Chapter 4 Time Response

Time Response
Introduction
The output response of a system is the sum of two responses: the forced response and the natural
response.
Although many techniques, such as solving a differential equation or taking the inverse Laplace
transform, enable us to evaluate this output response, these techniques are laborious and time-
consuming. Productivity is aided by analysis and design techniques that yield results in a minimum
of time. If the technique is so rapid that we feel we derive the desired result by inspection, we
sometimes use the attribute qualitative to describe the method.
The use of poles and zeros and their relationship to the time response of a system is such a
technique. Learning this relationship gives us a qualitative ‘‘handle’’ on problems. The concept of
poles and zeros, fundamental to the analysis and design of control systems, simplifies the evaluation
of a system’s response.

First Order Systems

A first-order system without zeros can be described by the transfer function shown in Figure 4.4(a).
If the input is a unit step, where R(s) = 1/s, the Laplace transform of the step response is C(s), where
a
C ( s )  G ( s ) R( s )  (4.5)
s s  a 
Taking the inverse transform, the step response is given by
ct   c f (t )  c n (t )  1  e  at (4.6)
where the input pole at the origin generated the forced response cf(t)=1, and the system pole at -a, as
shown in Figure 4.4(b), generated the natural response cn(t)=e-at. Equation (4.6) is plotted in Figure
4.5.
Let us examine the significance of parameter a, the only parameter needed to describe the transient
response. When t= 1/a,
e  at  e 1  0.37 (4.7)
t 1 / a
Or
ct  t 1 / a  1  0.37  0.63 (4.8)
We now use Eqs. (4.6), (4.7), and (4.8) to define three transient response performance
specifications.

Time Constant, Tc
We call 1/a the time constant of the response.

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

1
Tc 
a
From Eq. (4.7), the time constant can be described as the time for e-at to decay to 37% of its initial
value. Alternately, from Eq. (4.8) the time constant is the time it takes for the step response its final
value (see Figure 4.5).

FIGURE 4.5 First-order system response to a unit step.

The reciprocal of the time constant has the units (1/seconds), or frequency Thus, we can call the
parameter a the exponential frequency. Since the derivative of e-at is -a when t=0, a is the initial rate
of change of the exponential at t=0. Thus the time constant can be considered a transient response
specification for a first order system, since it is related to the speed at which the system responds to
a step input.
The time constant can also be evaluated from the pole plot (see Figure 4.4(b)). Since the pole of the
transfer function is at -a, we can say the pole is located at the reciprocal of the time constant, and
the farther the pole from the imaginary axis, the faster the transient response.
Let us look at other transient response specifications, such as rise time, Tr, and settling time, Ts, as
shown in Figure 4.5.

Rise Time, Tr
Rise time is defined as the time for the waveform to go from 0.1 to 0.9 of its final value. Rise time is
found by solving Eq. (4.6) for the difference in time at c(t)= 0.9 and c(t)=0.1. Hence,
2.31
c(t ) t  0.9  1  e  0.9 a 
a
0.11
c(t ) t  0.1  1  e  0.1a 
a
2.31 0.11 2.2
Tr    (4.9)
a a a

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

Settling Time, Ts
Settling time is defined as the time for the response to reach, and stay within, 2% of its final value.
Letting c(t)=0:98 in Eq. (4.6) and solving for time, t, we find the settling time to be
1  e  aTs  0.98
4
Ts  (4.10)
a

First-Order Transfer Functions via Testing

Often it is not possible or practical to obtain a system’s transfer function analytically. Perhaps the
system is closed, and the component parts are not easily identifiable.
Since the transfer function is a representation of the system from input to output, the system’s step
response can lead to a representation even though the inner construction is not known. With a step
input, we can measure the time constant and the steady-state value, from which the transfer function
can be calculated.
Consider a simple first-order system, G(s)=K/(s+a), whose step response is
K K /a K /a
C ( s )  G (s ) R( s )    (4.11a)
s s  a  s sa

K
c(t )  1  e  at
a
 (4.11b)

FIGURE 4.6 Laboratory results of a system step response test.

If we can identify K and a from laboratory testing, we can obtain the transfer function of the system.
For example, assume the unit step response given in Figure 4.6. We determine that it has the first-
order characteristics we have seen thus far, such as no overshoot and nonzero initial slope. From the
response, we measure the time constant, that is, the time for the amplitude to reach 63% of its final
value.
Prepared by: Dr. Mohammad Abdul Mannan
 Chapter 4 Time Response

Since the final value is about 0.72, the time constant is evaluated where the curve reaches
0.630.72=0.45, or about 0.13 second. Hence, a=1/0.13=7.7.
To find K, we realize from Eq. (4.11) that the forced response reaches a steady-state value of
K/a=0.72. Substituting the value of a, we find K=5.54. Thus, the transfer function for the system is
G(s)=5.54/(s+7.7). It is interesting to note that the response of Figure 4.6 was generated using the
transfer function G(s)=5/(s+7).

Second Order System

To become familiar with the wide range of responses before formalizing our discussion in the next
section, we take a look at numerical examples of the second-order system responses shown in
Figure 4.7. All examples are derived from Figure 4.7(a), the general case, which has two finite
poles and no zeros. The term in the numerator is simply a scale or input multiplying factor that can
take on any value without affecting the form of the derived results. By assigning appropriate values
to parameters a and b, we can show all possible second-order transient responses. The unit step
response then can be found using C(s)=R(s)G(s), where R(s)=1/s, followed by a partial-fraction
expansion and the inverse Laplace transform.

Overdamped Response, Figure 4.7(b)

For this response,
9 9
C ( s)   (4.12)

s s 2  9s  9  s( s  7.854)(s  1.146)

This function has a pole at the origin that comes from the unit step input and two real poles that
come from the system. The input pole at the origin generates the constant forced response; each of
the two system poles on the real axis generates an exponential natural response whose exponential
frequency is equal to the pole location. Hence, the output initially could have been written as
c(t )  K1  K 2 e 7.854t  K 3 e 1.146t
This response, shown in Figure 4.7(b), is called overdamped (So named because overdamped refers
to a large amount of energy absorption in the system, which inhibits the transient response from
overshooting and oscillating about the steady-state value for a step input. As the energy absorption
is reduced, an overdamped system will become underdamped and exhibit overshoot.).

We see that the poles tell us the form of the response without the tedious calculation of the inverse
Laplace transform.

Underdamped Response, Figure 4.7 (c)

For this response,
9 9
C ( s)   (4.13)
 2
s s  2s  9  s ( s  1  j 8 )(s  1  j 8 )
  
c(t )  K1  e  t K 2 cos 8t  K 3 sin 8t  
 K1  K 4 e  t cos 8t    
K 
Where, K1  K 4  K22  K 32 ;   tan 1  3 
 K2 

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

FIGURE 4.7 Second-order systems, pole plots, and step responses.

This function has a pole at the origin that comes from the unit step input and two complex poles that
come from the system. We now compare the response of the second-order system to the poles that
generated it. First we will compare the pole location to the time function, and then we will compare
the pole location to the plot. From Figure 4.7(c), the poles that generate the natural response are at
s  1  j 8 .
Comparing these values to c(t) in the same figure, we see that the real part of the pole matches the
exponential decay frequency of the sinusoid’s amplitude, while the imaginary part of the pole
matches the frequency of the sinusoidal oscillation.

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

Let us now compare the pole location to the plot. Figure 4.8 shows a general, damped sinusoidal
response for a second-order system.

The transient response consists of an exponentially decaying amplitude generated by the real part of
the system pole times a sinusoidal waveform generated by the imaginary part of the system pole.
The time constant of the exponential decay is equal to the reciprocal of the real part of the system
pole. The value of the imaginary part is the actual frequency of the sinusoid, as depicted in Figure
4.8.
This sinusoidal frequency is given the name damped frequency of oscillation, d. Finally, the
steady-state response (unit step) was generated by the input pole located at the origin. We call the
type of response shown in Figure 4.8 an underdamped response, one which approaches a steady-
state value via a transient response that is a damped oscillation.

Example 4.2: By inspection, write the form of the step response of the system in Figure 4.9.

Solution: First we determine that the form of the forced response is a step. Next we find the form of
the natural response. Factoring the denominator of the transfer function in Figure 4.9, we find the
poles to be s=-5j13.23. The real part, -5, is the exponential frequency for the damping. It is also
the reciprocal of the time constant of the decay of the oscillations. The imaginary part, 13.23, is the
radian frequency for the sinusoidal oscillations. Using our previous discussion and Figure 4.7(c) as
a guide, we obtain
c(t )  K1  e 5t K 2 cos13.23t   K 3 sin13.23t   K1  K 4 e t cos13.23t   
K 
Where, K1  K 4  K22  K 32 ;   tan 1  3  , and c(t) is a constant plus an exponentially
 K2 
damped sinusoid.

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

Undamped Response, Figure 4.7(d)

For this response,
9 9
C ( s)   (4.14)
 2
s s 9  s( s  j3)( s  j3)

This function has a pole at the origin that comes from the unit step input and two imaginary poles
that come from the system. The input pole at the origin generates the constant forced response, and
the two system poles on the imaginary axis at j3 generate a sinusoidal natural response whose
frequency is equal to the location of the imaginary poles. Hence, the output can be estimated as
c(t )  K1  K 4 cos13.23t   
his type of response, shown in Figure 4.7(d), is called undamped. Note that the absence of a real
part in the pole pair corresponds to an exponential that does not decay. Mathematically, the
exponential is e-0t=1.

Critically Damped Response, Figure 4.7 (e)

For this response,
9 9
C ( s)   (4.15)

s s 2  6s  9  s( s  3) 2

This function has a pole at the origin that comes from the unit step input and two multiple real poles
that come from the system. The input pole at the origin generates the constant forced response, and
the two poles on the real axis at -3 generate a natural response consisting of an exponential and an
exponential multiplied by time, where the exponential frequency is equal to the location of the real
poles. Hence, the output can be estimated as
c(t )  K1  K 2 e 3t  K 3te 3t
This type of response, shown in Figure 4.7(e), is called critically damped. Critically damped
responses are the fastest possible without the overshoot that is characteristic of the underdamped
response.

We now summarize our observations. In this section we defined the following natural responses and
found their characteristics:

1. Overdamped responses
Poles: Two real at -1, -2
Natural response: Two exponentials with time constants equal to the reciprocal of the pole
locations, or
c(t )  K1e  1t  K 2e  2t

2. Underdamped responses
Poles: Two complex at -d  jd
Natural response: Damped sinusoid with an exponential envelope whose time constant is
equal to the reciprocal of the pole’s real part. The radian frequency of the sinusoid, the
damped frequency of oscillation, is equal to the imaginary part of the poles, or
c(t )  Ae  d t cos d t   

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

3. Undamped responses
Poles: Two imaginary at  j1
Natural response: Undamped sinusoid with radian frequency equal to the imaginary part of
the poles, or
c(t )  A cos1t   

4. Critically damped responses

Poles: Two real at -1
Natural response: One term is an exponential whose time constant is equal to the reciprocal
of the pole location. Another term is the product of time, t, and an exponential with time
constant equal to the reciprocal of the pole location, or
c(t )  K1e  1t  K 2te  1t

The step responses for the four cases of damping discussed in this section are superimposed in
Figure 4.10. Notice that the critically damped case is the division between the overdamped cases
and the underdamped cases and is the fastest response without overshoot.

Exercise 4.3: For each of the following transfer functions, write, by inspection, the general form of
the step response:
Transfer Function Answer
400 6t
c(t )  A  Be cos19.08t   
G ( s) 
2
s  12s  400
900 c(t )  A  Be 78.54t  Ce11.46t
G ( s) 
2
s  90s  900
225 c(t )  A  Be 15t  Cte 15t
G ( s) 
s 2  30s  225
625 c(t )  A  B cos25t   
G( s) 
s 2  625

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

4.5 The General Second-Order System

Now that we have become familiar with second-order systems and their responses, we generalize
the discussion and establish quantitative specifications defined in such a way that the response of a
second-order system can be described to a designer without the need for sketching the response. In
this section, we define two physically meaningful specifications for second-order systems. These
quantities can be used to describe the characteristics of the second-order transient response just as
time constants describe the first-order system response. The two quantities are called natural
frequency and damping ratio. Let us formally define them.

Natural Frequency, n
The natural frequency of a second-order system is the frequency of oscillation of the system without
damping.
For example, the frequency of oscillation of a series RLC circuit with the resistance shorted would
be the natural frequency.

Damping Ratio, 
Available definition for this quantity is one that compares the exponential decay frequency () of
the envelope to the natural frequency ( n). This ratio is constant regardless of the time scale of the
response. Also, the reciprocal, which is proportional to the ratio of the natural period to the
exponential time constant, remains the same regardless of the time base.
We define the damping ratio, , to be
Exponential decay frequency, (σ d ) 1 Natural period ( s)
ξ 
Natural frequency (rad/s), (ωn ) 2π Exponential Time cons tan t

The general second-order system can be transformed to show the quantities  and n. Consider the
general system
b
G ( s)  (4.16)
2
s  as  b
Without damping, the poles would be on the j -axis, and the response would be an undamped
sinusoid. For the poles to be purely imaginary, a=0. Hence,
b
G( s)  (4.17)
2
s b
By definition, the natural frequency, n, is the frequency of oscillation of this system. Since the
poles of this system are on the j-axis at jb,
n  b (4.18)
Hence
b   n2 (4.19)

Now what is the term a in Eq. (4.16)? Assuming an underdamped system, the complex poles have a
real part, , equal to –a/2. The magnitude of this value is then the exponential decay frequency.
Hence,
Exponential decay frequency, (σ ) a / 2 a
ξ   (4.20)
Natural frequency (rad/s), (ωn )  n 2 n
from which
a  2ξ n (4.21)
Our general second-order transfer function finally looks like this:

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

 n2
G (s )  (4.22)
s 2  2 n s   n2

Now that we have defined  and vn, let us relate these quantities to the pole location. Solving for
the poles of the transfer function in Eq. (4.22) yields
s1,2    n   n  2  1 (4.24)
FromEq. (4.24) we see that the various cases of second-order response are a function of , they are
summarized in Figure 4.11.

Example 4.3: Given the transfer function of Eq. (4.23), find  and n.
36
G (s ) 
2
s  4.2 s  36
Solution:
Here, a=4.2, b=36;  n  b  36  6 ;
a 4. 2
a  2ξ n i.e. ξ    0.35
2 n 2  6
Comparing Eq. (4.23) to (4.22),  n2  36 , from which n=6. Also, 2ξ n  4.2 . Substituting the
value of n, ξ  0.35 .

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

Example 4.4: For each of the systems shown in Figure 4.12, find the value of  and report the kind
of response expected.

Solution: First match the form of these systems to the forms shown in Eqs. (4.16) and (4.22). Since
a=2n and n=b,
a
ξ (4.25)
2 b

Using the values of a and b from each of the systems of Figure 4.12, we find  = 1.155 for system
(a), which is thus overdamped, since  > 1;  = 1 for system (b), which is thus critically damped;
and  = 0.894 for system (c), which is thus underdamped, since  < 1.

Exercise 4.4: For each of the following transfer functions, do the following: (1) Find the value of 
and n, (2) characterize the nature of the response.

Answer
Transfer Function (1) (2)
 n Nature of Response
400
G ( s)  0.3 20 Underdamped
s 2  12s  400
900
G ( s)  1.5 30 Overdamped
s 2  90s  900
225
G ( s)  1 15 Critically damped
s 2  30s  225
625
G( s)  0 25 Undamped
2
s  625

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

Step Response of an Underdamped Second-Order System

Let us begin by finding the step response for the general second-order system of Eq. (4.22). The
transform of the response, C(s), is the transform of the input times the transfer function, or
 n2 K1 K1s  K 2
C ( s)    (4.26)

ss 2
 2 n s   n2  s s  2 n s   n2
2

where it is assumed that  < 1 (the underdamped case). Expanding by partial fractions, using the
methods described in Section 2.2, Case 3, Eq. (4.26) yields

s   n   n 1   2
1 1 2
C ( s)  
s s  n 2  n2 1   2 
Taking the inverse Laplace transform, which is left as an exercise for the student, produces
 

c(t )  1  e  n t cos  n 1   2 t  sin  n 1   2 t  (4.28a)
   1 2   
 
1  
c(t )  1  e  n t cos  n 1   2 t    (4.28b)
1  2   
 
1   
Where,   tan
 1 2 
 
A plot of this response appears in Figure 4.13 for various values of , plotted along a time axis
normalized to the natural frequency. We now see the relationship between the value of  and the
type of response obtained: The lower the value of , the more oscillatory the response. The natural
frequency is a time-axis scale factor and does not affect the nature of the response other than to
scale it in time.

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

We have defined two parameters associated with second-order systems,  and n. Other parameters
associated with the underdamped response are rise time (Tr), peak time (Tp), percent overshoot
(%OS), and settling time (Ts). These specifications are defined as follows (see also Figure 4.14):
1. Rise time, Tr: The time required for the waveform to go from 0.1 of the final value to 0.9 of
the final value.
2. Peak time, TP: The time required to reach the first, or maximum, peak.
3. Percent overshoot, %OS: The amount that the waveform overshoots the steady-state, or
final, value at the peak time, expressed as a percentage of the steady-state value.
4. Settling time, Ts: The time required for the transient’s damped oscillations to reach and stay
within 2% of the steady-state value.

Evaluation of Peak Time, TP

TP is found by differentiating c(t) in Eq. (4.28) and finding the first zero crossing after t=0. This task
is simplified by ‘‘differentiating’’ in the frequency domain by using Item 7 of Table 2.2. Assuming
zero initial conditions and using Eq. (4.26), we get
 n2
L c(t )  sC (s )  (4.29)
s 2  2 n s   n2
Completing squares in the denominator, we have
n
n 1   2
 n2 1 2
L c(t )   (4.30)
s   n 2  n2 1   2  s  n 2  n2 1   2 
Therefore,
n  
c(t )  e  n t sin   n 1   2 t  (4.31)
1 2  
Setting the derivative equal to zero yields

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

  1   2 t  n
 n  (4.32)
 
or
n
t (4.33)
2
n 1  
Each value of n yields the time for local maxima or minima. Letting n=0 yields t=0, the first point
on the curve in Figure 4.14 that has zero slope. The first peak, which occurs at the peak time, Tp, is
found by letting n=1 in Eq. (4.33):

TP  (4.34)
n 1   2

Evaluation of %OS
From Figure 4.14 the percent overshoot, %OS, is given by
c  cfinal
%OS  max  100 (4.35)
cfinal

The term cmax is found by evaluating c(t) at the peak time, c(TP).Using Eq. (4.34) for TP and
substituting into Eq. (4.28) yields
 
   / 1  2  

cmax  c(TP )  1  e   cos   sin  
 1  2 
 
 
   / 1  2 
cmax  c(TP )  1  e   (4.36)

For the unit step used for Eq. (4.28),

cfinal  1 (4.37)
Substituting Eqs. (4.36) and (4.37) into Eq. (4.35), we finally obtain
 
   / 1  2 
%OS  e    100 (4.38)
Notice that the percent overshoot is a function only of the damping ratio, .

Whereas Eq. (4.38) allows one to find %OS given , the inverse of the equation allows one to solve
for  given %OS. The inverse is given by

  ln%OS / 100 (4.38)
2
1
 2 2
 ln 2 %OS / 100
2
1
 
 2 2  1   2 ln 2 %OS / 100
ln 2 %OS / 100
2 
 2  ln 2 %OS / 100
 ln%OS / 100
 (4.39)
 2  ln 2 %OS / 100
Prepared by: Dr. Mohammad Abdul Mannan
 Chapter 4 Time Response

The derivation of Eq. (4.39) is left as an exercise for the student. Equation (4.38) (or, equivalently,
(4.39)) is plotted in Figure 4.15.

Evaluation of Ts
In order to find the settling time, we must find the time for which c(t) in Eq. (4.28) reaches and
stays within 2% of the steady-state value, cfinal. Using our definition, the settling time is the time it
takes for the amplitude of the decaying sinusoid in Eq. (4.28) to reach 0.02, or
1
e  n t  0.02 (4.40)
1 2
 
This equation is a conservative estimate, since we are assuming that cos  n 1   2 t     1 at
  
the settling time. Solving Eq. (4.40) for t, the settling time is
 ln 0.02 1   2 
Ts   
(4.41)
 n

You can verify that the numerator of Eq. (4.41) varies from 3.91 to 4.74 as  varies from 0 to 0.9.
Let us agree on an approximation for the settling time that will be used for all values of ; let it be
4
Ts  (4.42)
n

Evaluation of Tr
A precise analytical relationship between rise time and damping ratio, , cannot be found. However,
using a computer and Eq. (4.28), the rise time can be found. We first designate nt as the
Prepared by: Dr. Mohammad Abdul Mannan
 Chapter 4 Time Response

normalized time variable and select a value for . Using the computer, we solve for the values of
nt that yield c(t)=0.9 and c(t)=0.1. Subtracting the two values of nt yields the normalized rise
time, nTr, for that value of . Continuing in like fashion with other values of, we obtain the
results plotted in Figure 4.16.
Figure 4.16 can be approximated by the following polynomials:
 nTr  1.76 3  0.417 2  1.03  1
(maximum error less than 1/2% for 0<<0:9), and
=0.115( nTr)3-0.883(nTr)2+ 2:504( nTr)-1.738
(maximum error less than 5% for 0.1<<0.9).
The polynomials were obtained using MATLAB’s polyfit function.

FIGURE 4.16 Normalized rise time versus damping ratio for a second-order underdamped
response.

Example 4.5: Given the transfer function

400
G ( s) 
s 2  12s  400
find Tp, %OS, Ts, and Tr.

Solution: n and  are calculated as 10 and 0.75, respectively. Now substitute  and n into Eqs.
(4.34), (4.38), and (4.42) and find, respectively, that TP = 0.475 second, %OS = 2.838, and Ts=0.533
second.
Using the table in Figure 4.16, the normalized rise time is approximately 2.3 seconds. Dividing by
n yields Tr=0.23 second.

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

Evolution of Peak Time, Percent Over-shoot, and Settling Time Relate to the
Location of the Poles
The pole plot for a general, underdamped second-order system, is shown in Figure 4.17. We see
from the Pythagorean theorem that the radial distance from the origin to the pole is the natural
frequency, n, and the cos=.

Figure 4.17 Pole plot for an

underdamped second-order system.

Now, comparing Eqs. (4.34) and (4.42) with the pole location, we evaluate peak time and settling
time in terms of the pole location.
Thus,
 
TP   (4.44)
n 1   2 d
4 4
Ts   (4.45)
 n d
where d is the imaginary part of the pole and is called the damped frequency of oscillation,and d
is the magnitude of the real part of the pole and is the exponential damping frequency.

Equation (4.44) shows that TP is inversely proportional to the imaginary part of the pole. Since
horizontal lines on the s-plane are lines of constant imaginary value, they are also lines of constant
peak time.
Similarly, Eq. (4.45) tells us that settling time is inversely proportional to the real part of the pole.
Since vertical lines on the s-plane are lines of constant real value, they are also lines of constant
settling time.
Finally, since =cos, radial lines are lines of constant. Since percent overshoot is only a function
of , radial lines are thus lines of constant percent overshoot, %OS.
These concepts are depicted in Figure 4.18, where lines of constant TP, Ts, and %OS are labeled on
the s-plane.

Depicted in Figure 4.19(a) are the step responses as the poles are moved in a vertical direction,
keeping the real part the same. As the poles move in a vertical direction, the frequency increases,
but the envelope remains the same since the real part of the pole is not changing. The figure shows a
constant exponential envelope, even though the sinusoidal response is changing frequency. Since all
curves fit under the same exponential decay curve, the settling time is virtually the same for all
waveforms. As overshoot increases, the rise time decreases.
Let us move the poles to the right or left. Since the imaginary part is now constant, movement of the
poles yields the responses of Figure 4.19(b). Here the frequency is constant over the range of
variation of the real part. As the poles move to the left, the response damps out more rapidly, while
the frequency remains the same. Notice that the peak time is the same for all waveforms because the
imaginary part remains the same.

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

FIGURE 4.18 Lines of constant peak time, TP, settling time, Ts, and percent overshoot, %OS. Here:
Ts2<Ts1; TP2<TP1; %OS1< %OS2.

Figure 4.19 Step responses of second-order underdamped systems as poles move: (a) with constant
real part; (b) with constant imaginary part; (c) with constant damping ratio.

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

Moving the poles along a constant radial line yields the responses shown in Figure 4.19(c). Here the
percent overshoot remains the same. Notice also that the responses look exactly alike, except for
their speed. The farther the poles are from the origin, the more rapid the response.

Example 4.6: Given the pole plot shown in Figure

4.20, find ; n; TP; %OS, and Ts.
Solution: The damping ratio is given by
=cos=cos[arctan(7/3)]= 0.394. The natural
frequency, n, is the radial distance from the origin
to the pole, or n=(72+32)=7.616. The peak time is
 
TP    0.449 sec (4.44)
d 7
The percent overshoot is
 
   / 1 2 
%OS  e    100  26% (4.38)

The approximate settling time is

4 4
Ts    1.333 sec (4.45)
 n 3

Example 4.7: Given the system shown in Figure 4.21, find J and D to yield 20% overshoot and a
settling time of 2 seconds for a step input of torque T(t).

Solution: First, the transfer function for the system is

1/ J
G (s )  (4.49)
D K
s2  s 
J J

From the transfer function,

K
n  (4.50)
J
And
D
2 n  (4.51)
J

But, from the problem statement,

4
Ts  2  (4.52)
 n
Or  n  2 hence

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

D
2 n  4  (4.53)
J

Also, from Eqs. (4.50) and (4.52),

4 J
 2 (4.54)
2 n K

From Eq. (4.39), a 20% overshoot implies  = 0.456. Therefore, from Eq. (4.54),
J
 2  0.456 (4.55)
K

Hence,
J
 0.052 (4.56)
K

From the problem statement, K = 5 N-m/rad. Combining this value with Eqs. (4.53) and (4.56),
D=1.04 N-m-s/rad, and J = 0.26 kg-m2.

Exercise 4.5: Find ; n; Ts; TP; Tr, and %OS for a system whose transfer function is
361
G( s) 
s 2  16s  361

Solution:
Here, a = 16, b=361;  n  b  361  19 ;
a 16
a  2ξ n i.e. ξ    0.421
2 n 2  19
4 4
Ts    0.5 sec (4.45)
 n 0.421 19
 
TP    0.182 sec (4.44)
2 2
n 1   19  1  (0.421)
 
Tr  1.76 3  0.417 2  1.03  1 /  n  0.079 sec
   
   / 1  2    0.421 / 1 0.4212 
%OS  e    100  e    100  23.3% (4.38)

Damping frequency:  d   n 1   2  19  1  (0.421) 2  17.2342 rad

Exponential damping frequency:  d   n  0.421  19  8

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

System Response with Additional Poles

Let us now look at the conditions that would have to exist in order to approximate the behavior of a
three-pole system as that of a two-pole system. Consider a three-pole system with complex poles
and a third pole on the real axis.
Assuming that the complex poles are at n  j n 1   2 and the real pole is at -r, the step
response of the system can be determined from a partial-fraction expansion. Thus, the output
transform is
A B( s   n )  C d D
C ( s)    (4.57)
s 2 2 s r
( s   n )   d
Or, in the time domain,
C (t )  Au (t )  e  n t B cos  d t  C sin  d t   De  r t (4.58)

The component parts of c(t) are shown in Figure 4.23 for three cases of r. For Case I, r = ar1 and
is not much larger than n; for Case II, r = ar2 and is much larger than n ; and for Case III, r =
.

Figure 4.23 Component responses of a three-pole system: (a) pole plot; (b) component responses:
Nondominant pole is near dominant second-order pair (Case I), far from the pair (Case II), and at
infinity (Case III)

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

Let us direct our attention to Eq. (4.58) and Figure 4.23. If r>>n (Case II), the pure exponential
will die out much more rapidly than the second-order underdamped step response. If the pure
exponential term decays to an insignificant value at the time of the first overshoot, such parameters
as percent overshoot, settling time, and peak time will be generated by the second-order
underdamped step response component. Thus, the total response will approach that of a pure
second-order system (Case III).
If r is not much greater than n (Case I), the real pole’s transient response will not decay to
insignificance at the peak time or settling time generated by the second-order pair. In this case, the
exponential decay is significant, and the system cannot be represented as a second-order system.

The next question is, How much farther from the dominant poles does the third pole have to be for
its effect on the second-order response to be negligible? The answer of course depends on the
accuracy for which you are looking. However, this book assumes that the exponential decay is
negligible after five time constants. Thus, if the real pole is five times farther to the left than the
dominant poles, we assume that the system is represented by its dominant second-order pair of
poles.

What about the magnitude of the exponential decay? Can it be so large that its contribution at the
peak time is not negligible? We can show, through a partial-fraction expansion, that the residue of
the third pole, in a three-pole system with dominant second-order poles and no zeros, will actually
decrease in magnitude as the third pole is moved farther into the left half-plane.

Assume a step response, C(s), of a three-pole system:

bc A Bs  C D
C ( s)     (4.59)
 
s s 2  as  b s  c  s s 2  as  b s  c
where we assume that the nondominant pole is located at-c on the real axis and that the steady-state
response approaches unity. Evaluating the constants in the numerator of each term,
ca  c 2
A  1; B (4.60a)
c 2  b  ca
ca 2  c 2 a  bc b
C ; D (4.60b)
c 2  b  ca c 2  b  ca

As the nondominant pole approaches ; or c  ,

A  1; B  1; C   a; D0 (4.61)
Thus, for this example, D, the residue of the nondominant pole and its response, becomes zero as
the nondominant pole approaches infinity.
In this case, the control systems engineer can use the ‘‘five times’’ rule of thumb as a necessary but
not sufficient condition to increase the confidence in the second-order approximation during design,
but then simulate the completed design.

Example 4.8: Find the step response of each of the transfer functions shown in Eqs. (4.62) through
(4.64) and compare them.
24.542
T1 (s )  (4.62)
s 2  4 s  24.542
24.542
T2 ( s)  (4.63)
2

s  10 s  4s  24.542 
73.626
T3 ( s)  (4.64)

2
s  3 s  4s  24.542 
Prepared by: Dr. Mohammad Abdul Mannan
 Chapter 4 Time Response

Solution: The step response, Ci(s), for the transfer function, Ti(s), can be found by multiplying the
transfer function by 1/s, a step input, and using partial-fraction expansion followed by the inverse
Laplace transform to find the response, ci(t).
With the details left as an exercise for the student, the results are
C (t )  1  1.09e 2t cos4.532t  23.8 (4.65)
C (t )  1  0.29e 10t  1.189e 2t cos4.532t  53.34 (4.66)
C (t )  1  1.14e 3t  0.707e 2t cos4.532t  78.63 (4.67)
The three responses are plotted in Figure 4.24.

Figure 4.24 Step responses of system T1(s), system T2(s), and system T3(s).

Notice that c2(t), with its third pole at -10 and farthest from the dominant poles, is the better
approximation of c1(t), the pure second-order system response; c3(t), with a third pole close to the
dominant poles, yields the most error.

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

System Response With Zeros

Now that we have seen the effect of an additional pole, let us add a zero to the second-order system.
We saw that the zeros of a response affect the residue, or amplitude, of a response component but
do not affect the nature of the response—exponential, damped sinusoid, and so on.
Starting with a two-pole system with poles at (-1  j2:828), we consecutively add zeros at -3, -5,
and -10. The results, normalized to the steady-state value, are plotted in Figure 4.25. We can see
that the closer the zero is to the dominant poles, the greater its effect on the transient response. As
the zero moves away from the dominant poles, the response approaches that of the two-pole system.
This analysis can be reasoned via the partial-fraction expansion. If we assume a group of poles and
a zero far from the poles, the residue of each pole will be affected the same by the zero. Hence, the
relative amplitudes remain appreciably the same.

Figure 4.25 Effect of adding a zero to a two-pole system.

An interesting phenomenon occurs if zero (a) is negative, placing the zero in the right half-plane.
From Eq. (4.70) we see that the derivative term, which is typically positive initially, will be of
opposite sign from the scaled response term. Thus, if the derivative term, sC(s), is larger than the
scaled response, aC(s), the response will initially follow the derivative in the opposite direction
from the scaled response. The result for a second-order system is shown in Figure 4.26, where the
sign of the input was reversed to yield a positive steady-state value. Notice that the response begins
to turn toward the negative direction even though the final value is positive. A system that exhibits
this phenomenon is known as a nonminimum-phase system.

Figure 4.26 Step response of a nonminimum-phase system.

Prepared by: Dr. Mohammad Abdul Mannan
 Chapter 4 Time Response

Effects of Nonlinearities Upon Time Response

Effect of Saturation

(a)

(b)

Figure 4.29 (a) Simulink block diagram, (b) Effect of amplifier saturation on load angular velocity
response.

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

Effect of Deadzone

(a)

(b)

Figure 4.30 (a) Simulink block diagram, (b) Effect of deadzone on load angular displacement
response; (figure continues).

Prepared by: Dr. Mohammad Abdul Mannan

 Chapter 4 Time Response

Effect of backlash

(a)

(b)

Figure 4.31 (a) Simulink block diagram, (b) Effect of backlash on load angular displacement
response.

References
[1] Norman S. Nise, “Control Systems Engineering,” Sixth Edition, John Wiley and Sons, Inc,
2011.

Prepared by: Dr. Mohammad Abdul Mannan