Experiment 6

Modeling and investigation of Modular

Servo System (MSS)

Objectives
The objective of this lab is to give a brief introduction to Modular Servo
System (MSS), its components and mathematical modeling. The student will
be able to conduct different experiments on this hardware in the forthcoming
labs.

Theory
The Modular Servo System (MSS) consists of the Inteco digital servomech-
anism and open architecture software environment for real-time control ex-
periments. The main concept of the MSS is to create a rapid and direct path
from the control system design to hardware implementation. The MSS sup-
ports the real-time design and implementation of advanced control methods
using MATLAB and Simulink tools and extends the MATLAB environment
in the solution of digital servomechanism control problems.
Servomechanism
In control engineering a servomechanism, sometimes shortened to servo, is
an automatic device that uses error-sensing negative feedback to correct the
action of a mechanism. It usually includes a built-in encoder or other posi-
tion feedback mechanism to ensure the output is achieving the desired effect.

The integrated software supports all phases of a control system develop-

ment which are given below:

54
 • On-line process identification

• Control system modeling, design and simulation

• Real-time implementation of control algorithm

Features
• The set-up is fully integrated with MATLAB/Simulink and operates in
real-time in MS Window 2000/XP/W7

• Real-time control algorithms can be rapidly prototyped. No C code

programming is required. The software includes complete dynamic
models

Modular Servo System Setup

Modular servo system setup consists of following parts
1. Tachogenerator

2. DC motor

3. Inertial load

4. Backlash

5. Encoder

6. Gearbox with disc

7. Magnetic brake

Tachogenerator
An electromechanical generator is a device capable of producing electrical
power from mechanical energy, usually the turning of a shaft. When not
connected to a load resistance, generators will generate voltage roughly pro-
portional to shaft speed. With precise construction and design, generators
can be built to produce very precise voltages for certain ranges of shaft speeds,
thus making them well-suited as measurement devices for shaft speed in me-
chanical equipment. A generator specially designed and constructed for this
use is called a tachometer or tachogenerator.
By measuring the voltage produced by a tachogenerator, we can easily
determine the rotational speed of whatever it’s mechanically attached to.

55
 Figure 6.1: Construction of Techogenerator

One of the more common voltage signal ranges used with tachogenerators is
0 to 10 volts. Obviously, since a tachogenerator cannot produce voltage when
its not turning, the zero cannot be live in this signal standard. Tachogenera-
tors can be purchased with different full-scale (10 volt) speeds for different
applications. Although a voltage divider could theoretically be used with a
tachogenerator to extend the measurable speed range in the 0-10 volt scale,
it is not advisable to significantly overspeed a precision instrument like this,
or its life will be shortened.

Tachogenerators can also indicate the direction of rotation by the polar-

ity of the output voltage. When a permanent-magnet style DC generator’s
rotational direction is reversed, the polarity of its output voltage will switch.
In measurement and control systems where directional indication is needed,
tachogenerators provide an easy way to determine that. Tachogenerators are
frequently used to measure the speeds of electric motors, engines, and the
equipment they power: conveyor belts, machine tools, mixers, fans, etc.

Backlash
In mechanical engineering, backlash, sometimes called lash or play, is clear-
ance or lost motion in a mechanism caused by gaps between the parts. It
can be defined as, The maximum distance or angle through which
any part of a mechanical system may be moved in one direction
without applying appreciable force or motion to the next part in
mechanical sequence.

Working
The MSS setup consists of several modules mounted at the metal rail and
coupled with small clutches. The modules are arranged in the chain. The DC
motor together with tachogenerator opens the chain. The gearbox with the

56
 Figure 6.2: Explaination of Backlash

output disk closes the chain. The potentiometer module is located outside the
chain. For example the DC motor can drive activates the following modules:
inertia, backlash, encoder module, magnetic brake and the gearbox with the
output disk. The rotation angle of the DC motor shaft is measured using an
incremental encoder. Anywhere the rotational angle measurement is required
we can place the encoder. A tachogenerator is connected directly to the DC
motor and generates a voltage signal proportional to the angular velocity.

Figure 6.3: Setup of MSS

The servomechanism is connected to a computer where a control algo-

rithm is realized based on measurements of angle and angular velocity. The
system has no inner feedback for dead zone compensation. The accuracy
of measurement of velocity is 5% while the accuracy of angle measurement
is 0.1%. The armature voltage of the DC motor is controlled by PWM

57
signal. For this reason the dimensionless control signal is the scaled input
voltage, u(t) = v(t)/v max. The admissible controls satisfy |u(t)| ≤ 1 and
Vmax = 12V .
The measurement system is based on RTDAC/PCI acquisition board equipped
with A/D converters. The I/O board communicates with the power interface
unit. The whole logic necessary to activate and read the encoder signals and
to generate the appropriate sequence of PWM pulsesto control the DC mo-
tor is configured in the Xilinx chip of the RT-DAC/PCI board. All functions
of the board are accessed from the Modular Servo Toolbox which operates
directly in the MATLAB/Simulink environment

Mathematical Model of Servo System

Linear Model
A DC motor with a negligible armature inductance is given in Fig.6.4 is
described by two classical equations:

Figure 6.4: Linear Model of Servo System

v(t) = Ri(t) + Ke ω(t) (6.1)

Eq. 6.1 is electrical, while Eq. 6.2 is mechanical
Jdω
= Km i(t) − βω(t) (6.2)
dt
Where,
v(t) Input voltage
i(t) Armature current
ω(t) Angular Velocity of rotor
J Moment of inertia of moving parts
β Damping coefficient due to viscous friction

58
Ke ω(t) Back EMF
τ = Km i(t) Electro-mechanical torque

This model is linear because the static and dry kinetic friction, as well
as the saturation are neglected. By combining the electrical and mechanical
equations we obtain the equation of a first order inertial system.

Ts ω̇ = Ksm v(t) − ω(t) (6.3)

Motor gain Ksm ad motor time constant Ts are given by:

RJ
Ts = (6.4)
βR + Ke Km
Km
Ksm = (6.5)
βR + Ke Km
The transfer function has the form:
ω(s) Ksm
Gs = = (6.6)
v(s) Ts s + 1

The transfer function for the motor position has the form:

α(s) Ksm
Gs = = (6.7)
v(s) s(Ts s + 1)
Here Eq. 6.7 is Angle transfer function while Eq. 6.6 is Velocity trans-
fer function
This linear model can be expressed in the state space form as:

ẋ1 = x2

ẋ2 = ax2 + bu
Where

−1
a= <0
Ts
Ks
b= >0
Ts
The matrix state space notation has the form

59
 ẋ = Ax + Bu

y = Cx
Where,
   
0 1 0
A= , B = Ks , C = I
0 − T1s Ts
The system can be classified as a multivariable (SIMO) because it has
two measurable state variables and one control variable. The parameters Ts
and Ks must be identified by a user.

Nonlinear model
Very often small changes of the state variables are assumed. Therefore, the
control system can be considered as a linear one. However, in some applica-
tions nonlinearities in the control loop have to be taken into account. This
includes non-linear static characteristics such as hysteresis and saturation,
which may occur if the following devices are applied: operational amplifiers,
actuators, finite word length in A/D and D/A converters. Often the signal
constraint first appears for the control variable.
We will assume a nonlinear model of the DC motor in the form:

ẋ1 = x2

ẋ2 = c(u − g(x2 ))

The function g is the inverted steady state characteristics of the system,
which can be determined experimentally. The original steady state character-
istics is obtained from measurements. The results of measurements undergo
a preliminary treatment consisting of scaling (to express them in appropri-
ate units) and a shift (to remove the bias). The function g is presented. An
interesting property of the g function is that it is discontinuous at zero and
shows distinct effects of dry friction in a vicinity of the origin as shown in
Fig. 6.5

60
 Figure 6.5: Inverted static characteristics of the servo

Lab Tasks
1. Explain following in detail and their functions in modular servo system.

• Tachogenerator
• Backlash
• Inertial load
• Magnetic brake
• Gear box

61
2. Put the values of Ksm and Ts from Eq.6.4 and Eq.6.5 in the following
transfer function

ω(s) Ksm
G(s) = = (6.8)
v(s) Ts s + 1
(a) Simulate it for step and ramp response for the parameters taken
as: J = 3.2284e−6 , β = 3.5077e−6 , K = Ke = Km = 0.0274N −
m/Amp, R = 4Ω
(b) Now take J = 2 ∗ j + Roll − number and re-drive transfer function

62
3. Take into account the effect of Inductance L which was neglected in
Fig.6.4 for linearization

(a) Re-drive the transfer function G(s)

(b) Simulate it for step and ramp response for the parameters taken
in Lab task 1.
(c) Is there any difference between the two responses

63
 Kg
4. If the Techogenerator T (s) = Rf +sLf
block is connected in series with
the Motor block G(s)

(a) which model (angle or velocity) of G(s) should be used to find

voltage transfer function VVout (s)
in (s)

(b) If J = 2 ∗ j + Roll − number, determine Vout for a step input by

taking Rf = 1.2Ω, Kg = 0.274, Lf = 7µH in T (s)

64
5. Use the Velocity transfer function of motor derived in Eq. 6.6 and take
the same parameter values as given in task 2(a)

(a) Incorporate the Motor’s dead-zone block of ±2 to obtain the Si-

nusoidal response of a system
(b) what is the difference between the two responses. Give your com-
ments
(c) If Motor drives a 5:10 gear train that has backlash of 0.02 rad,
Simulate the system again with sinusoidal input to compare the
two responses

65