__,.!.

.!
\-- H "'zu LINEAR SYSTEM SIMULATOR
z'©
!~ J
0
_1. OBJECT
0 To study the time response: of a varic:ty of simulated linear sys1ems and 10 corrcl:lle 1he
i
studies with theoretical results.
2. EQUIPMENT DESCRIPTION
The present sci up has been designed 10 provide a convenient inc:ans-Of verification or
\ "'5 the transient response studies of linear syslcms. Special anention has been givCn at ·every stage
to make the system operationully simple and ,traightfo?"Ward without compromising tl~c
participatjon of the sludcnt in choosing/wiring I.he lest signal, system conliguration, dist.urbuncC'
points etC. Block diagram approach has been employed a'S\.a standard lcc.hnique, with Fig. I
' depicting the features available to the user. A description of the: v.!rious sub-sys1cms 1s
"'X presented next, an underslanding of which is necessary for both -· conducting the routine
..f o 0---11,, experiments anJ designing more advanced ones.
2.1 Signal Sources
w
:r ":'; There a.re three built-in sources in the unit with the following specific2tions:
;:
..
:,·
I -~ o-- ·- ~ i "
0: (a) Square Wave: frequency 40-90 Hz (variable)
..,
0
\
P·P :impli1ude 0-2 Voll (variable)
I
"" Tri~gular Frequency 40-90 Hz (variable)
oci
(b)
$
I
' 0. p·p amwlitude Oc2 Volt (variable)
t,. '" (c) Trigger Fr(..--quency 40-90 Hz (variable)
,· ::=
I
amplitude ±5 Volt (approx.)
I
All the above sourcCs are derived from the s~me .b:1sic c1rcui1 an<l are 1hrrtfor1.:
I synchronized. They are uncalibrated bolh in frequency and ampli1udc, are zero balanrcd :111,I
o---0; \. grounded 10 a common ~anh.
2.2 Building Blocks

·
l ·A dynamic system of thi: desired confi&ura1ion ma; be: coustiucted b°y a. ~11i1abk
-@
interConnection of the ba.siC blocks available. To avoid'uMcccssary complicalions, lhesc block:-;
arc till· pre-wired; they nil have foo:d tram fer functions and huridle !'lignals referred lo :;i common
I "'u

·ground 1 !hese urc lhcrcfo'l'c: ·e:usily identified with !he flllJlilior i,iclorinl reprcscn1·a1iun of:. block
diagr:irn, with one diffrre111.:c:· viz., all blocks in this sys1cm have lransfrr function with :1
· ·ne&ative sizn. Following 1ypes of blocks have: been included in tl 1e system
..
. (:.i) Error Di:1ectvr-Cum-Cui11: The bloi:k h3s 1.hree inputs (c 1,c 2,eJ) :.ind one output (e0) which
z are related by tht expression ·
.2·
:,:
· a: ';t eo = -Klc1+t:rt-i:J)
"' 0
d "'...
"
~
I
- - - -, .' .
1
 IJ-
-~·--·- -- -·~~=~r-"-·" - -·-- -j .. ··---
'
ri
y
· · · · · · d, · O I JO by a 1cn-1uno po1cn11orndcr
wl1cre K ·;s lhe gam. The value ofK may be vane ,ram . 0 . . h .
UJ having a ca.libra1ed dial. Thi• block may .be used as a simple error dclecior_ (wol_ one signal
·V h
·z inverted) in a mul1iple loop ,ys1Cm or simply.as an :,dder. A cal.ibrated gam w,l negalivc
"'a:
"'UJ
a: 0 sign is·prer.cnt in all cases.- 1
::, 0 nd
:;; <I (b) Integrator: The· inlcgralor blpck has an approximale 1ransfer func1ion of lhe fonn -K ,Is ' .
0 .· is used in sirnulaling ,}'PC~t sys1Cms having a pole at 1hc o_rigin. .
0
{c) Time. Co11Mn11t: The two time conslanl blocks in the system have transf~~ ftm~ti~ns ~ th_e
form -K /(sT~J) each. The .second blocks has a XS option which results on• higher gain of
2
0
wa:· necessary.
;'! ::- ... (d) Dis1rrrbn11ce Adder: Thi• is a iwo inpul (e 1,c1), one ou1pu1 (c0) block having a defining
"'z :i: -
•· equal ion of lhe form
,, :i:.
0-'
0 u a.
u z
UJ
c0 = -(c 1+e 2)
:. The block can have applications similar lo lhc.error dc1cclor.
¢
;::
I
I 1: ...z .( (c) Uncommi11cd Amplifier: While complcling 1hc feedback path, one mighl need 10 invcn lhc
Ci .
...
0
signal so 11\.11 the rcsulling system is a ncga1ivc feedback system .. The uncommiucd
l. "'0Z · - . amp Ii ricr is used specifically for this purpose.
·..i
n
V
w
%
;::
o--- .,
, '-0~
I

-'

2.3 Power Supply

t~e sci-up.hos an internal Jcl2V JC rc~ulalcd supply .which is pcnnancn1ly conncc1cd 10
•II lho circ_uils. The power supply and all lhc circuils arc short circuil pro1ec1cd and will no, gel
damaged u-nd~r nonnal ,opctation. _No external d.c. supply should be connCc1cd 10 ·,he unil.
I
1' I
W(th the heir, of the above sub-sysrems a very large -v ariety of simulalcd syslcms can be
Q Q~
I
l:A a:
I
c?nstructcd~-OCiides singlc-inpU.t Si'181c-autput syslcms this would include multi-inpu1 syMem5.
........
0
systems ,~ilh ·rc_i:d-_fon.\:ilr? po1th arld ·syslems wit_h minor loops as well. Considering 1h_e 1irn~
con~lrnins in lh_c ·l~bof;11ory class, a limited number of hasic experiments have been proposed in
a: 1h~ follqwing pages which arc compatible with lhc· average c~ursc content or the subject.
"'...
UJ
I
I
,\_ ·· goo(J qual_iJy dual 1racc · mcasuring · oscilloscopc is the only external cquipmcnl
z I
s o----0
I
required for thC_·cxpcrimcnt.
3. BACl<GROUNo·suMMARY
A 1horo11gh. unders1and1ng of transient response analysis is a r-rc-rcquisitc for 1his
---0
"' ·cxpcrimcnl. Ailhough tJj(; :opit IS covered in great dclail .in all the ICXI books on ou1oma1ic
,onlrol, a.simplilicd ·dcs~ription of the pOllions rclcvanl to lhe _experiment is 1;;cscn1cd below.
I
a:
...u
0
_3.1 First O_rder Syslems
A fi·rst orJcr ·systcm is eh.?ractcriscd by one pole and/or .i zero. A pure integra1or and a
w ..
single 1i111c co11s1a111, having 1rai1srcr funclion of the forni K/s and K(sT +I), arc 1hc 1wo
or
0 "
a, commonly s1udicd rcprcscn1a1ives 1his class of sysieins,· Many. thermal systems and electrical
0 :i:
er::, sys1cms wi!h ·R-C/L-R clcmcnl arC 1hc examples ·or rirSJ order $ystcms.
~v
'
t+
r
'
,a . •
- ~--- ·- - ··-
L .- ..... --
 I'
L-- l Unil step r.esponse of the systems are computed ~s follows and 3re sh9wn in fig . 2(a ).
-tfJ--· ~· I
Y(t) C(\)l / . .. JfC(s)/ R(s) = G(s) = K/s lh en for R(s) = 1/s
1 C(s) = K/s
2
, and c(1j=K1 (I)
_Aga in ifG(s) = K/(sT+I) then wilh R (s) = 1/s,
. C(s)=KJs(sT+l), and _c (t)=K(l ;e"') : (2)
Tim~ constant of the system is defined from Eq. (2) al t=T, which gives
1
c(T) = K9.J -c ) = 0.632 KG(s)
' "··c
C(IL This is an impo,.;ant characteristics of' the system which j5 also ~efined in terms of the sl9pe of '
. K the response curve at t::::110. . · . . ·
•632K F~r a proper viewin8 on 2 · cRo, the step input may be repl~ced by a square w:ive or
J
sufficiently low frequency (IQ allow c(I) in Eq. (2) 10 reach up10 99'%', of ils final value). This is
1 . shown in Fig. 2(b). With a peak-10-peak ompli1udc of uni1y, lhe · Eqs. (I) and (2) remain
i ~ unchanged . Jt m;1y further be seen th:it if th e square wa.ve js of frequ~nc y f and peak-lo-pc3.k
I' t amplitude 1 V, the pe'2k -lo-peak amplitude of the .triangular· wave at the output of the pure
I' integrator is gi vcn as K/4f: ·
~1 i'
:•
FIG. 2 (a) UNl1 SlEP RESPONSE OF FIRST ORDER 7RANSF 3.2 Second Order Systems
FUNCltONS. These sys1ems are characlc"rjsed b'y twO po1es- and upro two zeros. for the purpos e of
rI '

transient response: s1udic~. zeros an: usually no·1 considered primarily to· have simplic i1y in
caku/a1io11s an d also because 1he zeros do'Cho1~·ffee1 the inte~al modes of the sy stems. A great
deal of anal y 1ica l resul1s regarding second order systems are available in the 1·e xt books: This
forms 1he basis of studying higher order syst em; many of which can be approx'fma1ed to second
order.
: 1: A sc:i:onJ ordc-r system is repres ented in the standard fo~ as
G(s) = ---~ '- -
s2 + 2Swns+w/
__)Dc:J=. L ~·. I
_ . where ti is called the damping ratio a.nd wn thl! :mdampcd. n:i.tural 6-cquency. Depc~djng upon
i
:
.
.

.
.-:s.
K'
l ••
. .·
r-.1. l""A ·
_rtL...!:,-- cs the valul! o( 0 , the poles of the sysacm may be real, repeated or complex corijuga_1c . which is
reflected in the natur-e ori1 ... step response. Results obtained for va'r ious cases 3l'e : ·
(a) U111/atla111p e ,I L' llU (0 < 6 < l)
i' ori
·· K
~ . . e-•w,, , . .
I
ll!
-I~
. /1 /7
---1 '-' L... .- ·- - . . .J O \ ; _ C(l) -:: !- r ==Sm(wdl-t-l~ln ---) (3)
51
.
+ 1
. . . . ,1 (1 - 0 2 1 Ii .
•.
/: where W i,1 =w 11 J{j___62} is 1efT11td as !he damped natw ~I .frequency. A sketch of th e i.;nit
Slt'p respo nse for various values of 6 is ~vailabl e in !he 1ext books .
'i FIG. 2 (b) SQUARE WAYE RESPONSE OF FIRST
ORDER TRANSFER FUNCllONS,
,-
.,
,a . -. - --
I
 (b) C,ritlcnlly ,lnmpctl cn.tc ( 6•·1)
c(t) c I - ci.>R, (J'+ w n 1) (4)
(c) Overdomped case (O>J)
c(I) = It--"'-"-- _e_ - _e_
2Jc&_2- ._1J [ ,,-,,, -~•] s2
where s 1 = ( 6 + JcS'~1)) o," and s 2 = ( 6 - J(</~-ij) w" ·
J.J c ·1~sed loop SHlems
( Closed loop or feedback systems involve a 'measurcmcnl' of lhc output of rhc system
.ind gencralion of _con1i-0J signals·, which arc based on 'decision making" under the influen!=C of
Ji· a 'Command' ~r 'r~fefenCe', and the measured valu~ of oulpul (Fig. 3). Such systems arc of
~. DECISION
MAKING SYSTEM ·1 grcal inlcrC.St . 1<,- con1ro;1 cn"g{nccrs du·c io features like automatic corrcclion, dislurbancc
+1:: "''
rejection, inimuni1y to "noise-and parameter ·variation c1c. A s111dy of the pcrfonnancc of closed
I.
loop syStcm is the! basiC objcclivC: of lhis experiment: 1t mily be e.is.ily appreciated 1hat although
REFERENCE
OR - ·-
C(s) a
the ma·1hcm4!tic~I dfs~ripiion of cTOSC9 -io;p system is no different from thal discussed in
sections 3.J and J.2, _the fa.cl that variati911 off~rward path gain shifls the pole loC:alion, changes
COMMAND
I !he silunlion draslic:illy and makel di~cl CompulalioO of respon:-e of little Value. Referring to
rig. J, 1!1e closed loop lrnnsfcr funclion Ior different open loop functions arc shown below :
(i) For G(s) = ~ . CC•l = -- ~----.!_.
s · R(~) _s :f· K : +I .
which gi:vcS a rcsPorise.similar to. Eq. (2) with time con~1aut dccr~ning as K increases.
(ii) For G(s) =---~-, =·__K__ =~:_15_!_..
· sT+-1 R(s) .s T+K+l sT/(l+K)+I
FIG. 3 BLOCK DIAGRAM OF A -CLOSED LOOP SYSTEM. whiCh has a s~cp response similar in nature ~s obtain:d above.
(iii) For G(s) = ___ K_-_ , C(s) = - - - _K__ = _
_ s(sT+I) R(.s) s2T+s+K s 2 ·+·s/T .+K/.T
which gives response simi!a( .10 EqS. (J), (4) or (5_) dep_i:nding upon lhc value of K .
. Thus·thc Tl!spoi1sc Or a cl'~sed loop system can be altered by varying Jh~ open loop gain
K ?nd as a ccmscqu_cncc ·ii should be possible to chose K · fo obtain. a •Suitable' performance.
This leads lo i_he -concCpl of _perfonnancc cha;ractcristics ~s · dcfmcd on the stcP response o.f an ·
111
_Hl~rdampcd second order system in Fig. 4. It musl be_noted that these specifications arc not
rc_S1nc1ed lo sccc,nd order s~sterns;. alihough 1he mathematical cxpressionsidefinitions given
bclow_arc valid and compu1a11~nally practica~I~ for second order syslcms Only.
(i ) D;'"f ,;,;,,: 1,, is defined as lhe lime needed for lhe. response 10 reach 50% of 1hc fmal
value. ·
'°· ...
I
 I
'! ''·
i
~i j
l (ii) Riu time, I,, is Iii, :imc l:iken for lhe response 10 reach I00~:, of 1he final value for 1hc lirsl
lime, Ti,is is given by
t, =--·,where
n -Jl
D= 1a11
_,fH2i
- --
wd Ii
..\ (iii) Peak time,~. is the rime Jakcn for lhe response 10 reach ihe first peak of the ovcrsho, 11 :md
is given by
n
Ip ::,1(1 -o i' 2
,! 0
(ivJ Mn.ri11111111 01'f!r.\ /Joo r. Mr, is de fined by
Mp , l_l_~_-:._~L"') x I 00%
e(a:J
''
.

,
Mp
-........,

:
I

/ 5'/, TOLERENCE BAND·

/J_ ess
1,., 11s value i11dica1cs.1he rcJarive s1abiJi1y of lhc system.
Seitli11g time, 1)1 is the time required b.y the sys 1cm rcsp ori:,e 10 reach and sray within a
prescribed 10Jerancc baud which i~ usu ally taken ;is ±2% or ±5%. An appro.xir~1a1c
f - - -___;- I · L ~- ca/cu/J1io11 bascJ on lhc envelops of1hc rcsponsl' for~ low d~mping rntio sys1cm yidds,
I\/_~ I '
I tl(± 5% loleri.lncc b;rnd) = 3/0(J>n
I
I , ll(± 2%, lol cr:rncc banJ) = 4/60> 11
0.5 .I
I
V
I J\nnlhcr imponar.1 ch;ir:iclcrist ics of ;i closed loop system is the src:nly sl:th: error. c~ . for
u11i1y feedback S)'$1cms, _c_,~ is defined as 1
;:i , /1
i I [ I c.,, Jim c(IJ Jim (r(1)-c(1))
td 1, tp I 15 t...,, Q;> 1..... w
1·
' A !impler way to c;:i/c111.ltc steady slJle error wi1hou1 ;:clualJy compuling the rime
n::;ponse is availn~Je in lhc complex frequency domuin . Applicaiion oflhc.final value theorem
FIG.4 STEP RESPONSE OF AN UNDERDAMPED SYSTEM. , of Laplace Transfom1 10 uni ty feedback S)'jtCm gives,
c., , Jim c(I) = lim ,; E(s) 2 Jim 2!.l.l:./..
1--,,r S•lli Htll+G(s)
S1i:.1d,Y Mole C'nor m~ be ob t:.i inrd for v:.irious inputs (sl cp, ramp, par.i.holic) :ind sysrems or
j,
"• 'ar:1ms lype. m.1111bc1s (i1wnhe, of poles at origin). A s11mm:iry of lhc rcsulls· of rhc .tbove
r.i lc:t{l:11io11s is gi v1:n in 1hr: 1.ihlc below:
T)1 1cno _-1· Step inµu: Ralllp lnpu1 /' _!'arabolic Jnpu1
i 0 . 1/l <K - 1·
I!,
1'1
i::
I _ __2_

I
, 0

U I~
o _J 1/K
ll:
Ii
U/
J.:;.
\
\ '1- ,0 •
 -,i
To facilitale lhe calcula1ions, error cocf/icicnls are defined
Pcisilion error coefficient ." K = lim G(s)
0
,-o

as l
.
Vcloci1 y error cocfricictll, K..,
-
= ~o
Jim sG(~)
Accclcralion crtor coefficient, K, = lim s 2G(s)
. _ _· . , · _ 1-tO ·
The steady slate errors may ·be obiained in terms of these error cocfficienls.
4. EXPERIMENTAL
. WORK
.
Som·• of the experiments which may be performed on the .unil arc dcl~ilcd below, The
· normal amount of time available in the laboratory has·becn kepi in mind while listing 1hesc
cxpcrimc111s. Suggc~1io11s have also bcCn included for .additiorial labora1ory work as a second
lcl'cl special problem or project woi·k. '
4. 1' Open Loop· Responso .
As a first step, lhe open loop 'transfer function of all the blocks viz. integralor, lime
constant, uncommilled amplifier and error de1cc1oriaddcrs arc to be dctennined experimen1ally:·-- - · -
All measurcmenls arc done wilh lhc help of a measuring oscilloscope and the signal source is
lhc built~in square· wave generator in each case . .furthcf, to gel · ? properly synchronizarl •
wal'eforrn, especially for small values of signal, ii will lie convenient to use the built-in trigger
source keeping lhe ·CRO in 'cxtem,1} _lriggering' mode, A double beam CRO for the
simultanco11~ vi<;win~ of_input ~nd ·output is recomme_ndCd. ·
(n) Error Dett!C'/or Cum ·Vn'rinhlt! Gnt'n
, _Ap.ply J'OOmV. sq11are wave signal to n;,y of lhc lhrcc inputs.
• Sci the gain sC11ing pOtenpomet_cr to I0.0
M_casurc. lhc p·p· o_~tp~t. voll.'!gc _and not~ it .fil&.1!· Cal~ulatc the gain. This is the
r"!}a=<imum.vah.lc of!i"afn pOSSi_blc for this block.
.Repeal f~r the ~iher two inpulS one by one.
Write ihe equation o_f this block and verify by connecting lhe signal 10 all 11,ree
i;,puJs.
I
i. (/J) Di.<f11rba11i;c ,1,/,{a
This..sccrion mfty he tcs.tcd cxoclly in the S;1111c milnncr os ~n) except that there arc Only
l\\·o i_npu1s a;1d_there is no _gain Sc_ni:ng potcntionicler.
(c) U11co11t111iu~d Amplifier
·Apply a ·1 volt p•p ~quarc. wave input.
J. • ··11o1eitsurc \ioc p-p outpu_l voliage and note its sig•n.
6
 ___ __
. . -- - -·· ~- ...,
' --
C(s)
(1/) /111egrator

.••,,, ,,,, •• ""'°" •,,... L. ,., , .•.

• App!y a ·J voJr P.·P square wave foi,u, of known frequency (frequency meast,ired by
I
rhe CRO).
Me3su~e the p-p ou·1pu1 -volr-dge-of"rhC- friangUlar·:-,,:1"."e_.a~~ als~ !}Ole its phase.
'I CaJCU!a1e the gain constant K of illfegrator as discussCd in secriori J. I and wri1e 1he
iransfer funclion of lhis block. ·
'/. C(si (e) Time Co1Jstn111s
Iii. Apply a 100 mV p-p square wave of known frequency (measured by rhe CRci). Far
this experiment, 1hc frequency should _be selected towards the lower end ro ensure
1ha1 steady state is nearly reached.
i:i • Find on lhe lrace the time 1=T a1 which the response re~ches 63:2¾. This is lhe lime.
constant.
Find on -1he trace 1i1e sready state value of lhe response. The .value Of K is &iven by
I rhe ratio ofp-p sleady-s1ate output lo the p-p input amplitude.
I,
I FIG. 5 CLOSED LOOP OPTIONS FOR FIRS·T ORDER SYSTEMS • Wri1e trarisfer function of the block as discussed in section 3. J
The wavcfom1 on l~e CRO may_be !raced On a tracing paper for analy~is.
IiHi
4.2 Closed Loop Response • First or_der system
11: Two forms of firs I order closed loop system, as sho_wn in Fig. "5 are possible. Make
!; proper conncc1ions for the co~figufat ion chosen.
}i,,. -
,.
Apply a J volt p-p square wave inpur and trace lhe output wav~fOnn on- a lr;:i~ing
pclper for K = 0.5, J .0, 1.5, ... Calculate the lime cons1an1 in each case· and compaie ·
.fl wirh 1hc:ore·1ical results of sectio~ J.J . ·
• Also calculate the slcady_ state errors f~r lhc above «:3sCs and compare wi1h
fi 1Jii:ore1ical results.
Jf Jht: ·opcn looJl [rans.fer func: tion in lhe chosen conlig~ration .was of [ype-1 , rhc
steady stale error abov~ wo uld be zero for a srep inpul. Jo .find steady. slate error fo j- J
ramp inpu1, apply a l vo/1 p-p lri::in g u/ar W3\"C input•. Keeping· the CRO in X-Y
mock, connect s7s1cr11 inpul lo the X input and rhe sys"1eri1 ourput [O the .Y input. A
1racc as shown in F!g . G wjjJ be seen on the C~O in which the venicaf displacemem
between !he two curves is the steady sr--Ji°e e1Tor.
• Repea1 the measurcmc::nts for s1e";.idy~sta1c error_ for different. values of. K and
compare 1he results with theoretical calculations.
I •~
/ 4.3 Closed Loop RcsponSe . Second Order System ·
FIG. 6 STEADY-STt'.TE ERROR FOR RAMP INPUT. !, • Choose anti wire a suiloble second ord e r syskm configLJra1ion from Fig. 7.
I
II
I
I
l
\ - --· 4 •
 \·
• ApJ)ly a l vol( P·P sq~are wave inpu~ and tr,acc lhc ourpul wavcfom~ on a lracine
pnpcr for diffcrcn1 values of K. Obi am. peak pcrccnf ovcrshoor, scu/ing rime. rrs~
. K1 C(s) . rime nnd_s1c~dy stale crror(rom !he lracmg, and ca/cula1c 6 and w11 • Compare wiih
thcorc_rical rcsu.J1s.
ST, +1
\ rn ihc ·,casc of n· lyptl...m!mJ,_ob1ain_steady....s1are-crror-10-r;imp inpu1 as suggcsicd
.,j in s_cctio~ 4.2.
·4.4 Closed Loa'p Response· Third Order System
Connecl lhe third order 1y1lc:m shown in Fig. 8..
I , Skc!ch 1he Root Locus dfogr~":'! and, for 'O=O.~, caJcUJa1e lhc value of Kand m
j: Rcpc-al for 0=0 (sus1aincd OSCillations case). ft •
I .
Ki C (s) Verify 1hc above by selling !he gain polcnt!omclcr at" ca!culatcd value of K and
sr,-;;- applying a ._
square wave inpul of I volt p•p.
,, ' 4.5 Disturbnncc Rejection
• Co~ncc1 the s~cond ordcfsys1c~ of Fig. 9.
r l, :,_..
Apply ;i low frequency disturbar1cc sign.al (sincwa"c of _l V) from some cxitm /
;1 · s~Hlrcc at _N1 and·scc ils cffr.cr on rhc oulpul. · 2
FIG. 7 CLOSED LOOP- OPTIONS FOR SECOND ORDER SYSTEMS,
i:; No\:· apply the same disturbance Signal at N2, and observe its effect on the O!Jlp u:.
I:• J11s11fy Ilic results: · . . •. ·
I
iii' i 4.6 Additional Expo_rimonts
.Som~ _other corTlpl~x syslcms wilh feed f~rward branch.cs, minor loops or mu !riPk
i: inpullou1pu·1 can also be ~ilTlulatcd in this 1c1 up and studied. However ,these may no1 :be
included at the basic' Jevcl." ·. · · · • ' ·
5, TYPICAL RES UL TS
below: .Some 1ypic;il rcs11l1s
· (qr second' order lype, I sysrc:m in closed loop arc giv en rn tab le
(,1) Input : square w;ivc, I volt p-p
Kl Time in ~C_c and~ in rad/sec,
5~ S.No. K M, ,, s•
"' / · I.
2.
10
- i---, .•
0.173 '
2:4 . .3.6 ,
'·
•10 . 0,487
"·
999

> b'
0.5 10
,.: I
979
7 0.12
1· 3.2 4.4 ~10 0.559 861
J. 5 0 609 61 9
I 0.06 4.4 6.0 •10 0.667 70i 0.721 693
Calculated from c.xpcrimcn1a/ values of ~1r :ind Ip.
I/if
,::, ,,

FIG. 8 CONNECTION FORT~ ORDER SYSTEM •

.:., # C;iiculai~d form 1)pcn ioop lr3nsfcr funclinn dc1cm1incd earlier.
- jli!~:::
1l LL~,~--- -
. :11/ -
., --- <I
 I'
1.
. I
Snmplc calcul:1tio11 for da1a a1 S. No. 2: The above ohserva1ions are for rht.· 1ypc-l second
order configuration shown in Fig. 7. The cakuforions here would require some resulis ob1ained
c.arJier from open-loop measurements. which are given ,:iext: ·
(I) lni~grotor: Given :1 IV p-p square wave input oftjme period T, the integrating capacitor in
this block charges from a constan1 current of o:5/R, and hence lhe p-p amplitude of rhe
rriangula:r output is given by ·
V =~_2_.=_: _!_=_:K 1 ,so1ha1K 4YP· P
•-• R 2C 4 RC 4 1
T
Mc:_1s11rel..l v:iluc orvP·Pfor 1=25 msfc is 60 mV. Hence K 1is 9.6.
N,(s)
.N,(s) (hJ. Time Co11~~u111 : The block ha~ a Ira sfer function of !he fonn Kz!{sT 1+ I) . Open-loop
measurements g1vc: -K1= JO and T 1=l mst."c , . . · .
Cakulalion for .0-:> and w." :The ~~imum ~vcrshoot M,. as a fraction of s1eady-s1a1e v:i ll;e is
related 'to dampillg coefficier1t lfby
f
·K2 _. __K_,_
~!l:;:;i sr,7T ST,• 1
·( -,6 - ) . (111 M,) ]
2
:
I •: ,
M
O
•
0
exp /(H , ) , so 1ha1 8- yl ~,:- ('; .j'
M
J I"/
. Tht;S for Mp=-0: l 2, 1hc- value of damping coefficient 5 by ·measurcmcnl is 0.559. Using
,·c1n1ion bc_rwce1~ <v 11 and ,,,
/Ii(.:
1 I·, --,1
;:i:
: !,/ <,>n .. -~=---== n • _ = 86 J rad/ sec
,.j(1-(6•J'.)
e-::
44xrn-'J0-(o.559J') . -
l,;,.;i:.
!i ..
FIG. 9 CONFIGURATION FO~ OTST:l)RBJINCE REJECTION STUDIES.
11k11l111io11 for 0.., :ind rn/: Open-loop transfer func1ion ~flhe total system may be written as
1/1
: qi : KJSJ<i . __ 7x 9.6x l0 _ 672x IO.) forK - ?
ii{ s(sr, + 1) 1sx 10- 3 1) i"'o')'
··Jllscd loOp tran sfe r ftrnc1i o t1 G(s) is· lhl'll wriue n :is
6 72~J0 3 .
G(s) - -~ - -·- -
• s:? "f' JO·\+ 672>- !0 3
' c_.rnparing cocffi cic r:: s in!!):'. l!cnominatoc of1hc expressio n for (i{sJ,
f.
.. wn~=672xJOJ, or w,."=S19rnd/scc
20w 11 == !OJ, or ;s • = 0.609 .
I
i J
·'
\ . . ·- -- ,, .
 r. ' ,_I
f
(b. lar wave, I volt p-p, r(I) =2Rt
· S.No: K. RV/see t'n ~.V cu' mV
r-·
-
so· o./6 167
I:,' I.
2.
,.
10 ·
BO 0.26 238
!· 3. ' 5 · 80 0.38 JJJ
I O Mc0$ured on CRO
·,, _CaJ~·u1:·1red from open loop tnmsfcr funclion .dcrcnnincd carJicr
i
I , ·s~;,,,;t.c calc.ulation for dnta at S.No,. 2 : The data here pertain lo the measurement o( the
!j srcady,slale error for ramp input. Since a ramp is n"I prac·rica, , rhe measuremcnr is being clone
i: I' .,, using a triangular input For a .ramp inpur, ·r(l)~Rr; where R is • the slope. However, for' a
III• ;
I
' . lrfongular inpu.r, rJie effective slope ·is double due 10 change in slope sign so rhar r/r)=2Rr. Foi a
I. , I V p-p inpur of a rimcperiod of 25 msce,
I,.j ;;
'ii i . ·, .
j . R 1112S• 80 vo/Js/scc.'._
I
r;
!
Effective slope = R, ·.. 2R = J60 vo/Js/scc.
(hf 11:, ('_· The s!cady~sra1c error .css is given ~j, 1;
1, ·h
i/ ·; ,•I
·, ' .
•· .R, 160 ·
c., =KK K .=----=238mV /!
/IiI r II .· ;I,. I I l 7 ,x 9.6 X I 0
u i : ·:1
J1
0
0. L/M/TAT/ONS OF THE SYSTEM !!
ii ! The accuracy of rcsulrs in an. experirhcnta/ measurement is always limited. Jr is ·good ro
_ll;. ;I:; / know and appreciate the limilotions ofll1e system so as Jo be able lo_'inrcrprer 'the resulls within
those /imitations. Some such foctors limiting the performance of the systems arc : ·
i ., ;:1
:!t ,' //i, I (a) Mo" measurements ore done using• CRO which moy nor be be11cr then :t 5%.
: ; i /!!'
(b) The sysJem'll.s c, JC ompliflcrs and in all analysis and catcularions, the bias currents I
and other error volrogc/currcnts have heel! neglected (though the design tends ro . i·
. ! :i
' I ,
· m'inimll.:c Jhcm). Furlhc.•,, lhc frequency response of all ampli ficrs has been assumed ·
' ! ';: nar which mny he f111e for a limilccJ low frequency range only.
(c) The ga,n of crror_dcreclor-cum-gain hlock is adjustable ii, rhc range 0- l O, T_hc r.ingc
g m"; be Joo lar£c_ for s_om~ configuraric:ns (rhe system stans oscillating · for g2iJ1
be)ond some hm11J whtlc 11 may be a b11 loo small for some or her resrricrinQ rhc ·
i § - number of observation scrs possible. - ·
: l. REFERENCES
ii:
/ ·i ,. '

I [ I J Modern Control Engineering•/(. Ogara, Pren rice Hall of lndia

[2] Au1om,11ic control systems - B.C. Kuo, Prentice Hall of India Pvr . Lru.
Pvr. Lid.
' 'i
:
I
I
JO
; i Ii.
: 'r I 11
' !
I
; ii
-½i. . -- - -- ,0 J
\