]);
w r sa co pro is

-4.5228 - 2.2595i
slow_pilot = pole(sys)
o eir is rk
th nd wo

0.2793 + 2.0314i
%
a his

0.2793 - 2.0314i
% Maximum pilot time delay, tau = 0.2045 sec
T

-0.3937
%
tau=0.2045; tau1=2; K=1; tau2=0.5;
nump=-K*[tau1*tau tau-2*tau1 -2];
max_pilot_delay =
denp=[tau2*tau tau+2*tau2 2];
sysp = tf(nump,denp);
-10.0433 + 2.2684i
sysa = series(sysp,syse);
-10.0433 - 2.2684i
sysb = series(sysa, sysg);
-4.3153
sys = feedback(sysb,[1]);
0.0001 + 2.6040i
max_pilot_delay=pole(sys)
0.0001 - 2.6040i
-0.3783

FIGURE CP6.5
Closed-loop system poles for an aircraft with a pilot in-the-loop.

CP6.6 The closed-loop transfer function is

1
T (s) = .
s3 + 5s2 + (K − 3)s + K + 1

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 276 CHAPTER 6 The Stability of Linear Feedback Systems

Utilizing the Routh-Hurwitz approach, for stability we determine that

K>4.

When K = 4, the roots of the characteristic equation are

s1 = −5 and s2,3 = ±j .

The m-file script which generates a plot of the roots of the characteristic
equation as a function of K is shown in Figure CP6.6.

K=[0:0.1:5];
n=length(K);
for i=1:n
numg=[1]; deng=[1 5 K(i)-3 K(i)];
sys_o = tf(numg,deng);

)
eb
sys_cl = feedback(sys_o,[1]);

er or in ing
ed id n
W
no the iss tea s

itt W tio
w
t p W em ch
p(:,i)=pole(sys_cl);

e
d on g. in t la

m ld a
an ing rnin tors igh

end

.
r
or ud a uc y

plot(real(p),imag(p),'x'), grid
w cl le tr p
e in nt ns co

text(-0.9,0.95,'K=4 -->'); text(-0.2,1.3,'K=5'); text(0,0.2,'K=0')

th k ( de f i es
of or stu e o tat

% From a Routh-Hurwitz analysis we find that

ity s w g us d S

% minimum K for stability is K=4

is
te f t ss th nite

Kmax=4;
e rt ss fo U
gr hi in e
th a a ly by

numg=[1]; deng=[1 5 Kmax-3 Kmax]; sysg = tf(numg,deng);

in o e r
y y p d le d
ro n an o te

sys_cl = feedback(sysg,[1]); pole(sys_cl)

st f a s d s ec
de o rse de ot
ill le u vi pr

1.5
w r sa co pro is
o eir is rk
th nd wo

K=5
a his
T

1 K=4 -->

0.5

K=0

-0.5

-1

-1.5
-6 -5 -4 -3 -2 -1 0 1

FIGURE CP6.6
Roots of the characteristic equation as a function of K, where 0 < K < 5.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Computer Problems 277

CP6.7 The characteristic equation is

p(s) = s3 + 10s2 + 14s + 12 .

A=[0 1 0;0 0 1;-12 -14 -10]; b=[0;0;12]; c=[1 1 0]; d=[0];

sys = ss(A,b,c,d);
%
% Part (a)
%
p=poly(A) p=
%
% Part (b) 1.0000 10.0000 14.0000 12.0000
%
roots(p) ans =
%
% Part (c) -8.5225
% -0.7387 + 0.9286i

)
eb
er or in ing
step(sys) -0.7387 - 0.9286i

ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r
or ud a uc y

Step Response
w cl le tr p
e in nt ns co

D
th k ( de f i es

1.4
of or stu e o tat
ity s w g us d S

is
te f t ss th nite
e rt ss fo U

1.2
gr hi in e
th a a ly by

k
in o e r
y y p d le d
ro n an o te
st f a s d s ec
de o rse de ot

1
ill le u vi pr
w r sa co pro is
o eir is rk
th nd wo
Amplitude

a his

0.8
T

0.6

0.4

0.2

0
0 1 2 3 4 5 6 7 8 9
Time (sec)

FIGURE CP6.7
Characteristic equation from the state-space representation using the poly function.

The roots of the characteristic equation are

s1 = −8.5225 and s2,3 = −0.7387 ± 0.9286j .

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 278 CHAPTER 6 The Stability of Linear Feedback Systems

The system is stable since all roots of the characteristic equation are in
the left half-plane. The unit step response and associated m-file script are
shown in Figure CP6.7.
CP6.8 The characteristic equation is

s3 + 10s2 + 10s + 5K1 = 0 .

(a) The Routh array is

s3 1 10
s2 10 5K1
100−5K1
s1 10

so 5K1
From the Routh-Hurwitz criterion, we obtain the limits 0 < K1 < 20

)
eb
er or in ing
for stability.

ed id n
W
no the iss tea s

itt W tio
w
t p W em ch

e
d on g. in t la

m ld a
(b) The plot of the pole locations is 0 < K1 < 30 is shown in Fig-
an ing rnin tors igh

.
r
or ud a uc y
w cl le tr p

ure CP6.8. As seen in Figure CP6.8, when K1 > 20, the pole locations
e in nt ns co

D
th k ( de f i es

move into the right half-plane.

of or stu e o tat
ity s w g us d S

is
te f t ss th nite
e rt ss fo U
gr hi in e
th a a ly by

Root Locus
in o e r
y y p d le d
ro n an o te
st f a s d s ec

4
de o rse de ot
ill le u vi pr
w r sa co pro is

3
o eir is rk
th d wo
Imaginary Axi s an his

2
T

1 k=20
0

?-1

?-2

?-3

?-4
?-12 ?-10 ?-8 ?-6 ?-4 ?-2 0 2
Real Axi s

FIGURE CP6.8
Pole locations for 0 < K1 < 30.

CP6.9 (a) The characteristic equation is

s3 + 2s2 + s + k − 4 = 0 .

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.
 Computer Problems 279

The Routh array is

s3 1 1
s2 2 k−4
6−k
s1 2

so k−4
For stability, we obtain 4 < k < 6.
(b) The pole locations for 0 < k < 10 are shown in Figure CP6.9. We see
that for 0 < k < 4 the system is unstable. Similarly, for 6 < k < 10,
the system is unstable.

Root Locus
2

)
k=10

eb
pole locations when k=0

er or in ing
ed id n
W
no the iss tea s

itt W tio
1.5 w
t p W em ch

e
d on g. in t la

m ld a
an ing rnin tors igh

.
r

1
or ud a uc y
w cl le tr p

pole location when k=0

e in nt ns co

D
th k ( de f i es
Imaginary Axi s

0.5
of or stu e o tat

k=10 increasing k
ity s w g us d S

is
te f t ss th nite

0
e rt ss fo U
gr hi in e
th a a ly by

k
in o e r
y y p d le d

?-0.5
ro n an o te

k=4
gk
st f a s d s ec

k=6
de o rse de ot

sin
ill le u vi pr

?-1
rea
w r sa co pro is
o eir is rk

inc
th nd wo

?-1.5
a his
T

?-2
?-3 ?-2 ?-1 0 1
Real Axi s

FIGURE CP6.9
Pole locations for 0 < k < 10.

© 2017 Pearson Education, Inc., Hoboken, NJ. All rights reserved. This material is protected under all copyright laws as they currently
exist. No portion of this material may be reproduced, in any form or by any means, without permission in writing from the publisher.