UNI-T.

~
DESIGN OF D -ISCRETE - TIME CONTROL
SYSTEMS Bv CoNVENTIONAL·· ·METHODS

Syllabus
DESIGN.OF DISCRETE-TIME CONTROL SYSTEMS BY CONVENTIONAL METHODS: Transient and steady state specifications
- Design using frequency _response in thew-plane for lag·and lead compensators.- Root locus t~que in the z-plane.

(LEARNING OBJECTIVES)

r:::? Transient and steady st~te specifications

r:::? Design of Lag and-Lead co'!1pensators

r:::? Root locus technique in z-plane

r ___
.___ INTRCDUC.T ICN J
The process of _altering the design procedure of a system to meet desired specification is known as compensation. However,

this is achieved by introducing <;:1n additional d~vic~ called compensator. Tt\e various_types of compensators employed are lag
. ,
compensator, Lead compensator and lead-lag compensator. Compensation Involve system redesign by modification of the structure

~r by employing suitable devices so as to meet the performance speciflcations. This is a process which compensates for the deficient

performance of the original system. Modification of the system can be done in any ·form.

• l-1 ,,

· '. SPECTRUM ALL-IN-ONE JOURNAL FOR ENGINEERING STUDENTS

 196 DIGITAL CONTROL SYSTEMS [JNTU-KAKINAOA)
I
!I
I 11,
i'
I
i~f; •1-·RAN,SIINT', AMD -.......,~.-..,,., s,,,.. (Ill) Peak Time (tP)
It is.defined as the time taken to reach the peak value for
I \;;;Spi~lflCATIO
' . ',1'11\r NS , ' '' ', ' ' I,:' : '
r,: :, , the first tim~.
The expression for the peak time is,
Q1. L.ist,out the transient response specifications.
7t 7t .
----~==
OR
Sep.•20, Sot-1 , Q1(e) t
_P - (J)d - Wn g
·, List out the transient response specifications (iv) Maximum (or) Peak Overshoot
-;~? .and explain In brief. , Aprll-18, Set-2,ae(b)
. ·,}-;'.' !\: Q~ , . I ',,; , It is defined as the amount by which response of the
·1 ·1,\ ·, j !· :.'
system overshoots from reference steady value '(final value) for ,
' ,,
the first time.
Define transient response specifications,
. ' . . '

Let, peak (maximum) value= C(t)

(i) .· Delay time (td)
-• -~: : . (ii) Rise time (t) ,;, . Final value = C(00 )
, ;·.... . .. r. . . i :,· • .. ·-~
Then peak overshoot,
(Iii) , Peak time (t ) . , . . t . · :'..
(ivf: Ma·ximurri (cir)'peafc,\jifsh~-of':'. (:>\ C(tp)-C(oo)
' • , ', • r '•
MP= C(oo)
(v):-. Settling-time of second
·. . ,.,.
order system.
. . . . ,,.
, ",, ', , , , ; , .,, ·.:Aprli/May-16, Set-4, Q7(a) Percentage peak overshoot (%MP)
· :- •"; ·:·... v~.. -.· _-;,:_-~i :\ ~:~~~-<- /.. -.. _._ .
= [C(tp)-C(oo)]xlOO
•
1
. . ·::- .. 111·e transient'
,,Write . response specifications.
~

ns: · ,:: ·.,··

.
C(oo)
Aprll/May-14, Set-1, Q7(a)
(i)' De!ny .Time (t11).· , . · . . ·... . . The expression for peak overshoot is,
... ,f.: ; __ .·' . . •.. ,,

'./:> ': ltis,defmed as the time taken for the response to reach -II~

·o~~~r
sih1: dfthe constipt J~ady stat~ ~alu~ (final value), for the firs
! . \ . ~ : { , ,.. \}~)t /.):;(-~~-~--?t~):~~<>. ~.:~:~: ;;-.··· -~,: ·
1 · Mp=e°R

The expression for delay time is, (v) Settling Time of Second Order System

\td =~I It is defined as the time taken for the response to reach
and stay within specified percentage range of its final value
(within tolerable limits).
Where,
The expression for settling time (t5) is,
. s = Damping ratio 4
ffi,.= Undamped natural frequency. ts =;;---for 2% error
~(J)n
(ii) Rise Time (t,)
3
. It is qefined as the time taken for the response raise from ts =-r- for 5% error
':, (l)n
0 to 100% of the final value for underdamped systems.
It is defined as the time taken for the response to raise Q2. Derive the expressions for rise time, peak time,
10 to 90% offinal value for ov~rdamped systems. peak overshoot and settling time of a second
It is defined as the time taken for the response to raise order system subjected to a step input.
from 5 to 95% of final value for critically damped system. Ans: . Model Paper-I, .09
The expression for the rise time is, . (i) Rise Time

~
It is defined as the time taken by the response to reach
I00% of its final steady state value.
Let,
Where, t, = Rise time
0 = Damping angle in radians (!)" = Undamped naturat frequency
Damped frequency of oscillati~n
ffid =
e = tan- 1 .J1-s 2
t, = Setting time
. s' s
= Damping r~tio ,
ffid = Damped frequency of oscillations, %mP = Percentage overshoot
w=wR. d II
C(t) = Output of the system
tP= Peak time.

WARNING: Xerox/Pho~ocopying of this book is a CRIMINAL act. Anyone found guilty is LIABLE to face LEGAL proceedings.J ·. : ,
 ~
q
- lJNIT-6 (Design of DilCfete-Time Control Systems By Conventional Methods)
-- The output of a system subj ected to unit step inp~t for undcrdomped bond ition is given by.
e -c,n,,
197

>r
C(t) =I- ~ sin(WJ + ♦) ... ( I )
vl - ~2

We know that at time ' "" ',· lhc rcs r o n~c is cq11nl to its illcncly Alnto vo luc i. e ., I .
= C(oo) ,,. I
Hence at r = I,, C( r)
Substituting the oh\wc vnlucs in cquntion (I ), we get,
(' - (ii,., , .
I "" 1 - ~ s m(w; + ♦)
e "1 - ~2 ,
1r
e -(p, ,,,
- ~ sin(w; + ♦) = O , .. (2)
vl - ~2 ,

~ that the
Ln '-''""' · wJ, + , ) must be zero. B ut we k now th at - e -r:.p_,. ,./
· · abo ve equa 1·100 (2) to be zero, e1"th er - e - C,,, •' ' or sm(
ri-:-~1
-

,-.,,nnoe be zero. Therefore, sin(w;, + ♦) must be zero .

i.e.. sin(O>/, + ♦) = 0 ... (J)
We know that sine is zero at angles equal to multiples of n .
Therefore, equation (3) becomes,
ro;, + ♦ = n
ro;, = rt - ♦
,=-
,
-
rod
Jt- ♦

Where, ♦ = cos-• ~
(H) Peak nme
It is defi~ as the time taken by the response to reach the first peak for the very first time .
From equation (I ), we have ,
e-o,,.,
e ,.--;i sin(WJ + ♦)
C(I) = I -
vl - ~2
Since. the ~sponse is maximum at,,,· Therefore, in order to determine,,,. differentiating C(t) with rc~pcct to t and equa ting
it to zero.

We know that.

j
(J)d = (J),. J1 - ~2

=
-e-v»~
, (~0>) sin(W/ + ♦) -
w,.H
~ e
-(;;>, •
• cos(_c.d j + ♦)
9 R y l -~- .

h r~ sin(w; + ♦) - R cO';/..Wj + ~)] ... ( 4)

From fl.• shown in fi gure,

/1 . •'

SPECTRUM All-Uf-OIE JOURMAL FOR ENGINEERING STUDENTS

 198 DIGI TAL CON TRO L SYS TEM S [JNT U-KA KINA
DA)
cos <1>= r,
sin< l>= H
Subst itutin g the aboye values in eq uation ( 4). we
dC(t)
get,
.
[cos <I> s in(W./ + <!>) - s in <I> cos(W J + <!>)] . .. (5)
dt

We know that,
,
sin(A :... B) = sin A cos B - cos A sin B
Using the above eq{iation, equati~n (5) can be writte
n as.
dC(t) w e-C,w,, r
" s in(w; + <l> - <I>)
dt
R - '

(J) e-C,w,,r
~ sinw. ;
vl - t,2
Subst itutin g t = t · in above equat ion and equat ing
p .
it to zero, we get,
Cl) e -C,w,rP .

I I " ~ smw ; = O
vi-r. ,- ,,
I·\ p,:I '
e
-r,w;," sin Wj = 0
P
1
In order that above equat ion is zero. either e-c;w,,,
·ifjlj\ must be zero i.e.,
,, or sin CJJJ,, must be zero . We know e - r.,w.," can't be zero
. Henc e, sin WJP

sin WJP = 0

'' The value of sine is zero at all the multiples of n i.e

.. nn. Takin g the small est possib le value i.e., n(n
of the above equat ion, we have, = l) to get the solution
·

7t
t = -sec s.
P (J)d .

(iii) Peak Overshoot

. It is define d as the amou nt l>y which the respo nse
first oversh oot.
of the system overshoots from reference steady state value during the
. C(tp) -C(oo )
%m = - ~ - - - x lOO
P C(oo)
... (6)
Wher e, C(t) is the respo nse of the system at peak
of the system att = oo. time i.e ., at r = tP and C(?>) is the steady. state respo nse i.e. ; the response
. •
.
From equat ion (I),

e-C,w,,r .
C(t)= 1- ~sm (w; + <I>)
vi .- s2 ,
Subst itutin g,

. -r,w __n_ _
7t 7t e -t;w,,r,, . " w, ✓ 1 - r,2
t=t = -(.t) = I- ~ s in(Wj , + <I>) = I - e
p
. d v1-s 2 , p ·
- r;i/ J1- ,:,2 .
C(t) = I - e
p
H (;;2
1-
sin(rt + (j))

WARNING: Xerox/Photocopying of this·book is a CRIMINAL act. Anyone found

guilty is LIABLE to face LEGAL proceedings.
 Conve ntiona l Metho ds) 199
l)NIT- 5 (Design of Discre te - lime Contro l Syste ms By
But, sin(1t + <I>) - - sin <I> Q4. What Is chara cteris tics equat ion? Expla in
its

e
-r.,c/J1-~2 . signif icanc e.
C(t ) =I+ ~ sm<j)
I ~·t;2 .vi Ans:

From figure, Characteristic Equation

sin< j)=H Characteristic equati on· is an algebr aic equati on that

gives insight about the inherent nature of a linear tim«: invariant
Substituting the value of sin q> in above equation, we system.
have. The general form of transfer function ofa system is given
-r.,c/J1-~2 ~
e by,
=l + ~ Xvl- l;-
\fl- ~2 0 n n-1 +a z n-2 + ... .+anz0
a z +a 1z 2 . ... (I)
C(t) = 1 + e -r.;c/ R T(z)= b z m .+ b z m- 1 + b z m-2 + ....+ b mz 0
... (7) 0 1 2

C(oo) = I ... <sr Equation (I) reveal s that transfe r functio n is a ratio
of numerator polynomial and denom jnator polyno mial. The
[ ·: Input is a unit step signal]
denominator polynomial of the transfer function is very significant
Substi tuting equa6 ons (7) and (8) in equati on (6), and when equated to zero, gives the charac teristic equati on.
we get,
I -r.,c/ J1,-~2 The closed loop transfe r fup.ction of a system with
. +e
%mp = - - - -1- - - x 100
-1 forward path gives G(z) and feedba ck path gain H(z) is given
as,
= e -r.;c/✓i -~ x 100
2
%m · · G(z)
p ... (2)
T(z) = 1.± G(z)H (z)
(iv) Settling Time
It is defined as the time taken by the response to reach From equation (2), the characteristic equati on in terms
the final stay state value and stay within the tolerable limits. of loop tran~fer function is given as,
4 I ± G(z) H(z) = 0 ... (3)
t = - r - for 2% error
s ~con
The roots of the cnaraoteristic equati ons are called a_
3
t = -r- for 5% error oles whi~hplay a very iniportanfrole in the stability assess men
s ~con fthe system. · ·· · : :: ·•.•.., .'·:' ..·
. '·, ·

For any value of error, t, is given by,

ln(¾e nor) ·
Significance of Cbaracteristic Equation
' '.
-. . /
•.· . .
.,~,. ·-.. ,.•
I=
s ~con l. Using characteristic -~quation, we can dclem1ine the typ
and order bf ~syste m. ;,
. ;, :. .
Q3. Sketch the trans ient respo nse of a secon,d order ·
system and deriv e the expre ssion for rise time ., The roots of the cbara<;teristic eq\lati on gives the pole
and peak overs hoot. of system .. ·. :.;."''
the ·.•,
'. ' .,
·'
.•

. : '. '. . ..
Ans: 3. From the:kn owled ge of locatio n of poles. the stabili ty
of the systemcan ~e d':term ined.

The type of damping of the system can be detennine~

usiog ·c haract eristi.c equation.

5. Using chara¢ tei·is~ tJ:. _equatio1i, the outpu t respon se

e determ ined .
..., se)
(transient n:spon .,_.''cau:b·
.

0 '· ~ 1., 6. Using ~hara~t~~f~ii~-:~\1a1fon, the variou s specification.'

of the systen i'i.e.;~i:lorhain and freque ncy doma i
Figure: Transient Response of I Second Order System
specification can be determ ined.
The above figure showi ng transie nt respon se of second
' .· , ' .I
J

0rd By modil ying·thc,~ba~ncteriiltic equati on tad.ding poles

er system usuall y exhibi ts dampe d oscilla tion. 7.
Expression for Rise Time an origin all •iinsta ble svs ·

E For answer refer Unit-V, QI , Topic: Rise Time. QS. Expla in the<stead y-sta te error analy sis o
,
lpression for Peak Overs hoot conti nuou s data contr ol and discr ete dat
For answe r refer Unit-V, qI, ,Topic: Peak Overshoot. control s~stems.
·
SPECTRUM ALL-IN-ONE JOURNAL FOR ENGINEERING STUDENTS
 'v
200 DIGITAL CONTROL SYSTEMS [JNTU-KAKINADAJ
Ans: · Steady State Error Analysis of Discrete Data Control Systein
(Model Paper-II, Q9 I April-15, Set-4, Q7(a))
To determine the transfer function of steady state error
Steady State Error
of a closed loop system.
Steady state error gives an idea in determining the · Let, R(s) be the input signal
accuracy of a system. These errors ·are caused in the system E(s) be the error signal
due to nature of input (namely ~it ·step, ramp and paraboli Y(s) be the output signal.
inputs), type or number of system and the disturbances cause
The block diagram of discrete data closed loop system
ue to system components. It is .defined as the value obtaine
Y error signal e(t), whe~ the val~e t>f
't'tends to 'co'.
can be drawn as shown in figure (2).
Y(1)
~
Mathematically expressed as, e,, = ,~~ e(t) •------------ - I
, T .
~!---:-+--~y(t)
Steady State Error Analysis of Continuous Data Control G,(s)
System .___~' Y (s)

''~- --------------- -------- -.''

To detennine the transfer function of steady state error
i---- G(s) -----1
of a closed loop system.
Let, R(s) be the input signal
C(s) be the output signal Figure (2): Block Diagram of Discrete-data Control System
E(s) be the error signal From figure (2),
C(s}H(s) be the feedback signal. Error signal,
Continuous data tI::ie block diagram of the closed loop
system can be drawn as shown in figure (I).
E(s) = Input signal - O utput signal

Error signal E(s) = R(s) - Y(s)

+------ C(s)
Output signal,
Y(s) =Goh Gp (s)
C(s}.H(s) H(s) The error analysis performed here is only for unity
Feedback
path feedback systems with H(s) = 1. Thus, the error signal is
represented by,
Figure 11)
From figure (I), E*(s) or E(K]). That is,

Error signal, E*(s) = R*(s) - Y*(s)

E(s) = I~put signal - Feedback signal E(K]) = R(K]) - Y(K1)
E(s) = R(s) - C(s)H(s) ... (I) At the sampling instants, the steady state error is defined as,
Output signal,
e;s = It E*(s) = It E (KT) ... (I)
· C(s) = E(s).G(s) .. . (2) K ➔oo K ➔oo

By substituting equation (2) in equation (I}, we get, By applying final value theorem ofZ-transform, we have,
E(s)= R(s) - E(s)G(s)H(s)
e;s = It £*(KT)= It (l-z- 1)E(z) ... (2)
E(s) + E(s}G(s)H(s) = R(s) K ➔- K➔I

E(s)[ I + G(s)H(s)] = R(s) Representing E(z) in terms .o f R(z) and Y(z) from figure
(2), we get,
R(s)
: . E(s) = 1+ G(s)H(s) ... (3) E(z) = R(z)- G h G/z) E(z) 0

E(z) (l + .G G/z)) = R(z)

0
,;

We know that,

e = Lt e(t) E(z) = l GR(z) ... (3)

ss t ➔ oo
+ oh Gp(z)
· By applying final value theorem of Laplace transfonn, Subtituting equat_ion (3) in equation (2) we· get,
we have,
e;s = It E* (KT) = It (l-z- 1) .R (z)
e = Lt s.E(s) K ➔- z➔I l+GhG·
p
(z}
ss s ➔O 0

... (4)
s.R(s) From equation ( 4) we conclude that the steady state error
Lt---'-'---
e,, = s➔o l + G(s)H(s) ( :: From equation (3)) depends on both reference input R (z) as well as transfer
function Goh G/z)
qf forward path.

WARNING: Xerox/Photocopying of this book is a CRIMINAL act. Anyone found guilty is LIABLE to face LEGAL.proceedings.
uNIT-5 (Design of Discrete - Time Control Systems By Conventional Methods) 201
The transfer function of controlled process of figure (2) When input is unit step,
be of the fonn,
I
G(s)= K(l+Tas)(l+Ths) .. .. (l+Tms) R(s) =-
s
P s 1 (1 + 7; s)(l +_T2 s) .. .. (I+ r,,s)
I
s
Where,j = 0, 1, 2, ... = Lt - - - -
e = Lt s
·" .1· ➔0J+G(s)H(s) . -►ol+G(s)H(s)
The transfer function of discrete data control system is,

G,,.,,G ,(z)=(l-z-- 1)z

· [G (s)l
_P_
-----= - -
1+ Lt ·G(s)H(s)
s ➔O
l+K
P
' I S

. G G(z)=(l-z--I)z[ K(l+Tas)(1+T6 s) ... (1+T111 s)

•• on P
s
j+l (l
+ J;s)(l + T2s) ... (I+ T,;s)
l Type O System
[·:KP= Lt G(s)H(s)]
.,➔ O

K = Lt G(s)H(s) I '
., ➔O
Q6. Derive an expression for steady state error for P

parabolic input.

An•: Sep.-20, Set-4, Q1(e)

Determination of Steady State Error for Unit Parabolic

Input, ·

e = Lt sR(s)_
" s-o I+ G(s)H(s)

For parabolic input, R(s) = __!_

. s3
Type l System
s K = Lt G(s)H(s)
3 P s ➔O
e = Lt s
" s-0 1 + G(s)H(s) = Lt K (s+z 1)(s+z 2 )(s+z 3 )• · ·
1 s ➔o s(s+ p 1)(s+ p 2 )(s+ p 3 ) · ··
= Lt-----
2 +s 2_G(s)H(s)
s-0 s = 00

i I
e,, = l + K = l + oo = O
. Lt s 2 G(s)H(s) Ka p
s -0
Type 2 System
Type OSystem

. K = Lt s 2 G(s)H(s)
0
KP= fjo G(s)H(s)
s!. 0
_ Lt k)s + z )(s + z 2 )
1

- s ➔O s 2(s+ pi)(s+ Pi)

= 00

I l
e =-= - =oo e,~ = l+K = l + oo = 0
" Ka 0 I'

Q7. Derive an expression for .steady state error for QB. Derive an expression for ste~dy state error for
ramp Input.
step input. ·
Ana: Ans: Sep.-20, Set-3, Q1(e)
Sep.-20, Set-2, Q1(e)
Determination of Steady State Error .for Unit Ramp Input •
l>eter~ination of Steady State Error for Step Input
Stead Ystate error Steady state error,

·L sR(s) sR(s)
e - t e - LI
,, - s ➔o 1+ G(s)H(s) ,, - .v ➔ o I+ G(s)H(s)

SPECTRUM ALL-IN-ONE JOURNAL FOR ENGINEERING STUDENTS

 ' \,
202
DIGITAL CONTROL ·svsTEMS [JNTU-KAKINAoAi
-;:::: :-=:- ----- ~~:.: ::~== ----- ...;_- -
For ramp input,
(II) T= I sec,
K 2

R(s) =~ K.=
K
T = T =2
2·
s

sx.L
s2
Steady state error for ramp input,
... e = Lt .---:--=--:~c---
.u ., - o I+ G(s)H(s) .
eramp
ss
= _1_
KV
= ½ 0.5 =
l
= Lt . s (iii) T= 2 sec, K =2
s-o 1 + G(s)H(s)

1 K . -- K-~=I
T - 2
= Lt
s + sG(s)H(s)
s-0
Steady state error for ramp input,
= 1 1 . '
l =J
.

Lt sG(s)H(s)
s-0 ·
=K V
ramp __I_ -
e,s - Kv - I
Q9.
The feed forward pulse transfer function given From the above 3 cases it, can be concluded that as
G(z)= · 1<z(1-e-T) r
the sampling time ·• increases, the steady state erro~ also
(z-1 )(z...;e-T) ,,. increases i.e.; steady state accuracy decreases.

Investigate the effect of the sampling period T Q1 O. ·A unity feedback system is characterized by the
on the steady state accuracy of the unit ramp open loop transfer function. ·
response for the following (I) T = 0.5 sec,
= =
K 2 (ii) T 1 sec, K = 2, (iii) T 2 sec, K 2. = = · 0.2385(z + 0.8760)
Write comment on the above cases. G"°G(z) = (z-1)(z-0.26 44)
Ans: Sep.-2O, Set-2, Q&(b)
Given that, The sampling period T = 0.2 sec, Determine
steady state errors for following (i) Unit Step
Feed forward pulse transfer function, (II) Unit ramp (iii) Unit parabolic
kz( 1-e-T) (Model Paper-I, 010 I Sep.-20, Set-1, Ql(b))
Ans:
. G(z) = ( z - 1 )( z-e-T)
Given that,
In order to obtain steady-state accuracy of the unit ramp
response, consider the steady state error for ramp input. Open loop transfer function of unity feedback system is,
. . . 1 )
Steady state error for ramp input, e';mp = y G G(z) = 0.2385 (z + 0.8760)
1
V
'° (z-l)(z-0.2644 )

Where ' KV = l..

T
Sampling period, T= 0.2 sec
..

_ Require~ to detennine,

Steady state error for,

(i) . Unit Step

Steady state error for unit step input is given by,

s 1
e rep=-wher e K =limG(z)
" 1 +Kp '
T = 0.? sec, K = 2
P
(i) z-1

K. = i K
P
= Jim
• z- 1
0.2385 (z + 0.8760)
(z- l}(z-0.2644)
= ls = (0.2385)(1 + 0.8760)
=4
(1-1)(1- 0.2644)

Steady state error for ramp input, = 00

1 l
eramp =,;, = 4 = 0.25 estep, .= ~ = .-1_. = o·
ss A,, , . ss . 1 +K
· p 1 +co ·
. ' ,• ,::;;. ·:
· :-- ·,/ .\ i" WARNING: Xarox/Photocopying •of this,book is a CRIMINAL act. Anyone found guilty is LIABLE to iaCe LEGAL proceedings.
 uHtT-5 ((?esign of Discrete ~ llme Control Systems By Conventional Methods) 203
• ne: Aprll-18, Set-4, 06
")
('-'
Unit Ramp ..
Given contro l system i/1 shown in fig ure.
Steady state erro r fo r ramp input . e romr = - ~
·•·• K, .
R(z) K(z:+ f)
I i,........-C(z)
Where K, ,= T !.i~1[(z- 1)G(z)] (Z l )(l 0,6065 )

, 1.
· K = - ,m
l(z 1)0.2385(: + 0.8760) ] Figure 111
.. ' T ,- , (..: - I )(z - 0.2644)
T = 0.1 sec
= _I_,.,0.2385 I+ 0.8760) ]
x(
C, = 0.5
0.2 I - 0 .2644
To dete~ine.
= ~.04 1
· C riti ca l va lue of ga in K •fo r stability = ?
I I
c.,"""' = Kv = 3.041 = 0 .328 Va lue o f K fo r [, va lue of0.5 = ?

wd for Kat C, val ue o f0 .5 = ?

(lill Unit Parabolic ' Numbe r o f samples per cyc le o f da mped s inusoidaJ
oscillation =?
Stead y state error fo r paraboli c input e para = _l_
' ss Ka From figure , we have,
I K(z + 1) ... ( I )
Where K = - 2 lim [(z- 1)2 G(z)] G(z) = -(z___l_)(..z.:. .---0...:...
.6-06_5_)
" T Z -· I
2 The characteristic equatio n is g iven as.
K = _!__lim l(z- 1) 0 .2385(z + 0.8760)]
" 2
T =- • (z - I )(z - 0.2644) 1.+ G(z) = 0

= _I lim
r2 : - I
I (z- l )0.23 85 (z + 0.8760)]
(z - 0,2644)
I+ K(z+ l)
(z- l) (z- 0 .6065)
=O

z2 + (K - I ,6065)z + 0.6065 +K =0 .. . (2)

= _ I_
0 .5 2
I(
1- 1)(0 .2385)( I + 0.8760) ]
( I - 0 .2644) Critical Value of gain K for Stal~llity
From Jury's st,lbilityCritcrion we huvc the
=0 characteristic equatjon as,, _•
I I .,. (3
e para =-
Ka
=-
0
= oo
..
, By comp~ring ..:qua~ic,n (2) und (3), we get.
011 . Consider the digital control system shown in
figure, plot the root loci as the gain K is varied • ,~0 ~}.::fi,;:f ~·.tl ,~~~• ~'z '"' 0,6065 +- K
from Oto""· Determine the critical value of gain
We also have the stabi lity condi tions as fo ll ows,
K for stability. The sampling period is 0.1 sec, or
T = 0.1. What value of gain K will yield a damping .. . l -l )
ratio ~ of the closed loop poles equal to 0.5?
Wtth gain K set to yield = 0.5, determine the s P( I ) > 0 ... (5)

damped natural frequency wd and the number ... (6)

P(- 1) > 0
of samples p-er cycle of damped sinusoidal
oscillation. On substitut ing a.,, t i 1and t i , vn h1l•-; in equa tion ( -l ), (5 )
C(z) and (6), we ge t, ·
R(z) K(z i I )
From cqw1t io11 l.l ), !0 .606 5 1 Kl < I
(z - I )(z - 0.606 5)
K - 0.393:'i ... (7)

From cqiu,tiun (4) , P( I ) ~ I t- K - 1.6065 + 0.6065 +K > 0

t
~ 2K > O ... (8)
Figure

SPECTRUM All•IN-ONE JOURNAL FOR ENGINEERING STUDENTS

 \
204 DIGITAL CONTROL SYSTEMS [JNTU-KAKINAD~J

From equati~n ( 5), P(- 1) ⇒ l - K + l .6065 + 0.6065 + K > 0

⇒ 3 . 213>0
... (8)

From equati~ns (7) and (8), we get,

0 < K < 0.3935

Critical value of gain K for stability = 0.3935
I ,
Value of gain K for ~ value of 0.5
Phase angle of equation (I) is given as,
LG(z)_= L (z + l) - L (z- l )-L (z - 0·.6065)
On substituting z = a + jro in the above equation, we get,
180° = L ((! + jro + l) ...:. L (cr + jro-1)".:... L (a+ jro - 0.6065)
180° + L [(a - 0.6065) + jro)] = L [(cr + l) + jro)]-L. [(cr-1) + jro]
180° + tan-:1.[ ro · ] =tan_; [ ~ ] - ta~- 1 [ ~ ]
cr-0.6005 a+) · cr-1
By substituting t~n on both_sides in the above eq~ation, we get,
0) (i)
(i)
cr+ l · cr-1 . tanA - tanB ]
a - 0.6065 [ ·.·tan(A-B)= l+tanAtanB

ro -2ro
a - 0.6065 cr 2-l + ro 2
By solving_above equation, we get,
cr 2 + 2cr + ro 2 = 2.213

(cr 2• + 2cr + l) + ro 2 = 2.213 + 1

(cr +1)2 + ro2 = 3,213

(cr +1)2 + ro 2 = (l.7925) 2

. We have,
lzl = e-IT.w,,
Jzl = Tro,, /1 -(, 2 · •
By substituting T and t;: values in the _abQve equations we get,
lzl = e --0.1 X o._s .,,, = e --0.05 "'n

Lzc=0.l x ro,,/(-(0.5) 2 =0.0866ro,,

jro

Figure (21

. WARNING: Xerox/Photocopying of this book is a CRIMINAL act. Anyone found guilty is LIABLE to face LEGAL proceedings._i..
r

205

-
UNIT-5 (Design of Discrete - :nme Control Systems By Conventional Methods)
By substituting co.= 4 rad/sec in lzl and L.z, we get
Jzi = e---0.o5 • 4 = 0.8187 Explain the design
OR
proced ure in the w.plan e.
!
(Aprll/May-16, Set-1 , Q7(a) I Aj>rll/May-16, Set-4, Q7(b) I
Jzl = 0.0866 x 4 = 0.346 rad= 19.847° April/May-14, Set-1, Q7(b)) ·!
I

z = lzl Lz= 0.8187 L.19.847° G(z)

[ro. = 4 rad/sec since by trial and error z lies on root I I

I
locus] I

z = 0.7701 +j0.2780 ·
By substituting z value in. the
'
magnitu de condition,
IG(z)I = l, Figure
We get, Ans: .

·.
.
K(0.770 1 +j0.2780 + 1) I The given digital control system is shown in figure,

I(0.7701 +j0.2780 -1)(0.7 701 +j0.2780 - 0.6065) == l

By solving above equation , we get,
K = 0.3606 x 0.3226
1.7918
K=0.06 49
' Figure
ro• fork at<; value of 0.5
The various steps involved in the design procedu re in
We have, the w plane. for the given digital control system are as follows,
cod = co n ✓1 - r' 2 Step-I
Initially evaluate the z transfor m of the plant precede d
By substitu ting co. ands values in the above equation , by a hold i.e., G(z). Now, by using the bilinear transfor mation
, we get, transform q z) in to the transfer function G(w).
Where,
cod= 4 /I-(0.5 ) 2
Bilinear transfor mation equation is given by,
cod= 3.464 14-(I\v
Number of Samples per Cycle of Damped Sinusoidal Oscillatim1. ~
We have,'
Number of samples per cycle of the damped
z= 1-(~r ... (I)

360° .
sinusoid al oscillat ion= · [ ·.·· L.z = 19.847°]
19.&47° the samplin
ote: Care :should::b_e':i~eri';,vhile:selecting
= 18.14 samples per cycle

.;;.:;:.~J.i;{if;~J;;:\,iJ~.::;.,. i .. ·. .
:\.'-$:cfor ,' G(ji•) i.e.;, by substitu tin~

;;:fts;\r=tf&,/l~";i." .. ·•. . ·
., ,_., ,.,;.Ibe,,1:i9de P;l,ot~

, .. The requjr,e.cf;~~Iu~fHe,1f sll!tic error constants, phase

Q12. Write design· proce~ ure In thew-p lane. nargit1 an.d gain i'.nafgib·afe:noied '.tJo~:the bode plot. , ..
Sep.-20, Set-1, Q6(a) sf,ep~4:::. :t ·}> /{\-~}tf)};1~iz1?~{? ·, .. . .
OR . . Eval~at e)~e\ sy}t~t# ,:gai~:~~ pertbe requirement~ fo
a given static error cqilstant. Here, $e low frequency gam o
Write brief note on desiOn proced ure in the iscrete time.controlle'(mtlst:b.eas$Uined to be unitv. After th
w-plane. April-18, Set-2, Q1(e) etennii1ation' o(s1stenfg~in ~:valuailthe
poles and ~eros of th
igital controllertransfer f'unctirin 'with.the help of conventional
OR .Ol~tim;ou's}i~ne_c(mtrol system. Finall)
de_sign ,technique for .C
Explain the design proced ure In the co-plane of
lag com_,ansator: -.;/i)ff('.\'.~~-,-· " · )ff\~)/: .• ·
the open loop transfer funcfion.for ,thede~igned system will b
o_btained as. ,.. ·. ..
April/May -17, Set-2, Q6(a) GD(w) q<,._,_' •. ,· 'a' ,. •.· )
' ,,;-;.~1 t ,~'"'.ff;J)'\;.(t\..;:J....i.,~ ,.. .. {,, ~,.,~'
• ~ :',, • ·,..... "' "Z..i:R-~ '
t

,, '
1
'· · ·_.< . _., ;SPECTRUM All-IN-ONE JOURNAL FOR ENGINEERING STUDENTS
206 DIGITAL CONTROL SYSTEMS (JN
TU-KAKINADA]
Where the value of E rang es betw een
50 to 120.
By using.bilinear transformation equation
trans fer the controller transfer w) into Go<z).
shown below
Go< 1-P
sin «I>,.= l+P
Whe re,
4. Dete rmin e the ney.' gain cros s over
Bilinear transformation is given by, frequency which
satisfies the conditions give n below,
2 z-1
. w= - - -
... (3)
T z+l Mloom =-1 0 log (~)
Go<z) = Go<w) ... (4) Where, ..ro,. is the new gain cross over frequ
Step -6 ency given by,

With the· help of a computational algo 1

rithm analyze the w,. = T Jj3
pulse trans fer function Go<z).

. Q13 . Wri te the design procedure 5. Using the value of~ . obta ined in step
_of lag compensator 3, determine the
in ro-plane. value ofT, roe, and roe~ usin g the relat
ion shown below,
Ans : llllllll!l!lllll!llllll!IIII 1
April-18, Set-3, Q1(e)
1. ~"-. •.u ....... " " " .,...,, ..... vali.1e of gain, k of w'" = T Jj3
the
that it satisfies the'r equi re~e nts i.e., give system such
n steady stat
erro r .specifications; / \ . ;:;;tt,\ : .
l 1
Draw the bode plot i.e., magnitude and roe, = T' roc2= PT
phase plot with
the valu es of k obta ined irt'abo~e-step _'
. ·
From the bode plot, .de~6nnh1.e the gai'n 6. Determine the valu e of K for the com
crossover fre C pens ated system
quen cy that gives requfred transient resp using relation,
on·se. .
Dete nnin e the tal~ agn itud t't him gain
plot required K= Kc~
.to shift the gain plotJ<> have'.:ftie 'new
·gain crossove
frequency. · i/,\'·'· · .. ___ · · . >·· 7. Draw the Bod e plot takin g t)le tran sfer
function of the
Select,the upper C?!l lerfr eque nifsi i~h! · compensated syst em and henc e chec k
?at_it is 1 decad the specifications .
·.belo'<vthe'freqtiency'o btain ed i.fi.step,3/
/ - .... _ The transfer function of com pens ated
syst em is given
6. The first com er frequency is give·n by, by,
KcP
G(s )- - -(l+T
(J) _ (l) gc, new s)
c, - 10 --
c (l+P Ts)
7. The seconq com er frequency is given
by, Q15. Con side r the tran sfer fuct
ion sys tem shown.
I The sam plin g peri od Tis ass ume
(J) =- d to be 0.1 sec
. C2 pr obtain G(w). ·
8. Determine the lag compensator.transfer

given as,
function. Usually
the transfer function of such compensator
in ro-plane is
___,;/ - 11-e-T• I\ I 10
6T . . s > s+10
L'
1
: l+w T
D(co) = - - P > 1: T > O (Mod el Pape r-Iii, Q9 I April -18, Set-2
, Q6(a) I
t+pr or Ans : April /May -14, Set-4, Q7(b))
Q14. Write the design procedure Give n syst em is, .
of lead compensator

Ans:
in co-plane.

Procedure for Drawing Bode Plots of

a Lead Coinpens ator
April-18, ~et-4; Q1(e)
-- -., ;_/ _ -
6y
1-e- T•
s
-

Figure
- -
10
-
s+10
-
1. Determine and set the value of!( whic
h satisfies given Sam plin g period, T = 0.1 sec
steady state error specifications .
G(w) =?
2. Draw the Bode plot with the value of
K obtained and
dete rmin e the phase margin(«!> ). The trans fer function of the syst em is
given as,
pm
3. Dete rmin e the maximum lead angle «I>' 1-e - Ts 10
" and hence the G-(s) =
.value of~ using the relation, s . s+lO
tt. =tt. - (j> + E
'f' m '+' spec ified 10
pm G(s ) = l-e-n· - - -
. s(s+ l0)
WARNING: Xerox/Photocopying of this book is a CRIMINAt
act. Anyone found guilty is LIABLE to face LEGAL proceeding
s.
 UHIT-6 (Design of Discrete - Time Control Systems By Conventional Methods) 207

· 10 Substituting s = :..._10 in equation (5), we get,

Let G 1 (s) = (
s s+IO) - I0=A(O)-B10
- •• G (s) = (I - e-n) G 1(s) ... (I) => B=-1
G (s) = G 1 (s) - e-r, G 1(s)
Let, X 1 (s) = e-r, G 1 (s) ... (2)
1.-.B=-l I
Assume, e-T•= G0(s) Substituting s = 0 in equation (5), we get,
==> X1(s) = G0(s)G 1(s)
I0=A.10
en by,
smce Lu Ko (<)go (t- t)dt} = G0 (s )G, (s) => A=l ·

1.-. A =I I
1 1
r:te the ·· r {X 1(s)} =r {G0 (-s)G1(s)}
IO I - I
- - - = - + - -·
,elow,
J
I
s(s+IO) s s+lO
x 1(t) = g 1('t)g 0 (t-t)dt
0
I I
G 0 (s) = e-rs G1(s)=----
s s+IO

r- {G (s)} = r:•~-~}
1
0 On applying z-Transform cin both sides, we get,

ystem
g 0(t) = o(t - T) ·
1
z{G1(s)}= z{.!.s __ l_}
s+IO
·: L"' {1}=o(t) }
{ => r•~-Ts }=o(t-T) G 1(z)-
1 1
- -
J - -I - J -e ~I OT z -I
-z
fg
I
of the
x.(t) = 1(t)o(t-t-T)d't 1 1
1tioris. ·:--z
0 (s+a) +-> (1-e-aT z-1)
given
o(t - t-T) = I fort - t - T = 0
z
= 0 otherwise.
z-e -IOT

-
1own.
1 sec
x 1(t) =g 1 (
On applying z-Transfonn on both·sides,we get,

x (z)=z- 1G (z) {·:x(t-KT)=z-KX(z)}

I - T)

... (3)
1
·z
G(z)= _:...___1
z-1 z-·e-
z
(.".T=0.I sec)

1 1 z z .
- G(s)=G (s)-e-T•G 1_(s) (Fromequation(l)) G 1 (z)= z-1- z-0.3678 ... (6)
1
On applying z-Transfonn on both sides, we get, From equations (4) and (6), we get,
z{G(s)}= z{G1(s)}-z~-T•a.(s)}
ai;(a)I (From equation _(2)) G(z) = (1 :_ z--t ){zz-1 z }
Q7(b)) z{G(s)}= z{G1(s)}-z{X1 (s)} z-0.3678

z{g(t)}= z{g 1(t)}-z~1(t)}

G(z) =(z-1 ')f_:_ z }
G(z) = Gi(z)-z-•a1 (z) (From equation (3)) . z Jlz-1 z-0.3678

4
G(z)=(l-z- 1 )G1(z) ... ( ) z-1
G(z) = I
. 10 · z-0.3678
·:G(s)= s(s+IO)
· z-0.3678-z+l
_On applying partial fraction, we get, G(z) = z-0.3678
10 .A B
---=-+- 0:6322 ... (7)
s(s+IO) s s+l0
G(z) z-0.3678
... (5)
_10 = A (s + 10) + Bs
. · SPECTRUM ALL-18-ONE JOURNAL FOR ENGINEERING STUDENTS,
 208 DIGITAL CON TRO~ SYST EMS [JNTU-KAKINA
DAJ
By using biline ar transformation, we get,

z= 1-{f }
z = 1+(0~ 1 r,
1-(0 ~1}

l+0.0 5w
z=---
l-0.05w
Hence to obtain G(w), substi tute the value of z in equati
on (7), we get,
G(w) = 0.6322
1+ 0.05w
03678
l-0.0 5w -
G w _ 0.632 2(1- 0.05w) _ 0.632 2(1- 0.05w)
. ( )- (1 +0.05 wh 0.367 8(1-0 .05w) - _1+0.05w - 0.3678
+ 0.3678(0.05w)

-, ----
0.632 2(1-'0:05w)
-'-- ----' --- =- 0.632
-~ 2(1-0 .05w)
- -~
0.6322 + 0.05w + 0.0185w 0.6322 + 0.06839w
0.632 2(1- 0.05w) 9.244 (1- 0.05w)
= =
0_0683J w+ 0.6322 )
~l 0.06839 w+9.244

_._ G(w) = 9.244( 1-0.0 5w) I

I_ w+9.244 .
Q16. Consider the digital' control system shown
in figure 6. Design a digital controller in the w-pla
that the phas e marg in is 50°, t~e gain marg in is at ne such
least 10 dB, and the static veloc ity e'rror cons tant
is 2 sce- • Assu me that the samp ling perio d is 0_
1
.2 sec.
K.
·
1-eTS K c(t)
s S(S + 1) c(z)

Ans: Figure
S.p.-2 0, Set-'. Q6
Desig n a digita l contro ller in the w plane such that the
phase margi n is 50°, the gait\ margi n is fgJ OdB and stati~
onstan t, K is 2sec· 1• vell1c1t)
y
· ' • • ··
' •
Assum e T = 0.2 -· · · . .,· .· · • ·
....,., ' \

Solution

G(z) = z [ l - e--0.2s K - ] = 0.0 I

873 [ K(z + 0.935 6) ]
s s (s+ l) . (z- l)(z - 0.81 87)

G(w) =. G(z)f, yo, w -

J- 0. JW ~
K[~+
w(
f-*l W )
I + 0 .997

WARNING: Xerox/Photocopying of.this book is a CRIMINAL ac-t. Anyone found guilty is

°LIABLE to face LEGAL proceedings} .. .
 Methods) 20 9
trol Systems'By Conventional
ONIT-5 (Design ·of Discrete -Ti me Con
~ Poles at w = 0 and w= 0~997
LHP zero at w = 300.6 and RH P zero at w =•Jo

,. - j

i',, ,,!'
\ I I.
"
. -·-_ :-;:I . .

i
I
_· \ i f
I.

Try a load com pen sato r

w
I+ -
G (w) = _a _
D W
l+ -
13
ENGINEERING STUDENTS
'"/.v~ ·: .. ·;,, .. 1
•• d!;:C.-\~{l\'-,t'•,...-A<. --.'
SPECTRUM Alt-IN-ONE JOURNAL FOR
 210
DIGITAL CONTROL SYSTEMS [JNTU-KAKINAD'A)
Need to adjust K , a .and 13 to satisfy specifications. Adjust Q17. Exp1,1n the dlgltal lmplementatlon of lead lag
K to meet static accuracy specification. controller
. I '
In detall with suitable diagram.
.
Open loop transfer function is, Ans:
. w K ( -. w- + I) [1-w ))
. I .+- igit:d lmplemcntalion of Lead Lug Controller
-
G (w) G(w) = _ _a_ 300.6 IO The digital impkmcnlat ion of kad lag l.'<•ntrollcri
D
I+ -W [ •ompensator combines the effects of both lead 1.:omp1:11~ator
w -W- + I )
13 0.997 1 nd lag compensntor. Lead compensator is designed to give the
Require K,, = 2 esired transient to response and stability to the system whereas
Where Ing compensator is designed to improve the steady state error of
he lead compensator. Thus. lead lag compensator results with
K . = limwG 0 (w)G(w)
w- 0 etter transient response, steady state response and stability.
⇒ K=2 The transfer function of lead lag compensator is given
With this· value of K , we can read the gain and phase by,
margi!)s. We find 30° phase margin and 15.5 dB gain margin. G(z)l1eac1-11g = G(z)llead x G(z)l1ag ... (I)
To give a boost in the phase margin, we adjust the parameters ·
of the lead network a and p In order to detemiine the digital implementati on oflead
. (I) lag compensator of digital control system, first lead and then
I+- Jag compensator has to be evaluated. The transf~ function of
G (ro)= _a_ continuous control system is converted into discrete control
D (I)

I +p system by using three difference approximatio n rpethods. They

are,
I. Backward approximati on .
ro =a

L Gp(ro) ( b.- .
It is solved by su st1tutmg S . =-- z-1)
zT'
2. Forward approximatio n.

we decide on
w
l + --
( It is solved ~y substituting S = z -,
. T.
l) .
0.997 3. Bilinear approximati on.
w
I+ 3.27
⇒ 50° phase margin and 14 dB gain margin_. ( It is solved by substituting s =
·
~ ( z - l ))
T' z+l
Now transform the controller to the z phme Lead Compensator

Go<z)= GD(w) lw=10..::.!.. The tt ansfer function oflead compensator in continuous

z+l control system is given by,
z - 0.8187
⇒ . Go<z) = 2.718 z..:. 0.5071 1 S+ -
J
I The open-loop pulse transfer function of the compensated
system is,

Go(z) G(z) = O.IOI 8 (z -I) (z - 0.5071)

z +0.9356 Where,
.
G{s)= - (
~ S+ - 1
r1 ·
~T

~ - Phase lead compensator constant.

The closed-loop transfer function _is,
T - Phase lead compensator time constant.
C(z) = 0 .1018 (z + 0 .9356) Now,
R(z) (z - 0.7026 +J0.3296)(z - 0.7026 - J0.3296)
The transfer function of lead compensator using
closed-loop poles z = 0.7026 ± J0.3296 backward approximation is,
⇒ s= 0.5 ; G(z)I = .!_ (za3 +a2 )
2n . l<od ~ za3 +bi
We find that w, = T = 14.3 wd
Where,
where ro)s the damped natural frequency of these poles .
a = ( -T'
r - 2 ) , a 3 =· ( 2 + - T')
r and b = (-~r
T' -1·)
wd = wn /1.- s2 2 . 2

T' - integration time,

· ~1;l}· :/,:::; WARNING: Xerox/Photocopying of this book is a CRIMINAL act Anyone fo~nd guilty is LIABL.Eto -fa;e LEGAL proceedings:} f\
·;/:,. ., '
 211
uNIT-5 (Design of Discrete - nme Control Systems By Convent ional Methods )
The transfer function oflcad con)pcnsalor usi11g forward Lead Lag C ompensa tor
(ii)
approximation is. By substituting G(z) values of hac k ward, forward and
approximation with respect lo lead ;ind lag compensator
G(z)/,,.,Ji [::~I]
=
bi linear
in equation ( I). the transfer function of lead lag compens ator
arc as follows.
Wherc.
By Backward Approximation
a,=(~' -1) and b 1 = (;; - .,)
( i)

1 ( za 3 + a 2 ) l ( zc:... I )
(iii) The transfer function oflead compensator using bilinear G(z)l,.,,.,.1;,~= 13 za3 +b2 x a zd - I
approximation is,
(ii) By Forward Approximation
G(zt.d = t( ::: ::: ) Gz --I --x -
( ll1c,d-lu~- ~ z+b1
I --
(J. z+ dt
(z+a1) (z+c1)
Where,

a,=(!_'
-_ T
_-2), a =2+ !_'_T andb = (£-2)
3~T 2
(iii) By Bilinear Approximation

Lag Compensator
The transfer function of lag compens ator is continuoi.t~
control system is given by, Q18'i.ll':: Oiscuss 'the'ir:ev.lew.<>f Rllase lag, lead and lag
II!''' lead coinpen,~tor.. :' _ -.,:,?: ·
·_+ ~. ~ :Jjf;~:/,. >, . .
- . ..\ .,.. . ·•·.

J
•·

s+ - ,. · · · ... , r- , '· Q7(b{(~pril/May-14, Set-3, Q7(b))

~ f ' - ~~;-~'.:11r:i ~':, ~-' ...

. I
G(s) =
[ s+ -
a.T que is used to reduce the
Where,
encies but the sysLem gai
a - Phase lag compens ator constant reduced.

(i) ·
T - Phase lag cm1~pensator time constant.
The transfer function oflag compens ator using backward
. ~i e
reduces the bnndwidth
· as slower speed ot
approximation is,

G(z)/Jag = a.
I(zdzc-1)
- I
all system gain can
gain is low, therefore .
ca~ed which improves
Where,

c =(l+ ~') and d=(l+:T) ualed in phase la

(ii) The transfer function of lag compens ator using forward

approximation is,
.
) .,

,' proves the stability

~ ( ;:~: )
;'lr

~(z)/,,s =
•a,;es the bandwid th o
Where, aster spe\!d of response .

c
I
= (..CT --i) and d = (__c_-1
a.T I
) · in high frequenc y noist:
, i frequency gain.
(iii) The transfer function of lag compens ator using bilinear
approximation is ·
' ~.....l"'l\i':( . . ~
tor i; .fonm:d bv• casca<lin
.

~ fog · ·a_sc ,}e_acl co_mpensator.

.. '
ii) - -• S!'J~l~r}o~ frcqucncy gain
1

c )' f lOte_accuracy of th
Where, s j . .
T'
c2 = ( y-2
) ~ c3 = ( 2 + TT') and d~=
(a.T
1'' - 2) iii)
nc aseo.
1argin -of a lag lead
: - ·:_:/ -

SPECTRUM ALL-IN-ONE JOURNA

.
L FOR ENGINEERING STUDENTS
'
 CON 1 ".., _ ctton of trequency.
. DIGITAL tun .
21 2 rtodlC ·
. · tem Is a pe
a '
dis cre te sys
res pon se of
Q19· Show that the frequency
Ans:
Let,
h(n) = Impulse response
x(n) = Input sequence.
put is given as,
By using convolution sum the out
00

y(n) = L,h (k) x(n -k)

k= -

= ejw,I

I
Let, x(n)

· => ;(n ) ~ :t:_h(k~eftd.•-•J ~,... [:t:_h(k).e_,.. ~e"'" [H(e")]

Where, H(e"")= f
k= -
h(k).e- i~

... (I)
= Frequency response of the system
. y(n )= ti""'. [H(e"")]

H(e"") is also written as, ·

H(ei"') = HR,(e"") + H,,.,(ei"')
a
Consider sinusoidal, equation,
we get,

x(n) =A cos(co0n + 0)

. · [ei<~+a) +e-i<~+a)
x(n) =A
·
]
2

-,.~n.eJ8 +-Ae- 1·,,,._ft

A 1"" ...,,.._e
-J8
= -e"
2 2
From equation ( l ), we have,
y(n )= eiwn [H(e'"')]
A 1. .eJ8 ,
~
Lety 1(n) is the response to -e
2
. A 1.· J8
(n) = H(e1"lo )-e
2
1Jlon .e
.
YI
n .e-J8 then,
Let y 2(n) is the response to ; e-J~

y (n) = H(e1"lo); e-J!Jlon_e-J8

2

The total response,

= H(e i"lo) A·( ei~ _e1fl +e- i!Jlon .e-J8)

2
= A V!( ei"lo )j cos(co0n + 0)
. .
. -
is a CRIMINAL act . Anyone found UUil
· WARNING: Xerox/Photocopying of this book t .15
y LIABLE to face LEGAL proceedings,
.
.
 &JNIT-6 (Design of Discrete - Time Control Systems By Conventional Metho~s) '213
· We know that,
.. '

H( e"") .=·2, h(k ).e - iox

t=-

Where, [H{ei"')] is a continuous and periodic function of oo with period 211 and e 112 n H•J• = e'",1;.
The frequency response of the system is same at oo and oo + 2n. ·
Hence the frequency response of a discrete sys.tern is a periodic' function of frequency.

· 5.3 · Roar Loe~• TecttN•~~ IN T~E .Z-~LAN~

Q20. How do you map constant damping loci from a-plane to z-plane? (Model Paper-IV, 091 Aprll-18, Set-1, 05(•))

OR
Explain about the conceptof constant damping ratio loci.
Aft9: Aprll/May~16, s.t-3, Q6(a)

Basically, the constant damping ratio· loci in S-plane corresponds to a spi;al in the Z-plane as depicted in figure {i).
jro

Constant S-plane Im
damping j<i)d
ratio line
- j(J\
. ro,
J-
2

(1) Clftlt111t D1mpln1 Ratio Line in S-pl•ne ibl Corresponding Locus in Z-plane
Fi,ure (i)

lbe expression for constant damping ratio line in the ~-plane is,

s = ~wn +100n R
~ S =- ~O)n + jO>d
Where,
. ,.----;:i
00;= 00nvl-~

And the z-transform relation is,

~O>n + jw) = exp(- ~wnT+ J0>l)
Z = ers = exp1(-
.
(·.-T= !: )
(·.- ro. = it,)
Lz = 21t ood
ro.,

SPECTRUM All-IN-ONE JOURNAL FOR ENGINEERING STUDENTS

 214 '
Ol svs TEI VIS [JNTU-KAKINADA]
~·. . .;;;;;;;;::====IP.!;-M-,,,,-.,--------~D~IG~IT~A~
- ---- · - - - - ... v,,1 ""-•
L~C
~ ..., .. .,
~O~N~T~R~~~mmiimimmiimiiDiim---...
ulc magn1tu c nf= and incrcnscs the ang 1•01 - u v .... , ... -
f <0,1 on any poin t of the spiral can be easil c -·
' y calculated. .
The spira . . . . . c lies in the seco nd or third uadr-J n
Th, l hes w1thm lhc unit circl e in Z-plane. prnv idcd the d'101 ping rnUO 1IJl 4
ratio me r1es .mt I1e' hr,,
cons tant , . . ,. .
f S-pla ue. Conv ersel y, the spira l lies onstu nt dam ping
outsi de the unit circl e in Z-pla ne, when th 0
h i: ou1t h quad rant of S-pk mc. c c -
,aJ The locus of cons tant dam ping ratio for 0.8 and I lying in the Z-pl ani: is a~ sho,v
diffe rent values i.e., ~ = 0, 0.2. 0.4, 0 -6 · n
in figur e ii).

Im

Z-plane

,,
/(

_ J_ _1_ --1. -1-1 .,1-: :..-- -..a. .-Re

-1.0 - 0.5 0 0.5
Figure Iii): Constan.t Damping Ratio Loci in the Z-pla
In the S-pl ane, the cons tant dam ping ratio
ne .
loci is usua lly norm al to the cons tant W
In the Z-pl ane, inter secti on of cons tant w 0 loci as depi cted -
~ fi~ e (iii) (~~:
loci with cons tant damp ing ratio spira ls occu
n . rs at right angl es as depi cted ID figure (m)
,
(b).
. jro

S-plane

Cons tant damp ing ratio loci

'\..---" G""· Cons tant Ci\, loci

(a) Orthogonality of Constint Damping-Ratio Loci

Im
.
and Constant ro ·Loci in S-plane
'

Z-plane

°"=0.6f
(I)

------ °"= o.4 -t,. (I)

s=O
s=o.2
s=o.4
s=o.6
s=o.s
°'1= 0.2 ro,
- 2

s=l
(bl Constant Damping Ratio Loci and Constant
co. Loci in Z-plane
Figura (iii)

WARNING: Xerox/Photocopying of .this book is a CRIM!NAL act. I

Anyone fourid guilty is LIABLE to face LEGAL proceedings. I

II
uNIT-5 (Design of Discrete-Time Control Systems By Conventional Methods) 215
Q21. Write about the general r~les for constructing Root Loci. Sep.-20, S.t-2, Q6(a)

OR
State the rules for the construction of root locl of a sampled data control system.

~n•= Aprll/May-17, S.t-2, Q6(b)

The rules to be followed for the construction of root loci of a sampled data control system are as follows,
Rule t: The nwnber of root loci, N is equal to the number of poles, p of a open loop transfer function.
Rule 2: Root loci always starts with poles and terminates through zeroes. Hence, poles are known as starting points and zeroes
are known as terminating points. · ·
Rule 3: The locus made by the roots must be symmetrical about the real axis of s-plane.
Rule 4: A point on real axis will alWjlYS lie on the root locus to the left hand side of an odd nwnbered poles and zeroes. Hence,
mark the poles and zeroes and poles on the real axis.

Rule S: Angles ~ade by asymptotes(n -m) are given by,

± 180°[2q + 1]
$a~ (n-m)

Where, q = 0,1 ,2, ....... , (n-m-1).

Rule 6: All asymptotes intersect with the real axis at a point called centroid is given by,

.d · (Sum of real parts of poles' )-(Sum of real .parts of zeroes)

Centro1 , cr = ----=--=--~--=-_..:_~__:___:__ _.L1 . -~----:....
A Number of poles(n)- Number of zeros(m)

L Poles of GR( s) - L Zero~s of Gil( s)

= n-m
Rule 7: The breakaway points and break-in points are the roots of the equation,
dk .
ds =O

The angle at which root locus branches breakaway point is as given as,
180°
·+=±- r
Where,
r = nwnber of branches approaching or leaving the break poiJ\t.
The point of intersection of the root locus branches with the un~t_circle and the critical value ofk is found using Jury's
stability criterion. After det~ination of critical value of k, the root loci 1s drawn.
Q22. The open loop transfer function of a unity feedback dlgltal control system is given as,
k(z + O.S)(z + 0.2) .
G(z) = (z-1)(z-0.5) .
'sketch the root loci of the system for O< k < ""· Indicate all important Information on root loci.
Ans: (Model Paper-II, Q10 I May-10, Set-2, Q2)

Given open loop transfer function,

k(z + 0.5)(z + 0.2)
G(z) = (z-l)(z-0.5)

Step 1: Locate Poles and Zeros

Loc~ting poles and zeros of open loop system in complex plane,
Locating zeros
(z + 0.5) (z + 0.2) = 0
z =-0.5 and z =...:. 0.2
z = - 0.5, - 0.2

. 'SPECTRUM All-IN-ONE JOURNAL FOR ENGINEERING STUDENTS

l6
DIGIT AL CONTROL SYSTEMS [JNTU-KAKINAP1'J

jv

joo

-4 ·_ 3 - 2 - I 0 ! 2 3 4 5

Figure (1): Pole-zero Plot

Step 3: Locating Asymptotes
Number asymptotes = n - m = -2 - 2 = 0
Therefore, we have no asymptotes ,
Step 4: Determination of Break-away and Break-in Points
The characteristics equation of the system is,
k(z + 0.S)(z + 0.2)
G(z) = (z - lXz -0.5) ... (A)

The expression for calculating break away and break-in point is,

k =~ A(z)
B(z)
... (B)
From equation (A), we have,
A(z)= (z - l) (z - 0.5)
B(z)= (z + 0.5) (z + 0.2)
On substituting equation (C) into equation (B), we get, ... (( )
2 '
k = -(z -0.5Xz -l) = - (z - z -0.5z +0.5)
(z + 0.5Xz + 0.2) . z 2 + 0.2z + 0.5z + 0.1
2
-(z - l.5 z + 0.5)
k= 2 ... (1)
z +0.7z+0.l
dk
-=0
dz

WARNING: Xerox/~hotocopying of this book is a CRIMINAL act. Anyone found guilty is LIABLE to face LEGAL proceedings.
UNIT-6 (Design of Discrete -Time Control Systems By Convention al Methods)
-~~---""-:""~---------...:..:::=.:.::...:;__:~==:.:::..:.:~~--------
217
Differentiating equation (I) with respect to z.
2
dk d (-z + l.5z-0.5)
dz = dz z 2 +0.7z+0.l

2 2
(z +0.7z+O.l ) dd (-z +l.5z-0.5) - (- z 2 +l.5z-0.5) ~(z 2 +0.7z+0. I)
= z dz
2 2
(z +0.7z +0. 1)
2 2
dk _ (z +0.7z+0.IX -2z + l.5) - (- z +l.5z -0.5X2z +0.7)
dz - (z 2 +0.7z +0.1)2

dk ·
- =0
dz

(zl + 0.7z + 0. 1)(- 2z + 1.5) - (-z2 + l.5z - 0.5) (2z + 0.7) =0

- ~ -1 - I .4z2 - 0.2z + I .5z2 + ( 1.5) (0. 7) =+ (0.1) ( 1.5) - [- 2z-l + 3z2 - z - 0. 7 z2 + 1.05 z - 0.35 ) = 0

- 2zl + O. lz2 + 0.85z + 0. 15 - [- 2z1 + 2.3.:-l + 0.05z - 0.35] = 0

- 2zl - O. lz2 + 0.85z + 0. 15 +2..-1 - 2.3.:-l - 0.05z + 0.35 • 0

-:- 2.2.z2 + 0.8z + 0 .5 = 0

2.2.z2 - 0.8z - 0 .5 =0

-b± ✓b 2 -4ac
2a

- (0.8) ± ✓< .:..0.8) -4(2 ..2X-0.5)

2

2(2.2)

-(0.8) ± .Jo.64 + 4.4

= 4 .4

-(0.8) ± 2.24
4.4

· 1.44 -3.04
4 .4 ' 4.4

z = 0.32, - 0.69

Substituting z = 0.32 into equation ( l ),

2
2 )~_-_1._s(~o_
--~<(~0-=-3~ ) +_ o_.5_)
.3_2'-
k = · (0.32)2 + 0. 7(0.32) + 0._l

-(0. l 024 - 0_.48 + 0.5)

=
0. 1024+ 0.224+ 0.1

-0. 1224
0.426k

Ik = - 0.287 I
SPECTRUM All-II-ONE JOURNAL FOR ENGINEERING STUDENTS
21 8 YSTEMS [JNTU-KAKINADA}
DIGITAL CONTROL .S
Substituting z = - 0.69 into equatio -
n ( l ), we get,
k = -((0 .69)2 - l.5( 0.6 9)+ 0.5 )
(-0. 69) 2 +0.7(- 0.6 9)+ 0.l
= -(0. 476 1+ l.03 5+0 .S)
0.4'761-:-0.483 +O. l
-2.0 111
= 0.0931

\ k=- 21. 60 \
For z = 0'. 32, k is real and negativ
e i'. e., k = - 0.287
For z = - 0.69, k is reai and negativ
e i.e., k = - 21.60
Since the gain is negative there is
no break-away and break-in point.
St~p S: Determination .o f Angle of
Departure or Angle of Arrivai
Since there ·are no complex pol~s
are zeros; so, no angle of departure
or !Ingle of arrival occur.
Step 6: Determination of Intersec
tion Point of Root Locus with Ima
ginary Axis
The characteristic equation of the
syst em is,
l + G(z)= 0
i +k r(z+ O.S Xz+ 0.2 )1 =0
l (z-lXz-0.5) j
(z - l) (z.: 0,5) +k [(z + 0.5) (z +
0.2)] =0
z2 - 0.Sz - z + 0.5 +.k [z2 + 0.2z + O.Sz
+ 0.1] =, 0
z2 - l.5z + 0.5 + k [z2 + 0.7z +. 0.1]
=0 •.. (2)
Substituting z = jv into equation (2),
we get, , •
(jv)2 - 1.5 (jv) +O S+ k [(jv)2 + 0.7
(jv) + .0.l] =0
-v2 - 1.5 Jv + 0.5 + k [-v2 + 0.7 Jv+
0.1 ]( -v2 - kv2 + 0.1 k + 0.5) + j (-
1.5 v + 0.7 vk) = 0
Separating real and imaginary part

-v2 -kv 2 +0. lk+ 0.5 =0

• -l.S v+0 .7vk = 0
-v2 (l+k )::.: -0.5 -0. lk
v(-l .S+ 0.7 k) = 0
(0.5 +0. lk)
v2 =~ -- -'- -1. 5+0 .7k =0
(l+ k)
k= -1.5
-0.7

+v 2 =-(0.5 + O. \.
"-- ---lk)
'- k=2 .l4
. . (l+ k)

v2 = 0.5 +0. lk
l+k ... (3)
Put k=2 .14
0.5+0.1(2.14) 0.5+ 0.21 4
v2 = 1+ 2.14 3,14 .
v2 = 0.227
V = ± 0.47
The crossing point of root locus is
± j 0.47. The value k at this cross
point is 2.14. The complete root locu
in graph. s plot is shown

: J:\. WARNING: Xerox/Photocopyillg of this book is a CRIMINAL act. Anyone

found Ql!ilty is LIABLE to face LEGAL proc
aading~.. : ,, '
uNtT-5 '(Design of Discrete - Time Control Systems By Conventional Methods) 219
1-'. _j •: .
I •

i·::~~:-
..J

<~:__: •· ,} -
·--·r----t---t---1-.,.-i· ----t

-·- .
J _ ,· 1--+_
-l- - -'-'-1---t
- '- i -+---1-'-+-.;...+-t-+-t-,--
-l-- -
~ --t-+-1--+----1 --L -

··t-- 1+·-+-,-1---,-iT
1<m]" ·<i-·~i~
- - Tt-,--~+-+-t-'lr-t--t-1-

~ "l--'--'ll=-'~'v ' ··" · ,..J,. - - 16 -, ~

~~~=::1.:~!I ;
. l l
I_

r ···r- - -
. '

r~~-7-.: ·i·····j':. .:.:.,

1~-·-t··:+-+ ,. bi_t~.: .-:.:.·.: ::.•.·~
:. .· ~~~ -~~~~
Q23. The open loop transfer function of a un~ty feedback digital control system'is given as,
K (z '+ 0.. 5) (z + 0.5)
G(z)-
- (z-1)(z 2-z+0.5)
Sketch the root loci of the system for O < K < 00 • Indicate all important information on the root loci.
Ans: (Model Paper-Ill, Q10 I April/May-17, S.t-3, QI)
· Giv.en open loop transfer function,
· · K(z + 0.5)(z + 0.2)
G(z) =
(z- 1)(z 2-z + 0.5)
To,
sketch the root loci of the system·for O<K< oo
Stepl: Locate ,Poles a·nd Zeros
Locating zeros,
(z + 0.5)(z + 0.2) =0
z = - 0.5, - 0.2
:. The numbenof zeros (m) = 2 located at z = - 0.5, -0.2
.
Locatmg /
poles,
(z -1 )(z2- z + 0.5) =0
(z - I )(z-(0.5+)0. 5)) (z - (0.5 - J0.5)) = 0
z = l , 0. 5 + J0.5, 0.5 - J0.5
:. The number of poles, n = 3 located at z =i, 0.5 + J0.5 and 0.5 - J0.5 Hence, these branches starting from
2
"' l , 0.5 ± J0.5 wi ll terminate at z = --0. 5, - 0.2

SPECTRUM ALL-IN-ONE JOURNAL FOR ENGINE~RING STUD~NTS