INTRODUCTION TO AUTOMATIC CONTROL

Control systems are an integral component of any industrial society and are necessary to provide
useful economic products for society.
Control engineering is based on the foundation of feedback theory and linear system analysis, and
integrates the concepts of network theory and communication theory. Therefore, control
engineering is not limited to any engineering discipline but is equally applicable for aeronautical,
chemical, mechanical, environmental, civil, and electrical engineering. For example, quite often a
control system includes electrical, mechanical, and chemical components. Furthermore, as the
understanding of the dynamics of business, social and economical systems increases, the ability to
control these systems will increase also.

A control system is an interconnection of components forming a system configuration that will

provide a desired response. The automatic toaster, the thermostat, the washer and dryer, the air
conditioner, the computer, the microprocessor, the space vehicles, the robots are some of the familiar
control systems.

The methods used to accomplish the control objectives can be classified, very generally, as
open-loop controls and closed-loop or feedback controls.

1.1 Open-loop control systems

Systems in which the output quantity has no effect upon the input quantity are called open-loop
control systems. Open-loop controls are, in general, calibrated systems. An open-loop control
system utilizes a controller or control actuator in order to obtain the desired response.
One of the earliest open-loop control systems was Hero’s device for opening the door of a temple.
The command input to the system (see Figure 1.1) was lighting a fire upon the altar. The
expanding hot air under the fire drove the water from the container into the bucket. As the bucket
became heavier, it descended and turned the door spindles by means of ropes, causing the
counterweight to rise. Dousing the fire could close the door. As the air in the container cooled and
the pressure was thereby reduced, the water from the bucket siphoned back into the storage
container. Thus the bucket became lighter and the counterweight, being heavier, moved down,
thereby closing the door

1. 1
 Figure 1.1 Hero’s device for opening temple doors

Other examples of open-loop control systems are:

Toaster
Assume that a toaster is set for the desired darkness of the toasted bread. The setting of the
darkness, or timer knob, represents the input quantity, and the degree of darkness of the toast
produced is the output quantity. If the degree of darkness is not satisfactory, because of the
condition of bread or some other reason, this condition in no way automatically alter the length of
time that heat is applied. The heater portion of the toaster, excluding the timer unit, represents the
dynamic part of the overall system.

1. 2
 Figure 1.2 Automatic toaster
Automatic washers and dryers
These devices change operating modes and eventually shut themselves off according to a
preprogrammed schedule, which is usually based on elapsed time.

DC Motor
Another example is the dc shunt motor. For a fixed armature voltage, the field rheostat resistance
setting can be increased to produce the desired increase in the motor speed.

Figure 1.3 DC motor

1. 3
A functional block diagram, as shown in Figure 1.4, can represent the above examples just cited
symbolically.

Figure 1.4 Functional block diagram of an Open-loop control

The desired darkness of the toast or the desired speed of the motor is the command input; the
selection of the value of time on the toaster timer or the value of the voltage applied to the motor
armature is represented by the reference-selector block; and the output of this block is identified as
the reference input. The reference input is applied to the dynamic unit that performs the desired
control function and the output of this block is the desired output.

1.2 Closed-loop control system

Systems in which the output has an effect upon the input quantity are called closed loop control
systems. The feedback action controls the input to the dynamic unit. Closed-loop or feedback
controls operate according to a very simple principle:
1. Measure the variable to be controlled.
2. Compare the measured value with the desired (commanded value) and determine the
difference (error).
3. Use this difference (amplify if necessary) to adjust the controlled variable so as to reduce the
difference (error).
The events of everyday life give evidence of some sophisticated feedback control systems. For
example, we eat when we are hungry, and we stop when we are full. This behavior suggest the
operation of a feedback loop much like the loops used to control the level of fluid in a tank. In
both cases the set point is the desired level of fullness, the process variable is some measure of
actual fullness, and the control variable is the rate of intake.

Figure 1.5 Feedback control of eating and appetite.

1. 4
Water level float regulator
The first historical feedback system claimed by the Russian is the water level float regulator
invented by Polson in 1765. The level regulator system is shown in Figure 1.6 (a). The float
detects the water level and controls the valve that covers the water inlet in the boiler.

Steam engine fly ball governor

James Watt’s fly ball governor for controlling speed, developed in 1769, can be considered the
first widely used feedback control system for controlling the speed of a steam engine. The
governor measured the speed of the output shaft and utilizes the movement of the fly ball with
speed to control the valve and therefore the amount of steam entering the engine.

Figure 1.6 (a) Water level float regulator; (b) Watt fly ball governor

1. 5
Examples of feedback control are numerous: For example in thermostat control of room
temperature a person selects the desired room temperature (command input) and adjusts the
temperature setting on the thermostat. A bimetallic coil in the thermostat is affected by the actual
room temperature (output) and the reference selector setting. If the room temperature is lower than
the desired temperature, the coil strip alters its shape and causes a mercury switch to operate a
relay, which in turn activates the furnace to produce heat in the room. When the room
temperature reaches the desired temperature, the shape of the coil strip is again altered so that the
mercury switch opens. This deactivates the relay and in turn shuts off the furnace. In this example
the bimetallic coil performs the function of a comparator. The switch, the relay, and furnace are
the dynamic elements of this closed loop control system.

Figure 1.7 Thermostat arrangement

The block diagram of a closed-loop control system is shown in Figure 1.8

Figure 1.8 Functional block diagram of a closed-loop control system

1. 6
Other examples of control systems are:

Electric supply voltage and frequency control systems

Voltage and frequency are measured at the generating station and compared with the desired
values. If the voltage is in error, the field current of the generator is adjusted; if the frequency is
not correct, the turbine speed is adjusted. Figures 1.9 and 1.10 shows the two major control loops,
with which most large generators are equipped. In Figure 1.9 the automatic voltage regulator
(AVR) loop controls the magnitude of the terminal voltage V. This voltage is continuously sensed,
rectified and smoothed. This dc signal, being proportional to V , is compared with a dc reference
Vref. The resulting error voltage after amplification serves as the input to the exciter, which
delivers the voltage V f to the generator field winding. The change in the field excitation will
affect the generator terminal voltage.

Figure 1.9 The AVR control loop for a synchronous generator.

1. 7
The automatic load-frequency control (ALFC) loop regulates the megawatt output and frequency
(speed) of the generator. The ALFC loop shown in Figure 1.10 will maintain control only during
normal (small and slow) changes in load and frequency. When load on the generator is suddenly
increased by a small amount this demand is supplied by the kinetic energy of the rotating system
causing the reduction of the speed. The speed governor senses this reduction of speed; the
governor action after amplification will operate the turbine throttle valve to admit more steam
such that the input is increased to maintain the power balance. The speed and frequency will
maintain their steady-state value according to the governor drooping characteristic. Additional
secondary control loop is required to maintain the fine adjustment of the frequency.

Figure 1.10 The ALFC loop for a synchronous generator.

1. 8
Antiaircraft radar tracking control systems
An early example of military application of a feedback control system is the aircraft radar tracking
control system shown in Figure 1.11. The radar antenna detects the position and velocity of the
target airplane, and the computer takes the information and determines the correct firing angle
for the gun. This angle includes the necessary lead angle so that shell reaches the projected
position at the same time as the airplane. The output signal of the computer, a function of the
firing angle, is fed into an amplifier, which provides power for the drive motor. The motor then
aims the gun at the necessary firing angle.

Figure 1.11 Antiaircraft radar tracking control systems.

Temperature control system

Figure 1.12 shows a schematic diagram of temperature control of an electric furnace. A
thermometer measures the temperature in the electric furnace, which is an analog device. The
analog temperature is converted to a digital temperature by an A/D converter. The digital
temperature is fed to a controller through an interface. This digital temperature is compared with
the programmed input temperature, and if there is any discrepancy (error), the controller sends out
a signal to the heater, through an interface, amplifier, and relay, to bring the furnace temperature
to a desired value.

1. 9
 Figure 1.12 Temperature control system.

Robot control system

Industrial robots are frequently used in industry to improve productivity. The industrial robot must
have some sensory devices. In low-level robot, micros witches are installed in the arms as sensory
devices. The robot first touches an object and then, through the micros switches confirms the
existence of the object in space and proceeds in the next step to grasp it.
In a high-level robot, an optical means (such as a television system) is used to scan the
background of the object. A digital computer acting as a controller recognizes the presence and
orientation of each mechanical part by a pattern recognition process that consists of reading the
code numbers attached to it. Then the robot picks up the part and moves it to an appropriate place
for assembling.

Figure 1.13 Robot using a pattern recognition process.

1. 10
Also, there have been many applications of control system theory to biomedical experimentation,
diagnosis, prosthetics, and biological control systems. The control systems under consideration range
from the cellular level to the central nervous system, and include temperature regulation and
neurological, respiratory, and cardiovascular control.
The closed loop system can be designed to provide extreme accuracy in the steady-state; also its
response time can be reduced by appropriate design. Unfortunately, the feedback link introduces
serious problem stability! Feedback systems readily become oscillators; that is, the controlled
variables fluctuate continuously at some periodic rate (frequency) and never reach the desired
steady-state condition. For any specific system, stable operation can be achieved by proper design,
but the design problem is not always an easy one to solve.

1.3 SYSTEM CLASSIFICATIONS AND LINEARIZATION

Feedback control systems may be classified in a number of ways, depending upon the purpose of
the classification.
Static or Dynamic Systems: - Static systems are composed of simple linear gains or nonlinear
devices and described by algebraic equations, and dynamic systems are described by differential
or difference equations.
Continuous-time or Discrete-time Systems: - Continuous-time dynamic systems are described
by differential equations and discrete-time dynamic systems by difference equations.
Linear or nonlinear Systems: -Linear dynamic systems are described by differential (or
difference) equations having solutions that are linearly related to their inputs. Equations
describing nonlinear dynamic systems contain one or more nonlinear terms.
Lumped or Distributed parameters: - Lumped-parameter, continuous- time, dynamic systems
are described by ordinary differential equations, and distributed parameter, continuous-time,
dynamic systems by partial differential equations.
Time-varying and Time invariant Systems: - Time-varying dynamic systems are described by
differential (or difference) equations having one or more coefficients as functions of time. Time
invariant (constant-parameter) dynamic systems are described by differential (or difference)
equations having only constant coefficients.
Deterministic or Stochastic Systems: - Deterministic systems have fixed (nonrandom)
parameters and input, and stochastic systems have randomness in one or more parameters or
inputs.
The scope of this course is limited to a consideration of linear, lumped parameter,
continuous-time, dynamic systems having only deterministic parameters and input, both
time-varying and time-invariant systems are modeled, but the analysis and design emphasis is on
time-invariant cases. Nonlinear dynamic systems may often be linearized about either static or
dynamic operating points. Static operating points, referred to as equilibrium points, occur from
the application of a constant input, such as a constant force or a constant voltage. Dynamic
operating points occur when the input varies with time to yield a nominal response as a function
of time. Examples are a nominal flight path of an aircraft and a nominal trajectory.

1. 11
1.4 The control problem
A control engineer’s first step is the formation of a suitable model of the dynamic system or
process to be controlled. The model may be validated by analyzing its performance for realistic
input conditions and then by comparing with field test data taken from the dynamic system in its
operating environment. Model validation is often achieved with the aid of computer simulation in
which the effects of varying parameters can be determined more easily. Further analysis of the
simulated model is usually necessary to obtain the model response for different feedback
configurations and parameters settings. Once an acceptable controller has been designed and
tested on the model, the feedback control strategy is then applied to the actual system to be
controlled.
When we wish to develop a feedback control system for a specific purpose, the general procedure
may be summarized as follow:
1. Choose a way to adjust the variable to be controlled; e.g., the mechanical load will be
positioned with an electric motor or an electrical resistance heater will control the temperature.
2. Select suitable sensors, power supplies, amplifiers, etc., to complete the loop.
3. Determine what is required for the system to operate with the specified accuracy in
steady-state and for the desired response time.
4. Analyze the resulting system to determine its stability.
5. Modify the system to provide stability and other desired operating conditions by redesigning
the amplifier/controller, or by introducing additional control loops.

1.5 Transfer function

In control theory, functions called transfer functions are commonly used to characterize the
input-output relations of components or systems that can be described by linear, time invariant,
differential equations.

The transfer function of a linear, time-invariant, differential equation system is defined as the ratio
of the Laplace transform of the output to the Laplace transform of the input under the assumption
that all initial conditions are zero. Consider the linear time-invariant system defined by the
following differential equation:
d nc d n −1c dc d mr d m −1r dr
an n
+ an −1 n
+" + a1 + a0 c = bn m
+ bn −1 m
+" + b1 + b0 r (1.1)
dt dt dt dt dt dt
Where c is the output, and r is the input of the system. The transfer function is obtained by taking
the Laplace transform of both sides of equation (1.1), under the assumption that all initial
conditions are zero, or

1. 12
 L[output]
Transfer function =
L[input] zero initial condition

C ( s ) bm s m + bm −1s m −1 +" b1s + b0

G ( s) = = (1.2)
R( s ) an s n + an −1s n −1 +" a1s + a0

Note that the denominator polynomial of (1.2) is the coefficient of C(s). This polynomial set equal
to zero is the characteristic equation of the differential equation (1.1). It will become apparent in
the next chapter that the ai coefficients in (1.1) are parameters of the physical system described
by the differential equation. It follows; therefore, that the characteristic equation does indeed
characterizes the system, since its roots are dependent only upon the system parameters.

1.6 Linearization of nonlinear systems

A great majority of physical systems are linear within some range of the variables. However all
systems ultimately become nonlinear as the variables are increased without limit. A system is called
linear if the principle of superposition applies. The principle of superposition states that the response
produced by the simultaneous application of two different forcing functions is the sum of the two
individual responses. Hence, for the linear system, treating one input at a time and adding the results
can calculate the response to several inputs. In the dynamic system if the cause and effects are
proportional, thus implying that the principle of superposition holds, and then the system can be
considered linear.

Consider a nonlinear system y (t ) = f ( x) shown graphically as in Figure 1.14.

Figure 1.14 A graphical representation of a nonlinear element

1. 13
If the variation of x about x0 are small enough, the nonlinear curve can be approximated by its
tangent at a.
The Taylor series expansion about the operating point is
A good approximation to the curve over a small range
df
f ( x) = f ( x0 ) + ( x − x0 )
dx x = x0
or
y − y0 = m( x − x0 ) (1.3)
df
Where m = is the slope at the operating point
dx x = x0

1. 14
 Chapter 2

Mathematical Modeling of Physical Systems

Introduction
The first step in the analysis and design of control systems is mathematical modeling of
the system. The two most common methods are the transfer function representation and
the state equation approach. The state equations can be applied to portray linear as well
as nonlinear system and will be discussed in the next chapter.
All physical systems are nonlinear to some extent; in order to use the transfer function
and linear state equations the system must first be linearized. Thus, the first task of the
control systems engineer is to make proper assumptions and approximations so that a
linear mathematical model can characterize the system.
The dynamic performance of the physical systems is obtained by utilizing the physical
laws of electrical, mechanical, fluid and thermodynamic systems. We generally model
physical systems with linear differential equations with constant coefficients when
possible. Other models can be derived from the differential equation.

2.1 Electrical Systems

Electric systems are often used extensively in a control system to filter or modify a signal
in order to obtain satisfactory performance of the system. The model of electrical systems
is determined by the application of nodal and loop analysis. The Laplace transform of the
resulting integral-differential equation are obtained with zero initial conditions. A
detailed transfer function block diagram can be obtained by denoting the relationship
between variables graphically by means of the block diagrams. Alternatively, the
s-domain equations may be solved for an output/input ratio, which is the desired transfer
function.

Example 2.1
Find the transfer function relating the capacitor voltage Vc ( s ) , to the input voltage, Vi ( s) .

R L
i (t ) +

vi (t ) C vc (t )

-
Figure 2.1 Circuit of Example 2.1
2.1
Applying KVL
di (t )
Ri (t ) + L + vc (t ) = vi (t ) (2.1)
dt
and
dvc (t )
i (t ) = c (2.2)
dt
Taking Laplace transforms with zero initial conditions, we obtain
[ R + Ls ]I ( s ) + Vc ( s ) = Vi ( s ) (2.3)
and
I ( s ) = CsVc ( s ) (2.4)
Substituting from (2.4) into (2.3), the system transfer function is
[ LCs 2 + RCs + 1]Vc ( s ) = Vi ( s )
or
Vc ( s ) 1
=
Vi ( s ) LCs + RCs + 1
2

or
1
G ( s) = LC (2.5)
R 1
s2 + s +
L LC

Vi ( s ) 1/ LC Vc ( s )
R 1
s2 + s +
L LC
2.2 Block Diagrams

The block diagram is a short hand, pictorial representation of the cause and effect
relationship between the input and output of a physical system. It provides a convenient
and useful method for characterizing the functional relationships among the various
components of a control system. The relation V2 ( s ) = G ( s )V1 ( s ) is represented by a
transfer function block as shown:
V1 ( s ) V2 ( s )
Transfer function G( s) V2 ( s ) = G ( s )V1 ( s )

2.2
 V1 ( s ) V2 ( s)
Summing point V2 ( s) = V1 ( s) ± F ( s)
±
F ( s)

V1 ( s ) V1 ( s )
Pickoff point V1 ( s ) = V1 ( s) = V1 ( s)
V1 ( s )

R (s) E (s) C (s)

Simple feedback loop G (s)
− F ( s)
H (s)

Consider the above simplified feedback system. Where

G (s) is the transfer function of the forward path,

H (s) is the transfer function of the feedback path,
R(s) is the Laplace transform of the reference input r (t )
E (s) is the Laplace transform of the actuating signal or error signal e(t )
C (s) is the Laplace transform of the controlled variable or output c (t )

From the above block diagram the following three equations may be written
C (s) = G (s) E ( s)
F ( s ) = H ( s )C ( s )
E (s) = R(s) − F (s)
Eliminating E ( s ) and F ( s ) will result in
C ( s) G ( s)
= (2.6)
R( s) 1 + G ( s) H ( s)
A block diagram representation for the electric system of Example 2.1 with Vi ( s ) as
input and Vc ( s) as output is obtained by utilizing Eqs. (2.3) and (2.4), i.e.
1
[Vi ( s ) − Vc ( s )] = I ( s )
R + Ls
and
1
I ( s ) = Vc ( s )
Cs
Applying the above pictorial representations to these equations will result in the
following block diagram.
2.3
 Vi ( s ) 1 I ( s) 1 Vc (s)
R + Ls Cs
−

Figure 2.2 Block diagram of Example 2.1.

The above feedback system where H ( s ) , is called a unity feedback system. Applying
(2.6) we have
1
V ( s) Cs ( R + Ls ) 1
T ( s) = c = =
(1) LCs + RCs + 1
2
Vi ( s ) 1 + 1
Cs ( R + Ls )

This is the same as the transfer function obtained in (2.5).

2.3 Block Diagrams Reduction

Example 2.2

Consider the block diagram shown in Figure 2.3

G3
+
R( s) E1 ( s ) E2 ( s) + C ( s)
G1 G2
− −
H2
H1

Figure 2.3 Block diagram of Example 2.2

The equation describing the above block diagram is (Write one equation for the output of
each summing point)
E1 ( s ) = R( s ) − H1C ( s ) (2.7)
E2 ( s) = G1 E1 ( s ) − H 2C ( s) (2.8)
C ( s ) = G3 E1 ( s ) + G2 E2 ( s ) (2.9)
Which describe the system completely with variables E1 ( s ) , E2 ( s) , and C ( s ) .

2.4
Method 1 Back substitution

Substituting (2.8) into (2.9) gives

C ( s ) = G3 E1 ( s ) + G2 [G1 E1 ( s ) − H 2C ( s )] (2.10)

Finally, substituting (2.7) into (2.10) and simplifying yields the desired transfer function
C ( s) G1G2 + G3
=
R ( s ) 1 + G2 H 2 + G1G2 H1 + G3 H1

Method 2 Matrix form and Application of Cramer’s Rule

Rewriting (2.7)-(2.9), we have

E1 + H1C = R ⎡1 0 H1 ⎤ ⎡ E1 ⎤ ⎡1 ⎤
G1 E1 − E2 − H 2C = 0 ⇒ ⎢ G −1 − H ⎥ ⎢ E ⎥ = ⎢0 ⎥ R
⎢ 1 2⎥⎢ 2⎥ ⎢ ⎥
G3 E1 + G2 E2 − C = 0 ⎢⎣G3 G2 −1 ⎥⎦ ⎢⎣ C ⎥⎦ ⎢⎣0 ⎥⎦

Solving for C

1 0 1
G1 −1 0
G3 G2 0 C (s) G1G2 + G3
R( s ) ⇒ =
1 0 H1 R( s ) 1 + G2 H 2 + G1G2 H1 + G3 H1
G1 −1 − H 2
G3 G2 −1

2.4 Block Diagram Transformation Rules

Block diagrams of complicated control systems may be simplified using easily derivable
transformations.

2.5
 Block Diagram Transformation Rules

1. Combining R( s) C (s) R( s) C ( s)
cascade blocks G1 G2 ⇒ G1G2

2. Combining R( s) C (s) R( s) C (s)

parallel blocks G1 ⇒ G1 ± G2
±
G2
C ( s ) = G1 R ( s ) ± G2 R( s ) ⇒ C ( s ) = [G1 ± G2 ]R( s )

3. Moving a R( s) C (s)
summing point G ⇒ R( s) G
C ( s)

ahead a blocks ± B( s) ± 1 B(s)

G
C ( s) = GR( s ) ± B( s ) ⇒ C ( s ) = G[ R( s ) ± 1 B( s )]
G
4. Moving a R( s) C (s) R( s) C ( s)
summing point G ⇒ G
beyond a blocks N ( s ) ± N (s) ±
G
C ( s ) = G[ R( s ) ± N ( s )] ⇒ C ( s) = GR( s ) ± GN ( s )

5. Moving a R( s) C (s) R (s) C (s)

pickoff point G ⇒ G
Y ( s) Y ( s)
ahead of a blocks G
Y ( s) = C ( s) = GR( s) ⇒ Y ( s) = GR( s)

6. Moving a R( s) C (s) R (s) C (s)

pickoff point G ⇒ G
Y ( s) Y ( s) 1
beyond a blocks
G
Y (s) = R( s) ⇒ Y ( s ) = 1 C ( s ) = 1 GR( s ) = R( s)
G G

7. Eliminating R( s) E (s) C (s)

G R( s) G C ( s)
single feedback ⇒
− 1 + GH
loop H

Figure 2.4 Block diagram transformational rules

2.6
Example 2.3

Apply the block diagram transformation rules to example 2.2 to determine the system
transfer function

G3
+
R( s) E1 ( s ) E2 ( s) + C ( s)
G1 G2
− −
H2
H1

Moving the summing point ahead of a block

1
G3
G2
R( s) C (s)
G1 G2
− −
H2
H1

Eliminating parallel block and the feedback loop

R( s) G3 G2 C ( s)
G1 +
− G2 1 + G2 H 2
G G + G3
= 1 2
G2
H1
Eliminating the feedback loop

R( s) G1G2 + G3 C ( s)
1 + G2 H 2 + G1G2 H1 + G3 H1

2.7
Example 2.4
Using block diagram transformation Determine C / R for the following systems

H3

− +
R( s)
G1 G2 + G3
C ( s)
− −
H2
H1

Moving takeoff point beyond block G3 and eliminating a feedback loop

R(s) G2G3 C ( s)
G1
1 + G2 G 3 H 3
− −
H2
G3
H1
G3

Eliminating the next feedback loop

R(s) G2G3 C ( s)
G1 1 + G2 G3 H 3 + G2 H 2
−

H1
G3

Eliminating the last feedback loop

R( s) G1G2G3 C ( s)
1 + G1G2 H1 + G2 H 2 + G2 G3 H 3

2.8
Example 2.5

C ( s)
Determine
R( s)

H3

− C (s)
R(s)
G1 G2 G3 G4
− −

H1 H2
Moving one takeoff beyond G4 and eliminating the feedback loop G3G4 H 2

H3
G4
− G3G4 C (s)
R( s)
G1 G2 M1 =
1 + G3G4 H 2
−

H1

Moving takeoff point beyond block M 1 and eliminating the top feedback loop

R( s) G2 G 4 M 1 C (s)
G1 M2 =
G4 + G 2 H 3 M 1
−
H1
M1

Substituting for M 1 , we have

R( s) G1G2G3G4 C ( s)
1 + G3G4 H 2 + G2G3 H 3
−

H1 + G3G4 H1 H 2
G3G4

Eliminating the feedback loop

R( s) G1G2G3G4 C ( s)
(1 + G1G2 H1 )(1 + G3G4 H 2 ) + G2G3 H 3

2.9
2.5 Transfer Function (Multivariable System)

The definition of transfer function can be extended to a system with a multiple number of
inputs and outputs (multivariable system). Consider a system with two inputs and two
outputs as shown in Figure 2.5.
R1 ( s ) C1 ( s )
Multivariable
R2 ( s ) system C2 ( s )

Figure 2.5
When dealing with the relationship between one input and one output, it is assumed that
all other inputs are set to zero, e.g.,
C1 ( s) C1 ( s)
G11 = G12 = (2.11)
R1 ( s) R2 = 0 R2 ( s) R1 = 0

C2 ( s) C2 ( s )
G21 = G22 = (2.12)
R1 ( s ) R2 = 0 R2 ( s ) R1 = 0
Since the principle of superposition is valid for linear system, the total effect on any
output variable due to all the inputs acting simultaneously is obtained by adding up the
outputs due to each input acting alone.
C1 ( s ) = G11 R1 ( s ) + G12 R2 ( s )
(2.13)
C2 ( s ) = G21 R1 ( s ) + G22 R2 ( s )

Representing the above by matrix equation we have

C (s) = G (s) R(s) (2.14)
Where

⎡G G12 ⎤
G = ⎢ 11 ⎥
⎣G21 G22 ⎦

2.10
Example 2.6
C ( s)
Determine
R2 ( s )

R2 ( s)
H3

−
R1 ( s) C (s)
G1 G2 G3
− −
H2
H1

Redrawing the block diagram with R1 = 0 and moving the second summing point beyond
block G2

R2 ( s)
G2 H 3

−
C (s)
G3
− −
H2
G1G2 H1

Combining parallel blocks

2.11
 R2 ( s )
G2 H 3

−
C (s)
G3
−

G1G2 H1 + H 2

Eliminating the top feedback loop

R2 ( s )

1 C (s)
G3
1 + G2 H 3 −

G1G2 H1 + H 2

Combining cascade blocks

R2 ( s ) C (s)
G3
−
G1G2 H1 + H 2
1 + G2 H 3
Eliminating the feedback loop

R2 ( s ) G3 (1 + G2 H 3 ) C ( s)
1 + G2 H 3 + G3 H 2 + G1G2G3 H1

2.12
2.7 Signal Flow Graph

A signal Flow Graph is a pictorial representation of the simultaneous equations

describing a system. It graphically displays the transmission of signals through the
system. It is easier to draw and therefore easier to manipulate than the block diagram.
The properties of signal flow graph are as follow:

Let us first consider the simple equation

Y = AX (2.15)
Where A is a mathematical operator mapping X into Y , and is called the transmission
function. The signal flow graph for the above equation is

Node A Node
X Bracnh Y

Figure 2.6

Variables X and Y are represented by a small dot called node, and the transmission
function A is represented by a line with an arrow, called branch. Signals travel along
branches only in the direction described by the arrows of the branches. The branch
directing from node X to node Y represent the dependence of variable X upon Y , but not
the reverse.

2.8 Signal Flow Graph Definitions

The following terminology is frequently used in signal flow graph theory.

Node Nodes on a signal flow graph represent system variables.

Branch Branches are unidirectional paths that connect the nodes. An arrow is
assigned to indicate the direction of cause and effect.

Input node An input node has only outgoing branches

Output node An output node has only incoming branches.

Path A path is a continuous connection of branches with arrows in the same

direction.

Loop A loop is a path that starts and ends on the same node with all other nodes
in the loop touched only once.

Common Common node is a node that is contained in two or more loops.

node

2.13
Nontouching Nontouching loops are loops that have no common nodes.
loop

Forward path A forward path starts at an input node, ends at an output node, and
touches no node more than once. A forward path may traverse one or
more feedback branches in proceeding from input to output nodes.
Gains Gains for paths and loops are defined as the products of branch gains for
the paths or loops.

Signal Flow Graph Algebra

1. Addition Rule: - The value of a variable designated by a node is equal to the sum of
all signals entering the node, i.e. equation
Y = a1 X 1 + a2 X 2 + a3 X 3 + a4 X 4
is represented by
X3
X4
X2

a3
a2
a4

X1 a1
Y
Figure 2.7
2. Transmission Rule: - The value of a variable designated by a node is transmitted on
every branch leaving that node. In other words, the equation
Y1 = b1 X , Y2 = b2 X , Y3 = b3 X
is represented by
Y1
b1

b2
X Y2

b3

Y3
Figure 2.8
2.14
3. Parallel Branches: - Parallel branches in the same direction connecting two nodes can
be replaced by a single branch with gain equal to the sum of the gains of the parallel
branches

4. Multiplication Rule: - A cascade connection or series connection of unidirectional

branches can be replaced by a single branch with gain equal to the product of branch
gains.

2.9 General Gain Formula (Mason's Gain Rule)

It is possible to simplify signal flow graph in a manner similar to that of block diagram
reduction. But it is also possible, and much less time-consuming, to write down the input-
output relationship by inspection from the original signal flow graph. This can be
accomplished by the use of general gain formula, known as Mason’s Gain Formula.
C (s) n M k ∆ k
G ( s) = =∑ (2.16)
R ( s ) k =1 ∆
where
r = Input node variable
c = Output node variable
G = Gain between input r and output c
n = Total number of forward paths
M k = Gain of the k th forward path
∆ = 1 - (sum of all individual loop gains) + (sum of gain products of all possible
combinations of two nontouching loops) - (sum of the gain products of all possible
combinations of three nontouching loops) + . . .

∆ k = The ∆ for the part of the signal flow graph which is nontouching with the k th
forward path. If all loops touches path k , ∆ k = 1 .

It must be emphasized that the gain formula can be applied only between an input node
and an output node.

2.15
Example 2.7

Determine the close loop transfer function for the block diagram of example 2.2 by using
Mason's gain formula.

The following signal flow graph is obtained from Example 2.2 block diagram or the
following equations describing the block diagram.

E1 ( s ) = R ( s ) − H1C ( s )
E2 ( s ) = G1 E1 ( s ) − H 2C ( s )
G3
C ( s ) = G3 E1 ( s ) + G2 E2 ( s )

1 G1 G2 1
R( s) C ( s)
E1 ( s ) E2 ( s ) C (s)

−H2

− H1

There are two forward paths with gains

M 1 = G1G2
M 2 = G3
There are three individual loops, with gains
L1 = −G2 H 2
L2 = −G1G2 H1
L3 = −G3 H1
Since these three single loops have at least one common node, there is no combination of
two nontouching loops. Therefore ∆ is given by

∆ = 1 − ( L1 + L2 + L3 )
= 1 + G2 H 2 + G1G2 H1 + G3 H1

Both paths have at least one node in common with every single loop. Therefore
∆1 = ∆ 2 = 1 .

C ( s) G1G2 + G3
=
R ( s ) 1 + G2 H 2 + G1G2 H1 + G3 H1

which agrees with the result in example 2.2.

2.16
The above equations describing the system of Example 2.7 can be written in MATLAB
and using the Symbolic Toolbox “solve” function, the transfer function can be obtained
as follows:

syms R E1 E2 C G1 G2 G3 H1 H2 % Define the system variables and gains

eq1 = E1 + H1*C -R;
eq2 = G1*E1 - E2 - H2*C;
eq3 = G3*E1+ G2*E2 - C;
S = solve(eq1, eq2, eq3, E1, E2, C);
G=S.C/R % Obtains the transfer function C(s)/R(s)

The result is

G=

(G3+G2*G1)/(G3*H1+G2*G1*H1+G2*H2+1)

2.17
Example 2.8

For the signal flow graph shown, use the general gain formula to find the transfer
C (s)
function
R( s )

G4

R( s) 1 G1 G2 G3 1 1
C (s)
E2 ( s ) E3 ( s ) E4 ( s ) C (s)
E1 ( s )
− H1 −H2

−H3

Forward paths:
M 1 = G1G2G3
M 2 = G1G4

Loops:
L1 = −G1 H1
L2 = −G3 H 2
L3 = −G1G2G3 H 3

Nontouching loops:

L1 L2 = G1 H1G3 H 2

Therefore ∆ is given by

∆ = 1 − (−G1 H1 − G3 H 2 + G1G2G3 H 3 ) + G1 H1G3 H 2

M1 touches all loops, therefore ∆1 = 1

L2 does not touch forward path M 2 , therefore ∆ 2 = 1 + G3 H 2
The transfer function is

C (s) G1G2G3 + G1G4 (1 + G3 H 2 )

=
R ( s ) 1 + G1 H1 + G3 H 2 + G1G2G3 H 3 + G1 H1G3 H 2

The following statements

2.18
syms R E1 E2 E3 E4 C G1 G2 G3 G4 H1 H2 H3
eq1 = E1 + H1*E2 + H3*E4 - R;
eq2 = E2 - G1*E1;
eq3 = E3 - G2*E2 + H2*E4;
eq4 = E4 - G3*E3;
eq5 = C- E4 - G4*E2;
S = solve(eq1, eq2, eq3, eq4, eq5, E1, E2, E3, E4, C);
G=S.C/R % Obtains the transfer function C(s)/R(s)

Results in

G=

G1*(G3*G2+G4+G4*H2*G3)/(1+H2*G3+G1*H1+H1*G1*H2*G3+H3*G3*G2*G1)

2.19
Example 2.9
A system with several feedback loops and forward paths is shown. Using Mason's rule,
find the transfer function of the system

G7 G8

R ( s) 1 G1 G2 G3 G4 G5 G6 1
C (s)
C (s)

−H4
− H1
−H 2

−H3

The forward paths are

M 1 = G1G2G3G4G5G6 ⇒ ∆1 = 1
M 2 = G1G2G7G6 ⇒ ∆ 2 = 1 − L5 = 1 + G4 H 4
M 3 = G1G2G3G4G8 ⇒ ∆3 = 1

The feedback loops are

L1 = −G2G3G4G5 H 2
L2 = −G5G6 H1
L3 = −G8 H1
L4 = −G2G7 H 2
L5 = −G4 H 4
L6 = −G1G2G3G4G5G6 H 3
L7 = −G1G2G7G6 H 3
L8 = −G1G2G3G4G8 H 3

Loop L5 does not touch loops L4 and L7 . Also L3 does not touch L4 .
∆ = 1 − ( L1 + L2 + L3 + L4 + L5 + L6 + L7 + L8 ) + L5 L4 + L5 L7 + L3 L4

C ( s) M1 + M 2 ∆ 2 + M 3
G ( s) = =
R( s) ∆
Substitute and obtain the transfer function

2.20
2.10 Modeling of Mechanical System

Most feedback control systems contain mechanical as well as electrical components. The
motion of mechanical elements are either translational, rotational, or a combination of
both. The equations governing the motions of mechanical systems are formulated from
Newton’s law of motion.

2.11 Translational Motion

This is the motion of a rigid body along a straight line. Newton’s Law of motion states
that the algebraic sum of forces acting on a rigid body in a given direction is equal to the
product of the mass of the body and its acceleration in the same direction. i.e.

∑ forces = Ma (2.17)

For translational motion, the following elements are usually involved.

Mass- Mass is considered as an indication of the property of an element, which stores the
kinetic energy of translational motion. It is analogous to inductance of electrical
W
networks, M =
g
g is the acceleration of the body due to the force of gravity, g = 9.81 m/sec2 or 32.2
ft/sec2.
The force equation is written as
x(t )

M f (t )

d 2 x(t ) du (t )
f (t ) = Ma(t ) = M 2
=M (2.18)
dt dt
Spring- A spring is considered to be an element that stores potential energy. It is
analogous to a capacitor in electrical network. Its behavior may be approximated by a
linear relationship.

x2 (t ) x1 (t )
K
f (t ) f (t )
f (t ) = K [ x1 (t ) − x2 (t )] (2.19)

where K is the spring stiffness (or constant) measured in N/m.

Friction for Translational Motion: - Whenever there is motion between two elements,
2.21
frictional forces exist. In practice, frictional forces may be assumed to be viscous friction.
Viscous friction represents a retarding force that is a linear relationship between the
applied force and velocity.
x2 (t ) x1 (t )
B
f (t ) f (t )

⎡ dx (t ) dx (t ) ⎤
f (t ) = B ⎢ 1 − 2 ⎥ (2.20)
⎣ dt dt ⎦
where B is the viscous frictional coefficient in N-s/m

Example 2.10
Find the transfer function for the system shown in Figure 2.9 (a).

K x(t )
d 2 x(t )
M
dt 2
M f (t ) dx(t ) M f (t )
B
dt
Kx (t )

B (a) (b)

Figure 2.9
We draw the free-body diagram as shown in Figure 2.9 (b). The applied force is directed
to the right. Forces due to the acceleration of mass, friction and spring impede the motion
and act to oppose it. These forces point to the left as shown. Using Newton’s law of
motion we have
d 2 x(t ) dx(t )
f (t ) = M 2
+B + Kx(t )
dt dt

Taking the Laplace transform, we have

F ( s ) = [ Ms 2 + Bs + K ] X ( s)
Taking the displacement of the mass x(t ) as the output and the applied force f (t ) as the
input, the transfer function is
X ( s) 1
G ( s) = =
F ( s) Ms + Bs + K
2

2.22
Analogous Systems

We can obtain an analogous electrical system for the above mechanical system.

Mechanical System
u (t )
u (t ) x(t ) u (t )
B x(t ) K x(t )
f (t ) M f (t )
f (t )
f = Kx
f = Bx f = Mx
= ∫ Kudt
= Bu = Mu

Electrical System
R i L i C i
di
v = Ri v=L 1
C∫
dt v= idt

Figure 2.10

From the above relations we can observe the following analogy

Mechanical system Electrical system

Force f emf v
Velocity u Current i
Frictional coefficient
B Resistance R
Mass M Inductance L
1
Compliance Capacitance C
K
From the force equation it is easy to verify that the above mechanical system is analogous
to a series RLC circuit.
1 K
M K Ms s
B B
u U (s)
+ +
f (t ) F (s)
− −

(a) Time-domain (b) s-domain

f = Bu + Mu + K ∫ udt K
F ( s ) = [ Ms + B + ]U ( s )
s
= [ Ms + Bs + K ] X ( s )
2

Figure 2.11
2.23
Example 2.11
X 2 (s)
The system shown is initially at rest. Find the transfer function G ( s ) =
F (s)

u2 (t ) u1 (t )
B x2 (t ) x1 (t )
K
M f (t )

To draw the electric circuit analogy with velocity analogous to current, note that the
spring is moving with velocity (u1 − u2 ) . Clearly, we have two loops and the circuit is as
shown.
M B

+ u1 1 u2
f (t )
− K

f (t ) = K ∫ (u1 − u2 )dt ⇒ f (t ) = K ( x1 − x2 )

Mu2 + Bu2 + K ∫ (u2 − u1 )dt = 0 ⇒ Mx2 + Bx2 + K ( x2 − x1 ) = 0

Taking Laplace transform, we get

F ( s ) = K [ X 1 ( s ) − X 2 ( s )]
[ Ms 2 + Bs + K ] X 2 ( s ) − KX 1 ( s ) = 0
Alternatively, we can draw the s-domain equivalent circuit and then write the mesh
equations.

Writing mesh equations by inspection in matrix form, we get

Ms B

⎡ K K ⎤ ⎡ U1 ( s ) ⎤ ⎡1 ⎤
+ K ⎢ s − ⎥⎢ ⎥ ⎢ ⎥
U1 U2 s
F ( s) ⇒ ⎢ ⎥⎢ ⎥ = ⎢ ⎥ F ( s)
− s
⎢− K K⎥⎢ ⎥ ⎢ ⎥
Ms + B +
⎣⎢ s s ⎦⎥ ⎣⎢U 2 ( s ) ⎦⎥ ⎣⎢ 0 ⎦⎥

The transfer function with U 2 ( s) as output and F ( s ) as input can be found by using
Cramer’s rule

2.24
 K
1
s
F (s)
K K
− 0
U 2 ( s) = s = s F ( s)
K K K⎛ K ⎞ K2
− ⎜ Ms + B + ⎟ − 2
s s s⎝ s⎠ s
K K
− Ms + B +
s s

Reducing the above expression result in

U 2 ( s) 1
=
F ( s ) Ms + B
Since U 2 ( s ) = sX 2 ( s ) then the transfer function in terms of the output X 2 ( s ) is

X 2 ( s) 1
G ( s) = =
F ( s ) s ( Ms + B)

Example 2.12
The system shown is initially at rest. Draw the electric circuit analogy and write the s-
domain equation describing the mechanical system.

M1 M2 B2
K2 B2
B1
M2 + u1 u2 1
⇒ f (t ) 1
− K2
K1 B1 x2 u2 K1

M1 (b)

f (t )
x1 u1
(a)

To draw the electric circuit analogy with velocity analogous to current, note
that K1 and B1 are moving with velocity (u1 − u2 ) . Clearly, we have two loops and the
circuit is as shown in (b) above.

2.25
Or the s-domain circuit is
M 1s M 2s B2

B1
+ U1 ( s ) U 2 ( s) K2
F ( s)
− K1 s
s

(c)
Writing mesh equations by inspection in matrix form, we get
⎡ K1 ⎛ K ⎞ ⎤ ⎡ U1 ( s ) ⎤ ⎡1 ⎤
⎢ M 1s + B1 + s − ⎜ B1 + 1 ⎟ ⎥⎢ ⎥ ⎢ ⎥
⎢ ⎝ s ⎠ ⎥⎢ ⎥ = ⎢ ⎥ F ( s)
⎢ ⎛ K1 ⎞ K1 K 2 ⎥ ⎢ ⎥ ⎢ ⎥
⎢ − ⎜ B1 + ⎟ M 2 s + B1 + B2 + + ⎥ ⎣⎢U 2 ( s) ⎦⎥ ⎣⎢ 0 ⎦⎥
⎣ ⎝ s ⎠ s s ⎦
Or in terms of displacements, since U1 ( s) = sX 1 ( s) and U 2 ( s ) = sX 2 ( s ) , we have
⎡ M 1s 2 + B1s + K1 − ( B1s + K1 ) ⎤ ⎡ X 1 ( s ) ⎤ ⎡1 ⎤
⎢ ⎥ =
M 2 s + ( B1 + B2 ) s + K1 + K 2 ⎦ ⎢⎣ X 2 ( s ) ⎥⎦ ⎢⎣ 0 ⎥⎦
F (s)
⎣ − ( B1s + K1 )
2

Using Cramer’s rule we can find the transfer function.

Example 2.13
The system shown is initially at rest. Draw the electric circuit analogy and write the s-
domain equation describing the mechanical system.
u2 u1
x2 x1
M2 M1 f (t )
K2 K1
B2 (a) B1

To draw the electric circuit analogy with velocity analogous to current, note that K1 is
moving with velocity (u1 − u2 ) . Clearly, we have two loops and the s-domain circuit is as
shown in (b).
M 1s B1 M 2s B2

+ U1 ( s ) K1 U 2 (s) K2
F ( s)
− s s

(b)

2.26
 ⎡ K1 K1 ⎤ ⎡ U1 ( s ) ⎤ ⎡ 1 ⎤
⎢ M 1s + B1 + s s
− ⎥ ⎢ ⎥ ⎢ ⎥
⎢ ⎥⎢ ⎥ = ⎢ ⎥ F ( s)
⎢ M 2 s + B2 + 1 + 2 ⎥ ⎢ ⎥ ⎢ ⎥
K K K
− 1
⎢⎣ s s s ⎥⎦ ⎢⎣U 2 ( s ) ⎥⎦ ⎣⎢ 0 ⎦⎥

Or in terms of displacements since U1 ( s ) = sX 1 ( s ) and U 2 ( s ) = sX 2 ( s ) , we have

⎡ M 1s 2 + B1s + K1 − K1 ⎤ ⎡ X 1 ( s ) ⎤ ⎡1 ⎤
⎢ ⎥⎢ ⎥ = ⎢ ⎥ F (s)
⎣ − K1 M 2 s + B2 s + K1 + K 2 ⎦ ⎣ X 2 ( s) ⎦ ⎣0 ⎦
2

Using Cramer’s rule we can find the transfer function.

2.12 Rotational Motion

The variables generally used to describe the rotational motion are torque; angular
acceleration, α ; angular velocity, ω ; and angular displacement, θ . The following
elements are usually involved with the rotational motion.
Inertia: - Inertia, J , is considered as an indication of the property of an element, which
stores the kinetic energy of rotational motion. Moment of inertia of a given disk about its
axis is given by
1
J= Mr 2 (2.21)
2
where M is the mass of the disk and r its radius. Torque due to acceleration is
d 2θ dω
T = Jα = J 2 = J (2.22)
dt dt
Torsional spring: - the torque produced by a torsional spring is
K

τ (t ) θ1 (t ) θ 2 (t ) τ (t )

τ (t ) = K [θ1 (t ) − θ 2 (t )] (2.23)
where K is the spring constant in N-m⁄rad.
Friction for rotational motion: - As with the linear case viscous frictional torque for
rotational motion is

2.27
 B

τ (t ) θ1 (t ) θ 2 (t ) τ (t )

⎛ dθ1 (t ) dθ 2 (t ) ⎞
τ (t ) = B ⎜ − ⎟ = B[ω1 (t ) − ω 2 (t )] (2.24)
⎝ dt dt ⎠

Example 2.14
The rotational system shown consists of a cylinder of moment of inertia J. The viscous
friction coefficient is B and the spring has stiffness K . A torque is applied to the shaft as
θ ( s)
shown. Write the equations describing the system and obtain the transfer function
T ( s)

K
J
τ (t ) B
θ (t )

The s-domain electric circuit analogy is

B Js

Ω( s )
+ K
T ( s)
− s

K
T ( s ) = ( Js + B + )Ω ( s ) or
s
T ( s ) = ( Js 2 + Bs + K )θ ( s )

1
G ( s) =
Js + Bs + K
2

Example 2.15
Write the equations describing the system shown below.
K1 B3 K2
τ (t )
J1 J2
θ1 (t ) θ 2 (t ) B
1 B2 θ 3 (t )

2.28
The s-domain electric circuit analogy is
B1 J1 s J2s B2

+ K1 K2
T ( s) Ω1 ( s ) Ω2 ( s) B3 Ω3 ( s )
− s s

Writing mesh equations result in

K
T ( s ) = 1 (Ω1 − Ω 2 )
s
K1
( B1 + J1s )Ω 2 + B3 (Ω 2 − Ω3 ) + (Ω 2 − Ω1 ) = 0
s
K2
( B2 + J 2 s +)Ω3 + B3 (Ω3 − Ω 2 ) = 0
s
or in matrix form we have

⎡ K1 K1 ⎤⎡ ⎤ ⎡1 ⎤
⎢ s − 0 ⎥ ⎢ Ω1
s ⎥ ⎢ ⎥
⎢ ⎥⎢ ⎥ ⎢ ⎥
⎢ − K1 B1 + B3 + J1s +
K1
− B3 ⎥ ⎢Ω ⎥ = ⎢0 ⎥ T (s)
⎢ s s ⎥⎢ 2 ⎥ ⎢ ⎥
⎢ ⎥
⎢ 0 K ⎢ ⎥ ⎢ ⎥
− B3 B2 + B3 + J 2 s + 2 ⎥ ⎢ Ω ⎥ ⎢0 ⎥
⎢⎣ s ⎥⎦ ⎣ 3 ⎦ ⎣ ⎦

2.13 Electromechanical Systems

DC Machines
A machines is basically a torque transducer that converts electric energy into mechanical
energy or vise versa. Dc machines depending on the type of excitations are classified as
permanent magnet, separately excited, and self excited machines. Self-excited machines
are either series field excited, shunt field excited or compound excitation.
The emf induced in the armature of a dc machine is proportional to the angular speed of
the armature ω , and the field flux φ . When saturation is neglected, the field current is
proportional to the field flux (φ ∝ I f ) , and the induced emf can be written as

e = K f I fω (2.25)

The electromechanical torque produced in the armature is proportional to the product of

armature and field currents, i.e.

τ = K f Ia I f (2.26)
DC machines are used in control system as a generator to sense the speed, i.e., as
2.29
tachometer. Also, DC machine are used widely as motor in control systems, for example
to control the position of an antenna. Because of the linear characteristics of the dc motor
compared to the AC motor, DC motors are more suitable for control system applications.
With the development of the Alnico, ceramic and rare-earth permanent magnet, motors
offers very high torque to volume at a reasonable cost. Also, the advancement in the
brush-and-commutator technology has resulted in highly reliable DC motors with very
low inertia and very small time constant. These motors are used widely in computer
peripheral equipment such as tape drives, printers, disk drives, as well as in automation
and machine tool industries. Many other motors such as stepper motor, AC servomotor
(two-phase induction motor), and synchros are also employed in control systems. Below
the transfer function model of a DC motor is developed. Same procedure can be used to
obtain a model for other electromechanical devices.

DC Motor Model
Consider a separately excited (or a permanent magnet) dc motor with constant field
current as shown in Figure 2.12.
Ra La Fixed field
or PM
+
ia (t )

v(t ) ea (t ) M J
B
τ (t ) θ (t )
-

Figure 2.12 A Separately excited dc motor

For constant field current I f let K m = Kτ = K f I f , then from (2.25) and (2.26) the motor
back emf and the developed torque are
e = K mω (2.27)
τ = Kτ I a (2.28)
K m and Kτ are numerically equal, K m is expressed in Volt-sec/Rad, whereas Kτ is
expressed in N-m/A.
You may draw the electric circuit analogy for the mechanical system as shown in Figure
2.13.

2.30
 Ra La B J
+ ia (t ) ω (t )
+
v(t ) ea (t ) M τ (t )
-

- ea (t ) = K mω
τ (t ) = Kτ I a
Figure 2.13
Applying the KVL to the armature circuit and the electric circuit analogy, we have
di (t )
v(t ) = Ri (t ) + L + e(t ) ⇒ V ( s ) = [ R + Ls ]I a ( s ) + Ea ( s ) (2.29)
dt
dω (t )
τ (t ) = Bω (t ) + M ⇒ T( s ) = [ B + Js ]Ω( s ) (2.30)
dt
You may also draw the s-domain equivalent circuit and write the KVL equations directly
in s-domain. From (2.29), (2.28), (2.30), and (2.27), we have the following four s-domain
equations,
1
I a ( s) = [V (s) − Ea ( s)]
Ra + La s
T ( s ) = Kτ I a ( s )
(2.31)
1
Ω( s ) = T ( s)
B + Js
Ea ( s) = K m Ω( s)
dθ
The angular velocity ω = or in s-domain Ω( s ) = sθ ( s ) , therefore the angular position
dt
in s-domain is written as
1
θ ( s ) = Ω( s ) (2.32)
s
From (2.31) and (2.32), the dc motor block diagram is drawn as shown in Figure 2.14
V ( s) 1 I a ( s) T ( s) 1 Ω( s ) 1 θ ( s)
Kτ
Ra + La s B + Js s
−
Ea ( s )

Km

Figure 2.14 Block diagram representation of a dc motor

2.31
From the above block diagram, it is clear that the dc motor model has two time constants,
L
one associated with the armature circuit inductance, τ 1 = a , and the other the mechanical
Ra
J
time constant due to the motor inertia constant τ 2 = . In most cases the electrical circuit
B
time constant is much smaller than the mechanical circuit time constant and La may be
neglected.
From the block diagram the dc motor transfer function is
Kτ
θ ( s) s ( Ra + La s)( B + Js) Kτ
= =
V (s) 1 + Kτ K m s[( Ra + La s )( B + Js) + Kτ K m ]
( Ra + La s)( B + Js)
or
θ (s) Kτ
G (s) = = (2.33)
V ( s ) s[ JLa s + ( BLa + JRa ) s + ( BRa + Kτ K m )]
2

Gear Trains

Gear train, lever, or timing belt over a pulley is a mechanical device that transmits energy
from one part of the system to another in such a way that force, torque, speed, and
displacement may be altered. Gear trains in mechanical systems serve the same purpose,
as do transformers in electrical circuits; gear trains are thus mechanical transformers.
First we consider two ideal gears as shown in Figure 2.15. In the idea gear train the
inertia and friction of the gears are neglected. Later on we will consider these external to
the gears.

T1 θ1
r1

ω1 N1 r1 θ1
N2 ω2 θ
r2 2

r2
T2 θ 2

Figure 2.15 Gear train

T1 , T2 = Torque transmitted to gears
θ1 , θ 2 = Angular displacements
N1 , N 2 = Number of teeth
r1 , r2 = Radius of gears
2.32
ω1 , ω 2 = Angular velocities

The number of teeth on the surface of gears is proportional to the radius r1 and r2 , i.e.,

N1 r1
= (2.34)
N 2 r2
The arc distance traversed on the surface of gears of the two meshing gears is the same,
i.e.,
r1θ1 = r2θ 2 (2.35)
Assuming no losses, the work done by gear 1 is equal to the work done on gear 2.,
therefore
T1θ1 = T2θ 2 (2.36)
or mechanical power input is equal to the mechanical power output, i.e.,

T1ω1 = T2ω 2 (2.37)

From (2.34)-(2.37), we have

T1 θ 2 ω 2 r1 N1
= = = = =a (2.38)
T2 θ1 ω1 r2 N 2

The above is analogous to the ideal transformer

N1ii N 2
+ ω1 ω2 +

T1 T2

− −
T1 ω 2 N1
= = =a
T2 ω1 N 2

Figure 2.16
In practice, real gears have inertia and friction between coupled gears, and they are
included external to the ideal turns ratio.

Example 2.16
In the gear train shown J1 and B1 are the inertia and friction of gear 1, J 2 and B2 include
the inertia and friction of gear 2 and the load. A spring with a fixed end is connected to
gear 2. A torque T is applied to the input gear. Obtain a transfer function model with
T as input and θ1 as output.

2.33
 T J1
θ1 ω1 B1 N1
T1
N2
B2 K
J2
N1 θ 2 ω2

Figure 2.17
The s-domain electric circuit analogy is drawn as shown. This is analogous to a transformer on load.
J1 s N J2s
B1 a= 1 B2
N2
Ω1 ( s ) Ω2 (s)
+ K ( s)
T (s) T1 ( s ) T2 ( s )
- s

T1 Ω 2 ( s ) N1
= = =a
T2 Ω1 ( s ) N 2
Figure 2.18

Applying analogous KVL equation to the secondary, we have

K
T2 ( s ) = [ B2 + J 2 s + ]Ω 2 ( s ) (2.39)
s
Substituting for T2 , and Ω 2 in terms of the primary quantities, we get

N2 ⎛ K⎞N
T1 ( s) = ⎜ B2 + J 2 s + ⎟ 1 Ω1 ( s ) or
N1 ⎝ s ⎠ N2
2
⎛ K ⎞⎛ N ⎞
T1 ( s ) = ⎜ B2 + J 2 s + ⎟ ⎜ 1 ⎟ Ω1 ( s ) (2.40)
⎝ s ⎠ ⎝ N2 ⎠
Applying analogous KVL equation to the primary, we have
T1 ( s ) = [ B1 + J1s + ]Ω1 ( s ) + T1 (2.41)
Substituting for T1 in (2.41) from (2.40), we get
⎛ K⎞ 2
T1 ( s ) = [ B1 + J1s + ]Ω1 ( s ) + ⎜ B2 + J 2 s + ⎟ ( a ) Ω1 ( s ) or
⎝ s⎠
⎛ K⎞
T1 ( s ) = ⎜ B1 + J1s + a 2 B2 + a 2 J 2 s + a 2 ⎟ Ω1 ( s ) (2.42)
⎝ s⎠

2.34
Similar to the transformer equivalent circuit reflecting from gear 2 to gear 1 we can draw
the equivalent circuit as shown in Figure 2.18. Where the secondary parameters are
transferred to the primary with the secondary parameters multiplied by the square of the
turn ratios.

B1 J1 s a 2 B2 a2 J2s

+ Ω1 ( s )
K
T ( s)
- a2
s

Figure 2.19 The equivalent circuit of the gear train referred to the primary

Ω1 ( s ) 1
G(s) = =
T (s) ( J + a 2 J )s + ( B + a 2 B ) + a 2 K
1 2 1 2
s

With θ1 as output, the transfer function is

θ1 ( s) 1 1
G ( s) = = = (2.42)
T ( s) ( J1 + a J 2 ) s + ( B1 + a B2 ) s + a K J eq s + Beq s + a 2 K
2 2 2 2 2

where Beq = B1 + a 2 B2 is the equivalent friction referred to the primary gear, and
J eq = J1 + a 2 J 2 is the equivalent inertia referred to the primary gear.

2.17 Sensors
Sensing devices are needed to provide measurements of the plant variables for use in the
feedback control circuits. Some of the common sensors are as follow:

Potentiometers

Angular Displacement Potentiometers Rotory Potentiometer

Potentiometer is a transducer that converts mechanical energy into electrical energy. The
output is a voltage proportional to the displacement. The mechanical displacement can be

2.35
either linear or rotational.
V
− +
V RW
− + e(t ) = V RW = Kθ
RT
Kθ
e(t ) = V
θ RT
RW RT e(t ) = K Pθ
− e(t ) + − e(t ) +
Linear displacement Angular displacement
potentiometer potentiometer

Figure 2.20 Potentiometer used as a position indicator

The mechanical displacement can be either linear or rotational as shown in Figure 2.20,
in either case the potentiometer is represented symbolically as shown in Figure 2.21

+ θ (t )
θ (t ) e (t )
V KP
− e(t )

Figure 2.21 Potentiometer symbol and its transfer function

Two potentiometers in parallel will allow the comparison of two remotely located shaft
positions.

θ1 (t ) θ 2 (t ) θ1 (t ) e(t )
+ KP
V −
− θ 2 (t )
e(t ) = K P [θ1 (t ) − θ 2 (t )]
e(t )
Figure 2.22 Two potentiometers used to sense the position of two shafts.
Schematic diagram below shows a typical dc motor position control system, using dc
control.

2.36
 Amplifier e (t ) M J
a
B Load Servo motor &
DC Motor Load Transfer
e(t ) function
Pot
+ θ r (t ) e(t ) e (t ) θ L (t )
KP GA a Gm
V
Reference θ (t ) − θ L (t ) − Amplifier
r
input

Figure 2.23 A dc motor position control system with potentiometer as sensor

If there is a discrepancy between the load position and reference input, an error signal is
generated by the potentiometer. The error signal is amplified and drives the motor.

θr

0
t

Figure 2.23 Error, amplified error and the motor step response

2.37
Tachometers
Tachometers convert mechanical energy into electrical energy. DC Tachometers are used
in control system as velocity indicators or for speed control. Figure below shows the
block diagram of a typical velocity control system. The output is a voltage proportional
to speed.

ω Load
r (t ) e(t ) Power
Controller Amplifier M
− DC
Motor

+ ω
vt (t ) T
− Tachometer

Figure 2.24 Velocity control with tachometer

The transfer function of a tachometer is
dθ (t )
e(t ) = KT = KT ω (t ) ⇒ E ( s ) = KT sθ ( s ) = KT Ω( s )
dt
The output voltage of a tachometer is proportional to the rotor speed. The difference
between the two signals, or the error, is amplified and used to adjust the motor speed, so
that the speed will eventually reach the desired value. If the response is not satisfactory a
controller is designed to achieve the design specifications.

Incremental Encoder

A digital encoder is a device that converts motion into a sequence of digital pulses.
Encoders have both linear and rotary configurations, but the most common type is the
rotary incremental encoder. A typical rotary incremental encoder has a light source, a
2.38
rotary disc, and a sensor. An encoder with three tracks is shown in Figure 2.25. An
incremental encoder produces equally spaced pulses from one or more concentric tracks
on the code disk. The pulses are produced when a beam of light passes through
accurately placed holes in the code disk. Each track has its own light beam. The inside
track has only one hole, which is used to locate the home position on the code disk. The
holes in the middle track are offset from the holes in the outside track by one-half the
width of the hole. The purpose of the offset is to provide directional information.

Code disk

Three unit light

sensor assembly
(photovolatic,
phototransistor,
photodiode)
Three unit Three light
lights source beams
assembly

Square-wave signal compatible with digital logic are derived from the sensors.

1
Outer track
0

Middle track

Inner track

Time
The incremental encoder has three tracks. The inner track provides a reference signal to
locate the home position. The middle track provides information about the direction of
rotation. In one direction, the middle track lags the outside track; in the other direction,
the middle track leads the outside track.
2.39
 Chapter 3
State Variable Modeling

The differential equations of a lumped linear network can be written in the form

x (t ) = Ax (t ) + B u (t ) (3.1)
y (t ) = Cx (t ) + D u (t )
This system of first-order differential equations is known as the state equation of the
system and x (t ) is the state vector and u (t ) is the input vector. The second equation is
referred to as the output equation. A is called the state matrix, B the input matrix, C
the output matrix, and D the direct transition matrix. One advantage of the state-space
method is that the form lends itself easily to the digital and/or analog computer methods
of solution. Further, the state-space method can be easily extended to analysis of
nonlinear systems. State equations may be obtained from an nth-order differential
equation or directly from the system model by identifying appropriate state variables.
To illustrate how we select a set of state variables, consider an nth-order linear plant
model described by the differential equation
dny d n −1 y dy
n
+ an −1 n −1
+ " + a1 + a0 y = u (t ) (3.2)
dt dt dt

where y (t ) is the plant output and u (t ) is the plant input. A state model for this system is
not unique but depends on the choice of a set of state variables. A useful set of state
variables, referred to as phase variables, is defined as

x1 = y, x2 = y , x3 =  y, " , xn = y n −1
Taking the derivatives, we have
x1 = x2 , x2 = x3 , x3 = x4 , " , and xn is given by (3.2) (3.3)
xn = −a0 x1 − a1 x2 − " − an −1 xn + u (t )
or in matrix form

 x1   0 1 0 " 0   x1  0 
 x   0 0 1 " 0   x2  0 
 2    
 # = # # # # #  +  #  +  #  u (t ) (3.4)
       
 xn −1   0 0 0 # 1   xn−1  0 
 xn   − a0 −a1 −a2 − an −1   xn  1 

and the output equation is

y = [1 0 0 " 0] x (3.5)

3.1
Example 3.1

Obtain the state equation in phase variable form for the following differential equation.
d3y d2y dy
2 3 + 4 2 + 6 + 8 y = 10u (t )
dt dt dt
The differential equation is third order, thus there are three state variables as follows
x1 = y, x2 = y , x3 = y
and the derivatives are
x1 = x2 , x2 = x3 , and x3 = −4 x1 − 3x2 − 2 x3 + 5u (t )
Or in matrix form

 x1   0 1 0   x1  0 
      
 x2  =  0 0 1   x2  + 0  5u (t )
 x3   −4 −3 −2   x3  1 
 x1 
y = [1 0 0]  x2 
 x3 

The M-file ode2phv.m is developed which converts an nth-order ordinary differential

equation to the state-space phase variable form. [A, B, C] = ode2phv(ai, k) returns the
matrices A, B, C, where ai is a row vector containing coefficients of the equation in
descending order and k is the coefficient of the right-hand side.

ai = [2 4 6 8];
k = 10;
[A, B, C] = ode2phv(ai, k)

produces the following phase variable state representation

A= B = C =
0 1 0 0 1 0 0
0 0 1 0
-4 -3 -2 5

3.1 Equations of Electrical Networks

The state variables are directly related to the energy-storage elements of a system. It
would seem, therefore, that the number of independent initial conditions is equal to the
number of energy-storing elements. This is true provided that there is no loop containing
only capacitors and voltage sources and there is no cut set containing only inductive and
current sources. In general, if there are nc loops of all capacitors and voltage sources, and
nL cut sets of all inductors and current sources, the number of state variables is
n = eL + eC − nC − nL (3.6)

3.2
where
eL = number of inductors
eC = number of capacitors
nC = number of all capacitive and voltage source loops
nL = number of all inductive and current source cut sets

Example 3.2

Write the state equation for the network shown in Figure 3.1.

vc 0.5 H
+
iL
is (t ) 0.5 Ω 1F 2.5 Ω

Figure 3.1 Circuit of Example 3.2

Define the state variables as current through the inductor and the voltage across the
capacitor. Write a node equation containing capacitor, and a loop equation containing an
inductor. The state variables are vc , and iL .

The node-voltage equation is

dv dvc
1 c + 2vc + iL − is (t ) = 0 ⇒ = −2vc − iL + is (t )
dt dt
and the loop equation is
di di
0.5 L + 2.5iL − vc = 0 ⇒ L = 2vc − 5iL
dt dt
or
 vc   −2 −1  vc  1 
 i  =  2 −5  i  + 0  is
 L   L  

Example 3.3

Write the state equation for the network shown in Figure 3.2.

3.3
 4Ω 1H vc 2H
+
i1 i2
+
e(t ) 0.5 F 3Ω
−
−

Figure 3.2 Circuit of Example 3.3

Define the state variables as current through the inductors and the voltage across the
capacitors. Write a node equation containing the capacitor, and two loop equations
containing the inductor. The state variables are i1 , i2 , and vc .

The loop equations are

di di1
4i1 + 1 1 + vc − e(t ) = 0 ⇒ = −4i1 − vc + e(t )
dt dt
di di2
2 2 + 3i2 − vc = 0 ⇒ = −1.5i2 + 0.5vc
dt dt
Node equations is
dv dvc
0.5 c + i2 − i1 = 0 ⇒ = 2i1 − 2i2
dt dt
or
 i1   −4 0 −1   i1  1 
      
 i2  =  0 −1.5 0.5  i2  +  0  e(t )
 vc   2 −2 0   vc   0 

Example 3.4

Write the state equation for the network shown in Figure 3.3.
4 Ω vc1 2H vc 2
iL

+ 0.25 F
vi 0.5 F 1Ω is
-

0
Figure 3.3 Circuit of Example 3.4

Define the state variables as current through the inductor and voltage across the
capacitors. Write two node equations containing capacitors and a loop equation
containing the inductor. The state variables are vc1 , vc2 , and iL .

3.4
Node equations are
dv v −v
0.25 c1 + iL + c1 i = 0 ⇒ vc = −vc1 − 4iL + vi
dt 4
dv v
0.5 c 2 − iL + c 2 − is = 0 ⇒ vc 2 = −2vc1 + 2vc 2 + 2is
dt 1
and the loop equation is
di
2 L + vc 2 − vc1 = 0 ⇒ iL = 0.5vc1 − 0.5vc 2
dt
or
 vc1   −1 0 −4   vc1  1 0
 v  =  0 v 
 c2   −2 2   vc 2  +  0 2   i 
i
 iL   0.5 −0.5 0   iL   0 0   s 

3.2 Simulation Diagram

Equation (3.3) indicates that state variables are determined by integrating the
corresponding state equation. A diagram known as the simulation diagram can be
constructed to model the given differential equations. The basic element of the simulation
diagram is the integrator. The first equation in (3.3) is

x1 = x2
Integrating, we have
x1 = ∫ x2 dx
The above integral is represented by the time-domain diagram shown in Figure 3.4 (a)
similar to the block diagram, or the time-domain diagram shown in Figure 3.4 (b) similar
to the signal flow graph.
x2 (t ) 1 x1 (t ) s −1
x2 (t ) x1 (t )
s
(b)
(a)
Figure 3.4 Simulation diagram for integrator

It is important to know that although the symbol 1/ s is used for integration, the
simulation diagram is a time domain representation. The number of integrators is equal to
the number of state variables. For example, for the state equation in Example 3.1 we have
three integrators in cascade, the three state variables are assigned to the output of each
integrator as shown in Figure 3.5. The last equation in (3.3) is represented via a summing
point and feedback paths. Completing the output equation, the simulation diagram known
as the phase-variable control canonical form is obtained.

3.5
 5u (t ) x3 1 x3 1 x2 1 x1 y 1 s −1 s −1 s −1 1
1 5u (t ) x x1 y
- s s s x3 -2 3 x2
- -
2 OR -3
3 -4
4
(a) (b)
Figure 3.5 Simulation diagram for Example 3.1

3.3 Transfer Function to State-Space Conversion

Direct Decomposition

Consider the transfer function of a third-order system

Y ( s) b2 s 2 + b1s + b0
= (3.7)
U ( s ) s 3 + a2 s 2 + a1s + a0
where the numerator degree is lower than that of the denominator. The above transfer
function is decomposed into two blocks as shown in Figure 3.6.

U (s) 1 W (s) Y (s)

3 2 b2 s 2 + b1s + b0
s + a2 s + a1s + a0

Figure 3.6 Transfer function in (3.7) arranged in cascade form

Denoting the output of the first block as W ( s ) , we have

U ( s)
W (s) = 3 2
and Y ( s ) = b2 s 2W ( s ) + b1sW ( s ) + b0W ( s )
s + a2 s + a1s + a0
or
s 3W ( s ) = − a2 s 2W ( s ) − a1sW ( s ) − a0W ( s ) + U ( s )
This results in the following time-domain equation
w = −a2 w
  − a1w − a0 w + u (t ) and y (t ) = b2 w  + b1w + b0 w (3.8)
From the above expression we see that w has to go through three integrators to get w as
shown in Figure 3.7. Completing the above equations results in the phase-variable control
canonical simulation diagram.
b2
b1
+ +
u (t ) w 1 w  1 w 1 w + y
x b0
- s x3 s x2 s 1
- -
a2
a1
a0

Figure 3.7 Phase variable control canonical simulation diagram

3.6
The above simulation in block diagram form is suitable for SIMULINK diagram
construction. You may find it easier to construct the simulation diagram similar to the
signal flow graph as shown in Figure 3.8.
b2
b1
1 w s −1 w
 s −1 w s −1 w b0 1 y
u (t )
− a2 x3 x2 x1
− a1
− a0

Figure 3.8 Phase variable control canonical simulation diagram

In order to write the state equation, the state variables x1 (t ) , x2 (t ) , and x3 (t ) are assigned
to the output of each integrator from the right to the left. Next an equation is written for
the input of each integrator. The results are

x1 = x2
x2 = x3
x3 = − a0 x1 − a1 x2 − a0 x1 + u (t )
and the output equation is
y = b0 x1 + b1 x2 + b3 x3
or in matrix form

 x1   0 1 0   x1  0 
 x  =  0 0 1   x2  + 0  u (t ) (3.9)
 2 
 x3   − a0 − a1 − a2   x3  1 
 x1 
y = [1 0 0]  x2 
 x3 

It is important to note that the Mason’s gain formula can be applied to the simulation
diagram in Figure 3.6 to obtain the original transfer function. Indeed ∆ of Mason’s gain
formula is the characteristic equation. Also, the determinant of sI − A matrix in (3.9),
results in the characteristics equation. Keep in mind that there is not a unique state space
representation for a given transfer function.

The Control System Toolbox contains a set of functions for model conversion.
[A, B, C, D] = tf2ss(num, den) converts the system in transfer function from to state-
space phase variable control canonical form.

3.7
Example 3.5

For the following transfer function

Y ( s) s2 + 7s + 2
G ( s) = = 3
U ( s ) s + 9s 2 + 26s + 24
(a) Draw the simulation diagram and find the state-space representation of the above
transfer function.
(b) Use MATLAB Control System Toolbox [A, B, C, D] = tf2ss(num, den) to find the
state model.

(a) Draw the transfer function block diagram in cascade form

U (s) 1 W (s) 2 Y ( s)
s + 7s + 2
s + 9 s + 26s + 24
3 2

From this we have

s 3W ( s ) = −9 s 2W ( s ) − 26 sW ( s ) − 24W ( s ) + U ( s ) & Y ( s ) = s 2W ( s ) + 7 sW ( s ) + 2W ( s )
or in time-domain

w = −9w
  − 26w − 24w + u & y = w  + 7 w + 2w
The above time-domain equations yield the following simulation diagram
1
7
1 w s −1 −1 −1

w s w s w 2 1 y
u (t ) x3
−9 x2 x1
−26
−24

To obtain the state equation, the state variables x1 (t ) , x2 (t ) , and x3 (t ) are assigned to the
output of each integrator from the right to the left. Next an equation is written for the
input of each integrator. The results are

x1 = x2
x2 = x3
x3 = −24 x1 − 26 x2 − 9 x1 + u (t )
and the output equation is
y = 2 x1 + 7 x2 + x3
or in matrix form

3.8
  x1   0 1 0   x1  0 
 x  =  0 0 1   x2  + 0  u (t )
 2 
 x3   −24 −26 −9   x3  1 
 x1 
y = [ 2 7 1]  x2 
 x3 

(b) We write the following statements

num = [1 7 2]; den = [1 9 26 24];

[A, B, C, D] = tf2ss(num, den)

The result is

A= B= C= D=
-9 -26 -24 1 1 7 2 0
1 0 0 0
0 1 0 0

Note that MATLAB assigns x1 to the output of the first integrator, and x2 , and x3 to the
output of the second and third integrators.

Functions for Model creation

MALAB Control System Toolbox contains many functions for model creation and
inversion, data extraction, and system interconnections. A few of these functions for
continuous-time control systems are listed below. For a complete list of all functions type
help/control/control at MATLAB prompt.
tf Create transfer function models.
zpk Create zero/pole/gain models.
ss Create state-space models.
tfdata Extract numerator(s) and denominator(s).
zpkdata Extract zero/pole/gain data.
ssdata Extract state-space matrices.
append Group LTI systems by appending inputs and outputs.
parallel Generalized parallel connection (see also overloaded +).
series Generalized series connection (see also overloaded *).
feedback Feedback connection of two systems.
connect Derive state-space model from block diagram description.
blkbuild Builds a model from a block diagram.

The Control System Toolbox supports four commonly used representations of linear
time-invariant (LTI) systems: tf, zpk, and ss objects. To create an LTI model or object,
use the corresponding constructor tf, zpk, or ss. For example,
sys = tf(1,[1 0]) .

3.9
creates the transfer function H(s) = 1/s. The result sys is a tf object containing the
numerator and denominator data. You can then manipulate the entire model as the single
MATLAB variable sys. For more details and examples on how to specify the various
types of LTI models, type ltimodels followed by tf, zpk, or ss.

The functions tfdata, zpkdata, and ssdata are provided for extracting the parameters of
the tf, zpk and ss objects. For example the command [num, den] = tfdata(T, 'v') returns
the numerator and denominator of the tf object. The argument ‘v’ returns the numerator
and denominator as row vectors rather than cell arrays. The Control System Toolbox
contains seven more functions, which are useful for creating a single model out of its
components.

Example 3.6

Use feedback function to obtain the closed-loop transfer function and the tf2ss function
to obtain the closed-loop state-space model

R( s) 5( s + 1.4) 1 C (s)
s+7 s ( s + 1)( s + 4)
−
Gc ( s ) G p ( s)
H (s)

10

The following commands

Gc = tf(5*[1 1.4], [1 7]) % transfer function Gc

Gp = tf([1], [1 5 4 0]); % transfer function Gp
H = 10;
G = series(Gc, Gp) % connects Gc & Gp in cascade
T = feedback(G, H) % obtains the closed loop transfer function
[num, den] = tfdata(T, 'v'); % returns num & den as row arrays
[A, B, C, D]=tf2ss(num, den) % returns the A,B, C, D matrices of the state model
result in

Transfer function:
5s+7
---------------------------------
s^4 + 12 s^3 + 39 s^2 + 78 s + 70

A= B= C= D=
-12 -39 -78 -70 1 [0 0 5 7] 0
1 0 0 0 0
0 1 0 0 0
0 0 1 0 0

3.10
3.4 Solution of State Equation

Before presenting a formal solution for the state-space equation, let us review the solution
of the linear first-order differential equation due to initial condition. Consider

x(t ) = ax(t ) with initial value x(0) (3.10)

In solving this equation, we can assume a solution of x(t ) = Ke mt . For this to be the
solution it must satisfy the given equation, i.e.,
mKe mt = aKe mt or (m − a ) Ke mt = 0 ⇒ m=a
Thus
x(t ) = Ke at
Since at t = 0, K = x(0) , the solution is
x(t ) = e at x(0) (3.11)
We can also solve (3.10) using Laplace transform
sX ( s ) − x(0) = aX ( s )
or
1
X ( s) = x(0)
s−a
The inverse Laplace transform results in the same solution

x(t ) = e at x(0)

Now we consider the solution of the state equation

x (t ) = Ax (t ) + B u (t ) with initial value x (0)

As with a single first-order differential equation, the state equation has two solutions:
1. Solution due to initial condition (Homogenous solution) or zero-input response
2. Solution due to input in the absence of initial condition, known as the zero-state
response.

First we consider the zero-input solution

Homogeneous State Equation

x (t ) = Ax (t ) given x (0) (3.12)
Taking Laplace transform

s X ( s ) − x (0) = AX ( s )
or
[s I − A ] X ( s ) = x (0)
Thus the s-domain solution is

3.11
 X ( s ) = [ sI − A]−1 x (0) or X ( s ) = Φ ( s ) x (0) (3.13)
where
Φ ( s ) = [sI − A]−1 (3.14)
Taking the inverse Laplace transform, the time-domain solution is

x (t ) = φ (t ) x (0) (3.15)

where φ (t ) = L-1  ( sI − A) −1  is called the state transition matrix

Example 3.7

Consider the state equation for the circuit in Example 3.2.

vc   −2 −1  vc 
i  =  2 −5  i  , with initial conditions vc (0) = 1 , and iL (0) = −1
 L   L

Determine the zero-input response that is, the response due to initial condition in the
absence of the current source.

The zero-input response is given by

X ( s ) = [ sI − A]−1 x (0)
 s + 5 −1   s + 5 −1 
s + 2 1   2 s + 2   2 s + 2 
[ SI − A] =   ⇒ Φ ( s ) = [ SI − A]−1 =  = 
 −2 s + 5 ( s + 2)( s + 5) + 2 s 2 + 7 s + 12

The s-domain solution is

 s + 5 −1   1   s + 6
 2   
s + 2   −1  −s 
−1
X ( s ) = [ sI − A] x(0) =  =  
s 2 + 7 s + 12 ( s + 3)( s + 4)
or
 s+6   3 −2 
 ( s + 3)( s + 4)   + 
X ( s) =   = s + 3 s + 4
 −s   3 −4 
 ( s + 3)( s + 4)   s + 3 + s + 4 
 
Thus the zero-input response is

vc  3e −3t -2e −4t 

x(t ) =   =  −3t −4t 
 iL  3e -4e 

3.12
Solution of Non-homogenous State Equation

The solution of the linear non-homogenous state equation

x (t ) = Ax (t ) + B u (t ) with initial value x (0)

can be obtained by the Laplace transform approach.

s X ( s ) − x (0) = AX ( s ) + B U ( s )
[s I − A ] X ( s ) = x (0) + B U ( s )

Solving for X ( s ) , we get

X ( s ) = [ sI − A]−1 x (0) + [ sI − A]−1 B U ( s )

X ( s ) = [ sI − A] x(0) + [ sI − A]−1 BU ( s )
−1
(3.16)

 

 
zero-input response zero-state response

or
X ( s ) = Φ ( s ) x(0) + Φ ( s ) BU ( s )
 

 

zero-input response zero-state response

where
Φ ( s ) = [sI − A]−1
Taking the inverse Laplace transform in (3.16) the zero-input response and the zero-state
response are obtained

We can also express the above equation in terms φ (t ) and the convolution integral.

t
x(t ) = φ (t ) x(0) + ∫ φ (τ ) Bu (t − τ )dτ (3.17)
0
where

φ (t ) = L-1  ( sI − A) −1  is the state transition matrix

You are required to know and be able to find the zero-input response and the zero-state
response using Laplace transform method.

3.13
Example 3.8

Determine the state vector x(t ) for t ≥ 0 , for the system given below.

 x1   0 1   x1   0 
x  =  −2 −3  x  + 1  r (t )
 2   2  

r (t ) = 1u (t ) is a unit step input, x1 (0) = 1, x1 (0) = −1,

 s + 3 1  s + 3 1
 s −1   −2 s   −2 s 
[ SI − A] =  −1 
⇒ Φ ( s ) = [ SI − A] = 2  =  

 2 s + 3 s + 3s + 2 ( s + 1)( s + 2)

The zero-input response or the natural response is

X ( s ) = [ sI − A]−1 x (0)

 s + 3 1  1   s+2   1 
 −2 s   −1  −( s + 2 
X ( s ) = [ sI − A]−1 x(0) =    =   =  s +1 
 
( s + 1)( s + 2) ( s + 1)( s + 2)  −1 
 s + 1 
 e−t 
xn (t ) =  −t 
 −e 

The zero-state response or force response is

X ( s ) = [ sI − A]−1 BU ( s )

 s + 3 1  0  1  1/ 2 −1 1/ 2 
−1
 −2 s  1 
    1
s
   s + s +1 + s + 2 
X ( s ) = [ sI − A] BR ( s ) = = = 
( s + 1)( s + 2) s s ( s + 1)( s + 2)  1 −1 
+
 s + 1 s + 2 
 1 − t 1 −2t 
−e + e 
xs (t ) =  2 2
 −t −2 t 
 e −e 

The complete solution is

1 1  1 1 
 e − t   − e −t + e −2t   + e −2t 
x(t ) = xn (t ) + xs (t ) =  − t  + 2 2 = 2 2
 −e   e − t − e −2t   −e −2t 
   

3.14
3.5 State-Space to Transfer Function Conversion

Consider the state and output equations

x (t ) = Ax (t ) + B u (t )
y = Cx (t ) + D u (t )

Taking the Laplace transform

sX ( s ) = AX ( s ) + BU ( s ) ⇒ [sI − A] X ( s ) = BU ( s )
Y ( s ) = CX ( s) + DU ( s )

Substituting for X ( s ) in the second equation above, we get

Y ( s ) = C[ SI − A]−1 BU ( s) + DU ( s )
or
Y ( s)
G ( s) = = C[ SI − A]−1 B + D (3.18)
U ( s)
Example 3.9

A system is described by the following state-space equations

 x1   0 1   x1  0 
x  =  −6 −5  x  + 1  u (t )
 2   2  

 x 
y = [8 1]  1 
x2 
Obtain the system transfer function using the formula in (3.18)

 s + 5 1
 s −1   6 s 
[ SI − A] =   ⇒ Φ ( s ) = [ SI − A]−1 =  2
6 s + 5 s + 5s + 6

 s + 5 1  0  1
 6    [8 1]  
G ( s ) = C[ SI − A]−1 B = [8 1]  2
s  1 
= 2 s = 8 + s
s + 5s + 6 s + 5s + 6 s 2 + 5s + 6
Therefore
s +8
G ( s) = 2
s + 5s + 6
In MATLAB [num, den] = ss2tf(A, B, C, D, i) converts the state equation to a transfer
function for the ith input.

3.15
Example 3.10
A system is described by the following state-space equations

 x1   0 1 0   x1  10 
 x  =  0 0 1   x  +  0  u (t )
 2   2  
 x3   −1 −2 −3  x3   0 
 x1 
y = [1 0 0]  x2 
 x3 
Y ( s)
Find the transfer function, G ( s ) = .
U ( s)

The following statements:

A = [0 1 0; 0 0 1; -1 -2 -3]; B = [10; 0; 0];

C = [1 0 0]; D = [0];
[num, den] = ss2tf(A, B, C, D, 1)
G = tf(num, den)
result in

num =
0.0000 10.0000 30.0000 20.0000
den =
1.0000 3.0000 2.0000 1.0000

Transfer function:
10 s^2 + 30 s + 20
---------------------
s^3 + 3 s^2 + 2 s + 1

Also, [z, p] = ss2tf(A, B, C, D, 1) converts the state equation to transfer function in

factored form.

3.16
 Chapter 4

Time-domain Response

Assessing the time-domain performance of closed-loop system models is important because

control systems are inherently time-domain systems. The performance of dynamic systems
in the time domain can be defined in terms of the time response to standard test inputs. One
very common input to control systems is the step function. If the response to a step input is
known, it is mathematically possible to compute the response to any input. Another input of
major importance is the sinusoidal function. A sinusoidal steady-state output is obtained
when an asymptotically stable linear system is subjected to a sinusoidal input. Thus, if we
know the response of a linear time-invariant system to sinusoids of all frequencies, we have
a complete description of the system.
The following standard inputs are often used in checking the response of a system.

• Unit step (position) function r (t ) = u (t )

• Unit ramp (velocity) function r (t ) = tu (t )
1
• Unit parabolic (acceleration) function r (t ) = t 2u (t )
2
• Sinusoidal function r (t ) = A cos ω t

The time response of a control system is usually divided into two parts: the transient
response and the steady-state response,

c(t ) = ctr (t ) + css (t )

The steady-state response css (t ) is that part of the response, which remains after the
transient, has died out.

Since inertia, mass, inductance and capacitance are part of a physical system and we know
for some variable instantaneous changes cannot occur and we may go through a transient
period. During this period we have deviation between input and output and therefore
dynamic behavior of a system is very important. The steady-state response of a control
system is also very important. If the output does not agree with the steady-state of the input
exactly, the system is said to have a steady-state error. This topic will be discussed in the
next chapter.

4.1
4.1 Time Response of First-order Systems

Consider the first-order differential equation

dc(t )
+ a0 c(t ) = b0 r (t ) , with initial condition c(0) (4.1)
dt
Taking the Laplace transform, we have

sC ( s ) − c(0) + a0C ( s) = b0 R( s )
or
c(0) b
C (s) = + 0 R(s) (4.2)
s + a0 s + a0
This equation can be represented in block diagram as shown

R( s) b0
s + a0 + R( s) b0 C (s)
C (s)
s + a0
c(0) 1 +
s + a0
(a ) (b)
Figure 4.1 Block diagram representation of first-order system

The block diagram shown in Fig. 4.1 (b) is for the zero initial condition. It is only sufficient
to obtain the response due to the force function. For example, if the step response is
obtained, from property of linear system the response due to initial condition can be
obtained from the step response as follow

u (t ) ru (t ) Step
d d response
u (t ) ru (t )
dt dt d
c(0) ru (t )
⇓ ⇓ Impulse L [ C (0)] = c (0)δ ( t )
−1
dt
δ (t ) rδ (t ) response
Response due to initial condition
from step response

Figure 4.2 Finding response due to initial condition from the step response.

4.2
Unit Step response
b
C ( s) = 0 R(s)
s + a0
1
If the input is a unit step function r (t ) = u (t ) , i.e., R ( s ) = , the response is
s
b0 b0
b0 1 a0 a
C ( s) = = − 0
( s + a0 ) s s s + a0
Taking the inverse Laplace transform
b
(
c(t ) = 0 1 − e − a0t
a0
)
or
t
−
c(t ) = K (1 − e ) τ
(4.3)
b0
where τ is the time constant and K = . The response is shown in Fig. 4.3
a0

Fig. 4.3 Step response of first-order system

The response will settle to its final value in approximately four to five time constants. In
this course we take the settling time to be approximately four time constants, i.e., τ = 4ts .
The general first-order transfer function can be written in terms of its time constant as
follow
K
G(s) = & C ( s) = G ( s) R( s) (4.4)
τ s +1
The steady-state response to a unit-step input is K . This can also be obtained by applying
the final value theorem. lim f (t ) = lim sF ( s ) .
t →∞ s →0

css = c(∞) = lim sG ( s ) R( s ) (4.5)

s →0
4.3
 1
For unit step response R ( s) = , therefore the step response steady-state value is
s
1
css = lim sG ( s ) = G (0) = K (4.6)
s →0 s
This shows that the step response steady-state value is given by the transfer function gain at
s = 0 i.e., G (0) . This is called the system DC gain.

Ramp Response (First order system)

1
The Laplace transform of a ramp function input is L[tu (t )] =
s2
K
C ( s) = R( s) (4.7)
τ s +1
K /τ K − Kτ Kτ
C ( s) = 2 = 2+ +
s ( s + 1/ τ ) s s (s+1/τ )
Taking the inverse Laplace transform
c(t ) = Kt − Kτ + Kτ e− t /τ (4.8)
The response has the same time constant; this is true, regardless of the input. However the
amplitude of the exponential term is different by a factor of τ. If τ is much greater than 1,
the exponential term will have a major effect on the system transient response. This can be
a problem in designing system to follow ramp inputs. The steady-state response is given
by
css = Kt − Kτ
The ramp response for a system with τ = 0.5 sec and K = 1 is shown in Figure 4.

Fig. 4.4 Step response of first-order system

4.4
The steady-state error, i.e. the difference between input and the output is Kτ. We will
discuss the steady-state error in detail in the next chapter.

4.2 Time response of second-order systems

The standard form of the second-order transfer function is

Given by
C (s) ω n2
= G ( s) = 2 (4.9)
R( s) s + 2ζω n s + ω n2
The characteristic equation is given by
s 2 + 2ζω n s + ω n2 = 0 (4.10)
For ζ > 1 , the poles of the transfer function are
s1 = −ζω n + ω n ζ 2 − 1, s2 = −ζω n − ω n ζ 2 − 1 (4.11)

1
For the unit step response R ( s) = , and the s-domain step response can be written as
s
ω n2 1 1 K1 K2
C ( s) = = + +
( s − s1 )( s − s2 ) s s ( s − s1 ) ( s − s2 )
Thus the inverse Laplace transform is
c(t ) = 1 + K1e s1t + K 2 e s2t = 1 + K1e − t /τ1 + K 2 e− t /τ 2 (4.11)
1
That is, for ζ > 1 the response is overdamped with two time constants τ 1 = − and
s1
1
τ2 = − .
s2
For ζ = 1 , the poles of G ( s) are real and equal, s1 = s2 = −ζω n , and the response is
critically damped
c(t ) = 1 + e −ζωnt ( K1 + K 2t ) = 1 + K1e − t /τ + K 2te− t /τ (4.12)
1
The time constant for critically damped response is τ 1 =
ζω n
The transient response of a practical control system often exhibits damped oscillations
before reaching steady-state. For ζ < 1 , the roots of the characteristic equation are complex

s1,2 = −ζω n ± jω n 1 − ζ 2 = −ζω n ± jω d (4.13)

where ω d = ω n 1 − ζ 2 is the damped frequency of oscillation, and ω n is the natural
frequency. The natural frequency is the frequency of oscillation if all of the damping is
4.5
removed. Its value gives us an indication of the speed of the response. ζ is the
dimensionless damping ratio. The damping ratio gives us an idea about the nature of the
transient response. It gives us a feel for the amount of overshoot and oscillation that the
response undergoes.

The underdamped response ( ζ < 1 ) to a unit step input, subject to zero initial condition is
given by
ω n2 1 1 K K*
C ( s) = = + +
( s + ζω n − jω d )( s + ζω n + jω d ) s s ( s + ζω n − jω d ) ( s + ζω n + jω d )

1 1 1−ζ 2
K is found to be K = − ∠θ − 90D where θ = tan −1 , therefore
2 1−ζ 2 ζ

1
c(t ) = 1 − e−ζωnt (sin ω d t + θ ) (4.14)
1−ζ 2

The response is underdamped as shown in Figure 4.5.

Figure 4.5 Step response of the second order system for ζ < 1 .

Roots of the characteristic equations given in (4.11) are marked in the s-plane as shown in
Figure 4.6. s1,2 = −ζω n ± jω n 1 − ζ 2

4.2
 s1
jω n 1 − ζ 2

ωn

θ
−ζω n 0

s2
Figure 4.6 Pole locations s-plane

The vector length | s1 |= [(ζω n ) 2 + ω n2 (1 − ζ 2 )]1/ 2 ⇒ |s1 |= ω n

The time-constant of the exponentially damped sinusoid is given by the reciprocal of the
horizontal component, i.e.,
1
τ= (4.15)
ζω n
ζω
Also cosθ = n
ωn
That means the damping ratio is given by
ζ = cosθ (4.16)
The examination of the location of poles of the transfer function reveals the nature of the
response. For ζ < 1 , as poles move horizontally to the left the response time constant
1
becomes smaller ( τ = ), i.e., a smaller settling time ( ts  4τ ). If the poles vertical
ζω n
component increases, that is larger θ , it means smaller damping ratio (ζ = cos θ ) , which
results in larger overshoot. Ifθ is equal to zero (ζ = 1) , poles will be equal, resulting in the
critically damped response. For ζ > 1 , we have two distinct real roots and the response will
become overdamped. For ζ = 0 , that is if poles are on the jω -axis, we have an undamped
oscillatory response as is evident from (4.14). It is s very important to note that if ζ < 0 , the
poles real part becomes positive, i.e., the poles lies in the right half s-plane and the response
amplitude increase without limit. The response is said to be unbounded and the system is
unstable.

4.3
Locus of roots when ω n is held constant while the damping ratio is varied from −∞ to
+∞ is shown in Fig. 4.7.
jω n
ζ =0

0 > ζ > −1
0 <ζ <1
Unstable Region

∞ ←ζ ζ >1 ζ < −1 ζ → −∞
ζ >1 ζ = −∞ σ
ζ =1 ζ =∞ ζ = −1

ζ =0
Figure 4.7 Locus of roots of the characteristic equation with constant ω n , varying ζ

The step response comparison for various root locations in the s-plane is shown in
Figure 4.8.

4.4
Figure 4.8 Step response comparison for various root locations in the s-domain
jω

0 σ
Overdamped
s-plane
jω
jω

0 σ

s-plane Critically damped

jω

0 σ
Underdamped
s-plane
Undamped sinusoid
jω

0 σ

s-plane
jω

0 σ Unstable

s-plane
jω

0 σ
4.5
Unstable
s-plane
4.3 Step response using MATLAB Control System toolbox

Given a transfer function of a closed-loop control system, the function step(num, den)
produces the step response plot with the time vector automatically determined. num, and
den are the coefficients of the numerator and denominator of the closed loop transfer
function in descending order of power. If the above commands are invoked with the left-
hand argument c = step(num, den), the response c is returned. If the time array t is
specified, step(num, den, t) produces the step response plot over the specified time
interval. With the left hand argument c = step(num, den, t) returns the response for c,
over the specified time interval, and we need to use plot function to obtain the plot.
Similarly the functions impulse, initial, and lsim can be used to obtain the impulse
response, response due to initial condition and response to a defied forcing function.

Example 4.1
C (s) ω n2
Obtain the unit step response for the second-order system = G ( s) = 2
R( s) s + 2ζω n s + ω n2
For the following damping ratios ζ : 0.1 0.2 0.4 0.7 1 4 , and ω n = 5 .
We use the following commands

wn=5;
num = wn^2;
zeta =[0.1 0.2 0.4 .7 1 4];
for k=1:6
den = [1 2*zeta(k)*wn wn^2];
step(num, den);
hold on
end
legend('\zeta=0.1', '\zeta=0.2', '\zeta=0.4', '\zeta=0.7', '\zeta=1', '\zeta=4')

The result is shown in Figure 4.8.

4.6
Figure 4.8 Response of a second-order system to a step input for different values of ζ .

4.4 Time -domain performance specification

The performance criteria that are used to characterize the transient response to a unit step
input include rise time, peak time, overshoot and settling time. The underdamped step
response for a second order system is shown in Figure 4.9. We define the rise time tr as the
time required for the response to rise from 10 % of the final value to 90 % of the final
value. The time to reach the peak value is t p . The swiftness of the response is measured by
tr and t p . The similarity with which the actual response matches the step input is measured
by the percent overshoot and settling time ts . For underdamped systems the percent
overshoot P.O. is defined as
Maximum value - Final value M − Css
P.O. = ×100 = pt ×100 (4.17)
Final value Css
The peak time is obtained by setting the derivative of c(t ) to zero.
π
tp = (4.18)
ωn 1 − ζ 2
4.7
 Figure 4.9 Second order system step response for ζ < 1 .

The peak value of the step response occurs at this time and evaluating the response at
t = t p yields
1−ζ 2
c(t p ) = M tp = 1 + e −ζπ / (4.19)

Therefore, the percent overshoot is

1−ζ 2
P.O. = e −ζπ / × 100 (4.20)

Settling Time: settling time t s is the time required for the step response to reach within 2%
of its final value. These parameters are easy to find graphically once a plot of the response
is obtained. Regardless of the percentage used, the settling time will be directly
proportional to the time constant τ, i.e.
1
ts ≅ 4τ = 4 (4.21)
ζω n

4.8
Example 4.2

The block diagram of a position control servomechanism is as shown in Figure 4.10.

KP Servomotor
R( s) 20 C (s)
0.2 9
s ( s + 4)
− −

0.16s
Sensor
KD
0.2
Figure 4.10

Determine the step response percentage overshoot, P.O. , peak time, t p , and settling time t s .

Applying Mason’s gain formula, the closed-loop transfer function is

36
C (s) s ( s + 4) 36
= G ( s) = =
R( s) 3.2 s 36 s ( s + 4) + 3.2 s + 36
1+ +
s ( s + 4) s ( s + 4)
or
36
G ( s) = 2
s + 7.2s + 36
Comparing to the standard second order transfer function
ω n2
G ( s) = 2
s + 2ζω n s + ω n2
ω n2 = 36 ⇒ ω n = 6 rad/s , and
7.2
2ζω n = 7.2 ⇒ ζ = = 0.6
(2)(6)
The peak time is
π π
tp = = = 0.6545 sec
ωn 1 − ζ 2
6 1 − (0.6) 2
P.O. = e −ζπ / 1 − ζ 2 × 100 = e − (0.6)(π ) / 1 − (0.6) 2 × 100 = 9.478%

1 1
τ= = = 0.2777 sec
ζω n (0.6)(6)
ts ≅ 4τ = 4(0.2777) = 1.11 sec
For higher order system the time-domain specifications are obtained from step response
4.9
using MATLAB. A function named timespec is developed in “Computational Aids in
Control Systems Using MATLAB” for finding the time-domain specifications. However,
the new MATLAB Control System Toolbox has powerful ltiview program obtains the
system responses and the time-domain or frequency domain specifications. This is
demonstrated in the following Example 4.3. See Tutorial II MATLAB Functions for
Modeling and Time-domain analysis

Example 4.3

Use the ltiview to obtain the step response time-domain specifications for the system in
Example 4.2. We use the following statements

num = 36; den = [1 7.2 36];

G = tf(num, den) % transfer function G
ltiview('step', G)
The result is shown in Figure 4.11.

Figure 4.11 Sep response for Example 4.3.

The mouse right-click is used to obtain the time-domain specifications. From File menu
you can select Print to Figure option to obtain a Figure Window for the LTI Viewer for
editing the graph.

4.10
Example 4.4

The block diagram of a servomechanism is shown in Figure 4.12. Determine values of d

and e so that the maximum overshoot in unit-step response is 40% and the peak time is 0.8
sec.
R( s) d C (s)
s ( s + 1)
−

1 + es

Figure 4.12
P.O. = e −ζπ / 1 − ζ 2 × 100
e −ζ (π ) / 1 − ζ 2 = 0.4
−ζ (π )
= ln 0.4 = −0.91629
1−ζ 2
ζ 2π 2 0.8396
= 0.8396 ⇒ ζ2 = = 0.0784 ⇒ ζ = 0.28
1−ζ 2 10.709
π π
tp = = 0.8 ⇒ ωn = = 4.0906 Rad/s
ωn 1 − ζ 2 0.8 1 − 0.282

From the block diagram we have

C (s) d
= 2
R( s ) s + (de + 1) s + d
The characteristic equation is
s 2 + (de + 1) s + d = s 2 + 2ζω n s + ω n2
Equating the coefficients
d = ω n2 = (4.0906) 2 = 16.733
de + 1 = 2(0.28)(4.0906)
e = 0.077
The closed-loop transfer function is

C ( s) 16.733
= 2
R( s ) s + 1.288s + 16.733

We use the following statements to obtain the step response and time-domain
specifications.

4.11
 num = 16.733; den =[1 2.29 16.733];
G = tf(num, den) % transfer function G
ltiview('step', G)

The result is shown in Figure 4.13

Figure 4.13 Sep response for Example 4.4.

4.12
4.5 Frequency Response of systems

The frequency response of a system is defined as the steady- state response of the system
to a sinusoidal input signal. Consider a system with transfer function G ( s) and a
sinusoidal input

r (t ) = A cos ω t (4.22)
Substituting for the Laplace transform of r (t ) , the s-domain response C(s) of the system
output is

As As
C (s) = G(s) = G ( s)
s +ω
2 2
( s − jω )( s + jω )
(4.23)
The partial fraction expansion will result in
K K*
C ( s) = + + ∑ of terms generated by the poles of G ( s )
s − jω s + jω
where
As 1 1
K= G ( s) = AG ( jω ) = A G ( jω ) ∠θ (ω ) (4.23)
( s + jω ) s = jω
2 2
The poles of G ( s ) are the natural frequencies (or natural modes). They govern the
waveform of the transient component of the response. For the linear lumped network, the
terms generated by the poles of G ( s ) will not contribute to the steady-state response c(t ) .
The inverse Laplace transform of the partial fraction for complex poles is given by
−1  K K*  −α t
L  +  = 2 K e cos( β t + θ )
 s + α − jβ s + α − jβ 

Therefore, the steady-state response is given by the inverse Laplace transform of the first
two terms of C ( s) .

c(t ) = A G ( jω ) cos[ω t + θ (ω )] (4.24)

From the above equation it can be seen that system output has the same frequency as the
input and can be obtained by multiplying the magnitude of the input by G ( jω ) and
shifting the phase angle of the input by the angle of the transfer functionθ (ω ) . The
magnitude G ( jω ) and its angle θ for all ω constitute the system frequency response, and
they provide a significant insight for the analysis and design of control systems. G ( jω ) is
4.13
known as the Frequency Response Transfer Function. There is a correlation between
frequency and transient responses for the case of second-order systems. In practice, the
frequency response characteristic is adjusted by using various design criteria, which will
normally result in an acceptable transient response.

First-Order System
Consider first the frequency response of a first-order system with the following transfer
function
1
G ( s) = (4.25)
τ s +1
The frequency response transfer function is

1 1
G ( jω ) = = ∠θ (ω ), where θ (ω ) = − tan −1 (ωτ ) (4.26)
1 + jωτ [1 + ω 2τ 2 ]
A plot of G ( jω and θ (ω ) is given in Figure (4.14).

Figure 4.14 Frequency response of a first-order system.

The system bandwidth is defined as that value of frequency at which the magnitude of the
1
frequency response is reduced to 2 of its low frequency value. This frequency is denoted
by ω B . For the first-order system the bandwidth is given by
1
ωB = (4.27)
τ
Equation (4.27) is true only for the first-order systems.

4.6 Frequency response of second-order systems

Now consider the standard second-order transfer function

4.14
 ω n2
G ( s) = (4.28)
s 2 + 2ζω n s + ω n2
The frequency response transfer function is

ω n2 1
G ( jω ) = =
−ω 2 + j 2ζω nω + ω n2 ω 2
ω
1− + j 2ζ
ωn2
ωn
The frequency response transfer function amplitude is

1
G ( jω ) = 1
(4.29)
 ω 2   2
ω  
2 2

1 − 2  +  2ζ  
 ω n   ω n  

A plot of frequency response for ζ = 0.4 and ω n = 5 is given in Figure (4.15).

Mpω 1.4

1.2
|G(jω)|
1

0.8
Bandwidth
0.707
0.6

0.4

0.2

0 ωB 7
0 1 2 3 4 ωr 5 6 8 9 ω 10

Figure 4.15 Frequency response of a second-order system

For a constant ζ , increasing ω n causes the bandwidth to increase by the same factor. This
corresponds to a decrease in the peak time t p and the rise time tr the transient response.
Therefore, to increase the speed of response of a system, it is necessary to increase the
system bandwidth. For a particular system, an approximate relationship is given by
ω Btr = constant . Where this constant has a value in the neighborhood of 2 for the second-
order system, i.e.,
4.15
 ω B tr ≈ 2 (4.30)

For ζ < 0.707 , the amplitude exhibits a peak value. The frequency at which the peak
occurs is obtained by setting the derivative of G ( jω ) to zero. This frequency is known as
the resonance frequency. For ζ< 0.707, the resonance frequency ω r is given by
ω r = ω n 1 − 2ζ 2 (4.31)
The maximum value of the magnitude of the step response, denoted by M pω , is

1
M pω = (4.32)
2ζ 1 − ζ 2
The peak in the frequency response is directly related to the amount of overshoot in the
transient response. The greater the peaks, the more overshoot will occur. Thus, in control
system design M pω is usually restricted to a maximum allowable value.

Example 4.5

A system is described by the closed loop transfer function

25
G ( s) =
s + 2s + 25
2

(a) Find the resonant frequency ω r , peak amplitude M pω and the bandwidth ω B of the
system.

(b) Use MATLAB to obtain the frequency response and find ω r , M pω and ω B .
ω n2 = 25 ⇒ ω n = 5 Rad/s
2
2ζω n = 2 ⇒ ζ = = 0.2
(2)(5)
ω r = ω n 1 − 2ζ 2 = 5 1 − 2(0.2) 2 = 4.8 Rad/s

1 1
M pω = = = 2.55 Rad/s ⇒ dB (ω r ) = 8.136
2ζ 1 − ζ 2
2(0.2) 1 − (0.2) 2

The bandwidth frequency is found from

4.16
 2
1  1 
2 2
= 
 ωB2   ωB   2
1 − ω 2  +  2ζ ω 
 n   n 

This will result in

2
ω 
x − 2(1 − 2ζ ) x − 1 = 0 , where x =  B 
2 2

 ωn 
or
x = 1 − 2ζ 2 ± 4ζ 4 − 4ζ 2 + 2
2
ω 
For ζ = 0.2 , x = 2.2788 =  B  ⇒ ω B = 7.55 Rad/s
 ωn 
ζ = 0.5 x = 1.618034 = (
ω B )2 1⇒ ωB = 2.544
ωn
(b) We use the following statements.
num=25;
den=[1 2 25];
bode(num, den), grid
The result is shown in Figure 4.16

4.17
 Figure 4.16 The Bode plot for the system of Example 4.5.
To find the peak amplitude right click on the amplitude curve near the peak value, hold and
drag back and forth until you locate the peak value dB(ω r ) = 8.14 , and ω r = 4.8 Rad/s .
The bandwidth is the frequency where the amplitude is 0.707 or –3dB. Right-click on the
amplitude curve near –3 dB point, hold and move until gain is –3 dB. The bandwidth is
found to be 7.55 Rad/s.

4.7 Reduced order model

Since the poles of the closed-loop transfer function are the roots of the characteristic equation, they control the t
half s-plane close to origin are dominant poles since their time constants are large. These poles are referred to as
located far away to the left of the dominant poles they have very little effect on the transient response as these e
to the left than the dominant poles the system can be represented by reduced-order model.

4.18
Example 4.6

The closed-loop transfer function of a control system is described by the following third-
order transfer function

C (s) 750
= T (s) = 3 2
R( s) s + 36s + 205s + 750

(a) Find the dominant poles of the system

(b) Find a reduced-order model
(c) Obtain the step response of the third-order system and the reduced-order system on the same
figure plot, and find their time-domain specifications.
(d) Obtain the frequency (Bode Magnitude) of the third-order system and the reduced-order system
on the same figure plot, and find M pω , ω r , and ω B for each model.

(a) The command

den = [ 1 36 205 750];

r = roots(den)
result in
r=
-30
-3 + 4i
-3 - 4i
Therefore the transfer function is
750 25
T (s) = =
( s + 30)( s + 6s + 25) (1 + 0.0333s)( s 2 + 6s + 25)
2

The time constant of the real pole at s = −30 is τ 1 = 1/ 30 which is negligible compared to
the time constant of τ 2 = 1/ 3 for the dominant poles −3 ± j 4 . Therefore the approximate
reduced model transfer function is

25
T (s)  2
( s + 6 s + 25)

We use the following commands

num1 = 750; % third-order system num

den1 = [1 36 205 750]; % third-order system den
num2=25; % reduced-order system num
den2 = [1 6 25]; % reduced-order system den
4.19
sys3rd = tf(num1, den1) % third-order system
sys2nd = tf(num2, den2) % reduced-order system
ltiview('step', sys3rd, 'b', sys2nd, 'r')

Figure 4.17 Third-order and reduced-order step responses of Example 4.6.

With a simple right-click on the mouse the time-domain specifications are

obtained. Next we right-click on the mouse and select the Bode Mag plot. Place
the curser on the response near –3dB, left-click hold and move to display –3 dB.
Also right-click and chose Characteristics/peak value to display the peak.

4.20
 Figure 4.18 Third-order and reduced-order Bode plot of Example 4.6.

Table 4.1

tr - s t p -s P.O. ts - s M pω ω r - rad/s ω B - rad/s

Third-order 0.371 0.791 9.47 1.19 0.352 2.75 0.304
Model
Second-ord 0.378 0.828 9.33 1.22 0.352 2.75 0.304
Model

Time-domain and frequency-domain specifications are summarized in Table 4.1. As it

can be seen the reduced model step response and time-domain specifications are very
close to the third-order system. Also peak frequency response and the bandwidth of the
two systems are very close to each other. At very high frequency the reduced-order
model has more attenuation.

4.21
 1

EE-371 Frequency Response Analysis

Part I

Frequency Response Analysis and Design

The frequency response method and the root-locus method are simply two different
ways of applying the same basic principles of analysis. These methods supplement
each other, and in many practical design problems, both techniques are employed.
One advantage of the frequency response method is that the transfer function of
a system can be determined experimentally by frequency response tests. Further-
more, the design of a system in the frequency domain provides the designer with
control over the system bandwidth and over the effect of noise and disturbance on
the system response.
As we have seen the phase and gain margins of a control system are a mea-
sure of the closeness of the polar plot to the (-1, 0) point. Therefore, the open-loop
frequency response gain margin and phase margin may be used in the design crite-
ria.
The gain and phase margins along with ωpc and ωgc are obtained more easily
from the Bode plot. The phase margin may be read directly off the Bode plot at the
frequency which the amplitude curve crosses the 0 dB line, and the gain margin
may be read (in decibels) at the frequency at which the phase angle curve crosses
the −180◦ line. For satisfactory operation the phase margin should be between 30◦
and 60◦ , and the gain margin should be greater than 6 dB.
Additional performance specifications in terms of closed-loop frequency re-
sponse are the closed-loop system bandwidth, ωB , and the closed-loop system res-
onant peak magnitude, Mp . The bandwidth indicates how well the system tracks
an input sinusoid and is a measure of the speed of response. If the bandwidth is
small, only signals of relatively low frequency are passed, and the response is slow;
whereas a large bandwidth corresponds to a faster rise time. Therefore, the rise time
and the bandwidth are inversely proportional to each other.
The frequency at which the peak occurs, the resonant frequency, is denoted
by ωr , and the maximum amplitude, Mp , is called the resonant peak magnitude.
Mp is a measure of the relative stability of the system. A large Mp corresponds
to the presence of a pair of dominant closed-loop poles with small damping ratio,
which results in a large maximum overshoot of the step response in the time do-
main. In general, if Mp is kept between 1.0 and 1.7, the transient response will be
acceptable.
In summary, the frequency response design provides information on the steady-
state response, stability margin, and system bandwidth. The transient response per-
formance can be estimated indirectly in terms of the phase margin, gain margin,

EE-371 H. Saadat
2

and resonant peak magnitude. Percent overshoot is reduced with an increase in the
phase margin, and the speed of response is increased with an increase in the band-
width. Thus, the gain crossover frequency, resonant frequency, and bandwidth give
a rough estimate of the speed of transient response.
A common approach to the frequency response design is to adjust the open-
loop gain so that the requirement on the steady-state accuracy is achieved. This
is called the proportional controller. If the specifications on the phase margin and
gain margin are not satisfied, then it is necessary to reshape the open-loop transfer
function by adding the additional controller Gc (s) to the open-loop transfer func-
tion. Gc (s) must be chosen so that the system has certain specified characteristics.
This can be accomplished by combining proportional with integral action (PI) or
proportional with derivative action (PD).

Proportional Controller

The P controller is a pure gain controller. The design is accomplished by choosing

the gain KP for the uncompensated system to give the desired steady-state error.
When the gain KP is varied, the phase angle plot will not be affected. The Bode
magnitude curve is shifted up or down to correspond to the increase or decrease in
KP . Similarly, the effect of changing KP on the Nyquist diagram is to enlarge or
reduce it; the shape of the Nyquist diagram cannot be changed.

Example 1
A plant has a minimum-phase transfer function

K K
G(s) = = 3
s(s + 2)(s + 8) s + 10s2 + 16s

sketch the polar plot and determine the closed-loop system stability the phase
crossover frequency, and the gain margins for K = 50.6, K = 160, and K = 506

The frequency response transfer function is

K
G(jω) =
jω(2 + jω)(8 + jω)

For ω = 0, G(jω) = ∞6 −90◦ , and at ω = ∞, G(jω) = 06 −270◦ . To find the

real-axis intersection, we write G(jω) as

−jK
G(jω) =
ω[16 − ω 2 + j10ω]
EE-371 H. Saadat
 3

Multiplying numerator and denominator by the conjugate of denominator, we get

−jK[16 − ω 2 − j10ω] −K[10ω + j(16 − ω 2 )]

G(jω) = 2 2 2
=
ω[(16 − ω ) + 100ω ] ω[(16 − ω 2 )2 + 100ω 2 ]

To find the real -axis crossing and the phase crossover frequency, we set the imagi-
nary part of the numerator to zero, resulting in ωpc = ±4 rad/s. Therefore the gain
at ωpc is
−10K −K
G(j4) = 2
=
100(4) 160

The Nyquist function is used to obtain an accurate plot for the three values of K.
Nyquist plot
1

0.5

K = 506
−0.5
K = 160
K = 50.6
−1

−1.5
−4.5 −4 −3.5 −3 −2.5 −2 −1.5 −1 −0.5 0
The results for the specified controller gain are

K |G(jω)| G.M. G.M.dB Stability

50.6 0.31625 3.162 10 Stable
160 1 1 0 Marginally stable
506 3.1625 0.3162 -10 Unstable

Example 2
The bode (magnitude and phase angle) plots of the above open-loop transfer func-
tion for k = 50.6, k = 160, and k = 506 are as shown in the next page. From the
plots determine the phase and gain crossover frequencies, the gain margin and the
phase margin for the specified controller gains.
EE-371 H. Saadat
 4

Uncompensated and compensated bode plots

30
K = 106 K = 506
K = 50.6
20

10
dB

−10

−20

−30
0.1 0.2 0.4 0.6 0.8 1 2 3 4 5 6 7 8 910

−100

−120

−140
θ, degree

−160

−180

−200

−220
0.1 0.2 0.4 0.6 0.8 1 2 3 4 5 6 7 8 910
ω, rad/s
The results from the bode plots are

K ωpc ωgc G.M.dB P.M. Stability

50.6 4 2.1 10 29◦ Stable
160 4 4 0 0◦ Marginally stable
506 4 6.75 -10 −23.5◦ Unstable

EE-371 H. Saadat
 5

Example 3
Obtain the Bode plot, gain and phase margins for the feedback control system with
the open-loop transfer function

8 8
G(s) = = 3
s(s + 1)(s + 4) s + 5s2 + 4s

Determine the gain factor KP of a proportional controller such that the steady-
state error due to a ramp input will equal 0.25. Obtain the frequency response of
the compensated system and the new gain and phase margins.
The steady-state error specification requires

1
ess = = 0.25
Kv
or

Kv = 4

where the velocity error constant Kv is given by

8Kp
Kv = lim s = 2Kp = 4
s→0 s(s + 1)(s + 4)

or

Kp = 2

The compensated open loop transfer function to satisfy the steady state error is

2(8) 16
G(s) = = 3
s(s + 1)(s + 4) s + 5s2 + 4s

Using bode function, the open loop frequency response for the compensated and
uncompensated system is obtained as shown.

EE-371 H. Saadat
 6

Uncompensated and compensated bode plot

30

20

10
dB

0
GMc ≈ 2 dB
−10 GMu ≈ 8 dB

−20

−30
0.1 0.2 0.4 0.6 0.8 1 2 3 4 5 6 7 8 910

ωgcU ωgcC

−100

−120
θ, degree

−140

−160
P.M. −156 + 180 ≈ 22° P.M. ≈ 5°
−180

−200
0.1 0.2 0.4 0.6 0.8 1 2 3 4 5 6 7 8 910
ω, rad/s ω
pc
From the frequency response plot, ωpc , ωgc , gain margin and phase margin for the
uncompensated and compensated systems are:

Compensated ωpc = 2 ωgc = 1.2 G.M.dB = 8 P.M. = 22◦ ess = 0.5

Uncompensated ωpc = 2 ωgc = 1.79 G.M.dB = 2 P.M. = 5◦ ess = 0.25
It can be seen that increasing the gain values improves the steady-state behavior but
will decrease the phase and gain margins resulting in poor stability. It is then nec-
essary to redesign the system by using a suitable controller to alter the frequency
response so that the performance specifications will be met.

EE-371 H. Saadat
Operational Amplifier

The operational amplifier is an

integrated circuit (IC) composed of about thirty tran-
sistors, a dozen resistors and several capacitors man-
ufactured on a single piece of silicon crystal called a
chip. The Op amps are used in many electronic cir-
cuits for both linear and nonlinear operations. The cir-
cuit symbol for an op amp with its five essential termi-
nals is shown in Figure 4.1.

c c
c
+ve power

c c
......
.
.
. ........
.
. ...... .
supply
.
.
.
.
.
.
.
.
.
.
.
.
.................................................................................................
.
.
.
- ........
.
. ......
......
......

+
.
. ......
.
.
. .....
.
. ......
. .................................................................................................
.
. ......
.
. ......

+
.
.
.
. ........
.
. . .....
.
.
+ .
.. .
..

vn
.................................................
. .......
.
. ...... . .
. ..... .
+
.
. .

vo
...........
. .
.
.
. .
.

vp
.

ve power
supply
...................................................................................................................................................................................................................................................................

Figure 4.1 Circuit symbol for an op amp.

There are more terminals in a typical op amp, but we
are not concerned with the remaining terminals. Our
focus is on the terminal behavior of the amplifier as a
circuit element and not on the internal structure.
EE-253 H. Saadat
71
Thus, we treat the op amp as a black box with the
following essential terminals:
 The inverting terminal designated by - sign.
 The noninverting terminal designated by + sign.
 The output terminal.
 The positive power supply.
 The negative power supply.
When a voltage vn is applied to the inverting terminal
the output voltage is an inverted amplified signal of the
input vn. The voltage vp applied to the noninverting
terminal is amplified without inversion. Thus, with both
input voltages the output voltage is
v0 = Avn + Avp
= A(vn vp)
Where A is the open-loop gain of the op amp. A is
very large, typically in the order of 10 10
5– 6 for DC
and low frequency signals. The gain drops with in-
crease in the frequency and approaches unity with
frequency in the range of 50 MHz.
EE-253 H. Saadat
72
Voltage transfer characteristic
The transfer characteristics describes how the output
voltage varies as a function of the input voltage as
shown in Figure 4.2.
.

VCC
.
.
.
.
.
.
.
.
.
.
..
......... ......... ......... ..................................................................................................
.
. ..
.
.
. ...
.
. ..
.
.
. ....

v0
.
. .
..
.
.
. ...
. ..

Slope = A
.
. ..
.
.
. ..
.
.
. ...
.
.
.
. ...
.
. ..
.
. ..
.
.
.
. ...
.
. ..
.
. .
.
. ..
. ....
.
.
. ...
.
.
. .
. ...
.
.
. .. .
.

v vn)
....
.
.
..........................................................................................................................................................................
......................................................................................................................................................................

0 ( p
..
....
.
.
.
.... . .
.
.
... . .
.... . .
.
.
.... . .
.
.. .
.
. .
.
.
..
. .
.
.
.. .
.
.
.. .
.
.
. .
.
..
. .
.
.
.. .
... .
.
.
... .
.
.
... .
.
.
.. .
.
.

VCC
. .
.... .
.
.
...
. .
.
.
... .
.
................................................................................................. ........ ........ .........
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

j j
When vp vn is small the op amp behaves as a linear
device. Outside this linear range the op amp saturates
and behaves as a nonlinear device. For operation in
the linear range the net input voltage is constrained to
the voltage VCC applied to the power terminals. The
op amp output is given by
8
>
< VCC A(vp vn) < VCC
v0 = > A(vp vn) VCC  A(vp vn)  VCC
: VCC A(vp vn) > VCC
EE-253 H. Saadat
73
For a typical op amp operating on its linear region,
the voltage difference between the two inputs is very
small, less than 2 mV.

vx = vp vn  0
Also the input resistance is very large, and the output
resistance is negligibly small. For example, the com-
10
mon 741 op amp has Aapprox 5 , Rin 2
M ,
and Rout  40
. Because of the high input resis-
tance, negligibly small current flows into the two op
amp inputs, usually on the order of A.
In most practical applications, the circuit analysis can
be simplified with good accuracy by assuming the op
amp to be ideal, i.e.,
 A = 1, so that vn = vp
 Rin = 1, so that in = ip = 0
 Rout = 0
EE-253 H. Saadat
74
Inverting Amplifier
Since the open-loop gain of an op amp is infinitely
high to stabilize the operation a portion of the output
signal is returned to the input inverting terminal, re-
ducing the input voltage applied to this terminal which
in turn reduces the output voltage, and the op amp
operates on its linear region. Figure 4.3 shows the
circuit configuration for an inverting op amp.
R2
b
... ... ... ...
......................................................................................................................................... .
.
.... ... ... ... ...

b
.
.
.
.... .
.
.
... .
.

i-n +VCC
.
.
... .

b
.
.
... .
.
.
.... .
.
.
.
... .


.

R1
b
.
.
... .
.
... .
.
.
.
.... .
.
.
.
.
.
..............
. .

b
... .
.
. .
.
. .
. ... .
. .
.
... .. .. .. .
.
- . .
. ....... . .
.

vx
.......................... ........................................................................................ ....... .
.
. ... .. .. .. ...... .
.
.... ... .


...... .
... .... ...... .
.
...... .
... ... ...............................................
.
...........................................
.... .... .......
.
....
+
... ... ..
...
.. ......
...
+ ..

vi
. .
..
............................................ .......
.. .. ...... .
+ ... ........... .
.

vo
.... .
.
.
.... ..... .
.
.

VCC
... .

- ..
..
.
....
....
... ...
.... ...
... ...
.... ....
... ..
............................................................................................................................................................................................................................................................................................
..
.......................
...........
......

Applying KCL at the inverting node, we have

vx vi vx v0
R1
+ R + in = 0
2
Assuming ideal op amp in = 0, and vx = 0, we get
vi v0 R2
R1
+ R2
= 0 or v0 =
R1
vi

EE-253 H. Saadat
75
Noninverting Amplifier
The circuit configuration for a noninverting op amp is
shown in Figure 4.4
R2

b
.... ... ... ...
............................................................. ........ ........ ........ .............................................................
.
.

bb
.
... .
.
.
.... .
.
.
.

+VCC
.... .
.
.

b
... .
.
.
.
... .

i
.
.... .
.
.
.
.... .

R1 - -
.

n
.
.
.... .
.
... .
.
.
.
. .
... . .


..... . .
.

b
. . .
.... ... .......... . . .
.

R3 vn i-p +
. ...... . .
. .
.. .. .. .. .
. . . .
.
.. . . . . . ....... .
....................................................................................... ........ ........ ........ ....................... .............................................. ...... .
.
.
. ...... .
... ... ...... .
.
... .... ......
......
.
.
.
.
.... ... ...
..
..
..
..
..
..
..
..
...
..
..
.............................................................
... .... .......


.

+
.....

vp
.... ..
. ...
..
... ...... ..... ..... .....
. . . .
. .
.........
...................... .... .... .... .................... ............................................ ......
... ... .. .. .. .. ... .......... . .

vo
.
... ... ............ .
.
.
.
.... .... . .

vi
.

VCC
.
....
...
..
....
+
...
...
...
.
- ....
.... ...
.... ...
....................................................................................................................................................................................................................................................................................................................................................
...
......................
...........
......

Applying KCL at the inverting node, we have

vn vn v0
R1
+ R2
+ in = 0
At the noninverting terminal vp = vi R3 ip. Assum-
ing ideal op amp in = ip = 0, and vn = vp = vi.
Substituting in the first equation, we have
vi vi v0
R1
+ R2
= 0
or
R
v0 = (1 + 2 )vi
R1
EE-253 H. Saadat
76
Summing amplifier circuit Rf

b
... ... ... ...
........................................................................................................................................
.
. .
.. .. .. ..

Ra

 bb
.
. .
.
.
. .
.
.
. .
.
.
. .
.

i-n +VCC
.
. .
.
. .

b
. .
.
.
.
. .
.
. .
.
. . . . . . . . .
.
. .
.
....................................................................................................................................................................................................................................................... .
.
.
.
. .
.
.

Rb
.
. .. .. .. .. .
. .
. .
.
.
.
. .
. .
. .
.
.
. .
. .
. .
.
. .
. . .
.
.
. .
. .
. . .
.
.
. .
.
. .
.
. . .
.
. .
.
. . .


. . .

b
.
. .
. .
. .......... . .
. . . . . .
.

vn
. . . . ...... . . .
.
.
.
.
.
.
... .... .... ....
. . . .
........................................................................................................................ ..... ..... ..... ...............................
.
.
.
.
.
......................................................................................... -.
.
.
.
.. ...
.
.
.
.......
.
.
.
.
.

Rc
.
. . .. .. . . .. . . ......... .
.
. . . .
.
. ..
. .
.
. .
.
.
...... .
.
.
.
. .
. .
. .
. ...... .
...............................................

va +-
.
.
. .
. .
. .
. . . ...........................................
... .
.
.
.
.
. ..... ....

+
... .
. .
.
. . ....

vp
....
.
.
. .
. . ........
. . . . . . . . .
+. .

vb
. ................................................................................................. . .. .....
. ...
. .
...... ............
. .
. .
. .
. .
. ...
.
...
.......
. .. .. .. .. .
. .
. ......... . .
.
+ .... .

vo
. .
. .
.
. .......... .
.
.... .
.
.
.
.
.. .
.
.

vc
.

VCC
.
. .. .
. .
.
.
.
.
.
.
.
.
.
.
.
.
- .
+ .
.
.
.
.
.
.
.
.
.
. .... .
.
.
.
.
.
.
.
.
.
.
.
....
...
..
- ...
.
.
.
.
.
.
.
.
.
.
.
.
. . .
.
.
. ..
. .
.
.
.
.
.
.
.
. ..
. ..
.
.
.
.
.
........................................................................................................................................................................................................................................................................................................................
................................................................................................................................................................................................

vn va vn vb vn vc vn vo
.. .
.
.

+ + + + in = 0
.
.
......................
...........
......

Ra Rb Rc Rf
vn = vp  0; and in = 0; thus we have
va vb vc vo
Ra Rb Rc Rf
= 0
Rf Rf Rf
vo = ( Ra va + R vb + Rc vc)
b
So the output voltage is an inverted scaled sum of the
three input voltages. If Ra Rb Rc Rs, then = = =
Rf
vo = ( va + vb + vc)
Rs
and if Rf = Rs, we have
vo = (va + vb + vc)
77
Difference amplifier circuit
Rf

b
... ... ... ...
......................................................................................................................................... .
.
.... ... ... ... ...

b
.
.
.
.... .
.
.
... .
.

+VCC
.
.
... .

b
.
.
... .
.
.

i
.... .
.
.
.
... .

 b b
.

Ra - -
.

n
... .
.
.
.... .
.
.
.
... .
.
. .
.
.
.... ............. .
.
.
.
.
.
.
..... . .

Rb vn i-p +
. .
. .
. ...... . . .
.
.
.................. .......
............... ......
. .
. ......
...... .....
. ...
. .......
. ..... . ..
..
...
..
..
..
...
..
...... ........ ........ ........ .................................................................
. ....... .
.
.
. ... .... .... .... ..... .
.
. . .. ...... .
.


. . .
.
.
. ...
.
......
....... .
.
.
.
.
. ..
. ................................................
.
...........................................
.
.
. ..
. ...
.....
. .. .....
+
.
.
. .
. ....
. .
.....
.
. ... .... .... .... .. ..
.

va
.
. .
....................... ..... ..... ..... ................................................................ . .......
.....
. .. .. .. .. . .. ...... . .
+ ... .......... .
.

vo
.... .... .

Rc vp VCC
.
.... .... ..... .
.

vb
. ..... .
.

-
.
+ ...........
.......... .
..
....
.
.
. .....
.
.
.
.
.
.
.
.
.
.
- .
..........
......
....
..
.
.
. ..... ....
.
. ... ...
.
.
. . .
...............................................................................................................................................................................................................................................................................................................................................................................
.
. ..
.
......................
...........
......

vn va vn vo
Ra
+ Rf
+ in = 0

Rc
in = 0 = ip = 0; and vn = vp = R + Rc vb; thus
b

vo =
( Ra + R f )
vn
Rf
v
Ra Ra a
Rc(Ra + Rf ) Rf
= Ra(R + Rc) vb Ra va
b
Rf (Ra=Rf + 1) Rf
= Ra(R =Rc + 1) vb Ra va
b
EE-253 H. Saadat
78
The output voltage is equal to the difference between
scaled replica of the two inputs. If Ra=Rf = Rb=Rc,
we get
Rf
v0 = (vb va)
Ra
and if
Rf = Ra
we obtain
vo = vb va

EE-253 H. Saadat

79