Antenna Azimuth Position

Control System Verification

Dr. Jose Granda
ME27
!2"2!"7
Graduate Students#
Su$h%ir &undal
'ee (hao
(yrone (racy
)*(+,D-C(),*#
Control systems are inte.rated into many facets of life/ ran.in. from nature to
social issues to mechanical systems. A nature e0am1le 2ould %e our %odies# 2hen 2e
2ant to .ra% an o%3ect4 our eyes 2ill focus on it and relay the information to our %rain4
2hich 2ill send electrical si.nals to our muscles to 1osition our arms correctly. (here are
mechanical systems li$e controllin. 2ater flo2 from a tan$4 and $ee1in. the tan$5s 2ater
hei.ht at an a11ro1riate hei.ht. (here are an unlimited num%er of control e0am1les in
the 2orld4 %ut 2e 2ill focus on a 1articular electromechanical system.
*orman S. *ise in6esti.ates the desi.n of a control system for an antenna in his
te0t%oo$ Control Systems En.ineerin.. (he name of this 1ro3ect is Antenna Azimuth
Position Control System. (he %asic idea is that someone4 from a control to2er4 can ad3ust
a sim1le 1otentiometer %y hand and ultimately mo6e a lar.e antenna.
(his system is thorou.hly analyzed %y Dr. *ise. (he system schematics are .i6en
and transfer functions are o%tained %y hand. (his 2ill 1ro6ide us su11ort durin. our
analysis of the Antenna Azimuth Position Control System.
,'JEC()VE#
,ur .rou15s o%3ecti6e is to o%tain the same analysis results as the Control Systems
Engineering te0t%oo$ on the same Antenna Azimuth Position Control System. (he
te0t%oo$ uses hand analysis4 %ut 2e 2ill try to match the results usin. a com%ination of
Cam17G4 Matla%4 and Simulin$. (he items 2e 2ill com1are are transfer functions and
ste1 res1onses.
Cam17G should 1ro6ide the same results as lon. as our modelin. of the system is
correct. (his is 2here the hardshi1 may lie.
S8S(EM C,*CEP(#
9e must understand the systems o1eration so an e01lanation 2ill follo2.
Po2er is a11lied to a 1otentiometer. An o1erator 2ill rotate a 1otentiometer5s $no% a
certain amount. (he 6olta.e difference from the source and the ne2 1osition is then
measured. (hat ne2 6olta.e si.nal 2ill %e sent to a 1ream1lifier. (he 1ream1lifier 2ill
am1lify the si.nal %y a .ain :;<. (hen that result 2ill %e sent to a 1o2er am1lifier. (he
1o2er am1lifier is 1o2ered %y a much .reater 6olta.e to allo2 for si.nificant .ain. So
the out1ut of the 1ream1lifier 2ill %e am1lified %y the 1o2er am1lifier and that result 2ill
%e sent to a motor. (he motor is connected to a shaft and .ear =!. Gear =! is connected
to a lar.er .ear =2. Gear =2 is directly connected to the shaft of the antenna. ,nce the
motor 1o2ers .ear =! then .ear =2 2ill start to rotate the antenna. Gear =2 is also
connected to .ear =>. Gear => and .ear =2 ha6e the same ratio. Gear => is connected to
another 1otentiometer. (hat 1otentiometer is connected to the same 6olta.e source as the
first one. (he difference %et2een that 1otentiometer and the 6olta.e source is then
feed%ac$ to our ori.inal 1ream1lifier. (hen the 1ream1lifier 2ill ta$e the difference
%et2een the t2o 6olta.e si.nals and out1ut that result into the 1o2er am1lifier. (he
1o2er am1lifier then dri6es the motor to mo6e the antenna. (he 1rocess 2ill continue to
cycle as lon. as there is a 6olta.e difference %et2een the t2o 1otentiometers in the
system. A .ra1hical 6ie2 is 1ro6ided %y the te0t%oo$.
(E?(',,; SC&EMA()C#
(he antenna control system needs to %e con6erted from a 1hysical system to a %o0
dia.ram schematic. (his schematic is 1ro6ided %y the te0t%oo$.
&ere 2e can see all 1arts of our 1hysical system. (he 1otentiometer the o1erator
controls is at the 6ery to1. (he si.nal is sent to the 1ream1lifier then to the 1o2er
am1lifier. (hat result is sent to a motor @.yratorA then to a .ear4 2hich is connected to
another .ear to chan.e the 1osition of the antenna. Binally the antenna5s si.nal is
connected to another .ear and a 1otentiometer. (he feed%ac$ is easily seen .oin. %ac$
from the 1otentiometer into the differential 1ream1lifier. (he feed%ac$ is crucial in
determinin. 2hether our antenna is in the correct 1osition.
(he 1arameters of the system are .i6en to us %y the te0t%oo$. 9e 2ill use the same
1arameters 2hen com1arin. results.
(a%le !# Schematic Parameters
A*AC8S)S#
9e 2ill start %y %rea$in. do2n 2hat 1arts of the system to analyze. (here are
three main 1arts# differential 1ream1lifier4 1o2er am1lifier4 and motor 1lus .ears.
D)BBE+E*()AC P+EAMPC)B)E+#
(he %oo$ is 6ery strai.ht for2ard a%out this 1art. (hey assume that the am1lifier
ne6er reaches saturation and the dynamics of this 1art are ne.lected. (his is %ecause the
res1onse of this 1art is ty1ically much faster than the res1onse of the 1o2er am1lifier.
(he transfer function of this 1art is :;< 2hich is 3ust the .ain. ,ur 1ream1lifier is
modeled in Cam1G#
(he te0t%oo$5s transfer function and ste1 res1onse is sho2n %elo2#
,ur transfer function and ste1 res1onse is sho2n %elo2#
(he t2o transfer functions and ste1 res1onses are the same. (his 2as an easy 1art
%ecause all that is ha11enin. is a constant .ain is a11lied to a si.nal.
P,9E+ AMPC)B)E+#
(he 1o2er am1lifier ta$es the out1ut of the 1ream1lifier and am1lifies the si.nal
to send to our motor. (he o17am1 circuit is sli.htly different than the 1ream1lifier. 9e
had 1ro%lems 2ith deri6ati6e causality in our Cam1G model so 2e had to add a C
element. (he %ond .ra1h is sho2n %elo2#
(he te0t%oo$5s transfer function and ste1 res1onse is sho2n %elo2#
,ur transfer function and ste1 res1onse is sho2n %elo2#
(he transfer function and ste1 res1onses for %oth the te0t%oo$ and our results match
closely. 9e had to ad3ust resistor and ca1acitor 6alues in order to o%tain the matchin.
transfer functions. (his means that our 1o2er am1lifier should %e modeled correctly.
M,(,+ A*D C,AD#
(he motor and load setu1 has %een done se6eral times. (he schematic is in the
Cam1G user manual. 9e started 2ith that one %ut added an additional (B so that 2e can
account for our system5s feed%ac$. (he Cam1G model is sho2n %elo2#
(he te0t%oo$5s transfer function and ste1 res1onse is sho2n %elo2#
,ur transfer function and ste1 res1onse is sho2n %elo2#
9e 1ut the transfer function the %oo$ calculated into Simulin$ and .ot a 1lot. (hen 2e
too$ ours and entered it into Simulin$ and came u1 2ith almost the same e0act 1lot. (he
slo1e of the line is different due to the .ain of the system.
C,MPCE(E S8S(EM#
*o2 2e must com%ine all three 1arts and o%tain the system5s res1onse. Birstly4
our antenna system Cam1G is sho2n %elo2#
(he te0t%oo$5s transfer function and ste1 res1onse is sho2n %elo2#
,ur transfer function and ste1 res1onse is sho2n %elo2#
,nce a.ain4 2e used simulin$ to com1are the results re.ardin. the 2hole system. -sin.
simulin$ the transfer function and 1lot 2ere 6ery similar. ,ur 1ro%lem 2as matchin. u1
the Matla% 1lots.
C,D)*G
,ne of the ma3or issues 2e had 2ith this 1ro3ect 2as the Matla% codin.. 9e
entered in all the initial conditions that 2ere in the te0t%oo$4 and then entered in a 6alue
for the .yrator4 2hich 2as not .i6en. After that 2e 2ould .et errors and 1ro%lems 2ith
our results mostly for the Motor analysis and o6erall system analysis. Binally 2e fi.ured
out that 2e need to chan.e matrices dimensions in order to o%tain 2
nd
and >
rd
order
transfer function eDuations. Still 2e don5t understand the matla% code 2ell enou.h to
match u1 our calculated transfer function to that one in the %oo$.
C,*CC-S),*
,6erall this 1ro3ect 2as 6ery interestin.. )t incor1orated understandin. of se6eral
disci1lines. A system 2as .i6en to us that had already %een analyzed %y hand. 9e first
had to read a%out the system and conce1tually fi.ure out ho2 it 2as 2or$in..
(o start the analysis 2e had to %rea$ the system u1 into a fe2 sections/ the
differential 1ream1lifier4 1o2er am1lifier4 and motor. (hen 2e continued to model each
section 2ith Cam1G %ond.ra1hs. Cam1G hel1ed us o%tain transfer functions and ste1
res1onses. 9e also used simulin$ to 6erify these results. -ltimately 2e com1ared the
te0t%oo$s results 2ith ours and found that they 2ere close on some 1arts and not so close
on others. 9e56e s1ent many hours on these 1ro%lems4 %ut are una%le to mani1ulate the
Matla% code or system characteristics to o%tain the results 2e 2ant.
,ne other as1ect of this 1ro3ect 2e s1ent a lot of time 2ith is 1hysically %uildin.
the antenna control system on a %read%oard. (his 2as 6ery interestin. to see the second
order"feed%ac$ system at 2or$. 9e had to research all the electrical com1onents of the
model and connect e6erythin. to.ether. After %lo2in. u1 se6eral )Cs 2e came u1 2ith a
2or$in. 1hysical model. 9e ha6e 6ideo ta1ed the system in action and included it in the
folder that 2e 2ill turn in.
Coo$in. %ac$ 2e should ha6e started the 1ro3ect a little sooner4 %ut 2e56e learned
a lot in a short time4 and are confident that 2e understand the %asics of
Cam1G"Matla%"Simulin$ 6ery 2ell.