~ , ~ " - ' . .

PHOT
®
• FACT PUBLICATION 205B5

D C

D
PET Circuits
Rufus P. Turner, Ph.D.

~I HOWARD W. SAMS & CO., INC.

THE BOBBS-MERRILL CO., INC.
INOIANAPOLIS • KANSAS CITY • NEW YORK
 FIRST EDITION

FIRST PRINTING - 1967

Copyright © 1967 by Howard W. Sams & Co., Inc., Indianap-

olis, Indiana 46206. Printed in the United States of America.

All rights reserved. Reproduction or use, without express per-

mission, of editorial of pictorial content, in any manner, is pro-
hibited. No patent liability is assumed with respect to the use
of the information contained herein.

Library of Congress Catalog Card Number: 67-27005

Preface

The field-effect transistor is a comparatively new semiconductor

device. Its chief merit is its high input impedance, a feature which
makes it more adaptable than the conventional ( or bipolar) transis-
tor to tube-type circuits. But it also possesses those qualities .that
recommend all transistors: simplicity; small size; ruggedness; instant
operation; high overall efficiency; and freedom from hum, micro-
phonics, and, in most cases, generation of heat.
Now that the FET is commercially available at comfortable
prices, designers and experimenters at all levels are showing interest
in its applications. This book attempts to meet the demand for
practical information on the subject.
The numerous tested circuits described on the following pages
cover the field of applications from amplifiers to test instruments
and will be especially interesting to experimenters and hobbyists.
( A great many of the circuits contain only one or two FETs.) This
collection is made up of circuits which survived impartial tests to
eliminate those which gave only marginal performance.
 The reader may use these circuits singly or in such combinations
as his needs suggest. The present collection by no means exhausts
the possibilities of FET application; therefore, many of these cir-
cuits might also serve as idea generators.
The author is grateful to Siliconix Incorporated ( Sunnyvale, Cal-
ifornia) and to Texas Instruments, Inc. ( Dallas, Texas), who kindly
supplied operating data and generous samplings of their field-effect
transistors for use in developing the circuits described in this book.

RUFUS P. TURNER
Contenu

CHAPTER I

FUNDAMENTALS OF THE FET --------------------------------------------------------------- 9

The Field Effect-Background of the FET-FET Structure and Op-
eration-FET Performance-FET Ratings-FET Types

CHAPTER 2

AMPLIFIER CmCU1Ts ----------------------------------------------------------------------------- 19

Simple Small-Signal A-F Preamplifier-Single-Stage, Low-Gain A-F
Preamplifier - Single-Stage, Medium-Gain A-F Preamplifier - A-F
Source Follower-Two-Stage, R-C-Coupled A-F Amplifier-Two-Stage,
Transformer-Coupled A-F Amplifier-High-Impedance Input For Bi-
polar Transistor-Conventional Phase lnverter-Paraphase Phase In-
verter-Dual-Channel Audio Mixer-L-C-Tuned Bandpass A-F
Amplifier-R-C-Tuned Bandpass A-F Amplifier-R-C-Tuned Band-
Suppression (Notch) A-F Amplifier-Video Amplifier-455-kHz I-F
Amplifier-Auxiliary Headphone Amplifier-Simple Audio AGC Am-
plifier-Simple D-C Amplifier
 CHAPTER 3

OsCILLATOR Cmcu1Ts ........................................................................... 43

Transformer-Feedback A-F Oscillator-L-C-Tuned A-F Oscillator-
Phase-Shift A-F Oscillator-Single-Frequency Wien-Bridge A-F Oscil-
lator-Drain-Coupled Multivibrator-Source-Coupled Multivibrator-
Self-Excited R-F Oscillator-Conventional Crystal Oscillator-Pierce
Crystal Oscillator-100-kHz Crystal Oscillator-Colpitts A-F Oscillator
-Code-Practice Oscillator

CHAPTER 4

RECEIVER AND RECEIVER-AccESSORY CmcuITs ................................... 59

Superhet Broadcast Receiver-Regenerative Broadcast Receiver-All-
Wave Regenerative Receiver-Single-Frequency, Crystal-Controlled
Converter - Beat-Frequency Oscillator - Q-Multiplier - "Selectoject"
-Audio Squelch Amplifier

CHAPTER 5

TRANSMITTER AND TRANSMITTER-ACCESSORY CmcuITs ................... 73

Single-Stage Crystal-Controlled Transmitter-Single-Stage 0.8-Watt
Crystal-Controlled Transmitter - Two-Stage Transmitter - Push-Pull
Final Amplifier-100-MHz Grounded-Gate Amplifier-Single-Ended
Frequency Multiplier-Push-Push Doubler-Four-Band Exciter-Bal-
anced Modulator-C-W Monitor-Modulating the FET Transmitters-
Single-Control ("TNT") Transmitter-Phone Monitor

CHAPTER 6

CONTROL CmcuITs ............................................................................... 93

D-C Relay-A-C R-F Relay-Touch-Plate Relay-Amplified Photo-
cell-Photo-FET Relay-Timer-FET as Voltage Variable Resistor-
Constant-Current Source-Sensitive A-C A-F Relay-Sound-Operated
Relay- Modulated-Light Detector/ Amplifier-Phase Shifter
 CHAPTER 7

INSTRUMENT CIRCUITS ......................................................................... 111

Single-FET Electronic D-C Voltmeter-Balanced Electronic D-C Volt-
meter-Electronic A-C Voltmeter-Electronic A-C Voltmeter/Milli-
voltmeter-Crystal Dual-Frequency Standard-100-kHz/10-kHz Secon-
dary Frequency Standard-Direct-Reading Audio-Frequency Meter-
4.5-MHz Sound-Marker Generator-Variable-Frequency Audio Oscil-
lator-Dip Oscillator-A-F Milliammeter-Inductance Checker-Substi-
tution-Type Capacitance Meter-Impedance Meter-Heterodyne Fre-
quency Meter-Harmonic-Distortion Meter-A-F R-F Signal Tracer
-Sensitive Light Meter-Geiger Counter

APPENDIX A

TERMINAL GUIDE FOR DIODES AND TRANSISTORS ······························ 151

APPENDIX B

MANUFACIURERS OF SEMICONDUCTORS ····································•·········· 153

APPENDIX C

DIRECTORY OF MANUFACTURERS •························································ 155

INDEX······································································································ l57
 1

Fundamentals of the FET

The field-effect transistor, because it behaves very much like a

vacuum tube, opens new approaches to semiconductor circuit de-
sign. Some circuits in which regular transistors either operate poorly
or require special components now operate efficiently and without
modification with field-effect transistors.
We can wish that the field-effect transistor had come first. Many
transistor circuits then would have- been much simpler than at
present, and the transition from tubes to transistors would have
been less painful than it was. The regular transistor, as a later
development, would have had many special-purpose applications.
Nevertheless, the late coming of the field-effect transistor detracts
little from its usefulness.
Field-effect-transistor theory differs somewhat from that of the
regular transistor. However, it can be presented in simple terms.
Obviously, the circuit designer functions with increased ease and
profit when he understands how this new semiconductor device
works. This chapter presents the necessary elementary theory so that
this understanding can be developed.
9
 11IE FIELD EFFECT
For more than a century electrical men occasionally pondered
the idea of controlling current in a conductor by applying an
electrostatic field across the latter. And undoubtedly a great many
of them tried it-unsuccessfully. Indeed, sooner or later every
experimental-minded schoolboy gets around to mounting a resistor
or a length of insulated wire between the plates of a capacitor to
see if a voltage applied to the plates will affect the current in the
resistor or wire. But no practical device emerged from any early
experiments of this kind.
HIGH CURRENT

®-=- ®t;~
POWER SU PPLY
(A) Zero control voltage.

MEDIUM CURRENT

@ @
T.P@
B ~ONTROL vgLTAGE
(B) Medium control voltage.
-=-
POWER SUPPLY

LOW CURRENT

Q fiir' CONTROL VOLTAGE

~-=-~TB~/ (C) High control voltage.
L-.J2LJ@
POWER SUPPLY

Fig. 1-1. The field effect.

The electrostatic control which these experimenters sought is

appropriately termed the -fi,eld effect. This phenomenon is illustrated
by Fig. 1-1. Here A and B are metallic plates parallel to each
other and mounted close to, but out of contact with, some sort
of resistor ( Rl) which is sensitive to an electrostatic field. If a
voltage is applied to the plates through terminals 1 and 2, the field
set up between the plates penetrates the resistor. A power supply,
represented by battery Ml, passes current through the sensitive
resistor ( Rl) and a load resistor ( RL) in series. When the control
10
voltage is zero (Fig. 1-lA), no field exists between the plates, and
resistance Rl is a very low value. Consequently, there is a high
current through the load. When a finite control voltage subse-
quently is applied to the control terminals ( 1, 2), Rl will change
and so will the load current. Thus, when the control voltage
(battery M2) has some medium value (Fig. 1-lB ), Rl assumes a
somewhat higher value than before, and the load current decreases
to a medium level. Similarly, when the control voltage is high ( Fig.
1-lC), Rl becomes high and the load current is reduced to a low
value. If the control voltage is very high, Rl becomes infinite and
the load current falls to zero. In this way the load current is
modulated by the control voltage. Another way of thinking about
the field effect is to suppose that the control voltage decreases the
sensitive resistance.
The advantage of such a field-effect device is its voltage-re-
sponsive character. That is, ideally no current is drawn from the
control-voltage source, although the controlled current may be
sizable. This is the same advantage offered by the vacuum tube.
Unlike the tube, however, the field-effect device requires no heater.
The field-effect transistor is the first practical example of such a
device.

BACKGROUND OF THE FET

Early searchers for a crystal triode were frustrated in their at-
tempts to use an electrostatic field to modulate current in a semi-
conductor. They had hoped in this way to imitate the vacuum
tube, and their later studies seeking to find out why the arrange-
ment refused to work led to the invention of the regular transistor.
That this transistor became immensely useful is history. But its low
input impedance resulted in ( 1) inability to replace the tube in
some circuits, ( 2) need for tapped coils and step-down transformers,
and ( 3) difficulty for many persons to shift from voltage-amplifier-
tube thinking to current-amplifier-transistor thinking.
Despite the success of the regular transistor, work continued
along the original line. Accordingly, several experimental field-type
transistors appeared in the United States and in Europe during
the 1950's. (Dr. William Shockley, one of the inventors of the first
transistor, reported some of his work along this line in an article,
"A Unipolar Field-Effect Transistor," in the November, 1952, issue
11
of Proceedings of the I. R. E.) The modem field-effect transistor
( FET) is the culmination of this continued investigation.
The FET has supplied the high input impedance, through elec-
trostatic control, that was sought by the original researchers. Being
more compatible with the vacuum tube than is the regular transis-
tor, the FET gives promise of eventually replacing the tube-and
the regular transistor-in many circuits used in electronic equipment.

FET STRUCTURE AND OPERATION

Fig. 1-2 shows the basic structure of the FET; and while this is
not the actual configuration of some commercial models, it is
accurate enough for explanatory purposes. The heart of the device
is a thin bar or wafer of silicon ( infrequently, germanium) with
an ohmic ( nonrectifying) contact ( A, B) at each end. The end-to-
end electrical path through this bar is termed the channel. If the
silicon is N type, as in Fig. l-2A, a P region is processed into each
face of the bar in such a way that .each such region is parallel to
the other, and a connection ( C, D) is made to each. These regions
C

(A) N-channel.

C
P-TYPE SILICON BAR

A B (B) P-channel.

Fig. 1-2. Basic structure of a field-effect transistor.

12
are termed the gates. (In most commercial FET's, the gates, if
two are used, are wired together internally and connected to a
single terminal.) The completed transistor is given the name N-
channel field-effect transistor ( NFET). If, instead, the silicon is P
type, as in Fig. l-2B, then the gates are N regions and the transis-
tor is named P-channel field-effect transistor ( PFET).
When a d-c voltage is applied between A and B, the current
carriers ( electrons in the N-channel and holes in the P-channel
FET) flowing through the bar must pass through the channel
between the two gate electrodes. The anode terminal (A) is termed
the drain, and the cathode terminal ( B) the source. In a symmetri-
cal FET, either terminal may be the source, and the other the
drain. The drain is equivalent to the plate of a tube or the collector
of a regular transistor; the source is equivalent to the cathode of a

DRAIN
DRAIN
GATE~
GATE

SOURCE SOURCE
(A) N-Channel. (B) P-channel.

DRAIN DRAIN

GATE l ~ G A T E 2
GATE~BSTRATE
SOURCE SOURCE
(C)MOS FET. (D) Tetrode.
Fig. 1-3. FET symbols.

tube or the emitter of a regular transistor; the gate is equivalent

to the control grid of a tube or the base of a regular transistor.
For FET symbols, see Fig. 1-3. To emphasize the resemblance of
the FET to the tube and to eliminate new terms, some early
researchers called the gate, drain, and source by the comparable
familiar names of grid, plate, and cathode, but their example failed
of adoption.
It is the nature of a PN junction that a thin depletion layer is
present at the junction. This is a region in which there are no
available current carriers. The depletion layer for each FET junction
is indicated in Fig. 1-4 as the region inside the dotted lines at the
gate junctions. The depletion layer may be deepened by applying
13
a reverse voltage between the gate and source, the depth increasing
with voltage. Such an increase in depth narrows the channel through
which the current carriers must pass, thus increasing the resistance
of that path.
lo------
DRAIN
DEPL£TION LAYER
',,
' I
GATE I HIGH
.--------<1111 I
I I DRAIN
I \
CURRENT
(A) Low gate voltage.

RCE

lo------

' \ I,
,,,,,,
II
(B) High gate voltage.

Fi1, 1-4. FET action (N-channel shown).

Fig. 1-4 shows FET action. Here a reverse voltage, Vos, is applied
between gate and source. A second voltage, Vns, is applied between
drain and source. These are equivalent to the grid and plate
voltages, respectively, of a tube. An N-channel unit is shown; for a
P-channel, reverse both Vos and Vns, In Fig. l-4A, the shallow
depletion layers result from the low gate voltage, and the channel
between them therefore is wide ( permitting a large number of
electrons to How through the bar), so the drain current, In, is high.
In Fig. l-4B, the gate voltage is high and it deepens the depletion
layers, causing them to penetrate farther into the bar, narrowing
the channel and reducing the drain current. When the gate voltage
reaches a critical value, termed the pinch-off voltage, the depletion
14
layers meet, completely shutting off the current. Because the
control voltage, Vos, reverse-biases the gate junction, any gate
current, loss, is exceedingly tiny ( on the order of 0.1 nanoampere).
This accounts for the high input impedance of the FET and con-
sequently its behavior as a voltage-controlled device. Because the
resistance of the silicon bar is modified by the gate-voltage field,
the FET is a true field-effect device.
The FET is called unipolar from the fact that it uses only one
type of current carrier-majority carriers ( electrons in the N-channel
FET's, and holes in the P-channel). Similarly, the regular transistor
is called bipolar because it uses both types-majority and minority
carriers ( electrons internal, holes injected, in the NPN; holes inter-
nal, electrons injected, in the PNP).

FET PERFORMANCE

The control action in an FET is similar to that in a tube. To

illustrate this, Fig. 1-5 shows a family of curves obtained by varying
the drain-to-source voltage, Vos, at various levels of gate-to-source
voltage, Vos, and noting the drain current, 10 • These curves re-
semble those of a pentode tube. Note that the avalanche breakdown
of the gate junction will be reached if Vos is made high enough.
The sudden increase of drain current at the breakdown point is
indicated by the dotted extension of each curve. This breakdown
voltage decreases as Vos increases, since Vos acts in series with V Ds
to break down the junction.
. ..
I 11
: ::
I 11

I I
I I
I I
-0.5V
.
I J
I
I

_ _ __:-::.1.w~--mrrr)
-1.5V
-2.IH
-2.5V ,.,.)
I

3.IH.-+1
:

-4.IN I

DRAIN-SOURCE VOLTAGE Ide vl------

Fig. 1-5. Typical FET performance curves.

15
 Because the FET is a voltage amplifier, its performance, like
that of a tube, may be rated in terms of transconductance. In the
FET, forward transconductance for the common-source circuit
( equivalent to the grounded-cathode tube circuit) is the ratio of a
change in drain current to the change in gate-to-source voltage
which produces it:

gcs = d I» X 1000
d-V
GS

where,
grs is the forward transconductance, in µ.mhos,
In is the drain current, in milliamperes,
VGs is the gate-to-source voltage, in volts.

Depending on make, type, and d-c operating voltages of the

FET, the transconductance ranges from a low of 35 to a maximum
of 50,000 µ.mhos.
The FET lends itself readily to use in tube-type circuits. Fig.
1-6, for example, shows a common-source audio-frequency ampli-

+ 5to35W l
5K to -
L5meg

OUTPUT
INPUT

@ 0.5to lmeg

Fig. 1-6. Typical application (FET audio amplifier).

fier stage. This is equivalent to the grounded-cathode tube-type

ampli.6er. The following chapters of this book are devoted to the
description of such practical circuits.
16
 FET RATINGS

In their data sheets, FET manufacturers give minimum and

maximum values of various operating parameters. These include
the following, dependent on make and type:

1. Common-Source Forward Transconductance ( gc.). The ratio

of din to dVos- Similar to tube transconductance (gm). Given
for a specified Vns, Vos, and f = 1 kHz. Range: 35 to 50,000
µ.mhos.
2. Common-Source Output Conductance ( Input Shorted) ( g,, •• ).
Reciprocal of drain-source output resistance. Given for speci-
fied Vns, Vos, and f = 1 kHz. Range: 1 to 600 µ.mhos.
3. Common-Source Input Capacitance ( Output Shorted) ( Ciss).
Capacitance between gate and source. Given for specified Vns,
Vos, and f = 1 kHz. Range: 2 to 65 pf.
4. Common-Source Reverse Transfer Capacitance ( Crss). Given
for specified Vns, Vos, and f = 1 kHz. Range: 1.5 to 6 pf.
5. Drain Current at Zero Gate Voltage .(Inss), The current in the
drain-to-source circuit ( i.e., through the channel) when Vos
= 0. Given for specified Vns, Ranges: Vns = 5 v: 0.1 to 10 ma.
Vns = 8 v: 5 to 25 ma. Vns = 10 v: 0.03 to 0.6 ma. Vns = 15
v: 0.2 to 20 ma. Vns = 20 v: 0.4 to 7.5 ma. Vns = 35 v: 80 to
250ma.
6. Drain Cutoff Current ( InioFF> ) . Leakage current through the
channel when Vos has been adjusted for output cutoff. Given
for specified Vns and Vos, Range ( for Vns of 15 v) : 0.05 na at
Vos = 5 v, to 0.07 na at Vos = 10 v.
7. Gate-Drain Voltage (Von), Also called drain-gate voltage. The
maximum voltage which may appear between the gate and
drain electrodes. Range: 20 to 50 v at 25°C.
8. Gate Reverse Current (loss). Also called gate cutoff current.
The reverse current in the gate-to-source circuit. Given for
Vns = 0 and a specified Vos• Range: Vos 15 v: 2 na. Vos
20 v: 10 pa to 0.5 na. V0 8 30 v: 0.1 to 30 na.
9. Gate-Source Breakdown Voltage (BVoss), The voltage at
which the gate junction will enter avalanche. Given for 10 =
1 µ.a and Vns = 0. Range: 20 to 50 vat 25°C.
10. Gate-Source Pinch-off Voltage (VP). The gate-to-source volt-
age at which the field just closes the conduction channel.
17
 Given for In = 1 na, 10 na, or 1 µ.a, and a specified Vns ( e.g.,
5 to 15 v). Range: 0.6 to 50 v.
11. Gate-Source Voltage (Vos). Also called source-gate voltage.
The maximum voltage which may appear between the gate
and source electrodes. Range: 20 to 50 v at 25°C.
12. Noise Figure (NF). Internal noise generated. Given for Vos =
0, and a specified Vns ( e.g., 15 v) and frequency ( e.g., 1 kHz,
200-Hz bandwidth). Range: 0.5 to 3 db.
13. Total Device Dissipation ( P). Maximum power which can be
safely dissipated by the FET structure. Range: 200 mw to
0.8 w in free air at 25°C.

FET TYPES
It has already been noted that FETs may be classified as N-
channel or P-channel according to whether N-type or P-type semi-
conductor material is used in the channel. Alternate terminology is
NFET and PFET.
In addition to these two basic types, a newer FET employs a
metallic gate which is insulated from the semiconductor by a thin
oxide film. This device, which provides an extremely high input
impedance because of the near-zero loss and also gives good r-f
performance, is known by four names: IGFET ( insulated-gate field-
effect transistor), MOS (metal-oxide-semiconductor), MOS FET
( metal-oxide-semiconductor field-effect transistor), and MOST ( me-
tal-oxide-semiconductor transistor). The symbol for this FET is
given in Fig. l-3C.
Another development is the tetrode FET. This unit has two
gate terminals ( "front gate" and "back gate") and its symbol is
given in Fig. 1-3D. The tetrode, which acts like a variable-pinch-off
FET, provides a high gr,/C"'"" ratio and consequent superior per-
formance as a radio-frequency amplifier.

18
 2

Amplifier Circuits

The FET offers several immediate advantages over the bipolar

transistor in amplifier circuits. Circuit design techniques, for ex-
ample, are identical-or very nearly so-with the familiar ones used
for tube circuits. And the high input impedance of the FET permits
the simple cascading of r-c-coupled stages, without loss of gain
between stages. Also, high-impedance devices-crystal microphones
and pickups, piezoelectric transducers, and capacitance-type trans-
ducers-may be operated directly into FET amplifiers.
Representative amplifier circuits are described in this chapter.
The operating data given with them were obtained with the particu-
lar FET's used by the author. Therefore, the reader is cautioned
that individual performance may vary above or below these figures
because of the spread in, FET characteristics ( the transconductance
spread, for example, can be as high as 7.5:1, and the drain current
at zero gate voltage can spread over a range of 30: 1 ) . The single-
stage amplifiers shown here may be cascaded for higher gain.
The input resistance (impedance) of any one of the circuits may
be boosted, if desired, by substituting a higher value of gate
resistor. In most cases, up to 22 megohms is permissible.
19
 SIMPLE SMALL-SIGNAL A-F PREAMPLIFIER
Fig. 2-1 shows the circuit of a simplified, common-source a-f
amplifier employing a 2N3819 FET. The source is grounded di-
rectly, and this arrangement is suitable for a-f input levels below
0.6 volt rms.

+ D·C SUPPL1
(SEE TEXT)
15V
@ 4ma ~
2K

2N3819
s

A-f INPUT @ 2meg A·F OUTPUT

Fig, 2-1. Simple small-signal a-£ preamplifier.

At a d-c supply voltage level of 6 volts ( 2.5 ma), the single stage
gives a no-load voltage gain of 2.5. At 9 vdc ( 3.1 ma), the gain is
3.7. And at 15 vdc ( 4.0 ma), the gain is 8. The maximum signal
input before output-peak clipping is approximately 0.58 volt rms.
Frequency response, referred to 1 kHz, is flat within 2 db from
50 Hz to 50 kHz. All gain figures are given on the basis of a !-
megohm output load.
The 2N3819 is encapsulated in plastic and gives no installation
problems. All wiring must be kept as short and direct as practicable,
for reduction of stray pickup at the low signal levels at which this
amplifier can operate.
The 2-megohm gate resistor ( Rl) and 2000-ohm drain resistor
( R2) both are 1/2 watt.

SINGLE-STAGE, LOW-GAIN A-F PREAMPLIFIER

Higher gain than that afforded by the preceding circuit and
higher output voltage, as well, may be obtained by self-biasing the
20
FET. A low-gain circuit of this latter type, based on a 2N4868 FET,
is shown in Fig. 2-2. Here, source resistor R3 (bypassed by capacitor
C3) provides the required gate bias. Note the similarity of this
circuit to that of the familiar grounded-cathode tube amplifier or
the common-emitter bipolar-transistor amplifier. All resistors arc
1/2 watt, and electrolytic capacitor C3 is rated at 25 dcwv.

A-F INPUT @ lmeg A-F OUTPUT

@ @
5K SOmfd

Fig. 2-2. Single-stage,low-gain a-f preamplifier.

With a d-c supply of 9 volts ( 0.3 ma), the open-circuit voltage

gain is 20. The maximum signal input before output-peak clipping
is 50 millivolts rms, and the corresponding output signal is 1 volt
rms.
All wiring must be kept as short and direct as practicable to
minimize stray pickup at the low signal levels at which this ampli-
fier can operate. One pigtail of the 2N4868 is connected to the
metal case of this FET and should be grounded, for shielding.

SINGLE-STAGE, MEDIUM-GAIN A-F PREAMPLIFIER

Higher voltage gain than that afforded by the preceding low-gain
circuit is obtained with the self-biased, common-source a-f pre-
amplifier, based on a 2N4338 FET, shown in Fig. 2-3. Note here
also the similarity of this circuit to that of a grounded-cathode tube
amplifier or a common-emitter bipolar-transistor amplifier.
With a d-c supply of 15 volts ( 8 µ,a), the voltage gain ( open
circuit) is 45, and the maximum signal input before output-peak
21
clipping is 78 millivolts rrns. This results in a maximum signal
output of 3.51 volts rrns. Frequency response, referred to 1 kHz, is
flat within 0.5 db from 50 Hz to 20 kHz, and within 4 db from
50 Hz to 50 kHz.

+ 15V

80K
8µa
l @

r·
A-f INPUT
@

@ lmeg
G
D@
2N4338
s

@
·7 A-f OUTPUT

3K 50mfd

Fig. 2-3. Single-stage, medium-gain a-f preamplIBer.

The !-megohm gate resistor ( Rl), 80,000-ohm drain resistor

( R2), and 3000-ohm source resistor ( R3) are 1/2 watt. Electrolytic
capacitor C2 is rated at 25 dcwv.
All wiring must be kept as short and direct as practicable, ·to
minimize stray pickup at the low signal levels at which this ampli-
fier can operate. Care must be taken also in the location of the
2N4338, since the gate electrode is internally connected to the metal
case of this FET.

A-F SOURCE FOLLOWER

Fig. 2-4 shows the circuit of a source follower employing a
2N4340 FET. Note that with its unbypassed source resistor ( R2)
across which the output signal voltage is developed, this circuit
closely resembles its counterparts-the tube-type cathode follower
and bipolar-transistor emitter follower. Like the latter two circuits,
the source follower is invaluable for transforming a high-impedance
input to a low-impedance output. This circuit finds regular use,
therefore, in matching a high-impedance output to a low-impedance
22
 +

l
15V
Ima

@ D
@

r·
A-F INPUT
®
@
2meg

l.5V
G

@
2N4340
s

IK
@
0.1
(
l
A-F OUTPUT
(Z•7001'll

Fig. 2-4. A-f source follower.

transmission line, matching a high-impedance transducer to a low-

input-impedance transistor stage, etc. Fixed bias is provided by the
1.5-volt cell, Ml. The voltage gain and maximum permissible input-
signal voltage are somewhat less when this cell is omitted.
With the d-c supply of 15 volts ( 1 ma), the voltage gain is 0.5,
and the maximum signal input before output-peak clipping is 1.5
volts rms. This results in a maximum signal output of 0.75 volt rms.
Frequency response, referred to 1 kHz, is Hat within 1 db from 50
Hz to 100 kHz. The gain and frequency response are given here
on a no-load basis.
The output impedance of the source follower is somewhat lower
than the 1000 ohms of source resistor R2. In this instance, it is
700 ohms. This situation is similar to that which characterizes
the cathode follower and emitter follower. In the FET circuit,

where,
Ro is the output impedance (resistance), in ohms,
R. is the source resistance ( R2 in Fig. 2-4), in ohms,
gm is the forward transconductance of the FET, in mhos,
g is the output conductance, in mhos.
0

All resistors are 1/2 watt, and Rl may be raised as high as 22

megohms if a higher input impedance is desired. Input-circuit
23
wiring must be kept as short and direct as practicable to prevent
stray pickup by this high-impedance circuit. Care must be taken
also in locating and mounting the 2N4340, since the metal case
of this FET is internally connected to the gate electrode.

1WO-STAGE, R-C-COUPLED A-F AMPLIFIER

One of the advantages aHorded by the high input impedance of
the FET is the latter's adaptability to conventional r-c-coupled
multistage amplifiers. As in tube practice, there is no limiting loss
of gain between stages as there is in r-c-coupled bipolar-transistor
amplifiers. Fig.2-5 shows the circuit of a 2-stage, r-c-coupled ampli-
fier, employing two 2N4338 FETs (Ql, Q2). Cascaded, self-biased
common-source stages are used.

+
BV
200µa 1
@ sac@

·7
® @

r·
A-F INPUT @ lmeg
G

@
s

@
2N4338

@
A-F OUTPUT

@
5K 50mfd 50mfd

Fig. 2-5. Two-stage,r-o-coupled a-f amplifier.

The total d-c drain is 200 µa at 15 volts. With this d-c supply,
the overall voltage gain of the amplifier is 1000 when potentiometer
R4 is set for full output. At this level, the maximum signal input
before output-peak clipping is 4 millivolts rms, and this corresponds
to a maximum signal output of 4 volts rms. Frequency response,
referred to 1 kHz, is down 4 db at 50 Hz and at 50 kHz. Gain and
frequency-response figures are given in terms of a 1-megohm load
across the output.
24
 All resistors are 1/2 watt. Electrolytic capacitors C2 and C4 are
rated at 25 dcwv. All wiring must be kept as short and direct as
practicable, to minimize stray pickup at low signal levels. Care
must be taken also in the location of the 2N4338's, since the gate
electrode is internally connected to the metal case of this FET.

TWO-STAGE, TRANSFORMER-COUPLED A-F AMPLIFIER

The high input impedance of the FET makes practicable the
use of step-up coupling transformers between amplifier stages, as is
common in tube-type amplifiers. The additional voltage gain due
to the step-up turns ratio of the transformer thus is easily obtained.
( In transformer-coupled bipolar-transistor amplifiers, on the other
hand, the low input impedance of the bipolar unit necessitates a
step-down ratio.)
@

0.1
lOK

A-F A-F
RI l meg OUTPUT
INPUT +
@ @ 1.:Vma lOmfd
500 50mfd -

Fig. 2-6. Two-stage,transformer-coupled a-f amplifier.

Fig. 2-6 shows the circuit of a 2-stage amplifier employing trans-

former interstage coupling. Here, transformer Tl provides a 2:1
step-up, and may be any convenient tube-type interstage trans-
former. Note the usual primary ( P, B) and secondary ( G, F) labels.
This amplifier is based on two TIS14 FETs ( Ql, Q2); however,
the same scheme may be used with other field-effect transistors.
In fact, some designers will favor transformer coupling to overcome
the low voltage amplification provided by low-transconductance
FETs.
At 1000 Hz, the voltage gain ( with potentiometer R3 set for full
output, and the amplifier loaded with 1 megohm) is 3000. This
25
corresponds to a maximum signal input of 1 millivolt rms for 3
volts rms output before output-peak clipping. For high gain, a
higher turns ratio than the 2: 1 value of the author's transformer
may be used. Frequency response of the amplifier is dictated largely
by that of the transformer. Total current drain of the circuit is
1.6 ma at 9 vdc.
All leads must be kept as short and direct as practicable, to
minimize stray pickup and stray coupling, and the transformer case
should be grounded. A fourth pigtail of the TIS14 is connected to
the metal shell of this FET and should be grounded, for shielding.
All fixed resistors are 1/2 watt, and electrolytic capacitors C2, C3,
and C4 are rated at 25 dcwv.

IDGH-IMPEDANCE INPUT FOR BIPOLAR TRANSISTOR

A convenient application of the FET is the provision of high
input impedance for a high-gain bipolar-transistor amplifier. In
this instance, the FET serves as an input-impedance transformer.

15V
l.04ma 1
.,,,,,

@ 471)( @
3.8 meg
G (SEE TEXT) ~
C@ 0.1

2N333
8
A-F
INPUT @ 2meg
A-F
OUTPUT

@
25mld

Fig. 2-7. FET high-impedance input for bipolar transistor.

Fig. 2-7 shows such an arrangement. Here, a 2N4340 FET source

follower, similar to the one described earlier in this chapter, is con-
nected ahead of a high-gain stage employing a 2N333 silicon
transistor ( Q2). The 2N333 stage alone provides a voltage gain of
26
200 when bias resistor R3 is adjusted to an optimum value for the
particular transistor used. And since the 2N4340 stage gives a
voltage gain of 0.5, the overall gain of the complete circuit is 100
when the output is terminated with 1 megohm.
The maximum signal input before output-peak clipping is 20 milli-
volts rms, and this corresponds to a signal output of 2 volts rms.
The circuit draws approximately 1.04 ma at 15 vdc.
All resistors are 1/2 watt. Electrolytic capacitor C4 is rated at
25 dcwv. Either Rl or R2 may be made a potentiometer for gain
control, although varying R2 will introduce less noise.

CONVENTIONAL PHASE INVERTER

Fig. 2-8 shows the circuit of a conventional two-triode a-f phase
inverter based on two 2N4338 FETs (Ql, Q2). The two outputs
are 180 degrees out of phase with each other. Output 1 is 180
degrees out of phase with the input signal; Output 2 is in phase
with the input signal.

@
.15
@

r·
A-F
INPUT RI
1.5V
0.4ma
OUTPUT l

COMMON

OOTPUT2

@
.15

Fig. 2-8. Conventional phase inverter.

The circuit operates in the same manner as the equivalent tube-

type circuit. That is, the signal is amplified and inverted by transis-
tor Ql, and the output of Ql ( developed across resistors R7 and
R6) is delivered to the OUTPUT 1 terminal. A portion of this output
signal is applied to the gate of transistor Q2, and is amplified and
27
inverted by this latter transistor. The output of Q2 ( developed
across resistors RS and R6) is delivered to the OUTPUT 2 terminal.
Resistors R7 and RS may serve directly as the gate-to-ground
resistors of a succeeding push-pull FET stage, or as the grid
resistors if the push-pull stage contains tubes.
Like the equivalent tube circuit, this phase inverter is self-
balancing: The signal voltage developed across resistor R6 by the
output of transistor Q2 bucks the signal voltage developed across
this resistor by the output of Ql ( these two voltages being out of
phase with each other), and this action reduces the gate-to-ground
signal voltage of transistor Q2 and thus keeps the signal at OUT-
PUT 1 and OUTPUT 2 substantially equal. The inverter draws 0.4
ma at 15 vdc.
Capacitor C3 and resistor R4 form a decoupling filter which
promotes stability; these may be omitted in some layouts if there is
no trouble from motorboating. All resistors are 1/2 watt, and
electrolytic capacitors C2 and C3 are rated at 25 dcwv. In the
2N4338, the gate electrode is internally connected to the metal
case of this FET; hence, care must be taken in the placement of
this transistor.

PARAPHASE PHASE INVERTER

A single-FET ("paraphase") phase inverter, equivalent to the
single-tube circuit, is shown in Fig. 2-9. This circuit, which employs
a Ul83 FET, exploits the 180-degree phase difference between
source output and drain output ( equivalent to cathode output and
plate output, respectively), and provides the simplest arrangement
for converting from single end to push-pull without a transformer.
The signal at the OUTPUT 1 terminal is 180 degrees out of phase
with the input signal, that at OUTPUT 2 is in phase with the input.
And the equality of drain resistor R2 and source resistor R3, to-
gether with the large amount of degeneration supplied by the un-
bypassed source resistor, keeps these two outputs approximately
equal. ( They may be equalized exactly through close adjustment
of either R2 or R3, one with respect to the other.) The circuit
draws 0.25 ma at 15 vdc.
The maximum signal input before output-peak clipping is 1 volt
rms, and the voltage gain in each half of the circuit is approximately
0.8. Frequency response, referred to 1 kHz, is flat within 3 db from
28
 @
OUTPUT I

r,
QI
@
COMMON
G
@
OUTPUT 2
0.1
A-F INPUT @ Jmeg

.....
Fig. 2-9. Paraphase phase inverter.

50 Hz to 50 kHz. Gain and frequency-response figures are given for

1-megohm output load.
All resistors are 1/2 watt, and electrolytic capacitor C2 is rated
at 25 dcwv. A fourth terminal of the Ul83 is internally connected
to the metal case of this FET and should be grounded, for shielding.

+
20K
15V
Q35ma r @
@
r-~
A-F INPUT I
®
D®
2N4868
-~
s

@ 5K
@ SOmld
A-F OUTPUT

A-F INPUT 2
I meg
GAIN
CONTROL 2

Fig. 2-10. Dual-channel audio mixer.

29
 DUAL-CHANNEL AUDIO MIXER
Fig. 2-10 shows a conventional circuit for mixing two a-f signals
or for selecting one or the other. Like the equivalent tube or
bipolar-transistor circuit, it provides separate inputs and a common
output. Separate gain controls ( !-megohm potentiometers Rl and
R2) are provided, and the resistances of these controls may be
increased to 5 megohms each, if a higher input impedance is
desired.
Using two 2N4868 FET's (Ql, Q2), the circuit gives maximum
voltage gain of 10 per half. This corresponds to a maximum signal
input of 15 millivolts rms for an output of 1.5 volts rms before
output-peak clipping. Current drain is 350 µ,a at 15 vdc.
All resistors are 1/2 watt, and electrolytic capacitor C3 is rated
at 25 dcwv. A fourth terminal of the 2N4868 is internally connected
to the metal case of this FET and should be grounded, for shielding.

L-C-TUNED BANDPASS A-F AMPLIFIER

An amplifier which can be tuned for peak output at a desired

audio frequency is invaluable in bridge balancing, signal selection,
telemetering, c-w signal peaking, selective signaling, and electronic
control. Fig. 2-llA shows such a circuit which employs two 2N4340
FET's (Ql, Q2) and is tuned by means of inductance and ca-
pacitance. Fig. 2-llB shows the frequency response.
The first stage is a common-source amplifier with an unbypassed
source resistor ( R2). The latter supplies substantial degeneration
for stability and low distortion. The second stage is also a common-
source amplifier, with separate outputs from its drain ( high imped-
ance) and source ( low impedance) to accommodate various loads.
With potentiometer R3 set for maximum gain, the maximum input
signal before output-peak clipping is 7 millivolts rms. The cor-
responding maximum signal outputs are: A-F OUTPUT 1, 2 volts
rms; A-F OUTPUT 2, 0.45 volt rms. The curve in Fig. 2-llB is
based upon a !-megohm load for A-F OUTPUT 1. The circuit
draws 0.6 ma at 15 vdc.
The inductance and capacitance values shown in Fig. 2-llA are
given for 1-kHz operation: Ll is a U.T.C. Type VI-Cl5, 5.4-henry
adjustable inductor ( or equivalent), and C2 is a .005-mfd mica
capacitor. A tuning screw allows the inductance to be varied over a
30
 +
15V
0.6ma 1.,,,,.

r·
A-F
OUTPUT I
G
COMMON
A-F
OUTPUT 2
A-F
INPUT

(A) Circuit.

(B) Response. I
't
<

o-+-----+------<
500 l(XX) 2000
FREQUENCY (Hz)

Fig. 2-11. L-c-tuned bandpass a-f ampli&er.

narrow range to set the frequency exactly to 1 kHz. Other operating

frequencies may be obtained by changing the value of the coil in-
ductance (Ll) or the capacitance (C2). If necessary, both induct-
ance and capacitance can be varied.
All wiring must be kept as short and direct as practicable, to
minimize stray coupling and pickup. The metal case of the 2N4340
is internally connected to the gate electrode, so this FET must be
mounted clear of contact with other components. All fixed resistors
are 1/2 watt.
31
 R-C-TUNED BANDPASS A-F AMPLIFIER
Resistance-capacitance tuning of a bandpass a-f amplifier is a
more compact arrangement than the inductance-capacitance tuning
described in the preceding section. Fig. 2-I2A shows a I-kHz ampli-
fier which is r-c tuned. Fig. 2-I2B shows the frequency response.
The basic amplifier is a 3-stage unit, based on 2N4340 FETs
( QI, Q2, Q3). Unbypassed source resistors ( R4, RIO, and RI3)
supply a large amount of current degeneration for stability and

+ 15V 1
l.65ma .,,,,,

@ @
32@I
C5 .01
32K @) 20K

@""@ @ @

@
"'®
C2
~I.. \®.
.,,,,,
@ (§)
l.Omfd

@
2N4340
0.1

2N4340
@ 0.1

rt
A-f@ @ lmeg
A-f
OUTPUT
INPUT 2K

(A) Circuit.

~ -10
...
~ -20+---+-+--------1
0.
V'I
~ -30 --+----#---+--------< (B) Response.
::l
~
~ -40

1000 2000
FREQUENCY (Hz)

Fig. 2-12. R-c-tuned bandpass a-f amplifier.

$2
low distortion. The overall voltage gain ( with 1-megohm output
loading) is 42 db, with potentiometer Rll set for full output, and
the odd number of stages rotates the phase correctly for negative
feedback in the path through capacitor C6 back to the input.
Tuning is accomplished with a parallel-T network (C3, C4, C5,
R5, R6, R7). This is a null network ( r-c notch filter) connected
in the feedback loop between output transistor Q3 and input
transistor Ql. The 1-mfd capacitor, C6, provides d-c blocking for
this network, and 100,000-ohm resistor R2 limits the loading of the
network. Negative feedback through the loop is sufficient, at the
42-db gain of the amplifier, to cancel the amplifier gain. However,
the parellel-T network removes feedback voltage at the network
null frequency. The result is transmission by the amplifier very
sharply at that frequency, as shown by Fig. 2-12B. The values
given for capacitors C3, C4, and C5 and for resistors R5, R6, and
R7 are for 1-kHz tuning. Parallel-T networks may be set up for
other desired frequencies, provided that the following relationships
are maintained:

C,,
C3 = C4 = 2

Rs= Rs= 2R,

The pass frequency of the amplifier then is

1
f = 27rCR

where,
f is the pass frequency of the amplifier, in hertz,
C is the capacitance of C3, in farads,
R is the resistance of R5, in ohms.

The amplifier draws 1.65 ma at 15 vdc. All fixed resistors are 1/2
watt. The capacitors and resistors in the parallel-T network must
be rated at 1 percent accuracy.
All wiring must be kept as short, rigid, and direct as practicable,
for stable operation and minimum pickup of stray signals. The gate
electrode of the 2N4340 is internally connected to the metal case

33
of this FET, so the latter should be mounted out of contact with
other components.

R-C-TUNED BAND-SUPPRESSION (NOTCH)

A-F AMPLIFIER
An amplifier which can be tuned sharply to eliminate an audio
frequency is invaluable for suppressing a heterodyne in radiophone
reception, eliminating an undesired c-w signal, separating one fre-
quency from a mixture, and removing a single hum or noise com-
ponent from a signal complex. Fig. 2-13A shows the circuit of
a resistance-capacitance-tuned amplifier of this type, based on two

@ @
321<,,;;;\ 32K
~I-01
@.,,.@
@ ,oo;~oo;
~164K·
rt
A-F RI
A-F
OUTPUT
INPUT @ ®
50mfd 50mfd

(A) Circuit.

2.0
/
---....... /
'\. I
~
~ 1.0
\ , I
I

I (B) Response.
I-
ii:' I
5 I
0 I
't-
< 0
100 1,000 !OK
FREQUENCY IHZI

Fig. 2-13. R-e-tuned hand-suppression (notch) amplifier.

34
TIS14 FETs (Ql, Q2). Fig. 2-13B shows the typical frequency re-
sponse that can be expected of the circuit. Note how a sharp drop
in output of one band of frequencies can be designed into the cir-
cuit.
The tuning network is a parallel-T filter ( C4, C5, C6, R4, R5, R6)
connected between the amplifier stages. The values given for the
filter capacitances and resistances in Fig. 2-13A are for I-kHz
elimination. Similar filters may be set up for other frequencies, pro-
vided that the following relationships are maintained:

The null frequency then is

1
f = 2-n-CR

where,
f is the null frequency, in hertz,
C is the capacitance of C4, in farads,
R is the resistance of R4, in ohms.

If the filter capacitors and resistors are closely selected and the
capacitors are high Q, the notch point ( see Fig. 2-13B) will be
very close to zero. At some distance on each side of the notch
frequency ( say at O.lf and lOf), the a-f output will be 2 volts rms
for a maximum a-f input of 3.75 millivolts rms before output-peak
clipping. This assumes that potentiometer RS is set for maximum
gain and that the amplifier is terminated with a I-megohm resistive
load.
All wiring must be kept as short, rigid, and direct as practicable,
to promote stability and to minimize stray pickup and stray
coupling. A fourth pigtail of the TIS14 is connected to the metal
case of this FET and should be grounded, for shielding. All fixed
resistors are 1/2 watt, and electrolytic capacitors C2 and CS are
rated at 25 dcwv.
The amplifier draws 0.2 ma at 9 volts de.
S5
 VIDEO AMPLIFIER
Fig. 2-14 shows the circuit of a simple video amplifier using a
single 2N3819 FET. This circuit provides a voltage gain of 5: The
maximum signal input before output-peak clipping, at I-megohm
load, is 0.6 volt rms for 3 volts rms output. Stages may be cascaded
for increased gain.

Fig. 2-14. Video amplifier.

SIGNAL
INPUT @ lmeg
SIGNAL
OUTPUT

The frequency response is within +2 db from 50 Hz to 4 MHz.

Inductor Ll ( Miller No. 4508, or equivalent) is adjustable by
means of a screw-tuned slug from 24 to 35 µ,h and is set experi-
mentally for flat output throughout the 50-Hz to 4-MHz range with
a constant-amplitude signal applied to the SIGNAL INPUT termi-
nals. Total current drain is 8.6 ma at 15 vdc.
Short, rigid leads are important in this circuit, to minimize strays
and to stabilize operation. Resistors Rl and R2 are 1/2 watt.

455-kHz 1-F AMPLIFIER

Because of the reverse transfer capacitance of the FET ( equiva-
lent to grid-plate capacitance in a tube), the field-effect transistor
will oscillate in an amplifier in which 1-c tanks in the input (gate)
and output (drain) circuits are tuned to the same frequency, unless
the circuit is carefully neutralized. Such a circuit is the conventional,
transformer-coupled intermediate-frequency amplifier. Neutraliza-
tion can be a nuisance, since special tapped i-f transformers for
36
single-ended neutralization are not readily available except in the
step-down type required by the bipolar transistor.

1-f ~
~ 500K
@
INPUT
@ @
5K 0.1

Fig. 2-15. 455-kHz i-f amplifier.

The i-f amplifier shown in Fig. 2-15 avoids this difficulty by using
a 455-kHz ceramic filter (U.S. Sonics Type A25H4C, or equivalent)
instead of a transformer. This filter, being self-resonant, requires
no tuning, and is connected simply between two amplifier stages
in place of the usual coupling capacitor. Its voltage insertion loss
is approximately 1 db.
The amplifier employs two TIS14 FET's (Ql, Q2) and provides
an overall open-circuit gain of 400: The maximum i-f input is 2.5
millivolts rms before peak clipping occurs in the I-volt rms output.
Selectivity of the circuit corresponds very closely to that of the
filter alone, i.e., - 3 db at 2-kHz bandwidth to - 40 db at 160-kHz
bandwidth. Total current drain is 0.2 ma at 9 vdc.
All wiring must be kept as short, rigid, and direct as practicable,
to minimize i-f losses and to promote stable operation. A fourth
pigtail of the TIS14 is connected to the metal case of this FET and
should be grounded, for shielding.

AUXILIARY HEADPHONE AMPLIFIER

Often magnetic headphones must be connected to a receiver or
other instrument without appreciably loading the device. This
requires an auxiliary amplifier with high input impedance.
37
 @)

r·
A-F
INPUT
@
VOLUME
CONTROL
+
Fig. 2-16. Auxiliary headphone amplifier.

Fig. 2-16 shows the circuit of an amplifier with 2 megohms input

resistance, driving a pair of magnetic headphones having 2000-
ohms d-c resistance. This amplifier, based on a single TIS14 FET,
provides a voltage gain of 40: This corresponds to a maximum a-f
signal input of 75 millivolts rms before peak clipping in the 3-volt
rms output signal ( potentiometer Rl is set for maximum volume).
The circuit draws 1.2 ma at 9 vdc. This power may be supplied
by a self-contained battery or may be taken from the power supply
of the device which delivers the signal to the headphone amplifier.
Resistor R2 is 1/2 watt, and electrolytic capacitor C2 is rated at
25dcwv.

SIMPLE AUDIO AGC AMPLIFIBR

The gain of the FET amplifier stage is easily controlled by vary-
ing the d-c gate bias voltage, the gain varying inversely with the
voltage. When the control voltage is derived from some key signal
point, such as the output of a multistage amplifier ( through recti-
fication and filtering) and applied to one or more stages, automatic
gain control results.
Fig. 2-17 shows a simple, one-stage a-f amplifier employing this
principle. The d-c control voltage is applied at the CONTROL-
VOLTAGE INPUT terminals and acts as additional gate bias for the
2N4868 FET (Ql). Initial operating bias is supplied by the voltage
drop across source-bias resistor R4. The maximum current drawn
from the control-voltage source is 60 microamperes.
When the d-c control voltage is zero, the voltage gain of the stage
is 10, with a maximum undistorted output signal of 1 volt rms. When
38
the control voltage is 6 volts de, the stage output is reduced to 0.5
millivolt rms ( better than 90 db range). A lower d-c control voltage
may be used if less db change is needed. For minimum distortion,
the maximum a-f input signal reaching the gate of the 2N4868 must
be held to 0.1 volt rms. Total current drain is 0.5 ma at 9 vdc.
+ 'N -
o.sma.l
@) ON-OFF""

@f~D
2N4868 ----7
lOK @
0.1

A-F OUTPUT
A-F INPUT

• @
50mfd

+
CONTROL-VOLTAGE INPUT
(ll-6Y de)

Fig. 2-17. Simple audio age amplifier.

All fixed resistors are 1/2 watt. The 50-mfd source bypass capaci-
tor, C2, is a 25-dcwv electrolytic, and the 0.1-mfd input and output
coupling capacitors, Cl and C3, are 100-volt plastic or paper units.
A fourth pigtail of the 2N4868 is connected to the metal case of this
FET and should be grounded as shown, for shielding.
The circuit has a great many uses other than conventional auto-
matic gain control in an audio amplifier. Any application is feasible
in which an adjustable or fluctuating d-c voltage ( at virtually zero
power) is available for varying the output of an audio channel.
When the control voltage is positive with respect to ground, use a
P-channel FET (such as a 2N3578), changing the values of R3 and
R4 as required and reversing capacitor C2.

SIMPLE D-C AMPLIFIER

Fig. 2-18 shows the circuit of a simple one-stage d-c voltage am-
plifier, employing a 2N3578 FET ( Ql). This stage has a no-load
39
voltage gain of 5.65. Similar stages may be cascaded for higher
gain, provided proper arrangements are made for biasing the gate-
source junction of each stage.
When the d-c input signal is zero at the gate of the 2N3578, there
is a maximum drain current of almost 1 ma through drain resistor
R2. This produces a voltage drop across R2, which reduces the d-c
output voltage to approximately 0.35 volt ( "false zero"). If exactly
zero ouptut is desired, the residual voltage may be eliminated with
a common output-bucking circuit. When the d-c input signal is 1.5

- 1.
'N zero sk,Jnal: lma

l
@ ON-OFF
max. signal: Oma

G
+ D-C OUTPUT
D-C INPUT lmeg 0 to 8.5V
Otot5V @ GAIN
+

Fig. 2-18. Simple d-c amplifier.

volts at the 2N3578 gate, the drain current is cut off, and the output
rises to - 8.5 volts. The output values given above are for zero load-
ing. This condition obtains only when the amplifier feeds a high-
resistance device, such as an oscilloscope, vtvm, or other amplifier.
( At 10,000-ohms load, the output drops to 5 volts de).
Because a positive input signal gives a negative output signal,
the amplifier is an inverter as well, and this feature will be welcome
in a number of applications. The unit is also a kind of current ampli-
fier: virtually all of the signal-input current is accounted for in gain-
control potentiometer Rl, ( 1.5 µ,a at 1.5 volts input), whereas the
output current flows through an external load as the result of the
amplified d-c voltage. This is 500 µ,a in a 10,000-ohm load-a cur-
rent gain of 332. ( The current and voltage gains referred to here
indicate a d-c power gain of 1,000,000 plus).
A simple d-c amplifier of this type is useful in control systems as
well as in instruments. For example, it will drive a 10,000-ohm.
40
5-volt d-c relay, with an input signal of only 1.5 volts at 1.5 micro-
amperes. Various thresholds of operation may be obtained simply
by adjustment of gain control Rl.
Layout of the amplifier is not critical. However, the 2N3578 must
be mounted clear of wiring and other components, as the gate lead
is internally connected to the metal case of this FET. Drain re-
sistor R2 is 1/2 watt.

41
 3

Oscillator Circuitr

The high input impedance and high gain of the FET promote sim-
plicity and efficiency in many transistorized oscillator circuits. Fre-
quently, the FET may be used directly in tube-type circuits and
requires no special circuit components. This latter convenience is
important, especially in crystal-controlled, r-c-tuned, and capaci-
tance-feedback oscillators.
Negligible loading of 1-c-tuned circuits by the FET may result
in higher output and lower distortion than is usually obtained with
comparable bipolar transistors. This desirable characteristic also
removes the need to tap coils for transistor impedance matching,
and tends to preserve undisturbed the Q of the tank circuit. The
high gain obtainable with many FET's ensures that sufficient output
voltage will be available for efficient feedback.
FET's are proving to be useful in oscillators of all types. They
can be used across the frequency-spectrum-from low audio fre-
quencies to high radio frequencies. The oscillator output can be a
symmetrical sinusoidal waveform or a nonsinusoidal train of pulses.
43
 TRANSFORMER-FEEDBACK A-F OSCILLATOR
Fig. 3-1 shows the circuit of an audio-frequency oscillator which
uses inductance-capacitance tuning and inductive feedback and em-
ploys one 2N2608 P-channel FET. This simple arrangement is useful
for general-purpose, single-frequency applications, such as bridge
excitation, tone signaling, signal injection and tracing, and amplifier
testing.
6V
l.Oma 1
SECONDARY

A-F OUTPUT

!Omfd

Fig. 3-1. A-f oscillator with transformer feedback.

Feedback from the drain (output) circuit to the gate (input) cir-
cuit is provided by an interstage audio transformer, Tl, which sup-
plies a 2: 1 ( or higher) step-up turns ratio between primary and
secondary windings. This transformer must be polarized correctly
for regenerative feedback, by proper connection of the primary and
secondary.
The oscillation frequency is determined by capacitance Cx and
the inductance of the transformer secondary:

1
f= 21ryLC
where,
f is the oscillation frequency, in hertz,
L is the inductance of the transformer secondary, in henrys,
C is the capacitance of Cx, in farads.
44
Since transformer manufacturers do not usually specify the induc-
tance of windings, this value must be measured-or various ca-
pacitors may be tried as C, until the desired frequency is obtained.
Because of the self-capacitance of the secondary, oscillation will
occur at one frequency when no external capacitor is in the circuit.
and this is the highest frequency at which the circuit will operate.
The feedback amplitude is adjusted with potentiometer Rl. The
latter must be set for best sine-wave output shown by an oscillo-
scope connected to the A-F OUTPUT terminals. Excessive feed-
back overdrives the FET and clips the output-wave peaks; insuffi-
cient feedback causes instability and sluggish starting.
A U.T.C. Type S-8 transformer tested in this circuit gave a fre-
quency of 500 Hz with Cx = .002 mfd, and 1100 Hz with no ex-
ternal capacitor. The open-circuit output voltage was 1 volt rms at
500 Hz, and the current drain 1 ma at 6 vdc. Comparable operation
is possible with some of the smaller, transistor-type transformers.
Since the output is capacitance coupled from the drain circuit
which includes transformer Tl, the external load will tend to de-
tune the oscillator somewhat. This effect can be eliminated, how-
ever, by adding a source follower ( see Chapter 2) to isolate the
oscillator from the load.
The gate electrode of the 2N2608 is internally connected to the
metal case of this FET, so care must be taken in the mounting of the
FET to avoid contact between the case and other components or
wiring. Resistor R2 is 1/2 watt, and electrolytic capacitor Cl is rated
at25 dcwv.

L-C-TUNED A-F OSCILLATOR

In the oscillator circuit shown in Fig. 3-2, the frequency is de-
termined by capacitance Cx and the inductance of iron-cored induc-
tor Ll:

1
f= 21r-../LC
where,
f is the frequency of oscillation, in hertz,
L is the inductance of Ll, in henrys,
C is the capacitance of Cx, in farads.

45
 w
Q2ma I.,,,.

"®® @

® 5()()1(
G
@
0.1

@ 300K
G ·[ OUTPUT I
WAVEFORM
ADJUST
0.5 ®
0.1 OUTPUT 2
COMMON

Fig. 3-2. L-c-tuned a-f oscillator.

Here, L may be any convenient coil ( from a filter choke to a high-Q

adjustable inductor), and Cx may be selected to give a desired fre-
quency. The experimenter may start with a given coil and choose
a capacitor. or start with a given capacitor and choose an inductor.
The circuit is essentially a 2-stage r-c-coupled amplifier, employ-
ing 2N2608 FET's, with the tuned circuit (LCx) forming the plate
tank of the first stage, and with overall feedback for oscillation
provided by capacitor C2. An even number of stages is necessary
for the required positive feedback.
Potentiometer Rl allows the feedback amplitude to be adjusted
for best sine-wave output, as observed with an oscilloscope con-
nected to the output terminals ( either OUTPUT 1 or OUTPUT 2).
Excessive feedback overdrives Ql and distorts the output wave; in-
sufficient feedback causes instability and sluggish starting.
With Ll = 5 hy and Cx = .005 mfd, the frequency is 1000 Hz.
The open-circuit signal voltage at OUTPUT 1 (high-impedance out-
put) is 1 volt rms, and at OUTPUT 2 (low-impedance output) is
0.6 volt rms. Current drain is 0.2 ma at 9 vdc.
All wiring must be kept short, direct, and rigid for maximum
stability and minimum stray pickup, and the entire device should
be box shielded. The gate electrode of the 2N2608 is internally
connected to the metal case of this FET, so care should be taken
46
to avoid contact between the FET's and other components. Fixed
resistors R2, R3, R4, and R5 are 1/2 watt.

PHASE-SHIFT A-F OSCILLATOR

The phase-shift oscillator is well known for its excellent sine-wave
output and the simplicity and compactness of the resistance-ca-
pacitance network which determines its frequency. Fig. 3-3 shows
an FET version of this familiar circuit, employing a single 2N4338.
The frequency-determining r-c network consists of identical ca-

@
+ 22.5V
0.2ma
70K
l
G

A-F OUTPUT

@ 100<

Fig. 3-3. Phase-shift a-f oscillator.

pacitors Cl, C2, and C3, and identical resistors Rl, R2, and R3.
Each leg of this network ( i.e., Cl-Rl, C2-R2, and C3-R3) produces
a 60-degree phase shift. The total 180-degree phase shift thus
introduced in the feedback path from drain to gate of the 2N4338
is correct for the positive feedback needed for oscillation. Oscil-
lation occurs at that frequency at which the phase angle of each
leg of the network is 60 degrees:

1
f = 10.88 RC
where,
f is the frequency of oscillation, in hertz,
R is the resistance of Rl, R2, or R3, in ohms,
C is the capacitance of Cl, C2, or C3, in farads.
47
From this relationship, C = 1/ ( 10.88fR) and R = 1/ ( 10.88fC).
All capacitors and resistors must be rated at 1 percent or better
accuracy.
The network values given in Fig. 3-3 ( Cl = C2 = C3 = .002
mfd, and Rl = R2 = R3 = 100,000 ohms) give a signal frequency
of approximately 460 Hz. The open-circuit output voltage of the cir-
cuit at this frequency is 8 volts rms. Current drain is 0.2 ma at 22.5
vdc.
In this single-stage circuit, the FET must be a strong amplifier
( i.e., its transconductance must be high), or no oscillation will be
obtained. Specifically, the voltage amplification of the circuit must
be high enough to override the insertion loss of the r-c network;
otherwise, the feedback voltage reaching the gate will be insufficient
to set up oscillation.
As in other simple oscillators, the output here is capacitance
coupled from the drain circuit which includes the input end of the
phase-shift network. Because of this, the external load may tend
to detune the oscillator somewhat. This defect may be corrected,
however, by adding a source follower ( see Chapter 2) to isolate
the oscillator from the load.
All wiring must be kept short, rigid, and direct for maximum
stability and minimum stray pickup. However, complete shielding
is unnecessary unless the oscillator is operated in an environment
of strong magnetic fields. But care must be taken to mount the
2N4338 clear of other components and wiring, as the gate electrode
is internally connected to the metal case of this FET and is vulner-
able to capacitive pickup, as well as to short circuits and grounds.
All resistors are 1/2 watt, and electrolytic capacitor C4 is rated at
25 dcwv.

SINGLE-FREQUENCY WIEN-BRIDGE A-F OSCILLATOR

Another widely used r-c-tuned audio oscillator is the Wien-
bridge type. This circuit is somewhat simpler than the phase-shift
oscillator described in the preceding section, since the Wien bridge
uses only two resistors and two capacitors to set the frequency,
whereas the phase-shift oscillator requires three capacitors and three
resistors. However, the Wien-bridge circuit requires two FET's.
Fig. 3-4 shows a Wien-bridge oscillator circuit arranged for
single-frequency audio operation. This circuit, employing two

48
 @ ® 30K
@

'1
32K l.Omfd

_oo;
@
®
2N4340
l.Omfd
@
2N4340
G
@ 8K @ lmeg D
@ OUTPUT!
WAVEFORM G s
ADJUST f----:9.
l.Omfd OUTPUT 2
@ ,oo;
@ 32K
~ COMMON
\!:V 4K

Fig. 3-4. Single-frequency Wien-bridge a-f oscillator.

2N4340 FET's (Ql, Q2), is essentially a 2-stage r-c-coupled ampli-

fier with a Wien bridge ( Cl, C2, Rl, R2, operated against R4, R5)
connected in the feedback path from the output drain ( Q2) to the
input gate ( Ql). In this frequency-determining r-c network, Cl =
C2, and Rl = R2. At one frequency, determined by these R and
C values, the feedback voltage at the Ql gate is in the proper phase
for oscillation, and at all other frequencies degeneration produced
by the feedback-voltage drop across R4 + R5 cancels oscillation.
The output of the Wien-bridge oscillator therefore has an excellent
sinusoidal waveform. When Cl = C2, and Rl = R2,

1
f = 21rRC

where,
f is the oscillation frequency, in hertz,
R is the resistance of Rl, in ohms,
C is the capacitance of Cl, in farads.

From this relationship, Cl = 1/ ( 21rfR), and Rl = 1/ ( 21rfC).

The Cl, C2, Rl, and R2 values given in Fig. 3-4 are correct for
1000-Hz operation. Excellent sine-wave output is obtained at this
frequency when potentiometer R6 is set for the best waveform
49
observed with an oscilloscope connected to the OUTPUT 1 and
COMMON terminals. The no-load output is 6 volts rms at the
OUTPUT 1 (high-impedance) terminal, and 0.4 volt rms at the
OUTPUT 2 (low-impedance) terminal. Current drain is 0.8 ma at
22.5 vdc.
For maximum stability and minimum stray pickup, all wiring
must be kept short, rigid, and direct. And since the gate electrode
of the 2N4340 is internally connected to the metal case of this FET,
both transistors must be mounted clear of other components. All
fixed resistors are 1/2 watt. Since the high output is capacitance
coupled from the Q2 drain circuit which is also coupled to the Wien-
bridge network through capacitor C4, the output load may tend to
detune the oscillator somewhat. However, this defect may he
corrected, as recommended for the phase-shift oscillator, by adding
a source follower ( see Chapter 2) to isolate the oscillator from the
load.

DRAIN-COUPLED MULTIVIBRATOR

Fig. 3-5 shows the circuit of a free-running multivibrator, based

on two 2N4340 FET's (Ql, Q2). This arrangement is equivalent to
the plate-coupled, tube-type multivibrator.
The multivibrator is essentially an r-c--coupled amplifier with the
output of each of two stages driving the input of the other. Thus,
in Fig. 3-5 the drain of Ql is coupled through capacitor Cl to the

• 15V
@ llX< ll.4ma
® llX<
@ ® @

@D
2N4340
s
•002

G
.002

G
o@
s
2N4340
·1 OUTPUT

@ IOOK @ IOOK

Fig. 3-5. Drain-coupled multivibrator.

50
gate of Q2, and the drain of Q2 is coupled through capacitor C2
back to the gate of Ql.
As in the tube circuit, the generated frequency is governed by
the time constants of the cross-coupled r-c legs:

where capacitances are in farads and resistances in ohms. If the

components are closely matched ( i.e., Cl = C2, and R2 = R3),
the formula becomes simply

f = 1
2 C2R2

The coupling resistance and capacitance values in Fig. 3-5 give

2500-Hz operation. Some adjustment of these values may be re-
quired with individual FET's.
The output signal is a rectangular wave having a no-load ampli-
tude of 12 volts peak to peak when a 15-volt d-c supply is used.
Current drain is 1.4 ma. A synchronizing signal may be injected at
either gate or either drain.
Resistors R2 and R3 must match within 1 percent or better, and
so must capacitors Cl and C2. All resistors are 1/2 watt. Wiring
should be kept short and rigid, but lead dress is not critical except
at frequencies of 5 kHz and higher. The 2N4340's should be in-
stalled out of contact with each other or with other components,
since the gate electrode is internally connected to the metal case
in this FET.

SOURCE-COUPLED MULTIVIBRATOR

A somewhat simpler free-running multivibrator circuit than the

one described in the preceding section is shown in Fig. 3-6. This
source-coupled multivibrator is equivalent to the cathode-coupled,
tube-type circuit. In this arrangement, using two 2N4340 FETs
(Ql, Q2), only one interstage coupling capacitor (Cl in the for-
ward path from Ql to Q2) is needed. The coupling back from the
second FET (Q2) to the first (Ql) is obtained through the common
source resistor, R3.
51
 15V
l0.2
+

ma

@
.001 @
300K
OUTPUT
@ 300K
@ lOK

Fig. 3-6. Source-coupled multivibrator.

The generated frequency depends principally upon capacitance

Cl and resistance R4. The Cl and R4 values shown in Fig. 3-6 are
for 550-Hz operation. Increase Cl for lower frequencies, and de-
crease it for higher frequencies. The output is a rectangular wave
with an amplitude of 10.5 volts peak to peak when the d-c supply
is 0.2 ma at 15 volts de. A synchronizing voltage may be injected
at the gate of Q 1.
All resistors are 1/2 watt. As in the previous circuit, wiring
should be kept short and rigid, but lead dress is not critical except
at frequencies of 5 kHz and higher. The 2N4340's must be installed
free from contact with each other or with other components, since
the gate electrode is internally connected to the metal case in this
FET.

SELF-EXCITED R-F OSCILLATOR

The 2N3823 FET has an excellent high-frequency figure of merit
( ratio of transconductance to input capacitance), hence is useful
for r-f applications through the vhf region. Fig. 3-7 shows the cir-
cuit of a self-excited radio-frequency oscillator employing this FET.
This is a series-fed Hartley circuit, in which the tank circuit Ll-
Cl determines the generated frequency. The values of Cl and Ll
may be determined for a desired frequency, either in terms of an
available capacitance or available inductance:
1
C, = --~-
39.5 f 2 L 1
52
 RFC R-F OUTPUT
2.Smh

@
.002
l
------------------0 6V
5.3 ma

Fig. 3-7. Self-excited r-f oscillator.

1
Li= 39.5 PC

where,
C 1 is capacitance, in microfarads,
L 1 is inductance, in microhenries,
f is frequency, in megahertz.

The inductance and capacitance values may also be found with

the aid of a reactance slide rule or 1-c-f chart. For variable-frequency
operation, either Cl or Ll may be adjustable. The coil is center
tapped, but in some instances the tap may have to be displaced
somewhat from center.
Radio-frequency output may be capacitance coupled from the
drain, as shown in Fig. 3-7. In this case, the no-load r-f output
amplitude is 2.5 volts rms when the d-c input is 5.3 ma at 6 vdc.
Link coupling out of Ll also may be employed-the conventional
manner. The r-f potential across Ll is 1.3 volts peak.
All wiring must be kept short, direct, and rigid, especially at
frequencies above 1 MHz. The case of the 2N3823 is connected
to a fourth pigtail which should be grounded, as shown. All re-
sistors are 1/2 watt.

CONVENTIONAL CRYSTAL OSCILLATOR

Fig. 3-8 shows the circuit of a crystal-controlled r-f oscillator
employing a 2N3823 FET. This circuit is equivalent to the conven-
tional plate-tuned, tube-type circuit and is operated in the same way
as the latter.

58
 @
C2 and ll: ~-FOUTPUT
@
@+
See Text
2N3823
@
0-C INPUT Ml CURRENT R-f OUTPUT
G (volts) (ma) (v rms)
6 l.40ma 0.40
9 1.41 0. 52
15 2.00 0.80
XTAL @ 22.5 2.50 1.40
® 470K 0-5 0-C
MILLIAIMIETER

D-C INPUT

Fig. 3-8. Conventional crystal oscillator.

In this arrangement, capacitance C2 and inductance Ll are chosen

to resonate at the crystal frequency. These tank-circuit values may
be determined for a desired frequency either in terms of an avail-
able capacitance or available inductance:

where,
C2 is capacitance,in microfarads,
L1 is inductance, in microhenries,
f is frequency, in megahertz.

The inductance and capacitance values may be determined also

with the aid of a reactance slide rule or 1-c-f chart. Either C2 or
Ll, or both, may be adjustable. The circuit is set to the crystal
frequency in the conventional manner-by tuning either C2 or Ll
for drain-current dip indicated by d-c milliammeter Ml. Fig. 3-8
shows the no-load r-f output voltages obtained with various values
of d-c supply voltage and current.
A simple capacitance-coupled r-f output is shown in Fig. 3-8.
However, energy may also be inductively coupled out of the oscil-
lator, with a link coil coupled to the lower end of coil Ll.
54
 All wiring must be kept short, direct, and rigid. The case of the
2N3823 is connected to a fourth pigtail which should be grounded,
as shown. Both resistors are 1/2 watt. For best high-frequency
performance, Cl, C3, and C4 should be mica capacitors; and if C2
is fixed, it should be silvered mica.

PIERCE CRYSTAL OSCILLATOR

The Pierce crystal oscillator is easy to use, since it requires no
tuning. This circuit has a long record of convenience in calibrators,
marker generators, crystal-testing oscillators, and radio transmitters.
Fig. 3-9 shows a Pierce circuit and the voltages obtained with
various values of d-c supply. This is a vigorous oscillator.
XTAL

G ®·1RFC
2.5mh 0-C INPUT
(voltsl
6
INPUT
CURRENT
(mal
1.8
R-f OUTPUT
(v, peak)
5
R-f OUTPUT
@ lO(l( 9 2.0 6.6
15 2.6 11.0
D-C INPUT

Fig. 3-9. Pierce crystal oscillator.

The operator must be familiar with the type of crystal he uses in

this circuit, as the Pierce oscillator favors the fundamental fre-
quency. Harmonic-type crystals operate here at their fundamental
frequency only.
All wiring must be kept short and rigid. The case of the 2N3823
is connected to a fourth pigtail which should be grounded, as
shown. Resistor Rl is 1/2 watt, and capacitor Cl should be a mica
unit.

100-kHz CRYSTAL OSCILLATOR

A 100-kHz crystal oscillator is invaluable for use as a frequency
standard and crystal calibrator. Fig. 3-10 shows the circuit of a
simple oscillator of this type, employing a 2N3823 FET, which
requires no tuning.

55
 ~,!mid
RFC
2.Smh
I G
I
I
I R-F OUTPUT
I
I
I
I

100-kHz @:1~";.50pf @ IOOK

I
XTAL I 6V
I
I 0.8 ma
I
I
I
I
I
I
I

Fig. 3-10. 100-kHz crystal oscillator.

The circuit is a fast starter and vigorous oscillator. At 6 volts,

0.8 ma d-c input, the no-load r-f output is 5.4 volts rms, and this
output is sufficiently nonsinusoidal that the 100-kHz harmonics
may be detected far into the radio-frequency spectrum. If the
oscillator frequency is to be set against standard-frequency signals
from WWV or some other standard, the optional air trimmer ca-
pacitor Cl may be added. A 50-pf unit should allow the crystal
frequency to be shifted sufficiently for the zero beating of a suitable
harmonic with the high-frequency standard signal.
All wiring must be kept short, direct, and rigid. The metal case
of the 2N3823 is connected to a fourth pigtail which should be
grounded, as shown. Resistors Rl and R2 are 1/2 watt, and C2 and
C3 are mica capacitors. ( The capacitance of C3 may need to be
reduced in some applications to minimize loading of the oscillator.
In all instances, coupling should be as loose as possible. )

COLPITTS A-F OSCILLATOR

The Colpitts oscillator has the advantage that its inductor needs
no tap, and no transformer is required for feedback. In spite of this
simplicity, the oscillator is capable of a good output waveform and
will operate at radio and audio frequencies.
Fig. 3-11 shows the circuit of a Colpitts a-f oscillator. While this
is essentially a single-frequency unit ( 900 Hz), the frequency may
be varied over a narrow range ( 650-900 Hz) if a slug-tuned, 5-

56
henry inductor (United Transformer VIC-15, or equivalent) is used
as shown in Fig. 3-11.

@ lOK
@ I
@
5 hy
~lA-F OUTPUT
(4.5V RMS, NO-LOAD)

WAVEFORM
ADJUST
@ 4.7K

Fig, 3-11, Colpitts a-f oscillator.

This oscillator is built around a single plastic-encapsulated 2N3819

FET (Ql). Direct-current drain is 0.5 ma at 9 volts, and the cor-
responding no-load output signal is 4.5 volts rms. With the aid of
an oscilloscope or distortion meter connected to the A-F OUTPUT
terminals, rheostat R2 may be set for best sine waveform.
The frequency may be changed by substituting other values for
the Cl and C2 capacitances shown here; however, the 10:1 ratio
must be preserved between Cl and C2. A small shift of frequency
may be obtained, when a fixed inductor is used for Ll, by varying
only one capacitance. If, for example, Ll = 5 hy and C2 is changed
to .07 mfd, the frequency becomes 1 kHz. However, rheostat R2
must then be readjusted for the best output waveform.
Source resistor Rl is 1/2 watt. Frequency-determining capacitors
Cl and C2 must be high-grade units, for stability and good wave-
form. ( While mica is desirable, such capacitors are expensive in
the large capacitances shown here; good plastic units are almost as
satisfactory.) Output coupling capacitor C3 is a 100-volt plastic unit.

CODE-PRACTICE OSCILLATOR
An interesting application of the Colpitts circuit of the previous
section is the code-practice oscillator shown in Fig. 3-12. In this
unit, the magnetic headphones themselves supply the inductance of

57
the frequency-determining circuit, and no other coil or transformer
is needed.
This oscillator, which employs a single plastic-encapsulated
2N3819 FET (Ql), gives a hefty signal (a maximum output of
4.2 volts rms is developed across a pair of Trimm 2000-ohm mag-
netic headphones). Current drain is 0.8 ma at 9 vdc.

®
lOmeg
® VOWME G
2000 (l
MAGNETIC
HEAOPHO~S
@
@ 2.7K

'N
O.Bma
~----___, .1

Fig. 3-12. Code-practice oscillator.

With the Trimm phones and with Cl = .01 mfd and C2 = 0.1
mfd, the signal frequency is approximately 1000 Hz. This can be
changed by altering the Cl and C2 values while preserving their
10:1 capacitance ratio; however, limited frequency variation may
also be obtained, without killing oscillation, by varying only C2.
The lower the capacitances, the higher the frequency is and vice
versa.
The oscillator keys cleanly and has a good smooth tone. The
headphone volume may be varied, with only slight effect on the
frequency and waveform, by means of 10-megohm rheostat Rl.
Since there is no current unless the key is depressed, there is no
need for an ON-OFF switch.
Resistor R2 is 1/2 watt. The quality of capacitors Cl and C2 is
not critical in this application, so both capacitors may be of any
type that is convenient to the user. However, if frequency stability
and excellent waveform are demanded, Cl and C2 should be good
grade 100-volt plastic units.

58
 4

Receiver and Receiver-Accessory Circuits

Use of the field-effect transistor presents few problems in con-

ventional receiver circuits. In fact, some high-voltage FET's, such
as Crystalonics Type C632, will replace tubes directly at the original
plate voltages if cathode resistors are changed ( where this is re-
quired) and i-f and r-f stages are neutralized when these stages
oscillate with triodes. Lower-voltage FET's may be employed
readily in receiver circuits set up especially for their use.
The receiver circuits given in this chapter are of especial interest
to the hobbyist and experimenter. In addition to actual receivers,
some accessory devices for use with receivers are also described.

SUPERHET BROADCAST RECEIVER

Fig. 4-1 shows the circuit of a superheterodyne broadcast-band
receiver (540-1650 kHz). Three 2N3823 FET's are employed in the
combined oscillator-converter ( Ql) and first and second r-f ampli-
fiers ( Q2, Q3); a TIS14 FET ( Q4) is used in the a-f driver stage.
The second detector is a 1N295 (Xl), and the push-pull, class-B a-f
output stage employs 2N169A bipolar transistors ( Q5, Q6).
59
 @
®
2N3823
.Iii

INffi

@)
1t- 100

!:""'

8
r,t .~i; r.;;,,....
®t~ I ®1:01
l I~

[
I. I I ®±• "·
!I I .......... I
TUNING
@)!1- [ON~
l"
 The self-contained ferrite antenna, Ll, is a Miller 2004 ( or
equivalent). Because the FET produces negligible loading, Q 1 is
connected across this entire antenna coil, and an outside antenna
(when used) is connected to the low-impedance tap (ANT), giving
a signal step-up through the autotransformer thus created. The
oscillator transformer, Tl, is a Miller 2020 transistor-type unit ( or
equivalent). Tuning throughout the standard-broadcast band is ac-
complished with dual 365-pf variable capacitor Cl-C3. Trimmer
capacitors C4 and C6 are part of the tuning capacitor, but C2 is
a separate padder with 1200-pf maximum capacitance. During
initial alignment of the receiver, the antenna and oscillator are
trimmed by means of C6 and C4, respectively, and the oscillator
adjusted additionally with the aid of the tuning slug of TL
The converter is coupled to the i-f amplifier through a conven-
tional transistor-type i-f transformer, T2, which is reverse connected,
as shown. The i-f amplifier is fixed-tuned and needs no alignment
adjustments. The self-tuning of the circuit to 455 kHz is provided
by a Clevite TF-OlA ceramic filter (Ml) which replaces the bypass
capacitor across source resistor R4, and by a U. S. Sonics A25H4C
ceramic filter ( M2) inserted between the two i-f stages.
The 1N295 diode, Xl, is a combined second detector and age
rectifier. The volume-control potentiometer ( R9) serves the addi-
tional function of d-c load resistor for this diode, and the age
voltage is taken from the top of the potentiometer and fed back
to the first i-f stage through filter C9-Rl0. The audio component of
the detector output is taken from the potentiometer through cou-
pling capacitor Cl3 and is presented to the gate of the FET driver
(Q4). The class-B coupling transformers are miniature transistor-
type units: Input transformer T3 (Allied Radio 54-D-2389, or
equivalent) has a lOK primary and 3K center-tapped secondary;
output transformer T4 (Allied Radio 54-D-2358, or equivalent) has
a 500-ohm center-tapped primary and a 4-ohm secondary.
The receiver delivers audio-output power of approximately llO
milliwatts to the speaker. With a d-c supply of 9 volts, the current
drain is ll ma at zero signal; the drain is 260 ma when a maximum
signal is received.
All fixed resistors are 1/2 watt, and electrolytic capacitors Cl4
and Cl5 are rated at 25 dcwv. Each FET has a fourth pigtail which
is connected to the metal shell internally and must be grounded
as shown.
61
 REGENERATIVE BROADCAST RECEIVER
The regenerative broadcast-band receiver circuit shown in Fig.
4-2 will interest experimenters and hobbyists. This circuit, employ-
ing one 2N3578 FET, makes a sensitive receiver which is surpris-
ingly selective for a regenerative setup. It picks up local stations
with a self-contained ferrite loop antenna, Ll, and can bring in
weak and distant stations with an outside antenna and ground.
RFC

ANT @
00-----111----,
2~1 - -......- - - - 1 l----1r---t-4-I

@!i::
FERR 1TE ANTENNA IiII ________ ___
@ 365
TUNING
@
_..
5meg ~f

!!

2000
GND
REGENERATION

Fig. 4-2. Regenerative broadcast receiver.

The 365-pf variable capacitor, C2, tunes the loop antenna through-
out the standard-broadcast band ( 540-1650 kHz), and the 200-ohm
rheostat, R2, provides smooth control of regeneration from zero to
full oscillation. Very strong stations come in satisfactorily with no
regeneration at all ( i.e., with R2 set to its maximum resistance),
whereas weak stations are missing unless a good deal of regeneration
is used. When the listener is hunting for a very weak station, he
can use maximum regeneration, which will produce a whistle when
the station carrier is located. He then can reduce the regeneration
to the point at which the whistle just disappears.
The circuit operates at battery voltages between 9 volts and 22.5
volts, the higher voltages giving highest sensitivity and loudest sig-
nals. With 2000-ohm headphones in the circuit, the current drain is
3 ma at 9 volts, and 4 ma at 22.5 volts.
For increased volume, as for speaker operation, a suitable transis-
torized audio amplifier may be added to the receiver through a
coupling transformer substituted for the headphones.
62
 ALL-WAVE REGENERATIVE RECEIVER
The simplest possible sensitive receiver for all-wave general-
coverage and amateur-band phone and c-w reception is the regen-
erative type. Fig. 4-3A shows the circuit of a receiver of this kind
which tunes from 3.5 to 35 MHz with five plug-in coils.

@
n~
ANT
RFC

(A) Circuit.
l"-DIAA'ETER, 4-PIN FORM

(B) Coil Detail.

For coll winding data,

see Table 4-L

Fig. 4-3. All-wave regenerative receiver.

A simple detector with drain tickler feedback is employed. Ll is

the tickler coil, and L2 the tuning coil. Both coils are wound on a
I-inch-diameter, 4-pin, plug-in form, as shown in Fig. 4-3B. Variable
bypass capacitor C5 is the regeneration control and is a midget
350-pf single-section unit. Audio output is delivered by Tl, which
may be any convenient transistor- or tube-type interstage trans-
former with a step-up turns ratio. Either headphones or an a-f
amplifier may be operated at the A-F OUTPUT terminals.
63
 The d-c operating voltage may be any available value between
6 and 22.5 volts, the higher values giving best sensitivity and loud-
est signals. The current drain is 4 ma at 6 vdc, and 8 ma at 22.5 vdc.
Table 4-1 gives coil-winding data. Some experimentation with
the number of turns shown in this table may be necessary, since the

Table 4-1. Coil Winding Data for All-Wave

Regenerative Receiver"'

80-METER BAND (3.5-4 MHz)

Ll: 20 turns No. 30 enameled wire, closewound immediately at top of L2
L2: 50 turns No. 24 enameled wire, closewound

40-METER BAND (7-7.3 MHz)

Ll: 8 turns No. 30 enameled wire, closewound 1/8 inch from top of L2
L2: 17 turns No. 22 enameled wire, spaced to winding length of 3/4 inch

20-METER BAND (14-14.35 MHz)

Ll: 6 turns No. 22 enameled wire, closewound 1/4 inch from top of L2
L2: 10 turns No. 22 enameled wire, spaced to winding length of 3/4 inch

15-METER BAND (21-21.45 MHz)

Ll: 4 turns No. 22 enameled wire, closewound 1/4 inch from top of L2
L2: 7 turns No. 22 enameled wire, spaced to winding length of 3/4 inch

IO-METER BAND (28-29.7 MHz)

Ll: 3 turns No. 22 enameled wire, closewound 1/4 inch from top of L2
L2: 4 turns No. 22 enameled wire, spaced to winding length of 3/4 inch
0 All coils wound on 1-inch diameter, 4-pin coil forms (Millen 45004 or equivalent).
See Fig. 4-3(B).

FET and circuit capacitances will vary in individual layouts and

will alter tuning characteristics. Coils LI and L2 must be wound
in the same direction. The coils may be plugged into a standard
tube socket.
If the receiver is used for ham-band reception, C3 will be used
to set the receiver to the top of the band ( with C2 set to zero
capacitance) and bandspread tuning will be afforded by C2. For
general-coverage tuning, C2 must be set to minimum (zero) ca-
pacitance and the receiver tuned by means of C3. Antenna trimmer
Cl must be adjusted for minimum detuning by the outside antenna
used.
64
 SINGLE-FREQUENCY, CRYSTAL-CONTROLLED
CONVERTER
Fig. 4-4 shows the circuit of a crystal-controlled superhet con-
verter which may be used ahead of an available i-f amplifier/de-
tector/a-£ amplifier channel. This arrangement is based on 2N3823
FET's as mixer ( Q 1) and oscillator ( Q2). This arrangement is
intended for highly stable single-frequency reception, such as the
monitoring of one station in standby service for communication or
control purposes.

15V
l.6ma

XTAL

Fig. 4-4. Crystal-controlled converter.

The antenna input coupler is composed of coils Ll and L2, and

tuning capacitor Cl. The inductance of L2 is selected such that the
desired signal frequency (f s) can be reached by varying Cl, and
the frequency ( f x) of the crystal ( XTAL) is chosen lower than the
signal frequency by the amount of the intermediate frequency
(flf ), i.e., fx = fs - f;r- Similarly, i-f transformer Tl must be
tunable to the desired intermediate frequency. The primary coil
( Ll) of the antenna coupler will contain one-fourth to one-third
the number of turns in the secondary coil ( L2). The oscillator is
tuned to the crystal frequency by adjustment of coil L3.
65
 As a design example, consider that the converter is to operate
into a 455-kHz i-f channel and that the received signal will be 5
MHz. ( 1 ) The crystal frequency must be 5000 - 455 = 4545 kHz.
( 2) To tune to this frequency, with C6 = 100 pf, L3 must be ad-
justable to 12.2 µ,h, neglecting circuit and FET capacitances. ( 3)
For good leeway of adjustment, let the 5-MHz signal be tuned in
with Cl set to half-range, i.e., to 50 pf. This will require that L2
be 20 µ,h, neglecting circuit and FET capacitances. This value of
inductance is obtainable in commercial slug-tuned coils to which
a primary ( Ll) may easily be added by winding insulated wire
around the bottom of the coil. Alternatively, a 20-µ,h coil system
may be wound on a I-inch-diameter form, with L2 = 6 turns of
No. 22 enameled wire spaced to a winding length of 1/2 inch, and
Ll = 3 turns of No. 22 enameled wire closewound 1/8 inch from
the bottom of L2.
For stability, all wiring must be as short, rigid, and direct as
practicable. A fourth pigtail of the 2N3823 is internally connected
to the metal case of this FET, and for shielding purposes should be
grounded as shown. All resistors are 1/2 watt; and for maximum
stability, all capacitors should be silvered mica.
At 15 vdc, the total current drain is 1.6 ma.

BEAT-FREQUENCY OSCILLATOR
In a standard superhet receiver, a beat-frequency oscillator
( bfo) is needed for c-w reception. The usual procedure is to
operate the bfo at the intermediate frequency and to couple its
output into the second detector. A slight detuning of the bfo then
will cause a beat note with the i-f carrier and will make the c-w
signal audible. Fig. 4-5 shows the circuit of a bfo, using a Ul83
FET, which may be used with a receiver having no bfo of its own.
This Colpitts-type circuit operates at 455 kHz, the oscillation
frequency being determined by 2.5-mh inductor Ll ( which may be
a conventional r-f choke of that inductance) and the series-parallel
arrangement of capacitors Cl, C2, and C3. The frequency is set
initially, with C3 set to midcapacitance, by adjustment of the tun-
ing slug in coil Ll ( Miller 21A223RBI, or equivalent). Subsequent
tuning of variable capacitor C3 adjusts the pitch of the beat note
to suit the operator's ear.

66
 3-~ 1-FOUTPUT

@~
lK

®r"
I
@ PITCH CONTROL
(See text)
4- ® 5~
@ ~- ®
Ul83

+
6 to 15V

Fig. 4-5. Beat-frequency oscillator.

At 6 vdc, the current drain is 2 ma and the no-load i-f output is

5 volts peak to peak; at 9 vdc, the current drain is 2.1 ma and the
i-f output 7 volts peak to peak; and at 15 vdc, the current drain
is 2.2 ma and the i-f output 8.5 volts peak to peak. The bfo output
is coupled to the second detector in the receiver through a 3-30-pf
trimmer capacitor, C5. This capacitor must be set for optimum signal
for a smooth beat note without receiver blocking.
Resistors Rl and R2 are 1/2 watt, and capacitors Cl and C2 are
silvered mica. For stability, all wiring must be kept short, rigid, and
direct, and the fourth pigtail of the Ul83 ( which is internally
connected to the metal case of this FET) should be grounded as
shown.
The bfo may be adapted to a higher intermediate frequency by
using a lower-inductance coil in place of the 2.5-mh unit shown in
Fig. 4-5.

Q-MULTIPLIER
The Q-multiplier is a handy device for boosting the selectivity of
a receiver in either phone or c-w reception. This device is especially
valuable when the receiver has no crystal filter. Regeneration in the
Q-multiplier sharpens the response of the i-f channel so as to peak
a desired signal or reject an undesired one. Fig. 4-6 shows the circuit
67
 RFC

TO MIXERrN RECEIVER

@ 5~1 @

@ @~I@
50K I
~
lOKWW +
REGENERATl ON V
6
6ma

""'"RECEIVER GROUND

Fig, 4-6. Q-multiplier.

of a simple Q-multiplier employing a 2N3823 FET and designed

for receivers having a 455-kHz i-f channel. The output of this in-
strument is coupled, through capacitor Cl, to the plate, collector,
or drain in the mixer stage of the receiver ( depending upon whether
the receiver uses tubes, bipolar transistors, or field-effect transistors).
The Q-multiplier shown here is a 455-kHz Hartley oscillator with
controlled regeneration. The operating frequency is determined by
coil Ll ( Miller 6323, or equivalent) which is adjustable between
0.5 and 5 mh, and capacitors C2 and C3. Initially, the instrument
is set to 455 kHz, with C3 set to midcapacitance, by adjustment of
the tuning slug in the coil. Capacitor C3 then is used subsequently
to tune the multiplier sharply to the receiver intermediate frequency.
The amount of regeneration, and thus the selectivity of the re-
ceiver, is controlled by varying the FET d-c voltage with the
10,000-ohm potentiometer, R2. Two modes of operation are ob-
tained: with a certain amount of regeneration, the Q-multiplier acts
as a notch filter, eliminating a signal to which the receiver is tuned;
with more regeneration, the i-f response is peaked, tuning the
receiver sharply to one desired signal coming in on the set.
The shortest possible leads must be used between Q-multiplier
and receiver. Also, all wiring inside the Q-multiplier must be short,
rigid, and direct. Resistor Rl is 1/2 watt, and capacitors Cl, C2, and
C4 are silvered mica. The instrument should be shielded by en-
closing it in a metal box, and for additional shielding the fourth
pigtail of the 2N3823 ( being connected internally to the metal case
of this FET) must be grounded as shown.
At 6 vdc, the total current drain is 6 ma. The d-c operating power
for the Q-multiplier may be obtained from a self-contained battery
or from the receiver power supply.

68
 "SELECTOJECT"

The "Selectoject," like the Q-multiplier and the crystal filter, per-
mits either the selection or rejection of a signal but on the basis
of the latter's audio frequency. The "Selectoject" is connected be-
tween a-f stages of a receiver, or between second detector and
audio channel.
Fig. 4-7A shows the circuit of a transistorized "Selectoject" em-
ploying four TIS14 FET's. Fig. 4-7B shows performance of the
circuit. In this arrangement, which is essentially a sharply tuned
a-f amplifier, a dual rheostat ( R4) serves as the tuning control. One
rotation of this control tunes the "Selectoject" from 100 Hz to 10
kHz. Potentiometer R9 serves as the selectivity control ( for adjust-
ing the sharpness of response to suit individual conditions) and
dpdt switch Sl serves as the function switch ( for selecting either
the pass or reject function of the circuit).
With the function switch in its PASS position, the response at the
frequency to which the "Selectoject" is tuned is shown by the solid
curve in Fig. 4-7B. With this switch in its REJECT position, the
response is shown by the dotted curve. The height of the pass curve
or the depth of the reject curve, and the width of either curve, are
controlled by the setting of potentiometer R9, highest selectivity in
either case resulting when R9 is set for maximum gain in FET Q4.
The select/reject response of the circuit results from the fact that
signal through the Q4 stage ( from input to output of the entire
circuit) is negative when the function switch is thrown to REJECT
and therefore cancels the tuned-in signal, whereas this feedback
( this time, from the source of FET Q3 to the input of the circuit)
is positive when the function switch is thrown to PASS, and there-
fore boosts the tuned-in signal. If potentiometer R9 is advanced far
enough when the function switch is set to PASS, the circuit will
oscillate, delivering a sine-wave signal at the frequency to which
tuning control R4 is set.
The resistance of R2 must be selected equal to the resistance of
R3 within 1 percent. Similarly, the resistance of R5 must be selected
equal to the resistance of R6 within 1 percent. Also, the capacitance
of C2 must equal the capacitance of C3 within 1 percent. All fixed
resistors are 1/2 watt.
For stability, all wiring must be kept as short, rigid, and direct as
practicable. A metal case, to which the circuit is grounded, will

69
 + 'N
2ma
1.-
'S'"

A-f OUTPUT
@ 2meg
500K'L------- -------- ___ j /
A-f
INPUT TUNING 500K
@ 2.5K @ 5K

@e
PASS

REJECT REJECT

@ G
SEI.ECTIVllY 500K

(A) Circuit.

,,,, ... , .
/,/REJECT

(B) Performance.

Fig. 4-7. "Selectoject."

70
shield the "Selectoject" from external fields, and for additional
shielding the fourth pigtail of the TIS14 ( which is internally
connected to the metal case of this FET) must be grounded as
shown. Total current drain is 2 ma at 9 vdc.

AUDIO SQUELCH AMPLIFIER

Fig. 4-8 shows the circuit of a simple squelch amplifier which can
hold the audio channel of a receiver silent until the receiver signal
reaches a predetermined amplitude. This single amplifier stage may
be inserted at any convenient point near the front end of the audio
channel.

A-F INPUT
@ IOK
·7A-F OUTPUT

@• 50mfd

-o------
D-C CONTROi. ON-OFF
@s
VOLTAGE INPUT

I L
6V
+o------
w
+ 54µa -
D-C THRESHOLD D.5ma
VOi.TAG£ INPUT

Fig. 4-8. Audio squelch amplifier.

The circuit is a simple r-c-coupled amplifier, based on a 2N4868

FET (Ql), which is biased to cutoff by a d-c gate voltage applied
to the THRESHOLD terminals. A d-c voltage, applied to the D-C
CONTROL VOLTAGE terminals, bucks this bias and activates the
amplifier whenever it exceeds the threshold cutoff voltage. The con-
trol voltage may be derived from the i-f amplifier (by rectification)
or from the 2nd detector of the receiver. When the amplifier is in
operation, it has a voltage gain of 10, with a maximum undistorted
output of 1 volt rms. This means that the a-f input voltage at the
71
gate of the 2N4868 must not exceed 0.1 volts nns for any setting of
gain-control potentiometer Rl.
Main current drain is 0.5 ma at 9 vdc. Virtually all of the cur-
rent taken from the 6-volt threshold source Hows through resistor
R2 and is 54 microamperes. The d-c control voltage appears
across resistor R3, and this creates a current of 0.6 ma through
that resistor. While separate d-c voltage sources are shown for
the main supply and the threshold, these voltages may be ob-
tained from appropriate points in the receiver circuit (as along a
voltage divider in the power supply.)
If a particular receiver circuit gives voltage of the polarity op-
posite to that shown in Fig. 4-8, use a P-channel FET ( such as
2N3578), change the values of R4 and R5 as required, and reverse
capacitor C2.
Lead dress may be important in individual installations, so all
leads must be kept as short and direct as practicable. Route them so
as to avoid stray coupling in the receiver. A fourth pigtail of the
2N4868 is internally connected to the metal case of this FET and
must be grounded as shown, for shielding. Resistors R2, R3, R4,
and R5 are 1/2 watt. Capacitors Cl and C3 are 100-volt plastic units,
and C2 is a 25-dcwv electrolytic.

72
 5

Transmitter and
Transmitter-Accessory Circuits

The special features of the FET which recommend this transis-

tor for use in oscillators (see Chapter 3) also enable it to perform
efficiently in low-powered transmitter circuits. In fact, many tube-
type oscillator and r-f amplifier circuits for transmitters need no
change other than voltage reduction if the tubes are replaced with
suitable FET's. This means that standard coils, tuning capacitors,
and chokes may be used, and also that no loss results in capacitance-
type interstage coupling in transmitters and exciters. Furthermore,
the high input impedance of the FET tends to preserve the high Q
of the transmitter tank circuit.
Low-powered circuits offering high overall efficiency and relative
simplicity are possible in this area with FET's. The transmitter cir-
cuits described in this chapter are of especial interest to the radio
experimenter, novice, and ham. But they will also suggest applica-
tions to communications professionals. In addition to actual trans-
mitters, some accessory devices for use in transmitting stations are
described. Additional devices may be found in Chapter 7.

73
 Operation of transmitters is regulated by the Federal Communica-
tions Commission (FCC). Be sure you are complying with their
rules and procedures before you attempt to put a signal "on the air."

SINGLE-STAGE CRYSTAL-CONTROLLED TRANSMITTER

Fig. 5-1 shows the circuit of a simple crystal-controlled, oscillator
type c-w transmitter with a low-impedance output. This single-
stage unit, based on a 2N3823 FET, covers the amateur bands from
10 to 80 meters and is capable of 56 milliwatts d-c power input.

@ @
G
TUNING
50pf

C
m:wm
'
I
I

@ 47()1( +
+
KEY
c XTAL
.001
0-5 D-C
MILLIAMMETER
L
D-C INPUT

Fig. 5-1. Single-stage 56-mw crystal-controlled transmitter.

The circuit is that of the basic drain-tuned crystal oscillator. The

tuned tank circuit comprises 50-pf variable capacitor C2 and tank
coil Ll. Each plug-in "coil" consists of a tank coil (Ll) and link-
coupling coil ( L2) as a unit ( Millen low-powered, single-ended
40,000 series, or equivalents), and separate plug-ins are available
for the 10-, 20-, 40-, and 80-meter ham bands. The 20-meter unit
can be used also for the 21-meter band. Special coils wound to suit
individual demands will cover other frequencies.
The low-impedance output may feed a coaxial antenna feeder
directly, or may operate into a suitable antenna coupler. In portable
emergency applications, a whip antenna may be connected directly
to the top of coil Ll.
With the key depressed, the circuit is tuned, without antenna or
other load, by adjusting C2 for drain-current dip, as indicated by
0-5 d-c milliammeter Ml. The antenna then is connected and any

74
adjustments made to the antenna coupler ( if one is used) to "load"
the drain current up to the highest value at which the oscillator
starts readily each time the key is closed. Drain current is 1.40 ma
at 6 vdc, 1.41 ma at 9 vdc, 2.0 ma at 15 vdc, and 2.5 ma at 22.5 vdc.
With an active crystal, the oscillator can be made to act as a
doubler, i.e., tank Ll-C2 may be tuned to twice the crystal fre-
quency. The r-f output will be lower, however, than in straight-
through operation.
In addition to serving as a single-stage transmitter, this circuit
may also be used as the crystal oscillator in a multistage transmitter.
Either inductive or capacitive output coupling may be used in the
latter application.
All wiring must be kept as short and rigid as practicable. And
the fourth pigtail of the 2N3823 (being internally connected to the
metal case of this FET) must be grounded as shown. Resistors Rl
and R2 are 1/2 watt; capacitors Cl and C3 are mica.

SINGLE-STAGE 0.8-WATI CRYSTAL-CONTROLLED

TRANSMITIER
The U222 FET will provide a maximum of 1 watt r-f output at
frequencies as high as 200 MHz. This capability makes the U222
especially attractive for transmitter use.
Fig. 5-2 shows the circuit of a single-stage crystal-controlled c-w
transmitter using the U222. This circuit is similar to the one de-

@ 100<

Fig. 5-2. Single-stage 0.8-watt crystal-controlled transmitter.

75
scribed in the preceding section, except that the U222 circuit em-
ploys shunt feed of the d-c supply through a 2.5-mh r-f choke (LS).
This allows the bottom of the tank circuit, Ll-C4, to be grounded.
Plug-in coil sets (Ll-L2) permit the transmitter to be tuned to any
frequency for which a crystal is available. Each plug-in set consists
of a tank coil (Ll) and link-coupling coil (L2) assembled as a unit.
The 11illen 40,000 series of low-powered single-ended coils provides
sets for the 10-, 20-, 40-, and 80-meter ham bands (the 20-meter set
can be used also for the 21-meter band). For nonham frequencies,
other Ll-L2 coils may be wound to suit individual demands.
The circuit is tuned, with the key depressed and with no antenna
or other load, by adjusting C4 for drain-current dip, as indicated by
0-50 d-c milliammeter Ml. Then, the antenna is connected and any
adjustments made to the antenna coupler (if one is used) to "load"
the drain current up to 35 ma minimum. At this point, the d-c
power input is 0.8 watt. The low-impedance output may be con-
nected directly to a coaxial antenna feeder or may operate into a
suitable antenna coupler.
Like the preceding single-stage transmitter (Fig. 5-1), this one may
serve either as a complete c-w transmitter or as the oscillator of a
multistage b·ansmitter.
Resistors Rl and R2 are 1/2 watt, and capacitors Cl, C2, and
CS are mica. All wiring must be short and rigid, for stability. The
gate electrode of the U222 is internally connected to the metal case
of this FET which accordingly must be kept out of contact with
other components.

TWO-STAGE TRANSMITTER
Fig. 5-3 shows the circuit of a crystal-controlled oscillator-ampli-
fier type of transmitter, employing single-ended stages. A 2N3823
FET is used in the oscillator (Ql) and a U222 FET in the r-f ampli-
fier (Q2). The d-c power input of the amplifier is 0.8 watt.
The oscillator is tuned by tank circuit Ll-C2. Here, Ll is a plug-
in coil chosen to resonate with C2 at the crystal frequency. The
amplifier is tuned by tank circuit L2-C6, and here a plug-in coil
set (L2-L3) is used: L2 resonates with C6 at the crystal frequency,
and LS is a link-coupling coil. For ham bands, the oscillator and
amplifier coils both may be commercial low-powered units, such
as the Millen 40,000 series of single-ended units. For Ll, the main

76
 @
@ 2N3823 ,----------11--, ~-
@
0S CIUATOR D
C
@
~T
r:
I
I
I
I
I
I
RFC I
I
I
I
I
XTAL I
@ I
I
I
_ _ _ _...,._ _ _ _- _ _ _ _ _ __,.___ _ _ _ _ _,... _____ JI
.001 :

®
0-1 D-C 22.5V
MILLIAMMETER 38ma

'-------------+---0 +
+
-

Fig. 5-3. Two-stage transmitter.

coil is used without the link winding. The low-impedance R-F

OUTPUT terminals of the transmitter can feed a coaxial antenna
feeder directly or can operate into a suitable coupler for a higher-
frequency antenna.
A single 0-1 d-c milliammeter (Ml) is used for tune-up purposes.
When meter switch Sl is in its OSC position, the meter is switched
across 27-ohm resistor R3, the scale is converted to 0-3 ma, and the
meter reads oscillator current; when Sl is in its AMP position, the
meter is switched across 1.3-ohm shunt resistor R6, the scale is con-
verted to 0-43 ma, and the meter reads amplifier current. (Both
shunt resistors are stock values). The transmitter is tuned to the
crystal frequency, in the conventional manner, by adjusting first the
oscillator capacitor (C2) and then the amplifier capacitor (C6) for
drain-current dip. Common drain-current values are 2.5 ma for the
oscillator and 35 ma for the amplifier.
All wiring must be kept short, direct, and rigid as practicable.
Overall shielding, such as complete enclosure in a metal box, is ad-
vantageous but not mandatory. However, the fourth pigtail of the
2N3823 (being internally connected to the metal case of this FET)
must be grounded as shown. The gate electrode of the U222 is in-

77
 ;
en
~
l=- R-F
~ ! I rj
rl®
s @)A1~5~f @

~
~ITTPITT
'= :@u222 @a-r-~
cxJ e: I ,..--.... s
[ I G
D
!- l@
I
I
C4
"
L----o---
2.5mh

f
!"'
#.
15pl

@i.001
0-100 D-C
MILLIAMMETER +
+
22.5V 70ma
i·
ternally connected to the metal case of that FET, so the U222 must
be kept clear of all contact with other components and wiring. All
fixed resistors are 1/2 watt, and all fixed capacitors are mica.

PUSH-PULL FINAL AMPLIFIER

A push-pull final amplifier capable of approximately 1.6 watts
d-c power input is shown in Fig. 5-4. This amplifier may be driven
by the two-stage circuit shown in Fig. 5-3 or by the crystal oscillator
shown in Fig. 5-2. Two U222 FET'S are used.
The amplifier employs link-coupled input and output (low-imped-
ance coupling coils Ll and L4). For ham bands, plug-in coils (such
as the :'.\fillen 40,000 series, low-powered, center-tapped type) may
be used both for the input (Ll-L2) and output (L3-L4) tanks. Both
tanks are tuned to the incoming frequency.
In many instances, the amplifier will self-oscillate and will need
to be neutralized. The 15-pf variable capacitors (CS, C4) are indi-
cated for this purpose, but will not be needed in every layout. The
circuit should be tried first without them, noting if self-oscillation
occurs at any setting of hming capacitors ClA-ClB and C5A-C5B.
:'IJ'etutralizing is accomplished in a conventional manner.
All wiring must be kept short, direct, and rigid; and care should
be taken to avoid contact between the FET's or between them and
other components, since the gate electrode is internally connected to
the metal case of the U222. Resistors Rl and R2 are 1 watt, and
capacitors C2 and C6 are mica.
Current drain is approximately 70 ma at 22.5 volts, but may be
higher with individual FET's.

100-MHz GROUNDED-GATE AMPLIFIER

The grounded-gate FET amplifier is the counterpart of the
grounded-grid h1be amplifier. Neither of these amplifiers requires
neutralization; hence, both are usable at very high frequencies in
circuits which are simpler than they otherwise might be. The
grounded-gate ( and grounded-grid) amplifier has the additional ad-
vantage of somewhat higher output than that of the grounded
source, grounded cathode, or grounded emitter under similar op-
erating conditions, since most of the excitation power also appears
in the amplifier output.
79
 @ @ Ul83
8. 7pf

@
®
R-F
INPUT

o-10 D-C 15V

MILLIAMMETIR _ 4.5ma

Fig. 5-5. 100-MHz grounded-gate amplifier.

Fig. 5-5 shows the circuit of a grounded-gate r-f amplifier suitable

for 100-MHz use. Employing a Ul83 FET, this amplifier has 50-ohm
input and output (available through coaxial jacks Jl and J2, respec-
tively).
In this arrangement, the input (source-to-gate) circuit is tuned by
means of tank Ll-C2, and the output (drain-to-gate) circuit by
means of tank L2-C3. Variable capacitors Cl and C5 are used for
coupling and impedance matching. Each of the variable capacitors
(Cl to C5) has a maximum capacitance of 8.7 pf (E. F. Johnson
160-104 midget, or equivalent) and each of the tank coils ( Ll, L2)
is a slug-tuned 0.184-0.275-µh unit ( Miller 40A227CBI, or equival-
ent). The coil slugs are set to tune the tanks to 100 MHz with C2
and C3 at half-capacitance.
Current drain is 4.5 ma at 15 vdc; this corresponds to a d-c power
input of 67.5 mw. The r-f input (driving) power is approximately
4mw.
All wiring must be kept as short, direct, and rigid as practicable;
and the fourth pigtail of the Ul83, being internally connected to
the metal case of this FET, must be grounded as shown. Fixed ca-
pacitor C4 must be silvered mica.

SINGLE-ENDED FREQUENCY MULTIPLIER

Fig. 5-6 shows the circuit of a single-ended frequency-multiplier
stage. Although this circuit gives best efficiency as a frequency
80
 r~ RFC 2.5mh
R-F OUTPUT
R-F INPUT

@ 22.5V
_ 35ma

Fig. 5-6. Single-ended frequency multiplier.

doubler, it may also be used as a tripler. Operation is entirely con-

ventional: the output tank (Ll-C4) simply is tuned to twice or to
three times the frequency of the r-f input, as desired.
The circuit, which employs a single U222 FET, may be driven
by either of the crystal oscillators described earlier (Fig. 5-1 and
5-2), but substantially higher output as a doubler or as a tripler will
be obtained when the oscillator of Fig. 5-2 is used as the driver.
Inductance Ll is selected to resonate with capacitance C4 at the
desired multiple of the input-signal frequency. For the ham bands,
plug-in coils may be any of the commercial low-powered units,
such as the Millen 40,000 series. Either capacitive output coupling
may be used, as shown in Fig. 5-6, into a high-impedance load; or
low-impedance inductive coupling may be obtained with the link
coil supplied as part of the commercial plug-in coils (Ll).
The d-c power input is approximately 0.8 watt (35 ma at 22.5 vdc).
Drain-current dip, as indicated by 0-50 d-c milliammeter Ml, indi-
cates correct tuning of the multiplier.
All wiring must be kept short, direct, and rigid; and since the gate
electrode of the U222 is internally connected to the metal shell of
this FET, the latter must be kept out of contact with other compon-
ents and wiring. All resistors are 1/2 watt, and all fixed capacitors
are mica.

PUSH-PUSH DOUBLER
The push-push doubler, using two tubes or transistors, gives ap-
proximately twice the r-f power output of the single-ended doubler.
81
 @

®al4 R-F
@
--0-.- OUTPUT

Fig. 5-7. Push-push doubler.

Fig. 5-7 shows the circuit of a push-push doubler employing two

U222 FET's.
Note that, in the conventional manner, the gates are connected
in push-pull and the drains in parallel. Because of this arrangement,
a center-tapped input tank coil ( L2) and single-ended output tank
coil (L3) are required. (Both tanks are link coupled, the input
through Ll and the output through L4). The inductance of L2 is
selected to resonate with dual tuning capacitor ClA-ClB at the
input frequency, and the inductance of L3 to resonate with tuning
capacitor C4 at twice the input frequency. Input- and output-coil
sets may be low-powered commercial plug-in units, such as the
~1illen 40,000 series.
Tuning procedure is conventional: Tune the input tank to the
input frequency by adjusting ClA-ClB for peak deflection of milli-
ammeter Ml, then adjust the output tank for drain-current dip at
twice the input frequency, as indicated by milliammeter M2. The
d-c power input is approximately 1.6 watt (70-ma drain current at
22.5 vdc).
All wiring must be as short, direct, and rigid as practicable. Resis-
tors Rl and R2 are 1 watt, and all fixed capacitors are mica. The
U222's must be kept clear of contact with each other and with wir-
ing and other components, since the gate electrode is internally con-
nected to the metal case of this FET.
82
 ,--------------',..,2
1
0 1

,~ ! OUTPUT
@sEI.ECTOR

OSCILIATOR
@ u~
®LI @ 1ST DOUBI.ER
_ u_,@u~
@ 1
I~ .002

_ I(-,@ U222
€9
l .002
2ND D O U B ~ I . £ @ 3RD DOUBI.ER
.,..-t'o------.,

h@ U222i'-----i
14 0
@)
.002T ""
1-F
!
OUTPUT
1
.001 .oin I .001
D
~
qq @
;
~

gg l
l
~ XTAL
H. 001
~ I I I I I ( ~ I I I I ~ ~ I I I I ~ ~ I I I I @~l.3

METER SW ITCH
,___ _ _;.;;1D c;;2:.::,D_ _ _---J
0

osc 30
® ® 22.5V
u+

140mal
0-1 D-C MILLIAMMETER
 FOUR-BAND EXCITER
Fig. 5-8 shows the circuit of a standard four-band exciter for
ham radio. This arrangement consists of a crystal oscillator (Ql)
followed by three doublers (Q2, Q3, Q4), each stage containing a
single U222 FET. The output (fl) of the oscillator or that of the
doublers (f2, f3, f4) is selected, as desired, by means of the single-
pole, 4-position rotary switch, S2. All r-f outputs are capacitance
coupled (through C4, C9, Cl4, and Cl9, respectively). In a
popular application of this type of exciter, the crystal oscillator
operates in the 80-meter band, and the doublers in the 40-, 20-, and
IO-meter bands, respectively.
The various stages are tuned by means of separate tank capaci-
tors: C2 for the oscillator; C7, first doubler; Cl2, second doubler;
and Cl7, third doubler. Each stage is tuned, in the conventional
manner, starting with the oscillator and progressing through the
exciter, for drain-current dip. For economy in this tuning, a single
0-1 d-c milliammeter (Ml) is switched (by means of Sl) across a
shunt resistor (RS, R6, R9, Rl2) in any stage being adjusted. These
resistances are low enough that they introduce only negligible loss
in the exciter, but multiply the meter to approximately 0-42 ma full
scale. The scheme is much simpler than would be required if the
lines had to be broken to insert the meter and closed after comple-
tion of the measurement. The 1.3-ohm resistors are stock 5 percent
units. Where economy is unimportant, however, a separate 0-50 d-c
milliammeter may be installed in each stage.
The d-c power input is approximately 0.8 watt per stage (35 ma
at 22.5 vdc). Total current drain for the exciter is 140 ma.
All wiring must be kept as short, direct, and rigid as practicable.
All resistors are 1/2 watt, and all fixed capacitors are mica. The
U222's must be kept clear of contact with each other and with wir-
ing and other components, since the gate electrode is internally con-
nected to the metal case of this FET.

BALANCED MODULATOR
For single-sideband operation, the FET makes possible a triode-
type balanced modulator which is compact and low powered. Fig.
5-9 shows such a balanced-modulator circuit employing two
matched 2N4340 FE T's. Note that the a-f (modulating) signal is

84
applied to the gates in push-pull through transformer Tl and that
the r-f (carrier) signal is applied to the sources in parallel through
capacitor C3.
Cancellation of the carrier in the output circuit (only the two
sidebands which result from the modulation remaining) results from
@) 2N4340.--------,
D

G
@ 15K

@ CARRIER
TO
SIDEBAND
!OK BALANCE
FILTER

@2N4340
s @ 15K

G
®r1~,
0
15V
R-F INPUT Lima

l i
Fig. 5-9. Balanced modulator.

the symmetry of the circuit. To ensure this symmetry, there must be

close correspondence between the two halves of the circuit. This
means that the secondary winding of transformer Tl must be accur-
ately center tapped, the two FET's must be matched, Cl must
match C2, and R2 must match R4. Any slight unbalance due to
these components or to circuit strays may be corrected by adjust-
ment of the CARRIER BALANCE potentiometer, R3. Total cur-
rent drain is 1.1 ma at 15 vdc.
All wiring must be kept as short, direct, and rigid as practicable,
and the circuit should be enclosed within a grounded shield box. All
fixed resistors are 1/2 watt, and all fixed capacitors are silvered
mica. The 2N4340's must be kept clear of contact with each other
and with wiring and other components, since the gate electrode is
internally connected to the metal case of this FET.
The terminal arrangement of the FET's can be checked against
the drawings shown in Appendix A. Manufacturers may be ob-
tained from Appendix B, and the addresses are shown in Ap-
pendix C.

85
 C-W MONITOR
Fig. 5-10 shows the circuit of a simple monitor for c-w telegraphy.
Requiring no connection to the transmitter and containing its own
power supply, this device may be placed conveniently on the op-
erating desk. It will give a reliable aural check of signal quality, as
well as making the keying audible.

@) 2N3823 D

@ lOOK

'IV 3ma

Fig. 5-10. C-w monitor.

The circuit is that of a Colpitts-type oscillating detector employ-

ing a single 2N3823 FET. The signal is tuned in by the dual 100-pf
variable capacitor C2A-C2B. The antenna may be any convenient
whip; short, stiff, vertical wire; or flexible, insulated, scatter lead
which will pick up enough signal for a comfortable headphone
level. The tuning capacitor may be detuned above or below the sig-
nal for the most pleasing beat tone. Total current drain is 3 ma at
9vdc.
Inductance Ll is chosen to resonate with one-half C2A-C2B (i.e.,
with 50 pf) at the signal frequency. For ham-band operation, a set
of low-powered commercial plug-in coils, such as the Millen 40,000
series, may be used to cover the 3.5- to 30-MHz spectrum.
The volume delivered by a pair of high-impedance magnetic
headphones will be sufficient in most environments, especially with
medium- and high-power transmitters. However, a transistorized
audio amplifier may be added to the monitor if speaker operation or
louder headphone signals are desired.
86
 All wiring must be kept as short, direct, and rigid as practicable,
and the monitor circuit should be enclosed in a shield box to which
the circuit is grounded. For further shielding, the fourth pigtail of
the 2N3823 (which is internally connected to the metal case of this
FET) must be grounded as shown. Resistor RI is 1/2 watt, and ca-
pacitors Cl, C3, and C4 are mica. The terminal arrangement for
the FET is shown in Appendix A.

MODULATING THE FET TRANSMITTERS

The FET transmitters shown in this chapter are depicted for c-w
telegraph operation. They may be amplitude modulated for tele-
phony. however, by short-circuiting the telegraph key and con-
necting a suitable a-f speech amplifier/modulator unit in series with
the transmitter d-c supply, as shown in Fig. 5-11.
TO D-C INPUT TERMINAL OF FET
TRANSM lffiR

SPEECH
AMPLIFIER/ MODULATOR
Tm

MICROPHrO-NE----1 ~~~:]l l~~~~~~-i----------,

111111
AMPLIFIER I MODULATOR
i TRANSMITTER
D-C SUPPLY

O-C SUPPLY

Fig. 5-11. Amplitude-modulation arrangement.

The only requirements are that the a-f power output of the mod-
ulator must be at least twice the d-c drain power input of the
transmitter, and that the impedance ( Z.,) of the modulator output
transformer, T 111 , match the impedance of the transmitter. Table 5-1
shows the values of impedance (for each of the applicable h·ans-
mitters in this chapter) which transformer Tm in the modulator must
match. It also shows the a-f output power needed to modulate each
of the transmitters 100 percent. The primary impedance ( Z of the 111 )

transformer matches the impedance of the transistor(s) in the mod-

ulator.
87
 Table 5-1. A-M Modulator Data
Transmitter Transmitter A-f Power Transmitter Im-
Circuit D-c Voltage (Milliwatts) pedance (Ohms)
Fig. 5-1° 6.0 16.8 4290
9.0 25.4 6360
15.0 60.0 7500
22.5 56.2 9000
Fig. 5-2° 22.5 1600 645
Fig. 5-3 22.5 1600 645
Fig. 5-4 22.5 3200 32.2
Fig. 5-5 15.0 135 3300
Fig. 5-6 22.5 1600 645
Fig. 5-7 22.5 3200 32.2
"Modulated oscillators are prohibited in some services.

Note that the modulation must be applied to the final r-f stage
in the transmitter circuit given in Fig. 5-3.

SINGLE-CONTROL ("TNT') TRANSMITTER

Fig. 5-12 shows the circuit of a c-w transmitter. This is a varia-

tion of the tuned-plate, tuned-grid (TPTG) oscillator which uses
tuning only in the output circuit. An untuned coil is used in the
input circuit. This accounts for the descriptive term TNT ( "tuned,
non tuned").
In the arrangement shown here, the drain circuit is tuned by
means of 50-pf variable capacitor C2 and a plug-in coil. Each plug-
in coil set ( Millen 43,000 Series, or equivalent), consists of an un-
tapped main coil (U) and end link (L5). Three sets provide coverage
of the 3.5-, 7-, and 14-MHz amateur bands. Corresponding un-
tuned gate coils ( LI, L2, and L3) for these bands are selected by
means of single-pole, three-position switch SL The gate coils are
preadjusted by means of their screwdriver-operated tuning slugs.
(LI, 3.5-MHz band, Miller No. 40A475CBI. L2, 7-MHz band, Mil-
ler No. 40Al55CBI. L3, 14-MHz band, Miller No. 40A336CBI).
The U222 FET (QI) draws 35 ma at 22.5 volts de, a d-c input
corresponding approximately to 0.8 watt. Lower d-c voltages will
88
 BANDSWITCH @
@ @
50 pf ~ R-F
TUNING ~OOTPUT

MILLI AMMETER
KEY JACK @ JACK
.002
+
D-C POWER INPUT
(22.5V)

Fig. 5-12. Single-control ("TNT") transmitter.

produce correspondingly lower power input. A O to 50 d-c milliam-

meter plugged into jack J2 allows drain current to be read at in-
tervals, or it can be monitored continuously.
In operation, the transmitter may be tuned throughout each band
simply by adjusting variable capacitor C2. There is sufficient dis-
tributed capacitance and circuit capacitance across the gate coil (Ll,
L2, LS) in most layouts that this coil, in true TNT fashion, needs no
separate tuning. However, the transmitter output will be maximum
around the one frequency at which the gate coil tends to self-reson-
ate, and will fall off somewhat at higher and lower frequencies. The
optimum gate-coil inductance which will give best output through-
out the band may be selected by adjusting the gate-coil slug, vary-
ing capacitor C2, and noting the relative output as indicated by an
r-f vtvm connected to the R-F OUTPUT terminals, or by a 2-volt
pilot lamp (with a I-tum pickup coil) coupled to tank coil L4. In
most instances, the gate coil will be adjusted for peak output at
the center of the band. When maximum efficiency is desired at
some extreme frequency in any band, the appropriate gate coil
may be retrimmed for maximum output at that frequency.
The gate coils must be mounted at right angles to each other
and separated as widely as practicable, to prevent power absorption
by the unused two. For maximum isolation, Ll, L2, and LS may
be enclosed in separate shield cans.
Closed-circuit jacks are used both for the key (Jl) and drain
milliammeter (J2). This allows the circuit to restore itself auto-
matically when the key or meter is removed.
89
 To prevent parasitic oscillation, all leads must be as short, rigid,
and direct as practicable. The U222 must be mounted clear of
wiring, chassis, and all other components, since the gate electrode
is internally connected to the metal case of this FET. Resistor Rl
is 2 watts, and capacitors Cl and CS are mica. Selector switch Sl
must be ceramic insulated.
The TNT circuit may be used as a doubler ( 3.5 to 7 MHz, and
7 to 14 MHz) if a driving signal is fed into the gate coil and the
drain tank is tuned to twice the driving frequency.

PHONE MONITOR
A tuned diode detector with headphones is often used to moni-
tor radiophone transmissions. The semiconductor diode has the
advantage in this application that, at least for large signals, its
response is linear, so that it gives a faithful check of the monitored
signal. But its output is relatively low, so that the headphone signal
cannot compete with room noise.
The solution is to use a headphone amplifier following the de-
tector. Fig. 5-13 shows an arrangement of this sort. The advantage

WHIP ANTENNA @)

@
PLUG-IN !OK
COIL @
25mfd

Fig. 5-13. Phone monitor.

of a field-effect transistor (here a TIS14 unit, Ql) is the resultant

zero loading of the detector by the amplifier.
A short whip antenna will pick up adequate signal from the
transmitter. The signal is tuned in by means of 365-pf variable
capacitor Cl which resonates with plug-in coil Ll at the signal
frequency. Table 5-2 gives data for coils to cover the frequency
range of 0.26 to 100 MHz in five bands. The specified units are

90
 Table 5-2. Phone Monitor Coil Data

Coil Inductance Frequency

No. (µh) Range (MHz) Coil Type"

A 1000 0.26--1 40A000CBI

B 110 o.~3 40Al04CBI
C 16 2-8 40Al55CBI
D 1.3 7-28 40Al56CBI
E 0.1 16--100 40Al07CBI
~Catalog numbers of J. W. Miller Co.

miniature, ceramic, slug-adjusted coils which may be set exactly

to frequency during calibration of the monitor. For plug-in pur-
poses, these coils may be mounted inside 1 inch-diameter 4-prong
forms, such as Millen Type 45004. For narrower-band tuning than is
afforded by this coil assortment (example: amateur or citizens'
bands), a lower capacitance (for example, 50 or 100 pf) should be
used for Cl.
The detector is an inexpensive 1N34A germanium diode (Xl).
Audio output of the diode, developed across load resistor Rl, is
applied to the gate of the TIS14. Capacitor C3 blocks the d-c
component in the output of the diode. Since the load resistor (Rl)
is a potentiometer, efficient control of headphone volume is assured.
Current drain is 1.2 ma at 9 vdc.
Tuning capacitor Cl is a midget, broadcast-type variable. C2 is
a mica capacitor, and C3 a 100-volt plastic unit. C4 is a 25-dcwv
electrolytic. Resistors R2 and R3 are 1/2 watt.
All wiring must be as short, direct, and rigid as practicable. A
fourth pigtail of the TIS14 is internally connected to the metal
case of this FET and must be grounded as shown, for shielding.
The monitor may be calibrated by loosely coupling coil Ll to an
amplitude-modulated r-f signal generator, tuning in generator sig-
nals at selected frequencies by adjusting Cl for loudest headphone
response, and graduating the Cl dial accordingly.

91
 6

Control Circuits

The sensitivity of the field-effect transistor to low signal levels

and to small changes in level, together with its negligible loading
of timing circuits, signal circuits, and actuating devices, suits it to
a variety of electronic control circuitry where special provision
often must be made for the bipolar transistor. Usually, a simple
FET circuit in a control system offers straightforwardness and
economy not so easily obtained with either tubes or bipolar transis-
tors.
In solid-state control circuits, the FET allows some components,
such as high-resistance photocells and capacitive transducers, to
be operated with greater efficiency and convenience than is pos-
sible in such circuits containing bipolar transistors. And a special
member of the field-effect family-the photo-FET, a light-sensitive
FET-weds photocell and field-effect transistor.
This chapter describes several basic control circuits which are
intrinsically useful. Methods of operating relays from a variety of
signals-direct current, radio frequencies, audio frequencies, and il-
lumination levels-are shown. Perhaps these can also serve to gen-
erate ideas for other uses.

93
 D-C RELAY

Fig. 6-1 shows the circuit of an amplifier-type d-c relay based on

a single 2N3819 FET. This circuit has an input resistance of 10
megohms due entirely to resistor Rl, so it draws only 0.15 micro-
ampere from the d-c control-signal source.

@
500

G RELAY
10000
Ima

0-C INPUT ro':'\ TO

0-1.5v ~ IOmeg CONTROLLED
CIRCUIT
+

@ IK WIREWOUNO

BALANCE
@) @
lil•P--r---<f'
✓.0---~
6V ON-OFF

Fig. 6-1. D-c relay.

The 1000-ohm, 1-ma d-c relay ( Sigma 5F, or equivalent) is con-

nected in a resistance-bridge circuit whose four arms are R2, the
internal drain-to-source resistance of the FET, and the two halves
of potentiometer R3. With the D-C INPUT terminals open, R3 is
set to balance the static FET current out of the relay ( at balance,
the relay opens firmly), and the bridge should remain balanced
indefinitely without further attention.
Subsequent application of a d-c input signal of 1.5 volts will
unbalance the bridge by changing the FET drain-to-source resis-
tance; this will cause approximately 1.2 ma to flow through the
relay, closing the latter. Up to 20 megohms may be used at Rl,
thereby decreasing the drain on the signal source to as little as
0.075 microampere.
All fixed resistors are 1/2 watt. Maximum current drain from the
6-volt d-c supply (Ml) is 10 ma. The 2N3819 is a plastic-encapsu-
lated FET which, because of this construction, demands none of
the installation precautions occasioned by metal-cased units. The
plastic housing will withstand soldering temperatures.
94
 A-C R-F RELAY
The d-c relay described in the preceding section may be actuated
by an a-c or r-f control signal if a diode rectifier is operated ahead
of the circuit, as shown in Fig. 6-2.

r-
500

RELAY
IOOOQ
Ima

A-C R-F INPUT @ TO

CONTROLl.£D
IHl. 71 v rms
@ CIRCUIT
1N36«5

BALANCE ~

@ ~
l1I • ~
6V ON--OFF

Fig, 6-2. A-c r-f relay.

A shunt-diode rectifier circuit is used ( Cl-·Rl-Xl ). The main

advantage of the shunt diode is isolation from any d-c component
present in the signal source. In this arrangement, capacitor Cl is
charged to the peak value of the a-c signal ( less the small forward
voltage drop in the diode) on positive half-cycles of signal voltage,
and this peak d-c voltage ( across 10-megohm load resistor Rl) is
presented to the gate of the FET. The diode must be polarized as
shown for the d-c voltage at the gate to be negative. A silicon diode
( 1N3605) is required to work into the high load resistance, Rl; the
back resistance of germanium diodes is too low to support rectifi-
cation with such a high value of load. Resistor R2 and capacitor
C2 form a filter to remove any signal-voltage fluctuations from the
diode d-c output.
The 1000-ohm, 1-ma d-c relay ( Sigma 5F, or equivalent) is con-
nected in a resistance-bridge circuit whose four arms are R3, the
internal drain-to-source resistance of the FET ( 2N3819), and the
two halves of potentiometer R4. With the AC-INPUT terminals
open, R4 is set to balance the static FET current out of the relay
( at the balance point, the relay opens firmly), and the bridge should
then remain balanced indefinitely without further attention.
95
 Subsequent application of an a-c or r-f input signal of 0.71 volt
rms will unbalance the bridge by changing the FET drain-to-source
resistance, and this will cause approximately 1.2 ma to flow through
the relay, closing the latter.
All fixed resistors are 1/2 watt, and capacitors Cl and C2 are
mica. Maximum current drain from the 6-volt d-c supply (Ml) is
10 ma. The 2N3819 is a plastic-encapsulated FET which, because
of this construction, demands none of the installation precautions
occasioned by the metal-cased units.
The terminal arrangement of the FET's can be checked against
the drawings shown in Appendix A. Manufacturers may be ob-
tained from Appendix B, and the addresses are shown in Ap-
pendix C.

TOUCH-PLATE RELAY
The touch-plate relay has many commercial, industrial, and
household applications. Touch the metal plate lightly and the relay
closes; remove your finger and the relay opens. Call signals, safety
devices, intrusion alarms, and kindred devices are operated from
such relays.

METAL PLATE

RELAY
lOOOQ
lma

G
,.--------o
t TO
CONTROLLED

200
r
.____ _o
CIRCUIT

@
____,J
lOmeg ®

'N
ON-OFF

Fig. 6-3. Touch-plate relay.

Fig. 6-3 shows the circuit of a simple touch-plate relay based on

a single 2N3819 FET. In this arrangement, the touch plate is a
disc or plate of sheet metal or foil (2-inch minimum diameter) or
a metal ball. When the plate is undisturbed, the FET draws negli-

96
gible current through the relay ( Sigma 5F, 1000-ohm, 1-ma, or
equivalent). When the plate is touched, enough stray noise is
coupled into the 10-megohm gate circuit to boost the drain current
to approximately 1.7 ma and close the relay. (This assumes a firm
touch; a lighter touch will produce a lower current change, but the
relay will close as long as this current reaches a minimum value
of 1 ma.
If desired, a delayed dropout may be obtained by means of a
capacitor in parallel with the relay coil. With the 1000-ohm coil,
the hold-in is approximately 0.8 second per 1000 mfd of parallel
capacitance.
Resistors Rl and R2 are 1/2 watt. The 2N3819 is a plastic-
encapsulated FET which demands none of the installation pre-
cautions occasioned by metal-cased units.

AMPLIFIED PHOTOCELL
The silicon solar cell ( self-generating type) is basically a low-
voltage, high-current device. In some applications, especially where
high impedances are involved, more convenient control might be
effected if the d-c output voltage of the cell could be boosted and
the cell looked like a high-impedance source.

@
UK
G
'---------0+

D-C
OUTPUT
SOlAR
CELL
SlM

BAlANCE
@ ®
..._--l•l•~r------<l'r:r",
22• SV ON-OFF

Fig, 6-4. Amplified photocell.

Fig. 6-4 shows the circuit of a simple d-c amplifier based on a

single Ul83 FET. The maximum no-load output ( 0.4 volt) of the
solar cell ( International Rectifier SlM, or equivalent) is amplified

97
to 1.6 volts by this arrangement. A particular advantage of a cir-
cuit such as this is the high-resistance load which the FET input
offers to the cell. On rapidly changing light signals and those
modulated at high frequencies, this input avoids the high damping
caused by the low resistances into which such silicon cells usually
operate.
The output half of the circuit is a resistance bridge whose four
arms are Rl, the internal drain-to-source resistance of the FET, and
the two halves of potentiometer R2. With the solar cell completely
darkened, R2 is set to balance the static voltage from the D-C
OUTPUT terminals ( a d-c vtvm temporarily connected to the
output terminals will serve as a balance indicator). At the balance
point, the indicator will read zero, and the bridge should then
remain balanced indefinitely without further attention.
Subsequent illumination of the cell will unbalance the bridge
because the d-c output voltage of the cell changes the drain-to-
source resistance of the FET. This will cause 1.6 volts to appear at
the D-C OUTPUT terminals. ( This level of output voltage will
maintain itself across a 20,000-ohm load.)
Wiring is uncritical. Resistor Rl is 1/2 watt, and potentiometer
R2 is 1 watt. The fourth pigtail of the U183 is internally connected
to the metal case of this FET and should be grounded as shown.
Total current drawn from the 22.5-volt source ( M 1) is 24 ma.

PHOTO-FET RELAY
The photo-FET ( photoelectric field-effect transistor) has several
advantages over photodiodes and bipolar phototransistors. It com-
bines, for example, the functions of self-generating photojunction
and high-impedance-input amplifier. And it provides high ampli-
fication at low noise level ( 0.5-db typical noise).
Fig. 6-5 shows the circuit of a light-sensitive relay employing a
P-102 photo-FET. Note that this is a P-type field-effect transistor.
The P-102 has a built-in lens, which greatly simplifies application.
In this arrangement, the dark current in the photo-FET drain
circuit is approximately 300 µa. When the FET is illuminated at
only 10 foot-candles, the drain current increases to approximately
1.6 ma, more than enough to close the 2500-ohm, 1.4-ma d-c relay
( Sigma 5F, or equivalent). If desired, a delayed dropout may be
obtained by means of a capacitor in parallel with the relay coil.
98
With the 2500-ohm coil shown, the hold-in is approximately 2.75
seconds per 1000 mfd of parallel capacitance. Current drain on the
1.5-volt gate-bias source, Ml, is less than 0.1 microampere.

REVIY
2500Q
l.4ma
D
0

® lOmeg
s
i TO
CONTROLLED

@A -----ON-OFF ____ ----------------- @s

l CIRCUIT
0

@11-1,zv 1
@)-;;-w

Fig, 6-5. Photo-FET relay.

Wiring is straightforward, and resistor Rl is 1/2 watt. The gate

electrode of the P-102 is internally connected to the metal case of
this FET, so the latter must be kept clear of contact with wiring,
metal chassis, or other components.

TIMER
Resistance-capacitance-controlled timers have been built with
vacuum tubes, thyratron tubes, bipolar transistors of all types, and
with no tubes or other active devices whatever. The bipolar-transis-
tor circuits, while offering the advantage of compactness, small size,
light weight, freedom from the power line, and zero heat generation,
have been troublesome to design because the low-impedance tran-
sistors load the r-c timing circuit and impose undesirable limitations
on it. The high input impedance of the field-effect transistor, how-
ever, allows vacuum-tube performance to be attained in solid-state
electronic timers.
Fig. 6-6 shows the circuit of a timer employing a single Ul83
FET. This is essentially a bridge-balanced d-c amplifier, with a
relay in its output circuit, which receives its driving signal from a
charged capacitor discharging through an adjustable resistor. It
provides a hold-in of the relay that ranges from 1 second to 80
seconds, as selected by adjustment of the resistance.
99
 The timing interval is determined by the number of seconds taken
by 100-mfd capacitor Cl, after being fully charged, to discharge
through rheostat Rl to 0.4 of its charged voltage. This capacitor is
charged by momentarily depressing push-button switch SL This
connects Cl to a 1.5-volt tap on resistor R2. When released, the
push-button blade returns to its normal position against the upper
contact, and this connects the charged capacitor to rheostat Rl and
the gate of the FET.
The 1000-ohm, 1-ma d-c relay (Sigma 5F, or equivalent) is con-
nected in a resistance bridge whose four arms are R3, the internal
drain-to-source resistance of the FET, and the two halves of po-
tentiometer R4. With capacitor Cl completely discharged, R4 is

@
500
® G RELAY
10000
I Ima
PUSHBUTTON

TO
® + IOOmfd
lmeg
TIMING
CONTROLl£D
CIRCUIT

@ IK WIREWOUND
BALANCE

@ ®
'-----~l•i=-1*----✓
W ON--OFF
!Oma
Fig. 6-6. Timer.

set to balance the static FET current out of the relay ( at balance,
the relay opens firmly), and the bridge then should remain balanced
indefinitely without further attention.
To use the timer, depress Sl, charging capacitor Cl. Then re-
lease this switch, whereupon the relay will pick up immediately,
since the negative voltage presented to the FET gate by the
charged capacitor unbalances the bridge, causing current through
the relay. The relay will hold in until the capacitor has discharged
sufficiently ( through rheostat Rl) to reduce the gate voltage to
0.4 of its value at full charge of the capacitor. This time interval

100
depends on the resistance setting of Rl, and the rheostat accord-
ingly may be calibrated to read directly in seconds. This may be
done with a stopwatch, or the time ( t in seconds) may be calculated
approximately from the resistance setting ( Rl in ohms), the ca-
pacitance ( Cl in farads), and a multiplier ( 0.8 for the relay speci-
fied here). Thus:

t = 0.0001 ( 0.8) R 1 = 0.00008R 1 (sec)

With a I-megohm volume control having an audio taper for Rl,

time intervals from 1 to 80 seconds may be read on a single dial
scale.
Various combinations of capacitance and variable resistance other
than the 100 mfd and 1 megohm indicated in Fig. 6-6 may be used.
In any instance, the electrolytic capacitor used for the purpose
should be new and have a good power factor. For close reading
or for coverage of narrow bands, several switched-in rheostats
may be used against a single capacitor.
Resistor RS is 1/2 watt; R2 and R4 may be wirewound rheostats
with slotted shafts for screwdriver adjustment. Alternatively, R2
may be a stick-type resistor with slider. Electrolytic capacitor Cl
is rated at 25 dcwv. The fourth pigtail of the Ul83 is internally
connected to the metal case of this FET and should be grounded
as shown.

FETAS VOLTAGE-VARIABLE RESISTOR

Since at any value of drain-to-source d-c voltage, the drain cur-
rent is a function of the applied gate-to-source bias voltage, the
instantaneous value of internal drain-to-source resistance depends
on the d-c gate bias. This provides a convenient means of varying
resistance with an available variable voltage; and since the cor-
rectly polarized gate bias in the FET results in virtually zero gate
current, the process is essentially a voltage-controlled one.
Fig. 6-7A shows the simple setup of a U222 (high-current) FET
as a voltage-controlled resistor. Here, the control voltage ( as gate
bias) is applied to terminals A and B. Looking into terminals C
and D, one see~ a resistor varied by the voltage on A and B. In
use, this "resistor" is connected in series with a d-c supply and load
device, the supply being polarized so that C is positive and D

101
negative. Some devices in which a voltage-variable resistor is
advantageous are voltage regulators, signal translators, voltage-
tuned frequency-sensitive r-c networks, voltage-tuned audio oscil-
lators, remote-control devices, temperature transducers, and remote
volume controls.
Fig. 6-7B shows performance of the resistor when the voltage
between C and D is 15 volts. Note that the resistance varies from

G
- A C + (A) Circuit.
D-C RESISTOR
CONTROL
VOLTAGE TERMINALS
+ B D -

10

,____VDS • 15vdc
lma=
8
7.5ma
...
<!)
lOma
~6 I
0
>
_, 18ma
§
z4 2lma
8
c.,
I
Q

2 26ma

0 30ma
100 lK 111(
RES ISTANCE (OHMS)
(B) Performance.

Fig. 6-7. FET 85 voltage-variable resistor.

102
500 ohms ( when the control voltage is zero) to 15,000 ohms ( when
the control voltage is - 8 volts). Other resistance ranges are ob-
tained when the C-D voltage is some value other than 15 volts.
The power-handling capability of the variable FET resistor cor-
responds to the rated power dissipation of the FET. ( The manu-
facturer gives 300 mw as the absolute maximum device dissipation
for the U222 at 25°C free-air temperature). Note that the corres-
ponding drain-current values are given along the curve in Fig. 6-7B.
When using any FET as a voltage-variable resistor, keep well
within the maximum safe drain-to-source voltage, drain-current,
and device-dissipation ratings supplied by the FET manufacturer.
The "resistor" may be calibrated by successively applying the vari-
ous intended values of bias voltage and noting the corresponding
drain current. The resistance is calculated at each point as VD/In,
where VD is the instantaneous drain-to-source voltage ( in volts)
and ID is the corresponding drain current (in amperes).

CONSTANT-CURRENT SOURCE
The pentodelike drain-voltage/drain-current characteristic curves
of the FET suit this component for use as a constant-current
adapter. Fig. 6-8 shows the circuit of such an adapter employing a
TIX33 FET.
@
+

D-50 D-C
MILLIMIMrnR

ADAPTER
TERMINALS
B -
CURRENT
ADJUST

Fig. 6-8. Constant-current source.

The adapter is connected in series with an external d-c power

supply and the device operated from the supply, with the supply
polarized correctly to make terminal A of the adapter positive and
B negative. Potentiometer Rl then is set for the desired load cur-
rent, as indicated by 0-50 d-c milliammeter M2. ( With the TIX33,
103
this may be any value between 1 and 40 ma.) That current level
then will be maintained automatically within 6 percent, even if the
voltage of the external supply varies between 3 and 20 volts.
Current drain from the 22.5-volt gate bias supply, Ml, is 45
microamperes, which is due entirely to current through the 0.5-
megohm potentiometer, Rl. Resistor R2 is 1/2 watt.

SENSITIVE A-C A-F RELAY

Using two FET's (Ql, Q2) and one silicon bipolar transistor
( Q3), the circuit given in Fig. 6-9 closes a I-milliampere, 5-volt
d-c relay (Kl) in response to an input signal of only 0.1 volt rms

Go
r-c~~ROLI..ED
~ c u 1r

Fig. 6-9. Sensitive a-c a-f relay.

( at approximately 0.1 microampere ), when gain control R4 is set

to maximum. Good response is obtained throughout the spectrum
from 50 Hz to 50 kHz.
The input signal is amplified by a two-stage r-c-coupled section
based on two FET's: 2N4868 (Ql) and 2N4340 (Q2). The ampli-
fied signal then is rectified by the 1N34A diodes ( Xl, X2), and the
resultant direct current is amplified by a 2N2712 plastic-encapsu-
lated silicon bipolar transistor ( Q3) and applied to the 1-ma, 5000-
ohm d-c relay ( Kl, Sigma 5F, or equivalent).
With zero input signal, the 2N2712 draws virtually no current,
and the total circuit drain is 1.8 ma at 9 vdc. With a 0.1-volt input
signal (R4 set for maximum gain), the circuit drain rises to 2.8 ma.

104
 All fixed resistors are 1/2 watt. Capacitors Cl, C3, C5, and C6
are 100-volt plastic units, and C2 and C4 are 25-dcwv electrolytics.
For stability, all wiring must be kept as short, direct, and rigid as
practicable, and the entire unit should be enclosed in a grounded
shield box if interfering fields are present. A fourth pigtail of the
2N4868 is internally connected to the metal shell of this FET and
must be grounded as shown, for shielding. In the 2N4340, the gate
electrode is internally connected to the metal case, so this FET
must be mounted clear of chassis, wiring, and other components.
Signal input leads to the A-C INPUT terminals should be shielded.

SOUND-OPERATED RELAY
Fig. 6-10 shows the circuit of a sound-operated relay based on
two 2N4868 FETs (Ql, Q2) and one 0183 FET (Q3). With this
arrangement, a microphone output signal of 1.8 millivolts rms
( -52.5 db) will close the 1-ma, 1000-ohm d-c relay (Kl) when
sensitivity control R4 is set to maximum.

®)
• 'NI
ON~~~-,,,-

R9

TO
CONTROU£D
(§ CIRCUIT
IKW.W,
ZERO SET

Fig. 6-10. Sound-operated relay.

The first two stages form a high-gain r-c-coupled amplifier, the

output of which is rectified by the 1N3605 silicon diode, Xl. The
resultant d-c output voltage of the diode, developed across load
resistor R7, is applied to the gate of the 0183 FET ( Q3) through
the resistance-capacitance filter R8-C6. The 0183 acts as a d-c

105
amplifier which drives a 1-ma, IOOO-ohm d-c relay ( Sigma 5F, or
equivalent). Current drain is 12 ma at 9 vdc.
Initial adjustment of the circuit consists of balancing the static
drain current of the Ul83 out of the relay:

1. Short-circuit the microphone terminals.

2. Set potentiometer R4 for maximum sensitivity.
3. Adjust potentiometer RIO until relay Kl opens.

( This adjustment may be made very precisely if a d-c vtvm is

connected across the relay coil and RlO is adjusted for zero de-
flection of the meter. )
All fixed resistors are 1/2 watt. Potentiometer RIO is wirewound.
Capacitors Cl, C3, C5, and C6 are 100-volt plastic units, and C2
and C4 are 25-dcwv electrolytics. The high-impedance input of the
first stage will accommodate a crystal or ceramic microphone di-
rectly; for a dynamic microphone, use an appropriate input trans-
former.
For stability and minimum stray signal pickup, all wiring must
be as short, direct, and rigid as practicable. For complete shielding
in electrically noisy locations, the entire unit must be enclosed in
a grounded metal box. In each of the FET's, a fourth pigtail is
internally connected to the metal case and must be grounded as
shown, for shielding.

- 'N
5.5ma 1
+

®) OtHlFF .,,,-

®
·~

A-F OUTPUT
@ lmeg @ 330

C2 + 50mfd

Fig. 6-11. Modulated light detector/amplifier.

106
 MODULATED-LIGHT DETECTOR/ AMPLIFIER
The P-102 photo-FET is a good detector of modulated light
( either chopped or smoothly modulated), and its a-f output is
easily amplified. Fig. 6-11 shows a circuit composed of a P-102
photo-FET (Ql) and a 2N3578 P-channel FET (Q2) as an a-f
amplifier.
The sensitivity of this circuit is such that 15 foot-candles of light
modulated or chopped at a 1000-Hz rate will produce 1 volt rms
(no load) at the A-F OUTPUT terminals when potentiometer R4
is set for maximum gain. Current drain is 5.5 ma at 9 vdc.
The detector/amplifier has many uses, such as light-beam com-
munication, alarm signaling, intrusion alarm service, sound-on-film
pickup, and optoelectronic control. The virtually zero loading of the
photo-FET by the amplifier results in maximum efficiency of the
P-102, while the amplifier produces a useful output-signal ampli-
tude for direct application or further amplification.
All fixed resistors are 1/2 watt. Capacitors Cl and C3 are 100-volt
plastic units, and C2 is a 25-dcwv electrolytic. In both the P-102
and the 2N3578, the gate electrode is internally connected to the
metal case of the FET, so each of these units must be mounted clear
of other components, chassis, and wiring. The P-102 must be posi-
tioned so that its nose-end lens faces the incoming light. All wiring
must be as short, rigid, and direct as practicable.

PHASE SHIFTER
The circuit given in Fig. 6-12, combining a triode phase inverter
and an adjustable r-c circuit, provides continuously variable phase
shift from O to 180 degrees while leaving the signal amplitude
virtually untouched.
This arrangement, based on one 2N3578 FET (Ql), has equal
fixed resistances ( R2 and R3, matched within 1 percent) in both
drain and source legs. When rheostat R4 is set to one end of its
range, the output is taken from the source of Q 1 and is in the same
phase as the input signal ( Ql is then acting as a source follower);
therefore, there is zero phase shift between output and input. But
when R4 is set to the opposite end of its range, the output is taken
from the drain of Q 1 and is of opposite phase from the input ( Q 1
then is acting as a conventional triode amplifier); therefore, there
107
is then 180 degrees phase shift between output and input. At in-
termediate settings of R4, the various phase angles between 0 and
180 degrees are obtained. Two identical stages may be cascaded for
0 to 360 degrees phase shift.
The circuit provides good, full-range operation between 150 Hz
and 5 kHz. Below 150 Hz and above 5 kHz, inherent phase shifts
in the circuit and components restrict the range to limits somewhat
higher than zero degrees on one end and somewhat lower than
180 degrees on the other. However, some improvement may be
obtained by careful attention to lead dress and shortness of leads,
and to minimizing stray capacitances.
For undistorted operation, the maximum input-signal amplitude
is 2 volts rms. With potentiometer Rl set for maximum gain the
corresponding output-signal amplitude is 0.75 volt rms. Current
drain is 1 ma at 9 vdc.

@ OUTPUT
----..-<SOOK
INPUT
® PHASE SHIFT

•CLOSELY MATCHED

Fig. 6-12. Phase shifter.

Resistors R2 and R3 are 1 watt and are of the 1 percent non-

inductive, wirewound type. Capacitor Cl is silvered mica.
All wiring must be as short, rigid, and direct as practicable and
kept well away from the metal chassis if one is used. The input and
output circuits must be kept well separated. In the 2N3578, the gate
electrode is internally connected to the metal case, so this FET
must be mounted clear of other components, wiring, and chassis.
The dial of the phase-shift control, R4, may be calibrated in

108
degrees by means of Lissajous figures. For this purpose, connect
an audio-frequency signal generator to the INPUT terminals and
to the vertical-amplifier input terminals of an oscilloscope having
excellent phase characteristics, and connect the OUTPUT termi-
nals of the phase shifter to the horizontal-amplifier input of the
oscilloscope.

109
 7

Instrument Circuits

Solid-state test equipment has long been prized for its complete
portability, electrically clean operation, and total isolation from
the power line. In all classes of such equipment, the bipolar transis-
tor, often in company with the semiconductor diode, brought about
many improvements. However, its use generally necessitated design
compromises. But the FET, being more nearly like the vacuum tube,
demands few compromises in the transistorizing of standard in-
strument circuits.
The electronic voltmeter is a clear example of successful tran-
sistorization via the field-effect transistor. Here, the FET has faith-
fully preserved the high input resistance ( 1 megohm or more per
volt) which is characteristic of the vacuum-tube voltmeter, whereas
transistorization via the bipolar transistor usul,llly afforded an input
resistance of only 100,000 ohms per volt. Another example is the
highly useful dip oscillator; in this instrument, the FET provides the
same high sensitivity and excellent selectivity that is characteristic
of the grid-dip oscillator.
This chapter describes several simple instruments which are made
efficient, reliable, and completely portable by the FET. For circuits
lll
in other chapters, which may be used for instrumentation, see Figs.
2-11, 2-12, 2-13, 3-1, 3-2, 3-3, 3-4, 3-7, 3-8, 3-9, 3-10, 4-7, 5-10, and
6-6.

SINGLE-FET ELECTRONIC D-C VOLTMETER

Fig. 7-1 shows a transistorized version of the extremely useful
d-c vacuum-tube voltmeter. This instrument, which ( like its vtvm
counterpart) has a constant input resistance of 11 megohms on all
ranges, employs a single 2N4340 FET. ( I megohm of the input
resistance is accounted for by isolating resistor RI in the probe.)

.-------o0.5V
@ 5meg
IV
- J!@) 4meg @)
I +
I
I
5V
I 2K
I
I @ 500K
10,,@ @)
I
I
I Imeg G
@ CALI BRAT ION
I
I @ 400K RANGE s 50K
I
I
50,, WI REWOUND
iSHIELDED
i CABLE
I
@ 50K • ®r-002
I
I
HXN
I +
I
I
I
I
® 40K
500,,
I 0-50 D-C
I
I
I
@ 5K MI CROAMMETER
l(JO(JI/
I®

II
lmeg @ SK
II
II
II GND ZERO-SET

PROBE
- @ @
11•
1 ✓
9V ON-OFF
2.Sma
Fig. 7-1. Electronic d-c voltmeter (single FET).

The instrument covers Oto 1000 volts in eight ranges: 0-0.5, 0-1,
0-5, 0-10, 0-50, 0-100, 0-500, and 0-1000 selected with single-pole,
8-position, nonshorting rotary switch SL The range resistors ( R2
to R9) in the input voltage-divider string must be accurate to at
least I percent. Some of these ( 400K and 4 meg) are not stock
values and must be made up by series-connecting available resistors
( SOOK + IOOK, and 3 meg + I meg, respectively). Linearity of
response is within +2 percent of full scale.

112
 Resistor RlO and capacitor Cl form a filter to remove any stray
fluctuations picked up by the test leads. Rheostat Rll is the
CALIBRATION control mounted safely inside the instrument case,
and potentiometer Rl3 is the front-panel ZERO-SET.
The 0-50-microampere scale of meter Ml reads direct on the 50-
volt range and is read on the 0.5-, 5-, 50-, and 500-volt ranges
by mentally shifting the decimal point, as required. On the 1-, 10-,
100-, and 1000-volt ranges, however, the readings must be mentally
multiplied by 2 or some multiple of 2 (by 0.02 for the 1-volt, by
0.2 for the 10-volt, by 2 for the 100-volt, and by 20 for the 1000-volt
range).
To calibrate the instrument:

1. Set switch Sl to the 1-volt range.

2. Set rheostat Rll approximately to the center of its range.
3. Plug the test leads into jack Jl and the GND terminal.
4. Close switch S2.
5. Adjust potentiometer Rl3 to zero the meter.
6. Connect an accurately known 1-volt d-c source to the input
( negative to the shielded test lead, positive to the GND termi-
nal).
7. Adjust rheostat Rll for exact full-scale deflection of meter Ml.
This rheostat then need not be touched again until the instru-
ment is routinely recalibrated.

A diode rectifier operated ahead of the input voltage divider will

adapt the instrument for a-c voltage measurements but will require
special meter scales and calibrations, in addition to the d-c ones.
All fixed resistors are 1 watt, and variable resistors Rll and Rl3
both are wirewound ( the former is provided with a slotted shaft
for screwdriver adjustment). Capacitor Cl may be mica or ceramic.
The 2N4340 must be kept clear of contact with the chassis or other
components, since the gate electrode is internally connected to the
metal case of this FET. Current drain from the 9-volt battery, \12,
is 2.5 ma.

BALANCED ELECTRONIC D-C VOLTMETER

As in vacuum-tube voltmeters, a symmetrical (balanced) FET
d-c voltmeter is superior to the single-ended circuit described in the

113
 0.5V
@ 5meg
"1 lV
~·
--1
~
I

t!l
;-
f-
If
11
11
11
11
1®®
"@
4meg

500<
5v® (§
RANGE lmeg G
D
I I• t ~ t -., ~BALANCE
11

@ 1 .001
& 11
11 10\/ @

HI
0 11
::, 11

f-""
f-""
i;·
i:i.
I,
11
11
11
11 50\/
1
"
CALIBRATION
0-50 D-C
,I>,. 11
<
0 i iSHIELDED. 50K MICROAMMETER
i i CABLE
f 11
100\/ @) lOK @ lOK

.
t'I>
it
c"
11
11
11
11
11 500\/ ZERO-SET
@~
9V
e. 11 l.2ma
Ii>
::, "@
RI
;
(')
1000\/
lmeg
11
II 5K
11 GND
PROBE
• MATCHED PAIR
preceding section. The balanced circuit exhibits less temperature
drift, and therefore requires fewer zero resettings than in the
simpler circuit. It also draws less current from the power supply
during normal operation.
Fig. 7-2 shows a balanced d-c voltmeter circuit employing two
2N4340 FET's. For best results, these FET's must be matched,
and may be so obtained from the manufacturer. The circuit is
essentially that of a resistance bridge, the four arms of which are
(1) the internal drain-to-source resistance of FET Ql, (2) the in-
ternal drain-to-source resistance of FET Q2, ( 3) Rll and that part
of Rl4 to the left of the slider, and ( 4) Rl3 and that part of Rl4
to the right of the slider. Adjustment of potentiometer Rl4 balances
the bridge and thus zeroes the meter. Application of a test voltage
to the gate of Q 1 ( through the input terminals of the instrument)
changes the internal resistance of Ql, unbalancing the bridge, and
causes the meter to deflect.
The instrument covers 0 to 1000 volts in eight ranges: 0-0.5, 0-1,
0-5, 0-10, 0-50, 0-100, 0-500, and 0-1000 volts selected with single-
pole, 8-position, nonshorting rotary switch SL The range resistors
( R2 to R9) must be accurate to at least 1 percent. Some of these
( 400K and 4 meg) are not stock values and must be made up by
series-connecting available resistors ( SOOK + lO0K, and 3 meg + 1
meg, respectively). Linearity of response is within + 2 percent of
full scale. The input resistance of the instrument is constant at ll
megohms on all ranges ( 1 megohm of this is accounted for by the
isolating resistor, Rl, in the test probe).
Resistor Rl0 and capacitor Cl form a filter to remove any stray
fluctuations picked up by the test leads. Rheostat Rl2 is the
CALIBRATION control mounted safely inside the instrument case,
and potentiometer Rl4 is the front-panel ZERO-SET. Potentiometer
Rl5 ( also mounted safely inside the case) is used to balance the
circuit by varying the grid bias on Q2 before the instrument is
placed into service.
The 0-50--microampere scale of meter Ml reads direct on the
0-50--volt range and is easily read on the 0.5-, 5-, 50-, and 500-volt
ranges by mentally shifting the decimal point, as required. On the
1-, 10-, 100-, and 1000-volt ranges, however, the reading must be
multiplied by 2 or some multiple of 2 (by 0.02 for the 1-volt, by 0.2
for the 10-volt, by 2 for the 100-volt, and by 20 for the 1000-volt
range).
ll5
 The circuit must be initially balanced in the following manner:

1. Unplug the test leads from jack Jl and GND.

2. Set switch Sl to its 0.5-volt range.
3. Close switch S2.
4. Adjust potentiometer Rl4 to zero the meter, changing the set-
ting of rheostat Rl2 if necessary to get a readable deflection
of the meter when unzeroed.
5. Temporarily ground the gate of FET Ql; this should drive the
meter from zero. Without touching Rl4, adjust Rl5 to rezero
the meter.
6. Remove the ground from the Ql gate; this should drive the
meter from zero. Without touching Rl5, reset Rl4 to rezero
the meter.
7. Again, ground the gate of Ql and note if the meter reads off
zero. If it does, repeat Step 5.
8. Continue to alternate between these steps until the zero set-
ting of the meter is undisturbed by grounding the gate of
Ql. Potentiometer Rl5 then should be protected from tamper-
ing.

After this initial balancing adjustment, the instmment may be

calibrated in the following manner:

1. Set switch Sl to the 1-volt range.

2. Set rheostat Rl2 to approximately the center of its range.
3. Plug the test leads into jack Jl and the GND terminal.
4. Close switch S2.
5. Adjust potentiometer Rl4 to zero the meter.
6. Connect an accurately known 1-volt d-c source to the input
( negative to the shielded test lead, positive to the GND termi-
nal).
7. Adjust rheostat Rl2 for exact full-scale deflection of meter Ml.
This rheostat then need not be touched again until the meter
is routinely recalibrated.

A diode rectifier may be operated ahead of the input voltage di-

vider to adapt the instrument to a-c voltage measurements. How-
ever, this will require special meter scales and calibrations, in addi-
tion to the d-c ones.

116
 All fixed resistors are 1 watt, variable resistors Rl2 and Rl4 both
are wirewound, and variable resistor Rl5 ( because of its high re-
sistance) is a composition type. Rl2 and Rl5 each should have a
slotted shaft for screwdriver adjustment. Capacitor Cl may be
mica or ceramic. The 2N4340's must be kept clear of contact with
each other or with the chassis, wiring, or other components, since
the gate electrode is internally connected to the metal case of this
FET. Current drain from the 9-volt battery, ~2, is 1.2 ma.

ELECTRONIC A-C VOLTMETER

Fig. 7-3 shows the circuit of a high-impedance electronic a-c volt-
meter suitable for audio and ultrasonic tests and measurements. This
circuit, which is based on a single 2N4868 FET, has an input im-
pedance of 1 megohm in series with 0.1 microfarad on all voltage
ranges.

® ON-OFF
.-----<r
®
----.!,l•I•
o-o
'N 0.4ma
@ @ CALIBRATIO;

1.0 1N34A SK t,;;-:;,.

WIREWOUNO ~

SrN
lrN

50V @
SK
@ 1N34A
0-50 D-C
!OrN
+ MI CROAMMETER

@ 4K
50rN
@ 500 @)
lOOrN
+
@ 500 @ Sllmfd

Fig. 7-3. Electronic a-c volbneter.

The instrument covers 0 to 1000 volts rms in eight ranges: 0-0.5,

0-1, 0-5, 0-10, 0-50, 0-100, 0-500, and 0-1000, selected with single-
pole, 8-position, nonshorting rotary switch SL The range resistors

117
( Rl to RB) must be accurate to at least 1 percent. Two of these
( 400K and 4 meg) are not stock values and must be made up by
series-connecting available resistors ( 300K + lOOK, and 3 meg + 1
meg, respectively). Linearity of response is within 5 percent of
full scale. Frequency response, referred to 1 kHz, is down 3.5 db at
50 Hz and 2 db at 50 kHz.
The circuit is the conventional amplifier-rectifier arrangement: the
signal is amplified first by the FET and then is rectified by the two
germanium diodes ( Xl, X2). The d-c output of the diodes is more
than sufficient to drive meter Ml to full scale when a signal voltage
of 0.5 volt rms is applied to the A-C INPUT terminals. The meter
deflection is proportional to the average value of the a-c signal
voltage, but the meter may be calibrated to read rms voltage on a
sine-wave basis.
No zero adjustment is needed. Rheostat Rl2, the CALIBRATION
control, is mounted safely inside the instrument case to protect it
from tampering after the instrument has been calibrated.
The 0-50-microampere scale of the meter reads direct on the
0-50-volt range, and is easily read on the 0.5-, 5-, 50-, and 500-volt
ranges by mentally shifting the decimal point, as required. On the
1-, 10-, 100-, and 1000-volt ranges, however, the reading must be
mentally multiplied by 2 or some multiple of 2 ( by 0.02 for the 1-
volt, by 0.2 for the 10-volt, by 2 for the 100-volt, and by 20 for the
1000-volt range).
Calibrate the instrument in the following manner:

1. Set switch Sl to the 5-volt range.

2. Set rheostat Rl2 to full resistance.
3. Close switch S2.
4. Connect an accurately known 5-volt rms a-c source to the A-C
INPUT terminals ( any frequency between 60 Hz and 1 kHz).
5. Adjust rheostat Rl2 for exact full-scale deflection of meter Ml.
This rheostat then need not be touched again until the instru-
ment is routinely recalibrated.

All fixed resistors are 1 watt. Capacitor Cl is a 1000-volt ceramic

unit (Sprague 5HK-Pl0, or equivalent), C2 a 25-dcwv electrolytic,
and C3 a 200-volt metallized paper tubular (Aerovox P8292ZN, or
equivalent). Rheostat Rl2 is a wirewound unit which has a slotted
shaft for screwdriver adjustment. The fourth pigtail of the 2N4868
118
 is internally connected to the metal case of this FET and must be
grounded as shown. Current drain from the 9-volt battery, M2, is
0.4 ma.

ELECTRONIC A-C VOLTMETER/MILLIVOLTMETER

For more sensitive a-f and ultrasonic voltage measurements than
are permitted by the circuit described in the preceding section, the
two-transistor circuit shown in Fig. 7-4 is recommended. The addi-
tional amplification available in this latter arrangement permits full-
scale deflections down to 0.1 volt ( 100 mv) rms.

f'
A-C
INPUT
@
@
@
HN

100,I
@ 0.5meg @ IN34A
@
0-50 o-c
@ + MICROAMMETER
1000,I @ @ 3.3K
+
@ @ 50mfd

Fig. 7-4. A-f voltmeter/millivoltmeter.

Based on one 2N4868 FET (Ql) and one 2N4340 FET (Q2),
the instrument covers O to 1000 volts rms in five ranges: 0-0.1, 0-1,
0-10, 0-100, and 0-1000, selected with the single-pole, 5-position,
nonshorting rotary switch S1. The range resistors ( Rl to R5) must
be accurate to at least 1 percent. One of these ( R5) is not a stock
value and must be made up by series-connecting one 100- and one
11-ohm resistor. (This odd resistance makes it possible to hold the
other four resistors to stock values. ) Frequency response, referred to
1 kHz, is down 4 db at 50 Hz and 3 db at 50 kHz. Linearity is
within 5 percent of full scale on each range.

119
 Like the simpler a-c voltmeter (Fig. 7-3 ), this instrument uses
the conventional amplifier-rectifier arrangement: the signal is ampli-
fied by the two FET's (Ql, Q2), then is rectified by two germanium
diodes (Xl, X2) whose d-c output drives microammeter Ml. De-
flection of the meter is proportional to the average value of the
a-c signal voltage, but the meter may be calibrated to read rms
voltage on a sine-wave basis.
No zero adjustment is needed. Rheostat Rll, the CALIBRATION
control, is mounted safely inside the instrument case to protect it
from tampering after the meter has been calibrated.
Unless special scales are drawn for the meter, voltages must be
determined from the 0-50-microampere scale by mentally multiply-
ing the deflection by some multiple of 2 (by 0.002 for the 0.1-volt,
by 0.02 for the 1 volt, by 0.2 for the 10-volt, by 2 for the 100-volt
and by 20 for the 1000-volt range).
Calibrate the instrument in the following manner:

1. Set switch Sl to its 10-volt range.

2. Set rheostat Rll to full resistance.
3. Close switch S2.
4. Connect an accurately known 10-volt rms a-c source to the
A-C INPUT terminals ( any frequency between 60 Hz and 1
kHz).
5. Adjust rheostat Rll for exact full-scale deflection of meter Ml.
The rheostat then need not be touched again until the meter
is routinely recalibrated.

All fixed resistors are 1 watt. Capacitor Cl is a 1000-volt ceramic

unit ( Sprague 5HK-Pl0, or equivalent), C3 is a 200-volt metallized
paper tubular, and C2 and C5 are 25-dcwv electrolytics. Rheostat
Rll is a wirewound unit which has a slotted shaft for screwdriver
adjustment. The fourth pigtail of the 2N4868 (Ql) is internally
connected to the metal case of this FET and must be grounded as
shown, but the gate electrode of the 2N4340 ( Q2) is internally
connected to the metal case of the latter FET, which accordingly
must be kept clear of contact with wiring, chassis, and other com-
ponents. Current drain from the 9-volt battery, M2, is 0.8 ma.
The terminal arrangement for the FET's is shown in Appendix
A. Manufacturers of FET's can be obtained from Appendix B, and
the addresses of the manufacturers are shown in Appendix C.
120
 @
.0001

G
@
@ 470K 1.7-2. 9 µh 19-4lµh

l(XXJ kHz
IOOkHz R-f
@A _FREQUENCY SELECTOR _______________ _ OUTPUT

!OOkHz !OOOkHz
@s ®
ON-0~

XTAL-1
OOOkHzl .001

Q16V
\51-=- 0.8ma

Fig. 7-5. Crystal dual-frequency standard.

CRYSTAL DUAL-FREQUENCY STANDARD

Fig. 7-5 shows the circuit of a dual crystal oscillator with output
selectable at either 100 or 1000 kHz. Both signals give strong har-
monics for calibration purposes far into the high-frequency spec-
trum. The circuit is based on a single 2N3823 FET.
A dpdt FREQUENCY SELECTOR switch ( SlA-SlB) connects
the 100-kHz crystal and the first tank ( Ll-C5) into the circuit for
100-kHz output, or the 1000-kHz crystal and the second tank ( L2-
C6) into the circuit for 1000-kHz output. The tank coils are slug
tuned for maximum output signal, as indicated by an a-c vtvm or
transistorized electronic a-c voltmeter connected to the appropriate
output terminals. (Ll is a Miller No. 42A223CBI, or equivalent; L2
is a Miller No. 42A335CBI, or equivalent.) Each crystal frequency
is adjustable over a narrow range by means of a 50-pf trimmer ca-
pacitor ( Cl, C2) to standardize the outputs against WWV trans-
missions or other standard-frequency signals.
All wiring must be kept as short, direct, and rigid as practicable.
Also, wiring should be kept well away from metal chassis or case,
to minimize stray capacitance which will bypass higher harmonics

121
of the signals. Capacitors C3, C4, and C7 are mica; C5 and C6 are
silvered mica. The fourth pigtail of the 2N3823 is internally con-
nected to the metal case of this FET and must be grounded as
shown. Current drain from the 6-volt battery, Ml, is 0.8 ma.

100-kHz/10-kHz SECONDARY FREQUENCY STANDARD

The secondary frequency standard circuit shown in Fig. 7-6 has
a 100-kHz oscillator and 10-kHz multivibrator. This arrangement
provides 100-kHz harmonics far into the high-frequency spectrum,
and each 100-kHz interval is subdivided by harmonics of the multi-
vibrator. The oscillator is based on a single 2N3823 FET ( Q 1) and
the multivibrator on two 2N4340 FET's (Q2, Q3).
The oscillator is tuned to the crystal frequency by adjusting the
slug-tuned coil, Ll ( Miller No. 42A223CBI, or equivalent) and the
crystal frequency is adjustable over a narrow range, by means of
50-pf trimmer capacitor Cl, to standardize the 100 kHz against
WWV transmissions or other standard-frequency signals. The 100-
kHz output is capacitance coupled through capacitor C3.
Synchronizing 100-kHz voltage is injected into the multivibrator
through potentiometer R5 which, with R4, forms the gate resistor
of FET Q2. The 10-kHz output is capacitance coupled through
capacitor C6.
The multivibrator is best adjusted by connecting the 100-kHz
output of the instrument to a communications receiver, tuning in
two adjacent 100-kHz harmonics on the receiver, closing switch Sl,
and adjusting potentiometer R5 until exactly nine intermediate
( 10-kHz) signals can be counted between each two adjacent 100-
kHz harmonics. Adjustment of R5 will cause the intermediate points
to jump abruptly from one number to another as the multivibrator
synchronizes at 7, 8, 9, 10, or 11 kHz. When the 10-kHz harmonics
are not wanted, switch Sl may be opened, and only the 100-kHz
harmonics will remain.
The IO-kHz fundamental is separately available, through capaci-
tor C6, as a highly accurate audio frequency, as a square wave, and
as a synchronizing signal.
All wiring must be kept as short, direct, and rigid as practicable.
Also, wiring should be kept well away from metal chassis or case,
to minimize stray capacitance which will bypass higher harmonics
of the signals. All fixed resistors are 1/2 watt, and all fixed capacitors

122
 IOO-kH z OUTPUT ID-kHz OUTPUT
l'!j
qa·
~
I
!')
....
g (rii)
11 ®t«D~
@T.oooI
@
~® MIIITll/1::TnD ® ~~ M?

~
'....0
~ G " @.L @ IS~ I I '-/IS
N .001 1. 7-1
~ "'

~r !
(I)
I') 2.9ph

l @ 470K @.L ®
~ TOOl ~ @ 25K IO-kHz
ON-OFF
XTAL _
;f (IOOdizl
g Kc1 J:;::.s~
!
g
Q.,

~ l 1 -~ µ·"'
C2 J:OOl j l j j@
~~F
®
are silvered mica. Potentiometer R5 is wirewound and provided
with a slotted shaft for screwdriver adjustment; it should be
mounted safely inside the instrument case to protect it from tamper-
ing. The fourth pigtail of the 2N3823 is internally connected to the
metal case of this FET and must be grounded as shown. The gate
electrode of the 2N4340 is internally connected to the metal case in
the latter FET, however, so both 2N4340's must be kept clear of
contact with each other or with wiring, chassis, or other components.
Total current drain from the 6-volt battery, Ml, is 2.2 ma.

DIRECT-READING AUDIO-FREQUENCY METER

A direct-reading audio-frequency meter is a convenient instru-
ment. Feed an unknown audio frequency into its input terminals
and read the frequency directly from the meter scale. No adjust-
ments of any kind need be made; only the frequency range must be
switched in the same way that a voltmeter range is switched.
Fig. 7-7 shows the circuit of such an instrument, based on one
2N4868 FET (Ql) and one 2N4340 FET (Q2). This instrument
covers 0 to 100 kHz in four ranges: 0-100 Hz, 0-1 kHz, 0-10 kHz,
and 0-100 kHz. The frequency is indicated on the scale of microam-

CALIBRATION CONTROLS

A.
RANGES
OtolOOHz
@ @
B. oto 1 kHz f--:o A
C. Oto 10 kHz @0.1
D. Oto 100 kHz f--<, B
2K @)"°1

----I
f:ool°C RANGE
@ D ,,-----7

r-. @
s
®
.0001

ON--OFF@
@A

l
SIGNAL 1N34A
INPUT
@ lmeg@ 5K @_ 9V
@ -uma
50mfd

Fig. 7 -7. Direct-reading audio-frequency meter.

124
meter Ml, and the indication is independent of signal amplitude
from 1.7 volts rms upward and is independent of waveform over a
wide range. The response is linear; hence, only one point need be
calibrated in each frequency range.
The circuit exploits the same principle employed in some tube-
type frequency meters: It is essentially two overdriven amplifier
stages in cascade. The output of the last stage accordingly is a
square wave which is applied to an r-c circuit ( R6-R9 and C4-C7)
and diode rectifiers Xl and X2. Since the square wave is limited
( i.e., of constant amplitude), the deflection of meter Ml depends
only on the number of pulses passing through this meter during
each second, so is directly proportional to the pulse frequency.
The frequency meter must be calibrated at one point in each of
its four ranges ( the best point is the top frequency in each range).
Rheostats R6, R7, RS, and R9 are the CALIBRATION controls,
and these units are provided with slotted shafts for a screwdriver.
They are also mounted safely inside the instrument case to protect
them from tampering. Follow this calibration procedure:

1. Close switch Sl.

2. Set range switch S2 to position A.
3. Connect an accurate audio oscillator to the SIGNAL INPUT
terminals and set its output to 2 to 3 volts rms.
4. Set the signal frequency to 100 Hz and adjust R6 for full-
scale deflection of meter Ml.
5. Set S2 to position B.
6. Set the signal frequency to 1 kHz and adjust R7 for full-scale
deflection of the meter.
7. Set S2 to position C.
8. Set the signal frequency to 10 kHz and adjust RS for full-scale
deflection of the meter.
9. Set S2 to position D.
10. Set the signal frequency to 100 kHz and adjust R9 for full-scale
deflection of the meter.

All wiring must be kept as short, direct, and rigid as practicable.

All fixed resistors are 1/2 watt. Capacitors Cl and CS are 400-volt
metallized paper tubulars, and C2 is a 25-dcwv electrolytic. Ca-
pacitors C4 and C5 are 200-volt metallized paper tubulars, and C6
and C7 are silvered mica. While the range capacitors ( C4 to C7)

125
should be reasonably accurate, any deviation in capacitance will be
compensated by adjustment of the calibration controls ( R6 to R9).
The fourth pigtail of the 2N4868 is internally connected to the
metal case of this FET and must be grounded as shown. In the
2N4340, however, the metal case is internally connected to the
gate electrode, so this latter FET must be kept clear of contact
with wiring, other components, or chassis. Current drain from the
9-volt battery, M2, is 1.4 ma.

4.5-MHz SOUND-MARKER GENERATOR

Fig. 7-8 shows the circuit of a 4.5-MHz crystal oscillator for
sound-channel alignment in tv receivers. This oscillator, based on
a single 2N3823 FET, is tuned to the crystal frequency by adjusting

XTAL
G ,!1
@
19-41 µh

4.5MHz R-F OUTPUT

@ 470K

@ 2K + 6V

@ @
.001 • 001

Fig. 7-8. Television sound-marker generator (4.5-MHz).

the slug-tuned tank coil, Ll ( Miller No. 42A335CBI, or equivalent)

for maximum output, as indicated by a communications receiver
or an r-f voltmeter connected to the R-F OUTPUT terminals.
The open-circuit signal amplitude at the R-F OUTPUT terminals
is 4.0 volts rms when the d-c supply is 6 volts at 1.4 ma. The signal
strength may be continuously adjusted by making C4 a 50-pf midget
variable capacitor.
Resistors Rl and R2 are 1/2 watt. Capacitors Cl, C2, and C4 are
mica, and capacitor CS is silvered mica. The fourth pigtail of the
2N3823 is internally connected to the metal case of this FET and
must be grounded as shown.
126
 RANGES
A. 20TO 200Hz @ @ @ @
B. 200 TO 20<Xliz 20 2meg 200K 20K @
C. 2T020cHz meg 30K
D. 20TO 200<Hz A
C
@A' B
'\
I
I
iRANGE
I

I G

'
A' ,,
I
I
I
B C D

FEEDBACK
@
20meg
@@
200K 2(1( @) 4K
ADJUST

Fig. 7-9. Variable-frequency audio oscillator.

VARIABLE-FREQUENCY AUDIO OSCILLATOR

Fig. 7-9 shows the circuit of a variable-frequency r-c-tuned audio
oscillator employing two 2N4340 FET's. Using the Wien-bridge
frequency-determining network, this instrument covers 20 Hz to 200
kHz in four ranges: 20-200 Hz, 200-2000 Hz, 2-20 kHz, and 20-200
kHz. Tuning in each range is accomplished with a dual 365-pf
variable capacitor (C2), and ranges are changed by switching
identical resistors in pairs (Rl-R5, R2-R6, R3-R7, and R4-R8). These
resistances must be accurate to at least 1 percent.
In this circuit there are two feedback paths, both through ca-
pacitor C4. One is through the bandswitched Wien bridge to the
gate of FET Ql, the other is through FEEDBACK ADJUST rheo-
stat Rll to the source of Ql. At the resonant frequency of the Wien
bridge, the feedback to the Ql gate is positive and creates oscilla-
tion. The feedback through Rll, however, is negative and, since it
cancels the oscillation on all frequencies except the bridge resonant
frequency, is responsible for the very low distortion of this type of
circuit.
The amplitude of the negative feedback is adjustable with rheo-
stat Rll, and the amplitude of the driving voltage presented to the
feedback/ output amplifier ( Q2) is adjustable with potentiometer
127
Rl2. Optimum adjustment of these two controls results in lowest
distortion of the output signal.
The low-impedance, source-follower output is adjustable with
potentiometer Rl4 from approximately zero to 0.4 volt rms. Current
drain is 0.8 ma from the 22.5-volt battery, Ml. Higher output ampli-
tude ( at the same current drain for the oscillator) may be obtained
with an additional output-amplifier stage.
The tuning capacitor ( C2) must be insulated from the metal
chassis and should be enclosed in a grounded shield box. The lower
section of this capacitor is grounded, but the upper section floats
and must be compensated with the 50-pf trimmer capacitor, Cl.
Solid construction must be employed throughout, and all wiring
must be short, direct, and rigid.
The oscillator may be calibrated in the conventional manner
against any available standard audio frequency, using an oscillo-
scope set up for Lissajous figures. Another method makes use of a
calibrated audio-frequency meter ( see an earlier section of this
chapter) connected to the A-F OUTPUT terminals. With switch Sl
set to the proper range, capacitor C2 is adjusted to resonance with
the standard frequency ( or to the desired frequency, as indicated
by the audio-frequency meter). The dial of this tuning capacitor
then is set to that frequency. Trimmer capacitor Cl is adjusted
for continuous oscillation throughout the range.
Rheostat Rll is adjusted for constant output over the entire range
of the oscillator. as indicated by an electronic a-c voltmeter or
oscilloscope connected to the A-F OUTPUT terminals. Potentiom-
eter Rl2 is adjusted for minimum distortion ( as indicated by an
oscilloscope or distortion meter connected to the A-F OUTPUT
terminals). After adjustment, these two variable resistors need not
be touched again until the instrument is routinely rechecked.
All fixed resistors are 1 watt. Capacitors C4 and C5 are 50-dcwv
electrolytics, and capacitor C3 is a 200-volt metallized paper tubu-
lar. Rheostat Rll and potentiometer Rl2 have slotted shafts for
screwdriver adjustment and are mounted safely inside the instru-
ment case to protect them from tampering. Trimmer capacitor Cl
is also inside the case, being securely fastened to the frame of
tuning capacitor C2. The gate electrode of the 2N4340 is internally
connected to the metal case of this FET, so both 2N4340's must be
kept clear of contact with each other, other components, wiring, or
chassis.
128
 DIP OSCILLATOR
The tube-type grid-dip oscillator is a well-known and useful
instrument which needs no theoretical explanation nor a recitation
of its merits here. Previously, some transistorized versions of this
device, while offering freedom from the power line, have been less
satisfying than the tube type, since the bipolar transistor reduced
sensitivity and selectivity of the circuit. With an FET, however, the
dip oscillator behaves just as it does with a tube, because the gate-
to-ground impedance of the FET is comparable to the grid-to-
ground impedance of the tube. Response is sharp and fast, and the
indicating meter need not be connected in some special part of the
circuit different from the grid circuit of the tube.

@ ©
IK
PWG-IN COIL
lSEE TABLE 7-U

0-50 D-C
+ MI CROAIMIETER

Fig. 7-10. Dip oscillator.

Fig. 7-10 shows the circuit of a dip oscillator, based on a 2N3823

FET. With six plug-in coils, this instrument covers the range 1.1
to 95 MHz: (A) 1.1-2.5 MHz, (B) 2.5-5 MHz, (C) 5-11 MHz, (D)
10-25 MHz, ( E) 20-45 MHz, and ( F) 40-95 MHz. The circuit is
that of a Colpitts oscillator tuned with a dual 100-pf variable
capacitor, Cl.
Dip is indicated by the 0-50 d-c microammeter, Ml, in the gate-
to-ground leg, and the meter sensitivity is adjustable with rheostat
Rl. Downward deflection is sharp when the tank (Ll-Cl) is tuned
to the resonant frequency of an external 1-c circuit to which the
tank is loosely coupled.
129
 Like other dip oscillators, this one is useful for checking signals
as well as dead 1-c circuits. A closed-circuit jack, Jl, is provided
for high-resistance magnetic headphones used for this purpose.
,vhen the headphones are plugged in, the microammeter is auto-
matically disconnected from the circuit.
The instrument may be calibrated in either of two ways: ( 1)
Loosely couple Ll to a succession of dead 1-c tank circuits whose
frequencies are accurately known, tune Cl successively for dip at
the various frequencies, and mark the dial of this variable capacitor
correspondingly, or ( 2) With headphones plugged into jack Jl,
tune in a number of accurately known signals, adjusting Cl for
zero beat at each frequency and marking the dial of the tuning
capacitor correspondingly.
For stability, the instrument must be solidly constructed, and the
shortest, most direct, and most rigid wiring practicable must be
used. All coils are tightly wound on I-inch-diameter plug-in forms
(see Table 7-1 for coil-winding data). The number of turns in
the coils may need some adjustment to compensate for stray ca-
pacitance in an individual instrument and for the particular gate-
to-source capacitance of an individual FET. Resistor R2 is 1/2
watt, and capacitors C2, CS, and C4 are silvered mica. The fourth
pigtail of the 2N3823 is internally connected to the metal case of
this FET and must be grounded as shown. Current drain, from the
6-volt battery, M2, is 5 ma.
The steady deflection of the microammeter does not have the
same value on all frequency ranges and may even change to a
higher or lower value as the instrument is tuned through a single

Table 7-1. Coil-Winding Data 0

Coil Frequency Range Specifications

A 1.1-2.5 MHz 150 turns No. 32 enameled wire, closewound

B 2.5-5 MHz 77 turns No. 28 enameled wire, closewound
C 5-11 MHz 35 turns No. 22 enameled wire, closewound
D 10-25 MHz 17 turns No. 22 enameled wire, spaced to
winding length of 1 inch
E 20-45 MHz 8.5 turns No. 22 enameled wire, spaced to
winding length of 1 inch
F 40-95 MHz 41,i! turns No. 22 enameled wire, spaced to
winding length of 1 inch
0 All coils wound on Millen No. 45004 1-inch, 4-pin coil forms, or equivalent (use two
of the pins ) .

130
range. Adjustment of rheostat Rl, however, enables the operator to
set the pointer of Ml to a desirable point on the scale before tuning
for dip. This rheostat also provides some degree of volume control
when the instrument is used with headphones.

A-F MILLIAMMETER
Fig. 7-11 shows the circuit of an audio-frequency milliammeter,
based on a single 2N4868 FET ( Q1), which covers the current
range 0 to 500 ma in seven full-scale ranges: 0.5, 1, 5, 10, 50, 100,
and 500 milliamperes. A 2-pole, 7-position, rotary selector switch
( Sl) switches in the appropriate values of shunt resistance ( Rl to
R7). The double-pole switching guarantees that the circuit is never

RANGE
,------------------, ®
J +

9V
0.4ma1
ON-OFF.
-

@A/ @s/
® 500 0.5ma@
Ima I
meg
@ 400
5ma
@ 50
A-F !Oma
INPUT @ 40
50ma
® IN34A
0-50
@5
o-c
Ml CROAMMETER
!OOma
@4
500ma
® 5K
.@
@ 1.0
50mfd

Fig. 7-11. A-f milliammeter.

exposed to excessive current between switch positions. Linearity of

response is within 5 percent of full scale. Frequency response, re-
ferred to 1 kHz, is down 3.5 db at 50 Hz and 2 db at 50 kHz. Cur-
rent drain is 0.4 ma at 9 vdc.
Input-resistance values are: 0.5 ma, 1000 ohms; 1 ma, 500 ohms;
5 ma, 100 ohms; 10 ma, 50 ohms; 50 ma, 10 ohms; 100 ma, 5 ohms;
and 500 ma, 1 ohm.
Like the a-f millivoltmeters described earlier in this chapter, this
circuit uses the conventional amplifier-rectifier arrangement: the
131
signal is developed by current through the selected shunt resistors
and is first amplified by the FET and then rectified by the two
germanium diodes ( Xl, X2). The d-c output of the diodes is more
than sufficient to drive microammeter Ml to full scale. The meter
deflection is proportional to the average value of the input a-f
current, but the meter may be calibrated to read rms current on a
sine-wave basis.
No zero adjustment is needed. Rheostat Rll, the CALIBRATION
control, is mounted safely inside the instrument case to protect it
from being turned by accident after the instrument has been com-
pletely calibrated.
The Oto 50 d-c microampere scale of meter Ml reads directly on
the 50-ma range, and is easily read on the 0.5-, 5-, and 500-ma
ranges by mentally shifting the decimal point as required. On the
1-, 10-, and 100-ma ranges, however, the reading must be mentally
multiplied by 2 or some multiple of 2 ( thus, by 0.02 for the 1-ma,
by 0.2 for the 10-ma, or by 2 for the 100-ma range of the a-f
milliammeter).
Calibrate the instrument in the following manner:

1. Set switch Sl to the 50-ma range.

2. Set rheostat Rll to full resistance.
3. Close switch S2.
4. Connect an accurately known 50-ma a-c source to the A-F
INPUT terminals ( any frequency between 60 Hz and 1 kHz).
5. Adjust rheostat Rll for exact full-scale deflection of meter Ml.
This rheostat then need not be touched again until the instru-
ment is routinely recalibrated.

The instrument is now in calibration on all ranges if accurate values

have been used for resistors Rl through R7.
All fixed resistors are 1 watt. ( The range resistors-RI to R7-
must be accurate to at least 1 percent.) Capacitor Cl is a 25-dcwv
electrolytic; C2 is a 100-volt plastic unit. Rheostat Rll is a wire-
wound unit provided with a slotted shaft for screwdriver adjust-
ment.
For stability, all wiring must be as short, rigid, and direct as
practicable; and the fourth pigtail of the 2N4868, which is internal-
ly connected to the metal case of this FET, must be grounded as
shown, for shielding.
132
 INDUCTANCE CHECKER
Any inductance between 60 µ,h and 60,000 hy may be checked
by the resonance method, using the circuit shown in Fig. 7-12 with
a variable-frequency audio oscillator. With the unknown inductance
connected to terminals X-X, the oscillator is tuned for peak deflec-
tion of meter Ml, whereupon the inductance is determined ( in
terms of the .01-mfd capacitance, CS) by calculation or by reference
to the graph in Fig. 7-13.
In this circuit ( Fig. 7-12), the U 183 FET ( Q 1 ) serves as a source
follower between the oscillator and the measuring circuit. The ad-
vantages of this arrangement are (a) the high input impedance of
the resultant circuit reduces oscillator loading to virtually zero, and
( b) the follower isolates the measuring circuit ( Ml-Xl-C3-Lx) from
the signal source.
Potentiometer Rl allows adjustment of the gain for exact full-
scale resonant deflection of meter Ml. This is an added convenience
even when the audio oscillator has its own output control. With Rl
set for maximum gain, an oscillator output of 1.5 volts rms will
drive meter Ml to full scale with most inductors which will be
connected to terminals X-X. D-c microammeter Ml is made to
respond to alternating current simply by connecting the 1N34A
germanium diode ( Xl) in parallel with it. Current drain of the
circuit is 3.3 ma at 9 vdc.
Resistor R2 is 1/2 watt. Capacitor Cl is a 100-volt plastic unit;
C2 is a 25-dcwv electrolytic. For stability and accuracy, CS must be
a silvered mica capacitor as close as practicable to the rated .01 mfd,
since it is with reference to this capacitance that the unknown in-
ductance is determined.
The checker must be built mechanically solid, and all wiring must
be as short, rigid, and direct as possible. A fourth pigtail of the Ul83
is internally connected to the metal case of this FET and must be
grounded as shown, for shielding.
To use the checker:

1. Connect a variable-frequency audio oscillator ( tuning range:

20 to 200,000 Hz) to the OSCILLATOR INPUT terminals.
2. Connect the unknown inductance ( Lx) to terminals X-X.
3. Close switch Sl, and switch on the oscillator.
133
OSCILLATOR INPUT @ 100 TO UNKNOWN INDUCTANCE
X (Lxl
1N34A

Fig. 7-12. Inductance checker.

INDUCTANCE (HENRYS)

0.01 1 100 10, --

\
\

'\ ~

\
0.001 0.1 10 1000

0.0001 0.01 1 100

\ ..

0.00001 0.001 0.1 10

10 100 200
100 1000 2000 FREQUENCY
1000 !OK 20K } (Hz)

!OK lOOK 200K

Fig. 7-13. Inductance chart.

134
 4. Starting at the botom of the lowest frequency range, tune the
oscillator slowly throughout its ranges until a sharp deflection
of meter Ml occurs. Adjust the oscillator output control and
gain-control potentiometer Rl to bring this deflection close to
full scale of the meter.
5. At this point of peak deflection, read the frequency from the
oscillator dial.
6. For close results, calculate the inductance:

1
L, = -(-3-95-f~2 -X-l~0--9 ~)

where,
Lx is the unknown inductance, in henrys,
f is the oscillator frequency, in hertz.
7. For approximate results, eliminating calculations, use the chart
given in Fig. 7-13.

SUBSTITUTION-TYPE CAPACITANCE METER

Small capacitances are accurately measured by means of the sub-
stitution process, since this method of measurement automatically
compensates for stray capacitances in the test setup. It is ad-
vantageous to measure small capacitances (0.1 to 400 pf) at radio
frequencies. Fig. 7-14 shows the circuit of a substitution-type ca-
pacitance checker operated at 1 MHz.

@ MAIN ADJUSTMENT

X
TO UNKNOWN
@ CAPACITANCE
. 0001 X

@
® ON-OFF

XTAL

1MHz
@~,,,
t -_ _ _....,_ _ _ ._
001_ _
12ma
_ _ _ _ _ _ _ _ _ _ __,MICROAMMETER
0-50 D-C

Fig. 7-14. Substitution-type capacitance checker.

135
 This arrangement employs a 1-MHz crystal oscillator based on a
2N3823 FET (Ql). The fix-tuned oscillator tank (Ll-C2) is link-
coupled to the resonant measuring circuit, which is composed of
inductor L4, trimmer C4, variable capacitor C5, and unknown ca-
pacitance Cx connected to terminals X-X. A simple r-f voltmeter
( consisting of germanium diode Xl, rheostat RS, microammeter
Ml, and r-f bypass capacitor C6) is connected across the measuring
circuit as a resonance indicator. Both C4 and C5 are provided with
calibrated dials reading directly in picofarads. The oscillator coil
( Ll) and the measuring-circuit coil ( L4) are slug-tuned units
(Ll, adjusted to 250 µ,h, Miller No. 20A224RBI; L4, adjusted to 60
µ,h, Miller No. 41A685CBI). L2 and LS are link-coupling coils which
must be provided by the builder; each consists of 2 turns of insu-
lated hookup wire closewound near the bottom ends of Ll and L4,
respectively.
All fixed resistors are 1/2 watt. Capacitors Cl, CS, and C6 are
mica; C2 is silvered mica. All wiring must be short, direct, and
rigid. Moreover, the wiring in the measuring circuit ( i.e., between
L4, C4, C5, and terminals X-X) must be straight, solid No. 14 bus.
A fourth pigtail of the 2N3823 is internally connected to the metal
case of this FET and must be grounded as shown, for shielding.
Total current drain is 12 ma at 15 vdc.
After the assembly has been completed and the wiring verified,
the crystal oscillator must be initially adjusted:

1. Connect an r-f vtvm between ground and the top of LL

2. Close switch SL
3. Adjust the Ll slug for peak deflection of the vtvm. This slug
need not be touched again except during routine recalibration
of the instrument.

Next, initially adjust the measuring circuit:

L Set variable capacitor C4 to half capacitance and C5 to maxi-

mum capacitance.
2. Set rheostat RS to maximum resistance.
3. Close switch Sl.
4. Adjust the L4 slug for peak deflection of meter Ml, reducing
the resistance of RS if necessary to ensure a readable deflec-
136
 tion. This slug need not be touched again except during rou-
tine recalibration of the instrument.

To use the checker:

1. Close switch Sl, and resonate the circuit by setting C5 to

maximum capacitance and trimming with C4 for peak deflec-
tion of meter ML Thenceforth, do not touch C4.
2. Record the capacitance setting of C5 as Ca,
3. Connect the unknown capacitance ( Cx) to terminals X-X by
the shortest possible leads. This will detune the measuring
circuit, and the meter deflection will either decrease or disap-
pear.
4. Retune C5 to restore resonance, as indicated by peak deflection
of Ml.
5. Record this new capacitance setting of C5 as Cb.
6. Calculate the unknown capacitance:

When long leads must be used to connect the unknown capacitor

to the instrument, install the leads first ( in the same position they
will take when the capacitor later is connected to them) and then
perform Steps 1 and 2. This automatically compensates for the lead
capacitance. Then connect the unknown capacitor to the far end
of the leads and perform Steps 3 to 6 in the same manner described
above.
When very small capacitances (0.1 to 10 pf) are measured, all
readings ( Ca and Cb) should be taken from the dial of C4, and
capacitor C5 should be used only to trim the circuit to resonance
in Step 1. This step then consists of the following procedure:

1. Set C4 to maximum capacitance. Close switch Sl, and tr.m

with C5 for peak deflection of meter Ml. Record this setting
of C4 as Ca, and thenceforth do not touch C5.

IMPEDANCE METER
Fig. 7-15 shows the circuit of an impedance meter which employs
a widely used principle of measurement: an a-f voltage is applied
137
to an unknown impedance ( Z) in series with a calibrated variable
resistor ( R), and the voltage drops ( Ez and Er) across Z and R,
respectively, are checked separately with an electronic voltmeter
while R is varied. When the voltage drops are equal ( i.e., Ez =
Er), Z = R and may be read directly from the calibrated dial of
the variable resistor.

MAIN
CONTROL
®
® @ IN34A l.Omfd @@ A

@ 1----o---:-B

"811
!OK W.W. .I
INPUT lmeg
@ IN34A
@.
®
ON-
@ 5K
.@@
lmeg
SENSITIVITY
OFF 50mld
MI CROAMMETER

X X +
9V
TO 0.4ma
UNKNOWN IMPEDANCE

Fig. 7-15. Impedance meter.

In Fig. 7-15, rheostat Rl is the calibrated variable resistor ( 0 to

10,000 ohms) which is in series with the unknown impedance con-
nected to terminals X-X. A signal voltage, derived from an external
audio oscillator, is applied to this series combination through input
transformer Tl. The electronic voltmeter is composed of an a-f
amplifier, based on a 2N4868 FET (Ql), followed by a 2-diode
rectifier ( Xl, X2) and a d-c microammeter (Ml). When switch S2
is in Position A, this voltmeter is connected across variable resistor
Rl and reads E .. ; when Sl is in Position B, the voltmeter is con-
nected across the unknown impedance and reads Ez. The meter
needs no voltage calibration, since it serves only to indicate refer-
ence points; the d-c scale of microammeter Ml may be used directly
for this purpose. (It is only necessary to adjust Rl so that the de-
flection is the same for each position of switch S2; the magnitude
of the deflection is inconsequential.) The input potentiometer ( R5)
of the voltmeter circuit allows the meter to be adjusted for deflection
near full scale to increase accuracy. Current drain of the circuit is
0.4 ma at 9 vdc.

138
 All fixed resistors are 1/2 watt. Rheostat Rl is wirewound and
must have a dial reading directly in ohms. Capacitors Cl and C3
are 100-volt plastic units; C2 is a 25-dcwv electrolytic. Transformer
Tl may be any convenient unit having a 1:1, 2:1, or 3:1 turns ratio,
and preferably should be shielded.
All wiring must be as short, rigid, and direct as practicable.
Overall shielding is desirable but not mandatory. A fourth pigtail
of the 2N4868 is internally connected to the metal case of this FET
and must be grounded as shown, for shielding.
Use this instrument in the following manner:

L Connect an audio oscillator to the OSCILLATOR INPUT

terminals ( common test frequencies are 400, 500, and 1000
Hz), switch it on, and set its output to a level which will not
saturate transformer TL
2. Connect the unknown impedance (Zx) to terminals X-X.
3. Set potentiometer R5 for maximum sensitivity.
4. Close switch SL
5. Throw switch S2 to Position A, and adjust rheostat Rl and
potentiometer R5 for a deflection near full scale of meter ML
Note the meter reading.
6. Throw switch S2 to Position B and note the meter reading.
7. Adjust Rl to make the deflection the same as it was in Step 5.
8. Throw switch S2 back to Position A, note the deflection and
adjust Rl again, as in Step 7.
9. Continue this procedure until there is no change in meter
reading as S2 is thrown back and forth between A and B.
10. At this point, read the value of the unknown impedance from
the ohms dial of rheostat RL

HETERODYNE FREQUENCY METER

By means of the zero-beat method radio frequencies far into the
spectrum may be spotted and measured with a calibrated low-
frequency r-f oscillator. This is made possible by zero beating be-
tween the unknown frequencies and harmonics of the calibrated
oscillator. When a mixer and calibrated oscillator are combined for
this purpose in one unit, the result is a heterodyne frequency meter.
Fig. 7-16 shows the circuit of a heterodyne frequency meter con-
sisting of a 1 to 2-MHz oscillator (Ql), an untuned mixer (Xl),
139
and an a-f (beat-note) amplifier ( Q2). The oscillator employs a
2N3823 FET; the mixer, a 1N34A germanium diode; and the a-f
amplifier, a TIS14 FET.
The oscillator is of the Hartley type and is continuously tunable
over the range of 1 to 2 :\1:Hz. The oscillator tank comprises coil Ll,
365-pf tuning capacitor C4, and 50-pf trimmer capacitor C3. Coil
Ll consists of 65 turns of No. 28 enameled wire closewound on a
I-inch-diameter form, with a tap taken at the 20th tum from the
ground end. L2 is a coupling coil consisting of 10 turns of No. 28
enameled wire closewound around the center of Ll and insulated
from the latter.

® 0

RFC 1 2.5 mh

@
TUNING

® ON-Off

I
+ 'N -
6.5ma

Fig. 7-16. Heterodyne frequency meter.

The oscillator output is coupled, through L2, into the diode mixer
simultaneously with the unknown r-f signal ( coupled through ca-
pacitor C5). A beat note between the two signals is generated in
the mixer and is transmitted to the a-f amplifier through coupling
capacitor C7. The latter also serves to block the diode d-c com-
ponent from the TIS14 gate. Magnetic headphones, plugged into
jack Jl, complete the TIS14 drain circuit and make the beat note
audible. Other beat-note indicators, such as a vtvm, oscilloscope,

140
magic-eye tube, or audio-frequency meter, may be used, provided
Jl is shunted by a 2000-ohm resistor to complete the TIS14 d-c
circuit.
The instrument draws 6.5 ma at 9 vdc. This value is constant for
zero signal and maximum signal.
All fixed resistors are 1/2 watt. Capacitors Cl, C5, and C6 are
mica; C2 is silvered mica. C7 is a 100-volt plastic unit, and C8 a
25-dcwv electrolytic.
The oscillator tuning unit ( Ll-C3-C4) must be stoutly built, and
the entire oscillator stage should be enclosed in a grounded shield
box. All wiring must be as short, direct, and rigid as practicable. A
fourth pigtail of both the 2N3823 and the TIS14 is internally con-
nected to the metal case of these FET's and must be grounded as
shown, for shielding.
Main tuning capacitor C4 must be provided with a dial calibrated
to read directly in frequency-1000 to 2000 kHz. One way to cali-
brate this dial is to feed an accurate, unmodulated r-f signal-
generator output into the R-F SIGNAL INPUT terminals, set C3 to
midcapacitance, set the generator successively to as many frequen-
cies as practicable between 1000 and 2000 kHz, and at each
frequency set C4 to zero beat ( as indicated by headphone signal)
and inscribe the C4 dial with the corresponding frequency. This
calibration then may be checked more precisely by replacing the
generator with a 100-kHz crystal oscillator, which will set up an
accurate spot frequency each 100 kHz across the C4 dial. To
standardize the oscillator in the future, it is only necessary to feed
in the 100-kHz oscillator, set the C4 dial to its 1000-kHz point, and
adjust trimmer C3 for zero beat. Since C3 needs no constant adjust-
ment, it should be mounted inside the instrument case, where it
will be safe from disturbance, and should be provided with a slotted
shaft for screwdriver adjustment.
On second harmonics, the 1 to 2-MHz C4 dial covers 2 to 4 MHz,
on twentieth harmonics 20 to 40 MHz, and so on. Similarly, on the
second "subharmonic" the dial covers 500 to 1000 kHz.

HARMONIC-DISTORTION METER
A harmonic-distortion meter is convenient for measuring the total
distortion of an audio amplifier, a component, or a network. A pure
sine-wave signal is applied to the input of the device under test and
141
 I
+ 12V -

®C· 2.2mal

--------------------r------------7
VOLTMETER Q
CALIBRATION l!9,
~

®fQ( W.W.

@.20meg ©
!:l
~
-;-t
I ®~@l ~A

.... @ @@

r·
~ ~c G
@)20meg I !meg

i ,i.
•-······-1 •B@ I
I
400 pf
@2meg B
'VOLTMETER
RANGE
8 @ @ZOQ( @s
51 A-F
lmeg C
a INPUT
SENSITIVITY
~ f @
@)
JQ(
Il)
®t-i%. lt~m
1\/Yy

I1 ~
1.0
mid
@~2K

l RANGES
@~2K

f
A. 20-200 Hz
. t
@~2K

I
B. 200-2000 Hz
C. 2-20 kHz
the distortion meter is connected to the output. The meter is tuned
to suppress the fundamental frequency of the test signal. The self-
contained electronic-voltmeter section of the instrument then is used
to sample the output-signal voltage before and after the suppression.
The first voltage contains the fundamental and harmonics, whereas
the second one contains only the harmonics ( distortion components
introduced by the device under test). The total distortion therefore
is the ratio of the second voltage to the first, and can be expressed
either as a decimal or as a percentage.
Fig. 7-17 shows the circuit of a complete distortion meter employ-
ing these principles. This instrument covers the frequency range of
20 Hz to 20 kHz in three bands: (A) 20-200 Hz, ( B) 200-2000 Hz,
and ( C) 2-20 kHz. Ranges are selected by double-pole, 3-position
switch S2. The filter circuit is composed of range resistors R7 to R12
and dual 400-pf tuning capacitor C3. Within the amplifier comprised
by FET's Ql, Q2, and Q3, this r-c network nulls sharply at the fre-
quency to which it is tuned, and allows signals at other frequencies
to pass relatively untouched in amplitude. The r-c network alone
tunes broadly, but its response is sharpened by overall negative
feedback around the amplifier through capacitor C4 and resistor
Rl3. The voltmeter section comprises FET Q4, input range selector
S4 and R16-R19, and rectifier-type microammeter ( Ml-Xl-X2).
Total current drain is 2.2 ma at 12 vdc.
The amplifier section is based on three 2N4340 FET's ( Ql, Q2,
Q3). This section is followed by an a-f voltmeter based on a single
2N4868 FET ( Q4). When switch S3 is in Position A, the voltmeter
checks the signal voltage before the filter; and when S3 is in Posi-
tion B, the meter checks the signal voltage after the filter.
All fixed resistors are 1 watt. Resistors R2, R3, R5, R6 to Rl2, and
R14 to R19 must be rated at 1 percent. In an individual layout, Rl3
may need adjustment for sharpest null response of the circuit. Ca-
pacitors Cl, C2, C4, C5, and C7 are 100-volt plastic units; C6 is a
25-dcwv electrolytic. Each section of the dual 400-pf tuning capaci-
tor ( C3) has a built-in trimmer for initial adjustment.
After the instrument has been completed and the wiring verified,
first the voltmeter section should be calibrated:

1. Set potentiometer Rl for maximum sensitivity.

2. Set switch S3 to Position A.

143
 3. Set switch S4 to its 5-volt position.
4. Connect an accurately known 5-volt (rms) source to the A-F
INPUT terminals.
5. Close switch Sl.
6. Adjust rheostat R23 for exact full-scale deflection of microam-
meter Ml.

The voltmeter now is calibrated on all ranges if resistors Rl6 to Rl9

are accurate. ( R23 need not be touched again except during routine
recalibration).
Next, the filter (tuner) should be adjusted:

1. Set potentiometer Rl for maximum sensitivity.

2. Set switch S2 to Position A.
3. Set switch S3 to Position B.
4. Set switch S4 to fts I-volt position.
5. Connect a high-grade audio oscillator to the A-F INPUT termi-
nals and set it to 200 Hz at I-volt rms output.
6. Set tuning capacitor C3 to minimum capacitance.
7. Close switch Sl.
8. By adjusting the trimmers on C3A and C3B, tune the circuit
for full dip (null) of the Ml deflection. ( Readjust Rl and S4,
as required, for readable meter response.)
9. Tune C3 successively to various other frequencies down to 20
Hz, inscribing the C3 dial accordingly. The filter now will
automatically be calibrated on the two other bands ( B and C)
if resistors R7 to Rl2 are accurate.

Use of the instrument is straightforward:

1. Connect a high-grade sinewave audio oscillator to the input of

the device under test.
2. Connect the distortion meter to the output of the device.
3. Set switch S2 to the appropriate frequency range.
4. Close switch Sl.
5. Set switch S3 to Position A.
6. With switch S4 in the appropriate voltage position, adjust po-
tentiometer Rl for full-scale deflection of meter Ml.
7. Record this voltage as E1,
144
 8. Set switch SS to Position B.
9. Tune capacitor CS for null, and set S4 to successively lower
ranges for accurate reading of the voltage at null.
10. Record the residual null voltage as E 2 •
11. Calculate the distortion percentage:

D % = 100 _§_
E1
Under certain conditions, the instrument may be direct reading in
distortion percentage, requiring no calculations. Thus, if E 1 is 1
volt, this point may be regarded as 100 percent. The distortion per-
centages at null then may be read directly; e.g .. 0.01 volt = 1 per-
cent, 0.1 volt = 10 percent, and so on.

A-F R-F SIGNAL TRACER

The high input impedance of the FET makes possible for the first
time a transistorized signal tracer which is fully competitive with
the tube-type tracer. Basically, the signal tracer is a high-gain a-f
amplifier which drives a visual signal-strength indicator and a
speaker, and to which may be attached either a simple test probe
for a-f input or a demodulator probe for a-m r-f input.
Fig. 7-18 shows the circuit of a signal tracer employing two field-
effect transistors ( Q 1, Q2 ) and three NPN bipolar transistors ( QS,
Q4, Q5). High sensitivity is provided by the front end, which con-
sists of an r-c-coupled audio amplifier based on one 2N4868 FET
(Ql) and one 2N4340 FET (Q2). The a-f probe or r-f probe, as
required, is plugged into this amplifier through concentric jack Jl.
The front-end amplifier drives meter Ml through germanium diodes
X2 and X3. This provides the visual indication of signal strength.
For aural indication of signal strength, a speaker is driven by the
power amplifier consisting of a driver stage (based on a 2Nl69A
transistor, QS) and a push-pull class-B output stage (based on two
2N169A transistors, Q4 and Q5). The latter stage delivers 100 milli-
watts to the speaker.
Switch Sl allows the speaker section to be turned off when the
silent, visual indication provided by the meter is sufficient. At 9
vdc, the signal tracer draws a zero-signal current of 11 ma whether
the speaker is on or off, and a maximum-signal current of 25 ma
when the speaker is on and 11 ma when it is off.
145
 WITH SPEAKER WITHOUT SPEAKER
SH/DCABLE
CONCENTRIC ZERO SIGNAL Hrna Hrna
MAX. SIGNAL 25rna Hrna

. . .,. 4'"'"
-j =1]~PLUG
~F

I -- -- - ---I ,t--- - 3 _j]

GROUND CLIP A-F PROBE
::l
l[Q @
...?>
-.,t
I

ON-1+
lOK
I-'
,I>,.
a:,
.t
,!,.,
tw
"'
fi" l.Omfd 1N34A
®
e.
.,i
:" INPUT
@
0-50 o-c
E
llcd 8f1
SPEAKER
® + MICROAMMETER @
/ @ lmeg
@ 27K
CON-
CENTRIC
JACK
@ 5K @
1N34A

@
2700
(§Lo B
 All fixed resistors are 1/2 watt. Capacitor Cl is a 500-volt ceramic
unit, while C2, C4, C6, and C7 are 100-volt plastic or paper units.
C3 and C5 are 25-dcwv electrolytics.
The layout is not critical; however, all wiring must be as short
and direct as possible, to minimize stray coupling and interlerence.
For shielding, the entire unit must be enclosed in a metal box.
The instrument needs no calibration; since response of the meter
section is linear, the readings of microammeter ~l may be taken
directly as indications of signal strength, as long as the setting of
gain-control potentiometer R5 is undisturbed during a measurement.

SENSITIVE LIGHT METER

The P-102 photo-FET is the basis of the sensitive light-meter
circuit shown in Fig. 7-19. In this instrument, full-scale deflection
of the O to 100 d-c microammeter (Ml) is obtained with only 5
foot-candles of illumination, when Rl is 10 megohms. Other ranges
may be obtained by changing the value of this resistor-the lower
the resistance of Rl, the higher the range.

@
lK

0-100 o-c
@ lOmeg Ml CROAMNCTER

@ lOK W.W.
ZERO SET
o ®Af-----~--Off
+
l.5V
------
---- @a
_
....
o------o+ 'N -
dark current: 1. 2ma
max. signal current: 2ma

Fig. 7-19. Sensitive light meter.

Since there is some static drain current through the P-102, po-
tentiometer R2 has been provided for setting the meter to zero. The
operation consists simply of darkening the P-102 and adjusting R2
for zero deflection of meter M 1.
147
 Current drain of the circuit is 1.2 ma at 9 vdc when the P-102 is
darkened, and 2 ma when the light input deflects the meter to full
scale.
Layout of the instrument is not critical. However, the P-102 has
a lens in its nose, so this photo-FET must be mounted so as to
point toward the light source. And since the gate electrode is in-
ternally connected to the case of the P-102, the FET must be
mounted clear of wiring, chassis, and other components. Both fixed
resistors are 1/2 watt, and potentiometer R2 is a wirewound unit.

GEIGER COUNTER
Fig. 7-20 shows the circuit of a simple Geiger counter based on
one 2N4338 FET (Ql) and one 2N4868 FET (Q2). Basically, this
instrument is a high-gain r-~oupled a-f amplifier followed by an
a-c meter circuit ( coupling capacitor C4, germanium diodes Xl and
X2, microammeter Ml, and high-capacitance meter capacitor C6).

f o~~ma Y
®i/' 00-0FF -,1,-
@
PHEADPHONE JACK

0-50
0-C
MICROAMNETER

IN34A

•DANGER! HIGH VOLTAGE

Fig. 7-20. Geiger counter.

A jack (Jl) also is provided for alternate use of headphones; and

since this is a closed-circuit jack, withdrawal of the phone plug
automatically switches the meter back into the circuit. Current
drain is 0.6 ma at 12 vdc.
148
 The amplifier receives its input signal from a Victoreen 1B85 G-M
tube (Vl) biased by 900 vdc (Less than 10 ma). This is a beta-
gamma particle counter. The type of self-contained supply selected
for this high voltage will depend on the expense to which the
builder can go. It can vary from three 300-volt photoflash batteries
( Eveready No. 493, or equivalent) connected in series to one of the
factory-built, miniature, dry-battery-operated high-voltage modules.
Between these limits are numerous build-it-yourself units operated
from flashlight cells ( 1.5 to 6 volts) of the transistor-oscillator,
buzzer, and push-button-and-capacitor types. The ON-OFF switch
in the high-voltage supply must be ganged with the instrument
switch, SL
All fixed resistors are 1/2 watt. Capacitor Cl is a 1000-volt cer-
amic unit; CS and C4 are 100-volt plastic, and C2, C5, and C6 are
25-dcwv electrolytics. All wiring must be short and direct. The gate
electrode is internally connected to the case of the 2N4338 (Ql), so
this FET must be mounted clear of wiring, chassis, and other com-
ponents. A fourth pigtail of the 2N4868 ( Q2) is internally connected
to the metal case of this FET and should be grounded as shown,
for shielding.
The count-rate range of the instrument may be determined by
exposing the 1B85 tube to various calibrated radioactive samples,
noting the corresponding deflections of meter Ml, and graduating
the scale of the meter accordingly.

149
 APPENDIX A

TerminaJ Guide for Diodes and Transistors

1N34A 1N295 1N36<r; 2Nl69A

v ~STRIPE • ---------~RIPES /TRIPE
E B C
~ ~ ~ <·. ·)

Bo Go Go
2N333 2N260! 2N2712 2N3S78
C 0 D

(CASE) (CASE)
E s s

2N3823 2N4118 2N4338

2N3819 G G G(CASE)

V
S GD
DOCASE DOCASE oC)
s s s

151
 2N4340 2N4868 P-102 C632

Do G(CASEI

s
G

•OCASE
s
(CAS~lo
D

s
s
,Q,
SIM TIS14 TIX33
G G
REDn

BLACK~
DOCASE •OCASE
s s

U222

.Q"
s

152
 APPENDIX B

Manufacturers of Semiconduaors

1N34A GE, Sylvania 2N4118 Siliconix

1N295 GE, Sylvania 2N4338 Siliconix
1N3605 GE, Sylvania 2N4340 Siliconix
2Nl69A GE 2N4868 Siliconix
2N333 GE, Raytheon, P-102 Siliconix
Texas Instruments C632 Crystalonics
2N2608 Siliconix SIM International Rectifier
2N2712 GE TIS14 Texas Instruments
2N3578 Siliconix TIX33 Texas Instruments
2N3819 Texas Instruments 0183 Siliconix
2N3823 Texas Instruments 0222 Siliconix

153
 APPENDIX C

Directory of Manufacturers

AEROVOX. Aerovox Corporation. New Bedford, Massachusetts

02741.
ALLIED RADIO. Allied Radio Corporation. 100 N. Western
Avenue, Chicago, Illinois 60680.
AMPEREX. Amperex Electronic Corporation. Providence Pike.
Slatersville, Rhode Island 02876.
CLEVITE. Clevite Corporation, Piezoelectric Division. 232 Forbes
Road. Bedford, Ohio 44014.
CRYSTALONICS. Crystalonics, Incorporated. 147 Sherman Street.
Cambridge, Massachusetts 02740.
EVEREADY. Union Carbide Corporation, Electronics Division.
270 Park Avenue. New York, N. Y. 10017.
GE. General Electric Company, Semiconductor Products Depart-
ment, Electronics Park, Syracuse, New York.
INTERNATIONAL RECTIFIER. International Rectifier Corpo-
ration. 233 Kansas Street, El Segundo, California 90245.
MILLEN. James Millen Manufacturing Co., Inc. 150 Exchange
Street. Malden, Massachusetts 02148.
155
MILLER. J. W. Miller Company. 5917 So. Main Street. Los
Angeles, California 90003.
RAYTHEON. Raytheon Co., Semiconductor Operation, 350 Ellis
Street. Mountain View, California.
SIGMA. Sigma Instruments, Inc. 170 Pearl Street. Braintree, Mas-
sachusetts 02185.
SILICONIX. Siliconix Incorporated. 1140 W. Evelyn Avenue.
Sunnyvale, California 94086.
SPRAGUE. Sprague Electric Company. 233 Marshall Street. North
Adams, Massachusetts 01247.
SYLVANIA. Sylvania Electric Products, Inc. Semiconductor Di-
vision. 100 Sylvan Road. Woburn, Massachusetts 01801.
TEXAS. Texas Instruments, Incorporated. P.O. Box 5012. Dallas,
Texas 75222.
TRIMM. Trimm, Incorporated. 400 Lake Street. Libertyville, Illi-
nois 60048.
U. S. SONICS. U. S. Sonics Corporation. 63 Rogers Street. Cam-
bridge, Massachusetts ()<2742.
UTC. United Transformer Corporation. 150 Varick Street. New
York, N. Y. 10013.
VICTOREEN. Victoreen Instrument Division. 10101 Woodland
Avenue. Cleveland, Ohio 44104.

156
Index

A source follower, 22-24

Age amplifier, 38-39
A-c a-f relay, 104-105
All-wave receiver, 63-64
r-f relay, 95-96 Amplified photocell, 97 -98
voltmeter, electronic, 117-118, 119- Amplifier, a-f
120 bandpass, 30-34
Action, FET, 14 r-c-coupled, 24-25
A-f amplifier transformer-coupled, 25-26
bandpass, 30-34 Amplifier
band-suppression, 34-35 audio age, 38-39
1-c-tuned, 30-31 audio squelch, 71-72
notch, 34-35 auxiliary, 37
r-c-coupled, 24-25 bandpass, 30-34
r-c-tuned, 32-34 band-suppression, 34-35
transformer-coupled, 25-26 bipolar-transistor, 26-27
A-f milliammeter, 131-132 d-c, 39-41
oscillator, 56-57 final, 78-79
1-c-tuned, 45-47 grounded-gate, 79-80
phase-shift, 47-48 headphone, 37-38
transformer-feedback, 44-45 i-f, 36-37
Wien-bridge, 48-50 modulated-light, 107
preamplifier, 20-22 push-pull, 78-79
relay, 104-105 r-f, 79-80
signal tracer, 145-147 video, 36

157
Audio age amplifier, 38-39 Depletion layer, 13
mixer, 30 Detector, diode, 90-91
oscillator, 127 -128 modulated-light, 107
squelch amplifier, 71-72 Diode detector, 90-91
Audio-frequency meter, 124-126 Dip oscillator, 129-131
Auxiliary amplifier, 37-38 Dissipation, 18
Distortion meter, harmonic, 141-145
B Doubler, push-push, 81-82
Drain-coupled multivibrator, 50-51
Background, FET, 11-12
Balanced modulator, 84-85
Bandpass a-f amplifier, 30-34 E
Band-suppression a-f amplifier, 34-35
Effect, field, 10-11
Beat-frequency oscillator (bfo), 66-67
Electronic a-c voltmeter, 117-118,
Bipolar, definition of, 15
119-120
Bipolar-transistor amplifier, 26-27
Breakdown voltage, 17 d-c voltmeter, 112-113, 114-117
Broadcast receiver, 59-61, 62 Exciter, four-band, 84

C F

Capacitance, input, 17 FET action, 14

reverse transfer, 17 background, 11-12
Checker, capacitance, 135-137 operation, 12-15
inductance, 133-135 performance, 15-16
Code-practice oscillator, 57-58 ratings, 17-18
Colpitts a-f oscillator, 56-57 structure, 12-15
Conductance, output, 17 terminal guide, 151
Constant-current source, 103-104 FET, symmetrical, 13
Converter, crystal-controlled, 65-66 tetrode, 18
Counter, Geiger, 148-149 Field effect, 10-11
Crystal frequency standard, 121, 122- Field-effect transistor; see FET
123 Final amplifier, 78-79
oscillator, 53-56 Forward transconductance, 16, 17
Crystal-controlled converter, 65-66 Four-band exciter, 84
transmitter, 74-75, 76, 77-79 Frequency meter
Current, dark, 98 audio, 124-126
Current source, constant, 103-104
heterdoyne, 139-141
Cutoff current, 17
Frequency
C-w monitor, 86-87
multiplier, 80-81
transmitter, 74-75, 76, 77-79
standard, crystal, 121-122
D
G
Dark current, 98
D-c amplifier, 39-41 Geiger counter, 148-149
relay, 94 Generator, sound-marker, 126
voltmeter, electronic, 112-113, 114- Grounded-gate amplifier, 79-80
117 Guide, terminal, 151

158
 H N
Harmonic-distortion meter, 141-145 NF; see noise figure
Headphone amplifier, 37-38 NFET, definition of, 13
Heterodyne frequency meter, 139- Noise figure, 18
141 Notch a-f amplifier, 34-35

I 0

I-f amplifier, 36-37 Operation, FET, 12-15

IGFET, 18 Oscillator, audio, 56-57, 127 -128
Impedance meter, 137-139 1-c tuned, 45-47
Inductance chart, 134 phase-shift, 47-48
checker, 133-135 transformer-feedback, 44-45
Input capacitance, 17 Wien-bridge, 48-50
Inverter, paraphase, 28-29 Oscillator
phase, 27-28 beat-frequency, 66-67
code-practice, 57-58
Colpitts, 56-57
L crystal, 53-56
Layer, depletion, 13 dip, 129-131
L-c-tuned oscillator, 45-47 Pierce, 55-56
Light meter, 147-148 r-f, 52-53
variable-frequency, 127-128
Output conductance, 17
M
'.\fanufacturers, 153-156 p
'.\farker generator, sound, 126 Paraphase phase inverter, 28-29
Meter, audio-frequency, 124-126 Performance, FET, 15-16
capacitance, 135-137 PFET, definition of, 13
distortion, 141-145 Phase
frequency, heterodyne, 139-141 inverter, 27-28
harmonic distortion, 141-145 paraphase, 28-29
impedance, 137-139 shifter, 107-109
light, 147-148 Phase-shift oscillator, 47-48
Milliammeter, a-f, 131-132 Phone monitor, 90-91
Millivoltmeter, a-c, 119-120 Photocell, amplified, 97-98
Mixer, audio, 30 Photoelectric field-effect transistor;
'.\fodulated-light amplifier, 107 see photo-FET
detector, 107 Photo-FET, 98-99
Modulating transmitters, 87-88 relay, 98-99
Modulator, balanced, 84-85 Pierce crystal oscillator, 55-56
Monitor, c-w, 86-87 Pinch-off voltage, 14-15, 17-18
phone, 90-91 Preamplifier, a-f, 20-22
MOS FET, 18 Push-pull final amplifier, 78-79
MOST, 18 Push-push doubler, 81-82
Multiplier, frequency, 80-81
Mulitivibrator, drain-coupled, 50-51 Q
source-coupled, 51-52 Q-multiplier, 67-68

159
 R T
Ratings, FET, 17-18 Terminal guide, FET, 151
R-c--controlled timer, 99-101 Tetrode FET, 18
R-c--coupled a-f amplifier, 24-25 Theory, field-effect, 10-11
Receiver, all-wave, 63-64 Timer, 99-101
broadcast, 59-61, 62 "TNT" transmitter, 88-90
regenerative, 62, 63-64 Touch-plate relay, 96-97
superheterodyne, 59-61 Tracer, signal, 145-147
Regenerative receiver, 62, 63-64 Transconductance, forward, 16, 17
Relay, a-c, 95-96 Transfer capacitance, reverse, 17
a-c a-f, 104-105 Transformer-coupled a-f amplifier, 25-
d-c, 94 26
photo-FET, 98-99
Transformer-feedback oscillator, 44-
r-f, 95-96
45
sound-operated, 105-107
touch plate, 96-97 Transmitter, crystal-controlled, 74-75,
Resistor, FET as variable, 101-103 76, 77-79
Reverse current, 17 single-control, 88-90
transfer capacitance, 17 "TNT," 88-90
R-f amplifier, 79-80 Transmitters, modulating, 87 -88
oscillator, 52-53
relay, 95-96 u
signal tracer, 145-147
Unipolar, definition of, 15
s
V
"Selectojet," 69-71
Shifter, phase, 107-109 Variable phase shifter, 107-109
Signal tracer, a-f r-f, 145-147 resistor, FET as, 101-103
Single-control transmitter, 88-90 Variable-frequency oscillator, 127-128
Sound-marker generator, 126 Vldeo amplifier, 36
Sound-operated relay, 105-107 Voltage, pinch-off, 14-15
Source, constant-cwrent, 103-104 Voltage-variable resistor, 101-103
Source follower, a-f, 22-24 Voltmeter, electronic, a-c, 117-118,
Source-coupled multivibrator, 51-52 119-120
Squelch amplifier, 71-72 d-c, 112-113, 114-117
Standard, frequency, 121, 122-123
Structure, FET, 12-15
Superheterodyne receiver, 59-61
w
Symmetrical FET, 13 Wien-bridge oscillator, 48-50

160