NAMIBIA UNIVERSITY OF SCIENCE AND TECHNOLOGY

FACULTY OF ENGINEERING
----------------------------------------------------------------------------------------------------------------
Control System COS620S
Bachelor of Technology in Mechanical Engineering
----------------------------------------------------------------------------------------------------------------
Laboratory Report 1
BALL AND BEAM
----------------------------------------------------------------------------------------------------------------
BONGE KARLUSH 219054630
Lecturer: Mr. Kaputu Tulipale
Lab Technician: Mr. Conrad January
Due Date: 12 November 2021

1
Table of Contents

1. Abstract ............................................................................................................... 3

2. Introduction .......................................................................................................... 3

3. Apparatus ............................................................................................................ 3

4. Experiment 1: Basic Tests and Familiarisation .................................................... 4

5 Experiment 2: Proportional Control of Beam Angle and Manual Control of ball

Position ...................................................................................................................... 8

6. Experiment 3: Phase advance control of the ball position ................................ 12

7.1 Experiment 4: Proportional Plus Derivative Control of the Ball Position ........... 16

8. Experiment 5: Measurement of the System Characteristics and Alternative PD

Design ...................................................................................................................... 19

5. RESULTS .......................................................................................................... 21

6. Discussion ......................................................................................................... 26

7. Conclusion ......................................................................................................... 27

8. Recommendations ............................................................................................... 27

6. References. ....................................................................................................... 27

2
1. Abstract
The CE106 Ball and Beam apparatus is part of an elite suite of products created for
the theoretical study and practical exploration of basic and advanced control
engineering principles. This covers the use of analogue or digital techniques to
analyse static and dynamic systems. The device is designed to solve difficulties
involving the control of unstable systems, which are common in the aerospace and
related control industries. It can also be used as a hands-on introduction to control
system design, operation, and application in general.

2. Introduction
A self-contained bench mounting apparatus designed to allow students of all
academic levels to investigate basic and advanced control principles, including
control of naturally unstable systems. The Ball and Beam Apparatus demonstrates
the control problems of unstable systems, such as a rocket or missile during launch,
which requires active control to prevent the missile from becoming unstable and
toppling over. A steel ball is free to roll on two parallel tensioned wires in the
apparatus. The wires are suspended from a beam that pivots at its centre. The beam
angle is controlled by a servo motor, and sensors measure the beam angle and ball
position. The fundamental control problem is to vary the beam angle to maintain
control.(TecQuipment Ball and Beam Apparatus – CE106 – AYVA Educational
Solutions, n.d.)

3. Apparatus

Fig 1: (Ball-and-Beam-Apparatus, n.d.)

3
4. Experiment 1: Basic Tests and Familiarisation

4.1 Objective
 The goal of this experiment is to test basic apparatus functions, familiarize the
user with the Ball and Beam apparatus, and calibrate the transducers,
specifically the beam angle and ball position transducers.

4.2 Apparatus
 CE106 Ball and Apparatus
 CE120 Control

4.3 Part 1: Beam Drive Motor Test

4.3.1 Connections

Fig 2: Basic Tests and Familiarisation Connection diagram

4
4.3.2 Setting
 CE120 – Potentiometer in mid-position, such that the voltmeter reads 0 V
 CE106 – remove the ball from the beam

4.3.3 Procedure
 Turn the potentiometer slowly clockwise, noting that at a certain non-zero
voltage, the motor will start to turn, and the bean will tilt back and forth as
the drive cam rotates through one revolution. By setting the potentiometer
to 3 V, the phenomenon can be observed. The drive motor will now spin,
alternately speeding up and slowing down.

 Turn the potentiometer anti-clockwise to apply a negative voltage of

approximately -3V to ensure that the beam motor turns in the opposite
direction. As the cam-follower spring aids and opposes the motor, it
should turn, alternately speeding up and slowing down. This ensures that
the drive motor works properly.

5
4.4 Part 2: Beam Angle Transducer Calibration
4.4.1 Connection

Fig: 3 Beam angle transducer calibration connection diagrams

4.4.2 Settings
 CE120 – Potentiometer in mid-position such that the voltmeter reads 0 V.
 CE106 – Ensure that the equipment is on a level, horizontal surface.

4.4.3 Procedure
 As with all experiments with the CE106, it is first necessary to ensure that the
beam transducer reads zero volts when indicated as being horizontal (0°).

6
  Use the CE120 potentiometer to adjust the drive motor until the centre arrow
on the beam boss aligns with the (0°) of the front panel scale.
 With the beam horizontal, observe the beam angle transducer output (θ) as
indicated on the CE120 digital voltmeter. If it does not read zero volts, this
may be set to zero using the beam adjust potentiometer at the rear of the
CE106 front panel.
 Remove the ball from the beam.
 Manually rotate the beam by placing your hand underneath the right-hand end
and raising or lowering it as required to set and then measure the beam angle
transducer output voltage for the angles ±2.5° and ±5° with the interval of
±2.5.
 Plot a graph of your results and use this to ensure that increasing the beam
angle beyond ±5° produces a proportionate increase in transducer output over
the whole range of achievable beam angle (approximately +7.5° to -10°).

4.5 Part 3: Ball Position Transducer Calibration.

4.5.1. Connections.

Fig 4: Ball Position Transducer Calibration connection

7
4.5.2 Procedures.
 Place the ball at the position on the beam and read the ball position
transducer output from the voltmeter.
 Move the ball +10 cm along the beam and record the transducer output in
table.
 Repeat the above procedure at 10 cm intervals along the entire length of the
beam.

5 Experiment 2: Proportional Control of Beam Angle and

Manual Control of ball Position

5.1 Objective
 The object of this experiment is to place the beam angle under closed loop
proportional control and illustrate the difficulty of manually controlling the ball
position.

5.2 Apparatus
 CE106 Ball and Beam Apparatus
 CE120 Controller
 Chart recorder

8
5.3 Part 1: Beam Angle Control – Error
5.3.1 Connection

Fig 5: Beam Angle Control Connection Diagram

9
5.3.2 Settings
 CE120 – Function generator set to square wave output, frequency of 0.1 Hz,
offset zero and level zero. The proportional block should have a gain of 1.

5.3.3 Procedure
 Set the proportional gain of 1 and slowly increase the function generator
level to obtain a square wave amplitude of 2 V.
 Now remove the voltmeter from the function generator output and connect it
to the beam angle error signal. As the beam tilts clockwise and anti-clockwise.
 Repeat the above procedure for = 2, 3, 4, 5, 6, 7, 8. Note that the error
should decrease as is increased, indicating more accurate control. Also,
the speed of the of response should increase as is increased.

10
5.4 Part 2: Beam Angle Control – Speed of Response
5.4.1 Connection
 Connect the beam angle output to the chart recorder.

Fig 6: Beam Angle Control Connection Diagram

11
5.4.2 Settings
Set the proportional gain of 1 and set the function generation level to 2 V.

5.4.3 Procedure
 Plot the response to the square wave (suggested time base 5 mm/sec)
 Repeat the exercise for = 3, 5, 8. Note that as the increases, the speed of
response increase with overshoot starting to occur for gains greater that =
5.

6. Experiment 3: Phase advance control of the ball position

6.1 Objective
 The object of this experiment is this experiment is to place the position under
phase advance block on the CE120 controller and illustrate pre-filtering of the
signal.

6.2 Apparatus
 CE106 Ball and Beam
 CE120 Controller
 Chart Recorder

12
6.3 Part 1: Manuel Adjustment of the Reference Ball
Position
6.3.1 Connections

Fig 7: Phase Advance Control Connection Diagram

6.3.2 Settings
 CE120 upper potentiometer turned to 0 V
 Beam angle controller gain =8
 Proportional gain = 0.1
 Phase advance controller = 4, = 0.1

13
6.3.3 Procedure
 Place the ball carefully on the beam at the center position.
 Use the potentiometer to vary the reference ball position.
Varying the Phase Advance Gain

 Turn the potentiometer to move the ball to various positions on the beam and
notice that the error in beam position can be quite large.
 Increasing to 0.3 and repeat the above procedure.
 Repeat the above procedure for = 0.5.

Varying the Phase Advance T, α Parameters

 Move the ball from the zero position to 20 cm and back again by adjusting the
potentiometer.
 Switch the phase advance controller α to 0.005 and 6
 Repeat the above procedure
 Switch the phase advance controller α to 0.5 and 0.5 1
 Repeat the above procedure

14
6.4 Part 2: Transient Response of the Phase Advance
Controller

6.4.1 Connection

Fig 8: Transient Response Connection Diagram

15
6.4.2 Settings
 CE120 – set the potentiometer to 4 V
 Set the controller gain to = 8, = 0.3, = 4, = 0.05 =1

6.4.3 Procedure
 Gently place the ball on the mid – point of the beam
 Make the dotted connection in fig 7, this introduces a step change in position
of 4 V.
 Observe the transient response of the ball by plotting it on the chart recorder
 Repeat the procedure with = 2, = 0.1

7. Experiment 4: Proportional Plus Derivative Control of the

Ball Position

7.1 Objective
The object of this Experiment is to place the ball position under phase advance
control using the proportion plus derivative block of the controller on the CE120
controller and illustrating high frequency filtering in the feedback loop.

7.2 Apparatus
 CE106 Ball Beam
 CE120 Controller
 Chart Recorder

16
7.3 Part 1: Manual Adjustment of Reference Ball Position
7.3.1 Connections

Fig 9: Manuel Adjustment of the Reference Ball Position Connection Diagram

7.3.2 Settings
 CE120 – upper potentiometer turned to 0 V
 Beam angle controller gain =8
 PID proportional block gain = 0.5
 PID derivative block gain = 0.5
 High frequency filter = 4

7.3.3 Procedure
 We placed the ball carefully on the beam at the center position
 We used the potentiometer to vary the reference ball position

17
  The high frequency filter cut-off frequency is controlled by the integrator gain
.
 Decrease back to 4 and repeat the exercise, noting the difference in
response speeds.

7.4. Part 2: Transient Response of the Reference Ball

Position
7.4.1 Connections
 Disconnect the potentiometer from the summer block and connect the
function generator output to the summer block, , input.

Fig 10: Manuel Adjustment of the Reference Ball Position Connection Diagram

18
7.4.2 Setting
 CE120 – as for part 1 and select a square wave output with frequency 0.04
Hz.

7.4.3 Procedure
We placed the at centre of the beam and slowly increase the function generator level
until the square wave voltage is 4V. The ball will now move between approximately
±20 with a frequency of 0.04 Hz
Influence of Derivative Action
 We Connected the voltmeter to the monitoring point on the output of the PID
controller derivative block. The derivate block output increases as the ball
gets faster and falls to zero as it stops at either ends. The action of the
derivative block is to control the damping factor of the transient response.
From the theory, the closed loop damping factor is proportional to .
=
2
=
Where G is the ball and beam gain is the damping factor and is the
natural frequency
 We verified by plotting on a chart recorder the square wave response with
= 0.5 = 0.8.
Influence of Proportion Action
 The action of the proportional block is to control the natural frequency of the
closed process, checking this by plotting the transient response for =
0.5 =1

8. Experiment 5: Measurement of the System

Characteristics and Alternative PD Design

8.1 Objective
 The objective of this experiment is the to measure the ball and beam transfer
function and hence design the coefficients of a PD controller using an
alternative PD implementation.

8.2 Apparatus
 CE106 Ball and beam
 CE120 Controller
 Chart recorder
 Stopwatch

19
 

8.3 Part 1: Measurement of System Characteristics

8.3.1 Connections

Fig 11: Measurement of System Characteristics Connection Diagram

8.3.2 Procedure
 We placed the ball carefully in the center of the beam, ideally, in this
configuration, the ball and beam should be an oscillator and the ball should
oscillate with a constant amplitude and angular frequency, given
=
Where G is the gain of the ball and beam transfer function ( ) given by:
( )=

20
 Due to extra phase lag in the loop, the oscillation will grow slightly at each
period. Furthermore, the of oscillation will still be close to where:
2
ω =
 We used the stopwatch to measure and calculate the gain G of the ball and
beam transfer function.

5. RESULTS

Experiment 1: Part 2
Beam angle (o) Beam angle Transducer Output
(volts)
+5 3.50
+2.5 -0.85
0 -2.52
-2.5 -4.84
-5 -9.74

Experiment 1: Part 3
Ball position (cm) Ball position transducer output (volts)
45 9.00
40 8.00
30 6.02
20 4.04
10 2.05
0 -1.95
-10 -3.96
-20 -4.40
-30 -5.94
-40 -7.96
-45 -8.94

Experiment 2: Part 2
Controller gain kp1 Maximum Beam angle-error (1volts)
Up Down
1 -0.82 -0.14
2 -0.45 -0.11
3 -0.31 -0.09
4 -0.23 -0.08
5 -0.18 -0.07
6 -0.17 -0.08
7 -0.11 -0.11
8 -0.07 -0.11

21
22
Experiment 4

23
Experiment 5: Part 2
Damping factor Natural frequency Kd kp
0.6 1 0.7 0.6
0.3 1 0.4 0.6
0.6 2 1.5 2.5(error)
0.3 2 0.7 2.5(error)

Lab two tables

Level required to make air valve turn 2.46
(positive)
Level required to make air valve turn -2.84
(negative)

Engine speed (rmp) Engine speed sensor output (v)

498 2.02
711 2.87
943 3.81
1262 5.08
1576 6.31
1814 7.29

Function generator Dead-band width Dead-band width

Dither (positive) (negative)
Level setting
0 9.70 -10.66
1 9.86 -10.54
4
1 10.14 -10.57
2
3 10.59 -10.76
4

Gain k1 Effective Dead-band Typical observed error

1 0.83-1.25 (-0.04)-1.48
S 0.57-2.58 0.18-0.23

24
10 0.69-2.70 0.09-0.09

Master controller Error in speed control Speed sensor output (v)

proportional gain kp2
0.5 -3.15 6.19
1 -2.54 5.55
3 -1.46 4.49
5 No constant speed No constant speed

25
6. Discussion
Experiment 2
Part 1: The CE120 Controller has analog knobs (such as the one-off setting the
values of Kp and Ki), therefore the values set might be slightly different from the
required values, which might cause a slight difference between the expected
results and the ones obtained
Part 2: The graphs generated on the plots from the angle output after feeding it
with 2V from the function generator are typical square waves and the shape of
these waves’ changes depending on the gain. It was observed that the response
speed has a linear relationship with the gain; thus, as Kp1 was increased the
response speed also increases. Another observation made was that as the gain
Kp increases the overshoot on the waveform also increases. Since Kp=8 was the
highest gain used, overshoot was higher at this gain.
Part 3: As mentioned in the procedures, the process of manually controlling the
ball is characterized by initially reducing the control input as you approach the
halfway mark of your destination or final desired position. This exercise gives the
operator a bit of prediction which is required to avoid any overshoot
Experiment 4
Part 1: It was observed that the beam response was much slower and had low
stability when K was incremented to its new value. The beam had faster response
at previous, smaller values of K and it was more stable. This experiment proved
that the response of the beam is inversely proportional to a specified value of K. It
was also noticed that if the optimum beam response is not attained, the system will
be unstable.
Part 2: When the ball is set at the required position and moved manually, it will
move back rapidly to its initial position. At new values of alpha and T, the system
is noticeably faster, hence making these values favorable for faster response. The
control of the ball in this experiment is better compared to experiment 2. It was
impossible to get the ball back to initial position during manual control. The system
came very close to setting ball back to initial position in this experiment.

26
7. Conclusion
In conclusion, these series of experiments on the ball and beam balance apparatus
made us realize and apply the theory that was taught on control systems in general.
Especially on designing a controller for a system. From this practical exercise we
were able to realize the importance of controllers because we were able to witness
how it is almost impossible to manually control any system. The functions of the
derivative and integral part of the PID controller in minimizing the error were realized
in this experiment. Although there some challenges such as lack of enough
appliances so that the experiments could be done individually, and the appliances
used were very old thus it sometimes malfunctioned, and it is also susceptible to
errors; this laboratory exercise was successful as all the objectives were met. The
set of experiments were successful and important as they helped students
understand ball and beam system effectively. Theoretical control systems knowledge
made more sense after conducting experiment

8. Recommendations
 Lab technician assistants and equipment so that students will not be cluttered
on one experiment and wander off.
 A more advanced technique can also be incorporated to expose the students
to recent day technology, the vast possibilities and limitations that can be
encountered in the industry.
 The CE120 Controller has analog knobs (such as the one-off setting the
values of Kp and Ki), therefore the values set might be slightly different from
the required values, which might cause a slight difference between the
expected results and the ones obtained.

6. References.
ball-and-beam-apparatus. (n.d.).
TecQuipment Ball and Beam Apparatus – CE106 – AYVA Educational Solutions.
(n.d.).

27
28