3.

2 Transient Response
For a step input Vr (s) = R/s, Eq. 2 yields
R KAkm/T
o(s) =
s s+(KAKMKb +l)/T
Taking inverse Laplace transform

o(t)= R.KAKM 1-e

K KmKr +1,.t
K KmK +1 (4)

where effective time constant Tefr may be defined as,

Ter - T/(KAKuKrtl) (5)
The transient response has an exponential character similar to a capacitor chargingl
through a resistor. Further, the effective time constant Tefr decreases with increasing
KA making the motor response faster.

The effective time constant may be determined from a recording of the step response
using either a pen recorder or a storage CRO. The step response for various values of KA
obtained through a storage oscilloscope, Tektronics, Model: TDS-210is shown in Fig.4 (a).
It may further be observed that for large gains (260) the speed of response becomes constant
due to saturation of amplifier and/or motor. The initial portion of the response is therefore a
straight line. Time constant may also be computed using an ordinary CRO as explained next.
Consider a general first order, type-0 transfer function of the form
C(s) K
R(s) St +1

which may represent both open loop and closed loop speed control systems defined by Eçs.
(1) and (2). Its response to a step input, R ut) may be seen to be
c() = R.K.(1-exp(-t/r))
For a square wave of p-p value of R as input, refering to Fig. 4 (b) it is easy to see that

T= (6)
2f
In
C(p-p) 1
R(p-p) K
method for
Where f is the frequency of the squarewave. The above equation suggests a
computing the motor time constant T.
shaft speed o which will
In the present electromechanical system, however, it is the
square wave excitation, i.e.
g0 through a triangular wave type of variation in response to a
given by
CoP-p). Since we are measuring the shaft speed using tachogenerator, o(p-p), equation.
the
VTlP-p)/KT, and VM(D-p), the motor input, the time constant may be found trom

Techno Instruments DCS-01 Page 4

 T
2r Vr (p-p)) KM KT for open loop system 7(a)
VM (p -p))
Where VM (p-p) = Ka-Vs (p-p)
and

T=
2f for closed loop system 7 (b)
V(p-p)
V (p-p) KMKb
where K includes only KyKT product from Eq. (1).
3.3 Disturbance Rejection
One of the important features of a feedback control
system is its ability to reduce the
effect of external disturbances. Fromn Fig. 2, the disturbance transfer
be written as
function for VR=0, may
o(s) sT +1
Op($) 1+G(s)H(S) sT+1+ KaKyKT
For a unit step disturbance, oD(s)=lp/s, the steady state output speed is given by

Oss K KyKT +1 (8)

Thus, the steady state speed change caused by an external disturbance should reduce
as the gain KA is increased. Also, the performance should be much superior to the
open loop case, i.e. with feedback disconnected (Kr=0)
In the experimental unit, the external disturbance is created by an eddy current brake.
Ihe pole pieces of a permanent magnet are inserted to varying depths into the rotating
aluminium disk. The eddy currents induced in the disk result in power loss and thereby load
the motor.

4. EXPERIMENTAL WORK

The experiments suggested in this section start with a study of the open loop system
and its subsystems. This is followed by the perfor1nance evaluation of the closed loop system
1or various operating conditions like forward path gain and disturbance.
4.1 Subsystem Performance
Various subsystem blocks are shown in Fig. I and 2. The characteristics of motor,
lachogenerator and square wave source are determined first. The FEEDBACK terninals are
leit open during this experiment, Note that KA may be varied from 0 to 100 using a 10
turn potentiometer. Thus one turn of the potentiometer corresponds to gain variation
from 0 to 10.

Techno Instruments DCS-OI Page 5

 (a) Signal and reference
Set KA=0. Conncct DVM to measure the range of variation of reference voltage
VR
Switch ON the square wave signal s and measurc its amplitude and frequency
using a calibrated CRO. The frequency of this sigral is about IHz,which makes
the CRO display very inconvenient for measureiments. It is suggested that the
ampliude may be measured with time-base swiched OFF, and for frequency,
simply count the number of pulscs (as scen on CRO scrcen), in say 60 seconds,
using a watch.
8'splay reoaí inne reod'
a(b) Motor and Taclhogenerator Goin
Set V=1 Volt and Ka=3. The motor may be running at a low speed. Record
specdN in rpm, and the Tachogenerator output V. Sal

Repeat with V=l and Ky=4..5,..,. 10, and tabulate mcasured motor voltage VM
(=VRKA). steady state motor speed N in rpm (or o,s-Nx2r/60 in radians/sec.) and
tachogenerator output V.
Plot N vs. VM. and V vs. N. Obtain KM and K from the linear region of the
curves(see Fig. S).
shaft speed in rad/ sec, @ss
Motor gainconstant, Ky= , and
Motor voltage, VM
Tachogenerator gain, K,=T Volt-sec
resdri
Oss rad

To calculate motor time constant, with square wave signal Vg ON, set VR and K
so that the peak-to-peak variation of VM lies between 3-8V. This wouldensure a
reasonably linear operation of the motor. Use Eq.(7) to calculate the motor time
constant T.

(Caution: The CRO must be kept in'dc-input' mode for this measurement.)
Oblain the motor transfer function using
KM
G(s)= sT +1
oFfF
Switt
(c) Disturbanc o {Or Is )
Set KA=5 and adjust the reference VR to get a speed reading close to 1200 rpm.
The brake setting should be at 0 i.e. no braking.
Record and tabulate the motor speed variation for different settings of the eddy
current brake.
Calculate percentage decrease in spced at each setting of' the brake, starting trom
no braking.

Ve 0.72 to

Techno Instruments DCS-01 Page 6

 4.2 Closed Loop Performance
Performance of the closed loop system is evaluated in terms of steady state error and
disturbance rejection as functions of forward gain. The FEEDBACK terminals are connected
together for this experiment.
(a) Steady state error
Set VR=l volt and KA-5. The motor may be running at a low
speed.
record speed N in rpm, tachogenerator voltage V, and the steadyMeasure and
state error
es (=VR-V).
Repeat above for KA=5, 10, 15, 20, ...... 100
Compare in each case the value of steady state error computed from Eq. (3) 1.e.
Css
1+KaKMKT
Comment on the results.
(6) Transient Perforinance
Set V=0.5 Volt and Ka=5. Switch ON the square wave signal and
measure peak
to-peak amplitudes of Vs and V. Use Eq. (7) to calculate system time
constant
Tar with R=Vs(p-p) and C-V(p-p). The value of K may be obtained from (2) as
K= KAKM : keff
KKMKT +1
(Note: Although the method suggested above is theoretically valid, in the
experimental unit theresults are likely to be erroneous. As the closed loop system
encounters a step input, the motor is driven harder (increased input) so as to force
it to respond faster. The current limitation and saturation in the
amplifier does not
permit this in a linear fashion. However for small values of K, say less than 5, the
results are reasonable.)
(c) Disturbance Rejection
With KA=5, FEEDBACK terminals shorted and the brake setting at 0, adjust
reference VR to get a speed close to 1200 rpm.
Record and tabulate the variation in speed for different settings of the eddy
current brake. Calculate percentage decrease in speed at each setting of the brakc.
Repeat above for KA=10, 50, 100.
Compare the percentage decrease in speed at various brake settings for open loop,
closed loop with KA=5, and closed loop with KA=10. Comment on the results.

5. RESULTS
Typical results obtained on an experimental unit are given below for guidance.

Techno Instruments
DCS-01 Page 7
 (a) Motor and Tachogenerator Characteristics
VR=l volt
S.No. KA N rpm V, Volt VM Volt Experimental
Setting K =Vw/VR
3 560 0.31 3.01 3.01
2 4 1453 0.81 4.19 4.19
3 2264 1.27 5.45 5.45
4 6 2440 1.38 6.46 6.46
7 2445 1.39 7.44 7.44
6 8 2445 1.39 8.55 8.55
7 2445 1.39 9.59 9.59

Graphs of N vs. VM, and V, vs. N are shown in Fig. 5. From the linear region, KM and KT are
obtained as, KM=36.468, K-0.00523,and Kuk-0.19
Time constant was obtained as
1
T= = 1.997 sec.
2f VT(p-p) 1 0.04 1
In 1 2x0.82 In 1-
L 0.8 0.19
VM(p-p) KMKb.
Where f=0.82Hz., VM=0.8V(p-p), Vr= 40mV (p-p)
and the motor transfer function as
36.468
G(s) = KM
sT+1 1.997s +1
Note that the motor speed-voltage characteristics is rather non-linear. This is because the
motor fails to start at very low voltages and at higher voltages its speed saturates due to
internal speed limiter.

(b) Closed loop performance

(i) Steady state error
VR=l volt

S.No. KA Nrpm V, Volt es(Vg-V) ess-1/(1+KaKmKr)

Experimental Theoretical
5 648 0.37 0.63 0.516
10 1083 0.62 0.38 0.348
3 15 1276 0.73 0.27 0.262
4 20 1377 0.79 0.21 0.210
25 1442 0.82 0.18 0.175
50 1581 0.90 0.10 0.096
7 75 1635 0.93 0.07 0.066
100 1656 0.94 0.06 0.0506

Techno lnstruments DCS-01 Page 8

 Observe that the numerical values of theoretical and experimentally obtained steaay
state error do not match, though the pattern of variation is same ie it decreases with increase
in forward gain. The mismatch is due to the fact that the motor gain
constant KM does n0t
remain constant due to non-linearity of the motor.
(i)System timeconstant
VR=0.5 Vd.c.
f =0.82 Hz

KA=5, Vp = 480 mV (p-p), V,= 120 mV (p-p)

Fromequation 7(b)
Keff = K¡ KM KT
5x0.19
-=0.487, and
I+KA KM KT 1+5x0.19
1
Terr= = 846 msec
2x0.82 120
0.487 480

(iüi) Disturbance rejection

Speed = 1200rpm (approx.)
The table below shows the variation of speed under various conditions of feedback and thus
illustrates theeffectiveness of speed control system in rejecting disturbance.
Brake Setting 2 3 4

Open Loop 1204 1200 1136 812 596 563

Speed, rpim
Closed Loop 1205 1203 1182 1023 910 886
(KA5)
Closed Loop 1208 1205 1189 1107 1010 991
(KA=10)
Closed Loop 1186 1186 |183 1156 1134 1128
(Ka=50)
Closed Loop 1196 1195 1192 1178 1167 1164
(Ka=100)

6. REFERENCES
[1] Control system Engineering-1.J. Nagrath and M. Gopal, Wiley Eastern Limited.
[2] Modern Control Engineering - K. Ogata, Prentice Hall of India Pvt. Ltd.
[31 Automatic Control System - B.C. Kuo, Prentice Hall of India Pvt. Ltd.

Techno Instruments DCS O1 Page 9