Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

INDIAN INSTITUTE OF TECHNOLOGY, ROPAR

DESIGN LAB-II (ME306) LAB REPORT

Lab - 4

Submitted by: Group C(Wednesday)

Ankur Bharali - 2020MEB1267


Apeksha - 2020MEB1268
Arun Kumar - 2020MEB1269
Aryan Prajapati - 2020MEB1270
Ashish Jagarwar - 2020MEB1271

Submitted on: 08/02/2023


Title : DC Motor Control (Frequency Response and Principles of Feedback)

Aim / Objective
● To measure the parameters of a plant using frequency tests
● To describe the frequency characteristics of a first order lag
● To understand why Bode Plots are preferred over other frequency plots
● To identify the significant parts of a feedback control system and manipulate transfer
function blocks
● To derive the closed loop and error transfer functions from the forward and open loop
transfer functions
● To describe the effect of closed loop on the steady state performance and transient state
performance.

Experimental Setup

a) MS15 DC Motor module


i) DC Motor
ii) Tacho generator
iii) Output potentiometer (produces -5V to 5V)
iv) Calibrated disk
v) Brake (has three positions)
vi) Ports (for power supply)
vii) Analog control
1) Motor drive
2) Tacho generator output
3) Potentiometer output
b) Digital control (for input and output)
c) AS3 command potentiometer
d) Virtual Control Laboratory (Software)
e) Connecting Leads (4 mm)
f) Power supply

Introduction
The experiment consists of two portions : Frequency Response and Principles of Feedback.
In the frequency response portion, we will plot Bode Plots and will try to find the cut-off
frequency or break frequency using three different methods : slope method, -3dB point method
and -45° method, respectively.
In the principles of the feedback system, we will look at the working of a closed loop system
with feedback and will try to find out the closed loop transfer function and error transfer
function. We will also observe the closed loop performance in steady-state and transient state.

Theory
Frequency analysis is an important aspect in control engineering. There are a number of analysis
and design techniques which are solely based on the frequency response of the plant. These are
graphical techniques and require only a knowledge of the measured frequency response of the
plant to be controlled. While studying the frequency response, it is observed that the output
signal obtained has an amplitude different from the input signal with some phase lag but the
nature of the output will be the same as of input signal. Bode plot is the plot where the amplitude
ratio and the phase are plotted separately against log frequency. Using this plot, cut-off frequency
or break frequency is calculated. Bode plot is very useful as it is quite easy to translate the
information obtained from the plot into any of the graphical forms used in control engineering
design.
The closed loop system with feedback is as shown below:

𝑂𝑢𝑡𝑝𝑢𝑡 𝐶
For this closed loop system, the transfer function (G) is : 𝐼𝑛𝑝𝑢𝑡
= 𝑅
𝐸
Error transfer function (E) is : 𝑅
Let K be the gain of the controller, Kp be the plant gain and H be the gain of feedback. Then, a
closed loop transfer function (CLTF) is given by:
𝑂𝑢𝑡𝑝𝑢𝑡 𝐶 𝐾·𝐾𝑝
CLTF = 𝐼𝑛𝑝𝑢𝑡
= 𝑅
= 1+𝐾·𝐾𝑝·𝐻
For unit feedback system, H = 1
𝐶 𝐾·𝐾𝑝
So, CLTF = 𝑅
= 1+𝐾·𝐾𝑝
Error transfer function (ETF) is given by :
𝐸 1
ETF = 𝑅
= 1+𝐾·𝐾𝑝·𝐻
For unit feedback system, H = 1
𝐸 1
So, ETF = 𝑅
= 1+𝐾·𝐾𝑝
Also, when tp is plant time constant, then the time constant of the closed loop system ( tcl) is
given by:
𝑡𝑝
tcl = 1+𝐾·𝐾𝑝

Procedure
Following steps are taken to perform the experiment.
Part-1: Frequency response of the system
● Make the proper wiring.
● Load the desired setup.
● Set the reference input to internal
● Choose Sine wave input with 60% level.
● Choose the frequency as 10 mHz.
● Make sure to ON the input and velocity to get them displayed on the graph.
● Switch ON and enable the plant.
● Freeze the display and click frequency ON.
● Measure the peak to peak voltage.
● Select phase D and measure the phase difference by moving the vertical line until it
intersects with the peak value.
● Repeat the above steps by varying frequencies.
● From the data obtained, make the Bode plot and calculate cutoff frequency.

Part-2: Feedback
● Load the desired setup.
● Set the reference input to internal.
● Choose DC signal as input with 50% offset.
● Make sure to ON the input, error, and velocity to get them displayed on the graph.
● Check that the gain K in the gain box on the on-screen mimic is set to 1.
● Switch ON the motor and enable the plant.
● Set the display as ‘meter’.
● Vary the value of gain and note the error and output (velocity).
● Again, load the setup and set the reference input to internal.
● Choose Step Input as input with 20% level.
● Make sure to ON the input, error, and velocity to get them displayed on the graph.
● Check that the gain K in the gain box on the on-screen mimic is set to 1.
● Switch ON the motor and enable the plant.
● Freeze the display and measure the time constant using ‘63% method’
● Measure the time constant for different gain values.

Observation Tables

Part-1: Frequency Response of DC Motor


Frequency Output Phase Lag Amplitude Ratio 20𝑙𝑜𝑔10𝐴 (dB)
(Volts p-p) (Degrees) A

10 mHz 5.8 1 0.9667 -0.2941

20 mHz 5.8 2 0.9667 -0.2941

50 mHz 5.8 5 0.9667 -0.2941

100 mHz 5.7 12 0.95 -0.4455

200 mHz 5.50 19 0.9167 -0.7554

500 mHz 4.50 40 0.75 -2.4987

1 Hz 3 60 0.5 -6.0206

2 Hz 1.70 106 0.2833 -10.9551

5 Hz 0.8 94 0.1333 -17.5034

10 Hz 0.4 246 0.0667 -23.5175

Part-2: Closed Loop Performance - Steady State


We have, Kp = 0.75 , R = 2.5V
Also, we know that :
(𝐶/𝑅)𝑡ℎ = 𝐾. 𝐾𝑝/(1 + 𝐾. 𝐾𝑝) and
(𝐸/𝑅)𝑡ℎ = 1/(1 + 𝐾. 𝐾𝑝)

Gain K Velocity (C/R)exp (C/R)th Error (V) (E/R)exp (E/R)th


(V)

1 1.17 0.468 0.428 1.32 0.528 0.571

2 1.62 0.648 0.6 0.88 0.352 0.4

5 2.06 0.824 0.789 0.44 0.176 0.2105


10 2.26 0.904 0.882 0.24 0.096 0.1176

20 2.37 0.948 0.937 0.13 0.052 0.0625

50 2.42 0.968 0.974 0.05 0.02 0.0260

100 2.46 0.984 0.987 0.02 0.008 0.01316

Part-3: Closed Loop Performance - Transient


From the previous lab, we calculated the time constant as tp = 0.24 s and gain, Kp = 0.75
We know that, for a closed loop system, theoretical value of time constant can be calculated
using:
𝑡𝑐𝑙 = 𝑡𝑝 / (1 + 𝐾. 𝐾𝑝)

Gain K Theoretical Time Constant Measured Time Constant

1 0.1371 0.107

2 0.096 0.075

3 0.0738 0.07

4 0.06 0.07

5 0.0505 0.05

10 0.0282 0.03

Plots
● Bode Plot
Calculations

Time constant

● 3 dB method
In this method, a horizontal line at amplitude -3 dB is made to intersect on the bode plot
and the point of intersection is the cut-off frequency.
Here the line intersects at fc = 0.552 Hz
Now,
ω𝑐 = 2π𝑓𝑐 = 2(3. 14)(0. 552) = 3. 467 𝑟𝑎𝑑/𝑠
1 1
τ1 = ω𝑐
= 3.467
= 0. 288 𝑠𝑒𝑐
● -45o

In this method, a horizontal line at phase -45o is made to intersect on the bode plot and
the point of intersection is the cut-off frequency.
Here the line intersects at fc = 0.594 Hz
Now,
ω𝑐 = 2π𝑓𝑐 = 2(3. 14)(0. 594) = 3. 734 𝑟𝑎𝑑/𝑠
1 1
τ2 = ω𝑐
= 3.734
= 0. 268 𝑠𝑒𝑐
● Slope method

In this method, a line of slope -20 dB/decade is drawn at higher frequency on the bode
plot and its x-intercept is the cutoff frequency.
Here the line intersects at fc = 0.594 Hz
Now,
ω𝑐 = 2π𝑓𝑐 = 2(3. 14)(0. 6166) = 3. 874 𝑟𝑎𝑑/𝑠
1 1
τ3 = ω𝑐
= 3.874
= 0. 258 𝑠𝑒𝑐
τ1+ τ2+ τ3 0.288+0.268+0.258
Average of the three methods, τ𝑎𝑣𝑔 = 3
= 3
= 0. 271 𝑠𝑒𝑐

Error in determining CLTF and ETF :

1. For Closed loop performance: Steady-State


| (𝐶/𝑅)𝑡ℎ − (𝐶/𝑅)𝑒𝑥𝑝 |
We know that, % Error in CLTF = | | × 100
| (𝐶/𝑅)𝑡ℎ |
| (𝐸/𝑅)𝑡ℎ − (𝐸/𝑅)𝑒𝑥𝑝 |
% Error in ETF =| | × 100
| (𝐸/𝑅)𝑡ℎ |

Gain K % Error in CLTF % Error in ETF

1 9.346 7.531

2 8 12
5 4.436 16.39

10 2.494 18.367

20 1.174 16.8

50 0.616 23.077

100 0.304 39.210

2. For Closed loop performance : Transient state


| (𝑡𝑐𝑙)𝑡ℎ − (𝑡𝑐𝑙)𝑒𝑥𝑝 |
% Error in determining time constant is : || (𝑡𝑐𝑙)
| × 100
|
| 𝑡ℎ |

Gain K % Error in tcl

1 21.955

2 21.875

3 5.14

4 16.667

5 0.99

10 6.383

Results and Discussion ( By Ankur Bharali - 2020MEB1267)

In this experiment, the closed loop system and the relationship between its output and error value
and the change in gain (K) value are first examined. We saw that as the gain was raised, the
output value rose and the error value dropped. The theoretical tendency and the observed trend
are consistent. In the second experiment, we calculated the time constant of the closed loop
system using both the theoretical calculation based on the plant time constant and the 63%
technique from the graph. The differences between the measured and estimated time constant
trends with varying gain K indicate a problem with the instrument. The final component of this
experiment looks at open loop system frequency. By adjusting the frequency value, we assessed
the amplitude and phase angle between the input and output. Using the ratio of amplitudes and
phase angle, we drew the bode plot. Three alternative methods—the slope method, the -3dB
method, and the 45° method—were used to determine the cut-off frequency from the bode plot.
Results and Discussion ( By Apeksha - 2020MEB1268)
In this experiment, we study the frequency response of the first order lag system and how a Bode
plot is plotted. Using the Bode plot, cut-off frequency of the system is determined using several
methods and for lower frequency, straight line represents gain. Thus, we can say that a first order
lag is characterised , in frequency domain, by its low frequency gain and its cut-off frequency. At
cut-off frequency the actual curve is 3dB below the low frequency level.

In the closed loop control system with feedback, we observed that the steady-state error
decreases as loop gain is increased. However, with high gain the output becomes noisy and the
motor in the system will start making peculiar noises. Here, we used a unit feedback system to
get exact errors so that the controller can be tuned such that the error decreases and thus
obtaining the required output for a particular input. We also observe that error transfer function
(ETF) value decreases as the gain is increased. It implies that increasing the gain of the controller
reduces the error in steady state and the output obtained is nearly the same as input when gain is
high. It can also be noted that with increasing gain, time constant also decreases. This implies
that with increasing gain, time to achieve the desired output decreases, i.e., desired output is
achieved faster with higher gain.

Results and Discussion ( By Arun Kumar - 2020MEB1269)


● We figured out the stability of the system using a bode plot.
● As we increase the gain, time constant decreases.
● As we increase the loop gain the steady state error decreases.
● Error transfer function value decreases as the gain is increased.
● We did a frequency test to determine gain and time constant.
● Steady state error is inversely proportional to gain.

Results and Discussion ( By Aryan Prajapati - 2020MEB1270)

In the first part of the experiment, the frequency response of the DC Motor (plant) is measured
and parameters like low-frequency gain and time constant are measured by plotting the Bode plot
of the system. A first order lag is described, in frequency domain, by its low frequency gain and
its cut-off frequency. Bode plot shows the variation of amplitude ratio and phase difference with
respect to the frequency. For the calculation of cutoff frequency, three methods were used
namely, -3 dB method, -45o method and slope method. From this cutoff frequency, the time
constant of the system is calculated. At cut-off frequency the actual curve is 3dB below the low
frequency level and the phase lag is 45o. The time constant calculated from time response and
frequency response were very close and thus, add the layer of verification in the results obtained.
In the second part of the experiment, unity feedback is also added to the system, making it a
closed loop control system. First, the steady state performance of the system is characterised by
𝐶
measuring the steady state error. It was observed that increasing the gain will make 𝑅
→ 1 and
𝐸
𝑅
→ 0 indicating that steady state error will decrease as the loop gain is increased. However this
comes at the cost of noisy output at higher gain. One important detail to mention is that the unity
feedback is used to get errors directly. For characterising the transient state performance, time
constant of the closed loop system was measured. It was observed that time constant decreases
with increasing gain indicating that it will take less time for the system to achieve steady state at
higher gains.

Results and Discussion ( By Ashish Jagarwar - 2020MEB1271)


In this experiment we learnt how to find the transfer function and error and if we increase the
gain then error will decrease. Considering Closed loop system in steady state, it was observed
that E/R tends to 0 and C/R tends to 1 as gain increases. As the gain continues to increase the
noise in the output increases and the steady state error decreases. BODE plots were used in this
experiment to analyse the results. With the use of the Bode plot, the cut-off frequency of the
system is identified using a variety of techniques, and for lower frequencies, straight line denotes
gain. As the value of K is increased from 1 to 2, the time constant gets decreased proportionally
with increasing gain. Although, for K exceeding 2, the time constant does decrease. We have
learnt about the cut-off frequency and time constant and learnt the relation between them. Time
constant is inversely proportional to the cut-off frequency.

You might also like