Case Studies 363

SOLUTION:
a. The simplified block diagram for the system is shown on the front endpapers. The
steady-state error is given by
sR…s†
e…∞† ˆ lim sE…s† ˆ lim (7.107)
s®0 s®0 1 ‡ G…s†
From the block diagram, after pushing the potentiometer to the right past the summing
junction, the equivalent forward transfer function is

6:63K
G…s† ˆ (7.108)
s…s ‡ 1:71†…s ‡ 100†

To find the steady-state error for a step input, use R…s† ˆ 1=s along with
Eq. (7.108), and substitute these in Eq. (7.107). The result is e…∞† ˆ 0.
To find the steady-state error for a ramp input, use R…s† 1=s2 along with
Eq. (7.108), and substitute these in Eq. (7.107). The result is e…∞† ˆ 25:79=K.
To find the steady-state error for a parabolic input, use R…s† ˆ 1=s3 along with
Eq. (7.108), and substitute these in Eq. (7.107). The result is e…∞† ˆ ∞.
b. Since the system is Type 1, a 10% error in the steady state must refer to a ramp input.
This is the only input that yields a finite, nonzero error. Hence, for a unit ramp input,
1 …1:71†…100† 25:79
e…∞† ˆ 0:1 ˆ ˆ ˆ (7.109)
Kv 6:63K K
from which K ˆ 257:9. You should verify that the value of K is within the range of
gains that ensures system stability. In the antenna control case study in the last chapter,
the range of gain for stability was found to be 0 < K < 2623:29. Hence, the system is
stable for a gain of 257.9.

CHALLENGE: You are now given a problem to test your knowledge of this chapter’s
objectives: Referring to the antenna azimuth position control system shown on the front
endpapers, Configuration 2, do the following:

a. Find the steady-state errors in terms of gain, K, for step, ramp, and parabolic inputs.
b. Find the value of gain, K, to yield a 20% error in the steady state.

Video Laser Disc Recorder: Steady-State

Error Design via Gain
As a second case study, let us look at a video laser disc focusing system for recording.

PROBLEM: In order to record on a video laser disc, a 0.5 μm laser spot must be focused
on the recording medium to burn pits that represent the program material. The small laser
spot requires that the focusing lens be positioned to an accuracy of  0:1 μm. A model of
the feedback control system for the focusing lens is shown in Figure 7.22.

Power Motor &

Desired Detector Filter amplifier lens Actual
lens lens
position + K1(s + 800) K3 position
0.12 K2
_ (s + 40,000) s2
FIGURE 7.22 Video laser
disc recording: control system
for focusing write beam
364 Chapter 7 Steady-State Errors

The detector detects the distance between the focusing lens and the video disc by
measuring the degree of focus as shown in Figure 7.23(a). Laser light reflected from the
disc, D, is split by beam splitters B1 and B2 and focused behind aperture A. The remainder
is reflected by the mirror and focuses in front of aperture A. The amount of light of each
beam that passes through the aperture depends on how far the beam’s focal point is from
the aperture. Each side of the split photodiode, P, measures the intensity of each beam.
Thus, as the disc’s distance from the recording objective lens changes, so does the focal
point of each beam. As a result, the relative voltage detected by each part of the split
photodiode changes. When the beam is out of focus, one side of the photodiode outputs a
larger voltage. When the beam is in focus, the voltage outputs from both sides of the
photodiode are equal.
A simplified model for the detector is a straight line relating the differential voltage
output from the two elements to the distance of the laser disc from nominal focus. A
linearized plot of the detector input-output relationship is shown in Figure 7.23(b)
Mirror

A, pinhole P, split
aperture photodiode
B2, 50-50
beam splitter

Q–X/Y plate

D-Disc
L1, condensing
B1, polarizing
lens
beam splitter
L2, recording
objective lens
(a)

Differential
voltage

0.6

Distance from
–5 5 nominal focus
(μm)

–0.6

(b)

FIGURE 7.23 Video disc laser recording: a. focus detector optics;6 b. linearized transfer function
for focus detector6

6
Isailovic ́, J. Videodisc and Optical Memory Technologies, 1st Edition,  1985. Reprinted by permission of
Pearson Education, Inc., Upper Saddle River, NJ.
 Summary 365

(Isailovic ́, 1985). Assume that a warp on the disc yields a worst-case disturbance in the
focus of 10t 2 μm. Find the value of K1K2K3 in order to meet the focusing accuracy required
by the system.

SOLUTION: Since the system is Type 2, it can respond to parabolic inputs with finite
error. We can assume that the disturbance has the same effect as an input of 10t 2 μm.
The Laplace transform of 10t 2 is 20=s3 , or 20 units greater than the unit acceleration used
to derive the general equation of the error for a parabolic input. Thus, e…∞† ˆ 20=K a .
But K a ˆ lim s2 G…s†.
s®0
From Figure 7.22, K a ˆ 0:0024K 1 K 2 K 3 . Also, from the problem statement, the
error must be no greater than 0.1 μm. Hence, e…∞† ˆ 8333:33=K 1 K 2 K 3 ˆ 0:1. Thus,
K 1 K 2 K 3  83333:3, and the system is stable.

CHALLENGE: You are now given a problem to test your knowledge of this chapter’s
objectives: Given the video laser disc recording system whose block diagram is shown in
Figure 7.24, do the following:

a. If the focusing lens needs to be positioned to an accuracy of  0:005 μm, find the
value of K1K2K3 if the warp on the disc yields a worst-case disturbance in the focus
of 15t 2 μm.
b. Use the Routh-Hurwitz criterion to show that the system is stable when the conditions
of a. are met.
c. Use MATLAB to show that the system is stable when the conditions
of a. are met.

Power Motor &

Desired Actual
Detector Filter amplifier lens
lens lens
position + K1(s + 600) K3 position
0.2 K2
_ (s + 20,000) s2

FIGURE 7.24 Video laser disc recording focusing system

Summary
This chapter covered the analysis and design of feedback control systems for steady-state
errors. The steady-state errors studied resulted strictly from the system configuration. On the
basis of a system configuration and a group of selected test signals, namely steps, ramps,
and parabolas, we can analyze or design for the system’s steady-state error performance.
The greater the number of pure integrations a system has in the forward path, the higher the
degree of accuracy, assuming the system is stable.
The steady-state errors depend upon the type of test input. Applying the final value
theorem to stable systems, the steady-state error for unit step inputs is

1
e…∞† ˆ (7.110)
1 ‡ lim G…s†
s®0
366 Chapter 7 Steady-State Errors

The steady-state error for ramp inputs of unit velocity is

1
e…∞† ˆ (7.111)
lim sG…s†
s®0

and for parabolic inputs of unit acceleration, it is

1
e…∞† ˆ (7.112)
lim s2 G…s†
s®0

The terms taken to the limit in Eqs. (7.110) through (7.112) are called static error
constants. Beginning with Eq. (7.110), the terms in the denominator taken to the limit are
called the position constant, velocity constant, and acceleration constant, respectively.
The static error constants are the steady-state error specifications for control systems. By
specifying a static error constant, one is stating the number of pure integrations in the
forward path, the test signal used, and the expected steady-state error.
Another definition covered in this chapter was that of system type. The system type
is the number of pure integrations in the forward path, assuming a unity feedback
system. Increasing the system type decreases the steady-state error as long as the system
remains stable.
Since the steady-state error is, for the most part, inversely proportional to the static
error constant, the larger the static error constant, the smaller the steady-state error.
Increasing system gain increases the static error constant. Thus, in general, increasing
system gain decreases the steady-state error as long as the system remains stable.
Nonunity feedback systems were handled by deriving an equivalent unity feedback
system whose steady-state error characteristics followed all previous development. The
method was restricted to systems where input and output units are the same.
We also saw how feedback decreases a system’s steady-state error caused by
disturbances. With feedback, the effect of a disturbance can be reduced by system gain
adjustments.
Finally, for systems represented in state space, we calculated the steady-state error
using the final value theorem and input substitution methods.
In the next chapter, we will examine the root locus, a powerful tool for the analysis
and design of control systems.

Review Questions
1. Name two sources of steady-state errors.
2. A position control, tracking with a constant difference in velocity, would yield how much
position error in the steady state?
3. Name the test inputs used to evaluate steady-state error.
4. How many integrations in the forward path are required in order for there to be zero
steady-state error for each of the test inputs listed in Question 3?
5. Increasing system gain has what effect upon the steady-state error?
6. For a step input, the steady-state error is approximately the reciprocal of the static error
constant if what condition holds true?
7. What is the exact relationship between the static error constants and the steady-state
errors for ramp and parabolic inputs?
8. What information is contained in the specification K p ˆ 10;000?
9. Define system type.