-- --- -" v r u v A u L A V A W

- Paiia 1.N 6.
#I:
(i)
The block diagram of the system is:
; ..
.. , ;
-;a -:..
. ,
.
- .
.
.
.r
1 ..
.~ ,

-- .a
.
,.'
.. ... > .,-
.
..- * - .
.
-
'
..-+.
- 0
., - '4.
.-,> -
.,:. ,*
k:>p:: . 2 ; $ .:.
V'! ? p.
,-., .

-- z;
...-
dLl.{,,cfia;,:
!;".:'
. . .
i . h
.
. . b ?:
Q\! - 2 ;.<:
. .x'l:kt.;
:-. :,.<-;.
- . .
; 91 I :!I

:*-, ;;
C.A

.:-:..
.;;<,f.,...:<
. ... .. . .
i,*,'d a .
f . ' 1.. a - - Y
'
1 -
,
-.*.ti:
.&10-1* 100
@! 100 + s l s(l + 025s)
-'

:! ..
XF I:
*
& =lqM

and T(s)= as>

- -- 400 - 400 -
1+ G(s)H(s) 400 + S(S+ 4) (s+ 100) s3 + 10qS2+ 400s + 400
1 -
(iii) The p z diagram of the original inon-approldmated) GH is sgown below:

The OLTF pole a t s= -100 ( due to the torque-motor) is xZ5 away than the
pole at s=- 4 ( due to the servoactuator). Furthe re, for K=400, the CLTF
poles originating from the OLTF poles at s=O & s= will be dominant (since
the correspoding root locus branches will veer to the right ) .Thus, we may
=T
neglect the OLTF pole a t s=-100 as far as its effect on the transient response
of the closed-loop system is concerned. ( A / o f i f fi ~5 A h L2)
(iv) The GH function may be approximated as

-
ie 2(0,=4 =2rakndssec ;6 = 1 or the Closed-loop system
and on=&
is Critically damped -

The CE is : 2s2+ 2 s + 1 + s3-<xs2-s,+ a=O

or s 3 + ( 2 - a ) s 2 + s + (1+ a l = O
 Page 2 of 6
The R H array is

- .
For stability, we must have 2 a > 0 A =*l-la/> 0 and 1+ a > 0
These conditions imply a < 2 ,7a < 0.5 and a z - 1mspectively
(i) Since a > 0 , the range of a for stabaty is 0 < a < 0.5
(ii) If a = 0.6 ,a row of zeros results irr the REI array :The auxiliary equation
h 1.5~2+ 1.5 = 0 or s = 2 jl , ie o,,, = 1radianlsec

(B) (i)The OLTF' is G(s) = X (s+2) 1 s (e+l) (s+3)

n = 3 , m =I n-m=2 -- 2 asymptotes at A 90" and
O,= [(O-13)-(-2)]/2= -1

The Breakout Point between s=O and s=-1 can be found by trial & error
by locating the value of s ,between 0 & -1 ,for which the magnitude of K
1 K I = I s (s+l)(s+3)/ (s+2) I ,reaches a maximum :
Using trial values,

-0.3 i Maximum here

-0.4 j indicates approximate i
-0.5 : location of BP
iat s =-0.5
-0.6 0.41 ..----...-......--.-...-.-...*..-..--.
-0.7 0.37
t

The root locus is shown below: I

*90° asymptotes
at s = -1

-3
V
4%
-
C

BP Q
.-
= -0.5
-

(ii)If the point s l = 4.588 +j0.658 is to lie on &e root-locus, the angle criterion
must (Lzeros) - C (Lpoles) = * 180" must be satisfied at the point.
ie Referring to the p z diagram shown , we must have :
 Page 3 of 6
6 = B3 - 8 - B2 - O4 = - 1800 where the angles are as shown.

o., , 3 1 , 8 O
I
: * +
3i <-

4, 57.7 c. ......-
....
.
.--::-• ,p,.-.-----.
:
%
...-• ;
-
e3: 3s00Q fi
a...

- .....- ...-......-....-I&
u.. Tea *
...-*$ B i t A'
~1 :
4* - .

:. The angle criterionis satided and s lies on the root locus.

Using the magnitude criterion : K (63 s,) = A B.C I D 11.0995
(i&
/
Settling time T, = 4 /Go, = 440.588 = 6.8027 sec

(i)Reducing the complete system (& with D(s) = O), the

TR= Y* /R is
K
5 , S(S +1) - K
R l + L ( l + ~s*+s(l+K&kK ~ ~ )
s(s+l) # .
I

The characteristic eq~ation(CE) is :s2 + (1+ KKPs + K = 0

Since the closed-loop pole locations are complex ,we know that the
sqTtem is underdamped and the roots of the CE are :
- 50, t jod = -con +jwnJ(lx2) = - 1.61 t j 3.14
Thus, Pn= 1.61 and o d = onJ(1*) = 3.14

and w, = 4 [(<wn)2 + = = 3.53 K = (on)2112.48

i
Also, 1+ KKl = 250, = 2 (1.61) = 3.22 or KK, = 2.22
:. Kl=2.22/12.46=0.178
( { = Sw, 1 w, =1.61/3.53= 0.456)

(ii) For the unity feedback system, G(s) = K 1 s(s+l+EEl) + TYPE 1

system
The 'position' error constant is Hp = limG(s) = =
s-0
The trelocity' e m r constant i s = lim s G(s) = K/( l+KK1
s4 0
& = 12.46/[1+12.46(0.178)]= 3.872
For R(s) = Us + Us?, e, = I./ (1+ I$) + 1 K,, = 0 + Y3.872 = 0.258
 Page 4 of 6

(iv) For T = TR = Y B ST : K ~= &m a T / , q .

#4 : (i) e,, (unit ramp)= 1 I 12 corresponds to I(y = 12 = & = W40 .: K = 480 - -

.: 'Kt,(dB) = 20 log 12 = 21.6 dB

at col = 5.85 rakandsec is

(ii)The phase shift of G(jo)
6(01)= - 90"- tan-1(0~14)- tam1 (oln0) = - 90"- 55O.64- 30°.33
- 176"
I
-
ie Phase Margin 0 = 180°- 17C =

(iii) ..
p l u plot I
.nd G(a
j),=-
- 1402+j(40*wos)
Setting ImG(jo)= 0 yields phase crossover
o, = 440 = 6.32 radiandsec
. and A = 4801(14) (40)= 0.857 or GM = 1 / 0 . 8 5 7 = m
\ G(jo ,

(or 1.34 dB)

 Page 5 of 6
#5 : (A)
(i)For the uncompensated system I$,= 14040 = 7
emu.= l/(l+Wg,b
b
and -iB= 0.125
.- .
,.,.'.. : .> , . ,
- ., ..;: . z . 3'r.
, .
9 ,. . sp,
, . . .: 1 ;
. .=. I/(1+5).
, .

p We require en,, = epU A0 = 1i.80

constant & = (zc1p> I$,= 7zc)pc .
. ,;.w e the new error-
. ... . :. .

(B) (i)The p-z diagram of the uncompensated system is shown with the
- dominant pole positions indicated by A [only the upper pole is rho= here 1

From the magnitude criterion, 4.

K = a.b.c = \/(1.22+22).4(2.82+22) .\/(4.82+22)~(2.3324)(3.4409) (5.2) = 41.73
1
The damping ratio 5 = cos59".03= 0.515 , Settling time T,=411.2 = 3.33 sec

(ii)To reduce T, by a factor of 3 while keeping f unchanged ,we need to have

dominant poles at s, = - (4ll.11)*j (411.11) tan 59O.03 = - 3.6 i j6

Referring to the pz diagram above, the angle LG(s,) = - [ $, + $, +

where $, , $* and are the angles shown above.
From the diagram, = 180" - tan-W3.6) = 121°, 4 = tan-l(WO.4)= 86O.2
and (0 tan-l(612.4)= 68O.2 and .-.LG(sl) = - 275O.4 * -180"
Therefore ,the exess (positive) angle to be supplied by the compensator
 Page 6 of 6

must be 275O.4 -180" = 8 5 O . 4 . For a PD compensator, this excess angle is

supplied by a suitably located zero (at s = - zc).In fibs case, zccan be found
5 - -
from the equation : tan(l8O0- 959.4) = 6 / (3.6 zJ or 2, = 3.6 0.567 = 3.03 -
The p-zdiagram of the compensated system is shown below:

With the compensator added the system - - gain a t s=sl is found from the
magnitude criterion as gl&= ABC I D : -
l<,K = d(3.62+62).J(0.41+62).4(2.42+62)14(0.572+62)
=(6.997)(6.013)(6.462)16.027 = 271.9 16.027 = 45.11
The compensator gain is therefore Kc = 45.11 / 41.73 =
(iii) The root-locus of the compensated system is sketched below:
n - m = 2 .............i 90" asympto&?sat o = (-lo+ 3.03) I2 = - 3.5
I

Now ,the OLTF zero at s= -3.63 is als; a CLTF zero and the presence of
the CLTF pole at s=-2.8 provides an approximate p z cancellation {(&
the residue for the (s+2.8)pole term in the inverse Laplace transform of
Y(s) is likely to be small ) Thus the pole-pair at - 3.6 * j6 is likely to be
dominant.