Download Design and construction of an electronic gain

Scholars' Mine
Masters Theses
Student Research & Creative Works
1949
Design and construction of an electronic gainphase meter
Gabriel G. Skitek
Follow this and additional works at: http://scholarsmine.mst.edu/masters_theses
Part of the Electrical and Computer Engineering Commons
Department:
Recommended Citation
Skitek, Gabriel G., "Design and construction of an electronic gain-phase meter" (1949). Masters Theses. 4878.
http://scholarsmine.mst.edu/masters_theses/4878
This Thesis - Open Access is brought to you for free and open access by Scholars' Mine. It has been accepted for inclusion in Masters Theses by an
authorized administrator of Scholars' Mine. This work is protected by U. S. Copyright Law. Unauthorized use including reproduction for redistribution
requires the permission of the copyright holder. For more information, please contact [email protected].
DESIGN AND CONSTRUCTION OF AN ELECTRONIC GAIN-PHASE METER
BY
GABRIEL GEORGE SKITEK
A
THESIS
submitted to the faculty of the
SCHOOL OF MINES AND 'Mb"'TAIJJURGY OF THE UUIVERSITY OF MISSOURI
in partial fulfillment of the work required for the
Degree of
MASTER OF SCIENCE IN ELECTRICAL ENGINEERING
Rolla, Missouri
1949
Approved by
~~~.-~~/_""J
Professor of Electrical Engineering
_
ii
TABLE OF CONTENTS
Page
List of Illustrations •
...
·..
·.
Introduction. •
·.....·......
• • • •• · . . . . . .
Review of Literature.
Description of Gain-Phase Meter • • · . . . . . .
List of Tables.
ii
iii
1
3
6
Circuit Analysis
a. Circular Sweep Generator ••
b. Pulse Generator • • • • • •
c. Cathode-ray Tube Indicator.
10
.....
·....
16
21
Design Considerations
a. Pulse Generator
·.. ...·......
b. Circular Sweep Generator. •
c. Cathode-ray Tube Indicator.
·..
·......
...·....
.. . .. ..· ....
·. . .........
Summary and Conclusions •
Bibliography•
...
Vita• • • • •
• • • • • • • • •
25
25
26
Calibration of Gain-Phase Meter •
Discussion of Sources of Errors •
23
-.
• •
• • • • • • •
36
42
43
44
iii
LIST OF ILLUSTRATIONS
Figure
1
2a
2b
3
4a
4b
Sa
Page
Block Diagram of Gain-Phase Meter.
·.
·..
· .. . . . .... . • • •
Gain Scale. . . . . · . . . . . . . . . · . . .
Circular Sweep Generator Circuit. . . . · . . . · . . . .
Phase Angle Scale •
·.·....
Vector Plot of Voltages in Phase Shifter. •
..
Phase Angle Measurement •
. . . . . .. ... . .
Gain Iieasurement. • • • •
..... ..
Bridge Phase Shifting Network • • •
7
8
8
11
II
1.4
15
15
7a
·.
Sinusoidal Voltage Input to Pulse Generator • . . . . . .
7b
Voltage at the output of the First Clipper Stage. • • • •
20
7c
Voltage at the Output of the Third Clipper stage. •
20
6
7d
7e
8
9
10
12
Pulse Generator Circuit.
11
20
. . ...... .... .....
output Pulse. • • • · . . . . . . . .
.....
Variable Phase Voltage Generator•• • •
·......
Block Diagram of Phase Angle Test Circuit. • • · . . . .
28
Block Diagram of Cireuit Used to Measure Gain and. Phase
Angle of an Audio Frequency Amplifier • • • • • • • • ••
38
Plot of Gain and Phase Angle Versus Frequency for a
Transformer Coupled Audio Amplifier • • • • • • • •
39
Differentiated Wave • •
Polar Plot of Gain and Phase Angle of a Transformer
Coupled Audio Amplifier • • • • • • • • • • • • • •
...
20
20
21
40
iv
LIST OF TABLES
Table
I
II
Page
Accuracy Check of hase Angle Reading
About the Circular Sweep • • • • • • •
30
Minimum Width of Blanked-out Portion of
32
Circular Sweep • • • • • • • • • • •
III
IV
v
Phase Angle Changes Due to Deviation of
Pulse Generator Input Voltage From 0.5
Volt • • • • • •
Gain Calibration • • • • • •
... ..
Gain and Phase Angle Data of a 6c5
Transformer Coupled Audio Frequency
AInplifi er. . . . . . . . . . . . . .
..
33
35
1
INTRODUCTION
There are InallY' communication and electronic circuits, such as £11ters, phase-shifting networks, matching networks, voltage amplifiers
with and without feedback, whose proper usage with associated equipment
depends upon the knowledge of the variation with frequency of the voltage ratio and phase angle between the output and input sinusoidal voltages or currents.
A connnon laboratory technique for obtaining the phase angle between two voltages is to appq the voltages to a cathode-ray oscilioscope and from the dimensions of the elliptical pattern formed,
calculate the phase angle. (1)
This technique gives poor accuracy,
(1) F. E. Terman, Measurements in Radio Engineering, page 324,
McGraw-Hill Book Company, 1935.
because unequal phase shift may be introduced through the vertical and
horizontal voltage amplifiers in the oscilloscope.
The ratio of the
output to input voltage is commonly obtained from the reading of a
voltmeter, such as a vacuum tube voltmeter.
Neither of the above
measuring teclmiques are well suited for class roam demonstration or
student laboratory work covering phase angle and gain measurements,
because a good mental picture of phase angle and gain cannot be
established through elliptical patterns and voltmeter readings.
It is the purpose of this thesis to design and build an electronic instrument that 'Vtill measure the voltage gain and phase angle
between two sinusoidal voltages or currents
of collllIIWlication
cir-
cuits, in such a manner as to aid the student in visualizing how these
2
quantities vary with frequency, and to increase the accuracy of the
phase angle measurements.
This instrwnent, the gain-phase meter, will
be designed to give a phase angle accuracy of ±1.0 degree or better,
and gain accuracy of
:!:5% or better, for input voltages from .5 to 62.5
volts over a frequency range from 40 to 10,000 cycles per second.
Though several high accuracy electronic phase meters have been
constructed, the author of this thesis feels that none are of the type
that fulfill the above objectives.(2)(3)
(2) Edwin F. Florman and A. Tait, An Electronic Phasemeter, Proceedings of the IRE, Volume 37, Page 207, 1949.
(3) Alan Viatton, Modulated-Beam Cathode-Ray Phasemeter, Proceedings of the IRE, Volume 32, Page 268, 1944.
3
REVIE'I[ OF LITERATURE
Search into literature does not reveal any electronic instrument
such as the proposed gain-phase meter for measuring voltage gain and
phase angle over a wide range of audio frequencies.
Although several
electronic instrwnents have been designed and built to measure phaseangle, none have combined gain and phase angle measurements, and none
have indicated such measurements in a manner that would be suitable for
classroom demonstration and student laboratory work.
In
1942 an electronic phase angle meter, 'whose phase angle read- .
ings were indicated by a vacuum tube voltmeter, was designed and built
by Edward L. Ginzton. (4)
This instrument, the author claims, can meas-
(4) Edward L. Ginzton, An Electronic Phase-Angle Meter, Electronics, Volume 15, Page 60, 1942.
ure phase angles from 0 to 180° directly.
Vlth the use of a trans-
former, whose phase shift is 180°, the range can be extended to 3600 •
The author claims that the test voltages, applied to the instrument,
must be betvreen
0.5
and
5.0
volts for proper operation.
The aceurac,y
of angle measurement and frequency range were not discussed in this
article.
In this instrument, the two voltages whose phase angle is to be
measured, are applied to two identical amplifier-clipper channels that
convert the sinusoidal voltages to square wave voltages of sarne phase
displacement.
The two square wave voltages are fed into a voltage
adding circuit, whose voltage output is indicated by a vacuum tube
voltmeter.
Theoretically, it can be shown that the sum of the two
square wave voltages is a linear function of the phase angle betw-een
4
the two sinusoidal voltages under test.
In 1944 a phase angle meter, whose angle readings were indicated
by a cathode-ray tube, was built by JUan Watton. (5)
(5) Watton,
2£.
This instrument,
Cit.
as claimed by the author, will read angles from 0 to 360 0 with an
accuracy of ~. ~ •
TIle two voltages, whose phase angle is to be meas-
ured, are applied separately to a circuit that converts the sinusoidal
input voltage to a square wave voltage.
This square wave voltage is
then allorred to modulate the electron beam of a cathode-ray tube whose
deflection plates are excited by quadrature voltages.
The modulation
of the grid blanks out one-half of the circular pattern on the screen
of the tube and leaves a semi-circular trace.
Another semi-circular
trace is produced by the second voltage under test.
The phase angle is
obtained from the angular displacement of the two semi-circular
patterns on the screen of the cathode-ray tube.
In 1948 E.
(6) E.
o.
o.
Vandaven designed a polyphase meter(6) for use in
Vandaven, Phase Meter, Electronics, Volume 21, Page 142,
1948.
measuring the phase angles of voltages applied to a type 2H21 phasitron
tube.
Phase angle indications are produced on a cathode-ray tube
screen, ..lhose deflecting plates are excited by quadrature voltages, and
its grid is modulated by pulses derived from the volta es under test.
In 1949 Florman and Tait(7) designed apha.se-meter having a voltage
(7) Florman and Tait, Q£. Cit.
5
range of 1 to 30 volts over a frequency range of 100 to 5,000 cycles
per second with an accuracy of ~o.~. This instrument operates on a
principle similar to the principle of the instrument constructed by
Ginzton in 1942.
meter.
The phase angle is indicated by a vacuum tube volt-
6
DESCRIPTION OF GAIN-PHASE METER
The gain-phase meter discussed in this thesis incorporates the use
of a cathode-ray tube for indicating the gain and phase angle of two
sinusoidal voltages or currents.
A
cathode-r~
tube was chosen for the
intlicating device, because it lends itself well to classroom demonstration, and leaves a mental picture with the student of gain and phase
angle variations with frequency as found in communication and electronic circuits.
A complete block diagram of the gain-phase meter is shown
in Fig. 1.
The phase angle between two voltages,
~
and E2, is indicated by
the angular displacement of two blanked-out port,ions of a circle on
the cathode-ray tube as shown in Fig. 2a.
The circular sweep is ob-
tained by applying to the cathode-ray tube equal quadrature deflecting
voltages that are derived through a phase shifting network from one of
the sinusoidal input voltages as shovm in Fig. 1.
The blanked-out
portions of the circle are produced by applying to the control grid of
the cathode-ray tube negative pulses of very short time duration and
steep front.
These negative pulses are generated by electronic cir-
cuits that amplify, clip, and differentiate the two sinusoidal. voltages El and E2 as shovm in Fig. 1.
The time separation of the pulses,
derived from El and E2, depends upon the phase angle between the two
sinusoidal voltages El and E2.
Angles from 0 0 to 360°, lead or lag, can be read directly' by this
type of indicator without ambiguity.
Lead or lag of angle,s i,s deduced
from the direction of travel of the spot on the screen in its circular
path.
_.~-----.Ltr
: :
I
I
I
.... 1 1•
ITTII amp.
~6~
-
(1)
6SJ7
'
(
I
1_ _-
6SS7
6SJ?
-\f\---
I
1:'
~
--'1!7--
<) ,
. L
..
605
(3)
Parar:-·hase
ar.::n.
Dumont
241
1.""",;~-JiI'---~
Oscillof,r8Dh
Brict.~e
Ph~8C Snlltter
-
<
-- '-\)- ----~ ----- ~- ------CQ--- I
I
--:\}
I
I
I
I
I
t
I
I
I
I
I
I
I
I
-
E1
:
I
(,
--lb- -----lb------%----9:J---y.U
I
'
I
,
I
I
•
I
I
t
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
,
I
I
•
I
'
,
6H6
1st
anD.
To Z Axis
-
Circular Sween
2nd
2nd
Cl i!YDer
Pulse Generator
ar::rp.
3rd
amp.
3rd
C11puer
t-~
6HB
I--
f Voltage
~elector
BLOCK DIAGRAl-1 OF GAIN-PHASE HETER
Figure 1.
.....
8
'\
\
"-
-
"
-
PBASE ANGLE SCALE
FIGUHE 2a.
GAIN SCALE
FIGURE 2b.
~-~
Circle
. - - - E1
~lrcle
9
The voltage ratio or gain between the sinusoidal voltages, E2 and
El is indicated by the radius of the circular pattern on the screen of
the cathode-ray tube as shol'm in Fig. 2b.
Voltage gain may be read
directly on a radial scale attached to the screen of the ca.thode ray
tube.
Three gain ranges are obtained through the use of the attenuator
at the input of the circular sweep generator.
ments are not taken simultaneously.
Gilln and phase measure-
10
CIRCUIT ANALYSIS
The gain-phase meter can be considered as having three main elements; circular S1veep generator, pulse generator, and cathode-ray tube
indicator.
Circular SweeR Generator
The circular sweep is generated by an electronic phase shifting
circuit shown in Fie. 3.
To the input of this circuit is applied
either of the two sinusoidal voltages, El or E2, depending upon which
is to be taken as the reference.
This voltage is amplified to the de-
sired level by the 6SJ7 resistance coupled amplifier(8) of conventional
(8) F. E. Terman, Radio Engineers Handbook, Pages 354-366,
McGraw-Hill Book Co., 1943.
design and fed into a 6c5 paraphase amplifier. (9)
Two equal voltages
(9) L. B. Arguimbau, Vacuwn Tube Circuits, Page
Book Co., 1948.
355,
JaM Wiley
that are 180 degrees out of phase are obtained from this tube stage,
one voltage from cathode to ground and the other from a tap on the
plate resistor to ground.
A phase splitting arrangement of this type
was chosen because it supplies equal voltages whose phase angle does
not deviate appreciably from 1800 over a frequency range from 40 to
10,000 cycles per second.
The balanced output of the paraphase ampli-
fier is applied to a bridge type of
hase shifting network(lO) as
(10) staff Membe s of .I.T. adar School, Principles of Radar,
Chapter 3, Page 31, McGraw-Hill Book Company, 1946.
C 20
6 SJ7
+300 yolt.
Yolto
Q)
Re
6S J 7
:.:..::
~
E,
----0
f
EZ
~
6SJ7
Fro.
1
R I4
o.tI. ~ ttIMI
e'
®
PI.t
D__••
141
4
...1••
........ or
O. c
'"
-=-
-:-
-:-
-=-
-::-
6SJ7
@
C
CIRCULAR
SWEEP
GENE'RATOR
Fiqure
3
CIRCUIT
zT
iI." ....
12
PARTS LIST FDR CIRCULAR S'NEBP GENERATOR
Resistors
R
= 0.1
Meg. Ohnl potentiometer - 1 watt
R2
= 0.2
ohm potentiometer - 1 watt
M:eg. ohms - 1/2 watt
= hOK ohms
R4 =10K ohms
R3
R,
=1.0 Meg.
R14' Rl" R16,
Rl , R13 ,
- 1/2 watt
- 1/2 watt
= 20K ohms - 1/2 watt
R6, R21' R22' R3l, R32
= 0.1 Meg.
ohms - 1/2 watt
R7, ~7' RIB , R27 , R28 = 600 ohms 1/2 watt
= 0.1~7 Meg. ohms - 1/2 watt
R34 = 0., Meg. ohms - 1/2 watt
RS, R19' R20, 11.29' R30
~, R 23' R 24, R33'
R10, R12
Rll
!=
= 2.5K
R2.5, R26
6.50 ohms 1/2 watt
ohms 1/2 watt
= 1.0 Meg.
ohm potentiometer - 1 watt
Condensers
Cl' C.5, C6, C7' C20, C21' C26, C27
= 0.1 ~f.
(paper) condenser
C2, C3' C4, C16, C17' C18, C19' C22' C23' C24' C25
(electrolytic)
C8' ~5
• .00015 ~f. (mica)
C9 , C14· .00025 ~f. (mica)
C10, C13 = .002 !-Lf. (mica)
Cll, C12 = .01!-Lf. (mica)
=8.0~f.
13
shown in Fig.
3.
A nore detailed illustration of this bridge phase shifting network
and a vector plot of the voltage relationships is shown in Figs. 4a and
l.~b.
From this figm e it can be seen that four equal sinusoidal volt-
ages, that are 90 degrees displaced from each other, are produced if Xc
is made equal to R.
These four voltages EoA' EoB' ~OD' and EoE are in-
jected into conventional resistance coupled amplifiers using type
vacuum tubes as shoyVIl in Fig. 3.
6SJ7
The voltage appearing between the
plate of tube No. 1 fed by EoA' and the plate of tube No.2, fed by
t.oB, is 90 degrees out of phase with the voltage between the plates of
tubes 3 and
4
that are fed by EOD and EOE respectively.
This phase
relationship between the voltages appearing from plate to plate for the
two pairs of amplifier stages, can be obtained in the same manner as
shown in Fig. 4b for the voltages from grid to grid, EnE and EBA., of
the same pair of tubes.
The quadrature output voltage is then applied
to the two pairs of deflecting plates to produce a circular sweep on
the screen of the tube as shown in Fig. Sa.
The amount of signal fed to each of the four voltage amplifiers
can be controlled separately in order to correct for unequal amplification of each stage, and anY phase shift from 180 degrees that may
develop in the paraphase amplifier stage.
Four values of Capacitance
(C) found in the bridge phase shifter were required to cover the range
from 40 to 10,000 cycles per second in conjunction with two variable
resistors (R) of 100,000 ohms each.
Gain measurements are also made through the use of the circular
sweep.
To measure the gain or ratio of E2 to
Er,
first be calibrated by the smallest voltage .c.l.
the instrwnent must
This voltage is
11
A
A Tube
,...------------,!L----------------O
No.1
r -_ _ ~Tube
No.2
E: Tube
'-----t""I
No.3
D TUbe
r---<l
No.4
BRIDGE PHASE SHIFTING NETWORK
FIGURE 4a.
DE ~
-OD
" .....
,
\
\
\
\
\
,
\81) or
B
I
...
B
T-------~
\
A
I
\
\
.
\
\
I
,,
.._ r:
/
/
/
.....
/
....
",
('I
-
_~/...
--
VECTO . PL.)T OF VOLTAG7S IN
FI'} R_: 4b.
P~.A
E SHIFTER
15
e
n
ure
e 5
n
16
selected at the input of the pulse generator, amplified, and fed to the
circular sweep generator by sw:i tch position Eo at its input.
The mag-
nitude of this voltage is adjusted by Rl' and/or Rk' to give a circular
pattern of 1/2 inch in diameter on the cathode-ray tube. when the range
selector at the input to the sweep generator is set to position
1-5.
With the instrument calibrated, the larger voltage (E 2 ) is applied at
the input of the pulse generator and the gain can be read directly from
a radial gain scale.
This scale is l~ inches long and is divided into
25 equal divisions of 1/20 of an inch in width.
set to 1-5, a maximum gain of 5 can be read.
With the range switch
The gain scale can be in-
creased to 25 on the 1-2.5 scale setting and to 125 on the 1-125 scale
setting.
meter.
Gains from -12.5 to +12.5 can be measured 'with the gain-phase
Fig.
5b shows the circles for gain measurements produced by
E:t
and E2 simultaneously through double exposure.
Pulse Generator
The blanked-out portions of the circular sweep used for phase
angle indication are produced by the pulse generator shown in Figs. 1
and 6.
Since only one pulse generator is used, one blanked-out por-
tion of the circular sweep appears for each voltage selection made by
the switch at the input of the pulse generator.
one pulse generator instead of two, one for
E:t
The reason for using
and the other for E2,
will be discussed under "Design Considerations".
The voltage at the input of the pulse generator is passed
through the 6SN7 cathode follower stage ,vhose output voltage appears
across the low cathode impedance Rk and ground.
The amount of voltage
fed to the first 6AC7 amplifier stage of the pulse generator must be
.. 300 volta
t300 nih
R
To
R
7
Rig
I3
ot
Ci 3
I
6H6
6SN7
2
E,
2
Rg
~C,
RI !!
6H6
RZI
6H6
DulllOll' 241
fiH6
+j
0----<>
O.cillo~w. . .
J
Co
E2
To Z Aai. of
~,
Cg
A
CII5
A
R
27
-:::::::-
300 yolta
PULSE
GENERATOR
Fiqure
6
CIRCUIT
A
VR-IOI
To Laboratory
R.I.lo t. d
Po .. ., Suppl Y
18
PARTS LIST FOR PULSE GENERATOR
Resistors:
R1
=1.0 Meg.
Rk
= 20K
R3
= 0.5 Meg. ohms - 1/2 watts
ohms - 1/2 watt
ohms - 10 watt potentiometer
R4' R10' R16 • 190 ohms - 1 watt
R5, R6, R11, R12' R17' R1 8
=10K ohms
- 2
R7, R13' R19
= 60K
RS, R1.4' R20
= 0.5 Meg. ohms - 1/2 watt
~,
R15' R21
=.22 Meg.
R23
=0.2 Meg.
R24
=6K ohms
o~~ -
y~tts
1 watt
ohms - 1 watt
ohms - 1 watt
- 10 watts
R26 = 200 ohms - 1 "watt
R27
=3.25 K ohms -
10 watts
Condensers:
Gl, C2
C3'
=0.1
G4' C5'
~.
C6' C7' C8' C11 , C1 5
=8.0
j.1f.
(electrolytic)
C , C12 , G , C c 1.0~. (paper)
13 14
9
GIO = 28 - 60 - 120 - 300 - 675 ~. (mica condensors)
19
adjusted to approximately
selector switch.
0.5
volt for both settings of the voltage
This adjustment of input voltages is made through the
use of the circular sweep generator with the cathode ray tube as an
indicator.
With potentiometer Rl of the circular sweep generator set
to tile pre-calibrated position that will give a 2" circular pattern for
an input voltage of
0.5
volt,
the potentiometer Rk in the cathode
follower circuit of the signal generator is adjusted to produce a 2"
circular pattern for the two voltages under test.
The reason for using
a cathode follower stage and making the input voltage to the first 6AC1
volt will be discussed under IIDesign Considerations ll •
The output voltage of the first 6AC7 amplifier is fed into a full
wave 6H6 diode clipper that clips off the top and bottom portions of
the sinusoidal input voltage as shO\m in Figs. 7a and ?b.
This clipped
wave is amplified by the second 6AC7 amplifier and clipped again by the
second full wave clipper.
This process of amplifying and clipping is
repeated in the third amplifier and third clipper stages respectively.
The output of the third clipper stage is a square wave that has a steep
rise and fall as shown in Fig. 7c.
This square wave voltage is applied
to an R-C differentiating circuit that produces narrow, steep-fronted
positive and negative pulses as shown in Fig. 1d.
This differentiated
voltage is applied to a voltage selector(ll) that clips off the nega(ll) B. Chance, V. Hughes, E. F. Ma.cNichol, D. Sayre, andF. C.
VTilliams, Waveforms, pp. 44-55, McGraw-Hill Book Company,
1949.
tive pulse and selects only the top portion of the positive pulse as
shown in Fig. 1e.
These positive pulses are applied to the g axis
21
amplifier of a type 241 Dumont Oscillograph (12) where the polarity of
(12) Manufactured by Allen B. Dumont Laboratories, Inc., Passaic,
N. J.
the pulse is reversed through amplification.
These negative pulses are
then applied to the control grid of a SJPI cathode-ray tube cutting off
the electron beam for the duration of the pUlse; thereby, producing a
blanked-out portion of the circular sweep.
Since the circular sweep is
of the same frequency as the repetition rate of the pulse, the blankedout portion will remain stationary on the circular sweep.
Due to un-
equal changes in phase shift in the circular sweep and pulse generator
with frequency, the zero reference will change positions on the circular sweep.
This requires the phase angle scale to be rotated to make
the zero of the angle scale correspond with the reference voltage
blanked-out portion of the circular sweep.
From the above discussion on the formation of the pulse, it can
be readily seen that the separation expressed in degrees between two
pulses, one evolved from E1 and the other from E2, will be equal to the
phase angle between the sinusoidal voltages E1 and E2•
Fig. Sa shows,
through double exposure, two blanked-out portions produced by voltages
Cathode-rq Tube Indicator
The SJP1 cathode-ray tube, g axis amplifier, and associated power
supply of the type 241 Dumont Oscillograph were used in conjunction
with the circular sweep and pulse generators in order to make gain and
phase angle measurements.
22
The output of the circular sweep generator is ted directly to the
deflecting plates of the 5JPl to produce an Wldistorted circular
pattern
3-1/4 inches in diameter. The
average sensitivity of the de-
flecting plates was found to be equal to approximately 32 volts d-c per
inch.
23
DESIGN CONSIDERATIONS
In the design of this gain-phase meter, special attention had to
be paid to circuits and measuring technique to reduce or prevent errors
in phase angle and gain measurements.
Pulse Generator
A pulse generator of the clipper-differentiator type, as found in
Fig. 6, was used to produce the narrow blanking pulse because of its
simplicity and ability to generate voltage pulses of desired width.
ide band amplifier tubes and circuits(13) were used in the square wave
(13) F. E. Ter,man, Radio Engineers Handbook, Pages 413-416,
McGraw-Hill Book Co., 1943.
and pulse amplifier stages of the pulse generator to produce steep
front square wave and pulse voltages.
A steep fronted pulse voltage is
required to produce a sharp cut-off of the electron beam.
Two pulse generators, one for voltage
E1
and the other for voltage
E2, were not used, because it was found upon building two identical
pulse generators, that there existed a maximum phase difference of
approximately three degrees between the generators.
enee varied vdth the frequency of the input voltages.
This phase differAlthough this
phase difference could have been corr:pensated for by the use of a phase
shifting network, a much simpler instrument is had with only one pulse
generator.
The voltage input to the first 6AC7 amplifier must be adjusted to
approximately
0.5
volt for the two input voltages El and E2.
This
adjustment of input voltage is required, because it has been found
24
(see Table III) that a phase shift, in the order of 0., degree, is induced in the pulse generator for an input voltage variation of ~,O%
from 0., volt.
This phase change is associated with changes in the
amo"Wlt of distortion produced in the first 6AC7 stage.
By
using the
technique for voltage adjustment as discussed "Wlder Circuit Analysis,
the error in phase angle readings due to unequal input voltages is reduced to practically zero.
A potentiometer located in the cathode
circuit of the preceding cathode
follol~r
is used to equalize the input
voltages to the first 6AC7 amplifier.
A cathode follower stage was used to prevent
~
large phase
change due to the adjusting device used for the voltage equalization.
Another good characteristic of the cathode follower is that it offers
a low sh"Wlt capacitance to the input signal. (13)
A ,OO,OOO-ohm
(13) Lawrence B. Arguimbau, Vacuum. Tube Circuits, page
354, John
Wiley Book Co., 1948.
potentiometer located at the input of a convElltional triode 6C, amplifier was found experinlentally to produce a maximum phase change of
approximately four degrees.
Three stages of amplification and clipping were found sufficient
to produce the desired square wave over a frequency range from
10,000 cycles per second.
40
to
The increase in rise and fall time of the
square wave at high frequencies was compensated for by the decrease of
the capacitance in the differentiating circuit.
A regulated laboratory power sup!)ly was used to supply d-c voltage
to the pulse and circular sweep generators.
De-coupling in the plate
leads of the 6AC7 tube stages in conjunction with a regulated power
25
supply prevents coupling between stages.
Circular Sweep Generator
A paraphase circuit using a plate loaded 6c5 tube with its cathode
resistor un-bypassed was used to obtain t,vo equal voltages that are 180
degrees out of phase.
This circuit was found to produce equal voltages
of opposite polarity over a larger range of frequencies than a two-tube
phase inverter.
A balanced output voltage is used in the circular sweep to produce
a large undistorted voltage for the production of an undistorted circular trace on the cathode-ray tube.
The 6SJ7 balanced resistance
coupled voltage amplifiers are designed to produce a large undistorted
output voltage.
To reduce the loading effect of the gain-phase meter during gain
and phase measurements, a cathode follower stage was used at the input
of the pulse generator and a pentode stage at the input of the circular
sweep.
Cathode-Ray Tube Indicator
The 5JPl and associated g axis amplifier of the type
241
Dumont
Oscillograph was used in conjunction with the pulse and circular sweep
generators.
A cathode-ray tube was used as the indicating device be-
cause it lends itself well to classroom demonstration and student
laboratory work.
A type 5JPl cathode-ray tube 'was used because its
screen size, 5 inches in diameter, is sufficient for small classroom
demonstrations and its deflection sensitivity is reasonably high.
26
CALIBRATION OF GAIN-PHASE METER
To check the accuracy and calibrate the gain-phase meter, several
tests were performed.
TIle accuracy of the phase angle reading at various points on the
circular sweep )'{as checked by applying to the gain-phase meter two
voltages that were separated by constant phase rr.i.th respect to each
other, but of variable phase with respect to the input voltage of the
circular sweep generator.
With these two voltages it was possible to
produce blanked-out portions of the circular sweep, of constant phase
separation, on any portion of the circular pattern on the cathode-ray
tube.
Any inaccuracy in phase angle reading was evidenced by the dif-
ference in measured phase displacement of the two blanked-out portions
of the circular sweep, as the angle between the two input voltages and
the circular sweep voltage was varied from 0 to approximately 360
degrees.
It should be noted that the phase angle between the two test
voltages is not required to be
main constant.
mown,
but it is important that it re-
The circuit used to produce the two test voltages of
constant phase separation is shown in Fig. 8.
A block diagram showing
its conneetion to the gain-phase meter is shown in Fig. 9.
The variable phase angle
be~7een
the
t~
test voltages of constant
phase separation and the circular sweep voltage was obtained through
the use of t,vo phase splitting bridges (14) as shown in Fig. 8.
The
(14) F. A. verest, Phase Shifting up to 360 degrees,
Electronics, Vol. Ih, pp. 46-49, 1941.
constant phase rolgle between the two test voltages was obtained from
605
o
C4
E
r---,'-UII--
0
¢l
2
~~--1~~06 •
HewlettPackp..rd
Audio
Osc.
2000
+300
.
T
I
II
I
I
0'5---0 E¢
C
.. . 5 Meg. ohm Pot .
= .1 Meg. ohms
• 2.0 Meg. ohms
::I:
5 K ohms
1\
R1
lI\2, ~'3
F· 4 , 1\5
R
'P
6'
°3'
C
4'
0 , 0,., ,
6
IT!
...
= .5 Meg. ohms
= 400 ~t.
a
.01 J,Lt.
,.,
0
0 1 , O2 ,
+300
.01
° = 8.0
J-Lf.
1J,t'.
b
:::I
= Gener81
ltad10 Company
TrHnsforoer b78-A
VARIABI.Z PHASE VOLTAGE GEr-lERATOl\
Figure 8
ro
-..:z
HewlettPackard
Auo.i0
Oscillator
-0~
Sweep
Generator
·200C
o-J
u--
.n
C1I;cular
Dumont
241
Oscillograph
.E~1
I....-
Variable
Phase
Voltage
. Generator
0
-
Pulse
--
Generator
~
Z Axis
4-
E¢
BLOCK
PHASE
E
1
E
Z
DIAGFUU~
~GLE
Z
OF
TEST CI1\CUIT
Figure 9
!:"
en
29
the R-C network.
Absolute phase angle calibration from any fixed reference point
was not used because the reference point of the gain- hase meter
changes with frequency.
The gain-phase meter was tested for angle accuracy at 100, 1000,
and 5000 cycles per second.
The results of the test, as shown in
Table I, show a maximum deviation in angle measurement of :to.5 degree
over the range covered.
The minimum width of the blanked-out portion of the circular
pattern varies with frequency as shown in Table II.
For wide blanked-
out portions of the circle, the angle measurements were made between
the ends of the blanked-out portions that correspond to the steep front
of the pulse applied to the grid of the cathode-ray tube.
The amount of phase angle shift in the pulse generator, due to the
input voltage of the first 6AC7 stage deviating from 0.5 volt, was
found at 1000 and 10,000 cycles per second and is tabulated in Table
III.
This data was obtained for t"l"0 voltages of zero phase separat' on
applied to the terminals of the pulse generator.
The variations in the
magnitude of the input voltage to the first 6AC7 tube were produced by
the potentiometer in the cathode circuit of the preceding cathode follower stage.
from
0.5
From this data it can be seen that a deviation of :t50%
volt will produce an error of approximately
±D., degree..
If
the method outlined under "Circuit Analysis ll is used for adjusting the
two input voltages, the error can be reduced practically to zero.
Gain measurement accuracy wa
obtained by checking the gain read-
ings on the cathode-ray tube scale against the gain calculated from the
)0
TABLE I
Accuracy Check of Phase Angle Readings About the Circular Sweep
"l(-f
= 100
cycles per sec.
*f
= 1000
cycles per sec.
53.5°
~
5.SO
15.0°
18°
53.0°
15.5°
15.0°
36°
53.0°
25.0°
61°
53.5°
35.5°
15.0°
15.0°
87.5°
53.5°
45.0°
15.0°
104.5°
53.5°
55.5°
15.0°
112.0°
53.5°
65.5°
15.0°
127.0°
53.0°
15.0°
144.5°
53.0°
65.5°
75.0°
151.5°
53.0°
85.5°
15. SO
164.0°
53.0°
94.0°
15.5°
107.5°
15.5°
119.0°
15.5°
130.0°
15.5°
140.0°
15.0°
150.0°
15.0°
160.0°
15.0°
170.0°
15.0°
178.SC
15.0°
187.0°
15.0°
198.0°
15.0°
209.0°
15.0°
~
-&
0°
...e-
15.0°
31
TABLE I (Cont.)
-:<f
= 5000
~
15.0°
cycles per sec.
-e17.5°
¢ = The
angle of displacement of
with zero reference, taken
at 3 o'clock on the cathoderay tube screen.
~1
35.5°
17.0°
47.5°
11.5°
62.0°
17.0°
77.0°
17.0°
91.0°
11.0°
108.0°
17.0°
123.0°
16.5°
J.42.00
16.5°
159.5°
16.5°
118.0°
16. SO
19h.Oo
16.5°
216.0°
11.0°
229.5°
11.0°
284.5°
17.0°
298.0°
17.0°
310.0°
17.0°
323.5°
11.0°
343.0°
17.0°
355.0°
17.5°
..e- = Angle between E1 and E2 as
read on the angle scale on
the front of the cathode-r~
tube.
* = Yith
0.5 volt input to first
6AC7 amplifier stage of the
pulse generator.
32
TABLE II
M:i.nimum Width of Blanked-out Portion of Circular Sweep
Frequency
Width (Inches)
100
1/100
1000
l/SO
5000
1/16
10,000
1/4
33
TABLE III
Phase Angle Changes Due To Deviation of Pulse Generator Input
Voltage from 0.5 volt.
f
(Voltage measured at the grid of first 6AC7
amplifier tUbe.)
= 1000
cycles per second
E2
El
~
~
.50
.5
83.0
83.0
.80
.5
83.0
82.5
1.2
.5
83.0
81.5
.3
.5
83.0
83.5
f
= 10,000 cis
E2
El
~
~.
.5
.5
102
102
.75
.5
102
101.5
1.35
.5
102
104.0
.3
.5
102
102.5
34
ratio of the input and output voltage as read by a vacuum tube voltmeter.
The gain scales were calibrated by applying
0.5
volt to the El
terminals of the pulse generator and adjusting, by the controls Rl and
Rb the diameter of the C"Lrcle to read 1 on the
1-5
gain scale.
The
maESIlitude of the voltage applied to the E2 terminals of the pulse gener tor 'i'fas varied from 0.5 to 62.5 volts to check the three gain
scales, 1-5, 1-25, and 1-125, on the screen of the tube.
of this test are tabulated in Table IV.
seen that
~1e
The results
From these results it can be
accuracy decreases with the larger readings on each
scale, with a maximum error of -4% at the highest gain reading.
From
the method used for gain measurement it can be seen that the gain
scale calibration obtained for El equal to 0.5 volt can be used for
any other value of El that is within the linear voltage response of
the circular sweep generator and the cathode follower stage of the
pulse generator.
20 volts.
This linearity was found to exist up to value of
This means that the gain scale can be calibrated with any
magnitude of voltage up to 20 volts input to the pulse generator.
For
El values greater than 20 volts, the error in gain measurement will
increase and the circular pattern will become distorted.
To prevent
this error and distortion, an external attenuator could be used, common to El and E2' so as to reduce El below 20 volts.
35
TABLE IV
Gain Calibration
F equency
a
1000 cycles per second
Measured
Calculated
Gain
%Deviation~~'
El*
E2*
0.5
0.5
1.0
1-5
1.0
0
0.5
1.0
2.0
1-5
2.0
0
0.5
1.5
3.0
1-5
3.0
0
0.5
2.0
3.9
1-5
4.0
2.5
0.5
2.5
4.8
1-5
5.0
4.0
0.5
2.5
5.0
1-25
5.0
0
0.5
5.0
10.0
1-25
10.0
0
0.5
7.5
15.0
1-25
15.0
0
0.5
10.0
20.0
1-25
20.0
0
0.5
12.5
24.0
1-25
25.0
4.0
0.5
12.5
25.0
1-125
25.0
0
0.5
25.0
50.0
1-125
50.0
0
0.5
37.5
75.0
1-125
75.0
0
0.5
50.0
98.0
1-125
100.0
2.0
0.5
62.5
120.0
1-125
125.0
4.0
Gain~H<
Scale
* = Voltages measured by Ballentine Model
** = The precision measure of the line
'Y'
of full scale reading.
(See Fig. 2b..)
300 vacuwn tube voltmeter
•
n scale is equal to :!:2%
36
DISCUSSION OF SOURCES OF ERRORS
The error in angle readings due to non-orthogonality of the
cathode-ray tube deflecting plates and non-quadrature deflecting voltages is reduced to zero i f a precisely circular pattern is formed on
the cathode-ray tube. (15)
This means that the non-orthogonal"ty of the
(15) Sollar, Stan, and Valley, Cathode-Ray Tube Displays, M.l.T.
RaQtation Series, page 299, McGraw-Hill Book CompaQY, 1948.
deflecting plates is compensated by an equal departure from quadrature
of the two voltages app1jed to the deflecting plates.
Errors due to ellipticity of the circular sweep are reduced to a
ver,y small value since small ellipticities are easily detected by the
eye when a reference circle is provided on the screen of the tube. (16)
(16) Sollar, Stan, and Valley, Ope Cit., page 301.
The greatest error associated with the circular sweep is due to
distortion of the circular pattern when harmonics are present in the
quadrature voltage output of the circular sweep generator.
This dis-
tortion causes the angular velocity of the spot on the cathode-ray tube
to depart from constancy; thereby, producing an error in the angle
measurements.
Analysis of this type of distortion is quite compli-
cated, for it not only is a function of the magnitude of distortion,
but also of the harmonic content of the quadrature voltages.
To reduce
this type of error the gain-phase meter should not be operated below or
above the frequency band that will produce a highly distorted circular
sweep.
A frequency range from 40 to 10,000 cycles per second was found
37
to produce a circular pattern with slight distortion.
The harmonic content of the quadrature voltage also produces an
error in the gain measurements, for it may distort the circle at the
location of the gain scale.
Error in phase angle measurement, due to distortion in the first
6AC7 amplifier stage of the pulse generator, are reduced to a minimum
by equalizing the input voltage to this stage as discussed in "Circuit
Analysis".
The poor accuracy of gain measurements is largely attributed to
the inability to interpret a reading closer than 1/40 of an inch by a
1/40 of an inch wide circular sweep trace.
of error are; non-linearity of
cathode-r~
Other contributing sources
tube deflection, increases
in capacitance shunting effect of the range attenuator with frequency,
and slight non-linearity of sweep generator amplifiers.
To demonstrate the gain-phase measurement ability of the gainphase meter, the gain and phase shift of a transformer coupled audio
frequency amplifier was obtained over a frequency range from 40 to
10,000 cycles.
10.
A block diagram of the circuit used is shown in Fig.
A plot of gain and phase angle shift about the middle frequency
value of 1800 is shcrnn in Figs. 11 and 12.
test are given in Table V.
Tabulated results of this
Hew1ettFa ckflrcl
AUdio
Oscillator
200C
El
'!'rpnsformer
Cou.Dled
AUdio
Amplifier
E2
----
~
:.:.
1
~
t-+
~~<'.,
E
l
~
~
E2
Cir-cular
Sweep
Gerlerator
:o~
~
I
Pulse
Gel!erator
Dumont
241
OacillogrslJh
)
BLOCK DIAGRAH OF CIRCUIT USED TO f-lEASURE GAIN
AND PHASE AiGLB OF AN AUUIO FREQUENCY AMPLIFISR
Figure 10
~
l
S
.2
': )( J
~
."
"!~
0:
u
0:
:j
~]
;}...
w
l~-T"1-~:T--r--;-iTIT-,-~~-r;--;:-;-..,...-;--;-:-...,..,.---:------,.---
~
-I-H-'--.-.l
-
~-t-
T
T• i
11;--
,I'
;::r-_ -:-
-
~.
1T
-
.
'l
\
l'
L
.
-
.-
\
\
. ~
;
~
:
...
-
1
.
\.
,
ol[,)
+
.
n
m
.
rl
II
~
,.;o~
:i~
....
I(/) ,r!0
r~ . . . t
Q>
~
.::l
• • en
11')
§
L-+----l
I
l-
--
-4----
~
8_1'"~ii'...~~t-_!+-t----:
t~Vr-?_-_"+-.~..~i:...:~...:.·~·~~J_-J'_ I- -:-I-t-~~R,,-=-)<---1/-7/Li-_-;~'*.~
;' ;~-_2..~.t--- J~: ~
3
2
1
9
8
7
6
5
4
II
1(');1
:::K
.
~K
-BK
/
/
/
,
/
/
,
\
0
\\
.
I
\
\
I
I
I
1°
I
t· .
I'l)
lr-f
I
i.
I.
/
- - --.--
/
,"
.
TABLE V
Gain and Phase Angle Data of
A 6c5 Transformer Coupled Audio Frequency Amplifier
Frequency
~
l'.:2
Gain El
Phase Angle Ee")
-&--
1800
40
20.5
223.5
+ 53.5
60
23.0
215.5
+
80
24.0
208.5
+ 28.5
100
25.0
202.5
+
22.5
200
25.5
193.0
+
13.0
400
26.0
184.5
+
4.5
600
26.0
179.5
-
0.5
800
26.0
177.0
-
3.0
1000
26.0
174.0
-
6.0
2000
26.0
163.4
- 16.6
3000
25.5
155.0
- 25.0
4000
2L~.5
145.5
- 34.5
6000
22.5
128.0
- 52.0
8000
20.0
112.0
- 68.0
10000
18.0
97.0
- 83.0
held constant at .5 volt.
35.5
42
SUMMARY AND CONCLUSIONS
The gain-phase meter, when properly adjusted, will give unambiguous phase angle readings from 0 to 360 degrees with a variation of
degree and a gain reading of -125 to +125 with a deviation of
:!.5
4% for
sinusoidal input voltages of frequencies from 40 to 10,000 cycles per
second.
To obtain the above measurements, the two test voltages or
currents must establish a minimum of 0~5 volt at El and E2 terminals.
Input voltages greater than 20 volts require a common attenuator ahead
of the gain-phase meter.
It has been found that the time required to
adjust the circular sweep and to calibrate the gain measuring scale,
which has to be carried- out with each frequency change, is reduced considerably when the operator becomes farJiliar with the usage of the
gain-phase meter.
The gain and phase-angle measurements of two sinusoidal voltages
or currents are indicated by the gain-phase meter in such a manner that
one can easily correlate the results with the vector plot of the same
voltages or currents.
A 5-inch cathode-ray tube, used for the gain and
phase-angle indication, makes it possible to use this instrument for
classroom demonstration.
The author of this thesis feels that this instrument, the gainphase meter, will be of great value in classroom demonstration i'fOrk and
will allow rapid and accurate laboratory measurements of gain and
phase-angle of voltages or currents as found in communication and electronic circuits.
Although this instrument was designed primarily for
communication work, it can be used to measure the gain and
between any two or more voltages or currents.
phase-an~le
43
BIBLIOGRAPHY
Arguimbau, L.B., Vacuum Tube Circuits, N.Y., Wiley, 1948, pp. 354-356.
Chance, B., Hughes, V., I!acNichol, E.F., Sayre, E., and Williams, F.E.,
Vvaveforms, N.Y., McGraw-Hill, 1949, pp. 4h-h5.
Everest, F.A., Phase Shifting up to 360 Degrees, Electronics, Vol. 14,
pp. 46-49, 1941.
Florman, E.F., and Tait, A., An Zlectronic Phasemeter, Proceedings of
the I.R.E. Vol. 37, Feb. 19M, pp. 207-210.
Ginzton, E.L., An Electronic Phase-Angle Meter, Electronics Vol. 15,
1942, pp. 60-62.
May
M.I.T. Radar School Staff Members, Principles of Radar, N.Y., McGrawHill, 1946, Chapter 3, pp. 31-33.
Sollar, T., starr, M.A., and Valley, G.E.,Jr., Cathode-R~ Tube
N.Y., McGra~Hill, 1948, pp. 299-302.
Displ~s,
Terman, F.E., Measurements in Radio Engineering, H.Y., MCGraw-Hill,
1935, pp. 324-325.
Terman, F.E., Radio Engineers Handbook, N.Y., McGraw-Hill, 1943,
pp. 354-366.
Vandaven, E.O., Phase Meter, Electronics Vol. 21, June 1948, pp. 1421!~3.
Watton, A., Modulated-Beam Cathode-Ray Phasemeter, Proceedings of the
I.R.E. Vol. 32, May 1944, pp. 268-272.
VITA
Gabriel George Sldtek was born on October 25, 1919, at
st. Joseph,
Missouri, the son of tIr. and 1Irs. Felix John Sldtek.
His pre-college education was obtained in the public schools of
St. Joseph, Missouri between the years 1926 and 1938.
From 1938 to 1939 he ormed and
opera~ed
an electric service shop
that brought about financial aid to the college work begun in 1939.
On September 15, 1939, he entered the 1ussouri School of ,·nes
and Metallurgy where he received a degree of Bachelor of Sdence in
b~ectrical
Engineering in Januar,y 1943.
From January 1943 to July 1943, he
VlaS
employed by the Missouri
School of Mines and Metallurgy as an instructor in the E.S.M.W.T. Program.
On July 19, 1943, he reported for duty at the Army Air Force
Officers Candidate School in Miami Beach, Florida and on November 13,
1943 he was commissioned second Lieutenant in the Air Force Reserve.
He was assigned to the Air
1~teriel
Command, Engineering Division at
Wright Field, Dayton, Ohio, "mere he remained until discharge from the
A'rrJW.
He became a Captain in the Officer's Reserve Corps upon his
separation from tlle Army on November 2, 1946.
In November, 1946, he became an Instructor at the School of Mines
and lletallurgy of the University of Missouri in which capacity he still
serves.