radio handbook - americanradiohistory.com€¦ · the transistor radio handbook 1st edition table...

THE TRANSISTOR

RADIO HANDBOOK

--"- Theory

Circuitry

- Equipment

at -

THE TRANSISTOR

RADIO HANDBOOK(First edition)

- Theory

- Circuitry

- Equipment

\c

by Donald L. Stoner and L. A. Earnshaw

Copyright, 1963, by Editors and Engineers, Ltd.

Published and distributed to the electronics trade by

EDITORS and ENGINEERS. Ltd. Summerland , CaliforniaDiaIsm: Electronic distributors, order from us. Bookstores, libraries, newsdealers order from Baker LTaylor, Hillside, N.J. Export (exc. Canada), order from H. M. Snyder Co., 440 Park Ave. So., N.Y. 16.

Other Outstanding Books from the Same Publisher(See Announcements at Back of Book)

THE RADIOTELEPHONE LICENSE MANUAL

THE SURPLUS RADIO CONVERSION MANUALS

THE WORLD'S RADIO TUBES ( RADIO TUBE VADE MECUM

THE WORLD'S EQUIVALENT TUBES ( EQUIVALENT TUBE VADE MECUM

THE WORLD'S TELEVISION TUBES ( TELEVISION TUBE VADE MECUM

THETRANSISTOR

RADIO HANDBOOK1st EDITION

TABLE OF CONTENTS

Chapter One - INSIDE SEMICONDUCTORS 91-1 What is Matter 9

1-2 Building Blocks of the Universe 101-3 Atomic Structures. 12

1-4 Crystal Lattice Structures 12

1-5 Impurities 13

1-6 Junctions 15

1-7 Diode Action . 17

1-8 Transistor Action 191-9 Transistor Construction 23

Chapter Two - AUDIO AMPLIFIERS 282-1 Circuit Configuration 282-2 Bias Considerations 292-3 Stabilization 302-4 Transistor Impedances 322-5 Interpretation of Transistor Data 332-6 Load Lines for Transistors 342-7 Amplifier Considerations 362-8 Miscellaneous Audio Circuits 392-9 Semiconductor Speech Amplifier 402-10 A Deluxe Audio Compressor. 432-11 The Mini -Amplifier -Modulator 462-12 The Mini -Amplifier #2 472-13 A 10 -Watt Amplifier/Modulator 482-14 A Kit 10 -Watt Modulator 502-15 A 5 -Watt Class A Amplifier 522-16 Sliding Bias Amplifiers 53

5

Chapter Three - R.F. CIRCUITS 553-1 Alpha Cutoff 553-2 Impedance Matching 563-3 Neutralization 573-4 RF and IF Amplifier Stability 583-5 Detection 593-6 The Mixer and the Converter 603-7 Automatic Gain Control. 623-8 Self -Excited Oscillators 633-9 Crystal Controlled Oscillators 663-10 Transmitting RF Amplifiers 683-11 Modulating RF Circuits 69

Chapter Four - RECEIVERS 714-1 The Superregenerative Detector 71

4-2 Simplex Crystal Set 734-3 Simplex Audio Power Pack 744-4 The TR One 764-5 The TR Two 774-6 The Solar Two 794-7 A Regenerative Receiver 794-8 The Super Three 81

4-8a Constructing IF Transformers 854-9 A Four Transistor Superheterodyne . 864-10 The Mobileer 894-11 The Product Detector 924-12 A Professional Communications Receiver 944-13 Crystal Filters for a Communications

Receiver 1034-14 A Mechanical Filter for the

Communications Receiver 1054-15 A Transistor All -Band Converter 1064-16 Autodyne Converters 111

4-17 A Transistorized Six -Meter Converter .. 1134-18 A Transistorized Two -Meter Converter . . 1154-19 A 220 Mc. Transistorized Converter 117

6

Chapter Five - RF POWER AMPLIFIERS 1205-1 RF Power Amplifiers . 1205-2 RF Oscillators 1235-3 Linear Amplifiers 125

5-4 Modulation. 1275-5 A Transistor Phasing Exciter 1285-6 A Transistor Filter Exciter 1365-7 40 Meter SSB Transceiver 1385-8 A VFO for AM, CW, or SSB 1435-9 A 3 -Watt C.B. or 10 -Meter Transmitter . 1465-10 A Two -Meter Transmitter 1475-11 A Potpourri of Circuits 149

Chapter Six - POWER SUPPLIES. 1566-1 A.C. to D.C. Conversion 1566-2 D.C. to D.C. Conversion 161

6-3 D.C. to A.C. Conversion 171

6-4 A Low Cost Power Converter 171

7

FOREWORD TO THE FIRST EDITION

In all probability this is the first transistor manual written by two authors intwo widely separated countries. Transistors are transistors the world over, andin each country they function the same.

Different environments engender different ideas, however. One cannot imaginean Eskimo inventing a plow, a Bolivian inventing a kayak, or an American in-venting a set of chopsticks. However, if one gave an American a set of chopsticksand said "Improve on these," no doubt he would.

Through the medium of amateur radio, the authors have discussed the sub-ject of transistors back and forth across the Pacific ocean. We sincerely hope thatthe synthesis of our ideas have produced worthwhile results. While we make noclaim to having invented a new kind of plow or a better kayak, we do feel that byadding small bowls to the ends of the chopsticks and calling them spoons, wehave improved on something. Bowls are not new, nor are chopsticks, but acombination of both is a very useful tool.

The authors particularly wish to thank the following individuals and companieswho have provided assistance and encouragement in this project.

Mr. Jules RubenMr. Stanley IssacsMr. Irving SeligmanMr. Bill CourtneyMr. E. P. KellyMr. Ed. KingMr. John VadjaMr. William WilsonMr. C. D. SimmondsMr. Frank O'BrienMr. John FischerMajor Gilbert

Allied RadioLafayette RadioIrving Electronics

J. IV Miller Coil Co.StancorAmperex SemiconductorsMotorola SemiconductorsInternational Rectifier Corp.Philco SemiconductorsPacific SemiconductorsRadio Corporation of AmericaTexas Instruments

8

CHAPTER ONE

Inside Semiconductors

The reader need not understand all themysteries pertaining to holes, covalencebonds, etcetera, to construct transistor cir-cuits and equipment. All that is needed isa good magazine article and a few clearphotos. However, this basic informationis essential either to troubleshoot equip-ment properly or to design circuits your-self. If you are content to follow the leader,omit this chapter.

Before delving into the operation of atransistor, it would be useful to review abit of high school chemistry regarding thenature of matter.

1.1 What is Matter?

That's simple. Matter is just abouteverything! Typical examples are the airwe breathe, the water we drink, and the foodwe eat. These examples also illustrate thethree basic forms of matter: gaseous, liquid,and solid. Matter may be found in theseforms as either elements or compounds. Anelement is defined as abasic structure whichcannot be separated into substances ofother kinds. Scientists have been able toisolate over 100 elements. Some of themore familiar ones are copper, aluminum,hydrogen, and oxygen. Lesser known ele-ments, which we shall study later, includingsilicon and germanium.

A compound, on the other hand, is apure substance which contains two or moreelements and has a constant composition.It usually exhibits properties different fromthe elements of which it is composed. Avery abundant compound is water. It iscomposed of two parts hydrogen and onepart oxygen and is known by chemicalshorthand as H20.

The Elements are thought to be com-Molecule posed of molecules. A molecule

is the smallest physical part ofan element or compound. If one dividesa piece of copper many times into the small-est part possible and still has copper, thissmall part would be a molecule.

The Scientists believe that moleculesAtom are constructed of various ar-

rangements of atoms. Themolecule may contain one or more likeatoms in an element or two or more differ-ent atoms in a compound. Figure 1.1-Ashows the relationship between the atomand the molecule. Two of the simplehydrogen atoms combine with a singleatom of oxygen to form one molecule ofthe compound water.

Since there are over 100 elements, it isreasonable to expect that there exist thesame number of atomic structures whichcorrespond to these elements. As will

10 Inside Semiconductors The Transistor

® =ONE MOLECULE Of WATER

TWO ATOMS OF HYDROGEN PLUS ONE OXYGEN A TOM . H2O

Figure 1.1-ATHE RELATIONSHIP BETWEEN ATOMS

AND MOLECULES

soon be shown, the atom is made up ofsubatomic particles. The same particles makeup all the atoms in the universe. Only themanner in which they are arranged varies,and it is the arrangement which gives eachatom its own characteristics. This may beeasier to comprehend if one compares theatom to a fingerprint. The fingerprint iscomposed of simple loops and whirls, butthey can be combined in an infinitenumberof ways, and no two fingerprints are alike!Criminologists are convinced that all thefingerprint types in the world have notbeendiscovered. Similarly, scientists assume thatthere are more elements which have not beenidentified.

1.2 Building Blocksof the Universe

For the purpose of illustration, let thereader equip himself with a more power-ful microscope than exists in the world.With the microscope, examine a piece ofaluminum. Aluminum is usually associatedwith chassis and panels. It is smooth and

MAG. 100 X

Figure 1.2-ACRYSTALLINE SURFACE OF ALUMINUM

0000001 INCH

MAG. 10,000,000 X

Figure 1.2-BVAGUE OUTLINE OF THE ATOMIC

STRUCTURE OF ALUMINUM

shiny, and may be drilled, punched andformed into angles. If, however, the sur-face is magnified 100 times, it becomesobvious that the aluminum is not smooth,but actually has a rough crystalline surfaceas shown in figure 1.2-A.

If the surface is magnified 10 milliontimes, one begins to see vague outlines ofroughly spherical shape. These tiny blobsare the atoms of aluminum, and may be"seen" in figure 1.2-B.

No one has seen an atom. However,with X-ray techniques and advanced mathe-matics, it is possible to make some ratheraccurate guesses about their appearance.Scientists believe the atom looks like thedrawing in figure 1.2-C. It consists of acentral core with one or more tiny particlesorbiting around it as planets orbit the sunin the solar system.

To simplify explanations and drawings,

MAG. 100,000,000 X

Figure 1.2-CSCIENTISTS BELIEVE THE ATOM LOOKS

LIKE THIS.

Radio Handbook Atoms 11

the atom is usually depicted as shown infigure 1.2-D; in this case it is the "chemicalschematic" of an atom of aluminum. No-tice that this atom of aluminum is morecomplex than the hydrogen atom shown infigure 1.1-A.

Composition Referring to figure 1.2-D, theof the Atom central core of the atom is

termed the nucleus. If it werepossible to examine the nucleus of the alum-inum atom under the imaginary micro-scope, one would see something similar tofigure 1.2-E. The nucleus is composed ofneutrons with no charge and protons which havea positive charge. These components makeup the heart of the atom. Note in figure1.2-D that the nucleus of the atom consistsof 13 protons and 14 neutrons.

Each of the tiny particles orbiting thenucleus is a negatively charged electron. Theelectrons whirl at tremendous speedsaround the nucleus in orbits or rings, asthey are more commonly called. Note fur-ther that the aluminum atom has two elec-trons in the first ring, eight in the second,and three in the outer ring, for a total of13 electrons.

Size It might be helpful to describethese particles in definite terms.The diameter of an electron is

believed to be 0.00000000000022 or22 x 10'4 inches and is about three timesthe size of the proton. Despite its smallersize, the mass of the proton is about 1,850times that of the electron. The comparisonis something like a cubic foot of lead to acubic foot of feathers.

1ST RING. 2 ELECTRONS

2ND RING, 6 ELECTRONS

NUCLEUS CONTAINING13 PROTONS AND 14NEUTRONS.

3RD RING, 3 ELECTRONS

Figure 1.2-DA SINGLE ATOM OF ALUMINUM

.000000000001INCH

MAG. 1.000,000.000, 000 X

Figure 1.2-ETHE NUCLEUS OF THE ATOM IS BE-

LIEVED TO LOOK LIKE THISIt contains both positive protons, and

neutrons which have no charge.

If you can comprehend a copper pennybeing enlarged to the size of the earth'sorbit around the sun, then you would beable to visualize the electrons in the copper.They would be roughly the size ofbaseballsand might be spaced approximately threemiles apart! The nucleus, thecentral "sun,"would be composed of a group of neutronsand protons about the size of walnuts!

All matter is composed of atoms. Bythe same token, all atoms are built up fromelectrons, neutrons, and protons. Onlythe way in which the parts are arranged andtheir quantities will differ. Thus, theseparticles may be considered the fundamentalbuilding blocks of the universe.

Charges The positive charge on thepro-ton and the negative charge onthe electron are exactly equal in

the aluminum atom discussed earlier. The13 protons equal the 13 electrons, and thisaluminum atom is said to be electricallybalanced.

These charges are believed to be thesmallest that exist, and therefore they areconsidered the fundamental unit of electricalcharge. However, this quantity is much toosmall to deal with, and the coulomb iscommonly used in its place. Onecoulombof electricity contains more than six million,million, million, or 6.28 x 1018 electrons!


1.3 Atomic Structure

The most important feature of the atomis the fickle outer ring of electrons. Theatom is somewhat similar to the commonhousehold onion in that it is composed ofmany rings. It is relatively easy to removepieces of the outer ring simply by peelingthem off. The inner rings do not fall off,for the outer rings hold them in place.

The inner rings of the atom are tightlybound to the nucleus, but the outer ringmay be stripped off. More commonly,subtle modifications are made to change itscharacteristics.

It is also known that the electrons in theouter ring may skip from one atom toanother in a random manner due to thermalagitation (heat). Also, electrons may beadded or removed from the outer ring.Electrons able to move in this manner areknown as valence electrons. Any matter witha large number ofvalence electrons is knownas a conductor, and matter which has fewloosely held electrons is termed an insulator.Materials referred to as semiconductors liesomewhere between these two extremes.Germanium and silicon are two excellentexamples of semiconductors. These twoelements are not very good conductors intheir pure states. When properly modifiedthey can be made to conduct electricity,much the same as a piece of wire.

The atomic schematics for germaniumand silicon are shown in figure 1.3-A and1.3-B, respectively. For simplicity, the neu-trons found in the nucleus are not shown.Since neutrons have no charge, they donot affect our discussion. Note also thesimilarity between the germanium and sili-

1ST RING- 2 ELECTRONS

2ND RING - 8 ELECTRONS

3RD RING -18 ELECTRONS

4TH RING - 4 ELECTRONS

Figure 1.3-AGERMANIUM ATOM (Ge)

IST RING- 2 ELECTRONS2ND RING- 8 ELECTRONS

3RD RING- 4 ELECTRONS

Figure 1.3-8SILICON ATOM (Si)

con atoms. Each atom has four valenceelectrons in the outer ring. Since the actionoccurring in the atom takes place in thevalence ring, the atomic schematic may befurther simplified by showing only the val-ence electrons, as in figure 1.3-C. Whenreduced to this form, both germanium andsilicon appear to be identical, and as amatter of fact they are equally useful inmanufacturing transistors.

The reader should understand that dif-ferent materials may contain a differentnumber of valence electrons. However, justbecause two atoms contain the same numberof valence electrons, they do not necessarilyconstitute the same material. Note that inaluminum the valence ring contains threeelectrons, as does gallium, which is alsoused in manufacturing semiconductor pro-ducts. Two other elements useful in pre-paring semiconductors are antimony andarsenic, each having five valence electrons.

1.4 Crystal LatticeStructures

Most substances when examined undera microscope will exhibit a crystalline struc-ture, even though the surface may appearsmooth to the eye. All crystalline sub -

Figure 1.3-CA SIMPLIFIED DRAWING SHOWING

ONLY THE VALENCE ELECTRONSThis could be either the germanium orsilicon atom since they both have four

valence electrons.

Radio Handbook Atomic Structures 13

stances have an identifying characteristic.The most common example would be theelusive snowflake, which is composed of aninfinite number of geometric patterns madeup of 60° angles. Common householdsalt will form tiny cubes, while some mater-ials form needles, rhomboids, or variationsof hexagonal or rectangular structures.

By scientific deduction, it appears thatthe rotation of the valence electrons in oneatom is closely coordinated to the electronmotion in adjacent atoms. Thecoordinatedrotation forms an electron pair bond or acovalence bond. This bond is countermandedby the repulsion of the positively chargednucleus and a state of equilibrium exists,forming a single crystal of the material.

The lattice structure of a single german-ium crystal is shown in figure 1.4-A. Forsimplicity, only one "layer" ofthe structureis shown. Although it is not possible toillustrate the fact with two dimensional draw-ing, each germanium atom is bonded byelectron pairs to four adjacent atoms (orwould be if an infinite number could beillustrated). This is due to the fact thateach atom contains four electrons in theouter or valence ring.

In the element copper, the valence elec-trons are not tightly bound as just des-cribed, but are free to move about. Thus,when subjected to an electric potential, thevalence electrons move in an orderly man-ner. Such material is said to be a goodconductor. On the other hand, material

> COVALENT BONOS

Figure 1.4-AA STABLE LATTICE STRUCTURE CON-

TAINING ONLY GERMANIUM ORSILICON ATOMS.

EXCESS ELECTRON

Figure 1.5-AA "DONOR" IMPURITY HAS BEEN

ADDED TO THE PURE GERMANIUMINTRODUCING AN EXCESS ELECTRON

like polystyrene is a poor conductor, andexamination shows that its valence electronsare very tightly bound together.

1.5 Impurities

Earlier it was stated that silicon and germ-anium could be made into conductors bymodifying their structure. These structureswere considered to be pure, that is, contain-ing nothing but germanium or siliconatoms. To be usable in semiconductors,germanium and silicon must be refinedto thepoint where only one impurity will befound in 10 billion parts of the pure mater-ial.

When this state of purity is achieved, acontrolled amount of impurities may be addedto modify the structure in a desired manner.Such impurities might be antimony, arsenic,aluminum, or the gallium mentioned earlier.

A typical example might be the puregermanium structure discussed in section1.4. Impurities "dope" the element byadding one part antimony (five valence elec-trons) to 10 million parts of germanium.An examination of the germanium structurenow will show an extra atom in the group(figure 1.5-A0, a single atom of antimony.A closer look will show that the atom isaccompanied by an excess electron. Re-member that germanium has four valenceelectrons, while antimony has five. Sincethere can be no more than four electronsin any of the valence rings, this excesselectron must drift through the structure


looking for a gap to occupy. Fortunatelyfor the transistor industry it never findsone! The excess electron may combinewith one of the valence rings, but it willalways drive out an electron, leaving oneextra electron in the structure for each ex-cess electron added.

The impurity that has donated the extraelectron is logically called a donor impurity.Arsenic could also be the donor since ithas one extra electron when compared togermanium or silicon. Although the dis-cussion has dealt with single structures anda single excess electron, the reader shouldbe aware that millions of these excess elec-trons are present in millions of lattices,even in a piece of germanium too small tobe seen with the naked eye.

Germanium to which a donor impurityhas been added is called N -type germanium,since it contains an excess of electrons whichcarry a negative charge.

Acceptors Just as we have added the anti-mony and arsenic impurities, itis possible to "contaminate"

pure germanium with aluminum or gal-lium. As mentioned earlier, these two im-

Figure 1.5-BAN "ACCEPTOR" IMPURITY IS SHOWN

ADDED TO PURE GERMANIUMValence electrons try to fill this space,

but a gap always remains in thestructure.

purities have only three valence electrons.What happens to the atomic structure whenthese impurities are combined with thegermanium or silicon? There is no longeran excess of electrons, but rather a lack ofthem. Thus, this impurity is able to acceptelectrons from the germanium, leaving abole in the valence rings. This type of im-purity is referred to as an acceptor impurity.

To illustrate what might happen in asingle crystal of germanium when "doped"with acceptor impurities, refer to figure1.5-B. The impurity atom might withdrawan electron from the adjoining germaniumatom in a futile effort to neutralize the

SEMICONDUCTORRECTIFIERS AND TRAN-SISTORS TAKE MANYSHAPES AND SIZES.

These are only a few ofthe package styles youwill find in electronicequipment. (Courtesy,

General ElectricCompany)

Radio Handbook Junctions 15

structure. However, this leaves a gap in theadjacent atom's valence ring which might befilled. Our incomplete atom is compelledto become whole again and it, in turn,withdraws an electron from another atom.This occurs continuously, and in a ran-dom manner. The hole is simply a gap ina valence ring, which has no electron to fillit. Since there is continuous agitation inthe structure (due to ever-present heat) thegap or hole is continuously changing posi-tion. As before, there are millions of theseimpurity atoms in a tiny speck of german-ium, and, consequently, there are also mil-lions of holes seeking an electron. Althoughthe holes are moving, they are evenly dis-tributed and never leave the confines of thegermanium. A piece of germanium dopedin this manner is called P -type germanium,since it lacks electrons and therefore mustcarry a positive charge.

An analogy which might help clarify thefunction of the hole is the classic garagefull of cars illustration. As shown in figure1.5-C, the 12 -car garage is nearly filled tocapacity with 11 automobiles. Naturally,the only empty space is at the rear of thegarage. When one drives up to store a car,the attendant moves the front car forwardinto the space. He then moves the secondcar up a space. Finally, after moving eachcar up a space, he makes way for the 12thcar. Although the attendant has only movedcars forward, in effect, he has moved the

C[21 C

CAR MOVEMENT

C C

V.

CCJSPACE "HOLE" MOVEMENT

C C

C CI

121

Figure 1.5-CCLASSIC EXAMPLE SHOWING MOVE-

MENT OF HOLES

THESE POWER DIODES WILL DO THESAME JOB AS A 5U4 RECTIFIER

But they will provide more output volt-age for an indefinitely long time.

(Courtesy, General Electric Company)

space (the hole) toward the entrance.

When a potential is applied to the P -typegermanium, the holes will move in a veryorderly fashion and in one direction, whichconstitutes an electric current. In N -typegermanium, the electrons move from thenegative to the positive source while in P -type material, the holes traverse from posi-tive to negative. The holes (P) or electrons(N) are termed the majority carriers since theymake up the majority of moving particles.

1.6 Junctions

The reader with a fair imaginationshould be able to visualize what happensinside a piece of impurity -doped german-ium or silicon, whether type N or type P.

Consider now what happens whenblocks of N -type and P -type materials are"sandwiched" together to from a P -N junc-tion (figure 1.6-A.) The situation pro-duced by joining of P and N type materialis somewhat complex. Remember thatwhen manufactured, P -type material is elec-


PROBABLY THE WORLD'S SMALLESTSEMICONDUCTOR DIODE

The "Tiny Tim." That's not a 16 -pennynail; it's a common household pin!(Courtesy, Pacific Semiconductors, Inc.)

trically neutral, as is also the N -type mater-ial. Because the atoms in each are sharingelectrons or holes, there is no actual surplusof electrons or holes. A surplus or a de-ficiency of electrons would mean that thegermanium (or silicon) was electricallycharged. When the two types of germaniumare placed together the reader might im-mediately assume that the excess electronsin the N -type material would join the holesin the P -type material. This is not the case.Although a few recombinations do occur, theseare caused by thermal agitation. The twosides actually set up a barrier, preventingtotal recombination. If an electron crossesfrom the N zone to the P zone, both elec-trically neutral pieces will now take on acharge. The N zone will have lost an elec-tron and the P zone will have gained anelectron. The P zone will thus becomenegative with respect to the N zone. Thus,the negative P zone will now repel the entryof further negative particles. Remember thatlike poles repel and unlike poles attract. Infact, a barrier is set up between the twojunctions, and it takes a potential from a

P -TYPE N -TYPE

<0 .0 O. a.0 0 0 0 0 0 ATOMS

0'0 0 0 I 0 O 0 0 o HOLESo

I0 0 0 I O 0 0 ELECTRONS

*(EXCESS)

DEPLETIONREGION

SHOWINGRECOMBINATIONS

Figure 1.6-ATWO "BLOCKS" OF GERMANIUM ARE

JOINED TO FORM A JUNCTION

battery to force reluctant electrons acrossthe barrier.

The physical width of the barrier is ra-ther indefinite and is usually measured interms of the voltage necessary to drive anelectron across the depletion region. Withno external potential applied, the barrierwidth may be as high as 0.6 volts. This iscalled the barrier potential.

ForwardBias

If electrons can be made totraverse the barrier with an ap-plied potential, it is reasonable

to assume that the P -N junction will con-duct a current.

To obtain forward bias, the positive term-inal of a battery is connected to the P -typegermanium and the negative terminal of thebattery is connected to the N -type german-ium. If one could examine the inside ofthese slabs of germanium (figure 1.6-B) itwould be seen that the holes are beingrepelled from the positive terminal and the

P -TYPE N -TYPE

0 ATOMSCS. OOC.*

O 0 0 Co O O O

0- O 000; 0 0 -0

o HOLES

ELECTRONSO O 0 00 "" O 0 O (EXCESS)

Cw, 0000 ELECTRONSFROM VAL-ENCE BONDS

Figure 1.6-BA GERMANIUM JUNCTION WITH FOR-

WARD -BIAS APPLIED

Radio Handbook Diode Action 17

electrons repelled from the negative term-inal. This electric pressure against the de-pletion region greatly reduces it, allowinglarge quantities of holes and electrons torecombine. For each recombination thatoccurs, an electron from the negative term-inal of the battery enters the material. Sim-ilarly, an electron breaks its bond and entersthe positive terminal of the battery. Atthe same time an electron from the negativeterminal enters the junction to repair thebond. The action is continuous. Recom-bination about the depletion region occursso long as the battery is across thejunction.Connected in this manner, an electric cur-rent flows in the external circuit.

Reverse An entirely different effect takesBias place when the battery is re-

versed. For a reverse bras con-dition, the positive battery terminal is con-ected to the N -type germanium and thenegative terminal is connected to the P-

type germanium.

Once again inspecting the interior ofthe germanium (figure 1.6-C), one findsthat the holes in the P -type germanium areattracted to the negative end ofthe structure,while the electrons are drawn to the posi-tive end. This has the effect of stretchingthe depletion region, which increases thepotential needed to overcome this barrier.In actual practice, the barrier width in-creases to the potential of the applied volt-age and conduction does not occur.

The reverse -biased junction may bethought of as having a very high resistance.

P -TYPE N -TYPE

..0 0: 0 0

0- 0 0 o 0 0 0 -o0 0 01 0 0 0

DEPLETIONREGION

Figure 1.6-CA GERMANIUM JUNCTION WITH RE-

VERSE -BIAS APPLIED

TYPE -P TYPE -N

LOAD

FILTER CAPACITORACTION

Figure 1.7-ASINGLE P -N JUNCTION USED AS A

RECTIFIER

Actually the resistance does not go to in-finity, for a small amount of recombinationdoes occur. This represents a tiny reverseleakage current caused by the minority currentcarriers. With a small voltage applied, theleakage current will be only a few micro-amperes. If excessive voltage is applied,the atomic stability of the lattice may be im-paired, which breaks down the junction.This effect does not damage the junctionunless excessive heat is generated due toheavy current flow.

1.7 Diode Action

From the preceding explanation thereader can see that the diode will conductwhen the applied potential is in theforwarddirection and (for all practical purposes)will not conduct when reverse biased.

If connected as shown in figure 1.7-A,the junction becomes a rectifier when an a.c.signal is applied. During the positive half-

cycle the depletion region is almost non-existent (once the barrier potential is over-come), and current flows virtually unop-posed. During the negative half -cycle theconditions are identical to reversing thebattery (reverse bias) as explained earlier insection 1.6. During the entire half -cyclethe depletion region expands and contractsin proportion to the appliedvoltage. Thus,during the entire negative half -cycle, thebarrier places a very high resistance in thecircuit. The voltage appearing across theload, then, is a series of positivehalf-cycles.These may be used to charge up an electro-lytic capacitor to supply nearly pure d.c. tothe load.


Figure 1.7-BTHE SYMBOL FOR SEMI -CONDUCTOR

DIODEIn the case of a junction rectifier thearrow corresponds to the anode which

is of type -P material. The bar is thecathode and is of the N -type. The

arrow points INTO the direction of

electron flow.

Silicon The P -N junction makes a veryPower effective rectifier in high voltageDiodes and high current supplies, par-ticularly when silicon is used. Silicon isable to perform its duties in the presenceof high temperatures and therefore is some-times preferable to germanium.

Silicon diodes are becoming extremelypopular with amateurs and experimentersand have virtually replaced selenium recti-fiers in all but the most inexpensive com-mercial equipment. The reason for theirpopularity is the depletion region or poten-tial barrier mentioned earlier. It requiressome 0.6 volts to cross over the barrier.Once this potential is exceeded, the diodeacts as a virtual short circuit.

Examine the power supply shown in fig-ure 1.7-A. Apply 150 volts a.c. and con-nect a 500 ma. (0.5 ampere)load to theoutput. As soon as the barrier is overcome( 0.6 volts ), current surges through the loadon the positive half -cycle. At the peak of150 volts, the load is drawing 500 ma.What about the diode losses which generate

REVERSE

POTENTIALBARRIER

FORWARD

Figure 1.7-CZEN ER DIODE VOLTAGE/CURRENT

CHARACTERISTICS

heat and limit the operation? Only 0.6volts is "lost" inside the diode, and at 500ma. one finds that only 0.3 watts is absorbed(and must be dissipated) by the diode.Actually, this is slightly utopian, for thediode resistance does not drop to zero.Even so, the losses are not likely to exceedone watt, which may easily be dissipatedwithout cooling fins.

The selenium rectifier, on the otherhand, has comparatively high losses. Thiscan be proved without an atomic explana-tion simply by examination. A typical 500ma. selenium rectifier might be three inchessquare, one inch thick, and contain five orsix plates with selenium oxide spread onthe cooling fins. Compare that area to thetiny 0.5 ampere diodes shown in the photo-graphs!

The P -N junction contained in a 0.5ampere silicon diode might be the size ofa pin head. The junction is securely fast-ened to the metal case (which would fit in-side a thimble) to provide all the heat dissi-pation necessary.

4.*

GERMANIUM USED IN THE MANU-FACTURE OF TRANSISTORS MUSTHAVE NO MORE THAN ONE PART

IN TEN BILLION OF IMPURITIES.The amount of impurity then intention-ally added to "dope" the pure ger-manium in order to form a N or Ptype is roughly comparable to twopeople in the population of the UnitedStates. (Courtesy, CBS Electronics)

Radio Handbook Transistor Action 19

Zener Earlier it was mentioned thatDiodes excessive reverse voltages could

break down the P -N junction.Certain diode types are cultured to exhibitthis effect at a very low reverse voltage. Whenthe breakdown voltage is exceeded, thelattice structure collapses and the electronsavalanche through the junction. Thus, nearlytotal recombination occurs.

Through careful manufacturing tech-niques it is possible to provide diodeswhich break down at discreet voltages, say3.9, 4.7, 5.6 and so on up to several hun-dred volts. Such a diode is known as azener diode, or more correctly as an avalanchediode. A diode which breaks down at someexact voltage (figure 1.7-C) makes an excel-lent regulating device.

Junction Although not particularly appli-Capacity cable to this discussion, an in-

teresting point is the capacityexhibited by the depletion region. As thereverse bias on the diode junction is in-creased, the depletion region expands. Theaction is similar to moving the plates of a

capacitor farther apart. This, of course,effectively reduces the junction capacitancewhich in turn reduces the terminal capaci-tance of the diode. By the same token,lower reverse voltages narrow the deple-tion region, thereby decreasing the capaci-tance. This variable capacitance effect may

TYPE- PEMITTER

TYPE -N TYPEBASE COLLECTOR

- P

0 0 0 0 0 0 0

0 0 0 0 0 0 00

\1tDEPLET ON

REGIONS

Figure 1.8-AP -N -P TRANSISTOR WITHOUT APPLIED

POTENTIALThe holes and electrons drift aimlessly

about.

THIS TINY VARIABLE CAPACITY DIODEWILL REPLACE THE CAPACITORS

SHOWNThese "Varacap" diodes will undoubt-edly be widely used in amateur equip-ment. (Courtesy, Pacific Semiconduc-

tors, Inc.)

be used to advantage by applying diodes toadjust the resonant frequency of tuned cir-cuits. Such a device is known as a variablecapacitance diode.

1.8 Transistor Action

The diode (or P -N junction) providesan excellent foundation of blocks uponwhich to build a transistor. The simplejunction transistor may be thought of as atriple -deck germanium sandwich.

When a block of P -type german-ium is sandwiched between twoblocks of N -type germanium,

two potential barriers develop ---one at eachjunction.

The block diagram of a P -N -P transistoris shown in figure 1.8-A. From left toright, the blocks are termed emitter, base, andcollector. The recombined electrons andholes in the depletion regions are not

TheP -N -P

Transistor


TYPE -PEMITTER

TYPE -NBASE

TYPE -PCOLLECTOR

Figure 1.8-BFORWARD BIAS

Forward bias of the emitter -base junc-tion allows holes to recombine in thecollector, creating a current flow in thereversed -bias base collector junction.

shown due to space limitations. For thepurpose of explanation, connect two bat-teries as shown in figure 1.8-B. Wired inthis manner, the P -N junction (emitter tobase)is forward biased, and recombinationoccurs. The holes that would normallyspeed off to the battery are instead driventhrough the P -N junction (base to collector),due to the greater attraction of the negativecollector for the positively charged holes.If the forward bias battery is reduced tozero, the emitter base barrier increases andreduces the amount of holes reaching thecollector. The action is quite similar tothat of the grid in a vacuum tube, whichcontrols the electron flow between cathodeand plate.

The The action of this transistor isN -P -N identical with the P -N -P tran-Transistor sistor except that the blockscontain the opposite type of impurities, andthe majority current carriers are electronsrather than holes. Since electron carriershave a negative charge, all the battery polar-ities in figure 1.8-B must be reversed.

Before proceeding, it might be wise toreview two points which are important to thefollowing discussion and which should bekept firmly in mind.

1. The emitter base junctions must beforward biased in any transistor, or the major-ity current carriers (holes P -N -P and elec-

trons N -P -N) cannot be propelled to thecollector region.

2. In either transistor type, the collectormust be reverse biased to prevent all the holes(P -N -P) or electrons (N -P -N) from goingdirectly to the collector. If the collectorbase junction were forward biased, thebar-rier would be greatly reduced, and the emit-ter to base junction could not control cur-rent flow.

Transistor At this point the reader mayGain relegate to memory the elusive

atoms, electrons, and holes todeal with the more tangible external effectscreated by the transistor. The illustration,figure 1.8-C, shows a more schematic -likeversion of figure 1.8-B. As before, theemitter base junction is forward biased andthe base collector junction is reverse biased.Recall that a forward biased junction be-haves relatively like a medium resistance,since the base emitter circuit allows someelectrons (or holes) to exit at the collectorconnection.

Under the bias conditions just described,an external milliammeter will show thatapproximately one ma. is flowing in theemitter circuit. Further investigation willshow that approximately 95% of this cur-rent (0.95 ma) is flowing in the collectorcircuit. This may be hard for the reader toreconcile since it has been emphasized thatthe emitter base junction is equivalent to alow resistance, and the base collector junc-tion to a much higher resistance. Whydoesn't more current flow in the emitterbase junction? Although not stated prev-iously, the collector is at a much higher

1.0 I

MA.+

E(P)

B(N) (P)C

10.05

+0.95I MA.

Figure 1.8-CCURRENT FLOW IN A P -N -P TRANSISTOR

Radio Handbook Transistor Action 21

potential than is the base, with respect tothe emitter. An inspection of figure 1.8-Cwill make this fact obvious. It can be seenthat the collector -to -emitter voltage sourceis actually two batteries in series, whilethere is only one battery between emitterand base. In practice, the emitter base po-tential may be only 0.2 volts, while 1.5 to100 or more volts may exist between col-lector and emitter.

Another consideration is the width of thebase region. The base width is quite nar-row, and the number of recombinations ofelectrons from the emitter circuit and holesfrom the base circuit is very small, withonly about 5% of the electrons leaving theemitter. The remaining 95% of the emitterelectrons will traverse the emitter collectordepletion region and reach the collector.This accounts for the division of currentas shown in figure 1.8-C.

Amplification If, in our experimental cir-cuit (figure 1.8-C), the bat-

ery started to weaken, the terminal volt-age would drop. The current in the emitterbase junction might drop from 0.05 ma.to 0.03 ma. We know that this currentdrop would widen the potential barrier andreduce the amount of electrons (or holes)flowing in the collector circuit. For exam-ple, in practice, the collector current mightdrop from 0.95 ma. to 0.75 ma. Thus, achange of 0.0 2 ma. in the emitter base j unc-tion has caused a 0.2 ma. change in the col-

COLLECTORBASE

EMITTER

(PNP)

BASECOLLECTOR

EMITTER

(NPN)

Figure 1.8-DSCHEMATIC REPRESENTATION OFP -N -P AND N -P -N TRANSISTORS

Note that the direction of the arrowspoints INTO the current flow.

Ily constant voltage v,1. Zener diodes, tilered in the reverse dirt

1 as a voltage regulatment.

+1

ZEN ER DIODES ARE MANUFACTUREDTO "BREAKDOWN" AT DISCREET

VOLTAGE.They thereby provide excellent regu-lating characteristics. Note the "Z"shape of the voltage/current curve.

(Courtesy, International RectifierCorporation)

lector circuit. Stated another way, the tran-sistor exhibits current amplification. Simplemathematics will show that this transistorhas a current gain of 10. The current gainratio is termed beta (p) and is defined as thechange in collector current divided by thechange in base current. The beta of a trans-istor is always more than one (assuming itis not defective) and may reach one hundredor more.

In the preceding example, when thebasecurrent dropped from 0.05 ma. to 0.03 ma.,the collector current dropped from 0.95 ma.to 0.75 ma. The emitter current, then, musthave dropped from one ma. to 0.78 ma.Even though the margin between collectorand emitter current is less, it can neverreach unity, or one. This brings us aroundto the term alpha (a) which is expressed asthe change in collector current divided by thechange in emitter current. Alpha can neverexceed one, and is usually closer to 0.95.


100 ft

Figure 1.8-EA SIMPLE EXPERIMENT DEMONSTRAT-ING GAIN IN A COMMON -BASE CIR-

CUIT.

10000.

Voltage To illustrate how voltage amp li-Gain fication may be obtained, con-

nect the transistor as shown infigure 1.8-E. It is similar to the previousfigure, but the schematic symbol is used.Resistors have been inserted in series withthe emitter and the collector. Also, asource of alternating current is connectedin series with the forward bias battery.Since the emitter -base junction representsa low resistance, this series resistor is small(100 ohms ). The collector -to -base junctionappears to be a greater resistance, and there-fore this resistor is larger (1000). Assumethat additional batteries have been added tocompensate for the resistors, and the cur-rent flowing in the circuit is the same asbefore ( 0.05 ma. base and 0.95ma.collector ).

The a.c. is at very low voltage, however.The negative peaks aid the forward biascurrent causing the transistorto draw morecollector current. Positive peaks oppose theforward bias, which in turn reduces thebase current. As before, this changing cur-rent is amplified by the transistor and ap-pears as a large variation in the collectorcircuit.

The transistor has a current gain ( p )of 10. For every unit of current change inthe base circuit there will be 10 units ofchange in the collector circuit. Similarly,the transistor will exhibit a voltage gain. Inother words, one unit of current throughthe base resistor will cause one unit ofvoltage drop across it, and 10 units of cur-rent through the collector resistor will cause10 units of voltage drop across it. Theamplified voltage can be used to drive a

following transistor for additional gain.The transistor emitter, which may be

thought of as emitting the electrons (orholes), is similar to the vacuum tube cath-ode. The base, which acts as the controlcenter for the electron flow, may be likenedto the tube's grid. The collector, which gath-ers up the electrons (or holes), acts muchthe same as a vacuum tube plate.

The transistor may be compared to thevacuum tube in yet another way. If thepositive peak of the driving signal is toogreat, it opposes the bias current to thepoint where the transistor collector currentwill be cut-off If the negative peak is high,it aids the bias in forcing almost all of theelectrons (or holes) into the collector, pro-ducing saturation.

This section has presented some of thehighlights of transistor action. Voltage andcurrent applification, bias, stabilization, cir-cuit configurations and impedance match-ing are much too diverse subjects to includein this chapter. These subjects are coveredmore fully in Chapter 2.

A CRYSTAL BEING GROWN BY AHIGH TEMPERATURE -METHOD

The molten material is cooled as it iswithdrawn from the crucible to form agermanium crystal ingot. (Courtesy,

CBS Electronics)

Radio Handbook Transistor Construction 23

1.9 Transistor Construction

The physical construction of transistorsis interesting, but knowledge of the subjectis not necessary for design of circuits usingtransistors.

Purification There are so many difficultiesand variables connected withmanufacturing a transistor that

it is difficult to believe that one can be man-ufactured to sell for less than $1.00. But itis done, and the manufacturer, distributor,and supplier each make a profit!

Take, for example, the purity required.To make a transistor, the impurities shouldbe reduced to one part in 10' , or aboutone grain of sand in a boxcar load of sugar!This is accomplished by zone refining thegermanium or silicon. The contaminatedmaterial is placed in a carbon boat or sup-ported between carbon holders. Carbon forthis purpose is inert, and its impurities donot affect the material being refined.

The boat containing the slender rod ofsemiconductor material is passed throughan induction coil which melts a narrow zoneof material. The boat moves slowly throughthe coil allowing the melted zones to solid-ify again. The impurities are forced alongahead of the melting area much the same asone might wipe dirt off a mobile whip an-tenna with a cloth. The end result is a prac-tically pure ingot of germanium or silicon.

E B C

Figure 1.9-ASIMPLIFIED CONSTRUCTION OF A

POINT -CONTACT TRANSISTOR

CLAIMED TO BE THE WORLD'S LARG-EST TRANSISTOR

This demonstrator dynamically showsthe flow of holes and electrons in ajunction transistor. (Courtesy, Texas

Instruments.)

PointContactTransistor

transistor.

The invention that heralded thedawn of the semiconductor erain 1948 was the point contactThe junction transistor was not

invented until late in 1949. Point contacttransistors are not now used or manufac-tured. In fact, they are considered collect-or's items.

Internally, the point contact transistorlooks very much like the 1 N2 1B radar di-ode, but it has two very closely spaced "cat-whiskers", as shown in figure 1.9-A. Thecontact points rest on a slab of N -type ger-manium. By itself, this device would notwork at all. However, in manufacture, avery short pulse of high current is passedthrough the contact and the germanium.This is known as forming and produces afilm a few atoms thick which exhibits thesame characteristics as a type -P germanium.Since the contacts are very close together,the emitter base junction is capable of con-trolling the current flow in the base col-lector junction.


INITIAL INTERMEDIATECONDITIONS STAGE

FINISHEDSTRUCTURE

=MB \ i 1

ALLOY IE=I(IMPURITY CONTACT) WWW/71.. If W

DOTSDOTS APPLIED

t PI

HEAT MELTS DOTS RECRYSTALLIZE

...-DIFFUSION

(IMPURITY CONTACT) .0 f .,DOUBLE ETCHING

GASEOUS DOPING DIFFUSION EXPOSESAGENTS APPLIED COMPLETED BASE

RATEGROWING

(GROWN)..:"--''' 7...

HEAT

--

HEAT REAPPLIED

IF i tIJ''''k.-J1

..f k.k oREMOVAL

CYCLE JUSTCOMPLETED GI

V GROWTHESRAPI D T O FORM

JUNCTIONS

MELTBACK e(GROWN)"fDOUBLE 0 TIP FREEZING

DOPED PELLET HEAT MELTS TIP FORMSJUNCTIONS

AGROWN

' .DIFFUSED - ,... ,,H1 :7;'ir:H.'W:til(GROWN) ?'. t Al. i t t k'.

MOLTEN METAL BASE IMPURITY EMITTERDOPED wITH DIFFUSES REGION ALONEEMITTER AND RAPIDLY INTO CONTINUESBASE IMPURITIES COLLECTOR TO GROW

Figure 1.9-BALLOY AND GROWN TRANSISTOR

JUNCTIONS


These early point contact transistorshave apas low as 2.5. They have the furtherdisadvantage of being very delicate, since asharp blow could dislodge the fragile cat -whiskers. As they could not be made inproduction quantities, they were priced veryhigh.

To its credit, the point contact transistorcould operate to approximately 100 Mc.The input impedance was ofmediumvalue,while the output impedance was of the or-der of 20,000 ohms.

Alloy and Impurities may be added to theGrown pure germanium or silicon inTransistors one of several ways. Some ofthe methods are shown in figure 1.9-A.The alloy transistor technique was probablya major contribution to the mass produc-tion of transistors.

In this operation dots of impurities areplaced in contact with the pure material andheated. The applied heat melts the dotsand the impurities alloy into thebasemater-ial. When the dots recrystallize, we findthe impurities in the formerly pure mater-ial.

The impurities may also be added in agaseous state as in the diffusion method.The impurities are allowed to envelop thepure material. It is then cut up and etchedto expose the base region.

Another way to make transistors is togrow them! The pure material is placed ina graphite crucible (also inert) which isinserted in a crystal -growing furnace. The mater-ial is melted by induction and apull rod witha seed on the end is dropped into the moltenmetal, then withdrawn slowly. The heatmay be applied intermittently causingspaced junctions on the single crystal pulledfrom the molten metal. A more commonmethod is to add impurities to the moltenarea just below the crystal. With carefultiming and temperature changes, it is pos-sible to grow a long series of P -N -P orN -P -N junctions.

Upon completion of the growing cycles,the ingot (containing the junctions) is cutinto thin wafers. The wafers, in turn, arediced up into tiny blocks and mounted inthe transistor case. Leads are connected tothe elements, usually by a system of thermo-compression bonding. Finally the air is re-

moved, an inert gas inserted, and the tran-sistor is sealed.

The completed transistor is then testedand graded. The grading process is neces-sary, since not all transistors from the samebatch will test the same. The best units arelabeled with a particular 2N number, whilethe remaining transistors are again graded.The transistor run is sorted into specifica-tion groups, such as the maximum fre-quency of amplification, and the remainingfew are either destroyed or sold as salvage.

Radio Handbook Transistor Construction 25

The Drift The drift transistor, pioneeredTransistor by Radio Corporation of Amer-

ica, is manufactured in such amanner as to accelerate the electrons on theirjourney through the junction. This short-ens the time it takes for the electrons totraverse the material. This shortened transittime increases the high frequency perform-ance much the same as in a vacuum tube.

Surface The surface barrier transistor wasBarrier developed by the Philco Corp-Transistor oration. In this device, two pitsare etched into opposite sides of the basematerial. The spacing between the pit bot-toms is only a few tenths of one -thousandthof an inch (a fraction of one mil). Next, thesurface of the bottom of the pit is electro-plated with emitter and collector materialand electrodes are attached. This closespacing between emitter and collectorgreatly reduces the transit time, which im-proves the high frequency performance ofthe unit.

M.A. D.T. The micro -alloy diffused transistor,a product of the Philco Corp-oration, is similar to the sur-

face barrier transistor. The base width,diffusion layer width, and collector dia-meter are individually controlled by electro-chemical etching. Since these factors arelargely responsible for determining thetransistor's high frequency properties, pre-cise control results in a device of excellentuniformity. Several current M.A.D.T. typetransistors, such as the Philco 2N1745, arevery inexpensive, yet have an f max as highas 1 300 Mc.

Mesa The mesa transistor is well suitedto high frequency, high powerapplications. Using photolith-

ograph techniques, gaseous impurities arediffused into the semiconductor wafer toform two P -N junctions. The semicon-ductor wafer is etched away from the base

NEW METHODS OF GROWING SILI-CON AND GERMANIUM ACCOUNTFOR MUCH OF THE PROGRESS IN

SEMICONDUCTORSSilicon diodes are the end products ofthe crystal ingot pictured here. (Court-

esy, CBS Electronics)

emitter area, leaving these elements on aplateau or mesa. The mesa transistor ex-hibits low saturation resistance, high cur-rent beta, and, in addition, many usefulv.h.f. characteristics. Its high dissipationmakes the design of high -power amplifiersquite feasible.

Epitaxial The epitaxial transistor is a lateTransistors development in semiconductor

technology. The device is simi-lar to the mesa, but the semiconductor layeris formed by an epitaxial (Greek for "set-tling") process and evaporated on a basesubstrate. These transistors feature all thefavorable characteristics of the mesa, butare somewhat easier to construct. Theyshould play an important part in futurehigh -power communications equipment.


PROCESSDESIGNATION

GEOMETRICALSHAPEB. BARD. DOUBLE

SIDED WAFERS. SINGLE

SIDED WAFER

CROSS-SECTIONALVIEW

SHOWINGJUNCTIONS

(Nor TO SCALE )

RATE GROWN B

E B C

I II I

MELT BACK B EDMELTBACK-DIFFUSED B ED

GROWN -DIFFUSED BI II

DOUBLE DOPED BI I

ALLOY D

DIFFUSED ALLOY(DRIFT)D

ALLOY DIFFUSED SCE

DIFFUSED BASE(MESA) S

B

DIFFUSED EMITTER -BASE(MESA)

S B

E L

SURFACE BARRIER D gMICRO ALLOY D

MICRO ALLOY DIFFUSED D #

Figure 1.9-CCHART SHOWING JUNCTION PRO-

FILES FOR VARIOUS TRANSISTORTYPES


Power The construction and manu-Transistors facturing techniques for power

transistors are similar to those ofalloy transistors. In this device, however,more material is used to increasethe powerhandling ability. In addition, several stepsare taken to remove heat from the junctions.Since the power transistor has more mass,its high frequency performance is reduced.

The transistors discussed represent onlya few of the popular types. A more com-plete listing is shown in figure 1.9-C.

In section 1.6, it was mentioned thatthermal agitation was responsible for theaction of the minority carriers (leakagecurrent). This points up a basic flaw ofthe transistor. Changing temperature willaffect the transistor by influencing both themajority and minority carriers. To illus-trate the point, bias a transistor to drawone milliampere in the collector circuit. Asa soldering iron or match is brought nearthe case, the heat causes increased thermalagitation in the unit and the collector cur-rent increases beyond one ma. This effectis inherent in transistors and must becom-pensated for in circuit design. Thetemper-ature influence on current is minimized inthe better transistors but is always present.

Abnormally high temperatures maymake the transistor useless or actually des-troy it. Germanium transistors are usuallyrated to 85° C (185° F. ), which is morethanadequate for most applications. Militaryequipment, on the other hand, must have awide margin of safety. This type of equip-ment almost always employes silicon tran-sistors which may be used in external temp-eratures as high as 350°F.

High The high frapency tinnsirtorFrequency places special emphasis onPerformance reducing the base width toshorten transit time and to increasethe highfrequency performance. In addition, theemitter and collector junctions are made astiny as practicable to minimize junctioncapacity, which also inhibits the high fre-quency performance.

These techniques do increase the highfrequency operation but have one seriousdrawback. As the junction size is reduced,so is the heat -dissipating ability. For thisreason most high frequency transistors havevery small dissipation rating. One exceptionis the mesa transistor.

A power transistor has a relatively largemass and junction area. Although it canbe made to dissipate hundreds of watts, theoperating frequency is reduced when com-

Radio Handbook 27

pared to smaller transistors. Some high -power audio transistors do not exhibit flatresponse between 20 cycles and 20 kilo-cycles, and feedback correction in ampli-fier circuits using these devices must beincorporated.

Noise The early transistors were verynoisy. Although this gave tran-sistors a very bad name, modern

devices are the equal of tubes insofar asnoise generation is concerned. In theory,at least it should be possible to make atransistor which is quieter than a vacuumtube, due to the lower operating temper-ature of the transistor. Currently, transis-tors are replacing tube preamplifiers inapplications where quality and performanceare of paramount importance.

Transistor It is impossible to make a flatLife statement as to the life of a

transistor. If it is adequatelysealed to prevent external contaminationand operated conservatively it might lastforever! The fact that it has no filamentweighs heavily in its favor.

Transistors are far more rugged thanare vacuum tubes. Military transistors have

HIGH POINER ATHIGH FREQUENCIESThese transistors have 125 watts dis-sipation and an alpha cut-off of 50 mc.and are the smallest types in the ser-ies. These triple diffused mesa trans-istors are available up to 1200 wattsdissipation. (Courtesy, Pacific Semi-

conductors, Inc.)

actually been shot from a gun into a tele-phone book to advertise dramatically howwell they are constructed. When the tran-sistors were reconnected into the circuitthey worked perfectly!

CHAPTER TWO

Audio Amplifiers

Fundamentally, the transistor is a valvewhich controls the flow of electricalcharges. We call thesechargescutrentcarrters.As current carriers pass through the tran-sistor they are controlled as easily as thecurrent carriers which pass through an elec-tron tube.

It is sometimes advantageous to com-pare transistors with vacuum tubes, for inmany ways they are similar, but there aremajor differences to be pointed out as weproceed.

The transistor may be compared with avacuum tube by considering the actions ofthe two devices. The emitter, which maybe thought of as emitting electrons (or"holes"), is similar to a tube cathode. Thebase, which influences electron flow, maybe likened to the tube grid. The collector,which gathers the current carriers (elec-trons or holes), acts in a manner similarto the vacuum tube plate.

2.1 CircuitConfigurations

We can make further comparisons be-tween the transistor and vacuum tube byexamining the circuit configurations. Thereare several circuit arrangements for intro-ducing a signal into a transitor, and forextracting the amplified signal from thecircuit.

Common -Base Many have used the cir-Amplifiers cuit in which a carbon

microphone is connec-ted in series with the cathode ofa tube amp-lifier. This is a simple grounded -grid audiostage.

transistor equivalent, shown infigure 2.1-A, the signal is introduced intothe emitter -base circuit, is then amplifiedby the transistor, and is extracted fromthe collector -base circuit. Since the baseelement is common to the input and out-put, this circuit is termed a common -baseconfiguration. The emitter requires a pos-itive voltage, and the collector requires anegative potential (for a P -N -P transistor),and, in this circuit, two separate batteriesare required.

Figure 2.1-AA COMMON -BASE AMPLIFIER COM-PARED TO THE GROUNDED -GRID

VACUUM TUBE AMPLIFIER

Bias Considerations 29

Figure 2.1-BA COMMON -EMITTER AMPLIFIERCOMPARED TO A GROUNDED -CATH-

ODE VACUUM TUBE AMPLIFI ER

Common -Emitter This is the most com-Amplifiers mon configuration in

transistor (and vacu-um tube) equipment and is shown in fig-ure 2.1-B. Not only does it exhibit morepower gain than the other circuits, butlike the vacuum tube equivalent, it is theonly configuration which provides signalphase inversion between the input and out-put circuits.

In this configuration both the base andcollector are biased with negative poten-tials. Since both potentials are negativewith respect to the emitter, a single batterymay be used. Internally, the base -emitterjunction forms a voltage divider in con-junction with R1, which provides the nec-essary low value of forward bias voltage.

Common -Collector The cathode followerAmplifier is a circuit which

operates with itsplate "grounded" for the signal frequency.The transistor equivalent is a common col-lector stage or emitter -follower, and is shownin figure 2.1-C. Contrary to other circuits,a cathode or emitter follower may havepower gain but has no voltage gain.

Symbols In figure 2.1-D will be foundthe symbols for P -N -P and N-

P -N transistors. Note that in both casesthe arrows point in the direction of thehole flow, and opposite to the movementof electrons.

Figure 2.1-CA COMMON -COLLECTOR AMPLIFIER,OR EAMTTER FOLLOWER, IS SIMILARTO THE VACUUM TUBE CATHODE

FOLLOW ER

2.2 Bias Considerations

A transistor which is designed to havea negative d.c. voltage applied to its col-lector is a P -N -P type (positive -negative -

positive). An N -P -N transistor has a pos-itive potential on the collector (with respectto the emitter).

The circuit of figure 2.2-A is one of thesimplest transistor audio amplifiers. Ignor-ing resistor R , for the moment, the sig-nal is applied between the base and emitterof the transistor. This is equivalent todriving the grid and cathode in a vacuumtube circuit. The emitter (cathode) isgrounded, and the collector (plate) is con-nected through the headphones to the neg-ative terminal of the battery.

PNPTRANSISTOR

(A)

NPNTRANSISTOR

E

NOTE ARROWHEADREVERSED

(B)

Figure 2.1-DSTANDARDIZED SYMBOLS FOR THEPNP (a) AND NPN (b) TRANSISTORS

30 Audio Amplifiers The Transistor

The input signal controls the currentflow between the emitter and collector.Unlike a vacuum tube, a transistor requirescurrent bias. In other words, the base ismade to draw current so that it can controlthe flow of carriers between emitter andcollector. This is accomplished by apply-ing a negative potential to the base until itdraws the correct current. This is the pur-pose of R1, which is known as the forwardbias resistor, and its value may be determinedby Ohm's Law. The required base currentis given in the transistor data and charac-teristic charts.

If the audio amplifier must handle alarge input signal, the transistor is biasednear the center of its collector current curveso that the instantaneous collector poten-tial may swing an equal amount in each di-rection. If the amplifier is biased near thebottom of its curve it will operate class -B.

The circuit of figure 2.2-A is a simpleand practical one, and many similar cir-cuits will be found in this and other books.It might be considered as an audio "pack-age" which could be added to crystal setsor transistor regenerative receivers.

2.3 Stabilization

It is an unfortunate characteristic of thetransistor that when it is heated the collector

current rises. This larger current flow willin turn heat the junction further. (A com-

parable phenomenon in the economicworld is termed inflation, which if allowedto continue unchecked will cause a struc-tural collapse!) Most of us have, at onetime or another, inadvertently made the

plate of a tube glow cherry red, yet haveseen the tube go on working year afteryear. One cannot do this with a transistor.Once a transistor is over heated, its charac-teristics are altered and the unit often isrendered inoperative.

If the radio equipment containing thetransistor is heated (in a closed car, forexample, that has been standing in the hotsun) the current flow through the transistormay well rise to startling figures. Somemeans must be adopted to check the currentand heat rise. This means, or measure ofcontrol, is known as stabilization.

D.C. Feedback The circuit of figure2.3-A is identical with

that shown in figure 2.1-B, except that theearphones have been replaced with a loadresistor. When the current through thetransistor increases, voltage is dropped a-cross the resistor R2. If the forward -biasresistor R , is connected to the collector asin figure 2.3-B, a measure of stabilizationis obtained. As the current through the

PNP TRANSISTORS0071,2N109, 2N117, GT109, 2N1380

SIGNALINPUT

+1.5 TO 3V.

Figure 2.2-AA SIMPLE TRANSISTOR AMPLIFIER

A simple transistor amplifier showingthe easiest method of applying bias.

IN. 2

(-IA OUT.

Figure 2.3-AThis circuit is similar to fig. 2.2-A, butuses a collector load resistor. It is sus-

ceptible to thermal disturbance.

Radio Handbook Stabilization 31

I N.

8-

a

OUT.

Figure 2.3-BConnecting the bias resistor to thecollector provides a degree of stabil-

zation.

transistor rises and voltage at the collectordecreases, the available base bias is de-creased. As a result, the flow of currentthrough the transistor is decreased. Thiscircuit (or variations of it) is used to a con-siderable extent in small radios and hearingaids where the number of componentsmust be held to a minimum.

Emitter The circuit of figure 2.3-C isBias identical with that of figure 2.2-

A except that the componentsC1 and R3 have been added. When thecollector current rises, an increased voltagedrop occurs across R3. This, in turn,has the effect of lowering the difference inpotential between the base and the emitter.In other words, an increase in collectorcurrent reduces the bias on the base and"drags" the collector current down again.

Capacitor C, is used to bypass the signalaround R3, for without it severe signal de-generation would result. The value of R3is chosen to create a voltage drop of one-third to one -sixth of the supply voltage.This is a simple Ohm's Law calculationobtained by dividing the required voltagedrop across R3 by the collector current ofthe stage concerned.

Bias Voltage In figure 2.3-D, anotherDividers resistor, R4, has been ad-

ded to the circuit. A volt-age divider is formed by the two resistors,R 1 and R4, and the junction is connectedto the base. Due to improved regulationof the base voltage the bias is unable tochange. The system shown in figure 2.3-D will provide superior stabilization. It isthe most widely used bias system of alland will be found extensively in this book.

Whether they are audio amplifiers, os-cillators, or class C amplifiers, most tran-sistor circuits will have a means of obtain-ing stabilization and bias as depicted infigure 2.3-D. The larger the value of R3,and the smaller R and R4, the better willbe the stabilization. A point to rememberis that if R 3 is made too large, the voltagebetween the collector and emitter will be-come too small. This causes the transistorto saturate (as explained later in this chapter)and will create severe signal distortion.

OUT.

IN.

Figure 2.3-CAn emitter resistor also helps stabil-ize the transistor to temperature vari-

ations.

OUT.

IN.

Figure 2.3-DPREFERRED CIRCUIT FOR BIASING A

TRANSISTORThis circuit is almost immune to tem-

perature effects.


Th erm istor The resistor, R4, may be aStabilization thermistor. A thermistor

is a special resistor whoseresistance decreases when heated. Whenused in the circuit of figure 2.3-D (inplaceof R ) it will reduce the base bias whenthe temperature increases. Thermistorsare used when very precise stabilization isrequired or when transistors are workingvery close to the maximum collector dis-sipation rating.

2.4 TransistorImpedances

One of the major differences betweentransistors and tubes is that transistors arecurrent amplifiers and tubes are voltage amp-lifiers. A small increase in current in thebase of a transistor will cause a large in-crease in current in the collector circuit. Sincethe transistor base draws current, its impe-dance will be low. This is quite the oppo-site of a tube.

The transistor may have a high outputimpedance. The common emitter circuit(figure 2.1-B) may have an input imped-ance of 1,000 ohms and an output imped-ance of 50,000 ohms. In some instancesthe output impedance may be more than amegohm.Impedance Matching The differing im-In Transistor pedances may beCircuits used to advantage.

A low impedancemicrophone may be fed directly into theemitter of a transistor whose high imped-ance output will match the grid of a follow-ing tube. Not only does the transistor actas a matching device, it provides consider-able gain. The use of a high impedancecrystal microphone creates an immediateproblem because of the low transistor inputimpedance. When shunted across the mi-crophone, the transistor will reduce con-siderably the output voltage and distort thefrequency response of the microphone. If

a step-down transformer is not used, it isnecessary to employ a different input cir-cuitry. An example of this is shown in fig-ure 2.4-A. This circuit is an emitter fol-lower which is the transistor equivalent ofthe cathode follower which was discussedin section 2.1. Like the cathode follower,it has a high input impedance, approxi-mately 100,000 ohms in this version. Theshunting effect of the transistor on the mi-crophone is further reduced by placing aresistor, R1, in series with the microphone.Although some gain is lost, the over-allresult is improved.

Capacitive To obtain maximum transferCoupling of power from onetransistor

stage to another, it is neces-sary to match the output impedance of theone to the input of the other. This mayreadily be done with transformers. Trans-formers, however, are expensive and usuallyoccupy considerable space. Often it is ad-vantageous to use resistance- capacitancecoupling and accept the reduced gainbrought about by the impedancemismatch.

Negative If feedback is used around anFeedback amplifier, the input impedance

may be raised. A cathode fol-lower has a high input impedance becauseof its complete feedback.

680

2 N408, 2N1380,0071. GT109

1-0 AUDIO OUT.101( 81.IF

1 -

Figure 2.4-AAN EMITTER -FOLLOWER STAGE

Suitable for use with high impedancemicrophones.

Radio Handbook Transistor Impedance 33

1G. IN. C

0075, 2N408, ETC. 8-, a-120.

100 K

100K r. 47K

8.1.18

ISO OUT.1

) Z 0(LOAD)

Figure 2.4-BAn emitter -follower stage with feedback

to further raise the input impedance.

Figure 2.4-B shows the circuit ofan am-plifier which employs feedback to raise theinput impedance. Capacitor C , is thefeed-back component. Without C , (and with-out a load) the input impedance is around100,000 ohms. With C , the impedanceis around 500,000 ohms. But, if a 1,000ohm load is placed across the output term-inal of this circuit, the input impedanceagain drops to 100,000 ohms. WithoutC , it drops to about 51,000 ohms. Thesefigures tell an interesting story. They re-veal that the input impedance ofa transistorstage is dependent upon the output load.Therefore, the stage following this circuitshould also have a high input impedance.

The table of figure 2.4-C shows the var-ious impedances which may be obtainedfrom different circuit configurations. Thecommon -base configuration (which is e-

quivalent to grounded -grid) has a low in-put impedance and a high output im-pedance. Quite often it is advantageous tofollow one configuration with another suchas following a common base with a commoncollector circuit.

2.5 Interpretation OfTransistor Data

In the figures published by transistormanufacturers concerning their products,the following information is usually listed:

(a) Maximum collector voltage.(b) Maximum collector current.(c) Maximum collector dissipation.

If the maximum current is multiplied bythe maximum voltage, the result will bemuch greater than the maximum collectordissipation figure. All values given are ab-solute maximums, and if the maximumvoltage is used, the maximum current can-not be used, and vice versa. The figureproduced when the operating voltage ismultiplied by the current must not exceed thedissipation figure.

If there is a transfomer, pair of head-phones, coil, choke, or any inductance inthe collector circuit, the peak voltage presentwill be twice the battery voltage. As a fur-ther caution, it must be realized that thesudden make or break of the circuit mayproduce voltage "spikes" many times themaximum collector break -down rating.For this reason, never remove or insert atransistor in a circuit when voltage is ap-plied.

Figure 2.4-CA CHART COMPARING THE VARIOUS TRANSISTOR CONFIGURATIONS

Tube Analogy Grounded Grid Common Cathode Cathode FollowerTransistor Equiv. Common Base Common Emitter Emitter FollowerCurrent Gain Approx. Unity High, 40 to 60 High, 40 to 60Input Impedance Low Medium HighOutput Impedance High Medium LowMax. Frequency Higher than

AlphaLower thanAlpha

Varies withCircuitry


Current If the current in the collectorGain circuit rises one milliampere as

the base current is made to riseten microamperes, the transistor has a cur-rent gain of .100. This assumes that thereis no resistance in the collector circuit.

The symbols for current gain are alpha( a ) or beta (13). Both terms are used,but alpha applies to common -base circuitryand will be less than unity. Beta applies tocommon -emitter circuitry, with typical fig-ures being between 20 and 100. In thecommon -base circuit, a one milliampererise in emitter current mightproducea cor-responding rise of 0.98 ma. in the collector.In themselves, these figures appear to giveno gain. The impedance of the emittercircuit in which the one milliampere changetakes place is very low (possibly 50 ohms).However, the 0.98 milliampere change inthe very high impedance output circuitshows that the grounded -base circuit iscapable of very high values of current gain.

Alpha Alpha cut-off, or lab as it is written,

Cut-off is the frequency at which the gainof the transistor decreases to

0.707 of the gain at 1,000 cycles. As theterm implies, this ratio is measured in thecommon -base configuration. The transis-tor will perform beyond its alpha cut-off fre-quency, but at that frequency, its gain isfalling. Consequently, from this figure, oneis able to ascertain with some assurancewhether or not a certain transistor will doa certain job. Suppose one desires to build

an r. f. amplifier at a frequency of 4 Mc.,with either a 2 N 3 31, a 2 N 1 3 8 0, or a2N371. Which would be the best transis-tor to use? The data on these transistorsshows that the 2N331 has an alpha cut-offof 1.6 Mc. The 2N1380 has an alpha cut-off of 2 Mc., and the 2N371 has an alphacut-off of 30 Mc. Obviously, the 2N371 isthe transistor for this application.

If the transistor is operated as a ground-ed emitter, it is well to choose a device withan alpha cut-off of five times the operatingfrequency. The 2N1 39 has an alpha cut-

off of 6.8 Mc. and will therefore make anefficient grounded -emitter amplifier up toa little more than 1 Mc.

2.6 Load LinesFor Transistors

The good designer does not sit downwith a box of resistors and capacitors be-fore him, trying different values in the tran-sistor amplifier until the speaker gives offthe desired sound. Rather, he obtains chartscontaining characteristic curves, and fromthese curves he is able to work out the var-ious circuit component values. Thedesign-er may do all this without having even seenthe components. The collector voltage -currentcurves yield much information to those whotake the trouble to read them. Figure 2.6-Ashows the circuit of a simple transistor am-plifier. Follow the step-by-step process ofcircuit design, just as a designer might do.

Dissipation It is wished to operate theIntersects amplifier class A and obtain

the maximum possible sig-nal output from it. Figure 2.6-B is a plotof the collector current for different collect-or voltage values at various base currentsfor a transistor. Assume that the transistorhas the following ratings: maximum col-lector voltage = -12 volts; maximum collect-or current = -40 ma.; maximum collectordissipation at 25°C = 80 mw. (milliwatts).

OUTPUT

Figure 2.6-AA TYPICAL TRANSISTOR AMPLIFIERThe circuit values are determined in

the text.

Radio Handbook Load Lines 35

100

V)-60

a.U.1

_J5 60

0I -

U

0-1-40

ICMAX.

20

0

COMMON -EMITTER CIRCUIT, BASE INPUTAMBIENT TEMPERATURE = 25.* C

4/'MAXIMUM DISSIPATION LINEtI -1.2i

717-7.7F.--'.i...71-is0I"--"-t-"-----0. 9

0.6

I. -0.7-0. is

-0 4

..--....OPERATINGLOAD LINE POIN7-

BASE MI L1AMPERES4 -0.2e

4.-

rHuro POINT SECOND POINT F/ Sr POINT

---"'---,-..rr

VC MAX../..

4,,,SUPPLV VOLTSI

1 NO SIGNALOPERATING PD/NT

-2 -4 -6 -6 -10 -12COLLECTOR -TO -EMITTER -VOLTS

-14

Figure 2.6-BBASE CURRENT CURVES FOR A TYPICAL TRANSISTOR

A "family" of base current curves for a typical transistor showing how to deter-mine the operating conditions for the transistor. The component values may also

be determined by correctly interpreting this data.

-16

First, draw intersects showing the max-imum dissipation of the transistor. Startingat 10 volts, 80 mw. divided by 10 volts = 8ma. Find and mark the point on the graphwhere the 10 volts and 8 ma. intersect. Thisis marked "First point" on graph. Do thesame at 8 volts. 80 mw. divided by 8 volts= 10 ma. (Second point) Repeat at 6 ma.and so on along the graph. Connect thepoints with a dotted line. The area to theleft of the dotted line represents all thepoints which are within the rated transistordissipation (under the maximum dissipa-tion line).

Next, decide upon a supply voltage. A9 -volt battery is a common size, so in thisexample assume 9 volts. Mark 9 volts onthe base line ("supply volts"). Draw alinefrom this point toward the vertical currentscale so that it just touches the dotted line. This

line crosses the collector (Y) axis of thegraph at approximately 37.5 ma. Had theline higher than this been taken the tran-sistor would operate in the overload portionof the graph. By just touching the maximumdissipation line, it is possible to realize themaximum output from the transistor. Hadmaximum output not been required, theline could have been taken to a lower col-lector current point. This would have low-ered the standing collector current and, ofcourse, limited the collector swing.

Operating Now select the operatingPoint point. This should be ap-

proximately halfway alongthe load line. The left-hand limit is setby curvature of the "base milliampere" lineat the left hand edge of the graph and theright-hand limit by the I cc, or collector


current that continues to flow even whenthe base current is zero. If the operatingpoint is at the center of the load line, thecollector voltage can swing from approxi-mately 0.8 volts to 8.5 volts. Four voltshas been chosen as the operating point.This is not exactly the center of the loadline, but near enough for the purpose. Byselecting a figure of 4 volts, the collector toemitter potential may swing down to 1 voltand swing up to 8 volts without runninginto non -linearity. The base current will fluc-tuate from 0.04 ma. to 0.4 ma. The staticcollector current will be -20 ma. The loadresistor equals:

RI Collector voltageCollector current

4 volts0.02 amps

= 200 ohms

The base current at the operating point= 0.22 ma. From Ohm's Law, resistor Rbis:

Rb 9 volts (supply volts)0.22 ma.

= 41,000 ohms.

From the graph, one may obtain the cur-rent gain for d.c. The base current variesfrom 0.04 to 0.4 ma., or a total variationof 0.36 ma. The collector current variesfrom 4 to 32 ma., or a total of 28 ma. Thed.c. gain of the transistor is 28 ma. dividedby 0.36 ma. or a Beta of 78.

Thus, from one graph we have obtainedthe following: static collector voltage andcurrent, minimum and maximum collectorcurrents and voltages, collector load re-

sistance, forward bias resistance, staticbasecurrent, minimum and maximum base cur-rents, and the d.c. current gain.

2.7 AmplifierConsiderations

In figure 2.7-A, with the componentvalues shown, the collector current has beencalculated to be one ma. with a collectorvoltage of 6 volts.

Input To calculate the input imped-Impedance ance use the formula:

Input Impedance = (1+0 ) hib

The term h fe is the forward current transferratio with the output shorted for a.c. The e

after the h, signifies common emitter con-figuration. The forward current transferratio is the base current gain and may be

found in the transistor data. For the2N190, this figure is 43.

The term h ,b is the input resistance withoutput shorted for a.c. The b after h. in-dicates common base operation. This fig-ure is also given in the transistor data andis 29 ohms for the transistor in question.

At one ma. of collector current, the in-put resistance for the circuit shown in

figure 2.7-A is 43 x 29 or approximately1250 ohms.

Output The output impedance is

Impedance given by the formula:

12 V

OUT.

+12 V

Figure 2.7-A

A PROPERLY DESIGNED AUDIO AM-PLIFIER STAGE

Impedance calculations are discussedin the text.

Radio Handbook Amplifier Considerations 37

This is simply Ohm's Law and says in thiscase that the voltage on the collector di-vided by the current flowing in the collec-tor will give the resistance of the load, R.,.

Power The maximum power output isOutput again Ohm's Law.

Power output - V" x i`2

The When the load is a trans -Transformer former, the turns ratio has

to be adjusted to suit thespeaker impedance or the impedance ofthe following stage. The formula for thisratio (N) is:

N -.slOutput impedanceLoad impedance

Assume the output impedance is 4,000ohms and the load is 1,000 ohms. Theformula becomes:

N - 40001000

- 2

Thus a ratio of 2:1, in the step-down di-rection, is required.

Class B Unlike a tube, a transistorAmplifier cannot be completely cut off.

Even when the base current iszero there will still be a small flow of col-lector current. This is known as the lc..Class B transistors must be biased a littleabove the cut-off point. This is the no -sig-nal operating point shown on the graphof figure 2.6-B.

Input The calculation of the inputImpedance impedance is more difficult

now because the base imped-ance is varying widely with signal. Thegreater the input signal, the lower will bethe base impedance. The impedance of thedriver transformer may be made other than

XR-2P Right View

HEATH XR-2PThe Heath XR-2P uses a class B push-pull audio output stage. (Photo by

Heath(

optimum in order to give the signal sourcea higher impedance. This has the effectof reducing the variation in input resis-tance at low current values, which in turnmakes a sizable reduction in distortion.

Transformer The output impedance forOutput a push-pull Class B stageImpedance is given by the formula:

2 (1,7')2- p0

PowerOutput

The maximum power output ofa push-pull Class B stage for agiven load and supply voltage is:

po 2(V`e)2

The mer The transformer -turnsTransformer ratio may be calculatedin the same manner as for a class A stage.

Efficiency The efficiency of the class Bstage is considerably higher than

for class A configuration. The standingcollector current is small, and the peak


O INCORRECT ® CORRECT

Figure 2.7-B

AUDIO -AMPLIFIER VOLUME CON-TROL CONNECTION

Correct and incorrect methods of con-necting the volume control in a tran-

sistor audio -amplifier are shown.

collector current is large. As a rule, theclass B stage may be designed to deliverpower equal to five times the collector dis-sipation of one transistor. That is, a tran-sistor with a dissipation of 80 mw., whenoperated in push-pull class B (with anothertransistor, of course), may deliver 400 mw.of power output.

Class -B operation is preferred for port-able receivers so that the standing currentdrawn from the battery is low.

The Volume The gain control circuit ofControl figure 2.7-B(a) is poorly

designed because the cur-rent flowing in the control will make itnoisy. The circuit of (b) should be used.If the amplifier follows a receiver, the gaincontrol may be the diode detector load.

The Tone Transistor tone controls willControl differ little from tube designs.

A variable resistor in serieswith a fixed capacitor connected from baseto ground of one of the audio stages willgive ample treble cut. There are so manysystems of bass and treble tone controlsit is impossible, in the space available, tocover them. In general, they are identicalto those used with tubes. Differences willbe brought about due to the lower transis-

tor impedances, and as a consequence ca-pacitors may be larger and resistors smal-ler.

Inverse This technique also followsFeedback tube practice. The resistor

R in figure 2.7-C feeds backenergy from the output stage to the driverstage. Its value in this instance may beabout 33K. Inverse feedback is helpful inreducing distortion and in stabilizing thegain of a transistor.

An unbypassed emitter resistor will pro-duce negative feedback. If carried too far,as with tubes, the stage gain will sufferseverely.

Saturation This is more important thanmight be supposed. The

term saturation could well explain the reasonthe amplifier or receiver that was built"didn't work the way the book said it

should." In figure 2.7-D, assume that theforward bias resistor R, has a resistanceof 10,000 ohms. Under these conditionsthe 2N1380 will draw about 8 ma. whenthe output transformer has a d.c. resistanceof 300 ohms. The potential between thecollector and the emitter will be about 1.5volts. If a pair of earphones whose d.c.

a -

Figure 2.7-C

A SIMPLE METHOD OF APPLYINGINVERSE FEEDBACK TO A TWO -

STAGE TRANSISTOR AMPLIFIERThis technique increases thefrequencyresponse and decreases the distortion.

Radio Handbook D.C. Amplifier 39

V -9V

Figure 2.7-DAn amplifier such as this will saturateeasily if the collector load resistancevaries too far from the nominal value.

resistance measures 2,000 ohms is con-nected in place of the transformer the cur-rent will drop to 3 ma. and the collector -

emitter potential will drop to 0.1 volts.The large current flowing through the highresistance earphones has reduced all theavailable collector voltage. The result issevere distortion and very low output.

If R is increased to 47K, the collectorcurrent will fall and the voltage on the col-lector will rise, allowing the transistor tooperate normally. The rule to be inferredfrom this study is that the d.c. resistanceof the output transformer must bethe sameas that specified to obtain the properresults. Otherwise, the experimenter mustbe prepared to change the bias conditionon the transistor. The lower the d.c. re-sistance of the output transformer, thegreater will be the power output that maybe obtained from the stage.

Complementary N -P -N and P -N -P tran-Amplifiers sistors may both be

used in the same am-plifier stage. Figure 2.7-E shows how apush-pull complementary circuit that doesnot use an input or output transformermay be constructed.

Although single -stage amplifiers maydrive a speaker directly, without an outputtransformer, the steady d.c. collector cur-rent displaces the speaker cone from its

Figure 2.7-EA COMPLEMENTARY AMPLIFIER

The amplifier operates class B, yet re-quires no input or outputtransformers.

normal position and the result is inferioroperation. A complementary amplifier config-uration such as that shown in figure 2.7-Ecauses the two collector currents to opposeeach other, and the speaker cone is not dis-placed. The two transistors must be close-ly matched in their characteristics.

2.8 MiscellaneousAudio Circuits

There are many transistor circuits foundin communications equipment which aredifficult to classify. Some of them are dis-cussed here.

D.C. Amplifier A dc. amplifier is exactlywhat the name implies:

d.c. voltage applied to the input is ampli-fied and presented in the output circuit.It may or may not appear in the samephase. The a.g.c. circuit of figure 3.7-A(Chapter 3) is really a d.c. amplifier. Ther.f. is converted to d.c. by the base -emitterdiode, and the d.c. causes.the collector volt-age to vary. The voltage drop across theresistor at the collector will be consider-ably greater than the voltage which pro-duced it. In this case it will be of the op-posite polarity.


0-1

Figure 2.8-AA SIMPLE TRANSISTOR

D.C. AMPLIFIER

The d.c. amplifier of figure 2.8-A is acurrent amplifier. A small increase in cur-rent at the base (input) of the transistorwill cause a large increase in currentthrough the meter. This circuit may beused to increase the sensitivity of a meter.

D.c. amplifiers are considerably moredifficult to temperature -stabilize than area.c. amplifiers. Temperature changes maymake necessary frequent adjustments to thezero setting in circuits such as in figure2.8-A.

S -M eterAmplifier

plifier is

The circuit of figure 2.8-A isshown as an S -meter ampli-fier in figure 2.8-B. This am -

little affected by temperaturechanges. The potentiometer R adjusts themeter zero, and R adjusts the sensitivity ofthe meter. The circuit has been redrawn inB to show its operation. From this, it willbe seen that the transistor and its associatedcomponents are a bridge circuit, of whichTR is a variable arm. Any change of TRcauses unbalance of the bridge and a con-sequent meter reading.

Reflex A reflex stage is one which ampli-Circuits fies one signal twice, on two sep-

arate frequencies. For example,an i.f. amplifier may also be an audio am-plifier. After the signal has left the i.f. am-plifier, it is detected and then fed backthrough the same transistor again as anaudio voltage. Reflexing can only be donewhen the two frequencies are sufficientlyfar apart that no interaction between thetwo signal components can occur. If an

i.f. amplifier is reflexed it cannot be con-trolled with a.g.c. voltage without depre-ciating the audio signal.

2.9 SemiconductorSpeech Amplifier

High impedance crystal and dynamicmicrophones are the phone operator'sstock -in -trade. But, as pointed out in sec-tion 2.4, such microphones are heavilyloaded due to the low input impedance ofa transistor. When the microphone doesnot "see" the proper load impedance, thelow frequencies are severely attenuated. Thenet effect makes even the most resonantvoice sound like a young boy. Even underthe best possible conditions, the mismatchwill cause a loss of gain.

The solution to this dilemma, alsoshown in section 2.4, is to connect the in-put stage in a grounded collector (or emit-ter -follower) configuration and employfeedback to raise the input impedance.

How It A speech amplifier featuring aWorks high -input impedance is shown

in figure 2.9-A. The input tran-sistor, TR1, is an emitter followerwith feed-back between emitter and base. The outputof this stage drives a common emitter am-plifier, TR2, which has the necessary com-ponents for proper temperature stabiliza-

TR, Ri

Figure 2.8-BS -METER AMPLIFIER

A circuit similar to fig. 2.8-A, butmodified for use as an S -meter am-

plifi er.

Radio Handbook S -Meter Amplifier 41

HI ZMIC.

R1

1)--M."--ff47-100K 8(1JF

005

25 F

TR I

100 K

81JF

100 K 47K

81JF

100K 10K 100K 10 K

TR2

IOK I

+T8LF

TR3

OK 1

F

12*

500 500

2K

Figure 2.9-AMICROPHONE PREAMPLIFIER

A properly designed microphone preamplifier which will bring a -54 db micro-phone up to about one volt. The coupling and bypass components are selected toamplify only those frequencies in the voice spectrum (300-3,000 cycles). Any ofthe transistors in the parts list may be used without component substitution.

Parts ListT, - Interstage audio transformer, 10K to 2K c.t. (Triad TY-56X or Stancor TA -35).TRH, TR2, TR3 - Use either 2N217, 2N406, 2N1380, GT109, or 0071.

+12V.

tion. This stage, in turn, drives a similarstage, TR3, with a matching transformeras its collector load. A 10K resistor isconnected across the primary of the match-ing transformer to help provide a constantload for TR3. This resistor may not benecessary in some audio applications. Theoutput winding will match a transistorpush-pull load such as the Mini -amplifierdescribed in section 2.11. This speechamplifier may also be used to drive an audiophase shift network or diode balanced mod-ulator in s.s.b. phasing and filter trans-mitters.

Audio feedback through the power sup-ply circuit is prevented by decoupling thefirst and second stages from the outputtransistor. A 1K resistor and a 25 Ai.capacitor are used for this purpose. Feed-back due to stray fields is minimized byusing an R -C filter network in the micro-phone lead.

The amplifier has more than adequategain for most applications, and it may benecessary to install a volume control. Thismay be accomplished by substituting a 10Kpotentiometer for the stabilization resistor

in the base circuit of TR3. The electrolyticcoupling capacior would connect to the armof the potentiometer. This is shown infigure 2.9-A by dotted lines.

Construction The amplifier can be con-structed on a piece ofterm-

inal strip material. The transistors mountby their leads to the terminals, and thecom-ponents are laid across the board, betweenthe terminals. The layout is entirely up tothe builder, and is not particularly critical.There is no tendency for the speech ampli-fier to oscillate due to proximity of com-ponents. The builder should be cautionedto use shielded cable for the microphonelead, however. Due to the high input- im-pedance, the transistor speech amplifier isjust as likely to pick up stray a.c. as is itsvacuum tube counterpart! Noise on thesupply line should not be a problem dueto the large decoupling components (1Kand 25 //id. ).

Note that the amplifier is designed withboth a common plus and a common minusline. This will allow the builder to use theunit in a mobile installation with either apositive or negative electrical system.


SEMI -CONDUCTOR SPEECH AMPLI-FIER

This transistor speech amplifier hasan input impedance of approximately100K and provides an excellent matchto crystal and dynamic high impe-dance microphones. It occupies aspace of only 4" X 2" x 3/8" whenmounted on the etched circuit board.

Operation The best way to test this am-plifier is to connect it to a

transmitter with known audio characteris-tics. Connect a 470 -ohm resistor acrossone-half of the secondary of T1. In parallelwith this, connect a 100 to 1 divider (a100K and 1K resistor will do nicely). Us-ing shielded cable, connect the 1K resistoracross the high impedance microphone in-put on your transmitter. Ground the speechamplifier to the transmitter (either plus orminus line) and connect the microphoneyou will be using with the unit. Have somereliable person check your audio qualitywhile you adjust resistor R for the most

pleasing balance between highs and lows.In most cases you will find that no resis-tance is necessary, but a minimum of 4.7Kshould be used to act as an r.f. filter. Thisis necessary to prevent stray r.f. picked upby the microphone lead from being rec-tified by TRI and producing audio feed-back.

When you are satisfied with thefrequen-cy response, disconnect the divider whichwas temporarily placed across the output.Measure the audio voltage across the 470 -ohm resistor while speaking into the mi-crophone. You should be able to obtainalmost one volt of audio before noticeabledistortion takes place. This is more thanadequate drive for s.s.b. audio phase -shiftnetworks and is just about ideal for diodebalanced modulators which operate withabout 6 volts of r.f. signal level.

The noise generated by the input tran-sistor is approximately 30-40 db downfrom full output. This figure may be im-proved considerably by substituting an in-put transistor intended for low -noise audioapplications.

Radio Handbook Audio Compressor 43

2.10 A DeluxeAudio Compressor

An audio compressor, when used aheadof public address amplifiers or commercialand amateur transmitters, will provide amajor gain in "talk -power." The averagevoice level may be held high, but the peaks(which would normally overload an am-plifier or transmitter) are compressed to asafe or desirable level. In tube compressors,use is made of the variable-fi characteristicof the control grid to adjust automaticallythe gain of the amplifier.

Ordinarily, the transistor gain is variedby (a) reducing the emitter current, or (b)by reducing the collector voltage. The twosystems are discussed in Chapter 3, under'the heading automatic gain control. It willbe realized that automatic gain control ismuch the same whether the signal beingcontrolled is audio, r.f., or i.f. energy.Both of these control systems drive the tran-sistor into a non-linear portion of its char-acteristic curve in an effort to reduce stagegain. This, while perhaps permissible inr.f. applications, is not desirable in audiocircuits. Considerable experimental work(both with characteristic curves and com-pressors) based on the above principleswas undertaken, and the conclusions con-firmed that distortion of a serious naturedid take place.

A New Principal- Apart from the twoHow It Works methods mentioned,

the gain of a stage maybe controlled by shunting its output elementwith a low -value resistor. A resistor placedacross the output transformer primary ofa tube receiver will reduce the audio voltageacross the transformer to an extent depend-ent upon the value of the resistor. If theresistor is reduced in value whenever thesignal in the transformer exceeds a pre-determined voltage, the audio (as deliveredby the speaker) will never rise above a cer-tain loudness level.

...

THE COMPRESSORThe compressor was constructed in a

small L-shaped chassis box. The knobis the compression control.

In this case the resistor is a nearly -sat-urated transistor connected in parallel withthe load resistor of an audio amplifier. Thecontrol transistor is biased to operate justabove the knee in its curve. Signal is ex-tracted from the amplifier, further ampli-fied, converted to negative d.c., and appliedto the base of the control transistor. Theoperating point of the transistor "slidesdown" the curve, and the internal resistancerapidly falls. As the connection between


LO ZM IC. 81JF

TR,100n

o+12V. -

OUTPUT

TR 3 PRI. SEC.

P 811F 5.66 ION TI 2 N,C.T.BLUE

RED

Ill

YELLOW

-1 10011E

22 11j:55 'T 100

'UP

TRa

RITON APPROX.(SEE TEXT)

Figure 2.10-ASCHEMATIC DIAGRAM FOR THE DE -

LUX AUDIO COMPRESSOR

Parts ListDi - General purpose germanium di-

ode, 1 N34A, 1 N64G, 0A85, etc.R - Approximately 70K. See text.T - Interstage matching transformer,

10K to 2K c.t. (Stancor TA -35 orTriad TY-56X)

TR12 3 4- General purpose transistors,prototype model worked success-fully with 2N1380, 2N408, GT-81R, and OC 75.

Note that the two 25 /Ad. electro-lytics are connected in the mannershown to eliminate polarization, notto increase breakdown voltage.

A single 15 /dd. capacitor cannot beused as a substitute. Ceramic, mica,paper, and electrolytic capacitor no-tations are given below the values.

the control and the controlled transistor isvia a large value capacitor, the a.c. part ofthe signal is shunted to ground, leavingthed.c. working point of the controlled stageunchanged. Clipping of the signal cannotoccur unless the amplifier is considerably

)31

COMPRESSIONCONTROL

overdriven. A large signal at the collectorof the controlled stage may conceivably berectified by the controlling stage. However,with normal operation this undesirable con-dition does not exist (figure 2.10-A).

Construction The unit may beconstruct-ed in any convenient form.

The main point to observe is the matterof shielding, which must be complete. R.f.from the transmitter should not be allowedin the compressor unit.

The resistor R may be left out untilthe unit is completed. Transistors con-siderably different from those suggestedmay require a different value for this resis-tor. A potentiometer could be substitutedand then replaced with a resistor of themeasured value.

Adjustment After the wiring hasAnd Operation been completed and

checked, measure thebattery current, which should be between3 and 4 ma. Next, turn the compressioncontrol off and connect an audio generatorto the compressor (you can whistle into themicrophone). Adjust the temporary po-tentiometer R , until the output signal juststarts to fall. This indicates that the bias

Radio Handbook Audio Compressor 45

TaCRYSTAL ORDYNAMIC MIC

C

TR t

Figure 2.10-BMATCHING CIRCUITRY FORHIGH IMPEDANCE MICRO-

PHONESSchematic showing how a trans-former may be connected to al-low crystal or dynamic micro-phones to be used. TransformerT2 may be either a Stancor TA -

46 or a Triad TY-50X.

has been adjusted to place the transistoron the knee of its characteristic curve. Afurther reduction in resistor R will veryrapidly reduce the output which indicatesthat the correct point has been passed. Ifthe resistor is too high in value, and thebias too low, the rapid change in the col-lector voltage of the control stagewill causea pronounced thump at the beginning ofeach word and will show on the scope asa "blip" that exceeds the compressed level.

Excessive compression is to be avoided.An overcompressed signal will becomeharsh and unpleasant. Background noise

will become excessive, and a voice control-led rig will trigger every time a fly movesa foot on the ceiling!

The transformer T,, if different fromthat suggested, may require phasing. Thatis, the input or the two output leads willhave to be connected in the right manner.Incorrect phasing will show spikes at thebeginning of each audio cycle and a roughhum in the monitor. The remedy is simple:reverse either the primary or the secondaryleads, but not both.

The input impedance to the compressoris low and suitable for a dynamic micro-phone without a matching transformer. Acrystal microphone will require a matchingtransformer or a matching stage such asthe one shown in figure 2.4-B of thischapter.

INSIDE VIEW OF TRANSISTORIZEDSPEECH COMPRESSOR

The size of the unit could have beengreatly reduced if subminiature com-

ponents had been used.


2.11 The Mini -Amplifier -Modulator

The Mini -amp -modulator is designed tofollow a receiver which in itself has insuf-ficient amplification for comfortable listen-ing. The amplifier may be used in con-junction with many of the receivers de-scribed in this book. It is also a permanentpart of the communication receiver de-scribed here -in. By substituting a suitabletransformer, it may be used to modulatea 1/2 to 1 -watt walkie-talkie. The Mini --amplifier should do much to dispel theincorrect impression gained through listen-ing to too many 2 -inch speakers. Tran-sistor quality can be as good as tube quality!

How Refer to figure 2.11-A. TheIt Works signal is fed to the bases of the

two transistors via the step-down transformer T1. The input impe-dance was calculated for transistors with arather low h fe so that a variety of transis-tors may be used with only small alterationsto the circuit.

The transistors operate in class B andare biased just sufficiently to prevent cross-over distortion due to curvature of the tran-sistor characteristics. A lower output trans-former primary impedance would allowgreater output without saturation and dis-tortion, but the transistors might then ex-ceed their ratings. Output is much morethan sufficient for the average room.

Negative feedback is applied over the stagevia feedback resistor R2. This resistor maybe omitted if desired. The result will be aslight increase in gain and distortion.

Construction Any practical layout maybe followed. Printed cir-

cuit boards as used in the prototype makefor compactness, yet allow open, easy -to -get -at wiring.

9-12V

Figure 2.11-ATHE MINI -AMPLIFIER

The Mini -amplifier has a power out-put of approximately 350 milliwatts,which is more than adequate for ama-teur radio work. The various taps onT2 provide correct speaker matching.If the amplifier oscillates when R2 isconnected, connect it to the other col-

lector.

Parts List- Interstage transformer, 10K to 2K

c.t. (Triad TY-56X, Stancor TA -35,or equiv.).

T2 - Speaker matching transformer,500 ohms c.t. to 4, 8, or 16 ohms(Triad TY-45X, Stancor TA -42, or

equiv.).TR1-TR2 - 2N1380 (Texas Instruments).

Other transistors may be used ifR ' is adjusted for minimum cross-over distortion. The 2N109, GT109, 2N408 and 0072 were allused in the prototype with good

results.

Operation If transistors other than the2N1 380's are used, the re-

sistor R should be adjusted to give anidling collector current of around 3.0 ma.If the idling current is too low, crossover dis-tortion will be apparent at low volume levels.If it is too high, battery power will be wastedfor no useful purpose.

If the amplifier is operating from a bat-tery to which the other equipment is alsoconnected, "motorboating" or oscillationmay take place. This is particularly true if

Radio Handbook Mini -Amplifier 47

BATTERY

Figure 2.11-BIt may be necessary to add decoup-ling components as shown when theMini -amplifier is used with accessoryequipment. A 100 pfd. capacitor isalready built into the Mini -amplifier.

the battery is partly discharged or if thereis resistance in the battery circuit. Theclass B current of the amplifier may reachvery high peaks which can pull down thebattery voltage. This, in turn, will reducethe signal. The result is a "phut -phut -phut -

phut," not unlike the sound of a motor-boat. If there is very little capacity in thecircuit the "motorboat" may reach jet speedand give forth the appropriate sound! Asthe cause is almost always due to couplingthrough the power supply, adequate de-

. coupling is essential. A 100 /dd. capacitordirectly across the battery should be suf-ficient to prevent the voltage variation. Ifthe motorboating continues, it may be nec-

THE MINI -AMPLIFIERThe completed Mini -Amplifier with drivermounted on an etched

circuit board

essary to add decoupling resistors as shownin figure 2.11-B.

Modulator A small r.f. power amplifier(either tube or transistor)

can be modulated by substituting a 500 -to -5,000 ohm transformer (such as thethe Triad TX -49X or Stancor TA -4) forT2. It may be necessary to re -adjust the biasdivider network for minimum crossoverdistortion.

2.12 The Mini-Amplifier #2

Many applications call for more gainthan provided by the Mini -Amplifier alone.When this is the case, a suitable driverstage (figure 2.12-A) may be added. Thedriver should draw approximately 1 ma.of collector current to match correctlytransformer T1.

Note the polarity of the coupling capaci-tor C1. If the input side of this capacitoris connected to the collector of a precedingstage, the capacitor must be reversed. Sincethe collector will be more negative than thebase of TR1, the negative end of the capaci-tor must be to the input side.

The motorboating condition describedearlier cannot occur in the Mini -amplifier


TR ICI81JF

SiToll

MINI-AMPLIFIER

Figure 2.12-AMINI -AMPLIFIER

Many other transistors may be usedin this circuit without component valuealteration. Resistor 121 may b e adjustedto give a collector current of approxi-mately 1 ma. when the transistor char-acteristics are considerably different

from the types specified.

No. Two due to the added 100 ohm/100gfd decoupling network in the supply lead.If the amplifier is used with accessory radioequipment ( receiver, i.f. strip, etc.), theB- buss should be connected to the pointmarked "A" in figure 2.12-A.

2.13 A 10 WaftAmplifier/Modulator

Because it uses so few components (andthese are inexpensive), this amplifier shouldbe very popular for public address or modu-lator applications. The amplifier has beendesigned around the Motorola 2N554, a$1.35 transistor. Transformers, usuallythe most expensive part of an amplifier ormodulator, have been kept to a minimumand direct coupling has been employed be-tween the driver and the output stage.

How TR is a conventional commonIt Works emitter amplifier and requires

very little mention. If an audiogain control is required on this stage it maybe wired as shown in figure 2.13-B. Tran-sistor TR, is coupled to the push-pull

driver stage (TR2, TR3-2N1380) by a min-iature transformer. The driver stage is inturn direct -coupled to the class B amplifier.The emitter current of the driver is actuallythe base current for the class B stage. Ac-tually, both stages operate class B in orderto deliver the maximum amount of outputpower. The driver stage operates veryclose to its maximum dissipation rating,and a heat sink is recommended if the am-plifier is to be operated in high temperatureenvironments. The heat sink may consistof a small piece of copper platebent aroundthe transistor case and bolted to the chassis.The amplifier may be used as a modulatorsimply by substitution of the output trans-former. No component value changes arerequired.

Adjustment Before the amplifier is con-nected to the battery ascer-

tain that the potentiometer P, is at maximumresistance. With a tone fed into the input,and an oscilloscope connected across theoutput, adjust r, for minimum crossoverdistortion consistent with minimum2N554 collector current. If a scope is notavailable, adjust P1 until the two driver tran-

TOP VIEW OF AMPLIFIER/MODULA-TOR

The remaining transistors are mountedon terminal strips under the chassis.

Radio Handbook 10 -Watt Amplifier 49

TR12N1380

INPUT

T1

Figure 2.13-AA 10 -WATT EMITTER -FOLLOWER AM-

PLIFIER AND MODULATORExperimenters could use this samecircuit as a 40 -watt amplifier/modu-lator by using the 2N554's as emitterfollowers to drive a pair of 5 amperetransistors. A Triad TY-66A (modu-lator) or TY-67A (amplifier) transform-er would be used, and T2 would beeliminated.

PARTS LIST

P1 - 4 -watt wire -wound potentiometer.T - Interstage transformer, 10K to 2K

- c.t. (Stancor TA -35 or Triad TY-56X).

T Ra.TR32 N1380

TR4.TR52N 554

Tz

SPEAKEROR

CLASS "C.STAGE

T2 Collector to speaker, 24 ohmsc.t. to 4,8, or 16 ohms (Triad TY-29X). Collector to 20 -watt modu-lated stage, 32 ohms c.t. to 4K,6K, or 3K (Triad TY-65Z).

A universal push-pull plate to speakertransformer can be reverse -connected

as a modulation transformer.

If such a transformer is used, connectthe 16 ohm tap to one collector, the4 ohm tap to B-, and the common leadto the other collector. The secondarywill then be twice the original value

of impedance.

sistors draw 7 ma. of idling current. Thisshould provide a 2N554 idling current ofapproximately 100-150 ma. The idling cur-rent may fluctuate a little, according to tem-perature, but should not differ greatly fromthis figure.

Application A phonograph cartridge feddirectly into the input stage

will provide plenty of output if it offers areasonable impedance match to the tran-sistor. A crystal pickup will require match-ing, which can be provided by an emitter -

follower or 100:1 stepdown transformer.A dynamic microphone will drive the am-plifier to full output with close talking, andwhen a matching transformer is used itwill provide a reserve of gain.

The input impedance of the amplifier isaround 1,000 ohms.

Other Other transistors may beTransistors used in place of those sug-

gested. However, cautionshould be used in the selection of substi-tutes. The two driver transistors must havea dissipation of at least 150 mw. each if they

TR8.UF

INPUT 0 -14. -

Figure 2.13-13VOLUME CONTROL CONNECTIONA method of connecting a volumecontrol to the 10 -watt amplifier/mod-

ulator.


Figure 2.13-CTypical currents for the 10 -watt ampli-

fier/modulator

NO SIGNAL

TR1

TR2, TR3

TR,s, TR5

1.0 ma.7.0 ma. total100 - 150 ma total(measured at T2 C.t.)

SIGNALMaximum current

TR1

TR2, TR3

TR4, TR5

I 0 ma.60 ma. total1 ampere total (meas-ured at T2 c.t.)

are to drive a pair of 2N554's, or similartransistors. The pre -amplifier transistormay be readily replaced with a great numberof similar transistors, for it is in no waycritical. The driver 2N1380's and the out-put 2N554's are particularly suited to eachother, and it is recommended that thiscombination be retained.

For maximum output, the 2N554's (orsubstitutions) must be matched to eachother and to the load. If the output tran-sistors are unbalanced, one will run warmwhile the other remains relatively cool. Therecommended transformers provide theproper impedance match either to a speakeror to a 20 -watt class C final r.f. amplifier.

2.14 A Kit10 -Watt Modulator

Figure 2.14-A

THE HEATHK1T PUBLIC ADDRESS AMPLIFIER MODEL AA -80

AUXILIARYINPUT

+I Go50 1JFD

T1

+IC9 +IC8Rs

0 KIs V. 1 -50 LIFO R7215V.50 .1.1F D

2.7

R43680

RI R2 C3500 K 100K .11.1FDAUXGAIN "

CI40.UFD +

25V.

R6[4.7K

R922K

Q, CS 02s0

.6 101/FD vill48.4

11;7047 e1

Csswot,

Rs R7 - Ca RI° R e13 R,s - C7"C6.80 10 +TOLIF 4.7 K

.n.470 3.3K 2211 -1

V.17501./FD.

15

R,46.6K

Q32

.1%W4

1180

2 2

R,6

RIB33011

as

Q4

TO 8 OR160. SPKR

MALECONN.

72

c.2r;I LIFO

CHASSIS ACGROUND

R3MIC. GAIN

r -DYNAMICMICROPHONE

PUSH -TO-TALK 1

-I SHIELD

`(YELLOW

J

PV.

ON

OFF

3 ASLO-BLO

* PAIR MATCHED FOR POWER GA/N.** Q3 MUST BE INSULATED FROM SUBCHASSIS.

1. Q I, Q2 2N 4082. Q3,Q4, Q5 CDT 13373. ALL VOLTAGES ARE MEASURED /N RESPECT TO THE

POSITIVE BUS BAR USING A HEATHMIT MODEL MM -I VOM.4. R,7 IS A 2 OHM THERMISTOR USED FOR TEMPERATURE COMPENSATION.5. VOLTAGES SHOWN ARE AVERAGE READINGS OBTAINED US/NG A 72 V

BATTERY AND WITH NO SIGNAL APPLIED TO THE AMPLIFIER.

sn

rc,

IpF

L COM.FEM.

CONN'S.

KI

12-I5VDC

MALE FEM.CONNECTORS K2

Radio Handbook AA -80 Amplifier 51

HEATHKIT AA -80 PUBLIC ADDRESS AMPLIFIER

Top view showing the layout of components.

Those who like to integrate commercialkits into their equipment will be interestedin the Heathkit AA -80 10 -watt Public AddressAmplifier. It can be used to modulate a 20 -

watt transistor transmitter with no modi-fications. In addition, it contains interest-ing circuits which can be adapted to exist-ing equipment.

The circuit diagram of the AA -80 isshown in figure 2.14-A. No input impe-dance matching stage is included since thisunit is designed for use with a low -impe-dance dynamic microphone. Impedance

matching is required, however, between thisstage and the very low base impedance ofthe driver transistor, Q3. The driver stagesupplies sufficient power to swing the classB output stage into saturation.

The circuit can be duplicated easily byexperimenters. Transformer T1 is similarto the Triad TY-61X, while T2 is almostidentical to the Triad TY-29X. An experi-menter's transistor, such as the 2N234A,2N307, or 2N554, is suitable for Q3. Amatched pair of transistors, rated at 10


FRONT VIEW OF HEATHKIT AA -80 PUBLIC ADDRESS AMPLIFIER

Amplifier can be used as a modulator with no modifications.

watts or more, can be used in place of Q4and Q.

In transistor -transmitter applications,the 16- ohm connection would drive the finalamplifier, while the 8 -ohm winding wouldsupply modulation power to the transistordriver stage. This technique is describedin more detail in Chapter 5.

2.15 A 5 -WattClass A Amplifier

When battery consumption and outputstage efficiency are not considerations, con-structors may wish to use a class A ampli-fier in receivers and modulators. Theaver-

age collector current of a class A stage isconstant. Therefore, audio excursions areless likely to frequency -modulate the localoscillator in a communications receiver.

The circuit shown in figure 2.15-A wasnot constructed or duplicated, but appearedin the Pbilco Application Report #309, a de-pendable source of construction informa-tion. The input impedance of this ampli-fier is 100 ohms, and at 5 watts output ithas a total distortion of 5.4% (combined2nd and 3rd harmonic). The 3 db band-width is 110 cycles to 14 kc. The 39 -ohmresistor across the secondary of the drivertransformer is necessary to reduce distor-tion at high levels of power output. Thisdistortion occurs because the input impe-

Radio Handbook 5 -Watt Amplifier 53

dance of the T-1041 output transistor risessharply when the collector current is re-duced to very low levels. The effect causesa peaking of the output waveform on one-half of the audio cycle. At output levelsbelow 4 watts, the effect of this resistor isnegligible. The 0.27 -ohm resistor in theemitter of the T-1041 output stage is neces-sary to insure thermal stability of the unit.The prototype of this amplifier employeda U.T.C. type LS -33 five -to -one interstagetransformer and a Freed type QGA-37 forthe 2:1 output transformer. It would ap-pear that the more common Stancor TA -48and TA -50 would make excellent substi-tutes for these transformers. The outputtransistor, which is not commonly available,can be replaced with a 2N234A (Bendix),2N301 (RCA), or 2N554 (Motorola).

2.16 SlidingBias Amplifiers

The 5 -watt amplifier just describeddraws a larger amount of collector currentsince it is biased for class A operation. Thismeans the transistor must dissipate largeamounts of power in the form of heat.

CENTERLAB AUDIO AMPLI-FIER

This Centerlab audio amplifier, whichis about the diameter of a dime, con-tains 4 transistors and associated com-ponents. It features 80 db of voltage

gain.

A more efficient "pseudo -class A" am-plifier is shown in figure 2.16-A. Thistrick circuit, developed by Phillips of Hol-land, is called a slidhig-bras amplifier. Someof the audio output signal is rectified andfed back to the transistor in the form of ad-ditional forward bias. Thus, at higher audiolevels, the forward bias increases, permit-

-12 V

INPUT

100

Figure 2.15-AA 5 -WATT CLASS A AMPLIFIER

Transformer commercial equivalentsare given in the text.

500

500 -I2V 27

1000 I T -13

-13

LOAD

3.6

54 Audio Amplifiers

ting the transistor to handle the larger am-plitude signal with a minimum of distor-tion. The bias, and therefore the operatingpoint, slides up and down thecharacteristicslope of the dynamic curve in proportionto the signal amplitude. The sliding -biasamplifier is much more efficient than classA for it consumes only enough power fromthe battery for proper operation. In thisregard it is similar to the class B ampli-fier. This circuit produces a higher level ofdistortion products than the class A or Bsystems.

In figure 2.16-A the rectified audio volt-age adds to the bias. Thus both the biasvalue and amount of audio must be ad-justed for best performance. A 2N301 orany experimenter power transistor can beused in place of the OC-16. Note that thephasing of the output transformer is im-portant for proper operation.

Figure 2.16-B shows an adaptation ofthe Phillips circuit which is used in the Auto-crat (New Zealand) car radio. A separatewinding is used for the sliding -bias sourcepermitting the speaker to be grounded.Note that the phasing of the bias windingis important. It should be pointed out thatthe sliding -bias circuit should always beadjusted with the aid of an oscilloscope andaudio oscillator.

0C-16

2200T.

Figure 2.16-ASLIDING BIAS CIRCUIT

The switch Si adjusts the point at whichrectification occurs.

2N301

-12 V. +

Figure 2.16-BA SECOND SLIDING BIAS CIRCUITPrinciple is similar to that of Fig.

2.16-A.

CHAPTER THREE

RF Circuits

In principle there is very little differencebetween an audio and an r.f. amplifier. Thecomments made in Chapter 2 regardingbias, stability and so on, are equally appli-cable to r.f. circuits.

3.1 Alpha Cutoff

One factor to consider when dealingwith r.f. circuits is the alpha cutoff frequencyof the transistor. No doubt you have seenthis term in technical articles and transis-tor literature. What exactly is this myster-ious attribute which separates sometransis-tors from other transistors?

Transistors, like vacuum tubes, exhibitreduced performance as the operating fre-quency is increased. There are two majorreasons for this. Like the tube, the currentcarriers of the transistor (whether holes orelectrons) take time to traverse the regionbetween the emitter and collector. This isknown as the transit time. When the lengthof time is appreciable in relation to theperiod of a full cycle at the frequency beingamplified, the high frequency performanceof the transistor is impaired.

High frequency performance is alsolimited by the base -emitter capacitance andthe base resistance, Rb, which is in serieswith the source. These two componentsform a 'built-in' phase -shift network whichcannot be compensated for externally. Theeffect of the capacitance can be minimizedsomewhat by reducing the source imped-ance.

If a transistor is driven with a 1,000cycle audio signal, the circuit will exhibita certain amount of gain (expressed indecibels). If the applied frequency is thenincreased, a decrease in the output voltagewill be noted. At some frequency the out-put of the amplifier will have decreased by3 db from the signal level at 1,000 cycles.This is known as the alpha cutoff frequency.However, this is not an absolute frequencyabove which the transistor quits working.Transistors will work well beyond the alphacutoff frequency, but above this point theoutput declines. For this reason a transistorshould not be used in r.f. amplifier cir-cuits much beyond alpha cutoff. A trans-istor will usually oscillate up to 4 or 5times its alpha frequency.

R.f. transistors are constructed speciallyto minimize the transit time and junctioncapacitance. The construction of thesevarious transistor types is discussed undersection 1.10.

56 RF Circuits The Transistor

3.2 Impedance Matching

As stated earlier, r.f. and audio amplifiers are essentially the same. An r.f. am-

plifier has a tuned circuit in place of the audio transformer. It is necessary to select

a transistor whose alpha cutoff will allow efficient amplification at the frequency concerned.

As with the audio amplifiers, an imped-

ance match must be provided to obtain the

greatest possible gain. In r.f. circuits, matching also prevents the low impedance

of the transistor from loading the tuned circuit, which would cause poor selectiv-

ity in addition to reduced gain.

Transformer Coupling from one trans - Coupling istor to another may read-

ily b e accomplished b y auto -transformer action. This is shown in Fig. 3.2-A.

The collector of the driver transistor tapped down the tuned circuit to reduce

loading, and the base of the following stage is fed from a link winding. Instead of a

link, the base may be tapped on the coil through a suitable coupling capacitor. The

advantage of the link is that the bias vol-

tage is fed in at the bottom of the winding, whereas in the tapped coil arrangement the

bias components are shunted across part of the tuned circuit. This reduces both gain

and selectivity. If a double -tuned trans -

Figure 3.2-A Link coupling provides an interstage

impedance match between transistor stages.

Figure 3.2-B

A capacitive voltage divider will also

match collector to base.

former is used, the base may be tapped down on the secondary and the bottom of

the secondary returned to the bias circuit. Common step-down figures range be-

tween 25,000 ohms and 5,000 ohms and from 100,000 ohms to 600 ohms. The

primary impedance is determined by the

output impedance of the collector of the driving transistor. If the preceding stage

is an r.f. mixer, its collector current will often be low, and the impedance will thus

be high. It is common to use an i.f. trans- a primary of

100,000 ohms following a mixer. If the secondary of the transformer feeds the base

of a following stage, the impedance will

be lower than if the secondary feeds a diode detector. The two quoted figures, 600 and

5,000 ohms are typical for a transistor base and a diode detector, respectively.

Capacitive The system of obtaining a ca - Coupling pacitive step-down impedance

ratio is shown in Fig. 3.2-B. This circuit will be recognized as being

similar to the pi coupler so often used in the

output stages of transmitters. If C2 and C

2 have the same capacitance value, the junction of the two will be equivalent to a

coil center tap. As C 2

is made larger and C smaller (to maintain resonance), the

junction of the two assumes a lower impedance. If the following stage requires

a very low impedance source (such as a grounded -base stage), it may often be more

easily obtained from such a capacitive step- down system. It may be impractical to pro-

Radio Handbook Neutralization 57

vide the very few turns, or part of a turn, which are required in the inductive match-

ing system in Fig. 3.2-A. This may be particularly true at very high frequencies.

Which The inductive system is gener- System? ally used at the lower radio fre-

quencies because it is simple and efficient. At the higher frequencies, espe-

cially in switched circuits, the capacitive

system wins favor. Both systems are excellent when used in the proper places.

They are, in fact, similar. Both accomplish the same job

- impedance transformation.

Tuned Circuits In some circuits it will be Without noticed that the base of

Step Down the following stage may be connected to the 'top'

of the previous tuned circuit (Fig. 3.2-C). This may be done where a very low Q circuit exists and selectivity is not required.

Though it is possible to raise the Q of the circuit by using a very high C to L ratio,

it is preferable to either tap the tuned circuit or provide a low impedance winding

coupled to the coil.

3.3 Neutralization

The triode tube and the transistor have

one thing in common. They each have internal capacitance between the input and

the output terminals through which high

Figure 3.3-A

Many commercial I.F. transformers

(T) provide a tap for neutralization voltage.

frequency energy may flow from the output to the input. This feedback causes circuit

oscillation when it is neither required nor desired.

Early tube experimenters found that oscillation could be prevented by feeding back

an out -of -phase component via an external path between the output and input circuit. This technique is called neutralization.

Neutralization may be applied to transistors as well as to vacuum tubes. In Fig. 3.3-A transformer T is tapped, and the lower end is connected back to the base

through the neutralization capacitor NC. The phase of the fed -back energy will be the opposite of that which has passed

through the transistor itself, and neutralization will be accomplished. The neutralization capacitor will be small, for in this

case the source feedback voltage is quite

Figure 3.2-C

When the transistor capacitance is

high the internal junction capacitance

may be used in place of C2.

Figure 3.3-B The neutralization voltage may be

taken from the link secondary instead of a primary tap.


high. If, however, the tap is moved down the coil, the capacitor NC will need to be

increased in value to obtain the same feedback voltage.

Instead of providing a special primary tap for this purpose, we may utilize thebase

winding which is a low impedance secondary. NC must then be increased in value

accordingly. This circuit is shown in Fig. 3.3-B. To calculate NC you need to know

the value of the capacitance between the base and the collector. This value is then

multiplied by the turns ratio of the transformer. For example, if the transformer

impedance is 25K to 1K the ratio will be:

5 1K

NC will therefore be five times the base -

collector capacity. The base -collector capacity may be found in the transistor data

provided by the manufacturer.

Unilateralization The path between the base and the emitter

may contain some resistance as well as capacitance. The combination of resistance

and capacitance imparts to the feedback energy a phase shift which will not be com-

pletely cancelled out by the energy that is deliberately fed back through NC. The in-

complete cancellation may not lead to oscillation unless the amplifier has very high

gain. It will, however, cause the base to exhibit a different impedance which will

then lead to loss of gain through impedance mismatch. The inclusion of a resistor in

series with NC will correct for the internal phaseshift, and the amplifier will act as it

was designed to act. This is known as unilateralization.

Not all transistors require neutralization or unilateralization. Many modern tran-

sistors have a very low internal base -collector capacitance and insufficient feedback takes place to cause oscillation. This is

2N1742

T.

Li = 3 r 418 ENA IA CLOSE WOUND ON BRASS SLUG 1/ 4" COIL FORM.

L2= IT #78 ENAM. WOUND AT COLD END OF L I .

L3 =6T. # 14, 3/8" LD., 7/6 LONG. -10 V. TAP AT AT. FROM COLLECTOR END. OUTPUT TAP AT 2.5 T. FROM COLLECTOR END.

Figure 3.3-C

A PRACTICAL EXAMPLE

OF A NEUTRALIZED R.F. AMPLIFIER

This 220 Mc. preselector has a power gain of 14 db and a noise figure of

5.5 db. The nominal bandwidth is 20

Mc. into a 3.5K load.

particularly so at the lower broadcast and intermediate frequencies. Among tran-

sistors of this nature are the Philco MADT types, the RCA Drift series, the mesa tran-

sistor, and many others. However, some circuits using these transistors do call for

neutralization or unilateralization. It is best to follow the specifications of the de-

signer.

3.4 RF and IF

Amplifier Stability

Because of low transitor impedances there is not as much tendency toward os-

cillation caused by coupling between components as there is with tube circuitry.

But one may still get coupling between the transistor elements through the ex-

ternal circuits. Adequate decoupling and bypassing must be used to keep the cir-

cuit stable. Capacitors used to bypass base and collector circuits should be con-

nected back to the emitter as shown in Fig. 3.4-A. These components are des-

ignated C, and C3. If returned to ground, the path back to the emitter would be

Radio Handbook Detection 59

SATELLITE TRANSMITTER This transmitter uses an overtone oscillator and pushpull amplifier, similar to

the circuits discussed in this chapter. (Photo courtesy of DuKane)

through C1. Capacitor C i

will have some reactance and this is often sufficient to con-

vert the amplifier into an excellent oscillator! Similarly, oscillation may take place

as a result of the common power supply impedance between stages, and adequate decoupling is necessary to prevent this.

These components are C2 and R1 in Fig. 3.4-A. With tube circuitry in particular,

the value of R1 is not critical. In trans-

AGC a -

Figure 3.4-A

For best circuit stability, bypass capacitors should be returned

to the emitter.

istor circuits, however, R1 should be chosen with care, for it may well be part of a for-

ward a.g.c. system (as explained later), and its value might be critical.

Decoupling capacitors for transistor circuitry may be large by tube circuitry standards. Typical values may be as high as

1,000 pfd. Usually, the decoupling resistor has a very low resistance value because

a few milliamperes through a decoupling resistor that is too large might well drop

half the power supply voltage. To obtain sufficiently long time constants with low impedance circuits, the decoupling capaci-

tance by necessity must be increased.

3.5 Detection

In Fig. 3.5-A, r.f. is applied to the diode and is rectified. Positive portions of the r.f. wave are able to flow through the

diode, but negative portions are rejected.


RF IN. C C

0

TO AUDIO

Figure 3.5-A A SIMPLE DIODE DETECTOR

Commonly used in transistor radios.

The resultant pulsating d.c. voltage will flow through the gain control. R.f. is sep-

arated from the audio by capacitor C1, which stores the d.c. component.

Fig. 3.5-B is a common -emitter transistor detector. Its operation is similar to that of the diode detector. Here the diode

is the base -emitter junction, and ft, is the diode load. Pulsating d.c. current flowing

through the junction and through the resistor R1 causes the current in the collector

circuit to vary. The transistor is therefore acting not only as a detectorbut as an audio

amplifier as well. Several receivers using this type of detector are to be found in this

handbook.

The Product Detector

The product detector (Fig. 3.5-C) is a special form of

detector, designed to mix a

carrier with the received signal as well as to detect it. The product detector may be

regarded as a mixer. It is operated on a curved portion of the transistor's dynamic

characteristic curve. Mixing between a sig-

(-0 AUDIO OUT.

a -

Figure 3.5-B

A TRANSISTOR IN FINITE -IMP EDANC E

DETECTOR

nal and the local carrier occurs, and side -

bands are created. Assume the b.f.o. is tuned to 455 kc. and a signal at 456 kc. is

fed to the base circuit. The two sidebands will be 1 kc. (456 kc. minus 455 kc.) and 911 kc. (455 kc. plus 456 kc.). The ca-

pacitor C from the collector circuit to ground bypasses the unwanted image (the

911 kc. component) and the 1 kc. side -

band is passed to the audio amplifier for further amplification. The b.f.o. voltage

is fed to the product detector via the emitter. Correct product detector action is

indicated if the b.f.o. is considerably detuned and audio output ceases. If there is audio

output it indicates that the detector is either overloaded or is not functioning correctly.

If the b.f.o. is turned off, audio output will

appear because the transistor biasing condition is upset. The product detector is

actually functioning as a common -emitter detector of the kind shown in Fig. 3.5-B.

Figure 3.5-C

A TRANSISTOR PRODUCT DETECTOR

The mixing signal is injected into the emitter circuit.

3.6 The Mixer and

The Converter

Generally a mixer is considered to be a transistor in which two signal components

are mixed. For example, these might be a signal from the antenna and the energy from

the local oscillator. The resultant, which is either the sum or difference of the two input

Radio Handbook Converter 61

ANT. COIL I

. F. T.

L

OSCILLATOR COIL

Figure 3.6-A

BIAS B-

A TRANSISTOR CONVERTER. WITH

BASE INJECTION

frequencies, is recovered from the collector circuit. The local oscillation usually takes

place in another transistor. A converter

stage is similar to the above, but the transistor is its own oscillator. These terms are often confused. A converter is often called

a mixer, and vice -versa. The mixer was explained in the preceeding paragraph on product detectors. The only difference here

is that the output frequency is now at the intermediate frequency instead of in the

audio range. The external oscillator may be coupled to the mixer in one of three

ways: through the base circuit, through the emitter circuit (Fig. 3.5-C), or through the collector circuit. All three systems are

effective. The two methods generally used

are base injection and emitter injection. The converter is similar to the old autodyne

circuit in which the triode tube was made to oscillate in the cathode circuit and a sig-

nal was injected into the grid. In Fig. 3.6-A the signal is injected into

the base, and the intermediate frequency is taken from the collector. The oscillator tuned circuit is also in the collector lead,

and the feedback winding is in the base. If the LE transformer is considered for a

moment as a very low resistance, it will be

seen that it joins the oscillator coil directly to the collector. Local oscillators will be

discussed later in this chapter. Fig. 3.6-B show the circuit of a differ-

ent converter configuration. The oscillator -

THE KNIGHT

5 -TRANSISTOR RADIO The Knight 5 -transistor portable uses

a converter circuit similar to fig. 3.6-B

(Photo courtesy of Allied Radio)

tuned circuit is in the emitter circuit, and the feedback winding is connected to the

collector. The very low impedance of the emitter necessitates that it be coupled to the

oscillator coil through a link. In some circuits a tap is made on the oscillator coil.

Both systems are used, and the authors are not able to state that one is better than the

other. In general, the adjustment of converters is more critical than that of mixers.

Converters are popular in portable radios because of the saving in space and compon-

ents. Coupled to large antennas, converters are more likely to generate spurious prod-

ucts called 'birdies', and in areas near strong stations these might be quite troublesome.

Figure 3.6-B

A TRANSISTOR CONVERTER WITH

EMITTER INJECTION


3.7 Automatic Gain Control

As discussed in section 3.5, a d.c. vol-

tage is recovered from the incoming signal in the process of detection. The stronger

the signal the greater is the voltage across the load resistor. This voltage may be used

to automatically control the signal level in the r.f. and i.f. amplifiers. There are two

ways whereby the gain ofan r.f. or i.f. transistor amplifier may be controlled. Most

transistors exhibit maximum gain at an emitter current of about 1 ma. If the emit-

ter current is reduced, the gain will also be reduced.

The simplest way to reduce the emitter current is to reduce the bias. In Fig. 3.7-A,

TR 1

is biased to have an emitter current of approximately 1 ma. A strong signal is

rectified by the diode. A positive d.c. vol-

tage is built up across the audio gain control which, in this instance, is the load re-

sistor. This voltage is fed back to the base of the i.f. amplifier stage or stages, reducing

the standing bias and thus the collector current. At the same time, since the bias con-

ditions on the i.f. transistor have been changed, the impedance matching between

the i.f. transformer and the transistor has been upset, bringing about a further re-

duction in gain. This system may be ex -

I. F.T. TRi I.F.T.

AUDIO OUT.

(.0 8.UF

10K

Figure 3.7-A I.F. STAGE

Controlled by diode detector agc.

FLO AUDIO OUT.

Figure 3.7-B Amplified agc. may be obtained from

the infinite impedance detector as shown here.

tended further by not only applying it to two i.f. stages, but by putting a resistor in the lead marked 'X' in only oneofthe stages.

A circuit may use the voltage drop across this resistor to forward bias a diode which

is connected in parallel with one of the tuned circuits. This diode causes a con-

siderable loss in sensitivity, thereby providing additional effective a.g.c. action.

The gain of a transistor may also be controlled by varying its collector voltage.

This may be done in two ways. If a resistor is inserted at the point marked `X' in

Fig. 3.7-A and a negative voltage is fed along the a.g.c. line from the diode detect-

or, the transistor will react by drawing heavy collector current which will, in turn, cause

a voltage drop across resistor X. The collector is thus deprived of voltage and the

gain is reduced. This is known as forward

a.g.c. The previous method is known as

reverse a.g.c. The negative bias required for forward a.g.c. is obtained simply by reversing the diode in the detector circuit.

The transistor detector of Fig. 3.5-B may also be used to supply an amplifuda.g.c.

voltage. When a signal is rectified it causes a d.c. voltage to build up across the load

resistor R1. (See section 3.5.) This voltage causes an increase in collector current.

Before the arrival of a signal the transistor is biased to almost cutoff, and the potential at the collector is about equal to the supply

voltage. An increase in collector current

Radio Handbook Oscillators 63

will cause a voltage drop across the collect-

or load which will swing more positive. This positive -going voltage may be used to

reduce the negative bias on the i.f. amplifier stage, or stages, and thus reduce

the system gain. This is the reverse a.g.c. system again.

These are the main systems in use. Most others are related to those already de-

scribed, and a little analysis will soon show

just what the relationship is.

3.8 Self -Excited Oscillators

Transitor oscillators are similar to tube oscillators. The differences are brought

about by the differing impedances and capacitances common to transistor circuitry.

The capacitances which exist between the transistor elements are variable and will be changed by temperature and supply vol-

tage. Whereas a one -degree increase in temperature is a small increasewhen compared

with the internal heat of a tube, it is a large increase when compared with the internal

heat of a transistor. Oscillators work on one principle and

one principle only. A portion of the r.f.

output of a transistor is fed back into the input where it is amplified, fed back to the input, amplified, and so on. The process builds up until it is limited by the loss in

the circuit. Proper oscillator design suggests that

the tuned circuit be as loosely coupled as possible to the tube or transistor. This is

even more important with transistors than with tubes because of the higher input and

output capacitance of the transistor. Pru- dent designers introduce as much external

capacitance into the tuned circuit as the oscillator will permit. Assume that the trans-

istor has an input capacity of 100Nifd. and the transistor base is connected to a resonant circuit which is tuned with a 100 mid.

capacitor. Across the coil will be a total of

Figure 3.8-A A tickler coil oscillator

with feedback between collector and base.

200 mfd., 50% of which is made up of transistor input capacity. The emitter -base

capacitance is subject to variation. If, therefore, the external capacitance is increased

to 1 ,000Rifd. ,

thetotal capacitance will then be 1,100 Nefd., of which less than 10% is

made up of the base -emitter capacitance. Obviously, variations in base -emitter capac-

itance will now have less effect upon the oscillator stability. Too much external C will

prevent oscillation altogether, however.

The Tickler This "old-time" circuit will

Coil be familiar to most readers. Oscillator The tube has merely been

replaced by a transistor. In Fig. 3.8-A, r.f. is taken from the collector and fed back to the coil in the base circuit. The vacuum tube's usual grid leak and

capacitor should not be used with the transistor. Bias is very easily obtained with the

forward bias resistors in the normal manner. If desired, the tuned circuit may be

Figure 3.8-B

A tickler coil oscillator with feedback between emitter

and collector.


connected to the collector and the tickler

coil to the base. Referring to Fig. 3.8-A, the tuned circuit is damped by the low base

impedance, and unless the base is tapped down the coil to reduce loading, the transistor may refuse to oscillate. Alternately,

a capacitive divider may be placed across the coil and the junction connected to the

base. This was discussed under r.f. amplifiers in section 3.2.

Fig. 3.8-B shows the tuned circuit in the collector lead and the feedback winding

connected to the emitter. To further reduce loading, the collector may be tapped

down the coil.

The Colpitts Oscillator

The similarity between the tube and the transistor Col-

pitts circuits, Fig. 3.8-C, will be apparent at a glance. The base -emitter

capacitance is effectively in parallel with C2. Thus C2 should be made large to reduce

the influence of the transistor capacitance upon the tuned circuit. Capacitor C1should

also be made large to reduce the effects of the collector capacitance upon the tuned cir-

cuit. The collector capacitance changes in value with variations in supply voltage. In

general, a high C -low L circuit is required to maintain a stable oscillation frequency.

RFC

Figure 3.8-C

THE COLPITTS OSCILLATOR

The Clapp Oscillator

A popular version of the Col- pitts is the Clapp circuit illus-

trated in Fig. 3.8-D. In this oscillator the parallel tuned circuit is re-

placed with a series -parallel tuned circuit. The remarks concerning the size of CI and

Figure 3.8-D THE CLAPP OSCILLATOR

C 2

above also apply to the Clapp oscillator. Note that that ground point has been

changed in this oscillator. In transistor circuitry there is no good reason why any

one point should not be connected to ground. One may even ground the collect-

or if desired. Obviously, the base and emitter lines would not be grounded at the

same time! Resistor R1 in Fig. 3.8-D may be a thermistor ifbetter stability is required.

Fluctuating temperatures cause a resistance change in the thermistor which in turn causes a compensating change in base cur-

ent.

The Hartley The Hartley, too, is an old Oscillator favorite circuit. Little need

be said except that the collector and the base, as with the previously

discussed oscillators, may be tapped down the coil to obtain better isolation from the

tuned circuit. The oscillator may be series or parallel fed.

The Ultra In tube circuitry, the Ultra Audion Audion, Fig. 3.8-E, is an os-

scillator cillator similar to the Colpitts, but no capacitive center tap is

made to the tuned circuit. The internal capacitances of the transistor (collector -to -

base and base -to -emitter) are the equivalent of C and C

2 of the Colpitts oscillator (Fig.

3.8- C ).

Fig. 3.8-F is a form of Ultra Audion oscillator in which the tuned circuit is con-

nected between the collector and the base, which is at the r.f. ground potential. C,

and C 2

(the latter representing the emitter -

to -base junction capacitance) form a capac-

Radio Handbook Oscillators 65

Figure 3.8-E

THE ULTRA-AUDION OSCILLATOR

itive voltage divider connected across the coil L1. At the lower frequencies it becomes

necessary to place a capacitor in parallel with C

2 to sustain oscillation. In the in-

terests of stability, C 1

and the external capacitor C, should have as large a value as

possible. This circuit has been extremely useful in grid dip meters.

RC Phaseshift In this oscillator, Fig. Oscillator 3.8-G, output is taken

from the collector, passed through the phase shift network N, and fed

back to the base to sustain oscillation. The

vs, OUT.

Figure 3.8-G A PHASE -SHIFT AUDIO

OSCILLATOR

time constant of the network components determines the frequency of oscillation.

This oscillator is mainly used at audio frequencies. For proper operation, a high

beta transistor should be employed in conjunction with the highest permissible col-

lector supply potential in order to overcome the losses in the phaseshift network.

The Multi- The multivibrator (Fig. 3.8-H) vibrator is an arrangement of two tran-

sistors wherein the output of one transistor is fed to the input of the

other. The output of the second transistor

Figure 3.8-F

AN ULTRA-AUDION OSCILLATOR This oscillator uses the base -emitter

capacitance of the transistor to complete the tank circuit.

is then returned back to the input of the first. Once started by inherent circuit im-

balance, the multivibrator rapidly builds up into a state of oscillation. Multivibrators may produce a variety of waveforms and

are especially useful for generating square and sawtoothed waveforms. The multi

- vibrator finds special application in test

equipment, TV transmitters and receivers, and a multitude of calculating machines. A

host of configurations have been evolved, each with different or special properties

which are far too numerous to describe here. However, multivibrators may be divided into two classes, those which require

a triggering pulse and those which do not. The oscillator which requires no pulse to start operation is said to be five running. If

the oscillator has one transistor which needs to be pulsed into operation, it is said to

have one stable element. If the oscillator requires that both transistors be pulsed, it

TRI, TRz 2N408, 2811380,

0071, ETC.

Figure 3.8-H

A SQUARE -WAVE GENERATOR (MULTIVIBRATOR)

jig


TRANSISTOR

CODE PRACTICE

OSCILLATOR

KEY

THE KNIGHT CODE PRACTICE OSCILLATOR

This oscillator is a simple tickler coil audio oscillator similar to Fig. 3.8-A.

(Photo by Allied Radio)

is said to be bi-stable. A trigger pulse fed

to the base of one transistor may cause it

to flip from a saturated condition to a biased -off condition, where it rests until an-

other pulse flops it back on again. Hence the term flip-flop.

The Tunnel The tunnel diode operates on Diode

transistor. exhibits a

an entirely different principle from either the tube or the Briefly, it is a conductor which negative resistance. Unlike a

piece of copper wire, through which a rise in voltage will cause a corresponding rise in current, in a tunnel diode a rise in volt-

age will cause the current to decrease. This is the state (negative resistance) required to begin and sustain oscillation. The cir-

R.F. OUT.

Figure 3.8-I

A TUNNEL DIODE R.F. OSCILLATOR

cuit of an oscillator using a tunnel diode is shown in Fig. 3.8-I. It is likely that the

tunnel diode will greatly benefit radio as

we know it. This is particularly true of the ultra -high frequencies, for the relatively low noise figure places the tunnel diode

second only to the parametric amplifier in weak signal response.

3.9 Crystal Controlled Oscillators

A crystal may replace the tuned circuit in most of the L/C oscillators described in

the previous section. Even the tickler feedback type of oscillator may be adapted to

use a crystal. The frequency of oscillation is now determined almost wholly by the

crystal. Capacitance in the circuit will have

some effect on the frequency of operation and a capacitor may be provided as a means

of vernier frequency control. Fig. 3.9-A shows the circuit of the tickler feedback

Figure 3.9-D Another oscillator of the Colpitts family, with thefrequency determining crystal in series with thef eedback path.

Radio Handbook Crystal Controlled Oscillators 67

Figure 3.9-A A TICKLER COIL

CRYSTAL CONTROLLED

OSCILLATOR

Figure 3.9-B

THE COLPITTS CRYSTAL

OSCILLATOR

Figure 3.9-C A MODIFIED COLPITTS

CRYSTAL -CONTROLLED OSCILLATOR

type of transistor oscillator with the crystal in the feedback path. The frequency of

operation is determined by the series resonant frequency of the crystal. Better stability will

be obtained if the collector is tapped down the coil. L and C are tuned to the same

frequency as the crystal. The oscillator circuit shown in Fig. 3.

9-B is a Colpitts. The coil has been replaced by the crystal operating in the paral-

lel resonant mode. Fig. 3.9-C is similar to the Colpitts and

is a popular transistor oscillator. Capaci-

tor C1 may be used to control the feedback.

Another popular oscillator which also belongs to the Colpitts family is that shown

in Fig. 3.9-D. The transistor operates in grounded -base configuration. Grounded

-

THE HEATH GC -1A TRANSISTORIZED

COMMUNICATIONS RECEIVER This receiver uses many of the circuits

discussed in this chapter. (Photo by Heath)


L 2N247, 00515, ETC.

Figure 3.9-E OVERTONE OSCILLATOR

If the crystal capacitance is properly neutralized, the circuit may be used

with fifth overtone crystals.

base operation extends the frequencies at which the transistor will operate and is preferred for the higher frequencies. If C

1

is

too large, the oscillator may super -regenerate

or squegg. If C1 is too small, the transistor will refuse to oscillate. The stability of this

oscillator is excellent when properly adjusted.

Overtone Fig. 3.9-E shows the circuit of Oscillator an overtone oscillator in which

the crystal is placed in the feedback path. L1 and C, are tuned to the crys-

tal harmonic frequency.

3.10 Transmitting RF Amplifier

An r.f. amplifier may be operated in class A, B, C, or anypoint in-between. The class of operation is determined entirely by

the bias. Class A amplifiers were discussed in Chapter 2.

Class B The bias on a class B transistor Amplifiers stage is set near the cutoff point.

If the stage is handling a signal with a carrier, the maximum ratings of the stage will be considerably reduced, as the presence of the carrier will cause a high

static collector current. If the class B amplifier is handling s.s.b., the standing cur-

rent is low, but the peaks (on voice) might reach very high figures.

A linear amplifier may amplify a single- sideband signal, but due to the absence of

carrier in the signal, the operating conditions will not be the same as those required

for amplitude modulated amplification.

Class C Class C transistor amplifiers Amplifiers are usually modulated in the

collector circuit. If an amplifier increases the level of a modulated signal, it is usually not operated class C. It

should be operated class A, B, or at some point in-between. Class C operation of an

amplifier clips the r.f. waveform, but the clipped portions are restored by the fly-

wheel action of the tuned circuit. A tuned circuit, however, cannot restore the clipped

portions of a modulated waveform! A class C amplifier will be rich in harmonics which, if not prevented from reaching the antenna,

may cause severe television interference. High power transistor transmitters often have modulated driver stages, followed by

a collector -modulated class C stage. This is done to increase r.f. drive and results in

higher modulation percentages.

Class C amplifiers are biased beyond cutoff. In the amplifier shown in Fig. 3.10-

A, the bias is determined by resistors R1 and R2. The stage is operated with the

base at ground potential for r.f. This type of circuit does not require neutralization.

In the past, the unavailability of suitable r.f. transistors has been a severe check on the development of high power semicon-

ductor transmitters. However, the state-of- the-art has advanced to the point where r.f.

Figure 3.10-A A PUSH-PULL CLASS C

R.F. AMPLIFIER

our.

Radio Handbook Modulators 69

transitors of a type known as triple diffused layer mesa and planar style have collector dis-

sipation figures as high as 125 watts. These transistors are capable of operation to sev-

eral hundred megacycles. It would seem that a new problem of providing suitable

power supplies might now emerge. Transmitting circuits are described in

greater detail in chapter 5.

3.11 Modulating RF Circuits

Modulation may be regarded as a mixing

process, and any of the usual mixer circuits may be used to modulate an r.f. carrier with an audio signal. However, plate or high level mixing ( or modulation) has found

considerable favor in communications applications for a very good reason. R.f. am-

plifiers may be operated class C and thus are much more efficient.

The process of modulation creates side -

bands which are transmitted in addition to the carrier. The carrier itself does not

change in amplitude during the modulation process. The term `amplitudemodulation'

is something of a misnomer, for nothing of the sort takes place unless the amplifier is incorrectly operated or overmodulated.

The High The circuit of Fig. 3.11-A will

Level deliver 12 watts of audio from Modulator the modulation transformer

T2. The secondary of T2 should match the impedance of the ampli-

fier it modulates. Theprimary of T1 should

Figure 3.11-B

LOW LEVEL MODULATOR

match the driver. This is a class B audio amplifier, and its operation has been cov-

ered in Chapter 2.

The Low The low level modulator is Level little different from a conven-

Modulator tional mixer. The audio circuit replaces the r.f. oscillator.

The circuit is shown in Fig. 3.11-B. The modulated stage should be followed by a

linear amplifier.

The Diode The diode balanced modu- Balanc ed lator performs two functions:

Modulator (a) the r.f. is mixed with the modulation and (b) the car-

rier is balanced out. In Fig. 3.11-C, both the carrier and the audio are fed into the

balancing potentiometer P1 .

This control compensates for the unequal characteristics

of the two diodes. Because the diodes are low impedance devices, they load the circuit to which they are connected. As a con -

AUDIO IN.

C

2N301 (z) Tz Mein

Figure 3.11-A

HIGH LEVEL MODULATOR

AUDIO OUT.

Di Li

Figure 3.11-C

DIODE BALANCED MODULATOR

70 RF Circuits

sequence, CI and C2 are made large in an effort to raise the Q of the circuit. A typi-

cal figure for C1 and C2 at 4 mc. is 0.002 ufd. each. The resistance of the diodes

should be matched in the forward direction, and the diodes should have low forward

resistance. Because the impedance of the circuit is low, the diodes need not have a

high reverse resistance.

The A transistor balanced modu- Transistor lator circuit is shown in Fig. Balanced 3.11-D. The action of each

Modulator transistor is similar to that of a mixer. The two transistors

opposing each other balance out the carrier. For better stability, the resistor R

1

should be left unbypassed for audio. The carrier is balanced out by the potentiometer P1.

The impedance of the tuned circuit is considerably higher than that of Fig. 3.11-C,

and the capacitors C1 and C2 may have smaller values.

CARRIER IN.

AUDIO IN.

Figure 3.11-D TRANSISTOR

BALANCED MODULATOR

SIDE BAND OUTPUT

CHAPTER FOUR

Receivers

Several of the construction projectswhich follow are not really of a radio am-ateur nature. However, the authors feelthat the construction of these simple broad-cast radios will be good preparation forthe more elaborate equipment to follow.

Not only quite sim-ple, but the authors have tried to build -inguaranteed performance. The little radioshave been duplicated on both sides of theglobe, using different components, andthey work as advertised.

In getting circuits to work properly,all too often overseas readers are left indoubt regarding the exact description ofcomponents. Component descriptionsneeding clarification are the correct num-ber of turns on coils, turns ratio of trans-formers, and approximate inductive values.Sufficient information of this type is in-cluded here so that reasonable substitu-tions can be made.

4.1 The "Super Regen"Detector

One of the most popular circuits con-structed by amateurs and experimentersalike is the superregenerative detector. Withonly one transistor it is possible to amplify

a radio frequency signal from two or threemillionths of a volt up to headphone vol-ume! Truly, the sup erregen erative detectorcreates a minor miracle!

One of the first high -frequency tran-sistors available was the Philco surface barriertype, and many circuits have been devisedto employ this unit. The Philco SB-100,the AO -1, 2N232, 2N346, and the T-1768

Lt

2 .001

20K HIGH ZPRIMARY

T T+1 5V - + 3V.-

RFC

Tt

1 -COLLECTOR2- BASE3- EMITTER

OZ root

AUDIO OUTPUT

Figure 4.1-AA SUPERREGENERATIVE

DETECTORThis detector uses a surface barriertransistor. Coil Li is a two turn linkplaced at the lower end of 12. CoilL2, for 28 Mc., is 7 turns, #22, 1 inchdiameter, L1/8 inch long. Transform-er T1 is a 10K to 2K transistor inter -

stage transformer.

72 Receivers The Transistor

!' Figure 4.1-BA S EN smv E

SUPERB EGEN ERATIVE DETECTORFor 10 meters, coil Li is 7 turns, #22,1 inch diameter, 1-1/8 inch long. Coil12, the antenna link, is 2 turns at thebottom end of Li . Coil 13, the feed-back winding, is 3 turns at the top endof coil Li . Transformer Ti can be a

10K to 2K transistor interstagetransformer.

all work on the 27 Mc. Citizens' band andthe 10 -meter amateur band. The SB-101,SB-102, SB-103, T-1324, T-1767, and the2N588 all operate as high as 6 meters,with several varieties operating as high as90 Mc. The 2N500 and 2N588 will op-erate in the 2 -meter band. The newer Phil -co MADT devices, such as the 2N1742series, will also work in circuits specifyingsurface barrier transistors.

Fig. 4.1-A shows a simple superregen-erative detector which works well with mostsurface barrier transistors. With thevaluesshown, the receiver will tune both theCitizens' band and 10 meters. The poten-tiometer is adjusted for best sensitivity, asare the antenna coupling coil and theemitter tap.

A more sensitive circuit is shown inFig. 4.1-B. Potentiometer R1 is adjustedfor maximum sensitivity and should bereset for each station. Potentiometer R2 is

also set for maximum sensitivity. How-ever, once the best setting has been found,no further adjustment is required. Capa-citor C1 controls the quench frequency andmay have values between 0.015 /ad. and0.1 /Ad., depending upon the transistorused. This circuit has been operated witha 12 -foot antenna and provides excellentperformance, even without an externalground. Amateur stations on the10-meterband were received from distances as greatas 2,000 miles.

Another superregenerative detector isshown in Fig. 4.1-C. Notice the similarityto Fig. 3.8-F (the Ultra Audion oscillator).This circuit differs from the previous super -regenerative detectors in the method ofantenna coupling. Note that the antennais connected to the emitter through a trim-mer capacitor.

The circuits just described will work on11, 10, and 6 meters. But what about 2meters (144 Mc.)? The circuit shown inFig. 4.1-D was designed by Philco for usewith their 2N1742 transistor for operationbetween 85 and 185 Mc. This includesthe f.m. aircraft, 2 -meter, mobile services,and television bands.

In this circuit the bias control is setfor a collector current of 1.1 ma. Capa-citor C1 is the tuning control, and capa-

Li

AUDIOOUTPUT .7)

JJF

S 131 -ANT.

L

Figure 4.1-CThis circuit will work on the 6 -meteramateur band with the Philco T-1324,T-1657, and 2N299 surface barriertransistors. Coil Li is 5 turns, #16 wire3/8 inch inside diameter, 1 inch long.

Radio Handbook Simplex Crystal Set 73

2 N1742"G/MMICK. 1.0 ULF

C5 25UOFTUN/NC

Hi'.005 .

AUDIOOUTPUT

C24 -30 ILUFREGENER.CONTROL

27

0

2

1 -COLLECTOR2 -BASE3 -EMITTER

Figure 4.1-DTHE SUP ERREG EN ERATIV E

DETECTOR

Shown here is designed for operationon the two -meter band and will workto at least 180 Mc., using the Philco2N1742 MADT transistor. Coil Li is3 turns, #12 self supported, 1/2 inchinside diameter, and 1 inch long. Coil.L2 consists of 18 turns, #30 space woundon a 3/16 inch form, 1/2 inch long.The "gimmick" is a few turns of wire

wound around the collector lead.

citor C2 is adjusted for maximum sensitiv-ity. Although not specifically stated in thePhilco Application Report #505, this circuitcould be moved up to 220 Mc. for use intiny hand-held transceivers.

4.2 Simplex Crystal Set

Excluding the headphones, this crystalset uses four inexpensive components: thecoil, the tuning capacitor, a germaniumdiode, and a 0.005 pfd. fixed capacitor.Yet the circuit offers much in the way ofknowledge and experience to those new tothe fascinating hobby of radio. There ismuch to be learned from the humble crys-tal set. The circuit is shown in Fig. 4.2-A.

REAR VIEW OF THESIMPLEX CRYSTAL SET

How itWorks

The antenna couples a multitudeof signals to the coil and variablecapacitor, which together select

the desired station. The chosen signal isthen rectified by the diode. This meansthat the audio component is recoveredfrom the radio wave and passed to theheadphones. In the headphones the fluc-tuating current (the recovered audio)flowsthrough a coil of wire, creating a mag-netic field. This in turn vibrates a softiron diaphragm. The end result is a move-ment of air which creates sound. It is im-

FRONT VIEW OF THESIMPLEX CRYSTAL SET


PHONES

Figure 4.2-ATHE SIMPLEX CRYSTAL SET

Parts ListC1 - Tuning capacitor, any value be-

tween 300 and 385 Mufd.CR1 - Any germanium diode, such as

1 N34A, 1 N48, 1 N60, etc.Li - One inch closewound winding

of #34 enameled wire on 1-1/4inch form. Tap 1/4 inch fromground end of coil. (J. W. Miller#2004 or similar loopstick mayalso be used.)

portant to realize that the flow of currentmust complete the circuit. It must flowfrom the tuned circuit (coil and capacitor)through the crystal, through the head-phones, and back to the tuned circuit.The direction of flow is governed by theway the crystal is connected. Reversingthe crystal will reverse the direction offlow.The sound, however, will be unaffected bythe reverse connection. The capacitoracross the phones bypasses any r.f. whichmanages to get past the crystal.

Construction The tuning capacitor mayHints be obtained from a discard-

ed broadcast set. The ca-pacitor in the photograph is a two gangunit, which means there are actually twocapacitors on a common shaft. Only oneof the capacitor sections is used.

Don't forget to scrape the enamel in-sulation from the wire whenever a con-nection is made to the coil!

Experiments If the antenna is connectedto Perform near the top of the coil, it

will detune the circuit. Thestations will cover large portions of thedial, one station often interfering with an-other. If the antenna is connected near thebottom of the coil, the selectivity of the cir-cuit will be immensely improved but thevolume will be less. The tap shown hasbeen adjusted to give a compromise be-tween selectivity and gain.

4.3 Simplex AudioPower Pack

A capacitor, a resistor and an inexpen-sive transistor are all the components re-quired to add an amplifier (power pack)to the Simplex Crystal Set. The amplifierwill provide pleasant listening, and even

GROUND

GROUNDED TO PANEL

ANTENNAFigure 4.2-B

CHASSIS LAYOUT OF THESIMPLEX CRYSTAL SET

Note that the phone jack and thetuning capacitor are connected toget-her through the metal panel. If an in-sulated panel is used, a wire will haveto be connected from the phone jack

to the ground post as shown(dotted lines).

Radio Handbook Simplex Audio Amplifier 75

WIRING CONNECTIONS FOR THE SIMPLEX AUDIO POWER PACK

weak stations will be brought up to goodheadphone strength. The circuit is shownin Fig. 3.4-A.

How It In this receiver the crystal diodeWorks is connected to the antenna tap

on the coil. If the diode is con-nected to the top of the coil, the stationswill have a very broad tuning range andwill interfere with each other to a consid-erable extent. The other end of the diodeis connected to the base of the transistor,which amplifies the signal and passes it onto the headphones. The resistor R1places

a small amount of bias on the transistorbase. This bias is necessary for correctoperation.

Construction The transistor may be sol-Hints dered directly into the cir-

cuit, or as an alternative, itmay be plugged into a transistor socket. Ifthe wires are soldered, leave the leads long,but cover them with vinyl tubing topreventshorts. Be careful not to let the heat fromthe iron flow up the leads and ruin the tran-sistor. One end of the crystal will bemarked with a red, white, or black dot or


TR I81,476-220N

+1.510 60 -

PHONES

TRANSISTORBASES

Figure 4.3-ATHE SIMPLEX AUDIO POWER PACK

Adjust value of R i to suit transistorused. The tuned circuit is the same asin Fig. 4.2-A. The transistor is notcritical and a 2N107, 2N109, 2N408,

2N1380, or 0072 may be used.

line. This end connects to the base of thetransistor. If the panel is made of metal itis important that the headphone jack be in-sulated from the panel to avoid shortingout the battery.

The transistor will operate from a dis-carded flashlight cell, but best results willbe obtained from a small 6 -volt battery.

No switch is used in the amplifier cir-cuit because the current drain from the bat-tery is so small it will last the shelf life.

Experiments Try connecting the antennato Perform and the base of the transis-

tor to different taps on thecoil. Sometimes different transistors willwork better on different taps.

Try a different value of R but do not usea value smaller than 47,000 ohms. A lowvalue of R will cause heavy collector cur-rent which may ruin the transistor. Al-

most any type of PNP audio transistorwill work in this ciruit. If you connecta 0-1 ma. meter in series with the battery,resistor R may be adjusted for a currentof approximately 0.25 ma.

4.4 The TR One

This receiver uses only one transistor.Yet is is "a receiver with a difference" be-cause it will give good results with only afraction of a volt of power supply. Even ifthe battery circuit is opened and one leadheld in each hand, allowing the circuit tobe completed through the body, the receiv-er will still give good volume! There is no

point in using more than 1 1/2 volts sup-ply with a receiver of this nature!

Using The TR ONE will give excellentA Solar results when powered with a sin-

Cell gle solar cell. The results arecomparable with those from a

battery, even in the shade. The cell usedwith the prototype receiver was in Interna-tional Rectifier Co. #SAS -M, costing less than$2.00. The receiver works equally wellwhen the cell is under an electric light, sobatteries are not necessary.

How itWorks

The transistor operates withoutbias and is a common -emitter de-tector. Its operation has been

described in Chapter 3. The base to emit-ter junction behaves exactly like the diodein the Simplex Crystal Set, and the remain-der of the transistor acts as an amplifier.

Figure 4.4-ASCHEMATIC DIAGRAM

OF THE TR-ONEThe tuned circuit values are the sameas in Fig. 4.2-A. Almost any transis-tor, except the drift types, may b e used

(see Fig. 4.3-A).

Radio Handbook The TR One 77

ConstructionHints

The same baseboard, pan-el, tuning capacitor, andcoil used in the previous

receivers may be used again. The transis-tor is mounted on a small terminal strip.No special layout is necessary. Completethe receiver before soldering in the transis-tor. It is always agood ideato have a meterin the battery lead so if anything is wrongit will immediately show on the meter. Forexample, if the meter indicates an excessivecurrent after the battery isconnected, some-thing is amiss. Prompt disconnection ofthe battery would likely save the transistor.With this receiver, as with the last, thebattery current is so low that a switch isunnecessary.

Experiments As in the case of the previ-to Perform iously described receivers,

the coil tap may be variedto advantage. It is also interesting to ob-serve the effects of heat on the transistor.Connect a 0 - 1 ma. meter in the minusbattery lead. The current will normallybe very low, possibly between 30 and 150

#a Now hold the transistor allowing heatfrom the fingers to flow into the transistor.Watch the collector current rise rapidly.Note how little time it takes for the heatto affect the current flow. This experimentindicates the reason why, in later receivers,extra resistors are used in heat stabilizingnetworks. Without a means of limitingthe current caused by heat, the transistorcould destroy itself. Imagine how highthe current would rise if the receiver wereleft in a closed auto that had been standingin the hot sun!

The effects of rectification may be ob-served by tuning in a station with the meterstill connected in the B -lead. A strong sta-tion will cause the battery current to riseconsiderably. A weak station will cause itto move only a little. In fact, the meter maybe used to measure the relative strength ofthe stations received.

4.5 The TR Two

A transistor, two resistors, and a capac-itor are the only parts required to add anamplifier to the "TR ONE". Taking advan-

WIRING VIEW OFTHE TR-ONE


tage of the fact that the "TR ONE" will giveexcellent results from a very low batteryvoltage, the detector transistor has beencoupled directly to the amplifier as shownin Fig. 4.5-A.

How It By examining the circuit andWorks ignoring the amplifier for a mo-

ment, it can be seen that the de-tector transistor is supplied with collectorvoltage through resistor R1. Rememberthat in previous experiments when a stationwas tuned in the collector current increased.In this circuit, when the collector currentrises it causes a voltage drop across R1.The changing d.c. voltage is applied tothe base of the transistor amplifier. Thesignal in the headphones is considerablyamplified and will even drive a speakeron strong stations. Resistor R2 is neces-sary to raise the emitter voltage to nearlythe same as the base voltage. The capaci-tor C1 is necessary to provide a path toground for the signal.

Apart from the need to provide anotherterminal strip on which to mount the fewextra components, there is little more tothe construction of this receiver.

TR2Ci TRi AUDIO

DETECTORtie PHONES

0-1

5 TO 6V

Figure 4.5-ATHE TR-TWO

The tuned circuit is the same as inpreceding receivers. The transistorsmay be 2N139, 2N408, 2N1380, GT-

109, or 0072.

Experiments A 0-1 ma. meter insertedto Perform in the battery lead at "X"

will show that heat doesnot affect the current flow in this circuit.The reason for this is simple. When theapplied heat causes a greater flow of cur-rent through resistor R1, the voltage onthe detector decreases so that over-all cur-rent consumption remains the same. Ifthe amplifier transistor is heated, a largeramount of current will flow through R2,decreasing the bias between the emitter andthe base and thus reducing the collector

REAR VIEW OF THETR-TWO

Radio Handbook Regenerative Receiver 79

current. This action, known as temperaturestabilization, is a very important function intransistor circuitry. The total current drawnby the TR TWO is approximately 0.5 ma.at 6 volts.Amplifier If more output is required,

the Mini -Amplifier in Chap-ter 2 may be added to this circuit with ex-cellent results.

4.6 The Solar Two

The SOLAR TWO is an adaptation ofthe TR TWO. This little receiver has beendesigned to operate from one or two In-ternational Rectifier Co. #SA5-M solar cells.In this circuit the values have been changedto suit the high current, low voltage outputof the solar cell. Note that R1 has beenreduced to 10K and that the resistor andcapacitor in the emitter lead of TR2 havebeen eliminated. The SOLAR TWO mustoperate with low impedance headphonessuch as dynamic types. Alternatively, high

LI CI TR'DETECTOR

TR2AUDIO

5.45- AISOLAR CELL

IK 100K

Figure 4.6-ATHE SOLAR TWO

The tuned circuit and transistors arethe same as in preceding receivers.The solar cell is made by InternationalRectifier Corporation and is availableat most jobbers or mail order supplyhouses. Transformer T1 is a 100K highimpedance microphone to 1K transis-tor base transformer, reverse con-nected. For additional output, twosolar cells connected in series could

be used.

impedance headphones may be connectedthrough a step-up transformer to obtainoptimum performance. When the receiveris operated in bright sunlight or under anelectric light its performance is remarkable.

4.7 A RegenerativeReceiver

This is a recommended circuit whetherthe would-be constructor is a young ex-perimenter or an old timer new to tran-sistors. Selectivity is such that two broad-cast stations may be copied alongside eachother without interference. Although ituses only two transistors, the receiver hasan audio output equal to that of many su-perheterodynes which have more transis-tors and additional components. Thehighsensitivity, selectivity, and gain are obtainedthrough the use of regeneration appliedto the detector stage and the incorporationof band-pass tuned circuits.

FRONT VIEW OF THEREGENERATIVE RECEIVER

A National dial was used for ease oftuning and a metal panel eliminateshand capacitive effects. The volumecontrol is to the right of the dial, andthe regeneration control is below thedial. The phone jack is located to the

right of the regeneration control.


REAR VIEW OF THE REGENERATIVE RECEIVER

Note that the coils are at opposite ends of the chassis to prevent any undesiredcoupling. The remainder of the components are mounted on the fiber terminal

strip.

How It A tuned circuit, L1 and C1, isWorks top -coupled via capacitor C3 to

another tuned circuit, L2 and C2.The value of C3 has been carefully chosento give high gain consistent with goodselectivity. Transistor TR1 is a detectorsimilar to that in the TR- ONE receiver.In this case, however, regeneration is ad-ded to increase the gain of the stage. Thismeans that a little of the output has beenreturned to the input of the TR1 via thetickler winding on L2 , causing the signal

to build up to a high value. If the signalis increased too much, the stage oscillates,and the speaker emits sounds akin to apet shop! The audio stage TR2 is a con-ventional amplifier designed to feed intohigh impedance headphones.

Construction The construction may takeHints almost any form. In the

photograph it will be no-ticed that the authors used a 5" x 7" chas-sis to support the tuning capacitor and the

Radio Handbook Super Three 81

ANT. 4 REGENERAricwC3

L, 51111F3

EARTHGROUND

2

34704

Lt 4 5 2 oIIII11111

4 3

ITI

3

strip of phenolic material 'containing mostof the components. A metal panel is an im-portant adjunct to a regenerative receiver toprevent hand capacity effects when the re-ceiver is near oscillation. The coils arewound according to the data given. Notethat it is essential that all the coils be woundin the same direction. That is, winding 4-5on L2 must be in the same direction aswinding 3-1, in order to permit the de-tector to regenerate. If the windings arein the opposite direction, it will be neces-sary to reverse wires 4 and 5 to sustain os-cillation. Keep the coils away from metala distance at least equal to their diameter.

Operation Set the trimmer capacitorson C1 and C2 to about center

and tune a station in with the regenerationcontrol P1 turned down (arm of potenti-ometer at the emitter end). Now adjustthe trimmers one at a time while rockingthe tuning capacitor back and forth. Sev-eral maximum settings will be found, butone setting in particular will be louderthan the others and this setting is the onewhich should be chosen. The alignmentshould be done with the tuning capacitorsomewhere near the center of its range.

A drift, MADT, or other type of highfrequency transistor, will give superior re-sults at TR1. Ordinary i.f. amplifier typesare not recommended. Almost any generalpurpose transistor may be used for TR2.

VOL UME

-8JJF+ TR2-6 8 V.

-F 9 V. -

PHONES

1. 0 MA

ONOFF

Figure 4.7-ASCHEMATIC DIAGRAM OF THE

REGENERATIVE RECEIVERParts List

C1, C2 - Tuning capacitor, 365 ,ufd.per section (Miller #2112).

L1 Primary (1 and 2) 40 turns,#32 enamel. (Use this sizewire for all windings.) Secon-dary (3 and 4) 85 turns, leave1/8" between windings.

12 Tickler (4 and 5) 35 turns.Secondary (1 and 3) 80 turns,tapped at 28 turns from #3end, spaced 1/4" betweenwindings.

TR1 2N247, 2N274, 2N371, 2N-1745, 0C44.

TR2 2N408, 2N1380, GT81R, OC-75.

It will be found that the setting of P1which gives maximum output is just belowthe point at which oscillation commences.This setting is also the point at whichgreatest selectivity occurs. If the receiver isallowed to oscillate, output will be low anddistorted.

4.8 The Super Three

As its name implies, the Super Three isa three -transistor superheterodyne whichwill drive a speaker on strong stationsand will operate from theloopstick antenna


EXTERNAL ANT.

5-Li

all3

2 4

1(70T.

AV5.6K

322K

TR, I FT11-

si

6//

C2 I

12

470 10

331F_ANv2 K

I FTaTRa F-4- -

51

if.01 3

C R TR3

1_00

2,

'

.05

105

+ 61JF-

3:10 K I

8-

R476

ABC

Figure 4.8-ASCHEMATIC DIAGRAM FOR THE SUPER THREE

Resistor R 1 may be adjusted to suit different transistors in the output stage. Its

value will determine the collector current of TR3, which should provide a collectorvoltage of approximately 4.5 volts. Thetotalbattery current is 6 to 8 ma. If phones

are used in place of T1, resistor R1 may need adjustment to provide the propercollector voltage.

100 .1J F

100101

Ti

020

alone when the stations are not too faraway. With an external antenna, stationsmore than 500 miles away come throughat good headphone strength. The receiveris equipped with an automatic gain controlcircuit so that stations maintain a fair aver-age loudness.

How It The first stage (TR1 in Fig.Works 4.8-A) is a converter. It is both

mixer and oscillator combined.The oscillator beats against the incomingsignal and produces an intermediate fre-quency.

Transistor TR2 is an i.f. amplifier of455 kc. which increases the signal fed tothe diode detector ( CR1). Here the signalis rectified and fed back to the i.f. amplifieras a.g.c. voltage. It also is used to drivethe audio stage. The i.f. amplifier is neu-tralized as explained earlier in Chapter 3.

Construction A little more care is neces-H ints sary in the construction of

the Super Three than isnecessary with some of the simpler re-ceivers. While layout is not critical, it

should follow some sort of order so thatthe input and output circuits are reason -

9 V.

Parts ListC1, C2 - Two gang tuning capacitors,

antenna section 290Oscillator section 130 pfd.(Miller #2112 with plates re-moved to track, or Alps #-72)

CR1 - 1 N64, 1 N295, 0A85, or sim-ilar germanium diodes.

IFT1 - 25K to 600 ohms i.f. trans-former (Miller 2041 or simi-lar).

IFT2 25K to 1,000 ohms i.f. trans-former (Miller 2042 or simi-lar).

Li - Antenna loopstick. For wind-ing data see text. (Miller 2003with a few turns removedfrom the primary)

L2 - Oscillator coil, for windingdata see text.

- Collector to speaker trans-former, 500 or 1,000 ohmsto 4, 8, 16 ohms (Stancor TA -9 or TriCKI TY-44X).

TR1,TR2- 2N1380, 2N371, 2N410, GT-139, 00615 or similar.

TR3 - 2N408, 2N1380, GT109,0C-72 or similar.

SPKR

Radio Handbook Super Three 83

TERMINAL STRIP

TO SPEAKER TO CI

MEM

6

UNDER -CHASSIS LAYOUT

Figure 4.8-BUNDER CHASSIS LAYOUT OF THE

SUPER 3 RECEIVER

ably well separated. The receiver maybe condensed into a very small packageof about the size of a pack of cigarettes ifthe Miller subminiature i.f. transformersare used. The prototype model was delib-erately spread out so that the constructioncould be clearly seen and followed (fig.4.8-B). It is advised that constructors usethe larger layout unless they have had con-siderable experience building receivers.

The chassis is bent up from a piece of#18 gauge aluminum to a length of 6 1/2"a width of 3", and a depth of 1", but acommercial 5" x 7" chassis may be sub-stituted. The panel is made from a scrap oflaminated plastic 31/2" high. A small ply-wood cabinet, covered with imitation lea-ther, is used to give the receiver a profession-al appearance. The gain control was laterchanged to a switch -type potentiometer, andthe switch is used to disconnect thebattery.

The Coils The coils are wound in thefollowing manner. On coil

L2, a 1/2" slug -tuned form, wind 4 turnsof #32 enameled wire. Space the windingout to occupy about 1/4". This is theemitter winding marked 1 and 2 on theschematic. Next, closewind 13 turns overthe previous winding. The ends of thiswinding are 3 and 4. On top of this wind130 turns. The start of this winding is #6on the schematic. The latter winding is

CLOSEUP VIEW OF THEHOME-MADE I.F. TRANSFORMER

The performance is comparable to theminiature Miller series, but this ver-

sion requires a shield.

jumble would to a width of about 1/2".Melt a little candle wax over the coil to pre-vent it from coming undone.

The antenna coil L1 is wound on a loop -stick which is a long ferrite rod or slab.Different loopsticks will call for a differentnumber of turns. As a guide, closewind28 turns of #32 enameled wire about onethird of the distance from one end. Theends of this winding are 1 and 2. Wind-ings 3 and 4 consist of 5 turns of hookupwire wound alongside the larger winding.The two windings should both be woundin the same direction and 2 and 4 should beadjacent to each other. The loopstick isconnected by 6" leads so that is may be

SEC.

Figure 4.8-CConstruction details for th ehome-mad e

455 kc. transistor i.f. transformers.


attached to the top of the chassis away frommetal work. The J. W. Miller #2022 can beused for the oscillator coil L2 and the#2004 can be used for the loopstick Liif you prefer not to make your own.

Adjustment After the wiring has beenfinished and thoroughly

checked, insert a milliameter in one ofthe battery leads and check the current.If it is greater than 6 to 7 ma. switch offthe receiver and check the wiring. If thecurrent reads correctly, place the receivernear a broadcast radio and listen for thecarrier from the transistor oscillator as thetuning capacitor in the SUPER THREE isrotated. Place the receiver near the antennafor this. The oscillator beat note shouldtune 455 kc. higher than the capacitor

UNDER CHASSIS VIEW OF THETHREE -TRANSISTOR

BROADCAST RADIO

setting. If no carrier is heard, reverse thetwo leads 3 and 4 on coil L2. An r.f. probe,if available, will determine whether thisstage is oscillating. If the heterodyne isheard on 1500 kc., the oscillator tuningcapacitor should have the plates approxi-mately half way out, indicating a receivingfrequency of about 1050 kc. If no signalgenerator is available, the i.f. transformerswill have to be peaked on a weak station.Once a signal is received, the rest will beeasy. All stages should now be alignedwith the tuning capacitor set at approxi-mately center position. If the trimmer isfully meshed and cannot peak the signal,the loopstick secondary winding requiresmore turns. The loopstick may be pre -adjusted to frequency by using it and itsassociated tuning capacitor as a wavetrapand nulling out a station on the broadcastreceiver. For example, if the nulled -outstation is on 1200 kc., then the setting ofthe tuned circuit is also 1200 kc.

Radio Handbook IF Transformers 85

%kir

REAR VIEW OF THETHREE -TRANSISTOR

BROADCAST RECEIVER

Constructing IFTransformers

J. W. Miller i.f. transformers have beenused in this section of the book becausethey are inexpensive, compact and readilyavailable. However, overseas readers mayhave trouble locating these units or anytransistor i.f. transformers. When thisis the case, home -built transformers canyield the same performance as can com-mercial units.

Remember that core material and sizemay change the inductance of the windingsconsiderably. Therefore the data to followmust be considered only a guide. Theimportant parameter of these transformersis the turns ratio, and if it is necessary to in-crease or decrease the number of turns onthe primary, the secondary and also thetaps should be varied in the same propor-tions.

The home -built transistor i.f. transfor-mer shown in the schematic (Fig. 4.17-A)and accompanying photograph was woundon a Neosid 3/8" form. These forms aregenerally available throughout the BritishEmpire. Similar forms should function aswell.

The The primary should consist ofPrimary 180 turns of fine wire, #32 to

#38 swg enamelled, jumblewound to a width of approximately 3/8"on a 3/8" form. Subject to the conditionsmentioned above, the total primary turns

will be the same in all cases for a frequencyof 455 - 465 kc. This winding, in con-junction with the 200 mfd. resonating ca-pacitor will have an impedance of approxi-mately 90,000 ohms (based on a reason-able working Q).

Drift transistors are designed to workinto an impedance of approximately 100Kand therefore may be connected across theprimary without using a tap. Other junc-tion transistors generally will require animpedance of 10K to 25K for reasonableimpedance matching.

Transformer Place a tap on the180-turnDesigned primary at 60 turns upto Match from the battery end of10,000 ohms the winding (terminal #1 ).

Fora 600 -ohm secondary,to match the transistor base, use a 9 -turnlink of the same wire. For a 2,000 ohmsecondary, to feed the diode detectors,wind an 18 -turn link over the primary.

Transformer Place a tap on the 180 -turnDesigned primary at 95 turns upto Match from the battery end of25,000 ohms the winding (terminal #1).

For a secondary, use thesame turns ratio as for the previous trans-former.

The secondaries are wound with thesame gauge wire as the primaries and are

FRONT VIEW OF THE SUPER THREEBROADCAST BAND RECEIVER


wound directly over the primaries. Theend of the primary nearest the core is ter-minal #4. The end of the primary farthestfrom the core is terminal #3. If Litz wireis used to wind the primary, the transfor-mer will have a higher Q. The transfor-mer should be shielded by a can with adiameter of at least one inch.

4.9 A Four TransistorSuperheterodyne

The four transistor superheterodynehasbeen designed around the popular midgetMILLER coils and i.f. transformers. The

cans of these components are about 1/2"square and 3/4" high. With care, it ispossible to build a receiver into a verysmall space indeed. The wristwatch radioof the comic strips is almost a reality!

However, the compression of a receiverinto a 2" x 4" area calls for a certainamount of experience if unnecessary coup-ling and consequent undesired oscillationare to be avoided. A larger layout on a

TOP VIEW OF THE FOUR -TRANSISTORBROADCAST RECEIVER

A piece of insulating board was usedfor the chassis.

metal chassis will eliminate many of thetroubles that otherwise may be encoun-tered (Fig. 4.9-A).How It The loopstick and oscillator coilWorks output links are in series, and

oscillator injection is to the baseof transistor TRi . Feedback takes placevia the lead from the collector circuit tothe tap 5 on oscillator coil L2 .

The forward bias resistor RI, on thesecond i.f. stage, is' fed back to the emitterof the first stage. The voltage drop acrossR2 supplies the forward bias to TR3.

When positive a.g.c. voltage is fed to thebase of TR2 , the current through TR2 de-creases. The voltage across its emitterswamping resistor R also decreases, andthe forward bias to TR3 decreases. Thisis a means of offering good a.g.c. actionover two stages with a considerable re-duction in the number of components inthe a.g.c. line.

The neutralizing capacitors, C2 and C39may not be necessary with many transis-tors, particularly the drift type. However,with other transistors, neutralization maybe desirable even though oscillation is nottaking place. The presence of feedbackthrough a transistor may modify the inputimpedance of the stage and reducethe gain.This subject was discussed in section 3.3.

Radio Handbook 4 Transistor Superhet 87

3

5

27

TR, IFTI

r

51

16

K

_ _1 3

C2 -.SEE TEXT- C 3

TRa

05

81JF-u+

I(

IFT2 TR3r-I I `,=4 12

L4_ - -13)F-5

.05 .05

R,1R.26

470 75

IFT3 TR4

4 L

S 1.2 K

I CR,211

J3

120K

22 6 ACC

Figure 4.9-ASCHEMATIC DIAGRAM FOR THE

FOUR TRANSISTOR SUPERHETERODYNE

Parts ListCI 0,b- Tuning capacitor, antenna sec -C2 Lion 10-130 /4ufd., oscillator sec-

tion 10-78 mfd. (Miller #2110or similar)

C3 - Varies with transistor type. For2N1380, CK760, 2N139, GT139,use 15 NM. For 2N136 use 5Nifd. For 2N371 no capaci-tance is required.

CR, - R.f. diode, 1N48, 1N60, 1N64,1 N295, 0A85, or similar types.

!FT] - 25K to 600 ohms (Miller #2131or similar type).

IFT2 - 10K to 600 ohms (Miller #2031or similar type).

IFT3 - 25K to 1,000 ohms (Miller#2042or similar type).

- Loopstick antenna (Miller #2005,or see Fig. 4.8-A).

12 - Oscillator coil (Miller #2021, orcircuitry of Fig. 4.8-A may besubstituted for handwound coil).

- Collector to speaker trans-former, 500 or 1,000 ohms to4, 8, or 16 ohms (Stancor TA -9or Triad TY-44X).

TR1,2- 2N136, 2N139, 2N371,TR3 - 2N1380, CK760, GT139.TR4 - 2N107, 2N408, 2N1380,

GT109, 0070.

Li

1011 8

/

1,= ÷ 10011F

+ 9V. - SPKR

5.6K220

10 K

Q9r

The i.f. signal is rectified by CRi andfed to TR4, which, in turn, passes theamplified audio signal via the output trans-former to the speaker.ConstructionHints

ing 2" x 4".

The prototype receiver wasbuilt on a piece of 1/16"thick phenolicboard meas-The oscillator coil and i.f.

transformer are mounted along the rearfrom left to right. The tuning capacitoris at the left in front of the oscillator coil,and the 2" speaker is at the right. Thereis just room for the frame of the speakerto project over the i.f. transformers. Themidget audio gain control is under theboard at the left and does not show inthe photograph. The oscillator and i.f.cans are mounted with their lugs fore andaft. The lugs are pushed through smallholes in the board and bent over. Twopieces of #22 tinned wire are run alongthe board in line with the lugs to whichthey are soldered. A third wire is runalong the front of the board. The endsare bent over and pushed through smallholes in the board material. A coupleof cross members are now added form-ing a grid, and all the cross -over pointsare soldered. The purpose of the grid isthree fold: (a) it holds all the cans firmly


to the board (they tend to flop about other-wise); (b) all the cans are connected tothe common line; (c) the tuning capacitor,gain control case, resistor and capacitor,returns are connected to the wire grid.This is really a busbar hook-up similar tothose used in receivers years ago, beforethe point-to-point wiring technique be-came popular.

A tie point is made by pushing com-ponent pigtail wires through holes in theboard, bending the ends over on the otherside, cutting the ends off so that they arenot too long, and then soldering themtogether. Some of the solder knobs maybe seen on the top side of the board inthe space normally occupied by the speaker.The output transformer feet are solderedto two of the ground point knobs on thetop of the chassis.

The receiver may be installed in a plasticcase. The case must not be metal, as metalwill render the loopstick useless. Theoriginal case used once housed a woman'snecklace. The interior of the receiver(though a thing ofbeauty to theperson whobuilds it) may not necessarilybe consideredbeautiful by those who know nothing aboutradio. Consequently, the transparent casewas sprayed inside with a paint bomb. Be-fore the painting, a small dial card was gluedto the inside, face up. Also, a number ofsmall holes were made in the lid to let thesound come out. The shaft was cut off ofthe gain control potentiometer. Both endswere tapped, and then the shaft was re-placed through a hole in the side of thecabinet. A case a little larger than the pro-totype would make this step unnecessary.

It is not essential that you use transistorsockets. It is a simple matter to drill threeholes underneath the transistor and maketie points as described above. This wasdone with the output transistor in theprototype model and may be seen in thephotograph. Sockets were generally pre-ferred in the prototype because this made

Figure 4.9-BTHE FOUR TRANSISTOR

SUPERHETERODYNEUnder chassis drawing of the FourTransistor Superheterodyne. The

heavy lines are the ground busbar.

it easy to try different transistors and com-pare their performance.

It is strongly recommended that unlessyou have had considerable experience build-ing compact receivers, you construct thereceiver first on a chassis similar to thatused to house the Super Three previouslydescribed. Then, when you feel that youhave mastered any problems which mayarise, have the feel of the alignment op-eration, and are confident enough to gofurther, remove the parts from the chassisand start the condensed model. Draw thelayout on a piece of paper the same sizeas the phenolic board and try various lay-outs until everything fits in without thereceiver taking on the unkempt appearanceof a sparrow's nest.

Alignment A signal generator is inval-uable when it comes to the

alignment of a superheterodyne. Thosewho do not possess one may follow thealignment procedure suggested for theSuper Three.

To proceed with the alignment, couplethe generator (tuned to 455 kc.) througha small capacitor of 5 to 20 Nrfd. to thestator of Cia . Switch on the generatortone and peak the i.f. transformers for max-imum volume. Reduce the generator gain

Radio Handbook The Mobileer 89

until the signal is just audible. A stronggenerator signal will create an excessivea.g.c. voltage which will cause false indica-tions.

Next, set the generator to 1500 kc. Setthe receiver capacitor to approximately thisfrequency and adjust the trimmer on theoscillator section of Cl for maximum sig-nal. Set the generator at 600 kc. and adjustthe oscillator coil slug until Cl peaks at thecorrect position. Set the generator and re-ceiver to 1100 kc. and adjust the trimmeron the loopstick portion of Cl for maxi-mum. It may be necessary to repeat thisprocedure once or twice because there is acertain amount of interaction between thevarious adjustments. Finally, with the loop -stick in position and the generator outputlead held near the loopstick, adjust Ciatrimmer and the i.f. transformers again formaximum volume.

In an area far removed from stations,an external antenna may be connected tothe fixed plates of Cia through a 5/4Ufd.capacitor. Volumeshouldthenbe sufficientfor all normal listening.

4.10 The Mobileer

Many have wished to copy 75 meterstations on the auto radio, or perhaps onthe XYL's broadcast band receiver. Thewish can be fulfilled for the price of aweek's supply of cigarettes and the timeyou would have spent smoking them.

Here are a few of the MOBILEER'sexcellent qualities. (1) In a side by sidetest with a commercial 75 meter commun-ications receiver the MOB ILEER held itsown. There was no noticeable differencein sensitivity between the two. (2) If thebattery voltage is dropped from 12 to 6volts the crystal oscillator will move onlya few cycles. (3) Image and spurious sig-nals are negligible and certainly no morethan a tube counterpart. (4) The MOBIL-EER uses only two transistors, employs awar surplus crystal, and draws 3 ma. froma 12 volt battery.

TOP VIEW OF THEMOBILEER CONVERTER

This simple device will convertthe3.5-4.0 Mc. ham band down to the broad-cast receiver range. The knob at theleft is the antenna trimmer. Abovethat is the antenna coil slug. Thecrystal (near the center of the chassis)is between the mixer coil and the

oscillator coil.

The selectivity or ability to separatestations is determined by the receiver towhich the MOBILEER is connected, andnothing can be done in the converter itselfto improve this.

The MOBILEER is designed to operatewith 9 to 12 volt batteries, although goodperformance, with reduced gain, is obtain-able from as low as 6 volts.

How It The heart of the MOB ILEER isWorks the combined mixer and crystal

oscillator. Considerable workwent into the design of this important partof the converter. A great number of cir-cuits were tried but all suffered from verybad image and spurious signal break-through. So bad in fact were these un-desired signals, that the whole idea of aconverter using only two transistors wasnearly abondoned. Careful thoughtbrought the realization that the spurioussignals were due to incorrect operation ofthe mixer -oscillator stage. The transistorhad been operating over the entire part ofthe dynamic transfer characteristic curve.


ANT..---WAVETRAP

' (SEE TEXT)Ll TRI

- 6 T012 V.1 -

Figure 4.10-ASCHEMATIC DIAGRAM FOR THE

MOBIL EER CONVERTERThis is a 3.5 - 4.0 Mc. converter forcar or house radio. The Mobileeruses only Iwo transistors, yet has aperformance equal in every respect

to its tube counterpart.

Parts List

C1 - 35 to 50 Aufd. antenna trimmer.C2 - 100 nafd. mica.C3 - 0.001 pfd. mica or ceramic.Li - 70 turns, #32 enamel jumble

wound on 3/8" form, 3/8" long.Antenna tap 5 turns from bottom,emitter link 2 turns, small hook-up wire over primary. (Miller#4400 or similar slug tuned form)40 turns #32 enamel, same asLi. Emitter link 3 turns of hookupwire over primary.90 turns, #36 s.c. wire, otherwisesame as L1. Link 3 turns hookupwire over primary.

12

L3

The peaks of the oscillator waveform werebeing rectified, producing exactly the sameeffects one gets when a diode is connectedin series with the antenna lead in the pres-ence of a strong station (cross modulation).With this realization, the circuit design wasmodified, the cross modulation stopped,and the MOBILEER took its present form,Fig. 4.10-A.

The oscillator is a Colpitts circuit withthe capacitor's C2 and C3 forming the ca -

001

CARRADIO

TR1 -22N1745, 2N371,0C170, 00615

pacitive voltage divider. This is explainedin section 3.2. The values of C2 and C3

should not be changed. The oscillatoroperates in the common -emitter config-uration. The mixer part of this stage op-erates common base with the signal inject-ed into the emitter.

The r.f. amplifier also operates commonbase. This configuration was chosen forsimplicity and to prevent the need for neu-tralization, which might have been necess-ary if transistors other than those suggest-ed were used. If other transistors are in-corporated in the present circuit theyshould have a high alpha cut-off frequency.

Both stages are fully stabilized to allowsafe operation in the high temperaturesoften encountered in closed automobiles.

The The crystal for the MOBILEERCrystal need cost no more than a dollar.

A vast number of crystals areavailable on the surplus market for muchless than this. Any crystal between 4550and 5100 kc. will operate in this circuit.If a 4550 kc. crystal is chosen, the broad-cast receiver is tuned between 550 and1050 kc., which represents the frequenciesbetween 4.0 and 3.5 Mc. respectively. A5.1 Mc. crystal will place 3.5 Mc. at thehigh end (1600 kc.) of thebroadcastband.A crystal frequency between these limitswill cause the 75 -meter band to cover themiddle portion of the receiver dial.

A point to note is that the broadcastreceiver tunes backwards. That is, the lowfrequency end of the 75 -meter band will

Radio Handbook The Mobileer 91

L,1COLORDOT

MILLER CO/LTERMINALS

Figure 4 10-BIf strong broadcast stations "break-through" the converter, a wavetrapmay be inserted in the antenna lead.The best position for the trap is insidethe Mobil eer case close tothe antennaconnection. The coil may be any os-cillator coil, such as the Miller #2022.Tune the slug until the offending sta-

tion nulls out.

be on the high frequency end of the re-ceiver dial and vice -versa. However, ifeither of the suggested crystal frequenciesare used, the stated calibration marks on

the receiver dial will hold for 75 and 80meters.

Construction Construction may take anyHints desired form. The main

points to observe are thatthe input and output circuits should notbe placed too close to each other, and com-plete shielding of the whole unit is requir-ed, including the lead to the broadcast re-ceiver. Incomplete shielding will allowbroadcast stations to be picked up and tocompete with the 75 -meter signals. If theunit is constructed on a flat plate, as shownin the photograph, it may be easily droppedinto a small U-shaped chassis and securedin place.

If desired, a switch may be built intothe unit for bypassing the converter andallowing normal operation of the receiver,Fig. 4.10-C. If a switch is used, take carewith the lead dress and shield the wires

UNDERCHASSIS VIEW OF THE MOBILEER

Note that the components are mounted on a length of terminal strip stock. Itshould be possible to use the Mobil eer on 40 meters by using a 6.0 Mc. crystal

and removing a few turns from the coils.


Figure 4.10-CA double pole, double throw switchwill remove the Mobileer from thecircuit when broadcast reception isrequired. A third deck on the switchmay be used to break the battery

voltage.

right up to the switch. Since the input andoutput leads oftheMOB MEER arebroughtto the same switch, this is a likely place forthe transfer of broadcast stations past theconverter.

Home The MOBILEER may be operatedStation from a large outdoor antenna a -U se head of a domestic receiver, but

it is possible that in areas close tostrong broadcast stations some break-through may be experienced. If this is thecase, the crystal frequency may be alteredto shift the 80 -meter band away from thatportion of the dial. Alternatively, a wave -trap may be inserted in the antenna leadclose to the converter. Suitable componentsfor a wave trap are shown in Fig. 4.10-BA receiver with a built-in antenna shouldnot be used with the converter, for it willbe almost impossible to prevent broadcastsignals from QRM'ing 75 and 80 metersignals.

Figure 4.10-DThis layout was used in the prototype,which left room for the switch andwave -trap at one end. Any similar

layout may be used.

Here is another interesting possibility.If coil L3 is replaced by a 2.5 mh. r.f. choke,the MOBILEER can be used with the fa-mous "Q-5'er" Command Set receiver(BC -453A), which tunes 190-550 kc. Notonly will this make an excellent home sta-tion for the Novice or General Class Am-ateur, but it will also provide a stable mo-bile receiver. The MOBILEER could belocated in the trunk of the car, along withthe transmitter, and only the "Q-5'er"need be positioned near the driver.

4.11 The ProductDetector

As mentioned elsewhere in the book,a product detector is a mixer combiningthe i.f. signals with the b.f.o. to producea separate i.f. in the audio range. Mostmixer circuits may be converted to a prod-uct detector with little alteration. However,a mixer is used at the front of the i.f. stripwhere the signal level is very low, and aproduct detector is used at the rear of thei.f. where the signal level is very high.

To prevent overload ofthe detector whenthe normal mixer is used, the input signalhas to be attenuated and the consequentloss in recoverable audio is considerable.This necessitates additional amplification tobring the signal to par with a.m. detection.

A slightly different product detector(though it may appear to differ very littlefrom the usual) has large signal handlingcapabilities and very linear output. Thecircuit of this detector is shown in simpli-fied form in Fig. 4.11-A. The detectoroperates with little or no applied collectorvoltage and will handle as much as onevolt of input signal at the base.

A large input signal requires that therebe a large b.f.o. component applied to theemitter circuit. Assume that the requiredsignal is applied to the emitter from theb.f.o. Also assume that the bottom endof R2 is grounded. Positive half cycles of

Radio Handbook Product Detector 93

An etched circuit version of the product defector shown in Fig. 4.11-B.

b.f.o. voltage will cause the base to be neg-ative with respect to ground, and the basewill be forward biased. The base -emitterpath will be a very low impedance, and con-sequently, the base will assume a positivepotential with respect to ground. Thebase-collector path will now be a low impedance,and the collector will also assume apositivevalue. This will cause the capacitor C1 tocharge and actually maintain a steady volt-age. The transistor has developed its owncollector voltage. This voltage is a prod-uct of the b.f.o. voltage and is negativewith respect to the base although positivewith respect to ground. Thus, a signal at.the base will be superimposed upon the

2N1380

Figure 4.11-ATHE BASIC PRODUCT DETECTOR

collector voltage, causing it to vary in am-plitude. Negative b.f.o. cycles applied tothe emitter will increase the base emitterimpedance, and the collector current willcease to flow. However the charge on Clwill remain.

In practice, when the product detectoris built into a receiver, and where the b.f.o.and signal voltages are not excessive, opti-mum output requires that a small d.c. volt-age be applied to the collector. The tran-sistor is thus biased part way up the kneeof its characteristic curve. This means thatsince the transistor is drawing current, thecollector -base impedance is low. The out-put impedance of the product detector isconsequently very low.

Adjustment Fig. 4.11-B shows the circuitof a transistorized product

detector which may be added to an existingreceiver. The input from the receivershould be adjusted so that overload of thedetector and frequency modulation of theb.f.o. does not occur.

When the b.f.o. is detuned 10 kc. awayfrom the intermediate frequency, there


1

I.F. INPUTTI TRi 0.251JF

AUDIO OUTPUT

I 8-9 V.0

REG.

.005EA. .001

SILVER MICA

2.711 .005

47 K

5.8 K

Figure 4.11-BA SIMPLE PRODUCT DETECTOR

This product detector can be added toexisting receivers. For maximum sta-bility, silver mica capacitors are used

in the oscillator circuit.

Parts ListLi - 90 turns, #34 enamel scramble

wound on 3/8" slug tunedform.Link 10 turns of the same wire,wound over primary.

T - 455 kc. transistor i.f. transformer(J. W. Miller #2041).

TR] , - 2N247, 2N274, 2N371, 2N-TR2 1745, 0C139.

should be no audio output. Any outputunder these conditions indicates that theproduct detector is being overloaded.However, when the b.f.o. is turned off,a.m. detection should normally take place.If the circuit is to be used for a.m. recep-tion, a small amount of bias should beapplied to the detector base as in Fig.4.11-B to prevent distortion at low signallevels. The bias will not affect the productdetector action.

The value of the collector load resist-ance of TRI is not critical. However, theoperating point may be changed if an in-correct value is used. A potentiometercould be temporarily inserted in the cir-cuit to substitute for the detector load.

With the b.f.o. on, the potentiometer isadjusted for maximum output consistentwith linear operation. In many tests car-ried out by the authors, optimum collectorvoltage seemed to be around 0.25 volts.The collector current under these con-ditions will be around 2 ma. with a supplyof 1 2 volts.

If a high audio output level is requiredfrom the product detector, it may be neces-sary to use a b.f.o. amplifier to increasethe injection voltage. The signal at thebase of TR1 should never exceed in ampli-tude the b.f.o. voltage at the emitter. A

large signal at the base may frequencymodulate the b.f.o. unless an isolatingstage is used.

4.12 A ProfessionalCommunications Receiver

This receiver has amplified a.g.c., aproduct detector, and provision for an Smeter. The receiver covers the basic fre-quency range of 3.5 to 4 Mc, and can beused in conjunction with a converter forthe other bands. Then the receiver be-comes a tunable i.f. and the stability, evenon 10 meters, is exactly the same as on80 meters. To make the project less com-plicated, the converter was not built intothe receiver. However, there is no reasonwhy this cannot be done, and the only pre-caution to observe is in shielding. Thisis necessary to prevent receiver oscillatorharmonics from mixing with other prod-ucts and producing a multitude of "birdies"in the converter.

How it Because many of the principlesWorks used in the receiver arenot widely

known, a fairly complete descrip-tion of the design and function ofcompon-ents are given in Fig. 4.12-A.

The r.f. stage transistor TRI is oper-ated grounded base. This is the equiva-lent of grounded grid in tube circuitry.

Radio Handbook Communications Receiver 95

Originally, the r.f. stage was operatedgrounded emitter. This was changed whenit was observed that the action of the a.g.c.on the r.f. stage was causing detuning ofthe local oscillator. When the a.g.c. varied,the flow of collector current and the out-put impedance of the stage also changedand reflected a change in mixer input im-pedance. The input impedance of thislatter stage, as far as the oscillator is con-cerned, is highly capacitive and is actuallya capacitor connected across part of theoscillator coil, L3 . On strong stations the

CLOSEUP VIEW OF THECOMMUNICATIONS RECEIVER

I.F. AMPLIFIER STRIP

frequency variations were quite severe. Al-though unnoticeable on a.m., s.s.b. stationsgave a gargling sound. Because the out-put impedance was very high with ground-ed -base configuration, an impedancechange was but a small percentage of thewhole. Thus, the reflected change acrossthe oscillator coil was negligible. Anotheradvantage offered by grounded base is free-dom from the need for neutralization.

FRONT PANEL VIEWOF TH ETRANSISTOR

COMMUNICATIONSRECEIVER

The controls, fromleft to right, are an-tenna trimmer, r.f.gain, on -off switch,audio gain, and

b.f.o. tuning.


A '0

A2,0

120 K1

.005

L_ _ 2201.c

L

lit RF AMPLIFIER, 3.5-4.0 MC. MIXER, 3.4-4.0 MC.TR i La TR 2 IFT1

4r 1

I

le13

I T -ii ;

1 ,----- ,a,_ _,

r. 0 V. 5 -6.6K1-)I--L 0

01 .005 1 IAT -7.6V

* =SILVER MICA

OA

.0, 470

ADC

Figure 4.12-A

SCHEMATIC DIAGRAM FOR THE PROFESSIONAL COMMUNICATIONS RECEIVER

Note that voltage data is included, and readings are from the points given to

chassis ground.

IF AMP., 455 EC.Nc Rc

--v."TR4

C O AGC*1

r iFrz 7

d I

I 1

I a-05o

L _,.01 -7.2

A=SILVER MICA

/F AMP., 455 MC.Nc Rc

TR5 r 1FT3 7

SIDEBANDSELECTOR 2.7K 2.7K 1

PROD. DET. /AM DET.AGC AMP.,435 MC.390

TRe -3.51/ TR7

0 Az

OC

3.11F

.11% DAUDIO

-6.4 V.

100

POINT A(SEE TEXT)

-4-1.2 V RF


AUDIO

C10,

16,1c

8.4V.I ENERDIODE

TRIO T1

Figure 4.12-A cont.

Parts List

- Tuning capacitor from BC -455receiver. Modified per text.

IFTI, - 25K to 600 ohms, 455kc. tran-2,3 sistor i.f. transformer (J.W. Mil-

ler 02041).Li - 27 turns, 034 enamel, on 3/8"

diameter slug tuned form (Mil-ler 04400 or similar), antennatap at 7 turns, emitter link 2turns small hookup wire woundbeside or over primary.

L2 - 27 turns, 034, form as above,base link 3 turns as in Li.

- 16 turns, 034 enamel, on 3/8"form as in Li . Link, 1 turn atground end.

- 90 turns, 036 silk covered on3/8" form, tap 30 turns frombottom.

L3

14,5

Ti

Ti 10K to 2K interstagetransform-er (Stancor TA35 or Triad TY-56X).

500 ohms to speaker (StancorTA35 or Triad TY-45X).

TR1 - 2N247, 2N274, 2N371, OC-3,8 619.TR2 - 2N412, 0C45.TR4,5 - 2N247, 2N371, 2N410,

GT139, 0C44.TR6 - 2N1380, 2N408, GT81R,

0072.TR79 - 2N412, 2N1380, 0C44.TRio, - 2N188A, 2N408, 2N1380,11,12 0072 (see Mini -Amp data,

Chapter 2).Z 8.2 volt zener diode (Inter-

national Rectifier 1 N151 1 orsimilar).

T2

No doubt the observant reader has no-ticed an unusual connection between thebase of the r.f. stage and the emitter of thesecond i.f. stage in Fig. 4.12-A. This isanother way of obtaining amplified a.g.c.voltage. Because of the presence of thei.f. emitter resistor, the emitter of TR, isnegative with respect to ground. Auto-matic gain control voltage decreases thei.f. stage emitter current, and the voltageacross the resistor also decreases. Thestage produces amplified a.g.c. voltage.

The Mixer The mixer is conventional,and with the oscillator signal fedOscillator into the emitter. The oscil-

lator is a very high "C" Col-pitts circuit (see Chapter 3). As this is theheart of the receiver's stability, consider-able care went into the design. The resultis an oscillator which shifted only 150cycles in 4 days of continuous operation,with a maximum excursion noted of 400cycles. The operating point of the oscil-ator has been very carefully chosen to min-


imize frequency shift due to temperaturechanges and supply voltage variations. Al-though the oscillator will withstand a cer-tain amount of supply voltage change, itmust be voltage stabilized, nevertheless. Thisis necessary to prevent the very large classB current demands of the output stagefrom modulating the oscillator. An 8.2 -volt zener diode regulates the supply to thereceiver, with the exception of the audiostages.

Mechanical considerations will governthe stability of the receiver perhaps morethan will electrical. For this reason theexcellent variable capacitor from an ARC -5surplus receiver was chosen to tune thetransistor receiver. This capacitor has abuilt in gear mechanism of unsurpassedquality. Unfortunately, its plates are cutfor straight line frequency when straightline capacity would have been better forthis circuit. However, by not using thewhole range of the capacitor, and by re-moving many of the moving plates, excel-lent spread and good tracking were ob-tained.

The I.F. The i.f. amplifiers are con -Amplifiers ventional and arebuilt around

the compact Miller i.f. trans-formers. A duplicate i.f. strip was alsoconstructed using home -built transform-ers, and its performance was also excellent.Data for these homebuilt transformers issupplied at the end of this chapter. Thei.f. stages are a.g.c. controlled by reducingthe negative base bias and thus the emittercurrent. The r.f.-i.f. manual gain systemcontrols the same bias line.

The AGC The a.g.c. amplifier ( TR6) isAmplifier in reality a detector with ahigh

value load resistor. It operatesnear cutoff, and the standing collector volt-age is high. Consequently, the a.g.c. linefrom its collector to the i.f. amplifiers ishighly negative. A signal at the base isrectified by the base -emitter portion of the

TR1 0.6 ma.TR2 0.2 ma. (200fiamp.)TR3 1.0 ma.TR4 0.8 ma.TR5 1.0 ma.TR6 Nil.TR7 2.0 ma. (b.f.o. on)

30 amp. (b.f.o. off)TR8 2.0 ma.TR9 1.0 ma.TRIo 1.0 ma.TRH 12 3.5 ma. (approx.)

Figure 4.12-BChart showing current drawn by PRO-FESSIONAL COMMUNICATIONS RE-CEIVER transistors. For these meas-urements, the r.f. gain is maximumand no antenna is connected to the

receiver.

transistor, and a voltage drop is developedacross the 1K load resistor in the base cir-cuit. This voltage, being negative in value,causes a collector current to flow and a re-sultant drop in collector voltage. Bias tothe i.f. amplifier (and indirectly to the r.f.amplifier) is reduced. Note that the vary-ing collector current may also be used tooperate an S meter. This adaptation isdescribed later.The Product The product detector ( TR7)and is a mixer circuit, and itsAM Detector similarity to the mixer stage

at the front of this receiverhas been described in section 4.11. Thebase of the detector is fed through the 390/rend. capacitor which was chosen to pre-sent the correct signal level to the base.

A low collector voltage provides a curvebest suited to the purpose. It is signifi-cant that when the b.f.o. is detuned, theproduct detector has no output. The com-ponent values are somewhat critical, andit is recommended that the transistor in-dicated be used. However, information will


be given later so that builders may adjustthe components to suit different transis-tors. The 0.01 pfd. capacitor from thecollector to ground is an essential part ofthe product detector circuit and shouldnot be omitted.

When the b.f.o. voltage is removed, thecollector current of TR7 drops to near-

zero. The transistor is forward biasedjust enough to overcome distortion whichis the result of the knee at the bottom ofthe characteristic curve.

The BFOandAmplifier

nal at thechange in

Except for the difference in fre-quency, the b.f.o. ( TR8) is thesame as the local oscillator de-scribed earlier. The large sig-base of the detector causes aimpedance which, if no b.f.o.

buffer is used, causes a change in the beatfrequency. If the input signal is reducedin value, the detector output becomes low,and under certain circumstances transistorhiss level can become a problem. Theauthors turned to the use of a b.f.o. am-plifier and buffer ( TR9) only as a last re -

0411,*

11$41J AO.C.*O cg C

&Ms .r`Ns ...,,ro,---

sort. The effect was remarkable. With ther.f. gain wide open and the a.g.c. pumpingviolently on the strongest of sideband sig-nals, not a cycle of shift could be detected.Even c.w. gained a musical quality not pre-viously noticed.

The Audio Following the detector comesSection a class A audio amplifier

(TRIO ). This amplifier drivesa class B push-pull stage (TRII-12) similarto the MINI -AMP described in Chapter 2.Because of its low idling current and theconsequent saving of battery power, classB operation was chosen.

UNDER CHASSIS VIEW OF THECOMMUNICATIONS RECEIVER

Note the construction technique usedto fabricate the i.f. amplifier. The beatfrequency oscillator is mounted in acan above the chassis, at the lowerright corner. Components for the re-ceiver local oscillator are located be-low the chassis, just to the left of center.


1

A "gremlin" distorted the signal evenin this, the simplest of circuits. This dis-tortion was traced to an above -audibilityaudio oscillation and was remedied byplacing a 0.05 /dd. capacitor across theprimary of the output transformer.

The audio output is greater than 250milliwatts and is of excellent quality whenconnected to the headphones or to an ex-ternal speaker. The small inboard speakerwill handle only 100 milliwatts and leavessomething to be desired in the way ofquality.

Techni- The zener diode, like the VRcalities tube in a vacuum tube sideband

receiver, is an absolute must. Itsvalue is not particularly critical. Any valuearound 8 volts is satisfactory.

A great number of different transistorswere tried in the various circuits to ascer-tain that others than those specified couldbe used without too much circuit alteration.In most cases it is only necessary to plugthe substitutes into the sockets. The twoimportant transistors in the receiver arethe two oscillators. Of those tried, thedrift and MADT transistors gave superiorresults. Those tested were the RCA 2N371 and 2N274, the Philco 2N1745, andthe Amperex (Phillips) 0C139. If the b.f.o.is tuned to 455 kc., the 8th harmonic isat 3640 kc. and can be heard. However,by placing a small trap in the base lead ofthe product detector, the harmonic is re-duced to negligible proportions. The con-nections for the trap are shown in Fig.4.12-C. To adjust the trap, tune in theb.f.o. harmonic with a VTVM connectedto the a.g.c. line. Adjust the trap for mini-mum negative voltage. This means mini-mum signal from the b.f.o. Repeak thelast i.f. transformer.

Construction The prototype receiver wasHints built upon a standard 5" x

11" x2" aluminum chassis.A cut-out at the rear houses the i.f. stripwhich was built upon a 6 1/2" x 2" piece

TR7AUDIO OUT.

Figure 4.12-CSCHEMATIC DIAGRAM FOR THE

B.F.O. TRAP USED IN THECOMMUNICATIONS RECEIVER

Note that the trap must be groundedat the same point as the collector by-pass for TR 7. Coil L consists of 30turns, #34 enamel., jumble wound on

a 1/4" diameter form.(J.W. Miller 20A00ORB1)

of phenolic sheet for the sake of conveni-ence. Components are mounted on top,and the wires are brought through smallholes as in an etched circuit board. In lieuof a ground strip around the edge, a pieceof #18 tinned wire was run down the twolong sides of the board and held to thechassis by the mounting bolts. End wiresof the components are bent over and sol-dered to the closest points. Coupling be-tween sections does not result as one mightexpect. The i.f. cans arelikewisegroundedto the busbars. Note that the Miller i.f.transformers shown in the picture havepinconnections slightly different from thosepresently made by that company. Thenum-bered connections shown on the schematicare for the newer transformers.

The b.f.o. and its amplifier are housedin a surplus i.f. can measuring 2"x 1 3/8"wide and 4" high. A slightly larger canwould have allowed the r.f. choke, in thecollector of the b.f.o. amplifier, to be in-cluded. This is desirable where possible.The b.f.o. and amplifier, like the i.f. strip,are built on phenolic board of a size nec-essary to make a neat fit in the can. Thecan is lined with thick paper to preventshort circuits, and the ground leads are


taken to a single lead that is insulated fromthe can. That lead is brought below the chassisby insulated wire and soldered to a lug onthe chassis alongside the product detector.The 0.01 /Ad. disc ceramic, bypassing theproduct detector collector, also is solderedto this point, along with the ground con-nection for the b.f. o. trap. If the b.f.o.harmonic is to be reduced to a minimum,these connections must be strictly ob served.Similarly, the B- lead entering the can isalso bypassed at this point. The b.f.o.tuning capacitor must also be shielded sothat no harmonic energy may escape to thefront of the receiver.

The Tuning The tuning capacitor is re -Capacitor moved from a surplus BC -

455 -B receiver and is modi-fied in the following manner. Remove allthe rotor plates except two from the gangfarthest from the worm drive. Begin at thesplit plate end and work back. Do this bybending each plate back and forth, not bypulling on the plates. The authors wellremember the frantic hunt for the smallsteel balls which shot out of the end bear-ing when the plates were pulled and therotor jumped out of the bearings! With asmall hacksaw blade cut the strap connect-ing the moving plates together. Becauseit is without trimmers, this section of thegang is used in the r.f. stage. Connectedacross this section is an auxiliary antennatrimmer capacitor which is adjusted fromthe front panel. The center capacitor sec-tion is connected to the mixer and is mod-ified in the same manner. Eight movingplates are left in the oscillator section.

Coils Constructors may wish to use coilforms differing from those speci-

fied. If this is the case, the coils may begrid -dipped to frequency. The presentturns to tap ratios should be retained. Re-member that the oscillator coil tunes 455kc. higher than the r.f. and mixer stages.

The b.f.o. may be set to frequency bymeasuring the frequency separation be-tween harmonics which result when theb.f.o. is coupled to a broadcast receiver.The difference between the harmonics isthe fundamental frequency of the b.f.o.

The r.f. oscillator section was built ontoa small plate and mounted below the chas-sis because there was insufficient room onthe top of the chassis. The empty space atthe end of the if. strip is reserved for eithera mechanical or a crystal filter. The oscil-lator components must be rigidly mounted.Only ceramic materials should be used forthe coil form, and the tiepoints must be ofgood quality wherever they support por-tions of the tuned circuit.

The r.f. stage is built on a small phenol-ic strip and mounted below the chassis atone end. The coil (Li) is mounted on thechassis proper, and the adjustment screwis accessible from the top.

Battery When the battery is new, the idl-ing current of the entire receiver

is around 20 ma. and drops to 17 ma.when the battery is down to 10.5 volts.The average current is around 30 to 50ma., depending on receiver volume. Peaksmay reach as high as 100 ma. If the re-ceiver is used in an automobile and is op-erating from the car battery, the stabilitywill be excellent, even though the voltagemay fluctuate as thegenerator output varies.When the battery voltage falls below thezener diode control level, frequency shiftwill appear on c.w. and s.s.b., and distor-tion will be present on all signals.

Protecting If there is danger of r.f. fromthe the transmitter entering theFront End receiver and burning out the

r.f. stage transistor, a smalldiode may be connected directly across thefirst tuned circuit. Because most diodes(especially the silicon variety) do not con-duct until a certain voltage level is reached,rectification will not occur on the signal.


Even better protection is afforded by twodiodes connected in parallel but each facingin an opposite direction.

OtherTransistorsandN e utr al -

ix ation

If transistors other than thedrift type are used in the i.f.stages, neutralization may berequired. The neutralizationcomponents are shown dot-ted in the schematic, but the

resistors and capacitors may not be neces-sary. The subject of neutralization has beencovered in Chapter 3. If 2N139's or 2N410's are used, Itc will be 470 ohms andNc approximately 100 Nifd. If in doubt,make Nc variable and later replace it witha fixed value. The b.f.o. amplifier may beany of the lower frequency transistors, andthe a.g.c. amplifier may be any general pur-pose audio transistor having low IC. anddrift.

Alignment The process of alignment isidentical to that used with tube

receivers. The coil slugs align the low fre-quency end of the band, and the trimmersare used for alignment at the high end.The i.f. stages are aligned to 455 kc. Av.t.v.m. connected to the a.g.c. line is avery

sensitive alignment indicator. To tune tolower sideband, the b.f.o. capacitor is setabout 1 kc. lower in frequency than zerobeat. Conversely, on upper sideband, theb.f.o. is set about 1 kc. higher.

To adjust the b.f.o. injection voltage,tune in a strong a.m. station until it is zerobeat with the b.f.o. Next, detune the b.f.o.coil slug off -frequency until the heterodynecannot be heard. Output from the receivershould drop considerably. Adjust the am-plifier coil slug (L5) until the station nulls,indicating the correct injection voltage andcurrent. The b.f.o. slug is then returnedto the normal position.Using the Because the a.g.c. action is ex -Receiver ceptional, a strong pumping

action will take place on strongc.w. or s.s.b. stations. This is caused bythe a.g.c. holding back the receiver gain onstrong signals but opening up the receiversensitivity the moment the transmitter op-erator pauses for breath. The audio gainshould be turned up and the r.f. set sothat the a.g.c. just operates on the weakeststation. The tuning may be difficult if ther.f. gain is high and every weak signal isbrought up to the same level and allowedto compete with the stronger stations.

.4....... -

----- 7--- Lisniiilli, w-iiiio) 007)

114111"1111,111 0

lafl

it 1 I Illifil li it k /' I 0110 ' I.laii p,

,,

TOP CHASSIS VIEWOF THE

COMMUNICATIONSRECEIVER

The Mini -Amplifier(described in Chap-ter 2) is located atthe left, behind thespeaker. The spaceat the right rear cor-ner of the chassis isreserved for the

crystal i.f. filter

Radio Handbook Crystal Filters 103

An I.F. Without an i.f. filter, the selectiv-F ilter ity of the receiver is poor. Three

tuned circuits are insufficient tokeep out an intruding neighbor. However,the receiver was designed to be used witha filter, and a suitable circuit is shown inthe following section.

4.13 Crystal Filtersfor a

Communications Receiver

Three tuned circuits in the i.f. sectionof any receiver designed to cover the ama-teur bands are insufficient. Double tunedcircuits will improve the selectivity to amarked extent. However, unless ten totwelve tuned circuits of high Q and correctcoupling factor are used, selectivity, by to-day's standards, will still be poor. Modernreceivers, especially those designed forsingle sideband reception, use crystal,mechanical, or complex tuned filters toobtain a high order of selectivity. Betterfilters improve passband shape by provid-ing steep skirts and a flat top passband.

The readily available surplus crystals inthe i.f. region offer the chance to constructa filter costing no more than a few dollars,yet comparing favorably with filters costingten to twenty times as much. The crystalfilter shown in Fig. 4.13-A uses four 50cent surplus crystals in a back-to-back half -lattice arrangement. Adjustment is merelya matter of aligning transformers T1 andT2 and capacitor Cl for maximum outputmidway between the two crystal frequencies.Skirt selectivity is excellent.

Two filters are described in this section.Each is shown connected into the receiverjust described. These types of filters maybe used with other transistor receivers toobtain additional selectivity.

A Collins Mechanical Filter can also beused. The input circuitry is the same asin Fig. 4.13-A except that the primarytransducer coil is resonated with a 100-150

capacitor. The output transducer is

Figure 4.13-ASCHEMATIC OF THE 3KC. CRYSTAL

FILTER FOR THE PROFESSIONALCOMMUNICATIONS RECEIVER

Crystal Y1 is a surplus channel 327,and V2 is channel 329. Capacitor CIis a 3-30 pfd. compression padder,used to tune the 2.5 mh. r.f. choke(this value is critical). See text for

transformer and alignment data.

connected to the base of the first i.f. tran-sistor through a 100-1 5 0 Nifd. capacitor.Both of these resonating capacitors shouldbe adjusted for maximum gain. Use ofthis filter is described in more detail in sec-tion 4.14.

Filter #1 The filter shown in Fig. 4.13-Ahas a passband of about 3.5 kc.

Insertion loss is very small and no extrastages are required to overcome filter loss.The filter may be constructed on the chas-sis in the extra space at the input end of thei.f. strip, or can be built on a separate plate.This sub -assembly can be surface, sub -chassis, or above -chassis mounted. Thei.f. transformers shown in the photographare not normally available and are, in fact,modified broadcast receiver types. How-ever, J. W Miller Co. should have double -tuned i.f. transformers suitable for thisapplication. Other transformers may beused providing they satisfy one or two con-ditions. Choose transformers which arefixed -tuned with a 100 to 200 Nefd. capac-itor. The capacitor in the secondary wind-ing of T1 and the capacitor in the primaryof T2 are removed and replaced by two


THE COMMUNICATIONS RECEIVERCRYSTAL FILTER FOR

TWO KILOCYCLE BANDWIDTH

250 pfd. capacitors series -connected asshown in the schematic. Transformerswith different L/C ratios may give consid-erably different results.

Choosing Crystals should be fairly wellthe matched before being used inCrystals the filter. Connect one pin of

the crystal to a frequency meteror signal generator, and with the r.f. probeof a VTVM on the other pin, tune the os-cillator through the crystal frequency. At

the series resonant frequency, the meterwill show a large peak and a deep null a-longside it. Choose two crystals of thesame channel numbers with peaks of thesame height and similar frequencies. All

four crystals should show the same peakamplitude. Crystals are readily availablefrom various sources ready -matched andpaired.

Alignment Adjust the frequency meter orsignal generator to midway be-

tween the two peaks as read on the volt-meter. The generator must be midpoint

between these two frequencies. Align alltransformer slugs for maximum output.Connect a VTVM to the receiver a.g.c. line.Tune the generator across the passband.If there is a large peak at one side of thepassband, it indicates that the generatorwas not set at mid -frequency, or that thecrystals were not matched for amplitude(or Q). It is possibletominimizethis con-dition by realigning to the lower frequencyside until the two peaks are equal in height.

Next, adjust C1 for more capacitance,(approximately one-half turn) until theskirts are at their maximum steepness.This point should coincide with a dip ap-proximately 7% of total passband ampli-tude. If the trimmer is screwed in too far,the dip will become excessive. The selectiv-ity of the filter is such that the carrier of ana.m. station may be dropped right off thepassband and the signal actually turned in-to s.s.b. On a.m., high frequency responsewill be considerably down. To obtain max-imum fidelity, the receiver should be detuned a little to favor one sideband.

Care must be used to prevent signalfrom leaking around the filter. You havea shielding problem if the signal is stillheard when the crystals are removed.

F filter #2 The passband ofthe filter shownin Fig. 4.13-B is only 2.0 kc.

wide. Thus, the recovered audio in thecommunications receiver will tend towardbass. If the ultimate quality in selectivityis desired, this is the filter to use. How-ever, the high audio frequencies attenuatedto this degree may be objectionable.

The alignment procedure is the sameas for the filter just described, except thatthe trimmer C1 is tuned for maximum out-put and left at that setting.

Installing There are several ways ofthe mounting the filter in the re -F ilter ceiver. If the receiver is already

built, the filter assembly maybe placed on end in the space saved nearthe r.f. end of the i.f. strip. The filter

Radio Handbook Mechanical Filter 105

Figure 4.13-BWIRING DIAGRAM FOR THE 2KC.

PASSBAND RECEIVER FILTERCrystals V1 and V2 are war surplustypes for channel 327 and 328 re-spectively. All other component val-ues are the same as in Fig. 4.13-A.

should be connected to the receiver througha short length of small coaxial cable.

Transformer T1 in the receiver shouldbe disconnected or removed. No otherwiring in the receiver is altered. An altern-ative position for the filter is on the i.f.strip itself, with the mixer moved to makespace for it.

Which Those who have neither used norFilter? constructed filters will no doubt

be in a quandary as to which filterto build. Filter #1 is best for those whoprefer a.m., yet it will allow good s.s.b. orc.w. copy. Filter #2 is essentially for thec.w. or sideband man, and a.m. may becopied only on one sideband. Of course,QRM with the narrower filter is consider-ably reduced.

4.14 A Mechanical Filterfor the

Communications Receiver

Those fortunate enough to possess amechanical filter may wish to install it inthe receiver just described. The methodof connection is applicable to other receiv-ers such as in Fig. 4.14-A. The secondi.f. transformer is removed and replacedby the mechanical filter. The primary cir-cuit is otherwise undisturbed. It is neces-

sary to use series tuning on the secondaryside of the filter in order to obtain an im-pedance match to the base of thetransistor.Becausethis means inserting a capacitor inseries with the filter output winding, it isnecessary to add an r.f. choke to completethe d.c. circuit. The small 22Nifd. capaci-tor across the secondary brings the windingto resonance. It is possible that othertran-sistors with different input capacities mayrequire a different value, in which casethe mfd. capacitor may be replaced with a3-30 mfd. trimmer. A number of transis-tors have been tried with little variation ingain. No other changes need be made inthe base circuit of the i.f. amplifier.

Note that the filter is installed betweenthe first and second i.f. amplifiers, not be-tween the mixer and the first i.f. amplifieras is customary. It was found that a betterimpedance match was possible when themechanical filter was driven by an i.f. am-plifier stage (in preference to the mixer),and a considerable increase in gain result-ed. Ground the case and make sure thatno signal is getting around the filter in-stead of through it.

Both the 3.1 kc. and 2.1 kc. plug-intype Collins Filters gave excellent results inthe Communications Receiver. The 2.1 kc.(type 455 J 21) is now a permanent partof the device.

151- IF AMP. MECH. FILTER

14

1u00 2ND I F AMP.

72 K

TO AGC

Figure 4.14-ASCHEMATIC DIAGRAM FOR COLLINS

MECHANICAL FILTERI.F. AMPLIFIER MODIFICATION

(Project 4.12)


There is ample gain in the receiver inspite of the relatively high filter attenuation,and no further amplicication stages arenecessary.

The b..fo. capacitor setting is quite crit-ical with this type of filter.

4.15 A TransistorAll -Band Converter

The converter described in this sectionis crystal controlled. This means that thestability on any band is determined onlyby the receiver to which it is connected. Ifthe receiver is stable on 80 meters, then10 -meter signals will be equally stable.The converter can be used ahead of the re-ceiver described in section 4.12, or withany receiver that covers 3.5 to 4.0 Mc.The circuit is shown in Fig. 4.15-A.

How it The r.f. amplifier (TRI ) is acon-Works ventional grounded -emitter

stage. A capacitive divider acrossthe tuned circuit establishes a correct im-pedance match to the transistor. The mix-

er ( TR2) also operates in the grounded -emitter configuration. Both the oscillatorvoltage and the signal are fed into the basethrough a capacitive impedance transfor-mation arrangement. The oscillator ( TR3)is a form of Colpitts, and is a very activecircuit. This circuit gave no difficulty inthe several converters which were built.However, it is important to watch that theoscillator is actually operating at the fre-quency of the crystal. Oscillation at otherfrequencies may take place when the slugcoil in the collector circuit is adjusted in-correctly.

The output circuit on 3.5 to 4.0 Mc. isloaded with resistance to reduce the gainof the converter. The gain was originallyso high it caused cross modulation in thereceiver r.f. stage.

Note that on 80 meters the antenna isconnected through coil L to the base of themixer ( TR2 ) which, in the absence of anoscillator signal, acts as an r.f. amplifier.There are several advantages in using the80 -meter connection shown in the sche-

LIALIB 1,4, 5 3

CURRENTS

TR 1-1.6 MA.TR2- 1.0 MA.TR3- 1.4 MA.

-12V.

.005

220L2A10.!

L

L2A TR 2 L4

.005

39K of

411Xie

.005 2 1C.7- quo0 M.

XIF

1010

1111

L3

Figure 5.15-A

SCHEMATIC DIAGRAM OF THE ALL -BAND TRANSISTOR CONVERTER

Transistors can be 2N384, 00619, 2N371 or 2N1745. All capacitors marked Mshould be mica, others may be disc ceramics (d) or similar.

001D

Radio Handbook All -Band Converter 107

L 65 turns, #36 scc jumblewound on 1/4" form, tap at 10 turns.

11 40 meters, 32 turns, #36 scc jumble or pi wound, tap at 5 turns.120 Same as Li but tap at 8 1/2 turns.13. 15 turns, #22 enam., closewound.

11b 20 meters, 20 turns, #26 enam. closewound, tap at 3 1/2 turns.126 Same as Lib, tap at 7 1/2 turns.13b 15 turns, #26 enam. closewound.LI, 15 meters, 14 turns, #26 enam. closewound, tap at 2 turns.12, Same as Li tap at 7 turns.13, 8 turns, #26 enam. closewound.11d 10 meters, 9 turns, #26 enam. closewound, tap at 1 1/2 turns.12d 10 turns, #26 enam. closewound, tap at 4 1/2 turns.13d 4 turns, #26 enam. closewound.L4 40 turns, #36 scc jumble or pi wound, 1/4" wide, link 8 turns.Xtals International Crystals, type FA9.XI. 40 meters - 11.0 Mc.

Xlb 20 meters - 10.5 Mc.X1, 15 meters - 17.5 Mc.Xld 10 meters (28.0 to 28.5 Mc.) 24.5 Mc.Xi e 10 meters (28.5 to 29.0 Mc.) 25.0 Mc.X11 10 meters (29.0 to 29.5 Mc.) 25.5 Mc.

Figure 4.15-B

COIL AND CRYSTAL CHART FOR THE CONVERTER

SHOWN IN FIG. 4.15-A

matic. (1) No extra wafer is required onthe bandswitch. (2) The gain on 80 metersis approximately equal to that obtained onother bands. (3) Wiring is extremelysimple.

Note that the 10 -meter crystals eachmake use of the same oscillator coil. Also,only one 10 -meter coil is used in the r.f.and mixer stages. This system has provenadequate. The extra coils which were orig-inally in the circuit were removed and theswitch sections were rewired.

It is emphasized that in numerous teststhe transistor converter was as good as sev-eral tube counterparts. In addition heatervoltage is not required. The only powerrequirement is a supply of 12 volts at 4Mc. This may be a simple power supplyas shown in Fig. 4.15-D.

Construction The construction shownHints in the photograph is prob-

ably the ultimate in sim-plicity. At the same time it is effective,small, yet easy to work on. A piece ofaluminum measuring 5 1/2" x 5" is thebase plate upon which everything is mount-ed. The various components are mountedupon a terminal strip measuring 5" x 2".The method of mounting components isshown in Fig. 4.15-C. This allows a cleanboard layout which slips easily beneath thewavechange switch. The ground connec-tions are made by drilling through the luginto the base plate, tapping the hole, andinstalling a suitable bolt. These bolts servealso to hold the terminal strip in position.If a terminal strip is used with rivets whichgo through the board, insulating material


NI

CHASSIS VIEW OFTHE CONVERTER

BEFORE THE BAND -CHANGE SWITCH

WAS INSTALLED

must be placed between the board and thealuminum.

It is suggested that the transistors firstbe temporarily mounted as shown in thephotograph. This leaves the openings be-tween the switch sections clear and allowseasy access with a soldering iron. If thecoil forms are different from those speci-fied and it becomes necessary to cut andprune to get the coils right, the switch maybe left out and the coils temporarily con-nected in the circuit. They may then be ad-justed, sealed, and later connected to theswitch. Several converters have been builtwith the coil data provided. Each func-tioned without coil pruning of any sort.

Transistor sockets are not necessaryand were used in the prototype convertermerely to allow experimentation with dif-ferent types of transistors. If the converteris to be a unit separate from a receiver, itmay be encased by a U-shaped cover. Theconverter shown was built into a receiver.The convenient size allows either under-

chassis or above -chassis mounting.

Other The observant reader will haveBands noticed the provision for another

band. There is no reason whycoils for the 10 Mc. section of the short-wave bands could not be installed to permitreception of WWV. Even 6 meters may betuned, although some experimentation may

13- TO I-2 a-

IP I IFigure 4.15-C

COMPONENT LAYOUT ON THETERMINAL STRIP

The strip stock is similar to J.W. Miller#460.

Radio Handbook All -Band Converter 109

TOP VIEW OF THE COMPLETED CONVERTER

be required with the capacitive matching toachieve a good noise figure.

Adjustment Switch to the higher 10 -

meter crystal, and adjust the10 -meter coil Lad until stable oscillationis obtained. Signals should be apparentimmediately at any setting of the r.f. andmixer coil slugs. The signals should notshift more than a few cycles when the oscil-lator slug is tuned. If the slug is too farinto the coil, spurious oscillation may re-sult. There is no doubt about the correctposition once it has been obtained. Itshould only be necessary to align the r.f.and mixer stage for maximum output onthe other 10 -meter position. These two

stages are best aligned to the center of theband. The r.f. trimmer will take care ofany peaking that is necessary at the twoends of the band.

Alignment of the other bands is accom-plished by adjusting the respective oscil-lator coils to obtain oscillation, and thenadjusting the r.f. stage and mixer coils formaximum converter output.

The output coil should be adjusted tothe center of the band covered by the re-ceiver, with the 1K resistor temporarilydisconnected. After the resistor is con-nected, the tuning is so broad that it is dif-ficult to tell when the coil slug is at thecorrect setting.


The prototype converter was designedto work into a receiver covering the range3.5 to 4.0 Mc. The top part of the 10 -meter band was omitted in order to makeavailable the spare switch section requiredfor the 20 Mc. band mentioned earlier. Ifthe remainder of the 10 -meter band is re-quired, it is necessary to provide one morecrystal socket.

If the receiver covers a greater rangethan 500 kc., some of the 10 -meter crystalsmay be eliminated.

Transistors RCA 2N371's may be usedin the oscillator and mixer

stages. However, the RCA 2N3 84 or Phiko2N1745 is recommended for the r.f. stage

SIDE VIEW SHOWING THE RELATION-SHIP OF THE COILS TO THE BANDSWITCH

if 10 -meter operation is desired. Althoughthe difference in gain is slight, there is aworthwhile improvement in the noise fig-ure at this frequency. AVOW 0C169'salso functioned very well in the converterand were those in use when the photo-graphs were taken. No doubt other high -frequency transistors will also performwellin the circuit without alteration.

The Placement of the output coil wasOutput a problem. Finally, it was placedCoil under the switch between the

mixer and the oscillator switchwafer. This proved to be an ideal location,as it permitted very short leads withoutcramming.

It is very important that shielded wirebe used between the output coil and thereceiver. Unshielded wire will cause pick-

Radio Handbook Autodyne Converters 111

0 +I6.3 V.A.C. 25JJF

12 WVDC

D

CATH,

02CAT H.

100 LIF12 WVDC

Figure 4.15-DThis circuit will be extremelyuseful when operating the con-verter in conjunction with a vac-uum tube receiver. The simplevoltage doubler supplies 12

volts d.c. by rectifying the 6.3volt filament potential.

12 V.

up of numerous 80 -meter signals whichwill ride over those coming through theconverter.

4.16 AutodyneConverters

The autodyne converter is a "some-thing -for -nothing" circuit. That is, a singletransistor is used for both the mixer andoscillator. This is relatively easy to accom-plish in vacuum tube circuitry because thetube exhibits nearly constant gain, and fair-ly high impedances are involved. However,the transistor has a decreasing gain withfrequency. Its terminal impedances varyconsiderably, and there is much less input-output isolation than the vacuum tube e-quivalent. Thus, the adjustment of tran-sistor autodyne tends to be critical, andoccasionally the device refuses to oscillatefor no apparent reason. Even with its faults,it is useful in certain applications and shouldnot be neglected.

One example of a transistor autodyneconverter for 27 - 30 Mc. is shown in Fig.4.16-A. Signals from the antenna appearacross LI and are coupled to the transistorbase. Energy in the collector circuit is

coupled to L2 through the feedback wind-ing. The emitter, which is connected toL2, reinjects signals into the transistor

where they are again amplified, creating afeedback loop. Thus, the stage oscillatesat a frequency determined by L2 and itsassociated components. The oscillatorenergy combines with the incoming sig-nal to produce an intermediate frequencyin the manner of all superheterodynes.

The remainder of the circuit is straight-forward with the exception of the outputtransformer T1 . Pin 4 was used ratherthan the customary pin 5 (the tap) for thecollector connection. This permits capaci-tor Cl to act as both a bypass for the endof the collector feedback winding and as atuning capacitor for the primary of T1.Capacitor C1 is built into the transformer.

The Autodyne Converter is quite use-ful in the VHF region where compact, low-cost equipment is often required. SeveralGerman portable radios are being market-ed in this country and employ novel frontends for the 88 - 108 Mc. f.m. band. En-terprising experimenters could easily adaptthese circuits for other VHF applications,even though no coil winding data is avail-able.

23MC. INPUT

+10V. -10 V.

L1=12 T. *24 INSULATED, SPACED TO OCCUPY 5/16'. TAPPED AT 3 T.LS -6 FORM 1/4 DIA.

L2= 86W MIN/DUCTOR 03004 11T. FEEDBACK COIL 3T. *24 INSU-LATED, SPACED DIA. OF WIRE. WOUND OVER TOP END 0E413004.EMITTER TAPPED AT 3/4 TURN FROM COLD END.

T 1= 0.455 MC. ME/SSNER *16-9014 OR N. MILLER 2041.

RLF 4700.

Figure 4.16-ASCHEMATIC DIAGRAM FOR A

TUNABLE 27-30 Mc.AUTODYNE CONVERTER


WHIP ANTENNA

6,75 MC. TRAPL4

VHF TRAP1/4 A STUB

50

0061521,43842N1745 L3

500 -9V

Figure 4.16-BA ONE -TRANSISTORV.H.F. CONVERTER

This converter could be adapted toeither two or six meters. The circuitis neutralized to prevent self -oscilla-

tion at the signal frequency.

500

Fig. 4.16-B is the circuit for an extremelyingenious single transistor converter withan output i.f. of 6.75 Mc. Signals are fedinto the input tuned circuit through a series6.75 Mc. i.f. trap and a shunt stub traptuned to the image frequency (or interfer-ing television signals). The autodyne cir-cuit is very similar to the unit is Fig. 4.16-A. However, the output i.f. transformeris resonated by circuit stray capacitance andacts as an r.f. choke at the signal and oscil-lator frequency. Feedback occurs betweenthe collector and emitter. Notice that theoscillator coil contains a neutralizationwinding. The input (base) circuit is neu-tralized through a 14 mfd. capacitor toprevent the converter from oscillating atthe signal frequency.

An autodyne converter and r.f. ampli-fier are shown in Fig. 4.16-C. Both tran-sistors operate in a common -base configu-ration. The autodyne stage does not re-quire neutralization because L2 is verylightly coupled to the emitter.

The autodyne principle can also be ap-plied to crystal -controlled converters. Theoscillator -mixer shown in Fig. 4.10-A wasdeveloped by ZL1AAX to simplify circuitryand minimize the number of transistors

used in the device. The circuit differs fromthose just discussed in that feedback occursbetween collector and base, while the sig-nal is applied to the emitter. The outputcoil L3 , although tuned, represents a fairlyhigh impedance at the oscillator frequencyand permits the stage to oscillate at thecrystal frequency.

A third overtone autodyne circuit de-veloped by W6TNS is shown in Fig. 4.16-D. The circuit is similar to 4.10-A. How-ever, a resonant circuit tuned to the thirdovertone of the fundamental crystal fre-quency is inserted in series with the col-lector signal path. This raises thecollectorimpedance at the third overtone, permittingthe stage to oscillate at this frequency.

The circuit has one unusual featurewhich is the subject of a patent application.Although not immediately obvious, the os-cillator coil is in series with the i.f. signaland the image. If the crystal is on the highfrequency side of the signal, coil L2 can beused as an image trap, in addition to itsnormal function. It is possible to obtainan image rejection approaching 40 db. at30 Mc., even with a 455 kc. i.f. system. Inproduction equipment, rejection ratios ex-ceeding 30 db. can easily be obtained.

006152 0384201742

276

3-12 10

500 2.56 4.7KqT Hit40 470

160

006152N3842N1745

T1

7211.470

Figure 4.16-CAUTODYNE F.M. CONVERTER

This converter employs an r.f. ampli-fier and is designed to operate from

a 6 -volt source.

Radio Handbook Six -Meter Converter 113

Li21 2N1745

.005

.1-12 V.

27N 100I

M

Ti

L

Figure 4.16-DA THIRD OVERTONE

AUTODYNE CONVERTERThe values shown are correct for the27-30 megacycle band. Both coils are13 turns, #26 on a 5/16" form. Thebase tap on Li is one turn up from

the ground end.

4.17 A TransistorizedSix -Meter Converter

This converter employs three PhilcoMADT v.h.f. transistors and operates at asupply voltage of 12 volts. A communi-cations receiver capable of tuning the 7 to11 Mc. frequency range can be used as thei.f. system.

How it A Philco 2N1742 transistor isWorks employed in the neutralized r.f.

amplifier stage (see Fig. 4.17-A). It operates as a common -emitter stagewith the incoming signal being applied tothe base through the input tuning networkconsisting of coil L1 and shunt capacitorC2. It works with either a 50 or 70 -ohmantenna system. The output utilizes adouble tuned circuit made up of coilstuned to the desired frequency. Neutrali-zation is provided by capacitors C7 and C4.The tuned circuits are sufficiently broad-band so that they can be fixed tuned. Astandby/receive switch is inserted in theemitter lead of the r.f. amplifier.

Manual forward gain control is used toreduce the gain on strong signals. Thismethod of r.f. gain control is accomplishedin the following manner. As the collector

current is increased, the voltage betweenthe collector and emitter drops due to theseries resistors R4 and R5. Hence, thegain of the stage drops. The Micro AlloyDiffused Transistor is well suited for thistype of gain control and provides muchbetter overload performance than the con-ventional reverse gain control method.R.f. gain control R2 provides maximumstage gain when set fully clockwise. Byvarying the gain control counterclockwise,the value of collector current increases,causing the gain to drop. Emitter resistorR4 and the bias dividing resistor networkconsisting of RI, R2 and R3, provide thenecessary d.c. stabilization for ther.f. stage.

A 2N1743 is used as a mixer with thesignal being applied to the base througha tap on the coil L3. The output trans-former T1 tunes to about 8.5 Mc. andcouples the output from the collector tothe output connector. A loading resistorR9 is placed across the primary windingL4of transformer T1 to flatten the LE re-sponse of the output circuit with somesacrifice of converter gain. Emitter resistorR8 and bias dividing resistor R6 and R7provides the d.c. stabilization for the mixerstage.

Emitter injection is obtained by tappingthe emitter capacitor Cio on the oscillatortank coil L6 . An injection voltage of 0.15to 0.25 volts r.m.s. should be measuredat the emitter terminals of the mixer. If anr.f. voltmeter is not available, a test of thelocal oscillator injection may be accom-plished in the following manner. Place a3 ma. d.c. meter in series with emitter re-sistor R8 and positive 12 volts. This isdone by unsoldering the lead from R8 tothe positive 12 volts, and inserting themeter between these two points. The pos-itive side of the meter is attached to thepositive 12 -volt point. Now, increase theloading on the oscillator coil L6 by movingthe point where Cie taps onto the coil to-wards the collector end and readjust capac-itor C12 . A point will be found where the


LiANT. jeINPUT

rd 2

Cis

R.F. GAINCONTROL

R.F. AMP.PH/LCO2N1742

C4

OF -CASES GROUNDED TO CHASSIS

SHIELD

3

10.91\1 ON -OFFMA. SWITCH

-O12 V. BAT

Ti

MIXERPH I LCO

e 2N1743

L6

MA. 112 1131 .71 -

CI4

TORECEIVER

LOC. OSC.PHIL CO

2N1744

Rio

Figure 4.17-A

A TRANSISTORIZED SIX METER CRYSTAL CONTROLLED CONVERTERThe difference between total current and individual transistor current is

drawn by the resistive voltage dividers in each stage.

Parts List

C1 - 0.0054d. 70V disc ceramicC2 - 1-18 ,ufd. piston capacitor VC -

40 JFDC3 - 0.0015 /Ad. stand off ceramic

capacitor, Sprague 508CC4 - 1.2 Aufd axial ceramicC5 - 0.005 pfd stand off ceramic ca-

pacitor, Sprague 508CC6 - 1-18 ivfd. piston capacitor VC -

4G JFDC7 - 40 Nifd. ± 5%C8 - 1-18 nufd. piston capacitor, VC -

40 JFDC9 - 0.005 pfd. 70V disc ceramicC10 - 0.005 pfd. 70V disc ceramicC11 - 3 nufd. ± 5% silver mica

C12 - 1-18 /gad. piston capacitor, VC -4G JFD

C13 - 5.0 Awfd. ± 10% silver micaC14 - 470 Amfd. disc ceramicC15 - 0.005 #fd. 70V disc ceramicRI - 1200 ohm carbon 1/2 WR2 - 10K ohm potentiometerR3 - 2700 ohm carbon 1/2 WR4 - 1000 ohm carbon 1/2 WR5 - 1000 ohm carbon 1/2 WR6 - 12K ohm carbon ± 5% 1/2 WR7 - 2700 ohm carbon ± 5% 1/2 WR8 - 1200 ohm carbon + 5% 1/2 WR9 - 3900 ohm carbon 1/2 WR10 - 10K ohm carbon ± 5% 1/2 WR11 - 3300 ohm carbon ± 5% 1/2 WR12 - 1500 ohm carbon ± 5%1/2 W

emitter current begins to rise due to theincrease in the injection voltage. Any fur-ther increase will cause the emitter currentto rise higher. A tap point to use on L6 is

the one just below the point that causes theemitter current to rise rapidly. Emitter re-sistor R10 and bias dividing resistors R11and R12 provide d.c. stabilization in thelocal oscillator stage. A 43 Mc. overtone

crystal is used to control the frequency ofthe local oscillator. A Philco 2N1744 isused for the local oscillator.Performance With the gain control set

for maximum, an overallpower gain of 46.0 db was obtained at51.5 Mc. The gain drops off to about 43db at 50 and 53 Mc. The noise figure ofthe converter is about 4.0 db with a 0.4

Radio Handbook Two -Meter Converter 115

ANTENNA COILLi = 8 TURNS OF BOW M/N/DUCTOR #3003. BASE TAP 3 TURNS

FROM COLD END. ANTENNA TAP 2 TURNS FROM COLD END.

INTERSTAGE COILMADE FROM A SINGLE SECTION OF BOW M/N/DUCTOR #3003.THE COIL IS BROKEN AT 8 1/2 TURNS. THE LEADS ARE THEN UN-WOUND A HALF TURN IN BOTH DIRECTIONS AND ARE USED TOCONNECT THE COLD ENDS OF L2 AND L3 TO THE PROPER POINTS.

Bd W 43003-imuumumuiMUM Minn

TO L2 L3 TOCOLLECTOR BASE

TO TO TO TO TO

CR G7 GND G9 CsL2= &TURNS OF BOW AIIIVIDUCrOP #3003L3= 8 TURNS OF BOW M/ANDUCTOR #3003. MIXER BASE TAP

AT 2 TURNS FROM GROUND.

OUTPUT TRANSFORMER TI

L4 =CLOSEWOUND#30 NYCLAD COPPER WIRE ON A 1/2" DIAM.FORM TO OCCUPY A WINDING LENGTH OF 15/18.

L5= IS TURNS 0E4028 NYCLAD COPPER WIRE CLOSEWOUNDOVER COLD END OF L4.

TO GND.

TO COLLECTOROf MIXER

TO GND.

TO OUTPUTCONNECTOR

FORM =CAMBION L57IRON CORE=CAM5/ONX2018D RED DOT.

OSCILLATOR COILL6 = 9 TURNS 880/ M/N1DUCTOR *3003. MIXER EMITTER CAPA-

CITOR TAP 1/3 TURN FROM GROUND END.

Figure 4.17-BSIX METER CONVERTER COIL DATA

,v, sensitivity for 10 db signal to noiseratio.

This converter was designed by J.Specialny, Jr. and described in Philco Ap-plication Lab Report #667.

4.18 A TransistorizedTwo -Meter Converter

This 144 to 7 Mc. converter providesexcellent results in the 2 -meter band. Tran-sistors are used throughout, and the onlysupply voltage necessary is a 12 -volt battery.

How It The circuit (Fig. 4.18-A) is con -Works ventional and no difficulty should

be experienced in duplicating it.A Philco 2N1742 is employed in the r.f.amplifier stage which is fixed neutralized

by capacitor C5. Capacitance dividers C1and C2 provide a 50 -ohm match to the in-put circuit. Coil Li and capacitor C3 formthe input tuning. The base of the amplifieris tapped on Li to match 75 ohms. CoilL2 and capacitors C7 and Cg tune the out-put of the amplifier. A portion of L2 to-gether with neutralizing capacitor C5 formthe neutralizing network. The base of thePhilco 2N1743 mixer is tapped down onL2. The output of the mixer is coupledfrom the collector by capacitor C10 and out-put coil L3 at 7 Mc. Winding L4 providesan output at 50 ohms to permit couplingto the input of a communications receiver.

A Philco 2N1744 is employed as a localoscillator and operates 7 Mc. higher thanthe signal frequency. Coil L5 and capaci-tors C12 and C13 form the tank circuit.

The local oscillator signal is injectedinto the mixer emitter through capacitorC1 iby tapping the oscillator coil L5.

Operation The r.f. bandpass is about4 Mc. at the 3 db point. A

communications receiver capable of tuningthe 7 Mc. band should be used as the i.f.

L I. 4 TURNS Isle BARE COPPER WIRE 1/4.I.D., WINDINGLENGTH 1/4.. BASE TAP I TURN FROM GROUND END OF LI.

L2 = 8 TURNS *16 BARE COPPER WIRE 1/4" WINDINGLENGTH 4/104.GROUND TAP 4 TuRN5 FROM COLLECTOR END.OUTPUT TAP 3/4 TURN FROM GROUND TAP.

L3= 8.30 NYCLAD CLOSEWOUND TO OCCUPY I/2" OF WINDINGSPACE ON A 3/8" COIL FORM (CAME/ON LS -5), SEE BELOWFOR CONSTRUCTION DETAILS. (RED Dor cox, )

TO GND.TO OUTPUT

SEC.

TO GND.

TO COLLECTOR

PRI,

L4= TURNS1430 NYCLAD OVER COLD END OF L3.L5. 41/2 TURNS 4418 BARE COPPER WIRE 1/4" I.D., SPACED TO

OCCUPY 5/8".EMITTER TAP 1/8 TO 1/4 TURN FROM GROUND END.

NOTE I IN TUNING THE 7 MC. OUTPUT CO/L, THE POWDERED /RONIS VAR/ED 50 THAT IT MESHES ONLY THE COLLECTOR ENDOP L3.

Figure 4.18-BTWO METER CONVERTER COIL DATA

(Courtesy of Philco Corp.)


PH/LCO2N1742

LI

1=6 Ral.66.8K 3.9K 1.5K 5000

9k -CASE GROUNDEDNOTE -CAPACITY VALUES /N LLUF.

PN/LCOLa 2N1743

12 VOLTS

50I1OUT

Figure 4.18-A

SCHEMATIC FOR THE TWO METER TRANSISTORIZED CONVERTERThe variable frequency local oscillator permits use of virtually any i.f.

(Courtesy of Philco Corp.)

R91.8K

C1

Paris List- 6.8 ##fd. mica ± 5%

C14

C17

- 1.2##fd. =601 ceramic- 1.0 Mid. mica

C2 - 22 Aufd disc ceramic R2 - 3.9K 1/2 waft carbonC3,8 - 1.0 - 8.0 Hifd. Tublar trimmer R1 - 6.8K 1/2 watt carbon

Erie #532-8 R3 - 1.5K 1/2 watt carbonC4, 6, 9, 11, 15, 16

- .0054d. disc ceramic 70VR4

R5

- 15K 1/2 watt carbon- 4.7K 1/2 watt carbon

C5 - 5.0 ##fd mica + 5% R6,9 - 1.8K 1/2 watt carbonC7 - 5.0 #fifd. mica + 5% R7 - 18K 1/2 waft carbonCio - 30 Mufd. mica ± 5% for 7 mc. i.f. R8 - 4.7K 1/2 watt carbon

output TR1 - 2N1742C12 - 6.0 #fifd. silver mica ± 5% TR2 - 2N1743C13 - 1.5-3.0 ##fd. air variable TR3 - 2N1744

system. If a fixed -tuned converter oper-ation is desired, the tuning range will belimited to about 2 Mc. with the mixer out-put coil used. The frequency range of 144to 146 Mc. can be tuned without touchingthe converter once the local oscillator fre-quency has been set. The i.f. system thentunes from 6 through 8 Mc.

If continuous tuning of the converter isdesired, a vernier dial and a panel can beadded to the converter. The communi-

cations receiver in this case is a fixed tunedi.f. system operating at 7 Mc.

A standby/receive switch should be lo-cated in the positive leg of the 12 -volt sup-ply. The coaxial antenna switching relayshould be located as near as practical tothe input terminals of the converter.

Performance The power gain at 146 Mc.is about 30 db and falls off

to 27 db at 144 and 148 Mc. The noisefigure of the particular 2N1742 used was

Radio Handbook 220 Mc. Converter 117

5.0 db at 200 Mc. and the overall noisefigure of the converter should be no greaterthan 5.0 db at 144 Mc. It is believed thatthe newer Phiko MADT type T2028 r.f.amplifier will provide a noise figure sub.stantially below 4.5 db at this frequency.

This converter was designed by J.Specialny, Jr. and was originally describedin Philco Application Lab Report #651.

4.19 A 220 Mc.Transistorized Converter

This converter operates at a supply volt-age of 12 volts and works into acommuni-cations receiver capable of tuning the 10to 15 Mc. frequency range.

How it The unit (Fig. 4.19-A) employsWorks five transistors, all operating in

Transistor TR1 , operating as a neutralizedr.f. amplifier, is coupled to the mixerthrough a double -tuned circuit. This meth-od of interstage coupling is preferred be-cause of its ability to reject signals outsidethe r.f. bandpass and to minimize feed -

through at the if. frequency. The antennais coupled to the amplifier through a tapon the input coil L1. Shunt capacitor CItunes the input circuit to the proper fre-quency. A series matching capacitor C2 ap-plies the incoming signal to thelow imped-ance base which typically is about 60 ohms.Neutralization is provided for by a capaci-tor network consisting of C3 and C6. Neu-tralization provides an increase in r.f. pow-er gain of approximately 3 db as well asgood circuit stability, although the ampli-fier would be stable if neutralization werenot used.

The r.f. amplifier output circuit is tunedby inductor L2 and capacitors C4 and C5.Manual r.f. gain control is incorporated toreduce the gain on strong signals. Themethod used here, forward gain control,is used because of its excellent overload

characteristics. The term forward comesfrom the fact that the collector current isincreased to reduce the stage gain, ratherthan decreasing the collector current as isdone in the reverse method. A resistor R5is inserted in series with the output circuitand the negative terminal of the powersupply. As the current increases by adjust-ment of the gain control potentiometer R3,the voltage available between the collectorand emitter of the r.f. stage decreases. Thiscauses the gain to drop. This drop inpower gain is nearly linear, as the collector -to -emitter voltage is varied from 8 volts to0.5 volts. Resistor R4 provides emitterstabilization and resistors RI, R2 and R3determine the biasing level. The value ofcollector current varies from 2.5 to 6 ma.depending on the setting of R3. The nor-mal operating value is 2.5 ma. for maxi-mum gain. A standby/receive switch is in-corporated in the emitter lead.

The output of the r.f. amplifier iscoupled to the mixer transistor TR2 byloosely coupling mixer coil L3 to amplifiercoil L2 (see coil data, Fig. 4.19-B). Capac-itor C7 tunes coil L3 and the value of capac-itor C8 is selected to match the input re-sistance of the mixer. The local oscillatorpower is injected into the emitter terminalby returning bypass capacitor C13 to groundthrough a tap on coil L8 . An i.f. frequencyof 10-15 Mc. was selected. Coil L4 andcapacitor C9 tune the collector output tothis frequency range. The output is coupledto the load through coil L5, which is woundover the cold end of coil L3 . The 3 db i.f.response of the converter is about 3 Mc.Since most of the activity is centered around221 Mc., the i.f. response was peaked to11 Mc. The r.f. response at the mixer baseis quite flat from 219.5 to 225.5 Mc.

Emitter resistor R8 provides d.c. stabili-zation and resistors R6 and R7 determinethe operating point.


INPUT220-225 MC.

R. F. AMPLIFI ERPH ILCOT1832

2N17422 TR I !SHIELD

R.F. GAINCONTROL

1 1114 T5-,- =

0..".R4 '---1

.152.5 MA.ti

RsSTANDBY-- =RECEIVE ---SWITCH

TC14

ON/OFF S

12 V. BATTERY

L3

IB

OVERTONE XTAL, 52.5 MC.-101-C2C

F= 52.5 MC.TR

R9 'Rio

TR42N1744

T1859T1695

C 6

MA. R12 R R14 IC17

F=105 MC.

MIXERPHI LCOT1833 -2N1743TR a

TR52N1744

T1859T1695

C21

F 1 F =10-15 MC.

" RECEIVERTCI-1-t_r

- LI=C13

SHIELD

011

1.12 11 2.0MA.

F = 210 MC.

HARMONIC GENERATORSECTION

iR151R16'R171115

is

Figure 4.19-A

A TRANSISTORIZED CONVERTER FOR 220 Mc.The local oscillator energy is furnished by a threetransistor harmonic generator.

The harmonic generator section pro-vides at least 180 millivolts r.m.s. of in-jection voltage to the emitter terminal ofthe mixer TR2 . The local oscillator fre-quency is on the low side and the outputfrequency is 210 Mc. This high frequencyoutput is obtained through the use of twostages of frequency doubling and a singlestage overtone oscillator operating on afrequency of 52.5 Mc.

Transistor TR3 is used in the crystalcontrolled oscillator circuit. Coil L6 andcapacitor C10 are tuned to 52.5 Mc., theovertone frequency of the crystal. The os-cillator output drives TR4 through cou-pling capacitor C20 The output is tunedto a frequency of 105 Mc. by coil L7 and

capacitor C11. The 105 Mc. output fromfrequency doubler TR4 is used to driveanother frequency doubler TR5 throughcoupling capacitor C21. The output fre-quency of TR, is tuned to 210 Mc. by coilL8 and capacitor C12

Emitter resistors R11 , Rig and R17 pro-vide the necessary d.c. stabilization andbiasing resistors R9, R10, R12, R15 andR16 determine the biasing current of theirrespective stages.

The actual collector current flowing intransistors TR4 and TR5 is influenced tosome extent by the level of r.f. excitationfrom the oscillator TR3 , since a combin-ation of fixed and self biasing is employedin these stages.

Radio Handbook 220 Mc. Converter 119

10

TOCOLLECTOR

TOMIXER

I

TO GND.

pql M.4

TOCOLLECTOR

SEC L5TO GND.

TO OUTPUTCONNECTOR

PLS5FORM

LI-B TURNS 420 TINNED COPPER WIRE 3/16" I.D. 1/2. WINDINGLENGTH. ANTENNA TAP 1 TURN FROM GROUND END.

L2-3 1/2 TURNS*26 TINNED COPPER 5/32.I.D., 1/4. W,L.L3m4 1/2 TURNS*24 TINNED COPPER 5/32"1.0., 1/4" W.L.L2 AND L3 FORM A DOUBLE TUNED CIRCUIT. AIR WOUND IN THESAME DIRECTION. SPACING BETWEEN COILS AS NOTED ABOVE.

L4.72 TURN5*34 NYCLAD COPPER WIRE CLOSEWOUND. W.L.ABOUT 5/B" ON 3/B" CERAMIC FORM (CAAIB/ON TYPE PCS-52C42) POWDERED IRON SLUG.

L5=.15 TURNS ik.34 NYCLAD COPPER WIRE CLOSEWOUND OVERGROUND OF L4 (SEE SKETCH ABOVE)

L6=.9 TURNS 03005 MIN/DOCTOR ( Raw) OR A/R DUX #476r.L7=3 TURNS OF*3003M/N/DUCTOR (B&W)OR A/R OUX #416T.LB -U TURNS * 18 TINNED COPPER WIRE 1/4" I. D., 1/2" W.L.

TAPPED ABOUT 1/4 TURN FROM GROUND.

Figure 4.19-B220 Mc. CONVERTER COIL DATA

Alignment The sweep generator methodof alignment is suggested in

up the converter. However, theunit can be tuned up fairly well by peaking

it up on a carrier from the transmitter orfrom an r.f. signal generator.

If a variable capacitor is used for C2,alternately adjusting C1 and C2 for maxi-mum output should peak the input prop-erly. The point -of -best -noise figure shouldcoincide very nearly to the point of maxi-mum power gain. The noisefigure shouldbe in the vicinity of 5.5 to 6.5 db.

Performance The overall power gain isabout 22.0 db. An addi-

tional 1.5 to 2.0 db can be realized by in-serting a series -tuned 11-12 Mc. trap be-tween the mixer base and ground becausethe input circuit does not completely shortthe 12 Mc. input admittance of the mixer.The additional tuning procedure involveddid not warrant the addition of the trap.

This converter was designed by J.Specialny, Jr. and was originally describedin Phi/co Application Lab Report 079.

CHAPTER FIVE

Transmitters

When this book was started, high pow-er r.f. transistors were still laboratory cur-iosities. Shortly thereafter companies suchas Pacific Semiconductors, Texas Instru-ments, Fairchild, RCA, and many othershad developed devices capable of generatingmany watts of r.f. at very high frequencies.The work of these companies has openedthe way for developing new designs whichwill not only render obsolete present tubeequipment, but will provide considerablesavings in bulk, weight, power consump-tion, and, in many cases, cost.

To properly utilize the r.f. power tran-sistor, it is necessary to readjust one'sthinking to new concepts in transistortheory. It is the purpose of this chapterto introduce these new concepts.

This photograph shows the waveformat the collector of a class B

r.f. amplifier.

5.1 R.F. Power Amplifiers

An r.f. power transistor, whether oscil-lator or amplifier, is rarely matched to theload. The load value is a function of thedesired power output and is found fromthe formula:

RL = 0.5 Ke2Po

where:RL = the load resistanceVice = d.c. collector voltagePo = required power output

The transistor may be considered aswitch which turns the d.c. on and off atan r.f. rate. Assume that 5 watts is re-quired from an amplifier and that the powersupply is 12 volts. Fromtheaboveformula,

RL - 0.5 x 122 14.4 ohms5

In order to develop the 5 watts of power,the transistor must "look into" a load of14.4 ohms. If, however, the load has animpedance of other than 14.4 ohms, animpedance transformation must be madeusing transformers, pi networks, or othermatching devices.

The output from a single transistor op-erated in Class B or Class Cwillbe a seriesof half -waves, as illustrated in figure 5.1-A.

RF Power Amplifiers 121

One purpose of a tuned circuit is to, by-virtue of its flywheel action, restore themissing half -cycle. A waveform other thana sine wave contains a great number ofhar-monics. These harmonics are eliminatedwhen the waveform is converted to a sinewave by means of resonant circuits. Torestore the waveform to a sine wave effi-ciently, the output circuit must have a rea-sonable Q. From the formula:

Q = RLX,

it will be seen that to maintain Q at a givenfigure, a reduction in RL necessitates a de-crease in X, . The term X, , in this in-stance, represents the reactance of eitherthe coil or the capacitor. When the loadresistance is very low, "C" becomes veryhigh and is often impractical in value. Rep-resentative Q figures range from 3 to 15.

By transposing the formula for Q, wefind that:

XC = RL

Q

Assuming a Q of 5 and a load resistanceof 14.4 ohms, X, will have a value of 2.8ohms. At a frequency of 5 Mc. this wouldbe represented by a tuning capacitor of ap-proximately 0.01 pfd. This is impracticaland would result in a coil so small that itwould be made up almost entirely of thecapacitor lead inductance.

A low value of load resistance may beeffectively transformed to a higher value bytapping the collector down the coil. If thecollector is tapped midway down a coil,the 14.4 ohm load resistance will be trans-formed to 57.4 ohms (turns ratio2). Anyvalue of load resistance may be transformedto almost any higher value, and the L/Cratio of the tuned circuit may be chosenaccordingly.

The preceding discussion has conveni-ently ignored the unloaded Q of tuned cir-cuit components. A complete explanationof unloaded and loaded Q is beyond thescope of this book. However, in mostpractical cases, if the unloaded Q of thecomponents is at least 10 times the loadedQ, the preceding formulas may be con-sidered sufficiently accurate. It is a simplematter to obtain unloaded Q's of 60 ormore. Coils should be largeand airwoundor, as an alternative, wound on some ofthe newer very high Q ferrite toroids.

Bias The primary differenceConsiderations between the various clas-

ses of transistor opera-tion is that of bias. A Class B amplifieris operated at the bottom of its Vbe -I,curve. Because the transistor is not con-ducting during one half of the input cycle,heating is reduced and efficiency, comparedwith Class A, is increased. Efficiency maybe further increased by biasing the transis-tor even beyond the cutoff point so thatconduction takes place only for a portionof one-half cycle.

Bias may be supplied from a fixedsource or be self -developed. The circuitshown in figure 5.1-B obtains its bias fromself -rectification in which a voltage is de-veloped across the 100 -ohm resistor in thebase circuit.

Figure 5.1-BLarge signal NPN r.f. amplifier with

collector impedancematching to the tank coil.

122 Transmitters The Transistor

OUT

Figure 5.1-CA bandpass coupler will greatly

attenuate harmonics with little or noinsertion loss.

The Input The input impedance of anCircuit r.f. power transistor is usu-

ally quite low, and suitableimpedance matching must be employed formaximum power transfer from the driver.Matching may be accomplished by any oneof the usual methods, as discussed in Chap-ter Three. Some germanium r.f. powertransistors have a fairly low base -to -emitterbreakdown voltage. In this case, care mustbe exercised to avoid high values of resis-tance in the base circuit, across which ex-cessive self -bias may be developed. This isquite important. Several transistor break-downs in some of the authors' experi-mental breadboard units occurred beforethis fact was realized.

The Output As mentioned earlier, theCircuit output of the Class B and

C transistor stage is not asine wave and the output circuit should beof high Q to prevent the transfer of har-monics. This is particularly importantwhen the tank circuit feeds an antenna. Notonly will the harmonic radiation cause in-terferance with other stations and services,but it will also subtract from the availablefundamental power. Figure 5.1-A is a

photograph of the waveform at the collect-or of a 10 -watt r.f. power amplifier. A watt-meter connected to the output of this trans-mitter would give an entirely false reading.The harmonic output would be very high.In addition, where the power is read bymeasuring the r.f. voltage across a knownvalue of resistance, the actual voltage read-

ing could be greatly in error. This is be-cause the half -wave type of VTVM measuresonly one half of the waveform. Readingswould differ according to whether the pos-itive or the negative half were being read.

To convert the pulse -type waveform offigure 5.1-A to a sinewave, it is often nec-essary to incorporate an additional tunedcircuit between the transistor and the an-tenna in the form of a band-pass coupleras shown in figure 5.1-C. As an alterna-tive, the pi coupler may be used to removethe harmonics. However, when the im-pedance of the load is similar to that re-quired by the transistor, it may be impos-sible to construct a single pi network ofsufficient Q. In this case a double pi, inwhich the input impedance is stepped upto a higher value (about 10 x RL) and thendown to the antenna impedance, may beused. The waveform at the antenna may beviewed conveniently on an oscilloscope fordistortion, a direct indication of harmoniccontent. Preferably the oscilloscope shouldhave provision for direct connection to thedeflection plates unless it is capable of am-plification of frequencies at least six timesthe amplifier output frequency.

Neutral- The subject of neutralizationization and unilateralization was dis-a nd cussed in Chapter Three. Im-U nilateral- proper neutralization of anization r.f. power amplifier may cause

a severe loss of transistorgain. At radio frequencies, feedback maybe degenerativeand yet become regenerative

Figure 5.1-DA neutralized version of the

r.f. amplifier shown in Fig. 5.1-B.

Radio Handbook RF Oscillators 123

when the output circuit is tuned. In addi-tion to the neutralization methods discussedin Chapter Three, neutralization may beaccomplished as shown in figure 5.1-D.In this instance, a tuned circuit consistingof the inductance L and the base -collectorcapacitance is resonated at the operatingfrequency. At resonance, Xc = XL. It willbe seen that the internal capacitance of thetransistor (between base and collector) iseffectively cancelled. The arrangement issimilar to that employed with tube cascodeamplifiers.

5.2 R.F. Oscillators

Almost all the preceding remarks con-cerning amplifiers apply to the oscillator.This is true whether the latter is crystal -controlled or self-excited. An oscillatormay be considered as an amplifier, inwhich energy is extracted from the outputand returned to the input. The fact thatthe input is supplied from the output cir-cuit is of little consequence to the transis-tor. The input of the transistor should bematched to the driver. The driver here isthe output circuit and, accordingly, an at-tempt should be made to effect a match be-tween the base and the output circuit. Infigure 5.2-A, this is accomplished by ad-justing the ratio of the feedback windingturns to that of the collector winding. In

the Hartley circuit of figure 5.2-B, the im-pedance is adjusted by varying the B feed -

point. This point, as far as r.f. is con-cerned, is connected to the emitter via thebypass capacitor and the common ground.Thus, the coil is divided each side of theemitter. The impedance of the windingfeeding the base will be reduced as the tapis moved toward the bottom of the coil.The Colpitts circuit is matched in a simi-lar manner by selecting the ratio of the twocapacitors across the tuned circuit. Thisis shown in Fig. 5.2-C.

Crystal To convert a self-excited oscil-Control lator to crystal control, it is nec-

essary to insert the crystal inseries with the feedback path. The crystal,to all frequencies except that at which it isseries resonant, will represent a high im-pedance. Oscillation at these frequenciesnormally will not be possible. At reso-nance, the impedance of the crystal may bein the range of 50 to 2,000 ohms. In thethree preceding oscillator circuits, the crys-tal would be operating in the series reso-nant mode. To use the crystal in the paral-lel mode, it would be connected in place ofthe inductance in Fig. 5.2-C. In this case,feedback would still be controlled by theratio of C1 to C2.

In these circuits, crystals can also beoperated on overtones, or odd multiplesof the fundamental frequency. For ex -

Figure 5.2-AA tickler coil providesfeedback in this oscil-

lator circuit.

Figure 5.2-BThe Hartley oscillatoremploys a tapped coil.

Figure 5.2-CThe Colpitts circuituses a capacitivevoltage divider toestablish the level of

feedback.


aFigure 5.2-D

A grounded -collector r.f. am-plifier which is oft en confusedwith the common -collector

configuration.

ample, a 9 Mc. crystal can be made to os-cillate at 27 Mc. by resonating the tunedcircuit at approximately the third harmonicof the crystal frequency. Crystals whichhave a fundamental frequency above 25 Mc.are extremely difficult to make, for they be-come quite fragile. At frequencies above15 Mc., overtone operation is desirable,because it avoids the use of expensive crys-tals and eliminates the components of afrequency multiplying stage. Since un-wanted frequencies can feedback throughthe holder capacitance, crystals used in-

overtone service may require neutralization.The effect can be eliminated by placing aninductance across the crystal and resonat-ing the holder capacitance slightly abovethe overtone frequency.

If the feedback voltage is not exactly inphase with the input, it will be necessaryto increase the amplitude of the feedbackin order to maintain oscillation. At thehigher frequencies, undesirable phase shift

aFigure 5.2-E

The grounded collector r.f.amplifier is redrawn here toshow the actual circuit more

clearly.

may take place in the transistor, necessi-tating increased feedback. A point isreached, however, where all the availableoutput energy is being fed back to the in-put, leaving little power to be delivered tothe load. In circumstances such as these,it is usual to correct the phase of the feed-back with an L/C or other network so thatthe feedback arrives in phase w ith the signal.

Circuit When analyzing crystal os-Configuration cillator circuit diagrams,

it is helpful to break theoscillator down into its basic type. Gen-erally, it may be stated that all the well-known oscillator configurations, such asColpitts, Hartley, or Ultra Audion, maybe found in transistor circuitry. However,it should be remembered that the transis-tor circuits may not, at first evaluation, ap-pear to be any known configuration. Thisis brought about by the fact that the capac-itance which exists between the elements ofthe transistor may eliminate the need foran external capacitor. In the circuit of Fig.5.2-C, for example, C2 may be made upentirely ofbase-emitter junction capacitance.When changing a transistor for anothertype in an existing circuit, it is well to re-member this point. The replacement tran-sistor may have considerably different in-ternal capacitances, and may result in im-proper operation. This is true in any r.f.circuit which employs transistors. In high-er power r.f. oscillator and amplifier cir-cuits, there is a notable tendency to op-erate transistors in the common -emittermode and thus take advantage of the high-er power gain. However, near the limitingfrequency of the transistor, greater powergain will usually be obtained from thecommon -base connection.

The common - collector connectionshould not be confused with the grounded -collector common -emitter stage. This isshown in Fig. 5.2-D. The circuit is re-drawn in Fig. 5.2-E, and inspection willshow that the circuits are identical and areof the same configuration as in 5.1-B. The

Radio Handbook Linear Amplifiers 125

fact that the collector is grounded does notaffect the transistor configuration. Oftenit is convenient to connect the collector ofthe transistor directly to the chassis in ord-er to obtain better heat dissipation.

5.3 Linear Amplifiers

Transistor linear power amplifiers poseproblems not always encountered with oth-er stages. As the name implies, the stagemust amplify the signal presented to the in-put of the transistor in a linear manner.Linear amplifiers may be operated Class Aor B but never Class C. The single -endedClass B stage may be confusing to thoseused to audio circuitry. In this case themissing half cycle is replaced by the tunedcircuit. It is essential, therefore, that thetank circuit of a Class B linear amplifierhave a high Q.

Low level Class A stages are readily de-signed to amplify a signal in a linear man-ner. An example is the LE stage in a tran-sistor receiver wherein the transistor is op-erated well within its capabilites and at cur-rents very close to 1 ma.

When the transistor operating level isincreased, non -linearity of transistor char-acteristics becomes a problem. Fig. 5.3-Ashows a curve for collector current plottedagainst collector voltage. The base currentis the running parameter on which a loadline has been drawn. The subject was dis-cussed in Chapter Two. The base currentmust swing from 0 to 10 ma. to obtain thegreatest output. Note the shape of the out-put waveform over this range of base cur-rent. The effect is due to the crowding ofthe base lines as the base current increases.It is pointed out that this is an exaggeratedcase but none -the -less valuable as a lesson.The set of characteristic curves, in this case,belongs to a silicon transistor but is plot-ted at low voltage levels. For comparison,the high voltage characteristics are shownin Fig. 5.3-B using the same transistor.Note the evenly spaced base lines and linearoutput. In selecting a transistor for usein high -power linear amplifier service, onemust make a close study of the transistor'scharacteristic curves.

The linearity of the amplifier may beconsiderably imp roved by applying negative

500

450

400

350CC

CC 300

U

O

LOW VOLTAGE

, BASECURRENT

EMMEN' MA Min.5 MA

MIEMEN 4-4MmWA IW/ 1WA250

- 200U 150O 100

U 50

MA

rzsmoommilm1/11111MENIMME

IMMEMENE I B 0

MA

0 I 2 3 4 5 6 7 6 9 10

VCE -COLLECTOR VOLTAGE -VOLTS

Figure 5.3-ABASE CURRENT CURVES FORLOW COLLECTOR VOLTAGES

Note the severe non -linearity.

OUTPUTCOLLECTORCURRENT


feedback around the stage. Negative feed-back has the effect of spacing the base linesevenly. However, in some amplifiers, un-wanted phaseshifts in associated compon-ents may make the application of negativefeedback difficult, especially at the higherfrequencies.

The EmitterFollower

Considerable output may beobtained from a transistorconnected as an emitter fol-

lower if due attention is paid to correct im-pedance matching. The one -hundred per-cent inherent feedback makes it necessaryto apply considerably more drive to thestage than is required in the common -emit-ter connection.

Bias If the linear amplifier is op -Co nsid er- erated Class A, the stage mustations be biased so that the collector

current will swing equally ineach direction. This corresponds to the15 ma. base current in Fig. 5.3-B. ClassB operation demands that the stage bebiased close to cutoff. As in class B audiostages, a small fixed bias should beappliedto the transistor to lift the operating pointabove the curved portion of the transistorcharacteristics.

When the transistor stage is operatedclass B, it must be provided with a stiffsource of bias. Any resistance in the basecircuit will allow a self -bias to be developed

F0UT

Figure 5.3-CIn linear amplifier service, a stablebias point can be obtained by employ-ing a zener or silicon diode, as shown

here.

in addition to the fixed bias. On large sig-nals, the self -bias may cause the stage toenter the class C region, thus creating se-vere distortion of the waveform.

In the interests of temperature stabili-zation, an emitter resistor is mandatory.Unfortunately, an emitter resistor will causea bias which is subject to variation withchanging signal levels. This may be over-come by shunting a zener diode across theresistor so that the bias level is held con-stant.

However, a zener diode connected atthis point will subtract from the usefulnessof the resistor as a temperature stabili-zation control. When a low emitter bias isrequired, the barrier potential of a german-ium or silicon power diode may be usedto hold the voltage constant. This arrange-ment is shown in Fig. 5.3-C. Thepotential

60

55

4 5

I._ 45zLLI 40CC

CC 357

30cr0 25

141

_J0

0

15

10

61 IOU VOLTAGE

A

MNIIMMENNEMINNI

IPERMINIENSEINII10111.102:31.21M11111111EMIles- 2° MIA

15 MAI.1MAA05 MA

iimiNUINIMIEN111IB= 0

0 5 10 15 20 25 30 35 40 45 SO 55 60VC E - COLLECTOR VOLTAGE- VOLTS

Figure 5.3-BThe same transistor, as in Fig. 5.3-A,when operated at high collector volt-ages. Notice the improvement in

linearity.

Radio Handbook Modulation 127

drop across the diode will be approxi-mately 0.6 volts for silicon. Two or morediodes may be connected in series. Whendifficulty is experienced in finding a ger-manium diode capable of carrying the cur-rent, the base -emitter or base -collector junc-tion of a germanium power transistor maybe utilized.

The transistor linear amplifier shouldbe driven from a higher impedance sourceso that base -emitter impedance variation isbut a small percentage of the total imped-ance in the circuit. This will prevent dis-tortion of the signal due to thenonlinearityof the base -emitter junction.

5.4 Modulation

Transistor r.f. power amplifiers may beeither high or low level (or efficiency) mod-ulated. The low level system, though in-efficient, requires only a small modulatingpower. High level modulation is effieientand generally is more easily put into op-eration. However, modulator power out-put requirements are high. More than50% of the r.f. power amplifier d.c. inputis required for 100% modulation.

The Low An r.f. power amplifier mayLevel be low-level modulated by ap-System plying the modulating signal

to the base of the stage. Anaudio signal at the base has the effect of in-creasing and decreasing the base bias withaudio signal. It will be obvious that thecollector current must be able to follow thebase excursions if modulation is to takeplace without distortion. The transistorcollector current must have a mean valueabout which the current swing may takeplace. Thus, the transistor is not able tooperate at maximum level and its efficiencyis low.

If the current gain of the transistor isnot constant over the entire range of col-lector current, the r.f. envelope will not be

MODULATOR FINAL DRIVER

Figure 5.4-AWhen high-level modulat-ing transistor transmitters,it is usually necessary toalso modulate the driver

and even the oscillator.

an exact replica of the modulating waveform, and distortion will occur. In Fig.5.3-A it was shown that decreased spacingbetween the base lines, especially at thehigher current levels, created harmonic dis-tortion. A low level modulated stage, dur-ing the higher peaks of modulation, willalso have a non-linear output. Accordingly,the stage should be operated within thelimits of the transistor. The transistorcharacteristics illustrated in Fig. 5.3-B showexcellent linearity over a wide range of col-lector currents.

The High High level modulation is a

Level process in which an externalSystem voltage supplied by the mod-

ulator is added to and sub-tracted from the r.f. amplifier collector volt-age. Because extra power is supplied bythe modulator, the r.f. amplifier is operatedat greater efficiency than is possible withthe low level system. However, on positivecycles of modulation (NPN case), the col-lector voltage is doubled during 100%modulation peaks, and the collector -to -basebreakdown voltage is equal to four timesthe supply voltage. Linearity of the stagewill be affected by crowding of the baselines on the Ic / Vce characteristic curves.Positive modulation peaks should not beallowed to run the transistor into the non-linear area. Similarly, on negative peaks ofmodulation, the transistor should not be


MODULATOR FINAL AMP.

AUDIO CHOKE

Figure 5.4-BAn audio choke will improve quality

by preventing saturation of themodulation transformer core.

allowed to saturate, for distortion will oc-cur. The transistor contains some resis-tance between the emitter and the collector,and therefore the collector voltage cannotbe reduced to zero. The above require-ments make it difficult, if not impossible,to obtain 100% modulation when the tran-sistor is operated at or near maximum dis-sipation. In practical cases, 70%-85%modulation is normal.

Modulation percentage maybe increasedby simultaneously applying power from themodulator to a driver stage. See Fig. 5.4-A. This is accomplished by providing ei-ther a tapped or a separate secondary wind-ing on the modulation transformer.

No hard and fast rule can be statedabout the division of modulator power be-tween the final and the driver. This is de-pendent upon a number of factors. It isbetter to apply as much modulation aspos-sible to the final. This is because modu-lation applied to the driver causes the finalto be partly efficiency -modulated. Thelimitis set by the factors outlined above.

The The modulator transistorsModulator must be capable of supplying

an output power of at least50% of the r.f. power amplifier d.c. input.In practice, due to transformer losses, con-siderably more power than this is usuallyrequired for 100% modulation. The mod-ulation transformer must be capable of sup-plying the power to the transmitter withoutcore saturation. Transistor r.f. amplifiers

draw heavy currents, and the transformercore size must be adequate for this. Thisis an important but often overlooked point.

To prevent saturation of the core, causedby the flow of d.c. current through thetransformer, the secondary may be isolatedfrom the supply as shown in Fig. 5.4-B.The choke must offer a high impedance tothe audio frequencies.

5.5 A Transistor Phasing Exciter

This exciter has a performance equal inevery respect to its tube counterpart. Theoutput is sufficient to drive a pair of 68Q5'sto 10 watts on either 80 meters or 20meters. The exciter has no tuning controlsand has only one switch to change whengoing from one band to the other. Theentire exciter, including batteries, measuresonly 12" long, 2 3/4" high, and 5" deep.The total weight is less than 4 1/2 lbs. Theunit uses only ten low-cost transistors anddraws only 12 ma. at 12 volts. Standardcomponents were used throughout, and noattempt was made at super miniaturization."Barefoot," the rig has worked severalcountries on 20 meters with a cubicle quadthirty-five feet high. Admittedly, the con-ditions were excellent and free from QRM

Measuring only 12" long, 2 3/4" highand 5" deep, this little s.s.b. exciterhas worked several countries running"barefoot". Controls are, from left toright, 2 carrier balance potentiome-ters, sideband selector, band change

and on/off switch.

Radio Handbook Phasing Exciter 129

TRANSISTOR PHASING EXCITER

Liberal use of tag boards has allowed a compact construction. Large space at theleft is the 12 -volt battery compartment. The small space is for the v.f.o.

at the time, but it does indicate that the ex-citer has creditable performance.

How It Apart from the first stage, theWorks audio amplifier preceding the

audio phaseshift network is con-ventional transistor circuitry and was de-scribed in Chapter Two. The first stagewas designed to be driven by a crystal mi-crophone, and it was necessary to raise theamplifier input impedance. This was doneby applying feedback via C1 to the base ofa common collector stage. The 100K re-sistor in series with the microphone alsoincreases the input impedance. In addition,it acts as an r.f. filter in conjunction withthe 5K bypass capacitor. The capacitor al-so restricts the high frequency response.This is necessary to meet present day band-width requirements.

The audio phaseshift network was con-structed with 1% high stability resistors and2% silver mica capacitors. The compon-ents were taped together and placed in analuminum can which may be seen beneaththe 9 Mc. crystal in the center of the photo-graph. The values shown in Fig. 5.5-Aare also used in the Central ElectronicsPSN, which can be substituted for thehomemade network. Other networks havethe wrong impedance for use in conjunc-tion with transistors.

The 0.01 //id. capacitor across the pri-mary of transformer T1 is necessary to fur-ther limit the high frequency response.This value was found to be most effectivewith the particular transformers and micro-phones used. It is possible that a differentmicrophone or other makes of transform-ers may require a slight change if identical

100K

MIC.

5 K

Di

C12

Q3

T00

7526

109

2610

96/

100K

47K

10K

SW

IR

FC

A50

0.U

H

A47

0 P

F

5000

1J'H

-1-1

,R

FC

B

r-T

-1K

qau

--1

Qa

2N 2

74

4 M

C. =

100K

L3Q

92N

371

100

04-

-

1 5M

I.5M

+ -

100K

8

D1,

2,3,

4,P

31K 47

0 P

F

2,17

5 P

F

LaB 12

V.

B-1

2 V

. 100

K

500

PF

3001

1.Q

10O

N 2

74-

Q4/

Q5

OC

75

'S

B 12 V

.

8-12

V. L6 20

06/0

710

0KT

26/

1

100

K B-1

2 V

.T3

6/

R.F

. OU

T.

7

INO

UT

l_60

7.5

AU

DIO

PH

AS

E-S

H I

FT

NE

TW

OR

K

Fig

ure

5.5-

AT

his

two

band

pha

sing

type

s.s

.b. e

xcite

r w

ill d

rive

two

EL

84/6

1305

tube

s to

10

wat

ts in

put u

sing

onl

y 10

low

-cos

t tra

nsis

tors

.

100K

,218

133.

3 K

Phasing Exciter 131

Parts List - Figure 5.5-A

D1,D2, 03, D4 - Germanium diodes,1N34, 1N35, 1N198, 0A85 orsimilar.

Li - 0.8 microhenry. 9 turns #18enamel wound on the threadedend of a 1/4" bolt. Remove boltand screw in a powdered ironcoil slug which has an externalthread.

12 - 15 turns #22 enamel 3/4" longon 1/2" diameter slug tunedform. Link 5 turns #22 enamelclose wound over cold end.

Total 6 turns. Each winding 3turns. #22 enamel to a windinglength of 3/4" on a 1/2" slugtuned form.

L4 3/4" length of #30 on 3/8" form,slug tuned. Secondary, 7 turns#30, close wound, cold end.

15 Pi or jumble wound coil #26 ny-lon covered wire. 1/4" widewinding in the center of a 3/8"slug tuned form. (Wire length 6')Link 8 turns of the same wire asclose as possible to primary.

13

16 - 16 turns #22 enamel wire spacedto occupy 3/4" on 3/8" slugtuned form. Link 4 turns overcold end.

P1 , P2, P3, P4 - Carbon or wirewound.

01,02,03,04,05,06,07 - May bereplaced with 2N1274, 2N408or similar.

T1,T2,T3 - 10,000 ohm primary. 6/1

turns ratio.

R1 - 100 ohms, 1 watt, carbon. Selectvalue on ohmmeter as close tothis as possible.

C2 - Made up of 150 pfd. and 25sifd. capacitors in parallel. En-deavor to match these compon-ents within 5%.

T, B -

Figure 5.5-BThe audio preamplifier compon-ents are mounted on the termi-nal board as shown. The stripfollowing the PSN isconstructed

in a similar manner.

results are to be obtained. A tone fed intothe microphone jack should produce ap-proximately 1.5 volts of audio at the secon-dary of Ti well before distortion occurs.The total d.c. current drain of this sectionis 2.7 ma. at 12 volts. Production spreadin the transistors and components maymodify these figures somewhat, but largedifferences in the readings would indicatesomething wrong. Potentiometer P1 is theratio control, pertinent to most phasingrigs whether they are tube or transistortypes. The control may be either carbonor wirewound. Its purpose is to dividethe audio into two sections of unequal am-plitude. This allows the audio phaseshiftnetwork, which has more attenuation inone leg than the other, to deliver two out-puts of equal amplitude.

Following the PSN are two emitter fol-lowers, Q4 and Q5. Emitter followers wereused here because of their high input im-pedance. They are followed by Q6 and Q7,which are amplifiers operating with a highnegative feedback. This has the dual pur-pose of stabilizing the transistor gains andraising the input impedance. If the inputimpedance of Q6 and Q7 is raised, thisnaturally increases the value of the load im-pedance on Q4 and Q5. The input imped-ance of Q4 and Q5 is beta times the load


impedance. Thus, the impedance changeis reflected through to the audio PSN. Itis necessary to avoid shuntingthe PSN witha load, for this will depreciate the sidebandsuppression considerably.

The potentiometer P2 corresponds tothe audio balance control in a tubecounter-part and insures that the outputs from thetwo channels are equal. The output of T2and T3 is more than ample to drive the di-ode -balanced modulators.

Transistor Q8 is a 9 Mc. crystal oscil-lator. The associated components havebeen very carefully adjusted to provide op-timum output. This oscillator feeds ther.f. phaseshift network, which is made upof C2 , Li , and RI. When a capacitor andresistor in series are shunted across asource of r.f., a 90° phaseshift occurs onlyif the circuit is unloaded. When the cir-cuit is loaded (in this case by balancedmodulators), it is necessary to insert re-actance of the opposite sine in the networkto correct the phaseshift and maintain equalamplitudes. The slug in Li is a vernierand its adjustment is not critical.

Select the diodes by measuring theirforward resistance on an ohmmeter. Theresistances should be within 10%. Thecombining coil in the output is bifilar

wound, and the two capacitors across itshould be matched on a bridge. If thebridge is not available, the capacitors may

Figure 5.5-C

THE BIFILAR-WOUND

BALANCED MODULATOR COIL

TO DIODES

1K1 RFC

41111.---F-1

L3

L-1-

1.

Note how th e startof one winding isconnected to thefinish of theother.The link is woundover the center.

be selected by placing them across an os-cillator and noting the frequency shift thatoccurs.

The link winding from the balancedmodulators provides a good impedancematch to the base of the 2N371 mixer. The2N274 5 Mc. crystal oscillator has beendesigned to provide optimum mixer in-jection. The r.f. signal at the collector ofthe mixer should be approximately 4 voltswith full carrier insertion. This is morethan sufficient to drive the tube amplifiershown in Fig. 5.5-E. In the initial stagesof developement, a 2N371 r.f. amplifierwas added to the mixer. This proved notonly unnecessary but undesirable. It com-plicated the switching and made swampingof the mixer output necessary to preventflat topping. The increase in output wassmall.

The second audioboard. This boardfollows the audiophaseshift network.The input is at the

right


Construction The mechanical construc-tion of the exciter is un-

orthodox but has many advantages overnormal construction. With this system,miniaturization is accomplished withoutcreating a "rat's nest", and sections areaccessible for testing or alteration. If alarger physical size can be tolerated, thereis no reason why conventional constructioncannot be employed.

Where possible, sections are built uponterminal strips which are in turn bolted toaluminum shield partitions 2 1/2" high.The terminal strips are held away from themetal by small spacers slipped over thebolts. This leaves sufficient space to clearthe components, which are mounted onboth sides. A layer of thick paper is gluedto the partitions wherever there is a possi-bility of shorts occuring. The shield par-titions are held to the base plates by twoself -tapping screws. A section may be re-moved easily by undoing these two screwsand then lifting up the strip and aluminum.Whenever practical, the wires are connect-ed at one end of the board and are longenough to allow the sections to hinge.Audio The audio amplifier, com-Amplifiers prising Ql, Q2, and Q3 is

on a strip at the left rear ofthe chassis. The layout is shown in Fig.5.5-B. Small diameter coaxial cable con-nects the crystal microphone jack (leftfront panel) to the base of Qi through the8 #fd. capacitor and the 100K resistor,which are mounted on the terminal strip.The coaxial cable may be seen forming ahalf circle near the 9 Mc. crystal. Thecable enters the audio section at the righthand end. Transistor Qi is at theleft handend of the strip. The strip material onwhich the components are mounted meas-ures 2" wide, with lugs along the twoedges spaced 3/8" apart. Grounded ter-minals are connected to the mounting bolts.

The potentiometer at the end of thestrip and close to the battery compartmentis the audio balance control, P2. The ratiocontrol, P1 , and the transformer, T1 , are

beneath P2 . The audio phaseshift networkis in the aluminum can beneath the crystal.The wires are of sufficient length so thatwhen the strips are removed the wires maybe left connected and the circuit operatedin this manner.

Transistors Q4 , Q, , Q6, and Q7 are onthe 3 1/2" long strip second from the rear.The wires connecting T2 and T3 may beseen passing through the rubber grommetat the left hand end of the partition. It isnecessary to undo these three wires whenremoving the strip. The wires at the righthand end are of sufficient length to allowthe strip to hinge upwards. T2 and T3 aremounted on a small bracket attached to theshield partition, allowing them to be re-moved as a unit.

The small terminal strip at the right ofT2 and T3 contains the r.f. phaseshift net-work and the 9 Mc. crystal oscillator com-ponents. The coil is mounted in the com-partment beneath the 2N274 transistor andis attached to the bracket which holds thecrystal. The bracket is fixed to the shieldpartition. This whole unit may belifted outby removing the two self -tapping screwsat the bottom.

The bifilar-wound balanced modulatoroutput coil L3 (immediately behind the mi-crophone socket) is in the front compart-ment along with the germanium diode -bal-anced modulators. These diodes are onthe terminal strip behind the carrier bal-anced modulators. These diodes are onthe terminal strip behind the carrier bal-ance potentiometers. The control near thecenter is the sideband selector switch.

Figure 5.5-DTHE OSCILLATOR COIL, 12

TO PSN

011111117

4*0Lz

8 -

The link is woundover the cold end.Note that the hotend is closest to

the chassis.


A coaxial cable is connected from L3 tothe mixer Q9 and is taped to the micro-phone cable. Transistor Q9 may be seenbehind the band switch, which is the knobto the left of the crystal socket. TransistorQio, the 5 Mc. crystal oscillator, is behindthe crystal socket. The slide switch nextto the crystal removes the B minus fromthe unit.

The mixer and 5 Mc. oscillator compart-

ment were the most difficult to assemble.Eventually, a new split front panel wasmade. Now, by undoing two self -tappingscrews and removing the nuts and boltsholding the controls to the panel, it can beremoved and the compartment completelyexposed.

The space at the right front of the chas-sis was left for the v.f.o. Initially, each por-tion of the rig was breadboarded while the"bugs" were ironed out. This practicewas used with the v.f.o. also. However,the lack of a suitable dial and capacitorgearing mechanism that would fit the spaceavailable has made the v.f.o. a project forthe future. The space available is 2 1/2"long, 2 1/2" high, and 2" deep. Experi-ments were conducted with variable capac-itance diodes in order to use a potentiom-eter as the tuning component. The prob-lem is a mechanical one and not electrical,for an excellent v.f.o. was built separatelyand is described in section 5.7.

Coils It is recommended that coils Li ,

L2, and L3 conform with the sup-plied data. A small shield, not visible inthe photograph, was placed between the20 -meter coil L6 and the 5 Mc. coil L4, mid-way between the 5 Mc. crystal socket andthe wavechange switch.

Balanced The particular diodes usedModulators were of English manufac-

ture. Other diodes may besubstituted without circuit alteration. Thetwo disc ceramics are mounted on the ter-minal strip containing the diodes and areconnected across the carrier balance po-tentiometers.

Adjustments Adjust P2 to the center ofits range. Use an r.f. probe

or a receiver and adjust L2 for oscillation.Turning the carrier balance potentiometersshould not stop the oscillator, although aslight change in frequency when thepotenti-ometers are at an extreme is permissible.If the frequency change is too great or theoscillator stops altogether, the link is toolarge or the coupling too tight. No troublewas experienced in three duplications ofthis portion of the circuit. With the car-rier inserted, resonate L3. Next, adjust theslug in L4 until the crystal oscillates. Thereshould be 0.25 to 0.5 volts of r.f. appliedto the emitter of Qg. Resonate the 20 and80 -meter coils for maximum output. The

The microphonepre-amplifier audio

board is fitted to asmall aluminum par -titian. The input isatthe left. Also shownare the ratio and

audio balancepotentiometers.


6BQS/EL84

Figure 5.5-EThis amplifier, if used after the exciter,will deliver a large signal to the load.Coils may be bandswitched or pluggedin as desired. Make certain that theinput and output circuits are well

shielded from each other.

tuning of these stages is fairly broad duetothe loading of the transistor. Nothing isaccomplished by tapping the collectordown the coil, for it then becomes neces-sary to load the coils with resistance tobroadband the tuned circuit. There shouldbe no difficulty in obtaining 4 volts of r.f.at the grid of a tube amplifier if a step-upmatching device, such as a link coupling,is used.

Balance out the carrier using first onecontrol and then the other, returning againto the first, and so on. The carrier shouldbalance out completely, and there shouldbe a negligible 5 Mc. component in theoutput. No trouble with self -oscillation ofany stage was experienced, and it is logicalto suppose that with such low impedancecircuits, none should occur.

Next, short out the microphone inputsocket and feed a tone through a large cou-pling capacitor to the collector of Q3. If thetone output is very low it may be fed intoan earlier stage. It is of the utmost im-portance that no overloading occurs. Over-loading creates a distortion which, whenviewed on the scope, is indistinguishablefrom that due to poor sideband suppres-sion. Adjust the tone gain so that approx-imately 2 volts of r.f. is developed at thecollector of the mixer.

Adjust the slug of Li to half way in thecoil. Adjust P1 for minimum ripple onthe scope. Switch sidebands and readjustPt . If a different setting is obtained, set ithalf way between. Go through the sameprocedure with P2. Repeat the adjustmentswith Pi . When no further improvementcan be obtained, adjust Li and switch side -bands as before. Repeat Pi and P2 adjust-ments if necessary. There should be nonoticeable ripple on a two inch pattern,and 40 db of suppression is readily ob-tainable. It is important, when makingthese adjustments, that no sideband isfavored.

Do not be concerned if P2 is not in thecenter. This might be caused by a differ-ence in the gain of the transistors. How-ever, too much unbalance is undesirable,and an attempt should be made to matchthe transistors more closely.

Batteries The battery consists of eightflashlight cells connected toget-

her and then suitably insulated from themetal compartment. This compartment islocated at the right rear corner of the cab-inet. The space around the batteries isfilled with corrugated cardboard to preventthem from moving about.

Amplifier An amplifier has been built inbreadboard fashion. It per-

forms very well and develops more than 10watts input with a B supply of only 250volts. Thought has been given to buildingthe amplifier into the battery compartmentalong with a transistor power supply. Thiswould be ideal for mobile work where anexternal battery is already available. It doespresent some difficulties concerning con-trols on the front panel. No doubt eachconstructor will have his own ideas on this.

Gain In order to keep the number ofControl controls to a minimum, no gain

control was incorporated. Thishas disadvantages, and a miniature potenti-ometer could be built in. The 10K resistor


in the collector of Q2 may be a potentiome-ter, with the moving arm connected to thebase of Q3 through the 8/Ad. capacitor. Inlieu of this, gain can be adjusted by de -tuning the grid circuit of the amplifier.

5.6 A Transistor Filter Exciter

The exciter shown in Fig. 5.6-A is pre-sented for those who prefer the filter sys-tem of s.s.b. generation. Many of the com-ments made on the phasing rig just de-scribed also apply to this unit. The exciterdevelops approximately the same poweras the phasing rig and can be used to drivethe 6BQ 5 final shown in Fig. 5.5-E.

How It The modulation section is simi-Works lar to the phasing exciter but

uses only two transistors. Theaudio amplifier consists of a common col-lector stage with additional feedback toraise the input impedance to a high figure.The stage is designed to be driven by acrystal or ceramic microphone. The sec-ond stage is a conventional common -emit-ter amplifier feeding a step-down trans-former.

Transistor TR1 is a conventional crystal -controlled oscillator which supplies thecarrier. The r.f. voltage across the second-ary winding of T1 should be at least onevolt. Different transistors or matchingtransformers may require readjustment ofthe value of capacitance connected betweenthe base of TR1 and ground.

The balanced modulator circuit is a con-ventional half -lattice. As in the phasingcircuit, the diodes should be matched forforward resistance. The 1 K potentiometerbalances out the carrier by electrically bal-ancing the bridge. In actuality only a resis-tive balance is obtained, and some carrierwill remain unless an attempt is also madeto obtain a capacitive balance. This is thepurpose of the capacitor connected between

The carrier oscillator, balanced mod-ulator and grounded -base amplifierboard (left to right). The potentiome-ter is the carrier balance control. Thetrimmer capacitor at the top of theboard also aids in balancing out the

carrier.

one side of the balanced modulator circuitand ground. With other transformers thevalue may have to be changed or even con-nected to the other side of the bridge.

The amplifier following the balancedmodulator is connected as a common -basestage in order to raise the output impe-dance. A high output impedance is re-quired to prevent severe loading of thei.f. transformer, which would result in apoorly loaded Q. The filter circuit hasbeen described in Chapter Four in con-junction with the transistor communica-tions receiver. It should be pointed outthat as the surplus crystals have aged, manyof them have changed frequency over theyears. It may be necessary to select carriercrystals to match the filter. Carrier crystalsshould be chosen to operate about 20 dbdown the filter slope.

The output of the filter feeds an ordinaryi.f. amplifier which, in turn, drives a mixerstage. Note that the mixer, without v.f.o.injection, has no bias. Bias is obtained byrectification of the v.f.o. signal. Correct

Fig

ure

5.6-

A

TR

12

N41

0-

2N13

81T

i...

6,T

j._

.r17

/10,

...-L

III--

MIL

LER

330

Cr

o--I

101

gt!L

O.C

17-1

-T

DI 4 D2

IR

FC

S 1

11.0

0111

1 ID

S1.

052.

5 M

H

2N12

7400

75N

-Io

oK10

0 K

M1C

.10

011

1005

2N12

742N

1380

SC

HE

MA

TIC

DIA

GR

AM

FO

R T

HE

FIL

TE

R E

XC

ITE

R

The

v.f.

o. is

sho

wn

in s

ectio

n 5.

8.

TR

z2

N37

10C

169

R.F

. AM

P.

3O

K -

soon

10K

GA

IN

FIL

TE

R. S

EE

SE

CT

/ON

4.1

3

9-12

V.

TR

i=

.6 M

A.

TR

z =

.6 M

A.

TR

3 =

1 m

a.

TR

a =

1.7

MA

.

Par

ts L

ist -

Fig

ure

5.6-

A

TR

32N

371

0C16

9T

zLL

ER

204

1

2I

.2.0

5 V

V.R

F

9-12

V.

TR

a2N

371

0C16

9

VO

LTA

GE

S A

RE

UN

DE

R F

ULL

CA

RR

/ER

CO

ND

ITIO

NS

,

Di ,

D2

- 1N

34, 0

A85

or

sim

ilar.

Li-

30 tu

rns

#30

enam

el o

n 1/

4" fo

rm.

Link

4 tu

rns

hook

up

wire

. Mill

er20

A00

0RB

1.

, T2

- M

iller

204

1 45

5 kc

. i.f.

tran

sfor

mer

. 10,

000

ohm

s to

600

ohm

s.


collector current, as shown in the chart,indicates proper v.f.o. drive.

The output of the mixer also containsthe v.f.o. signal, and the tuned circuitsfollowing this stage should be of high Q.

Construction The construction techni-H ints que is similar to that used

in the Phasing Exciter.This subject has been covered in the pre-ceding section. The accompanying photo-graphs show the layout of the terminalboards. The photograph of the completedexciter was not satisfactory for reproduc-tion. Unfortunately, this was not discover-ed until after the unit had been dismantled.In constructing the exciter, none of thecircuits were found to be critical. The lay-out, using the terminal strip technique, canbe fabricated to suit the builder.

Alignment Connect a VTVM to one sideof the carrier balance poten-

tiometer and adjust Ti for an indication ofr.f. output. The slug in T1 should be setso that the drive is approximately the samefor both crystals. The oscillator shouldstart immediately upon the application ofpower. If the activity of the crystals is dis-similar, it may not be possible to obtainequal drive. Next, balance out the carrierby alternately adjusting the 1K potentio-meter and associated 3-30 Nifd. capacitor.

The audio boardtakes up very littleroom under the chassis.

THE I.F. AMPLIFIERAND MIXER BOARD

A complete carrier null should beobtained.Unbalance the carrier slightly and adjustthe filter circuit as described in section4.13, using the carrier rather than a re-ceived signal as the signal generator. Peakthe secondary of transformer T2 and theoutput coil Li to the center ofthe operatingfrequencies.

The circuitry of the v.f.o. is not shownbut is described in section 5.8. The v.f.o.can be mounted on the same chassis as theexciter or fed to the exciter through a linkas shown.

To test the exciter, connect a microphoneto the input or apply a suitable audio tone.

If the diodes are not properly balanced,the carrier may be reinserted with changesin audio or drivelevels. Imbalance is causedby the differing currents flowing throughthe diodes with changes in operating con-ditions. Since the unequal diode resis-tance causes a different voltage drop acrosseach, it unbalances the bridge.

5.7 40 Meter SSB Transceiver

The authors are indebted to Mr. JoEmmet Jennings, W6EI, for the designand construction information presentedbelow.

Radio Handbook 40 Meter SSB Transceiver 139

This neat little 40 -meter transceiverpackage was designed and built by

Jo Emmet Jennings, W6EI.

The transistorized transceiver is basic-ally made from components which arereadily available from the International Cry-stal Company. Some modifications arenecessary, and these are covered in the fol-lowing paragraphs and photographs.

The Audio The microphone pre -ampli-fier contains an emitter fol-

lower driving a common -emitter stage. Theemitter follower is necessary to create ahigh input impedance, thus allowing theuse of a ceramic microphone. A variety oftransistors, other than those shown, willperform well in this circuit.

On TRANSMIT, the output of the mi-crophone pre -amplifier is connected to theinput of an unmodified International TRA-2 sub -assembly. This assembly acts as amodulator in the TRANSMIT conditionand as an amplifier to,the speaker in theRECEIVE condition. Audio is more thanample to drive the balanced modulators.The balanced modulators are conventionaland differ little from those used with tubecircuitry. Output from the balanced mod-ulator combining coil is fed to the inputof a modified TRB-1 board.

VIEW OF THE V.F.O.

Note how the components are made firm by keeping the wires short and makinguse of tie points.

.1

RE

C. I

01

EX

TE

RN

AL

FIN

AL

SW

A

CO

NT

RO

L MO

DIF

IED

CR

C-

I

7. 2

MC

1ST

R.F

. AM

P.

A01

T4

100

PF

T2-

4.71

39

330

2N24

80.

1I

JJF

15V

.

MO

DIF

IED

TR

T-

2A

LIN

EA

RT

2 M

CT

2 M

CT

.2M

CA

01T

I2N

248

Ta

2N11

43T

3

TR

sw

Im =.0

05

B.F

. O.

=2.

5 M

H

TR

O-1

H17

0M

OD

IFIE

D

7.2

MC

2N

D R

.F. A

MP

.

TR

sw

I-

. 50

PF

!TR

IM

1500

1711

-17

TU

RN

S A

WG

15

TA

PP

ED

'I

5 T

UR

NS

FR

OM

TO

P,

3 T

UR

NS

FR

OM

BO

TT

OM

,I

7 T

UR

NS

FR

OM

BO

TT

OM

,T

OR

OID

FO

RM

.

TII

_ _J

TI

A01

A01

1

L.

3900

1800

PF

01T

R s

w

V. F

. O. A

MP

.

I/0

75

I

470

200

1.'6

1

L00

It 422

00

.01

SA

RN

ES

-rA

RZ

IAN

320-

A

MO

DIF

IED

TR

B-

tA

-15V

.

"ME

CH

AN

ICA

Lb

FIL

TE

R01

ISS

KC

150

PF

.002

7

0-11

,

SP

EE

CH

AM

P.

2N10

92N

109

'RO

B15

V.

T25

0/W

TR

sw

T12

10 J

J10

270

=1

OF

AU

DIO

LJ

XM

IT9A

UD

IOI

-P

RO

DU

CT

DE

T -

RE

C.

2N14

0M

IX -

XM

IT

I

TIO

rl 1-

4:s

firc

001

7 2

MC

g

CA

RR

IER

BA

L.

1 N

35 25-7

5F

RO

M O

UT

PU

T O

FT

RA

-P

FS

PE

EC

H A

MP

.X

AU

DIO

LE

XIS

TIN

GT

RA

NS

IST

OR

TR

AN

SIS

TO

RS

TT

R s

wT

IN

TR

sw

y

Ile°

It

UN

US

ED

20 IJ

F

ON

-O

FF

,IS

v

SO

=

"00

371

41

zR15

0I r

tt

210'

150

11-11

RI'II

I (1

T.0

03 BA

LAN

CE

DI

MO

DU

LAT

OR

II.

T T

5001

14:

M.

0500

8

2111

I

Id\

-15V

.R

EC

. AU

DIO

.G2010

AU

DIO

TA

PE

R

---

NO

TE

S: r

Rsw

SH

OW

N IN

RE

CE

llaP

OS

IT IO

N.

if -D

EP

EN

DS

ON

FIL

TE

R U

SE

D.

Fig

ure

5.7-

AS

CH

EM

AT

IC D

IAG

RA

M F

OR

TH

E C

OM

PLE

TE

40

ME

TE

R S

.S.B

. TR

AN

SC

EIV

ER

40 Meter Transceiver 141


Coil #1 - 30 turns #26 wire. Tap 5turns from bottom. Coil iswound on original formwhich is taken from the In-ternational unit.

Coil #2 - 29 turns #26. Tap 4 turnsfrom bottom.

Coil #3 - 29 turns #20. Centertapped. Wound on origi-nal coil form. The capaci-tor used to tunetheantennaon the International boardis now used to tune this coil.As seen, there is a two turnlink.

Coil #4 - 35 turns #30 wire. Tap 5turns from bottom. Link 3turns. Use original coilform.

Coil #5 - 22 turns #28 wire. Centertapped. Link 4 turns. Useoriginal coil form.

Coil #6 - 25 turns #28 wire. Tapped6 turns from top and 4 turnsfrom bottom. Use originalcoil form.

Coil #10 - 25 turns #28 wire. Tap 4turns from bottom. Centertapped. Use original oscil-lator coil form.

Coil #11 - See schematic.

Coil #12 - 25 microhenry r.f. choke.

Coil #13 - 25 turns #28 wire. Centertapped. Link 4 turns. 3/8"slug tuned form mountedin i.f. can.

Coil #14 - Miniature transistor i.f.

transformer. 15,000 ohmsto 600 ohms. The capacitorinside the can is removed.

TRANSMIT/RECEIVE switch - Minia-ture ceramic wafer. DP8T.

The TRB-1 Modifications to the boardare as follows. Remove the

squelch amplifier from the board. In itsplace build the product detector. OnTRANSMIT, the product detector func-tions as a mixer to 40 meters. Next, re-move the first and last i.f. transistors andreplace them with those indicated on theschematic. The original transistors havethe base internally connected to the tran-sistor case, which, in the higher gain unit,is undesirable. The second r.f. collectorcoil is now rewound to resonate at 40 me-ters, and the second i.f. transistor is re-moved. The transistor socket is used toconnect the leads of a small sub -assemblycontaining the transistors and the CollinsMechanical Filter into the circuit. The os-cillator portion of the board is now re-moved, and the unused section oftheboardcut off.

The TRC-1 This board is modified byremoving all but one r.f.

stage. If desired, a new board maybe builtcontaining just the one amplifier shown onthe schematic. The latter is perhaps thebetter course, because the modifications tothe TRC-1 are somewhat drastic. TheTRC-1 board is used on RECEIVE only.

The TRO- The b.f.o. board is not1 H, BFO altered electrically exceptfor the addition of some decoupling com-ponents and a change of value of the out-put capacitor. The additions and changesare clearly shown on the schematic. Theexact frequency of the cyrstal will be depen-dent upon the type of mechanical filterused. Note that in the TRANSMIT con-dition, the b.f.o. is the carrier generator.Its output is connected to the diode bal-anced modulator. In the RECEIVE con-dition, the output of the b.f.o. is connectedto the emitter of the product detector.

The V.F.O. The v.f.o. is the familiarColpitts oscillator followed

by an isolation stage and an amplifier. Theoscillator tank coil is wound on a toroid


The TRA-2 board isshown at the left.The other boardsmay be seen at the

left of the chassis.

form. This results in a reduced physicalsize and a smaller electric field, allowingthe v.f.o. unit to be housed in a small con-tainer. The v.f.o. is operated from a 6 -volttap on the battery. A silicon diode is in-serted in the battery lead to prevent theflowof current when the transceiver is turnedoff. A separate switch, ganged to the mainon/off switch, may replace the diode.

V.f.o. construction is clearly shown inthe photograph. Other layouts and con-tainers may be used if due regard is paid tostability requirements. Mount all compo-nents firmly by short stiff wires. However,be careful to avoid stress on thecomponent

parts due to over -tight wires.Duringthe TRANSMIT condition, the

output from the v.f.o. amplifier is connect-ed to the emitter of the product detectorwhich now functions as a mixer. Outputfrom the mixer is connected to the inputof a modified TRT-2 board.

The TRT-2

LinearAmplifierBoard

Several modifications are re-quired on this board. Themodifications may be madein the following order:

a) Remove the crystal socket and cutoff that end of the board.

Another view show-ing the placement ofthe various boards.

Radio Handbook V.F.O. 143

b) Change the first transistor to anA01 (Philco) and the second to a2N248.

c) Remove the coils and replace with40 meter coils wound according tothe data provided.

d) Add forward bias resistors to thethree stages to operate the transis-tors class A. The resistor values areshown on the schematic.

Note that the last stage is neutralized toprevent oscillation. The output coil shouldbe wound exactly as specified so that neu-tralization is accomplished without the needfor coil or component alteration.

Alignment Alignment is most readily ac-complished if a signal gen-

erator is available. In lieu of this, the v.f.o.and b.f.o. may be set on frequency by lis-tening to their respective outputs in a re-ceiver tuned to the required frequencies.In any case, a receiver should be availableand tuned to the output of the transceiver.Coils, i.f.'s., etc., may then be peaked formaximum S meter reading. The lengthof the antenna on the external receivershould be adjusted to prevent overloadingand false meter readings.

Output from the transceiver, when thestages are aligned for maximum signal, isapproximately 1/2 watt - sufficient to drivea 6BQ5 or similar tube linear.

5.8 A VFO for AM, CW,orSSB

The v.f.o. to be described was engineer-ed to have better than average stability andlinear dial reading. As it so happens, therequirement for a linear reading dial is thatthere be very high fixed C in the oscillatortuned circuit, preferably as much as 4 timesfixed to variable. This is also a conditionof high electrical stability in a v.f.o. HighC circuits are necessary to swamp out var-iations of capacitance in the transistor it-self.

C2

L1 0-50 PRESETVAR BLE

C3220M

COLLECTOR CURRENT

TR,. 1.5 MA.

TR2=4mA.

5,

=

5.6 I(

GA -51469 NTCOPTIONAL

LOW ZOUT

2.5 PAM RFC

_L-3 '2:1

Figure 5.8-ADIAGRAM OF THE V.F.O.

Note that several volts of r.f. can beobtained by using the high impedance

output.

Parts List - Figure 5.8C1 - 385 iridd. variable. Semi -circu-

lar plates and ceramic insulation.End plates should be slotted foralignment adjustments. Broad-

cast type capacitor.- National APC-50.- 9 turns #22 enamel on 3/8" di-

ameter ceramic form. Miller,catalog #4400. Wind link of 1

turn of hook up wire over coldend of Li .

12 28 turns #30 enamel on 1/4"form. Miller 20 A000RB1 orsimilar. Link 4 turns hook upwire over cold end.

TR1 - 2N384, 2N371,0C169,2N1745TR2 - 2N371. Other transistors may

be used if the neutralizing ca-pacitor C3 is suitably adjusted.

R1 Negative temperature Resistor.Approximately 4,000 ohms.(Thermistor) GA -51489 or simi-lar.

Resistors marked 1 watt may be carbonor wire wound.

11


The v.f.o. is designed to cover either3.5 Mc. to 4 Mc. or 3950 kc. to 4450 kc.The latter range is suitable for use with455 kc. crystal or mechanical filter SSBexciters.

The r.f. voltage at the collector of theoutput stage is in excess of 2.5 volts whenloaded, which is sufficient to drive mostsmall pentode tubes. The output is morethan sufficient to drive the average s.s.b.exciter.

How It The oscillator is aparallel-tunedWorks Colpitts. This is followed by a

lightly coupled tuned amplifier.The amplifier is neutralized to prevent vari-ations in output loading from being re-flected back to the oscillator. The valuesof C3 and C4 should not be changed fromthose shown on the schematic diagram.

The The tuning capacitor must beCapacitors of the straight line capacitance

variety (semicircular plates)if a linear dial reading is to be obtained.The capacitor may be a standard broadcastunit and should have ceramic insulation,if possible, and slotted end plates. In theprototype model, a two gang capacitor wasused only because it happened to be in thejunk box. One section was left unconnect-

ed. Remove any mica trimmers which arepart of the capacitor.

All fixed capacitors in the tuned circuitmust be silver mica. The trimmer capacitorC2 is a ceramic unit such as the NationalAPC-50.Construction The construction of theHints v.f.o. must be mechanically

sound. The best electri-cally designed v.f.o. in the world is only asgood as, the mechanical construction whichgoes into it. Avoid stresses and strains oncomponents, particularly the coil, thevaria-ble capacitors, and the dial mechanism. Ifthe dial mechanism does not line up withthe capacitor shaft, it will be necessary tokeep one hand on the tuning knob at alltimes in order to stay on frequency. Con-struction may take almost any form as longas it is sturdy. If the v.f.o. can be droppedfrom a height of three inches without thefrequency of oscillation moving more thana few cycles, the v.f.o. is mechanically stable.Naturally, the unit is mounted in its cabinetfor this test. Tie points, when they supportcomponents "hot" to r.f., should be ofcer-amic or similar material.

At first, no attempt was madeto stabilizethe oscillator against thermal variations,but it was found that thefrequency shift wasconsiderable when a change in room tern -

UNDERSIDE VIEWOF

THE V.F.O.The oscillator is at

the left and the am-plifier at the right.

Radio Handbook V.F.O. 145

The tiny transmitter in the foreground is intended for satellite service and usesthe mercury batteries shown at the right. An equivalent tube transmitter would

use the batteries at the left. (Courtesy of DuKane)

perature took place. Most ofthe movementwas traced to the coil. Frequency move-ment due to the transistor itself was verysmall. After a little experimentation, a ne-gative temperature co -efficient resistor(thermistor) was added to the base circuitof TR1. This component reduces the fre-quency movement to negligible propor-tions for all but excessive variations. It isrecommended that the oscillator be totallyenclosed so that the temperature changestake some time to soak through to thecom-ponents. In this circuit, the NT C resistoractually overcorrects the transistor, and bydoing so takes into account the movementof the coil and other components.

Voltage A reduction in supply vol-Stabilization tage from 12 to 8 volts

moves the frequency of theoscillator by 800 cycles. Ifthev.f.o. is to beused in mobile work the supply should bestabilized with a zener diode. The stabilized

supply should be fed to the oscillator andthe amplifier. For normal fixed stationwork, no stabilization is necessary.

Alignment To obtain a linear dial read-ing, the slug of L1 should be

adjusted for the low end of the band andthe preset trimmer set for the high end.One adjustment will affect the other, andthe builder must work back and forth be-tween the two. If a 100 kc. crystal calibra-tor is available, align the low end to the 3.6Mc. check point and the high end to the 3.9Mc. check point. If there are variations atthe intermediate check points, these maybecorrected by bending the slotted vanes ofC1. On the prototype model seen in thephotograph, v.f.o. dial readings are accur-ate to within 1 kc. except at the very highend of the range, where the minimum capa-citance of the capacitor is too high. Onlythe last 30 kc. are affected in this manner.All other readings, right down to 3.5 Mc.,are exact.


2N706 PT880 X00 PT879'sP7716

- 12V. -1 -

Figure 5.9-AA completely transistorized 6 -watt 10 -meter transmitter uses silicon devices man-ufactured by Pacific Semiconductors, Inc. Tomodulate the circuit, break the 12 -volt

line at the points shown.


Li - 12 turns #22 1/4" form. Secon-dary 4 turns #26.

- 9 turns #22 1/4" form. Secon-dary 3 turns #26.

- 7 turns #20 on CambridgeTh erm-ionic form 227-14.

L2

L3

14 - 5 turns #20 on Cambridge Therm -

ionic form 227-14.15 - 4 turns #20 on CambridgeTherm-

ionic form SPC2-J-21.

C1 - 8-60 ,u/rfd. ARCO 404.C2 - 90-400 mufd. ARCO 429.C3 - 90-400 ,u/dd. ARCO 429.Xtal - 3rd overtone.

5.9 A 3 -Watt C.B.or 10 -MeterTransmitter

For some time, Pacific Semiconductors,Inc. have had available a line of silicontransistors which have made possible an in-expensive transmitter capable of delivering3 watts of r.f. to the antenna. With thevalues shown in the schematic ofFig. 5.9A,the transmitter is capable of operating overboth the Citizen and the 10 -meter hambands.

The The oscillator is a PT716 operat-Circuit ing as a Colpitts. The overtone

crystal is in the feedback path.Following the oscillator is a PT880 bufferamplifier. The stage is not provided withfixed bias but depends instead upon recti-fication of the signal to develop a self bias.Operation is in the region between classB and class C and is a function of drive.

Similarly, the two paralleled PT879's de-pend upon the developed bias to establishthe operating point. The use of separatebias resistors tends to compensate for un-equal transistor gains and prevents eitherone of the transistors from doing most ofthe work.

The Output The output circuit consistsNetwork of a tuned r.f. choke (low

Q resonant circuit), follow-ed by a double L network. The networkhas been designed to present a specific loadto the transistors and at the same time pre-vent harmonics from reaching the antenna.Due to the low impedances involved, thisportion of the circuit is perhaps morecomplex than would be required of tubetransmitters. However, the circuits tunebroadly and merely need peaking for max-imum output.

Input to the final is approximately 6watts and efficiency is about 50%.

Radio Handbook Two Meter Transmitter 147

MODULATORTO BUFFER TOAND FINAL OSCILLATOR

TRANSMITTER

I

Figure 5.9-BAll three stages must be modulated toreach the 100% level, as shown here.

If the transmitter is to be modulated,The modulation transformer secondaryshould be center tapped and modulationapplied simultaneously to the oscillatorfrom the tapped winding. Both the bufferand the final are modulated from the top ofthe winding. The connections are shownin Fig. 5.9-B.

5.10 A Two Meter Transmitter

It is hard to believe that low cost re-ceiving type transistors can develop suffi-cient power on two meters to light a pilotlamp, but this is exactly the case with thecircuit shown in Fig. 5.10-A. This canonly be accomplished by careful construct-ion. Thought must be given to layout and

lead lengths. Each component and circuitmust operate at top efficiency to preventcircuit losses from destroying the smallpower generated by the transistors.

Selecting transistors of low cost with ahigh alpha cutoff and low collector capaci-tance also proved to be a problem. TheRCA 2N384 type transistor proved to bean excellent oscillator and multiplier, butdid not provide sufficient power outputas a 144 Mc. amplifier. The final ampli-fier transistor which provided the greatestpower output in this application was select-ed from the available types.

Three transistors are used in a 72 Mc.oscillator, doubler, and power amplifierconfiguration as shown in Fig. 5.10-A.The oscillator, a 2N384, is connected inthe common -base circuit and employs a5th overtone crystal, which oscillates on72.02 Mc. Feedback for the emitter cir-cuit is taken from a low impedance tap onthe coil in series with the crystal. A neu-tralizing coil Li insures oscillation at thecrystal fifth overtone. The oscillator out-put coil L2 feeds 72 Mc. energy to thedoubler through a modified pi -network.

The highest doubler power output isobtained when the stage is operated be-tween class B and C. The 2N384 doublerstage operates without bias, and positive -going r.f. cycles initiate conduction.

UNDERSIDE VIEWOF THE TWO METER

TRANSMITTER

Note the spacingand coupling of coilL4. The adjustment

is rather critical.


Qt

2N384 L,Q2

2N384

SF= SELF -RESONANT FREQUENCY (SEE TEXT)

Q32N1407

-12 V. MOD.

Figure 5.10-ASCHEMATIC FOR THE TWO METER TRANSMRTER

Note that only the final stage is modulated. Some improve-ment may be obtained by also modulating the doubler.

Parts ListC1 , C2 - 15 sifd. variable

(Hammarlund MAPC 15)C3 - 5 mufd. variable (E. F. John-

son 5M11)C.4, C5 - 7-50 it#fd. rotary ce-

ramic (Centralab)Li - 21 turns, #28 enamel,

closewound on 3/16" form(1 meg., 1 watt resistor)

12 - 10 turns, #22, 5/8" diam-eter, 8 turns per inch, xtaltap 7 turns from bottom,B- tap 5 turns from bottom(AirDux #508)

ANT.

13 - 5 turns, #14, 3/8" diameterL4 - 4 turns, #16, 3/8" diame-

ter. Adjust coupling andspacing for max. drive.

L5 - 5 turns, #14, 3/8" diameterL6 - 4 turns #24 enamel, inter -

wound in L5. Cement to L5when correct coupling isobtained.

01,02 - 2N384 (RCA)Q3- 2N1407(Texas Instruments)Xtal- 72.02 Mc., 5th overtone

(International Crystal Mfg.Company)

The 144 Mc. r.f. drive from the doubleris applied to the power amplifier (a TexasInstruments type 2N1407) through a linkand impedance matching capacitor. Thecircuit is series tuned and resonated at 144Mc. This stage is not forward biased.Therefore, positive r.f. cycles cause con-duction, producing near class C operation.The output of this stage appears across L5

which is also resonated at 144 Mc. A res-onated link coil L6 couples r.f. energy tothe antenna. The series capacitor tunes outthe reactance of the link and serves as aloading control. R.f. chokes, self resonantat the operating frequency, are used forisolation in all circuits.

Construction There are no particularH ints problems in constructing

the two meter transmitter,but the buider is cautioned to keep allleads extremely short. The photographsshow one of several layouts used by theauthors. Although not especially attractivethe component crowding permits shortlead lengths. The 2N384 transistors weremounted in clips drilled from an old fuseholder. The crystal, oscillator transistor,and bias components are mounted on therear apron. The 15/i/efd. tuning capacitors,the doubler transistor, and coils are mount-ed on the top of the chassis box. Thepower amplifier transistor (mounted in a

Radio Handbook Two Meter Transmitter 149

heat sink clip), capacitor C3, antenna jack,and matching components are on the re-maining apron of the box.

The only critical component is the neu-tralizing coil L1 connected across the cry-stal. This coil must resonate with the cry-stal capacitance at 78 - 80 Mc. After wind-ing the coil, clip the leads short, solder itacross the crystal pins and check the reso-nant frequency. Add or subtract turns asrequired to tune 78 - 80 Mc.

The r.f. chokes are made by winding20-30 turns of very fine wire on a onemegohm, one watt carbon resistor. Checkand adjust the resonant frequency to either72 or 144 Mc. by adding or subtractingturns. The exact frequency is not critical.

Testing A grid dip meter is mandatorywhen working with high fre-

quency circuits. First, apply 12 volts to theoscillator only, and check for operationwith the dipper switched to the absorptionwavemeter mode. Check the second har-monic on a receiver to insure that the sig-nal is crystal controlled. Next, connect amilliameter in series with the doubler sup-ply lead and peak the oscillator capacitorCi for maximum collector current. Con-nect the supply lead for Q2 directly to the12 volt line, and connect the milliameter in

VIEW SHOWING THE OSCILLATORCOIL 12

A COMPLETELY TRANSISTORIZEDTWO -METER TRANSMITTER

This transmitter will supply sufficientpower to illuminate the #49 pilot lamp.

series with the final supply lead. Peak thedoubler tuning for maximum final current.

Finally, observe the output with a griddip meter, field strength meter or receiverS meter. Repeak all adjustments for maxi-imum r.f. output. The circuit values havebeen selected so that excessive dissipationin any stage cannot occur. The final stagewill draw about 10 ma. when fully loaded.It should be possible to see a noticeableglow in the #49 pilot lamp used as adummy load (see photo).

5.11 A Potpourri of Circuits

Although the authors feel that amateursshould know how to design their own cir-cuits, this is not always feasible. Engineersworking in plants manufacturingtransistorequipment are no different from amateurs.Everyone borrows circuits for projectswhen it will save time and get the job done.

This section presents a group of inter-esting and useful circuits which have been


1- 001

(Ti)2N716

500

(Ti)2N716

(Ti)soo 2N716

RFC g.1.2.1.)

RFC1. 2JJH

- 500C

(Ti)Soo 2N716

L300000

2 h.C4 1.5 1.2

1.1H 2H

RFC, 1.211H

Figure 5.11-A

.5_0eL4

000001.--t-1

Ce

RFC, I. 2 LH

CIRCUIT FOR THE QUARTER WATT TRANSMITTER DESIGNED FOR140 Mc. OPERATION

Parts List

C1, C2, C3, C4, C6 - 1.8-13 Aufd. trimmerC5 - 7-45 rotary ceramic trimmerC7 - 3.2-50 NM. tuning capacitorLi - 8 turns of #408 AirDux, tapped 2 turns from ground end.

12 - 4 turns of #12 enamel, 1/2" diameter, 3/4" long.13 - 3 turns of #12 enamel, 1/2" diameter, 7/16" long.1.4 - 4 turns of #12 enamel, 1/2" diameter, 3/8" long.Xtal - 72 Mc., 5th overtone type (International Crystals Mfg. Co.)

RFC211H

4-300.

extracted from company application notesand instruction books. With minor adapta-tions (in some cases) they are all useful inamateur applications. Although the circuitshave not been tried by the authors, they donot appear to present any construction oralignment problems.

VHF Circuits Fig. 5.11-A shows the cir-cuit for a 140 Mc. trans-

mitter which can be returned to 144 Mc.with no coil or tuning capacitor modifica-tions. The circuit was designed by TexasInstruments.

The oscillator, a T1 type 2N716, is a5th overtone circuit on 70 Mc. In -phasefeedback occurs between a tap on the out-put coil and the emitter through thecrystal.It is interesting to note that 70 Mc. drivefor the doubler is taken from the low im-pedance emitter in the oscillator circuit.The doubler, another T1 2N716, operatesin the grounded -base configuration, and

its output network consists of a pi sectionfeeding a 2N716 common -base amplifier.The final stage, a 2N716 common -emitterpower amplifier, drives the antenna througha pi network. Note the liberal use of r.f.chokes to prevent unwanted coupling be-tween stages. This appears to be the secretin building high efficiency v.h.f. transmitt-ing equipment. R.f. chokes are used asd.c. returns in all circuits to avoid EI lossesin decoupling resistors. The power out-put of this transmitter is 1/4 watt (250milliwatts) at 140 Mc.

Fig. 5.11-B is a common -base poweramplifier for the 100 - 200 Mc. range. Un-like the preceding circuit whiich uses allsilicon devices, this circuit employs a ger-manium diffused v.h.f. transistor. Threedifferent types are given for a variety ofpower outputs. R.f. is applied to a suitabletap on the input circuit (depending on thedrive impedance) and is coupled to Q1,through a capacitive voltage divider. This

Radio Handbook V I-1 F Circuits 151

5031_

IN

1.6-133111F Q1 0.25 LH 50 CLOUT5.0 ILUF

5 50 LAUF30 V.

2N1141 =1110 MW. OUT.2N1142= 170 MW. OUT.2N1143=160 MW. OUT.

Figure 5.11-BPOWER AMPLIFIER

This power amplifier uses low costgermanium transistors.

serves to impedance match the base of thetransistor to the tuned circuit. The operat-ing point of the transistor can be adjustedwith the 1 K potentiometer, which variesthe base bias. The output circuit, whichmatches the collector to the antenna, is api network. It is particularly important tonote that the d.c. return for the amplifieris through the load. The circuit shouldnot be operated without the load connect-ed, as excessive base bias, due to self rec-tification, may develop.

The new Motorola epitaxial mesa family,types 2N1141 through 1143 and 2N1195,make excellent power amplifiers in the twometer band. The circuit shown in Fig.5.11-C will deliver 1/4 watt to a load at 160Mc. and should develop slightly more at144 Mc. At this level, the power input is560 milliwatts (14 volts at 40 ma.) for an

I+ 14 V. -2N1142

0.11 LH

160 MC. POWER AMPLIFIER

Figure 5.11-C

QUARTER WATTOUTPUT

POWER AMPLIFIERAmplifier uses new epitaxial mesa

device.

T1657e. 3RD OVERTONE

MC.XTALRL50 n

+12V. 1C= 2.5 MA.P.O. = 7.5 MW.

CRYSTAL OVERTONE OSCILLATOR CIRCUIT

Figure 5.11-DA THIRD OVERTONE OSCILLATOR

Can be used in conjunction with warsurplus 8 Mc. crystals.

overall efficiency of 45%. The power gainis 10 db at 160 Mc., indicating that a drivelevel of 25 milliwatts would be required,neglecting circuit losses. The input cir-cuit consists of a capacitive divider similarto Fig. 5.11-B. However, base bias isprovided by self rectification of the r.f.current flowing through R h. The trimmercapacitor in the emitter circuit series re-sonates the emitter inductance, providinga true ground. The r.f. choke simply pro-vides a d.c. return. The output circuit isthe familiar pi network to match a varietyof antenna impedances. The collector isshunt fed for d.c. through an r.f. choke.

Amateurs who prefer not to use 72Mc. fifth overtone crystals will be interest-ed in a circuit described in Philco Applica-tion Lab Report 646. This report describesa circuit for operating 8 Mc. crystals ontheir third overtone. The circuit( Fig. 5.11-D) employs a Philco type T1657 transis-tor, which is believed to have been replacedby the JEDEC type 2N1866. The outputtank, consisting of L1 and C1, is tuned tothe third overtone of thecrystal, or approxi-mately 24 Mc. One and one-half turns ofcoil L1 provides a degree of regenerativefeedback for the crystal, since it is not cutfor overtone operation. Capacitor C2

matches the transistor collector to the loadimpedance. Coil L1 consists of 12 turnsof #3003 B & W "Miniductor" tapped at


Nollulu nummar

SILICON

Figure 5.11-E

GALLIUM ARSENIDE

Tunnel diode characteristic curves for silicon and gallium arsenide. (Courtesy ofGeneral Electric)

1 1/2 turns for feedback. Since the valuespecified for C2 is not readily available, a3-30 trimmer could be used. CapacitorsC1 and C2 would be adjusted alternatelyfor maximum power output, which is givenas 7.5 milliwatts. The overall efficiency ofthe oscillator is approximately 28%. Thedrive from this circuit could be applied toa suitable tripler-doubler circuit for outputon 144 Mc.The When a semiconductor diode isTunnel forward biased, it does not be -Diode have in the same manner as a

good conductor (such as a pieceof copper wire). In a wire, one electronwill "bump" the next one, which in turn"bumps" the next, and so on. This move-ment is conducted down the wire eventhough the individual electrons may havemoved only a short distance. In a semi-conductor, current flow takes place due toan exchange of electrons from one orbit toanother, which slows down the action ap-preciably.

The tunnel diode exploits thefirst meth-od of conduction. The electrons move sofast that they seem to actually "tunnel"

TUNNEL DIODESThese tunnel diodes are currentlyavailable to experimenters. The Gen-eral Electric 1N2939 (front) and theRCA TD100 (rear) are compared to acommon straight pin to illustrate their

compact size.

Radio Handbook The Tunnel Diode 153

TUNNEL0100E

Figure 5.11-F

TUNNEL DIODE TRANSMITTERThis transmitter can be used on 80

and 40 meters.

through the potential barrier of the diode.See Chapter One. The effect is conductionsimilar to that which takes place in a wire.The quantum mechanical tunneling wasfirst discovered by a Japanese scientistnamed L. Esaki . The tunnel diode is oftenreferred to as the Esaki diode.

The important feature of the tunnel diodeis its negative resistance quality. Unlike apure resistance, current in the tunnel diodecan decrease with increasing applied vol-tages. Characteristic curves of voltage ver-sus current for two different types of diodesare shown in Fig. 5.11-E. The breakup ofthe negative resistance slope is probablycaused by parasitic oscillation in the testjig. This is one of the tunnel diode'scharacteristics and it is difficult to preventthem from oscillating!

A simple tunnel diode transmitter forthe 80 or 40 meter bands is shown inFig. 5.11-F and the accompanying photo-graphs. The only parts required are thetunnel diode, a crystal for either band, anr.f. choke (2.5 mh.), and a potentiometer.

A solar operated 80 meter c.w. trans-mitter. The solar cell, in the fore-ground, is an International Rectifier

SD -1020-A.

The transmitter can be powered by eithera 1.5 -volt penlite cell or a single solar cellas shown in the accompanying photograph.

The tunnel diode requires less than one-half volt and draws less than 5 ma. of cur-rent. Obviously, it is not capable of gener-ating much power, at least in its presentform. However, Lester A. Earnshaw, whenoperating from ZL1AAX in Warkworth,New Zealand, was ableto contact Jack KingZL1AOF, in Whakatane, N.Z., a distanceof 160 miles. His RST 339 signals werelater confirmed with a QSL and tape re-cording.

TUNNEL DIODETRANSMITTER

This transmitter wasused by ZL1AAX tocontact stations 160

miles distant.


THE HEATH GW30May be easily converted to 10 -meteroperation by replacing the crystal andretuning the receiver and transmitter

coils. (Courtesy of Heath)

A Ten Meter The current crop of Citi-Phone izen's Band equipment hasTransmitter resulted in numerous wal-

kie-talkie and handie-talkiedevices. These units may be easily con-verted to 10 meters by replacing the trans-mitter crystal with a third overtone type forthe desired operating frequency. If the re-ceiver is crystal controlled, this crystalshould be replaced with a third overtonetype which is displaced from the operatingfrequency by the i.f. value. The r.f. tunedcircuits will have to be repeaked for bestperformance.

A 30 milliwatt 10 -meter phone transmit-ter, which uses inexpensive transistors, wasdescribed in Philco Lab Report #610. Thecircuit is given in Fig. 5.11-G. PhilcoMADT (Micro Alloy Diffused) transistorsare used in the r.f. section and alloy junc-

tion types in the audio section. The trans-mitter is suitable for short range civiliandefense and amateur activities on the10-meter band.

A crystal controlled oscillator drives anunneutralized r.f. amplifier, which deliverspower to the 50 -ohm load. TransistorTR3 provides a high impedance input foracrystal microphone, while TR4 serves asan audio driver for the modulator transis-tor TR5. The modulator is a series typewhich does not use a matching transformer.

The transmitter should be tuned for c.w.conditions. Point A of the emitter resistorin the final amplifier is first connected tothe battery. Adjust C1 in the oscillator formaximum drive to the final, using the cur-rent meter as an indicator. Next, tune C2for minimum final collector current, whileadjusting C3 for desired loading. The col-lector current should be 5.5 to 6.0 ma.This tuning procedure does not necessarilydeliver the maximum output to the load buthas been found to be satisfactory. A bettermethod would be to use a sensitive fieldstrength meter located near the antenna.The power input for c.w. is approximately72 milliwatts, and the measured power out-put is 52 milliwatts.

For phone operation, point A ofthefinalis tied to A , the collector ofthemodulator.Bias resistor R is adjusted to a value whichcauses the collector current to drop to one-half the c.w. value at about 3 ma. This re-sults in about 32 milliwatts input to the finalr.f. stage. The modulation envelope appearsclean at full modulation (about 85%), andon -the -air reports indicate a good qualityphone signal. Care must be exercised notto remove the antenna from the final withthe transmitter on!

Although the transistor specified is anearly type, the T1657, it is believed that theJEDEC type 2N1866 makes an excellentreplacement and may provide even morepower output.

Radio Handbook Ten Meter Transmitter 155

TR11657

50.U.UF PH/ LCO0HIC

4715

HI Z 1.01JFMIC. 7 -

GAINCONTROL

L

0.5-201J1J F

TR211657PHILCO

470 1JUFL 2 10-75 5011

11.1JF OUTL

330 I.00I .0OtI 1270 1.00101C= 3 MA.

2.515

A 12V.-VCC

TRa2N 207

Z150 KR

° TR52N 223

211F t4.7H IS fL4.7

21JF _LTT = DAW 3003 12 T.

COLLECTOR TAP 8 T.FROM COLD END.COUPLING TAP 2.5 T.FROM COLD END.

L2= t3aW 3004 10T .L3 = 4 T. *24 NYCLAO WOUND

OVER COLD END OF L2.X = 27-30 MC. 3RD OVERTONE

CRYSTAL

1111141H_

NOTES:

Figure 5.1 1-G

THIRTY MILLIWATT 10 -METER PHONE TRANSMITTER

The transmitter uses low cost germanium transistors.

CHAPTER SIX

Power Supplies

6.1 A.C. to D.C. Conversion

If transistorized equipment is used in alocation where alternating current is avail-able, there is no advantage in using bat-teries when an inexpensive power supplymay be easily constructed.

A simple transistor power supply isshown in fig. 6.1-A and is suitable for ex-perimenter applications. The regulationof the supply is ample for most needs andonly high power audio amplifiers and mod-ulators will require a larger power supply.When connected to thetransistor communi-cations receiver described in Chapter 4, theoutput voltage does not change more than0.4 volts, even when the receiver is operatedat maximum audio output.

CH

taV DC

Figure 6.1-AA SIMPLE POWER SUPPLY FOR

TRANSISTOR WORK

The transformer and choke are de-scribed in the text. The diodes maybeRCA 1N1764 or any similar 500 ma.

units.

A SIMPLE POWER SUPPLY FORTRANSISTOR EXPERIMENTS

Even without regulation the stabilityof the output voltage is excellent.

The transformer, T1 , used in the supplyis a discarded 12 -volt vibrator transformer.However, a commercial unit such as theTriad F -92A which is capable of supplyingapproximately 16 volts at more than 500ma. will be suitable.

How It The circuit is a full -wave bridgeWorks rectifier which utilized both halves

of the a.c. cycle. The rectified d.c.,which is pulsating at 120 cycles per second,requires filtering to provide pure d.c. Thefilter, which consists of the input and out-

A.C. to D.C. Conversion 157

UNREGULATEDINPUT

REGULATEDOUTPUT

Figure 6.1-BTHE ZEN ER OR AVALANCHE

DIODEThis diode can be used to reg-ulate specific voltages. The di-odes are manufactured in dif-

ferent voltage groups.

put capacitors and the filter choke, formsa single pi -network. Increasing the size ofthe input filter capacitor will increase theoutput voltage and improve regulation. In-creasing the capacitance of the output filtercapacitor will decrease the ripple and alsoimprove the regulation.

Construction Any form of constructionand layout is practical and

no special precautions need be observed.The choke may be constructed by windinga small 3 -watt audio -output transformercore full of #24 enamel wire. The wiremay be layer or jumble wound on a home-made bobbin. Once the core is strippedit should take only 10 minutes or so towind the wire and re-insert the core.

BURGESS BATTERIESThe battery is still a convenient powersupply for transistor experimenters.These Burgess units are sealed re-

chargeable nickel cadmium type

MINIATURE SILICONZEN ER VOLTAGE

REFERENCE ELEMENTSCourtesy of International Rectifier

Corporation)

When wiring the supply, do not connecteither the positive or negative lead to thechassis. The two circuits should be inde-pendent of the chassis, as shown in fig.6.1-A. This prevents transformer leakagefrom damaging transistor equipment. Inaddition, a short from the power supplychassis to the equipment being tested coulddamage the supply if this precaution is notobserved. It is advisable to insert a 1/2ampere fuse in series with the secondarywinding to protect the power supply in theevent of an external short circuit.

Voltage One of the simplest devicesRegulation for regulating a varying

source of d.c. is the zenerdiode. A property of the zener diode isthat when the voltage across its terminalsexceeds a certain level, the zener diodeconducts. The result is a regulated outputvoltage much the same as that accomplished

158 Power Supplies The Transistor

with a gaseous voltage regulator tube, al-though the mechanism for accomplishingregulation is quite different. Note that thediode is connected in the reverse direction(Fig. 6.1-B). At the breakdown potential(analagous to ionization in the voltage re-gulator tube) the current carriers avalancheand the diode assumes a relatively low re-sistance.

The circuit for a simple zener diode re-gulator is shown in fig. 6.1-B. The resis-tor, R1, prevents excessive current flowthrough the diode when the load is re-moved.

A transistor can also be used to regulatea power supply as shown in fig. 6.1-C. Asthe output voltage is decreased with in-creasing loads, thetransistor bias increases.This causes the collector -emitter resistanceto decrease. More voltage is thenpermittedto develop across the load. Thus the outputvoltage will be regulated at approximatelythe battery potential.

A heavy duty supply using a small batteryfor stabilization is shown in fig. 6.1-D.This circuit, described in a Motorola appli-cation bulletin, uses inexpensive Motorola2N554 transistors. The supply deliverscontinuously variable voltage at currentsup to 4 ameres. As in fig. 6.1-C, the out-put voltage is determined by the stabiliza-

o 0

UNREGULATED i REGULATEDINPUT OUTPUT

o T o

Figure 6.1-CA battery can also be used toestablish a specific bias voltage

for a transistor regulator.

tion battery voltage, which acts as a refer-ence for the transistor regulators. The a.c.ripple on the output d.c. will vary between4 and 28 millivolts as a function of loadvoltage and current.

This type of regulator is commonly re-ferred to as an emitter -follower wherein Q3and Q4 serve as variable resistances inseries with the load. The terminal impe-dance of the regulator is less than one ohm.A regulation curve for the supply is alsoshown in Fig. 6.1-D.

The series -regulator transistors shouldon a large aluminum heat

sink as they dissipate up to 60 watts withthe control set for low voltages. The con-trol transistors, Qi and Q2, may also re-quire heat sinking as they each dissipateapproximately 1 watt.

Note that when regulation is employedless elaborate filtering circuits are required.A simple explanation is that the ripple ap-

UNDERSIDE VIEWOF THE

POWER SUPPLY

Radio Handbook Regulated Power Supply 159

<

12

10

2N554(4)

(EMITTERS NOT USED)

C a, No:..ur

MAX. VOUT

? 6

6. 4

O 2Rz SET AT ÷ ROTATION

02 3 4

OUTPUT CURRENT -(A.)

2N554(4)

Qi Q2 Q3 Q4

0-12 V., 0-4 A.(RIPPLE 50 MV.AT 2 A.)

Figure 6.1-DA HEAVY DUTY POWER SUPPLY

A heavy duty power supply using a battery stabilized emitter follower whichdrives two parallel connected series control transistors.

pears as a varying load voltage which iscounteracted to a large extent by the regu-lator.

A zener diode may replace the batteryin the circuits described above. The actionof the zener diode is the same as in thesimple regulator in Fig. 6.1-B. That is, itestablishes a fixed potential across its ter-minals. Thus, it functions as a battery, butdoes not require periodic replacement.

A Regulated When the load on a recti-Power fier-type 12 -volt supply isSupply varied, a change in the out-

put voltage is to be expected.If the load variation is sever, as when aclass -B amplifier is connected to the supply,the voltage may fall as much as 50%.

The schematic of fig. 6.1-E shows an in-expensive regulated supply which will pro-vide excellent regulation over a wide rangeof currents to 1 amp. Even greater currentsmay be supplied by cascading regulatorstages, as outlined later.

The transformer furnishes approxi-mately 24 volts to the rectifier. The rectifiermust be capable of carrying the peak load

current, and its peak inversevoltage shouldbe at least 75 volts. Due to the capacitiveload, the voltage is higher than that nor-mally encountered in this type of circuit.

Resistor R1 is used to limit the peakcurrents to a safe figure. When the supplyis first switched on the sudden surge ofcurrent into the 2,000 pfd. capacitor mayreach many amperes without R1. The cur-rent is limited by the resistance in the cir-cuit, which is composed of the d.c. resis-tance of the transformer and the resistorR1. Without knowing all the circuit com-ponent values it is difficult to statea definitevalue for R1. However, the specified valueof 1/2 ohm will give adequate protectionfor most applications. Resistor R1 willalso provide protection against accidentalshort circuits at the output of the supply.

A regulated output is available at pointA in fig. 6.1-E. In a great number of ap-plications the current available at this pointmay be sufficient. This is especially trueif the zener diode DI is capable of pass-ing large currents. however, large zenerdiodes are somewhat expensive, particular-


ly for experimental or amateur applica-tions. Less expensive diodes may be em-ployed by using the amplification abilityof the power transistor connected in thecommon -collector or emitter -follower con-figuration. A regulated voltage applied tothe base of the transistor will be amplifiedby the transistor action. If the transistorhas a current gain of 50 and a small zenerdiode rated at 50 ma. is used to stabilizethe base, the transistor will be capable ofstabilizing the voltage at currents up to2.5 amperes. The transistor must be ca-pable of carrying the current and must beable to dissipate the power dropped acrossthe transistor.

A 1,000 ohm load resistor is used tomaintain a small but constant load at theoutput terminals.

If further control is required the out-put of the transistor may be connected tothe base of a second emitter follower. Thecontrol will then be amplified by the cur -

TT 1N1525 2N234A

Ti -110 V. -20V. AT I A.(TRIAD F-92.41

2N1381 2N234A

(AMPLIFIED REGULATOR MODIFICATION)

Figure 6.1-EA SIMPLE REGULATED

POWER SUPPLYThe series current -passing transistor isstabilized by a zener diode. Thediodeis made by International RectifierCorp. The transistor is a Bendix 2N -234A. Also shown is a simple modi-fication to increase regulation with

high current loads.

TRIADF -40X

2N301'sIA. A

ITIMPI73zoo

alPra 1000

II

OF tr 27R 1.ikN91

Figure 6.1-FAN UNUSUAL REGULATED

POWER SUPPLYWITH VARIABLE OUTPUT

The circuit was designed by C.T. Gageof Sigma Instruments.

rent gain of the second transistor. Alsoshown in fig. 6.1-E is a two -stage regula-tor made by adding a 2N1 381 transistorto the circuit. This simple modificationresults in a constant output voltage fromno load to a full load of 2 amperes.

An unusual circuit for an adjustableregulated supply is illustrated in fig. 6.1-F.It may be constructed by using two inex-pensive RCA 2N301 power transistorsand a 1N91 diode in addition to the semi-conductor rectifiers. The supply features0-25 volts output at up to 500 ma. Theoutput impedance varies between 0.5 and-2 ohms as a function of outputvoltageandload. The output ripple is less than 100millivolts peak -to -peak over the entire vol-tage and current range.

The circuit is conventional except forthe use of negative current feedback in thepositive lead. The output current passesthrough the 1N91 regulator. Thus, in-creasing load currents lowers the bias ofQ2 due to increased voltage drop acrossthe diode. This action increases the for-ward bias applied to Qi , lowering its re-sistance and maintaining the load voltageconstant. The regulation linearity is a

function of diode characteristics and anumber of units must be tried to obtainbest regulation.

Radio Handbook D.C. to D.C. Conversion 161

6.2 D.C. to D.C. PowerConversion

The power transistor is finding increas-ed use as a replacement for the vibrator ind.c. to d.c. power conversion applications.The high efficiency, lack of moving partsand freedom from interference due to arcingof the vibrator points make the transistorpower converter a useful device.

Transistor The discussion will be con -Power fined to the PNP power tran-Cony erters sistor. Except for reversed

voltage polarities, the oper-ation of the PNP and the NPN transistorsare identical.

In power -converter service operationthe transistor is nothing more than an elec-tronic switch. The collector -to -emitter re-sistance is controlled by the influence ofthe base. When the collector is negativewith respect to the emitter, and the base iszero or positive, the junction resistancewill be high. The collector -emitter junc-tion will remain a high resistance until thebase is ma de more negative than the emit-ter. With sufficient negative bias appliedto the base, the collector -emitter junctionwill saturate and exhibit a very low resis-tance. It should be stressed that a PNP

transistor, unlike the usual vacuum tube,will not conduct, and therefore not amplifyuntil a small amount of negative bias is ap-plied to the base. In a transistor powerconverter this is called the starting bias.

It can be seen that if two transistors aresubstituted for the vibrator, the only prob-lem is to turn them on and off by control-ling the base bias. The circuit shown inFig. 6.2-B accomplishes this job and iscalled a power multivibrator. As in Fig.6.2-A the battery is connected to the center -tap of the primary winding. The ends ofthis winding are connected to the collectorsof the two transistors. The circuit is com-pleted by returning the emitters to the pos-itive battery lead. A few turns are added tothe transformer to provide a source offeed-back voltage to sustain oscillation. In ad-dition, a voltage divider consisting of tworesistors is connected across the battery.This provides the starting bias so that thetransistors will conduct initially. These re-sistors play an important part in the op-eration of the supply and are explained indetail under the heading, "How to Set theStarting Bias".

When the battery is connected to thecircuit the transistor with the lower junc-tion resistance will conduct first. For the

A MINIATURETRANSISTOR

POWER CONVERTERThis converter is suit-able for receiversand QRP transmitterapplications. (Cour-tesy of Mini-Verters,

Los Altos, Calif.)


TO RECTIFIERFILTER SYSTEM

BATTERY

Figure 6.2-ATYPICAL VIBRATOR

ALTERNATOR CIRCUITNote that electromagnetic ac-tion switches the contact backand forth connecting the endsof the primary alternately to

the battery.

purpose of explanation, assume that theupper transistor, Qi , is the first to con-duct. Current flows from the negative bat-tery terminal through the upper half of the

winding, then through the transistor col-lector -emitter winding and back to the pos-itive battery terminal. The lines of forceemanating from this half of the primaryinduce a voltage in all the other windings,including the other half of the primary.The base impedance of the transistor ismuch lower than the collector impedance.Thus, the base winding requires fewerturns than the collector winding to providethe necessary step-down ratio. The basewinding leads must be connected so thatthe polarity of the voltage to the base of Qiwill be negative -going when the collectorvoltage is positive -going. It will be seenthat as the collector current increases, thebase voltage also increases. This forces thecollector to draw even more current. Whenthe current reaches a predetermined ampli-tude the transformer core will begin tosaturate. This means the corewill not mag-netize further and the current fed back tothe base from the winding will begin tofall. Thus the condition is reversed; thebase voltage will decrease and as a conse-quence the collector current will decrease.At this point Qi is no longer in a heavycurrent condition and Q2 begins to con-duct.

Notice that during most of the heavycollector current condition of Qi , the tran-sistor was saturated and that over the sameperiod Q2 was cut off.

The cycle that has just been describedwill continue to repeat until the powersource is turned off. Although many stepsare involved, the whole process happensvery quickly. The entire cycle usually takesplace in about 1/400th to 1/ 2000th of asecond.

In addition to inducing a voltage in thebase winding, the lines of force generatedby the collector current induce a voltage inthe secondary that is a function of the pri-mary -to -secondary turns ratio. It shouldbe pointed out that when one transistor isconducting and the other is cut off the col-lector current of the conducting transistorwill induce an equal voltage in the otherhalf of the primary. Thus, the end -to -endvoltage across the primary will betwo timesthe supply voltage. This explains the term2E -( E = supply voltage) that is found intransformer formulas. For example, if asupply is operating from 12 volts, and 144volts are desired from the secondary, it isnecessary to have six (not twelve) times thenumber of primary turns onthesecondary.

The number of secondary turns may beadjusted to give the desired output voltagefrom the rectifier -filter system. The wave-form applied to the rectifiers will be a near -


BATTERY

Figure 6.2-BTransistor powerconverter

equivalent of the vibrator cir-cuit. The transistors replace the

moving contact.

Radio Handbook D.C. to D.C. Conversion 163

A TYPICAL MOBILEPOWER CONVERTER

Made by Universal Transistor ProductsCorp.

ly perfect square wave because of the rapidswitching action of the transistors. A sim-ple full -wave or bridge rectifier will convertthis signal to almost pure d.c. and verysmall filter capacitors may be used to re-move the remaining "spikes".

Efficiency The efficiency of a transistorpower converter may be ex-

pressed as the ratio of power output topower input. Thus, for a given input, thehigher the power output the more efficientthe power supply. The losses always showup in the form of heat, and are referred toas IR losses.

When one of the switch transistors iscut off, it will have only a small IR loss be-cause the junction resistance is so highthat only a small current can flow. By thesame token, when the transistor is saturatedthere will be little IR loss because there isonly the small resistance of the junction.Thus the only time major IR losses occur

is when the transistor is switching from theon state to the off state, or vice versa. Itshould be obvious that the faster the switch-ing action occurs, the more efficient is thesupply. Since the efficiency figure is direct-ly related to transistor heat, it can be as-sumed that the faster the switching time,the greater the power which can be handledby a given pair of transistors.

It can be shown that toroid tape woundcores using Deltamax or similar alloys pro-duce the fastest switching times, and areused in the most efficient supplies. Cer-tain powdered iron cores are only slightlyless efficient than the tape wound style, butare far less expensive. Keep in mind thatpoor efficiency always means heating (usu-ally in the transistors) which necessitatesa reduction in the maximum power han-dling capabilities. The transistors alwaysproduce heat, and it is the junction temper-ature that determines the power handlingability of the transistors.

In effect, the transistor only works dur-ing the instant of switching and rests be-tween these periods. It would be correctto assume that the transistor will handlemore power than it would under condi-tions of continuous operation such as classA. In actual practice, it has been foundthat any transistor can safely switch fromfive to eight times its class A power rating.If the system for transferring heat is veryefficient, this rule -of -thumb may be in-creased to ten times the class A powerrating. Thus, a transistor that is rated atfive watts may switch as much as 50 watts.Naturally, two transistors would be re-quired for the power converter and theywould be able to handle 100 watts of pri-mary power.

Rectifiers Vacuum tube rectifiers areveryinefficient devices and are very

rarely used in transistor power converters.Silicon rectifiers, by virtue of the fact thatthey have a fixed 0.6 volt internal drop andno filament are ideal for this application.Silicon rectifiers may be connected in series


SWITCH INC RECTIFIED FILTERED

OTYPICAL A.G. POWERSUPPLY WAVE FORMS

0 TYPICAL VIBRATOR 0_ MWAVE FORMS

l-0 vT EY,PrI °GRA LWAPOWER

WAVEFORMS- j_

Jm

Figure 6.2-CTYPICAL WAVEFORMS

Typical waveforms encountered in

various power supplies.

to obtain a higher breakdown voltage, andbalancing resistors are not needed. Thisapplication is explained more fully in thesection devoted to rectifiers and filter sys-tems.

Circuit The circuit shown in Fig.Variations 6.2-B is called a common -

emitter configuration. Avari-ation of this circuit is the common -collectorshown in Fig. 6.2-D. In this circuit theprimary is connected between the emitters,and the collectors are connected together.For PNP transistors the common -collectorconnection must be returned to the nega-tive battery lead.

The theory of operation is the same asthe common -emitter circuit discussed ear-lier. However, this circuit has the advan-tage that the transistor case (which is thecollector) may be bolted directly to thegrounded chassis when the supply is usedwith a negative -ground automobile ignitionsystem. This eliminates the need for in-sulation of the collectors, as in Fig. 6.2-B,and allows the most effieient transfer ofheat to take place. In this configuration,the emitter impedance is lower than thebase impedance, and the transformer mustprovide a step-up of approximately 1:1.25.

If the automobile power source has anegative ground, by all means use thecommon -collector configuration and boltthe transistors directly to the chassis. Ifthe automobile has a positive ground, theneither configuration may be used. Thetransistors will have to be insulated fromthe chassis in either case.

A typical circuit for the popular Triadtoroid transformers is shown in Fig. 6.2-E.Note that the emitter -to -base impedancestep-up is accomplished by connecting theindividual base windings in series with theemitter winding. This provides an auto -transformer action. The starting bias re-sistors must be installed in series with eachbase winding.

The earlier Triad ferrite -core trans-formers were designed for common -emitterapplications exclusively. However, thesetransformers, except the TY-69S, may bemodified for use in common -collector cir-cuits. Notice that the base winding con-sists of enameled leads with two paralleled

MINIATURE SILICON RECTIFIERSThese miniature silicon rectifiers arerated at 400 volts, 500 ma. and areideal for power converter applications.

(Courtesy of International RectifierCorporation)

Radio Handbook Setting the Bias 165

conductors. Although not shown on sche-matics, the base winding is actually con-nected as shown in Fig. 6.2-F. Separatethe two outside leads ( B1 and B2 ). Withan ohmmeter establish the two windingsand determine which one is center -tapped.Tape up the center tap as this lead will notbe used. Connect an a.c. source (between6.3 and 30 volts) to one half of the secon-dary, and experimentally connect the threewindings as shown in Fig. 6.2-E. Connectan a.c. voltmeter across the full primaryand alternately reverse the two base leadson each base winding until a maximumreading is obtained on the voltmeter. Thisprocedure will phase correctly allwindingsand the transformer may be connected asshown in Fig. 6.2-E.

How to Set Recall that the transistorsthe will not conduct initiallyStarting Bias unless a small amount of

forward bias is applied tothe bases. In the common -emitter circuitthe bias is connected to the center tap ofthe base or feedback winding. The voltageis obtained from a resistor voltage dividerconnected across the battery. These resis-tors are shown as R1 and R2 in Fig. 6.2-D.

If the transistors do not have enoughbias, the supply will not start when con-nected to a heavy load. If too much bias


BATTERY

Figure 6.2-DCOMMON -COLLECTOR

CONFIGURATIONCommon -collector configuration forgrounded negative automotive ig-

nition systems.

is applied, the efficiency will be reduced.Excessive bias can also damage the transis-tors due to additional heat. More impor-tant, with excessive bias, the supply maynot be self -protecting. When the supply isoperating properly, a short or overload inthe output circuit will absorb the feedbackvoltage and the supply will stop oscillating.The battery current will drop to a low val-ue, thereby protecting the transistors fromburn out. The bias should be set so thatnone of the transistor ratings can be ex-ceeded.

You may want to substitute transistorsor optimize designs. The bias may be setin this manner. First, short the output sothat the supply will not oscillate. Then in-sert typical values, such as 2,000 ohms and20 ohms, in the R1 and R2 positions. Toa great extent, the larger resistor controlsthe static current (current drawn when thesupply is not oscillating). The smaller ofthe two resistors controls the base currentwhen the supply is oscillating. As a roughrule -of -thumb, the resistors should be se-lected so that the transistors draw theirnormal class A current when the supply isnot oscillating. To show atypical example,consider the Sylvania 2N307. It is rated attwo watts and a pair will handle four watts.To comply with our rule, the bias resistorsmust be set so that the transistors consumefour watts of d.c. from the power source.Thus, for a 12 -volt system, the transistorsshould draw approximately 333 ma. staticcurrent. Typical values for the two resis-tors, with these Sylvania transistors, mightbe 820 ohms and 10 ohms.

The next step in setting the bias is toremove the short and replace it with a re-sistor that represents the normal full load.Upon applying power, the convertershould oscillate instantly and produce fullpower in the load. If it does not, try re-ducing the value of both resistors by half.Keep the value of these two resistors ashigh as possible, consistent with the con-ditions outlined earlier. If the larger value


22011

2200.

10 LW

IOLF


TRIAD TOROIOTRANSFORMER

Figure 6.2-ECOMMON -COLLECTOR CIRCUIT

Common -collector circuit with auto -transformer stepup.

resistor is reduced too far, excessive staticcurrent will flow. If the lower half of thedivider is too small it is unable to limitbase current during signal peaks. Eitherexcess can damage the transistors.

If it is still not possible to obtain fullpower, or easy starting, this indicates thatthe transistors have very low Betas whensaturated. If this is the case, it will be nec-essary to parallel additional transistors.

If the supply does not start easily, butcan be started by removing the smaller re-sistor, a silicon power diode may be sub-stituted for the smaller resistor. The cath-ode of the diode will connect to the tap onthe base winding. In the common -collec-tor circuit, the diode can be added in serieswith, or in place of, the lower value baseresistor. See R2 in Fig. 6.2-D. In oper-ation, the diode will not conduct until thesupply has started. The forward resistanceshould be about the same as the resistor itreplaced. If R2 is normally many times theforward resistance of the diode, it shouldbe connected in series with the diode.

There is another trick that can be usedto improve starting under load. If the sup-ply refuses to start until the larger resistoris reduced in value, try this remedy. Cal-culate the value of the large resistor neededto start the supply (even though it maydraw excessive static current). Then locate

a pilot lamp, or series of lamps, that meas-ure about the same resistance. Substitutethe lamp, or series of lamps, for the largevalue resistor as shown in Fig. 6.2-G. Theidea behind this scheme is that the cold re-sistance of the bulb(s) will be low enoughto start the supply. But when it starts, thecurrent through the bulb(s) will increasethe filament resistance, thereby decreasingthe base current.

Rectifier Although most transformersand are designed to work into aFilter bridge rectifier system, you needSystems not confine yourself to this cir-

cuit. Most power -convertertransformers will work equally well withhalf -wave, full -wave, or voltage doublerrectifiers.

As a practical example, let us assumethat you have purchased a Triad TY-79transformer. In a bridge rectifier config-uration, this transformer will develop 300volts at 200 ma. Following are some ofthe things that can be done with this trans-former and a discussion of the advantagesand disadvantages of the various rectifiersystems shown in Fig. 6.2-H.

Half Wave With a half -wave rectifier sys-tem (a), you will obtain the

same voltage and current as with thebridgeconfiguration. However, only half cyclesare rectified and the output is more difficultto filter. You will save on rectifiers, foronly half as many are required (two inseries, with the Ty -79).

Full Wave The full -wave rectifier is

shown in (b). In compari-son to (a) this configuration will produceone-half the d.c. output (150 volts), buttwice as much current (400 ma.). Theamount of filtering required is the same asfor the bridge circuit. Only two rectifiersare required, one in each leg. Smallamounts of negative bias (10 to 20 volts)can be obtained by connecting a resistor in

Radio Handbook Filter Systems 167

series with the grounded secondary center -tap connection. This resistor must be by-passed with an electrolytic capacitor, 50/dd., 50 w.v.d.c., positive end grounded.

Voltage The voltage doubler circuit (c)Doubler is very popular as it reduces

the total number of rectifiersneeded for a given output voltage. If youuse the voltage doubler with the TY-79, itwill be possible to obtain twice the voltage(600 v.) at one-half the current (100 ma.).The filtering will not be as good as thebridge system, and it will be necessary touse capacitors that are four times as largeas in the bridge and full -wave circuits.For short duty cycle applications, such aspower for a s.s.b. or agated final amplifier,you can use very large filter capacitors (80/Ad.) and obtain peak currents equal to thebridge system (200 ma.). A power con-verter suitable for use with a s.s.b. linearamplifier is shown later in the chapter.

The voltage doubler may also be put togood use when 12 -volt power convertertransformers are operated from a 6 -voltpower source. Approximately 90% of thenormal out -put voltage will be obtained be-cause the transistor losses will be higherwhen used on 6 volts.

Filter Normally only a minimum ofSystems filtering is required when com-

pared with 60 -cycle systems (10/dd., or so) because of the high operatingfrequency and fast switching time of thetransistor power converters. If high peakcurrents are drawn from the supply, it isadvisable to install a large capacitor (80/Ad.) across the output as an energy stor-age device.

As a safety factor, the voltage rating ofthe filter capacitors should be at least 25%higher than the no-load output voltage.Higher -than -normal battery voltage coulddamage a capacitor with a marginal voltagerating. For output voltages in excess of400 volts (no-load), two 450 -volt electro-lytic capacitors should be connected inseries. The voltage across these capacitorscan be equalized by shunting them with100K ohm, 2 -watt resistors. When usingthe bridge configuration, another methodof equalization is to connect the secondarycenter -tap to the junction of the two filtercapacitors. One-half voltage exists at thispoint and it will automatically equalize thecapacitors.

In some applications it may be necessaryto include r.f. chokes and disc ceramic ca-pacitors in the B+ circuit. Harmonics,contained in the square wave, can reach the

AN INEXPENSIVEPOWER CONVERTER

TRANSFORMERUSING

FERRITE CORES

(Courtesy of TriadTransformer Corp.)


2

COLLECTORWINDING

Figure 6.2-FACTUAL CONNECTIONS

FOR EARLY

TRIAD TRANSFORMERSThe modifications for com-mon -collector autotransfor-mer connection are de-

scribed in the text.

r.f. circuits and cause beat notes or "bird-ies" on received stations. A 2.5 mh. chokein series with the B+ lead and bypassed toground at each end with 0.001 /it'd. discceramic capacitors will eliminate this pos-sible source of trouble.

Selecting There are many combina-Components tions of components that

may be used in conjunctionwith commercial power -converter trans-formers. The transistors and rectifiersused in this chapter are not necessarilyrecommended. They only demonstratethat these, or similar parts, may be used.You may want to use your favorite type oftransistor, or those in your parts stock.Here are some of the things you shouldknow when substituting transistors or rec-tifiers.

Transistors There are three importantthings to consider when

selecting the switch transistors fora supply.These are power handling ability, break-down voltage ( Vce ), and saturation resis-tance.

The primary power flows through thetransistor junction and it must be capableof handling the current and dissipating theheat generated in the process. As explainedearlier, transistors can be used that wereoriginally intended for class A or class Baudio applications. The transistors usuallyhave the ability to switch eight times theirclass A audio rating, or four times the dis-sipation rating.

You must consider the breakdownrating as well. When one transistor con-

ducts and the other is cut off, a voltageequal to the supply will be induced in thenon -conducting half of the transformerprimary winding. The conducting transis-tor is a virtual short. Therefore, two timesthe supply voltage will appear across thecut-off transistor and the full primary. Thecollector -emitter breakdown rating ( )

must be high enough to withstand thispeak potential. Occasionally the break-down rating listed in the transistor litera-ture is only an average. As a safety factor,it is an excellent idea to allow a larger mar-gin. For example, assume a 12 -volt mobilesupply is to be constructed. As just ex-plained, a minimum potential of 24 voltswill develop across the transistors. Con-sidering that the battery terminal potentialmight reach 14 volts, a transistor with a30 volt collector -emitter rating would bemarginal. It is interesting to note that youcan rectify the full primary voltage to makea 6 -to -12, or 12 -to -24 volt converter. Insuch a unit the secondary would be unused.

In an improperly designed supply, it ispossible to have "spikes" or switchingtransients on the leading edge ofthe squarewave. These transients can create peak po-tentials that are well in excess of twice thesupply voltage. If in doubt, when check-ing a new supply, operate it at reduced in-put voltage and observe thetransistorwave-

4- 441 PILOT LAMPSSERIES -CONNECTED

1 N91

Figure 6.2-GTwo tricks to prov ide for easier startingunder heavy load. Four pilot lampsand a rectifier replace the resistorsshown in Fig. 6.2-D as explained inthetext. These components would be sat-isfactory for a high power supply using

Delco 2N278's.

Radio Handbook Rectifiers 169

Figure 6.2-HTypical rectifier systems used with

transistor power converters.

form with an oscilloscope. Small "knobs"at the beginning of the square wave flattopmay be seen, but they should be less than2.5% of the total amplitude. However,spikes must be eliminated. A 1600 -voltbuffer capacitor across the secondary willremove the transients, but will reduce theswitching time.

The saturation voltage mentioned earliermay be somewhat confusing. If a transis-tor were a perfect switch it would havezeroohms resistance when conducting. Un-fortunately, this is not the case. Even whensaturated, the junction will always have asmall resistance. The larger the resistance,the greater the voltage developed acrossthe junction. Since the junction is in serieswith the primary, the voltage droppedacross the junction cannot appear acrossthe primary and must be considered lost.For example, assume that a transistor has0.1 ohm saturation resistance and the 12 -volt converter in which it is used draws 10amperes. Thus, by Ohm's Law, we findthat one volt will be lost across the transis-tor. Only 11 volts will appear across theprimary, and the secondary voltage will below. If a commercial power -convertertransformer is used and the supply has lowoutput voltage, this is probably the reason.Low output can be minimized by selectingtransistors that have a low saturation resis-tance or by paralleling transistors. A goodtransistor will have a saturation resistanceless than 0.1 ohms. Some manufacturersrate the saturated junction by resistance,while others list the voltage drop at a stand-

ard current of one ampere. In either case,the lower the figure the better the transis-tor.

When paralleling transistors, the emit-ters and the collectors may be connectedtogether. However, the base of each tran-sistor should have a small (4.7 ohm) seriesresistor to equalize the feedback voltage.

Rectifiers Like the transistor, the siliconrectifiers has certain specific

ratings such as breakdown voltageand cur-rent handling ability which must be con-sidered when constructing a power supply.Most silicon rectifiers will handle 500 ma.with ease, and this rating usually is not ofimportance in amateur equipment. Recti-fiers which handle less current may bepur-chased at a slightly lower cost. However,be certain that they are adequate for thejob.Whenever there are two separate currentpaths, such as the full -wave and bridge cir-cuits, the rectifier system will be able tohandle twice the current rating of one recti-fier.

The peak inverse voltage rating must beclosely followed, for the rectifier can easilybe ruined by exceeding this voltage. A

common value rectifier with a PIV of 400volts can withstand between 150 and 200volts of square -wave energy. For example,in a 300v., 200 ma. transformer, end -to -end secondary voltage is 300 volts, peak -to -peak. If this transformer were used inthe half -wave circuit, it would be necessaryto connect two rectifiers in series. In a fullwave circuit it would be necessary to use arectifier on each side of the center -tap. Inthe bridge configuration, the diodes areconnected across the secondary and tworectifiers must be connected in series.Since two current paths exist, it is neces-sary to use a total of four rectifiers. Thesame conditions are present in the voltage -doubler circuit, and a total of four recti-fiers are required.

The rectifier junctions are self -equaliz-ing and any number of silicon rectifiersmay be connected in series without external


equalization. The current rating remainsthe same no matter how many rectifiers areconnected in series. The PIV rating willincrease with each rectifier that is added.

Transient Transient protection is oneProtection other power converter con-

sideration which should bediscussed. An improperly designed powerconverter can provide satisfactory opera-tion for a long period oftime and then sud-denly, for no apparent reason, blow a tran-sistor. Examination will show there is nofront -to -back ratio between the collectorand emitter. This indicates the junctionhas punched through or broken down.

What causes this to happen? As ex-plained earlier, the autotransformer actionresults in a collector -to -emitter potential ofapproximately two times the supply. In 12 -volt mobile applications this amounts toapproximately 24 volts. Even with theslight overshoot normally encountered inswitching, the peak -collector voltage usuallywill not exceed 30 volts. Thus, a transis-

60

40

20

it Ajik 7/. _

28 V CI C

(4 EACH)1N539 2011,1W 4-

250 DC75 MA.

ao-V.

VCB MAX.

40-

20-

0

'rkZENER VOLTAGEOF DIODES

Figure 6.2-ITransients in power converters can bedestructive. A double anode clipper,such as the International Rectifier"Klip-Sel" will short circuit transients

before they reach destructivepotentials.

tor with a BK rating of more than 30volts should be adequate.

Consider, however, what occurs in amobile installation when the starter is ener-gized with a transistor power converterconnected across the battery. The starterwinding represents a large inductive loadand the back e.m.f. can generate transientsacross the battery. Any induction load onthe supply can be a source of transients.Although the average energy contained inthe transient may be low, the peak voltagecan reach many times the supply voltage.The transient, if allowed to reach the tran-sistors, can add to the 2E collector po-tential and destroy the transistor. Even

when high -voltage transistors ( BV of60 volts or more) are used, external tran-sients which occur during the switchingperiod can add to the leading edge over-shoot and break down the junction.

In the circuit of Fig. 6.2-I, note thattwo double -anode breakdown diodes areconnected between the collector and emitterof each transistor. The diodes are designedto conduct whenever the diodeterminal po-tential exceeds 60 volts. As can be seen inFig. 6.2-I, any transient which could de-stroy the transistor is effectively clipped bythe diode. Transient suppressors shouldbe used in any high reliability power con-verter application to minimize the possi-bility of destruction due to transients. Theunits are made by General Electric underthe trade name Thyrector and by Internation-al Rectifier Corporation as Klip-Sets, andare available with a clipping level as low as25 volts. As a minimum precaution, anypower converter should have a large elec-trolytic across the supply terminals to ab-sorb as much transient energy as possible.

A capacitor connected across the secondarywill absorb transient energy, particularlythat produced by transformer leakage re-actance, but it also reduces the switchingspeed. Generally it is best to avoid thi stransient suppression technique, since theslower switching time increases transistordissipation.

Radio Handbook D.C. to A.C. Conversion 171

6.3 D.C. to A.C. Conversion

Direct current power may be convertedto alternating current by the technique de-scribed in section 6.2. The system is simi-lar to the transistor power converter. How-ever, no rectifier -filter system is employedand, of course, the frequency is reduced to60 cycles. In military aircraft installationsthe operation frequency is 400 cycles. Ineither case the conversion device is usuallyreferred to as an inverter.

The circuit for a 120 -watt inverter isshown in Fig. 6.3-A. It operates from asource of 12 volts d.c. and delivers 115volts and 60 cycles at 1 ampere maximum.Note that the common -collector configu-ration is used to eliminate the need for in-sulating the collectors from the chassis.This is compatible with modern automo-biles which use negative ground, 12 -voltignition systems.

Resistor R1 applies starting bias to theswitching transistors, while R2 completesthe divider network. Capacitor C1, acrossR2, minimizes attenuation of the base drivesignal. A 14 ampere fuse protects the pri-mary circuit.

RI'

Ra

Ry

ERN

87.K-GRNM

BLA1

Y EL .C)

BROWN

11511. - 60'b1 A,

Figure 6.3-AA power inverter which develops 117volts a.c., 60 cycles from a 12 volt d.c.supply. Caution must be exercisednot to exceed a one -ampere load cur-rent. For additional protection, con-nect a 14 -ampere fuse in the supply

line.

TRANSISTOR RI =RI, R2=R2, C1zC.1, C3

DELCO 2N278 2000. 211. 2 .UF 111F

05506 10W. 5 W 50 W.V. 200 W.V.

2000. 511. 2 lif 1.1./FBENDIX 2N678 10W. 2 W. 50 W.V.

CLEVITE 2N7746 10011 2 0. 2 /JF 111F20 W. 5 W. 50 W.V.

DELCO 2N773 200.0.20 W.

SD.5 W.

111F50 W.V.

1.1.1F

Figure 6.3-B

A buffer capacitor is connected acrossa portion of the secondary to absorb tran-sients generated during the switching cycle.This capacitor is mandatory when usingthe Triad TY-75-A transformer. The sec-ondary also includes two additional outputconnections to compensate for low batteryvoltage, etc.

The transistors should be mounted onheat sinks with a minimum surface area ofat least 50 sq. in. Caution must be exer-cised to prevent exceeding the 1 amperemaximum secondary current rating. Fur-ther, the transistor heat sinks must be vent-ilated during prolonged operation.

A chart of the recommended transistortypes and the associated component values,is shown in the chart, Fig. 6.3-B.

6.4 A Low Cost PowerConverter

This popular power converter was orig-inally developed by John Specialny of Phil -co Corporation (Application Report

#317). It uses no special switching orstep-up transformers. Rather, two 6.3 voltcenter -tapped filament transformers areused, one for voltage step-up and the otherfor impedance matching of the base drivesignal. This technique is particularly use-ful in power converters delivering 10 to40 watts output.

Transformer T1 , in conjunction withR1 and R2 , provides the feedback currentnecessary to sustain oscillation. The feed-

172 Power Supplies

Tir

.02600 V.

R2,1.2 I<

2N386's A

RECTIFIERS:SARNES-rAPZIAN M-500SILICON DIODES

+12V.- + 250-300 -V. D.C.

Figure 6.4-AA low power converter for receiver orQRP transmitter applications. Theout-put ratings are nearly ideal forpower-ing the Command Set receiver series.

back polarity, from points A and B, mustbe correct or the supply will neither startnor operate properly. As shown, thepower supply oscillates at approximately200 c.p.s. Resistors R3 and R4 providestarting bias for the transistors. Trans-former T2 steps up the primary voltage,which is essentially a square wave, by anamount determined by the turns ratio. Asimple bridge rectifier system providesnearly pure d.c.

The required load current will deter-mine how much feedback voltage is re-quired. The feedback can be increased byreducing the size of R1 and R2 while keep-ing both values equal.

As shown, the supply has the followingcharacteristics. At 60 ma. the output volt-age will be 310 volts (18.6 watts), whilethe primary current will be 2.5 amperesfor an efficiency figure of 60%. At 110 ma.,the output voltage will be 265 volts (29.2watts), and the supply will consume 4 am-peres for an overall efficiency of 67%.

The output transformer will run warmdue to IR losses. The transistors run rela-tively cool indicating an ability to switchgreater current levels.

The transistors should be mounted onheat sinks with a minimum surface area of5 sq. in. Since the transistors are connect-ed in the common -emitter configuration,mica or anodized aluminum washersshould be used to insulate the transistorsfrom the heat exchanges.

If desired, the power converter can beused at lower input voltages. The outputvoltage and available power will be almostproportional to the supply voltage.

The circuit is not at all critical with re-gard to the transistor type used. Anypower transistor with two or more wattsdissipation could switch the 4 ampereswhen the supply is operated at full loadcurrent.

There are a few precautions whichshould be observed when working withpower converters of this type. The leak-age reactance of the transformers intro-duces severe spikes on the leading edge ofthe switching waveform. These transientshave been measured with a variety of trans-formers and found to be almost exactlyfour times the supply potential, peak -to-

peak. Unless transistors with a peak col-lector rating of 50 volts or more are used,the transient will punch through the thinjunction and destroy the transistors. Insuch cases the builder has no alternativebutto use high voltage transistors or to installsome form of transient protection. It isusually easiest to place a 0.02 /dd. buffercapacitor across the primary ( 117 volt side)of the feedback transformer. As pointedout earlier, this does slow the switchingtime and increase transistor heating. How-ever, at the 30 to 40 watt level more thanadequate dissipation is available with thetransistors listed. As a further precaution,the value of R1 and R2 should not be re-duced without checking the base current.Reducing the resistor value increases thebase drive and excessive current can dam-age the transistors.

II

INDEX

AA.C. to D.C. Conversion. 156A.g.c. 62All -band converter 106

Construction of 107Adjustment 109Transistors 110Output coil 110

Alpha 21, 34Alpha cut-off 34, 55Amplification. 21

Current 21

Voltage 22Amplifier 28

Audio 28Class A, 5 -watt 52Class B . 37, 68Class C 68Common base 28Common collector. 29Complementary 39Considerations 36D.C. linear 39Linear 125Modulator, Mini 46Modulator, 10 -watt 48Neutralized, r f 58R.F. Power 120Semi -conductor speech 40, 42Sliding bias 53

S -meter 48Stability, r.f. and i f 58Transistor 34Vacuum tube 29

Atom 9, 11, 12, 15, 16Audio amplifiers 28Audio amplifier stage 36Audio circuits 39

Audio compressor . 43, 44, 45Autodyne converters 111

F.M.. 112Third overtone 113V.H.F. 112

Automatic gain control 62Avalanche diode 19, 157

B

Barrier 16Barrier potential. 16, 17Base current curves 35, 125Base injection 61

Beta 21, 34Bias 29

Emitter 31

Starting 165Voltage dividers 31

Bifilar wound balanced modulator coil .. 132Bond. 13

Broadcast receiver . 86

C

Capacitive coupling 32Cathode follower 29Circuits 55

R.F. 55Sliding bias 54VHF 150

Circuit, vibrator alternator 162Clapp Oscillator 64Class A amplifier, 5 -watt 52Class B amplifier 37, 68Class C amplifier 68Coil data for converters 115Colpitts oscillator . 64Common -collector amplifier 29Common -emitter amplifiers 29Communications receiver. 94, 103

A.g.c. amplifier 98Alignment 102Audio section 99Battery 101

B.F.O. and amplifier 99Coils 101

Construction hints 100Crystal filters 103, 105Front end protection 101

Heath GC -1A . 67IF amplifiers 98IF filter 103Mechanical filters 105Mixer and oscillator. 97

174

NuetralizationProduct and A.M. detectorTechnicalitiesTuning capacitorUse of receiver

Complementary amplifiersCompoundsCompressor, deluxe audioConversion

10298

100101

10239

943

Variable capacitanceZener

Diode balanced modulatorDiode detectorDissipationDividers, bias voltageDoping

E

18,19

19, 2169601831

13

A.C. to D.0 156 Efficiency 37

D.C. to A.0 171 Elements 9

D.C. to D.0 161 Emitter bias 31

Converters. 60 Emitter -follower 29, 32Autodyne 111 Emitter -follower amplifier 49Coil data 115 Emitter injection 61

Low power 172220 Mc 117, 119 F

Power. 171 Feedback, D.0 30Six -meter 113 Feedback, negative 32

Stage. 61 Filter systems 166, 167Transistor all -bond. 106 Filter exciter, Transistor 136, 137Two -meter. 115 Alignment 138

Coupling, capacitive 32 Construction hints 138

Coupling, transformer 56 Operation of 136

Crystal controlled oscillator 66 Flip-flop 66

Crystal set, simplex 73 F.M. autodyne converter 112

Crystal filters 103 Forming 23Crystal lattice 12 Forward bias. 16, 20Current amplification 21 Forward bias resistor 30

Current bias 30 Full -wave rectifier system 166

Current carriersCurrent gain

2834 G

Cut-off, alpha 34, 55 Gain control, Auotmatic 62Gain, current 34

D Germanium diode . 13, 14D.C. amplifierD.C. to A.C. conversion

39171 H

D.C. to D.C. conversion 161 Half -wave rectifier system 166D.C. feedback 30 Hartley oscillator 64Decoupling capacitance 59 Heath Kit Amplifier Model AA -80 .. 50,51, 52Detection 59 Heath GC -1A 67

Detector High level modulator. 67Diode. 60Infinite impedance 60 i

Product 60, 192 I.F. Stage 62Superregenerative 71, 72 I.F. Transformers, Construction of 85

Diode Impedance matching 32, 56Avalanche 19, 157 Impedance matching transformer ..... 85

Germanium 13, 14 Impedance, input..37Sil icon 13 Impedance, output . 37Tunnel 60 Impedance transformation 57

175

Impedances, Transistor 32Impurities . 13, 14, 23, 24Infinite impedance detector . 60, 62Injection 61

Base. 61

Emitter. 61

Input impedance 37Inverse feedback 38

J

Junction 15, 17, 18, 20Junction capacitance 19

ILeakage current 17, 26Linear amplifiers 125

Emitter follower 126

Bias considerations 126Link coupling. 56

Load lines for transistors 34Low level modulator 69

MMatching, impedance 32, 56Matter 9

Mechanical filter. 105

Mobileer 89Converter 90, 91Construction of 91

Microphone preamplifier 41

Mini -amplifier -modulator 46Mini -amplifier #2 47Mixer 60Modulation 127

Low level system 127

High level system 127The modulator 128

Modulating R.F. Circuits 69

Modulator 47Diode balanced 69High-level 69Low-level 69Transistor balanced 7010 -watt kit 50

Molecule 9

Mu ltiv ibrator 65

NNegative feedback 32

Neutralization 57

N -P -N type transistor 29

176

OOne -transistor receiver 76Operating point 35

OscillatorClapp 64Colpitts 64, 67Crystal -controlled 66Hartley 64Overtone 68R.C. Phaseshift 65

R.F. 123

Self-excited 63Tickler coil 63, 67Ultra audion 64

Output impedance 37

Output, Power 37

Overtone oscillator 68

P

Phasing Exciter, Transistor 128

Adjustments . . ............ . 134Amplifier. 135

Audio amplifiers 133

Balanced modulators 134Batteries 135

Coils 134Construction 133

Gain control 135

Operation of 129

P -N -P type transistor 29Potpourri of Circuits 149

Power amplifier 151

Power converter efficiency 163

Power converter, low cost 171

Power output 37

Power pack, Simplex Audio 74, 75Power Supplies. 156Power supply, Heavy duty 159

Preamplifier, microphone 41

Product detector 60, 92Professional communications

receiver 94, 103

R

Receivers 71

Receiver, regenerative ......... 79, 80, 81Recombination 16

Rectifier 17

Selenium 18

Rectifiers for power converters 169

Rectifier systems . 166Rectifier system, half wave 166Rectifier system, full wave . 166Rectifiers, vacuum tube 163

R.C. Phaseshift oscillator 65Reflex circuits 40Regenerative receiver 79, 80, 81Regulated power supply 159, 160R.F. Circuits 55

R.F. circuits, Modulating 69R.F. oscillators 123

Circuit configuration 124Crystal controlled. 123

R.F. Power Amplifiers 120Bias considerations 121

Input circuit 122Neutralization and Unilateralization .122

Third overtone autodyne converter. 113Operation 113Performance. 114

Third overtone oscillator. 151

Three transistor superheterodyne 81

Tickler coil oscillator. 63Tone control 38Transceiver, 40 -meter s.s.b

AlignmentThe AudioTRB-1.

TRG1 .TRO-1H

138, 140143139141

141

141

TRT-2 Linear amplifier board ... 142VFO 141

Transformer coupling 56Transformers, I.F., construction of . 85

Output circuit 122 Transformer output impedance...37Resistor, Forward bias 30 Transient protection 170

Reverse bias 17, 20 Transmitters 120

STransmitter, C.B. or 10 -meter

The circuit146146

Saturation 31, 38 The output network 146

Selenium rectifier 18 Transmitter, Two -meter 147

Self-excited oscillator 63 Construction hints 148

speech amplifier. 40, 42 Testing 149

Silicon diode. 13, 19 Transmitter, Tunnel diode . 153

Simplex Audio Power Pack 74, 75 Transmitter, 10 -meter phone 154

Simplex Crystal Set 73 Transmitting R.F. amplifiers . 68Sliding bias amplifiers 53 TransistorSliding bias circuit 54 Action 19

S -meter amplifier 40 Alloy 24

Solar cell 76 Balanced modulator 70

Square -wave generator 65 Base 19, 22

Squegg 68 Collector 19, 22Stability, R.F. and I.F. amplifiers 58 Drift . 25

Stabilization.. 30 Emitter 19, 22

Temperature 79 Epitaxial . 25

Thermistor 32 Gain 20

Stage, reflex 40 Grown.. 24Standardized symbols, PNP and NPN High frequency 26

transistors 29 Impedances . 32

Starting bias, Setting 165 Life 27

Superheterodyne, 4 -transistor 86, 88 Load lines for 34

Superregenerative detector 71,72 MADT 25

Super Three. 81,83 Mesa 25Noise 27

TNPNPoint contact

20, 2423

Temperature stabilization 79 Power 26Thermistor stabilization 32 Power converter 161, 162

177

1

PNP 19, 24Surface barrier 25

Transistor data, Interpretation of..Transistors for power supply .. .. . .. . 168Transistor vs. Vacuum tube 28Transit time 55TR One 76TR Two 77Tunnel diode 66, 152Tunnel diode transmitter 153Two -meter converter 115Two -transistor receiver 77

UUltra audion oscillator ... ...... 64Unilateralization 58

V

Vacuum tube amplifier 29Vacuum tube rectifiers 163VFO for A.M., C.W., or SSB 143

Alignment 145Construction hints 144Diagram of VFO 143Operation of 144The capacitors 144Voltage stabilization 145

V.H.F. autodyne converter 112Vibrator alternator circuit 162Voltage amplification 22Volume control. 38Voltage doubler 167Voltage regulation 157

WWaveforms 164

ZZener diode 18, 19, 21, 157

178

USEFUL radio books from E. & E.

"HOW -TO -BUILD" DATARadio Handbook

(16TH EDITION)

New amplifier designs New transmitter designs New receivers and transceivers

Gives extensive, simplified theory. Provides the latest design and con-struction data on a wide range of advanced radio amateur equipment,attractively styled. Broadest "How -To -Build" coverage in the field.Completely revised and up to date. Clearly indexed. 805 pages, all text,with hard covers. En* Book #166

$9e0U at your distributor (foreign $10.50)

CONVERT SURPLUSRADIO GEARINTO AMATEUR &C. B. EQUIPMENTA wealth of conversion dataIn 3 volumes shows you how.

Data includes instructions, photos, and diagrams . . . covers the most commonly availablesurplus items. Each conversion shown yields apractical piece of equipment -proved by testing.Items covered are listedbelow:SURPLUS RADIO CONVERSION MANUALS -3 Volumes -$3.00 ea. (foreign, $3.50)VOLUME I -BC -221 Freq. Meter; BC -342 Rcvr.; BC -312Rcvr.; BC -348 Rcvr.; BC412 Radar Oscilloscope; BC -645Xmtr./Rcvr.; BC -946 Rcvr.; SCR -274 (BC -453A Series)Rcvr.; SCR -274 (BC -457A Series) Xmtrs.; SCR -522 (BC -625,624) Xmtr./Rcvr.; TBY Xcvr.; PE -103A Dynamotor; BC-1068A/1161A Rcvr.; Electronics Surplus Index; Cross Index ofA/N Vac. Tubes; Amateur Freq. Allocations; Television andFM Channels. Book #311VOLUME II -BC -454 or ARC -5 Rcvrs.; AN/APS-13 Xmtr./Rcvr.; BC -457 or ARC -5 Xmtrs.; ARC -5 V.H.F. Xmtr./Rcvr.;GO9/TBW Xmtrs.; BC -357 Marker Rcvr.; BC -946B Rcvr. asTuner; BC -375 Xmtr.; Model LM Freq. Meter; TA -12B BendixXmtr.; AN/ART-13 (Collins) Xmtr.; Simplified Coil -WindingCharts; Selenium -Rectifier Power Units; AVT-112A Light Air-craft Xmtr.; AM-26/AIC to a HiFi Ampl.; Surplus BeamRotating Mechs.; ARB Rcvr. Diagram Only. Book #322VOLUME 1II-APN-1; APN-4; ARC -4; ARC -5; ART -13; BC -191, 312, 342, 348, 375, 442, 453, 455, 456-459, 603,624, 696, 1066, 1253; CBY-5200 series; COL -43065; CRC.7; DM -34; DY-2; DY8; FT -241A; LM Power Supply; MBF;MD-7/ARC-5; R-9/APN-4; R-28/ARC-5; RM-52-53; RT19/ARC -4; RT-159; SCR274N, 508, 522, 528, 538; T-15 toT-23/ARC-5; URC-4; WE -701-A. Schematics only: APA10;APT -2; APT5; ARR-2; ASB-5; BC -659; BC1335A; CPR-46ACJ. Book #333

THE SURPLUS HANDBOOK (Receivers and Transceivers) -$3.00 ea. (foreign $3.50)VOLUME I -Schematic Diagrams and large photographsonly-APN-1; APS-13; ARB; ARC -4; ARC -5 (L.F.); ARC5(V.H.F.); ARN-5; ARR-2; ASB-7; BC -222, 312, 314, 342,344, 348, 603, 611, 624, 652, 654, 659, 669, 683,728, 745, 764, 779, 794, 923, 1000, 1004, 1066, 1206,1306, 1335; BC -AR -231; CRC -7; DAK-3; GF-11; Mark II;MN -26; RAK-5; RAL-5; RAX-I; SCR -522 Super Pro; TBY;TCS; Resistor and Capacitor Color Codes; Cross Index of A/NV.T. and Commercial Tubes. Book #510

RADIOTELEPHONELICENSE MANUAL

$5.15 Book#030

(foreign $6.25)Helps you prepare for all U.S.A. commercialradiotelephone operator's license exams. Pro-vides complete study -guide questions and answersin a single volume. Helps you understand everysubject needed to obtain an operator's license.

Every Electronic Tube in the World Listed - 3 Volumes$8.00- ea. (foreign $8.50)

WORLD'S RADIO TUBES(Brans' Radio Tubes Vade Mecum). World's most completeand authoritative book of v.t. characteristics.

Book #471

WORLD'S EQUIVALENT TUBES(Brans' Equivalent Tubes Vade Mecum) Over 43,900 com-parisons and replacements. Book #. 493

WORLD'S TELEVISION TUBES(Brans' Television Tubes Vade Mecum). Characteristics ofall TV picture and cathode ray tubes ... also special purposeelectronic tubes. Book -,';-.1482

'Order from your favorite electronic parts distributor.

If he cannot supply, send us his name and yourremittance, and we will supply.

EDITORS and ENGINEERS, Ltd.Summerland, California 93067

Dealers: Electronic distributors, order from us.Bookstores, libraries, newsdealers order from Baker &

j Taylor, Hillside. N. J. Export (exc. Canada), orderfrom H. M. Snyder Co., 440 Park Ave. So., N.Y. 16.

radio handbook - americanradiohistory.com€¦ · the transistor radio handbook 1st edition table...

Documents