Documente Academic
Documente Profesional
Documente Cultură
AMC
AMC
AMC
PAMPHLET
PAMPHLET
PAMPHLET
PAMPHLET
DARCOM-P 706-136
DARCOM-P 706-137
DARCOM-P 706-138
DARCOM-P 706-139
SERVOMECHANISMS
SECTION 1, THEORY
SECTION 2, MEASUREMENT AND SIGNAL CONVERTERS
SECTION 3, AMPLIFICATION
SECTION 4, POWER ELEMENTS AND SYSTEM DESIGN
APRIL 1965
AUGUST 1963
AMCP 706-136
AMC PAMPHLET
ENGINEERING DESIGN
HANDBOOK
SECTION 1, THEORY
HEADQUARTERS,
APRIL
1965
HEADQUARTERS
UNITED STATES ARMY MATERIEL COMMAND
WASHINGTON, D. C . 20315
30 A p r i l 1965
A M C P 706- 136, S e r v o m e c h a n i s m s , S e c t i o n 1 , T h e o r y , f o r m i n g
p a r t of t h e A r m y M a t e r i e l Command E n g i n e e r i n g D e s i g n Handbook
S e r i e s , is p u b l i s h e d f o r t h e i n f o r m a t i o n a n d guidance of all c o n c e r n ed.
( AMC RD)
F O R THE COMMANDER:
SELWYN D. SMITH, J R .
M a j o r G e n e r a l , USA
Chief of Staff
OFFICIAL:
A-
STAN EY J. SAWTCKI
C o l o n e l , GS
C h i e f , A d m i n i s t r a t i v e Office
LIST OF ILLUSTRATIONS
Fig .No .
Page
Title
1-1
1-1
1-2
1-2
1-3
......................
1-4
1-4
14
1-6
1-4
2-1
2-3
2-2
2-6
2-3
2-6
2-4
3-1
3-4
3-2
3-3
Cubic chart
..................................................................................
3-7
3-4
3-8
3-5
3-11
3-6
3-7
3-8
3-25
3-9
................................................
3-26
3-26
.....................
......................
..............
..................................
2-8
3-6
3-19
3-22
3-28
Title
Page
+ s2 + 1.4s
3-31
3-32
3-19 Trapezoidal approximation for R e [ W ( j w )3 ...__.............____.___
3-20 Impulse response from Floyd's method W ( s ) = 1.4s + 0.14/
(93 + 8' + 1.49 + 0.14) ............................................................
3-32
3-21 Gain-phase loci of constant real part of W = G/(1+ G) 3-33
_________
3-22 Elements of analog computers . . . _ _ _ _ _ . . _ _ _ _ . _ . _ _ _ . . . _ _._ _3-40
3-41
4-1
41
4-2
4-5
4-3
LOCUS
of [ I ( I + 1)2]-1
4-5
4-4
Distortion of locus of [ I ( I
4-6
Locus of-(-)1
s
4-6
l+s
1-9
for I = jp
..
+ 1)2]-1
........................................
4-6
('+7
4-6
1-9
.. _ .., _ ....__,
_ . _ __.._ . . _.._
_ .._._....
4-7
Root-loci plots _ _ _ _
4-8
49
4-6
_ _, ,
............
..........
..................................................................
4-10
+ 301 _ . _ .
6-2
6-3
6-4
Bandwidth measure from magnitude of error-to-input frequency response E (jo)/ R (jo) ................................................ 6-3
6-6
62
OF ILLUSTRATIONS
LIST
Fig. No.
5-6
5-8
Title
5-5
O n
on
[(j-)'+2Q*-++]-'
5-9
Page
jw
5-7
(cont)
..........................................................
5-6
5-7
+ 2 - + 1 3 ............................................................ 5-8
c0.2 j w + 1 3
Magnitude plots for G(jw) = K
j w [ ( i ;)
+
'
0.6 i - + 11 5-9
10
(0.2 jo + 1)
5-9
Angle plots for G ( i w ) = K
i w [( j
0.6 j 10 + 11
(0.2 j w + 1)
5-10
Gain-phase plot of G(iw) = 6.5
0
W"
On
[ ( j - )'
5-11
5-12
$)2+
5-13
(jw)
[( i :)*+
0.6i
- +1]
10
. . . . . .. . . . . . . . ..
5-11
5-15 Closed-loop response construction on the G-1 plane .. . . .... . . . ... 5-11
....................................
..............................
5-13
5-13
......
5-13
5-13
5-14
. . . . . . . . . . . , . . . . . . . . . . . . . . . . . . .. . . . . . . . .. . . . . .. . .. . . . .. .
5-15
5-15
5-16
5-16
Page
Title
.........................................
5-19
6-20
=K
i.[( j ; ) *
+ 0.6j; + 1 ]
5-30 Approximate closed-loop magnitude response of unity-feedback system, G ( j o ) = K [jo(jo 1) 3-1 ................................
5-22
5-25
6-1
62
6-3
6-4
6-5
6-6
6-7
6-8
69
...
6-1
6-4
....
6-4
6-5
......................
6-11
6-17
Page
Title
6-19
6 1 9 Inertia damper
6-20
........................................................................
6-20
621
......
6-21
6-22
6-22
7-2
7-3
7-4
Normalized curves yielding time-interval ratios of the transient response corresponding to combinations of various time
constants ............................................
..................................
7-3
7-6
Normalized curves yielding the time- interval between 10and 80-percent response of the transient corresponding to
combinations of various time constants .................
...... 7-4
7-6
7-6
7-7
7-6
7-8
7-6
7-9
7-6
7-10
......................
7-7
7-11 Dimensionless transient error-response curves of a secondorder servomechanism to a unit-ramp input ............................
7-17
7-12 Transient error-response curves of a second-order servomechanism to a unilistep input ................................................
7-18
xiv
Title
Page
7-13
7-14
7-15
7-16
7-17
7-18
Comparison of steady-state frequency response characteristics and transient response following a step function of
input as a function of ol/w, ....................................
................
7-19
7-21
7-24
7-42
8-1
8-3
8-2
8-3
8-4
8-6
....................
8-6
8-6
8-8
9-1
9-2
9-3
9-4
Train of unit impulses which represents the carrier A (t) .... 9-3
Action of sampledclamper ................................................
9-3
9-5
9-3
9-6
9-3
9-7
9-6
9-8
9-6
9-9
9-10
. . ....................................
criterion
9-12
. 9-1
.........................
..............
9-2
9-6
9-6
lication of Nyquist
.................................
9-7
. .
Page
Title
Fig No
..............................................................
9-11
9-11
10-2
10-3
10-4
10-5
..........................................
10-4
10-6
..........................................
10-4
10-7
10-8
Hysteresis nonlinearity
10-9
..........................................
.......................
.............................................................
10-3
10-5
10-5
......................................................
10-2
..................
10-5
10-7
10-7
10-14 Degree of stability variation with input amplitude for contactor servomechanism ............................................................
10-7
10-15 Phase portrait of linear second-order system with { = 0.5
....
10-9
10-10
..........................
10-10
....................
10-10
lO-lX
10-12
10-12
10-12
LIST
Fig . N o .
OF ILLUSTRATIONS
Page
Title
11-1
11-2
11-3
11-4
11-5
........................
11-6
11-6
11-6
11-7
11-8
11-9
Straight-line approximation of smooth nonlinear function no change in slope sigr, ........................................................... 11-9
........................
11-1
................................................
11-2
11-10 Resistance and shaft angle increments for straight-line approximations to nonlinear function - 3 regions of unchanging slope sign .......................................................................... 11-9
11-11 Resistance diagram and resultant circuit diagram for the
tapped nonlinear function of Fig. 11-10 ................................ 11-10
11-12 Equivalent resistance diagram showing a resistance branch
for each region of unchanging slope sign ............................... 11-10
11-13 Tapped potentiometer for generating a nonlinear function
11-14 Circuit for measuring equivalent noise resistance, ENR
..
11-11
......
11-13
11-17
....................
........
11-2C
..
11-21
..
11-23
......
11-24
...........
11-26
..............
11-28
Title
Page
11-37
. . _ . _ _ _ _ , 11-38
____._
11-29 Connections for booster amplifier to primary winding of resolver ............................................................................................ 11-39
11-30 Induction potentiometer
. . . . . ... . .. . . . . . . . . _ _. . . . . . .... .
11-41
.....................
11-31 Magnetomotive force and flux relationships of a. toroidwound rotary transformer ........................................................ 11-42
11-32 Voltage distribution in a toroid-wound rotary transformer Il-ArJ
11-33 Toroid-wound rotary transformers in back-to-back circuit 11-44
11-34 Winding schematics of typical microsyns
11-35 Linear variable differential transformer
11-45
..I_.._.._....._.._.....___...
.__..........__...
. . . , . . . . . . . ,..
11-52
._..
11-53
11-41 Pictcrial diagrams of the three basic types of gyro units ..._ 11-62
11-42 Plot for determining the value of the earth-rate component
along the local vertical at any latitude .................................... 11-67
11-43 Single-axis feedback system using an integrating gyro unit
to drive the gyro supporting member according to an angular velocity command signal . . . . . . .. . . . . . .. . . . .. . ... . . . . . .. . . .. . . . .. . . . . .. 11-72
,
..................................................
xiii
..._...._.._.._..,_
11-86
Title
Page
..............
11-87
....
11-88
11-89
.....................................................
11-90
................................................................
12-1
Modulator waveforms
12-2
Chopper elements
12-3
Chopper modulator
12-4
12-5
12-6
Half-wave connection
12-7
...........................................
12-5
12-8
12-5
12-9
............................................................
12-1
......................................................................
12-2
................................................................
12-2
12-5
.............................................................
12-6
.....................
12-6
..................................
12-7
...............................................................
12-8
.................................................................
12-9
.................................................................
12-9
....................................................................
12-9
..............................................................
...............................................
12-11
12-13
..................................
12-13
.................................................................
12-13
...........................................................
12-14
............................................
....
12-14
12-15
...................................................
12-15
..................................................
12-16
............................................
12-17
..........................................
12-17
..................
12-18
Page
CHAPTER 13 (cont)
13-8.8
13-8.9
13-8.10
13-8.11
13-187
13-8.12
13-191
13-8.13
13-191
13-8.14
13-191
13-8.15
13-8.16
13-195
13-8.17
13-195
13-8.18
13-196
LIST OF ILLUSTRATIONS
Fig No .
Page
Title
13-1
13-2
13-3
13-4
13-5
13-6
Triode amplifier
....................................
................
13-1
13-2
13-3
..............................................
13-3
.........................................................................
13-3
13-7
134
13-8
134
13-9
........................................
..................................
13-5
13-6
Page
Title
, ,,
13-'7
.............................
13-14 Series tube amplifier ..............................................
13-8
13-!9
..............
13-!3
13-17 Equivalent circuit of a cathode follower and its manipulation .............................................................................................. 13-10
13-18 White cathode follower
..............................................................
13-111
13-11
........................................................
..........................................
13-1.3
................................................................
13-1.3
......................................................................
13-1.4
........
13-111
13-16
............................
..............................................
13-17
13-17
....................................
13-17
..................................................
13-18
................................
13-18
....................
13-19
..............................................
13-20
..............................................
13-20
.............................................
13-20
............................
13-21
....
....................................................................
13-25
13-29
..............................
13-30
..............................................
13-31.
Title
Page
13-32
13-35
13-36
13-37
13-38
......................................
...................................
13-43
................................................
13-44
...................
13-45
.........................
............
13-48
13-51
13-53
............................
................... 13-54
........................
................................
13-52
............
....................................
........................
13-55
13-57
13-59
13-59
13-61
13-62
13-63
13-64
Title
Page
.. . .
13-69
13-74
............................................
. .. . . .
..................
13-75
13-86 Position control illustrating use of single-stage phase-sensitive relay amplifier _ _ _ _ . . _ _ _...
. . .....................................
.___.
13-93
13-87 Position control illustrating use of two-stage phase-sensitive relay amplifier ................................................................
13-98
13-88 Static characteristics of idealized relay
. .... ., , , , . . . .. . , . , . . ., , . . . . . .. . . .
13-99
13-89 Error response of contactor servo with large input change 13-100
13-90 Circuit for measuring relay pull-in and drop-out time .... .... 13-101
.__.._____.
13-91 Typical waveforms observed during relay test . _ _ _ _ _ . _ _ _ _ 13-102
13-103
............................................................
13-93 Nomograph and equations for use in calculating the component values for an arc suppression circuit _ _ _ _ . . _ _ _ _ . . . _ . _ . .13-104
_.._....
13-94 Three-way spool-valve amplifier ._ _ . ... .... ...... .. . . . .. .......
13-95 Single nozzle-baffle amplifier
..
13-107
13-108
................................................
13-108
. ... . .. .. ... . .. . . .
13-109
13-109
Page
Title
13-109
......
13-112
....................
13-114
Title
Page
S
I1
13-122 Plot of
CndX,
[XI
and 5.0
P
P,
vs -with-as
X,
.......................................................................................
13-125
............
13-130
............................................................
13-138
13-134 Four-way spool-valve amplifier and load ; equivalent hydraulic-circuit representation .......................................................... 13-139
13-135 Equivalent circuit representation of load having mass and
opposing force .......................................................................... 13-146
13-136 Equivalent circuit representation of load having mass and
spring and opposing force ....................................................... 13-147
13-137 Equivalent circuit representation of load having mass,
spring, viscous damping, and opposing force ......................... 13-148
13-138 Piston with unequal working areas
xx
........................................
13-148
Page
Title
..........
13-152
13-144 Block diagram of electric amplifier, torque motor, and firstand second-stage hydraulic amplifier ...................................... 13-154
13-145 Relative compressibility coefficient as a function of pressure
and amount of entrained gas .................................................... 13-157
13-146 Simplified circuit diagram for illustrative example .............. 13-161
13-147 Simplified block diagram for illustrative exa- Tple
................
13-162
13-164
13-165
._..
..............................
13-166
LIST
OF ILLUSTRATIONS (cont)
Fig .No .
Title
Pa.ge
....................................................................
............................................................
................
..............
13-176
13-181
13-182
13-183
....................................
13-1!)0
....................................
13-193
........................................................
13-194
..................
xxii
LIST OF ILLUSTRATIONS
Fig.N o .
Title
Page
14-1
14-2
14-2
14-4
14-3
14-5
14-4
14-6
14-5
14-6
14-7
14-8
xiii
14-8
LIST
OF ILLUSTRATIONS
Fig. No.
(cont)
Title
Page
14-9
14-10
14-11
14-12
14-22
14-23
14-25
14-26
14-29
14-30
14-20
14-22
14-31
14-31
14-32
14-35
14-44
14-45
xiv
Title
Pug e
14-45
14-34
14-45
14-35
14-45
14-46
14-48
14-49
14-50
14-38
14-40
14-51
14-44
15-1
15-2
15-2
Types of gears
15-4
15-3
15-14
15-4
Geared differentials
15-16
15-5
Differential lever
15-17
15-6
Typical linkages
15-18
15-7
15-20
15-8
Coupling types
15-23
15-9
15-25
15-25
15-28
15-28
15-30
15-14
15-31
15-31
15-16
15-31
Title
Page
16-1
16-16
16-2
16-17
16-3
16-17
16-4
16-18
16-5
16-19
16-6
16-21
16-7
16-22
16-8
16-22
16-9
16-23
16-24
7-1
17-2
17-3
17-4
17-5
17-6
Final block diagram of power control system for M-38 FireControl System
17-9
18-1
Gear pump
18-2
18-2
Vane pump
18-3
18-3
18-5
18-4
18-5
18-6
18-6
18-7
18-7
Pressure-regulating valve
18-7
18-8
18-8
18-9
Gravity accumulator
18-8
xvi
17-6
Title
............................................
18-9
........................................................................
18-11
................... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
18-12
Page
.........................................................
18-13
...............................................................................
18-13
..............................................................................
18-14
.................................................
................................................
18-16
......................................................
18-17
18-18
..
Positive
stop
..................................................
18-18
...........................................................................
18-19
...................................................
18-19
........................................................
18-19
...........................................
18-19
19-1
19-2
19-3
20-1
20-2
20-3
20-4
Cantilever spring
..................................................
............
...........................
19-9
19-12
........
20-6
..........
20-7
....................................................................
xvii
19-8
20-11
LIST OF TABLES
Table No .
Title
Page
3-1
3-14
3-2
3-18
3-3
3-30
5-4
5-1
6-12
6-1
6-6
6-2
7-1
746
8-1
8-4
8-2
8-3
m
i
8-6
9-4
9-13
LIST OF TABLES
Table N o .
Title
Page
11-1
11-2
11-3
11-4
11-5
11-6
11-7
11-8
11-9
......................................
11-19
11-33
........................
11-36
....................
11-40
........
11-47
...........
11-59
..........................................................................
xv
11-8
......................
....
................
11-80
11-83
LIST OF TABLES
Title
Table No.
13-1
Symbols
....................................................................................
13-1
13-2
13-46
13-3
13-49
13-4
13-5
13-6
13-7
13-8
13-9
............
13-94
..................
..........................
.....
................
13-101
13-105
....................................................
13-99
13-99
..................................
13-67
13-106
13-15 Four-way spool valve with appreciable radial clearance equivalent-source flow gain and conductance parameters .... 13-123
13-16 Transfer functions of four-way spool-valve amplifier and
load . . __
........................................................................ 13-140
13-17 Transfer functions of rotary amplifier and load ....
_ _ _13-142
_.
13-18 Circuit parameters for rotary hydraulic amplifier and load 13-144
13-19 Dynamic and static characteristics of commercially available electrohydraulic servo control valves ..............................
13-151
. .
13-156
....................................................
13-157
13-167
........
13-175
....................................................
......
....
13-188
13-191
_ _ _ _ _ _ 13-195
LIST OF TABLES
Table No.
Title
Page
14-1
14-21
14-2
14-24
14-3
14-36
14-4
14-5
14-6
14-40
14-7
14-41
14-8
14-41
14-9
14-42
14-42
14-1 1
14-52
14-53
14-13
15-1
15-2
15-3
15-4
15-9
15-5
15-10
15-6
15-11
15-7
15-12
15-8
Sheave diameters
15-21
15-9
15-26
15-6
15-28
Title
Page
15-13
15-14
16-1
Comparison of servomotors
17-1
7 9-1
19-3
19-2
19-4
19-3
19-5
19-4
19-6
19-5
19-7
19-6
20-1
20-2
20-3
20-5
20-4
20-8
20-5
20-10
20-6
20-7
20-14
20-8
20-16
........................................
15-33
16-7
xix
........................
17-10
20-2
TABLE OF CONTENTS
Paragmph
Page
CHAPTER 1
PROPERTIES OF FEEDBACK CONTROL SYSTEMS
1-1
1-2
1-3
1-4
1-2
CHAPTER 2
DYNAMIC RESPONSE
2-1
INTRODUCTION
2-2
LINEARIZATION
2-3
TRANSIENT RESPONSE
2-4
FREQUENCY RESPONSE
2-5
FORCED RESPONSE
2-6
STOCHASTIC INPUTS
,.
..
..
2-1
. . . . . . . . ..... ......................
....... .
2-1
.. . . . . . . . . . . . . . . . . . _ . . . .. . . . . . . . . . . . . . . . . . . .
2-5
... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2-7
. . . . . . . . . . . . . . . . . . . . . . . . . _. _. . . _ . _ _ . . . . . _ . . . . . . . . 2-8
.
..
...
. . . . . . . . . . . . . _... . . . . . . . . . . , , . . . . . . .. . . . . . , .
2-8
CHAPTER 3
METHODS O F DETERMINING DYNAMIC RESPONSE OF LINEAR SYSTEMS
3-1
T H E DIFFERENTIAL EQUATIONS . . . . . . .
3-2
...... . .
. . . . . . .. . . . . .
3-1
3-2.1
FACTORING
3-2.2
.....
3-2.3
First Order
3-2.4
Second Order
3-2.5
Third Order
3-2.6
Fourth Order
. .... .
. . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . .. .. . . . . . ... . . . . .. . , , , ,
......................
..................................
..... ................
.. . . ......, . , , .. . . , , , ...... , , . . .. , , , .. .
. ..................................................................
T H E CONVOLUTION INTEGRAL
3-4
3-4
3-5
3-5
3-5
. . . . . . . . . . . . . . ................................
3-3
3-2
.. _ . . . .. ... . . . . .... . . . . . .. .. . . .. .. .
. . , _... . . . . .. . . . . . . .
3-11
3-12
Paragraph
CHAPTER 3 (cont)
3-4.1
GENERAL
3-4.2
THEOREMS
3-4.3
3-4.4
FREQUENCY RESPONSE
3-5
..........................................................................
......................................................................
........
..............................................
3-12
3-12
3-13
3-17
3-5.1
BLOCK DIAGRAMS
3-5.2
SIGNALFLOW GRAPHS
..........................................................
3-17
3-25
...............................
3-6
3-7
3-8
3-9
41
INTRODUCTION
42
ROUTH CRITERION
43
NYQUIST CRITERION
44
ROOT-LOCUS METHOD
..................................................................
............................................................
........................................................
......................................................
4-1
4-2
4-4
4-7
CHAPTER 5
G A I N DETERMINATION
5-1
......
5-1
5-1.1
GENERAL
5-1.2
GAIN
5-1.3
VELOCITY CONSTANT
5-1.4
ACCELERATION CONSTANT
5-1.5
TORQUE CONSTANT
5-1.6
5-2
5-1.7
BANDWIDTH
5-2
5-1.8
..........................................................................
5-1
....................................................................................
5-1
............
..................................
............
............................
....................................................................
vi
5-1
5-1
5-2
5-3
Paragraph
CHAPTER 5 (cont)
5-2
5-3
5-2.1
GENERAL ..........................................................................
5-3
5-2.2
5-3
5-2.3
5-4
6-3
5-3.1
GENERAL ......................................
5-3.2
..........
5-3.3
5-5
5-3.4
5-7
5-3.5
510
5-10
5-4
5-4.1
5-4.2
5-4.3
5-4.4
5-5
5-11
(NICHOLS
..............
5-15
5-15
GENERAL
5-5.2
5-6
5-6.1
5-6.2
5-6.3
5-6.4
5-6.5
5-7
5-7.1
5-7.2
5-7.3
5-7.4
5-7.5
5-13
5-5.1
5-5.3
5-10
......................................
...................
5-15
5-16
5-18
5-20
5-20
5-20
5-20
5-21
5-21
5-23
5-23
..............
5-23
......
5-23
5-23
vii
5-24
Page
CHAPTER 6
COMPENSATION TECHNIQUES
6- 1
INTRODUCTION
6-2
..................................................................
6-2.1
GENERAL
6-2.2
LAG COMPENSATION
6-2.3
LEAD COMPENSATION
6-3
..........
..........................................................................
6-3.2
LAG COMPENSATION
6-3.3
LEAD COMPENSATION
6-2
6-2
................................................
6-4
62
.....- .................................
........
6-3.1
6-1
......
..........................................................................
....................................................
................................................
6-4
6-5
6-6
............
..............................
......................
6-6
6-6
6-7
6-7
6-10
6-15
6-16
6-6.1
D-C ELECTRIC
..................................................................
6-16
6-6.2
A-C ELECTRIC
......................................
6-17
6-6.3
6-19
6-6.4
6-20
6-6.5
6-21
CHAPTER 7
PERFORMANCE EVALUATION
7-1
7-1.1
GENERAL
7-1.2
7-1.3
7-1.4
..........................................................................
7-2
ERROR COEFFICIENTS
7-3
PERFORMANCE INDICES
....................................................
viii
................................................
7-1
7-44
7-45
Paragraph
CHAPTER 8
8-1
CRITERIA O F PERFORMANCE
8-2
......................................
8-2.1
TRANSIENT INPUTS
8-2.2
..............
......
8-1
8-2
8-6
8-3
8-4
........
8-11
CHAPTER 9
SAMPLE-DATA SYSTEMS
9-1
GENERAL THEORY
9-2
............................................................
9-2.1
T H E z TRANSFORM
9-2.2
T H E w TRANSFORM
9-3
.......................................................
..................................
OPERATIONAL METHODS
9-1
9-4
9-5
..............................................
9-6
..........................................................................
9-6
9-3.1
GENERAL
9-3.2
9-3.3
9-4
DESIGN TECHNIQUES
9-5
PERFORMANCE EVALUATION
............................
....................................
9-9
9-10
CHAPTER 10
NONLINEAR SYSTEMS
10-1
INTRODUCTION
..................................................................
10-1
10-2
10-1
103
PHASE-PLANE PROCEDURES
10-8
10-4
LIMITATIONS, COMPENSATION, A N D O T H E R
10-13
METHODS ............................................................................
IX
........................................
TABLE OF CONTENTS
Page
Paragraph
CHAPTER 11
SENSING ELEMENTS
11-1
INTRODUCTION
11-2
POTENTIOMETERS
11-2.1
...............................................................
..........................................................
11-1
........................................................................
11-1
11-2.2
11-2.3
Types of Potentiometers
11-2.4
Principle of Operation
11-2.5
Use
11-2.6
Construction Features
...............................................
....................................................
..................................................................................
....................................................
LINEAR POTENTIOMETERS
11-2
11-3
11-3
11-4
.....................................................
11-4
Types of Linearity
11-2.9
..............................................
11-2.10
NONLINEAR POTENTIOMETERS
11-2.11
APPLICATION FACTORS
.............................
11-6
11-7
..............................................
11-11
................................................................................
11-11
11-2.12
Noise
11-2.13
Power Rating
11-2.14
Environmental Effects
11-2.15
Life
11-2.16
Mechanical Loading
...........................................................
11-13
...............................................
11-13
..................................................................................
11-14
. .
Lubrication
.....................................................
......................................................................
ROTARY TRANSFORMERS
................
11-3.1
GENERAL DESCRIPTION
................
11-3.2
GENERAL CLASSIFICATIONS
11-3
11-2
.....................................
11-2.8
11-2.17
11-1
.......................
..
Definition
11-2.7
11-1
...................
11-14
11-15
11-15
....................................
11-15
11-3.3
............................................
11-15
11-3.4
............................................
11-15
11-3.5
11-3.6
11-3.7
SYNCHROS
................................
11-15
........................................
11-16
......................................................................
11-16
11-3.8
Stator Construction
.......................................................
11-16
11-3.9
Rotor Construction
.....................................................
11-16
Paragmp h
CHAPTER 11 (cont)
11-3.10
Synchro Supply
11-3.11
Nomenclature
11-3.12
11-3.13
Transmitter Characteristics
11-3.14
Receiver Characteristics
11-3.15
Transformer Characteristics
11-3.16
11-3.17
11-3.18
Synchro Capacitors
11-3.19
11-3.20
Zeroing
11-3.21
11-3.22
11-3.23
11-3.24
Torque Relationships
11-3.25
Torque gradient
11-3.26
Performance prediction
........................................................
.......................................................
Synchro Accuracy
11-3.28
Static errors
11-3.29
Dynamic errors
11-3.31
11-18
11-19
11-20
11-22
.................................................
11-25
................................
11-25
11-25
....................
11-26
...................................
11-26
............................
11-26
...................................................
11-27
.......................................................
11-27
.............................................
....................................................
.........................................................
....................................................
11-27
11-29
11-29
11-29
11-31
............................................
11-37
..............................................................
11-37
............................................................
11-37
Design Principles
11-3.34
Applications
11-3.35
Booster Amplifiers
.........................................
..........................................................
11-39
..............................
11-39
..................................................................
11-39
INDUCTION POTENTIOMETERS
................................
11-41
...................................................................
11-41
11-3.39
Construction
11-3.40
Characteristics
11-3.41
.................
11-18
11-22
INDUCTION RESOLVERS
11-3.33
11-3.38
................................
11-16
.......................
Military Specifications
Basic Operation
11-3.37
....................................
................................................................
11-3.32
11-3.36
.................................
. . . . . . . . . . . . . . . . . . . ....................................................
11-3.27
11-3.30
..............................
11-16
...............................................................
........
11-41
11-41
Page
CHAPTER 11 (cont)
11-3.42
Construction
11-3.43
Principles of Operation
11-3.44
Electrical Characteristics
11-3.45
MICROSYNS
....................................................................
...........
....................................................................
Principles of Operation
11-3.47
General Classification
11-5.48
11-41
..... 11-43
..............................
11.3. 46
11-41
11-44
..................................................
11-44
....................................................
11-44
11-48
.................
11-3.49
Double-winding type
.................................................
11-45
11-3.50
Single-winding type
....................................................
11-45
11-3.51
11-3.52
11-4
............................................
11-46
11-4.1
GENERAL DESCRIPTION
11-4.2
DESIGN CHARACTERISTICS
...........................................
.....................................
11-4.3
Sensitivity Rating
11-4.4
11-4.5
Phase Angle
11-4.6
Output Impedance
11-4.7
Loading
11-4.8
Construction
11-4.9
APPLICATION
11-5
..........................
11-46
........................................................
11-47
11-48
11-48
................................
11-49
...................................................................
11-49
.............
........
..........................................................................
11-49
11-49
....................................................................
11-50
................................................................
11-50
TACHOMETER GENERATORS
........................................
11-51
............................................
11-51
11-5.1
GENERAL DESCRIPTION
11-5.2
11-5.3
Theory of Operation
11-5.4
........ 11-51
......................................................
11-51
11-5.5
11-5.6
11-5.7
11-5.8
11-5.9
3peed Error
-1
....................................................................
........ 11-54
11-54
11-56
Page
CHAPTER 11 (cont)
11-5.10
Frequency Errors
11-5.11
Harmonics
11-5.12
11-5.13
11-5.14
Wobble
11-5.15
Acceleration Error
11-5.16
.........................................................
11-55
.............................................................
11-55
................
...........................
11-56
.........................................................
11-56
..............................................
11-5.17
Comme;.cial Units
11-5.18
Selection
.....................................................................
11-5.20
11-5.21
.....................
.................................
...............
......................................................
11-6.1
INTRODUCTION
11-6.2
......................................................
11-6.3
Description
11-6.4
11-6.5
..................
.......................................................
11-6.6
Component Performance
11-6.7
Drift Rate
.................................
11-58
11-60
11-60
11-60
11-61
11-61
11-61
11-66
11-6.10
.................................
11-68
.........................................
11-68
...........
...........................................
11-6.12
Geometrical stabilization
11-6.13
Space integration
11-6.14
11-68
11-69
..........................................
11-69
...................................................
11-69
..................
11-69
...........................
11-69
...................................................................
11-69
11-6.17
SINGLE-DEGREE-OF-FREEDOMGYRO UNITS
11-6.18
11-6.19
11-58
..........................................
TYPICAL APPLICATIONS
Nutation
11-58
11-66
11-6.9
11-6.16
11-57
.........................................
11-6.15
11-57
11-66
.......
Typical Applications
11-57
...............................
11-6.8
11-6.11
.....
................................................
11-5.19
11-6
11-55
..............................
......
...........................
11-70
11-70
11-70
Page
CHAPTER 11 (cont)
11-6.20
11-6.21
Design Considerations
....................................................
11-6.22
Operating Arrangement
11-6.23
..............
11-71
11-71
................................................
11-71
Modes of operation
......................................................
11-71
11-6.24
Ideal performance
.....................................................
11-73
11-625
11-6.26
APPLICATION FACTORS
........................
11-73
..............................................
11-73
11-6.27
11-6.28
11-6.29
Impedance Levels
11-6.30
....................................
11-73
.......................................
11-74
...........................................................
11-75
.............................
11-6.31
11-6.32
11-7
ANALOG-TO-DIGITAL CONVERTERS
11-81
..............................................................
11-81
INTRODUCTION
11-7.2
NUMERICAL REPRESENTATION
..............................
11-81
.............................................................
11-81
11-7.3
Binary Numbers
11-7.4
11-7.6
11-7.7
11-7.8
11-7.9
11-7.10
11-7.11
11-7.12
11-8
11-8.1
11-76
..........................
11-7.1
11-7.5
..................
11-75
.................................
..........
11-81
11-83
11-83
11-85
11-86
11-83
11-8.2
DIFFERENTIALS
11-8.3
11-8.4
............................................................
ix
11-88
Page
CHAPTER 12
SIGNAL CONVERTERS
12-1
12-1.1
INTRODUCTION
TYPES
..................................................................
12-1
...............................................................................
12-1
12-1.2
Modulators
12-1.3
Demodulators
12-1.4
Form of Modulation
12-2
12-2.1
MODULATORS
, .......................................
12-1
. . . . . . . . .....................................................
12-1
....................
.....................................................
12-2
..................................................................
12-2
CHOPPER MODULATORS
...........................................
12-2
12-2.2
. .
Description
12-2.3
Characteristics
...............................................................
12-3
12-2.4
Phase shift 4
...............................................................
12-3
12-2.5
Drive-voltage frequency
12-2.6
Drive voltage
12-2.7
Contact rating
...........................................................
12-4
12-2.8
Chopper noise
...........................................................
12-4
12-2.9
Temperature effects
12-2.10
12-2.11
12-2.12
Packaging
................................................................
...........................................
.............................................................
12-2
12-3
12-3
...................................................
12-4
.......................................................................
12-4
............................................................
12-4
. .
Practical Circuits
MAGNETIC MODULATORS
12-2.13
Principle of Operation
12-2.14
Operating Circuits
..........................................
12-4
....................................................
12-4
..........................................................
12-7
....................................................................
12-7
12-2.15
...
Sensitivity
12-2.16
Harmonic Attenuation
12-2.17
123.18
Core Material
12-2.19
12-2.20
...............................................................
ELECTRONIC MODULATORS
Operation
Bridge modulator
12-2.22
Typical curve
12-3
12-3.1
..
Stability
.................................
12-7
12-7
12-7
.........................................................
12-2.21
12-2.23
.................................................
........................................................
............................................................
12-10
12-10
......................................................
ELECTRONIC DEMODULATORS
DIODE DEMODULATORS
X
....................................
..............
12-11
.. 12-11
Page
CHAPTER 12 (cont)
12-3.2
Operation
12-3.3
Ripple
12-3.4
Unbalance
..........................................................................
12-11
............................................................................
12-11
........................................................................
12-3.5
Transfer function
12-3.6
Output stability
Full-Wave Demodulator
12-3.8
TRIODE DEMODULATORS
Keyed Demodulator
12-3.10
Operation
12-3.11
Transfer function
12-3.12
12-3.13
12-3.14
12-3.15
12-4
...................................
............................................
........................................................
....................................................
Dual Use
12-12
12-12
12-14
........................................
12-14
..........................................................................
Output filtering
12-12
...........................
..................
CHOPPER DEMODULATORS
Operation
12-12
.............................................
12-3.7
12-3.9
.................
..........
12-12
............................................................
..........................................................................
DIGITAL-TO-ANALOG CONVERSION
12-14
12-15
12-15
.........
12-4.1
GENERAL
..........................................................................
12-16
12-4.2
Networks
..........................................................................
12-16
12-4.3
Operation
......................................................................
12-4.4
Accuracy
..................................................
12-45
Servomechanism
..............................................................
xi
12-16
12-18
TABLE OF CONTENTS
Paragraph
Page
CHAPTER 13
AMPLIFIERS USED IN CONTROLLERS
13-1
13-1.1
13-1.2
ELECTRONIC AMPLIFIERS
VACUUM TUBES
Diodes
...................................
13-1.4
Diodes as rectifiers
13-1.5
13-1
..............................................
13-2
......................................................
13-2
...................................................................
13-3
13-1.6
Plate characteristics
13-1.7
Graphical analysis
13-1.8
Linear approximations
13-1.9
Region of operation
13-1.10
13-1.11
13-1.12
13-1.13
Phase shift
..........................
13-3
........
13-3
......................................................
..............................................
13-4
....................................................
13-4
13-4
......................................
13-5
........................
13-5
..............................
13-5
..................................................................
................................
13-5
................................................
13-6
13-1.15
Plate characteristics
13.1.16
13-1.17
Graphical analysis
......
Interelectrode Capacitance
13-1.19
Tube Specifications
13-6
................................
13-6
..
............................
13-1.18
13-1.20
13-1
.........
..............................................................................
13-1.3
13-1.14
13-1
..............................................
............................................
13-6
........................................................
13-7
13-1.21
13-1.22
Simple Amplifier
13-1.23
13-1.24
Cascode Amplifier
13-1.25
Cathode Followers
13-1.26
13-1.27
Differential Amplifiers
......................................
13-7
............................................................
13-7
............................
............................
....................
13-8
13-8
13-9
............................
................
......................
V
.................
13-11
...................
13-12
Page
CHAPTER 13 (cont)
13-1.28
13-1.29
13-13
. . . . . _ _ . . _ _ _ _ _ . _ _ . _ . . _13-13
_......,____
13-1.30
13-1.31
. . . . . _ _ . . _ _ _ _ _ . _ _ . _ . . _13-13
_......,____
. . . . . .. .. . .. . . . . . . . .. . . . . . , . . .. ... . . . , ., . .
13-14
13-1.32
13-1.33
,,
. . . . . _ . . . . . . _ _ . . _ . . _ _13-15
.._..
13-1.34
13-1.35
13-1.36
Direct-Coupled Amplifiers
. . _ . _ _ _ . _ _ _ _ . _ _ . _ _ . 13-15
.__....__....
... .
13-1.37
13-1.38
13-1.39
Bridge circuits
13-1.40
13-1.41
13-1.42
............................................................
. , . .. . . .
13-19
13-21
13-23
13-1.43
13-1.44
Disadvantages ..................................................................
13-1.45
,,
, ,
, ,
13-25
13-1.46
Reliability
13-1.47
Construction
13-1.48
13-1.49
13-1.50
13-27
13-1.51
13-28
........................................................................
13-26
_ _ . _ _ . . _ _ _ . . _ _ _ . _ .13-27
.13-27
....._,..._.
. .. . .. . . . . . . . .
13-1.52
13-28
13-1.53
13-28
13-1.54
13-30
13-1.55
13-30
13-1.56
13-31
13-32
13-1.57
13-1.58
13-1.59
. .
................
Inductive loads .... ..............................
Resistive loads
vr
13-33
13-33
TABLE OF CONTENTS
Paragraph
Page
CHAPTER 13 (cont)
13-1.60
13-1.61
13-1.62
13-1.63
....................................................
................................
13-35
13-35
13-1.64
Types of Rectifiers
13-1.65
Power-Supply Circuits
13-1.66
......................................
13-36
13-1.67
......................................
13-37
..............................................
13-38
........................................................
13-38
13-2
13-2.1
..........................................................
..........................
TRANSISTOR AMPLIFIERS
BASIC PRINCIPLES
....................
13-35
13-35
13-2.2
13-2.3
Transistor Operation
......................................................
13-2.4
13-2.5
High-frequency operation
13-2.6
13-2.7
High-power applications
13-2.8
Switching applications
13-2.9
Summary
13-2.10
........
..........................................
....................................
............................................
..............................................
..................................................
13-39
13-39
13-39
13-39
13-40
13-40
13-40
13-2.11
Electron Current
13-2.12
Hole Current
13-2.13
Material Types
13-2.14
Junctions
..........................................................................
13-41
13-2.15
13-41
13-42
13-2.16
13-2.17
13-2.18
13-2.19
13-2.20
13-2.21
............................................................
13-40
13-40
....................................................................
13-40
................................................................
13-41
13-42
13-42
..............................................
13-44
Page
CHAPTER 13 (cont)
13-2.22
13-2.23
13-2.24
13-2.25
13-2.26
13-2.27
13-2.28
13-2.29
13-2.30
13-2.31
13-2.32
13-2.33
Grounded-Base Amplifier
13-2.34
Grounded-Collector Amplifier
13-2.35
13-2.36
13-2.37
13-50
13-50
13-51
......
13-51
13-51
13-52
..............................................................................
13-52
13-2.40
Direct-Coupled Amplifiers
13-2.41
State of the A r t
......................
13-54
............................................
13-54
..............................................................
13-55
..............................
13-55
...............................................
13-58
BASIC CONSIDERATIONS
13-58
...........................................
13-3.2
13-3.3
13-3.4
13-3.5
Temperature Limitations
13-3.6
Design Difficulties
13-3.8
13-50
............................................
13-3.7
13-50
13-50
13-50
................................................
13-2.39
13-3.1
13-47
13-47
13-50
.......................................
13-2.38
13-3
13-46
..............................................
. .
Biasing
13-2.42
13-44
13-46
..
.................................
...........
13-58
13-58
13-58
............................................
13-58
.........................................................
13-58
PRINCIPLES O F OPERATION
.....................................
............................
13-3.9
13-3.10
13-3.11
Reset period
13-59
....................
13-59
.............................
13-60
...............................................................
viii
13-59
13-60
Page
CHAPTER 13 (cont)
13-3.1-2
13-3.13
Control characteristics
...........................
13-61
13-3.14
13-61
13-3.15
13-62
13-3.16
13-3.17
13-3.18
13-3.19
13-3.20
13-3.21
13-3.22
13-3.23
.......... 13-60
.. 13-62
........................................
..................................
13-62
13-63
13-63
....................
13-63
13-3.24
Dynamic response
13-3.25
Accuracy of prediction
13-3.26
Analytical Representation
13-64
13-65
......................................................
13-65
..........................
13-65
..................
13-3.27
13-3.28
13-3.29
Conclusions
13-3.30
13-62
13-65
..................
..................................................................
13-3.31
13-3.32
Control Characteristics
13-3.33
Figure of Merit
......
....
13-66
13-66
13-66
..............................
13-66
..................................
13-66
................
13-66
13-3.34
13-68
13-3.35
13-69
13-3.36
Largest factor
13-70
13-3.37
Effects of dimensions
13-3.38
Leakage effects
13-3.39
Temperature effects
13-3.40
13-3.41
13-3.42
..............................................................
................................................
13-70
............................................................
13-70
..........................
......................
......................................
.................................
IX
......
13-70
13-70
13-70
13-71
TABLE
OF CONTENTS
(cont)
Page
Paragraph
CHAPTER 13 (cont)
13-3.43
13-3.44
13-3.45
Rectifiers
13-71
....................
13-71
..................................................................
13-72
13-3.46
13-3.47
13-3.48
13-72
13-3.49
13-72
13-3.50
13-3.51
Specifications
13-3.52
Approach to Design
13-4
13-4.1
13-72
......................
..................
..............
13-72
................................
13-72
.......
13-73
........................................................
13-73
................................
13-72
........
13-73
13-4.2
Basic Principles
13-4.3
13-4.4
Basic Features
13-4.5
13-73
..............................................................
13-74
................................................................
13-74
13-4.6
Excitation
13-4.7
Single-shunt-winding machine
13-4.8
Armature reaction
13-4.9
13-76)
13-4.10
13-76;
13-74:
....................................................................
..................................
13-76,
......................
13-76
........
........................................
13-4.11
13-4.12
13-4.13
13-4.14
13-4.15
13-4.16
Results of saturation
13-4.17
13-4.18
13-4.19
13-4.20
13-4.21
............
......................
................................
Fundamental Requirements
X
13-77
13-78,
....
13-77
......
..........................................
13-79
13-79
13-79
13-791
13-80
13-801
Page
CHAPTER 13 (cont)
13-4.22
Stored energy
13-4.23
13-4.24
Dissipated power
13-4.25
13-4.26
Power amplification
13-4.27
13-4.28
13-4.29
13-4.30
13-4.31
..............................................................
.........
..................
......................
..............................
RELAY AMPLIFIERS
................
....................................................
13-83
13-86
13-87
.........................................................
13-87
......................................................................
13-87
13-5.1
DEFINITION
13-5.2
13-5.3
RELAY CHARACTERISTICS
13-5.4
Description of Operation
13-5.5
Usage of Relays
..................
.......................................
13-87
13-88
13-88
...................
..............................................................
13-88
13-5.7
Typical Amplifiers
13-5.8
Operation
..........................................................
.....................................................
.............13-90
Servomechanism Applications
13-5.11
......................................
............................................................
13-88
................ 13-89
13-5.10
13-5.13
13-82
13-5.12
13-82
13-4.33
13-5.9
13-82
13-5.6
..............
.............................................
13-4.32
13-5
13-82
13-90
13-90
13-90
................................
13-90
................................................................
13-90
13-5.14
Typical types
13-5.15
Page
CHAPTER 13 (cont)
13-5.16
13-5.17
....................................
13-91
....................................................
13-93
13-5.18
Purpose
13-5.19
13-5.20
............................................................................
13-5.21
Idealized Relay
13-5.22
Sensitivity
13-5.23
Contact Rating
13-5.24
13-5.25
...
........................................................................
13-98
................................................................
General Considerations
Relay Operating Time
Relay Life
13-5.33
False Operation
13-100
13-100
13-100
13-100
13-101
..................
13-101
...................................................
13- 02
......................................................................
13-102
.............................................................
HYDRAULIC AMPLIFIERS
INTRODUCTION
.............................................
13-6.3
Characteristics
13-105
13-105
....................................................
13-105
................................................................
13-103
..........................
........................
13-6.2
13-6.5
13-99
13-5.32
13-6.4
..............................................
....................................................
13-6.1
..................................................
PARAMETER MEASUREMENT
13-5.31
13-6
..........
...............................................................
13-5.27
13-5.30
13-98
13-98
13-5.29
................
13-98
................................................................
13-5.26
13-5.28
....................................
13-93
......
............................................................
13-109
13-110
13-110
13-6.6
13-6.7
13-6.8
13-6.9
Page
CHAPTER 13 (cont)
13-6.10
13-6.11
13-6.12
13-6.13
Plate-Valve Type
............................................................
13-128
13-6.14
13-129
13-6.15
13-6.16
13-6.17
13-6.18
13-6.19
13-6.20
13-6.21
............
................................................................
......................
13-6.22
13-6.23
Block Diagram
13-158
13-136
...........
13-130
..................
................................................................
13-136
13-137
13-138
13-148
13-6.24
13-6.25
13-6.26
13-6.27
HYDRAULIC-CIRCUIT ELEMENTS
13-6.28
ILLUSTRATIVE EXAMPLE
............................
......................................
13-155
13-158
13-6.29
13-6.30
....................................
13-158
13-6.31
....................................
13-159
13-6.32
13-6.33
13-7
..............
................................
13-158
13-159
....................
13-160
..............................................
13-160
..............................................................
13-162
PNEUMATIC AMPLIFIERS
13-7.1
INTRODUCTION
13-7.2
PNEUMATIC VALVES
....................................................
xiii
13-163
13-7.3
13-7.4
Orifice Flow
13-7.5
Nondimensional Flow
13-7.6
Equivalent Source
13-7.7
13-7.8
..................
13-170
13-7.9
..................
13-1'70
...................................................................
13-163
.............................
13-165
..........................................................
13-168
........
................................
13-169
13-7.10
13-7.11
13-7.12
Resistance
13-7.13
Capacitance
13-7.14
Inertance
13-7.15
Time constant
..................................................................
13-1'73
..................................................................
13-1'75
......................................................................
13-1'75
..............................................................
13-1'76
13-7.16
13-7.17
13-7.18
13-7.19
13-7.20
13-7.21
Advantages
13-7.22
Disadvantages
13-8
13-8.1
13-8.2
................................
13-180
13-180
..................................................................
13-181
13-8.4
Cone-and-disc-amplifier
13-8.7
13-177
13-181
Ball-disc integrator
13-8.6
...................
13-176
..........................................
13-8.3
13-8.5
....................
..................................................................
MECHANICAL AMPLIFIERS
BASIC TYPES
......................
............
....................................................
..............................................
..........
13-181
13-182
13-182
13-182
..................................
13-184
Page
CHAPTER 13 (cont)
13-8.8
13-8.9
13-8.10
13-8.11
13-187
13-8.12
13-191
13-8.13
13-191
13-8.14
13-191
13-8.15
13-8.16
13-195
13-8.17
13-195
13-8.18
13-196
xv
TABLE OF CONTENTS
Paragraph
Page
CHAPTER 14
POWER ELEMENTS USED IN CONTROLLERS
14-1
INTRODUCTION
14-1
14-2
DIRECT-CURRENT MOTORS
14-1
14-2.1
14-1
14-2.2
14-7
14-2.3
14-7
14-2.4
14-10
14-2.5
14-10
14-2.6
14-10
14-2.7
14-11
14-2.8
14-12
14-2.9
14-12
14-2.10
14-13
14-2.11
14-15
14-2.12
14-16
14-2.13
14-17
14-2.14
14-17
14-2.15
Armature Resistance
14-18
14-2.16
Armature Inductance
14-18
14-2.17
Field Resistance
14-18
14-2.18
Field Inductance
14-18
14-2.19
Motor Constant
14-19
14-2.20
Armature Inertia
14-19
14-2.21
Viscous Damping
14-20
14-2.22
14-20
14-2.23
14-20
14-2.24
14-20
14-2.25
Static Characteristics
14-20
14-2.26
Dynamic Characteristics
14-22
14-2.27
Modified Servomotors
14-22
14-2.28
14-3
14-23
14-25
Page
CHAPTER 14 (cont)
14-3.1
14-3.2
14-25
14-3.3
14-29
14-3.4
14-3.5
14-30
14-31
14-3.6
14-31
14-3.7
Relay Servos
14-31
14-3.8
14-3.9
14-31
14-3.10
Alternatives
14-33
14-3.11
14-34
14-3.12
Nondimensional damping
14-34
14-3.13
Motor equations
14-35
14-3.14
FIGURE O F MERIT
14-37
14-3.15
14-37
14-4
HYDRAULIC MOTORS
14-37
14-4.1
INTRODUCTION
14-4.2
14-4.3
14-4.4
14-4.5
14-4.6
14-4.7
14-37
PNEUMATIC MOTORS
14-44
14-5.1
PRINCIPAL T Y P E S
14-44
14-5.2
STATIC BEHAVIOR
14-45
14-5.3
DYNAMIC BEHAVIOR
14-45
14.5
vi
Page
CHAPTER 14 (cont)
14-5.4
14.6
DIFFICULTIES ENCOUNTERED
MATIC MOTORS
WITH
PNEU-
MAGNETIC-PARTICLE CLUTCHES
14-46
14-47
14-6.1
DESCRIPTION
14-47
14-6.2
METHODS O F USE
14-47
14-6.3
ADVANTAGES
14-47
14-6.4
DISADVANTAGES
14-47
14-6.5
STATIC BEHAVIOR
14-50
14-6.6
DYNAMIC BEHAVIOR
14-51
14-6.7
14-6.8
Voltage-to-Position Relationship
LIFE EXPECTANCY
14-51
14-51
CHAPTER 15
MECHANICAL AUXILIARIES USED IN CONTROLLERS
GEAR TRAINS
15-1
15-1.1
PURPOSE
15-1
15-1.2
DEFINITIONS
15-1
15-1.3
GEAR TYPES
15-3
15-1.4
DESIGN FUNDAMENTALS
15-5
15-1
15-1.5
Backlash
15-5
15-1.6
Dynamic Load
15-5
15-1.7
Gear Accuracy
15-7
15-1.8
15-8
15-1.9
15-8
15-1.10
15-1.11
Inaccuracies
15-12
15-1.12
Friction
15-13
15-1.13
Inertia
15-13
15-1.14
Backlash
15-13
15-1.15
Compliance
15-15
15-2
15-2.1
MECHANICAL DIFFERENTIALS
PURPOSE
15-16
15-16
vii
Page
CHAPTER 15 (cont)
15-2.2
GEARED D I F F E R E N T I A L S
15-2.3
D I F F E R E N T I A L LINKAGES
15-3
.................................
.....................................
15-17
...............................................
15-17
..........................................
15-17
15-3.1
BASIC PURPOSE
15-3.2
PRACTICAL APPLICATIONS
15-3.3
EXAMPLES
15-3.4
NONIDEAL CHARACTERISTICS
..............................
15-17
........................................................
15-19
.......................
15-19
.........................................
15-19
15-3.5
15-3.6
Compliance
15-3.7
15-4
.....................................................
15-20
......................................
15-20
...................................................................
15-20
....................................................................
15-20
SHEAVES AND T A P E S
PURPOSE
15-4.2
STRESS
15-4.3
TENSION
15-4.4
SHEAVE SIZES
15-4.5
COMPLIANCE
...............................................................
15-21
................................................
15-21
15-5.2
KEYS AND S P L I N E S
. . . . . . . . . . . . . . . . . . 15-22
......................................................
15-5.3
Keys
15-5.4
Splines
15-5.5
Involute Splines
.............................................
............................................................
BEARINGS
15-21
..................................................
15-5.1
15-6
15-20
.................................
15-4.1
15-5
15-16
..........................................................................
15-22
15-25
15-25
15-25
..............................................
15-29
.........................................................................
15-29
15-6.1
ROLLER BEARINGS
.....................................................
15-29
15-6.2
BALL BEARINGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15-29
15-6.3
LUBRICATION
................................................................
15-32
15-6.4
FRICTION
.....................................................................
15-32
15-6.5
S L E E V E BEARINGS
15-6.6
MISCELLANEOUS BEARINGS
.......................................................
viii
...............................
15-33
15-34
Page
CHAPTER 16
TYPICAL PROCEDURE
16-1
INTRODUCTION
16-1
16-2
GATHERING O F SPECIFICATIONS
16-2
16-3
16-4
16-3.1
GENERAL
16-4
16-3.2
16-4
16-3.3
16-5
16-3.4
16-4
16-5
16-6
16-7
16.8
16-13
O F A SERVO
16-14
16-8.1
16-14
16-8.2
DESIGN O F T H E SERVO
16-15
16-8.3
Accuracy Determinations
16-15
16-8.4
Antenna characteristics
16-15
16-8.5
Synchro accuracies
16-15
16-8.6
Dials
16-16
16-8.7
16-16
16-8.8
Dynamic errors
16-16
16-8.9
16-8.10
16-16
. . . . . . . . . . . . . . . . . .. . . .
16-18
16-8.11
Detailed analysis
16-18
16-8.12
16-21
16-8.13
16-23
IX
Paragraph
CHAPTER 17
REPRESENTATIVE DESIGNS
17-1
INTRODUCTION
17-1
17-2
17-2.1
GENERAL
17-1
17-2.2
PU R P0SE
17-1
17-2.3
OPERATION
17-1
17-2.4
17-3
17-2.5
NOISE
17-4
17-2.6
DESIGN CHARACTERISTICS
17-4
17-3
17-3.1
GENERAL
17-5
17-3.2
OPERATION
17-5
17-3.3
17-8
17-3.4
DESIGN CHARACTERISTICS
17-9
CHAPTER 18
AUXILIARIES ASSOCIATED W I T H SERVOMECHANISMS
18-1
AUXILIARY PUMPS
18-1
18-1.1
PURPOSE
18-1
18-1.2
T Y P E S O F AUXILIARY PUMPS
18-1
18-1.3
Gear Pumps
18-1
18-1.4
Vane Pumps
18-2
18-1.5
Piston Pumps
18-3
18-1.6
MAINTENANCE
18-3
18-1.7
18-3
18-1.8
COST
18-4
18-2
HYDRAULIC AUXILIARIES
18-4
18-2.1
18-2.2
CHECK VALVES
18-2.3
18-4
18-4
X
Paragraph
CHAPTER 18 (cont)
18-2.4
PRESSURE-RELIEF VALVES
18-2.5
PRESSURE-REGULATING VALVES
18-2.6
ACCUMULATORS
......................................
....................................................
18-2.7
Gravity Accumulators
18-2.8
Hydropneumatic Accumulators
18-2.9
18-3
18-3.1
UNLOADING VALVES
ROTARY JOINTS
.................................................
18-8
18-9
..........................................................
18-12
DYNAMIC SEALS
.......................................................
18-12
.........................................................
18-12
....................................................................
18-12
18-3.3
0-Rings
18-3.4
18-3.5
Shaft Seals
18-3.6
Face Seals
18-3.7
High-pressure Seals
18-3.8
Friction
.........................................
18-12
.............................................................
18-12
................................................................
18-13
......................................................
18-14
..............................................................
18-15
18-4.1
PURPOSE
18-4.2
CHARACTERISTICS
18-4.3
LIMIT STOPS
18-4.4
POSITIVE STOPS
Buffers
18-8
18-10
Glands
18-4.5
..................................
18-6
................................................
18-3.2
18-4
..........................
18-6
...........................
.............................................................
18-15
18-15
............................................
18-15
............................................................
18-16
...........................................................
18-18
.............................................................
18-18
CHAPTER 19
CONSTRUCTIONAL TECHNIQUES
19-1
BASIC CONSIDERATIONS
19-2
COMPONENT LAYOUT
..............................................
19-1
..................................................
19-1
19-2.1
PHYSICAL ARRANGEMENT
19-2.2
.....................................
19-1
19-2.3
Radiation
19-2.4
19-4
19-2.5
Conduction
19-6
...............................................................
xi
19-4
Paragraph
CHAPTER 19 (cont)
19-2.6
19-3
VIBRATION ISOLATION
19-4
SHOCK ISOLATION
.....................................................
...........................................................
19-8
19-10
CHAPTER 20
SUPPLEMENTARY TABLES. FORMULAS. A N D CHARTS
20-1
.........................................
20-1
.....................................................................
20-1
20-1.1
DEFINITION
20-1.2
20-1.3
PARALLEL-AXIS THEOREM
20-1.4
20-1.5
PRODUCT O F INERTIA
20-1.6
..............................................
...................................
.............................
...............................................
..................................................................
20-1
20-1
20-1
20-1
20-4
20-1.7
20-1.8
COMPLICATED SHAPES
............................................
20-4
.........................................
20-4
.............................................................
20-4
20-2
20-2.1
VISCOSITY
20-2.2
Definition
20-2.3
20-2.4
Kinematic Viscosity
20-2.5
Effect of Temperature
20-2.6
FRICTION
...................................................................
..................................
....................................................
20-4
20-5
...............................................................
20-8
20-2.8
20-3.1
20-4
20-8
Coefficient of Friction
SPRINGS
20-4
................................................
20-2.7
20-3
................
..................................................
...................
20-8
20-9
.........................
20-9
.........................................................
20-9
...............................................................................
20-9
..............................................
HELICAL SPRINGS
20-3.2
Stress
20-3.3
Deflection
20-3.4
Torsional Elasticity
20-3.5
Design Table
.......................................................................
20-9
......................................................
20-9
....................................................................
20-9
xii
TABLE
OF CONTENTS (cont)
Paragraph
Page
CHAPTER 20 (cont)
20-3.6
CANTILEVER SPRINGS
...........................................
20-11
......................................................................
20-11
20-3.7
Definition
20-3.8
20-3.9
20-3.10
20-11
.............................................
20-11
.........................................................................
20-11
20-3.11
20-3.12
20-3.14
20-4
20-11
.............................................
..
Definition
20-3.13
......................................................
............................................................
VIBRATION IN SPRINGS
Natural Frequency
20-11
..............................................
20-14
.........................................................
20-14
20-4.1
BULK MODULUS
20-4.2
COMPRESSIBILITY
20-4.3
SPECIFIC GRAVITY
20-4.4
..........................................................
20-15
..................................................
20-15
.............................................
20-15
xiii
..................
20-15
CHAPTER 1
INPUT + NOISE
ERROR
AMPLIFIER
MEASURED OUTPUT'
MEASURING
ELEMENT
C COMPARATOR
CONTROLLED
ELEMWT
--
*OUTPUT
r-
1-1
THEORY
c
Fig. 1-2 Elements of an open-/oop control system.
able or input
n =noise
e =actuating
variable
m = manipulated
variable
1-3
upset
c
i = desired or ideal
output
y, = system error
THEORY
In many cases, the representation of Fig.
1-3 can be simplified. If the measuring and
feedback elements are ideal and have no
dynamic lag, it is possible to redraw the figure
(see Ch. 3) so as to have no elements in the
feedback path of the system.
A system in which the unmodified controlled variable is fed back directly for comparison with the input is called a unity-feedback system. The main loop of a unity-feedback system is shown in Fig. 1-4.
DISTURBANCE
VARIABLE
@
QHIPENSAnNG
VARIABLE
AND
AMPLIFYING
aatmrs
VARIABLE
CONTROLLED , ELEMENTS
i
MEASURING
AND FEEDBACK.4
ELEMENTS
DESIRED OUTPUT
IDEM
ELEMENTS
7-
MLIFYIMC
control system.
CONTROLLED
ELEMENTS
-ROLLED
ELEMENTS
ELEMENTS
1-4
CHAPTER 2
DYNAMIC RESPONSE*
2-1 INTRODUCTION
The dynumic response of a component or
system is the output response to an input that
is a varying function of time. The steadystate response of a component or system is
the output response to an input that is constant with time.
Paragraph 1 3 indicates that dynamic response is a basic determinant of system stability as well as an imporQmt element of
system performance. All design theory for a
feedback control system is centered on the
study, analysis, and manipulation of the
dynamic response characteristics of the system and of the components that are part of
the system. Because of its fundamental importance, the dynamic response of any physical device or system is classified according to
the nature of the input time variation that
occurs. In some cases, the input time variation may be entirely artificial since it may not
ordinarily occur in practice (for example, a
sinusoidal signal). In other cases, the input
variation may be one that is known to occur
in practice (for example, a step change). In
the former case, the artificial input time function is used primarily to facilitate analysis,
design, and testing. In the latter case, the
actual response of the system to the known
input function is an important measure of
performance which both the designer and
user need to know in order to verify that the
system meets the performance specifications.
In either case, a clear understanding of the
nature of the input and of the methods for
findiw the response to it are necessary for
successful design.
2-2 LINEARIZATION
The basic tool used to describe the dynamic
performance of a device is the set of differential equations that serve as a mathematical
model for the actual physical device. Since
quantitative techniques are imperative for
analysis and design, a mathematical description is necesaary. However, when going from
the physical device to the differential equation
model, one must resort to approximations if
usable results are to be expected from a
reasonable expenditure of time and effort. If
the physical situation is such that it is possible to describe the device with a set of
2- 1
THEORY
.. 5.1
= Cal
+ Cd + ... + C z , , h ~
A f (XI,
22, * * *
- f (210,
(2-1)
(2-2)
where
(zl0,210,. . ,zm0
= steady-state operating
point
( k = 1,2,.. ,n )
(k = 1,2,.
Axr = x r ; - z z x 0
Csk=
aaXkf
210, XLW,
* *
,G
. .,n )
In the approximation [Eq. (2-1) 1, the deviation or increment of the function from the
operating point has been expressed as a linear
function of the deviations (from the operating point) of its independent variables. The
constant coefficients C,, are called the paOtial
coefficients of f with respect to ua, 2 2 , .
2,
at the operating point ( X ~ O x, z 0 , . . ,Z.O).
Example. A shunt d c motor is speedcontrolled by the vacuum-tube circuit shown in
Fig. 2-1. Assuming that hysteresis is negligible, the basic equations of the system are as
follows :
E, = E , (Em4 )
2-2
..
(2-3 )
DYNAMIC RESPONSE
E. - Ep - N
Zf =
P'
dF
fdt
Rf
F =F(Zf)
Eb
+- I
M = k,FI.
dN
M = J -+ML
dt
where
E, = plate voltage
E, = grid voltage
Zf = field current
E, = supply voltage
Eb = motor back emf
La= armature inductance
Ra = armature resistance
N = motor shaft speed
k, = motor back emf constant
N f = number of field turns
F = field flux
Rf = field resistance
Z. = armature current
M = motor torque
k, = motor torque constant
J = total moment of inertia
ML = load torque
It is assumed that mechanical friction is
negligible.
To linearize Eq. ( 2 3 ) , let+
Ep:
E,,
e, = Epo AE,
(2-10)
E, I Ego e, = E ,
AE,
(2-11)
If :
If,,+if = 11, AZf
(2-12)
where Epo,Em, and Zfo represent values at the
steady-state operating point and e,, e,, and if
represent the deviations of the values of E,,
E,, and I f from their corresponding steadystate operating-point values.
+
+
+
+
+Symbol t
- is defined as "equals by definition".
d-c motor.
(-PI
e,+
(R,) if
(2-13)
= -peg
+ R, if
(2-14)
F d Fo+f=Fo+AF
(2-16)
+ Cf if
(2-16)
ZZ
CJi,
eb = k, (Not
(2-17)
+ iJ=
En,+ e, - (Epo+ e p )
ZJo
-N
d
r(F0
dt
+ CJ4)
(2-18)
or
(2-19)
+ Fan)
(2-24)
,z
E80
- kJf10
(2-25)
Ra
and the incremental equation for the armature circuit is
di.
La --k Raia= e, - k,NoC,iJ - k z 0 n
dt
(2-26)
where
Za
where
& Zao
ia
+ kmLof
m = kmFoia
(2-27)
where
e, = increment in E,
MAMo+m
It can be seen that the equation for the operating point of the field circuit is
(2-20)
and the incremental equation for the field
circuit is
Mo
= MLO
dn
m =J -
(2-22)
where
Eb A Ebo
+ eb
N $ No+n
(2-23)
(2-29)
di.
(2-21)
(2-28)
dt
+ mL
(2-30)
where
DYNAMlC RESPONSE
(2-20)
e,, = -Beg
Epo iEP
Zao
Fo 4
woo,
If,)
E80 - WON0
Ra
F (110)
+ Rpif
(2-14)
(2-31)
(2-21)
(2-25)
L, di,
-+ R,ia = e, - kcNoCfif- k,Fon
(2-32)
Mo
=MLO
(2-29)
Mo
4 kmFJao
(2-33)
dt
(2-26)
Jdn
dt
+ m, = k,Foi, + kmZaoC&f
(2-34)
[+-s,,(t -ajdt = 1
(2-35)
J -m
(2-36)
hO(t- a ) = 0, t
(2-37)
THEORY
*o+
- t
A. UNIT IMPULSE
8. UNITSTEP
C UNIT RAMP
0.6
0.4
-0.2
.,
6
8
t (SEC) -b
10
2-6
12
dx
d2Y
dt2 '
- 1
DYNAMlC RESPONSE
-a)
8 - , ( ~- U
) ~ X
(2-40)
d2y
dt*
(2-42)
(2-41)
e ( t ) = tan- - t
R
where
CONSTANT VELOCITY
(2-43)
NTENNA
V = target velocity
Fig. 2-4 Straight-line crossing course.
DYNAMIC RESPONSE
Atistical characterization of the output signal ensemble.
2-9
CHAPTER 3
serve as the mathematical model for the physical system. This chapter deals with methods
of determining the dynamic response of physical systems from the differential equations
that describe them. The type of response
sought depends upon several factors: the
specifications of the system; the design procedure adopted ; and the limitations imposed
by test conditions encountered when seeking
experimental verification of the design performance.
Differential equations may be classified as
follows :
(a) Linear differential equations with constant coefficients
(b) Linear differential equations with
time-varying coefficients
(c) Nonlinear differential equations
Of the three classes, constant-coefficient linear
differential equations are, by far, the most
widely used and the best understood. The subject matter of this chapter is focused exclusively on methods of solving equations in this
class. Chapter 9 deals with time-variable
linear differential equations of a specific type
that have a wide application. Chapter 10 considers nonlinear equations and some of the
techniques for treating them.
The general form of a linear differential
equation with constant coefficientsis
I
f=o
d'x
a-=
dt'
z
j=o
djy
bj-
dt'
z(t)=zp(t) +
(3-1)
*ByL.A . Gould
3-1
Akeprt
(3-2)
*=I
THEORY
1)
2 aipi = 0
=o
(33)
3-2.1 FACTORING
In most cases, the solution of a linear differential equation requires the determination
of the roots of the characteristic equation
[Eq. (3-3)]. Unfortunately, if the order of the
equation is high, the process of factoring the
equation to find the roots becomes extremely
tedious. For such cases, special techniques
have been developed (see Pars. 4-4, 5-7, and
7-1). This section covers some general factoring procedures applicable to any algebraic
equation. In addition, the characteristics of
first- and second-order equations are discussed and graphical methods for determining the roots of third- and fourth-order
equations are presented.
The factoring of rational polynomials is
covered by many authors.(5,6.7.8.9.10.33) The
method presented here is one that is very
convenient.
The general algebraic equation can be written as
p2
c2
e,
(3-4)
...
+
+ . . . + c1p +
.................
+ CI p + co l p .
v,"*+
CO
C"-p-1
c'2p2
C'2P2
+ c ' i ~+ co
+ +
C'lP
remainder
3-2
c'o
(3-7)
c'l
+P
+-
8 2
Go
(3-8)
c'2
+ l0.66@
+89.0@ +15.50p
+ 27.0 = 0
(3-9)
Ql(P) = P + - 16.60P
27'0 - p2
Dividing f (p) by g, (p) produces
89.0
+ 89.0
p2
+ 10.48~+ 86.9
++
(3-10)
+ 0.1742~+ 0.303
+ 27.0
0.17~~ 0 . 3 ~ ~
10.48~~
+ 88.7~~
+ 16.60~
10.48~~ 1 . 8 ~ ~ 3.18~
second trial divisor86.9~~
+ 12.3210 27.0
+ 16.14~+ 26.3
86.9~~
remainder- 2.82p+ 0.7
3-3
THEORY
27.0
-pz
+-12.32
P+-86.9
86.9
(3-11)
+ 0.1418~+ 0.311
Dividing f ( p ) by g, ( p ) yields
+ 1 0 . 5 1 ~+ 87.2
+~89.0~'+ 1 5 . 5 0 ~+ 27.0
p 2 + 0.1418~+ 0.311 / + 1 0 . 6 5 ~
+ 0.14~:'+ 0 . 3 ~ ~
10.51+
~ ~88.7~'+ 1 5 . 5 0 ~
+~ 1 . 5 ~ +
" 3.277,
10.51~
third trial divisor
+ 87.2~'+ 12.232, + 27.0
87.2~'+ 1 2 . 3 6 ~+ 27.1
p2
'p4
p4
> - 0.13~
- 0.1
remainder
(3-12)
27.0
+= p 2 + 0.1403~+ 0.310
87.2
+ 0.1403~+0.310)
(p'
+ 1 0 . 5 1 ~+87.2)
(3-13)
3-2.2
CHARACTERISTIC PARAMETERS OF
RESPONSE MODES
+ c,)
IC
z
?i aaI
=0
z ( t ) = Ae-c.t
(3-14)
j 7.72
06
0.4
(3-15)
2 0.2
0.4
0.8
1.2
1
IS
21)
2.4
2.8
3.2
3-4
3
.
6
41)
e4.
Figure 3-2 shows plots of o, versus 5 for constant values of a2 and al. Figure 3-3 shows
plots of a2 versus a1 for constant values of 5
and or. From these charts and the thirdorder equations, the roots of the cubic. can
be determined.
'P
+ 2bnP +
on2
=0
(3-16)
where
{=--
C1
_-
2 dco
o1=
=
'd
damping ratio
P'
+ 4)
p = co'/'h
1'
where
01
= q,+l
- (2 = damped
(3-20)
(3-17)
COB ( ~ d t
C2P2
+ cspa + + ClP + co = 0
(3-21)
where
frequency of
transient oscillation
If 5 2 1, the second-order factor can be factored into two first-order factors so that the
response consists of two first-order modes.
as = ca/c01/4
a2 = c2/c01/2
a1 = c l / c P
52.5 Third O d d 1 )
P9 C2P2 ClP co = 0
+ + +
is az A2
+ al 1 +1 = 0
(3-18)
or, alternatively
where
a2 = c2/co
and a1 = cl/c@2/8
or
of reference component
(3-19)
t, A
where
a1 =
p,
A or2/or1
= ratio
pt
w(,=ratio
25 + or2
o r
wr = a reference frequency
3-5
of undamped natural
frequencies of components
of damping ratios of
components
THEORY
3-6
3-7
+ 9 A +
a2 A2
a, A
+ 1 = 0
10.1
2 02
30.3
40.4
60.6
n
a
Y
80.8
K)o.Io
>
200
300
400
800
3-8
'1+
dt'
a,
d3x
dt3
+ a2 d 3 + a ,
dt2
drr
dl
+ aox
-o
1.0
1.0
0.9
0.9
0.8
0.1
0.6
C
0.5
5
Z
2I
0
0.4
U
EK
w
U
w
03
&!
Y
0
Q
P
C
4
ul
ul
W
Y,
w
Z
0.I
0.1
RATIOS OF DAMPING RATIOS p
I?
51
3-9
THEORY
sr.
N 4
+ aI2- 4a2
aa2
sr
az3
1
2
- and - (a3 + a l )
a3
a1
(3-24)
or = -
Vx-
P u p ) -1
Pg
=
Pu-
(8-25)
(3-26)
($)
4. , - ( P1w + $ )
{
a2
(3-28)
(3-29)
[*+
] ,V
(3-30)
(4+ al)
SCALE OF 1/2
CHART I
SCALE
OF
x(t) =
ly
d t , w ( t l ) ~ (- tt l )
w ( t ) =0 fort
lr
dtl
(3-33)
x(t) =
lo
- tl)
(3-34)
t l ) y(t1)
(3-35)
dtl w(t1) Y ( t
or
(3-31)
x(t)=
or
x(t) =
<0
(3-32)
This integral applies in every case and is useful for graphical time-domain studies of system performance. In many situations, however, evaluation of the convolution integral
is tedious, so more refined procedures are
used (see Pars. 3-4,3-6,3-7,3-8,and 3-9).
3-1 1
d t l w(t
(3-36)
or
x(t) =
dtl w ( t - t l ) Y ( t l )
(3-31)
Next Page
THEORY
If
(d) J+m
( t ) I d t is finite
-m
The conditions for the existence of the Laplace transform of a function are identical
with those for the Fourier transform except
that the fourth condition is relaxed to
~ f ( It e-r'
) d t is finite for a finite c
3-4.2 THEOREMS
J-m
(3-38)
where s = o
+ jw
[Inverse]
The following theorems are useful for applying the Laplace and Fourier transforms to
the solution of differential equations :
(3-39)
L [ f ( t ) ]$ F ( s ) &
(c)
d t e - * ' f ( t ) [Direct]
[-]
(O+)
dtn
(3-40)
- s"-'f'(O+)
-
[Inverse]
- . .. - s f (
f('1-1)(O+)
(3-44)
11-2
(O+)
(3-41)
+ jco.
s"'
Is'
.Jt
+
--
--OO
;(t)dt]
(3-45)
--
(3-46)
= F, ( s ) F , ( s )
(3-47)
(R) L [ f ( t- a l l = e-"#F(s)
if f ( t - a) = Ofor O < t < a
(3-48)
3-12
Previous Page
( h ) - C [ f ( t a ) 1 = eaaFFs)
i f f ( t + a ) = Ofor -a < t < 0
( i ) lim s L [ f ( t ) ] = lim f ( t )
x+
(3-49)
(3-50)
t+ m
( j ) lim s - C [ f ( t ) ]= lim f ( t )
(3-51)
I+ 0
R-,
(k) L [ f I( t ) f . ( t ) I
(3-56)
Theorems ( a ) , ( b ) , (e), ( f ) , ( i ) , ( k ) , and
(1) also apply to the Fourier transform.
Theorem (c) is called the r e d differentiation
theorem and theorem (d) is called the real
integration theorem. Theorem (e) is used to
change the time scale of a problem and is
called the normalization theorem. Theorem
( f ) is the real convolution theorem and, if applied to the convolution integral [Eq. (3-35) 3 ,
yields the very important result
= W ( s )Y (s)
(3-54)
Y ( s )+ B ( s )
(3-55)
system
input
(function) (transform
+(
initial condition)
function
(3-57)
3-13
THEORY
f(t),t 2 0
a,-, ( t )
1
1
-
( t ) , unit step
1
-
( t ) unit ramp
82
1 e-t/T
-
Ts + 1
s2
unit impulse
sin ot
w2
cos wt
1
(2)
5 = 1: te+
1
-
tt1-1
(n-l)!
S"
+ 1)"
(Ts
(n-l)!
tn-1
e-t/T
T"
3-14
(3-58)
+ 10.65-+d3x
where
dt4
dt3
89.0-+
d2x
dt2
15.50 dx
dt
+ 27.02 = 2 7 . 0 ~
(3-62)
(3-59)
X(s)=
where
Kkj
27.0
1
(j-l)!
s4
Let
and m,,is
s = s/,.
2"
D(s) =
(S
(s
(3-64)
e ( t ) = Y(t) - x ( t )
Then, transforming Eq. (3-64) and substituting for X ( s ) from Eq. (3-63), E ( s ) becomes
E(s) =
(3-65)
3-15
(3-66)
THEORY
Y ( s )= s2
(3-67)
E(s)=
S[S
+ 15.501
[S
[S
+ 5.26 + j7.721
(3-68)
E(s)=2
K
s
Kz
K,
+ 0.0702 - j0.552
+ 0.0702 + j0.552
K3
+ 5.26 - j7.72
(3-69)
K3
+ 5.26 + j7.72
(S:{+ 10.65s:
K, =
[S
(s
+ 89.0s i- 15.50)
= -(I
0702
+j
0.918 e-nl
1412~
~ . ~ ~ , 2
(3-71)
K:, =
+
+
,.=-5 . ~ ~ + 1 7 . T 2
= 2.89 x 10-4 p - 1 0 2 i i ~
(3-73)
+ 0.918 e-0.0702t
- 1.89) + e-J(0.5X2f
- 1.89) 1
(3-75)
3 = IW(jw)I
(3-78)
A
where .4, is the output amplitude, A,, is the
input amplitude, and W ( s ) is the transfer
function of the system. The phase angle of
the output relative to the phase angle of the
input is given by?
4Jx
- 4Jy
= L W(jw)
(3-79)
(22*23~24~25~26127~2*)
3-17
THEORY
Symbol
+ K(Bl1,+
TL
Operation
variable
operator
Y=AX
summinc point
Y=X--U
splitting point
x=x
multiplier
Y=XZ
-@I--
Tm = (Jms2 fhs) 0,
K ( e , - et) = JLsZO,
Description
(3-80)
(3-81)
where
K = shaft stiffness
3-18
ORIGINAL DIAGRAM
EQUIVALENT DIAGRAM
, + y + I-
3-19
3)
THEORY
__
RULl
ORIGINAL DIAGRAM
EQUIVALENT DIAGRAM
i-tl
IY
W
t x
I0
3-20
ORIGINAL DIAGRAM
EQUIVALENT DIAGRAM
t'
-t
WHERE A,
AC
- ED
WHERE A ? = 1
- ABCD
3-21
THEORY
Y+
J,sz
L
f, s
1
J,sz
ern
C. USE OF RULE 3 OF F k . 3 6
3-22
0,
1
J,r2+f,s
K
ern
1
Jms2+ 1,s
F1
K
ern
TL
3-23
8,
Next Page
THEORY
--
1
JmS2
+ f,s
0,
JLs2
1
Jms2 +
fnS
ITLK
I. USE OF RULE 1 OF FIG. 3.6
3-24
=.
1
W JLS2 + K
--
OL
CHAPTER 4
4-1 INTRODUCTION
+ Gr ( s ) G ( s ) H ( s ) = 0
(4-2)
Iu
C(s) = Gi ( s ) GP(s) R ( s ) - G z ( s ) V ( S )
1 G ( s )G,(s) H ( s )
(4-1)
* B y L . A . Gould
4- 1
THEORY
ity of the - 1 j 0 point, the Nyquist criterion provides the servo engineer with a relatively straightforward and extremely powerful tool for analysis and design.
The root-locus method is a graphical technique for revealing the position of the poles
of the response transform C(s) in the s plane
as a gain factor of the open-loop function
G l ( s ) G,(s) H ( s ) is varied. The primary
advantage of this method for stability determination is that the closed-loop pole locations
are kept in evidence a t all times. Thus, it is
easy to see when the poles move into the
right-half plane as the gain factor is varied.
There are two primary disadvantages connected with the root-locus method. First, the
location of the poles and zeros of the openloop function must be specified. This often
requires some sort of analytical approximation to the experimental test data. Second, the
plotting of the paths of the closed-loop poles
involves a trial-and-error procedure that can
be quite tedious. In spite of these disadvantages, however, the root-locus method is quite
useful in that it immediately places in evidence the closed-loop pole-zero configuration
for any particular design (stable, of course).
Thus, the characteristics of the time response
of the system are easily ascertained and the
verification of performance specifications in
the time domain is a straightforward matter.
a,sn
+ a,-,s"-' + . . . + a1s + a, = 0
(4-3)
(1)
a,
a"+
an-4
(2)
an-1
a,,+
an-5
.. .
...
(4-4)
Then, the array is extended by taking appropriate cross-products to determine the elements in the third row
(3)
an-1 a n - 2 - a n
an-1
an-3
%-I
a"-4 - a n
an-:
...
%-l
(4-5)
4-2
+ Ks3 + 2s2 + 4s + M = 0
where K
(4-6)
(1)
(2)
K
(K-2)
KM/2
0 (row multiplied by K / 2 )
(3)
(4)
[ (K-2) - K2M/8]
0 (row multiplied by 1 / 4 )
(5)
(K-2-K2M/8)
0 (row multiplied by 2 / K M )
(6)
> 0,
K-2 > 0,
K
(4-8)
(4-9)
and
[(K-2) - K2M/8] > 0
(4-10)
4-3
(4-7)
THEORY
(1)
89
(2)
46
260
(3)
-1
-1.95
(4)
6.17
(5)
(6)
(7)
(4-12)
(s + 4 ) (s - 2 + j3) (s - 2 -j3)
(5
+ 1 + j2) + 1 - j2)
(4-13)
(8
s,, s2 = + 2
+ j3
(4-14)
The Nyquist criterion is a graphical procedure by which one can determine whether
any of the roots of the equation
1
+G(s)=0
2 = number of zeros of 1
G (s) that lie
in the right-half s plane
(4-15)
lie in the right-half s plane. Only the following information is required in this procedure :
(1) the magnitude and phase angle of G ( j w ) ;
(2) the behavior of G (s) at the poles of G (s)
that lie on the imaginary axis or at the origin
of the s plane; and (3) the number of poles of
G (s) in the right-half s plane. (NOTE : For
nonunity feedback loops, one tests for the
zeros of the function 1 + G ( s ) H ( s ) where
G ( s ) is the forward transfer function and
H ( s ) is the feedback transfer function.)
The Nyquist criterion can be expressed
mathematically as
Z=N+P
right-half s plane
(4-16)
4-4
semicircular detour is made into the righthalf s plane a t each pole of G (s) on the imaginary axis. Thus, the G ( j w ) locus will describe a large semicircle instead of becoming
infinite. If the pole on the imaginary axis is
of multiple order, the G(jw) locus will describe one semicircle for each order of the
multiple pole.
A convenient rule for determining the
direction of turn of the G(jw) locus a t the
imaginary-axis poles of G (s) is
Turn to the right by 180" for each order of
the pole as the frequency increases.
I f G(s) has no poles in the right-half s
plane o r on the imaginary axis (except a t the
origin), the Nyquist stability criterion simplifies t o t
when L G (jo)= -180"
I G(jo)i< 1
(4-17)
KT
(4-19)
A (A
Plot (l/KT)G(A) on the complex plane for
h = ju. Such a plot is sketched in Fig. 4-3.
Since encirclements depend only upon the
topology of the plot, the locus can be distorted
to facilitate the counting of encirclements
(Fig. 4-4).
IMAGINARY AXIS
!
A
Examples.
( a ) Consider the function
K
s ( T s 1)2
For what range of K will 1
stable roots?
G(s) =
tSymbol
+ G(s)
2'1 I
(4-18)
REAL
u>o
have
denotes "angle"
Fig. 4-3 Locus of
1
G(A)
KT
1
for A = ip.
-
h(h
4-5
PLANE
A(l
+ 1)2'
THEORY
IMAGINARY
AXIS
Nature of
Stability
stable (2= 0)
I 0 I 2 I 2 I unstable ( Z # 0)l
0
unstable (Z#O)
-:>.
G ( s ) = K (1
1 1+
Fig. 4-5 LOCUS of s (1
s)
s ( 1 - s)
For what range of K will 1
stable roots?
(4-20)
+ G ( s ) have
N
0
-1
Nature of
Stability
unstable
stable
unstable
IMAGINARY
AXIS
+as) PLANE
-co<K<-~
4-6
1 (;3
fi (s +
G(s) = K O
a =1
,
n (S +
S J
(4-21)
sj)
j=l
(4-22)
x A, - x A, = -180"
where
from the zeros of G (s) to the point
in question
(c) The real-axis position of the locus always has an odd number of poles and zeros
to the right of the s point for KO > 0.
The value of the constant KO that is associated with each root-locus point is found
from the relation
(e) For two adjacent zeros, the locus enters the real axis from the complex plane a t
the point of minimum KO.
(4-23)
where
1180" - X A,
+ 2 Apl
[- 180"
4-7
+ Z A , - Z A*].
THEORY
( n = 0, 1,2,. . .)
(4-24)
x, =
f ( 6 ) = 6(-
C sj -PC= l s k
j=1
where
p = number of poles
sj = jtbpole
z = number of zeros
Sk
+ 3006
f ' ( 6 ) = 36' - 806 + 300 = 0
f (6) = 6" - 40a2
= kthzero
6 = 20.8, 6.3
s(s
Kol
02
01)
(5
0 2 )
(angle condition)
S+
6.3
+ 30)
+ / S + 10 + / S + 30 = 180"
I S I I s + 10 I I
+ 10) (-
K = 1.84
(4-25)
P--z
A=
30 I = 300 K
180" f 360" n
n
- z .r
(magnitude condition)
-60"
2.
- [O
= Z sj - Z S P -
v-z
+ 10 + 303 - [O]
3-0
= 13.3
Therefore
tan-'
- + tan-'
O C
0,
01
0 2
+ 90'
= 180'
4-9
THEORY
io
0
A
PLANE
A. TYPE T(1,O)
6. TYPE T(1,l)
io
C TYPE T(2,O)
D. lYPET(2.1)
4-10
I of 5 )
Iio
+
X
D. TYPE 1(2,1)
EQUILATERAL
TRIANGLE
E. TYPE T (3,O)
4-1 1
THEORY
E. TYPE T(3.0)
I
I
F. TYPE T(3.1)
4-12
j0
POLE
C. TYPET(3.2)
H. TYPE T(4.O)
4-13
THEORY
\--/
/=\
I
LX
d
H. TYPE T (4,O)
4-1 4
300K
s(s
+ 10) (s + 30)'
K1
1.84
/-
/
Fig. 4-9 Asymptotes and real axis behavior for
300K
G(s) =
s(s
+ 10) (s + 30)
'
300K
s(s.
+. 10) (s +. 30)
. .
4-15
THEORY
BIBLIOGRAPHY
4-16
CHAPTER 5
GAIN DETERMINATION*
5-1.1 GENERAL
The velocity constant of a system is a measure of the steady-state error if the input to
the system is a constant velocity. The velocity
constant is defined by the relation
( 5-3 )
where
mi
5-1.2 GAIN
[s G (s)]
K = lim
s+o
G(s)
(5-4)
s+o
(5-1)
(5-2)
where
a, = constant input acceleration
*By L. A . G o d d
5- 1
THEORY
Several commonly used upper-frequency values for unity-feedback systems (Fig. 5-1)
are given below :
(a) oR, resonant frequency -frequency
at which the closed-loop frequency response
C (j w ) / R (jo) has its peak magnitude M,
(Fig. 5-2).
An analysis of Eq. (5-6) shows that the acceleration constant of a single-loop unityfeedback system is finite and nonzero only
if the open-loop transfer function C(s) / E (s)
has exactly one double-order pole at the origin (two integrations).
(b) w,-frequency
at which the magnitude of the closed-loop frequency response
C (jo)/ R ( j w ) is unity (Fig. 5-2).
(c) cob-frequency at which the magnitude of the closed-loop frequency response
C ( j w ) / R ( j w ) is 0.707 (Fig. 5-2).
(5-7)
where
STATIC ACCURACY
5-2
GAIN DETERMINATION
(g) wc, asymptote crossover frequency frequency at which the -10 dg/dec asymptote
of the open-loop frequency response C ( j w ) /
E (jo)crosses 0 dg (Fig. 5-5 ; see Par. 5-3 for
terminology).
5-1.8 PEAK MAGNITUDE
( 5-8 1
(5-9)
0.
(*-
or
Fig. 5-5 Bandwidth measures from open-loop
frequency response.
5-2.1 GENERAL
THEORY
each frequency by : (1) multiplying the magnitude of G(jw) by the magnitude of the cascaded function to give the magnitude of the
combination; and (2) adding the phase angle
of the cascaded function to the phase angle of
G ( j o ) to give the phase angle of the combination.
G(jw) =
j w [ (jo)'
+ 0.6jo + 1)
(5-11)
140.
2204
1
G-1 ( j w ) = (5-10)
G h)
The polar coordinates of G-l (jo)a t each frequency are given by : (1) the reciprocal of the
magnitude of G(jw) ; and (2) the negative of
the phase angle of G (jo). Increasing the gain
of G ( j w ) shrinks the inverse locus in the
radial direction. If G ( j w ) is cascaded with
another transfer function, the resulting polar
coordinates of the inverse of the combination
are obtained a t each frequency by : (1) multiplying the magnitude of the inverse G ( j w )
locus by the inverse of the magnitude of the
cascaded function to give the magnitude of
the inverse of the combination; and (2) adding the negative of the phase angle of the
cascaded function to the phase of the inverse
G (jw) locus to give the phase angle of the inverse of the combination.
I 60
1500
2100
P W
I 70
I900
I800
moe
1900
1700
20w
I 6OS
210.
I500
140'
22cT
I:
2
jw[(iw)'+
0.6jw
+ 11'
5-3.1 GENERAL
GAIN DETERMINATION
(Tjo+ 1) (first order)
+ 2, j
+1
(5-14)
(second order)
(5-15)
G ( j o ) = K (jo)?"
(5-12)
Only three general types of algebraic factors
appear in Eq. (5-12).
The three factor types, which may occur in
any rational function, are the following:
(5-13)
20ds
I Odg
OL
019
LL
0
Y
n
a
C -1Odg
z
0
4
x
-20
dg
01
02
05
2
Tw
I0
20
Fig. 5-7 Asymptotes and true magnitude curves for the firsf-order factor (Tim+
5-5
50
+
l)+'.
I00
THEORY
043
;
Y
L
Y
-6 d9
-Q)
w-
Fig. 5-8 Asymptotes and true curves for the second-order factor
0"
[( j z); 2 v"+
0.
56
an
-1
11
10
GAIN DETERMINATION
[( i - 3 2 + 2 5 j w + 1 ] - '
On
[(i
$ ) 2 + 2 5 i 2O+n1
t90.
+49
(L
YY
Y
-I
0'
z
<
m
W
<
-48
0. I
0.2
0.5
T*
10
5-7
20
(Ti0 + 1)*'.
50
I00
THEORY
(0.2jo + 1)
G(jo) =
(jw)
[ ( i %)".
150'
120.
I-
U
4
Y
LL
0
W
-I
90'
0
Z
4
W
v)
4
I
0
6 0.
3 0'
0'
0. I
10
5-8
0.6 i w 1
10
(5-16)
[( j %) + <
+ I] .
2 j
w,
GAIN DETERMINATION
[0.2 jo
=K
io
0. I
0.2
0.5
[( i :)+'
+ 11
w + 11 *
0.6 i 10
20
I0
a-
(0.2 jo
iw
5-9
+ 1)
[( i : ) 2 + ~ . 6 i k + 1 ]
50
100
THEORY
WOdr
o,
5
5!
-lob
-184
(0.2jw
+ 1)
Odr
-160.
-144
-124
-100'
L ffid +
-8P
-60.
(0.2 j
o
6.5
iw
+ 1)
[( i ~ ) ' + o . 6 i ~ + I'1
(5-17)
The relations that exist between the closedand open-loop responses of a unity-feedback
system can be obtained by considering the
diagram in Fig. 5-1. In this diagram, it is
clear that the open-loop responses is given by
(5-18)
GAIN DETERMINATION
because it usually proves to be tedious. Instead, various aids for performing the
G-to-W plane transformation are used. These
are presented next.
54.2 POLAR-PLANE TECHNIQUE
OB=G
(5-21)
OA = -1
(5-22)
OB - OA = AB=I
+G
OB = G-
-
OA = -1
(5-27)
TO
+ OB = AB=1 + G-
(5-28)
L W ( ~ W ) =L O A B = +
(5-23)
Then, the closed-loop response at each frequency can be determined from the construction of Fig. 5-14 as follows:+
$= L
(5-25)
(5-30)
M = iW(jo)l
(5-24)
L W(~O)L
= ABO = 4
(5-26)
W(j0)
N=tan$
(5-31)
(5-32)
(5-33)
5-11
THEORY
The M and 4 contours aid greatly in performing the transformation from open- to
closed-loop frequency response and are used
to facilitate the design of a system when the
shape of the G function is to be altered so as
to improve performance.
G Plane
7-
M contours
M contours
y
G-' Plane
1~ = Zm (G-I)
= Zm(G)
x = Re (G-I)
x =Re(G)
center:
radius:
( -1,O)
1
-
1-M
intercept nearest origin : M
M
intercept nearest origin : -M+1
N contours
N contours
N=tar 4
N =tan4
(~+0.5)'+ 2 ~ -
center:
1 2'
2 N )
"+
1
1
= 4 ( 7 )
t LJX+ N = 0
-)
1
1
-2 ' 2 N
5-12
GAIN DETERMINATION
M- 1
LOCUS OF H CONTOUR
<b -\k.
CPLANE
G- PLANE
CONTOURS
I\
C# in the G- plane.
54.3 GAIN-PHASE PLANE TECHNIQUE
(NICHOLS CHART)
C PLANE
-360
-270
-180
-90
5-13
THEORY
5-14
GAIN DETERMINATION
The closed-loop response of the nonunityfeedback system (Fig. 5-21) can be written as
follows :
(5-34)
Since the bracketed portion of the right-hand
side of Eq. (5-34) has the same form as the
right-hand side of Eq. (5-20), the Nichols
chart (or the polar M-N contours) can be
G H ) from a plot of
used to find G H / ( 1
G (j w ) H (jo). The closed-loop response
C ( j w ) / R ( j w ) can then be found by multiplying G H / ( 1
G H ) by H-' at each frequency.
If the closed-loop response of a nonunityfeedback system is sought, a slight modification of the procedure used for the unityfeedback system will enable the designer to
use the Nichols chart and the polar M-N
contours as well.
5-15
THEORY
(%)
5-16
GAIN DETERMINATION
locus is the resonant frequency of the closedloop system having the specified M, (oRin
Fig. 5-24).
9 = sin-' (j-j-)
i ~+) 0.6jo
~ + 11
jo[(
Solution.
( a ) Direct-plane procedure :
(1) The G(io) locus is plotted
(&)=
230.
130.
240.
120.
250.
110.
260.
100.
(Fig.
5-25).
MP.
220.
140.
(5-35)
270.
90.
5-17
280.
80.
290.
7P
300.
60.
= 1.6, GQo)=
320.
4 0'
310.
50.
jo [Qw)'
+ 0.6 jo + 1 ]
38.7'.
THEORY
220'
230.
140.
130.
240.
120.
250.
110.
260.
100'
270.
9Q
280.
80.
290.
70.
60.
320
310.
300.
40.
50'
jo [(jo)' 0.6 jw
+ 11
GAIN DETERMINATION
+I0 dg
Solution.
(a) The- (jo) locus is plotted (Fig. 5-28).
K
(b) The G ( j w ) / K locus is placed over the
M, = 1.5 contour, the phase-angle coordinates
/ I
,g
z
Odg
Locus
-180.
-160.
-140.
-120.
-100.
-8Q
G(j0)K
G(@) = K
(0.2 io
+ 1)
$[(i~>;o.6i;+1]
5-19
-6Q
THEORY
The peak magnitude of the closed-loop response is not the only measure of the degree
of stability that is commonly used. More
direct, but less reliable, descriptions of the
approach of the G(jw)locus to the point
-l+jO
are available. These measures of the
degree of stability are called phase margill
and gain margin.
5 6 . 2 Phase Margin (Fig. 5-29)
- -1800
is determined.
-180.
A. POLAR PLANE
+ p.m.
-90.
8. GAIN-PHASE PLANE
5-20
(5-39)
GAIN DETERMINATION
magnitude curve) for a specified phase margin is a good approximation to the gain determined from the corresponding M, criterion. The M p criterion corresponding to a
given phase margin may be found from the
relation
IG
lolog,,, -- -4.5 dg
p.m. = sin-'
1
(%)
(5-45)
(5-41)
5 - 6 3 APPROXIMATE CLOSED-LOOP
RESPONSE
(5-42)
To have 1 G ( j w ) I = 1 a t w = 11 rad/sec
10 log,,, K = 4.5 dg, or K = 2.82
(5-43)
Note, in this example, that the use of the
asymptotic curve to estimate K gives a poor
result. The magnitude of the asymptotic approximation for G ( j w ) /K at w = 11 rad/sec
is -7.5 dg. This would give an approximate
value of K = 5.61 for a 45" phase margin.
The error of approximation is a factor of two,
which is too large to be acceptable. One should
note further that, for this system, the gain
margin is infinite since the negative phase
shift never exceeds 180".
The phase-margin criterion used as a measure of the degree of stability is a good substitute for the M p criterion provided that the
G function does not have low damping-ratio
quadratic factors (c 5 0.3) with natural frequencies in the range where
-135'
<L
G ( j w ) < -225"
(5-44)
(5-48)
(5-49)
THEORY
Fig.
5-30 Approximate
G(jo)=
io(iw
5-22
+ 1)
'
GAIN DETERMINATION
_-
G (io)= 0.707.
1, K
= 1.
(c) Therefore, K = 1.41 and o,.,,,
(b) At w
1
=- -
s = - <on?
jI
o,d 1 - t'
cos odt
where
(To.) -* = time constant of envelope of mode
md
= o;dl-
<* =damped
frequency of
transient oscillation
5 = damping ratio
o, = undamped natural frequency
Zm(s) =
_+
o d
= f 0. q 1 -
IS/ = O n
s = f cos-'
5-23
<'
(5-52)
(5-53)
r f 180"
(5-54)
THEORY
behavior of the system becomes immediately
evident once the damping-ratio criterion is
satisfied. The disadvantage of the root-locus
method is that considerable time is consumed
in constructing the locus. Adjustment of the
gain K for a specified dominant-root damping ratio is best demonstrated by a specific
example.
Example. The open-loop function G ( s ) of
a unity-feedback system is
K (0.2 s 1)
G(s) =
(5-55)
s(s
1)(0.1 s 1 )
Find the gain K and the closed-loop pole-zero
configuration for a dominant-root damping
ratio ~ 0 . 5 .
Solution
( a ) The open-loop function is placed in
the standard form of the root-locus method as
follows :
L(s+5)L s - L(S+l)
Thus, lines drawn parallel to the imaginary
- L (S
10) = --180"
(5-57)
axis are loci of constant-envelope time con(c) The magnitude condition is
stant (<on)-*.
Lines drawn parallel to the real
axis are loci of constant damped frequency of
Is
51
2K
(5-58)
oscillation od.Circles centered at the origin
Is1 1s 11 1s 101 =
are loci of constant natural frequency o,,.
(d) The locus of the roots is constructed
Radial lines emanating from the origin are
from the angle condition by choosing trial
loci of constant damping ratio 5. The various
points in the s plane and checking back to see
s-plane loci for the second-order factor are
whether
the angles of the vectors of the openshown in Fig. 5-31.
loop poles and zeros add up according to Eq.
(5-57).A curve drawn through the trial points
5-7.5 G A I N DETERMINATION IN THE s PLANE
that satisfy this equation is the root locus. To
The usual degree-of-stability criterion for
determine the gain K associated with each
determining the gain K from the root locus
locus point, Eq. (5-58) is used. The value of
of a system is:
s corresponding to a given locus point is substituted into this magnitude equation and the
The dominant roots are adjusted to
equation is then solved for K . The locus for
satisfy a specified damping ratio
this problem in the upper half of the s plane
The advantage of the root-locus procedure
is shown in Fig. 5-32.
over the M, criterion of t h e frequency(e) The line corresponding to ;= 0.5 ( L
response method becomes evident if a truly
s = 2120") is drawn and the intersection
dominant pole pair exists. In this case, all
with the locus is noted. The intersection occurs
other poles are f a r from the origin, and the
at s = -0.60 2j1.04.
transient response of the system is dominated
by the transient response mode associated
( f ) Using the magnitude condition, it is
found that K = 1.38.
with the dominant pole pair. Thus, the time
<.
5-24
GAIN DETERMINATION
W ( s )=
2.84 (s
(s
+ 5)
+6
+4
+2
-10
-8
-6
-4
-2
s(s
5-25
K(0.2~+ 1)
+ 1) (0.1s + 1)
THEORY
BIBLIOGRAPHY
5-26
CHAPTER 6
COMPENSATION TECHNIQUES
6-1 INTRODUCTION
(c) From the input specifications, the
error of the system is found when K is adjusted as in (b), and this error is checked
against the error specification.
(d) If the error does not meet specifications, a more complicated form for G,(s) is
introduced. The system is then adjusted to
satisfy the specified degree of s,tability, and
the error specification is again checked.
(e) The procedure is continued, trying different or more complicated compensation
functions, until the error falls within specifications (that is, if the specifications can be
met).
In practice, the forms of compensation normally employed are kept simple. This is due
in part to the fact that the fixed elements are
usually limited in their range of linear operation, and the introduction of complex compensation functions merely reduces the range
over which the linearity assumption applies.
In addition, it is found that the theoretical
advantages that may accrue with complex
compensation are not realized in practice because the theoretical model no longer fits the
physical system.
61
THEORY
6-2.1 GENERAL
G,(jo)= K
aTcjw
a>l
+1
(6-1)
- 8 -0
-10
-I2
-14
0
-180.
-90.
-2
-8
-10
0
-2
*"
-
-4
0
c
-6
-80
-70
-60
-50
ANGLE C,
-40 -30
-20
-10
IN DEGREES
8.
6-2
-8
COMPENSATION TECHNIQUES
T,= 0%
A lag function with a = 1 U is used to compensate the system. The lag function is to
contribute -5' of phase shift at the uncompensated resonant frequency. Design the compensation when 10 log,, M, = 1.5 dg.
Solution. The frequency response G,(jo) is
plotted in Fig. 6-4 as Curve A. For the specified M,,(see construction), the point of tangency of the M,,contour with Curve A occurs
a t the point where aR1
= 2.4 rad/sec; from
the displacement downward of the M,,
contour by 4.1 dg (i.e., 10 log K = 4.1) one
gets K = 2.57 for the uncompensated system
when G,(jw) = K. From the a = 10-plot of
Fig. 6-3, -5" of phase shift occurs at 0, = 10.
Therefore,
(6-3)
$ = 4.170
6-3
0 is, therefore,
(6-7)
THEORY
= T,o
( 6-9 1
In using the lead function to reshape an openloop frequency response, advantage is taken
of the fact that the lead functi* exhibits
positive phase shift. Thus, by adjusting the
time constant T,,it is possible to add positive
phase angles to the fixed-element response
Gf(jo) in a region where the negative phase
shift of the fixed elements is too great to secure
an M,-contour tangency. Hence, the lead func-
0
6
d
- 4
8
-I
0 2
I0
-0
-
8
6
-I
0 4
2
0
14
12
10
- 0
'
u
0
- 4
0
0
10
20
30
40
50
80
W U E G,, DEGREES
6-4
70
00
COMPENSATION TECHNIQUES
M,
(b) The first-trial choice of the lead-function time constant T,is found by determining
the frequency at which the following angle
relation holds :
(6-12)
C G,(j~l)
= -180" + 'II, - +m
The frequency o1 which satisfies Eq. (6-12)
be found directly from the gain-phase plot of'
G,( i w ) .Then, the first choice of T,is given by
--
T Cl
@Ad
.4
8
4
2
(6-13)
0 1
where Om is the frequency at which the maximum phase shift (&) occurs on the normalized lead-function curve of Fig. 6-5. The frequency B,, can be found from the relation
Bm
=-
(6-14)
dy
-240. -220.
-200.'-K)O.
-160.
-140.
-120.
L G -
6-5
-100.
-189.9"
(6-15)
fl = 0 . 0 4 4 ~
(6-16)
Compensation Adjustment
Tc (set)
(of,-(rad/sec)
0.044
8.2
6.68
0.020
5.6
8.65
2.4
2.57
No compensation
A rough picture of the effect of compensation can be obtained if the magnitude asymptotes are used in conjunction with the phasemargin criterion for the degree of stability
(see Par. 5-6). Using a 45' phase-margin criterion, the asymptotic method gives good results provided there are no low-damping-ratio
quadratic factors in the open-loop transfer
function C ( j ( o ) / E ( j c o ) . If the 45" phasemargin criterion is assumed, then the asymptote crossover frequency tov (defined in Par.
COMPENSATION TECHNIQUES
T --Pm
'l-
(6-19)
01
6-7
THEORY
phase shift, and a = 10, p4 = 10 from Fig.
6-3. Therefore, using Eqs. (6-17) or (6-18),
T, = 4.8 see. The scale change from p to w is
given by
(3 = 4.80
and asymptote curves for the lag-compensated system are drawn in Fig. 6-7, and the
composite phase-angle curve is drawn in Fig.
6-8. Using the 45" phase margin, the magni= 1.74 rad/sec,
tude-crossover frequency or,,,
and the corresponding gain is 20. The asymptote-crossover frequency o, = 1.74 rad/sec
and the corresponding gain is 17.8.
(6-20)
+IOdg
-t
--
0 dg
5
._
2 -IOdg
0,
-20dg
-30dg
-40
0.1
0.2
0.5
I0
20
50
0-
Fig. 6-7 Magnitude curves for lag-compensation procedure employing phase margin.
Fig. 6-8 Phase curves for lag-compensation procedure employing phase margin.
6-8
I00
COMPENSATION TECHNIQUES
Gf(jO1)= -189.9"
= 0.0440
(6-21)
+I0
--t
- -I0
0
-t
(6-22)
-20
-30
Fig. 6-9 Magnitude curves for lead-compensation procedure employing phase margin.
Fig. 6-70 Phase curves for lead-compensation procedure employing phase margin.
6-9
THEORY
= 7.0
the magnitude-crossover frequency o,,,,
Compensation
Adjustment
K for w,.,,,
No compensation ;
true magnitude
used
2.34
No compensation ;
asymptote used
Lag compensation ;
true magnitude
used
1.74
20
Lag compensation ;
asymptote used
Lead compensation ;
true magnitude
7.0
used
6.02
Lead compensation ;
asymptote used
(2t3*4)
COMPENSATION TECHNIQUES
G,(.i(t))H, ( j ( u )
then
C (j(o) A(jcl))
1
GI( j(u)
==
1 iG , ( j w ) H ,( j c o )
H,(jw)
(6-24)
If
G , ( j ( t l ) H ,(;(I))
(6-23)
>> 1,
<< 1 ,
then
(6-25)
< (0 < o1
iul<
(I),,
< LII,,
< (u <
when G, (jcu)H, ( j w )
<1
(6-26)
when G , ( j w ) H , ( j o ) > 1
(6-27)
when G , ( j w ) H , ( j w )
<1
(6-28)
6-1 1
THEORY
(6-33)
C,(S) = K
(6-34)
. .
KCS
(Tn
(6-35)
.
I
Gi(s)= ~ ( 0 . 3 +
s 1) (0.1s + 1)
+ 1) (0.3s + 1) (0.1s + 1)
(6-32)
Feedback compensation is to be used to improve the performance of the system in conjunction with a pure gain cascaded with the
6-12
COMPENSATION TECHNIQUES
dg
-10
dg
As a result of the application of the gainphase plane construction, the true magnitude
curve of C(jo)/A( j t o ) is shown in Fig. 6-14
can be obtained. The corresponding phase
angle curve appears in Fig. 6-15. The shape
of the true magnitude curve shows no severe
resonance effects so that acceptable closedloop performance may be expected. Using the
45" phase-margin criterion to adjust the cascade compensation G,.(s) = K , the magnitude= 8.6 rad/sec and the
crossover frequency o,.,,,
corresponding gain I; = 40. Thus, the performance is quite good. (Compare the results
of cascade compensation for this same system
in Pars. 6-2 and 6-3.) The only drawback to
the design is that the equivalent cascade lag
effect is at a fairly low frequency. This would
produce somewhat excessive peaking in the
transient response of the system which would
be followed by a long transient tail. Improvement in performance could be achieved by
further trial, e.g., by decreasing the feedback
compensation time constant T,, and attempting to increase the minor-loop gain K,. The
resultant system would then have a more
acceptable tran s ien t behavior, but the
magnitude-crossover frequency w,.,,, and the
major-loop gain K might be reduced. However, only further trial-and-error analysis
would show what actually occurs.
-20dg
-30dg
0.01
I11111
I I I 0.05
0.1
0.02
I I I111111
0.2
0.5
10
20
6
,
-
6-13
50
100
THEORY
* 2 0 dp
+I0 do
Odo
-10 dq
-20 dp
-30 dp
-40 dp
0.01
0.02
0.05
0.1
0.2
0.5
10
20
50
100
GJ-
RADIANYSECOND
-90.
- 120.
-135.
-210.
-240.
-270-
0. I
0.2
0.5
I
I'
I0
20
50
__c
6-14
100
COMPENSATION TECHNIQUES
6-1 5
<
THEORY
E(s)
then
Q(s)
(6-37)
Ei(s)
aTs+ 1
where
T = R2C,and
a = l + - R,
R2
E,(s)
--
E,(s)
_I, a T s + 1
a
Ts+l
(6-43)
where
T=
(6-41)
(6-42)
R1R2 ) C , a n d
R1+ Rz
a=l+
6-1 6
R,
-
R2
(6-44)
(6-45)
COMPENSATION TECHNIQUES
(2) Synthesis: Given a desired data-frequency (d-c) network, what is the equivalent
carrier-frequency network ?
If H ( j o ) is the frequency response of a
carrier network, the frequency response of the
data-frequency equivalent is given by
(6-46)
where
A = IH+I cos ( L H + ) IH-I
COS(
LH-)
(6-47)
B = - IH+I sin
( LH,)
+ IH-I sin ( L H - )
(6-48)
V ( t ) = e ( t ) coso,t
where e ( t ) is the true data signal, and o, is
the carrier frequency. Networks which are
designed to operate on the data of czrriermodulated signals are called a-c or carrierfrequency networks. If i t is necessary to compensate a system employing carrier-modulated signals a-c networks are required since
d-c networks will not work because they effectively operate on zero-frequency-carrier signals.
There are two questions involved in treating
the compensation of carrier-modulated signals :
(1) Analysis: Given a network which
operates on a carrier-modulated signal, what
is the data-frequency (d-c) equivalent network?
q d
B
= tan- A
H- 4 H Ej(%
(6-49)
-Wd)I
(6-51)
= data frequency
R1
B.
A. LAGNETWORK
6-17
LEAD NETWORK
THEORY
(6-52)
L H + = - L H-,
(6-53)
+ l
13 -+2
2jol
(6-54)
CARRIER-FREQUENCY EQUIVALENT
DATA-FREWENCY ELEMENT
Z jodL
u 2
,
"
,
6-1 8
COMPENSATION TECHNIQUES
The major difficulty in using resistancecapacitance notch filters is that they must be
tuned by successive adjustments of several
circuit elements; otherwise, high-precision
elements must be used.
H . 3 MECHANICAL DAMPER
T=
(i %)+ 4 (i
5)+
On
sz
j 0 d
iod 20,
(6-61)
Jm+Jd)
(6-62)
Jm
T, = motor torque
TI,= load torque
(6-59)
= RC
Jm)
Jd
J,
where
1
B(Js+
a=l+
(6-58)
1
Jd(Js
(6-60)
THEORY
The disadvantages are :
(1) damper must be designed and built
for each specific application.
(2) peak acceleration of the damper-motor
combination is reduced relative to that of the
motor alone because of the added apparent
inertia produced at the motor shaft by the
damper mechanism.
r SHELL
J.
SERVOUOTOR
WTPUT SHAFT
I TL
x., OUTPUT
FOLLOW-UP SLEEVE
POWER
PISTOW
'"'
El,
-PILOT
VALVE
INPUT
6-20
COMPENSATION TECHNIQUES
b
a
0
(') = -a ( a - 1)
xi(s)
where
(6-63)
K,
Kl
(6-64)
a=l+---=-
and
x, = flapper motion
l + Ts
( 1+ a T s )
(6-65)
T = B/(K, K I )
= actuating signal
66.5 PNEUMATIC
(See Par. 13-74
The general schematic of a typical pneumatic controller is shown in Fig. 6-22. In this
figure,
r = motion of set point (reference input)
c
= motion of pen (controlled variable or
output)
Ic
PNEUMATIC
FEEDBACK
NETWORK
'Ib
6-21
PILOT
RELAY
p,
THEORY
(6-68)
where
T R hRxCx
RI1 = integral resistance
CB = capacitance of tank
Kfb= sensitivity of proportional bellows
(6-67)
AIR
SUPPLY
=il
1-1
X1,
II
I
LOIMNCSPRHC
It
11
RELAY
OUTPUT, P,
A
B
C
D
AIR
SUPPLY
---
RESET BELLOWS
PROPORTIONING BELLOWS
RELAY
SUPPLY RESTRICTION
6-22
COMPENSATION TECHNIQUES
(6-72)
(6-69)
(6-73)
where
a],= 1
+ KlknlkprKlb
A
!
(6-74)
- l + L
A,
Rd = derivative resistance
c d = capacitance of tank
Ad = area of derivative bellows
A, = area of proportional bellows
K l b = sensitivity of proportional bellows
The transfer function of this controller is
b
(6-70)
B
C
D
E
--
PRWORTIONINC
BELLOWS
DERIVATIVE
BELLOWS
RELAY
RESTRICTION
VARIABLE
DERIVATIVE
RESISTANCE
6-23
THEORY
(greater than 50), then b > > 1, and the response is given approximately by
1
(KlknJkprKJb)
is very high
b
(greater t h a n l o ) , t h e response of t h e
proportional-plus-derivative controller is approximately
If the product-
Pe (8)
(s) e (1 - K j )
T , s + 1 1)
KIKJb
($s+
mz
- (1 - K l )
( 1 + Tas)
e (s)
KIKJb
(6-78)
(6-77)
BIBLIOGRAPHY
3 G. S. Brown and D. P. Campbell, Principles of Servomechanism, pp. #195235, 262-292, John Wiley & Sons, Inc.,
New York, N.Y., 1948.
4 H. Chestnut and R. W. Mayer, Servomechanisms and Regulating System
Design, Vol. 1, pp. #245-290, 327-357,
John Wiley & Sons, Inc., New York,
N.Y., 1951.
6-24
COMPENSATION TECHNIQUES
6-25
CHAPTER 7
PERFORMANCE EVALUATION*
7-1 RELATIONS BETWEEN FREQUENCY RESPONSE
AND TRANSIENT RESPONSE
approaches a straight line whose slope corresponds to the maKnitude of the dominant
time constant. T h a t is, if the dominant transient component is
7-1.1 GENERAL
c,, ( t )
Ac I / T ,
then
log, P , , ( I ) -- log, A
(7-2)
(7-3)
A n extrapolation of the straight-line aaymptot(. of log, I ' ( , ( I ) hack to zero time yields t h t
lomrithm of the ;tmplitudc A of t h e dominant
transient component. Thus, the dominant
transient component is completely determined
:ind can tw srrt)tract~d
from P , ( I ) . A plot of t h e
logarithm of the ~ I ~ I - ( w I I ~IfB,(/)
(~
r*,, ( I ) I
versiis time / protlrices :L curve which ; ~ p riro:rchos ;I s t r : ~ i ~ h t - l i n ;rsymptote
e
whose
4 o j w corrwpontls t o thc t ime constant of the
c*xponc*nl
i;il componclnt having the nc.xt sm:tllfir limci consl:mt. ISxtr:rpol;ttinK t h i s crirve
h c k to w r o time- yic~ltlsthe 1oK:irithm o f t h e
:impIitiidvof t h stw)ntl;try
~
componwt, r', , ( / ) .
Nfuxt, thv func*tion I I ' , ( / )
P,,(/)
I',:(/)]
is tlt4crminecl. m t l t h r procws c:in tiv rcpc~:it~*tl
rrnt i l thf* limit of rnc:is~irt~mrnt:iccuracy is rcwhctl.
'IT
Th;tl-l,;Irsfw
Kivos :I met hod for t l d c q - m i n i n g ;Ipproxim:ilc t r:insftIr fiirict ions I):tsc~tl
on the :i~~~iroxim:ilion
o f :I nonoscill:itory stty
rwporisc- I P V t hv t r:tnsfcr friricl ion
I',(t)
/:(-A)
I.(/.)
(7-1 )
whr.rr* I' ( 7 , ) is ttw final valrrr* of the o r i t p r i t .
I' ( 1 ) is thr. str.p r w p o r ~ s f and
~ , 1', ( I ) is thc
('(s)
w(s)
(7-.1)
I' I,?
____-
(s
" 1 : I,.
~ A . f;o~tld
7-1
I I ) ( T , s I I ) fT,s I I )
THEORY
where T2and T3 are dimensionless time constants and tois the dead time.
Example. Let
tl = 0.97 sec
- 3.00
1-0.6-T
1-0.5-T
1-0.4-T
1 0.3 -T
1-0.2-T
1-0.1 -T
7-2
PERFORMANCE EVALUATION
- T : 1,
0.4, 0.135
( 3 ) Curve 1 - 7: I , 0.520
( b ) Entering Fig. 7-2 with the dimension-
( I ) Curve 1
I .755
( 2 ) Curve I
1.790
0.4
T for T
T for T
0.275:
0.135:
( 3 ) Curve 1
T for T 0.520: 1.890
( c ) Ihc time ( t 3 - t , ) froni the acliial
transient divided by t h e dimensionless time
( t 3 - t , ) I , yields t h e conversion factor I, by
which t h e relative time constants found in
the first step must be multiplied lo oblain the
I-U.b-T
1-0.5-1
1-0.4 -1
1-0.3-1
1-0.2-1
1-0.1 -1
D I W O N L E U TIME CONSTWT T
0.2
0.4
0.6
0.8
DIMENYOMESSTLUE CONSTANT T
1.0
11. 1955. from article entitled Freauency Response from Experimental Nonarcillatory Transient-Raponse Data. by H. ThalLawn.
7-3
THEORY
(d) To check fo r the necessity of a deadtime factor, enter Fig. 7-5 with the dimensionless time constants T from the second
step to determine t P / t 1 .
(1) Curve l - T - T T o r T = 0 . 2 7 5 ;
1.075
(2) Curve 1 - 0.4 - T for T = 0.135;
1.055
tl
e-0.080.
(3) W ( s ) =
fl.2668
(1.850s+l) (0.963~+1)
7-4
PERFORMANCE EVALUATION
rectangular
impulse reslroirse is divided h t o equal iiikrvals. lhrn a rectangular -pul.w apyroxirriation is r r d e to the impt.~lsc.rtrsponse I Fig.
7-7). I f 1 , is the cc!nter of the nth time Jiit.t~tval, c,, -the value of the impulw ~(~spoiistr
at
t,,, and At --the
length of the tinic iiittrrval,
then the frequency ~ ~ S ~ J isI given
I S ~ try
( 7-6)
Seamans e t
tit
a I . ( ~ ~ - p Y ~
w,.
1.0
c,, l t t r
I
2:
,I
t 7-5)
,/,
w (it,,)
-t
f-
0.5
0
0
1.0
20
to impulse
Fig. 7-8 Triangular approximation to time function.
7-5
THEORY
cubics, etc.) . The approximate function c* ( t )
is differentiated enough times ( n times) to
dn
make-c*
a sequence of impulses. Actually,
L i n ~ i l I ( ~ ~has
.~*
proposed
)
a method for improving the foregoing approximation procedure. In this method, an investigation is made
of the effect of varying the position of the
approximate poles and zeros on the difference
between the actual function and the first approximation obtained from fitting the asymptotes and t h e i r corresponding correction
curves. For example, if
dtn
F ( w ) = 10 loglo I G ( j w ) I
'(7-8)
(7-9)
and
where a, is the magnitude of the kth impulse,
t,, - the time of occurrence of the kth impulse,
and n - the total number of times the original function has been differentiated.If rational approximations are sought for
an experimentally derived frequency function
W ( j w ) , advantage can be taken of the fact
that the plot of 10 log,, 1 W ( j w )I vs logo is
easily representable by straight-line asymptotes having slopes of f 10n dg/dec ( n = 0,
1,2, . . . ) . By combining the straight-line approximation of the magnitude function with
the first- and second-order response curves
given in Par. 5-3 (Figs. 5-7 through 5-10),
curve fitting is possible. The easiest procedure is to use the magnitude curves to get a
rough approximation and then to refine the
approximation with the phase curves.
-At
G ( s ) = s' - 20,s
s" - 20's
+ +
012
w12
022
w22
+ +
(7-10)
+aF
Aw2
(7-11)
am,
7-6
PERFORMANCE EVALUATION
-4
-5
-6
d
I
-7
-e
-9
0.2
0.4
0.8
0.6
1.0
- 0,
20 LOG 116) - ~
, l
- 0.8
-1.6
-2.4
-1' a'
-3.2
-4.0
-4.8
-5.6
-6.4
-7.2
-8.00
~
12
16
20
24
aJ
-
-or
B. DtPtNDENCE OF COCARITHMC C A N
ON MOTION O F A REAL ZERO
F,
ZOLOG IbJ-u,l
7-7
THEORY
0
-
0c
F,
20 LOC
20
18
16
14
I2
10
0.
w
wc
20LOG :(Jw)'-
20,(]4
+*,I+
7-8
I .4
PERFORMANCE EVALUATION
7-9
THEORY
(u
ac
7-10
PERFORMANCE EVALUATION
7
G. DEPENDENCE OF PHASE ON REAL ZERO (OR POLE)
G, = ( I U - u , )
I2
16
20
24
-0.
7-1 1
THEORY
7-12
IO}
PERFORMANCE EVALUATION
3"
-.
Id
G,
111' 12
- 10<lJd+ *:
&?I
7-13
THEORY
7-14
PERFORMANCE EVALUATION
I000
5i
W
w
8a
a"
I
'9
0.6
10
12
1.4
1.6
2.0
0 5
- 2o,(jo)
+ o>+ acq
7-15
2.2
THEORY
6J
0.
[(id'
- 7-,tj6~) + oe2+0,'I
7-16
PERFORMANCE EVALUATION
e,
=0
(7-16)
1
el = (7-17)
It is desirable to be able to describe propK"
erties of the transient response of a. system
1 - 452
when one is given the closed-loop pole-zero
e2 = (7-18)
On2
configuration and vice versa. Usually, the designer is given specifications for some form
The error response curves for a unit-ramp
of the transient response of a system. As a reinput are given in Fig. 7-11. Note the steadysult, the conversion of the transient-response
state error for a unit-ram?; input to this sysspecifications to a desired closed-loop poletem is given by
zero configuration is a starting point in many
design procedures [see Par. 6-5 and refer(7-19)
e n c e ~ ( ~ ~Since
~ ~the~ usual
~ ~ assumption
~ ~ ~ ~ ~ ~ ~ ) ] .
in these design procedures is that the closedThe error response curves for a unit-step
loop performance of the system is primarily
input are given in Fig. 7-12. The output recontrolled by a dominant pair of complex
sponse can be obtained from these curves by
Roles (dominant quadratic factor in the desubtracting them from unity. The solution
nominator), only the characteristics of a n
time or settling time t, of the step response is
underdamped second-order system are prethe time for the output to reach 98% of its
sented here.
final value or for the error to fall to 2% of its
If the system being examined is a unityinitial value. For the second-order system,
feedback system with a pair of complex4
conjugate poles and no closed-loop zeros, the
t,= (7-20)
closed-loop transfer function relating output
Ton
to input is
The output response curves for a unit-step
input are plotted in Fig. 7-13.
wl?
W ( s ) =--c ( 8 ) (7-12)
s2 + 250.8 + on*
R(s)
The error-to-input transfer function is
1.75
E ( s ) - s (8 2bn)
f I.50
(7-13)
S' + 2<wns
wn2
R (s)
The open-loop transfer function is
3
J
1.25
8
g
1.00
In these equations,
on= natural frequency
and
,T = damping ratio.
25
am
g 0.25
0,
a75
(7-15)
4
6
e
10
DIMENSIONLESSTIME, ant
12
14
7-17
THEORY
Quantitative descriptions of the relationships between properties of the transient response and the frequency response of a
second-order system will now be given. The
resonant frequency of the closed-loop response W ( j o ) is
OR=%d/1-2252
(7-21)
iu,=
(7-22)
2c dT=p
2"t
(7-23)
O d = o n g / l - y
(7-24)
3.0
2.5
g-E
2.0
II
i
1.5
1.0
0
0.1
0.2
0.3
0.4
DAMPING RATIO,
2
3
4
DlMENSlONLESS TIME, ont
0.5
0.6
0.7
b G.
7-18
PERFORMANCE EVALUATION
of the dominant pair of poles.(8 6 6'12.18.10.28)
The pertinent relations are as follows :
'
&
pO D -- c v1-t'
(7-25)
If bandwidth is defined as the frequency wb
at which 10 log,, I W ( j w )I is down 1.5 dg from
the zero-frequency value, then
@b
- 2c2+ d2 - 4c2
+ 4c4]"2
(7-26)
I00
00
8'2
60
B
c
3 40
YW
e, = 0
20
42
a4
46
as
(7-27)
1.0
5.
(7-29)
THEORY
e, = 0
(7-30)
el = 0 (K,+m)
(7-31)
From the characteristics of the secondorder system and the general relations (a)
through ( n ) of the preceding paragraph, the
conversion of time-domain specifications to
a closed-loop pole-zero configuration becomes
a fairly straightforward matter. Truxal(28)
presents a very good description of a typical
procedure.
(7-32)
where KO is the acceleration constant, - p j
is the jth pole of W ( s ) , and -zj is the j t h
zero of W (s).
z-
tbu
( c ) K, 2 50 sec-l
(7-33)
%o
2= 0.30 to 0.45
(7-34)
=- 1
(7-35)
Kv
(m) If the rise time t, is as defined in (k),
(7-36)
3tbu to 5tbv
(7-37)
7-20
<
PERFORMANCE EVALUATION
-1 -K,,
1
-+
Pi
1
Lo.+
jo.dl-S2
1
1
-on- jondl- 5'
z1
(7-38)
Therefore, z1 = 3.40. The desired pole-zero
configuration for W (3) is given by
W ( s )=
2520 (s
(sz
+ 3.4)
(7-39)
( 7-40)
( ,",) .
Since most of the design techniques discussed in Pars. 6-2,6-3,and 6-4involve considerations of the open-loop frequency re/ E ( j w ) = G (jo), methods for
sponse C (jo)
relating the open-loop frequency response to
the closed-loop transient response will be presented here.
Harris et al.(') present an approximate
technique for determining the error response
e ( t ) to a transformable input r ( t ) . If o, is
defined as the frequency at which the openloop asymptotes cross 0 dg (asymptote crossover frequency; see Fig. 7-16), this method
7-21
THEORY
s = JW,
PIp
occurs.
wu
0 1
--<t<-
w,to
(7-42)
OPEN-LOOP TRANSFER
FUNCTION G
a\
W
OL
I
a
z
4
ol= 0
a, RADIANS PER SECOND
4Z
Om
1.0
1. SECONDS
0,
to the frequency
wc at which the straight-line approximation (asymptote) of the
open-loop response is 0 decibels.
(Note: 2 decilogs = 1decibel.)
7-22
PERFORMANCE EVALUATION
<< w,
(7-43)
(low-frequency
(7-44)
for 1s.l
< o,,and s1
so
(ad"
Sb
(7-47)
where
= 0.04
@ = 0.2
~1
03x1
= 16
KOis a proportionality constant whose value
is to be determined. The asymptotes of this
04
function are sketched in Fig. 7-19.The crossover frequency o, is chosen as the geometric
mean of o3and o4since this particular choice
tends to produce the lowest closed-loop M,,.
Near o, the asymptote is given by
(7-46)
.=.,
- c(s - sa)"G(S)]
for 18)) > O,, and 8 2
K O
--1
or
7-23
o.942
KO= ~
=4042
~ 0 4 2 do3w4 04*
THEORY
2.4
2.2
2 .o
I .4
1.2
I .o
7-24
PERFORMANCE EVALUATION
2.4
22
2.0
1.8
1.6
1.4
1.2
1.0
0.0 1
0.10
4%
7-25
1.0
THEORY
I.o
I .4
I .2
1.0
2
'3
g 0.8
K
*
u
Z
5 o.6
Y
0.4
0.2
0.10
0.01
O,/%
7-26
1.0
PERFORMANCE EVALUATION
1.0
7-27
THEORY
7-28
PERFORMANCE EVALUATION
I.4
1.2
1.0
a'
e
tI 0.8
-----
0.6
Y
0.4
0.2
0I
I1111
I 1 1 1 1
1 t I Ill
0.I
0.01
"I/"*
7-29
I .o
THEORY
0.1
0.01
Wl'%
7-30
1.0
PERFORMANCE EVALUATION
2.4
2.2
2.0
I .@
5
I .6
1.4
1.2
I .o
0.1
0.01
7-31
THEORY
7-32
PERFORMANCE EVALUATION
.-
1.0
0.01
7-33
THEORY
2.4
2.2
2 .o
I.8
5
1.6
I.4
1.2
1.0
1.4
0. I
'
--
"1 to 90 db/&cado
"3 to
60 d b / k o d .
---
1.2
1.0
u :
2K
0.8
tiZ
W
0.6
W
K
Y
a 4
0.2
0. I
0.01
Q+=
7-34
1.0
PERFORMANCE EVALUATION
7-35
THEORY
7-36
PERFORMANCE EVALUATION
7-37
THEORY
7-38
PERFORMANCE EVALUATION
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.01
I.o
0.1
4%
0'
I 1 1 1 1
0.01
I I I l l
0.I
I I I l l i
1.0
'1/4
7-39
THEORY
0.01
0.1
m,fme
7-40
PERFORMANCE EVALUATION
7-41
THEORY
Inspection of Fig. 7-19and Eq. (7-48) shows
that the low-frequency zeros of G (s) occur at
--02
and -w3 and that the high-frequency
poles of G(s) consist of a double pole at
-a4. Therefore, the poles of the first approximation to C ( 8 ) / R (s) occur at -w2 (singleorder), -w3 (single-order) ,and -a4 (doubleorder), plus one pole at -wc. The zeros of
C ( s ) / R ( s ) are the zeros of G ( s ) , --oz and
-ma. As a result, the first approximation to
C ( s ) / R ( s ) is
c (s)
-Y
R ( s )-
(8
(s
(0,)
0,)
(s
s = -0.2
E(s) R(s)
(8
s(s
0,) ( 8
+ w1I2
+ (s +
02)
02)
(7-54,
This quantity is reasonably large so that a
second approximation to S3 is made by evaluating the right side of Eq. (7-54) at s =
(-1) + (-0.253), instead of at s = -1,
yielding
'G"& 1'
+ G ( s ) = (s + 0.2)(s + 1) ( 8 + 4)
(s
+ w2
+ 0.0016 = -0.1984
- -0.253
= 0.0016
(7-49)
The first approximation to E ( 8 ) is found to
be
-Y
=-
(7-53)
Since this quantity is small, the true pole
lies at
m442
KO
(7-51)
s=-1.53
-0.373
(7-55)
Theshift is still not too well approximated
since the change from Eq. (7-54)to (7-55)is
significant. The third approximation to a3 is
obtained by evaluating Eq. (7-55) at s =
(-1) + (-0.373), instead of at s = -1.253,
yielding
PERFORMANCE EVALUATION
-(s
+ 16)ZG(s)]
= 59.6
.=-16
(7-57)
or
64 = k7.7
+ 16)2G(s)]
= k6.4
.=-23.1
(7-58)
+
+
C(s) -(s
R(s)
1014 (s
+ 0.2) (s +1)
(7-59)
(3) If o, is the asymptote crossover frequency, the tail of the response is approximately the tail of an exponential with time
constant T, starting from l/oc at t =O and
falling to l/Kc at t = 00, where K , is the velocity constant.
(h) The maximum time delay by which the
output response to a unit-ramp input lags the
input ramp is approximately equal to the rise
time (0.63-value) of the output response to a
unit-step input.
7-43
THEORY
Biernson ~ 3 has
)
suggested th a t the foregoing
difficulty can be resolved by examining
r , r', r",. . . for discontinuities and subtracting these discontinuities from the corresponding functions. The remaining functions will
all be continuous, and the expansion can be
applied over the entire time range of interest.
Then, the effects of the discontinuities in
r, r', r",. . . are added to the response. In this
procedure, a discontinuity in a function is considered to occur if the function rises ( o r decays) more abruptly than the corresponding
transient response to the discontinuity. In
comparing the rise rate of the two curves, a
convenient criterion is to compare the times
fo r the two curves to reach 63% of the initial
rise (or decay) of the curves.
A convenient procedure f o r determining the
error coefficients required to satisfy performance specifications is the following:
e ( t ) = e,r(t)
+ e l r ' ( t ) + e8r"(t) + . . .
(7-60)
+ +
e l s e2s2 . . .
(7-61)
R(S)
This expansion is valid everywhere except
where the input o r any of its derivatives are
discontinuous. Fo r practical purposes, only a
few terms of the expansion are used to evaluate the error response. However, the expansion cannot be used near points of discontinuity of r , 6,r", . . . if accurate results are
sought. Thus, f or example, if a step discontinuity occurs in the input r ( t ),the expansion is
invalid f or a time interval extending from the
instant toat which the step occurs to the time
( t o+ t J , where t, is the settling time of the
transient error response to the step (time for
the error transient to fall to 2% of its initial
value). Obviously, the step can be ignored if it
is small compared to the remaining terms of
the expansion in the interval t o< t < ( t,, t,) .
-E ( s ) - eo
7-44
PERFORMANCE EVALUATION
Once the error coefficients have been specified by means of a procedure similar to the
foregoing, the design of the system can be
carried out as follows :
(a) The system is designed to meet all other
specifications on transient response and frequency response such as bandwidth, M,,Po,,
etc.
(b) The error coefficients of the system
designed in (a) are found in terms of the
system parameters and the gain.
(c) If any of the error coefficients found in
(b) exceed their specified values, they may be
reduced by increasing the system gain, if
possible, by the introduction of low-frequency
dipole lead functions in the pole-zero configuration of the open-loop transfer function
G (s), or by the feedback-compensation technique suggested by King.(l)
(d) The specifications on transient re-
7 3 PERFORMANCE INDICES~406)
A performance index P is a single number
which is used as an indirect measure of system performance. Other measures of system
performance have already been considered,
such as the various commonly used parameters M,,bandwidth, rise time, peak overshoot,
etc. However, these parameters provide only
a partial description of performance since, in
a sense, only part of the corresponding response is described by each. To be sure, if
enough of these response parameters (for
want of a better term) are known, a n accurate description of the corresponding response is possible. That is, the response
parameters may be considered direct descriptions of the shape of their associated
responses. However, since a response function is continuous, theoretically an infinite
number of response parameters are necessary
to describe the response. To get around the
use of a multitude of response parameters, a
7-45
THEORY
Index
-LW
Input
Description or Name
Reference
P -
edt
Transient
Control area
1,2 i k
Pz =
cI e I d t
Transient
P3 = l m t e d t
Transient
Jo
-1
Transient
Integral-time-multiplied absolute
error (ITAE)
14,16,18
-l
e 2dt
Transient
9,18,33,34
P - -te2dt
-l
Transient
P , = P3/P12
Transient
p - wtleldt
P -
P8 = e2
1
Mean-square error (MSE)
Stochastic
34,35
BIBLIOGRAPHY
PERFORMANCE EVALUATION
10 M. J. Kirby and D. C. Beaumariage,
Relative Stability of Closed-Loop Systems, Trans. AZEE, Vol. 72, Part 11,
pp. #22-43, 1953.
11 G. Biernson, Quick Methods for Evaluating the Closed-Loop Poles of Feedback Control Systems, Tram. AZEE,
Vol. 72, Part 11, pp. #53-70, 1953.
14 D. Graham and R. C. Lathrop, The Synthesis of Optimum Transient Response : Criteria and Standard Forms,
Trans. AZEE, Vol. 72, Part 11, pp.
#273-288, 1953.
7-47
7-48
CHAPTER 8
In Par. 7-3 it was indicated that the conventional measures of performance such as
rise time, peak overshoot, solution time, M,,
etc. were merely partial descriptions of the
frequency response of a system or the shape
of a particular transient response. As a
result, an adequate description of system behavior requires a fair number of response or
performance parameters. To avoid using a
multiplicity of response parameters (numbers which describe the response such as M,,
peak overshoot, bandwidth, rise time, etc.)
attempts have been made to describe system
behavior in terms of performance indices. A
performance index is a single number which
can be used as a criterion of performance.
The pertinent performance indices used are
those directly related to system error since
error is the basic determinant of the goodness of a system. The most common performance indices are those listed in Par. 73.
When performance indices are used in system design, the usual procedure is to minimize the index if it is a direct measure of
error. With a given index one also associates
the specified input to the system. Several approaches can be used in carrying out the
minimization procedure.
In one approach it may be assumed that all
but a few of the system parameters are specified. Then, the optimization procedure involves the adjustment of the free parameters
so as to minimize the performance index.
+ B y L. A . Gould
8- 1
THEORY
First, the designer is able, through optimization, to decide whether a given set of specifi=tions is compatible. Second, the designer
can decide whether a compatible set of specifications can be satisfied when bounded by
constraints.
(8-4)
I,
q
:
y,*(t) dt
(8-1)
I , = I , (Pl, Pz, *
1
m
-J+
Y , ( s ) Y , ( - s ) ds
2x3
( 8-5 1
Pk)
-= 0 (i = 1 , 2 , . . . , k)
* *
apC
(8-3)
However, it is often better to find the minimum graphically by working directly with I,.
-m
G,(s) =
s(Tls
1
1) (Tms
where
+ 1)
OPTIMIZATION METHODS
where
co = 1
The desired output is the input. The configuration that describes the problem appears in
Fig. 8-1.
do = K ,
+ Tm2
~2
= TI2
~4
= T12Tm'
dl = 1
& = TI
+ Tm
d3 = TIT,
Z,=X]
2
- +1- - K . ,1
Tm
TI
I,=
5.25KV+ 125
Ye(s)
= E ( s ) = R ( s ) - C ( s ) ,and
-.S
NOW,R ( s ) = N4
So, by substituting the expressions given
originally for G,(s) and GI(s)into the equation for C (s) we find that the error transform
is
Ku2
+ 47.6 K , - 2980 = 0
or
K, = 35.8 or - 83.4
= 0.049 N:
As a point of interest, for K, = 35.8 sec-l, the
value of the peak magnification is 10 loglo
M p= 2.8 dg which is a reasonable value.
Another optimization criterion is presented in a series of papers by Graham and
Lathrop(3.4.5) in which they have applied the
integral-time-multiplied-absolute-error criterion (ITAE) to optimize the performance of
THEORY
standard-system forms. However, their procedure is limited to step inputs only. The
ITAE criterion is
(8-7)
J-a
c (3)
-R(s)-
+ qn-isn-l + -..+ q1 s + 1
(8-8)
HOWMYENSlONM TIME
OF DEFINITE INTEGRALS
where
8-4
OPTIMIZATION METHODS
qls
+1
R (s) - s + q,,-ls+l+ . ..
+ qzsz + qls + 1
( 8-9 )
NW4DIMENSIOMM FREWENCY
8-5
THEORY
~
(8-13)
-0
10
15
G(s) =
(s)
do
+ 1) ( T m s + 1)
=L
! (white noise)
l.3
s(T,s
G , ( s ) = Kv
@vn ( 5 ) = 0
s_',"
8-5,
@,,(s)
YeZ
Since a,,, (o)is an even function, the evaluation of the integral in Eq. (8-11) can be
carried out by means of the integral table
(Table 8-1).
In all other respects, the design procedure
for minimizing the mean-square error for
stationary stochastic inputs with a fixed system configuration parallels the procedure for
transient inputs outlined above.
NONMMENBONM TIME
GVv(s)
(8-11)
8-6
OPTIMIZATION METHODS
T,
= 0.01 S ~ C
Tm
= 0.04sec
all
= 10 milliradian/sec
= 0.1
Yn
(sec)-1
where
K , = 90 sec-l
where
:"I?[
0.095 milliradian
u = T/v = 0.001
From Eq. (8-12) the power spectrum of system error expressed in terms of u is found to
be
K , = 7.8 sec-l
[Z]'':2.21 milliradian
8-7
THEORY
tem (r-lrW(s)l)
(t) = autocorrelation function of r ( t )
#,J ( t )
(8-14)
=
0 for tl
(tl)
J -cn
. (8-16)
g, ( t)
r(s)= 2nG,(-~)
&(s)
@w(s)
In I
1
Fig. 8-6 Configuration for
I
M E
minimization
1-
-k
OPTIMIZATION METHODS
- A2 = lim (E
- A)
(E
+ A)
E+O
@"*(S)
= 'YI
&(8)
=o
uv = 10 milliradian/sec
= 0.1 Sec'
ya = 0.4 (milliradian)
- sec
i(t) = v ( t )
Normalize the frequency scale by letting
A=
-*
where
Then,
-m = 0.5 d 2~ 1 = 15.82
n = 0.5 d 2c -1 = 15.80
The factor of @--(A) having all its poles and
zeros in the right-half plane is
-UV2= a
v2
b ; ( m + jn - A)
*--(A) = -
andvy,=b
(-
( m - jn -A)
A) (1 - A)
Therefore,
= @,(a)
8-9
THEORY
or
Since,
W ( s )=
+ 0.613s
where
or, numerically,
c(s)=50
E (8)
'Iz
[z]
(0.613s+1)
s(l0s +1)
The mean-square error 2 can then be evaluated by using Eqs. (8-11) and (8-12)together
with Table 8-1. The r m s error due to noise
alone is found to be
[z]
[z]l/[Z]-+[z].
1'2=
or
[z]1'2
= 1.57 milliradian
= 1.35 milliradian
8-10
OPTIMIZATION METHODS
8-1 1
BIBLIOGRAPHY
2 F. C. Fickeison and T. M. Stout, Analogue Methods for Optimum Servomechanism Design, Trans. AZEE, Vol.
71, Part 11, pp. #244-250, 1952.
3 D. Graham and R. C. Lathrop, The
Synthesis of Optimum Transient Response : Criteria and Standard Forms,
Trans. AZEE, Vol. 72, Part 11, pp.
#273-288, 1953.
8-12
CHAPTER 9
Linvill states, A sampled-data control system is one wherein the signal supplied to one
or more parts of the system is not given continuously in time, but is supplied at discrete
values of the time variable, t. In such a system,
the part of the system being fed intermittently
might, for example, have an input signal applied to it a t t = 0, T,2T,3T, . . . (where T is
the length of time between samplings) with no
data a t all supplied in the intervals separating
these sampling instants. A control system
makes use of sampled data when i t is impossible to supply continuous data to all its
parts.?
+ B y L. A . Gould
SAMPLING CARRIER
SIGNAL
*
-
HOLDING
QRCUIT
9-1
MEMBER
THEORY
sample value at that value until the next sampling time is called a sampler-clamper. Its effect is shown in Fig. 9-4. This type of behavior
can be exactly represented by the combination
of a n impulse modulator a n d a filter whose
transfer function G , ( s ) is
(9-1)
I
CARRIER
CUJIVALENT
LINEAR
FILTER
OUTPUT
MEMBER
i
9-2
C(t)
or
A(t) =
+*
T ( ~ T6 0) ( t - - T )
(9-6)
n=-m
R* ( s ) =- z
T n=-m
R(s +j n ~ )
(9-7)
where
( 9-2 1
6,(t--T)
+m
~*(=
t)
2n
n= T
n=-m
(9-8)
and
( 9-9 1
R ( s ) =X[T(t)I
(9-3)
or
( 9-4 )
R*(s)=
~*(=
t )A ( t ) ~ ( t )
I:
n=O
r ( n T ) e-ITa
(9-10)
Note that a starred transform like R ( s ) represents the transform of a starred (sampled)
time function. Also from Eq. (9-7), starred
transforms are periodic functions of frequency, the period being j R. That is,
( 9-5 1
R ( S) =R (s
+j n ~ )
(9-11)
A
t0
tOtTt0+2T
+t
9-3
THEORY
expressions for z transforms should have z replaced by r1to make them correspond to the
notation adopted in this chapter.
The introduction of the z transform enables
one to treat sampled-data systems by all the
techniques available for continuousdata systems since it is evident that the process of sampling a time function can be represented by a
R* ( s ) = F ( e 4 T )
(9-12)
(9-13)
R(s)=
K,.
zs + a/,
Laplace
Transform :
F(s)
1.
2.
2"
3.
1-2
Tz
4.
(1 - Z)Z
5.
6.
7.
8.
9.
10.
A short table of z transforms and their equivalent Laplace transforms is given in Table 9-1.
For more extensive tables see references 5,
25, and 26. Unfortunately, several authors
This notahave adopted the relation z =
tion arose from the mathematics of difference
equations, but it is awkward and physically
deceiving in the present connection, since
e+XTcorresponds to ideal prediction. Therefore, when using the literature, care must be
taken to verify which particular notation is
being used. In reference 5, for example, all the
11.
12.
13.
9-4
z Transform :
F* ( z )
(9-14)
/.=I
then
TABLE 9-1
LAPLACE AND L TRANSFORM PAIRS
s2 +a2
F ( s +a)
edTF( s )
eaaF(s)
1
1
s+-i;
hoca
s2
+ a2
1
(s
+a)2
1
1-Ze-T
z(l -P T )
( 1 - 2 ) (1 - 2 e - q
z sin aT
1 - 22 cos aT + z:
F* (eT
z)
z F* ( z )
d a I T ) F* ( z )
1
1+w
1-zcosaT
1 - 22 cos aT
+ z2
T P T Z
(1 - z
e = ) 2
SAMPLED-DATA SYSTEMS
A. rPLANE
B. zPLANE
9-5
1 - W
(9-16)
l+w
is a bilateral transformation from the z plane
to the w plane, then
is defined as the w transform of r ( t ) .The advantage of introducing the w transform becomes evident when an attempt is made to
evaluate R* (s) for s = jo.Such an evaluation
requires an infinite vector sum, theoretically [Eq. (9-7)], or else evaluation of R*
through the use of e-*= = cos oT + j sin oT
and so it is fairly difficult to obtain in practice.
The use of the w transform, on the other hand,
simplifies the determination of the frequency
response of sampled-data systems. The w
transform maps the unit circle in the z plane
THEORY
into the imaginary axis in the w plane and restores some of the analytical advantages that
were lost through the sampling process. If
w=u+jv
(9-18)
= tan
(+)
(9-20)
then
v
2
tan-lv
(9-19)
The operatiorla1 definition of impulse modulation given in Par. 9-1 simplifies the study of
sampled-data systems.
9-3.2 BASIC RELATIONS OF SAMPLED
FUNCTIONS
C ( S )= G ( s ) R * ( s )
(9-21)
C*(S)= G * ( s ) R * ( s )
(9-22)
[R* ( s ) ] * = R* ( s )
(9-23)
C ( S )= G ( s ) R ( s )
(9-24)
C * ( S )= [ G ( s ) R ( s ) ] *
(9-25)
C* (s)+G* ( s ) R* (s) ! !
(9-26)
Fig. 9-9 Sampling a filtered continuous signal.
9-6
SAMPLED-DATA SYSTEMS
(9-31)
(9-30)
Example. Consider a simple servomechanism with block diagram shown in Fig. 9-10.
The physical device includes a samplerclamper (See Par. 9-l) , a servomotor having
a one-second time constant, and an ideal amplifier. The transfer function of the continuous portion [including the filter as in Eq.
(9-1) 3 is given by Eq. (9-32).
A. Gt (1) PLANE
Fig. 9-1 7 Relations between s and G*(s) for application of Nyquid criterion.
9-7
THEORY
=-m
1
(s + jm)* (1
'$
r*t = r(nT) = -
+ s +inn)
2n3
(9-33)
R* ( 2 ) ~ ' - dz
(9-39)
1
G*(z) =
t+O
z 4
1-w
(9-35)
l+w
the w transform of Eq. (9-32) is found to be
(1 - w) (0.632 0 . 1 0 4 ~ )
1
G*(w) =
K
(0.632 + 1 . 3 6 8 ~(2w)
)
(9-36)
when w = jv, the function in Eq. (9-36) is
easily handled by conventional techniques
since the relation between the w plane and the
G*(w) plane is the same kind of relation as
that which exists between the s plane and the
G ( s ) plane. If the real frequency o is to be
considered, then there is an added difficulty
in that Eq. (9-20) must be used to calibrate
the frequency locus. The asymptotic and
gain-phase plane techniques can be used with
a change only in the relation between v and o.
2=-
t+m
(9-41)
Z+l
.ha
(f) F o r s = k j - ( k
= 0, 1, 2 , . . .), G*(z)
2
isalwaysreal n = - .
(a) G * ( P T ) =
lim r ( t ) = lim
:+a
v+l
(2w ) R * ( w )
l+w
(9-42)
t+m
9-8
v*
(-) 220
l+w
R* (w)
(9-43)
SAMPLED-DATA SYSTEMS
9 4 DESIGN TECHNIQUES
The problem of designing a sampled-data
system is complicated by the fact that the
system can contain both discrete and continuous elements. In addition, direct application
of z-transform theory merely gives the response at the sampling instants, but the behavior during the sampling instants cannot
be determined by simple methods.
The insertion of a sampling device in an
otherwise continuous system to produce a
sampled-data system introduces the following
limitations :
(a) A greater tendency towards instability results.
(b) Ripple components arise in the output
at the sampling frequency and its harmonics.
(c) The usable bandwidth of the system is
reduced to a fraction of the sampling frequency R, the theoretical upper limit being
0/2.
To determine the gain necessary to stabilize the system for a specified Mp in the
closed-loop frequency response, the introduction of the w transform greatly facilitates
plotting the frequency locus as indicated in
Par. 9-3. Conventional continuous-system
techniques can be used.
A. CONTINUOUS COMPENSATION
8. DISCRETE W E N S A T I O N
THEORY
pensation. In Fig. 9-12A, G,(s) is the transfer function of a continuous network which
is used to improve the closed-loop behavior ;
G,(s) is the transfer function of the fixed
elements. In Fig. 9-12B, G,*(s) is the transfer function of a digital network (digital
program) used to improve system performance. For the case of continuous compensation, we have
E* (8)
R*(s)
1
+ CGc(s)Gf(s)1*
(9-44)
c*(2) =
0.186(z2 + 1.392)
(1 - z ) C0.5542 - 1.1082
+. .
+ 13
9-10
c*(s)
-R* (s)
+ + Ge-2sT+ . ..
+ bledT + bzcZdT+ . . .
UlPT
bo
(9-48)
SAMPLED-DATA SYSTEMS
c (nT)
0
1
0
10
12
nT --C
0
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
0
0
-0.75
-0.375
-0.562
-0.469
-0.516
-0.492
-0.504
3
4
5
6
7
8
9
0
1.5
0.75
1.125
0.938
1.031
0.984
1.008
0.996
1.002
-0.498
c*
1.6
1.52
R*(z)
1 + 0.52
If r ( t ) is a unit step, the tabular evaluation
of c (nT)can be carried out as follows. Crossmultiplying, we get
(2)
--
IA
1.2
0.0
0.6
0.4
c ( n T ) +.05c[(n-l)T-J
=1.5r[ (n- 1) TI
0.2
0
0
or
c(nT) = 1 . 5 r [ ( n - l ) T ]
- 0.5c[ (n- 1) T]
6
nT --t
10
9-11
THEORY
(9-50)
C
'
(2)
(z")
+R*G*
Gt* ( z )
(9-51)
(2")
+ G*
(9-52)
(2)
9-12
SAMPLED-DATA SYSTEMS
or
+ a2 +)$:
- Ta(&
+ .. .
I1
(9-55)
1-2
I
Tz2+ mz
(1 - z ) 2
ze-T
1 - zeo=
BIBLIOGRAPHY
1 W.
6 W.
R. Ragazzini,
Sampled-Data Processing Techniques
for Feedback Control Systems, Trans.
AZEE, Vol. 73, Part 11, pp. #236-247,
7 A. R. Bergen and J.
of a Sampling Servomechanism, J .
A p p l . Phys., Vol. 21, pp. #290-294,
1960.
3 J.
1954.
8 C.
#225-234, 1952.
4
1964.
#247453,
9-13
1954.
THEORY
11 G. V. Lago, Additions to z-Transformation Theory for Sampled-Data Systems, Trans. AZEE, Vol. 73, Part 11,
pp. #403-408, 1954.
12 E. I. Jury, The Effect of Pole and Zero
Locations on the Transient Response
of Sampled-Data Systems, Trans.
AZEE, Vol. 74, P a r t 11, pp. #41-48,
1955.
13 J. Sklansky and J. R. Ragazzini, Analysis of Errors in Sampled-Data Systems, Trans. AZEE, Vol. 74, Part 11,
pp. #65-71, 1955.
14 G. W. Johnson, D. P. Lindorff, and C.
G. A. Nordling, Extension of Continuous-Data System Design Techniques
to Sampled-Data Control Systems,
Trans. AZEE, Vol. 74, Part 11, pp.
#252-263, 1955.
trol, edited by A. Tustin, pp. #409418, Butterworths Scientific Publications, London, England, 1953.
21 C. Holt Smith, D. F. Lawden, and A. E.
Bailey, Characteristics of Sampling
Servo Systems, in Automatic and
Manual Control, edited by A. Tustin,
pp. #377-404, Butterworths Scientific
Publications, London, England, 1952.
22 R. C. Oldenbourg, Deviation Dependent
Step-by-step Control as Means t o
Achieve Optimum Control for Plants
with Large Distance-Velocity Lag, in
Automatic and Manual Control, Proceedings of Cranfield Conference, 1951,
edited by A. Tustin, pp. #435-447,
Butterworths Scientific Publications,
London, England, 1952.
23 H. Sartorius, Deviation Dependent Stepby-Step Control Systems and Their
Stability, in Automatic and Manual
Control, Proceedings of Cranfield Conference, 1951, edited by A. Tustin, pp.
#421-434, Butterworths Scientific Publications, London, England, 1952.
26 M. F. Gardner a n d J. L. Barnes,
Transients in Linear Systems, Vol. I,
pp. #354-356, John Wiley & Sons, Inc.,
New York, N.Y., 1942.
9-14
CHAPTER 10
NONLINEAR SYSTEMS*
10-1 INTRODUCTION
All of the techniques of system analysis
discussed in previous chapters are restricted
in their application to linear systems. This
restriction imposes two limitations on design.
First, components must be of high quality if
they are to operate in a linear manner when
amplitudes and frequencies of signals vary
widely. Second, the linearity restriction limits
the realizable system characteristics, the
types of systems, and the tasks that can be
accomplished.
Nonlinearities are generally of two types :
incidental and intentional. Incidental nonlinearities are secondary effects which limit performance in otherwise linear systems. Examples of phenomena that introduce incidental
nonlinearities include backlash, saturation,
dead zone, hysteresis, and coulomb friction.
On the other hand, intentional nonlinearities
are those introduced purposely to improve the
characteristics of systems or to alter them in
specified ways. The contactor (on-off o r relay)
servo is the most extreme example of such an
intentionally nonlinear system.
NONLINEAR SYSTEMS
(10-1)
then
y(t)
(ot
(10-2)
and
y,
N(X,o) = -
(10-3)
LN(X,m)
(10-4)
=+I
OUTPUT Y
INPUT X
-?
69. Part
10-2
THEORY
this contactor as functions of the input amplitude X appear in Fig. 10-3. Figure 10-4 shows
the input-output characteristics of a nonlinear
element containing both dead zone D and saturation S. No phase shift is associated with
this element since, in general, phase shift will
not occur for a single-valued nonlinearity. The
describing function for no saturation ( S + m )
is presented in Fig. 10-5. The describing
function for no dead zone ( D = 0) is presented in Fig. 10-6. Describing functions for various combinations of dead zone and saturation
appear in Fig. 10-7. Figure 10-8 shows the input-output characteristic of a nonlinear element characterized by hysteresis (backlash or
free play). The describing function for this
nonlinearity is presented in Fig. 10-9. Other
describing functions for more complex nonlinearities can be derived for the particular
case being considered. Additional describing
functions are given in the literature.(1R.*8.2')
The prbcedure for using the describing
function to predict the nature of the stability
of a nonlinear system follows. Referring to
1.5
a
KIA
e:
yF
1.0
S$
0-
ub
::
Y Z
no
3 Z
tz
.s
:;
ZB
n
.5
1.5
1.0
20
2.5
3.0
2.5
3.0
A. AMPLITUDE RELATIONSHIP
i"
K
g
-10-
<
U
-20U
:-30.
IY
3 -40J
:-so=
1.0
I. 5
6. PHASE RELATIONSHIP
2.0
Fig. 10-3 Plot of the describing function N (simple contactor with hysteresis ratio h / A ) .
Adapted by permission from Tramactim of the AIEE, Volume
69. Part I. 1960. from article entitled 'A Frequency Reaponse
Method for Analyzing and Synthesizing Contactor Servomechanisms'. by R. J. Kochenburger.
10-3
NONLINEAR SYSTEMS
I.o
I Y
0.8
N
-
0.6
k 1 0.4
0.2
By ermission from Automatic Feedback Control S y s t e m S y n t h e s k by J. G. Truxal, Copyright, 1955, McGraw-Hill Book
Company, Inc.
I .o
0.8
N
-
1.0
kl
0.6
0.4
0.8
0.2
kl
0.6
0.4
I
0
X
D
0.2
Fig.
10
X
D
By permission from Automatic Feedback Control S y s t e m S y n thesis, by J. G. Truxal. Copyright. 1955, McGraw-Hill Book
Company, Inc.
Adapted by permission from Automatic Feedback Control System Synthesis. by J. G . Truxal. Copyright, 1955, McGraw-Hill
Book Company. Inc.
THEORY
OUTPUT y
-180.
INPUT x
,' 0'
,'/'
1.0
RATIO WX
10-.5
NONLINEAR SYSTEMS
(corresponding to instability), and the amplitude of oscillation will tend to increase. Thus,
oscillation cannot be maintained at A. However, point B is a point of convergent equilibrium, as can be determined by letting the
amplitude X both increase and decrease relative to Xn.rn each case, the tendency will be
for the amplitude of the oscillation to head
back to point B. Thus, the convergent equilibrium point B determines an amplitude X e
(read off the - l / N locus) and a frequency
wn [read off the G(jw) locus] a t which a sustained oscillation occurs.
Case 3. The G(jw) locus intersects the
-l/N locus a t point C. This is a convergent
equilibrium point, as can be determined by
letting the amplitude X increase o r decrease
relative to the amplitude X, associated with
the intersection C. In each case, the amplitude
of oscillation will tend to return to X,.
(10-5)
where w,, is the frequency associated with the
point of M, tangency for each value of X
along the -1/N locus.
These methods can be extended to determine the entire frequency response of a system by noting the intersections of other M
contours with the G locus at each value of X
and using the following relation$ to determine the input amplitude (or amplitudes) R
associated with each value of X:
The describing-function procedure is useful in predicting the closed-loop frequency response of a system containing an incidental
nonlinearity when no oscillation can occur.
Thereby, peculiarities in measured characteristics can be explained, and quantitative
10-4
THEORY
Example. A relay (contactor) servomechanism employs a relay with a ratio of hysteresis h to dead zone A of 0.5. The describing
function of the relay is plotted in Fig. 10-12
as a function of the normalized input amplitude a, where a = X / A . The linear element
GI (jo)is a pure gain K (see Fig. 10-1). The
linear element G2(jw) is represented by the
relation
G,(jo) =
1
jo(0.05jw
+ 1)
IEl =
-180.
-160.
-14V
-lZO*
+00*
a
-
[K/AI
and
K / A = 142
For the stable case, curve ( B ) of Fig. 10-12
and Eq. (10-5) are used to determine the
variation of M, with the magnitude IRI of the
sinusoidal input. The results are plotted in
Fig. 10-14.
-1
Fig. 10-13 Contactor servomechanism.
4.0
I.o
0
0
0.006
-80.
0.oK)
0.015
IRI
o.on,
10-7
NONLINEAR SYSTEMS
x + a ( x , x ) x+ b ( x , x ) x= 0
(10-8)
x + 2i3Jnx + W"2
=0
(10-9)
dx
+2t&
wn2
- =0
Y
(10-10)
dY = -a(x,y)
dx
2
- b(x,y) -
(10-11)
Y
By setting the right side of this equation
equal to a constant, a curve connecting points
of equal slope is determined. The isoclines
thus obtained a r e plotted in the phase plane,
and the slopes of the various phase trajectories can be drawn directly on the isoclines.
If a large number of isoclines a r e drawn, the
phase trajectories can be accurately determined.
Once the phase portrait of a system has
been constructed, the behavior of the system
can be investigated. If the response of the
system fo r a given set of initial conditions is
sought, the corresponding phase trajectory
determines the response. The variation of
time t along the trajectory can be ascertained
from the relation
t=ssdx
(10-12)
THEORY
the first-order terms in expansion a r e neglected. Thus, the singular points a r e determined from the solutions of the equations
x =P(X?Y)
(10-13)
=Q(w)
(10-14)
P(x,y) = 0
(10-15)
=0
(10-16)
Q(X,Y)
1U
,(X -
y = bl(x
+ cls(y
a ) + b,(y
a)
b)
(10-17)
b)
(10-18)
10-9
5 = 0.5.
NONLINEAR SYSTEMS
+bd
<0
(a1 + bz)' > 4(aib, - U & l )
(a,
I'
(10-19)
(10-20)
>0
( a , + b z ) 2 > 4(aibz - azbi)
(a, +b*)
(c) Stable
(10-22)
<0
(a, + bz)' < 4(aib* - azbi)
(a,
(10-21)
b2)
(10-24)
BY permhion from Automatic Feedback Control Svstcm Swrthe&. by J . G. Truxd. Copyrigbt, 1966, McGraw-Hill Book
Company. Inc.
+ >0
+ b,)') < 4(Uib2 - U z b i )
b2)
(10-23)
(10-25)
(10-26)
b2)
(10-27)
=0
(Qbp - azb1)
>0
(10-28)
( a , + b,) = 0
(10-29)
( 10-30)
I'
Fig. 10-18 Portrait in the vicinity .of a
stable focus.
By ptrmiuion from Artowaatic Feedbuck Control Syatem SynUuis, by J. G. Truxd, Copyright. 1966, YeGraw-Hill Book
Company, Inc.
By permission from Automatic Feedback Control Svatem SyncAeai#. by J. G. Trux.1. Copyright. 1966. McGraw-Hill Book
Company. Inc.
10-10
I'
I'
By penniuion from Automatic F e e k k Contrd Svrtem SWthuia. by J. G. TNX~I.Copyright. 1966, YcGlrw-Hill Book
Company. h e .
Adapted by permission from Automatic Feedback Control Svatem S # u t h b . by J. G. Truxar, Copyriphf 1966. YeGraw-Hill
Book Company, Ine.
I"
Fig. 10-20 Portrait in the vicinity of a center.
By penniuion from Autometic Feedback Cmtrd S y a t m SWt h b . by J. G. Truxal copyright, 1966. Mffilrw-Hill Book
Company. Ine.
(10-33)
10-1 1
NONLINEAR SYSTEMS
- ( a l + b,)
SADDLE
POINT
CENTER
(al b?
- ajbl)
UNSTABLE
STABLE
LIMIT CYCLE
By permission from Automatic Feedback Control System Synthesis. by J. G. Truxal. Copyright. 1955. McGraw-Hill Book
Company. Inc.
By permission from Automatic Feedback Control System Synthesis, by J. G . Truxal. Copyright. 1955, McGraw-Hill Book
Company. Inc.
-6P6 x+ - SQ6y
By permission from Automatic Feedback Control System Synthesis. by J . G. Truxal, Copyright, 1956, McGraw-Hill Book
Company, Inc.
10-12
THEORY
10-13
BIBLIOGRAPHY
M. Yachter, and J.
Zauderer, Instrument Inaccuracies in
Feedback Control Systems with Particular Reference to Backlash, Trans.
AIEE, Vol. 68, Part I, pp. #778-788,
1949.
1 H. T. Marcy,
1962.
7 R. J. Kochenburger, A Frequency-
F. C. Lindvall, and C. H.
Wilts, The Effect of Coulomb Friction
on the Performance of Servomechanisms, Trans. AIEE, Vol. 67, Part I,
17 E. Levinson,
9 G. D. McCann,
18 V.
THEORY
19 R. J. Kochenburger, Limiting in Feed-
29 A.
30 T.
#329-336, 1953.
1954.
34 I. Bogner and L. F. Kazda, An Investi-
for Determining the Closed-Loop Frequency Response of Nonlinear Systems, Trans. AZEE, Vol. 73, Part 11,
pp. #217-224, 1954.
#232-235, 1954.
37 J. G. L. Michel and A. Porter, The
NONLINEAR SYSTEMS
10 16
t
u.
AMCP 706-137
AMC PAMPHLET
SERVOMECHANISMS
SECTION 2, MEASUREMENT AND
SIGNAL CONVERTERS
HEADQUARTERS,
U. S. ARMY MATERIEL C O M M A N D
APRIL 1965
HEADQUARTERS
UNITED STATES ARMY MATERIEL COMMAND
WASHINGTON, D.C. 20315
30 A p r i l 1965
SELWYN D. SMITH, J R .
Major G e n e r a l , USA
Chief of Staff
OFFICIAL :
STANLEY J. SAWICKI
DISTRIBUTION : Special
PREFACE
The Engineering Design Handbook S e r i e s of the A r m y
Materiel Command is a coordinated s e r i e s of handbooks
containing b a s i c information and fundamental data useful
in the design and development of A r m y m a t e r i e l and s y s t e m s . The handbooks a r e authoritative r e f e r e n c e books of
p r a c t i c a l information and quantitative f a c t s helpful in the
design and development of A r m y m a t e r i e l so that it will
m e e t the t a c t i c a l and the technical needs of the A r m e d
F o r c e s . The presenthandbook is one of a s e r i e s on Servomechanisms.
Section 2 of the handbook contains C h a p t e r s 11 and 12,
which d e s c r i b e those s e r v o m e c h a n i s m components used
a s sensing elements and signal c o n v e r t e r s . Chapter 11
c o v e r s v a r i o u s sensing elements s u c h a s p o t e n t i o m e t e r s ,
r o t a r y t r a n s f o r m e r s , l i n e a r v a r i a b l e differential t r a n s
f o r m e r s , tachometer g e n e r a t o r s , gyroscopes and analogto-digital c o n v e r t e r s . Chapter 1 2 c o v e r s t h r e e types of
signal c o n v e r t e r s : m o d u l a t o r s , demodulators and digitalto-analog c o n v e r t e r s
F o r information on o t h e r s e r v o m e c h a n i s m components and on feedback control theory and s y s t e m design,
s e e one of the following applicable sections of this handbook:
...
111
CHAPTER 11
SENSING ELEMENTS
1 1-1 INTRODUCTION*
This chapter deals with those devices that
measure or sense the input or output of a
servomechanism and express it in electrical
form. The more common types of sensing elements used in ordnance applications convert
motion, either translational or rotational,
into a corresponding electrical representation. They are part of a class of devices called
trunsducers, the function of a transducer
being to receive information from some
source and to express that information in a
different form for use by another operating
component.
The variety of sensing elements that have
1 1-2 POTENTIOMETERS *
variable output voltage is taken. In an actual
potentiometer, one terminal suffices for end
points B and D combined.
RESISTANCE ELEMEN1
MOVABLE SLIDER
e.
VARIABLE
OUTPUT VOLTAGE
eovt
11-1
There are two basic types of potentiometers :(*) rotary and translatory. The majority
of potentiometers produced today are intended for applications where the slider motion is rotary. The basic rotary type can be
broken down into two subtypes: the singleturn potentiometer and the multiturn potentiometer.
Single-turn potentiometers are designed
for slider-travel limits of a full revolution or
less (e.g., 300'). Multiturn potentiometers
are designed for slider-travel limits of several
revolutions (e.g., some multiturn potentiometers are now being made with as many as
60 revolutions between stops).
R ( 6 ) = kI6
Rw = ki6wm
Jo p ( e )de
(11-3)
and
~ ( 0=
)
(11-1)
where
= slider angle (i.e,, the angular displacement of the slider with respect
to the zero end of the pstentiometer)
R ( 6 ) = resistance between the slider and
the zero end of the potentiometer
(i.e., the resistance between points
C and D in Fig. 11-2)
(11-4)
dR (8)
=do
(11-5)
(11-8)
R ( z ) =k:!x
Fig. 11-2 Wire-wound element and slider.
R," = kz
11-2
x ~ O ,
(11-9)
for p (x)= constant = ks
( 11-10)
SENSING ELEMENTS
(11-11)
~ ( e=
) L'cos e de = sin e
(11-14)
(11-12)
where
K 2 = constant value of p ( x )
11-2.5 Use
p ( e ) = cos e
(11-13)
Hence, this particular nonlinear potentiometer can be used to express electrically the sine
of the slider position.
1 1-2.6 Construction Features
q z -+=xK,,K,,,
1
where
(11-15)
SENSING ELEMENTS
Figure 11-3 shows a typical plot of percent output voltage as a function of percent
slider position, with percent deviation from
linearity (D) Eout/Eesexaggerated. The maximum percent deviation of the actual curve
from the ideal curve is called t e m i m l lineari t y . Terminal linearities as low as 0.01 percent have been achieved, but linearities of
0.1 to 0.5 percent are more common in most
commercial units.
A different measure of potentiometer perf ormance commonly used is independent linearity, which is defined by the Radio, Electronic and Television Manufacturers Association (RETMA) as the deviation when the
slope and position of a straight reference
line are chosen to make the maximum deviations a minimum over the actual effective
travel or any specified portion thereof. The
straight reference line used does not coincide
with the line corresponding to ideal performance. The independent linearity figure is
smaller than the terminal linearity figure.
The best straight line for the actual characteristics previously shown is drawn in
Fig. 114, which also shows that the maximum deviation now occurs at a different
slider position and is smaller than that in
Fig. 11-3. Note that the best straight line
does not pass through the origin nor does it
intersect the actual performance curve at
the 100-percent point. The best straight line
can be made to coincide with the required
straight line by trimming techniques, so that
the actual performance deviation from an
ideal straight line will differ by only the
independent linearity figure. Trimming is accomplished by referring the slider voltage
not to the voltage applied across the resistance element but to a value that corresponds
to DE in Fig. 11-4.
A typical trimming circuit is shown in
Fig. 11-5. The upper and lower trimmer potentiometers are center-tapped and the voltage between the two center taps is taken as
the reference. The sliders on the two trimmer potentiometers are adjusted so that
when the slider of the precision potentiometer is at each end, the percentage output voltage differs from the best straight line by m
and n, respectively, of Fig. 11-4.
LCTUAL POTENTlOYETER
PmFORM(LNCE
I(
IDEAL POTENTIOMETER
PERFORYIKE
PERCWT W I Y W
P
c
a
a
I-
[TERMINAL LINEAWTYI
3
0
I-
w
z
U
W
K
a
NOTE: THIS PLOT APPLIES TO THE SINATION IN WHICH
THE EXCITATIOW VOLTAGE IS APPLIED DIRECTLY
TO THE TWO TERMINALS OF THE RESISTANCE
ELEMENT, SO THAT E~w,)mo~
= E,
11-5
(11-16)
Ex a S ~ n c m
VOLTAGE
where
POTENTIOMETER
R, = load resistance
R (e) = resistance between the slider and
the zero end of the potentiometer
R, = total winding resistance of the potentiometer
REFERENCE
VOLTAGE
VOLTAGE
0
LOWER TRIMMER
POTENTIOMETER
0.1
02
0.3
0.4
0.5
0.6
0.7 0.8
0.9
1.0
R(B)/R,
11-6
SENSING ELEMENTS
A second method consists of adding a series
resistor R, as shown in Fig, 11-7. Nettleton
and Dole(3)have shown that for Rt > > Rw (by
at least a factor of 10) the output voltage cf
the uncompensated circuit has the following
maximum error, expressed as a percent of
the maximum output voltage :
OUTPUT
V 0 L T A GE
100R,
11Rt Rw
(11-17)
17.8613,
9RI + 2Rw
(11-18)
A reduction in the loading error in the ratio
of about four- or five-to-one can thus be accomplished. However, in order to achieve a
desired output voltage for R ( e ) = R,, the
supply voltage for the compensated case must
be made higher than in the uncompensated
case. For R e = 0.28RW,the supply voltage
must be 28 percent higher than the desired
output voltage for R (e) = R,.
percent maximum error =
SERIES RESISTANCE
SUPPLY
VOLTAGE
Fig.
LOAD
RE 51 STANCE
Rr
*TAT[/
PRECISION
POTENTIOMETER
PARALLEL
RESISTANCE
(11-19)
POTENTIOMETER
PRECISION
EXClTAnON
VOLTAGE
I
R1 > 10
R,
R,
-
Rt
-
RP
Rw
Ri
1
2
3
5
10
0.73
0.73
0.73
0.735
0.74
0.281
0.304
0.315
0.312
0.305
0.74
0.311
R
W
over10
I I I
(D1Eo.t
Etolft,mz
0.025
0.011
0.0067
0.0038
0.002
0.019
(Rt/R=)
put versus rotation data. Terminal conformi t y is the maximum percentage deviation,
with respect to the excitation voltage, of the
actual electrical output a t any point from
the specified function curve drawn through
the end points. By use of trimmer potentiometers similar to those discussed under linear
potentiometers, the performance of a nonlinear potentiometer can be made to coincide
with its independent conformity.
Nonlinear as well as linear potentiometers
are subject to loading errors, and compensation for nonlinear units is accomplished by
the same circuits as those discussed under
linear potentiometers. For example, GilberW
shows that where parallel compensation only
is used, the tap should be located at 74 perand paralcent of total winding resistance RII.
lel resistor R, (connected between the tap
and the upper end of the total winding resistance) should be 31 percent of load resistance R,. However, since nonlinear potentiometers very often must be made up specially for a specific application, the custom
design can take into account the effect of a
specified load and compensate for it in the
element construction.
A tapped linear potentiometer can be used
to approximate nonlinear fiinctions by the
addition of shunting resistors. This technique is described as follows :
The first step in the design of a shunting
circuit is to draw the required nonlinear function as shown in Fig. 11-9.A series of connected straight lines are next drawn to best
represent the function. If the tapped potentiometer has already been constructed, the
straight lines must join at angular points corresponding to the tap points. If the tapped
potentiometer is to be customdesigned for
the particular nonlinear application, the best
straight-line approximation can be drawn and
the tap points placed at the resulting points
of intersection. It is usually necessary to
space the taps closely in the region of maximum function curvature.
An important characteristic of the function of Fig. 11-9 is that the slope dues not
change its algebraic sign throughout the entire function. The curve commences with a
positive slope and continues positive throughout the remainder of the function.
118
SENSING ELEMENTS
8I
II
I
I
END
miu
NlcllQ
wuc
AOsl
POTENTIOMETER
RESISTANCE ELEMENT
R.?
RESISTANCE
SLIDER
% SHAFT ANGLE
6. CIRCUIT DIAGRAM
A. RESISTANCE DIAGRAM
Fig. 1 1 - 1
l,,(I.F.
It is necessary only to choose the potentiometer having a winding resistanee R,, that is
greater than this defined minimum value.
The steps required in the design of a potentiometer shunting network can be summarized as follows :
( a ) Plot the desired function %R (or
% V) versus %e. These percentages are commonly expressed in decimal form with values
ranging from 0 to 1.0.
(b) Approximate the desired function with
connected straight lines, joined at selected
tap points or at tap points predetermined by
the potentiometer construction.
II WAFT ANGLE
11-10
SENSING ELEMENTS
R,. = ARCX B X Rd
(11-21)
R, = ARp X B X Rd
2
( 11-22)
00
1 -E2
"Quoted by permission from the Potentiometev Handbook, 1956, Technology Instrument Corp.
Nonlinear functions can also be approximated by loading the slider and adding series
resistors to a linear potentiometer. Some typical configurations and the corresponding
transfer functions are given in Table 11-3.
Another method of generating nonlinear
functions through the use of tapped linear
potentiometers is based on clamping the taps
a t the voltage levels that the desired function
has at these points. This is illustrated in Fig.
11-13, which shows the resultant straight-line
approximation of the desired function. It is
best to derive the clamping voltages from a
very low impedance bleeder as shown, for not
only does this eliminate the need for many
separate bias voltages, but it permits adjustment of the tap voltages when the slider load
resistance is connected. If the load resistance
is low, the resultant voltage segments between tap points will be concave and can be
made to coincide with the desired function
more closely by increasing the tap voltages.
If the function slope between taps is steep,
care must be exercised that the resultant
a nonlinear function.
1 1-2.12 Noise
There are two types of noise voltages, usually not exceeding a few hundred microvolts,
that appear a t the output of a potentiometer :
active and passive. The former is due to the
motion of the slider over the resistance element. The latter is due primarily to the fluctuating contact resistance between the slider
and the resistance element and may be reduced by increasing the pressure of the slider
against the element. However, this will
shorten the life of the potentiometer and increase the driving torque.
Total noise is expressed as a n equivalent
noise resistance ENR, which is measured as
shown in Fig. 11-14. This figure, together
11-11
CIRCUIT
TRANSFER FUNCTlON
TYPICAL FUNCTIONS
I .o
0.8
- =cy
c.0.75
I1
c+y-y'
0.6
4 0.4
c.o.1
0.2
0.0
C.0.25
0.2
0.6
1.0
Y
APPROXYUTES
1- Y
(8-b-CdO;dd2.4)
1.2
1.0
0. e
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.e
-1.0
-I .2
mu m w
I-yl
41
&I42
1 LARGE
APPROXIMATES
I.cV
I .o
0.8
0.6
4c
UaAtckl-V'
0.4
- 0.2
2
0.0
=-0.2
1.0
0.2
0.0
11-12
Y4.978
4 2 0.6
1.0
SENSING ELEMENTS
with the following explanation, is taken from
Altieri :(O)
A source of constant current is placed
across one arm of the potentiometer and the
slider. The slider and the other arm of the
potentiometer are joined to a voltage-measuring circuit having a high input impedance.
Owing to the high impedance of the voltage
output circuit, the entire source current .i
flows through the left-hand portion of the
potentiometer and through the slider.
If the source current is zero, the noise
voltage measured at the circuit output can be
attributed entirely to the active noise potential at the slider contact. When the source
current is applied, a voltage drop appears
across any passive resistance which may
exist at the slider point of contact. This p a 5
sive noise voltage adds directly to the active
noise voltage.
The total voltage output of the circuit is
the sum of active and passive residual noise
components. Although the two types of noise
are of quite different nature (one is a generated voltage and the other is a passive resistance) it is helpful to express both active
and passive noise components as ohmic resistances. The total of these resistances is
called the equivalent noise resistance and is
defined as the total residual noise voltage
divided by the source current (Fig. 11-14).
G-
EXCITATI4
CWRENl
EOUVUENlNOlSEREIISlANCE
-=+ENR
-a -
ACTIVE
PASSIVE
R,
11-13
The torque load of the slider on the driving shaft varies with individual designs and
ranges from several thousandths of an ounceinch to several ounce-inches. Most units require about one ounce-inch of starting torque
and half that amount for running torque.
Moment of inertia figures also vary greatly
for different designs and range from less than
a thousandth of a gram-square centimeter t o
several hundred gram-square centimeters.
11-14
SENSING ELEMENTS
In some designs, it has been found desirable to immerse the wire-wound resistance
element in a light mineral oil. The flushing
action of the oil reduces electrical noise, disperses wear products which otherwise would
tend to lodge between the wire turns, and increases life because of its lubricating action.
An example O f lubricated potentiometers can
be found in the NIKE AJAX Instruction
Manual, Vol. VI, Chapter 11.
*By A . Kuako
more rotary transformers are used in a subsystem or in conjunction with the servo system. In such cases, one or more of the devices
may be used to receive, modulate, or transmit
data while another may be used to produce a
torque that is a function of the input received
from the data-handling units. For the present
discussion, the torques produced by rotary
transformers will be considered to be in the
order of a few inch-ounces because, in a strict
sense, rotary transformers are small-torque
devices.
11-3.4 Miscellaneous System Uses
Rotary transformers are also used in systems that require modulation of electrical
waveforms, resolution of vectors into components, and vector combining processes. In
general, such uses are concerned only with
data-handling operations.
1 13.5 General Functional Classification
11-15
All standard synchros are designed to operate from one of the following supplies : (a)
26 volts at 400 cps, (b) 115 volts at 400 cps,
or (c) 115 volts at 60 cps.
113.7 SYNCHROS
11-3.1 1 Nomenclature
In general, the synchro stator is a cylindrical slotted and laminated structure with three
windings arranged in the slots at 120 spatial.
displacement from each other. In most units,
the slots are skewed one slot pitch to avoid
slot lock and resulting angular errors.
Units that do not have skewed stator laminations are constructed with skewed rotor laminations.
The stator windings are not 3-phase in the
usual sense because all induced voltages are
in time phase. They can be either Yconnected or delta-connected (Fig. 11-17) and
serve as the secondary winding of the synchro. Stator connections are usually brought
out as three leads and labelled SI,S2,and S,.
Function
Contdol
Torque
Succeedins
Letters
Rotors of standard synchros are of twopole construction, the most common type being the salient-pole rotor used in transmitter
or repeater units. The rotor is a slotted and
laminated structure of the dumbbell or
I type. The structure is mounted on a
shaft that turns on ball bearings. It cames a
First Letter
Function
Differential
Receiver
Transformer
X
B
Transmitter
Rotatable Stator Winding
11-16
SENSING ELEMENTS
(d) The succeeding digit indicates the f r e quency of the power supply in accordance
with the following:
Number
Frequency
60
400
(vs)
=2
A. I-CONNECTED STATOR
6. DELTLCONNECTEO STATOR
[
C
ELECT;ICAL
ELECTRICAL
ZERO
COUhTE R.
CLOCKWISE
CLOCKWISE
ROTOR ANGLE
11-17
The various services formerly used individual methods of nomenclature which will still
be encountered in old designs. The Army
method of nomenclature is much like the system described in the previous paragraph,
while the Navy method for 115-volt 60-cps
synchros differs greatly. The latter synchros
have a code designation such as 5 HCT Mark
2 Mod 3B, which identifies the approximate
size, special design features, function, type
designation, and manufacturer as follows :
( a ) The first digit indicates a size group
in the following table:
Approx
Wt
Size
(lb)
Approx
Length
(in.)
Approx
Dia
(in.)
3.9 - 4.2
2.26
5.2 - 5.51
3.1
6.0 - 6.8
6.4 - 7.5
4.5
18
8.9 - 9.2
5.75
3.39
3.625
S -special
(c) One or more letters following the special-fitting letters indicate the function of the
unit from Table 11-4.
(d ) The Mark number signifies the design
of the particular unit being specified.
(e) The Mod number designates the manufacturer as assigned by the Bureau of Ordnance.
1 1-3.1 3 Transmitter Characteristics
11-18
SENSING ELEMENTS
I Functional Designation
Function
Symbol
Synchro transmitter
Generates and transmits electrical signal corresponding to the angular position of the mechanically driven
rotor of the unit.
Synchro receiver
Synchro control
transformer
CT
Produces an electrical output signal having a magnitude proportional to the angle of rotation of the unit
rotor with respect to the magnetic field set up by the
unit stator. Indirectly, the output signal is a sine function of the angle between the rotor shaft position and
the shaft position of the transmitter that generates
the transformer input.
Synchro differential
transmitter
DT
Synchro differential
receiver
DR
rotor, there is but one stator voltage condition. The electrical zero position for Y-connected and delta-connected stators is shown
in Fig. 11-17.
1 13.14 Receiver Characteristics
A receiver has the same electrical configuration as a transmitter but the unit output is
an angular position of the rotor corresponding to a n electrical signal input. The rotor
winding of the receiver is excited from the
same single-phase a-c source as the transmitter rotor, and the receiver stator terminals
are connected to corresponding terminals of
the transmitter (Fig. 11-18).Assume for the
moment that the rotor winding of the receiver is open and that voltages are induced
11-19
fields are aligned. The receiver rotor, therefore, is forced to take up an angular position
corresponding to the position of the transmitter rotor.
A distinguishing feature of a receiver is a n
oscillation damper in the form of a fiywheel
that is free to rotate relative to the rotor
shaft within limit stops set about 45 apart.
The flywheel has approximately the same moment of inertia a5 the rotor. A friction coupling between the rotor shaft and the fiywheel
serves to dissipate energy when the rotor oscillates as it does when coming into a position of alignment. The added inertia furished by the oscillation damper also prevents the rotor from running away when
it is forced to follow the transmitter rotor
in a continual process. Without the oscillation damper, running away would occur at
high angular velocities when the torque becomes large enough to overcome the aligning torque and thereby prevents the transmitter from maintaining control.
I
I
tI
I
I
I
1
I
I
I
I
I
I
I
I
s1
I
I
I
A X SUPPLY
I
I
SENSING ELEMENTS
11-21
s,
17-22
SENSING ELEMENTS
- - - -1
11-23
.L
r
I
I
I
I
I
r
I
i
I
I
I
I
I
I
I
I
I
I
L
r
I
I
I
c
a
P
-c
a
_ _ _ _ .-J
I
I
11-24
SENSING ELEMENTS
Synchro capacitors are used with transformers, differential transmitters, and differential receivers to neutralize the lagging component of the magnetizing current drawn
from the transmitters supplying the stators of
the units concerned. For this purpose, three
matched capacitors are delta-connected as
shown in Figs. 11-19 to 11-21. The leading
current drawn by the capacitors at almost zero
power factor neutralizes the lagging component of the magnetizing current and thus reduces the total current drawn from the transmitter. To ensure proper operation, the capacitance of the three legs of the capacitor must be
balanced to within 1 percent but the absolute
value of the legs may vary over t 1 0 percent.
For maximum effectiveness, the leads connecting the capacitor bank to its associated
transformer o r differential unit should be as
short as possible.
11-3.19 Dual-Speed Synchro Systems
1 1-25
each other. For this purpose, a definite relative position of the rotor with respect to the
stator has been designated as the electrical
zero position f o r all standard synchro units.
This position is the relative angu!ar position
of the rotor and the stator that results in a
minimum, or null, output voltage. The process of rotating the stator or rotor with respect to one another to obtain the electrica!
zero point is called zeroiuy.
To facilitate the zeroing of synchros, the
units are constructed with standardized fianges concentric with the shaft. In some synchros, the flanges are located on the head end
of the frame; in others, near the center. In
either case, the synchros are mounted and held
in place by a ring retainer or by dogs acting on the flange. Thus, by loosening the ring
or the dogs, the stator may be rotated until
the synchro output corresponds to the zero
shaft position. When this procedure is followed for all synchros in a system, the system
will operate properly.
RATEOVOLTAGE
ANDFREQUENCY
I3
A. APPROXIMATE ZERO
RATEDVOLTAGE
AND FREQUENCY
8. EXACTZERO
or receiver.
Using the connections shown in Fig. 1124A, the rotor of a differential synchro unit
will turn by itself to a n approximate zero position. For a n accurate zero determination,
connect the unit as shown in Fig. 11-24B and
adjust until the voltage across R , and R3 is
zero.
11-26
SENSING ELEMENTS
JS'
i a VCLTS
VOLTMETER
WJ
s2
~~
't
s3
A. APPROXIWTE ZERO
78 VOLTS
8 . EXACTZERO
is helpful to predict the performance of a synchro system in which the receiver is connected
to a transmitter of different size and construction. The method used for such a prediction is
based on the following known operating characteristics :
( a ) The unbalanced stator voltages of a
standard transmitter or receiver are the same
regardless of the size of the units.
(b) The only electrical characteristic that
varies with the size of a synchro is the internal impedance of the stator windings.
(c) When two standard synchros of the
same size are connected together, the internal impedance determines the amount of current flowing and thereby the torque produced.
(d) The torque produced by the synchros
in (c) can also be determined by measuring
the torque gradient of the receiver. A torque
gradient obtained in this manner is called
the unit torque gradient and is inversely proportional to the internal impedance of the
stator windings.
In view- of the above, consider a receiver
driven by a transmitter having a unit torque
gradient equal to R times that of the receiver.
11-27
8. EXACTZERO
TORWE GRADIENT
SAFE CONTINUOUS
11-28
ROTORCURRENT
DISPLACEMENT
ANGLE
TORWEORCURRENT
SENSING ELEMENTS
360' YAMIFACTURINE
IRREGULARITIES
Y
(L
ROTOR ANGLE
G UUTS,
N
NUMBER OF SLO
in static error, the error of a transmitter-receiver system can be predicted by the method
used previously for determining the actual
torque gradient of a receiver. The static error
of a transmitter-control transformer system
may be predicted approximately from the noioad electrical errors of the individual units.
If the units are connected so that their shafts
turn in the same direction, the no-load electrical error of the system equals the electrical
error of the control transformer minus the
electrical error of the transmitter. Hence, system accuracy improves if the no-load electrical error curves of both units follow approximately similar variations of the shaft position.
The above methods cannot be used to determine the over-all error of a system employing a differential synchro because the electri.
cal error of such a unit is a function of the
rotor positions of other units connected to it
In general, the over-all error of a differential
synchro system is the sum of the errors of the
individual units.(")
11-3.29 Dynamic errors. The dynamic erroi
of a receiver and differential receiver is the
angle by which the receiver shaft lags behind
the transmitter shaft during a slow rotation
of the transmitter shaft. The chief sources of
this type of error are the friction of the re.
ceiver bearings and brushes and the inertia oi
1 1-29
Next Page
d8 = K E cos ot sin e
(11-24)
where
E = applied rotor voltage
K = winding and frequency constant f o r
he synchro
w = 2nf = radian frequency of the voltage
applied to the rotor
(11-26)
Transmitter
1G
Torque gradient
min (oz-in./deg)
7 HG
or
7 HG
18
18
18
0.40
1.2
3.4
88.2
91.8
88.2
91.8
5 G
or
5 HG
18
18
0.06
0.25
1 HG
Static accuracy
max error
(minutes)
6 G
or
6 HG
3 G
or
3 HG
or
Size
Receiver
90(1 F)
150(1 HF)
0.06
36(3F, 3B)
60(3 HF)
0.25
36(5F, 5B)
45(5 HF)
0.40
88.2
91.8
Primary current
max (amp)
0.30
0.40
0.60
1.3
3.0
0.30
0.40
0.60
4.8
5.5
7.0
15.0
22.0
4.8
5.5
7.0
Secondary load
current
max (amp)
0.20
0.35
0.70
1.50
0.04
0.20
0.35
Temperature rise
max ( O C )
50
50
50
50
50
50
50
Primary power
max (watts)
88.2
91.8
88.2
91.8
11-30
88.2
91.8
88.2
91.8
88.2
91.8
CHAPTER 12
SIGNAL CONVERTERS*
12-1 INTRODUCTION
12-1.1 TYPES
Three types of signal converters are discussed in this chapter: modulators, which
superimpose a signal on a carrier ; demodulators, which recover the signal t h a t has been
superimposed on a carrier ; and digital-toanalog converters, which express a number in
electrical form as a proportionate voltage or
as a shaft rotation.
12-1.2 Modulators
I
A. ERROR SIGNAL
B. CARRIER SIGNAL
12-1.3 Demodulators
UI"'
G M O W L A T O R OUTPUT
*ByA . K.Susskind
12-1
The form of modulation used in servomechanisms is called suppressed-carrier modulation. This means that the output of the
modulator contains frequency components
which are the sum and difference of the carrier and the signal frequencies, but contains
no frequency component equal to the carrier
frequency. Figure 12-1A shows an error signal and Fig. 12-1C shows the corresponding
output of a suppressed-carrier modulator.
The modulated signal is zero when the error
signal is zero, and the phase of the carrier
sinusoids reverses when the sign of the error
signal reverses. Therefore, demodulators used
12-2 MODULATORS
12-2.1 CHOPPER MODULATORS
Choppers are electromechanical modulators and, because of their simplicity and high
zero stability (very nearly zero output when
the signal to be modulated is zero), find wide
application in servomechanisms.
ing voltage. When V, the voltage to be modulated, is connected as shown in Fig. 123,
the square-wave output waveform has the
CONTACTS
12-2.2 Description
DRIVE
COIL
12-2
OUTPUT
WAVEFOIW
SIGNAL CONVERTERS
I 2
cos d X ( Z m - 1 )
sin [ o ( 2 m - l ) t ]
(12-1)
DRIVING VOLTAGE
CHOPPER OUTWT
12-4
SIGNAL CONVERTERS
f
Fig. 12-7 Push-pull half-wave connection.
12-5
EXCITATION
WINDINGS
eL LOAD
SIGNAL CONVERTERS
There are two categories of operating circuits employing the magnetic modulator : one
category is very similar to that shown in Fig.
12-9, in which the signal and load windings
are separated; in the second category, the
d-c signal and the load are connected to a
common winding, each core having only two
windings - an excitation winding and a signal-load winding.
12-2.15 Sensitivity
For maximum magnetic-modulator sensitivity, the control-winding impedance is usually matched to the impedance of the signal
source, and the load-winding impedance is
usually matched to the load impedance, although this may not be true with high-impedance load circuits, such as grid circuits of
vacuum-tube amplifiers. It is possible to secure individual load-impedance and signalimpedance matching with the circuit in which
the control and load windings are separated.
However, the common-winding circuit will
result in greater sensitivity and, thus, may be
more desirable in certain applications.
12-2.20 Operation
I60
120
40
eo
40
120
OCNALIN MICROAMPERES
RL
OUTPUT
1
Fig. 12-13 Diode modulator.
12-8
160
SIGNAL CONVERTERS
~~
output during one half-cycle of the sinusoidal carrier voltage e. During the other halfcycle of e, the output is connected to ground.
Therefore, the output waveform is a rectangle of amplitude V, with a duration of half a
period of the carrier and with a repetition
rate equal to the modulating frequency. Resistors R , are connected in series with all
diodes to limit the diode current when the
when the upper end of the transformer secondary is positive, and the rectangle occurs
during the other half-cycle of the carrier. In
a variation of this circuit, a second d c signal
is connected to the center tap of the transformer. The output signal then has a peakto-peak amplitude that is the difference between the two d c signals, and the modified
circuit can therefore be used as a combination error (difference) detector and modulator. In the back-to-back triode circuit of Fig.
12-15,a very large carrier signal is applied
to the grids, making one of the two triodes
conduct during approximately the entire positive half-cycle of the carrier signal. The left
tube conducts during the positive halfcycle
if V is positive and the right tube conducts if
V is negative. The grid signal is so large that
the tube drop is very small during conduction and, therefore, the output is nearly at
ground potential during conduction. In order for this to occur, R, and RL must be
large (e.g., 500,000 ohms). The circuit is
used only when large signals, in the order of
100 volts, are to be modulated.
12-2.21 Bridge modulator. Figure 12-16 differs from the other configurations in that
two rectangles of opposite polarity are generated during each cycle of the carrier. When
the carrier polarity is such that point A is
positive with respect to point B, V appears
across the output transformer with one polarity; if A is negative with respect to B, V
appears across the output transformer with
opposite polarity.
12-2.22 Typical curve. In the diode circuits,
the statement that the rectangles have a duration of half a period of the modulating signal is based on the assumption that the peak
amplitude of e is large compared with V.
Where e is small compared with V, saturation occurs. A typical curve is given in Fig.
12-17,showing the rms value of the output
voltage as a function of the magnitude of V
for a fixed amplitude of e. The knee in the
curve shows the occurrence of saturation,
12-10
SIGNAL CONVERTERS
123.2 Operation
Figure 12-18 shows the schematic of a diode circuit frequently used in servomechanisms for demodulating suppressed-carrier
signals ; i.e., signals in the form
(12-2)
where
o. = angular frequency of modulated signal (e.g., error signal)
or= angular frequency of carrier
are, respectively
(12-4)
e, = V, sin opt
(12-5)
v1
W
v2
(12-7)
where
1
where
2fc
T, =
1--
NI =
where
x
fr
nRC)*
(12-8)
2n
O C
f c =
stant
2f rRc
s = complex frequency of modulated signal
12-3.6 Output stability. The demodulator output should always be taken across P and R,
as shown in Fig. 12-18. If push-pull outputs
are taken across PQ and RQ separately, the
ripple a t zero input signal results in timevarying outputs and poor stability.
12-3.7 Full-Wave Demodulator
(12-10)
12-12
SIGNAL CONVERTERS
-R
z;
41
32 c
12-13
The ripple voltage is zero when the modulated signal is constant. When the modulated
signal is a sinusoid, the ripple voltage is a
saw tooth, amplitude modulated by that sinusoid. The ripple factor, which is lower
than that of the diode demodulator, is
rejection is equally good, however. for both the keved demodulator and
1-
RE
0
0
,:
c*L
(12-11)
where
Choppers can also be used as demodulators. Figure 12-23 shows the schematic diagram of a full-wave chopper demodulator.
12-3-13 Operation
12-14
SIGNAL CONVERTERS
Muse
12-15
the filter during one-half-cycle of reed motion and grounds the filter input on the
other half-cycle. This results in a unidirectional filter-input signal, the magnitude of
e, =
IR
3 )( 2"'
(12-13)
R --R
3
where
(12-14)
' I -
P
n
= number to be converted
+ 1 = number of stages
2. =2
12-16
(12-16)
SIGNAL CONVERTERS
Figure 12-27 shows a circuit that is particularly useful when the digital information
originates from relays. Here, relay Kjis not
energized (contacts in the position shown)
when the associated binary digit is a ZERO.
The contacts are in the other position, however. when the associated digit is a ONE.
The output voltage is given by
6 O
El
R + 2"+?- 1
(12-17)
Rc
and the output impedance is given by
(12-18)
Figure 12-28 shows another circuit that is
also useful when the digital information originates from relays. Hem again, relay Kfis
not energized (contacts in the position
shown) when the associated binary digit is a
ZERO,and relay Kjis energized (contacts in
the other position) when the associated binary digit is a ONE. The output voltage is
given by
e, = 2"+'* - 1
(12-19)
accuracy of the precision resistors. In addition, in Figs. 12-25 and 12-26, accuracy is affected by the magnitude of the outputs of the
sources and also by their output impedance.
The circuits of Figs. 12-27 and 12-28 are
therefore somewhat more accurate, but even
these can rarely be made more accurate than
Fip. 12-28
Qrmyekc.
12-17
DIGITAL-TS
ANALOG
CONVERTER
AMD AMPLIFIER
ANALOGTO.
DIGITAL
COWVERTEII
BIBLIOGEAPHY
1 W. A. Geyger, Magnetic Amplilim Circuits, Ch. 16, McGraw-Hill Book Company, Inc. New York, N. Y., 1954.
5 K. E. Schreiner, High-Performance
Demodulators for Servomechanisms,
Proc. N a t i m t d Ehxtnnrics Conference,
Val. 2, 1946.
S P. Detwiler, Phssesensltive-Detector
C h d r i s t i C S , AIEE MbceUaAeou8
Paper 51349.
AMCP 706-138
AMC PAMPHLET
SERVOMECHANISMS
SECTION 3, AMPLIFICATION
APRIL 1965
HEADQUARTERS
UNITED STATES ARMY MATERIEL COMMAND
WASHINGTON, D.C. 20315
30 A p r i l 1965
SELWYN D. SMITH, J R .
Major G e n e r a l , USA
Chief of Staff
OFFICIAL:
Y i i Colonel, GS
Chief, Administrative Office
PREFACE
The Engineering Design Handbook S e r i e s of the A r m y
M a t e r i e l Command i s a coordinated s e r i e s of handbooks
containing b a s i c information and fundamental data useful
i n the design and development of A r m y m a t e r i e l a n d s y s t e m s . The handbooks a r e authoritative r e f e r e n c e books of
p r a c t i c a l information and quantitative f a c t s helpful i n the
design and development of A r m y m a t e r i e l so t h a t it will
m e e t the t a c t i c a l and the technical needs of the A r m e d
F o r c e s . The p r e s e n t handbookis one of a s e r i e s on S e r v o mechanisms.
Section 3 of the handbook contains Chapter 13, which
c o v e r s the different a m p l i f i e r s u s e d i n s e r v o c o n t r o l l e r s .
This chapter d e s c r i b e s the following types of a m p l i f i e r s :
electronic, t r a n s i s t o r , magnetic, r o t a r y electric, r e l a y ,
h y d r a u l i c , pneumatic and m e c h a n i c a l . The advantages and
disadvantages of the v a r i o u s t y p e s a r e d i s c u s s e d to help
the c i r c u i t d e s i g n e r i n making a choice f o r h i s p a r t i c u l a r
application.
F o r information on o t h e r s e r v o m e c h a n i s m components
and on feedback control t h e o r y and s y s t e m design, s e e one
of t h e following applicable sections of this handbook:
AMCP 706-136 Section 1 T h e o r y ( C h a p t e r s 1 - 1 0 )
AMCP 706-137 Section 2 M e a s u r e m e n t a n d Signal
C o n v e r t e r s ( C h a p t e r s 11-12)
AMCP 706-139 Section 4 P o w e r E l e m e n t s and S y s t e m
Design ( C h a p t e r s 14- 2 0 )
An index for the m a t e r i a l i n all four s e c t i o n s is p l a c e d
at the end of Section 4.
E l e m e n t s of the U. S . A r m y M a t e r i e l Command having
need f o r handbooks m a y submit r e q u i s i t i o n s o r official
r e q u e s t s d i r e c t l y to Publications and Reproduction Agency,
Letterkenny A r m y Depot, C h a m b e r s b u r g , Pennsylvania
17201. C o n t r a c t o r s should s u b m i t s u c h r e q u i s i t i o n s o r
r e q u e s t s to t h e i r contracting o f f i c e r s .
Comments and suggestions on t h i s handbook a r e welcome and should be a d d r e s s e d to A r m y R e s e a r c h OfficeD u r h a m , Box C M , Duke Station, D u r h a m , North C a r o l i n a
27706.
iii
CHAPTER 13
+By A . K.Swrskind
........................................................................................
.........................................................................................................
E,,
Ebb
.....................................................................................
e,
..........................................................................
eb
.........................................................................
ib
...............................................
..................................................
...............................
Ebo
Ibo
e,
.............................
e,
.............................
ip
13-1
AMPLIFICATION
13-1.3 Control of electron flow. Useful control of electron flow from the cathode cannot
be achieved by varying the heater potential.
Therefore, heaters are always connected to
sources of fixed potential, and heater symbols
are frequently omitted from electronic-circuit
diagrams.
In a diode, the only useful way to contrcl
electron flow is by control of the plate voltage
eb (the potential difference between the plate
and cathode). The relationship between plate
voltage eb and plate current i b for constant
cathode temperature is shown in Fig. 13-2,
where the curve is drawn for positive values
of eb. For all negative values of eb, plate current i b is zero. Reference directions (indicated
by arrows in Fig. 13-1, for example) are chosen so that eb is a positive number whenever the
plate is actually positive with respect to the
cathode, and i b is a positive number when current flows from plate to cathode.
*-
-(+
EWI
Rb+R
--2
3rc
i -
b'
+ -sincot
1
2
2
cos 20t - - cos 40t - . . .
15x
(13-1)
Fig.
13-2
733 Diode
2n
ot
3,
used as redifier.
AMPLIFICATION
SLOPE
/-
-l/RL
t
ib
a
- k,
ebl
eb2
eb3
13-1.10 Linear equivalent circuits. The dynamic voltage-source linear equivalent circuit
for a triode is given in Fig. 13-8.Here, e,, ep,
and ip are time-varying quantities and the
grid-to-cathode impedance is shown as infinite. This is primarily justified by the assumption that the grid is never allowed to
become positive with respect to the cathode
and, as a result, no significant grid current
flows. In the equivalent circuit, the amplification factor p is defined as
aeb
--C
PLATE
iP
eb
p=-
Ebb
(13-2)
I\
ib=COMtant
(13-3)
13-4
1
GRID
PLATE
i,
-&-
L
I
I
I
CATHODE
+ e,
eb
=Ebo
ib
= zbo + i p
+ e,
e, = E,,
(13-5)
(13-6)
(13-7)
13-5
AMPLIFICATION
PENTODE
Fig. 73-7 7 Plate characteristics of a beampower tube with constant screen voltage.
= gm r p
(13-8)
is much higher for multigrid tubes than for
triodes.
13-1.16 Linear equivalent circuits. When
screen-grid and suppressor-grid voltages are
constant, and operation is above the knee of
the curves, the linear equivalent circuits of a
multigrid tube have the same form as the
V
the signal frequencies encountered in servomechanisms. Therefore, interelectrode capacitance may be ignored, except when amplifier stability is considered.
13-1.19 Tube Specifications
The tubes used in low-level voltage amplifiers have the following characteristics :
(a) Triodes have high values of p (e.g.,
l O O j and pentodes have high values of g,
(e.g., 4000 micromhos) .
(b) Maximum current ratings are moderate (e.g., 10 ma).
(c) Plate-dissipation ratings are moderate
(e.g., 1 watt).
13-1.22 Simple Amplifier
(13-9)
P,= El, I b
(13-10)
AMPLIFICATION
gain from each stage. Equation (13-11) indicates that the maximum possible gain equals
1.1 and can be attained by making RL much
larger than r,. To prevent this large value of
RL from producing plate-current values so
low as to make r, large (rP is inversely related to i b ) , a high plate supply voltage is
required, reachicg a value that is usually
impractical. Practical values of Ebb do not
exceed a few hundred volts and usually restrict the gain to a level that is only 60 to 80
percent of the amplification factor. Similar
restrictions apply to a pentode, since Q,,,is
approximately proportional to i b .
TI:"
ESC
VI
t-
~1
= ~2 = 70
Rk = 22,000 ohms
Ebb
A more practical circuit for realizing maximum possible gain is the cascode triode amplifier of Fig. 13-15. With tubes V1 and V2
identical, the gain of this circuit is
= 350 Volts
K=&=
E,, = 45 volts
fhn
-P (p
1) RL
(p
rv RL
+a
(13-12)
which reduces to
K = -g,R,
(13-13)
when
>> 1
P +Ebb
Successive manipulations of the linear equivalent circuit (Fig. 18-17) demonstrate the
nature of e,. When the generator voltage pek
(which has the same polarity as e k ) is considered as due to a fictitious resistor @ k and is
then lumped with Rk in the final step of Fig.
13-17, the identity of the cathode terminal
vanishes. Therefore, additional cathode loading cannot be applied to the final equivalent
circuit of Fig. 13-17. The equations for ek and
gain K are
In equivalent circuits, the symbol e, represents the variational component of grid-tocathode voltage. When the tube circuit has
internal feedback, e, does not equal the externally applied input voltage. For example, in
Fig. 13-16, which is the circuit schematic of a
cathode follower, e, is expressed as
e,
= e,, - ek
AMPLIFICATION
13-10
(13-17)
when
rp2
+Rk(1
p2)
>> rpl
and p1 >> 1
1 P1
These inequalities can exist in a practical circuit without reducing the V 1 plate current to
such a small value that rplis large.
13-1.26 Simple Feedback Amplifier
K=
Fig. 7 3- I8 White cathode follower.
'*
+ r~-+Rk(1
(13-18)
lLRL
RL
+p)
RI
R
2+p+l(l+p.+p2Q)
'*
'0
81"
'+++*
a -
R,
R2 + R3
ein
Where the voltage gain of a cathode follower must approach unity as closely as possible, the resistor Rr in Fig. 13-16 can be replaced by a second tube as in Fig. 13-20. The
gain expression then becomes
80
-F
AMPLIFICATION
(13-19)
Ro = r p Rk ( 1 p)
Because of t h e appearance of t h e t e r m
R k ( l p) and also because of the approximate constancy of p over a range of operating
conditions (also during aging of the tube),
the addition of Rk increases the gain stability
in the presence of variations in rp.It also lowers the gain and increases the effective output
resistance. Output voltage e, is expressed as
Rk(1
>> R L
+ rp
Then
(13-20)
when
RL = Rk = R
Thecircuitcan be used as a phase splitter;
i.e., a circuit with two outputs equal in magnitude but 180" out of phase. It can also be used
as a coupling circuit between a single-ended
input and a push-pull stage. The two output
voltages must always be less than the input
voltage.
13-1.27 Differential Amplifiers
(13-21)
(13-22)
where
+ +
+ +
p)
>> RL +
13-12
r p
f
Fig. 73-27 Plate-and-cathode-loaded amplifier.
P +Ebb
In the previous discussion of voltage-amplifier circuits, no attention was paid to the current levels in the amplifier load because the
power supplied to that load is of little consequence. However, in the final stage of a servo
amplifier, the amplifier load is the output
member of the servomechanism. Since the
load is usually a motor or a similar powerconsuming device, the final amplifier stage
must be a power amplifier with adequate output.
13-1.30 Tubes Used in Power Amplifiers
The tubes used for power amplification differ from those used for voltage amplification.
These differences include :
(a) Greater maximum plate-dissipation
ratings (e.g., 10 watts)
13-13
AMPLl FlCAT10N
W
Fig. 7 3-23 Push-pull amplifier.
13-14
particularIy when pentodes are used. Therefore, circuit values for this configuration are
more easily determined by experimental procedures.
13-1.34 Efficiency
When a single amplifier stage cannot prcvide the required gain, additional cascaded
stages must be supplied until the requirement
is met. In a cascaded amplifier, each successive
stage amplifies the output of the yeceding
stage. There are two methods of casaadiny,
each named according to the type of cotipling
circuit used between stages :
(a) Direct coupling, in which the totai instantaneous value of the output of one stagc
affects the next stage.
(b) A-c coupling, in which only the timevarying component of the output of one stage
affects the next stage.
13-1.36 Direct-Coupled Amplifiers
(13-26)
The voltage dividei R1,R, reduces the gain
but is needed to make the V2 grid negative
with respect to its cathode. The voltage divider
can be eliminated by coupling the plate of V1
tothe grid of V2 by means of a battery E
(Fig. 13-28). However, the separate battery
makes this coupling method impractical from
many viewpoints. Another way of eliminating
the voltage divider is to couple the plate of V1
directly to the grid of V2. The cathode of V2
must then be returned to a positive-potential
point, and RL2to a still more positive point.
13-15
AMPLIFICATION
PLATE
CHARACTERISTICS
OF TUBE V1
ec
= -k,
PLATE
CHARACTERISTICS
OF TUBE V 1
PLATE
CHARACTERISTICS
OF TUBE VZ
B
A.
PLATE
CHARACTERISTICS
OF TUBE V1
0
e; =
0;
16
PLATE
PLATE
CHARACTERISTICS
OF TUBE V2
CHARACTER1 TICS
OF TUBE V?
0.
C.
13-16
Ip
c1
ERROR
SICMAL
FIELD WINDING
OF A MAOiINE
IN
115 VAC
c2
REFERENCE
VOLTAGE e,.
-LJ
I
I
I
I l l
I I I
t--C
j+Ebb
1
t-)
13-17
AMPLIFICATION
~I~LzRLIRL~
(13-27)
(RL1 T P d (RL2 T p 2 )
The heavy drain on the power supply, due to
the large bleeder current, makes this method
undesirable in many applications. Figure
13-30 shows still another method of elimina t i n g t h e e f f e c t of t h e voltage divider.
Resistance R , in Fig. 13-27 is replaced by a
gas-discharge tube, such as a Type VRlOS
voltage regulator. A gas-discharge tube is
characterized by a very nearly constant terminal voltage over a limited current range
(e.g., 10 to 30 ma) so that it acts as an equivalent battery. Figure 13-30 is the simplest coupling circuit where signal levels are high
enough so that the gas-discharge-tube noise
(somewhat reduced by RN and C N )does not
reduce the ampiifier signaI-to-noise ratio below acceptable limits. Iannonec3) discusses in
detail the use of gas-discharge tubes as coupling devices in d-c amplifiers.
13-1.37 Problems encountered in directcoupled amplifiers. Not only are there practical difficulties in the design of coupling
circuits in directcoupled amplifiers, but the
very nature of direct coupling makes the output level of such an amplifier dependent
upon all amplifier components, tube characteristics, and power supply levels. Any variation
K=
13-18
+bb
e, - @t
e,,
2rp i- Rt
( 1 3-28)
13-19
( 13-29)
AMPLl FICAT10N
(13-31)
and
Fig. 73-33 Drift-compensated d-c amplifier.
13-20
Because the reduction of drift to a negligible amount often leads to considerable circuit
complexity, direct-coupled amplifiers are generally avoided in servomechanism design.
Instead, a-c coupled amplifiers are used. However, in an a-c coupled amplifier, low-frequency signal components are attenuated and,
since typical signal frequencies in a servomechanism occur in the 0 to 20 cps band, the
signal must first be modulated ; i.e., superimposed on an a-c carrier. Typical carrier
frequencies are 60 cps and 400 cps. The block
diagram of a servomechanism incorporating
an a-c coupled amplifier is shown in Fig. 13-34.
If the amplifier load responds to modulated
a-c signals, the demodulator is not used. An
example of such a load is a 2-phase induction
motor. If the input signal is already in the
form of a modulated a-c signal, such as the
output of a synchro control transformer, the
modulator is not used. Servo compensation is
shown ahead of the modulator in Fig. 13-34
because d-c compensating networks are simpler to design than a-c compensating networks.
COMPENSATION
WODULATOR
CARRIER FREPUENCYVOLTAGE
I
A-C
DEMODULATOR
AMPLIFIER
13-1.41 Two-stage a-c coupled amplifier. Figure 13-35 is the circuit schematic of a simple
two-stage a-c coupled amplifier. Resistors Rk
and capacitors C,furnish self-bias by raising
the average cathode potential above ground.
The grid circuit may then be returned to
ground potential, as shown, and the need for
a grid-bias supply is eliminated. Ckis large
P +Ebb
13-21
AMP11FlCATlON
enough (1to 40 microfarads) so that its reactance is negligible at the frequencies of interest, thereby eliminating the effect of R, on
dynamic response. For linear operation, the
value of R, is chosen so that
I&ct
(13-33)
where
(13-34)
C. HlGH.FREPUENCY EQUIVALENT CIRCUIT
(13-35)
(13-36)
13-22
where
where
dK' - per unit change in gain with
-K'
feedback
K = open-loop gain
down to 3 db)
-
f2
+ RL1
2nC,1 ( ~ p i R +
~ irplRg2+ R&L)
TPl
K
Since
then
feedback
1 - PK I > 1 for negative feedback,
dK'
< -d K
K
(b) Modification o f frequency response.
For example, addition of negative feedback to
a single-stage a-c coupled amplifier (Fig.
13-39) decreases the lower half-power frequency and increases the upper half-power
frequency. Thus,
K'
f2,
=f 2 ( 1
-B K m d
- BKmid)
(13-39)
(13-40)
fv
fl
1
g a b
f2
l+i-
feedback
Kmid= mid-frequency gain of amplifier
without feedback
(13-37)
where
= fl/(l
where
f3
f3
fl,
(1) For voltage feedback (signal proportional to output voltage fed back), the output
impedance is given by
1
2nCak
dK'
1
--K'
1-$K
dK
K
where
To= output impedance with voltage
feedback
(13-38)
13-23
AMPL I FICAT10N
----
I.o
w
CJ
4
I-I
0.8
0
>
0.6
I-
3
I-
,"
->
0.4
I-
ii
0.2
0 01
0.I
1.0
10
01
1.0
I00
10
ACTUAL F
CY
FREQUENCY FOR 7 0 . m w T RESPONSE
_ -f
12
0
u)
-I
W
-10
w
U
n
22
-20
0
W
$ -30
-I
W
a!
-40
I3.01
0.I
1.0
10
0.I
I .o
10
ACTUAL FREQUENCY
FREQUENCY FOR 70.mPOINT RESPONSE
T,
100
=I
f2
80
4
-I
I-'
LL
40
In
Y
I
4
PW
L
I-
40
n
4
80
0.01
0.I
10
1.0
0.I
CY
RESPONSE =
1.0
7;
13-24
100
10
ACTUALFRE UENCY
FREQUENCY FOR 70.7X!Olkt
RESPONSE
7;
2:
el
> zo
20 = r , + f i k
2, =Rk
KO =-p
z 0 -- TP
+ Rk(1 + p)
(3) When both voltage and current feedback are used, then
(13-43)
<
20.
2, = r,
$ =-1
KO = p
r,
2, = -
1+v
(2) For current feedback (signal proportional to output current fed back), the output
impedance is given by
Z, = Zo
+ 2,(1 - K O )
(13-42)
where
AMPLIFICATI0N
(b) Over-all gain is less than in an openloop amplifier with the same number of
stages. That is,
K=-- K
1- OK
(13-44)
13-1.46 Reliability
The greatest single cause of failure in electronic servo amplifiers is vacuum-tube breakdown. In many modern designs, vacuum tubes
have been replaced by transistors and magnetic amplifiers, but there are still cases
where dynamic requirements or environmental conditions dictate the use of vacuum
tubes. In these cases, the designer can do
much toward achieving amplifier reliability.
It is possible to design amplifiers that will
operate for thousands of hours without failure
by taking the following measures :
(a) Care in selection of tube types. There
are now available higher-quality tube types
than those originally developed for consumer
use. The highest-quality tubes that are welladapted to military applications are listed in
MIL-STD-200. Tube types should be selected
from this list whenever possible.
(b) Deruting of tubes. Particular care
must be taken to ensure that maximum tube
ratings are not exceeded. These ratings are
published in tube manufacturers catalogs and
in Military Specification MIL-E-1.Circuits
should be designed for operation well below
maximum tube ratings because tube life is
appreciably increased when the tubes are
thus derated. While there is no available analytical relationship giving tube life as a function of derating, ample experimental evidence
shows that derating of tube characteristics
leads to increased tube life. Maximum plate
dissipation, maximum plate voltage, maximum cathode current, and maximum bulb
13-26
13-1.47 Construction
In the physical construction of an amplifier, the dominant considerations are : adherence t o weight and s p a c e requirements;
freedom from unwanted electrical coupling
between components ; proper ventilation to
prevent excessive temperatures ; freedom
from mechanical vibration ; and accessibility
of compcnents for maintenance.
13-1.48 Maintenance
13-27
AMPLIFICATION
A thyratron is a thermionic tube containing a plate, a cathode, and one or more grids.
The thyratron envelope is filled with hydrogen, mercury vapor, or a noble gas such as
xenon. Hydrogen-filled thyratrons are used
only for high-voltage pulse work; they will
not be considered in the following discussion.
Mercury-vapor-filled thyratrons must be kept
between fairly narrow operating-temperature
limits, usually between 40C and 8O"C, thus
limiting the ambient temperatures in which
tlese tubes can be used. However, nearly all
13-28
0-
2!
R2
IM
0.1 TO 1.0
IC
PUT
---o
R27 ( 3 + 4
10K
2w
6
INPUT
GAI
CONTI
INPUT "8"
o--
vs
(8
+ 9)
+12Au7
+12ov
5%
47K
+3wv
I j ,
IdOK, LINEAR
R28
MAIN GAIN CONTROL
AMPLIFICATION
1600
1400
yI
>
0
1000
4
0
I-
800
1200
I-
a-
600
400
200
-20
-16
-12
-8
-4
0
+4
CRITICAL GRID VOLTAGE IN VOLTS
f8
13-30
resistive load. The output current of this amplifier is a pulsating direct current, the magnitude of which depends upon the grid voltage.
A common method of applying grid control is
shown in this figure, where a constant sinusoidal voltage is phase-shifted by an R-C network so that i t lags the plate voltage of the
thyratron by 90. This phase-shifted voltage,
frequently called the rider, is added to a variable d-c control voltage E , ; the sum of the
two voltages is then applied between grid and
cathode of the thyratron. A very small capacitor C2 is connected directly between grid and
cathode to prevent transient disturbances in
the plate voltage from producing excessive
voltages on the control grid through the plateto-grid capacitance.
Figure 13-43 shows the waveforms of plate
voltage and critical grid voltage of the thyratron as a function of time. The actual grid
voltage produced by the circuit of Fig. 13-42
is shown for two values of the control voltage
E,,together with the resulting load voltage.
If E, is positive (Fig. 13-43A), the grid voltage exceeds the critical grid voltage early in
the positive half-cycle of plate voltage. The
thyratron thus fires early, permitting the
plate voltage to be applied to the load over the
E,
D-C CONTROL
VOLTAGE
13-31
LOAD
AMPLIFICATION
tl
PLATE VOLTAGE
AA\
PLATE VOLTAGE
/
VOLTAGE
\
I
wt
/
/
/
\
/
\
+ I
' /
/
\.-./
Ad'
A. E, POSITIVE
6. E, NEGATIVE
respect to the plate voltage) varies in response to the control current of the magnetic
amplifier.
13-1.57 Thyratron-Amplifier loads
13-32
Next Page
SOURCE OF ALTERNATING
CURRENT WITH VARIABLE
PHASE ANGLE
THYRATRON
FIXED D-C BIAS
(IF NEEDED)
t-s
A X SUPPLY
B. SINGLE-PHASE FULL-WAVE
(SUPPLIES D-C TO LOAD)
A. BACK-TO-BACK
(SUPPLIES A-C TO L O W
TI
I-
3.PHASE SUPPLY
13-33
SERVOMECHANISMS
SECTION 4, POWER ELEMENTS AND
SYSTEM DESIGN
CONSISTING OF
C H A P T E R S 14-20
PREFACE
Section 4 of the handbook on Servomechanisms contains Chapters 14 through 20, which
discuss servo power elements and system design. The significant features of servo output
members are presented in Chapter 14 (Power Elements Used in Controllers) and in Chapter
15 (Mechanical Auxiliaries Used in Controllers). Chapter 16 is devoted to a discussion of the
typical procedure that can be used a s a guide when designing a servo system ; this is followed
by descriptions in Chapter 17 of two representative servo systems from existing Army
equipments. Supplementary design information is given in Chapter 18 (Auxiliaries Associated with Servomechanisms), Chapter 19 (Constructional Techniques) and Chapter 20
(Supplementary Tables, Formulas, and Charts).
For information on other servomechanism components and on feedback control theory,
see one of the following applicable sections of this handbook:
ORDP 20-136 Section 1 Theory (Chapters 1-10)
ORDP 20-137 Section 2 Measurement and Signal Converters (Chapters 11-12)
ORDP 20-138 Section 3 Amplification (Chapter 13)
An index for the material in all four sections is placed at the end of Section 4.
iv
CHAPTER 14
14-1 INTRODUCTION
GENERATOR
D-C MOTOR
POWER INPUT
3-OF ROTATION)
CONTROL
SIGNAL INPUT
(REVERSIBLE
POLARITY)
ARMATURE
POWER OUTPUT
(VARYING SPEED AND
REVERSIBLE
ROTATION)
D-c SUPPLY
(CONSTANT VOLTAGE)
A.
BASICTYPE
AMPLIDYNE
D-C MOTOR
COMPENSATING
WINDING
POWER INPUT
(CONSTANT SPEED
3-
OF ROTATION)
CONTROL
0
SIGNAL INPUT
(REVERSIBLE
POLARITY)
0
ARMATURE
POWER OUTPUT
(VARYING SPEED AND
REVERSIBLE
RoTATIoN)
D C SUPPLY
(CONSTANT VOLTAGE)
B. AMPLIDYNE TYPE
Fig. 74-1 Rotary electric amplifier circuits for control of d-c motor.
14-2
speed regulation than series motors. Permanent-magnet motors a r e easy to drive since n o
power is used f o r the field. A limitation is
that the field may be demagnetized if the motor is badly overloaded. Conventional d-c motors a r e unidirectional devices. The arm atu re
rotation of series and shunt motors can be
reversed by reversing the current flow
through either the armature o r field but not
both simultaneously. Permanent-magnet motors a r e reversed by reversing the direction
of armature c ur ren t; high torques on reversal a r e obtainable.
Many means a r e available fo r controlling
the speed or armature position of d-c motors ;
they a r e
Rotary electric amplifiers
High-vacuum tube amplifiers
Gas-filled tube amplifiers
Magnetic amplifiers
Relay amplifiers
The generator, a basic rotary electric amplifier (Fig. 14-1A), operates a t a constant speed
and produces a n output voltage th at is proportional to the input signal applied to
its field. The output voltage developed across
the low impedance of the generator arm atu re
excites the d-c motor armature with comparatively low power losses. The amplifying action of the generator produces power gains
up to 100. A variation of the basic rotary
electric amplifier is the amplidyne (Fig. 141 B ) . A short-circuited brush connection in
the armature of the amplidyne makes it possible to obtain high power gains of 3000 to
10,000. By this means, it is feasible to control
5- to 10-hp motors from a source capable of
delivering only 5 to 10 watts of input signal.
The higher power gain of the amplidyne is
achieved at the expense of a slower response.
Rotary electric amplifiers (see Par. 13-4) a r e
used when large motors a r e to be controlled ;
when drive motors less than 1/4-hp rating a r e
to be controlled, this type of amplifier becomes impractical.
14-3
0-C MOTOR
CONTROL
SIGNAL INPUT
(REVERSIBLE
POLAR1TY)
-3 -
0.c SUPPLY
+bb
---
WHERE
RD
1,
V,
R,
CURRENT-LIMITING RESISTANCE
RATED ARMATURE CURRENT
D-C SUPPLY VOLTAGE
AWATURE RESISTANCE
14-4
POWER
OUTPUT
I
I-
---
SERIES MOTOR
SPLIT
FIELDS
ARMATURE
A
I-
- 3 .-
.,,
-M+
- -POWER OUTPUT
WPPLY
A. SPLIT-FIELD SERIES MOTOR (NO PLATE TRANSFORMER)
I-
FIELD 1
---
SERIES MOTOR
<
ARMATURE
POWER
- 3 - - OUTPUT
SUPPLY
FIELD 2
- POWER
OUTPUT
14-5
of armature control powered fram a multiphase supply. He also states that, with 3phase power, a 10-hp velocity servo with a frequency response flat to 15 cps is possible;
position control with a flat response to 10 cps
with the same motor is claimed.
Magnetic amplifiers (see Par. 13-3) provide more reliable control of d-c motors than
do vacuum-tube or relay amplifiers. Figure
14-4 shows a typical magnetic amplifier with
two cores for controlling the position of a
series motor with split-field windings.(') By
reversing the polarity of the input signal,
either direction of motor rotation can be
obtained. If a separately excited motor with
constant-voltage field supply were used, excessive circulating current would result when
the motor operates in one direction.(8) Magnetic amplifiers can be used to control the
speed of d-c motors by field control o r armature control.(g)One circuit (Fig. 14-5) f o r unidirectional speed control uses a large inductance in series with the armature circuit to
I
0
CONTROL
SIGNAL INPUT
(REVERSIBLE
POLARITY)
CORE 1
r - i
SPLIT-FIELD
SERIES MOTOR
*A*+
FIELD 1
L - J
BIAS SUPPLY
IY
.D 2
'I
f=F
6
A-C SUPPLY
CORE 2
14-6
lii
1i
CONTROL
SIGNAL
INPUT
_--
2-
POWER
OUTPUT
ARMATURE
D-C SUPPLY
LCORE 2
L
m-
A-C SUPPLY P
Fig. 14-5 Magnetic amplifier circuit for unidirectional confrol of shunt motor.
OF
In the general case for all types of d-c motor control, the following equations express
the steady-state behavior of the motor :
Vt = IttR,,
+ KIfo
T = KIfI,,
(14-1)
(14-2)
The above general equations may be modified for the following particular types of motor control:
( a ) Armature control with constant field
current
where
(14-3)
14-7
POLARIZED RELAY
r---i
CONTROL
SIGNAL INPUT
(REVERSIBLE
POLARITY)
:%{
I
j
SPLIT WINDINGS
SUPPLY
A. SERIES MOTOR CONTROL (NO DAMPING)
RELAY 1
r - - -
CONTROL
SIGNAL INPUT
FOR MOTOR
CWROTAnON
'F&
POWER
OUTPUT
r--CONTROL
SIGNAL INPUT
FOR MOTOR
CCW ROTATION
I---RELAY 2
14-a
The stall torque T, which occurs a t the maximum armature terminal voltage V t , is found
from Eq. (14-4) when T = TJr, Vt = VfJI,and
0 = 0.
(14-5)
K
The no-load speed wy a t the maximum armature terminal voltageVtJ, is found from Eq.
(14-3) when w = wJI, Vt = ViJI, and T = 0.
wg
V,,
=K1
(14-8)
K,
Solving Eq. (14-8) for -
viu
K1 --
1
-
(14-9)
Vt,
Wlf
Substituting Eq. (14-9) into Eq. (14-7)
(14-10)
031
2n X 3000
60
= 314 radians/sec
At the point w/wM = 1 on Fig. 14-7, the no-
V #- 0
-T --(14-11)
TM
V ~ M 0.v
A plot of T/TM versus w / o x characteristics
with Vt/V,, as a parameter is shown in Fig.
14-7. The curves of this figure are nondimensionalized and based on Eq. (14-11).
If the manufacturer's data for a particular
motor are known, the curves can be dimensionalized for that motor. For example, a
motor has an armature resistance of 5 ohms
and a no-load speed of 3000 rpm at a maximum armature terminal voltage of 90 volts.
load speed o , is
~ 314 radians/sec. All other
points on the w / o Maxis are dimensioned proportionally.
Substituting 314 for ox and 90 for V f Min
Eq. (14-9) and solving for K, gives
K1 =-Vt,
%
I
-- 90
314
= 0.286
14-9
K,
T.ii = - V+.i,
R,
(14-5)
1.0
1.0
I IM
i
I
0.9
0.8
= 5.15 newton-meters
-t
0.7
0.6
0.5
0.5
I
&
0.4
0.3
0.2
0.1
T = KIfl,,
(14-2)
T.1,
KI,.iiI,
(14-12)
T T.ir
I!
-
(14-13)
Il.11
where
If = I,M
A plot of Eq. (14-14) is shown in Fig. 14-9.
14-2.6 Series Motor Control
Vt
I , Ra
T = KZ,'
14-10
+ KI, o
(14-15)
(14-16)
g
w
z
3
I0
-
s
0.0
-
-0.1
SPEED
(a-
P4
-O.*
(14-18)
v
14-2.7 D Y N A M I C CHARACTERISTICS
D-C M O T O R S
OF
14-1 1
where
The dynamic characteristics for the armature-controlled motor are expressed by the
equations
V , = IfR, + L,s I ,
T = TI, + BO + JSO
(14-22)
(14-23)
where
V , = field terminal voltage, in volts
TL
14-12
The dynamic behavior for the field-controlled motor with constant armature voltage
is expressed by the equations
V , = Z,R,
dZ
+ KZF + L 2
dt
T = KIiZa
(14-29)
a''
--
(14-24)
aIi
(14-2)
-=
(14-30)
-aT
-aT
-KZ,
a],
aza
KZ,
,
(14-27)
Fig.
KIi
ao
(14-25)
The terms KZ,w and KIiZ, contain the products of two variables, and introduce a nonlinearity. If the back-emf term KZiw is
denoted as vb, the variation in the quantities
v
b and T can be written as
-KO
Vto = Z a o Ra
+ KZto
0 0
TO = KZfJ,o
(14-32)
(14-33)
To = T L ~ Bwo
J4-J 2 Block diagram for field-controlled d-c motor, constant armature current.
14-13
(14-31)
(14-34)
T o = KZ~J,, = T,,
(14-35)
v, = vb,,+ hvb
= o,+ Ao
1, = I,,
+ Al,
Klfi
+ Aza
T = J - d(Aw)
dt
(14-39)
T o 1KZ,,I,,
J d(Aw)
dt
AZj
+L
('Ia)
+ KZ,,, Aw
dt
(14-41)
14-14
AT = KI,, AIf
=J
dt
+ KI,, A I ,
+ BAo + AT,,
(14-42)
= (Js + B ) A o
where
T = L/R,
+ AT,
(14-46)
T h e dynamic characteristics f o r th e
series-connected motor are expressed by the
equations
V , = IaR,
(14-44)
T,
+ KI,o + Lad,
+ BW+ JSO
(14-47)
(14-23)
The runaway speed of the series motor depends upon the load; for example, point X
'f
Fig. 7 4-13 Block diagram for field-controlled d-c motor, constant armature terminal
voltage, incremental behavior.
14-15
To compute the dynamic response of a rotary electric amplifier, its transfer function
must take the source into account. For the
v,=
a1V1 ( s )
l+t,s
a,t;(s)
1+TfS
(14-48)
(14-49)
where
(KZ,)
K O =-
RT
RTBT
= total armature resistance of generator and driven motor, in ohms
BT
14-16
JT
LT
W(5)
(14-50)
where
W"
da2+ b2
~
5 =a%
For the case of the amplidyne type of
rotary electric amplifier (Fig. 14-1B), the
expression for the generated voltage V, given
by Eq. (14-48) must be modified. Refer to
Par. 13-4.
R,
RT = R t + R ,
J
Td
=B
AVf
= amplifier gain = -
AE,
14-17
R, = armature resistance
L, = armature inductance
R, = field resistance
L, = field inductance
K = motor constant
= armature inertia
B = viscous damping
(14-52)
Precise measurements of the armature inductance L,lare often unnecessary because this
parameter does not affect appreciably the
principal time constant of the motor. Probably the most convenient method of measuring
La,and one of sufficient accuracy, requires a n
a-c voltmeter and a n a-c ammeter. A 60-cps
voltage is first applied to the armature which
is blocked to prevent rotation; the field is
excited at rated d-c field current. Measured
where
L, =
A
-
(14-53)
377
An alternate method of measuring the armature inductance is to observe the transient
growth of current when the armature is excited by a step voltage.(*)The armature should
be blocked to prevent rotation and the field
should be excited a t rated current. The step
change in voltage is applied by closing a
switch connected in series with the armature,
a series resistance R,, and a d-c source. The
armature inductance can be calculated from
the time-constant relationship of the voltage
rise across the series resistance as observed by
the aid of a cathode-ray oscilloscope ; that is,
L a = (T*)(Ra
+ R,)
(14-54)
where
14-18
where
rn
(14-55)
The stall torque should be measured at various angular positions of the armature, and a n
average of the maximum and minimum values
should be used. If the motor is used with a
constant field input, the value of normal field
current should be used during the test.
A second method of determining the motor constant is to measure the armature current and speed for a given terminal voltage
V,. Knowing the armature resistance, the
motor constant is then calculated from Eq.
(14-1)by solving for K to obtain
(14-58)
where
J , = known inertia of body added to armature, in kilogram-meters2
Solving Eq. (14-58)for armature inertia J
(14-59)
(14-56)
14-2.20 Armature Inertia
(z)
Tj
(14-60)
where
(14-57)
14-19
W = nplr2
= weight density, in lb/in.3
I = cylinder length, in in.
= cylinder radius, in in.
T
g = 386 in./sec2
The viscous damping B should be determined when the motor is operating without a
load and at rated field input. The armature
current should be measured when only
enough voltage is applied to the armature to
obtain rated motor speed. The total armature input power is then
where
where
14-20
Rated
Voltage
(volts)
0.008
115
Armature
Resistance
(ohms)
Inertia
(kilogrammeters')
permanent magnet
0.003
0.011
2.15
permanent magnet
0.009
0.0091
0.82
permanent magnet
0.022
0.0093
0.38
permanent magnet
0.054
26
0.015
2.64
permanent magnet
0.080
26
0.0143
1.85
shunt
0.6
80"
0.198
5.82
x lo-"
0.24 x lo-'
0.37 x
0.53 x
0.37 x 10."
0.23 x 10."
2.9 x 10-4
0.329
1.57
shunt
0.23
280
1.1
Stall
Torque
(newtonmeters)
Stall
Torque/
Inertia
(sec-')
Motor
Time
Constant
R.J/x'
0.095
8600
0.061
0.025
10,200
0.043
0.062
17,000
0.036
0.141
27,000
0.023
0.135
37,000
0.043
0.181
34,000
0.047
9300
0.043
13,200
0.016
2.72
loo**
shunt
*Armature voltage
"Field voltage
2.5
65*
110**
x 10-3
13.6
Motor
No.
Nominal
Size
(hp)
Motor
Constant
(newtonmeters/
ampere)
$,
K,
+ ( K , - K,) e,
(14-64)
where
c
Ebb
B. COIL CONNECTIONS
14-22
-30
- 60
- 90
-120
-150
- I80
-210
10
20
50
30
200
500
A. STATIC CHARACTERISTIC
B. FREQUENCY RESPONSE
1000
(14-66)
T, = K,, E,
(14-67)
where
Next Page
To Convert
English Units
Quantity
To
Metric (Mks) Units
Multiply
BY
mass (M)
lb-sec'/in.
kilograms
175
inertia (J)
lb-in.-sec'
kilogram-meters*
0.113
force (F)
lb
newtons
4.45
torque (T)
lb-in.
newton-meters
0.113
lb-in.-sec
newton-meter-sec
0.113
lb-in.
newton-meters
0.113
With zero speed and maximum current, ZfM, the stall torque T , is seen to be KZf,,[Vi/R,,.
Divide the expression for T from Eq. (1) by TdIon the left and the equivalent expression
on the right. There results.
Substitute Eq. (3) into Eq. (4) to arrive at the following [which is Eq. (14-14)]
14-24
Previous Page
115 VAC
I,
I 1
6 L6
0.Spf
:
:
SHADING
COILS
15h
15K
ERROR
SIGNAL
TOGEAR
TRAIN
220n
*
NOTE: ADDITIONAL AMPLIFICATION AND COMPENSATION MAY BE REWIRED.
14-25
0
X
"
14-26
Fig. 14-17 Amplifier control circuit for 2-phase servomotor.
OPEN FOR
DOUBLESPEED
HIGH SPEED
t
SINGLE SPEED IN
...
190V
FROM POWER SUPPLY
Next Page
';
R'8
OJ
L
IC
A.
C.
Fig. 14- 19 Full-wave output-sfage circuit of a magnetic servo amplifier having inherent
dynamic-braking properties: (A) with two separate control winding elemenfs N'c N',.
(6) with series-connected d-c control winding elements N'r N",.
( C ) with series-connected a-c control winding elements N , N",
By permission from M a g n e t i c - A m p l i f i r s Circurte. by W . A.
Geyger, Copyright 1957, McGraw-Hill Book Company, Inc.
CHAPTER 15
15-1.1 PURPOSE
The discussion of gear trains will be confined to gears with involute tooth form, because this tooth form is used almost exclusively for engineering purposes. Other tooth
forms which result in rolling action between
meshing teeth have been made, but these
special tooth forms are confined to very specia1 applications, and cutters for their manufacture are not normally available.
'By J . 0.Silvey
15-1.2 DEFINITIONS
15-1
PITCH
CIRCLES
DEDENWM
1
CENTER DISTANCE
WLAR
C
PITCH
FACE
WIDTH
\
PRESSURE ANGLE Q
PITCH D ~ ~ M E T E R
BACKLASH
15-2
(h) The pyessure angle is the angle between a tangent to the tooth profile a t the
pitch circle and a radius to this pitch point.
Pressure angle is also the angle of inclination of the sides of the teeth in the basic
rack.
(i) The center distance is the shortest distance between gear axes of a pair of mating
gears. For gear pairs with parallel shafts,
the center distance is the distance between
axes.
(j) The backlash is the play between mating teeth or the shortest distance between
the nondriving surfaces of adjacent teeth.
(k) The g e a r ratio of a pair of mating
gears is the ratio of the number of teeth on
the gear divided by the number of teeth on
the pinion.
15-1.3 GEAR TYPES
15-3
A. SPUR
U. HELICAL
C. INTERNAL
cd
D. BEVEL
15-4
gear pairs must also be great enough to prevent binding due to eccentricities of the bearings, shafts, and gear-pitch circles. Backlash
standards for general-purpose spur gearing
have been established by the American Standa rd s A~sociation,(2.3,~)
and a method of specifying backlash in fine-pitch gearing has also
been adopted.(:)
The required amount of backlash can be
obtained by cutting the teeth thinner than
the theoretical optimum dimensions. This is
achieved by cutting the teeth deeper than
normal. I n most cases, the teeth of each gear
of a mating pair a r e cut thinner by a n
amount equal to half the required backlash
to retain a s much strength a s possible in
each gear. However, where small pinions a r e
used, all of the backlash should be obtained
by cutting the teeth of the mating gear thinner by a n amount equal to the total required
backlash. The excess depth of cut X to provide the required backlash when both gears
a r e cut deeper is
x=-
4 sin 4
( f o r both gears)
(15-1)
B
(one gear only)
2 sin 4
Xz-
(15-2)
where
z excess
+W
(15-3)
where
w =
339000H P
( 15-4)
w
d
15-5
z total
dynamic load, in lb
C=
E,+z
(15-7)
E,
(15-5)
C=
0.107e
(for 14-1/2-deg tooth
1
1 form)
-El+ -
where
C=
0.115e
1
I
I
1
Material
Tooth Form
(deg)
14-1/2
14-1/2
14-1/2
400
550
800
20 (full depth)
20 (full depth)
20 (full depth)
415
570
830
20 (stub)
20 (stub)
20 (stub)
430
590
860
Adapted from Manuel of Ccar I l r s i g n . Vol. 11. by Earle Buckingham. The Industrial Pwasu. N e w York City, with permission.
15-6
1600
2200
3200
1140
1180
1720
0.003
0.004
0.005
2400
3300
4800
3200
4400
6400
4000
5500
8000
1660
2280
3320
4150
5700
8300
1720
2360
3440
4300
5900
8600
5160
reduced to even smaller amounts. The foregoing conditions of error in action are listed
in Table 15-2 and are designated as Classes
1, 2, and 3.
15-1.7 Gear Accuracy
Diametric
Pitch
Class 1
Class 2
Class 3
0.0048
0.0024
0.0012
0.0040
0.0020
0.0010
0.0032
0.0016
0.0008
0.0026
0.0013
0.0007
0.0022
0.0011
0.0006
0.0020
0.0010
0.0005
6
and finer
Adapted from Mnnual of Gear Design, Vol. 11, by Earle Buckingham. The Industrial Press. N e w York City, with permission.
II
V
(ft/min)
Error
(in.)
V
(ft/min)
Error
(in.)
V
(ft/min)
Error
(in.)
250
0.0037
1500
0.0019
2750
0.0010
0.0032
1750
0.0017
3000
0.0009
0.0028
2000
0.0015
3250
0.0008
1000
0.0024
2250
0.0013
3500
0.0007
1250
0.0021
2500
0.0012
4000
0.0006
Adapted from Manual of G a r lhv gn. Vol. 11. by Earle Buekinaham. The Industrial Press, New York City, with permission.
15-7
V
(ft/min)
Error
(in.)
4500
0.0006
5000
and
over
0.0005
StljFy
(15-8)
W,<= D F K Q
where
where
(15-10)
(15-11)
= load-stress factor
s , . sin
~
K =
1.400
where
(zF )
+
(15-15)
where
w
d
(15-13)
(15-9)
Q = ratio factor
(15-12)
s,.
Number
of Teeth
14-1/2' Composite
and Involute Form
20" Stub
Tooth System
0.067
13
14
15
16
17
~~~
~~
0.083
0.075
0.088
0.078
0.092
0.081
0.084
I
I
0.094
0.096
I
I
0.103
0.108
0.111
0.115
0.117
18
0.086
0.098
0.120
19
0.088
0.100
0.123
20
0.071
0.090
0.102
0.125
21
0.092
0.104
0.127
22
0.093
0.105
0.129
24
0.095
0.107
0.132
26
0.098
0.110
0.135
28
0.100
0.112
0.137
30
0.101
0.114
0.139
34
0.104
0.118
0.142
0.106
0.122
0.145
43
0.108
0.126
0.147
50
0.110
0.130
0.151
60
0.113
0.134
0.154
75
0.115
0.138
0.158
100
0.117
0.142
0.161
150
0.119
0.146
0.165
300
0.122
0.150
0.170
Rack
0.124
0.154
0.175
38
Adapted from Manual of Gear Design, Vol. 11. by Earle Buckingham. The Industrial Press, New York City. with permission.
15-9
Material
Brinnell Hardness
Number
Gray iron
Semi-steel
Phos. bronze
St
160
12,000
190
100
18,000
24,000
Steel
Steel
Steel
150
36,000
200
240
50,000
Steel
280
70,000
Steel
320
80,000
Steel
Steel
360
400
90,000
100,000
60,000
from their own specific properties. The surface endurance limit of gear material is a
measure of the load limit in psi that the material will tolerate before deformation occurs. The values of surface endurance limit
sc for steel apFear to vary quite consistently
with the Brinnell hardness number up to
about 400 Brinnel hardness. Values of s, for
hardness numbers from 150 to 400 are listed
in Table 15-6. When the Brinnell hardness
number is much higher than 400, the steel
does not appear to have a definite endurance
limit. The values given in Table 15-7 are suggested for use with steels harder than 400
Brinnell hardness. Table 15-7 also gives
values for the load-stress factor K for these
harder steels. The load-carrying ability of a
pair of metal spur gears, for example, may be
limited by either the beam strength of the
gear tooth, or by the surface endurance limit
of the material. The lower of these two
values should be used to establish the loadcarrying ability of any given pair of gears.
Spur gears designed to conform to the preceding requirements can be expected to operate successfully with average loads. However, if heavy inertia loads are to be coupled
rigidly to the gears, the dynamic load should
be calculated in accordance with the method
described by Buckingham.(3#*)
Other types of gears such as internal, helical, bevel, and worm gears are used so infrequently in control work that the particulars
of their design are not included here. If use
of any of these types is contemplated, the
following references should be consulted :
15-10
Material
in Pinion
Brinnell
Number
Material
in Gear
Brinnell
Number
Steel
Steel
Steel
150
200
250
Steel
Steel
Steel
150
150
150
50,000
Steel
Steel
Steel
200
250
300
Steel
Steel
Steel
200
200
200
70,000
Steel
Steel
Steel
Steel
Steel
Steel
250
300
350
Steel
Steel
Steel
300
350
400
Steel
Steel
Steel
3,
K
(14-1/2")
K
(20")
30
43
58
41
58
79
80,000
90,000
58
76
96
79
103
131
250
250
250
90,000
100,000
110,000
96
119
144
131
162
196
Steel
Steel
Steel
300
300
300
110,000
120,000
125,000
144
171
186
196
233
254
350
400
400
Steel
Steel
Steel
350
350
400
130,000
140,000
150,000
201
233
268
275
318
366
Steel
Steel
Steel
150
200
250
Cast iron
Cast iron
Cast iron
50,000
44
87
144
60
119
196
Steel
Steel
Steel
150
200
250
Ph. bronze
Ph. bronze
Ph. bronze
50,000
70,000
85,000
46
91
135
62
124
204
Cast iron
90,000
193
284
Cast iron
60,000
70,000
70,000
90,000
15-11
K
(14-1/2")
K
(20")
Brinnell
Number
s,
S,
(14-1/2")
(20")
450
188,000
421
575
450
170,000
344
470
500
210,000
525
718
500
190,000
430
588
550
233,000
647
884
550
210,000
525
718
600
255,000
775
1058
600
230,000
630
861
i
,
50,000,000 Repetitions of Stress
550
147,000
165,000
182,000
200,000
394
544
476
450
500
132,000
208
284
148,000
261
356
550
163,000
316
432
600
179,000
382
522
OF
than 0.0002 inch a t the pitch line are difficult to achieve, and a r e obtained only by a
high degree of accuracy in the machine on
which the gears a r e cut. Often, however, such
highly accurate gears a r e obtained from a
routine production run by careful selection.
In general, the more-precise profiles a r e obtained only at increased cost. Most gears fo r
use in control work can be obtained with
tooth-profile errors of less than 0.001 inch,
and gears of 20 diametric pitch or finer caii
be produced with tooth errors of less than
0.0005 inch a t little additional cost.
( b ) E r r o r s in tooth spacing. Tooth-spacing errors a r e the result of indexing errors,
and usually a r e a result of the inaccuracy of
the machine on which the gear is cut. Gears
with a maximum tooth-spacing e rro r below
15 seconds of a r c a r e seldom produced, the
usual maximum accumulated spacing errors
being about 3 minutes of arc.
15-12
Gearing inertia is often an important portion of the load on a servomotor. The inertia
of any gear in a gear train, referred to the
driving point, is
Jr=
N'
(15-16)
Backlash results in a region of input motion within which no variation of the output
can be detected. When placed within a control-system closed loop, this type of "dead
zone" may produce system instability, and
usually causes the system to oscillate at a low
frequency, through a n angle of one to five
times the backlash angle. Backlash can be reduced by making the driving gear large (to
reduce the angle of backlash) or by using
gears with small backlash. Backlash can be
eliminated by spring loading the gears.
Three methods of eliminating backlash by
spring loading are shown in Fig. 15-3. The
method shown in Fig. 15-3A uses a spring
15-13
COMPR ESSlON
SPRING
15-14
on the output shaft to exert enough continuous torque to maintain a load on one side of
the gear teeth. This method can be used only
when there is ample torque available t.0 drive
the load, and only when the angle of rotation of the gear is limited to a few revolutions. In Fig. 15-3B, the gear is split in half,
forming two identical gears. One gear is attached rigidly to the shaft, and the other is
free to rotate on the shaft when driven by
expansion of the spring. When the gears a r e
to be meshed with a pinion, the free gear is
rotated manually on the shaft to compress
the spring. After meshing, the free gear is
released and the spring presses the teeth on
the two halves of the split gear against opposite sides of the pinion teeth, thus maintaining bidirectional contact between pinion
and gear teeth. The method shown in Fig.
15-3C is similar to the split-gear method, except that it can be used to simultaneously
load several gear meshes. In this method, the
pinion on shaft A is split and spring loaded,
thereby loading the entire train. The gear
trains shown in Fig. 15-3B and 15-3C a r e
capable of revolving continuously and therefore have a constant loading torque. Any of
the three spring-loaded gear-train systems
shown in Fig. 15-3 adds to the torque required
to drive the gear train because it adds to the
load on the bearings and to deformation of
the gear teeth.
The optimum amount of spring loading
applied to a gear train can sometimes be determined only by trial-and-error. Theoretically, the spring load should be a t least equal
to the maximum torque the gears a r e required
to transmit. In practice, however, the spring
load can often be reduced below this value,
All gear trains have an effective spring constant as the result of elasticity of the materials. In the presence of heavy inertia or
coulomb-friction loads, the compliance (reciprocal of spring constant) of the output
gear train of a servomechanism may result in
a low system gain, o r a highly undesirable
resonant frequency. Gear-train compliance
comes from the following three sources:
(a) Elastic deformation of the gear teeth
(b) Torsion and bending of the shafts
(c) Deformation of the bearings and housing.
Normally, torsion and bending of the shafts
are the only significant factors. The total
compliance of a gear train can be computed
by adding the compliances of each gear shaft
after referring them to a common point. The
referred compliance of a shaft is
c, =C
(15-17)
where
C, = referred compliance
N is greater than 1.0 when the actual shaft
rotates slower than the reference shaft, and
is less than 1.0 when the actual shaft rotates
faster than the reference.
15-15
Next Page
and B. In both types of differentials, the relation between the rotation of the three
shafts is
20:1 = 01 f
02
(15-18)
r=l
A. AUTOMOTIVE-TI PE DIFFERENTIALS
B. BEVEL-GEAR DIFFERENTIAL
15-16
CHAPTER 16
TYPICAL PROCEDURE*
16-1 INTRODUCTION
16-1
16-2 G A T H E R I N G O F SPECIFICATIONS
The first step in the design of a servomechanism is to define what the system is
to do and how well it is to do it. As pointed
out in Ch. 1, a servomechanism may perform
any or all of the following functions:
( a ) Bring about a change in the actual
value of the output so th at it is in conformity
with a desired value a t all times.
( b ) Minimize the effect of disturbances;
i.e., variables other than the desired value of
the output.
( c ) Minimize the effect of varying component performance on the output.
The designer must first determine which one,
or combination, of these functions is to be
achieved.
The degree of excellence to which the servomechanism must carry out the above functions is established by the servo performance
specifications. These specifications, furnished
by the system designer, are based on a set
of system specifications established by the
ultimate user of the system.
Servo performance specifications must be
given in terms of the desired value of the
output of the servomechanism for a given
value of the input signal to the servomechanism. The input signal encountered may be
one of the following types:
( a ) Aperiodic, noise free
( b ) Aperiodic, with noise
( c ) Periodic, noise free
( d ) Periodic, with noise
( e ) Stochastic, noise free
( f ) Stochastic, with noise
Noise is regarded as any inpct-signal variation that is not a measure of the information
carried by the input. The desired value of
the output will be expressed in terms of a
signal consistent with the input signal and
will therefore be one of the types listed.
Some typical specifications f o r the six types
of input signals listed above are as follows:
TYPICAL PROCEDURE
or disturbance of given autocorrelation function shall not have an rms value in excess of
a specified value.
In addition to the specifications on system
performance discussed above, the servomechanism design is influenced by environmental conditions. The designer must know
the allowable values of such physical properties as over-all servo weight and size. Characteristics of the available source of power
must also be known. Important among
these characteristics are :
( a ) Form, such as 60-cps or 400-cps alternating current, hydraulic oil at 3000 psi,
compressed air at 10 psi, 440-volt direct current, etc.
(b) Capacity, such as 115 volts, 15 amperes, 3000 psi at a maximum of 2 ft3/min,
etc.
(c) Regulation, such as 115 f 1 volt at
400 & 2cps.
16-3
v.,,,.,
and the inertia associated with its output
shaft, J,,,. Other procedures for choice of output motor can be found in Chestnut and
Mayer(5) and Ahrendt.(G)
16-3.2 DETERMINATION OF M O T O R SIZE
16-2
where
*See Appendix on page 16-8 for equation derivations.
16-4
TYPICAL PROCEDURE
where
tMP
= motor time constant
(16-4)
16-5
(16-5)
(16-6)
(16-7)
TYPICAL PROCEDURE
TABLE 16-1
COMPARISON OF SERVOMOTORS
Servomotors
A-C
jervomotors
Pneumatic
Actuators
Hydraulic
Motors
MagneticParticle
Clutches
High
Low
High
High
Low
Commutators
and brushes
Problem
None
None
None
None
Filter required
Electrical
Radio Noise
Suppression
None
Remove
foreign
particles
Remove
foreign
particles
None
Low
Efficiency
Flammability
Low
Low
None
High
(unless
special oils
used)
Explosion hazard
High
Medium
Low
Low
Medium
Coulomb friction,
stiction,dead-spot
High
Low
Medium
Medium
Medium
Long
Medium
Medium
Short
16-7
TO CHOICE
Assumptions
Motor has capability of delivering peak torque T J r p
Motor can operate a t peak velocity vJrp
Motor has a n inertia JJI
Load requires a speed-independent torque T,,
Load requires a peak velocity
Load requires a peak acceleration a , [ ,
Load inertia is J ,
Gear-train reduction ratio from motor to load is R.
Units. Any self-consistent set can be used. Two examples:
Torque
Velocity
Alcelerat ion
Inertia
newton-meters
radians/sec
radians/sec2
kg-meter2
radians/sec
radians/sec'
lb-f t-sec2
~~
foot-lbs
Derivation
Peak motor torque must a t least equal total load torque referred to motor ; thus
Peak motor speed must at least equal required load speed referred to motor; thus:
V.VP
2 RV1.P
Eq. (1) sets upper and lower limits on R. Equation (3) sets a n independent upper limit or
R. In order f o r the motor to drive the load, the minimum ratio allowed by Eq. (1) musl
be less than or equal to the maximum value allowed by Eq. ( 3 ) . Solve Eq. (1)for minimun:
value of R ; thus :
TYPICAL PROCEDURE
Note that Eq. ( 5 ) involves motor capabilities and load requirements independent of gear
train. Two conditions can be found from Eq. (5). First, the radical must be real ; therefore :
( 5 ) as follows:
Jdf
2al,r X
VLP
VYP TMP
( T.iiP2/J.i~
1-
then the right side of Eq. (7) will be less than zero, and so long as the radical is real, Eq.
(7) will be satisfied. Thus, for Case I, no additional constraint is found from Eq. (7).
Case ZZ. If
when Eq. (7) is satisfied, then Eq. (6) will automatically be satisfied. To find the conditions for satisfying Eq. (7), square both sides. There results :
JM
To summarize :
Case I . If
I then
I Case ILIf
I then TMp*/JM must be greater than 4aLp ( J , aLp+
16-9
The requirements for Cases I and I1 can be rewritten by multipIying the equations respectively by
= vl.P/aI.p
Also define peak-load and peak-motor powers as :
t1.r
vI.r (J L a1.P
VJtr X
+ T L P )= PLP
T N P= PMP
TI,P
Case 11.
The following derivation is for the equations used in the choice of a gear ratio. As mentioned
in the chapter, it is always possible to find a motor to drive any load as specified. Then it is
only necessary to find a gear ratio. The procedure ensures that once the motor is found, a
realizable gear ratio will result.
Multiply Eq. (1) by the factor
TYPICAL PROCEDURE
The result is :
and
5 1
(16)
V.IIP/VI,P
a nondimensional gear ratio and combine Eqs. (15) and (16) so that the
Consider
vJIP/vLP
result is :
The lower value of the two upper limits sets the upper bound on R. Define a motor over
power factor as
where ' J ' ~ is the minimum allowable power ratio and &2E is the actual power ratio
P1.r
min
PLP
for the motor chosen. From Eqs. (13) and (14), it is seen that
TLP
and
= 4- TJlP
TLP
16-11
Next Page
P,, F . 4 CT = Fa-9 1 1 P
PL'P
PLP
min
and
Substitute Eq. (20) into Eq. ( 1 7 ) . The two minimum values of R / ( v y p / v L p are
) found
directly from the left-hand side of Eq. ( 1 7 ) . The maximum values of R/(v.,IP/vr,P)
are
found from either the upper or lower inequality of the right-hand side of Eq. (17). The
condition as to whether the upper o r lower inequality of Eq. (17) holds depends upon which
of the two is the greater. These results all appear as Eqs. (16-4) through (16-7).
16-12
CHAPTER 17
REPRESENTATIVE DESlGNS"
17-1 INTRODUCTION
The examples of servomechanism systems
described in this chapter have been selected
from Army equipments which ncw carry an
unclassified designation. The objectives of
presenting these designs is to illustrate the
analytical procedures discussed in the earlier
chapters and to show how various components may be integrated to meet a set of
over-all requirements. It should be borne in
ordinates as it moves through space. Another purpose of the servos is to provide for
rapid slewing of the tracking antenna to the
angular position of the associated acquisition-radar antenna. Still another purpose is
to provide alignment (in elevation) of the
optical system with the tracking-radar antenna. The two servos also supply angularposition data to a computer and to remote
indicators.
17-2.3 OPERATION
An 1800-rpm
drive motor attached to an off-axis antenna feed (see Fig.
17-1) spins the beam in space and, in so
doing, places 30-cps amplitude modulation on
the chain of echo signals received from a target. After amplification and detection by the
radar receiver, the 30-cps amplitude-modulated video echo pulses are applied to a pulse
demodulator where the modulation signal is
recovered from the video pulses. The 30-cps
signal is then applied to an elevation-angle
17-1
ANTENNAFEED
DRIVE MOTOR
REFLECTOR AXIS
DRIVE MOTOR
AND PHASE-SENSITIVE
DEMODULATOR
VIDEO-PULSE
DEMODULATOR
400-CPS W P L l F l
(1-WATT OUTPU
- - --
ELECTRICAL CONNECTIONS
MECHANICAL CONNECTIONS
Fig. 17-1 Simplified functional block diagram of servo system for controlling
M33 tracking-radar antenna in elevation.
A-C
TRANWITTER
REPRESENTATIVE DESIGNS
phase-sensitive detector (see Par. 12-4) together with a 30-cps signal from a referencesignal generator coupled to the conical-scan
drive motor a t the antenna. The reference
signal serves a s a phase reference f o r the
elevation angle of the antenna-beam axis
measured with respect to the parabolic-reflector axis. If the antenna is on target, no erro r
signal is generated in the phase-sensitive
detector. If the antenna is off target in elevation, there is generated a n erro r signal whose
magnitude is a . measu re of the amount the
antenna-beam axis is off target and whose
polarity with respect to the elevation reference signal depends upon whether the beam
axis is above or below target. Xence, a s the
antenna tracks a moving target, the output of
the phase-sensitive detector is a slowly varying positive or negative voltage, the magnitude and polarity of which a r e a measure of
the elevation tracking error.
z actuating
= radar noise
on,
z angle
= tachometer-generator feedback
signal
of servomotor shaft
na reflector
G , ( s ) = t r a n s f e r function of r a d a r set,
elevation phase-sensitive detector,
400-cps modulator, and low-power
400-cps amplifier
COUPLING LOOP
-T. t
I '
u'K
The rate feedback supplied by the a-c tachometer generator [ H I(s)] is imperfect because the antenna structure is not absolutely
rigid. The rate feedback therefore tends to
improve the performance of the servomotor
but is not completely effective in reducing the
mechanical resonance of the antenna structure. The dynamic behavior of the servomotor and the antenna structure cannot be
separated in the system, as indicated by the
mechanical coupling loop. However, if .the
resonant frequency of the antenna structure
is very high, only the inertia of the structure
is important; the rate feedback is then a
direct measure of the elevation angular velocity of the antenna reflector.
120 w
Maximum velocity
500 mils/sec
Maximum acceleration
1000 mils/sec2
0.04 mil
360 va
360 va
Resonant frequency
1 cps
MP
1.5 dg
17-2.5 NOISE
The presence of radar noise in the line-ofsight signal prevents the actuating signal e
17-4
REPRESENTATIVE DESIGNS
COMPUTER
VELOCITY
SIGNAL
FILTER
AND
LIMITER
ACCELERATION SIGNAL
TRANSMITTER
. ---I
DIODE
STROKE ANTIJITTER
FEEDBACK SIGNAL
RECTIFIER
AND FILTER
A-C SUPPLY
MAIN-LOOP
SUMMING
NETWORK
1:l
COARSE
POSITIONAL
DATA FROM
COMPUTER
r-----SUMMING
AMPLIFIER
POWER
AMPLIFIER
c
c
_J
4 I
FILTER
L--.- _ _ _ _ _ _ - - - - - - - - - - - - - - - - -
I
I
16:1
FINE
POSITIONAL
DATA FROM
COMPUTER
ERROR-RATE-SERVO
SUMMING NETWORK
I
I
TO LOAD
(OUTPUT
GEARING
AN0 GUN)
HYDRAULIC
PUMP
_ _ - _---A-C
A-CTACHOMETERGENERATOR
ERROR-RATE-SERVO
VELOCITY FEEDBACK
SIGNAL
FILTER
- - ---
- --
ELECTRICAL CONNECTIONS
MECHANICALCONNECTIONS
I
4
REPRESENTATIVE DESIGNS
errors greater than 25 mils and the fine error signal controls circuit operation for errors less than 25 mils.
The summing amplifier in Fig. 17-3 is similar to that described in Par. 13-1 ; the power
amplifier consists of a phase inverter and a
push-pull amplifier. The output of the latter
stage drives the control winding of a 2-phase
stroke-control servomotor, the direction and
speed of motor rotation being dependent
upon the amplitude and phase of the amplifier output signal. The servomotor controls
the position of a follow valve in a hydraulic
servo cylinder (see Par. 14-4) and thus regulates the volume and direction of fluid delivery to the fixed-displacement hydraulic
motor geared to the gun. Controlling fluid
delivery controls the speed and direction of
the hydraulic motor and hence the speed and
direction of gun travel.
The coarse and fine positional-signal-control channel described above contains several
internal feedback loops. In addition, an auxiliary error-rate servo with its own error and
feedback voltages, mixing circuit, amplifier,
and servomotor is included in the over-all
power control system. To reduce overshoot
and to eliminate hunting of the stroke-control servomotor, a stroke velocity signal is
fed back to the summing amplifier through a
phase-correcting network. The feedback voltage is obtained from an a-c eddy-current generator attached directly to the shaft of the
stroke-control servomotor. The amplitude of
the generator output voltage is proportional
to the speed of shaft rotation of the servomotor, and the phase of the voltage is dependent upon the direction of rotation.
A second feedback path from stroke-control servomotor to summing amplifier serves
to suppress small oscillations of the servomotor output shaft that would otherwise be
transmitted to the gun. The stroke antijitter
voltage is obtained from a modified synchro
control transmitter that is friction-coupled to
the stroke-control servomotor shaft. The
transmitter rotor is excited from a potentiometer concected to a 6.3-volt a-c source.
The potentiometer is adjusted to provide the
minimum voltage consistent with the eli., nation of gun jitter. The output of the tranemitter is taken from two series-connected stator windings, and thus the magnitude of the
output signal is a function of the displacement of the rotor. The rotor is limited by
means of stops t o 230" displacement. The
output voltage of the transmitter is then rectified, filtered, and differentiated in a rate circuit. All steady voltages are therefore
blocked, but transient signals are passed on
to the diode modulator circuit.
The diode modulator circuit supplies the
summing amplifier with square-wave signals
that are in phase with the coarse and fine
error signals and that vary in amplitude in
proportion to the sum of the stroke antijitter
signal and an acceleration signal. The latter
signal provides a means of anticipating sudden changes in speed or direction of the gun
and is obtained from a velocity signal in the
computer. The velocity signal is supplied as a
d-c voltage from a permanent-magnet generator. After filtering to remove commutator
ripple, the velocity signal is differentiated in
order to obtain a voltage that is proportional
to acceleration. The acceleration signal is
then combined with the stroke antijitter signal as indicated above.
The error-rate servo in Fig. 17-3 receives
its input signals from the 1 6 : l speed transmitter in the computer. These signals are amplified to drive the rotor of an a-c tachometer generator, the stator of which is driven
by the gun gearing. The error voltage thus
generated is applied, after being phase-synchronized and mixed with the other feedback
signals, to the main summing amplifier. As
long as the gun turns in synchronism with
the fine-speed data from the computer, there
is no relative motion between rotor and stator of the a-c tachometer generator and no
signal voltage is generated. Should the gun
begin to lead or lag the point of synchronization, the phase and magnitude of the errorrate signal will accelerate or decelerate the
gun and thus bring the gun velocity into
agreement with the velocity of the computer
synchro-transmitter rotor.
17-7
dr
w1
w2
b5
bl
G9(s)
S
MAIN CONTROL
II
ERROR-RATE
STABILIZATION SIGNAL
r - - - d
b4
I
I
ERROR-RATE SERVO
L - - - - - - - - - - i
I
I
I
_1
17-8
REPRESENTATIVE DESIGNS
Control power
requirements
200 w
Resonant frequency
l 1 *
ERROR-RATE
STAB1
LIZATION
SIGNAL
1
'+I
"1
ERROR-RATE
SERVO
5 hp
Maximum velocity
1050 mils/sec
Maximum acceleration
2000 mils/sec2
0.75 mil
12 kw
+ 1, I + (12'12
ANTICIPATORY
ACTION
STABILIZED MAIN
CONTROL LOOP
17-9
REPRESENTATIVE DESIGNS
BIBLIOGRAPHY
Fire Control System, AA, T33, Contract W30-069-ORD-4490, Vol. I I , Wiring Diagrams (prepared by Western
Electric Company), Dept. of the Army,
FCDD-261, May 1, 1950.
2 0. Wilsker, Cant Corrector Servomechanism for Range Finder, T42, ORDTX10 Technical Memorandum, Report No.
M-52-8-1, F i re Control Instrument
Group, F r a n k f o r d A rse n a l , P h i l a delphia, Pa., July 1, 1952.
3 M. D. Swartz, Fire Control f o r Tanks;
Instrument Servos for Tank Applications, ORDTX-10 Technical Notes,
Report No. TN-1011, Fire Control
Instrument Group, Frankford Arsenal,
Philadelphia, Pa., January 16, 1953.
4 H. A. Smith, Tabulation of Data f o r Servo
S y s t e m s in E x i s t i n g A n t i n i r c r af t
Material, ORDTX-10 Technical Notes,
Report No. TN-1051, Fire Control
Instrument Group, Frankford Arsenal,
Philadelphia, Pa., October 15, 1954.
5 Second Antiaircraft Fire Control Working Conference (notes compiled by F.
Q. Barnett and A. M. Paup), Report
No. 932, Ballistic Research Laboratories, Aberdeen Proving Ground, Md.,
May, 1955 (secret).
9 Notes on Development Type Materiel Fire Control System, AA, T33, Contract W30-069-ORD-4490, Vol. I I I ,
Schematic Diagrams (prepared by
Western Electric Company), Dept. of
the Army, FCDD-261, May 1, 1950.
10 Notes on Materiel - Fire Control System
T38 for 75-MM Gun T83El and 75MM A A Gun Mount T69, Vol. 1, Sky-
T38 f o r 75-MM Gun T83E1 and 75MM AA Gun Mount T69, Vol. I I I , Computer T27E2, etc. (prepared by Sperry
Gyroscope Company), Dept. of the
Army, FCDD-263, Sperry Publication
No. 14-8093, February 1, 1951.
6 Notes on Materiel-75
MM A A Gun
Mount T I 9 Skysweeper, Vol. I I , Computer T27, Director T41 (prepared by
Sperry Gyroscope Company), Dept. of
the Army, FCDD-248, Sperry Publication No. 14-8092, May 25, 1948
(secret).
7 Notes on Development Type Materiel Fire Control System, AA, T33, Contract W30-069-ORD-4490, Vvl. I V ,
Maintenance Instructions (prepared by
17-11
Weapon (Stinger), Vol. 111, Schematics, Servo Loop Diagrams, and Wiring
Lists (prepared by Sperry Gyroscope
Company), Dept. of the Army, FCDD291, Sperry Publication No. 14-8094,
November 1, 1953.
Range A A
14 Notes on Materiel-Short
Weapon (Stinger), Vol. 11, Maintenance Instructions (prepared by Sperry
Gyroscope Company), Dept. of the
Army, FCDD-291, Sperry Publication
No. 14-8094, November 1, 1953 (confidential) .
15 Notes on Materiel-Short
Range A A
17-12
CHAPTER 18
Any type of pump that supplies the necessary pressure and rate of flow is usually
adequate for auxiliary service in a servo system. The type generally used in practice is
the fixed-displacement pump driven at a constant speed. The fluid pressure is controlled
by means of automatic pressure regulators or
unloading valves and accumulators. Occasionally, variabledelivery piston-type pumps are
used to maintain a constant pressure at all
flow rates between zero and their rated maximum. When used for this purpose, however,
the pump is considered as merely one component of a pressure-control system. Variable-delivery pumps a r e discussed in Par.
13-6.The fixed-displacement pump types discussed in the following paragraphs a r e :
(a) Gear pumps
(b) Vane pumps
(c) Piston pumps
18-1.3 Gear Pumps
A gear pump comprises two meshed gears
enclosed by a close-fitting housing equipped
with an inlet (suction) port and a discharge
port (Fig. 18-1). One gear is shaft-driven
by an external motor. Oil is carried from the
inlet port to the discharge port in the spaces
between the gear teeth, and is forced from
these spaces where the gears mesh.
Usually, spur or helical gears are used in
gear pumps. Gear materials range from cast
*Eg
J.O. Silvey
18-1
p u m p m a n u f a c t u r e r s recommendations
should be followed when choosing a hydraulic fluid. If a particular hydraulic fluid has
been specified, a pump designed to operate
efficiently with that fluid should be used.
1-
DISCHARGE PORT
iron through hardened alloy steel. The material used depends upon the service expected
from the pump.
Capacities of gear pumps normally used
a s replenishing pumps range from 1 to 60
gpm (gallons per minute). Rated output
pressures range from 15 psi for replenishing
service to 2000 psi for intermittent heavy
duty.
Oil leakage from the high-pressure (discharge) side of the gears to the low-pressure
(inlet) side is held to a minimum by the close
fit of the housing along the sides of the gears
as well a s at the tips of the teeth. Because
clearances cannot be held lower in a small
pump than in a large one, internal leakage
is larger fraction of the output flow in a
small pump. For this reason, small-size gear
pumbs are less efficient than large pumps
and are rated at lower pressures. The maximum efficiency obtained from a large gear
pump is about 80 percent. Small-pump efficiencies are below this figure. Leakage is also
largely dependent upon operating pressure
and the viscosity of the hydraulic fluid. The
18-2
DISCHARGE PORT
It
It
INLET (SUCTION) PORT
18-1.6 MAINTENANCE
suction line because the vacuum of the suction line may entrain pump-case air in
the oil.
18-1.8 COST
18-4
REGULATOR
L
THROTTLING
REGULATOR
tI
I
I
I
I
SUMP
Fig.
p2
r-+-J
ACCUMULATOR
CDNSTANT.
PRESSURE
18-5
LOAD
SPRING B
PORT D
BALL
PORT
F
P= A
(18-1)
where
F = spring preload, in lb
As the flow increases, the pressure a t port C
(Fig. 18-5) rises somewhat, because the
spring must be compressed to accommodate
the higher flow and because the high velocity
of the oil between the ball and its seat reduces
the pressure applied to a small area of the
ball surface.
For prevention of damage, it is desirable
to limit the pressure in the two main lines of a
hydraulic system and to permit excess oil
18-6
PORT A
TWO-DIAMETER
PISTON
p1
PORT 0
SPRING
DRAIN
CHANNELIN PISTON
DAMPING
CHAMBER
'
TO SUMP
TWO-DIAMETER
PISTON
SPRING-
DRAIN
18-7
niw REGULATED
and make it available for use when the system demand suddenly increases, thereby
decreasing the average power demand of
the system.
18-2.7 Gravity Accumulators
LOWREGULATED
/-AIR
VENT
rCYLlNDER
,-WEIGHT
18-2.6 ACCUMULATORS()
18-8
P!A,
PA,
(18-2)
where
SINGLE-PIECE
PISTON
WUTOFF
VALVE
r CyL"oER
HYDRAULIC
. PRESSURE
PRESSURE
METALLIC
- CHAMBER
FLEXIME
BAG
STOP
'
HYDRAULIC
PRESSURE
6. BAGTYPE
18-9
Assuming that the gas expansion is isothermal, the energy available when the accumulator is discharged can be expressed by
W = PoVo ln
(18-3)
where
Low-pressure hydropneumatic accumulators have been made in which the gas and
fluid are in direct contact (no bag or piston
is used). This type of accumulator is not
generally recommended because the gas dissolves in the fluid at high pressure and comes
out of solution when the pressure is removed,
producing gas bubbles that may cause serious
difficulties in the hydraulic system.
18-2.9 UNLOADING VALVES
18-10
PORT P
(FROM PUMP)
BYPASS VALVE
M
PORT R
(RETURN
TO SUMP)
SPRING B
L P I S T O N CHAMBER L
18-11
Dynamic seals must be used when a rotating or translating shaft must pass through
the wall of a chamber in which fluid must
be confined. To obtain a reasonable life
expectancy from the seal, it must be lubricated by the confined fluid. For this reason,
successful seals always leak enough fluid to
keep the shaft moist at the sealing surface,
but this leakage should be extremely small.
18-3.2 Glands
-PACKING
PACKING-GLAND
TREW
O.RINCc\
.O-RING
O-RING
B. LIMITED.TRAVEL
VANE MOTOR SEALED
WITH O-RINGS
18-13
- 1 1
/
STATIC SEALS
18-14
18-4.1 PURPOSE
18-4.2 CHARACTERISTICS
18-15
Limit stops restrict the travel of a servomechanism by replacing the normal error
signal with a signal that is a function of
the system output. The replacement signal
can reverse the direction of output motion,
cut off output power, or cause the output to
stop until a signal to move in the opposite
direction is received. The simplest form of
limit stop merely disconnects power from
the control circuit. Limit stops of this type
are seldom used because of the difficulty of
starting the drive in the opposite direction,
away from the limit.
18-16
upon the magnitude of the saturating-amplifier output voltage. This system operates with
little difficulty if d-c voltages are being
summed. However, some difficulty can be
expected if a-c voltages are used, because
of the phasing of the two signals and because
the waveforms of the mixed signals differ
(the output of the saturating error amplifier
is clipped, while the output of the potentiometer network is essentially sinusoidal) .
A mechanical limit-stop system, a form
of position servo, is shown in Fig. 18-19.
Error motion a t a reasonably high force level
is transmitted through both the springloaded force limiter and the box around the
spring-centered limit-stop arm to the control
valve. The valve controls the direction of
rotation as well as the starting and stopping
of the hydraulic motor, thereby controlling
the position of the cam that actuates the
limit-stop arm. When the cam is well clear
of the limits, error motion is transmitted
without interference to the control valve,
which opens one valve port and causes
rotation of the motor. If the cam eventually
turns the limit-stop arm to one side, the boss
on the arm shifts its enclosing box, forcing
the control-valve piston back to neutral. This
stops the hydraulic motor. The control valve
POTENTIOMETER
NETWORK
SATURATING
ERROR
AMPLIFIER
OUTPUT
AMPLI fl ER
18-17
LIMIT.STOP
MECHANISM
18-18
VENT AN0
FILLING SPOUT
-. ...-
LORIFIC
LOUTPUT
SHAFT
POSITIVE
STOP
h'I.
@I-
WASHER DETAIL
18-19
BIBLIOGRAPHY
18-20
CHAPTER 19
CONSTRUCTIONAL TECHNIQUES"
19-1 BASIC CONSIDERATIONS
In the
the design
design ofof any
any servomechanism
servomechanism or
or
In
control system,
system, aa great
great many
manyfactors
factorsbesides
besides
control
the response characteristics must be considered* Many systems that have adequate
response characteristics are completely unsatisfactory because too many adjustments
are required, the adjustments must be made
too often,
often, maintenance
maintenance isis difficult,
difficult, or
or the
the
too
equipment cannot
cannot withstand
withstand environmental
environmental
equipment
conditions. In
In ordnance
ordnance equipments,
equipments, itit is
is parparconditions.
ticularly important
important that
that much
much attention
attention be
be
ticularly
paid to these de,ils, not only because the
equipment must be made reliable and simple
to operate under conditions normally imPosed upon equipment and operator, but also
to minimize
minimize the
the time
time required
required to
to train
train
to
operators
operators and
and maintenance
maintenance organizations.
organizations.
19-1
Many of the devices used in servomechanisms develop undesired heat which must be
dissipated. Where space is not a limitation
and air is free to circulate around the device,
heat dissipation to the atmosphere may be
accomplished with a temperature rise of only
a few degrees above the ambient temperature. However, if the volume and hence the
area of the devices is small, the resulting
temperature rise may be sufficiently high to
damage the equipment. The high ambient
temperatures that can be encountered by
ordnance devices should be considered when
designing control systems.
The sources of heat in various control systems differ so greatly from one another that
it is necessary to evaluate each source separately. Some of the most common heat sources,
as well as the representative rates at which
they liberate heat, are listed in Table 19-1.
If the temperature at which a control system
of a particular design is to operate appears
to be excessive, the heating rates of the
actual components should be obtained under
service conditions.
The heat generated in a control system
must be adequately dissipated to prevent
excessive temperature rises. In general, it is
preferable that auxiliary cooling devices such
as fans and heat exchangers should not be
used. However, in some instances, auxiliary
cooling devices may be required ( f o r
example, to avoid excessive size of the
control system).
The mounting of electronic components so
that air can flow freely past the heat sources,
particularly in a n upward direction, is a wellestablished method of cooling amplifiers.
Attaching a heat source (for example, a control motor) to a large plate or structural
member of high heat conductivity helps to
decrease the temperature rise of the heat
source. In hydraulic systems, the temperature
of electrical equipment can often be lowered
by submerging the electrical device in the
hydraulic fluid. Because the fluid is in motion
and has much greater thermal capacity than
air, the temperature rise of the, submerged
electrical device will be much less than it
would be in air. Normally, hydraulic fluid
can be kept adequately cool by using a sufficiently large sump which, in turn, is air
cooled.
In most ordnance control equipment, the
generated heat is transmitted to the surroundings by radiation or by convection of
the air. Unfortunately, the laws of convection
and radiation are so complicated that it is
not usually practical to establish a single
simple relation that will quickly give the
heat dissipated by a hot body.
19-2
CONSTRUCTIONAL TECHNIQUES
Rate of Heat
Liberation
Device
Vacuum-tube voltage-amplifier triode
4 watts
3 watts
Remarks
25 watts
Constant-speed 60-cps
induction motor
no load
9% output rating
full load
Armature-controlled
d-c motor
field only
6% power rating
no load
full load
no load
stalled
rated
voltage
Auxiliary pump
Hydraulic motor
15% of input,
minimum
Hydraulic amplifier
Variable-delivery pump
15% of input,
minimum
19-3
where
19-2.3 Radiation
Q2
(19-1)
where
Q1
in OR
(Temperature in degrees Rankine
= 460
temperature in degrees
Fahrenheit.)
rounding body, in
in OK
(Temperature in degrees Kelvin =
273
temperature in degrees
Centigrade.)
OR
0.78
Polished iron
0.24
Oil paint
0.94
(19-2)
Surface (100F)
Emissivity
Polished aluminum
0.04
Polished copper
0.03 (approx.)
White paint
0.90 to 0.95
Black paint
0.96
Aluminum paint
0.3 to 0.6*
CONSTRUCTIONAL TECHNIQUES
H , = A,ec5I4
(19-3)
Surface
Height
(in.)
where
h
log ( A , .- 50) = 2.22 - 20
h = height of surface, in cm
(19-4)
HF [HP in Btu/(hr)(ft%OF)]
OF
20F
40F
60F
80F
100F
120F
140F
180F
100
0.60
0.75
0.87
0.93
1.01
1.06
1.12
1.23
69
0.65
0.75
0.88
0.96
1.03
1.10
1.15
1.24
35
0.65
0.8
0.87
0.92
0.98
1.02
1.07
1.13
23
0.7
0.8
0.88
0.95
1.0
1.05
1.10
2.17
11.3
0.75
0.9
1.00
1.06
1.12
1.17
1.23
1.32
5.9
1.10
1.3
1.38
1.44
1.51
1.55
1.60
1.66
3.2
1.65
1.8
1.95
2.03
2.12
2.2
2.27
2.38
1.8
2.0
2.4
2.55
2.70
2.86
2.98
3.10
3.28
19-5
Material
Aluminum
Brass (70-30)
Copper
119
60
218
Lead
20
Silver
238
Steel AISI-1020
26
Stainless steel
AISI-304
9.4
0.09
19-2.5 Conduction
AK8
(19-5)
QT=
where
Q T
O F )
of material, in
L = thickness of material, in f t
19-6
CONSTRUCTIONAL TECHNIQUES
Multiply
number
of
490
1760
2050
3.424
12.2
14.2
2.04
x 10-3
0.292
3.58
4.16
Watts/(cm2) (CO)
5.68
x 10-4
8.19
x 10-2
0.279
1.16
4.23
Kg-calories/(hr) (cm2) ( C o )
4.88
x 10-4
7.04
x 10-2
0.240
0.859
3.60
1.355
x 10-4
1.95
x 10-2
6.67
x 10-2
0.236
0.277
Btu/(hr) ( i n 2 )(F")
6.94
x 10-3
Watts/ ( i n 2 ) ( F " )
144
19-7
51.3
15.0
2$+l
(08,
y=x
S?
-+2L--+1
where
LOAD
w,, L!
(19-6)
0,)
in f t
METAL
PLATE
1
BASE
if$
c
5=
= damping ratio
A.
fq+y
B.
g = 32.2 ft/sec2
acceleration of gravity
19-8
CONSTRUCTIONAL TECHNIQUES
10.0
9.0
a. 0
7.0
6.0
5 .o
4.0
3 .O
2 .o
I .o
0.9
0.8
0.1
0.6
4-
0.5
0.4
d
W
2
3
O3
0.2
0.1
0.09
0.08
0 .Of
0.06
0.05
0.04
0.03
0.02
0.01
FREQUENCY RATIO Z
urn
'21 = w / k
(19-7)
0.902
radians/sec = cps
VY
(19-8)
where y is in feet, or
3.13
w''
77cps
where y is in inches.
The analysis given here is simplified to
include only motion along a line perpendicular to the mounting surface, assuming loads
for which the center of mass is equidistant
from the isolation supports. In practice, it
is necessary to consider rotary vibration of
the load, particularly if vibrations can occur
parallel to the mounting surface or if rotary
forcing vibrations can occur about any axis.
Vibration isolation can be made unnecessary in some cases by eliminating the vibration a t its source. Various friction, viscous,
and tuned damping devices have been used
for this purpose. Static and dynamic balancing are also used to eliminate vibration
caused by rotating devices.
intensity in the majority of ordnance servomechanisms. Table 19-6 lists typical vibration
and shock values encountered in ordnance
equipment.
The physical construction of mounts for
shock isolation is identical with that of
mounts for vibration isolation. In general,
the underlying basis for both types of mounts
is the same in that they serve to lower the
acceleration and hence the force applied to the
device to be protected. The resonant frequency of loads supported by resilient shock
mounts is often set at approximately 20 cps
because a resonant frequency of 20 cps
provides adequate protection against most
19-10
CONSTRUCTIONAL TECHNIQUES
Equipment
Frequency
(CPS)
Source
springs
2-5
tires
8-15
structures
Maximum Acceleration
in gst
5-7
60-200
3 times speed
of tank (in
mPh 1
Tank
muzzle blast 6 f t
from equipment
impulse
1000-1500 (exterior)
300-1250 (interior)
moderate nonpenetrating
ballistic impact
impulse
400
heavy nonpenetrating
ballistic impact
impulse
vehicle
equipment
$Maximum damage due to ballistic impact appears to be caused by accelerations of approximately 3000 g.
19-11
PROTECTED DEVICE
MOUNTING
RESILIENCE
AND DAMPING
/I
.1
CHASSIS
t.
Fig. 19-3 Analytical model of shockmounted device.
BIBLIOGRAPHY
CONSTRUCTIONAL TECHNIQUES
6 C. E. Crede, Vibration and Shock Zsolation, John Wiley & Sons, Inc., New
York, N.Y., 1951.
7 W. W. Soroka, Shock Isolation, Product
Engineering, Vol. 26, p. #167, June,
1955.
8 H. S. Ryder and E. K. Gatcombe, Design-
19-13
CHAPTER 20
20-1.1
where
DEFINITION
(20-1)
20-1.5
PRODUCT OF INERTIA
The product of inertia of a body with respect to two mutually perpendicular planes is
defined as
Z = I, + M r 2
Gab
= .f ab dm
(20-3)
(20-2)
*By J . 0.Silvey
20-1
(20-4)
OF MASS M O M E N T OF INERTIA
Multiply
number
of __c
To
obtain
By
gram-cm2
oz-in.2
gram-cm2
1
182.9
15.467 x 10-3
lb-in.2
3.418 x 10-4
0.0625
x
1.349 x
4.340
7.375 x 10-8
2926
oz-in.2
slug-f t 2
lb-in.2
16
lb-ft2
I 2304
2.159 X lo-'
1 4632
I 32.17
1
OF INERTIA,
To
1 dyne-cm
1 gram-cm2
1 gram-cm
1 gram-cm2
980.7
1 oz-in.
1 oz-in.2
386
1 lb-in.
1 lb-in.2
1 lb-ft
1 lb-ft2
1 lb-ft
1 slug-ft2
32.17
20-2
7.412 x 104
0.03108
A torque of
1 144
I 6.944 x 10-3 I 1
10-5
1.356 x lo7
4.214 x 10-5
1 1
10-4
slug-f t2
rad/se@
1000
100
ETWEEN 0.1 W D
asu BY 1 0 1 AND
10
0.1
ALUMINUM
BRASS
STEEL
- 0.0959 I b h 3
0.313 l b / h ~ . ~
0.183 Ib/In.
I
I I I I I
0.01
0.00I
I
1 I I IIIIIIIJ
3
4
CYLINDER DIAMETER (INCHES)
9 1 0
Fig. 20-1 Moments of inertia about the principal axes of cylinders one inch long.
M3
The moment of inertia of a body with respect to any line passing through the coordinate origin is
J = J, COS" a + J, C O S ~$ + J , C O S ~y
- 2G,z,cos a
- 2G,,
cos
cos $ cos y
where
G,,, G,,, G,, = products of inertia with respect to the yz- and xz-planes, the xz- and xyplanes, and the xy- and yz-planes, respectively.
If the coordinate system is so chosen that
the x-, y-, and z-axes lie along the principal
axes of the body, the products of inertia are
zero and Eq. (20-5) reduces to
J = J,%C O S ~a + J ,
C O S ~i3
+ J , cos2 y
(20-6)
=-clav
20-2.4 K i n e m a t i c Viscosity
(20-7)
I?
From Eq. (20-7),the viscosity is p = fZ/av
and has the dimensions ML-lT-' or FL-'T.
Viscosity defined in this way is called absolute,
or dynamic, viscosity. The only unit of absolute viscosity that has a name is the poise,
which is one dyne-second per square centimeter, or one gram mass per centimeter-second. Table 20-3 lists the relations between
various units of absolute viscosity.
f
Kinematic viscosity is defined as the absolute viscosity of a fluid divided by its density.
When a dimension system of force, distance,
and time is used, the dimensions of kinematic
viscosity are L'T-l. The centimeter-gram-second (cgs) unit is the stoke, which has the
dimensions centimeter squared per second.
The English unit, which would be used with
the pound force-foot-second system commonly
employed in engineering work, is the foot
squared per second. The relations between the
three commonly used units of kinematic viscosity are :
100 = centistokes
(b) (ft'/sec) X 929 = stokes
(a) stokes
F:
8s
$ .5
va,
h h
0 6
"I
To
obtain
8 $
C h
I a ,
o a
poise*
(dyne-seconds per
980.7
4309
6.89
x 104
479
3.15
gram f orce-seconds
per square centimeter
1.020
x 10-3
28.35
ounce force-seconds
per square inch
2.33
x 104
0.03527
16
0.111.
pound f orce-seconds
per square inch
1.451
x 10-6
2.202
X 1o-s
pound force-seconds
per square foot
2.09
x 1o-s
144
0.317
454
1000
Id
10
20
10'
5
4
3
2
100
5
4
H
Y
e
*
i
5
zf
lo
30 40 50
1000
100
SAYBOLT SECONDS U(IVERSAL
20.6
40
06-
10 20 3 0 4 0 5 0 6 0 70 80 90 100
120
140
160
180
200 2 2 0 240 260 280 300 320 340 360 380 4 0 0 425
Fig. 20-3 Viscosity versus temperature for some oils and fluids.
450
The viscosity of all lubricating oils, hydraulic fluids, and damping fluids increases
when their temperature is lowered. However,
most hydraulic fluids contain additives that
reduce the change in viscosity below that ex-
hibited by pure petroleum-base oils. The relation between temperature and viscosity for
some oils and fluids is shown in Fig. 20-3.
20-2.6 FRICTION
Material
Condition
Coefficient
of
Kinetic
Friction
Coefficient
of
Static
Friction
dry
0.152
0.162
greased
0.08-0.10
0.16
dry
0.213
greased
0.132
dry
greased
0.44
0.08-0.10
0.11
0.18
0.19
dry
greased
Steel on steel
0.09
Steel on steel
0.03
Steel on bronze
dry
0.152
20-8
0.07-0.08
0.15
However, the coefficientof kinetic friction decreases a t extremely high velocities and increases a t extremely low velocities. Since the
coefficient of friction, either static or kinetic,
is dependent upon the condition of the surfaces and the presence or absence of small
amounts of lubricants, any quoted value
should be considered as an approximation^
The coefficients of friction for various metals
are listed in Table 20-4.
20-3 SPRINGS
20-3.1 HELICAL SPRINGS
20-3.3 Deflection
20-3.2 Stress
Neglecting end effects and curvature COFrection and assuming round wires, the stress
in loose-wound tension helical springs or compression helical springs is
=-8WD
nd3
(20-8)
y =
8W D k
E,dq
(20-10)
where
W = load, in Ib
s=
8WD
-
(20-9)
nd?
+-
20-9
OF
Elastic
Limit*
(Psi)
Recommended
Working
Stress$
(Psi)
Modulus
Rigidity
(Psi)
130,000
70,000
11 x 106
92,000
60,000
11 x 106
95,000
35,000
11 >( 106
67,000
30,000
6.2 x lo6
Beryllium copper
80,000
50,000
6.5 x
Material
Music wire
lo6
*An average value of elastic limit is given here. Actual values vary widely.
$The recommended working stress given here i s approximately 80y0 of the minimum elastic limit of the
material.
20-10
203.7 Definition
A cantilever spring (Fig. 20-4) is made by
clamping one end of a bar and applying a load
a t the other end. In effect, the spring is essentially a cantilever beam and ordinary beam
theory can be used to obtain stress and deflection.
20-3.8 Stress and Deflection
s=
MC
I
(20-11)
where
(20-12)
y=---4WP
Ebh3
2SP
3Eh
(20-14)
A cantilever beam bends instead of twisting. Therefore, the modulus of elasticity and
the allowable stresses are not the same as for
a helical spring. Table 20-7 lists the tensile
elastic properties of some spring materials.
20-3.10 TORSION BAR SPRINGS
where
20-3.11 Definition
Y=
WP
-
3EI
(20-13)
S=-
E,r6
1
(20-15)
where
20-11
5/32
3/16
1/4
5/16
3/R
7/16
1/2
5/R
3/4
7/8
-1/4
1-1/2
0.010
0.305
9.47
0.012
0.522
20.7
0.421
10.0
0.350
5.43
0.014
0.823
40.3
0.669
19.4
0.560
10.5
0.422
4.18
0.016
1.21
72.9
0.983
34.4
0.824
18.4
0.626
7.37
0.018
1.71
124
1.040
57.4
1.17
30.7
0.889
12.1
0.020
2.32
200
1.89
91.3
1.60
48.6
1.22
18.9
1.972
9.06
0.024
3.90
464
3.22
208
2.72
108
2.09
41.4
1.68
19.9
1.41
11.1
6.67
0.029
6.61
1160
5.55
496
4.71
253
3.64
94.1
2.94
44.6
2.60
24.6
2.13
14.9
1.87
9.75
0.033
7.97
916
6.85
460
6.29
166
4.31
17.5
3.61
42.4
3.11
25.6
2.74
16.1
2.21
8.26
0.037
10.9
1580
9.42
785
7.41
280
6.01
128
6.07
69.6
4.41
42.2
3.88
21.3
3.10
13.3
2.61
7.43
14.4
1618
11.4
559
9.33
249
7.91
134
6.82
80.0
6.01
51.6
4.89
25.0
4.10
14.0
3.62
8.60
0.051
18.5
1233
15.3
543
13.0
288
11.3
168
10.0
107
3.01
51.1
6.78
28.6
5.79
11.3
5.14
11.4
0.059
27.6
2509
23.2
1069
19.9
554
17.3
322
15.3
204
12.4
96.3
10.4
62.7
9.01
32.2
7.89
21.1
6.33
10.3
38.7
2563
33.4
1295
29.4
141
26.1
463
21.3
216
18.0
117
16.5
70.6
13.6
45.3
11.1
22.4
9.21
12.6
65.6
3083
49.2
1708
44.1
1063
36.1
478
30.5
256
26.6
153
23.3
97.5
18.8
47.4
15.8
26.6
0.043
0.071
0.085
1.21
1-3/4
2-1/2
3-1/2
4-11?
Wire
Diameter
(in.)
Outside Diameter
of Coil ( i n . )
1/8
6/32
3/16
1/4
6/16
3/8
7/16
1/2
6/8
3/4
7/8
69.3
4630
61.0
2490
66.1
1616
46.6
683
38.4
363
33.4
216
91.0
4973
81.9
2967
68.3
1298
68.1
679
130
6710
109
!El7
1-1/4
1-1/2
1-3/4
29.6
136
23.8
66.4
19.9
36.8
17.1
22.6
60.8
400
44.8
266
36.2
12 1
30.1
66.6
26.1
40.9
22.8
26.6
93.8
1436
81.9
832
72.6
626
68.8
246
49.4
136
42.06
82.0
37.6
63.3
30.2
26.2
0.148
161
3166
133
1802
119
1119
96.3
614
81.2
279
70.6
168
62.0
109
49.7
63.1
41.7
29.7
0.177
260
7486
222
4160
198
2626
163
1146
137
609
119
361
106
234
84.7
113
71.0
63.2
61.2
38.4
341
8867
307
6266
266
2320
211
1218
188
716
166
467
134
218
113
121
484
11806
406
4988
349
2662
306
1488
269
941
218
443
614
,0199
634
6164
467
2919
416
1820
0.331
824
10786
733
6048
0.376
1166
19966
4-1/2
97.2
14.1
86.2
48.6
76.0
33.6
183
243
168
148
139
96.9
125
66.1
338
849
286
460
241
276
216
180
194
123
662
3713
634
1693
463
911
391
643
343
348
308
238
1038
10903
932
6662
767
2960
662
1676
667
497
697
446
933
1687
23202
1433
13869
1201
6006
1019
3132
888
1836
783
1166
704
786
2073
26646
1742
11268
1600
6736
1310
3330
1160
2097
1041
1406
0.6626
2426
19863
2096
9967
1832
6686
1628
3546
1468
2369
0.626
3230
33289
2826
16386
2491
9240
2221
6710
1997
3773
0.092
0.106
0.126
0.207
0.244
0.283
0.4376
0.600
2-1/2
3-1/2
406
Wire
Hameter
(in.)
The torque is
EJ6
T = -1
1tE,r'6
21
(20-16)
20-3.14 Natural F r e q u e n c y
E*Q
f = - a!
2xr2n
(20-17)
)/3-~,
where
d = spring wire diameter, in in.
20-3.1 3 V I B R A T I O N IN SPRINGS
spring, in cps
= 0.285 lb/in.3
OF S O M E S P R I N G MATERIALS
Elastic
Limit*
(Psi)
Recommended
Working
Stress$
(psi)
200,000
100,000
Music wire
200,000
110,000
120,000
80,000
150,000
40,000
80,000
40,000
130,000
75,000
Material
Beryllium copper
Modulus of
Elasticity
(Psi)
x 106
30 x lo6
32
x lo6
28 x 106
15 x 106
17 x lo6
28
*An average value of elastic limit is given here. Actual values vary widely.
$The recommended working stress given here is approximately 80% of the minimum elastic limit of the
material
20-14
20-4.1 BULK M O D U L U S
B Y - AP
/ AV \
\vl
(20-19)
where
B = bulkmodulus
AP = change in pressure
AV = change in volume
20-15
Material
Yield
Stress
(Psi)
Ultimate
Strength
(Psi)
Tensile
Modulus of
Elasticity
(Psi)
Soft-drawn copper*
7000
33,000
15 X lo6
49,000
55,000
15 x lo6
3000
9000
10 x 106
30,000
45,000
40,000
65,000
40,000
90,000
Hard-drawn copper*
Soft aluminum*
10 x 106
x
29 x
29
106
lo6
*These materials have no well-defined yield point. Values are for 0.2% offset.
BIBLIOGRAPHY
Helical Compression a n d E x t e n s i o n
Spring Design, Trans. ASME, Vol.
69, p. #725, 1947.
7 E. F . S m i t h , H e l i c a l C o m p r e s s i o n
Springs, Machine Design, April 19,
1956.
8 K. W. Maier, Dynamic Loading of
INDEX
Index Terms
Links
A
Alternating-current motors
14-25
14-37
14-31
figure of merit
14-37
14-31
14-25
6-15
13-1
electronic amplifiers
13-1
hydraulic amplifiers
13-105
magnetic amplifiers
13-58
mechanical amplifiers
13-181
pneumatic amplifiers
13-162
relay amplifiers
13-87
13-73
transistor amplifiers
13-38
3-39
11-79
coding discs
11-81
numerical representation
11-79
voltage-to-digital coders
11-84
3-29
Approximate procedures
5-20
5-21
5-21
5-20
Index Terms
Links
18-1
auxiliary pumps
18-1
hydraulic auxiliaries
18-4
18-15
rotary joints
18-12
Auxiliary pumps
18-1
cost
18-4
18-3
maintenance
18-3
purpose
18-1
18-1
B
Bearings
15-29
ball bearings
15-29
friction
15-32
lubrication
15-32
miscellaneous bearings
15-34
roller bearings
15-29
sleeve bearings
15-33
3-17
block diagrams
3-17
signal-flow graphs
3-25
C
Closed-loop response determination
5-10
5-13
nonunity-feedback systems
5-15
polar-plane technique
5-11
Compensation networks
6-16
a-c electric
6-17
d-c electric
6-16
hydraulic amplifier
6-20
mechanical damper
6-19
pneumatic controller
6-21
Index Terms
Links
Compensation techniques
6-1
6-15
6-10
6-6
6-2
6-16
20-15
bulk modulus
20-15
compressibility
20-15
20-15
specific gravity
20-15
Constructional techniques
19-1
basic considerations
19-1
component layout
19-1
shock isolation
vibration isolation
Convolution integral
19-10
19-8
3-11
D
Damping and friction
20-4
friction
20-8
viscosity
20-4
10-1
Design techniques
9-9
Differential equations
3-1
Digital-to-analog conversion
Direct-current motors
12-16
14-1
14-23
14-20
14-11
14-17
14-7
14-1
Dynamic response
2-1
forced response
2-8
frequency response
2-7
Index Terms
Links
2-1
stochastic inputs
2-8
transient response
2-5
Electronic amplifiers
13-1
E
cascading amplifier stages
13-15
13-35
feedback amplifiers
13-23
13-7
power amplifiers
13-13
13-26
thyratron amplifiers
13-28
vacuum tubes
13-1
F
Factoring and characteristic parameters of
response modes
3-2
3-4
factoring
3-2
6-10
Forced response
2-8
Frequency response
2-7
G
Gain determination
5-1
approximate procedures
5-20
5-10
5-4
5-1
Index Terms
Links
5-3
root-locus method
5-23
5-15
5-15
5-18
polar-plane construction
5-16
Gear trains
15-1
definitions
15-1
design fundamentals
15-5
gear types
15-3
15-12
purpose
15-1
Gyroscopes
11-60
application factors
11-71
11-66
11-61
11-64
11-74
11-73
11-69
11-68
11-68
11-66
typical applications
11-66
H
Hydraulic amplifiers
13-105
13-149
13-150
hydraulic-circuit elements
13-155
illustrative example
13-158
13-137
Index Terms
Links
13-150
13-136
13-110
Hydraulic auxiliaries
18-4
accumulators
18-8
check valves
18-4
18-4
pressure-regulating valves
18-6
pressure-relief valves
18-6
unloading valves
Hydraulic motors
18-10
14-37
14-43
14-39
14-43
14-44
14-37
14-39
L
Laplace and Fourier transforms
3-12
frequency response
3-17
3-13
theorems
3-12
18-15
characteristics
18-15
limit stops
18-16
positive stops
18-18
purpose
18-15
Linearization
2-1
This page has been reformatted by Knovel to provide easier navigation
Index Terms
Links
11-47
design characteristics
11-48
15-17
basic purpose
15-17
examples
15-19
nonideal characteristics
15-19
practical applications
15-17
M
Magnetic amplifiers
13-58
analytical representation of
13-64
basic considerations
13-58
construction of
13-70
performance of
13-66
principles of operation
13-59
13-72
typical circuits
13-62
Magnetic-particle clutches
14-47
advantages
14-47
description
14-47
disadvantages
14-47
dynamic behavior
14-51
life expectancy
14-51
methods of use
14-47
static behavior
14-50
20-1
13-181
basic types
13-181
dynamic behavior of
13-187
13-191
13-195
static characteristics of
13-184
15-1
15-29
15-1
Index Terms
Links
15-17
15-22
mechanical differentials
15-16
15-20
15-22
couplings
15-22
15-25
Mechanical differentials
15-16
differential linkages
15-17
geared differentials
15-16
purpose
15-16
3-1
3-29
3-17
convolution integral
3-11
differential equations
3-1
3-34
3-2
3-12
3-36
3-39
Modulators
12-2
chopper
12-2
electronic
12-7
magnetic
12-4
N
Nonlinear systems
describing function procedures
10-1
10-1
10-13
phase-plane procedures
10-1
Index Terms
Links
O
Objectives of feedback control system
1-1
1-2
9-6
9-8
9-6
8-1
8-1
8-11
8-2
8-8
8-2
8-6
transient inputs
8-2
8-8
P
Performance criteria
8-1
5-1
acceleration constant
5-1
bandwidth
5-2
gain
5-1
peak magnitude
5-3
static accuracy
5-2
torque constant
5-2
velocity constant
5-1
Index Terms
Links
Performance evaluation
7-1
error coefficients
7-44
performance indices
7-45
7-1
7-45
6-6
lag compensation
6-7
lead compensation
6-7
Phase-plane procedures
Pneumatic amplifiers
10-8
13-162
13-180
dynamic behavior of
13-173
pneumatic valves
13-163
13-163
Pneumatic motors
14-44
14-46
dynamic behavior
14-45
principal types
14-44
static behavior
14-45
Polar-plane representation
5-3
5-3
5-4
Potentiometers
11-1
application factors
11-11
11-1
linear potentiometers
11-4
nonlinear potentiometers
11-7
14-1
14-25
14-1
hydraulic motors
14-37
magnetic-particle clutches
14-47
penumatic motors
14-44
Index Terms
Links
1-1
1-1
1-2
1-2
1-3
R
Relations between frequency response and
transient response
7-1
7-1
7-17
7-21
13-87
13-87
13-100
parameter measurement
13-101
polarized relays
problems encountered with
13-93
13-102
relay characteristics
13-88
13-90
13-88
13-98
17-1
17-5
17-1
6-2
lag compensation
6-2
lead compensation
6-4
Index Terms
Links
3-36
Root-locus methods
5-23
5-24
5-23
13-73
characteristics of
13-77
13-80
13-83
13-85
13-73
13-86
18-12
dynamic seals
18-12
Rotary transformers
11-15
general classifications
11-15
general description
11-15
induction potentiometers
11-41
induction resolvers
11-37
microsyns
11-44
synchros
11-16
11-41
Routh criterion
4-2
S
Sample-data systems
9-1
design techniques
9-9
general theory
9-1
operational methods
9-6
performance evaluation
z transform and the w transform
9-10
9-4
Index Terms
Links
Sensing elements
11-1
analog-to-digital converters
11-79
gyroscopes
11-60
11-47
11-86
potentiometers
11-1
rotary transformers
11-15
tachometer generators
11-51
15-20
compliance
15-21
purpose
15-20
sheave sizes
15-21
stress
15-20
tension
15-21
Shock isolation
19-10
Signal converters
12-1
digital-to-analog conversion
12-16
electronic demodulators
12-11
modulators
12-2
types
12-1
Springs
20-9
cantilever springs
helical springs
20-11
20-9
20-11
vibration in springs
20-14
1-2
4-1
Nyquist criterion
4-4
root-locus method
4-7
Routh criterion
4-2
Stochastic inputs
2-8
20-1
20-15
Index Terms
Links
20-4
20-1
springs
20-9
T
Tachometer generators
11-51
11-58
11-51
1-3
Transient response
2-5
Transistor amplifiers
13-38
13-55
13-43
basic principles
13-38
13-40
13-50
16-1
16-13
16-4
16-14
gathering specifications
16-2
illustrative example
16-14
16-13
16-14
V
Vibration isolation
19-8
W
w transform
9-5
Z
z transform
9-4
This page has been reformatted by Knovel to provide easier navigation