Topic 5.

Transfer Function Approach to

Modeling Dynamic Systems

Instructor: Prof. Kwang-Hyun Cho

Department of Bio and Brain Engineering

Korea Advanced Institute of Science and Technology

Web: http://sbie.kaist.ac.kr

BiS352 System Modeling in Bioengineering Fall 2013

Introduction

In this topic, we will learn the transfer-function approach to modeling

and analyzing dynamic systems.
In the field of system dynamics, transfer functions are frequently used
to characterize the input-output relationships of components or systems
that can be described by linear, time-invariant differential equations.

Transfer function
The transfer function of a linear, time-invariant (LTI) differential-equation
system is defined as the ratio of the Laplace transform of the output
(response function) to the Laplace transform of the input (driving function)
under the assumption that all initial conditions are zero.
Consider the LTI system defined by the differential equation:
(n) ( n 1 ) (m ) ( m 1 )
a 0 y a 1 y a n 1 y a n y b 0 u b1 u b m 1 u b m u ( n m )
where y(t) is the output of the system and u(t) is the input.

Kwang-Hyun Cho

1
 Introduction

The transfer function of this system is given by:

L [ output ]
T ransfer function G ( s )
L [ input ] zero initial conditions

Y ( s ) b0 s m b1 s m 1 b m 1 s 1 b m

U (s) a 0 s n a1 s n 1 a n 1 s 1 a n

By using the concept of a transfer function, it is possible to represent

system dynamics by algebraic equations in s.
If the highest power of s in the denominator of the transfer function is
equal to n, the system is called an nth-order system.

Kwang-Hyun Cho

Introduction

Comments on the transfer function

The applicability of the concept of the transfer function is limited to linear,
time-invariant differential-equation systems.
Still, the transfer-function approach is used extensively in the analysis and
design of such systems.

1) The transfer function of a system is a mathematical model of that system.

2) The transfer function is determined only by the property of a system
itself, unrelated to the magnitude and nature of the input or driving
function.
3) The transfer function includes the units necessary to relate the input to
the output, however, it does not provide any information concerning the
physical structure of the system.
4) If the transfer function of a system is known, the output or response can
be studied for various forms of inputs with a view toward understanding
the nature of the system.
5) If the transfer function of a system is unknown, it may be established
experimentally by introducing known inputs and studying the output of
the system.

Kwang-Hyun Cho

2
 Introduction

Example 4-1
The equation of motion for the system is:

m x kx b x f ( t )
Taking the Laplace transform of both sides of
this equation and assuming that all initial
conditions are zero yields:

( ms 2 bs k ) X ( s ) F ( s )
From the last equation, the transfer function for
the system is:
Figure 4-1
X (s) 1

F ( s ) ms bs k
2

Kwang-Hyun Cho

Introduction: impulse-response function

The transfer function of a linear, time-invariant system is:

Y (s)
G (s)
U (s)
It follows that the output Y(s) can be written as the product of G(s) and
U(s):
Y ( s ) G ( s )U ( s )

Now, consider that the output of the system to a unit-impulse input

when the initial conditions are zero. Since the Laplace transform of the
unit-impulse function is unity, or U(s) = 1, the Laplace transform of the
output of the system is:
Y (s) G (s)
The inverse Laplace transform of the output yields the impulse response
of the system:
L 1 [ G ( s )] g (t )
It is called the impulse-response function, or the weighting function, of
the system.
Kwang-Hyun Cho

3
 Impulse response function reflecting the correlation
between the input and output
The impulse-response function g(t) is thus the response of a linear
system to a unit-impulse input when the initial conditions are zero.

The Laplace transform of g(t) gives the transfer function.

Therefore, the transfer function and impulse-response function of a

linear, time-invariant system contain the same information about the
system dynamics.

It is hence possible to obtain complete information about the dynamic

characteristics of a system by exciting it with an impulse input and
measuring the response.

In practice, a large pulse with a very short duration compared with the
significant time constants of the system can be considered as an
impulse.
Kwang-Hyun Cho

Impulse response function reflecting the relation

between the input and output (cont)
For a linear time-invariant system, u(t) is an input signal, y(t) is an
output signal, and g(t) is the impulse response function.
Superposition
property u(t)

u()(t-)
Time-invariant
property

u(t) y(t)

0 t=
g(t)
t
=

u(1)(t-1) u(1)g(t-1)
u (t )

u ( ) ( t )d
g(t)
y (t ) u (t ) g (t ) u ( ) g ( t )d
+

u(2)(t-2) u(2)g(t-2)
g(t)
The output of a linear time-invariant
system is equal to the convolution between
+

...

the input and the impulse response

u( )(t- u( )g(t-
) )
function of the system.
g(t)
Kwang-Hyun Cho

4
 Impulse-response function and system transfer function
For an identical linear time-invariant system, we can either use impulse
response g(t) or transfer function G(s) to represent the property of the
system (with an implicit assumption of zero initial conditions).
The output y(t) equals to the convolution between the input u(t) and the
impulse response (time domain).
The Laplace transform of output Y(s) equals to the multiplication of the Laplace
transform of input U(s) and the transfer function G(s) (complex frequency
domain).
From a perspective of mathematics, the following representations of an
elevator system are equivalent.

y (t ) u (t ) g (t )

u ( ) g ( t )d

Y ( s ) G ( s )U ( s )
Kwang-Hyun Cho

Block diagrams

A block diagram of a dynamic system is a pictorial representation of

the functions performed by each component of the system and of the
flow of signals within the system.
In a block diagram, all system variables are linked to each other
through functional blocks.
The functional block, or simply block, is a symbol for the mathematical
operation on the input signal to the block that produces the output.
Note that a signal can pass only in the direction of the arrows. Thus, a
block diagram of a dynamic system explicitly shows a unilateral
property.
Figure 4-2 shows an element of a block
diagram.
Note that the dimension of the output
signal from a block is the dimension of
Figure 4-2
the input signal multiplied by the
dimension of the transfer function in the
block.
Kwang-Hyun Cho

5
 Block diagrams

In general, the functional operation of a system can be visualized more

readily by examining a block diagram of the system than by examining
the physical system itself.
A block diagram contains information concerning dynamic behavior, but
it does not include any information about the physical construction of
the system.
Summing point
Figure 4-4 shows a symbol that stands
for a summing operation.
It is important that the quantities being
added or subtracted have the same
dimensions and the same units.
Branch point
A branch point is a point from which the
Figure 4-4 signal from a block goes concurrently to
other blocks or to summing points.

Kwang-Hyun Cho

Block diagrams

Block diagram of a closed-loop system

Figure 4-4 is a block diagram of a closed-loop system.
Any linear system can be represented by a block diagram consisting of
blocks, summing points and branch points.
When the output is fed back to the summing point for comparison with the
input, it is necessary to convert the form of the output signal to that of the
input signal.
This conversion is accomplished by the feedback element whose transfer
function is H(s), as shown in Figure 4-5.

Another important role of the

feedback element is to modify the
output before it is compared with
input, so that we can make an
output with a desired performance.
Figure 4-5 The procedure of changing H(s) to
achieve a desired C(s) is called
G (s) the design of controller.
C (s) R(s)
1 G (s)H (s) Kwang-Hyun Cho

6
 Block diagrams

Note that in a block diagram the main source of energy is not explicitly
shown and that the block diagram of a given system is not unique.
Example 4-2
Consider again the system shown in Figure 4-1.
The transfer function of this system is:
X (s) 1
Eq. 4-4
F ( s ) ms bs k
2

A block diagram representation of the system is shown:

Figure 4-1

Figure 4-6 (a)

Notice that Eq. 4-4 can be written as:
( ms 2 bs k ) X ( s ) F ( s )
Kwang-Hyun Cho

Block diagrams

Rewriting the latter equation as: ( ms 2 bs ) X ( s ) F ( s ) kX ( s ) Eq. 4-5

Figure 4-6 (b)

Eq. 4-5 can also be rewritten as:

1 k b
F ( s ) [ kX ( s ) bsX ( s )] ms 2 X ( s ) Or F (s) X ( s ) sX ( s ) s 2 X ( s )
m m m

Figure 4-6 (c)

Kwang-Hyun Cho

7
 Partial-fraction expansion with MATLAB

MATLAB representation of the polynomials in transfer functions

The transfer function of a system is represented by two arrays of numbers,
which are the parameters of two polynomials of s.
For example, consider a system defined by: Y ( s ) 25

U (s) s 2 4 s 25
This system is represented by two arrays, each containing the coefficients
of the polynomials in decreasing powers of s. We can write the two arrays
as:

Partial-fraction expansion with MATLAB

MATLAB allows us to obtain the partial-fraction expansion of the ratio of two
polynomials,
B ( s ) b (1) s h b (2) s h 1 b ( h )

A ( s ) a (1) s n a (2) s n 1 a ( n )

where a(1) 0 , some of a (i) and b (i) may be zero.

Kwang-Hyun Cho

Partial-fraction expansion with MATLAB

Num and den are row vectors that specify the coefficients of the
numerator and denominator of B(s)/A(s):
num = [b(1) b(2) b(h)]
den = [a(1) a(2) a(n)]
The command
[r,p,k] = residue(num,den)
finds the residue parameters, poles, and direct terms of a partial-fraction
expansion of the ratio of the two polynomials B(s) and A(s).
The partial-fraction expansion of B(s)/A(s) is given by:
B (s) r (1) r (2) r (n)
k (s)
A(s) s p (1) s p (2) s p (n)

As an example, consider the function

B ( s ) s 4 8 s 3 16 s 2 9 s 6

A(s) s 3 6 s 2 11 s 6
Kwang-Hyun Cho

8
 Partial-fraction expansion with MATLAB

As shown in MATLAB program 4-1, we obtain residue parameters (r),

poles (p), and direct terms (k).
MATLAB program 4-1 MATLAB program 4-1 is the MATLAB
representation of the partial-fraction
expansion of B(s)/A(s):
B ( s ) s 4 8 s 3 16 s 2 9 s 6

A(s) s 3 6 s 2 11 s 6
6 4 3
s2
s3 s2 s 1
Note that MATLAB first divides the numerator
by the denominator and produces polynomial
in s plus a remainder.
The command
[num,den] = residue(r,p,k)
converts the partial-fraction expansion back
to the polynomial ratio B(s)/A(s).
Kwang-Hyun Cho

Partial-fraction expansion with MATLAB

Example 4-3
Consider the spring-mass-dashpot system mounted on a massless cart as
shown in Figure 4-7.
Let us obtain a mathematical model of this
system by assuming that both the cart and
the spring-mass-dashpot system on it are
standing still for t < 0.
In this system, u(t) is the displacement of
the cart and the input to the system.
The displacement y(t) of the mass relative
to the ground is the output.
After a mathematical model of the system
Figure 4-7 is obtained, we determine the output y(t)
analytically when m = 10kg, b = 20 N-s/m,
k = 100 N/m.
The input is assumed to be a unit-step input.

Kwang-Hyun Cho

9
 Partial-fraction expansion with MATLAB

Applying Newtons second law to the present system and noting that
the cart is massless, we obtain:

m y b( y u) k ( y u)

or m y b y ky b u ku

The latter equation represents a mathematical model of the system

under consideration. Taking the Laplace transform of the equation,
assuming zero initial conditions, gives:
Y (s) bs k

U ( s ) ms bs k
2

Next, we shall obtain an analytical solution of the response to the unit-

step input.
Substituting the given numerical values:
Y (s) 20 s 100 2 s 10

U ( s ) 10 s 2 20 s 100 s 2 2 s 10
Kwang-Hyun Cho

Partial-fraction expansion with MATLAB

1
Since the input u is a unit-step function: U ( s )
s
Then the output Y(s) becomes:
2 s 10 1 2 s 10
MATLAB program 4-3 Y (s) 3
s 2 s 10 s s 2 s 2 10 s
2

To obtain the inverse Laplace transform of Y(s),

we need to expand Y(s) into partial-fractions.
MATLAB program 4-3 is the MATLAB
representation of the partial-fraction expansion
of Y(s).
0 .5 j / 6 0 .5 j / 6 1
Y (s)
s 1 j3 s 1 j3 s

Since Y(s) involves complex-conjugate poles, it

is convenient to combine two complex
conjugate terms into one as follows:

Kwang-Hyun Cho

10
 Partial-fraction expansion with MATLAB

0 .5 j / 6 0 .5 j / 6 s

s 1 j3 s 1 j3 ( s 1) 2 3 2

Then Y(s) can be expanded as:

1 s 1 s 11
Y (s)
s ( s 1) 3
2 2
s ( s 1) 2 3 2
1 s 1 1 3

s ( s 1) 3
2 2
3 ( s 1) 2 3 2
The inverse Laplace transform of Y(s) is obtained as:
1 t
y (t ) 1 e t c o s 3t
e s in 3 t , t 0
3
This equation is an analytic solution to the problem.

This solution is obtained under a zero initial condition. We call this

solution as zero state response of the system.
Kwang-Hyun Cho

Partial-fraction expansion with MATLAB

Example 4-4
Consider the mechanical system shown in Figure 4-8.
The system is at rest initially. The displacements x and y are measured from
their respective equilibrium positions.
Assume that p(t) is a step force input and the displacement x(t) is the
output.
Assume that m = 0.1 kg, b2 = 0.4 N-s/m, k1 = 6 N/m, k2 = 4 N/m, and p(t)
is a step force of magnitude 10 N.

The equations of motion for the system are:

x k1 x k 2 ( x y ) p
m
k 2 ( x y ) b 2 y
Laplace transforming these two equations:
( ms 2 k1 k 2 ) X ( s ) k 2Y ( s ) P ( s ) Eq. 4-9
k 2 X ( s ) ( k 2 b2 s )Y ( s ) Eq. 4-10
Figure 4-8
Kwang-Hyun Cho

11
 Partial-fraction expansion with MATLAB

Solving Eq. 4-10 for Y(s) and substituting the result into Eq. 4-9, we get:

k 22
( ms 2 k1 k 2 ) X ( s ) X (s) P(s)
k 2 b2 s

or [( ms 2 k1 k 2 )( k 2 b2 s ) k 22 ] X ( s ) ( k 2 b2 s ) P ( s )

From this, we obtain the transfer function:

X (s) b2 s k 2

P ( s ) mb2 s 3 mk 2 s 2 ( k1 k 2 ) b2 s k1 k 2
Substituting the given numerical values for m, k1, k2, and b2, we have:
X (s) 0.4 s 4

P ( s ) 0.04 s 3 0.4 s 2 4 b2 s 24
10 s 100

s 10 s 2 100 b2 s 600
3

Kwang-Hyun Cho

Partial-fraction expansion with MATLAB

Since P(s) is a step function of magnitude 10 N:

10
P(s)
MATLAB program 4-4 s
Then, X(s) can be written as:
10 s 100 10
X (s) 3
s 10 s 100 b2 s 600 s
2

On the basis of the MATLAB output, X(s) can be

written as:
0.6845 j 0.2233 0.6845 j 0.2233
X (s)
s 1.2898 j 8.8991 s 1.2898 j 8.8991
0.2977 1.6667

s 7.4204 s

1.3690( s 1.2898) 3.9743 0.2977 1.6667

( s 1.2898) 2 8.89912 s 7.4204 s

Kwang-Hyun Cho

12
 Partial-fraction expansion with MATLAB

The inverse Laplace transform of X(s) gives:

x ( t ) 1 .3 6 9 0 e 1 .2 8 9 8 t c o s (8 .8 9 9 1t )
0 .4 4 6 6 e 1 .2 8 9 8 t s in (8 .8 9 9 1t ) 0 .2 9 7 7 e 7 .4 2 0 4 t 1 .6 6 6 7

From the preceding examples, we have seen that once the transfer
function X(s)/U(s) = G(s) of a system is obtained, the response of the
system to any input can be determined by taking the inverse Laplace
transform of X(s):

L 1[ X ( s)] L 1[G(s)U (s)]

Kwang-Hyun Cho

Transient-response analysis with MATLAB

MATLAB representation of transfer-function systems

Figure 4-9 shows a block with a transfer function. Such a block represents a
system or an element of a system.

In MATLAB, the statement

sys = tf(num,den)
represents the system.
Figure 4-9

For example, consider the system: MATLAB program

4-5
Y (s) 2 s 25

X ( s ) s 2 4 s 25

MATLAB program 4-5 produces the

transfer function of the system.

Kwang-Hyun Cho

13
 Transient-response analysis with MATLAB

Step response
If numerator (num) and denominator (den) of a transfer function are known,
the command such as
step(sys) or step(num,den)
will generate a plot of a unit-step response and display a response curve on
the screen.
The computational time interval and the time span of the response can be
determined by the statement: t = 0 : t : T ;

Example 4-5
Consider again the spring-mass-dashpot system mounted on a cart shown
in Figure 4-7.
The transfer function of the system is:

Y (s) bs k

U ( s ) ms 2 bs k

Figure 4-7
Kwang-Hyun Cho

Transient-response analysis with MATLAB

Assume m = 10kg, b = 20 N-s/m, k = 100 N/m, and the input u(t) is a unit-
step function.
Substituting the given numerical values into the transfer function, we have:
Y (s) 20 s 100 2 s 10

U ( s ) 10 s 2 20 s 100 s 2 2 s 10
MATLAB program 4-6 will produce the unit-step response y(t).

MATLAB program 4-6

Figure 4-10
Kwang-Hyun Cho

14
 Transient-response analysis with MATLAB

Example 4-6
Consider again the mechanical system shown in Figure 4-8.
The transfer function X(s)/P(s) was found to be:
X (s) b2 s k 2

P ( s ) mb2 s 3 mk 2 s 2 ( k1 k 2 ) b2 s k1 k 2
The transfer function Y(s)/X(s) is obtained as:
Y (s) k2

X ( s ) b2 s k 2
Hence:
Figure 4-8 Y (s) Y (s) X (s) k2

P ( s ) X ( s ) P ( s ) mb2 s 3 mk 2 s 2 ( k1 k 2 ) b2 s k1 k 2
Assume that m = 0.1kg, b2 = 0.4 N-s/m, k1 = 6 N/m, k2 = 4 N/m, and p(t)
is a step force of magnitude 10 N.
Since p(t) is a step force of magnitude 10 N, we can define p(t) = 10u(t),
where u(t) is a unit-step input.
Kwang-Hyun Cho

Transient-response analysis with MATLAB

Substituting the numerical values and p(t) = 10u(t) into the given transfer
functions, we obtain:

X (s) 100 s 1000 Y (s) 1000

U ( s ) s 3 10 s 2 100 s 600 U ( s ) s 3 10 s 2 100 s 600
MATLAB program 4-8 produces the responses x(t) and y(t) of the system on
one diagram.
MATLAB program 4-8

Figure 4-12
Kwang-Hyun Cho

15
 Transient-response analysis with MATLAB

Impulse response
The unit-impulse response of a dynamic system defined in the form of the
transfer function can be obtained using one of the follows:
impulse(sys) or impulse(num,den)
impulse(sys,t) or impulse(num,den,t)
Before discussing computational solutions of problems involving impulse
inputs, we present some necessary background materials.

Impulse input
The impulse response of a mechanical
system can be observed when the
system is subjected to a very large force
for a very short time.
In handling impulse functions, only the
magnitude (or area) of the function is
important; its actual shape is immaterial.

Figure 4-13
Kwang-Hyun Cho

Transient-response analysis with MATLAB

Law of conservation of momentum

The momentum of a mass m moving at a velocity v is mv. According to
Newtons second law:
dv d
F ma m (m v) Hence, Fdt d ( m v )
dt dt
Integrating both sides, we have:
t2 v2
t1
Fdt
v1
d ( m v ) m v 2 m v1

Momentum is a vector quantity. The direction of the change in momentum

is the direction of the force.
In the absence of any external force,
d (m v) 0
or
m v constant
Thus, the total momentum of a system remains unchanged by any action
that may take place within the system, provided that no external force is
acting on the system.
Kwang-Hyun Cho

16
 Transient-response analysis with MATLAB

Example 4-8 (we have solved the same problem in time domain
analysis in Topic 3-2)
Consider the mechanical system shown in Figure 4-14.
A bullet of mass m is shot into a block of mass M (where M >> m).
Assume that when the bullet hits the block, it becomes embedded there.
Suppose that the bullet is shot at t = 0- and that the initial velocity of the
bullet is v(0-).
The displacement x of the block is measured from the equilibrium position
before the bullet hits it. Therefore, x(0-) = 0 and dx(0-)/dt = 0.
Assume the following numerical values:
M = 50kg, m = 0.01 kg, b = 100 N-s/m, k = 2500 N/m, v(0-) = 800 m/s

The input to the mechanical system

(the force between the bullet and
the block) in this case can be
considered as an impulse function,
the magnitude of which is equal to
the rate of change of momentum of
Figure 4-14 the bullet. Kwang-Hyun Cho

Transient-response analysis with MATLAB

By the previous analysis in Topic 3-2, the equation of motion during 0-
<t<0+ is constructed as:

M x (t ) b x (t ) kx F (t ) [ mv (0 ) m x ( 0 )] (t ) Eq. T5-1

i.e., 50 x (t ) 100 x (t ) 2500 x [8 0.01 x (0 )] (t )

and the equation of motion during t>0+ is written as

( M m) x(t ) b x(t ) kx F (t ) 0 Eq. T5-2

50.01 x (t ) 100 x(t ) 2500 x 0, with x(0 ) 0.15997 and x(0) 0
Because the systematic structures are not identical between Eqs. T5-1 and T5-2,
we cannot take Laplace transform from t=0-. Eq. T5-1 is used to find dx(0+)/dt.
Eq. T5-2 can be employed to find the response x(t) for t>0+.

Following the same procedures in Topic 3-2, dx(0+)/dt = 0.15997.

Then the derived initial condition dx(0+)/dt = 0.15997 is used to solve Eq. T5-2.
Kwang-Hyun Cho

17
 Transient-response analysis with MATLAB

Taking the Laplace transform of both sides of Eq. T5-2, we see that:

( M m)[ s 2 X ( s ) sx (0 ) x (0 )] b[ sX ( s ) x (0 )] kX ( s ) 0

( M m )[ sx (0 ) x (0 )] bx (0 )
X (s)
( M m ) s 2 bs k

Note that x(0 ) 0.15997 and x (0 ) 0. Therefore,
50.01 0.15997 8.0001
X (s) Eq. T5-3
50.01s 100 s 2500 50.01s 100 s 2500
2 2

The inverse Laplace transform of X(s) gives the response x(t) for t>0+.

8.0001 6.9993
X (s) 0.02286
50.01s 100s 2500
2
( s 0.9998)2 (6.9993) 2

Kwang-Hyun Cho

Transient-response analysis with MATLAB

Taking the inverse Laplace transform of the last equation yields:

x ( t ) 0.02286 e 0.9998 t sin 6.9993t

Thus, the response x(t) is a damped sinusoidal motion.

Example 4-9
Referring to Example 4-8, obtain the impulse response of the system shown
in Figure 4-14 with MATLAB.
Suppose there is an impulse input to the system, then the equation of
system is taken as:

( M m) x(t ) b x(t ) kx u (t ) (t )

The impulse response of the system is calculated under the zero initial
condition. We take the Laplace transform of the above equation.

X (s) 1 1

U ( s ) ( M m ) s bs k 50.01s 100 s 2500
2 2

Kwang-Hyun Cho

18
 Transient-response analysis with MATLAB

MATLAB program produces the impulse response of the system.

-3
x 10 Impulse Response of System
2.5

MATLAB program 1.5

Impulse Response g(t)

1
>> num = 1;
>> den = [50.01 100 2500]; 0.5
>> sys = tf(num,den);
>> impulse(sys); 0

>> grid;
>> title('Impulse Response of System'); -0.5

>> xlabel('t');
-1
>> ylabel('Impulse Response g(t)');
-1.5
0 1 2 3 4 5 6
t (sec)

According to the expression of X(s), we can find the x(t) for t>0+ by just
changing num to 8.0001.

Kwang-Hyun Cho

Transient-response analysis with MATLAB

Obtaining response to arbitrary input

The command lsim produces the response of linear, time-invariant systems
to arbitrary inputs.
If the initial conditions of the system are zero, then:
lsim(sys,u,t) or lsim(num,den,u,t)
Here, u is the input and t represents the times at which responses to u are
to be computed.

Ramp response: Example 4-10

Consider once again the system shown in Figure 4-7.
Assume that m = 10 kg, b = 20 N-s/m, k = 100
N/m, and u(t) is a unit-ramp input.
The transfer function of the system is:

Y (s) 2 s 10
2
U ( s ) s 2 s 10
Figure 4-7
Kwang-Hyun Cho

19
 Transient-response analysis with MATLAB

The resulting response curves are:

MATLAB program 4-11

Figure 4-18

Kwang-Hyun Cho

Transient-response analysis with MATLAB

Response to initial condition: Example 4-11
Consider the mechanical system shown in Figure 4-19.
Assume m = 1 kg, b = 3 N-s/m and k = 2 N/m.

Assume that the displacement x of mass m is measured

from the equilibrium position and that at t = 0 the

mass is pulled downward such that:

x(0) 0.1m and x(0) 0.05m / s

The system equation is: m x b x kx 0
The Laplace transform of the system equation gives:

Figure 4-19 m[ s 2 X ( s ) sx(0) x(0)] b[ sX ( s ) x(0)] kX ( s ) 0
or
(ms 2 bs k ) X ( s ) mx(0) s m x(0) bx(0)
Solving the last equation for X(s) and substituting the given numerical
values gives:
mx(0) s m x(0) bx(0) 0.1s 0.35
X ( s) 2
ms 2 bs k s 3s 2
Kwang-Hyun Cho

20
 Transient-response analysis with MATLAB

This equation can be written as:

0.1s 0.35
X (s) 1
s 2 3s 2
Hence, the motion of the mass m is the unit-impulse response of the
following system:
0.1s 0.35
G (s)
s 2 3s 2
Response of System subjected to initial condition
0.12

0.1

MATLAB program 0.08

Output x(t)
>> num = [0.1 0.35]; 0.06

>> den = [1 3 2];

>> sys = tf(num, den); 0.04

>> impulse(sys)
>> grid 0.02

>> title('Response of System subjected to initial

condition') 0
0 1 2 3 4 5 6 7 8 9
>> ylabel('Output x(t)') Time (sec)

Figure 4-20
Kwang-Hyun Cho

Transient-response analysis with MATLAB

This equation can also be written as:

0.1s 2 0.35s 1
X (s)
s 2 3s 2 s
Hence, the motion of the mass m is also the unit-step response of the
following system: 2
0.1s 0.35s
G (s)
s 2 3s 2

MATLAB program 4-12

Figure 4-20
Kwang-Hyun Cho

21
 Transient-response analysis with MATLAB

The previous two methods (using the impulse and step function) are
indirect ways of finding the responses subject to initial conditions

Can we directly use Matlab command to solve the problems subject to

initial conditions?

We can do this with the command lsim(sys,u,t,x0)!

However, a state space system description is required for this purpose.

We will study this in Topic 6.

Kwang-Hyun Cho

Mathematical modeling of electrical systems in view of

transfer function
Complex impedance
Consider the system shown in Figure 6-19.
In this system, Z1 and Z2 represent complex impedances.
The complex impedance Z(s) of a two-terminal circuit is the ratio of E(s),
the Laplace transform of the voltage across the terminals, to I(s), the
Laplace transform of the current through the element, under the
assumption that the initial conditions are zero, so that Z(s) = E(s) / I(s).

Complex impedance for resistance R is R.

Complex impedance for capacitance C is 1/Cs.
Complex impedance for inductance L is Ls.
The general relationship
E (s) Z (s)I (s)
corresponds to Ohms law for purely
resistive circuits.
The impedance approach is valid only if
Figure 6-19 the initial conditions are all zero.

Kwang-Hyun Cho

22
 Mathematical modeling of electrical systems in view of
transfer function
Complex admittance

The admittance, Y(s), is defined as the reciprocal of impedance:

1 I ( s)
Y (s)
Z (s) E ( s)

In general, admittance is complex. The real part is called conductance and

the imaginary part is called susceptance.

When writing a nodal equation (see Control Systems Engineering book), it

can be more convenient to represent circuit elements by their admittance.

Kwang-Hyun Cho

Mathematical modeling of electrical systems in view of

transfer function
Example 2-11 (from Control Systems Engineering book)
Find the transfer function Vc(s) / V(s) by using nodal analysis.
For this problem we sum currents at the nodes rather than summing
voltages around the meshes.
Node 1

Figure 2.6

Reference Node

In this case, we can directly write the nodal equation of the circuit, which
corresponds to Node 1.
V ( s) V ( s) V ( s) V (s) 1 1 1 V ( s)
L

L
L
0 ( )V ( s )
R Ls 1 R Ls R 1 L
R
1
2
R 1
2
1

Cs Cs
Kwang-Hyun Cho

23
 Mathematical modeling of electrical systems in view of
transfer function
Rearranging and expressing the resistances as conductances (reciprocal of
resistances), G1 = 1/R1, G2 = 1/(Ls), and G3 = 1/[R2+1/(Cs)], we obtain:
G1
( G1 G 2 G 3 )V L ( s ) G1V ( s ) VL (s) V (s)
G1 G 2 G 3
1
In Figure 2.6 (b), we also have VC ( s ) Cs VL (s)
1
R2
Cs
Solving for the transfer function Vc(s) / V(s) yields:

VC ( s ) 1 G1G 3

V (s) C s G1 G 2 G 3
For a more detailed introduction to the nodal analysis method, which is
quite often applied to large scale circuit analysis, see Engineering Circuit
Analysis, W.H., Hayt and J.E., Kemmerly.

Kwang-Hyun Cho

Mathematical modeling of electrical systems in view of

transfer function
Deriving transfer functions of electrical circuits with the use of
complex impedances
The transfer function of an electrical circuit can be obtained as a ratio of
complex impedances.
For the circuit shown in Figure 6-20, assume that the voltages ei and eo are
the input and output of the circuit, respectively. Then:

Eo (s) Z 2 (s)I (s) Z 2 (s)

Ei ( s ) Z1 ( s ) I ( s ) Z 2 ( s ) I ( s ) Z1 ( s ) Z 2 ( s )

For the circuit shown in Figure 6-21,

Figure 6-20
1
Z 1 L s R and Z 2
Cs
Hence, the transfer function is:
1
Eo (s) Cs 1

Figure 6-21 Ei ( s ) 1 LC s RC s 1
2
Ls R
Cs Kwang-Hyun Cho

24
 Mathematical modeling of electrical systems in view of
transfer function
Example 6-5
Consider the system shown in Figure 6-22.

Figure 6-22

The circuit shown in Figure 6-22 can be redrawn as in Figure 6-23(a), which
can be further modified to Figure 6-23(b).

Figure 6-23
Kwang-Hyun Cho

Mathematical modeling of electrical systems in view of

transfer function
In the system shown in Figure 6-23(b), the current I is divided into two
currents I1 and I2. Note that:
Z 2 I1 ( Z 3 Z 4 ) I 2 ; I1 I 2 I
We obtain:
Z3 Z4 Z2
I1 I; I2 I
Z2 Z3 Z4 Z2 Z3 Z4
Observe that
Z (Z Z 4 )
E i ( s ) Z 1 I Z 2 I1 Z 1 2 3 I,
Z2 Z3 Z4
Z 2Z 4
Eo (s) Z 4I2 I
Z2 Z3 Z4
We get:
Eo (s) Z 2Z 4

Ei (s) Z1(Z 2 Z 3 Z 4 ) Z 2 (Z 3 Z 4 )
Kwang-Hyun Cho

25
 Mathematical modeling of electrical systems in view of
transfer function
Substituting Z1 = R1, Z2 = 1/(C1s), Z3 = R2, and Z4 = 1/(C2s) into the last
equation yields:
1 1
Eo (s) C1s C 2 s

Ei (s) 1 1 1 1
R1 ( R2 ) ( R3 )
C1s C2s C1s C2s
1

R1 C 1 R 2 C 2 s ( R1 C 1 R 2 C 2 R1 C 2 ) s 1
2

which is the transfer function of the system. Notice that it is the same as Eq.
6-23.

Kwang-Hyun Cho

Mathematical modeling of electrical systems in view of

transfer function
Transfer functions of nonloading cascaded elements
The transfer function of a system consisting of two nonloading cascaded
elements can be obtained by eliminating the intermediate input and output.
For example, consider the system shown in Figure 6-16(a). The transfer
functions of the elements are
X 2 (s) X 3 (s)
G1 ( s ) and G2 (s)
X 1(s) X 2 (s)
If the input impedance of the second element is infinite, the output of the
first element is not affected by connecting it to the second element.

Then the transfer function of the whole

system becomes:
X 3 (s)
G (s)
X 1(s)
X 2 (s) X 3 (s)
G1 ( s )G 2 ( s )
Figure 6-16 X 1(s) X 2 (s)
Kwang-Hyun Cho

26
 Mathematical modeling of electrical systems in view of
transfer function
As an example, consider the system shown in Figure 6-17. The insertion of
an isolating amplifier between the circuits to obtain nonloading
characteristics is frequently used in combining circuits.
Since amplifiers have very high input impedances, an isolation amplifier
inserted between the two circuits justifies the nonloading assumption.
Thus, in this case:

Eo (s) 1 1
(K )
E i ( s ) R1C1 s 1 R2C 2 s 1
K

( R1C 1 s 1)( R 2 C 2 s 1)
Figure 6-17

Transfer functions of cascaded elements (refer to the next slide)

Many feedback systems have components that load each other.
Consider the system shown in Figure 6-22.
Assume that ei is the input and eo is the output.
Kwang-Hyun Cho

Mathematical modeling of electrical systems in view of

transfer function
The capacitance C1 and C2 are not charged initially.

Figure 6-22

In the previous analysis, we have already obtained the transfer function

between E0(s) and Ei(s)

Eo (s) 1
Eq. 6-23
Ei ( s ) R1C1 R 2 C 2 s ( R1C1 R 2 C 2 R1C 2 ) s 1
2

Kwang-Hyun Cho

27
 Mathematical modeling of electrical systems in view of
transfer function
The term R1C2s in the denominator of the transfer function represents the
interaction between two simple RC circuits.

The analysis just presented shows that if two RC circuits are connected in a
cascade so that the output from the first circuit is the input to the second,
the overall transfer function is not the product of their transfer functions.

The reason for this is that when we derive the transfer function for an
isolated circuit, we implicitly assume that the output is unloaded. In other
words, the load impedance is infinite, meaning that no power is consumed
at the output.

Kwang-Hyun Cho

Analogous systems

Systems that can be represented by the same mathematical model,

but that are physically different, are called analogous systems.
Thus, analogous systems are described by the same differential or
integrodifferential equations or transfer functions.

The concept of analogous systems is useful in practice, for the

following reasons:
1. The solution of the equation describing one physical system can be
directly applied to analogous systems in any other field.
2. Since one type of system may be easier to handle experimentally than
another, instead of building and studying a mechanical system (or a
hydraulic system, pneumatic system, or the like), we can build and study
its analog, for electrical or electronic systems are, in general, much easier
to deal with experimentally.

Kwang-Hyun Cho

28
 Analogous systems

Mechanical-electrical analogies
There are two electrical analogies for mechanical systems: the force-voltage
analogy and the force-current analogy.

Force-voltage analogy (Series analog)

Consider the mechanical system of Figure 6-24(a) and the electrical system
of Figure 6-24(b).
In the mechanical system, p is the external force, and in the electrical
system, e is the voltage source.
The equation for the mechanical
system is:
d 2x dx
m b kx p
dt2 dt
The equation for the electrical
system is:
di 1
Figure 6-24
L
dt
Ri
C id t e
Kwang-Hyun Cho

Analogous systems

In terms of the electric charge q, the last equation becomes:

d 2q dq 1
L 2
R q e
dt dt C
Comparing both equations, we see that the differential equations for the
two systems are of identical form. Thus, these two systems are analogous
systems.
The terms that occupy corresponding positions in the differential equations
are called analogous quantities, a list of which appears in Table 6-1.
Mechanical systems Electrical systems
Force p (torque T) Voltage e
Mass m (moment of inertia J) Inductance L
Viscous-friction coefficient b Resistance R
Spring constant k Reciprocal of capacitance, 1/C
Displacement x (angular displacement ) Charge q
Velocity x (angular velocity ) Current i
Table 6-1 Force-Voltage analogy Kwang-Hyun Cho

29
 Analogous systems

Force-current analogy (Parallel analog)

Another analogy between mechanical and electrical systems is based on the
force-current analogy.
Consider the mechanical system shown in Figure 6-25(a), where p is the
external force. The system equation is:
d 2x dx
m 2
b kx p Eq. 6-26
dt dt
Consider next the electrical system shown in Figure 6-25(b), where is is the
current source.
Applying Kirchhoffs current law
gives:
i L i R iC i s

where 1
L
iL edt
e de
Figure 6-25 iR , iC C
R dt
Kwang-Hyun Cho

Analogous systems

Thus, it can be written as:

1 e de Eq. 6-28
L edt R
C
dt
is

Since the magnetic flux linkage is related to the voltage e by the equation:
d
e
dt
Eq. 6-28 can be written in terms of as:
d 2 1 d 1
C 2
is Eq. 6-29
dt R dt L
Comparing Eqs 6-26 and 6-29, we can see that the two systems are
analogous.
The analogous quantities are listed in Table 6-2.

Kwang-Hyun Cho

30
 Analogous systems

Mechanical systems Electrical systems

Force p (torque T) Current i
Mass m (moment of inertia J) Capacitance C
Viscous-friction coefficient b Reciprocal of resistance, 1/R
Spring constant k Reciprocal of inductance, 1/L
Displacement x (angular displacement ) Magnetic flux linkage
Velocity x (angular velocity ) Voltage e
Table 6-2 Force-Current analogy

Comments
Analogies between two systems break down if the regions of operation are
extended too far.
In other words, since the mechanical models on which the analogies are
based are only approximations to the dynamic characteristics of physical
systems, the analogy may break down if the operating region of one system
is very wide.
Kwang-Hyun Cho

Analogous systems

Nevertheless, even if the operating region of a given mechanical system is

wide, it can be divided into two or more subregions, and analogous
electrical systems can be built for each subregion.
Analogy, of course, is not limited to mechanical-electrical analogy, but
includes any physical or nonphysical system.
Systems having an identical transfer function are analogous systems.
The concept of analogy is useful in applying well-known results in one field
to another.

Example 6-6
Obtain the transfer functions of the systems shown in Figure 6-26(a) and
(b), and show that these systems are analogous.

Figure 6-26
Kwang-Hyun Cho

31
 Analogous systems

For the mechanical system shown in Figure 6-26(a), the equation of motion
is:
b ( x i x o ) kx o
or
bx i kx o bx o
Taking the Laplace transform of this last equation and assuming zero initial
conditions, we obtain:
bsX i ( s ) ( k bs ) X o ( s )
Hence, the transfer function between Xo(s) and Xi(s) is:
b
s
X o (s) bs
k
X i ( s ) bs k b
s 1 We see that
k the two
For the electrical system shown in Figure 6-26(b), we have: systems are
analogous
Eo (s) RC s

Ei (s) RC s 1
Kwang-Hyun Cho

Mathematical modeling of electromechanical systems

Now, we move to the systems that are hybrids of electrical and

mechanical systems, the electromechanical systems.
We will obtain mathematical models of dc servomotors.
To control the motion or speed of dc servomotors, we control the field
current or armature current, or we use a servodriver as a motor-driver
combination.

Armature control of dc servomotors

In Figure 6-27, a magnetic field is developed by stationary permanent
magnets or a stationary electromagnet called the fixed field.

A rotating circuit called the

armature passes through the
magnetic field at right angles
and feels a force.
The resulting torque turns the
rotor, the rotating member of
the motor.
Figure 6-27
Kwang-Hyun Cho

32
 Mathematical modeling of electromechanical systems

DC servomotor
Stator fixed magnet
Rotor A rotating shaft
supported by bearings and coil
windings.
Armature the rotor and its
windings. This is the power-
producing component of the
motor. It rotates and carries the
current.
Commutator two insulated, independent
pieces of copper. An electrical switch that
periodically reverses the current in the
motor.
Brush provides a stationary electrical
contact to the moving commutator.

Kwang-Hyun Cho

Mathematical modeling of electromechanical systems

Armature control of dc servomotors (cont)

Ra = armature resistance,
La = armature inductance, H
ia = armature current, A
if = field current, A
ea = applied armature voltage, V
eb = back emf, V
Figure 6-27
= angular displacement of the motor shaft, rad
T = torque of inertia of the motor and load referred to the motor shaft, kg-
m2
b = viscous-friction coefficient of the motor and load referred to the motor
shaft, N-m/rad/s

The torque T developed by the motor is proportional to the product of the

armature current ia and the air gap flux , which in turn is proportional to
the field current, or
K f if w here K f is constant

Kwang-Hyun Cho

33
 Mathematical modeling of electromechanical systems

The torque T can therefore be written as:

T K f i f K 1i a w here K 1 is constant

For a constant field current, the flux becomes constant and the torque
becomes directly proportional to the armature current. So,
T K ia
where K is a motor-torque constant.
Notice that if the sign of the current ia is reversed, the sign of the torque T
will be reversed, which will result in a reversal of the direction of rotor
rotation.

When the armature is rotating, a voltage proportional to the product of the

flux and angular velocity is induced in the armature.
For a constant flux, the induced voltage eb is directly proportional to the
angular velocity:
d
eb K b
dt
where eb is the back emf and Kb is the back-emf constant.
Kwang-Hyun Cho

Mathematical modeling of electromechanical systems

The speed of an armature-controlled dc servomotor is controlled by the

armature voltage ea.
The differential equation for the armature circuit is:
d ia
La R a ia e b e a Eq. 6-31
dt
The armature current produces the torque that is applied to the inertia and
friction; hence:
d 2 d
J b T K ia
dt2 dt
Assuming that all initial conditions are zero and taking the Laplace
transforms of above equations, we obtain the following equations:
K b s (s ) E b (s) Eq. 6-33

(La s Ra )Ia (s) Eb (s) E a (s) Eq. 6-34

( Js 2 b s ) (s ) T ( s ) K I a (s ) Eq. 6-35

Kwang-Hyun Cho

34
 Mathematical modeling of electromechanical systems

Considering Ea(s) as the input and (s) as the output and eliminating Ia(s)
and Eb(s) from given equations, we obtain the transfer function for the dc
servomotor:
(s) K
Eq. 6-36
Ea (s) s[ L a J s ( L a b R a J ) s R a b K K b ]
2

The inductance La in the armature circuit is usually small and may be

neglected. If La is neglected, then the transfer function given by Equation
(6-36) reduces to:
K
(s) K Ra J Eq. 6-37

Ea (s) s ( Ra Js Rab K K b ) R b KKb
ss a
Ra J
Thus, the back emf increases the effective damping of the system. Damping
term

Kwang-Hyun Cho

Mathematical modeling of electromechanical systems

Equation (6-37) may be rewritten as:

(s) Km Km K /( Rab KKb ) motor gain constant

Ea (s) s ( T m s 1) Tm Ra J /( Rab KKb ) motor time constant
The last equation is the transfer function of the dc servomotor when the
armature voltage ea(t) is the input and the angular displacement (t) is the
output.

Gear train
Gear trains are frequently used in mechanical systems to reduce speed, to
magnify torque, or to obtain the most efficient power transfer by matching
the driving member to the given load.

Figure 6-28
Kwang-Hyun Cho

35
 Mathematical modeling of electromechanical systems

If the radii of gear 1 and gear 2 are r1 and r2, respectively, and the
number of teeth on gear 1 and gear 2 are n1 and n2, respectively, then:
r1 n
1
r2 n2
Because the surface speeds at the point of contact of the two gears must
be identical, we have:
r1 1 r2 2
where 1 and 2 are the angular velocities of gear 1 and gear 2,
respectively. Therefore:
2 r n
1 1
1 r2 n2
If we neglect friction loss, the gear train transmits the power unchanged. In
other words, if the torque applied to the input shaft is T1 and the torque
transmitted to the output shaft is T2, then:

T1 1 T 2 2
Kwang-Hyun Cho

Mathematical modeling of electromechanical systems

Example 6-7
Consider the system shown in Figure 6-29. Here, a load is driven by a
motor through the gear train.
Assume that the stiffness of the
shafts of the gear train is infinite,
that there is neither backlash nor
elastic deformation, and that the
number of teeth on each gear is
proportional to the radius of the
gear.
Find the equivalent inertia and
equivalent friction referred to the
motor shaft and those referred to
Figure 6-29 the load shaft.
The numbers of teeth on gear 1 and 2 are n1 and n2, respectively, and the
angular velocities of shaft 1 and 2 are 1 and 2, respectively.
The inertia and viscous friction coefficient of each gear train component are
denoted by J1, b1 and J2, b2, respectively.
Kwang-Hyun Cho

36
 Mathematical modeling of electromechanical systems

By applying Newtons second law to this system, the following two

equations can be derived: For the motor shaft (shaft 1),
J 1 1 b1 1 T1 Tm
where Tm is the torque developed by the motor and T1 is the load torque on
gear 1 due to the rest of the gear train.
For the load shaft (shaft 2),
J 2 2 b2 2 T L T2
where T2 is the torque transmitted to gear 2 and TL is the load torque.
Since the gear train transmits the power unchanged, we have:
T1 1 T 2 2
or
2 n
T1 T 2 T2 1
1 n2
If n1/n2 < 1, the gear ratio reduces the speed in addition to magnifying the
torque.
Kwang-Hyun Cho

Mathematical modeling of electromechanical systems

Eliminating T1 and T2 from given equations yields:

n1
J 1 1 b1 1 J 2 2 b2 2 T L Tm
n2
Since 2= (n1/n2)1, eliminating 2 from the last equation gives:
n1
2
n1
2
n1
J1 J 2 1 b1 b2 1 T L Tm Eq. 6-42
n2 n2 n2
Thus, the equivalent inertia and equivalent viscous friction coefficient of the
gear train referred to shaft 1 are given by:
2 2
n n
J 1 eq J1 1 J 2, b1 eq b1 1 b2
n2 n2
The effect of J2 on the equivalent inertia J1 is determined by the gear ratio
n1/n2. For speed-reducing gear trains, the ratio n1/n2 is much smaller than
unity. Similar comments apply to the equivalent viscous friction of the gear
train.

Kwang-Hyun Cho

37
 Mathematical modeling of electromechanical systems

In terms of the equivalent inertia and equivalent viscous friction coefficient,

Eq. 6-42 can be simplified to give:
J 1 eq 1 b1 eq 1 nT L Tm
where n = n1/n2.

The equivalent inertia and equivalent viscous friction coefficient of the gear
train referred to shaft 2 are:
2 2
n n
J 2 eq J 2 2 J1 b2 eq b2 2 b1
n1 n1
So the relationships between J1 eq and J2 eq and b1 eq and b2 eq are:
2 2
n n
J 1 eq 1 J 2 eq b1 eq 1 b2 eq
n2 n2
And Eq. 6-42 can be modified to give:
1
J 2 eq 2 b2 eq 2 T L Tm
n
Kwang-Hyun Cho

Mathematical modeling of electromechanical systems

Now that we have evaluated the mechanical constants, what about the
electrical constants in the transfer function of Eq. 6-37?
We will show that these constants can be obtained through a dynamometer
test of the motor, where a dynamometer measures the torque and speed of
a motor under the condition of a constant applied voltage.

Let us first develop the relationships that dictate the use of a

dynamometer (from Control Systems Engineering book)
Suppose the test is performed after the transient responses of the system
finishes. That is that the main physical quantities goes to constant (when
t). We usually call the final state of the system as steady state.

In this case, a dc voltage, ea, is applied, and the motor will turn at a
constant angular velocity, , with a constant torque, T.
Hence, dropping the functional relationship based on time from Eq. 6-31,
the following relationship exists when the motor is operating at steady state
with a dc voltage input.

Kwang-Hyun Cho

38
 Mathematical modeling of electromechanical systems

Steady state
d ia
La R a ia e b e a R a ia e b e a
dt
d
Note T K ia , and eb K b K b
dt
We have

Ra
T K b ea
K

Solving for T yields:

KbK K
T ea
Ra Ra
Above equation is a straight line, T vs , at
steady state as shown in Figure 2.38.
This plot is called the torque-speed curve.
The torque axis intercept occurs when the
Figure 2.38
angular velocity reaches zero.
Kwang-Hyun Cho

Mathematical modeling of electromechanical systems

That value of torque is called the stall torque, Tstall. Thus,

K
T ea
Ra
The angular velocity occurring when the torque is zero is called the no-load
speed, no-load. Thus,
ea
n o lo a d
Kb
The electrical constants of the motors transfer function can now be found
from above equations as:
K T
s ta ll
Ra ea
and ea
K
b
n o lo a d
The electrical constants can be found from a dynamometer test of the motor,
which would yield Tstall and no-load for a given ea.
Kwang-Hyun Cho

39
 Mathematical modeling of electromechanical systems

Example 6-8
Consider the dc servomotor system shown in Figure 6-30. The armature
inductance is negligible and is not shown in the circuit.

Obtain the transfer function

between the output 2 and the input
ea.

The torque T developed by the dc

servomotor is:
T K ia
Figure 6-30
The induced voltage eb is proportional to the angular velocity:
d 1
eb K b
dt
The equation for the armature circuit is:
R a ia e b e a
Kwang-Hyun Cho

Mathematical modeling of electromechanical systems

The equivalent moment of inertia of the motor rotor plus the load inertia
referred to the motor shaft is: 2
n
J 1 eq J 1 1 J 2
n2
The armature current produces the torque that is applied to the equivalent
moment of inertia J1 eq . Thus,
d 2 1
J 1 eq T K ia
dt 2
Assuming that all initial conditions are zero and taking Laplace transforms of
above equations, we obtain:
K b s 1 (s) E b (s )
Ra Ia (s) Eb (s) Ea (s)
J1 eq s 2 1(s) K I a (s)
Eliminating Eb(s) and Ia(s) from above equations, we obtain:
KKb K
J1 eq s2 1(s) Ea (s)
Ra Ra
Kwang-Hyun Cho

40
 Mathematical modeling of electromechanical systems

Noting that 1(s)/2 (s)= n2/n1, we can write this last equation as:

K K b n2 K
J1 eq s2 2 (s) Ea (s)
R a n1 Ra

Hence, the transfer function 2(s)/Ea (s) is given by:

n1
K
2 (s) n2

Ea (s) n
2

Ra J1 1 J 2 s K K b s
n2

Kwang-Hyun Cho

Mathematical modeling of operational-amplifier systems

In this section, we briefly discuss operational amplifiers.

Operational amplifiers, often called op-amps, are important building blocks
in modern electronic systems.

Consider the operational amplifier shown in Figure 6-31.

There are two terminals on the input side, one with a minus sign and the
other with a plus sign, called the inverting and noninverting terminals,
respectively.
We choose the ground as 0 volts and measure the input voltages e1 and e2
relative to the ground (the input e1 to the minus terminal of the amplifier is
inverted).
The total input to the amplifier is e2 - e1. The
ideal operational amplifier has the
characteristic:

eo K ( e 2 e1 ) K ( e1 e 2 )
where the inputs may be dc or ac signals and
K is the differential gain or voltage gain.
Figure 6-31
Kwang-Hyun Cho

41
 Mathematical modeling of operational-amplifier systems

The magnitude of K is approximately 105 to 106 for dc signals and ac signals

with frequencies less than approximately 10 Hz (the differential gain K
decreases with the frequency of the signal and becomes about unity for
frequencies of 1 MHz to about 50 MHz.)
Since the gain of the operational amplifier is very high, the device is
inherently unstable. To stabilize it, it is necessary to have negative feedback
from the output to the input.

In the ideal operational amplifier, no current flows into the input terminals
and the output voltage is not affected by the load connected to the output
terminal.
In other words, the input impedance is infinity and the output impedance is
zero.
In an actual operational amplifier, a very small (almost negligible) current
flows into an input terminal and the output cannot be loaded too much.

In our analysis here, however, we assume that the operational amplifiers

are ideal.

Kwang-Hyun Cho

Mathematical modeling of operational-amplifier systems

Inverting amplifier
Consider the operational-amplifier system shown in Figure 6-32.
Assume that the magnitudes of the resistances R1 and R2 are of comparable
order.
Let us obtain the voltage ratio eo/ei. In the derivation, we assume the
voltage gain to be K >> 1.
Let us define the voltage at the minus terminal as e.
Ignoring the current flowing into the amplifier, we have:

ei e ' eo e '
0
R1 R2

From this, we get:

ei e 1 1 '
o e
R1 R2 R1 R2
Figure 6-32
Kwang-Hyun Cho

42
 Mathematical modeling of operational-amplifier systems

Thus, ei e
o
R R2 Eq. 6-49
e' 1
1 1

R1 R2
Also,
eo K e ' Eq. 6-50

Eliminating e from above equations, we obtain:

ei e
o
e R R2 1 1 1 ei
o 1 eo
K 1 1 K R1 K R2 R2 R1

R1 R2
Hence,
R2

eo R1 eo R
K 1
R2 2 Eq. 6-51
ei R
1 2 R1 ei R1
R1
1
K Kwang-Hyun Cho

Mathematical modeling of operational-amplifier systems

Eq. 6-51 gives the relationship between the output voltage eo and the input
voltage ei.
From Eqs. 6-49 and 6-51 we have:
ei e 0
o
R1 R2
e
'
0
1 1

R1 R2
In an operational-amplifier circuit, when the output signal is fed back to the
minus terminal, the voltage at the minus terminal becomes equal to the
voltage at the plus terminal. This is called an imaginary short.
If we use the concept of an imaginary short, the ratio eo/ei can be obtained
much more quickly than the way we just found it.

Consider again the amplifier system shown in

Figure 6-32, and define:

ei e ' e ' e o
i1 i2
R1 R2
Figure 6-32 Kwang-Hyun Cho

43
 Mathematical modeling of operational-amplifier systems

Since only a negligible current flows into the amplifier, the current i1 must
be equal to the current i2. Thus,
e i e ' e ' e o

R1 R2
Because the output signal is fed back to the minus terminal, the voltage at
the minus terminal and the voltage at the plus terminal become equal, or e
= 0. Hence, we have:
ei eo R2
eo ei
R1 R2 R1
Note that the sign of the output voltage is the negative of that of the input
voltage. Hence, this operational amplifier is called an inverted amplifier. If,
R1 = R2, then the circuit is a sign inverter.

Kwang-Hyun Cho

Mathematical modeling of operational-amplifier systems

Obtaining mathematical models of physical operational-amplifier
systems by means of equations for idealized operational-amplifier
systems
We derive mathematical models of operational-amplifier systems, using
the following three conditions that apply to idealized operational-amplifier
systems:

1. From Figure 6-31, the output voltage eo is the differential input voltage
(e2-e1) multiplied by the differential gain K. That is,
eo K ( e 2 e1 )
where K is infinite. In designing active filters, we construct the circuit such
that the negative feedback appears in the operational amplifier like the
system shown in Figure 6-32. As a result, the differential input voltage
becomes zero, and we have
V oltage at negative term inal = V oltage at positive term inal
2. The input impedance is infinite.
3. The output impedance is zero.

Kwang-Hyun Cho

44
 Mathematical modeling of operational-amplifier systems
The use of these three conditions simplifies the derivation of transfer
functions of op-amp systems.
The derived transfer functions are, of course, not exact, but are
approximations that are sufficiently accurate.

There are two criteria that can be applied to analyze operational-amplifier

circuits. They are imaginary short and imaginary open criteria. Virtual
short says e+=e- when the input voltages of operators are analyzed. Virtual
open means i+=i-=0 when the input currents of operators are analyzed.

e+
e-
eo

imaginary short e+=e- imaginary open i+=i-=0

Kwang-Hyun Cho

Mathematical modeling of operational-amplifier systems

Example 6-9
Consider the operational-amplifier circuit shown in Figure 6-33.

If the op-amp is an ideal one, then the output voltage eo is limited and the
differential input voltage becomes zero, or voltage e and voltage e, are
equal.

' '' R1
Thus, ei e e eo
R1 R 2

Figure 6-33
Kwang-Hyun Cho

45
 Mathematical modeling of operational-amplifier systems

From this it follows that:

R
e o 1 2 ei
R1
This op-amp circuit is a noninverting circuit. If we choose R1 = , then eo =
ei, and the circuit is called a voltage follower.

Example 6-10
Consider the operational-amplifier circuit shown in Figure 6-34.

We define:
- e1 e ' e2 e '
i1 i2
+ R1 R2

e3 e ' e ' e o
i3 i4
R3 R4
Figure 6-34
Kwang-Hyun Cho

Mathematical modeling of operational-amplifier systems

Noting that the current flowing into the amplifier is negligible, we have:
e1 e ' e 2 e ' e 3 e ' e o e '
0
R1 R2 R3 R4
Since the amplifier involves negative feedback, the voltage at the minus
terminal and that at the plus terminal become equal. Thus e = 0, and
e1 e e e
2 3 o 0
R1 R2 R3 R4
or
R4 R R
eo e1 4 e 2 4 e 3
R1 R2 R3
If we choose R1 = R2 = R3 = R4, then

e o ( e1 e 2 e 3 )
This circuit is an inverting adder.

Kwang-Hyun Cho

46
 Mathematical modeling of operational-amplifier systems

Example 6-11
Consider the operational-amplifier system shown in Figure 6-35.
Obtain the transfer function for the system. Then obtain the response of the
system to a step input of a small magnitude.
Let us define:
ei e ' d ( e ' e o ) e ' e o
i1 i2 C i3
R1 dt R2

Noting that the current flowing into the

amplifier is negligible, we have:
i1 i 2 i3
Hence,
ei e ' d ( e ' e o ) e ' e o
C
R1 dt R2

Figure 6-35
Kwang-Hyun Cho

Mathematical modeling of operational-amplifier systems

Since the operational-amplifier involves negative feedback, the voltage at

the minus terminal and that at the plus terminal become equal. Hence e =
0.
ei d eo e
C o
R1 dt R2
Taking the Laplace transform of this last equation and assuming the zero
initial condition, we have:
Ei (s) R Cs 1
2 Eo (s)
R1 R2
which can be written as
Eo (s) R 1
2 (6-54)
Ei (s) R1 R 2 C s 1
Eq. 6-54 is the transfer function for the system, which is a first-order lag
system.

Kwang-Hyun Cho

47
 Mathematical modeling of operational-amplifier systems

Next, we shall find the response of the system to a step input.

Suppose that the input ei(t) is a step function of E volts; that is:
ei (t )
0
E
fo r t 0
for t 0
where we assume 0 < (R2/R1)E < 10 V. The output can be determined from:

R2 1
Eo (s) Ei (s)
R1 R 2 C s 1
R2 1 E

R1 R 2 C s 1 s
R2 E 1 1

R1 s s 1 /( R 2 C )
The inverse Laplace transform gives:
R2 E
eo (t ) 1 e t /( R 2 C )
R1
Kwang-Hyun Cho

Mathematical modeling of operational-amplifier systems

Example 6-12
Consider the operational-amplifier circuit shown in Figure 6-36.
The voltage at point A can be derived as:
e A ei e eA 1
o eA ( ei eo )
R1 R1 2
The Laplace transform of this last equation is:
1
E A (s) [ E i ( s ) E o ( s )]
2
The voltage at point B is:
1
Cs 1
EB (s) Ei (s) Ei (s)
1 R2C s 1
R2
Cs
Since the operational amplifier involves
negative feedback, the voltage at the
minus terminal and that at the plus
Figure 6-36 terminal become equal.
Kwang-Hyun Cho

48
 Mathematical modeling of operational-amplifier systems

Thus,
E A (s) E B (s)

and it follows that:

1 1
[ E i ( s ) E o ( s )] Ei (s)
2 R2C s 1
or
1
s
Eo (s) R2C s 1 R2C

Ei (s) R2C s 1 1
s
R2C

Kwang-Hyun Cho

49