Home ›

File info: application/pdf · 291 pages · 9.44MB

THE RADIO AMATEUR'S HANDBOOK

License Manual The A.R.R.L. Antenna Book Hints and Kinks How to Become A Radio Amateur Lightning Calculators e America's oldest radio magazine and the ac-cepted journal of the amateur, QST reports each month in dire…

Full PDF Document

If the inline viewer fails, it will open the original document in compatibility mode automatically. You can also open the file directly.

Extracted Text

SPECIAL DEFENSE EDITION
THE RADIO AMATEUR'S HANDBOOK
A Manual for Radio Training Courses
PUBLISHED BY THE
AMERICAN RADIO RELAY LEAGUE $1.00

�
-

SP ECI�-I,` "DEFENSE EDITION

4kie

NINETEEN-FORTY-WO

THE RADIO

IAMATEUR'S

HA:\ DBOOK

BY THE HEADQUARTERS STAFF OF THE AMERICAN RADIO RELAY LEAGUE
PUBLISHED BY THE AMERICAN RADIO RELAY LEAGUE, INCORPORATED, WEST HARTFORD, CONNECTICUT

Other Pf`ticalioni BY THE AMERICAN RADIO RELAY LEAGUE

(DST
The Radio Amateur's Handbook
The Radio Amateur's License Manual
The A.R.R.L. Antenna Book Hints and Kinks
How to Become A Radio Amateur
Lightning Calculators

eAmerica's oldest radio magazine and the ac-
cepted journal of the amateur, QST reports each month in direct and analytical style the rapid developments in amateur radio. Its subscription price is $2.50 per year in U. S. and Possessions.
eYou are reading the Special Defense Edition.
The standard edition, in addition to telling the things needed for the comprehensive understanding of amateur radio, contains chapters on the design, construction and operation of many pieces of radio equipment and on operating procedure. Paper bound, $1.00; $1.25 outside continental U. S. A. Buckram bound, $2.50. Also available in Spanish, paper bound, $1.50.
eThis 40-page booklet gives the necessary in-
formation, including typical or paraphrased examination questions and answers, to enable one to pass the government examination for amateur operator's license. Price 25c.
eA comprehensive manual on the design and
construction of many types of antennas. 144 pages in this format, profusely illustrated. Price 50c.
eA compilation of hundreds of good radio ideas,
savers of time and money for the experimenter. 128 pages in this format. Price 50c.
11 This publication is recognized as the standard elementary guide for the prospective radio amateur. 48 pages, just revised. Price 25c.
C. Circular slide rules, on 8X x 11-in, bases, of special cardboard, with satisfactory accuracy. Two types:
Type A solves problems in frequency, wavelength, inductance and capacity. Price $1.
Type Bmakes computations involving voltage, current and resistance. Price $1.

THE AMERICAN RADIO RELAY LEAGUE, INC.
WEST HARTFORD, CONNECTICUT

^Or
44, e7
....J-orecuord
SINCE 1926 the American Radio Relay League has published The
Radio Amateur's Handbook, primarily as amanual for the tens of thousands of practitioners of amateur-radio communication all over the world. The League is a nonprofit society of such radio specialists and one of its functions has been to make reliable literature on the art available at modest cost. With most amateur stations closed down because of the war, and with the whole nation engaged in the great struggle, the Handbook now has anew mission to perform. With many hundreds of radio training classes in formation in the universities, colleges, high schools and community centers of the country, this special revision of the current edition has been prepared in the belief that it can serve as amore valuable but inexpensive text for defense training courses.
Devoted to afast-moving and progressive science, the Handbook has required sweeping and virtually continuous modification throughout its life. Its annual rewriting is amajor task in the family life of the headquarters group of the League at West Hartford, where most of the technically-skilled specialists of the staff participate in the work under the general technical editorship of George Grammer, technical editor of QST. Some measure of the book's acceptance is to be found in the fact that there have been nineteen previous editions, with thirty-four pressings and total sales of well over three-quarters of amillion copies.
In this special edition for training courses we have retained everything from the current standard edition that seems useful to the task in hand but have eliminated those portions that would serve only to distract or encumber the student. We have omitted everything that treated particularly of this League, the advertising that usually accompanies the Handbook, and the chapters that ordinarily deal with the home construction of specific pieces of apparatus for amateur stations. In place of the latter we have prepared aprofusely-illustrated new chapter calculated to acquaint the student with the appearance and the circuit diagrams of representative types of apparatus. A valuable new chapter gives fundamental instruction in the solution of formulas and the reading of graphs, something that has long been needed even in the Handbook. And because many defense courses will also incorporate instruction in the international telegraph code, the newly-added material contains achapter on learning the code and some suggestions for aclassroom code table. Our aim throughout has been to write an understandable text for busy, practical people of average education, employing aminimum of mathematics. A major objective has been to provide the answers to the questions that naturally arise in the course of study. The material has been so arranged as to make it readily possible to find what is wanted, amultitude of headings identifying subjects at aglance. Information has been presented concisely but with copious cross-references to permit the background always to accompany the subject under consideration. We have endeavored to employ cross-references in such quantity that no treatment of any subject can be considered "too technical," since the references will eventually lead the reader, if he needs it, to the applying fundamentals themselves. The sequence of presentation has been planned to lend itself to an ordered course of instruction. Necessarily compact (as is any good text), information is deliberately presented without sugarcoating, but every effort has been made to make it understandable and to avoid saying things in such away that they are intelligible only to those who already know the subject thoroughly!

. 4.0. 1r,. 1 4 A word about the reference system: It will enoted that each chapter is

divided into serially-numbered sections. The number takes the form of two digits

or groups connected by ahyphen. The first figure is the chapter number, the sec-

ond the aform

assec(t�i4o-n7)n,ufmobr eerxawmiptlhei,n

the chapter. Cross which means that

references in the text the subject referred

take such to will be

found discussed in Chapter Four, Section 7. Illustrations are numbered serially

in each chapter. Thus, Fig. 502 can be readily located as the second illustration

in Chapter Five. There is acarefully-prepared index at the end of the reading

pages.

We here shall be very happy if this special edition of the Handbook can be of

as much assistance in the national effort as its predecessors have been to licensed

amateurs.

K ENNETH B.W ARNER

W EST H ARTFORD, CONN.

Managing Secretary, A.R.R.L.

February, 1942

SPECIAL DEFENSE EDITION

First Printing, February, 1942

40,000 Copies

Second Printing, April, 1942

40,000 Copies

Third Printing, September, 1942

40,000 Copies

Fourth Printing, December, 1942

50,000 Copies

PRINTED IN U. S. A. BY
THE RUMFORD PRESS
CONCORD, NEW HAMPSHIRE

Conlenb

Frontispiece:

Schematic Symbols

Chapter 1 Formulas and Graphs Chapter 2 Electrical and Radio Fundamentals Chapter 3 Vacuum Tubes Chapter 4 Radio-Frequency Power Generation Chapter 5 Radiotelephony Chapter 6 Keying Chapter 7 Receiver Principles and Design Chapter 8 Power Supplies Chapter 9 Wave Propagation Chapter 10 Antenna Systems Chapter 11 Radio Equipment Chapter 12 Measurements and Measuring Equipment Chapter 13 Workshop Practice Chapter 14 Learning the Code Chapter 15 Miscellaneous Data INDEX

7 21 42 58 84 106 111 143 156 162 196 225 249 254 268 281

SCHEMATIC SYMBOLS USED IN CIRCUIT DIAGRAMS

ibilenne.

Wires Connected

Milliammeter

Ground
7 Fixed 'Condenser

Wires crossing but not connected

D00000(

Twisted Pair

Voltmeter A.C. Plug A.C.Receptade

zirz Condenser (movable plates)

Inductance (Fixed coil or R F choke)

-6b-

Shielded Wire
Shielding
Ileao;ohenes

Loudspeaker

--,1--
--

Fuse Rectifier
Lamp Bulb Neon Bulb or Voltage
Regulator Tube

Plate

Tapped Inductance

Single -Buttorz, Carbon. Microphone
Double-Button, Carbon Microphon-e

Diode Vacuum, Tithe
Filament

Plate

1iEr Air-Core Transformer (or two coils coupled to each. other)

Ribbon or Velocity

IIo

Microphone

Grid

Triode Vacuum Tube

Crystal

Filament

Air-Core Transformer (to tth, variable
coupling)

Microphone

--ce511-

key

Plate G,

Multi.-6n.d

Vacuum

nibe.The grids are

62 usually/wintered,G,

beau; that closest

to the cathode

[LIinndku-cCtoaunpcleesd

Jac-is 11

VACUUM -TUBE SYMBOLS: Filament or

Phone Pill

Heater

Powdered /ron Core Transformer
EIron-Core Trans former

-""D'se�

Switch, Single-Pole Doutte-Tnrow Switch

Double-Pole e-- Single-Throw Switch

Cathode Cold Cathode Grid Plate Diode Plate

poop_ /ron -Core ' Inductance or Choke

*-- Rotary Tap or BandSwitch

11>-

--vvvvvv,-- I?eFsitxsetdor

Variable and

11 1111111111-- Battery

rapped Resistors (rheostat, potentiometer

Single Cell

voltage divider etc.)

Ammeter

-cra

Quartz Crystal

Beanz- Forming Plates
Anodes Electron-Ray Tube Target Anodes Cathode-Ray Tube Deflecting Plates
Indicates Gaseous Tube

CHAPTER ONE

.7ormulcti an] Graphi

IN ORDINARY radio work m athematical applications reduce to the solution of formulas, or equations, and to reading (and sometimes plotting) graphs. Only elementary mathematical knowledge is required, but that knowledge should include a few fundamentals in a number of branches of the subject. This chapter is intended to cover those fundamentals in sufficient detail to give the reader an understanding of principles, as well as to demonstrate the methods of arriving at solutions to the formulas he will find in later chapters. Familiarity with elementary arithmetic is assumed, but areview of decimals and the method of extracting square root is included for the benefit of those whose daily mathematical experience is limited to the very simple arithmetical operations which most people find sufficient for non-technical pursuits.
�I-1 DECIMALS
Notation -- Our system of writing numbers is based on multiples of ten, with the position of the digit indicating the multiple to be used. Fundamentally we have only the digits 1, 2, 3, 4, 5, 6, 7, 8, 9 and 0; all our numbers are based on the positions of these digits in relation to others in the number. Any number is actually asum. Thus the number 4826 means the sum of
6units 2 tens 8 hundreds (10 X 10) 4 thousands (10 X 10 X 10)
which we write 4826
Each time we move one place to the left the digit in that place is multiplied by an additional factor of 10.
Since this system lends itself so readily to calculations, it is very convenient to extend it to numbers less than one -- in other words, to include fractions. Looking at the number 4826 from aslightly different viewpoint, it is evident that we divide by ten each time we move one place to the right. There is no reason why this process cannot be continued indefinitely. We need some method of distinguishing the units column, and this is accomplished by placing a period or decimal point to its right, thus marking off the whole numbers from the fractions. Continuing the division, the multiple for the first place to the right of the decimal point

(first decimal place) is 1divided by 10, or 1/10,

the second place is 1/10 divided by 10, or

1/100, the third place is 1/100 divided by 10,

or 1/1000, and so on. By converting fractions

to make their denominators 10 or multiples of

10 we can fit them neatly into the system.

Forming decimals -- Fractions can be con-

verted by multiplying both the numerator

and denominator by a number which causes

the resulting denominator to be a multiple of

ten. For example, the fraction

can be

converted as follows:

2-1 -55 --150

which is then written 0.5. The zero is used to show that there are no whole numbers and to add emphasis to the fact that the decimal point
is present -- the decimal point is small and sometimes is overlooked. A fraction with a
larger denominator would be converted to the next larger multiple of ten:

--225 X -44 = --1800 = 0.08

and similarly for still larger denominators. The zero is placed between the decimal point and the 8 to bring the latter into the "hundredths" column.
Multiplying the numerator and denominator of a fraction by the same number does not change the ratio of numerator to denominator and hence does not change the value of the fraction, so we can for instance, do the following:

--150 X --1100 = --15000 = 0.50

5 or --10 X

100 --100

500 --1000

0.500

That is, zeros may be added to or subtracted from columns on the right-hand side of a decimal fraction without changing the value of the decimal. We can just as readily put zeros to the right of the decimal point in anumber which has no fraction; 237 and 237.0000 are exactly the same number.
The multiplier by which a fraction can be converted to adecimal can be found by dividing the denominator into a multiple of 10. Thus in the example

2 --25

4 4 -

--1800

0.08

CHAPTER ONE

7h ePalio AmateuA -flani`ooh

the multiplier is 100 � 25, or 4. The 4is then immediately multiplied by 2 to give 8, and then the denominator (100) is dropped by writing the result 0.08. This process can be shortened to the very simple one of dividing the numerator of the fraction by its denominator. To do this it is first necessary to add decimal places after the 2so that we can perform the division by familiar methods. Thus we have 2.00 divided by 25, which for purposes of division can be considered the same as dividing 200 by 25. By short division,
.08
25) 2.00
Note that the decimal point in the quotient is placed directly above the decimal point in the dividend, and that the place intervening before the first digit is filled by azero.
Mixed numbers, such as 643, can be changed to decimals either by making an improper fraction (which in this case would be 2--47)and
then dividing as above, or by changing the fraction alone into adecimal and simply placing the integer (an integer is awhole number) to the left of the decimal point. Thus 3 % = 0.75; adding the 6gives 6.75.
Rules for placing the decimal point -- In addition and subtraction of numbers containing decimals the numbers are set down in a column with the decimal points aligned vertically. Thus to add 24.85, 218.003, and 4.6, the procedure would be
24.8 5 218.0 03
4.6
247.4 53
the addition being carried out as though the decimal point were not there.
Subtraction is handled similarly; for example, suppose 31.028 is to be subracted from 286.2:
286.2 00 -- 31.0 28
255.1 72
In this case addition of the zeros in the first number is helpful, to avoid subtraction from "unfilled" places. In both addition and subtraction the decimal point in the answer is placed directly below the others.
When two decimal numbers are multiplied, the number of decimal places in the result is equal to the sum of the decimal places in the original numbers. Thus,
23.6 5 X 4.8
18920 9460
113.5 20

The decimal plac'es in the result are counted off from the right, including any terminating zeros that may arise as aresult of the multiplication. It should not be hard to see the reason for this rule, remembering that the decimal portions of the numbers are really fractions. Without decimals, the fractional parts of the numbers in the example would be 65/100 and 8/10, respectively, and the normal process of multiplication would be to convert both mixed numbers to improper fractions, one having a denominator of 100 and the other 10. On multiplication the resulting denominator would be 1000, or three decimal places.
Division reverses the rule for multiplication. The number of decimal places in the quotient is equal to the number in the dividend minus the number of places in the divisor. Thus in dividing 163.122 by 52.62 we have
52.6 2)163.1 22(3.1 15786
5262 5262
Another example, 22.5 divided by 0.15:
0.1 5) 22.5 0 (1 50 15
75 75
o
illustrates the case where the number of decimal places in the divisor is greater than that in the dividend. There is asimple rule covering such cases: When the number of decimal places in the divisor is greater than that in the dividend, add zeros to the right of the dividend until both have the same number of places, when the quotient will come out as a whole number.
In most cases division can be simplified by using the following form:
3.1
52.6 2) 163.1 22 15786
5262 5262
Starting from the decimal point in the dividend, count off to the right the number of decimal places in the divisor, place the decimal point for the quotient directly above and after this place, then proceed as in ordinary division. If there are no decimal places in the divisor, the decimal point in the quotient goes directly above that in the dividend. This form has the advantage that the division can be continued conveniently in case the quotient is not "even." For example, to divide 23.84 by 13.3:

8 CHAPTER ONE

5ormulaJ and eraphJ

1.7 92 + 13.3) 23.8 400
133 1054
9 3 1 1230 1197
330 2 6 6
64

Zeros can be added to the dividend until the answer is carried out to any desired number of decimal places.

�1-2 POWERS ANI) ROOTS

Powers --The power of a number is the number of times it is used as afactor, or multiplier. Thus the number 3can be considered to be the product of 3 X 1and, having been used once as afactor, 3is said to be the first power of itself. If we multiply 3by 3we have used 3 twice as afactor, hence the product, 9, is called the second power of 3. Multiplying by 3again gives 27, which is the third power; once more gives 81, the fourth power; and so on. A number is said to be raised to acertain power when it is used that number of times as a factor or multiplier.
When anumber is to be raised to acertain power it is indicated by writing the number of the power as asuperscript to the right of the number itself; thus 3raised to the 4th power is written 34.The number which denotes the power is called an exponent. The second power is frequently called the "square," because if the number represents the length of a line, multiplying by itself is equivalent to finding the area of asquare having sides of that length. Similarly, the third power is called the "cube," since the multiplication would give the volume of a cube having edges equal in length to the number.
Roots -- The converse of a power is called the root of anumber. It is that number which when raised to the required power would be the number given. For example, 3is called the
third (or cube) root of 27, because 3raised to the third power is 27. Likewise, 4is the second or square root of 16, because 42 (4 raised to the second power) is 16. The fourth root of 16 is 2, because 2raised to the fourth power equals 16, and so on. When a root is to be found the

radical sign

is used, with the order or

number of the root (called the index) written

in the opening of the "V." For example, the

4th root of 16 is written

When the second or square root is to be taken,

the index usually is omitted; the order of the root in this case is understood to be 2.
Square root -- The numerical operation of extracting square root is frequently necessary in solving problems. The method can best be described by means of an example. Suppose the square root of 528.367 is to be found. Set down the number and point off groups of two places, starting with the decimal point and going both to the right and left:
-V5'2 8.3 6'7 0
Note that azero has to be added to the decimal to make two figures in the last group; this is necessary on the right-hand side but not on the left. Now find the largest integer whose square is contained in the first group, which in this case is the single digit 5. The largest number whose square is contained in 5 is 2, the square of which is 4(the square of 3would be 9, which is greater than 5). Write this root above the 5 and write the square below the 5, then subtract:
2 -V5'2 8.3 6'7 0
4
Now bring down the next group of numbers alongside the remainder, 1, and draw avertical line to the left of the 1:
2 -V5'2 8.3 6'7 0
4 I128
Double the root (2) already found and place it to the left:
2 -V5'2 8.3 6'7 0
4 411 28
Cover the last number (8) and see how many times the divisor, 4, will go into the remaining number, 12. The answer is obviously 3. Place the three in the root above the 28 group and also write it alongside the divisor, 4. Then multiply the complete divisor, 43, by 3 and write the result under the 128:
23 V5'2 8.3 61 0
4 43ITF8-
129
At this point it is necessary to subtract again, but unfortunately the subtrahend is just abit too large. If 3 gives a product which is too large, obviously the next step is to try 2, so we

CHAPTER ONE

Radio ...AmateuA -llanidook

substitute 2 for 3 and try again, getting a smaller number which can be subtracted:
22
V5'2 8.3 6'7 0 4
42
84 44
Bring down the next group, 36, and repeat the procedure -- double the root already found (22) and find the largest multiplier which can be used:
22 9
V5'2 8.3 6'7 0 4
42 rfr� 84
449 14 436 4041
395
Continue until all places are filled above the two-digit groups in the original number:
22.9 8
V5'2 8.3 6'7 0 4
42 84
449�Ffr3 4041
4588fttT3T5
3 6 7 0 4 2866
Notice that the decimal point in the root is placed directly above the decimal point in the original number. If desired, the solution can be carried out to additional decimal places by adding groups of two zeros to the right of the existing decimal, then proceeding just as before.
Should any of the divisors be too large to go into the remainder, azero is placed in the answer at that point, just as in the similar case in ordinary division. This is illustrated in the following example:
5
1/2 5'7 0.4 9 25
101-- 70-
On covering up the 0in 70 it is found that 10 is too large to be contained in 7, hence azero is put in the second place in the answer and the process continued as before:
5 0
1/2 5'7 0.4 9 25
1001 7049

an additional group (49) being brought down for the next division.
�1-3 FOIRMULAS
Literal notation -- The laws of electricity and electrical circuits are most conveniently expressed by formulas in which the various quantities are denoted by arbitrary symbols, generally letters of the alphabet. This is called literal notation, and is simply a compact method of writing. To cite an example, Ohm's Law for direct currents, afundamental law in electrical work, can be expressed as follows: "The current flowing in acircuit is equal to the impressed electromotive force divided by the resistance, using consistent units." In algebraic formula this is written
I E
in which the letter /stands for current, E for electromotive force and R for resistance. Expressing the law in this way has the additional advantage that anyone familiar with the elementary processes of algebra can see at a glance the further relations:
E = RI and R = E-I
either of which would have to be learned by rote or else reasoned out from the ordinarylanguage statement.
"Solving a formula" merely means substituting the proper numbers for the literal quantities whose values are known, and then performing the indicated arithmetical operations to find the unknown quantity.
Signs of operation-- The ordinary signs of arithmetic ( -, X, �, V -,etc.) are used. However, division is nearly always indicated by writing the dividend over the divisor in the
form of a fraction as -ER in Ohm's Law given
above; this means E divided by R. An alternative method of writing the same thing is EIR, where the diagonal bar has the same meaning as the horizontal fraction bar. Multiplication can be indicated by the usual sign (X) and also by adot (.), but the sign is frequently left out altogether, the two quantities to be multiplied simply being written side by side. Thus, RI means "R multiplied by I." However, the multiplication sign must be used when ordinary numbers appear together in a formula; for instance, 31 multiplied by 42 is written 31 X 42; the abbreviated method cannot be used in this case because neither 3142 nor 31.42 is the same thing as 31 X 42. But when an actual number and a literal number are multiplied together they can be written with-

10 CHAPTER ONE

..7ormulas and grapni

out the sign: 15E, for example, means 15 times E.
In an expression such as 15E the number 15 is called a coefficient of E. A coefficient is a factor, either literal or numerical, whose value is fixed. The symbol k is frequently used in electrical formulas where a coefficient, or "constant," is specified. When numerical coefficients are multiplied the ordinary numerical operation is usually performed; thus if we multiply 4E by 3, the product is written 12E, instead of 4 X 3E. The latter or algebraic form is used only when multiplication is by aliteral quantity; thus, I X 3E would be written 31E. It is usual to assemble the literal factors together and to put the numerical factor first.
Brackets--Bracketed symbols indicate a group that is to be treated as a unit. Thus if we write (a b) we mean that the bracketed quantity must be treated as though it were all one symbol. This is necessary to distinguish between expressions of the type ab - c and a(b - c). The former means "Multiply aby b and then subtract cfrom the product," while the latter means "Subtract cfrom band then multiply the difference by a." The two are not the same thing, as substitution of any numbers you may choose for a, band cwill readily show. Similarly, (a - b) 2 means "Multiply (a - b) by (a - b)" as distinguished from a - b2, which means "Multiply b by itself and subtract the product from a."
Fundamental rules-- In general, a series of additions can be performed in any order, since the same sum results no matter which number is put first or last. For example, 11 � 6 � 8gives the same total as 6 -I- 8� 11 or 8 � 11 � 6, etc. The same thing is true of multiplication; 5 X3 X 2 =2 X5X3 3 X 5 X 2, etc. However, in subtraction and division this is not the case in ordinary arithmetic: 12 - 9is not the same thing as 9 - 12, nor is 42 � 7 the same as 7 � 42. By considering subtraction as algebraic addition of a negative number this restriction can be overcome, likewise with division by considering every divisor as a fraction of the form 1/n, where n is the divisor, and then using it as a factor. Thus, 12 - 9 = - 9 � 12, the plus and minus signs being considered to be properties of the numbers. A number having no sign prefixed is always understood to be posi-
tive. Similarly, 42 � 7 = 42 X 7 -'which is the
same as -1 X 42. Therefore, by reducing sub-
traction to algebraic addition, and reducing division to multiplication by afraction, we can say that, in general, a sequence of the same operations can be performed in any convenient order.

When algebraic addition and multiplication

are both indicated, as they frequently are in a

single formula, the order of operations is very

important. Thus if we have ab - c, in which

a, b, and chave the values 4, 8and 3, respec-

tively, substitution would give 4 X 8 - 3.

This gives 29 as an answer, which is not the

same as if we first subtracted the 3from the 8

and then multiplied the difference by 4. The

general rule is that the operations indicated in

an equation must be performed in sequence

from left to right. It is important to note that

in aliteral expression factors may not be sepa-

rated by addition or subtraction. Thus in the

expression a bc, when numerical values are

substituted the multiplication must be per-

formed first. Using the previous values, a bc

does not mean 4 � 8 X 3, which is 36, but

4 � (8 X 3), which is 28. In other words, a

series of factors is equivalent to a bracketed

quantity, when numerical values are sub-

stituted for the letter symbols.

Powers and roots -- In some formulas it is

necessary to raise aseries of factors to acertain

power or to take aspecified root of agroup of

factors. If we want to square the expression

abc we have (abc) 2.This means that the whole

expression is multiplied by itself; that is,

abc X abc, or, dropping the multiplication

sign, abcabc. Since the order in which the

multiplication is performed does not matter,

this can be rewritten aabbcc. But aa = a2,

bb = b2,and cc = e2;therefore (abc) 2 = a2b2c2.

In other words, the square of aseries of factors

is equal to the product of the squares of each

factor, and similarly for higher powers. Identi-

cal reasoning will show that the same rule ap-

plies for roots: Vabc = Vit X

X V-c, or

Va-VT) Ve.

In expressions calling for raising the alge-

braic sum of two or more numbers to agiven

power, or for extracting agiven root, no such

simple rule applies. For instance, (4 � 3) 24 is

equal to 72,which is 49. It is not equal to

42 � 32,which is the sum of 16 and 9, or 25.

(Expressions of this kind should be multiplied

out when they are encountered.) Usually, in

ordinary radio formulas the actual values will

be substituted and it will be possible to take

the sum before finding the power or root. This

simplifies the computations.

Bracketed quantities-- One other type of

operation needs to be learned. When asum is

multiplied by a number, as in the expression

a(b + -1 - d) each part (called aterm) of the

sum must be multiplied by the factor. In the example this gives

a(b + -1 - d) ab + -a - ad

Notice that the signs of the terms in the right-

CHAPTER ONE

DheRadio AmateuA -licirtihoo4

hand part of the equation are the same as in the original bracketed quantity (both a and b are positive, since their signs are not written). This is always the case when the multiplier is positive. If the multiplier has a negative sign, as in the expression a -- b(c + d -- e), all the terms inside the brackets must be reversed when the multiplication is performed and the brackets eliminated. Thus the resulting expression would be a -- be -- bd + be.
If the bracketed quantity is to be added to another quantity, as a -I- (b -- c� d), removing the brackets requires no changes in the signs, so
a (b--c+d)=a+b--c+d
But if the bracketed quantity is preceded by aminus sign, all signs inside the brackets must be reversed when the brackets are removed:
a -- (b -- c+ d) =a--b+c--d
These rules are easily illustrated by means of numerical examples. In each case the distributed form is shown first, followed by the method of first collecting terms (performing the additions) inside the brackets. Multiplication by apositive number:
5 X (8 -- 4 -I- 1) = 40 -- 20 + 5 = 25 5 X (5) = 25
Multiplication by anegative number:
12 -- 3(4 -- 2) = 12 -- 12 + 6 = 6 12 -- 3(2) = 12 -- 6 = 6
Addition of abracketed quantity:
10 -F (8 -F 2 -- 3) = 10 + 8 + 2 -- 3 = 17 10 + 7 = 17
Subtraction of abracketed quantity:
10 -- (8 -F 2 -- 3) = 10 -- 8 -- 2 -I- 3 = 3 10 -- 7 = 3
Transposition -- In general, a satisfactory solution of a formula cannot be obtained unless the numerical values of all but one of the literal quantities are known. Taking Ohm's Law again as an example, the relation
I = --ER
will give the current, I, when known values of electromotive force (voltage) and resistance can be substituted for E and R in the formula. Using ordinary units, if E should be 80 volts and R 40 ohms, the current would be 80 � 40, or 2amperes. However, should only the voltage be known, we could learn nothing about the current that might flow, since every value of resistance would give a different value of current. Therefore, it must be possible to substitute known values for every quantity in the

formula or equation except one, which is called the "unknown." Usually, any one of the literal quantities may be the unknown, regardless of its position in the equation, and the equation can be transposed so that the unknown quantity appears on one side (usually to the left of the equality sign) and all the known quantities on the other side. It is convenient to arrange an equation in this way before actually substituting the numerical values for the known quantities.
Transposition is based on the principle that performing the same operation on each side of an equation does not alter the equality. Thus if we add the same number to both sides of an equation the resulting sums also are equal. A numerical example will quickly illustrate the point. Suppose we have the following equation:
5 X 3 = 11 -I- 4
and add 6to each side:
(5 X 3) -I- 6 = 11 + 4 6 15 -F 6 = 21
The same quantity subtracted from each side of an equation also leaves both remainders equal; similarly, multiplying both sides by the same number or dividing both sides by the same number does not change the fact of equality. Likewise, raising both sides to the same power Of taking the same root on both sides leaves the resulting quantities equal. Whatever the operation, it must be performed similarly on both sides and on the whole of each side. For example, if we wish to square both sides of the equation a = b c, the right-hand side must be considered as aunit so that the resulting equation is a2 = (b e) 2.As we have seen before, this is not the same thing as b e2 or even b2 e2.Careless errors can be avoided by keeping this point in mind.
An equation such as I = E/R can be transposed very readily. Suppose we know the current and resistance and want to find E. The object is to get E alone on one side of the equation and get all the other quantities on the other side. To get E by itself we must eliminate R on that side of the equation. Since E is divided by R, we can eliminate the R in the denominator if we multiply EIR by R, since a number divided by itself is 1. The common expression is that the R's "cancel out." However, to make the resulting expression an equality we must multiply both sides of the equation by R, so we have
I X R = --E X R
The R's on the right-hand side cancel, leaving us with IR = E, or, since we usually put the quantity we want to find on the left-hand side,

CHAPTER ONE

..7ormufai and graphs

E = IR.From this we can readily find R in terms of the other two quantities by dividing both sides of the equation by Ior (which is the same thing) by multiplying both sides by 1//:

E X -/ = IR X -/

In this case the /'s on the right-hand side can-

cel, leaving us

-EI

In more complicated equations it may be necessary to perform several operations of this nature in sequence before the desired arrangement of the equation is secured. For example, the equation for the resonant frequency of a tuned circuit is

f -

22-

1 VLC

where fis the frequency, L is the inductance and C the capacity in the circuit. (ir, the ratio of the circumference to the diameter of acircle, has its usual value of 3.1416 ...) Suppose we want to find C in terms of the other quantities. First we need to bring C out of the denominator, since we do not want the result in
fractional form. To do this, multiply both sides by 21-VLC. This gives

X 22--V/ - 1-- X2TVLC 22-'LC
or 22-f-VM = 1
The order in which the factors 2, ir and fare written together does not really matter, but that shown is customary. We next need to get rid of the radical, which we can do by squaring both sides of the equation:

(221-VrC) 2 = 12 or (22-f) 2 LC = 1

since 12 is still 1. The last step is to get C alone on one side of the equation, which is accom-
plished by dividing both sides by all the remaining factors in the left-hand side of the above expression:

(22-f) 2LC

(27r-f) 2L (2111) 2 L

when the leaving

(22-f) 2L's on the left 1
C - (221) 2 L

cancel

out,

It is useful to know one more thing in connection with transposition. If we have an equation of the form

and multiply both sides by a1c1-, we have

abd = bd,c
abc adc
The a's and b's on the left cancel out, as do the d's and c's on the right, leaving
-d c -=b -aob rad c-
That is, if both sides of an equation are inverted, the resulting expression is an equality. Any equation can be considered to be in this form, even though neither side is expressly fractional, because any whole number can be considered to be a fraction of the form n/1 (dividing a number by 1does not change its value). Such anumber inverted becomes 1/n.
Inversion can be used to advantage in transposing aformula such as the following

R -1 1

1 1

RI R2 /?3

which is the rule for finding the net resistance of resistances connected in parallel. The series

of dots indicates that similar terms (1-�where

n indicates the last number of a consecutive series) may be added until there is one term for each resistor in the actual group considered in the problem. Suppose we have two resistors
in parallel and want to add athird to make the total resistance, R,have aspecified value. We need then to transpose the equation to give R3,the third resistor, in terms of the first two, RI and /?2, and the total resistance, R. By the rule for inversion,

1 111 RI -R2 R3 = -R

1 1 Subtracting -R I � -R 2 from both sides we have

remaining

1 1 (1

1\ 1 1 1

Inverting again,

R3 -

11 1

I? 1?1 R2

Observe that the whole expression on the right hand side must be inverted, not simply the individual fractions.
Hints for solving formulas- In practical use of formulas the quantity which must be found may or may not appear alone on one side of the equation as given. If it does not, the equation must be transposed to bring the unknown on one side and all the known quantites on the other.
When the formula is in suitable form, sub-
stitute the known values for the appropriate

13 CHAPTER ONE

DheRadio Amateur'e -ilandLoh

letters, inserting the proper signs of operation.
Where two or more factors appear together, write the multiplication sign between them when the figures are substituted. Enclose factors in brackets when their product is to be added to or subtracted from apreceding quantity, and perform the necessary numerical operations to reduce bracketed quantities to asingle number first.
When the equation has several terms, perform the operations necessary to make each
term a single number before adding or subtracting.
These points are illustrated by the following
example, using aformula for finding the output voltage of apower supply:

E',, = 0.9E:

+ /L )(RI+ R2) 1000

E,.

Since the meanings of the symbols have no
particular bearing on the present discussion, we may simply assign the following values for
the known quantities: E't = 750; ./b = 25; /L,= 100; R1 = 75; Rs = 125; E,. = 15. Substituting, we have:

= (0.9 X 750)

(25 -I- 100) (75 + 125) 1000

Collecting terms:

E.= 675

125 X 200 1000

= 675 - 25 - 15

E. = 635

When fractions appear in the denominator
of aformula it is usually best to convert them to decimals. For example, the formula for three resistances in parallel previously mentioned:

R- 1

1 1

-I71 + 71 2 + T1 3

is more easily solved by decimals than by the method of finding a common multiple for the fractions in the denominator and then adding.
Thus, suppose that the values given for RI,R2 and R3 are 500, 250 and 100, respectively. Substituting gives

R- 1

1 1

500 250 100

The corresponding decimals are 0.002, 0.004, and 0.01. Substituting again:

R -

0.002 + 0.004 -F 0.01

Performing the division, R = 62.5

1 0.016

t4 CHAPTER ONE

�1-4 LAWS OF EXPONENTS

Multiplication and division -- ASwe have seen (� 1-2), the power to which anumber is raised is the number of times it is used as a factor. The number a alone means the first

power of a, or a1,but the exponent 1is seldom

written since anumber having no exponent is

always understood to be its first power.

Now if we multiply a2 by a, we really have

aaa, since a2 = aa. Also if we multiply a2 by a2

we have aaaa, or a4.Similarly, a3 X a2 (ordi-

narily written a3a2)is equal to aaaaa, or a�. In

other words, when two powers of the same

number are multiplied together, the product

is the number raised to apower which is equal

to the sum of the exponents. Thus, aa3 = fel;

a 3a 4 = a 7; b4b

b5,and so on. It is necessary

to remember that anumber with no exponent written always has the exponent 1.
Suppose now we have a3/a.This is the same as saying

aaa

in which one ain the numerator cancels one in

the denominator, leaving aa, or a2. Or,

b3/b2 = bbb/bb; the two b's in the denominator

cancel two in the numerator, leaving bas the

answer. When apower of anumber is divided

by another power of the same number, the

quotient is the number raised to apower equal

to the exponent of the dividend minus the

exponent of the divisor. Thus, a4/a = a3;

b5/b2 = b3;b6/b3 b3,etc. This fact opens the

possibility of writing a divisor in a different

way. Thus, a4/a = a3 can be written as a

multiplication in which the exponent of the

divisor is negative: a4a-1 = a3. Addition of

the negative exponent is the same as sub-

tarrea:ctbi�obn-2of

apositive exponent. Other examples b3 ; a 2a -1 = a, and so on.

Consequently, anegative exponent indicates

division by the same number with a positive

exponent. In other words, a-3 = 1/a3;b-5 =

1/b5;a-1 = 1/a, etc. This form is frequently

used in formulas where it is desired to avoid

writing the equation as afraction.

When anumber (other than 0) is divided by

itself the quotient is 1; that is, a/a = 1. Since

a = a1 and 1/a = a 1, a/a can be written

ala-1 = 1, or simply aa-1 = 1. By the law of

exponents this also can be written aa-1 = a�.

Similarly, a3a-3 = a�; b2b-2 = b�, etc. Dividing

anumber by itself is equivalent to raising the

number to the 0th power; therefore the ex-

ponent 0means that the number reduces to 1.

Significant figures; order of magnitude

-- When numerical values are very small or

very large it is common practice to make two

factors out of the number, one of which is the

"significant" part and the other a power of

ten, the "order of magnitude." For example,

..7ormulaJ and graph3

the number 568,000,000 would be separated into the two factors 568 and 1,000,000 and written 568 X 106;293,000 would be separated (factored) into 293 and 1000 and written 293 X 103;0.0016 would be factored to 16 and 1/10,000 (0.0001) and written 16 X 10-4 ; 0.000004 would be factored to 4and 1/1,000,000 (0.000001) and written 4 X 10 -6,and so on. The following table shows the powers of

ber, the processes of multiplication and di-
vision reduce to addition and subtraction of exponents, and the processes of raising to a
power and taking roots reduce to multiplication and division of exponents. Advantage can be taken of this fact to simplify numerical operations, provided all ordinary numbers can be expressed as exponents of some number chosen as the base. Such exponents are called

10 commonly used as factors in writing numbers in such form:

106 = 1,000,000

105

100,000

104 = 10,000

103 =

1,000

102 =

100

10-1 =

0.1

10-2 =

0.01

10-3 =

0.001

10-4 =

0.0001

10-5 =

0.00001

10-6 =

0.000001

The factor 10, when used, is of course simply written as 10.
In practical work it is sufficient in nearly all cases to obtain results only to three significant
figures. Subsequent figures can be dropped or rounded off, retaining only the appropriate power of ten to give the order of magnitude.
Powers and roots-- To square a number such as a3,we use it twice as afactor, thus:

logarithms, and in the common system of logarithms the base chosen is the number 10. The logarithm of anumber is the exponent of the power to which the base 10 must be raised to produce the given number.
Since 101 = 10, the logarithm of 10 is 1; since 103 = 1000, the logarithm of 1000 is 3,
and so on. Likewise, the logarithm of 1 is 0 (written log 1= 0), since 10� = 1; log 0.1 = - 1, since 10-1 = 0.1, etc. Some of the logs of the powers of 10 are as follows:

log 10,000 = 4

log 1000 = 3

log 100 = 2

log

10 = 1

log

1= 0

log 0.1 = -1

log 0.01 = -2

log 0.001 = -3

log 0.0001 = -4

The logarithms of all numbers between 1 and 10 must have values lying between 0 and

a3a3 = a6. The operation usually would be 1, since log 1 = 0 and log 10 = 1. They

indicated by writing it in the form (a3)2,in- are, for convenience in calculation, usually

dicating that the quantity inside the brackets is to be squared. Similarly, (a2)3 = a2 a2 a2 = (16; (b4)2 b4 b4 = b8,and so on. The exponent
in the result is equal to the product of the exponents in an expression such as (a2)4 = a8. It follows that in taking aroot of such an ex-

written as decimals. For example, 10 1 (or
its equivalent 100.5)represents �Vii5, the value of which, to two decimal places, is 3.16, there-
fore log 3.16 = 0.5. Similarly, -VI)I = 10�. 333 , the value of the number being 2.15, therefore

pression as

we should divide the ex-

ponent of the number under the radical sign

by the index of the root to find the exponent

of the result; �Zra-9 a3;Nfal = a2;.N4fct-4 = a;

= a2,etc. Since division of exponents is indicated in such cases, we can write an ex-

pression like .N/0 as 0, which equals

Va4 can be written a1,equalling a2;

= a, �Wi6 = a1 = a2, and so on.

a3; An

log 2.15 = 0.333; -V102 = 10 0.666 ;the number is 4.64, therefore log 4.64 = 0.666. The logarithms of the integers from 1 to 10 are given, to four decimal places, in the following table:
log 1 = 0 log 2 = 0.3010 log 3 = 0.4771 log 4 = 0.6021 log 5 = 0.6990 log 6 = 0.7782

4
expression such as a1 also can be written

log 7 = 0.8451 log 8 = 0.9031

(a4),

since 4

multiplication

the

exponents

gives a2,and similarly for the other examples.

In other words, a fractional exponent of the

form 1/n means the same thing as

log 9 = 0.9542 log 10 1.0000
Despite the fact that the table above is extremely limited in scope, it will serve to illustrate the use of logarithms.

�1-3 LOGARITHMS

Fundamental relations-- Remembering

Logarithms-- The laws of exponents (� 1-4) the laws of exponents, and that logarithms are

show that, when we deal with one basic num- actually exponents of powers of the base, we

15 CHAPTER ONE

n eRadio AmaleaA ilandtooh

can express the rules for their use by the equations:
log (a X b) = log a + log b log a = log a - log b

decimal part of the logarithm is called the mantissa and the integral part is called the
characteristic. When the characteristic is negative it is customary to add 10 to it and then indicate that 10 is to be subtracted from the

log a. = n X log a log a. = log a

whole logarithm. Thus 2.7782 would be written 8.7782 - 10, 1.7782 would be written 9.7782

- 10, etc. This avoids having a negative

Suppose we multiply 2 by 4. Using the first rule and substituting 2for aand 4for b,
log (2 X 4) = log 2 + log 4 = 0.3010 � 0.6021 = 0.9031

characteristic with a positive mantissa. It would be possible, of course, to subtract di-
rectly; thus the algebraic sum of 0.7782 and - 1is - 0.2218, but this form would be less useful, because a negative exponent indicates that the same number with apositive exponent

This is the logarithm of the product, and we find from the table that 0.9031 is the log of 8. Therefore 2 X 4 = 8.
Divide 10 by 5, using the second rule: Substituting 10 for aand 5for b,

is to be divided into 1. Thus 10 -0.2218 means

10'22'8'or - 1.6. 66 Performing the division gives

0.6 as the required number, but the same result

is more conveniently obtained by keeping the

log (10 � 5) = log 10 - log 5 = 1.0000 - 0.6990 = 0.3010

This is the logarithm of 2, therefore the

quotient is 2.

Raise 2 to the third power. Using the third

rule and substituting 2for aand 3for n,

e= log

3 X log 2

= 3 X 0.3010

= 0.9030

Within the limit of accuracy, this is the log of 8, therefore the cube of 2is 8.
Take the square root of 9. Using the fourth rule and substituting 9for aand 2for n,

mantissa positive and using the negative characteristic simply to indicate the number of decimal places in the result. Thus 0.7782 - 1 is the logarithm of 6 X 0.1, or 0.6. The number of zeros after the decimal point and before the first digit is one less than the characteristic
when the characteristic is negative. To place the decimal point when the characteristic is positive, point off from the left one more place than the number in the characteristic.
Logarithms are assembled in tables for ready reference. Only the mantissas are given; in any particular problem the characteristic must be supplied by taking the appropriate power of 10 as described above. A four-place table is given
in the Appendix; four-place tables give results

log

92!=

log 9 2

= 0.9542 � 2

= 0.4771

which are accurate to the third figure and approximately so to the fourth figure. This
usually is high enough accuracy in radio calculations. Fig. 101 shows the relationship between the numbers from 1 to 10 and their

This is the logarithm of 3, therefore the square logarithms, and the drawing can be used as a

root of 9is 3.

three-place table by reading the logarithm

The characteristic-Since we simply add opposite the given number. The scale of num-

the exponents when multiplying two powers bers is called a logarithmic scale, and is very

of the same base, it is unnecessary to determine frequently used in graphs, as discussed in later

the logarithms of any numbers other than sections.

those between 1 and 10. Every number between 10 and 100 is 10 times some number �I-6 FUNCTIONS

between 1and 10, every number between 100 and 1000 is 102 times some number between 1 and 10, and so on. Similarly, every number
between 1 and 0.1 is 1/10 of (or 10-1 times) some number between 1and 10; every number between 0.1 and 0.01 is 10-2 times some number between 1 and 10, etc. For example, the logarithm of 600, which factors to 6 X 102,is equal to the sum of the logarithms of 6 and 102,or 0.7782 � 2, which equals 2.7782. Similarly, the logarithm of 6000 is 3.7782, log 60 =
1.7782, log 0.6 = 0.7782 - 1, log 0.06 =

Variables-One quantity is said to be a function of another quantity (called the independent variable) when the value given the
latter determines the value of the former. Thus the area of asquare is afunction of the length of one side, since assigning avalue to the length of the side immediately determines the area. Similarly, the current flowing through an electrical circuit of given characteristics is a function of the applied electomotive force or voltage. That is, if
E

0.7782 - 2 (written i.7782) and so on. The

16 CHAPTER ONE

.5ormulad and graph�

Fig. 101 -- The right hand
scale gives the logarithms of the numbers shown on the left hand scale.

1.0

0.9

0.8

0.7

4 -- -1-- 0.6

0.5 3

2.5

0.4

generally, in aformula of this kind the value of the function, or dependent variable, is proportional to a power (in the example, the second
power) of the independent variable. Such a function is called apower function. Direct and
inverse proportions are really special cases of power functions where the exponents are 1and --1, respectively. The characteristic feature of a power function is that when the independent variable is increased in a constant ratio the dependent variable also increases in a constant (although not necessarily the same)
ratio. For example, when A is proportional to r2,A will become 4 times as great each time the radius is doubled.
Exponential and logarithmic functions -- Another type of function is one in which an increase in equal-value steps in one variable is accompanied by equal-ratio steps in the other
variable. This is called an exponential function. Thus if we have aseries of numbers, a, a2,a3, a4,a5,the exponent increases by 1in each step,
but the number itself increases by the factor a each time. The difference between this type of
function and the power function ean be illustrated by the following table:

1. 5

0.2

0.1
--.=
--

Independent Variable
1
2 3 4 5 6 7
9 10

Dependent Variable

Power

Exponential

Function

1
4
9 16 25 36 49 64 81 100

1
2
8 16 32 64 128 256
512 1024

Iis a function of E, when R is constant. We can say the same thing in other ways: Ivaries with E, or Iis proportional to E. In this case I is directly proportional to E, meaning that if E is changed by a certain percentage, I will change by the same percentage and in the same direction (that is, if E is made greater, I will also be greater).
If in the same formula we hold E constant and vary R, we say that Iis afunction of R. In this case Iis said to be inversely proportional to R, since if R is made larger I becomes smaller, and vice versa.
Power functions -- Other functions may have different modes of proportionality. For example, the area of acircle is
A = Tr2
If ris aconstant, the area, A, will be proportional to the square of the radius, r. Speaking

using 2as the factor for the dependent variable in each case. That is, the power function is proportional to the square (exponent = 2) of the independent variable, and the exponential function is doubled (powers of 2) for each equal-value step in the independent variable. If we call the independent variable x and the dependent variable y, the formulas for the two cases are as follows:
Power function: y = Exponential function: y = 2z
From laws of exponents (� 1-4) and the discussion on logarithms (� 1-5) it is evident that the exponential formula can also be written in logarithmic form:
x= log2 y
where the. subscript 2indicates that the logarithmic base is the number 2. In ordinary formulas the base 10 would be used; converting

/7 CHAPTER ONE

Parlio AmaleuA Jiandiooh

from one base to another is asimple process, but need not concern us here.
An important case of alogarithmic function in radio work is the loudness of signals. To the ear, equal steps in loudness are caused by equal-ratio steps in sound power. A justdetectable increase (or decrease) in loudness is called a "decibel," and the relationship between steps of loudness and power ratios is given by the formula:
Decibels = 10 X log --P2

where P1 and P2 are the first and second power levels, respectively. The logarithmic base is 10,
which is always the case when no other base is definitely specified.

� 1-7 GRAPHS
Coordinates-- A graph is apictorial means of expressing the relationship between afunction and independent variable. It shows at a glance the value of the function for acontinuous series of values of the independent variable over any range of values of the latter that may be desired. The values are shown on co�rdinate systems, the most common of which is the system of rectangular coordinates, so called because it consists of equally spaced series of
lines at right angles to each other. A graph of the relationship between E and I in the formula I = E/R, when R is constant, is shown in Fig. 102. The horizontal lines are called abscissas and the vertical lines ordinates. The horizontal base line is called the "X axis" and the vertical base line (at the left) is called the "Y axis." These names are used because it is customary to use the symbol y for the function and x for the independent variable, and to plot the values of the function on the ordinates and the values of the independent variable on the abscissas.
Plotting -- Graphs are plotted by assuming suitable values of the independent variable at
relatively small intervals, then solving the equation for the value of the function for each
value of the independent variable. Each pair of values represents a point on the graph, the position of the point being determined by the spot where the ordinate and abscissa representing those particular values cross. Thus in Fig. 102 the X axis is marked off in terms of voltage and the Y axis in current. With R constant at 100 ohms, we can make atable of the values of /as follows, using 10-volt steps for E:

When E = 0volts, / = 0amperes

E = 10

I = 0.1

E = 20 " I = 0.2 E = 30 " I= 0.3 �i`

E = 40 " I = 0.4 "

E = 50 " I = 0.5 "

1.0 0.9

a mommm Rm-1u o00mm 011iMm S zM mu z � uM am mWmun ErmIe . iM

�t�n�� 0.8 iliMMMIERMIMEM

0.7 MIMMIUMUMMIBM.
k0.6

0.5 iiii�1111111111MUMUMMI

0.4 MENEMBUMMUME cc 0.3 inirnarAMMIMMM.

0.2

MailnrIaMBIUMUM MrIMMIMMUMB.

0.1 BliIMMMiniMMB rIMIMMIIIM111111111111113111111�
00 10 20 30 40 50 80 70 80 90 100 ELECTROMOTIVE FORCE, E (VOLTS)

Fig. 102

and so on, all values being found from the formula I = E/R. The first point obviously lies on the intersection of the axis at 0 (this point is called the origin). The next point, E = 10, I = 0.1, is plotted by moving out along the X axis to the value 10, then moving up on the ordinate at that point to the value 0.1. The second point is plotted by locating E = 20 and then moving up on that ordinate to the value 0.2. The remaining points are plotted similarly. When the series of points has been plotted, asmooth line is drawn joining all of them together; this line is called acurve, even though it may be perfectly straight as in the illustration.
It can be seen from Fig. 102 that the graph of / as a function of E is a straight line. A function which is directly proportional to the independent variable always gives a straightline graph on rectangular coordinates.
Scales -- Scales of the type shown in Fig. 102, in which- equal segments on agiven scale have the same value no matter where they are taken, are called linear. Thus the length on the X axis between 0and 10 (10 volts difference) is the same as between 40 and 50 (again 10 volts difference). The same length of line always represents 10 volts. This type of scale is useful for many kinds of graphs, but in some cases the logarithmic scale is better because parts of some curves are very steep when plotted on a linear scale. That is, asmall change in the value of the independent variable causes a large change in the function, which makes it difficult to read the graph accurately. In general, it is advisable to choose scales so that the plot of the function will be a straight line, or nearly straight line, making an angle of about 45 degrees with either axis. This gives maximum readability.
Logarithmic scales for both abscissae and ordinates will give astraight-line curve for any power function. Thus in the formula P I2R,

18 CHAPTER ONE

..7ormulad an' �rctpX3

which gives the relationship between power, current and resistance, P is a function of /2 when R is constant. Using 10 ohms for R and selecting values for I, we tabulate as follows:
When I = 1, P = 10 I = 2, P = 40 / = 3, P = 90 I = 4, P = 160 I = 5, P = 250, etc.
Plotting these values on logarithmic scales (graph paper of this type is called "logarithmic" paper, or sometimes "log-log" paper) gives a curve of the type shown in Fig. 103. Note that azero point cannot be plotted, since a logarithmic scale never reaches zero. This type of scale also is useful for plotting functions which are inversely proportional to the inde-

1900000 800

700

600 500

Rr 10 OHMS

400

300

200

�k,c 19000 Q. 80 ci 70 4.1 60 Q�.. 50
40 30
20

10 1A

1.5 2 25 3 *4 5 6 78910
CURRENT, I(AMPERES)
Fig. 103

pendent variable, such as in the formula I E/R, when E is held constant and R is varied. In this case Iis afunction of 1/R.
Exponential functions are best plotted on semi-logarithmic paper, or paper having one scale linear and the coordinate scale logarithmic. The graph of such a function will be a straight line when the exponential variable is plotted to alinear scale and the linear variable to alogarithmic scale. When the formula is in logarithmic form the logs are plotted on the

linear scale. Fig. 104 illustrates the graph of the formula for the relationship between decibels and power ratios, the latter being the independent variable and the former the function. It is seen that the curve is astraight line.
In plotting graphs it is desirable to mark off the scales in units which will make the whole graph approximately square, when the "end" values are selected. Thus, suppose we are interested in Ias afunction of E, with R constant at 100 ohms, for all values of E between 0and 100 volts. The end values of E are 0and 100, and the end values of I, found from the formula, are 0and 1ampere. Using ordinary rectangular paper having 20 divisions to the inch, with every tenth line heavy, it is convenient to make each small division represent 1volt on the X axis, and each small division represent 0.01
ampere on the Y axis. The graph will then be square, as shown in Fig. 102. (To avoid difficulty in reproduction, the "units" lines are omitted in the figure, only the "fives" lines being shown.)
If alarge range of values has to be shown, logarithmic scales are to be preferred to linear scales. If we extended the graph of Fig. 102 to include all values of Iwhen E varies from 1to 1000 volts, it is evident that to maintain the same accuracy of reading, particularly in the
range shown in Fig. 102, it would be necessary to make the axes 10 times as long. If we kept the graph the same size and simply reduced the scales to accommodate the 10-times larger range of values, considerable percentage error would arise in reading the curve for values below 100 volts. This difficulty can be over-

1000

800

465000000

300

200

10o 1111111111111111111M11111111

o
e:

6o 4so0

mmmiulmimammmmimnummuummmmaruAmmmmu imimnunm

30 20

mIummmmoummmmoumnmioummmmummummmmommummummumm 11111.1111111111MMIIIIMMEM1111

�lo

65 BOIMMT.IIMMIZZUW1111WM11181 4
3 MIIIIMIIMMIIUMMMZMIUM1M1M1I91M1M1U11M8M1I11M1M8IMMMOMIIIIMIMMIM
2 Mr4111.111111MMIIMIIIIMM

td1111111111111111111111111111

DECIBELS

Fig. 104

/9 CHAPTER ONE

5hePalio ...Amateur's --llandloole

come by using logarithmic paper, since with
logarithmic scales the same percentage ac-

how the values vary compared to alinear scale, using ascale such as Fig. 101 for comparison.

curacy can be obtained no matter what the order of magnitude of the quantities. In the specific problem considered, we could use "three cycle" paper (one cycle is the scale from
1to 10, or 10 to 100, or 100 to 1000, etc.) for all values of E between 1volt and 1000 volts, as shown in Fig. 105. The range could be extended indefinitely in either direction by using more

pWohientn oinn adoluinbeta,r ebsastiismaatnedtthheenpomsietaisounreofoftfhea
corresponding fractional length on an enlarged logarithmic scale. This procedure will permit fairly accurate interpolation.
Polar coordinates -- When it is necessary to show graphically the variation in aquantity whose value is afunction of its direction with respect to a fixed reference line, the function

eo e

.1�11�MM�m�������1111.1M1 11..1�����1�...1. 1.....1., �11

can be plotted on polar coordinate paper. Fig. 106 shows such agraph. Polar coordinates con-

sist of aset of radial lines representing units of

4 3
11111011111 R =100 01484S IIIMIUM111111
2 .111.1111.1111..1.111.1. 1 1IIMUMBEIIIII

angle, and a series of concentris circles, both coordinates originating from a common point (the origin).

111.111111111111.111 1111 2. 111111111

The polar coordinate graph is especially

h o

lummmullazuni

14 0

k 02 IMIIIIIIIIIMIIIM02�11111�11111�1111�11.1111111

useful for showing the directional characteristic of an antenna system. The plot in Fig. 106 is the theoretical directional characteristic or
"pattern" for a straight, horizontal antenna when its length is considerably less than ahalf

0.10 :=ITBITIZ=ZZ;ri====re;;;;;Zel=1MICren:

fl.11.11.M.MM.1./1,/���1.1 11.1MAOMMI���IM.111MuMMe ��

0.06 005
004 21111.edill111.11.1.1111111.111111.1.1111
0.03

002

wavelength. If the antenna wire lies in the direction indicated on the graph, the relative field strength, or relative intensity of radiation, will vary with the direction with respect to the
wire as shown, assuming that the field strength is observed at the same distance in any direc-

tion. Note that the scale in such acase is purely

0.01

2 3 4 561810

20 30 405060 80100

200 300 4 5 6 8 1000

relative, and the maximum strength, which

E - VOLTS

occurs in the direction at right angles to the

Fig. 105

length of the wire, is arbitrarily assigned a

value of 1. If we measure the field strength

cycles or changing both scales in the same ratio directly off the end of the wire it will be zero;

of some power of 10. As stated before, we can-
not reach zero on this type of scale, but in many

cases this is no particular handicap.

Interpolation -- On graph paper the co-

ordinates must be spaced at definite intervals,

and if we are interested in values lying between

tpowloatec.ooIrndtienraptoelatliinoens

it is

is necessary to inter-
simply estimating the

value from the position of the point between

200' loo� Ise no* 160.

210"

150'

220�

140�

230*

13o�

240'

120'

250� 260� 27

110�
(Kr so-

two lines. If the scale is linear, half the distance between two lines will add to the value of the
lower line half the difference between the assigned values of the two lines. That is, if a point lies half-way between two ordinates marked "4" and "5", the difference between the two values is 1, hence the half-way position indicates 0.5. Adding this to the value of the lower ordinate gives 4.5 as the value at that

2
30 310� 320* 330* 340'350' 0. 10� Fig. 106

80'
70�
60'
50' 408 30'

point. It is usually possible to estimate to onetenth of aunit when the coordinate lines are spaced inch or more apart. If they are closer together it may be possible to estimate accurately only to unit.
Interpolation on logarithmic scales is alittle more difficult, but can be learned by studying

at 45 degrees it will be 0.71, or 71%, of its value
at right angles to the wire, and so on. If the field strength is uniform in all direc-
tions with respect to the antenna, the graph of
the function will simply be acircle concentric with the origin.

20 CHAPTER ONE

I' II A P T IE

TW O

electricaland Pali() .7unciamentaA

�2-1 FUNDAMENTALS OF A �RADIO SYSTEM
THE BASIS of radio communication is the transmission of electromagnetic waves through space. The production of suitable waves constitutes radio transmission, and their detection, or conversion at adistant point into the intelligence put into them at the originating point, is radio reception. There are several distinct processes involved in the complete chain. At the transmitting point, it is necessary first to generate power in such form that when it is applied to an appropriate radiator, called the antenna, it will be sent off into space in electromagnetic waves. The message to be conveyed must be superimposed on that power by suitable means, aprocess called modulation.
As the waves spread outward from the transmitter they rapidly become weaker, so at the receiving point an antenna is again used to abstract as much energy as possible from them as they pass. The wave energy is transformed into an electric current which is then amplified, or increased in amplitude, to a suitable value. Then the modulation is changed back into the form it originally had at the transmitter. Thus the message becomes intelligible.
Since all these processes are performed by electrical means, a knowledge of the basic principles of electricity is necessary to understand them. These essential principles are the subject of the present chapter.
�2-2 THE NATURE OF ELECTRICITY
Electrons-- All matter -- solids, liquids and gases -- is made up of fundamental units called molecules. The molecule, the smallest subdivision of a substance retaining all its characteristic properties, is constructed of atoms of the elements comprising the substance.
Atoms in turn are made up of particles, or charges, of electricity, and atoms differ from each other chiefly in the number and arrangement of these charges. The atom has a nucleus containing both positive and negative charges, with the positive predominating so that the nature of the nucleus is positive. The charges in the nucleus are closely bound together. Exterior to the nucleus are negative charges -- electrons -- some of which are not so closely bound and can be made to leave the vicinity of the nucleus without too much urg-

ing. These electrons whirl around the nucleus like the planets around the sun, and their orbits are not random paths but geometricallyregular ones determined by the charges on the nucleus and the number of electrons. Ordinarily the atom is electrically neutral, the outer negative electrons balancing the positive nucleus, but when something disturbs this balance electrical activity becomes evident, and it is the study of what happens in this unbalanced condition that makes up electrical theory.
Insulators and Conductors -- Materials which will readily give up an electron are called conductors, while those in which all the electrons are firmly bound in the atom are called insulators. Most metals are good conductors, as are also acid or salt solutions. Among the insulators are such substances as wood, hard rubber, bakelite, quartz, glass, porcelain, textiles, and many other non-metallic materials.
Resistance-- No substance is aperfect conductor -- a "perfect" conductor would be one in which an electron could be detached from the atom without the expenditure of
energy -- and there is also no such thing as a perfect insulator. The measure of the difficulty in moving an electron by electrical means is called resistance. Good conductors have low resistance, good insulators very high resistance. Between the two are materials which are neither good conductors nor good insulators, but they are nonetheless useful since there is often need for intermediate values of resistance in electrical circuits.
Circuits -- A circuit is simply a complete path along which electrons can transmit their charges. There will normally be a source of energy (a battery, for instance) and aload or portion of the circuit where the current is made to do useful work. There must be an unbroken path through which the electrons can transmit their charges, with the source of energy acting as an electron pump and sending them around the circuit. The circuit is said to be open when no charges can move, due to a break in the path. It is closed when no break exists -- when switches are closed and all connections are properly made.
�2-3 STATIC ELECTRICITY
The electric charge -- Many materials that have ahigh resistance can be made to acquire

21 CHAPTER TWO

DheRadio Amaleur'3 ilattigooL

acharge (surplus or deficiency of electrons)
by mechanical means such as friction. The familiar crackling when a hard-rubber comb is run through hair on a dry winter day is an example of an electric charge generated by friction. Objects can have either asurplus or a deficiency of electrons -- it is called anegative charge if there is a surplus of electrons; a positive charge if there is a lack of them. As with all things in nature, there must always be abalance, and for every negative charge there
will be found a similar positive charge, since each electron that leaves an atom to form a negative charge leaves the rest of the atom with a positive charge. The kind of charge is called polarity, anegative charge constituting anegative pole, apositive charge being aposi-
tive pole. Attraction and repulsion -- Unlike charges
(one positive, one negative) exert an attraction
on each other. This can be demonstrated by

adopted to explain the "action at adistance" of the charge. The field is assumed to consist of lines of force originating on the charge and spreading in all directions. The number of lines of force per unit area is ameasure of the intensity of the field.
Potential difference -- If two objects are charged differently, a potential difference is said to exist between them, and this difference is measured by an electrical unit called the volt. The greater the potential difference, the higher (numerically) the voltage. This potential difference or voltage exerts an electrical
pressure or force as explained above, and for this reason it is often called electromotive force
or, simply, e.m.f. It is not necessary to have unlike charges to have adifference of potential; both, for instance, may be negative so long as one charge is more intense than the other. From the viewpoint of the stronger charge, the weaker one appears to be positive in such
acase, since it has asmaller number of excess

electrons; in other words, its relative polarity

is positive. The greater the potential difference

the more intense is the electrostatic field be-

tween the two charged objects.
Capacity--If two metal plates are sepa-

rated ashort distance by ahigh-resistance ma-

terial, such as glass, mica, oil or air, it will be

found that the two plates can be given acharge

Fig. 201 -- Attraction and repulsion of charged objects, as shown by the pith-ball experiment.

by connecting them to a source of potential difference. Such adevice is called acondenser, and the insulating material between the metal

giving equal but opposite charges to two very plates is called the dielectric. The potential

light objects of insulating material (pith balls difference, or voltage, of the charge will be

are used in the classical experiment) and sus- equal to that of the source. The quantity of the

pending them near each other. They will be charge will depend upon the voltage of the

drawn toward each other, and if they touch the charging source and the capacity of the con-

charges will neutralize, leaving both objects denser. The value of capacity of a condenser

without charge. Charges of the same type, however, repel each other, and asimilar experiment with like-charged objects will show them tend-

is a constant depending upon the physical dimensions, increasing with the area of the plates and the thinness and dielectric constant

ing to swing apart.

of the insulating material in between. The

Electrostatic field -- From the foregoing it dielectric constant of air is 1, while for other

is evident that an electric charge can exert a insulating materials it is usually higher. Glass,

force through the space surrounding the for instance, has adielectric constant of about

charged object. The region in which this force is exerted is considered to be pervaded by the electrostatic field, this concept of a field being

4; this means, simply, that if glass is substituted for air as the dielectric in an otherwise identical condenser, the capacity of the condenser will be four times as great.

oflfinoersce--

Cbhoadryged
(

Capacity is measured in farads, aunit much too large for practical purposes, and in radio work the terms microfarad (abbreviated fd.) and micro-microfarad (1.41d.) are used. The

microfarad is one-millionth of afarad, and the

micro-microfarad is one-millionth of a micro-

farad. The electrical energy in acharged condenser

Fig. 202 -- Lines of force from acharged object extend outward radially. Although only two dimensions are shown, the field extends in all directions from the charge, and the field should be visualized in three dimensions.

is considered to be stored in much the same way that mechanical energy is stored in a stretched spring or rubber band. Whereas the mechanical energy in the spring can be stored

22 CHAPTER TWO

electrical and Radio Dundamenlat

because of the elasticity of the material, the electrical energy in a condenser is stored in the electrostatic field between the plates.
Condensers -- The construction of a condenser is determined by the use for which it is intended. Where the capacity must be continuously adjustable, as in tuning radio circuits, sets of interleaved metal vanes are used with air as the dielectric. In high capacity units
1T T

To soierf of em.t.

Symbols

Fig. 203 -- A simple type of condenser, consisting of two metal plates with dielectric material between. The diagrammatic symbols for condensers are shown at the right. The two at the top indicate condensers of fixed capacity, the two below, condensers whose capacity is variable. The symbols in the left hand column are more commonly used.

where adjustment is not required, the dielectric
may be thin paper or mica. The choice of adielectric and its thickness is determined by the capacity desired, the voltage for which the condenser is intended and, in many cases, by the losses in the dielectric, since the electrical stress caused by the electrostatic field is accompanied by consumption of energy which appears as aheating effect in the dielectric.

� 2-4 THE ELECTRIC CURRENT
Conduction -- If a difference of potential exists across the ends of aconductor (by connecting the conductor -- usually a wire -- to a battery or generator or other source of volt-
age) there will be acontinuous drift of electrons from atom to atom, and an electrical current is said to be flowing. The individual electrons do not streak from one end of the conductor to the other but the action is rather like a"bucket brigade" where, instead of firemen handing buckets down the line, atoms pass electrons

f Batiery

171-+-=+ -+
Fig. 204 --Electrolytic conduction. When an e.m.f. is applied to the electrodes, negative ions are attracted to the positively charged plate and positive ions to the negatively charged plate. The battery is indicated by its customary symbol.

down the line of the conductor. The current,

or total effect of the electron drift, travels

quite fast, close to the speed of light, but the

electrons themselves move only a short

amperes, distance. The current is measured in

and a

current of one ampere represents nearly 10 19

(ten million, million, million) electrons flowing

past a point in one second. On more familiar

ground, the current which flows through an

ordinary 60-watt lamp is approximately one-

half ampere.

Gaseous conduction (ionization) -- All

conduction does not necessarily take place in

solid conductors. If aglass tube is fitted with

metal plates at each end, and filled with agas

or even ordinary air (which is a mixture of

gases) at reduced pressure, an electric current

may be passed through the gas if ahigh enough

voltage is applied across the metal terminals.

When the voltage is applied across the tube,

the positively charged plate attracts a few

electrons, which acquire considerable velocity

because of the electric charge and the fact that

the reduced pressure in the tube (less gas) per-

mits the electrons to travel farther before

colliding with a gas atom. When one does

collide with an atom, it knocks off an outer

electron of the gas atom and this electron also

joins the procession towards the positive

plate, knocking off more electrons from other

atoms as it goes. The atoms that have had an

electron or two knocked off are no longer true

atoms but ions, and since they have apositive

charge (due to the electron deficiency) they

are called "positive ions." These positive ions,

being heavier than the electrons, travel more

slowly towards the negative plate, where they

acquire electrons and become neutral atoms

again. The net result is a flow of electrons,

and hence of current, from negative plate

(called the cathode) to positive plate (anode).

Current flow in liquids -- A very large

number of chemical compounds have the pe-

culiar characteristic that when they are put

into solution the component parts become

ionized. For example, common table salt or

sodium chloride, each molecule of which is made up of one atom of sodium and one of

chlorine, will, when put into water, break down

into asodium ion (positive, with one electron

deficient) and a chlorine ion (negative, with

one excess electron). This can only occur so

long as the salt is in solution -- take away the

water and the ions are recombined into the

neutral sodium chloride. This spontaneous

disassociation in solution is another form of

ionization, and if two wires with a difference

of potential across them are placed in the solu-

tion, the negative wire will attract the positive

sodium ions and the positive wire will attract

the negative chlorine ions, and a current will

23 CHAPTER TWO

Dheleach� Amcileur's -Warn/Loh

flow through the solution. When the ions reach the wires the electron surplus or deficiency will be remedied, and aneutral atom will be formed. The energy supplied by the source of potential difference is used to move the ions through the liquid and to supply or remove electrons. This type of current flow is called electrolytic conduction.
Batteries -- All batteries depend upon chemical action for the generation of a potential difference across their terminals. The common dry cell (which will not work when completely dry) depends upon zinc ions (the metal case of adry cell is the zinc plate) with apositive charge going into solution and leaving the zinc plate strongly negative. The electrical energy is derived from the chemical energy, and in time the zinc will be used up or worn away. However, in lead storage batteries, such as are used in automobiles for starting, the electrical energy is stored by chemical means and entails no destruction of the battery materials. The water that must be replaced from time to time is lost by evaporation.
It might be pointed out here that the term "battery" is used correctly only when speaking of more than one cell -- asingle cell is not abattery, but two or more connected together become abattery.
Current flow in vacuum -- If a suitable metallic conductor, such as tungsten or oxidecoated or thoriated tungsten, is heated to a high temperature in a vacuum, electrons will be emitted from the surface. The electrons are
Fig. 205 -- Illu,tratuag conduction by thermionic emission of electrons in avacuum tube. One battery is used only to heat the filament to a temperature where it will emit electrons. The other battery places apositive potential on the plate, with respect to the filament, and the electrons are attracted to the plate. The flow of electrons completes the electrical path, and current flows in the plate circuit.
freed from this filament or cathode because it has been heated to a temperature that activates them sufficiently to allow them to break away from the surface. The process is called thermionic electron emission. Now if a metal
plate is placed in the vacuum tube and given a

positive charge by connecting a battery between plate and cathode, this plate or anode will attract anumber of the electrons that surround the cathode. The passage of the electrons from cathode to anode constitutes an electric current. All thermionic vacuum tubes depend for their operation on the emission of electrons from ahot cathode.
Direction of current flow -- Use was being made of electricity for a long time before its electronic nature was understood, and while it is now clear that current flow is adrift of negative electrical charges or electrons toward a positive potential, in the era preceding the electron theory it was assumed that the current flowed from the point of higher positive potential to apoint of lower (i.e. less positive or more negative) potential. While this assumption turned out to be wholly wrong, it is still customary to speak of current as flowing "from positive to negative" in many applications. The practice often causes confusion, but this distinction between "current" flow and "electron" flow often must be taken into account. If electron flow is specifically mentioned there is of course no doubt as to the meaning, but when the direction of current flow is specified it may be taken, by convention, as being opposite to the true direction.
� 2-5 ELECTROMAGNETISM
The magneticfield -- The power that abar or horseshoe magnet possesses of attracting small pieces of iron to itself is known to everyone. As in the case of electrostatic attraction (� 2-3) the concept of afield of magnetic force is adopted to explain the magnetic action. The field is made up of lines of magnetic force, the number of which per unit area determine the strength of the field.
A moving electron generates amagnetic field of exactly the same nature as that existing about a permanent magnet. Since a moving electron, or group of electrons moving together, constitutes an electric current, it follows that the flow of current is accompanied by the creation of amagnetic field.
Conversely, when a conductor is moved through amagnetic field (or the field is moved past the conductor) electrons in the conductor are forced to move, producing a current. An electric current generates amagnetic field about it and, conversely, an electric current is generated by amagnetic field moving (or changing) past the conductor.
When a conductor carrying a current lie placed in a magnetic field, a force is exerted on the wire which tends to move it in adirection determined by the relative directions of the flux lines of the external field and that set up by the current flow in the wire. This is a corollary of the fact that acurrent is induced

24 CHAPTER TWO

electrical and Mello ,Junc1amentafS

in awire moving in amagnetic field, and is
the principle used in the electric motor.
Magnetomotive force -- When the conductor is a wire, the lines of force are in the
form of concentric circles around the conductor and lie in planes at right angles to the axis of the conductor. The magnetic field constituted by these lines of force exists only when current is flowing through the wire. When the current
Fig. 206 -- Whenever current passes through a wire, a magnetic field exists around the wire. Its direction can be traced by means of a small compass.
is started through the wire, we may visualize
the magnetic field as coming into being and
sweeping outward from the axis of the wire. On cessation of current flow, the field collapses
toward the wire and disappears. Thus energy is alternately stored in the field and returned to the wire. When aconductor is wound into the form
of acoil of many turns, the magnetic field becomes stronger because there are more lines of
force. The force is expressed in terms of magneto-motive force (m.m.f.) which depends on the
number of turns of wire, the size of the coil and the amount of current flowing through it. The same magnetizing effect can be secured with a great many turns and aweak current or with few turns and astrong current. If 10 amperes
ir
SYMBOLS
Fig. 207 -- When the conducting wire is coiled, the individual magnetic fields of each turn are in such a direction as to produce afield similar to that of a bar magnet. The schematic symbols for inductance are shown at the right. The symbols at the left in the top row indicates an iron-core inductance; at right, air core.
Variable inductances are shown in the bottom row.

flow in one turn of wire, the magnetizing effect is 10 ampere-turns. Should one ampere flow in
10 turns of wire, the magnetizing effect is also 10 ampere-turns.
Inductance -- When a source of voltage is connected across a coil, the current does not immediately reach its final fixed value. The
reason for this is that, as the current starts to
flow through the coil, the magnetic field around
the coil builds up, and as the field changes it induces avoltage back in the coil. The current
caused by this induced voltage is always in the opposite direction to the current originally passed through the coil. Therefore, because of this property of self-induction, the coil tends constantly to oppose any change in the current
flowing through it, and it takes an appreciable
amount of time for the current to reach its normal value through the coil. The effect can
be visualized as electrical inertia. After the cur-
rent has come to a steady value, the selfinductance has no effect, and the current is
only limited by the resistance of the wire in the
coil.
The inductance of a coil is measured in
henrys or, when smaller units are more conven-
ient, the millihenry (one thousandth of ahenry) or microhenry (one-millionth of a henry). The inductance of acoil depends on several factors,
chief of which are the number of turns, the cross-sectional area of the coil, and the material
in the center of the coil, or core. A core of magnetic material will greatly increase the inductance of a coil, just as certain dielectrics
greatly increase the capacity of a condenser (� 2-3). Even a straight wire has inductance, although small compared to that of acoil.
The inductance of a straight wire of given
length is less as the diameter of the wire is increased. In general, aconductor of large cross-
sectional area, or large surface, will have less inductance than one of small area but having the same length.
Magnetic circuits and units -- Unlike elec-
trostatic lines of force, magnetic lines of force must always be closed, forming circles or loops,
so that the complete magnetic path of the lines
of force must be considered in computing the effect of a magnetic core material on the inductance of acoil. The measure of the number
of magnetic lines of force set up in a closed magnetic path or circuit through agiven material for aspecified applied m.m.f. is called the magnetic permeability of the material. It is ex-
pressed as aratio to the number of lines set up
by the same coil with the same applied m.m.f.
with air as the core material, air therefore being
assigned a permeability of unity. If the mag-
netic circuit is partly through a magnetic
material and partly through a non-magnetic
material (as in the case of acoil wound on a straight bar of iron, where part of the magnetic

25 CHAPTER TWO

ledio AmaieuA --fluttcMooh

path must be through air) the permeabilities of both mediums must be taken into account.
Permeability corresponds to conductivity in conductors, and its reciprocal, reluctance, corresponds to resistance. Magnetic flux density, or lines of force per unit area, is in the magnetic
circuit equivalent to current in the electrical
circuit, while the magnetomotive force is
analogous to electromotive force or voltage.

� 2-0 FUNDAMENTAL RELATIONS
Ohm's law -- The current in aconductor is determined by two things, the voltage across the conductor and the resistance of the conductor. The unit of resistance is the ohm, and, by definition, an e.m.f. of one volt will cause a current of one ampere to flow through aresistance of one ohm. Since the three quantities are interdependent, if we know the values of any two we can easily determine the third by the simple relation known as Ohm's Law. When I is the current in amperes, E is the electromotive force in volts and R is the circuit resistance in ohms, the formulas of Ohm's Law are:

R = --E

,= --E

E = IR

The resistance of the circuit can therefore be found by dividing the voltage by the current: the current can be found by dividing the voltage by the resistance: the electromotive force or e.m.f. is equal to the product of the resistance and the current.
The resistance of any metallic conductor depends upon the material and its temperature, its cross-sectional area and the length of the conductor. When resistance is deliberately added to a circuit, as is often done to adjust voltages or limit current flow, the resistance is usually lumped in asingle unit and the unit is called aresistor.
Heating effect and power -- When current passes through a conductor there is some molecular friction, and this friction generates heat. The heat generated is dependent only upon the current in the conductor, the resistance of the conductor and the time during which the current flows. The power used in heating (which may be considered sometimes as an undesired power loss) can be determined by substitution in the following equations:
P = El, or P = I2R,
or P = E2

across aresistance and the current through it are known, the product of volts and amperes will give the power. Knowing the value of a resistor (ohms) and the applied voltage across it, the power dissipated is given by the last formula.
Likewise, when the power and resistance in a circuit are known, the voltage and current can be calculated by the following equations derived from the power formulas given above:
e E = VITR
I =

Units--Besides the fundamental units --
volt, ampere, watt -- fractional and multiple units frequently are convenient. Thus a milliampere is 1/1000 ampere and a microampere is 1/1,000,000 ampere. Millivolt and microvolt are corresponding fractional units of the volt. The kilovolt also is afrequently used unit; it is equal to 1000 volts. Resistance is frequently expressed in megohms (1 megohm = 1,000,000 ohms) and sometimes in kilohms (1000 ohms). Other units for power are the microwatt, milliwatt, and kilowatt, having equivalent meanings to those above. The watt-hour and kilowatt-hour are energy units, representing the total energy consumed when it is delivered at a given power rate for a given period of time;
the numerical values are equal to the product of power and time in the units named.
Unless otherwise specified, formulas are always given in terms of the fundamental units, so that fractional or multiple units must first be converted to the fundamental units before
an equation can be used. Resistances in series and parallel--Re-
sistors may be connected in series, in parallel or in series-parallel, as shown in Fig. 208. When two or more resistors are connected in
series, the total resistance of the group is

SERIES

I
PARALLEL R,

0A
S

$ R

P being the power in watts, E the e.m.f. in volts, and Ithe current in amperes.
It will be noted that if the current in a resistor and the resistance value are known, we can readily find the power. Or if the voltage

SERIES PARALLEL
Fig. 208 -- Diagrams of series, para11,1 and seric,� parallel resistance connections.

26 CHAPTER TWO

actricaf an] echo .7uniairienfidi

higher than that of any of the units. Should two or more resistors be connected in parallel, the total resistance is decreased. Fig. 208 and the following formulas show how the value of a bank of resistors in series, parallel or series-
parallel may be computed, the total resistance being that which appears between A and B in each case.
Resistances in series:

Total resistance = Ri � R2 + R3 + R4

Resistances in parallel:

Total resistance

1 +

1 1+ 1+

11.1

112

113

114

Or, in the case of only two resistances in parallel,

Total

resistance

R1R2
R1+ R2

Resistances in series-parallel: Total resistance =

R1+ Rg + R3 + R4 + 14 + R8 + R7 + R8 + Rg

This means that in series-parallel circuits the various groups of series resistors should
first be added, then each group treated as a single resistor, so that the formula for resistances in parallel can be used.
Voltage dividers and potentiometers -- Since the same current flows through resistors connected in series, it follows from Ohm's Law that the voltage (termed voltage drop) across each resistor of aseries-connected group is proportional to its resistance. Thus in Fig. 209-A the voltage El across R1 is equal to the applied voltage E multiplied by the ratio of Ra to the total resistance, or

Ra -F R2 + Ra

Similarly, the voltage E2 is equal to

+ R2 �E Ri � R2 + R3
Such an arrangement is called avoltage divider. When current is drawn from the divider at the

(A)

(B)

Fig. 209 -- The voltage divider or potentiometer.

various tap points the above relations are no longer strictly true, since the same current does not flow in all parts of the divider. Design data for such cases are given in �8-10.
A similar arrangement is shown in Fig. 209-B, where the total resistance R is equipped with asliding tap for fine adjustment. Such a resistor is frequently called apotentiometer, although the word is not used in its original sense.
Inductances in series and parallel -- The formulas for the total inductance of agroup of separate inductances connected in series, parallel, or series parallel are exactly the same as those given in the previous paragraph for resistances, provided only that the magnetic fields about the coils are not permitted to interact with each other.
Condensers in series and parallel -- The total capacity of a group of condensers connected in series, parallel or series parallel can be computed by formulas similar to those used for resistances and inductances, but with the series and parallel formulas interchanged. Thus, for condensers in parallel,
Total capacity =C1 + C2 + Cg + C4,etc.

For condensers in series,

Total capacity -- 1

1 1 1

t.,

+
1

or for two condensers in series

Total capacity -- C1C2 + C2
With condensers in series parallel, first compute the resultant capacity of the condensers in series in each parallel branch, then add the capacities so found for the various branches.
Time constant -- When a condenser and resistor are connected in series with a source of e.m.f. such as a battery the initial flow of current into the condenser is limited by the resistance, so that a longer period of time is required to complete the charging of the condenser than would be the case without the resistor. Likewise, when the condenser is discharged through a resistance, a measurable period of time is taken for the current flow to reach a negligible value. In the case of either charge or discharge the time required is proportional to the capacity and resistance, the product of which is called the time constant of the circuit. If C is in farads and R in ohms, or C in microfarads and R in megohms, this product gives the time in seconds required for the voltage across adischarging condenser to drop to 1le or approximately 37% of its original value. (The constant e is the base of the natural series of logarithms.)
A circuit containing inductance and re-

27 CHAPTER TWO

Dh e Radio Amaieur'� ilanitooL

/nstant of

TIME

Closin9Switoh

10,-�

c/onssibn2fnstwo'ftch

TIME

Fig. 210 -- Showing how the current in acircuit com-
bining resistance with inductance or capacity takes a finite period of time to reach its steady-state value.

sistance also has a time constant, for similar reasons. The time constant of an inductive circuit is equal to LIR, and when Lis in henrys and R in ohms gives the time in seconds required for the current to reach 1-1/e or approx-
imately 63% of its final steady value when a constant voltage is applied to the circuit.
Measuring instruments -- Instruments for measuring d.c. current and voltage make use of the force acting on acoil carrying current in a magnetic field (� 2-5), produced by a permanent magnet, to move a pointer along a calibrated scale. All such instruments are therefore current operated, the current required for full-scale deflection of the pointer varying from several milliamperes to a few microamperes according to the sensitivity required. If the instrument is to read high currents, it is shunted (paralleled) by a low resistance through which most of the current flows, leaving only enough flowing through the instru-
ment to give afull scale deflection corresponding to the total current flowing through both meter and shunt. An instrument which reads microamperes is called a microammeter or galvanometer; one calibrated in milliamperes is
called a milliammeter; one calibrated in amperes is an ammeter. A voltmeter is simply amilliammeter with a high resistance in series so that the current will be limited to a suitable value when the instrument is connected across a voltage source; it is calibrated in terms of the voltage which must appear across the terminals to cause a given value of current to flow. The series resistance is called a multiplier. A wattmeter is a combination voltmeter and ammeter in which the pointer deflection is proportional to the power in the circuit.

An ammeter or milliammeter is connected in series with the circuit in which current is
being measured, so that the current flows through the instrument. A voltmeter is connected in parallel with the circuit.

� 2-7 ALTERNATING CURRENT
Description-- In self-induction the induced voltage always opposes the voltage causing the original current flow (� 2-5). Similarly, if a closed wire is placed in an expanding magnetic field, the current induced in the wire by the changing field will flow in such adirection that the magnetic field set up in turn by this induced current opposes the field which originally caused it. Now if the original field is caused to collapse (moving toward the wire instead of outward from it) the induced current will change its direction so that its field again will be in opposition to the original field. If the primary field regularly builds up and collapses the current will change
direction correspondingly; in other words, it is an alternating current. Since current is only caused to flow by achanging magnetic field, it is easy to see why alternating currents are widely used; they are a natural result of the application of the principle of induction.
The simplest form of alternating current (or voltage) is shown graphically in Fig. 211. This chart shows that the current starts at zero
value, builds up to a maximum in one direction, comes back down to zero, builds up to a maximum in the opposite direction and comes back to zero. The curve followed is described mathematically as a sine curve; its wavelike nature causes it to be known as asine wave.
Frequency -- The complete wave shown in Fig. 211 is called acycle, or period. Each half of the cycle, during which the current is flowing in one direction, although its strength is varying, is known as an alternation. The number of cycles the wave goes through each second of time is called the frequency of the current. Frequencies vary from afew cycles per second
for power line alternating currents to many

,Peak value

I 0

A. C. meters read the

.707

effective (rm.s.)valves

ofcurrent and voltage

(rms=.707 of peak

value of sine wave

rl �
360 �

One cycle
I 0
Fig. 211 -- Representing sine-wave alternating current or voltage.

28 CIIAPTER TWO

electrical anti Pali� .7unglameniat

millions per second in radio circuits. For con-
venience, two other units, the kilocycle (1000 cycles) and the megacycle (1,000,000 cycles) also are used. The abbreviations for these are kc. and Mc., respectively.
Electrical degrees-- If we take afixed point on the periphery of arevolving wheel, we rind that 9t the end of each revolution, or cycle, the point has come back to its original starting place. Its position at any instant can be expressed in terms of the angle between two lines, one drawn from the center of the wheel to the point at the instant of time considered, the other drawn from the wheel center to the starting point. In making one complete revolution the point has travelled through 360 degrees, a half revolution 180 degrees, a quarter revolution 90 degrees, and so on. The periodic wave of alternating current may be treated simi-
larly, one complete cycle equalling one revolution or 360 degrees, one alternation (half cycle) 180 degrees, and so on. With the cycle
divided up in this way, the sine curve simply means that the value of current at any instant is proportional to the sine of the angle which corresponds to the particular fraction of the cycle considered.
The concept of angle is universally used in alternating currents. Generally, it is expressed in the fundamental form, using the radian rather than the degree as a unit, whence a cycle is equal to 27r radians, or ahalf cycle to 7r radians. The expression 27rf, for which the symbol co is often used, simply means electrical degrees per cycle times frequency, and is called the angular velocity. It gives the total number of electrical radians passed through by acurrent of given frequency in one second.
Waveform, harmonics -- The sine wave is
not only the simplest but in many respects the most desirable waveform. Many other waveforms are met with in practice, however, and they may differ considerably from the simple sine case. It is possible to show by analysis that any such waveform can be resolved into a number of components of differing frequencies and amplitudes, but related in frequency in such a way that all are integer multiples of the lowest frequency present. The lowest frequency is called the fundamental, and the multiple frequencies are called harmonics. Thus awave may consist of fundamental, 3rd, 5th, and 7th harmonics, meaning, if the funda-
mental frequency is say 100 cycles, that frequencies of 300, 500 and 700 cycles also are present in the wave.
Effective, peak and average values -- It
is evident that both the voltage and current are swinging continuously between their positive maximum and negative maximum values,
and it might be wondered how one can speak of so many amperes of alternating current

when the value is changing continuously. The problem is simplified in practical work by considering that an alternating current has an effective value of one ampere when it produces heat at the same average rate as one ampere of
continuous direct current flowing through agiven resistor. This effective value is the square root of the mean of all the instantaneous current values squared. For the sine-wave form,

Eeff = 1/%2E2max
For this reason, the effective value of an alternating current, or voltage, is also known as the root-mean-square or r.m.s. value. Hence, the effective value is the square root of M or 0.707 of the maximum value -- practically considered 70% of the maximum value.-
Another important value, involved where alternating current is rectified to direct current, is the average. This is simply the average of all instantaneous values in the wave, and for a sine wave is equal to 0.636 of the maximum (or peak) value of either current or voltage. The three terms maximum (or peak), effective (or r.m.s.) and average are encountered frequently in radio work. For the sine form they are related to each other as follows:

--= Eat X 1.414 = E.v. X 1.57

Eeff = E. X 0.707 = Bay, X 1.11

X 0.636 = Eeff X 0.9

The relationships for current are the same as those given above for voltage.
Phase -- It has been mentioned that in a circuit containing inductance, the rise of current is delayed by the effect of electrical inertia presented by the inductance (� 2-5). Both increases and decreases of current are similarly

(a) Current and Von'age "nonere'reth Pure fefisAince eircwt Time
(b) Current ?ayfine" Voltage with /1,e lnde/a/Ke en-ccnt
(c)corrent "key Yale rah Pure Ca,oaetance
,rt ercu
Fig. 212 -- Phase relationships between voltage and current in resistive and reactive circuits. The symbol at the left represents agenerator.

CHAPTER TWO 29

5heleach� AntateuA ilatullooZ

delayed. It is also true that a current must flow into a condenser before its elements can
be charged and so provide avoltage difference between its terminals. Because of these facts, we say that acurrent "lags" behind the voltage in acircuit which has apreponderance of inductance and that the current "leads" the
voltage in a circuit where capacity predominates. Fig. 212 shows three possible conditions
in an alternating current circuit. In the first, when the load is a pure resistance, both voltage and current rise to the maximum values simultaneously. In this case the voltage and current are said to be in phase. In the second instance, the existence of inductance in the circuit has caused the current to lag behind the voltage. In the diagram, the current is lagging one quarter cycle behind the voltage. The cur-
rent is therefore said to be 90 degrees out of phase with the voltage. In the third example, with a capacitive load, the voltage is lagging one quarter cycle behind the current. The phase difference is again 90 degrees. These, of course, are theoretical examples in which it is
assumed that the inductance and the condenser have no resistance. Actually, the angle
of lag or lead (phase angle) depends on the amounts of inductance, capacity and resistance in the circuit.
The phase relationships between two currents (or two voltages) of the same frequency are defined in the same way. When two such
currents are combined the resultant is asingle current of the same frequency, but having an instantaneous amplitude equal to the algebraic sum of the amplitudes of the two components at the same instant. The amplitude of the resultant current hence is determined by the
phase relationship between the two currents before combination. Thus if the two currents
are exactly in phase, the maximum value of the resultant will be the numerical sum of the maximum values of the individual currents; if they are 180 degrees out of phase, one reaches its positive maximum at the instant the other reaches its negative maximum, hence the resultant current is the difference between the two. In the latter case, if the two currents have the same amplitude the resultant current is zero.
The a.c. spectrum-- Alternating currents of different frequencies have different properties and are useful in many varieties of ways. For the transmission of power to light lamps, run motors, and perform familiar everyday
tasks by electrical means, low frequencies are most suitable. Frequencies of 25, 50 and 60 cycles are in common use, the latter being most widespread. The range of frequencies between about 30 and 15,000 cycles is known as the audio-frequency range, because when frequencies of this order are converted from a.c.

into air vibrations, as by a loudspeaker or telephone receiver, they are distinguishable as sounds, having a tone pitch proportional to the frequency. Frequencies between 15,000 cycles (15 kilocycles) and about 1,000,000,000 cycles (1000 megacycles) are used for radio communication, because with frequencies of this order it is possible to convert electrical energy into radio waves. The latter frequency is about the highest it is possible to generate at present, but does not necessarily represent the highest frequency that could be used for radio work.
The a.c. spectrum is divided into the following approximate classifications' for convenience in reference:
15-15,000 cycles Audio frequencies 15-100 kilocycles Low radio frequencies 100-1500 kilocycles Medium radio frequencies 1.5-6 megacycles Medium high frequencies
6-30 megacycles High frequencies Above 30 megacycles Ultra-high frequencies
�2-8 OHM'S LAW FOR ALTERNATING CURRENTS
�Resistance--Since current and voltage are always in phase through a resistance, the instantaneous relations are equivalent to those in direct-current circuits, and since by definition the units of current and voltage for a.c. are made equal to those for d.c. in resistive circuits, the various formulas expressing Ohm's Law for d.c. circuits apply without any change for a.c. circuits containing resistance only, or for purely resistive parts of complex a.c. circuits. The formulas are given in �2-6.
Reactance--In an a.c. circuit containing inductance or capacity, the current and voltage are not in phase (� 2-7) so that Ohm's Law cannot be applied directly. The current is not limited by resistance, as in d.c. circuits, but by a quantity called reactance, which expresses the opposing effect of the voltage of self-induction (� 2-5), in the case of an inductance, and the accumulation of charge in the case of acondenser. In circuits containing only reactance no energy is consumed, since the energy put into an inductance or capacity in one part of the cycle is stored in the electromagnetic or electrostatic field and is returned to the circuit in another part of the cycle. Thus in apurely reactive circuit it is possible to have both high voltage and high current without the consumption of any power. Of course in practice there is always some resistance in the wire of an inductance, or heating of the dielectric of acondenser, so that some energy may be lost, but it is usually negligible in well-designed components.
Reactance is expressed in ohms, the same unit as for resistance, since with a given reactance at agiven frequency the current that

30 CHAPTER TWO

electrical anti leach� Dunclamenlat

will flow is proportional to the applied voltage. Hence,
E = --I
for a purely reactive circuit. X is the symbol for reactance.
In circuits containing both resistance and reactance the values of each cannot be added directly because of the different phase relations.
Inductive reactance-- The greater the inductance of acoil, the greater is the effect of self-induction (� 2-5), or the opposition to a change in the value of current, hence the higher the reactance. Also, the higher the frequency the greater the reactance, since the greater the rate of change of current the more opposition the coil offers to the change. Hence, inductive reactance is proportional to inductance and frequency, or
XL = 2rfL
It will be recognized here that angular velocity, 2wf(� 2-7), expresses the rapidity with which the current changes.
The fundamental units (ohms, cycles, henrys) must be used in the above equation, or appropriate factors inserted in case other units are employed. If inductance is in millihenrys, frequency should be in kilocycles; if inductance is in microhenrys, frequency should be in megacycles, to bring the answer in ohms.
Capacitive reactance-- When acondenser is used in an a.c. circuit it is rapidly charged and discharged as the a.c. voltage rises and falls and reverses in polarity. This repetition of charge and discharge constitutes the flow of alternating current through the condenser. Since for agiven voltage the energy stored in the condenser is fixed by its capacity (� 2-3) it is obvious that the total amount of energy stored in the condenser (and subsequently restored to the circuit) in one second will be greater when the condenser is charged many times per second than when it is charged only a few times. Hence the current flow will be proportional to the frequency and to the capacity of the condenser, or conversely the reactance will be inversely proportional to the frequency and the capacity. Therefore
Xe = 1 eirfC
where 2rf again is the angular velocity or the rapidity with which the current changes. When fis in cycles per second and C in farads, Xc will be in ohms. If C is in microfarads, f must be expressed in megacycles to bring the resistance in ohms.
Impedance--In circuits containing inductive reactance the current lags the voltage while with capacity reactance the current leads (� 2-7). Hence the effects of inductive and

capacitive reactance are opposite in sense, or, as it is commonly expressed, inductive and capacitive reactances cancel each other. In series circuits having both inductive and ca-
pacitive reactance the net reactance is the difference between the two, and the current will either lead or lag depending upon which is
larger, capacitive or inductive reactance. Inductive reactance is considered positive and capacitive reactance negative, so that

X = XL -- Xe

The combined effect of resistance and reactance is termed impedance. The symbol for impedance is Z and, for a series circuit, it is computed from the formula:

Z = VR2 X2

where R is the resistance and X is the reactance. The terms 2, R and X are all expressed in ohms. Ohm's Law for alternating current circuits then becomes

I=--EZ � Z =--EI� E = IZ

The phase angle depends upon the relative amounts of resistance and reactance, becoming
more nearly zero (current and voltage in phase) when reactance is small compared to resistance,
and more nearly 90 degrees when resistance is small compared to reactance.
Power factor --The power dissipated in an a.c. circuit containing both resistance and reactance is consumed entirely in the resistance, hence is equal to PR.However, the reactance is also effective in determining the current or voltage in the circuit, even though it consumes no energy. Hence the product of
volts times amperes (which gives the power consumed in d.c. circuits) for the whole circuit may be several times the actual power used up. The ratio of power dissipated (watts) to the volt-ampere product is called the power factor of the circuit, or

Power factor

Watts Volt-amperes

Distributed capacity and inductance -- It should not be thought that the reactance of coils becomes infinitely high as the frequency is increased to ahigh value and, likewise, that the reactance of condensers becomes infinitely low at high frequencies. All coils have some capacity between turns, and the reactance of this capacity can become low enough at some high frequencies to tend to cancel the high reactance of the coil. Likewise, the leads and plates of condensers will have considerable inductance at very high frequencies, which will tend to offset the capacitive reactance of the condenser itself. For these reasons, coils for high-frequency work must be designed to have

31 CHAPTER TWO

Dfl e echo Anctieur'i ilancitooh

low "distributed" capacity, and condensers must be made with short, heavy leads to have low inductance.
Units and instruments --The units used in a.c. circuits may be divided or multiplied to give convenient numerical values to different orders of magnitude, just as in d.c. circuits (� 2-6). Because the rapidly reversing current is accompanied by similar reversals in magnetic field, instruments used for measurement of d.c. (� 2-6) will not operate on a.c. At low frequencies suitable instruments can be constructed by making the current produce both magnetic fields, one by means of a fixed coil and the other by the moving coil. Such instruments are used for measurement of either current or voltage. At radio frequencies this type of instrument is inaccurate because of distributed capacity and other effects, and the only reliable type of direct-reading instrument is the thermocouple ammeter or milliammeter. This is apower-operated device consisting of a resistance wire, heated by the flow of r.f. current through it, to which is attached athermocouple, or pair of wires of dissimilar metals joined together and possessing the property of developing a small d.c. voltage between the terminals when heated. This voltage, which is proportional to the heat applied to the couple, is used to operate a d.c. instrument of ordinary design.
� 2-9 THE TRANSFORMER
Principles --If two coils of wire are wound on a laminated iron core and one of the coils is connected to asource of alternating current, it will be found that there is an alternating voltage across the terminals of the other coil of wire, and an alternating current will flow through a conductor connecting the two terminals. The alternating current in the first coil, or primary, causes a changing magnetic field in the iron core, and this changing magnetic field induces an alternating current in the second coil, or secondary. This is simply an application of the principle of induction (� 2-5) with the induced voltage being caused by a varying magnetic field set up by a current

Pnmary

Secondary
3

flowing in a separate winding instead of the same coil.
Voltage and turn ratio -- For agiven varying magnetic field, the voltage induced in a coil in the field will be proportional to the number of turns on the coil. Since the two coils of a transformer are in the same field, it follows that the induced voltages will be proportional to the number of turns on each coil. In the case of the primary, or coil connected to the source of power, the induced voltage is practically equal to, and opposes, the applied voltage. Hence, the secondary induced voltage is very nearly equal to the voltage applied to the primary, multiplied by the ratio of the number of turns on the secondary to the number of turns on the primary.
Voltage and current relations -- A transformer cannot deliver more power to its secondary load than it takes from the primary source of power, since to do so would be to violate the principle of conservation of energy. Hence we find that transforming agiven voltage to anew value causes an inverse transformation in the current delivered to the load as compared to that taken from the line. For example, a transformer with a secondary-toprimary voltage ratio of 5will have acurrent ratio of X, which means that the primary current will be five times the secondary current. A voltage ratio of less than unity gives a corresponding increase in secondary-to-primary current ratio. Actually these ratios are not exact, since the transformer will have some losses both in the wire and in the iron core, and this additional loss appears as power taken by the primary which is not available for the secondary load. The efficiency, or ratio of power delivered to the load to power taken from the line, of small transformers may vary between 60% and 90%, depending upon the design.
Impedance ratio--In a properly designed iron-core transformer practically all the magnetic lines of force cut both primary and secondary coils, hence the relationship between secondary current and primary current described in the preceding paragraph. The only reactance present is that due to "leakage," or magnetic flux lines which cut one coil but not the other. Since the leakage reactance is small, atransformer having aresistive load on its secondary will also "look like" a practically resistive load to the power line which supplies its primary. The impedance is equal to E/I (� 2-8) and, neglecting losses, if n is the secondary-to-primary turn ratio, then

SYMBOLS
Fig. 213 -- The transformer. Power is transferred
from one coil to the other by means of the magnetic field. The upper symbol at the right indicates an iron. core transformer, the lower one an air-core transformer.

E,
ip =

E, n't

or n2 E,

That is, the impedance (E5/I5)presented by

32 CHAPTER TWO

electrical and Radio .7utulanzentat

the primary to the line (called the reflected im-
pedance or reflected load) is equal to the secondary load impedance (E8!!1)divided by the
square of the secondary-to-primary turns ratio.
The impedance ratio, or ratio of secondary load impedance to impedance presented by the
primary to the line, is therefore equal to the square of the turn ratio. This relation is very
frequently used in radio circuits.
Impedance matching -- Many devices require a specific value of load resistance (or
impedance) for optimum operation. The resistance of the actual load which is to dissipate
the power may differ widely from this value, hence the transformer with its impedancetransforming properties is frequently called upon to change the actual load to the desired
value. This is called impedance matching. From the preceding paragraph,

are equal and, since these two reactances cancel each other, the net reactance becomes zero, leaving only the resistance of the circuit to
impede the flow of current. The frequency at which this occurs is known as the resonant
frequency of the circuit and the circuit is said to
be in resonance at that frequency or tuned to
that frequency.
Constant Vortaqe Variab/e v.- 0 i FreeeenCy
A-SERiES RESONANCE
' Constant Vartaqe Variable Frequency
B-PARALLEL R�ESONANC E

where �N./N,, is the required secondary-toprimary turn ratio, Z, is the impedance of the
actual load, and Z, the impedance required for
optimum operation of the device delivering the power.

iron Core

Fig. 214-- The auto-

transformer. Line and load

Line

currents in the common winding (A) flow in oppo-

site directions so that the

resultant current is the dif-

ference between them.

MICIPEASIMO ekepuEe/CY ��-���4>. Fig. 215 --Characteristics of series-resonant and parallel-resonant circuits.
Series circuits-- The resonant frequency of a simple circuit containing inductance and
capacity in series is given by
-- .27,1-_vLe X US
where
f is the frequency in kilocycles per second
2n- is 6.28
L is the inductance in microhenrys (ph.) C is the capacitance in micro-microfarads

Load
The autotransforrner -- The transformer principle can be utilized with only one winding
instead of two, as shown in Fig. 214; the princi-
ples just discussed apply equally well. The autotransformer has the advantage that the
line and load currents in the common section
are out of phase, hence this portion of the winding carries less current than the remainder of the coil. This advantage is not very marked unless the primary and secondary voltages do not differ very greatly, while it is frequently disadvantageous to have a direct connection between primary and secondary circuits. For these reasons its application is usually limited to boosting or reducing the line voltage for voltage correction or similar purposes.
�2-10 RESONANT CIRCUITS
Principle of resonance -- It has been shown (� 2-8) that the inductive reactance of a coil and the capacitive reactance of a condenser are oppositely affected by frequency. In any combination of inductance and capacitance, therefore, there is one particular frequency for which the inductive and capacitive reactances

The resistance that may be present does not enter into the formula for resonant frequency.
With aconstant-voltage alternating current applied as shown in A of Fig. 215 the current
flowing through such a circuit will be maxi-
mum at the resonant frequency. The magni-
tude of the current will be determined by the
resistance in the circuit. The curves of Fig.
215 illustrate this, curve a being for low resistance and curves band cbeing for greater resistances.
Parallel circuits -- The parallel resonant
circuit is illustrated in B of Fig. 215. This also
contains inductance, capacitance and resist-
ance in series, but the voltage is applied in
parallel with the combination instead of in
series with it as in A. Here we are primarily interested in the characteristics of the circuit as viewed from its terminals, especially in the
parallel impedance it offers. The variation of parallel impedance of a parallel resonant cir-
cuit with frequency is illustrated by the same
curves of Fig. 215 that show the variation in
current with frequency for the series-resonant
circuit. The parallel impedance is maximum at resonance and increases as the series resistance
is made smaller.

33 CHAPTER TWO

Dhe Radio Antaleur'� -ilandioo`

In the case of parallel circuits, resonance may be defined in three ways: the condition which gives maximum impedance, that which gives maximum power factor, or (as in series circuits) when the inductive and capacitive reactances are equal. If the resistance is low, the resonant frequencies obtained on the three bases are practically identical. This condition usually is satisfied in radio work, so that the resonant frequency of aparallel circuit is generally computed by the series-resonance for-
mula given above. Resistance at high frequencies -- At
radio frequencies the resistance of aconductor may be considerably higher than its resistance to direct current or low-frequency a.c. This is because the magnetic field set up inside the
wire tends to force the current to flow in the outer part of the wire, an effect which increases with frequency. At high radio frequencies this skin effect is so pronounced that
practically all the current flows very near the surface of the conductor, thereby in effect
reducing the cross-sectional area and hence increasing the resistance. For this reason low resistance can be achieved only by using conductors with large surface area, but since the inner part of the conductor does not carry current, thin tubing will serve just as well as solid wire of the same diameter.
A similar effect takes place in coils for radio frequencies, where the magnetic fields cause a concentration of current in certain parts of the conductors, again causing an effective decrease in the conductor size and raising the resistance. These effects, plus the effects of stray currents caused by distributed capacity
(� 2-8), raise the effective resistance of a coil to many times the d.c. resistance of the wire.
Sharpness of resonance --The resonance curves become "flatter" for frequencies near resonance frequency, as shown in Fig. 215, as the internal series resistance is increased,
but are of the same shape for all resistances at frequencies farther removed from resonance frequency. The relative sharpness of the resonance curve near resonance frequency is a measure of the sharpness of tuning or selectivity (ability to discriminate between voltages of different frequencies) in such circuits. This is an important consideration in tuned circuits
used for radio work. Flywheel effect; Q -- A resonant circuit
may be compared to aflywheel in its behavior. Just as such a wheel will continue to revolve
after it is no longer driven, so also will oscillations of electrical energy continue in aresonant circuit after the source of power is removed. The flywheel continues to revolve because of its stored mechanical energy; current flow continues in aresonant circuit by virtue of the
energy stored in the magnetic and electric

fields of the condenser and coil. The sharpness of resonance, which is directly related to the flywheel effect, is determined by the ratio of energy stored to energy dissipated, hence is
proportional to the reactance in the circuit and inversely proportional to the resistance. This ratio of stored energy to dissipated energy is called the Q of the circuit.
In resonant circuits at frequencies below about 28 Mc. the resistance is practically wholly in the coil; condenser resistance may
be neglected. Consequently the �of the circuit as awhole is determined by the Q of the coil,
or its ratio of reactance to resistance. Coils for frequencies below the ultra-high frequency region may have Q's ranging from 100 to several hundred, depending upon their size and
construction. Damping, decrement -- The rate at which
current dies down in amplitude in aresonant
circuit site; the source of power has been removed is called the decrement or damping of the circuit. A circuit with high decrement (low Q) is said to be highly damped; one with
low decrement (high Q) is lightly damped. Voltage rise-- When avoltage of the reso-
nant frequency is inserted in series in a resonant circuit, the voltage which appears across either the coil or condenser is considerably higher than the applied voltage. This is because the current in the circuit is limited only by the resistance and hence may have arelatively high value; however, the same current flows through the high reactances of the coil and condenser and consequently causes large voltage drops (� 2-8). As explained above, the reactances, and hence the voltages, are opposite in phase so that the net voltage around the circuit is only that applied. The ratio of the reactive voltage to the applied voltage is -proportional to the ratio of reactance to resistance, which is the Q of the circuit. Hence the voltage across either the coil or condenser is equal to Q times the voltage inserted in series with the circuit.
Parallel-resonant circuit impedance -- The parallel-resonant circuit offers pure resistance (its resonant impedance) between its

Reactance

Resistance

\"
+ FREQUENCY
Fig. 216 -- The impedance of a parallel-resonant circuit separated into its reactance and resistance components. The parallel resistance is equal to the parallel impedance at resonance.

34 CHAPTER TWO

U cirical and Palio .7unfanzentat

terminals at the resonant frequency. If the internal or series resistance of the coil is low so
that the impedance of the inductance branch is
practically the same as its reactance, the cur-
rent through the coil is equal to the applied voltage divided by the reactance (� 2-8). The current through the condenser also is equal to E/X. Since the two reactances are equal at resonance the two currents also are equal, and since they are nearly 180 degrees out of phase
(� 2-7) they cancel each other in the external
circuit, or line. The small current which flows
in the line results from the fact that the resistance in the inductance causes the phase angle to be slightly less than 90 degrees in this branch
so that complete cancellation cannot take place. The impedance (Z = Elnis high because the line current is small, and,is resistive be-
cause the current is practically in phase with
the applied voltage. At frequencies off resonance the current increases through the branch having the lower reactance (and vice versa) so that the circuit becomes reactive, and the resistive component of the impedance decreases
as shown in Fig. 216.
If the circuit Qis about 10 or more the parallel impedance at resonance is given by the formula
Zr -= X2/R =- XQ where X is the reactance of either the coil or the condenser and R is the internal resistance.
Q of loaded circuits-- ln many applica-
-tions, particularly in receiving, the only re-
(A)
(3)
Fig. 217 --The equivalent circuit of a resonant circuit delivering power to aload. The resistor R represents
the load resistance. At B the load is shown tapped across part of L, which by transformer action is equivalent to using ahigher value of load resistance across the whole circuit.
sistance present in the resonant circuit is that of the circuit itself. Hence the coil is designed
to have as high Q as possible. Since, within
limits, increasing the number of turns raises the reactance faster than it raises the resistance, coils for such purposes are made with relatively large inductance for the frequency under consideration.
When the circuit delivers energy to a load,
as in the case of resonant circuits used in transmitters, the energy consumed in the cir-
cuit itself is usually negligible compared with that consumed by the load. The equivalent of
such a circuit can be represented as shown in
Fig. 217-A where the parallel resistor represents the load to which power is delivered.

Since Z = XQ, the Q of acircuit loaded with a resistive impedance Z is (neglecting internal resistance)
Hence for a given parallel impedance, the effective Q of the circuit including the load is proportional to //X, or inversely proportional to the reactance of either the coil or the condenser. A circuit loaded with a relatively low resistance (a few thousand ohms) must therefore have alarge capacity and relatively small inductance to have reasonably high Q.
The effect of a load of given resistance on the Q of the circuit also can be changed by connecting the load across only part of the circuit. The most common method of accomplishing this is by tapping the load across part of the coil, as shown in Fig. 217-B. The smaller the portion of the coil across which the load is
tapped the less the loading on the circuit; in
other words, tapping the load "down" is equivalent to connecting ahigher value of load resistance across the whole circuit. This is similar in principle to impedance transformation with an iron-core transformer (� 2-9).
However, in the high-frequency resonant cir-
cuit the impedance ratio does not vary exactly
as the square of the turn ratio because all the
magnetic flux lines do not cut every turn of the coil. A desired reflected impedance usually must be obtained by experimental adjustment.
L/C ratio--For a given frequency the product of L and C must always be the same, but it is evident that L can be large and C small, L small and C large, etc. The relation between the two for afixed frequency is called the L/C ratio. A high-C circuit is one which has more capacity than "normal" for the frequency; alow-C circuit one which has less than normal capacity. These terms depend to a considerable extent upon the particular application considered, and have no exact numerical meaning.
Piezo-electricity -- Properly-ground crystals of quartz and some other materials show a
mechanical strain when subjected to an elec-
tric charge and, conversely, will show adifference in potential between two faces when subjected to mechanical stress. This characteristic is called the piezo-electric effect. A properlyground quartz crystal is amechanical vibrator electrically equivalent to a series-resonant circuit of very high Q, and can be used for many of the purposes for which ordinary resonant circuits are used.
� 2-11 COUPLED CIRCUITS
Energy transfer; loading --Two circuits are said to be coupled when energy can be transferred from one to the other. The circuit delivering energy is called the primary circuit;

35 CHAPTER TWO

Dhe leach� AntaieuA ilattallooh

that receiving energy is called the secondary are brought closer to each other with their axes

circuit. The energy may be practically all coinciding.

dissipated in the secondary circuit itself, as in

Link coupling -- A variation of inductive

receiver circuits, or the secondary may simply coupling called link coupling is shown in Fig.

act as a medium through which the energy is 219. This gives the effect of inductive coupling

transferred to a load resistance where it does between two coils which may be so separated

work. In the latter case the coupled circuits that they have no mutual inductance; the link

may be considered simply as

a means of providing the

mutual inductance. Because

mutual inductance between

coil and link is involved at

A-inductive

B- Capacitive

C-Resistive

each end of the link, the total

DIRECT COUPLING METHODS

coupling between two link-

coupled circuits cannot be made as great

as when normal inductive coupling is

used, but in practice this is usually not

disadvantageous. Link coupling is fre-

quently convenient in the design of equip-

D-Indrect Capacitive

E-Transformer

ment where inductive coupling would be impracticable because of constructional

Fig. 218 -- Basic types of circuit coupling.

considerations.

The link coils generally have few turns com-

may act as a radio-frequency impedance pared to the resonant-circuit coils, since the

matching device (� 2-9) where the matching coefficient of coupling (see next paragraph) is

may be accomplished by adjusting the loading relatively independent of the number of turns

on the secondary (� 2-10) and by varying the on either coil.

coupling between the primary and secondary.

Coefficient of coupling --The degree of

Coupling by acommon circuit element -- coupling between two coils is a function of

The three variations of this type of coupling their mutual inductance and self-inductances:

(often called direct coupling) are shown at

A, B and C of Fig. 218, utilizing common in-

k = --v a 2

ductance, capacity, and resistance, respec-

tively. Current circulating in one LC branch where kis called the coefficient of coupling. It is flows through the common element (L,,,, C,,,, often expressed as apercentage. The coefficient

or R.) and the voltage developed across this

element causes current to flow in the other

LC branch. The degree of coupling between

the two circuits is greater as the reactance

(or resistance) of the common element is in-

creased in comparison to the remaining re-

actances in the two branches. The circuit at D shows electrostatic coupling
between two resonant circuits. The coupling

Fig. 219 -- Link coupling. Mutual inductance at
both ends of the link is equivalent to nnitual inductance between the two tuned circuits.

increases as the capacity of C. is made greater (reactance of C. is decreased).
Inductive coupling -- Fig. 218-E illus-

of coupling cannot be greater than 1, and generally is much smaller in resonant circuits.

trates inductive coupling, or coupling by

Critical coupling--When there is little

means of the magnetic field. A circuit of this coupling between two circuits tuned to the

type resembles the iron-core transformer same frequency (loose coupling) each behaves

(� 2-9) but because only asmall percentage of much as though the other were not present.

the flux lines set up by one coil cut the turns of As the coupling is increased, each circuit loads

the other coil the simple relationships between the other because of the energy transfer, an

turn ratio, voltage ratio, and impedance ratio effect which is equivalent to increasing the

in the iron-core transformer do not hold. The series resistance in each circuit (or reducing its

interlinking of the lines of force emanating parallel impedance). Hence the sharpness of

from one coil with the turns of the other is resonance, or selectivity, is decreased. At

measured by the mutual inductance between critical coupling, maximum energy is trans-

the two coils. Its value is determined by the ferred from one circuit to the other, and the

sell-inductance of each of the two coils and overall resonance curve shows a single fairly

their position with respect to each other. The broad peak. At still closer coupling (tight

mutual inductance increases as the two coils coupling) the energy transfer will drop off and

36 CHAPTER TWO

electrical and Radio ..7uncianzenIctA

the overall resonance curve will show two peaks' one on either side of the frequency to which the circuits are tuned. The tighter the coupling the greater the frequency separation of the two peaks.
Critical coupling is afunction of the Q's of the two circuits taken independently. A higher coefficient of coupling is required to reach critical coupling when the Q's are low; if the
Q's are high, as in receiving applications, a

coupling (link coupling, for instance) there may
be asmall amount of residual reactance in the
secondary circuit.
preSvheinetldcionupgl--inIgtbiestwfereenquetnwtolycinreccuietsssawrhyictho
for constructional reasons must be physically near each other. Capacitive coupling may readily be prevented by enclosing one or both
of the circuits in grounded low-resistance
metallic containers called shields. The electro-

coupling coefficient of afew percent may give static field from the circuit components does

critical coupling.

not penetrate the shield because the lines of

Effect of circuit Q -- With loaded circuits force are short-circuited (� 2-3). In many cases

it is not impossible for the Q to reach such low a metallic plate, called abaffle shield, inserted

values that critical coupling cannot be obtained even with the highest practicable co-
efficient of coupling (coils as physically close as

between two components may suffice to prevent electrostatic coupling between them, since very little of the field tends to bend

possible). In such case the only way to secure around such a shield if it is large enough to sufficient coupling is to increase the Q of one make the components invisible to each other.

or both of the coupled circuits. This can be

Similar metallic shielding is used at radio

done either by decreasing the LIC ratio or frequencies to prevent magnetic coupling. In

by tapping the load down on the secondary coil (� 2-10). One or the other of these methods
often must be used in link coupling, because

this case the magnetic field induces a current
(eddy current) in the shield which in turn sets
up its own magnetic field opposing the original

the maximum coefficient of coupling between two coils seldom runs higher than 50% or 60%, and the net coefficient is approximately

field (� 2-5). The induced current is proportional to the frequency and also to the conductivity of the shield, hence the shielding effect

equal to the products of the coefficients at each increases with frequency and the conductivity

end of the link. If the load resistance is known and thickness of the shielding material. A

beforehand, the circuits may be designed for a
Qin the vicinity of 10 or so with assurance that sufficient coupling will be available; if unknown, the proper Q's can be detter-

closed shield is required for good magnetic shielding; in some cases separate shields, one about each coil, may be required. The baffle

mined by experiment. Coupled resistance and react-
ance -- If the two circuits are tuned
to the same frequency, their effect on each other is resistive. For example, aloaded and resonant secondary will
cause an apparent increase in the
series resistance of the primary cir-

o--M000

input 0

1-71 01:tput
CT

Input

L- Section

LOW PASS

ILD Q0
TC 0 tput CT

scuiiptate(drepinrestehnetilnogad)thewhiecnherigny

disturn

causes the parallel impedance of the primary to decrease. It is by this

means that the parallel impedance of the primary can, by adjustment of

Output

Input

secondary loading and coupling, be
adjusted to the optimum value for the device furnishing the power
(� 2-9). Should the secondary circuit be

L- Section

HIGH PASS it-Section,

slightly off tune it will have areactive

as well as resistive component, and

the the

prreiacmtaarnycceiricsuliti.keSwiinsceectohuipsliend

into turn

throws the primary off tune, the latter

must be retuned to bring it back to

resonance. The reflected reactance

may be either inductive or capacitive.

This effect occurs frequently in trans-

mitters, where with certain types of

0--'vVvW

Input

CT Output

Input 0--

RESISTANCE --CAPACITY

Fig. 220 -- Simple filter circuits.

Output
CT
1 o

37 CHAPTER TWO

Dne leacho Amaieur'd -llanliooh

shield is rather ineffective for magnetic shielding, although it will give partial shielding if placed at right angles to the axes of, as well as between, the two coils to be shielded from each other.
Cancellation of part of the field of the coil reduces its inductance, and since some energy is dissipated in the shield, the effective resistance of the coil is raised as well, hence the coil Q is reduced. The effect of shielding on coil Q and inductance becomes less as the distance between the coil and shield is increased. The losses also decrease with an increase in the conductivity of the shield material. Copper and aluminum are satisfactory materials. The Q and inductance will not be greatly reduced if the spacing between the sides of the coil and the shield is at least half the coil diameter, and is not less than the coil diameter at the ends of the coil.
At audio frequencies the shielding container is made of magnetic material, preferably of high permeability (� 2-5) to short-circuit the external flux about the coil to be shielded. A non-magnetic shield is quite ineffective at these low frequencies because the induced current is small.
Filters -- By suitable choice of circuit elements, a coupling system may be designed to pass without undue attenuation all frequencies below and reject all frequencies above a certain value called the cut-off frequency. Such acoupling system is called afilter, and in the above case is known as alow-pass filter. If fre'quencies above the cut-off frequency are passed and those below attenuated, the filter is a high-pass filter. Simple filter circuits of both types are shown in Fig. 220. The fundamental circuit from which more complex filters are constructed is the L-section. Fig. 220 also shows 7-section filters, constructed from the basic L-section and frequently encountered in both low-frequency and r.f. circuits. The proportions of L and Cfor proper operation of the filters depend upon the load resistance connected across the output terminals, L being larger and C smaller as the load resistance is increased.
A band-pass filter is one designed to pass without attenuation all frequencies between two selected cut-off frequencies and to attenuate all frequencies outside these limits. The group of frequencies passed through the
filter is called the pass-band. Two resonant circuits with greater than critical coupling represent acommon form of band-pass filter.
The resistance-capacity filter shown in Fig. 220 is used where both d.c. and a.c. are flowing through the circuit and it is desired to provide greater attenuation for the alternating current than the direct current. It is usually employed where the direct current has a low value so

(A)

(C)

(E)

(F)

Fig. 221 -- Bridge circuits utilizing resistance, inductance and capacity, alone and in combination.

that the d.c. voltage drop is not excessive, or when ad.c. voltage drop actually is required. The time constant (� 2-6) of the filter must be large compared to the time of one cycle of the lowest frequency to be attenuated. In determining the time constant, the resistance of the load must be included as well as that in the filter itself.
Bridge circuits -- A bridge circuit is adevice primarily used in making measurements of resistance, reactance or impedance (� 2-8), al-
though it has other applications in radio circuits. The fundamental form of bridge is shown in Fig. 221-A. It consists of four resistances (called arms) connected in seriesparallel to asource of voltage E, with asensitive galvanometer M connected between the junctions of the series-connected pairs. When the equation

= T?:

is satisfied no current flows through M because no potential difference exists between points A and B since the drop across R2 equals that across R4, and the drop across R1 equals that across R3. Under these conditions the bridge is said to be balanced. If R3 is an unknown resistance and R4 is avariable known resistance, R3 can be found from the following equation, after R4 has been adjusted to balance the bridge (null indication on M):

38 CHAPTER TWO

ffJ fa A aciricai and Radio. nilam en

R3 = RR--12 R4

X -- 300,000

RIand R2 are known as the ratio arms of the bridge; the ratio of their resistances is usually adjustable (frequently in steps of 1, 10, 100, etc.) so that a single variable resistor R4 can serve as a standard for measuring widely different values of unknown resistance.
Bridges can be similarly formed with condensers, inductances, and combinations of resistance with either. Typical simple arrangements are shown in Fig. 221. For measurements with alternating currents the bridge must not introduce phase shifts which would destroy the balance, hence similar impedances should be used in each branch, as shown in Fig. 221, and the Q's of the coils and condensers should be the same. When bridges are used at audio frequencies atelephone headset is agood null indicator. The bridges at E and F are commonly-used r.f. neutralizing circuits (� 4-7); the voltage from the source E.c is balanced out at X.
�242 LINEAR CIRCUITS
Standing waves -- If an electrical impulse is started along awire it will travel at approximately the speed of light until it reaches the end. If the end of the wire is open circuited, the impulse will be reflected at this point and travel back again. When a high-frequency alternating voltage is applied to the wire acurrent will flow toward the open end, and reflection will occur continuously. If the wire is long enough so that time comparable to ahalf cycle or more is required for current to travel to the open end, the phase relations between the refleeted current and outgoing current will vary along the wire, and at one point the two currents will be 180 degrees out of phase and at another in phase, with intermediate values between. Assuming negligible losses, this means that the resultant current will vary in amplitude from zero to a maximum value along the wire. Such a variation is called a standing wave. The voltage along the wire also goes through standing waves, but reaches its maximum values where the current is minimum, and vice versa.
Frequency and Wavelength -- It is possible to describe the constants of such line circuits in terms of inductance and capacitance, or inductance and capacitance per unit length, but it is more convenient to give them simply in terms of fundamental resonant frequency or of length. In the case of a straight-wire circuit, length is inversely proportional to lowest resonant frequency. Since the velocity is 300,000 kilometers (186,000 miles) per second, the wavelength is

where Xis the wavelength in meters and fs. is the frequency in kilocycles. The lowest frequency at which the wire or line will be resonant is known as its fundamental frequency
or wavelength. It is common to describe lines (or antennas, which have similar current and voltage distribution) as half-wave, quarter-wave, etc., for a certain frequency ("half-wave
7000-kc. antenna," for instance). Wavelength is also used interchangeably
with frequency in describing not only antennas but also for tuned circuits, complete transmitters, receivers, etc. Thus the terms "highfrequency receiver" and "short-wave receiver," or "75-meter fundamental antenna" and "4000-kilocycle fundamental antenna," are synonymous.
Harmonic resonance -- Although a coilcondenser combination having lumped constants (capacitance and inductance) resonates at only one frequency, circuits such as antennas containing distributed constants resonate readily at frequencies which are very nearly integral multiples of the fundamental frequency. These frequencies are therefore in harmonic relationship to the fundamental frequency and, hence, are referred to as harmonics (� 2-7). In radio practice the fundamental itself is called the first harmonic, the frequency

2 &monk JrdNonnonic

Fundomenlo/ or /it Harmonic

a-eurrent Maxima (anti-node) Itcdel -- Current Nodes
Fg. 222 -- Standing-wave current distribution on a wire operating as an oscillatory circuit at its funda men al, second harmonic and third harmonic frequencies.
twice the fundamental is called the second harmonic, and so on.
Fig. 222 illustrates the distribution of current on a wire for fundamental, second and third harmonic excitation. There is one point of maximum current with fundamental operation, two when operation is at the second harmonic and three at the third harmonic; the number of current maxima corresponds to the order of the harmonic and the number of standing waves on the wire. As noted in the figure, the points of maximum current are called anti-nodes (also known as "loops") and the points of zero current are called nodes.

39 CHAPTER TWO

-7he Radio -Amateur'6 Jiancgooh

Radiation resistance--Since a line circuit
has distributed inductance and capacity, current flow causes storage of energy in mag-
netic and electrostatic fields (� 2-3, 2-5). At low frequencies practically all the energy so stored is returned to the wire during another part of the cycle (� 2-8) but above 15,000 cycles or so (radio frequency) some escapes --
is radiated -- in the form of electromagnetic waves. Energy radiated by aline or antenna is
equivalent to energy dissipated as in aresistor. The value of this equivalent resistance is known as radiation resistance.
Resonant line circuits--The effective resistance of aresonant straight wire such as an antenna is considerable, because of the power radiated. The resonance curve of such a straight-line circuit is quite broad; in other
words, its Q is relatively low. However, by folding the line, as suggested by Fig. 223, the
fields about the adjacent sections largely cancel each other and very little radiation takes place. The radiation resistance is greatly reduced and the line-type circuit can be made to have avery sharp resonance curve or high Q.

Siancane Ware

-- _ _

of art-rent

Low { Impedance

A/4

KO Impedance

ImpeKdoahnce I

X/4.
IC,' raute3

anductors Short-ci

Fig. 224 -- A concentric line resonant circuit.

the low radiation resistance and relatively large conducting surfaces, such lines can be made to have much higher Q's than are attain-
able with coils and condensers. They are most applicable at ultra-high frequencies (very short wavelengths) (� 2-7) where dimensions are small.

� 243 CIRCUITS WITH SUPERIMPOSED CURRENTS
Combined a.c. and d.c.-- There are many practical instances of simultaneous flow of alternating and direct current in a circuit. When this occurs there is a pulsating current and it is said that an alternating current is superimposed on adirect current. As shown in Fig. 225, the maximum value is equal to the d.c. value plus the a.c. maximum, while the minimum value (on the negative a.c. peak) is the difference between the d.c. and the maximum a.c. values. The average value (� 2-7) of the current is simply equal to the direct-current component alone. The effective value (� 2-7) of the combination is equal to the square root of the sum of the effective a.c. squared and the d.c. squared:

�
Fig. 223-- Standing wave and instantaneous current (arrows) conditions of afolded resonant-line circuit.
A circuit of this type will have a standing wave on it, as shown by the dash-line of Fig. 223, with the instantaneous current flow in each wire opposite in direction to the flow in the other, as indicated by the arrows on the diagram. This opposite current flow accounts for the cancellation of radiation, since the fields about the two wires oppose each other. Furthermore, the impedance across the open ends of the line will be very high, thousands of ohms, while the impedance across the line near the closed end will be very low.
A folded line may be made in the form of two concentric conductors, as shown in Fig. 224. The concentric line has even lower radiation resistance than the folded wire line, since the outer conductor acts as a shield. Standing waves exist, but are confined to the outside of the inner conductor and the inside of the outer conductor, since skin effect prevents the currents from penetrating to the other sides. Thus such aline will have no radio-frequency potentials on its exposed surfaces. Because of

I �VIJ where I is the effective value of the a.c. component, Iis the effective value of the combination and 'dc is the average (d.c.) value of the combination.
Beats--If two or more alternating currents of different frequencies are present in a normal circuit, they have no particular effect upon one another and, for this reason, can be separated again at any time by the proper selective circuits. However, if two (or more) alternating currents of different frequencies are present in an element having unilateral or oneway current flow properties, not only will the two original frequencies be present in the output but also currents having frequencies equal to the sum, and difference, of the original fre-
iat
0 r/ME ��111..�
Fig. 225 -- Pulsating current composed of alternating current superimposed on direct current.

40 CHAPTER TWO

electric-a/an' Palio .7unclamenlat

quencies. These sum and difference frequencies are called the beat frequencies. For example, if frequencies of 2000 and 3000 kc. are present in a normal circuit, only those two frequencies
exist, but if they are passed through a unilateral-element (such as a properly-adjusted vacuum tube) there will be present in the output not only the two original frequencies of 2000 and 3000 kc. but also currents of 1000 (3000 -- 2000) and 5000 (3000 � 2000) kc. Suitable circuits can select the desired beat frequency.
By-passing -- In combined circuits it is frequently necessary to provide alow-impedance path for a.c. around, for instance, a source of d.c. voltage. This can be done by using a bypass condenser, which will not pass direct current but will readily permit the flow of alter-
nating current. The capacity of the condenser should be of such value that its reactance is low (of the order of 1/10th or less) compared to the a.c. impedance of the device being bypassed. The lower the reactance, the better is the a.c. confined to the desired path.
Similarly, alternating current can be prevented from flowing through a direct-current circuit to which it may be connected by inserting an inductance of high reactance (called

achoke coil) between the two circuits. This will permit the d.c. to flow without hindrance, since the resistance of the choke coil may be made
quite low, but will effectively prevent the a.c.
from flowing where it is not wanted. If both r.f. and low-frequency (audio or
power frequencies) currents are present in a
circuit, they may be confined to desired paths by similar means, since an inductance of high reactance for radio frequency will have negligible reactance at low frequencies, while a condenser of low reactance at radio frequencies will have high reactance at low frequencies.
Grounds-- The term "ground" is frequently met in discussions of circuits, and
normally means the voltage reference point in
the circuit. There may or may not be an actual connection to earth, but it is understood that apoint in the circuit said to be at ground
potential could be connected to earth without disturbing the operation of the circuit in any way. In direct-current circuits the negative side is generally grounded. The ground symbol in circuit diagrams is used for convenience in
indicating common connections between various parts of the circuit, and with respect to actual ground usually has the meaning indicated above.

41 CHAPTER TWO

CHAPTER THREE
Vacuum ..7ute�

�3-1 DIODES

Space charge -- With the cathode tem-

Rectification --Practically all of the vac- perature fixed, the total number of electrons uum tubes used in radio work depend upon emitted is always the same regardless of the

thermionic conduction (� 2-4) for their operation. The simplest type of vacuum tube is that shown in Fig. 301. It has two elements, cathode and plate, and is called a diode. The

plate voltage. Fig. 301 shows, however, that less plate current will flow at low plate voltages than when the plate voltage is large. With low plate voltage only those electrons nearest the

cathode is heated by the "A" battery and plate are attracted to the plate. The electrons

emits electrons which flow to the plate when the plate is at apositive potential with respect to the cathode. Because of the nature of thermionic conduction, the tube is a conductor in one direction only. If a source of

in the space near the cathode, being themselves negatively charged, tend to repel the similarly-charged electrons leaving the cathode surface and cause them to fall back on the cathode. This is called the space charge effect.

alternating voltage is connected between the As the plate voltage is raised, more and more cathode and plate, then electrons will flow electrons are attracted to the plate until

only on the positive half-cycles of alternating voltage; there will be no electron flow during the half cycle when the plate is negative. Thus the tube can be used as a rectifier, to change alternating current to pulsating direct current. This alternating current can be anything from

finally the space charge effect is completely overcome and all the electrons emitted by the cathode are attracted to the plate, and a further increase in plate voltage can cause no increase in plate current. This is called the saturation point.

the 60-cycle kind to the highest radio frequencies.
Characteristic curves -- The performance of the tube can be reduced to easily-understood terms by making use of tube characteristic
curves. A typical characteristic curve for a diode is shown at the right in Fig. 301. It shows the current flowing between plate and cathode with different d.c. voltages applied between the elements. The curve of Fig. 301 shows that, with fixed cathode temperature, the plate current increases as the voltage
between cathode and plate is raised. For an actual tube the values of plate current and
plate voltage would be plotted along their
respective axes. The power consumed in the tube is the
product of the plate voltage multiplied by the plate current, just as in any d.c. circuit. In a vacuum tube this power is dissipated in heat

�3-2 TRIODES
Grid control --If a third element, called the control grid, or simply the grid, is inserted between the cathode and plate of the diode the space-charge effect can be controlled. The tube then becomes a triode (three-element tube) and is useful for more things than rectification. The grid is usually in the form of an open spiral or mesh of fine wire. With the grid connected externally to the cathode and with asteady voltage from a d.c.- supply applied between the cathode and plate (the positive of the "B" supply is always connected to the plate), there will be a constant flow of electrons from cathode to plate, through the openings of the grid, much as in the diode. But if asource of variable voltage is connected between the grid and cathode

developed in the plate and radiated to the bulb. there will be avariation in the flow of electrons from cathode to plate (a variation in plate

current) as the voltage on the grid changes

PL ATE CURRENT

(Sat.-at/kw Point

about a mean value. When the grid is made

less negative (more positive) with respect to

the cathode, the space charge is partially

neutralized and there will be an increase in

plate current; when the grid is made more

Increase

negative with respect to the cathode, the space

PLATE VOLTAGE

charge is reinforced and there will be a de-

Fig. 301 -- The diode or two-element tube and atypical characteristic curve.

crease in plate current. Amplification -- The grid thus acts as

CHAPTER THREE

Vacuum 5uie�

valve to control the flow of plate current, and
it is found that it has a much greater effect on plate current flow than does the plate voltage; that is, asmall change in grid voltage
is just as effective as a large change in plate voltage in bringing about a change in plate
current. When a resistance or impedance (load) is connected in series in the plate circuit,
the voltage drop across it, which is afunction of the plate current through it, can therefore be changed by varying the grid voltage as well as by giving the plate voltage a new value. Thus asmall change in grid voltage will cause a large change in voltage drop across the impedance; in other words, the grid voltage is amplified in the plate circuit.
So long as the grid has anegative potential with respect to the cathode, electrons emitted by the cathode are repelled (� 2-3) with the result that no current flows to the grid. Hence under these conditions the grid consumes no power. However, when the grid becomes positive with respect to the cathode, electrons are attracted to it and acurrent flows to the grid; when this grid current flows power is dissipated in the grid circuit.
Characteristics -- The measure of the amplification of which atube is capable is known as its amplification factor, designated by g (MO. Mu is the ratio of plate-voltage change required for agiven change in plate current to the grid-voltage change necessary to produce the same change in plate current. Another important characteristic is the plate resistance, designated r,,. It is the ratio, for a fixed grid voltage, of asmall plate voltage change to the
plate current change it effects. It is expressed in ohms. Still another important characteristic used in describing the properties of a tube is grid-plate transconductance, or mutual conductance, designated by g,,, and defined as the rate of change of plate current with respect
to a change in grid voltage. The mutual conductance is a rough indication of the design merit of the tube. It is expressed in micromhos (the mho is the unit of conductance and is equal to 1/R) and is the ratio of amplification factor to plate resistance, multiplied by one million. These tube characteristics are inter-
related and are dependent primarily on the tube structure.
Static and dynamic curves -- The operation of avacuum tube amplifier is graphically represented in elementary form in Fig. 302. The sloping line represents the variation in plate current obtained at a constant plate voltage with grid voltages ranging from a value sufficiently negative to reduce the plate current to zero, to avalue slightly positive. Grid voltage is specified with reference to the cathode or filament. Notable facts about this curve are that it is essentially astraight line (is linear)

Upper bend

1�:�" Operatio9 point

'P/ate Current ;/- Swiz!z"

elett
t11
LOWeri
bend

44--
I
I r

Ak. five
ei-itrTeas
r I

-- 0 + GRID BIAS, Ec Voitafe,e9
,,r-GridValtafeSwinr

Fig. 302 -- Operating characteristics of a vacuumtube amplifier. Class-A amplifier operation is depicted.
over the middle section and that it bends towards the bottom (near cut-off) and near the top (saturation). In other words, the variation in plate current is directly proportional to the variation in grid voltage over the region between the two bends. With afixed grid voltage (bias) of proper value the plate current can be set at any desired value.
Tube characteristics of the type shown in Fig. 302 may be of either the static or dynamic type. Static characteristics show the plate current that will flow at specific grid and plate voltages in the absence of any output device in the plate circuit for transferring the plate current variation to an external circuit, while the dynamic characteristic shows the behavior of the same quantities when there is aload in the plate circuit, and thus represents the actual operation of the tube as an amplifier.
Interelectrade capacities -- Any pair of elements in atube forms aminiature condenser (� 2-3), and although the capacities of these condensers may be only a few micrornicrofarads or less, they must frequently be taken into account in vacuum-tube circuits. The capacity from grid to plate (grid-plate capacity) has an important effect in many applications. In triodes, the other capacities are the gridcathode and plate-cathode. In multi-element tubes (� 3-5) similar capacities exist between these and other electrodes. With screen-grid tubes, the terms "input" and "output" capacity mean, respectively, the capacity measured from grid to all other elements connected together, and from plate to all other elements connected together. The same terms are used with triodes but are not so easily defined since the effective capacities existing depend upon the operating conditions (� 3-3).
Tube ratings -- Specifications of suitable operating voltages and currents are called tube ratings. Ratings include proper values for filament or heater voltage and current, plate voltage and current, and similar operating

CHAPTER THREE 43

Dhe Radio Arrectieur'i --Ilan 'Loh

specifications for other elements. An important rating in power tubes is the maximum safe plate dissipation, which is the maximum power which can be dissipated continuously in heat on the plate (� 3-1).
� 3-3 ANIPLIFICATION
Circuits--Besides the vacuum tube, a complete amplifier includes ameans for introducing the signal or exciting voltage into the grid circuit, a means (the load) for taking power or amplified voltage from the plate circuit, and sources of supply for bias voltage, power for heating the cathode, and d.c. power for the plate circuit. A representative circuit for audio-frequency amplification is shown in Fig. 303. The signal is introduced into the grid circuit in series with the bias voltage by means of transformer T1,and T2 serves as ameans for transferring the amplified signal from the plate circuit to the load. Battery supplies are indicated for simplicity.

Input

Grid Bias

PlateorAnode
Grid Filament or Cathode

:4" Battery (Filament)
11111.1 11-1 �B"Battery (plate)
Fig. .30.3 -- A typical audio-frequency amplifier using
a triode tube.
A single amplifier such as is shown in Fig. 303 is called an amplifier stage, and several such stages may be used in cascade, the output of one stage being fed to the grid circuit of the next, to provide large amounts of overall amplification.
Load impedance-- The load connected in the output or plate circuit of the tube is called the load impedance or load resistance, and designated Ra.It is the device in which the power output of the tube is consumed. In some types of amplifiers the load is an actual resistance, but in most cases it is a resistive impedance; that is, it shows resistance for a.c. but not for d.c. This type of load can be obtained with aresonant circuit (� 2-10) or by coupling through atransformer to apowerconsuming device (� 2-9). The impedance load has the advantage that there is no drop in d.c. plate voltage across the load as there would be in .the case of the resistor, since the latter has the same resistance for d.c. as for a.c.
In general, there will be one value of R,, which will give optimum results for a given type of tube and set of operating voltages;

its value also depends upon the type of service for which the amplifier is designed. If the impedance of the actual device used is considerably different from the optimum load impedance, the tube and output device can be coupled through atransformer having aturns ratio such that the impedance reflected into
the plate circuit of the tube is the optimum value (� 2-9).
Operating point and grid bias-- As indicated in Fig. 302, the relationship between varying grid voltage and plate current will be determined by the grid bias (� 3-2), which sets the operating point on the characteristic curve. The choice of operating point depends upon
the type of service in which the tube is to be used.
Distortion--With negative grid bias as shown in Fig. 302 the operating point comes in the middle of the linear region. If an alternating voltage (signal) is now applied to the
grid in series with the grid bias, the grid voltage swings more and less negative about the mean
bias voltage value and the plate current swings up (positive) and down (negative) about the mean plate current value. This is equivalent to
an alternating current superimposed on the steady plate current. At this operating point it is evident that the plate current wave shapes (� 2-7) are identical reproductions of the grid voltage wave shapes and will remain so as long as the grid voltage amplitude does not reach values sufficient to run into the lower- or upper-bend regions of the curve. If this occurs the output waves will be flattened or distorted. If the operating point is set towards the bottom or the top of the curve there will also be distortion of the output wave shapes because part or all of the lower or upper half-cycles will be cut off.
Whenever the bias is adjusted so that the tube works over a non-linear portion of its characteristic curve, distortion will take place and the output wave-form will not duplicate the wave-form of the voltage introduced at the
grid. This characteristic of non-linearity of an amplifier is useful in some applications but is an undesirable feature in others. The distortion will take the form of harmonics added to
the original wave (� 2-7). If the exciting signal is asine wave, the output wave, when distortion is present, will consist of the fundamental plus harmonics.
Another type of distortion, known as frequency distortion, occurs when the amplification varies with the frequency of the a.c. voltage applied to the grid circuit of the amplifier. It is not necessarily accompanied by harmonic distortion. It can be shown by afrequency response curve, or graph in which the relative
amplification is plotted against frequency over the frequency range of interest.

44 CHAPTER THREE

Z Cu urn

1te.3

Voltage amplification -- The ratio of the alternating output voltage derived from the
plate circuit to the alternating voltage applied to the grid circuit is called the voltage amplification of the amplifier. A voltage amplifier is one in which this ratio is the primary con-
sideration, rather than the power which may be taken from the output circuit. The load resistance for voltage amplification must be high to give alarge voltage across its terminals.
Power amplification -- The ratio of output power to a.c. power consumed in the grid circuit (driving power) is called the power amplification of the amplifier. A power amplifier is one designed to deliver power to a load circuit, the voltage amplification being incidental. The power amplification ratio may be practically infinite in certain types of amplifiers. The load impedance for power amplification is selected either to give maximum power with minimum distortion or to give a desired value of plate efficiency.
Plate efficiency -- The ratio of output
power to d.c. input power to the plate (plate current multiplied by plate voltage) is called the plate efficiency of the amplifier. Plate efficiency is generally low in amplifiers designed primarily for minimum distortion, but may be made quite high when distortion is per-
missible. Power sensitivity-- This is the ratio of
output power to alternating grid voltage, and is ordinarily used in connection with amplifiers operating in such a way that no power is consumed in the grid circuit. The same term also is used frequently in connection with radiofrequency power amplifiers, but in this case has the same meaning as power amplification ratio,
defined above. Phase relations in plate and grid circuits
-- When the exciting voltage on the grid has its maximum positive instantaneous value the plate current also is maximum (� 3-2) so
that the voltage drop across the impedance connected in the plate circuit likewise has
its greatest value. The actual instantaneous voltage between plate and cathode is therefore minimum at the same instant, because it is
equal to the d.c. supply voltage (which is unvarying) minus the voltage drop across the load impedance. Since the decrease in instantaneous plate voltage is negative in sense, this means that the alternating plate voltage is 180 degrees out of phase with the alternating grid voltage. Thus there is a phase reversal through an amplifier tube.
Input capacity-- When an alternating voltage is applied between grid and cathode of the amplifier tube an alternating current flows in the small condenser formed by these elements (� 3-2), just as it would in any other condenser (� 2-8). Similarly, an alternating current also

flows in the condenser formed by the grid and plate, but since the instantaneous voltage between these two elements is considerably larger than the signal voltage when the tube is amplifying, the current in the grid-plate capacity is likewise larger than it would be were no amplification taking place. Looked at from the grid circuit, the increased current is equivalent to an increase in input capacity of the tube, and the effective input capacity may be many times that which would be expected from consideration of the interelectrode capacities alone. The effective input capacity is proportional to the actual grid-plate capacity and to the voltage amplification.
Feedback -- Some of the amplified energy in the plate circuit can be coupled back into the grid circuit to be re-amplified, this process being called feedback. If the voltage induced in the grid circuit is in phase with the grid signal voltage, the feedback is called positive, and the resultant voltage is larger and hence the amplification is increased. Positive feedback, usually called regeneration, can effectively increase the amplification of a stage many times. If the fed-back voltage is opposite in phase to the exciting voltage, the feedback is called negative and, since the resultant grid voltage is smaller, the amplification is decreased. Negative feedback is sometimes called degeneration.
Positive feedback is accompanied by a tendency to give maximum amplification at only one frequency, even though the input and output circuits may not otherwise be resonant. It therefore increases the selectivity of the amplifier and hence is used chiefly where high gain and sharpness of resonance are both wanted.
PARALLEL
PUSH-PULL Fig. 304--Parallel and push-pull amplifier cornicelions.

CHAPTER THREE 45

Dhe

-Amaleuc:s -.11untILon

Negative feedback has the opposite characteristics. It tends to widen the frequency range of the amplifier, even with resonant input and output circuits. It also reduces distortion and makes the amplifier tube more tolerant of changes in load impedance. Hence it is used where low-distortion, wide frequency range amplification is wanted, as in some audio circuits, even though amplification is sacrificed.
Parallel operatien-- When it is necessary to obtain more power output than one tube is capable of giving, without going to a larger tube structure, two or more tubes may be connected in parallel, in which case the similar elements in all tubes are connected together as shown in Fig. 304. The power output will then be in proportion to the number of tubes used; the exciting voltage required, however, is the same as for one tube.
If the amplifier operates in such a way as to consume power in the grid circuit, the grid power required also is in proportion to the number of tubes used.
Push-pull operation-- An increase in power output can be secured by connecting two tubes in push-pull, the grids and plates of the two tubes being connected to opposite ends of the circuit as shown in Fig. 304. A "balanced" circuit, in which the cathode returns are made to the midpoint of the input and output devices, is necessary with pushpull operation. At any instant the ends of the secondary winding of the input transformer, 7'1,will be at opposite potentials with respect to the cathode connection, so the grid of one tube is swung positive at the same instant that the grid of the other is swung negative. Hence, in any push-pull-connected stage the voltages and currents of one tube are out of phase with those of the other tube. The plate current of one tube therefore is rising while the plate current of the other is falling, hence the name "push-pull." In push-pull operation the even-harmonic (second, fourth, etc.) distortion is cancelled in the symmetrical plate circuit, so that for the same output the distortion will be less than with parallel operation.
The exciting voltage measured between the two grids must be twice that required for one tube. If the grids consume power, the driving power for the stage is twice that taken by either tube alone.
The decibel-- The ratio of the power levels at two points in acircuit such as an amplifier can be expressed in terms of aunit called the decibel, abbreviated db. The number of decibels is 10 times the logarithm of the power ratio, or
db. = 10 log Pi
46 c II A l' "I ER Tu R EE

The decibel is a particularly useful unit because it is logarithmic, and thus corresponds to the response of the human ear to sounds of varying loudness. One decibel is approximately the power ratio required to make a just noticeable difference in sound intensity. Within wide limits, changing the power by a given ratio produces the same apparent change in loudness regardless of the power level; thus if the power is doubled the increase is 3 db., or three steps of intensity; if it is doubled again, the increase is again 3 db., or three further distinguishable steps. Successive amplifications expressed in decibels can be added to obtain the overall antplification.
A power loss also can be expressed in decibels. A decrease in power is indicated by a minus sign (e.g., -- 7 db.), and increase in power by aplus sign (e.g., � 4db.). Negative and positive quantities can be added numerically. Zero db. indicates the reference power level, or apower ratio of 1.
Applications of amplification-- The major uses of vacuum tube amplifiers in radio work are to amplify at audio and radio frequencies (� 2-7). The audio-frequency amplifier is generally used to amplify without discrimination at all frequencies in a wide range (say from 100 to 3000 cycles for voice communication), and is therefore associated with non-resonant or untuned circuits which offer a uniform load over the desired range. The radio-frequency amplifier, on the other hand, is generally used to amplify selectively at a single radio frequency, or over a small band of frequencies at most, and is therefore associated with resonant circuits tunable to the desired frequency.
An audio-frequency amplifier may be considered a broad-band amplifier; most radiofrequency amplifiers are relatively narrowband affairs.
In audio circuits, the power tube or output tube in the last stage usually is designed to deliver aconsiderable amount of audio power, while requiring but negligible power from the input or exciting signal. To get the alternating voltage (grid swing) required for the grid of such a tube voltage amplifiers are used, employing tubes of high g which will greatly increase the voltage amplitude of the signal. Voltage amplifiers are used in the radiofrequency stages of receivers as well as in audio amplifiers; power amplifiers are used in r.f. stages of transmitters.
�3-4 CLASSES OF AMPLIFIERS
Reason for classification--It is convenient to divide amplifiers into groups according to the work they are intended to perform, as related to the operating conditions necessary to accomplish the purpose. This makes identi-

Vacuum DuleJ

fication easy and obviates the necessity for giving a detailed description of the operation when specific operating data are not required.
Class A -- An amplifier operated as shown in Fig. 302 in which the output wave shape is a faithful reproduction of the input wave shape, is known as aClass-A amplifier.
As generally used, the grid of a Class-A amplifier never is driven positive with respect to the cathode by the exciting signal, and never is driven so far negative that plate-current cut-off is reached. The plate current is constant both with and without grid excitation.
The chief characteristics of the Class-A amplifier are low distortion, low power output for agiven size of tube, and ahigh power-amplification ratio. The plate efficiency (� 3-3) is relatively low, being in the vicinity of 20 to 35 percent at full output, depending upon the design of the tube and the operating conditions.
Class-A amplifiers of the power type find application as output amplifiers in audio systems. Class-A voltage amplifiers are found in the stages preceding the power stage in such applications, and as radio-frequency amplifiers
in receivers. Class B-- The Class-B amplifier is primar-
ily one in which the output current, or alternating component of the plate current, is proportional to the amplitude of the exciting grid voltage. Since power is proportional to the square of the current, the power output of a Class-B amplifier is proportional to the square of the exciting grid voltage.
The distinguishing operating condition in Class-B service is that the grid bias is set so that the plate current is relatively low without excitation; the exciting signal amplitude is such that the entire linear portion of the tube's characteristic is used. Fig. 305 illustrates Class-B operation with the tube biased prac-
tically to cut-off. In this operating condition plate current flows only during the positive half-cycle of excitation voltage. No plate current flows during the negative swing of the
excitation voltage. The shape of the plate current pulse is essentially the same as that of the positive swing of the signal voltage. Since
the plate current is driven up toward the saturation point, it is usually necessary for the grid to be driven positive with respect to the cathode during part of the grid swing. Grid current flows, therefore, and the driving source
must furnish power to supply the grid losses. Class-B amplifiers are characterized by
medium power output, medium plate efficiency (50% to 60% at maximum signal) and a moderate ratio of power amplification. At radio frequencies they are used as linear amplifiers to raise the output power level in radiotelephone transmitters after modulation
has taken place.

Fig. 305 -- Operation of the Clasa.B amplifier.
For audio-frequency amplification two tubes must be used. The second tube, working alternately with the first, must be included so that both halves of the cycle will be present in the output. A typical method of arranging the
tubes and circuit to this end is shown in Fig. 306. The signal is fed to a transformer Tb whose secondary is divided into two equal parts, with the tube grids connected to the outer terminals and the grid bias fed in at the center. A transformer T2 with a similarlydivided primary is connected to the plates of the tubes. When the signal swing in the upper half of T1 is positive, Tube No. 1draws plate current while Tube No. 2 is idle; when the lower half of T1 becomes positive, Tube No. 2 draws plate current while Tube No. 1is idle. The corresponding voltages induced in the halves of the primary of T2 combine in the secondary to produce an amplified reproduc-
tion of the signal wave-shape with negligible distortion. The Class-B amplifier is capable of delivering much more power for agiven tube size than a Class-A amplifier.
Class AB -- The similarity between Fig. 306 and the ordinary push-pull amplifier circuit (� 3-3) will be noted. Actually the circuits are the same, the difference being in the method of operation. If the bias is adjusted so that the tubes draw a moderate value of plate current the amplifier will operate Class A at low signal voltages and more nearly Class B at high signal voltages. An amplifier so operated is called Class AB. The advantages of this method are
low distortion at moderate signal levels and higher efficiency at high levels, so that relatively small tubes can be used to good advantage in audio power amplifiers.
A further distinction can be made between amplifiers which draw grid current and those which do not. The Class-A B1 amplifier draws
no grid current and thus consumes no power from the driving source; the Class-A B2 amplifier draws grid current at higher signal levels and power must therefore be supplied to its grid circuit.
Class C-- The Class-C amplifier is one op-
erated so that the alternating component of

CHAPTER THREE 47

Mtclio �nctieuA .._llandiooh

largely eliminated by the flywheel effect of the tuned output circuit.
Although requiring considerable driving power because of the relatively large grid swing and grid-current flow, the high plate efficiency (ordinarily 70-75%) of the Class-C amplifier makes it an effective generator of radio-frequency power.

Excitino ,,Venal'

Pkile C.7rreat Tube I

Output Wave Shape

�35 MULTIELEMENT AND SPECIAL.. PURPOSE TUBES
Radio-frequency amplification -- In a ra-

dio-frequency amplifier the input (grid) and

output (plate) circuits must be tuned to the

same frequency for maximum amplification and

Plate Current rube 2

selectivity. If atriode tube is used in such an arrangement the feedback through the gridplate capacity will sustain oscillation at radio

Fig. 306 -- The Class-B audio amplifier, showing how the outputs of the two tubes are combined to give distortionless amplification.

frequencies (� 3-7) so that the circuit becomes an oscillator rather than an amplifier. Although special circuits can be used to overcome oscilla-

the plate current is directly proportional to the plate voltage. The output power is therefore proportional to the square of the plate voltage. Other characteristics inherent to Class-C operation are high plate efficiency, high power out-
put, and a relatively low power-amplification ratio.
The grid bias for a Class-C amplifier is ordinarily set at approximately twice the value required for plate current cut-off without grid excitation. As aresult, plate current flows during only a fraction of the positive excitation cycle. The exciting signal should be of sufficient amplitude to drive the plate current to the saturation point, as shown in Fig. 307. Since the grid must be driven far into the positive region to cause saturation, considerable numbers of electrons are attracted to the grid at the peak of the cycle, robbing the plate of some that it would normally attract. This causes the droop at the upper bend of the characteristic, and also causes the plate current pulse to be indented at the top, as shown. Al-
though the output wave-form is badly distorted, at radio frequencies the distortion is

tion, it is preferable to use a tube in which such feedback is negligible. Such a tube can be made by inserting a second grid to act as an electrostatic shield between the control grid and plate and thus reduce the grid-plate capacity to anegligible value. The addition of the extra grid, called the screen grid or screen, makes the tube atetrode, or four-element tube.
The tetrode -- The screen grid increases the amplification factor and plate resistance of the tube to values much higher than are attainable in triodes of practicable construction, although the mutual conductance is about the same as that of an equivalent triode. The screen grid is ordinarily operated at a lower positive potential than the plate, and is bypassed back to the cathode so that it has essentially the same a.c. potential as the cathode.
Another type of tetrode, in which the electrostatic shielding provided by the second grid is purely incidental, is built for audio power output work. The second grid accelerates the flow of electrons from cathode to plate, and the structure has a higher power sensitivity (� 3-3) than is possible with triodes.

Secondary emission -- When an electron

travelling at appreciable velocity through a

tube strikes the plate it dislodges other elec-

trons which "splash" from the plate into the

interelement space. This phenomenon is called

secondary emission. In the triode, ordinarily

operated with the grid negative with respect

to cathode, these secondary electrons are re-

pelled back into the plate and cause no dis-

turbance. In the screen-grid tube, however,

(A:SInal

the positively charged screen grid attracts the

secondary electrons, causing areverse current

to flow between screen and plate. The effect is

Fig. 307 -- Class-C amplifier operation.

particularly marked when the plate and screen potentials are nearly equal, which may be the

CHArrER THREE

Vacuum ..7u`e�

case during part of the a.c. cycle when the instantaneous plate current is large and the plate voltage low (� 3-3).
The pentode-- To overcome the effects of secondary emission a third grid, called the suppressor grid, can be inserted between the screen and plate. This grid is connected directly to the cathode and repels the relatively low-velocity secondary electrons back to the plate without obstructing to any appreciable extent the regular plate-current flow. Larger undistorted outputs therefore
Sup G

PENTODE

TETRODE

Fig. 308 -- Symbols for pentode and tetrode tubes. H, heater; C, cathode; G, control grid; P, plate; S, screen grid; Sup., suppressor grid.

can be secured from the pentode, or five-element tube.
Pentode-type screen-grid tubes are used as radio-frequency voltage amplifiers, and in addition can be used as audio-frequency voltage amplifiers to give high voltage gain per stage. Pentode tubes also are suitable as audiofrequency power amplifiers, having greater plate efficiency and power sensitivity than triodes.
Beam tubes-- A "beam" type tube is a tetrode incorporating astructure which forms the electrons travelling to the plate into concentrated paths, resulting in higher plate efficiency and power sensitivity. Suitable design also overcomes the effects of secondary emission without the necessity for a suppressor grid. Tubes constructed on the beam principle are used in receivers as both r.f. and audio amplifiers, and are built in larger sizes for transmitting circuits.
Variable-mu and sharp cut-off tubes -- Receiving screen-grid tetrodes and pentodes for radio-frequency voltage amplification are made in two types, known as sharp cut-off and variable-2 or "super-control" types. In the sharp cut-off type the amplification factor is practically constant regardless of grid bias, while in the variable-g type the amplification factor decreases as the negative bias is increased. The purpose of this design is to permit the tube to handle large signal voltages without distortion in circuits in which grid-bias control is used to vary the amplification.
Multipurpose types-- A number of combination types of tubes have been constructed to perform multiple functions, particularly in

receiver circuits. Among the simplest are fullwave rectifiers, combining two diodes in one envelope, and twin triodes, consisting of two triodes in one bulb for Class-B audio amplification. More complex types include duplex-diede triodes, duplex-diode pentodes, converters and mixers (for superheterodyne receivers), combination power tubes and rectifiers, and so on. In many cases the tube structure can be identified by the name, and all the types are basically the same as the simpler element combinations already described.
Mercury-vapor rectifiers-- The power lost in a diode rectifier (� 3-1) for a given plate current will be lessened if it is possible to decrease the plate-cathode voltage at which the current is obtained. If a small amount of mercury is put in the tube, the mercury will vaporize when the cathode is heated and, further, will ionize (� 2-4) when plate voltage is applied. This neutralizes the space charge and reduces the plate-cathode voltage drop to a practically constant value of about 15 volts regardless of the value of plate current. Since this drop is much smaller than can be attained with purely thermionic conduction, there is less power loss in the rectifier. The constant voltage drop also is an advantage. Mercuryvapor tubes are widely used in power rectifiers.
Grid-control rectifiers--If a grid is inserted in a mercury-vapor rectifier it is found that with sufficient negative grid bias it is possible to prevent plate current from flowing if the bias is present before plate voltage is applied. However, if the bias is lowered to the point where plate current can flow, the mercury vapor will ionize and the grid loses control of plate current since the space charge is neutralized. It can assume control again only after the plate voltage is disconnected. The same phenomenon also occurs in triodes filled with other gases which ionize at low pressure. Grid-control rectifiers find considerable application in many circuits where "electronic switching" is desirable.
03.4 COMMON ELEMENTS IN VACUUMTUBE CIRCUITS
Types of cathodes-- Cathodes are of two types, directly and indirectly heated. Directlyheated cathodes or filaments used in receiving tubes are of the oxide-coated type, consisting of a wire or ribbon of tungsten coated with certain rare metals and earths which form an oxide. capable of emitting large numbers of electrons with comparatively little cathodeheating power.
When directly-heated cathodes are operated on alternating current, the cyclic variation of current causes electrostatic and magnetic effects which vary the plate current of the tube at supply-frequency rate and thus produce

49 CHAPTER THREE

DheRadio Antateur'� ilandlooh

hum in the output. Hum from this source is eliminated by the indirectly-heated cathode, consisting of a thin metal sleeve or thimble, coated with electron-emitting material, enclosing atungsten wire which acts as aheater. The heater brings the cathode thimble to the proper temperature to cause electron emission. This type of cathode is also known as the equipotential cathode, since all parts are at the same potential.
Methods of obtaining grid bias -- Grid bias may be obtained from asource of voltage especially provided for that purpose, as a battery or other type of d.c. power supply. This is indicated in Fig. 309-A. A second method is shown at B, utilizing a cathode resistor; plate current flowing through the resistor causes avoltage drop which, with the connections shown, has the right polarity to bias the grid negatively with respect to the cathode. The value of the resistor is determined by the bias required and the plate current which flows at that value of bias, as found from the tube characteristic curves; with the voltage and current known, the resistance can be determined by Ohm's Law (� 2-6):
= E X 1000 where R, = cathode bias resistor in ohms
E = desired bias voltage = total d.c. cathode current in milliamperes.
-8 +
- B +
-8+
Fig. 309 -- Methods of obtaining grid bias. A, fixed bias; B, cathode bias; C, grid-leak bias.

Screen- and suppressor-grid currents should be included with the plate current in multielement tubes to obtain the total cathode current, and also the control-grid current if the control grid is driven positive during operation. The a.c. component of plate current flowing through the cathode resistor will cause avoltage drop which gives negative feedback into the grid circuit (� 3-3) so to prevent this the resistor usually is by-passed (� 2-13), Ce being the cathode by-pass condenser.
A third method is by use of agrid leak, R, in Fig. 309-C. This requires that the exciting voltage be positive with respect to the cathode during part of the cycle so that grid current will flow. The flow of grid current through the grid leak causes avoltage drop across the resistor which gives the grid a negative bias. The time constant (� 2-6) of the grid leak and grid condenser should be large in comparison to the time of one cycle of the exciting voltage so that the grid bias will be substantially constant and will not follow the variations in a.c. grid voltage. For grid-leak bias,
-- E X 1000
1,
where R, = grid-leak resistance in ohms E = desired bias voltage I,, = d.c. grid current in milliamperes.
When two tubes are operated in push-pull or parallel and use acommon cathode- or gridleak resistor, the value of resistance becomes one-half what it would be for one tube.
Cathode circuits; filament center tap -- When a filament-type cathode is heated by a.c. the hum introduced can be minimized by making the two ends of the filament have equal
and opposite potentials with respect to a
center point, usually grounded (� 2-13), to
which the grid and plate return circuits are
connected. The filament transformer winding is frequently center-tapped for this purpose, as shown in Fig. 310-A. The same result can be secured with an untapped winding by substituting a center-tapped resistor of 10 to 50 ohms as at B. The by-pass condensers, C1 and C2, are used in radio-frequency circuits to avoid having the r.f. current flow through the transformer or resistor, either of which may have considerable reactance at radio frequencies.
The filament supply for tubes with indirectly-heated cathodes sometimes is centertapped for the same purpose; although frequently one side of the filament supply, and hence one terminal of the tube heater, is simply grounded.
�3-7 OSCILLATORS
Self-oscillation -- If in an amplifier with positive feedback the feedback or regeneration

sn CHAPTER THREE

�

Grid Return

A
Plate Return

Grid Return

Plate Return

Fig. 310 -- Filament center-tap connections.

Vacuum .216,3
in the case of the resonant circuits usually associated with oscillators, is very nearly the resonant frequency of the circuit.
Magnetic feedback -- One form of feedback is by electromagnetic coupling between plate (output) and grid (input) circuits. Two representative circuits of this type are shown in Fig. 311. That at A is called the tickler circuit. The amplified current flowing in the "tickler," L2, induces avoltage in L1in the proper phase when the coils are connected as shown and wound in the same direction. The feedback can be adjusted by adjusting the coupling between L1 and L2.
The Hartley circuit, B, is similar in principle. There is only one coil, but it is divided so that part of it is in the plate circuit and part in the grid circuit. The magnetic coupling between the two sections of the coil provides the feedback, which can be adjusted by moving the tap on the coil.
R F Choke

(� 3-3) is increased to acritical value, the tube will generate acontinuous alternating current. This phenomenon, called oscillation, occurs when the power transferred between plate and grid circuits becomes large enough to overcome the circuit losses and the tube provides its own grid excitation. The power consumed is of course taken from the d.c. plate supply.
It is not necessary to apply external excitation to such acircuit, since any random variation in current, even though minute, will
rapidly be amplified up to the proper value to cause oscillation. The frequency of oscillation will be that at which losses are least which,

A
R FChoke
9 Fig. 311 -- Oscillator circuits with magnetic feedback. A, tickler circuit; B, Hartley circuit.

11
C9
Fig. 312 -- Osci lator ercuits with capacity feedback. A, Colpitts circuit B, tuned-plate tuned-grid circuit; C, ultraudion circuit.
Capacityfeedback -- The feedback can also be obtained through capacity coupling, as shown in Fig. 312. At A, the Colpitts circuit, the voltage across the resonant circuit is divided, by means of the series condensers, into two parts. The instantaneous voltages at the ends of the circuit are opposite in polarity with respect to the cathode, hence in the right phase to sustain oscillation.

51 CHAPTER THREE

e Radio ...Amaieur'� -lian(goolz

The tuned-grid tuned-plate circuit at B utilizes the grid-plate capacity of the tube to provide feedback coupling. There should be no magnetic coupling between the two tunedcircuit coils. Feedback can be adjusted by varying the tuning of either the grid or plate circuit. The circuit with the higher Q (� 2-10) determines the frequency of oscillation, although the two circuits must be tuned approximately to the same frequency for oscillations to occur.
The ultraudion circuit at C is equivalent to the Colpitts, with the voltage division for oscillation brought about through the gridto-filament and plate-to-filament capacities of the tube. In this and in the Colpitts circuit the feedback can be controlled by varying the ratio of the two capacities. In the ultraudion circuit this can be done by connecting a small variable condenser between grid and cathode.
Crystal oscillators -- Since a properly-cut quartz crystal is equivalent to ahigh-Q tuned circuit (� 2-10) it may be substituted for a conventional circuit in an oscillator to control the frequency of oscillation. A simple crystal oscillator circuit is shown in Fig. 313. It will be
Xtal.

--B

Fig. 313 -- Simple crystal oscillator circuit.

recognized as the tuned-plate-tuned-grid circuit with the crystal substituted for the res-
onant circuit in the grid. Many variations of this fundamental circuit are used in practice.
Series and parallel feed-- A circuit such as the tickler circuit of Fig. 311-A is said to be series fed because the source of plate voltage and the r.f. plate circuit (the tickler coil) are connected in series, hence the d.c. plate current flows through the coil to the plate. A by-pass (� 2-13) condenser, Cb, must be connected
across the plate supply to shunt the radiofrequency current around the source of power. Other examples of series plate feed are shown in Figs. 312-B and 313.
In some cases the source of plate power must be connected in parallel with the tuned circuit in order to provide apath for direct current to the plate. This is illustrated by the Hartley circuit of Fig. 311-B where it would be impossible to feed the plate current through the coil because there is a direct connection be-

tween the coil and cathode. Hence the voltage is applied to the plate through a radio-frequency choke which prevents the r.f. current from flowing to the plate supply and thus
short-circuiting the oscillator. The blocking condenser Cb provides a low-impedance path for radio-frequency current flow but is an open circuit for direct current (� 2-13). Other examples of parallel feed are shown in Figs. 312-A and 312-C.
Values of chokes, by-pass and blocking condensers are determined by the considerations outlined in �2-13.
Excitation and bias -- The excitation volt-
age required depends upon the characteristics of the tube and the losses in the circuit, including the power consumed in the load. In practically all oscillators the grid is driven positive during part of the cycle, so that power is consumed in the grid circuit (� 3-2). This power must be
supplied by the plate circuit. With insufficient excitation the tube will not oscillate; with too-high excitation the grid losses, or power
consumed in the grid circuit, will be exces-
sive. Oscillators are almost always grid-leak
biased (� 3-6), which not only takes advan-
tage of the grid-current flow but also gives better operation since the bias adjusts itself to
the excitation voltage available. Tank circuit -- The resonant circuit asso-
ciated with the oscillator is generally called the tank circuit. This name derives from the storage of energy associated with a resonant circuit of reasonably high Q (� 2-10). It is applied to any resonant circuit in transmitting applications, whether used in an oscillator or amplifier.
Power output --The power output of an oscillator is the useful a.c. power consumed in aload connected to the oscillator. The load may be coupled by any of the means described
in �2-11. Plate efficiency --The plate efficiency
(� 3-3) of an oscillator depends upon the load resistance, excitation, and other operating conditions, and usually is in the vicinity of 50%. It is not as high as in the case of an amplifier, since the oscillator must supply its own grid losses, which are usually 10% to 20%
of the output power. Frequency stability -- The frequency sta-
bility of an oscillator is its ability to maintain constant frequency in the presence of variable
operating conditions. The more important factors which may cause a change in frequency are (1) plate voltage, (2) temperature, (3) loading, (4) mechanical variations of circuit elements. Plate-voltage variations will cause a corresponding instantaneous shift in frequency; this type of frequency shift is called dynamic instability. Temperature changes will

52 CHAPTER THREE

Vacuum

cause tube elements to expand or contract slightly, thus causing variations in the interelectrode capacities (� 3-2), and since these are unavoidably part of the tuned circuit till frequency will change correspondingly. Temperature changes in the coil or condenser will change the inductance and capacity slightly,
again causing ashift in the resonant frequency. Both these temperature effects are relatively slow in operation, and the frequency change caused by them is called drift. Load variations act in much the same way as plate voltage variations except when there is atemperature change in the load, when drift also may be present. Mechanical variations, usually caused by vibration, cause changes in inductance and/ or capacity which in turn cause the frequency to "wobble" in step with the vibratidn.
Dynamic instability can be reduced by using atuned circuit of high effective Qwhich means, since the tube and load represent arelatively low resistance in parallel with the circuit, that
a low L/C ratio ("high-C") must be used (� 2-10), and that the circuit should be lightly loaded. Dynamic stability also can be improved by using a high value of grid leak, which gives high grid bias and raises the effective resistance of the tube as seen by the tank circuit, and by using relatively high plate voltage and low plate current, which accomplishes the same result. Drift can be minimized by using low d.c. input (for the size of tube), by using coils of large wire to prevent undue temperature rise, and by providing good ventilation to carry off heat rapidly. A low L/C ratio in the tank circuit also helps because the interelectrode capacity variations have proportionately less effect on the frequency when shunted by a large condenser. Special temperature-compensated components also can be used. Mechanical instability can be prevented by using well-designed components and insulating the oscillator from mechanical vibration.
Negative-resistance oscillators-- If aresonant circuit were completely free from losses

+Es

-B

-SUP. +Er

Fig. 314 -- Negative-resistance oscillator. This circuit, known as the "transitron," requires that the screen be operated at ahigher d.c. potential than the plate of the tube.

Fig. 315 --The multivibrator circuit, or relaxation oscillator.
a current once started would continue indefinitely; that is, sustained oscillations would occur. This condition can be simulated in practice by cancelling the actual resistance in the circuit by inserting an equal or greater amount of negative resistance. Negative resistance is exhibited by any device showing an increase of current when the applied voltage is decreased, or vice versa.
The vacuum tube can be made to show negative resistance by anumber of arrangements of electrode potentials. One circuit is shown in Fig. 314. Negative resistance is produced by virtue of the fact that as the suppressor grid of a pentode is given more negative bias, electrons normally passing through to the plate are turned back to the screen, thus increasing the screen current, and reversing normal tube action (� 3-2). The negative resistance so produced is sufficiently low so that ordinary tuned circuits will oscillate readily at frequencies up to 15 Mc. or so.
The multivibrator -- The type of oscillator circuit shown in Fig. 315 is known as the multivibrator, or relaxation oscillator. Two tubes are used with resistance coupling, the output of one tube being fed to the input circuit of the other. The frequency of oscillation is determined by the time constants (� 2-6) of the resistance-capacity combinations. The principle of oscillation is the same as in the feedback circuits already described, the second tube being necessary to obtain the proper phase relationship (� 3-3) for oscillation when the energy is fed back.
The multivibrator is avery unstable oscillator, and for this reason its frequency readily can be controlled by a small signal of steady frequency introduced into the circuit. This phenomenon is called locking. Its output waveshape is highly distorted, hence has high harmonic content (� 2-7). A useful feature is that the multivibrator will lock with a frequency corresponding to one of its higher harmonics (the tenth harmonic is frequently used) and it can therefore be used as afrequency divider.

53 CHAPTER TIIREE

DheRadio Ainaleur'J fiancliooh

�3-8 CATHODE-RAT TUBES

retentivity of vision makes the path of the mov-

Principles -- The cathode-ray tube is a ing spot (trace) appear to be acontinuous line.

vacuum tube in which the electrons emitted

Electrostatic deflection, generally used in

from ahot cathode are accelerated to give them the smaller tubes, is produced by deflection

considerable velocity, formed into abeam, and plates. Two sets of plates are placed at right

allowed to strike a special translucent screen angles to each other, as indicated in Fig. 316.

which fluoresces, or gives off light at the point The fields are created by applying suitable volt-

where the beam strikes. A narrow beam of mov- ages between the two plates of each pair. Usu-

ing electrons is similar to a wire carrying cur- ally one plate of each pair is connected to anode

rent (� 2-4) and is accompanied by electrostatic No. 2to establish the polarities (� 2-3) of the

and electromagnetic fields. Hence it can be de- fields with respect to the beam and to each

fleeted (have its direction changed) by applica- other.

tion of external electrostatic or magnetic fields

Tubes intended for magnetic deflection have

which exert a force on the beam in the same the same type of gun, but have no deflection

way as similar fields do on charged bodies or on plates. Instead, the deflecting fields are set up

wires carrying current (� 2-3, 2-5). Since the by means of coils corresponding to the plates

beam consists only of moving electrons, its in tubes having electrostatic deflection. The

weight and inertia are negligibly small, hence coils are external to the tube but are mounted

it can be deflected easily and without any close to the glass envelope in the same rela-

appreciable time lag. For this reason it can tive positions occupied by the electrostatic

be made to follow instantly the variations in deflection plates.

fields which are changing periodically at very

The beam deflection caused by a given

high radio frequencies.

change in the field intensity is called the de-

Electron gun -- The electrode arrangement flection sensitivity. With electrostatic-deflection

which forms the electrons into abeam is called tubes it is usually expressed in millimeters per

the electron gun. In the simple tube structure volt, which gives the linear movement of the

shown in Fig. 316, the gun consists of the spot on the screen as afunction of the voltage

cathode, grid, and anodes Nos. 1and 2. The in- applied to a set of deflecting plates. Values

tensity of the electron beam is regulated by the range from about 0.1 to 0.6 mm/volt, depend-

grid in the same way as in an ordinary tube ing upon the tube construction and gun elec-

(� 3-2). Anode No. 1is operated at apositive trode voltages. The sensitivity is decreased by

potential with respect to the cathode, thus an increase in anode No. 2voltage.

accellerating the electrons which pass through

Fluorescent screens--The fluorescent screen

the grid, and is provided with small apertures materials used have varying characteristics ac-

through which the electron stream passes and is cording to the type of work for which the tube

concentrated into anarrow beam. This anode is intended. The spot color is usually green,

is also known as the focusing electrode. Anode white, yellow or blue, depending upon the

No. 2is operated at a high positive potential screen material. The persistence of the screen

with respect to cathode and further increases is the time duration of the afterglow which

the velocity of the electrons in the beam. The exists when the excitation of the electron beam

electron velocity and sharpness of the beam are is removed. Screens are classified as long-, me-

determined by the relative voltages on the dium- and short-persistence. Small tubes for

electrodes. In some tubes a second grid is in- oscilloscope work are usually provided with

serted between the control grid and anode No. medium-persistence screens having greenish

1 to provide additional accelleration of the fluorescence.

electrons.

Tube circuits -- A representative cathode-

Methods of deflection --The gun alone ray tube circuit with electrostatic deflection is

simply produces a small spot on the screen, shown in Fig. 317. One plate of each pair of de-

but when the beam is deflected by either mag- fleeting plates is connected to anode No. 2.

netic or electrostatic fields the spot moves Since the voltages required are normally rather

across the screen in proportion to the force ex- high, the positive terminal of the supply is

erted. When the motion is sufficiently rapid, usually grounded (� 2-13) so that the common

Cathode 4nede No./

deflection plates will be at ground potential. This places the cathode and other ele-

ments at high potentials above ground,

F1- 1- 1

hence these elements must be well insu-

lated. The various electrode voltages are

obtained from a voltage divider (� 2-6)

across the high-voltage d.c. supply. R3 is a

Grid

Ariede No.2

Fig. 316 -- Arrangement of

DePlllaetcetsIon elements in the

cathode-ray

tube with electrostatic beam deflection.

variable divider or "potentiometer" for adjusting the negative bias on the control grid and thereby varying the beam cur-

54 CHAPTER THREE

Vacuum Duhe�

of the small current requirements a rectified a.c. supply with half-wave rectification (� 8-3) and asingle 0.5 to 2-1.`fd. condenser as afilter (� 8-5) is satisfactory.

High-Vo/true D.CSupply

Input Voltage Input Voltaye

Fig. 317-- Cathode-ray tube circuit. Typical values for athree-inch (screen-diameter) tube such as the type 906:
RI, R2-- 120 10 megohms. R3-- 20,000�0h111 potentiometer. R4 -- 0.2-megohm potentiometer. Rs -- 0.5 megohm.
The high-voltage supply should furnish about 1300 volts d.c.
rent; it is called the intensity or brightness control. The focus, or sharpness of the luminous spot formed on the screen by the beam, is controlled by R4, which changes the ratio of anode No. 2to anode No. 1voltage. The focusing and intensity controls interlock to some extent, and the sharpest focus is obtained by keeping the beam current low.
Deflecting voltages for the plates are applied to the terminals marked "input voltage," Ri and R2 being high resistances (1 megohm or more) to drain off any accumulation of charge on the deflecting plates. Usually some provision is made to place an adjustable d.c. voltage on each set of plates so that the spot can be "centered" when stray electrostatic or magnetic fields are present; the adjustable voltage simply is set to neutralize such fields.
The tube is mounted so that one set of plates produces ahorizontal line when avarying voltage is applied to it, while the other set of plates produces a vertical line under similar conditions. They are called, respectively, the "horizontal" and "vertical" plates, but which set of actual plates produces which line is simply a matter of how the tube is mounted. Et is usually necessary to provide a mounting which can be rotated to some extent so that the lines will actually be horizontal and vertical.
Power supply--The d.c. voltage required for operation of the tube may vary from 500 volts for the miniature type (1-inch diameter screen) to several thousand for the larger tubes. The current, however, is very small, so that the power required is likewise small. Because

� 3�9 THE OSCILLOSCOPE
Description -- An oscilloscope is essentially acathode-ray tube in the basic circuit of Fig. 317, but with provision for supplying asuitable deflection voltage on one set of plates, ordinarily those giving horizontal deflection. The deflection voltage is called the sweep. Oscilloscopes are frequently also equipped with vacuum-tube amplifiers for increasing the amplitude of small a.c. voltages to values suitable for application to the deflecting plates. These amplifiers are ordinarily limited to operation in the audio-frequency range, and hence cannot be used at radio frequencies.
Formation of patterns--When periodically-varying voltages are applied to the two sets of deflecting plates the path traced by the fluorescent spot forms a pattern which is stationary so long as the amplitude and phase relationships of the voltages remain unchanged. Fig. 318 shows how such patterns are formed. The horizontal sweep voltage is assumed to have the "sawtooth" waveshape indicated; with no voltage applied to the vertical plates, the trace would simply sweep from left to right across the screen along the horizontal axis X-X' until the instant H is reached, when it reverses direction and returns to the starting point. The sine-wave voltage applied to the vertical plates would similarly trace a line along the axis Y-Y' in the absence of any deflecting voltage on the horizontal plates. How-
Y' Ma. 318 -- Showing the formation of the pattern from the horizontal and vertical sweep voltages.

55 CHAPTER THREE

Dhe Radio _AinctieuA ....11andLoh

ever, when both voltages are present the posi-
tion of the spot at any instant depends upon the voltages on both sets of plates at that instant. Thus at time B the horizontal voltage
has moved the spot a short distance to the right and the vertical voltage has similarly moved it upward, so that it reaches the actual position B' on the screen. The resulting trace is easily followed from the other indicated positions, which are taken at equal time intervals.
Types of sweeps -- A horizontal sweep-voltage waveshape such as that shown in Fig. 318 is called alinear sweep, because the deflection
in the horizontal direction is directly proportional to time. If the sweep were perfect, the "fly-back" 'time, or time taken for the spot to return from the end (H) to the beginning (I or
A) of the horizontal trace would be zero, so that the line HI would be perpendicular to the axis Y-Y'. Although the fly-back time cannot be made zero in practicable sweep-voltage generators, it can be made quite small in comparison to the time of the desired trace AH, at least at most frequencies within the audio range. The fly-back time is somewhat exaggerated in Fig. 318 to show its effect on the pattern. The line H'I' is called the return trace; with a linear sweep it is less brilliant than the pattern because the spot is moving much more rapidly during the fly-back time than during the time of the main trace. If the fly-back time is short enough the return trace will be invisible.
The linear sweep has the advantage that it shows the shape of the wave applied to the vertical plates in the same way in which it is usually represented graphically (� 2-7). By making the sweep time equal to a multiple of the time of one cycle of the a.c. voltage applied to the vertical plates, several cycles of the vertical or signal voltage will appear in the pattern. The shape of only the last cycle to
appear will be affected by the fly-back in such acase. Although the linear sweep is generally most useful, other waveshapes may be desirable for certain purposes. The shape of the pattern obviously will depend upon the shape of the horizontal sweep voltage. If the horizontal sweep is sinusoidal, the main and return sweeps each occupy the same time, and the spot moves faster horizontally in the center of the pattern

than it does at the ends. If two sinusoidal voltages of the same frequency are applied to both sets of plates, the resulting pattern may be a straight line, an ellipse or a circle, depending upon the amplitude and phase relationships. If the frequencies are harmonically related (� 2-7) a stationary pattern will result, but if one frequency is not an exact harmonic of the other the pattern will show continuous motion. This is also the case when a linear sweep circuit is used; the sweep frequency and the frequency under observation must be harmonically related or the pattern will not be stationary.
Sweep circuits -- A sinusoidal sweep is easiest to obtain, since it is possible to apply a.c. voltage from the power line directly or through asuitable transformer to the horizontal plates. A variable voltage divider can be used to regulate the width of the horizontal trace.
A typical circuit for alinear sweep is shown in Fig. 319. The tube is a gas triode or gridcontrol rectifier (� 3-5). The breakdown voltage, or plate voltage at which the tube ionizes and starts conducting, is determined by the grid bias. When plate voltage is applied, the
voltage across C1 rises, as it acquires acharge through RI,until the breakdown voltage is reached, when the condenser discharges rapidly through the comparatively low plate-cathode resistance of the tube. When the voltage drops to a value too low to maintain plate-current flow, the ionization is extinguished and C1 once more charges through RI.If RI is large enough, the voltage across Ci rises linearly with time up to the breakdown point. This voltage is used for the sweep, being coupled to the cathode-ray tube or to an amplifier through C2. The fly-back is the time required for discharge through the tube, and to keep it small the resistance during discharge must be as low as possible.
To obtain a stationary pattern, the "sawtooth" frequency can be controlled by varying C1 and RI,and by introducing some of the voltage to be observed (on the vertical plates) into the grid circuit of the tube. This voltage "triggers" the tube into operation in synchronism with the signal frequency. Synchronizing will occur even though the signal frequency is a multiple of the sweep frequency,

Sync. Volta9e- 1
-a

Fig. 319 -- A linear sweep-oscillator circuit.

Ci -- 0.001 to 0.25 ;dd.

C2 -- 0.5 dd.

C3 -- 0.1 dd.

C4 --25 dd., 25-volt electrolytic.

R1--0.3 to 1.5 megohms.

R2 -- 2000 ohms.

R4 -- 25,000 ohms.

R3 --0.25 megohm.

Rs-- 0.1 rnegohm.

"B" supply should deliver 300 volts. CIand RIare

proportioned to give suitable sweep frequency; the higher the time constant (f 2-6) the lower the frequency.

R4 ir3 aprotective resistor to limit grid-current flow dur-

ing the &ionizing period, when positive ions are attracted to the negative grid.

56 CHAPTER THREE

Vacuum ..7ute�

provided the circuit constants and the amplitude of the synchronizing voltage are properly adjusted.
The voltage output of the type of circuit shown in Fig. 319 is limited because the charg-
ing rate of the condenser is linear only on that portion of the logarithmic charging curve
(� 2-6) which is practically a straight line. A linear charging rate over a longer period of time can be secured by substituting acurrent-
limiting device, such as a properly-adjusted vacuum tube, for RI.
Amplifiers-- The usefulness of the oscilloscope is enhanced by providing amplifiers for both the horizontal and vertical sweep voltages, thereby insuring that sufficient voltage will be available at the deflection plates to give a pattern of suitable size. With small oscilloscope tubes (3-inch and smaller screens) the voltage required for a deflection of one inch
varies from about 30 to 100 volts, depending upon the anode voltages, so that an amplifier
tube capable of an undistorted peak output

voltage of 100 or so is necessary. (With such an amplifier the voltage difference, or total volt-
age "swing", between the positive and negative peaks is 200 volts.) A resistance-coupled
voltage amplifier (� 3-3) having apentode tube is ordinarily used because of the high stage
gain obtainable, and the amplifier should be designed to have good frequency response over as wide arange of audio frequencies as possible (� 5-9). Since a voltage gain of 100 to 150 or more is readily obtainable, full deflection of the beam can be secured with an input of one volt or less with such an amplifier.
Constructional considerations-- An oscilloscope should be housed in ametal cabinet,
both to shield the tube from stray electromagnetic and electrostatic fields which might deflect the beam, and also to protect the operator
from the high voltages associated with the tube. It is good practice to provide an interlock switch which automatically disconnects
the high-voltage supply when the cabinet is opened for servicing or other reasons.

57 CHAPTER THREE

CHAPTER FOUR
Paclio--7requency Power

eneratIt. on

�4-1 TRANSMITTER REQUIREMENTS
General Requirements --The power output of a transmitter must be as stable in frequency and as free from spurious radiations as the state of the art permits. The steady r.f. output, called the carrier (� 5-1), must be free from amplitude variations attributable to ripple from the plate power supply (� 8-4) or other causes, its frequency should be unaffected by variations in supply voltages or in-
advertent changes in circuit constants, and there should be no radiations on other than the intended frequency. The degree to which these requirements can be met depends upon the operating frequency.
Design principles --The design of the transmitter depends on the output frequency, the required power output, and the type of operation (c.w. telegraphy or 'phone). For c.w. operation at low power on medium-high frequencies (up to 7Mc. or so) asimple crystal oscillator circuit can meet the requirements satisfactorily. However, the stable power output which can be taken from an oscillator is limited, so that for higher power the oscillator is used simply as a frequency-controlling element, the power being raised to the desired level by means of amplifiers. The requisite frequency stability can be obtained only when the oscillator is operated on relatively low frequencies, so that for output frequencies up to about 60 Mc. it is necessary to increase the oscillator frequency by multiplication (harmonic generation -- �3-3), which is usually done at fairly low power levels and before the final amplification. An amplifier which delivers power on the frequency applied to its grid circuit is known as astraight amplifier; one which
gives harmonic output is known as afrequency multiplier. An amplifier used principally to isolate the frequency-controlling oscillator from the effects of changes in load or other variations in following amplifier stages is called a
buffer amplifier. A complete transmitter therefore may consist of an oscillator followed by one or more buffer amplifiers, frequency multipliers, and straight- amplifiers, the number being determined by the output frequency and power in relation to the oscillator frequency and power. The last amplifier is called the final amplifier, and the stages up to the last comprise the exciter. Transmitters are usually de-

signed to work in anumber of frequency bands, so that means for changing the frequency of resonant circuits in harmonic steps usually is provided, generally by means of plug-in in-
ductances. The general method of designing atransmit-
ter is to decide upon the power output and the highest output frequency required, and also the number of bands in which the transmitter
is to operate. The latter usually will determine the oscillator frequency, since it is general practice to set the oscillator on the lowest frequency band to be used. The oscillator frequency is seldom higher than 7 Mc. except in
some portable installations where tubes and power must be conserved. A suitable tube (or pair of tubes) should be selected for the final amplifier and the grid driving power required determined from the tube manufacturer's data. This sets the power required from the preceding stage. From this point the same process is followed back to the oscillator, including frequency multiplication wherever necessary. The selection of a suitable tube complement requires knowledge of the operating characteristics of the various types of amplifiers and oscillators. These are discussed in the following
sections. At 112 Mc. and above the ordinary methods
of transmitter design become rather cumbersome, although it is possible to use them with proper choice of tubes and other components. However, in this ultra-high-frequency (� 2-7) region the requirements imposed are less severe, since the limited transmission range (� 9-5) mitigates the interference conditions that determine the requirements on the longdistance lower frequencies. Hence simple oscillator transmitters are widely used.
Vacuum tubes -- The type of tube used in the transmitter has an important effect on the circuit design. Tubes of high power sensitivity (� 3-3) such as pentodes and beam tetrodes, give larger power amplification ratios per stage than do triodes, hence fewer tubes and stages may be used to obtain the same output power. On the other hand, triodes have certain operating advantages such as simpler power supply circuits and relatively simpler adjustment for modulation (� 5-3), and in addition are considerably less expensive for the same power output rating. Consequently it is usually more

58 CHAPTER FOUR

e cho-.7requency Potver �encra lion

economical to use triodes as output amplifiers even though an extra low-power amplifier stage may be necessary.
At frequencies in the region of 56 Mc. and above it is necessary to select tubes designed particularly for operation at ultra-high frequencies, since tubes built primarily for the lower frequencies may work poorly or not at all.
� 4-2 SELF-CONTROLLED OSCILLATORS
Advantages and disadvantages -- The chief advantage of a self-controlled oscillator is that the frequency of oscillation is determined by the constants of the tuned circuit, and hence readily can be set to any desired value. However, extreme care in design and adjustment are essential to secure satisfactory frequency stability (� 3-7). Since frequency stability is generally poorer as the load on the oscillator is increased, the self-controlled oscillator should be used purely to control frequency and not for the purpose of obtaining appreciable power output, in transmitters intended for working below 60 Mc.
Oscillator circuits-- The inherent stability of all of the oscillator circuits described in �3-7 is about the same, since stability is more afunction of choice of proper circuit values and adjustment than of the method by which feedback is obtained. However, some circuits are more convenient to use than others, particularly from the standpoint of feedback adjustment, mechanical considerations (whether the tuning condenser rotor plates can be grounded or not, etc.), and uniform output over a considerable frequency range. All simple circuits suffer from the fact that the power output must be taken from the frequencydetermining tank circuit, so that aside from the effect of loading on frequency stability, the following amplifier stage also can react on the oscillator in such a way as to change the frequency.
The electron-coupled oscillator -- The effects of loading and coupling to the next stage can be greatly reduced by use of the electron-coupled circuit, in which a screen-grid tube (� 3-5) is so connected that its screen grid is used as aplate, in connection with the control grid and cathode, in an ordinary triode oscillator circuit. The screen is operated at ground r. f. potential (�2-13) to act as ashield between the actual plate and the cathode and control grid; the latter two elements must therefore be above ground potential. The output is taken from the plate circuit. Under these conditions the capacity coupling (� 2-11) between the plate and other ungrounded tube elements is quite small, hence the output power is secured almost entirely by variations

L,
r C
A

cz
IC
- H V. t

- N.V.+
Fig. 401 -- Electron-coupled oscillator circuit. For
maximum stability the grid leak, RI, should he 100,000 ohms or more. The grid condenser should be of the order of 100 pfd., other fixed condensers from 0.082 pfd. to 0.1 pfd. Proper values for R2 and R3 may be determined from 8-10. For maximum isolation between oseillatot and output circuits, the tube should have extremely low grid-plate capacity.
in the plate current caused by the varying potentials on the grid and cathode. Since in a screen-grid tube the plate voltage has a relatively small effect on the plate current, the reaction on the oscillator frequency for different conditions of loading is small.
It is generally most convenient to use a Hartley (� 3-7) circuit in the frequency-determining part of the oscillator. This is shown in Fig. 401, where LiC2is the oscillator tank circuit. The screen is grounded for r.f. through a by-pass condenser (� 2-13) but has the usual d.c. potential. The cathode connection is made to atap on the tank coil to provide feedback. In the plate circuit a resonant circuit, L2C2, can be connected as shown at A; it may be tuned either to the oscillation frequency or to one of its liars-Ionics. Untuned output coupling is shown at B; with this method the output voltage and power are considerably lower than with atuned plate circuit, but better isolation between oscillator and amplifier is secured.
If the oscillator tube is a pentode with an external suppressor connection the suppressor grid should be grounded, not connected to cathode. This provides additional internal shielding and further isolates the plate from the frequency-determining circuit.
Factors influencing stability -- The causes of frequency instability and the necessary remedial steps have been discussed in �3-7.

59 CHAPTER FOUR

pit2 Pull� Amctieur'�

These apply to all oscillators. In addition, in the electron-coupled oscillator the ratio of plate to screen voltage has an important effect on the stability with changes in supply voltage; the optimum ratio is generally of the order of 3:1 but should be determined experimentally for each case. Since the cathode is above ground potential, means should be taken to reduce the effects of heater-to-cathode capacitance or leakage, which by allowing a small a.c. voltage from the heater supply to develop between cathode and ground may cause modulation (� 4-1) at the supply frequency. This effect, which is usually appreciable only at 14 Mc. and higher, may be reduced by by-passing of the heaters as indicated in Fig. 401 or by operating the heater at the same r.f. potential as the cathode. The lat-
Fil.
Fig. 402 -- Method of operating the heater at cathode r.f. potential in an electron-coupled oscillator. L2 should have the same number of turns as the part of Li between ground and the cathode tap, and should be closely coupled to Li (preferably interwound). By-pass condenser Cshould be 0.01 to 0.1 pfd.
ter may be accomplished by the wiring arrangement shown in Fig. 402.
Tank circuit Q --The most important single factor in Oetermining frequency stability is the Q of the oscillator tank circuit. The effective Q must be as high as possible for best stability. Since oscillation is accompanied by grid-current flow, the grid-cathode circuit constitutes a resistance load of appreciable proportions, the effective resistance being low enough to be the determining factor in establishing the effective parallel impedance of the tank circuit. Consequently, if the ends of the tank are connected to plate and grid, as is usual, a high effective Q can be obtained only by decreasing the L/C ratio and making the inherent resistance in the tank as low as possible. The tank resistance van he decreased by using low-loss insulation on condensers and coils, and by winding the coil with large wire. With ordinary construction the optimum tank capacity is of the order of 500 to 1000 maid. at afrequency of 3.5 Me.
The effective circuit Q can be raised by in-

creasing the resistance of the grid circuit and thus decreasing the loading. This can be ac-
complished by reducing the oscillator grid current, by using minimum feedback to main-
tain stable oscillation and by using ahigh value of grid-leak resistance.
A high-Q tank circuit can also be obtained
with a higher L/C ratio by "tapping down" the tube connections on the tank (� 2-10). This is advantageous in that acoil with higher
inherent Q can be used; also, the circulating r.f. current in the tank circuit is reduced so that drift from coil heating is decreased. However, the circuit is complicated to some extent
and the taps may cause parasitic oscillations to be set up (� 4-10).
Plate supply -- Since the oscillator frequency will be affected to some extent by
changes in plate supply voltage, it is necessary that the latter be free from ripple (� 8-4) which
would cause frequency variations at the ripplefrequency rate (frequency modulation). It is also
advantageous to use avoltage-stabilized power supply (� 8-8). Since the oscillator is usually operated at low voltage and current, gaseous
regulator tubes are quite suitable.
Power level-- The self-controlled oscillator should be designed purely for frequency control and not to give appreciable power output, hence small tubes of the receiving type may be used. The power input is ordinarily not more
than a watt or two, subsequent buffer amplifiers being used to increase the power to the desired level. The use of receiving tubes is
advantageous mechanically, since the small elements are less susceptible to vibration and
are usually securely braced to the envelope.
Oscillator adjustment-- The adjustment of an oscillator consists principally in observ- .
jog the design principles outlined in the preceding paragraphs. Frequency stability should
be checked with the aid of astable receiver, or
an auxiliary crystal oscillator may be used as a standard for checking dynamic stability and drift, the self-controlled oscillator being ad-
justed to approximately the same frequency so that an audio-frequency beat (� 2-13) can be obtained. If it is possible to vary the oscil-. lator plate voltage (an adjustable resistor of
50,000 or 100,000 ohms in series with the plate supply lead will give considerable variation)
the change in frequency with change in plate voltage may be observed and the operating conditions varied until minimum frequency
shift results. The principal factors affecting dynamic stability will be the tank circuit
L C' ratio, the grid-leak resistance, and the amount of feedback. In the electron-coupled
circuit the latter may be adjusted by changing
the position of the cathode tap on the tank
coil; this adjustment is quite important in its
effect on the frequency stability.

CHAPTER FOUR

leall0-.7r�ltlenCy Power generation

Drift may be checked by allowing the oscillator to operate continuously from acold start, the frequency change being observed at regular intervals. Drift may be minimized by using less than the rated power input to the plate of the tube, by construction which prevents tube heat from reaching the tank circuit elements, and by use of large wire in the tank coil to reduce temperature rise from internal heating.
In the electron-coupled oscillator having a tuned plate circuit (Fig. 401-A) resonance at the fundamental and harmonic frequencies of the oscillator portion of the tube will be indicated by a dip in plate current as the plate tank condenser is varied. This dip should be rather marked at the fundamental, but will be less so on harmonic frequencies.
� 1-3 PIEZO-ELECTRIC CRYSTALS
Characteristics -- Piezo-electric crystals (� 2-10) are universally used for controlling the frequency of transmitting oscillators because the extremely high Q of the crystal and the necessarily loose coupling between it and the oscillator tube make the frequency stability of acrystal-controlled oscillator very high. Active plates may be cut from araw crystal at various angles to. its electrical, mechanical and optical axes, resulting in differing characteristics as to thickness, frequency-temperature coefficient, power-handling capabilities, etc. The commonly used cuts are designated as X, Y, AT, V, and LD.
The ability to adhere closely to aknown frequency is the outstanding characteristic of a crystal oscillator. This is also its disadvantage, in that the oscillator frequency can be changed appreciably only by using a number of crystals.
Frequency-thickness ratio -- Crystals used for transmitting purposes are so cut that the thickness of the crystal is the frequency-determining factor, the length and width of the plate being of relatively minor importance. For a given crystal cut, the ratio between thickness and frequency is a constant; that is,
F �
where F is the frequency in megacycles and t is the thickness of the crystal in thousandths of an inch. For the X-cut, k = 112.6; for the Y-cut, k = 77.0; for the AT-cut, k =- 66.2.
At frequencies above the 7-Mc, region the crystal becomes very thin and correspondingly fragile, so that crystals are seldom manufactured for operation much above this frequency. Direct crystal control on 14 and 28 Mc. is secured by use of "harmonic" crystals, which are ground to be active oscillators when ex-

cited at the third harmonic of the frequency represented by their thickness.
Temperature coefficient of frequency -- The resonant frequency of a crystal will vary with its temperature, to an extent depending upon the type of cut. The frequency-temperature coefficient is usually expressed in cycles frequency change per megacycle, per degree Centigrade temperature change, and may be either positive (increasing frequency with increasing temperature) or negative (decreasing frequency with increasing temperature). X-cut crystals have anegative coefficient of 15 to 25 cycles/megacycle/degree C. The coefficient of Y-cut crystals may vary from -- 20 cycles/ megacycle/degree C. to -I- 100 cycles/megacycle/degree C. The AT, V and LD cuts have very low coefficients. Y-cut crystals frequently "jump" to another frequency when the temperature is changed rather than gradually changing frequency as the nominal coefficient
might indicate, and hence are rather unreliable under temperature variations.
The temperature of a crystal depends not only on the temperature of its surroundings but also on the power it must dissipate while
oscillating, since power dissipation causes heating (� 2-6, 2-8). Consequently the crystal temperature may be considerably above that of the surrounding air when the oscillator is in
operation. To minimize heating and frequency drift (� 3-7) the power used in the crystal must be kept to aminimum.
Power limitations -- If the crystal is made to oscillate too strongly, as when it is used in an oscillator circuit with high plate voltage and excessive feedback, the amplitude of the mechanical vibration will become great enough to crack or puncture the quartz. An indication
of the vibration amplitude can be obtained by connecting an r.f. current indicating device of suitable range in series with the crystal. Safe
r.f. crystal currents range from 50 to 200 milliamperes, depending upon the type of cut. A flashlight bulb or dial light of equivalent current rating makes agood current indicator. By choosing a bulb of lower rating than the current specified by the manufacturer as safe for
the particular type of crystal used, the bulb will serve as afuse, burning out before acurrent dangerous to the crystal is reached. The
60-ma. and 100-ma, bulbs are frequently used for this purpose. High crystal current is accompanied by increased power dissipation and heating, so that the frequency change also is greatest when the crystal is overloaded.
Crystal mountings -- To make use of the crystal, it must be mounted between two metal electrodes. There are two types of mountings, one having a small air-gap between the top plate and the crystal and the other maintaining both plates in contact with the crystal. It is es-

61 CHAPTER FOUR

Palio Arnaleuris --11anliooZ

sential that the surfaces of the metal plates in
contact with the crystal be perfectly flat. In the air-gap type of holder, the frequency of oscillation depends to some extent upon the size of the gap. This property can be used to advantage with most low-drift crystals so that by using a holder having a top plate with closely adjustable spacing a controllable frequency variation can be obtained. A 3.5-Mc. crystal will oscillate without very great variation in power output over arange of about 5 kc. X- and Y-cut crystals are not generally suitable for this type of operation because they have atendency to "jump" in frequency with different air gaps.
A holder having aheavy metal bottom plate with a large surface exposed to the air is advantageous in radiating quickly the heat generated in the crystal and thereby reducing temperature effects. Different plate sizes, pressures, etc., will cause slight changes in frequency, so that if acrystal is being ground to an exact frequency it should be tested in the holder and with the same oscillator circuit with
which it will be used in the transmitter.

tube of high power sensitivity (� 3-3), such as apentode or beam tetrode (� 3-5). Thus for a given crystal voltage or current more power output may be obtained than with the triode oscillator, or for a given output the crystal voltage will be lower, thereby reducing crystal heating. In addition, tank circuit tuning and loading react less on the crystal frequency because of the lower grid-plate capacity (� 3-3):
Fig. 404 shows atypical pentode or tetrode oscillator circuit. The pentode and tetrode tubes designed for audio power work are excellent crystal-oscillator tubes. The screen voltage is generally of the order of half the plate voltage for optimum operation. Small tubes rated at 250 volts for audio work may be
Xtal

� 4.4 CRYSTAL OSCILLATORS
Triode oscillators -- The triode crystal oscillator circuit (� 3-7) is showti in Fig. 403. The limit of plate voltage that can be used without endangering the crystal is about 250 volts. With the r.f. crystal current limited to a safe value of about 100 ma., the power output obtainable is about 5 watts. The oscillation frequency is dependent to some extent on the plate tank tuning because of the change in input capacity with changes in effective amplification (� 3-3).
Tetrode and pentode oscillators -- Since the power output of acrystal oscillator is limited by the permissible r.f. crystal current (� 4-3), it is advantageous to use an oscillator

-B

Fig. 403 -- Triode crystal oscillator. The tank con-
denser CImay be a100-ggfd. variable, with Li proportioned so that the tank will tune to the crystal frequency. C2 should be 0.001 �M. or larger. The grid leak, RI, will vary with the type of tube; high-g types take lower values, 2500 to 10,000 ohms, while medium and low-5 types take values of 10,000 to 25,000 ohms. Flashlight bulb or r.f. milliammeter (� 4-3) may be inserted at X.

--B

+SG.

Fig. 404 -- Tetrode or pentode crystal oscillator. Typical values: CI, 100 ispfd. with L wound to suit frequency; Ci, Cs,, 0.001 dd. or larger; Ca, 0.01 dd.; RI, 10,000 to 50,000 ohms, best value being determined by trial for the plate voltage and operating conditions chosen; R2, 250 to 400 ohms. 112 and Ca may be omitted, connecting cathode directly to ground, if plate voltage is limited to 250 volts. C5 (if needed) may be formed by
two metal plates about .1,4i inch square spaced about WI
inch. If the tube has a suppressor grid, it should be grounded. X indicates point where flashlight bulb may be inserted (� 4-3).

operated with 300 volts on the plate and 100-125 on the screen as crystal oscillators. The screen is at ground potential for r.f. and has no part in the operation of the circuit other than to set the operating characteristics of the tube. The larger beam tubes may be operated at 400 to 500 volts on the plate and 250 on the screen for maximum output.
Pentode oscillators operating at 250 to 300 volts will give 4or 5watts output under normal conditions. The beam type% 6L6 and 807 will give 15 watts or more at maximum plate voltage.
The grid-plate capacity may be too low to give sufficient feedback, particularly at the lower frequencies, in which case a feedback
condenser, C5, may be required. Its capacity should be the lowest value which will give stable oscillation.
Circuit constants-- Typical values for grid-leak resistance and by-pass condenser values are given in Figs. 403 and 404. Since the

CHAPTER FOUR

leacho-.7requency Power eneralion

PLATE CURRENT

Fig. 405 -- D.c. plate current vs. plate tuning capacity with the triode, tetrode or pentode crystal oscillator.

.MMMmmimms

Loaded
iC

Mar rew/vo ovwc/rr e�,
crystal is the frequency-determining element, the Q of the plate tank circuit has arelatively minor effect on the oscillator frequency. A Q of 12 (� 4-8) is satisfactory for average conditions, but departure from this figure will not greatly affect the performance of the oscillator.
Adjustment of crystal oscillators--The tuning characteristics and procedure to be followed in tuning are essentially the same for triode, tetrode or pentode crystal oscillators. Using a plate milliammeter as an indicator of oscillation (a 0-100 ma. d.c. meter will have ample range for all low-power oscillators), the plate current will be found to be steady when the circuit is in the non-oscillating state, but will dip when the plate condenser is tuned through resonance at the crystal frequency. Fig. 405 is typical of the behavior of plate current as the tank condenser capacity is varied. An r.f. indicator, such as a small neon bulb touched to the plate end of the tank coil, will show maximum at point A. However, when the oscillator is delivering power to aload it is best to operate in the region B-C, since the oscillator will be more stable and there is less likelihood that a slight change in loading will throw the circuit out of oscillation, which is likely to happen when operation is too near the critical point, A. The crystal current is lower in the B-C region.
When power is taken from the oscillator, the dip in plate current is less pronounced, as indicated by the dotted curve. The greater the power output the smaller the dip in plate current. If the load is made too great, oscillations will not start. Loading is adjusted by varying the coupling to the load circuit (� 2-11).
The greater the loading, the smaller the voltage fed back to the grid circuit for excitation purposes. This means that the r.f. voltage across the crystal also will be reduced, hence there is less crystal heating when the oscillator is delivering power than when operating unloaded.
Failure of a crystal circuit to oscillate may be caused by any of the following:
1. Dirty, chipped or fractured crystal 2. Imperfect or unclean holder surfaces 3. Too tight coupling to load 4. Plate tank circuit not tuning correctly 5. Insufficient feedback capacity Pierce oscillator -- This circuit is shown in Fig. 406. It is equivalent to the ultraudion cir-

cuit (� 3-7) with the crystal replacing the tuned circuit. The output of the Pierce oscillator is relatively small, although it has the advantage that no tuning controls are required.
The circuit requires capacitive coupling to a
following stage. The amount of feedback is de-. termined by the condenser C2. To sustain oscillation the net reactance (� 2-8) of the plate-cathode circuit must be capacitive; this condition is met so long as the inductance of the r.f. choke, together with the inductance of any coils associated with the input circuit
Fig. 406 -- Pierce oscillator circuit. Tubes such as the
6C5 and 6F6 are suitable, operating at plate voltages not exceeding 300 to prevent crystal fracture. When a triode is used, Ra and Ca are omitted. RIshould be 25,000 to 50,000 ohms. 1000 ohms is recommended for R. R3 is the screen voltage dropping resistance (75,000 ohms for the 6F6). Ci may have any value between 0.001 and 0.01 dd. C3 and C4 should be 0.01 pfd. the regeneration capacity, must be determined by experiment; usual values are between 50 and 150 ppfd. The capacity of Ca, usually 100 �dd., should be adjusted so that the oscillator is not overloaded.
of the following stage and the tube and stray capacities, forms a circuit tuned to a lower frequency than that of the crystal.
� 4-5 HARMONIC-GENERATING CRYSTAL OSCILLATORS
Tri-tet oscillator --The Tri-tet oscillator circuit is shown in Fig. 407. In this circuit the screen grid is operated at ground potential and the cathode at an r.f. potential above ground. The screen-grid acts as the anode of a triode crystal oscillator, while the plate or output circuit is tuned to the oscillator frequency or'for harmonic output, to amultiple of it.
Besides harmonic output, the Tri-tet circuit has the "buffering" feature of electron-coupling between crystal and output circuits (� 4-2). This makes the crystal frequency less susceptible to changes in loading or tuning and hence improves the stability.
If the output circuit is to be tuned to the same frequency as the crystal, a tube having low grid plate capacity (� 3-2, 3-5) must be used, otherwise there may be excessive feedback and danger of fracturing the crystal.
The cathode tank circuit, Lei, is not tuned to the frequency of the crystal, but to a con-

63 CHAPTER FOUR

5heRadio _Amateur'd

-B

+5uP +SG.. +8

oscillator, Fig. 408, the crystal is connected between grid and ground and the cathode tuned circuit C2RFC is tuned to a lower frequency than that of the crystal. This circuit gives high output on the fundamental crystal frequency with low crystal current. The output on even harmonics (2nd, 4th, etc.) is not as great as that obtainable wit the Tri-tet, but the out-
6V6G, 6L6,6L6G

Xtal

RFC
Is. HF-IH

NMA,-- R2

C5 C,

-8 +S G +8

Fig. 408 -- Grid-plate crystal oscillator circuit. In

(B)

the cathode circuit, RFC is a2.5-mh. r.f. choke. Other constants are the same as in Fig. 506. X indicates point

where crystal-current indicator may be inserted

(� 4-3).

-8

+SG. +B

Fig. 407 -- Tri -tot oscillator circuit, using pentodes (A) or beam tetrodes (B). Ci and C2, 200-gpfd. variable; C3, C4, Cs, Ce, 0.001 to 0.01 pfd., not critical; RI, 20,000 to 100,000 ohms; R2, 400 ohms for 400- or 500-volt operation.
Following specifications for cathode coils, Li, are based on acoil diameter of 1Y inches and length 1inch; turns should be spaced evenly to fill the required length. For 1.75-Mc. crystal, 32 turns; 3.5 Mc., 10 turns, 7 Me., 6turns. The screen should be operated at 250 volts or less. Audio beam tetrodes such as the 6L6 and 6L6G should be used only for second-harmonic output. Flashlight bulb may be inserted at X (� 4-3).
The L-C ratio in the plate tank, L2Ca, should be adjusted so that the capacity in use is 75 to 100 pad. for fundamental output and about 25 pmfd. for second harmonic output.

siderably higher frequency. Recommended values for L1 are given under the diagram. Ci should be set as near minimum capacity as is
consistent with good output. This reduces the crystal voltage.
With pentode-type tubes having separate suppressor connections, the suppressor may be connected directly to ground or may be operated at about 50 volts positive. The latter method will give somewhat higher output than with the suppressor connected to ground.
With transmitting pentodes or beam tubes operated at 500 volts on the plate an output of 15 watts can be obtained on the fundamental
and very nearly as much on the second harmonic.
Grid-plate oscillator -- In the grid-piate

put on odd harmonics (3rd, 5th, etc.) is appreciably better.
If harmonic output is not needed, C2 may be afixed capacity of 100 pedd. The cathode coil, RFC, may be a 2.5-mh. choke, since the in-
ductance is not critical. Output power of 15 to 20 watts may be ob-
tained at the crystal fundamental with atube such as the 6L6G at plate and screen voltages of 400 and 250, respectively.
Tuning and adjustment -- The tuning procedure for the Tri-tet oscillator is as follows: With the cathode tank condenser at about three-quarters scale, turn the plate tank condenser until there is a sharp dip in plate current, indicating that the plate circuit is in resonance. The crystal should be oscillating continuously regardless of the setting of the plate condenser. Set the plate condenser so
that plate current is minimum. The load circuit may then be coupled and adjusted so that the oscillator delivers power. The minimum plate current will rise; it may be necessary to retune the plate ccndenser when the load is coupled to bring the plate current to a new minimum. Fig. 409 shows the typical behavior of plate current with plate-condenser tuning.
After the plate circuit is adjusted and the oscillator is delivering power, the cathode
condenser should be readjusted to obtain optimum power output. The setting should be as far toward the low-capacity end of the scale
as is consistent with good output; it may, in

64 CHAPTER FOUR

kacho-recrency Power �eneration

Fig. 409 -- D.c. plate current vs. plate tuning capacity with the Tri-tet oscillator.

ccte, zoaded-C77.--
(Y

Ll�

an/oadeei

R
..����111e.
TUNING CAPACITY

fact, be desirable to sacrifice alittle output if so doing reduces the current through the crystal and thus reduces heating.
For harmonic output the plate tank circuit is tuned to the harmonic instead of the fundamental of the crystal frequency. A plate-current dip will occur at the harmonic. If the cathode condenser is adjusted for maximum output at the harmonic, this adjustment will usually serve for the fundamental as well. The crystal should be checked for evidence of ex-

cessive heating, the most effective remedy for
which is to lower the plate and/or screen voltage, or to reduce the loading. With this circuit maximum r.f. voltage across the crystal is developed at maximum load so crystal heating should be checked with the load coupled.
When a fixed cathode condenser is used in the grid-plate oscillator the plate tank circuit is simply resonated, as indicated by the platecurrent dip, to the fundamental or aharmonic of the output frequency, loading being adjusted to give optimum power output. If the variable cathode condenser is used, it should be set to give, by observation, the maximum power output consistent with safe crystal current. The variable condenser is chiefly useful in increasing the output on the third and higher harmonics; for fundamental operation the
cathode capacity is not critical and the fixed condenser may be used.

DRIVER

AMP

DRIVER

AMP

DRIVER

AMP

DRIVER

AMP.

AM P

DRIVER

AM. )

--C

(F)

-- C

Fig. 410 --Direct- or capacity-coupled driver and amplifier stages. Coupling condenser capacity may be from 50 gpfd. to 0.002 'lid., not critical except when tapping the coils for control of excitation is not possible. Parallel plate feed to the driver and series grid feed to the amplifier may be substituted in any of the circuits (i 3-7).

65 CHAPTER FOUR

Dh e leacho _Amateur'J fianargooh

� 44i INTERSTAGE COUPLING
Requirements -- The purpose of the interstage' coupling system is to transfer, with as little energy loss as possible, the power developed in the plate circuit of one tube (the driver)
to the grid circuit of the following amplifier tube or frequency multiplier. The circuits in practical use are based on the fundamental coupling arrangements described in �2-11. In the process of power transfer, impedance transformation (� 2-9) also is frequently necessary so that the proper exciting voltage and current will be available at the grid of the driven tube.
Capacity coupling -- Fig. 410 shows several types of capacitive coupling. In each case, C is the coupling condenser. The coupling condenser serves also as a blocking condenser
(� 2-13) to isolate the d.c. plate voltage of the driver from the grid of the amplifier. The circuits of C and D are preferable when a bal-
anced circuit is used in the output of the driver; instead of both tubes being in parallel across one side, the output capacity of the driver tube and the input capacity of the amplifier are across opposite sides of the tank circuit, thereby preserving abetter circuit bal-
ance. The circuits of E and F are designed for coupling to apush-pull stage.
In A, B, E and F, excitation is adjusted by moving the tap on the coil to provide an optimum impedance match. In E and F, the two grid taps should be maintained equidistant from the center-tap on the coil.
While capacitive coupling is simplest from the viewpoint of construction, it has certain disadvantages. The input capacity of the amplifier is shunted across at least aportion of the driver tank coil. When added to the output capacity of the driver tube, this additional capacity may be sufficient, in many cases, to
prevent use of adesirable L/C ratio in circuits for frequencies above about 7 Me.
Link coupling-- At the higher frequencies it is advantageous in reducing the effects of tube capacities on the L/C ratio to use separate tank circuits for the driver plate and amplifier grid, coupling the two circuits by means of a link (� 2-11). This method of coupling also has some constructional advantages, in that separate parts of the transmitter may be constructed as separate units without the necessity for running long leads at high r.f. potential.
Circuits for link coupling are shown in Fig. 411. The coupling ordinarily is by a turn or
two of wire closely coupled to the tank inductance at apoint of low r.f. potential such as the center of the coil of abalanced tank circuit, or
the "ground" end of the coil in asingle-ended circuit. The link line usually consists of two closely-spaced parallel wires; occasionally the wires are twisted together, but this usually causes undue losses at high frequencies.

DRIVER
(A)
DRIVER

AMP
1
AMP

(3)

DRIV ER

AMP

-c

Fig. 411 -- Link coupling between driver and amplifier.

It is advisable to have some means of varying the coupling between link and tank coils. The link coil may be arranged to be swung in
relation to the tank coil or, when it consists of alarge turn around the outside of the tank coil, can be split into two parts which can be pulled
apart or closed somewhat in the fashion of a pair of calipers. If the tank coils are wound on forms, the link may be wound close to the main coil.
With fixed coils, some adjustment of coupling can usually be obtained by varying the number of turns on the link. In general the proper number of turns for the link must be found by experiment.

� 4-7 R.F. POWER AMPLIFIER CIRCUITS
Tetrode and pentode amplifiers -- When the input and output circuits of an r.f. amplifier tube are tuned to the same frequency, it will oscillate as atuned-grid tuned-plate oscillator unless some means is provided to eliminate the effects of feedback through the plateto-grid capacity of the tube (� 3-5). In all transmitting r.f. tetrodes and pentodes, this capacity is reduced to asatisfactory degree by the internal shielding between grid and plate provided by the screen. Tetrodes and pentodes

66 CHAPTER FOUR

Pailio-Drequency Power �eneralion

designed for audio use (such as the 6L6, 6V6, 6F6, etc.) are not sufficiently well screened for use as r.f. amplifiers without employing additional means for nullifying the effect of the
grid-plate capacity. Typical circuits of tetrode and pentode r.f.
amplifiers are shown in Fig. 412. The high
power sensitivity (� 3-3) of pentodes and tetrodes, however, makes them prone to selfoscillate with very small values of feedback voltage, so that particular care must be used to. prevent feedback by means external to the tube itself. This calls for adequate isolation of plate and grid tank circuits to prevent undesired magnetic or capacity coupling between them. The requisite isolation can be secured by keeping the circuits well separated and mounting the coils so that magnetic coupling is minimized, or by shielding (� 2-11).
Triode amplifiers -- The feedback through the grid-plate capacity of a triode cannot be eliminated in the tube itself, and therefore special circuit means, called neutralization, must be used to prevent oscillation. A prop-
erly-neutralized triode amplifier then behaves

C2 Z

(A)

ciSricausits
enri *SCREEN *SUM VOLTAGE VOLT
SINGLE -TUBE OR PARALLEL

Output

C2 C3 C2

(B) 0747t

RFC

+SUP V.

+5V03

H V

PUSH-PULL Fig. 412-- Tetrode-pentode r.f. amplifier circuits. Ci -- 0.01 pfd.; C2 -- 0.001 dd. or larger; Cs-L -- see �4-8. In circuits for tetrodes, the suppressor-grid connection and by-pass condenser are omitted.

as though it were operating at very low frequencies where the grid-plate capacity feedback is negligible (� 3-3).
Neutralization -- Neutralization amounts to taking some of the radio-frequency current from the output or input circuit of the amplifier and introducing it into the other circuit in such a way that it effectively cancels the current flowing through the grid-plate capacity of the tube, thus rendering it impossible for the tube to supply its own excitation. For complete neutralization it is necessary that the two currents be opposite in phase (� 2-7) and equal in amplitude.
The out-of-phase current (or voltage) can be obtained quite readily by using a balanced tank circuit in either grid or plate, taking the
neutralizing voltage from the end of the tank opposite that to which the grid or plate is connected. The amplitude of the neutralizing voltage can be regulated by means of a small condenser, th neutralizing condenser, having the same order of capacity as the grid-plate
capacity of the tube. Circuits in which the neutralizing voltage is obtained from a balanced grid tank and fed to the plate through
the neutralizing condenser are termed gridneutralized circuits, while if the neutralizing voltage is obtained from abalanced plate tank and fed to the grid of the tube, the circuit is plate-neutralized.
Plate-neutralized circuits -- The circuits for plate neutralization are shown in Fig. 413 at A, B and C. In A, voltage induced in the extension of the tank coil is fed back to the grid through the neutralizing condenser C. to balance the voltage appearing between grid and plate. In this circuit the capacity required at C. increases as the tank coil extension is made smaller; in general, neutralization is satisfactory over only asmall range of frequencies since the coupling between the two sections of
the tank coil will vary with the amount of capacity in use at C.
In B the tank coil is center-tapped to give
equal voltages on either side of the center tap, the tank condenser being across the whole coil. The neutralizing capacity is approximately equal to the grid-plate capacity of the tube in this case. A disadvantage of the circuit, when used with the single tank condenser shown, is that the rotor of the condenser is above ground potential and hence small capacity changes caused by bringing the hand near the tuning control (hand capacity) cause detuning. In general, neutralization is complete at only one frequency since the plate-cathode capacity of the tube is across only half the tank coil; also, it is difficult to secure an exact center-tap. Both these cause unbalance which in turn causes the voltages across the two halves of the coil to differ when the frequency is changed.

67 CHAPTER FOUR

Dne

Analeur'� ..flancilooh

etOuplinq Pu. ' 8.1 Ceruit3

DRIVER

AMPLIFIER
C,

DRIVER

AMPLIFIER

t.
Co. RFC
Ct

Etput

rni f. C,

�MY

Output

ri � �
E

*...,�

�SIAS� f

Fig. 413 -- Triode amplifier circuits. Plate neutralization is shown in A, B and C; D, E and F show types of

grid neutralization. Either capacitive or link coupling may be used with circuits of A, B or C.

C-L -- See �4-8.

Ci -- 0.01 dd.

Co-Lo -- Grid tank circuit.

C2 -- 0.001 mfd. or larger.

Cm -- Neutralizing condensers.

The circuit of C also uses a center-tapped tank circuit, the voltage division being secured by use of a balanced (split-stator) tank condenser, the two condenser sections being identical. C. is approximately equal to the gridplate capacity of the tube. In this circuit the upper section of the tank condenser is in parallel with the output capacity of the tube, hence the circuit can be completely neutralized at only one setting of the tank condenser unless a compensating capacity (Fig. 414) is connected across the lower section. In practice, if the capacity in use in the tank circuit is large compared to the plate-cathode capacity the unbalancing effect is not serious.
Grid-neutralized circuits-- Typical circuits employing grid neutralization are shown
Fig. 414 -- Compensating for capacity unbalance in
the single-tube neutralizing circuit. C., the balancing
capacity, should be variable and should have a maximum capacity somewhat larger than the output capacity of the tube. It is adjusted to minimize shift in neutralizing capacity at C. as the frequency is changed.

in Fig. 413 at D, E and F. The principle of balancing out the feed-back voltage is the same as in plate neutralization. However, in these circuits the fed-back voltage may either be in phase or out of phase with the excitation voltage on the grid side of the input tank circuit
(and the opposite on the other side) depending upon whether the tank is divided by means of a balanced condenser or a tapped coil. Circuits such as those at D and E neutralized by ordinary procedure (described below) will be regenerative when the plate voltage is applied; the circuit at F will be degenerative. In addition, the normal unbalancing effects described in the preceding paragraph are present, so that grid neutralizing is less satisfactory than the
plate method. Inductive neutralizatIon-- With this type
of neutralization inductive coupling between the grid and plate circuits is provided in such a
way that the voltage induced in the grid coil by magn�tic coupling from the plate coil opposes the voltage fed back through the grid-plate capacity of the tube. A representative circuit arrangement, using acoupling link to provide
the mutual inductance (� 2-11) is shown in Fig. 415. Ordinary inductive coupling between the two coils also could be used, but is less convenient. Inductive neutralization is complete at only one frequency, since the effective mutual inductance changes to some extent with tuning, but is useful in cases where the grid-
plate capacity of the tube being neutralized is

68 CHAPTER FOUR

Paclio-Drequency Power generalion

Fig. 415--Inductive neutralizing circuit. The link coils should have one or two turns and should be coupled to the grounded ends of the tank coils. Neutralization is adjusted by moving the link coils in relation to the tank coils. Reversal of connections to one of the coils may be required to obtain the proper phasing.
very small and suitable circuit balance cannot be obtained with circuits using neutralizing condensers.
Push-pull neutralization-- With pushpull circuits two neutralizing condensers are used as shown in Fig. 416. In these circuits the grid-plate capacities of the tubes and the neu-
Output
Fig. 416--Push-pull triode amplifier circuits with "cross-neutralization." Either capacitive or link coupling may be used. C-L -- See �4-8. Co -- Neutralizing condensers.
-- 0.01 dd.
C.2 -0.001 ufd. or larger.

tralizing capacities form a capacity bridge (� 2-11) which is independent of the grid and plate tank circuits. The neutralizing capacities are approximately the same as the tube gridplate capacities. With electrically similar tubes and symmetrical construction (stray capacities to ground equal on both sides of the circuit) the neutralization is complete and independent
of frequency. A circuit using a balanced condenser, as at B, is preferred since it is an aid in obtaining good circuit balance.
Frequency effects-- The effects of slight dissymmetry in a neutralized circuit become more important as the frequency is raised, and may be sufficient at ultra-high frequencies (or even lower) to prevent good neutralization. At these frequencies the inductances and stray capacities of even short leads become important elements in the circuit, while input loading effects (� 7-6) may make it impossible to get proper phasing, particularly in single-tube circuits. In such cases the use of a push-pull amplifier, with its general freedom from the
effects of dissymmetry, is not only much to be preferred but may be the only type of circuit which can be satisfactorily neutralized.
Neutralizing condensers--In most cases
the neutralizing voltage will be equal to the r.f. voltage between the plate and grid of the
tube so that for perfect balance the capacity required in the neutralizing condenser theoret-
ically will be equal to the grid-plate capacity. If, in the circuits having tapped tank coils, the tap is more than half the total number of turns
from the plate end of the coil, the required neutralizing capacity will increase approximately
in proportion to the relative number of turns in the two sections of the coil.
With tubes having grid and plate connections brought out through the bulb, acondenser having at about half-scale or less acapacity equal to the grid-plate capacity of the tube should be chosen. If the grid and plate leads are brought through acommon base, the capacity needed is greater because the tube
socket and its associated wiring adds some capacity to the actual inter-element capacities.
When two or more tubes are connected in parallel, the neutralizing capacity required will be in proportion to the number of tubes.
The voltage rating of neutralizing con-
densers must at least equal the r.f. voltage across the condenser plus the sum of the d.c. plate voltage and the grid-bias voltage.
Neutralizing procedure -- The procedure in neutralizing is essentially the same for all tubes and circuits. The filament of the tube should be lighted and the excitation from the preceding stage should be fed to the grid circuit. There should be no plate voltage on the amplifier.
The grid-circuit milliammeter makes agood

69 CHAPTER FOUR

..7ne leacho Amaieur'es ilancgooh

neutralizing indicator. If the circuit is not completely neutralized, tuning of the plate tank circuit through resonance will change the tuning of the grid circuit and affect its loading, causing a change in the rectified d.c. grid current. The setting of the neutralizing condenser which leaves the grid current unaffected as the plate tank is tuned through resonance is the correct
one. If the circuit is out of neutralization, the grid current will drop perceptibly as the plate tank is tuned through resonance. As the point of neutralization is approached, by adjusting the neutralizing capacity in small steps, the dip in grid current as the plate condenser is swung through resonance will become less and less pronounced until, at exact neutralization, there will be no dip at all. Further change of the neutralizing capacity in the same direction will bring the grid-current dip back. The neutralizing condenser should always be adjusted with a screwdriver of insulating material to avoid
hand-capacity effects. Adjustment of the neutralizing condenser
may affect the tuning of the grid tank or driver plate tank, so both circuits should be retuned each time a change is made in neutralizing capacity. In neutralizing apush-pull amplifier, the neutralizing condensers should be adjusted together, step by step, keeping their capacities
as equal as possible. With single-ended circuits having split-stator
neutralizing, the behavior of the grid meter will depend somewhat upon the type of tube used. If the tube output capacity is not great enough to upset the balance, the action of the meter will be the same as in other circuits. With high-capacity tubes, however, the meter
usually will show agradual rise and fall as the plate tank is tuned through resonance, reaching a maximum right at resonance when the circuit is properly neutralized.
When an amplifier is not neutralized, aneon bulb touched to the plate of the amplifier tube
or to the plate side of the tuning condenser will glow when the tank circuit is tuned through resonance, providing the driver has sufficient power. The glow will disappear when
the amplifier is neutralized. However, touching the neon bulb to such
an ungrounded point in the circuit may intro-

duce enough stray capacity to unbalance the
circuit slightly, thus upsetting the neutralizing. A flashlight bulb connected in series with a
single-turn loop of wire 2% or 3 inches in diameter, with the loop coupled to the tank coil, will also serve as aneutralizing indicator. Capacitive unbalance can be avoided by coupling the loop to the low-potential part of
the tank coil. Incomplete neutralization -- If a setting
of the neutralizing condenser can be found
which gives minimum r.f. current in the plate tank circuit without completely eliminating it,
there may be magnetic or capacity coupling between the input and output circuits external to the tube itself. Short leads in neutralizing circuits are highly desirable, and the input and output inductances should be so placed with respect to each other that magnetic
coupling is minimized. Usually this requires that the axes of the coils should be at right angles to each other. In some cases it may be necessary to shield the input and output circuits from each other. Magnetic coupling can
be detected by disconnecting the plate tank from the remainder of the circuit and testing for r.f. in it (by means of the flashlight lamp and loop) as the tank condenser is tuned through resonance. The driver stage must be operating, of course.
With single-ended amplifiers there are many
stray capacities left uncompensated for in the neutralizing process. With large tubes, especially those having relatively high interelec-
trode capacities, these commonly neglected stray capacities can prevent perfect neutralization. Symmetrical arrangement of apush-pull amplifier is about the only way to obtain
practically perfect balance throughout the amplifier.
The neutralization of tubes with extremely low grid-plate capacity, such as the 6L6, is often difficult, since it frequently happens that the wiring itself will introduce sufficient capacity between the right points to "over-
neutralize" the grid-plate capacity. The use of a neutralizing condenser only aggravates the condition. Inductive or link neutralization as shown in Fig. 415 has been used successfully with such tubes.

Fig. 417 -- Inverted amplifier. The number of turns at L should be adjusted by experiment t. give optimum grid excitation to the amplifier. By-pass condenser C may be 0.001 pfd. or larger.

70 CIIAPTER FOUR

kaclio--7requency Power generation

The inverted amplifier --The circuit of Fig. 417 avoids the necessity for neutralization
by operating the control grid of the tube at ground potential, thus making it serve as a
shield between the input and output circuits. It is particularly useful with tubes of low
grid-plate capacity which are difficult to neutralize by ordinary methods. Excitation is applied between grid and cathode through the coupling coil L; since this coil is common to both the plate and grid circuits the amplifier is degenerative with the circuit constants normally used, hence more excitation voltage and power are required for agiven output than is the case with a neutralized amplifier. The tube used must have low plate-cathode ca-
pacity (of the order of 1 add. or less) since larger values will give sufficient feedback to
permit it to oscillate, the circuit then becoming the ultraudion (� 3-7). Tubes having sufficiently low plate-cathode capacity (audio
pentodes, for example) can be used without danger of oscillation at frequencies up to 30 Mc. or so.

sistance should be relatively high, but if only
limited excitation voltage is available greater power output will be secured by using alower value of load resistance. The latter adjustment is accompanied by a decrease in plate efficiency. The optimum load resistance is that which, for the maximum permissible peak

4 t
Voltaye

Max. /InnsstdtuarnttoarnieeoouussGPrliadtVeoltace94
Zero Bias

eie:1 1:;.9

cut-off Bias
R.F Grid Volta/id

-- 4tieo-operattag

� 4-8 POWER AMPLIFIER OPERATION
Efficiency -- An r.f. power amplifier is usually operated Class-C (� 3-4) to obtain a reasonably-high value of plate efficiency (� 3-3). The higher the plate efficiency the higher the power input that can be applied to
the tube without exceeding the plate dissipation rating (� 3-2), up to the limits of other tube ratings (plate voltage and plate current).
Plate efficiencies of the order of 75% are readily obtainable at frequencies up to the 30-60 megacycle region. The overall efficiency of the amplifier will be lower by the percentage of
power lost in the tank and coupling circuits, so that the actual efficiency is less than the plate efficiency.
Operating angle -- The operating angle is the proportionate part of the exciting gridvoltage cycle (� 2-7) during which plate current flows, as shown in Fig. 418. For Class-C operation it is usually in the vicinity of 120-
150 degrees which, with other operating considerations, results in an optimum relationship between plate efficiency and grid driving power.
Load impedance -- The load impedance (� 3-3) for an r.f. power amplifier is adjusted, by tuning the plate tank circuit to resonance, to represent apure resistance at the operating
frequency (� 2-10). Its value, which is usually in the neighborhood of afew thousand ohms, is adjusted by varying the loading on the tank circuit, closer coupling to the load giving lower values of load resistance and vice versa (� 2-11). The load may be either the grid circuit of afollowing stage or the antenna circuit.
For highest efficiency the value of load re-

Fig. 418 -- Instantaneous voltages and currents in aClata-C amplifier operating under optimum conditions.
plate current, causes the minimum instantaneous plate voltage (Fig. 418) to be equal to the maximum instantaneous grid eoltage required to cause the peak plate current to flow; this gives the optimum ratio of plate efficiency to required grid driving power.
R.f. grid voltage and grid bias -- For most tubes optimum operating conditions result when the minimum instantaneous plate voltage is 10% to 20% of the d.c. plate voltage, so that the maximum instantaneous positive grid voltage must be approximately the same figure.
Since plate current starts flowing when the instantaneous voltage reaches the cut-off value (� 3-2), the d.c. grid voltage must be considerably higher than cut-off to confine the operating angle to 150 degrees or less (with grid bias at cut-off the angle would be 180 degrees). For an angle of 120 degrees the r.f. grid voltage must reach 50% of its peak value (� 2-7) at the cut-off point. The corresponding figure for an angle of 150 degrees is 25%. Hence the operating bias required is the cut-off value plus 25% to 50% of the peak r.f. grid voltage. These relations are shown in Fig. 418. The grid bias should be at least twice cut-off if the amplifier is to be plate modulated so that the operating angle will not be less than 180 degrees when the plate voltage rises to twice the steady d.c. value (� 5-3). Because of their relatively high amplification factors, with most modern tubes Class-C operation requires considerably more than twice cut-off bias to make the operating angle fall in the region mentioned above.

71 CHAPTER FOUR

5h.e Radio -AmaleuA fiandtooi

Suitable operating conditions are usually
given in the data accompanying the type of
tube used. Grid bias may be secured either from abias
source (fixed bias), a grid leak (� 3-6) of suitable value, or from a combination of both. When abias supply is used, its voltage regulation should be taken into consideration (� 8-9).
Driving power -- As indicated in Fig. 418,
grid current flows only during asmall portion of the peak of the r.f. grid voltage cycle. The
power consumed in the grid circuit is therefore approximately equal to the peak r.f. grid voltage multiplied by the average rectified grid current as read by a d.c. milliammeter. The peak r.f. grid voltage, if not included in the
tube manufacturer's operating data, can be estimated roughly by adding 10% to 20% of
the plate voltage to the operating grid bias, assuming the operating conditions are as de-
scribed above. , At frequencies up to 30 Mc. or so the grid
losses are practically entirely those resulting from grid-current flow. At ultra-high frequencies, however, dielectric losses in the glass envelope and base materials become appreciable, together with losses caused by transittime effects (� 7-6), and may necessitate
supplying several times the driving power indicated above. At any frequency, the driving stage should be capable of a power output two to three times the power it is expected the grid circuit of the amplifier will consume. This
is necessary because losses in the tank and coupling circuits must also be supplied, and also to provide reasonably good regulation of
the r.f. grid voltage. Good voltage regulation (see �8-1 for general definition) insures that the waveform of the excitation voltage will not
be distorted because of the changing load on
the driver during the r.f. cycle. Grid impedance-- During most of the r.f.
grid voltage cycle, no grid current flows, as indicated in Fig. 418, hence the grid impedance is infinite. During the peak of the cycle, however, the impedance may drop to very low values (of the order of 1000 ohms) depending upon the type of tube. Both the minimum and average values of grid impedance depend to a considerable extent on the amplification factor
of the tube, being lower with tubes having large amplification factors.
The average grid impedance is equal to E'2/P, where E is the r.m.s. (� 2-7) value of r.f. grid voltage and P the grid driving power. Under optimum operating conditions values of
average grid impedance ranging from 2000 ohms for high-ti tubes to four or five times as much
for low-g types are representative. Values in the vicinity of 4000 to 5000 ohms are typical of modern triodes with amplification factors
of 20 to 30.

Because of the large change in impedance
during the cycle it is necessary that the tank circuit associated with the amplifier grid have fairly high Q so that the voltage regulation over the cycle will be good. The requisite Q may be obtained by adjusting the L/C ratio or by tapping the grid circuit across only part of the tank (� 4-6).
Tank circuit Q --Besides serving as a means for transforming the actual load resist-
ance to the required value of plate load impedance for the tube, the plate tank circuit also should suppress the harmonics present in the tube output as aresult of the non-sinusoidal plate current (� 2-7, 3-3). For satisfactory harmonic suppression aQ of 12 or more (with the circuit fully loaded) is desirable. A Q of this
order is also helpful from the standpoint of securing adequate coupling to the load or antenna circuit (� 2-11). The proper Q can be obtained by suitable selection of LIC ratio in
relation to the optimum plate load resistance for the tube (� 2-10).
For aClass-C amplifier operated under optimum conditions as described above, the plate load impedance is approximately proportional to the ratio of d.c. plate voltage to d.c. plate current. For a given effective Q, the tank capacity required at a given frequency will be inversely proportional to the parallel resistance (� 2-10), so that it will also be inversely proportional to the plate-voltage/plate-current ratio. The capacity required on various amateur bands for aQ of 12 is shown in Fig. 419 as afunction of this ratio. The capacity given is for single-ended tank circuits as shown in Fig. 420 at A and B. When abalanced tank circuit is used, the total tank capacity required is re-
duced to Yi this value because the tube is connected across only half the circuit (� 2-9). Thus if the plate-voltage/plate-current ratio calls for acapacity of 200 �dd. in asingle-ended circuit
at the desired frequency, only 50 gad. would be needed in abalanced circuit. If asplit-stator or balanced tank condenser is used, each section should have a capacity of 100 ��fd., the total capacity of the two in series being 50 lapfd. These are "in use" capacities, not simply the rated maximum capacity of the condenser. Larger values may be used with an increase in
the effective Q. To reduce energy loss in the tank circuit the
inherent Q of the coil and condenser should be high. Since transmitting coils usually have Q's ranging from 100 to several hundred, the tank transfer efficiency is generally 90% or more. An unduly , large CIL ratio is not advisable
since it will result in large circulating r.f. tank current and hence relatively large losses in the
tank, with aconsequent reduction in the power available for the load.
Tank constants-- When the capacity nec-

72 CHAPTER FOUR

kacho-.7requency Power generation

900 800 700 600 500 400
300
200
150 i k' 00
1`,a9o0 :70 i60 ' i50
40
30
20

e.,,,,,,

c e

4,o

10 9 a

3
RATIO

� 4 5 6 7 8 9 10
PLATE VOLTS
PLATE MA.

15 20

essary for aQ of 12 has been determined from Fig. 419, the inductance required to resonate at the given frequency can be found by means

Fig. 419 -- Chart showing tank-capacities required for "Q" of 12 with various ratios of plate voltage to plate current for various frequencies. In circuits F, G, H (Fig. 420), the capacities shown in the graph maybe divided by four. In circuits C, D, E, I, J and K, the capacity of each section of the split-stator cond enser may be one-half that shown by the graph. Values given by the graph should be used for circuits A and B.
of the formula in �2-10. Alternatively, the required number of turns on coils of various construction can be found from the charts of Figs. 421 and 422.
Fig. 421 is for coils wound on receiving-type forms having a diameter of 1Y� inches and ceramic forms having adiameter of 194 inches and winding length of 3 inches. Such coils would be suitable for oscillator and buffer stages where the power is not over 50 watts. In all cases the number of turns given must be wound to fit the length indicated and the turns should be evenly spaced.
Fig. 422 gives data on coils wound on transmitting-type ceramic forms. In the case of the smallest form, extra curves are given for double-spacing (winding turns in alternate grooves). This is sometimes advisable in the case of 14- and 28-Me, coils when only a few turns are required. In all other cases it is assumed that the specified number of turns is wound in the grooves without any additional spacing.
Ratings of components -- The peak voltage to be expected between the plates of a tank condenser depends upon the arrangement of the tank circuit as well as the d.c. plate voltage. Peak voltage may be determined from Fig. 420, which shows all of the comnionly used tankcircuit arrangements. These estimates assume

0 002

0002

OHO
A
0 002

D
0002

40V

+11V

41�

*N.V.

/Ito

.5he Radio Aniateur'd ilandtooh

SOO

.1M

400 .11.1 MI.I1M. II1M1 B1.11!1 --'1111M 11IMR EI7M 2=I MIU MMIM MI.M

350

300

mumouNnomminieen 250
200

IMMUMUNIMIMMIIMMeMan

150 1111110111k1111M111111\11

E7. MOM MM.M

In. 11 �Knm

� 40 uusmmiI-MrIxIuIemI.mMIriWnmKasIL`aImnIaIo.MlMmoI.Mw.Ws�.IuwMmaIswsoum1m0m1,1.o1`0mmMmlM oimmMimsWiw1ve10e1efkf4lll

SO 40

M umEmM umIiUkM uuUmM mM okIwN uf ~kfil emiiBmO mxX`I mM asMsI s

mummvusimoemwammui k,2

35 30

mtumwmnommtknmmtiomminawvwInnni

� 25

".4 IS

633 4 .5 6 7 8 9 10

15 20 25 30 35 40 SO 60 70 SO

NUMBER OF TURNS ON COIL

Fig. 421 -- Coil-winding data for recei v in g-t ype
forms, diameter 1% inches. Curve A -- winding length , 1 inch; Curve B -- winding length, PA inches; Curve
C -- winding length, 2 inches. Curve C is also suitable for coils wound on 134-inch diameter ceramic forms
with 3 inches of winding length.

leacho-.7requency Power �eneralion

amplifier grid tank. Its Q can be increased to a suitable value by adjustment of the L/C ratio or by tapping the load across part of the coil (� 2-10).
Whatever the type of coupling, apreliminary adjustment should be made with the proper

principally determined by the screen rather than the plate voltage.
With reasonably efficient operating conditions, the minimum plate current with the amplifier unloaded will be a small fraction of

bias voltage and/or grid leak, but with the

amplifier plate voltage off; then the amplifier should be carefully neutralized. After neutralization, the driver-amplifier coupling should be readjusted for optimum power transfer, after
which plate voltage may be applied and the

Unloaded

Loaded

Fig. 423 -- Typical behavior of d.c. plate current with tuning of an amplifier.

amplifier plate circuit adjusted to resonance

and coupled to its load. Under actual operating

TUNING CAPACITY

conditions the grid current decreases below the the rated plate current for the tube (usually a

value obtained without plate voltage on the fifth or less) since with no load the parallel amplifier and the effective grid impedance impedance of the tank circuit is high. If the ex-

rises, hence the final adjustment is to recheck the coupling to take care of this shift.

citation is low, the "dip" will not be very marked, but with adequate excitation the

With recommended bias, the grid current obtained before plate voltage is applied to the amplifier should be 25% to 30% higher than the value required for operating conditions.
If this value is not obtained, and the driver
plate input is up to rated value, the reason may

plate current at resonance without loading will be just high enough so that the d.c. plate
power input supplies all the losses in the tube and circuit. As an indication of probable efficiency, the minimum plate current value should not be taken too seriously, because

be either improper matching of the amplifier without load the Q of the circuit is high and grid to the driver plate or simply insufficient the tank current relatively large. When the

power output from the driver to take care of all losses. Driver operating voltages should be checked to assure they are up to rated values. If batteries are used for bias and are not strictly

amplifier is delivering power to a bad, the
circulating current drops considerably and the tank losses correspondingly decrease. High minimum unloaded plate current is chiefly

fresh, they should be replaced, since batteries which have been in use for some time often develop high internal resistance which effectively acts as additional grid-leak resistance.

encountered at 28 Mc. and above, where tank losses are higher and the tank L/C ratio is usually lower than ndrmal because of irreducible tube capacities. The effect is particu-

If a rectified a.c. bias supply is used, the larly noticeable with screen-grid tubes which

bleeder or voltage-divider resistances should have relatively high output capacity. Because

be checked to make certain that low grid current is not caused by greater grid-circuit resistance than is recommended. In this connec-

of the decrease in tank r.f. current with loading, however, the actual efficiency under load is reasonably good.

tion it is helpful to measure the actual bias when grid current is flowing, by means of a high-resistance d.c. voltmeter. There is also

With the load (antenna or following amplifier grid circuit) connected, the coupling between plate tank and load should be adjusted to make

the possibility of loss of filament emission of the tube take rated plate current, keeping the

the amplifier tube either from prolonged serv- tank always tuned to resonance. As the output

ice or from operating the filament under or coupling is increased, the minimum prlate cur-

over the rated voltage.

rent will also increase about as shown in Fig.

Plate tuning -- In preliminary tuning, it is 423. Simultaneously, the tuning becomes less

desirable to use low plate voltage to avoid sharp, because of the increase in effective re-

possible damage to the tube. With excitation and plate voltage applied, rotate the plate tank
condenser until the plate current dips, then set the condenser at the minimum plate-current point (resonance). When the resonance point

sistance of the tank. If the load circuit simulates aresistance, the resonance setting of the tank condenser will be practically unchanged with loading; this is generally the case since the load circuit itself usually is also tuned to

has been found, the plate voltage may be increased to its normal value.
With adequate excitation, the off-resonance
plate current of atriode amplifier may be two or more times the normal operating value.

resonance. A reactive load (such as an antenna or feeder system which is not tuned exactly to resonance) may cause the tank condenser setting to change appreciably with loading since reactance as well as resistance is coupled

With screen-grid tubes, the off-resonance plate into the tank (� 2-11).

current may not be much higher than the normal operating value since the plate current is

Power output -- As a check on the operation of an amplifier, its power output may be

75 CHAPTER FOUR

Dh e Radio Anzuieur'.5 ...fluncgooL

measured by the use of a load of known resistance coupled to the amplifier output as shown in Fig. 424. At A a thermoammeter M and non-inductive (ordinary wire-wound resistors are not satisfactory) resistance R are connected across acoil of afew turns coupled to the amplifier tank coil. The higher the resistance of R, the greater the number of turns required in the coupling coil. A resistor used in this way is generally called a "dummy antenna," since its use permits the transmitter to be adjusted without actually radiating power. The loading may readily be adjusted by varying the coupling between the two coils, so that the amplifier draws rated plate current when tuned to resonance. The power output is then calculated from Ohm's Law:
P (watts) = PR where Iis the current indicated by the thermoammeter and R is the resistance of the noninductive resistor R. Special resistance units are available for this purpose ranging from 73 to 600 ohms (simulating antenna and transmission-line impedances) at power ratings up to 100 watts. For higher powers, the units may be connected in series-parallel. The meter scale required for any expected value of power output may also be determined from Ohm's Law:
I=
Incandescent light bulbs can be used to replace the special resistor and thermoammeter.
Tank Circuit
(A)
Tank Czrcud
(B)
Tank Circuit
(C)
Fig. 424 -- "Dummy antenna" circuits for checking power output and making operating adjustments without applying power to the actual antenna.

The lamp should be equipped with a pair of leads, preferably soldered to the terminals on the lamp base. The coupling should be varied until the greatest brilliance is obtained for a given plate input. In using lamps� as dummy antennas, asize corresponding to the expected power output should be selected so that the lamp will operate near its normal brilliancy.
Then when the adjustments have been completed an approximation of the power output can be obtained by comparing the brightness of the lamp with the brightness of one of similar power rating in a 115-volt socket.
The circuit of Fig. 424-B is for resistors or lamps of relatively high resistance. In using this circuit, care should be taken to avoid accidental contact with the plate tank when the power is on. This danger is avoided by circuit C, in which aseparate tank circuit, LC, tuned to the operating frequency, is coupled to the plate tank circuit. The loading is adjusted by varying the number of turns across which the dummy antenna is connected on L and by changing the coupling between the two coils. With push-pull amplifiers, the dummy antenna should be tapped equally on either side of the center of the tank, when Fig. 424-B is used.
Harmonic suppression -- The most important step to take in elimination of harmonic radiation (� 4-8, 2-12) is to use an output tank circuit having aQ of 12 or more. Beyond this,
it is desirable to avoid any considerable amount of over-excitation of aClass-C amplifier, since excitation in excess of that required for normal Class-C operation further distorts the platecurrent pulse and increases the harmonic content in the output of the amplifier even though
the proper tank Q is used. If the antenna system will accept harmonic frequencies they will be radiated when present, consequently the antenna coupling system preferably should be selected with harmonic transfer in mind (� 10-6).
Harmonic content can be reduced to some extent by preventing distortion of the r.f. grid voltage waveshape. This can be done by using a grid tank circuit with high effective Q. Link coupling between the driver and final amplifier are helpful, since the two tank circuits provide more attenuation than one at the harmonic frequencies. However, the advantages of link coupling in this respect may be nullified wiless the Q of the grid tank is high enough to give good voltage regulation and thus prevent distortion in the grid circuit.
The stray capacity between the antenna coupling coil and the tank coil may be sufficient to couple harmonic energy into the antenna system. This coupling may be eliminated by the use of electrostatic shielding (Faraday shield) between the two coils. Fig. 425 shows the construction of such ashield, while Fig. 426
illustrates the manner in which it is installed.

76 CHAPTER FOUR

Pacho-.7requency Power Generation

The construction shown in Fig. 425 is used to prevent current flow in the shield, which would occur if the wires formed closed circuits since the shield is in the magnetic field of the tank coil. Should this occur there would be magnetic shielding as well as electrostatic; in addition, there would be an undesirable power loss in the shield.
Improper operation-- Inexact neutralization or stray coupling between plate and grid circuits may result in regeneration. This effect is most evident with low excitation, when the amplifier will show asudden increase in output when the plate tank circuit is tuned slightly to the high-frequency side of resonance. It is accompanied by a pronounced increase in grid current.
Self-oscillation is apt to occur with tubes of high power sensitivity such as the r.f. pentodes
Wo Connections here )
jCoonidnuecdthoerrse}
Fig. 425 -- The Faraday shield. It is made of parallel conductors, insulated from each other except at one end where all are joined. Stiff wire or small diameter rod may be used, spaced about the diameter of the wire or rod.
and tetrodes. In event of either regeneration or oscillation, circuit components should be arranged so that those in the plate circuit are well isolated from those of the grid circuit. Plate and grid leads should be made as short as possible and the screen should be by-passed as close to the socket terminal as possible. A cylindrical shield surrounding the lower portion of the tube up to the lower edge of the plate is sometimes required.
"Double resonance" or two tuning spots on the plate tank condenser, one giving minimum plate current and the other maximum power output, may occur when the tank circuit Q is too low (� 2-10). A similar effect also occurs at times with screen-grid amplifiers when the

screen voltage regulation (� 8-1) is poor, as when the screen is supplied through adropping resistor. The screen voltage decreases with an increase in plate current, because the screen current increases under the same conditions. Thus the minimum plate current point causes the screen voltage, and hence power output, to be less than when a slightly higher plate current is drawn.
A phenomenon known as "grid emission" may occur when the amplifier tube is operated at higher than rated power dissipation on either the plate or grid. It is particularly likely to occur with tubes having oxide-coated cathodes such as the indirectly-heated types. It is caused by the grid reaching atemperature high enough to cause electron emission (� 2-4). The electrons so emitted are attracted to the plate, further increasing the power input and heating, so that grid emission is characterized by gradually increasing plate current and heat which eventually will ruin the tube if the power is not removed. Grid emission can be prevented by operating the tube within its ratings.
�4-10 PARASITIC OSCILLATIONS
Description -- If the circuit conditions in an oscillator or amplifier are such that self- � oscillation at some frequency other than that desired exist, the spurious oscillation is termed parasitic. The energy required to maintain a parasitic oscillation is wasted so far as useful output is concerned, hence an oscillator or amplifier having parasitics will operate at reduced efficiency. In addition, its behavior at the operating frequency often will be erratic. Parasitic oscillations may be higher or lower in frequency than the operating frequency of the amplifier.
The parasitic oscillation usually starts the instant plate voltage is applied or, when the amplifier is biased beyond cut-off, at the instant excitation is applied. In the latter case, the oscillation frequently will be self-sustaining after the excitation has been removed. At other times the oscillation may not be self-sustaining, becoming active only in the presence of excitation. It may be apparent only by the production of abnormal key clicks (� 6-1) over awide frequency range or by the presence of similarly wide-spread spurious side-bands (� 5-2) with 'phone modulation.

Fig. 426 -- Methods of using the Fara-
day shield. Two are required with apushpull or balanced tank circuit. The shield should be somewhat larger than the diameters of the coupled coils, and should be inserted between them so each is completely unexposed to the other.

Tank Cod )
Shield
- - - Qout

rank Coil

77 CHAPTER FOUR

Dhe Radio _Amateur's -naniZooh

Low-frequency parasitics -- Parasitic oscillations at low frequencies (usually 500 kc. or less) are of the tuned-plate tuned-grid type, the tuned circuits being formed by r.f. chokes and associated by-pass and coupling condensers, with the regular tank tuning condensers having only a minor effect on the oscillation. The operating-frequency tank coil has negligible inductance for such low frequencies and may be short-circuited without affecting the oscillations. The oscillations do not occur when
no r.f. chokes are used, hence whenever possible in series-fed circuits such chokes should be omitted. With single-ended amplifiers it is usually possible to arrange the circuit so that
either the grid or plate circuit needs no choke. In push-pull stages where chokes must be used
in both plate and grid circuits, it is helpful to connect an unby-passed grid leak from the choke to the bias supply or ground, thus placing the resistance in the parasitic circuit and tending to prevent oscillation. When the driver plate circuit has parallel feed and the amplifier
grid circuit series feed (� 3-7) this type of oscillation cannot occur so long as no choke is
used in the series grid circuit, since the grid is grounded through the tank coil for the parasitic .frequency.
Parasitics near operating frequency -- In circuits utilizing a tap on the plate tank coil to establish aground for abalanced neutralizing circuit, such as Fig. 413-B, aparasitic oscillation may be set up if the amplifier grid is tapped down on the grid (or driver plate) tank circuit for adjustment of driver-amplifier coupling (� 4-6). In this case the turns between grid and ground, and between plate and ground, form with the stray and other capacities present a t.p.t.g. circuit (� 3-7) which oscillates at a frequency somewhat higher than the nominal operating frequency. Such an oscillation can be prevented by dispensing with the taps in either the plate or grid circuit. Balancing the plate circuit by means of asplitstator condenser, as in Fig. 413-C, is recom-
mended.
Ultra-high frequency parasitics -- Parasitics in the u.h.f. region are likely to occur with any amplifier having abalanced tank circuit, particularly when associated with neu-
tralizing connections. The parasitic circuit may be either of the t.p.t.g. or ultraudion type, and is formed by the leads connecting the various components.
The frequency of such oscillations may be determined by connecting a tuned circuit in series with the grid lead to the tube. A variable
condenser (50 or 100 gpfd.) may be used in
conjunction with three or four self-supporting turns of heavy wire wound in acoil an inch or so in diameter. With the amplifier oscillating at the parasitic frequency, the condenser is slowly

tuned through its range until oscillations cease. In case this point is not found on first trial, the
turns of the coil may be spread apart or aturn removed and the process repeated. While this
may not be the simplest cure in all cases, the use of such a tuned circuit as a trap is an almost certain remedy, if the frequency can be determined, and introduces little if any loss at the operating frequency.
An alternative cure which is feasible when the oscillation is of the t.p.t.g. type is to detune the parasitic circuit in either the plate or grid circuit. Since this type of oscillation occurs most frequently with push-pull amplifiers, it may often be cured by making the grid and plate leads to their respective tank circuits of considerably different length. Similar considerations apply to neutralizing connections in push-pull circuits. The extra wire length may be coiled up in the form of aso-called "choke," which in this case is simply additional inductance for detuning the parasitic circuit.
Testing for parasitic oscillations -- An amplifier always should be tested for parasitic oscillations before being considered ready for service. The preferable method is first to
neutralize the amplifier, then apply sufficient fixed bias to permit a moderate value of plate current to flow without excitation. (The plate
current should not be large enough to cause the power input to exceed the rated plate dissipation of the tube.) If the amplifier is free from self-starting parasitics the plate current will remain steady as the tank condensers are varied in capacity; also, there will be no grid
current and aneon bulb touched either to the plate or grid will show no glow. Care must be used not to let the hand come in contact with
any metal parts of the transmitter in using the neon bulb.
If any of these effects are present the frequency of the parasitic must first be determined. If r.f. chokes are used in both the plate and grid circuits one of them should be shortcircuited to determine if the oscillation is at a low frequency; if so, it may be eliminated by the methods outlined above. If the test indicates that the parasitic is not alow-frequency oscillation, the grid trap described above should
be tried for the u.h.f. type. The type which occurs near the operating frequency will not occur unless the plate and grid tank coils are both tapped, hence may be eliminated from
consideration if this is not the case in the circuit used. When it is possible for such an oscillation to be present, its existence can be detected very readily by moving the grid tap to include the whole tank circuit, when the oscillation will cease.
Some indication of the frequency of the parasitic can be obtained from the color of the
glow in the neon bulb. Usually it will be yellow-

78 CHAPTER FOUR

Radio-..7redluency Power generation

ish with low-frequency oscillations and violet with u.h.f. oscillations.
If the amolifier is stable under the conditions described above, excitation should be applied and then removed to ascertain if aselfsustaining oscillation is set up with excitation. If the plate current does not return to the previous value when the excitation is cut off, the same tests should be applied to determine the parasitic frequency.
As afinal test, the transmitter should be put on the air and a nearby receiver tuned over as wide afrequency range as possible to locate any off-frequency signals associated with the radiation. Parasitics usually can be recognized by their poor stability, as contrasted to the normal transmitter harmonics, which will have the same stability as the fundamental signal as well as the usual harmonic frequency relationship. Harmonics should be quite weak compared to the fundamental frequency, whereas parasitic oscillations may have considerable strength.
� 4-11 FREQUENCY MULTIPLICATION
Circuits -- A frequency multiplier is an amplifier having its plate tank circuit tuned to amultiple (harmonic) of the frequency applied to its grid. The difference between a straight amplifier (� 4-1) and afrequency multiplier is in the way in which it is operated rather than in the circuit. However, since the grid and plate tank circuits are tuned to different frequencies a triode frequency multiplier will not self-oscillate, hence does not need neutralization. A typical circuit arrangement is shown in Fig. 427-A. For screen-grid multipliers the circuit is the same as in Fig. 412-A. Under usual conditions the plate efficiency of a frequency multiplier drops off rapidly with an increase in the number of times the frequency is multiplied. For this reason most multipliers are used as frequency doublers, giving second harmonic output.
A special circuit for frequency doubling ("push-push" doubler) is shown in Fig. 427-B. The grids of the tubes are in push-pull and the plates in parallel, thus the plate tank circuit receive� two pulses of plate current for each cycle of excitation frequency. The circuit is similar in principle to the full-wave rectifier (� 8-3) where the ripple frequency is twice the applied frequency.
Push-pull amplifiers are suitable for frequency multiplication at odd harmonics but are unsuited to doubling or other even-harmonic multiplication because the even harmonics are largely balanced out in the tank circuit (� 3-3).
Operating conditions and circuit constants -- To obtain good efficiency the operating angle at the harmonic frequency must be

180 degrees or less, preferably in the vicinity of 150-120 degrees (� 4-8). In a doubler this means that plate current should flow during only half this angle of fundamental frequency. Consequently the r.f. grid voltage operating bias, and grid driving power must be increased considerably beyond the values obtaining for normal Class-C amplification. For comparable
plate efficiency the bias will ordinarily be four to five times the normal Class-C bias, and the r.f. grid voltage must be considerably larger to drive the tube to the same peak plate current. Since the plate and grid current pulses under these conditions have the same peak amplitudes but only half the time duration as in astraight amplifier, the average d.c. values should be one-half those for normal Class-C operation. That is, atube operated in this way
will have the same plate efficiency as aClass-C amplifier, but can be operated at only half the
plate input so that the output power also is
halved. The driving power required is usually about twice that for straight-through amplification with the same plate efficiency.
Greater output can be secured by using a
larger operating angle (lower grid bias) or lower plate load resistance to increase the plate

couTo nf and bias circuits
1-4�
BIAS
A

runed to harmonic
H V

To coup/1n9
and bias,nq circuits 1-1-1?
BIAS

Tuned to harmonic

Fig. 427 --Frequency-multiplying circuits. A is for triodes, used either singly or in parallel. The push. push doubler is shown at B. Any type of coupling may be used between the grid circuit and the driver. Ci should be 0.01 ad. or larger; C2, 0.001 pfd. or larger.

79 CHAPTER FOUR

..7he Radio Amaieur'd ilandiooh

current, but this is accompanied by adecrease in efficiency. Since operation as described above is below the maximum plate dissipation rating of the tube, the decrease in efficiency can usually be tolerated in the interests of securing somewhat more power output. Ordinarily the efficiency is 40% to 50%.
The tank circuit should have reasonably high Q (12 is satisfactory) to give good output voltage regulation (� 4-9) since aplate-current pulse occurs only once every two cycles of output frequency. A low-Q circuit (high LIC ratio) is helpful chiefly when the operating angle is greater than 180 degrees at the second harmonic. Such atank circuit will have relatively high impedance to the considerable fundamental-frequency component of plate current which is present with large operating angles, and thus aid in reducing the average d.c. plate current.
The grid impedance of a frequency multiplier is considerably higher than that of a straight amplifier because of the high bias voltage. The average impedance can be calculated as previously described (� 4-8). The LIC ratio of the grid tank circuit may be higher, therefore, for agiven Q. It is often advantageous to use a fairly high ratio since a large r.f. voltage must be developed between grid and cathode, so long as it is not made too high (Q too low) to permit adequate coupling between the grid tank circuit and the driver stage. In some cases it may be necessary to step up the driver output voltage to obtain sufficient r.f. grid voltage for the doubler; this may be done by tapping the driver plate on its tank circuit, when capacity coupling is used, or by similar tapping or use of ahigher CIL ratio in the driver plate tank when the stages are link-coupled (� 4-6).
Tubes for frequency multiplication -- There is no essential difference between tubes of various characteristics in their performance as frequency doublers. Tubes having high amplification factors will require somewhat less bias for equivalent operation, but the grid driving power needed is almost independent of the �, assuming tubes of otherwise similar construction and characteristics. Pentodes and tetrodes having high power sensitivity will, as in normal amplifier operation, require less driving power than triodes for efficient doubling, although more power will be needed than for straight amplification.
�4-12 ULTRA-HIGH-FREQUENCY OSCILLATORS
Linear circuits -- At ultra-high frequencies tube interelectrode capacities become of increasing importance, so that eventually the shortest possible straight wire connection between elements, in conjunction with internal

leads and capacities, represents the highest possible frequency to which the tube can be tuned. The tube usually will not oscillate up to this limit because of dielectric losses in the seals and other loading effects (� 7-6). With most small tubes of ordinary construction the upper limit of oscillation is in the region of 150 Mc.; for higher frequencies it is necessary to use special u.h.f. tubes having low interelectrode capacities and low internal lead inductance. Only a few types are capable of developing more than a few watts at frequencies of 300 Mc. and higher.
Although ordinary coil and condenser tank circuits can be used at frequencies as high as 112 Mc., the Q of such circuits is low at ultrahigh frequencies because of increased losses, so that both stability and efficiency are poor. For this reason special tank circuits of the linear type (� 2-12) are preferable. These may be
any multiple of aquarter wave in length, the Q increasing with the number of quarter waves. The quarter-wave line is generally used, however, because of the considerable space required for longer lines. At 112 Mc. it is also possible to build high-Q tank circuits with lumped constants, not in the form of ordinary coils and condensers but with large conducting surfaces to reduce resistance to the lowest possible value.
The oscillator circuits used are the same in principle as on the lower frequencies (� 3-7) although frequently modified considerably to compensate for inherent capacities and inductances which are negligible at lower frequencies.
Two-conductor lines --The quarter-wave two-conductor open line is equivalent to a resonant circuit (� 2-12) and can be used as the tank circuit (� 3-7) in an oscillator. It should be used as abalanced circuit to avoid unequal currents in the two conductors and consequent loss of Q because of radiation.
A typical oscillator circuit of the ultraudion type is shown in Fig. 428. The resonant line is usually constructed of copper tubing to give a large conducting surface and hence reduce resistance, and also to make a mechanicallystable circuit and thus minimize the effects of vibration on the oscillator frequency. The line should be approximately aquarter wavelength long, although the resonant frequency will decrease somewhat when the tube with its internal capacities is connected across it so that a somewhat shorter length is ordinarily sufficient. The frequency can be changed by means
of the shorting bar, which can be moved along the line to change its effective length.
The tube elements preferably should be tapped down on the line as shown to reduce the loading effect and thus prevent an undue decrease in Q. In general, these taps should be
as cloue to the shorted end of the line as is con-

CHAPTER FOUR

letcho-Drequency rower generation

Shorting Bar '

be determined by experiment, the coils being adjusted until optimum stability and power

7'14

output are obtained. The oscillation frequency may also be ad-

justed by connecting a low-capacity variable

condenser across the open end of the line. The

RFC

added capacity makes it necessary to shorten

the line considerably for a given frequency,

however, and this together with the additional

loss in the condenser causes amarked decrease

in the Q of the line. These effects will be less if

the condenser is connected down on the line

rather than at the open end. Tapping down

Fig. 428 -- Single-tube line oscillator. The grid condenser, CI, may be 50 pad.; grid leak, RI, 5000 to 50,000 ohms depending upon the type of tube. The choke, RFC, will in general consist of relatively few turns (20 to 50) wound to adiameter of 3 inch, although dimensions will change considerably with the frequency. By-

also gives a greater band-spread tuning effect ( 7-7).
Push-pull oscillators --It is often advantageous to use apush-pull oscillator circuit at ultra-high frequencies, not only as ameans to

pass condensers should be small in size to reduce lead inductance; 500 upfd. is asatisfactory value.

secure more power output than can be ob-. tamed from one tube but also because better

circuit symmetry is possible with open lines. sistent with reliable operation and satisfactory Fig. 429 shows a typical push-pull circuit of

power output, since the frequency stability will be better under these conditions.
The coils (L) in the filament circuit are frequently required at 112 Mc. and higher to compensate for the effects of the inductance of
connecting leads, which in many cases are long

the t.p.t.g. (* 3-7) type. The grid line is usually
operated as the frequency-controlling circuit since it is not associated with the load and hence its Q can be kept high. The same adjustment considerations apply as in the case of the single-tube oscillator described in the preced-

enough to cause an appreciable phase shift ing paragraph. The grid taps in particular

(� 2-7) which reduces the oscillator efficiency. should be tapped down as far as possible, thus

The effective length of the filament circuit to improving the frequency stability.

the points of connection to the lines should be

It is also possible to use alinear tank in the

approximately M wavelength to bring the grid circuit for frequency control in conjunc-

filament to the same potential as the shorted tion with a conventional coil-condenser tank

ends of the lines. The proper inductance must in the plate circuit, where the lower Q does not

have so great an effect on the sta-

bility.

Fig. 429-B shows apush-pull oscil-

lator having tuned plate and cathode

circuits, using linear tanks for each.

The grids are connected together and

grounded through the grid leak, RI;

ordinarily no by-pass condenser is

needed across R1.This circuit gives good power output at ultra-high fre-

quencies, but is not especially stable

unless the plates are tapped down on

the plate tank circuit to avoid too

(A)

great areduction in Q. Tapping on the

cathode line is not feasible for me-

chanical reasons, since one filament

Shortin9 Bar

lead must be brought through the tubing in order to maintain both sides

of the filament at the same r.f. po-

tential.

Concentric-line circuits-- At fre-

ri Sup

OTutpTut

+o8

quencies in the neighborhood of 300 Mc. radiation (� 2-12) from the open

line becomes so serious that the Q is

(B)

greatly reduced. This is because the

Fig. 429 --Push-pull line oscillator circuits. See Fig. 428 and, conductor spacing represents an ap-

text for discussion of cite* �osistunts,

preciable fraction of the wavelength.

81 CHAPTER FOUR

Dheecho ..4frnaleur'� ilanclloon

Consequently at these frequencies the concen-
tric line must be used. In this type the field is confined inside the line so that radiation is neg-
ligible; there is afurther advantage in that the outside of the line is "cold"; that is, no r.f. potentials develop between points on the outer surface. The concentric line also is advanta-
geous at lower frequencies, but as it is more
complicated to construct and length adjust-
ment and tapping both are difficult mechanically, the open lines are generally favored.
The concentric line is usually constructed of copper pipes arranged concentrically and shortcircuited at one end. The optimum ratio of inner diameter of the outer conductor to the
outer diameter of the inner conductor is 3.6. Taps are usually made on the inner conductor and brought through a hole in the outer conductor to the tube element, as shown in Fig. 430. The tube loads the line in the same way as described in the preceding paragraphs, hence the length is generally shorter than an actual
quarter wavelength. The length can be adjusted by asliding short-circuiting disc at the
closed end, a close fit and low-resistance contact being necessary to avoid reduction of the Q. It is also possible to make the inner con-

c,
RFC
nl

0-4(1:,utput
(A)
+B

(B)
Fig. 430 -- Concentric-line oscillator circuits. The
line, usually of tubular conductors, is shown in crosssection. See Fig. 428 and text for discussion of circuit constants.

Output

- B

Fig. 431-- High-Q lumped-constant tank circuit in au.h.f. oscillator. This drawing shows across-section of the tank, which is usually built of concentric cylinders. Ci and RIare the grid condenser and leak, respectively; see Fig. 428 for discussion of circuit constants.

ductor a pair of close-fitting concentric tubes

so that one may be slid in and out of the other

to change the effective conductor length.

The circuit of Fig. 430-A is at.p.t.g. (� 3-7)

oscillator using the concentric line in the grid

circuit for frequency control. An ordinary coil-

condenser tank is shown in the plate circuit,
but a linear tank may be substituted. The

filament inductances have the same function

as in the preceding circuits. The ultraudion

circuit is shown at B; the same considerations

apply. In this case the output is taken from the

line inductively by means of the half-turn

"hairpin" shown; coupling can be changed to

some extent by varying the position of the hair-

pin. Both circuits may be tuned by means of

the small variable condenser Cry, although this

condenser may be omitted and the tuning

accomplished by changing the line length.

For ease of construction, the concentric line
is sometimes modified into a "trough," in

iwshiinchthtehsehcarpoesso-fseactsiqounaroef

the outer "U," one

conductor side being

left open for tapping and adjustment of the

inner conductor. Some radiation takes place

with this construction, although not as much

as with open lines.

High-Q circuits with lumped constants --

To obtain reasonably high effective Q when a

low resistance is connected across the tank

circuit it is necessary to use a high CIL ratio

and a tank of inherently high Q (� 2-10). At

low frequencies the inherent Q of any well-

designed circuit will be high enough so that it

may be neglected in comparison to the effec-

tive Q when loaded, so that no special pre-

cautions have to be taken with respect to the

resistance of coils and condensers. At ultra-

high frequencies these internal resistances are

too large to be ignored, and areduction of the

82 CIIAPTER FOUR

leacho--7requency Power Generation

L/C ratio will not increase the effective Q
unless the internal resistance of the tank can
be made very small. The reduction in resistance can be brought about by use of large conducting surfaces and elimination of radiation. In such cases the inductance and capacity are generally built as aunit; several arrangements
are possible, one being shown in Fig. 431. The tank circuit consists of arod A (the inductance) inside two concentric cylinders B and C which form a two-plate condenser, one plate being connected to each end of the inductance. The resonant frequency is determined by the
length and diameter of A, and the length, diameter and spacing of B and C. The oscillator shown uses the tickler circuit (� 3-7) with

the feedback coil in the grid circuit; this inductance is the wire D in the diagram. Output is taken from the tank circuit by means of the hairpin coupling coil. The tank circuit may also be used in the ultraudion circuit, replacing the concentric line in Fig. 430-B. A variable condenser may be connected across the tank for tuning, if desired, although the Q may be reduced if a considerable portion of the tank r.f. current flows through it.
This type of circuit actually has lumped constants only when the length is small (10% or less) in proportion to the wavelength. At greater lengths it tends to act as a linear circuit, eventually evolving into the concentric line.

83 CHAPTER FOUR

CHAPTER FIVE
lecediodephonl

�5-1 MODULATION The carrier --The steady radio-frequency
power generated by transmitting circuits cannot alone result in the transmission of an intelligible message to a receiving point. It serves only as a "carrier" for the message; the intelligence is conveyed by modulation (a change) of the carrier. In radiotelefehony this modulation reproduces electrically the sounds it is intended to convey.
Sound and alternating currents -- Sounds are caused by vibrations of air particles. The pitch of the sound depends upon the rate of vibration; the more rapid the vibration the higher the pitch. Most sounds consist of complex combinations of vibrations of differing rates or frequencies; the human voice, for instance, generates frequencies from about 100 per second to several thousand per second. The problem of transmitting speech by radio is therefore one of varying the r.f. carrier in a way which corresponds to the air-particle vibrations. The first step in doing this is to change the sound vibrations into alternating electrical currents of the same frequency and relative intensity; the electromechanical device which achieves this translation is the microphone. These currents may then be amplified and used to modulate the normallysteady r.f. output of the transmitter.
Methods of modulation -- The carrier may be made to vary in accordance with the speech current by using the current to change the
phase (� 2-7) frequency or amplitude of the carrier. Amplitude modulation is by far the most common system, and is used exclusively on all frequencies below the ultra-high-frequency region (� 2-7). Frequency modulation, which has special characteristics which make its use desirable under certain conditions, is used to a considerable extent on ultra-high frequencies. Phase modulation, which is closely related to frequency modulation, has had little
or no direct application in practical communication.
�5-2 AMPLITUDE MODULATION
Carrier requirements -- For proper amplitude modulation, the carrier should be completely free from inherent amplitude variations such as might be caused by insufficient filtering of a rectified-a.c. power supply (� 8-4). It is

also essential that the carrier frequency be entirely unaffected by the application of
modulation. If modulating the amplitude of the carrier also causes achange in the carrier frequency, the signal wobbles back and forth with the modulation, introducing distortion and widening the channel taken by the signal. This causes unnecessary interference to other transmissions. In practice, this undesirable
frequency modulation is prevented by applying the modulation to an r.f. amplifier stage which is isolated from the frequency-controlling oscillator by a "buffer" amplifier. Amplitude modulation of an oscillator is almost always accompanied by frequency modulation. It is permitted on ultra-high frequencies above 112 Mc. because the problem of interference is less acute than on lower frequencies.
Percentage of modulation -- In the ampli-
tude-modulation system the audible output at the receiver depends entirely upon the amount of variation -- termed depth of modulation -- in the carrier wave and not upon the strength of the carrier alone. It is therefore
desirable to obtain the largest permissible variations in the carrier wave. This condition is reached when the carrier amplitude during modulation is at times reduced to zero and at other times increased to twice its unmodulated value. Such a wave is said to be fully modulated, or 100% modulated. Any desired degree of modulation can be expressed as a percentage, using the unmodulated carrier as a base. Fig. 501 shows at A an unmodulated carrier wave; at B the same wave modulated 50%, and at C the wave with 100% modulation,
using a sine-wave (� 2-7) modulating signal. The outline of the modulated r.f. wave is called the modulation envelope.
The percentage modulation can be found by
dividing either Y or Z by X and multiplying the result by 100. If the modulating signal is
not symmetrical, the larger of the two (Y or Z) should be used.
Power in modulated wave -- The amplitude values correspond to current or voltage, so that the drawings may be taken to represent instantaneous values of either. Since power varies as the square of either the current or voltage (so long as the resistance in the circuit is unchanged), at the peak of the modulation up-swing the instantaneous power in the wave

84 CHAPTER FIVE

1 li

Radiotelephony
I Modulation capability -- The modulation capability of the transmitter is the maximum percentage of modulation that is possible without objectionable distortion from non-

linearity. The maximum capability is, of

wave -slia,oe of Modulating Signal

course, 100%. The modulation capability should be as high as possible so that the most

effective signal can be transmitted for agiven

carrier power.

--x4--
(3)

Overmodulation -- If the carrier is modulated more than 100%, acondition such as is shown in Fig. 502 occurs. Not only does the peak amplitude exceed twice the carrier ampli-

tude, but there may actually be aconsiderable

period during which the output is entirely

cut off. The modulated wave is therefore dis-

Ware-sliape of Modulating Signal

torted (� 3-3) with the result that harmonics of the audio modulating frequency appear. The carrier should never be modulated more

than 100%.

Sidebands -- The combining of the audio

--4--
Y

frequency with the r.f. carrier is essentially a heterodyne process and therefore gives rise to

beat frequencies equal to the sum and differ-

ence of the a.f. and r.f. frequencies involved

(� 2-13). Therefore, for each audio frequency

appearing in the modulating signal two new

radio frequencies appear, one equal to the

carrier frequency plus the audio frequency,

the other equal to the carrier minus the audio

Fig. 501-- Graphical representaion of (A) unmodu-
lated carrier wave, (3) wave modulated 50%, (C) wave
modulated 100%.

frequency. These new frequencies are called side frequencies, since they appear on each side of the carrier, and the groups of side fre-

quencies representing a band or group of

of Fig. 501-C is four times the unmodulated modulation frequencies are called sidebands.

carrier power. At the peak of the down-swing Hence a modulated signal occupies a group

the power is zero since the amplitude is zero. of radio frequencies, or channel, rather than a

With asine-wave modulating signal, the aver- single frequency as in the case of the unmodu-

age power in a 100%-modulated wave is one lated carrier. The channel width is twice the

and one-half times the unmodulated carrier highest modulation frequency. To accommo-

power: that is, the power output of the trans- date the largest number of transmitters in a

mitter increases 50% with 100% modula-

tion. Linearity -- Up to the limit of 100% modu-
lation, the amplitude of the carrier should

Wave-shape of
Modulating

follow faithfully the amplitude variations of

the modulating signal. When the modulated

r.f. amplifier is incapable of meeting this con-

dition it is said to be non-linear. The amplifier

may not, for instance, be capable of quadru-

pling its power output at the peak of 100%

modulation. A non-linear modulated amplifier

causes distortion of the modulation envelope.

Modulation characteristic -- A graph showing the relationship between r.f. ampli-

carrier

tude and instantaneous modulating voltage is

called the modulation characteristic of the

modulated amplifier. This graph should be a

straight line (linear) between the limits of zero

and twice carrier amplitude. Curvature of the

line between these limits indicates non-line-

arity.

Fig. 502 -- An overmodulated wave.

85 CHAPTER FIVE

Dhe echo -AmaieuA

("Lai

given part of the r.f. spectrum it is apparent that the channel width should be as small as possible, but on the other hand it is necessary, for speech of reasonably good quality, to use modulating frequencies up to about 3000 or 4000 cycles. This calls for achannel width of 6to 8kc.
Spurious side bands -- Besides the normal side bands required by speech frequencies, unwanted side bands may be generated by the transmitter. These usually lie outside the normally-required channel and hence cause it to be wider without increasing the useful modulation. By increasing the channel width these spurious side bands cause unnecessary interference to other transmitters. The quality of transmission is also adversely affected when spurious side bands are generated.
The chief causes of spurious side bands are harmonic distortion in the audio system, overmodulation, unnecessary frequency modulation, and lack of linearity in the modulated r.f. system.
Types of amplitude modulation -- The most widely used type of amplitude modulation system is that in which the modulating signal is applied in the plate circuit of aradiofrequency power amplifier (plate modulation). In asecond type the audio signal is applied to a control-grid circuit (grid-bias modulation). A third system involves variation of both plate voltage and grid bias and is called cathode modulation.
�5-3 PLATE MODULATION
Transformer coupling --In Fig. 503 is shown the most widely-used system of plate modulation. A balanced (push-pull Class-A, Class-AB or Class-B) modulator is transformer-coupled to the plate circuit of the modulated r.f. amplifier. The audio-frequency power generated in the modulator plate circuit is combined with the d.c. power in the modulated-amplifier plate circuit by transfer through the coupling transformer, T. For 100% modulation the audio-frequency output of the modulator and the turns ratio of the coupling transformer must be such that the voltage at the plate of the modulated amplifier varies between zero and twice the d.c. operating plate voltage, thus causing corresponding variations in the amplitude of the r.f. output.
Modulator power --The average power output of the modulated stage must increase 50% for 100% modulation (� 5-2), so that the modulator must supply audio power equal to 50% of the d.c. plate input to the modulated r.f. stage. For example, if the d.c. plate power input to the r.f. stage is 100 watts, the sinewave audio power output of the modulator must be 50 watts.
Modulating impedance, linearity -- The modulating impedance or load resistance pre-

+8 +8 Fig. 503 --Plate modulation of aClass-C r.f. amplifier. The plate by-pass condenser, C, in the r.f. stage should have high reactance at audio frequencies. Avalue of 0.002 dd. or less is usually satisfactory.

sented to the modulator by the modulated r.f. amplifier, is equal to

Eb -- X

1000

where Eb is the d.c. plate voltage and I, the d.c. plate current in milliamperes, both measured without modulation.
Since the power output of the r.f. amplifier must vary as the square of the plate voltage (r.f. voltage proportional to applied plate voltage) in order for the modulation to be linear, the amplifier must operate Class-C (� 3-4). The linearity depends upon having sufficient grid excitation, proper bias, and adjustment of circuit constants to the proper
values (� 4-8). Power in speech waves -- The complex
waveform of a speech sound translated into alternating current does not contain as much
power, on the average, as there is in a pure tone or sine wave of the same peak (� 2-7) amplitude. That is, with speech waveforms the ratio of peak to average amplitude is higher than in the sine wave. For this reason,
the previous statement that the power output of the transmitter increases 50% with 100% modulation, while true for tone modulation, is not true for speech. On the average, speech waveforms will contain only about half as much power as a sine wave, both having the same peak amplitude. The average power output of the transmitter therefore increases only about 25% with 100% speech modulation. However, the instantaneous power output myst
quadruple on the peak of 100% modulation

86 CHAPTER FIVE

Radiotelephony

(� 5-2) regardless of the modulating waveform. Therefore the peak capacity of the transmitter must be the same for any type of modulating signal.
Adjustment of plate-modulated amplifiers --The general operating conditions for Class-C operation have been described (� 3-4, 4-8). The grid bias and grid current required for plate modulation are usually given in the operating data supplied by the tube manufacturer; in general, the bias should be such as to give an operating angle (� 4-8) of about 120 degrees at carrier plate voltage, and the excitation should be sufficient to maintain the plate efficiency constant when the plate voltage is varied over the range from zero to twice the d.c. plate voltage applied to the amplifier. For best linearity, the grid bias should be obtained partly from afixed source of about the cut-off value supplemented by grid-leak bias to supply the remainder of the required operating bias.
The maximum permissible d.c. plate power input for 100% modulation is twice the sinewave audio-frequency power output of the modulator. This input is obtained by varying the loading on the amplifier (keeping its tank circuit tuned to resonance) until the product of d.c. plate voltage and plate current is the desired power. The modulating impedance under these conditions will be the proper value for the modulator if the proper output transformer turn ratio (� 2-9) is used.
Neutralization, when triodes are used, should be as nearly perfect as possible, since regeneration may cause non-linearity. The amplifier also should be free from parasitic oscillations (� 4-10).
Although the effective value (� 2-7) of power input increases with modulation, as described above, the average plate input to a platemodulated amplifier does not change, since each increase in plate voltage and plate cur-
Excitation
-C +5t15 -B +C
+8
Fig. 504 -- Plate and screen modulation of apentode Class-C rd. amplifier. Plate and screen by-pass condensers, CIand C2, should have high reactance at audio frequencies (0.002 dd. or less).

rent is balanced by an equivalent decrease in voltage and current. Consequently the d.c. plate current to a properly-modulated amplifier is constant with or without modulation.
Screen-grid amplifiers -- Screen-grid tubes of the pentode or beam tetrode type can be used as Class-C plate-modulated amplifiers provided the modulation is applied to both the plate and screen grid. The method of feeding the screen grid with the necessary d.c. and modulation voltage is shown in Fig. 504. The dropping resistor, R, should be of the proper value to apply normal d.c. voltage to the screen
under steady carrier conditions. Its value can be calculated by taking the difference between plate and screen voltages and dividing it by the rated screen current.

CLASS -C AM P

ExRc.iFt.ation �=,
RFC

g Ant.

' Grid Leak

RFC

CLASS -A MOD

From
Speech Amp.

- -

- C -B

-a

Fig. 505 -- Choke-coupled plate modulation.

The modulating impedance is found by dividing the d.c. plate voltage by the sum of the plate and screen currents. The plate voltage
multiplied by the sum of the two currents is the power input figure which is used as the basis for determining the audio power required from
the modulator. Choke coupling -- In Fig. 505 is shown the
circuit of the choke-coupled system of plate
modulation. The plate power for the modulator tube and modulated amplifier is furnished from a common source through the modulation choke, L, which has high impedance for audio frequencies. The modulator operates as apower amplifier with the plate circuit of the r.f. amplifier as its load, the audio output of the modulator being superimposed on the d.c.
power supplied to the amplifier. For 100% modulation the audio voltage applied to the
r.f. amplifier plate circuit across the choke, L,

87 CHAPTER FIVE

..7he echo Amaieur'd fianclgooh

must have apeak value equal to the d.c. voltage on the modulated amplifier. To obtain this without distortion, the r.f. amplifier must be operated at a d.c. plate voltage less than the modulator plate voltage, the extent of the voltage difference being determined by the type of modulator tube used. The necessary drop in voltage is provided by the resistor RI, which is by-passed for audio frequencies by the condenser
This type of modulation is seldom used except in very low-power portable sets, because asingle-tube Class-A (� 3-4) modulator is required. The output of aClass-A modulator is very low compared to that obtainable from a pair of tubes of the same size operated Class-B, hence only asmall amount of r.f. power can be modulated. �5-4 GRID-BIAS MODULATION
Circuit--Fig. 506 is the diagram of atypical arrangement for grid-bias modulation. In this system the secondary of an audiofrequency output transformer, the primary of which is connected in the plate circuit of the modulator tube, is connected in series with the grid-bias supply for the modulated amplifier. The audio voltage thus introduced varies the grid bias and thus the power output of the r.f. stage, when suitable operating conditions are chosen. The r.f. stage is operated as a Class-C amplifier, with the d.c. grid bias considerably beyond cut-off.
Operating principles-- In this system the plate voltage is constant, and the increase in power output with modulation is obtained by making the plate current and plate efficiency vary with the modulating signal. For 100%
ExRciFtation
+B
Fig. 506 -- Grid-bias modulation of aClass-C amplifier. The r.f. grid by-pass condenser, C, should have high reactance at audio frequencies (0.002 �fd. or less in usual cases).

modulation, both plate current and efficiency must, at the peak of the modulation up-swing, be twice their carrier values so that the peak power will be four times the carrier power. Since the peak efficiency in practicable circuits is of the order of 70% to 80%, the carrier efficiency ordinarily cannot exceed about 35% to 40%. For agiven size of r.f. tube the carrier output is about one-fourth the carrier obtainable from the same tube plate-modulated. The grid bias, r.f. excitation, plate loading and audio voltage in series with the grid must be adjusted to give a linear modulation characteristic.
Modulator power--Since the increase in average carrier power with modulation is secured by varying the plate efficiency and d.c. plate input of the amplifier, the modulator need only supply such power losses as may be occasioned by connecting it in the grid circuit. These are quite small, hence a modulator capable of only a few watts output will suffice for transmitters of considerable power. The load on the modulator varies over the audio-frequency cycle as the rectified grid current of the modulated amplifier changes, hence the modulator should have good voltage regulation (� 5-6).
Grid-bias source--The change in bias voltage with modulation causes the rectified grid current of the amplifier also to vary, the r.f. excitation being fixed. If the bias source has appreciable resistance, the change in grid current also will cause a change in bias in a direction opposite to that caused by the modulation. It is therefore necessary to use a grid-bias source having low resistance so that these bias variations will be negligible. Battery bias is satisfactory. If arectified a.c. bias supply is used the type having regulated output (� 8-9) should be used. Grid-leak bias for a grid-modulated amplifier is unsatisfactory and
its use should not be attempted. Driver regulation-- The load on the driving
stage varies with modulation, and a linear modulation characteristic may not be obtained if the r.f. voltage from the driver does not stay constant with changes in load. Driver regula-
tion (ability to maintain constant output voltage with changes in load) may be improved by using a driving stage having two or three times the power output necessary for excitation of the amplifier (this is somewhat less than the power required for ordinary Class-C operation), and by dissipating the extra power in a constant load such as a resistor. The load variations are thereby reduced in proportion to the total load.
Adjustment of grid-bias modulated amplifiers--This type of amplifier should be adjusted with the aid of an oscilloscope to
obtain optimum operating conditions. The

88 CHAPTER FIVE

leadiotefephony

oscilloscope should be connected as described in f5-10, the wedge pattern being preferable.

CLASS-C AMP

A tone source for modulating the trans-

mitter will be convenient. The fixed grid ficit--otio

bias should be two or three times the cut-

RFC

off value (� 3-2). The d.c. input to the amplifier,

assuming 33% carrier efficiency, will be 134 times the plate dissipation rating of the tube or tubes used in the modulated stage, and the plate

Grid Leak

+5 G.

RFC

current for thisinput (in milliamperes,1000P/E,

--C +C

where P is the power and E the d.c. plate volt-

age) determined. Apply r.f. excitation and,

without modulation, adjust the plate loading

(keeping the plate tank circuit tuned to reso-

nance) to give the required plate current. Next, apply modulation and increase in the modulating signal until the modulation characteristic shows curvature (� 5-10). This will
probably occur well below 100% modulation,

w Suppressor Bias
Fig. 507--Suppressor-grid modulation of apentode r.f. amplifier. The suppressor r.f. by-pass condenser, C, should be 0.002 dd. or less.

indicating that the plate efficiency is too high.

Increase the plate loading and reduce the ex- pentode tube is shown in Fig. 507. The operat-

citation to maintain the same plate current, ing principles are the same as for grid-bias

apply modulation and check the characteristic modulation. However, the r.f. excitation and

again. Continue this process until the charac- modulating signals are applied to separate

teristic is linear from the axis to twice the car- grids, which gives the system asimpler operat-

rier amplitude. It is advantageous to use the ing technique, since best adjustment for proper

maximum permissible plate voltage on the excitation requirements and proper modulating

tube, since it is usually easier to obtain amore circuit requirements are more or less independlinear characteristic with high plate voltage ent. The carrier plate efficiency is approxi-

and low current (carrier conditions) rather than with relatively low plate voltage and high plate current.

mately the same as for grid-bias modulation, and the modulator power requirements are similarly small. With tubes having suitable

The amplifier can be adjusted without an suppressor-grid characteristics, linear modulaoscilloscope by determining the plate current tion up to practically 100% can be obtained

as described above, then setting the bias to the with negligible distortion.

cut-off value (or slightly beyond) for the d.c.

The method of adjustment is essentially the

plate voltage used and applying maximum same as that described in the preceding paraexcitation. Adjust the plate loading, keeping graph. Apply normal excitation and bias to

the tank circuit at resonance, until the ampli- the control grid and, with the suppressor bias

fier draws twice the carrier plate current, and at zero or the positive value recommended

note the antenna current. Decrease the exci- for c.w. telegraph operation with the particular

tation until the output and plate current just start to drop, then increase the bias, leaving the excitation and plate loading unchanged, until the plate current drops to the proper

tube used, adjust the plate loading to obtain twice the carrier plate current (on the basis of 33% carrier efficiency). Then apply sufficient negative bias to the suppressor to bring the

carrier value. The antenna current should plate current to the carrier value, leaving the be just half the previous value; if it is larger, loading unchanged. Simultaneously, the an-

try somewhat more loading and less excita- tenna current also should drop to half its maxi-

tion; if smaller, less loading and more excita- mum value. The amplifier is then ready for tion. Repeat until the antenna current drops modulation. Should the plate current not

to half its maximum value when the plate follow the antenna current in the same pro-

current is biased down to the carrier value. portion when the suppressor bias is made Under these conditions the amplifier should � negative, the loading and excitation should be modulate properly, provided the plate supply readjusted to make them coincide. has good voltage regulation (� 8-1) so that the plate voltage is practically the same at both � 5-5 CATHODE MODULATION

values of plate current during the initial testing.

Circuit -- The fundamental circuit for

The d.c. plate current should be substan- cathode or "center-tap" modulation is shown

tially constant with or without modulation in Fig. 508. This type of modulation is acom-

(� 5-3). Suppressor modulation --The circuit ar-
rangement for suppressor-grid modulation of a

bination of the plate- and grid-bias methods, and permits a carrier efficiency midway between the two. The audio power is introduced

89 CHAPTER FIVE

leach� Amaleur'� ilartiLJ

CLASS C AMP

R.F Excitation

RFC

Grid t
,Leak T

firn.t
RFC

+B
Mod
Fil. 18001 Trans
Fig. 508--Cathode modulation of a Class-C r.f. amplifier. The grid and plate by-pass condensers, C, should be 0.002 dd. or less (high reactance at audio frequencies).
in the cathode circuit, and both grid bias and plate voltage vary during modulation.
The cathode circuit of the modulated stage must be independent of other stages in the transmitter; that is, when filament-type tubes are modulated they must be supplied from a separate filament transformer. The filament by-pass condensers should not be larger than about 0.002 pfd., to avoid by-passing the audio.
Operating principles -- Because part of the modulation is by the grid-bias method, the plate efficiency of the modulated amplifier must vary during modulation. The carrier efficiency therefore must be lower than the efficiency at the modulation peak. The required reduction in carrier efficiency depends upon the proportion of grid modulation to plate modulation; the higher the percentage of plate modulation the higher the permissible carrier efficiency, and vice versa. The audio power required from the modulator also varies with the percentage of plate modulation, being greater as this percentage is increased.
The way in which the various quantities vary is illustrated by the curves of Fig. 509. In these curves, the performance of the cathode-modulated r.f. amplifier is plotted in terms of the tube ratings for plate-modulated telephony, with the percentage of plate modulation as a base. As the percentage of plate modulation is decreased, it is assumed that the grid-bias modulation is increased to make the overall percentage of modulation reach 100%. The limiting condition, 100% plate modulation and no grid-bias modulation, is at the right (A); pure grid-bias modulation is represented by the left-hand ordinate (B and C).

As an example, assume that 40% plate modulation is to be used. Then the modulated r.f. amplifier must be adjusted for a carrier plate efficiency of 56%, the permissible plate input will be 65% of the ratings of the same tube with pure plate modulation, the power output will be 48% of the rated output of the tube with plate modulation, and the audio power required from the modulator will be 20% of the d.c. input to the modulated amplifier.
Modulating impedance -- The modulating impedance of a cathode-modulated amplifier is approximately equal to
Eb
lb
where m is the percentage of plate modulation expressed as adecimal, Eb is the plate voltage, and lb the plate current of the modulated r.f. amplifier. This figure for the modulating impedance is used in the same way as the corresponding figure for pure plate modulation in determining the proper modulator operating conditions (� 5-6).
Conditions for linearity -- R.f. excitation requirements for the cathode-modulated amplifier are midway between those for plate modulation and grid-bias modulation. More excitation is required as the percentage of plate modulation is increased. Grid bias should be considerably beyond cut-off; fixed bias from asupply having good voltage regulation (� 8-9) is preferred, especially when the percentage of plate modulation is small and the amplifier is operating more nearly like agridbias modulated stage. At the higher percent-

100

40 4('1
20
\c

446

100

rez.- PER CENT PLATE MODULATION

Fig. 509 -- Cathode modulation performance curves, in terms of percentage of plate modulation against per cent of Class-C telephony tube ratings.
-- D.c. plate input watts in per cent of plate-modulation rating.
W. -- Carrier output watts in per cent of plate-modulation rating (based on plate efficiency of 77.5%).
W. -- Audio power in per cent of d.c. watts input. Np -- Plate efficiency in per cent.

CHAPTER FIVE

Pachoielphony

ages of plate modulation acombination of fixed and grid-leak bias

DRIVER

CLASS-8 MODULATOR

can be used since the variation in

rectified grid current is smaller.

The grid-leak should be by-passed

for audio frequencies. The percent-

age of grid modulation may be

regulated by choice of asuitable

tap on the modulation trans-

former secondary.

Adjustment of cathode-mod-

ulated amplifiers --In most respects the adjustment procedure is similar to that for grid-bias modulation

+8
Fig. 510 -- Class.B modulator and driver circuit.

(� 5-4). The critical adjustments are those of

antenna loading, grid bias, and excitation. The

proportion of grid-bias to plate modulation

will determine the operating conditions.

Adjustments should be made with the aid of
an oscilloscope (� 5-10). With proper antenna loading and excitation, the normal wedgeshaped pattern will be obtained at 100% modulation. As in the case of grid-bias modulation too-light antenna loading will cause flattening of the up-peaks of modulation (downward modulation), as will also too-high excitation (� 5-10). The cathode current will be practically constant with or without modulation when the proper operating conditions are reached (� 5-3).

where Z. is the Class-C modulating impedance and Zp is the plate-to-plate load impedance specified for the Class-B tubes.
Commercial Class-B output transformers usually are rated to work between specified primary and secondary impedances and are designed for specific Class-B tubes. In such a
case the turn ratio can be found by substituting the given impedances in the formula above. Many transformers are provided with primary and secondary taps so that various turn ratios can be obtained to meet the re-

� 541 CLASSAS MODULATORS

quirements of a large number of tube combinations.

Modulator tubes --In the case of plate
modulation, the relatively-large audio power needed (� 5-3) practically dictates the use of a Class-B (� 3-4) modulator, since the power can be obtained most economically with this type of amplifier. A typical circuit is given in
Fig. 510. A pair of tubes must be chosen which is capable of delivering sine-wave audio power equal to half the d.c. input to the modulated Class-C amplifier. It is sometimes convenient
to use tubes which will operate at the same plate voltage as that applied to the Class-C stage, since one power supply of adequate current capacity may then suffice for both
stages. Available components do not always permit this, however, and better overall performance and economy may frequently
result from the use of separate power supplies.
Matching to load --In giving Class-B ratings on power tubes, manufacturers specify the plate-to-plate load impedance (� 3-3) into which the tubes must operate to deliver

Driving power -- Class-B amplifiers are driven into the grid-current region, so that
power is consumed in the grid circuit (� 3-3).
The preceding stage (driver) must be capable of supplying this power at the required peak
audio-frequency grid-to-grid voltage. Both these quantities are given in the manufacturer's tube ratings. The grids of the Class-B tubes represent avariable load resistance over the audio-frequency cycle, since the grid current does not increase directly with the grid voltage. To prevent distortion, therefore, it is necessary to have a driving source which has
good regulation -- that is, which will maintain the waveform of the signal without distortion
even though the load varies. This can be brought about by using a driver capable of delivering two or three times the actual power
consumed by the Class-B grids, and by using
an input coupling transformer having a turn ratio giving the largest step-down in voltage, between the driver plate or plates and Class-B grids, that will permit obtaining the specified

the rated audio power output. This load im- grid-to-grid a.f. voltage.

pedance seldom is the same as the modulating

Driver coupling -- A Class-A or Class-AB

impedance (� 5-3) of the Class-C r.f. stage, (� 3-4) driver is used to excite a Class-B so that a match must be brought about by stage. Tubes for the driver preferably should

adjusting the turn ratio of the coupling transformer. The required turn ratio, primary to secondary, is

be triodes having low plate resistance, since these will have the best regulation. Having chosen atube or tubes with ample power out-

91 CIIAPTER FIVE

5heRadio -Atnuieur'� -llaticgoolz

put from tube data sheets, the peak output voltage will be, approximately,
= 1.4 V PR
where P is the power output and R the load resistance. The input transformer ratio, primary to secondary, will be
E.
where E. is as given above and E, is the peak grid-to-grid voltage required by the modulator tubes.
Commercial transformers usually are designed for specific driver-modulator combinations, and usually are adjusted to give as good driver regulation as the conditions will permit.
Grid bias --Modern Class-B audio tubes are intended for operation without fixed bias. This lessens the variable grid-circuit loading effect and eliminates the need for a grid-bias supply.
When agrid-bias supply is required, it must have low internal resistance so that the flow of grid current with excitation of the Class-B tubes does not cause a continual shift in the actual grid bias and thus cause distortion. Batteries or a regulated bias supply (� 8-9) should be used.
Plate supply--The plate supply for a Class-B modulator should be sufficiently well filtered (� 8-3) to prevent hum modulation of the r.f. stage (� 5-2). An additional requirement is that the output condenser of the supply should have low reactance (� 2-8) at 100 cycles or less compared to the load into which each tube is working, which is Yi the plate-toplate load resistance. A 4-gfd. output condenser with a 1000-volt supply, or a 2-pfd. condenser with a 2000-volt supply, usually will be satisfactory. With other plate voltages, condenser values should be in inverse proportion to the plate voltage.
Overexcitation-- When aClass-B amplifier is overdriven in an attempt to secure more than the rated power, distortion in the output waveshape increases rapidly. The high-frequency harmonics which result from the distortion (� 3-3) modulate the transmitter, producing spurious sidebands (� 5-2) which readily can cause serious interference over a band of frequencies several times the channel width required for speech. This may happen even though the transmitter is not being overmodulated, as in the case where the modulator is incapable of delivering the power required to modulate the transmitter fully, or when the Class-C amplifier is not adjusted to give the proper modulating impedance (� 5-3).
The tubes used in the Class-B modulator should be capable of somewhat more than the power output nominally required (50% of the

d.c. input to the modulated amplifier) to take care of losses in the output transformer. These usually run from 10% to 20% of the tube output. In addition, the Class-C amplifier should be adjusted to give the proper modulating impedance and the correct output transformer turn ratio should be used. Such high-frequency harmonics as may be generated in these circumstances can be reduced by connecting condensers across the primary and secondary of the output transformer (about 0.002 Ilfd. in the average case) to form, with the transformer leakage inductance (� 2-9) a low pass filter (� 2-11) which cuts off just above the maximum audio frequency required for speech transmission (about 4000 cycles). The condenser voltage ratings should be adequate for the peak a.f. voltages appearing across them.
Operation without load-- Excitation should never be applied to aClass-B modulator until after the Class-C amplifier is turned on and is drawing the proper plate current to present the rated load to the modulator. With no load to absorb the power, the primary impedance of the transformer rises to a high value and excessive audio voltages are developed across it -- frequently high enough to break down the transformer insulation. If the modulator is to be tested separately from the transmitter a load resistance of the same value as the modulating impedance, and capable of dissipating the full power output of the modulator, should be connected across the transformer secondary.
� 5-7 LOW-LEVEL MODULATORS
Selection of tubes-- Modulators for gridbias and suppressor modulation usually can be small audio power output tubes, since the audio power required is quite small. A triode such as the 2A3 is preferable because of its low plate resistance, but pentodes will work satisfactorily.
Matching to load--Since the ordinary Class-A receiving power tube will develop about 200 to 250 peak volts in its plate circuit, which is ample for most low-level modulator applications, a 1:1 coupling transformer is generally used. If more voltage is required, astep-up ratio must be provided in the transformer. It is usual practice to load the primary of the output coupling transformer with aresistance equal to or slightly higher than the rated load resistance for the tube in order to stabilize the voltage output and thus improve the regulation. This is indicated in Figs. 506 and 507.
� 5-8 MICROPHONES
Sensitivity--The sensitivity of a microphone is its electrical output for agiven speech intensity input. Sensitivity varies greatly with

92 CHAPTER FIVE

lea choie teen ony

microphones of different basic types, and also varies between different models of the same type. The output is also greatly dependent on the character of the individual voice and the distance of the speaker's lips from the microphone, decreasing approximately as the square of the distance. It also may be affected by reverberation in the room. Hence, only approximate values based on averages of "normal" speaking voices can be attempted. The values given in the following paragraphs are based on close talking; that is, with the microphone six inches or less from the speaker's lips.
Frequency response -- The frequency response of a microphone is its relative ability to convert sounds of different frequencies into alternating current. With fixed sound intensity at the microphone, the electrical output may vary considerably as the sound frequency is varied. For intelligible speech transmission only a limited frequency range is necessary, and natural-sounding speech can be obtained if the output of the microphone does not vary more than afew decibels (� 3-3) over arange of about 100 cycles to 3000 or 4000 cycles. When the variation in decibels is small between two frequency limits, the microphone is said to be fiat between those limits.
Carbon microphones -- Fig. 511 shows connections for single- and double-button carbon microphones, with a variable potentiometer included in each circuit for adjusting the button current to the correct value as specified with each microphone. The single-button
microphone consists of a metal diaphragm placed against an insulating cup containing loosely-packed carbon granules (microphone button). Current from abattery flows through the granules, the diaphragm being one connection and the metal back-plate the other. The primary of atransformer is connected in series with the battery and microphone. As the diaphragm vibrates its pressure on the granules
alternately increases and decreases, causing a corresponding increase and decrease of current

flow through the circuit, since the pressure
changes the resistance of the mass of granules. The change in current flowing through the
transformer primary causes an alternating voltage, of corresponding frequency and intensity, to be set up in the transformer secondary (� 2-9). The dpuble-button type operates similarly, but with two buttons in pushpull.
Good quality single-button carbon microphones give outputs ranging from 0.1 to 0.3 volt across 50 to 100 ohms; that is, across the primary winding of the microphone transformer. With the step-up of the transformer, a peak voltage of between 3and 10 volts across 100,000 ohms or so can be assumed available at the grid of the first tube. These microphones are usually operated with abutton current of 50 to 100 ma.
The sensitivity of good-quality doublebutton microphones is considerably less, ranging from 0.02 volt to 0.07 volt across 200 ohms.
With this type microphone and the usual push-pull input transformer, a peak voltage of 0.4 to 0.5 volt across 100,000 ohms or so can
be assumed available at the first speech amplifier grid. The button current with this type
microphone ranges from 5to 50 ma. per button. Crystal microphones -- The input circuit
for apiezo-electric or crystal type microphone
is shown in Fig. 511-E. The element in this type consists of apair of Rochelle salts crystals
cemented together, with plated electrodes. In the more sensitive types the crystal is
mechanically coupled to a diaphragm. Sound waves actuating the diaphragm cause the crystal to vibrate mechanically and, by piezoelectric action (� 2-10), to generate a corresponding alternating voltage between the electrodes, which are connected to the grid circuit of a vacuum tube amplifier as shown. The crystal type requires no separate source of
current or voltage. Although the sensitivity of crystal micro-
phones varies with different models, an output

D.BT.rain

//Mnxe.ttroans

Fig. 511 -- Speech input circuit arrangements for five generally used types of microphones. M I, single-button carbon; Ms, double-button carbon; Ms, condenser; M4, ribbon or velocity types; Ma, crystal type.
CHAPTER FIVE 93

a% Pali� Amaleur'� --llanclhooh

of 0.01 to 0.03 volt is representative for communication types. The sensitivity is affected by the length of the cable connecting to the first amplifier stage; the above figure is for lengths of 6 or 7 feet. The frequency characteristic is unaffected by the cable but the load resistance (amplifier grid r�sistor) does affect it, the lower frequencies being attenuated as the shunt resistance becomes less. Grid resistor values of 1megohm and higher should be used, 5megohms being acustomary figure.
Condenser microphones-- The condenser microphone of Fig. 511-C consists of a twoplate capacity with one plate stationary and the other, separated from the first by about a thousandth of an inch, athin metal membrane serving as adiaphragm. This condenser is connected in series with aresistor and d.c. voltage source. When the diaphragm vibrates the change in capacity causes a small charging current to flow through the circuit. The resulting audio voltage which appears across the resistor is fed to the tube grid through the coupling condenser.
The output of condenser microphones varies with different models, the high-quality type being about one-hundredth to one-fiftieth as sensitive as the double-button carbon microphone. The first amplifier tube must be built into the microphone since the capacity of a connecting cable would impair both output and frequency range.
Velocity and dynamic microphones -- In a velocity or ribbon microphone, the element acted upon by the sound waves is athin corrugated metallic ribbon suspended between the poles of a magnet. When made to vibrate the ribbon cuts the lines of force between the poles in first one direction and then the other, thus generating an alternating voltage.
The sensitivity of the velocity microphone, with asuitable coupling transformer, is about 0.03 to 0.05 volt.
The dynamic microphone is similar to the ribbon type in principle, but the ribbon is replaced by acoil attached to adiaphragm. The coil provides several turns of wire cutting the magnetic field, and thus gives greater sensitivity. A small permanent-magnet loud-speaker makes apractical dynamic microphone.
�5-9 THE SPEECH AMPLIFIER
Description--The function of the speech amplifier is to build up the weak microphone voltage to avalue sufficient to excite the modulator to the required output. It may have from one to several stages. The last stage nearly always must deliver acertain amount of audio power, especially when it is used to excite a Class-B modulator. Speech amplifiers for grid-bias modulation usually end in a power stage which also functions as the modulator.

The speech amplifier is frequently built as a separate unit from the modulator, and in such a case may be provided with a step-down transformer designed to work into a low impedance, such as 200 or 500 ohms (tube-toline transformer). When this is done, astep-up input transformer intended to work between the same impedance and the modulator grids (line-to-grid transformer) is provided in the modulator circuit. The line connecting the two transformers may be made any convenient length.
General design considerations--The last stage of the speech amplifier must be selected on the basis of the power output required from it; for instance, the power necessary to drive a Class-B modulator (� 5-6). It may be either single-ended or push-pull (� 3-3) the latter being generally preferable because of higher power output and lower harmonic distortion. Push-pull amplifiers may be either Class A, Class AB' or Class AB2 (� 3-4) as the power requirements dictate. If a Class A or AB' amplifier is used, the preceding stages may all be voltage amplifiers, but when aClassAB2 amplifier is used the stage immediately preceding it must be capable of furnishing the power consumed by its grids at full output. The requirements in this case are much the same as those which must be met by adriver for a Class-B stage (� 5-6), but the actual power needed is considerably smaller and usually can be supplied by one or two small receiving triodes. Any lower-level stages are invariably worked as purely voltage amplifiers.
The minimum amplification which must be provided ahead of the last stage is equal to the peak audio-frequency grid voltage required by the last stage for full output (peak grid-to-grid-voltage in the case of a push-pull stage) divided by the output voltage of the microphone or secondary of the microphone transformer if one is used (� 5-8). The peak a.f. grid voltage required by the output tube or tubes is equal to the d.c. grid bias in the case of a single-tube Class A amplifier, and approximately twice the grid bias for a pushpull Class-A stage. The requisite information for Class AB' and AB2 amplifiers can be obtained from the manufacturer's data on the type considered. If the gain is not obtainable in one stage, several stages must be used in cascade. When the output stage is operated Class AB2,due allowance must be made for the fact that the next-to-the-last stage must deliver power as well as voltage. In such cases suitable driver combinations are usually recommended by manufacturers of tubes and interstage transformers. The coupling transformer must be designed especially for the purpose.
The total gain provided by a multi-stage

94 CHAPTER FIVE

PaChOlefelIZOny

amplifier is equal to the product of the individual stage gains. For example, when three stages are used, the first having again of 100, the second 20 and the third 15, the total gain is 100 X 20 X 15, or 30,000. It is good practice to provide two or three times the minimum required gain in designing the speech amplifier. This will insure having ample gain available to cope with varying conditions.
When the gain must be fairly high, as when a crystal microphone is used, the speech amplifier frequently has four stages, including the power output stage. The first is generally apentode because of the high gain attainable with this type of tube. The second and third stages are usually triodes, the third frequently
having two tubes in push-pull when it drives a Class AB2 output stage. Two pentode stages
are seldom used consecutively because of the difficulty of getting stable operation when the gain per stage is high. With carbon micro-
phones less amplification is needed, hence the pentode first stage usually is omitted, one or two triode stages being ample to obtain full output from the power stage.
Stage gain and voltage output --In voltage amplifiers, the stage gain is the ratio of
a.c. output voltage to a.c. voltage applied to the grid. It will vary with the applied audio
frequency, but for speech work the variation should be small over the range 100-4000 cycles. This condition is easily met in practice.
The output voltage is the maximum value which can be taken from the plate circuit
without distortion. It is usually expressed in terms of the peak value of the a.c. wave (� 2-7)
since this value is independent of the wave-
form. The peak output voltage usually is of interest only when the stage drives a power amplifier, since only in this case is the stage called upon to work near its maximum capabilities. Low-level stages are very seldom worked near full capacity, hence harmonic distortion is negligible and the voltage gain of the stage is the primary consideration.
Resistance coupling -- Resistance coupling is generally used in voltage amplifier stages. It is relatively inexpensive, good frequency
response can be secured, and there is little danger of hum pick-up from stray magnetic fields associated with heater wiring. It is the
only type of coupling suitable for the output circuits of pentodes and high-� triodes, since with audio-frequency transformers a sufficiently high load impedance (� 3-3) cannot be
obtained without considerable frequency distortion. Typical resistance-coupled circuits are given in Fig. 512.
The frequency response of the amplifier will be determined by the circuit constants, particularly C3R4, the coupling condenser and
resistor to the following stage, and eel, the

followlno Stage'
(B) fa
Fig. 512 --Resistance-coupled voltage amplifier eir. cuits. A, pentode; B, triode. Designations are as follows:
Ci -- Cathode by-pass condenser.
C2 -- Plate by-pass condenser. Ca -- Output coupling condenser (blocking condenser). C4 -- Screen by-pass condenser. R; -- Cathode resistor. Ra -- Grid resistor. 118 -- Plate resistor. R4 - Next-stage grid resistor. Rs -- Plate decoupling resistor. Re -- Screen resistor.
Values for commonly-used tubes are given in Table I.
cathode bias resistor and by-pass condenser. For adequate amplification at low frequencies the time constant (� 2-6) of both these CR combinations should be large. Depending upon the type of tube used in the next stage, R4 may vary from 50,000 ohms (with power tubes such as the 2A3 or 6F6) to 1megohm; it is advantageous to use the highest value recommended for the type of tube used since this gives greatest low-frequency response with a given size of coupling condenser, Cg. A capacity of 0.1 1.4fd. at C3 will provide ample coupling at low frequencies with any ordinarily-used tube, load resistance (R3)and nextstage grid resistance (R4).
The reactance (� 2-8) of C1 must be small compared to the resistance of R1for good lowfrequency response. While with values of R1 in the vicinity of 10,000 ohms, mom or less, a condenser of 1dd. will suffice, it is more common practice to use 5- or 10-pfd. low-voltage electrolytic condensers for the purpose, since they are inexpensive and provide ample bypassing. A value of 10 �ifd. is usually sufficient with values of R1 as low as 500 ohms.

95 CHAPTER FIVE

phe Radio AmaleuA .-flandhoo4

For maximum voltage gain the resistance at R3 should be as high as possible without causing too great adrop in voltage at the plate of the tube. Values range from 50,000 ohms to 0.5 megohm, the smaller figure being used with triodes having comparatively low plate resistance. The value of R1 depends upon R39 which principally determines the plate current; in general, the grid bias is somewhat smaller than in circuits having low-resistance output devices (such as a transformer) because of the lower voltage effective at the plate of the tube. This is also true of the screen voltage, for similar reasons, and values of the screen resistor, R5, may vary from 0.25 to 2 megohms. A screen by-pass (C4) of 0.1 pfd. will be adequate in all cases.
Table Ishows typical values for some of the more popular tube types used in speech amplifiers. The stage gain and peak undistorted output voltage also are given. Other operating conditions are of course possible. The value of the grid resistor, R2, does not affect any of these quantities, but should not exceed the maximum value recommended by the manufacturer for the particular type of tube used.
The resistance-capacity filter (� 2-11) formed by C2R5 is called a decoupling circuit. It isolates the stage from the power supply so that unwanted coupling between this and other
(A)
fa
CB)
Fig. 513 -- Transformer-coupled amplifier circuits for driving apush-pull amplifier. A, resistance-transformer coupling; B, transformer coupling. Designations correspond to those of Fig. 512. In A, values can be taken from Table I. In B, the cathode resistor is calculated from the rated plate current and grid bias as given for the particular type of tube used ( 3-6).
96 CHAPTER FIVE

stages through the output impedance of the power supply is eliminated. Such coupling is a frequent cause of low-frequency oscillation (motorboating) in multistage resistance-coupled amplifiers.
Transformer coupling -- Transformer cou-
pling between stages is ordinarily used only when power is to be transferred (in such acase resistance coupling is very inefficient) or when it is necessary to couple between a singleended and a push-pull stage. Triodes having an amplification factor of 20 or less are used in transformer-coupled voltage amplifiers.
Representative circuits for single-ended to push-pull are shown in Fig. 513. That at A uses a combination of resistance and transformer coupling and may be used for exciting the grids of a Class-A or AB I following stage. The resistance coupling is used to keep the d.c. plate current from flowing through the transformer primary and thereby prevent a reduction in primary inductance below its nocurrent value (� 8-4). This improves the lowfrequency response. With triodes ordinarily used (6C5, 6J5, etc.) the gain is equal to that with resistance coupling (typical values in Table I) multiplied by the secondary-toprimary turn ratio of the transformer. This ratio is generally 2:1
In B the transformer primary is in series with the plate of the tube and thus must carry the tube plate current. When the following amplifier operates without grid current, the
voltage gain of the stage is practically equal to the et of the tube multiplied by the transformer ratio. This circuit is also suitable for transferring power (within the capabilities of the tube) as in the case of afollowing Class-AB2 stage used as adriver for aClass-B modulator.
Gain control -- The overall gain of the amplifier may be changed to suit the output level of the microphone, which will vary with voice intensity and distance of the speaker from the microphone, by varying the proportion of a.c. voltage applied to the grid of one of the stages. This is done by means of an adjustable voltage divider (� 2-6), commonly called a "potentiometer" or "volume control," as shown in Fig. 514. The actual voltage applied between grid and cathode will be very nearly equal to the ratio of the resistance between AB to the total resistance AC, multiplied by the a.c.
voltage which appears across AC. The gain control is usually also the grid resistor for the amplifier stage with which it is associated.
The gain control potentiometer should be near the input end of the amplifier so that there will be no danger that stages ahead of the gain control will overload. With carbon microphones the gain control may be placed directly across the microphone transformer
secondary, but with other types the gain con-

leichotelphony

Fig. 514 -- Gain control circuit.
trol usually will affect the frequency response of the microphone when connected directly across it. The control is therefore usually placed in the grid circuit of the second stage.
Phase inversion --Push-pull output may be secured with resistance coupling by using an extra tube as shown in Fig. 515. There is a phase shift of 180 degrees through any normally-operating resistance-coupled stage (� 3-3) and the extra tube is used purely to provide this phase shift without additional gain. The outputs of the two tubes are then added to give push-pull excitation to the next amplifier.
In Fig. 515, V1 is the regular amplifier, connected in normal fashion to the grid of one of the push-pull tubes. The next-stage grid resistor is tapped so that part of the output voltage is fed to the grid of the phase inverter, V2. This tube then amplifies the signal and applies it in reverse phase to the grid of the second push-pull tube. Two similar tubes should be used at V1 and V2, with identical plate resistors and output coupling condensers. The tap on R4is adjusted to make V1 and V2 give equal voltage outputs so that balanced excitation is applied to the grids of the following stage.
The cathode resistor, R6,commonly is left un-bypassed since this tends to help balance the circuit. Double-triode tubes are frequently used as phase inverters.

TABLE I--TYPICAL VOLTAGE AMPLIFIER II ATA

Peak

Tube Type

RS,

R6, /il, ohms Output

megohma megohms

Volts

Voltage Gain

6C5 615 1;9'5, 6SF5 617 68.17

0.1 0.1 0.25 0.25 0.25 0.5

-- -- -- 1.2 1.0 2.0

6000 3000 3000 1200
900 1300

104

140

167

200

Other values (Fig. 512): Ci, 10 pfd. (low-voltage electrolytic); CS, 8-pfd. electrolytic; CS. C4, 0.1-pfd. paper; RS, 0.1 to 1megohm; R4. 0.5 megohm; Rg, 10,000 to 50,000 ohms. Data are based on a plate-supply voltage of 300; lower values will reduce the undistorted peak output voltage in proportion, but will not.materially affect the voltage gain.

Output limiting --It is desirable to modulate as heavily as possible without overmodulating, yet it is difficult to speak into the microphone at a constant intensity. To maintain reasonably constant output from the modulator in spite of variations in speech intensity, it is possible to use automatic gain control which follows the average (not instantaneous) variations in speech amplitude. This is accomplished by rectifying and filtering (� 8-2, 8-3) some of the audio output and applying the rectified and filtered d.c. to acontrol electrode in an early stage in the amplifier.
A practical circuit for this purpose is shown in Fig. 516. The rectifier must be connected, through the transformer, to atube capable of
Fig. 515 --Phase inverter circuit for resistancecoupled push-pull output. With a double-triode tube (6N7) the following values are typical: Ri -- 0.5 megohm. Rs, R3 -- 0.1 rnegohm. R4, R5 -- 0.5 megohm. Re -- 1500 ohms. Cl, CS-- 0.1 �M .
R4 should he tapped as described in the text. The voltage gain with these constants is 22.
delivering some power output (a small part of the output of the power stage may be used) or else a separate amplifier for the rectifier circuit alone may have its grid connected in parallel with that of the last voltage amplifier. Resistor R4 in series with R5 across the plate supply provides variable bias on the rectifier plates so that the limiting action can be delayed until adesired microphone input level is reached. R2, R3, C2, C3,and C4 form the filter, (� 2-11) and the output of the rectifier is connected to the suppressor grid of the pentode first stage of the speech amplifier.
A step-down transformer giving about 50 volts when its primary is connected to the output circuit should be used. A half-wave rectifier can be used instead of the full-wave circuit shown, although satisfactory filtering is more difficult.
Noise -- It is important that the noise level in a speech amplifier be low compared to the level of the desired signal. Noise in the speech amplifier is chiefly hum, which may be the result of insufficient power-supply filtering or may be introduced into the grid circuit of a

97 CHAPTER FIVE

.9he Radio AmaleuA fiandloo4

picture of the modulated output of the trans-

6H6

mitter, and the waveform errors inherent in

other types of measurements are eliminated.

Two types of oscilloscope patterns may be

e 4
3

obtained, known as the "wave envelope" and

"trapezoid." The former shows the shape of

FIRST SPEECH AMP

the modulation envelope (� 5-2) directly, while the latter in effect plots the modulation char-

acteristic (� 5-2) of the modulated stage on

the cathode-ray tube screen. To obtain the

wave-envelope pattern the oscilloscope must

have ahorizontal sweep circuit. The trapezoid

Fig. 516 -- Output limiting circuit. Ci, Cs, Cs, C4 -- 0.1-pfd. paper. Bi, R2, Ra -- 0.25 niegohm. Rs --25,000-ohm potentiometer. Rs --0.1 megohm. T --See text.

pattern requires only the oscilloscope, the
sweep circuit being supplied by the transmitter itself. Fig. 517 shows methods of connecting
the oscilloscope to the transmitter for both types of patterns. The oscilloscope connections for the wave-envelope pattern, Fig. 517-A,

tube by magnetic or electrostatic means from are usually simpler than those for the trapeheater wiring. The plate voltage for the ampli- zoidal figure. The vertical deflection plates are

fier should be free from ripple (� 8-4), particularly the voltage applied to the low-level stages. A two-section condenser-input filter

coupled to the amplifier tank coil or an antenna coil by means of a pickup coil of a few turns connected to the oscilloscope through a

(� 8-5) is usually satisfactory. The decoupling twisted-pair line. The position of the pickup

circuits mentioned in the preceding paragraphs also are helpful in reducing platesupply hum.
Hum from heater wiring may be reduced

coil is varied until a carrier pattern, Fig. 518-B, of suitable height is obtained. The sweep voltage should be adjusted to make the width of the pattern somewhat more than

by keeping the wiring well away from un- half the diameter of the screen. It is frequently grounded components or wiring, particularly in helpful in eliminating r.f. harmonics from the

the vicinity of the grid of the first tube. Com- pattern to connect a resonant circuit, tuned

plete shielding of the microphone jack is

advisable, and when tubes with grid caps instead of the single-ended types are used the

Ant Circuit

OS C

caps and the exposed wiring to them should

be shielded. Heater wiring preferably should
run in the corners of ametal chassis to reduce the magnetic field. A ground should be made

Final Tank

either on one side of the heater circuit or to

the center-tap of the heater winding. The

shells of metal tubes should be grounded;

(A)

glass tubes require separate shields, especially when used in low-level stages. Heater connec-

>ce

tions to the tube sockets should be kept as far

as possible from the plate and grid prongs,

and the heater wiring to the sockets should be

(B)

kept close to the chassis. A connection to a

Alternative Input Connectiens Ant Circuit

good ground (such as a cold water pipe) also

is advisable. The speech amplifier always

should be constructed on ametal chassis.

When the power supply is mounted on the

same chassis with the speech amplifier, the

power transformer and filter chokes should be

well separated from audio transformers in the

amplifier proper, to reduce magnetic coupling.

� 0-10 CHECKING 'PHONE TRANS. MITTER OPERATION
Modulation percentage-- The most reliable method of determining percentage of modulation is by means of the cathode-ray oscilloscope (� 3-9). The oscilloscope gives a direct

Fig. 517-- Methods of connecting an oscilloscope to the modulated amplifier for checking modulation.

98 CHAPTER FIVE

elide/1,4.y

(A) NO CARRIER
1111 111111 -z
(B) CARR IER ONLY

(c)

LESS THAN

(H)

100% moDuLArioN

(D) 100% MODULATION

(E)

OVER MODULATION

(J)

Fig. 518--Wave-envelope and trapezoidal patterns under different conditions of modulation.

to the operating frequency, between the vertical deflection plates, using link coupling between this circuit and the transmitter tank circuit.
With the application of voice modulation a
rapidly-changing pattern of varying height will be obtained. When the maximum height of this pattern is just twice that of the carrier alone, the wave is being modulated 100% (� 5-2).This is illustrated by Fig. 518-D, where the point X represents the sweep line (reference line) alone, YZ is the carrier height, and PQ
is the maximum height of the modulated wave. If the height is greater than the distance PQ, as illustrated in E, the wave is overmodulated in the upward direction. Overmodulation in the downward direction is indicated by agap in the pattern at the reference axis, where a single bright line appears on the screen. Overmodulation in either direction may take place even when the modulation in the other direction is less than 100%. Assuming that the modulation i8 symmetrical, however, any

modulation percentage can be measured directly from the screen by measuring the maximum height with modulation and the height
of the carrier alone; calling these two heights 11. 1 and h2,respectively, the modulation percentage is

hi -- hz h2

X 100

Connections for the trapezoidal pattern are shown in Fig. 5I7-B. The vertical plates are similarly coupled to the transmitter tank circuit through a pick-up loop; the tuned input circuit to the oscilloscope may also be used. The horizontal plates are coupled to the output of the modulator through a voltage
divider (� 2-6) R1E2, the latter resistance being variable to permit adjustment of the audio voltage to a suitable value to give a satisfactory horizontal sweep on the screen. R2 may be a 0.25-megohm volume control resistor. The value of R1 will depend upon the audio output voltage of the modulator. This voltage is equal to VPR, where P is the audio power output of the modulator and R is the modulating impedance of the modulated r.f. amplifier. In the case of grid-bias mo�lulation with a 1:1 output transformer, it will be satisfactory to assume that the a.c. output voltage of the modulator is equal to 0.7E for a single
tube, or 1.4E for a push-pull stage, where E is the d.c. plate voltage on the modulator. If the transformer ratio is other than 1:1, the
voltage so calculated should be multiplied by the actual secondary-to-primary turn ratio. The total resistance of R1 and R2 in series should be 0.25 megohm for every 150 volts of modulator output; for example, if the modula-
tor output voltage is 600, the total resistance should be four (600/150) times 0.25 megohm,
or 1megohm. Then with 0.25 megohm at R2, R1 should be 0.75 megohm. The blocking condenser C should be 0.1 pfd or more and its voltage rating should be greater than the
maximum voltage appearing in the circuit. With plate modulation, this is twice the d.c.
voltage applied to the plate of the modulated amplifier.
The trapezoidal patterns are shown in Fig. 518 at F to J, each alongside the corresponding
wave-envelope pattern. With no signal, only the cathode ray spot appears on the screen. When the unmodulated carrier is applied a vertical line appears, and its length should be adjusted by means of the pickup coil coupling to a convenient value. When the carrier is
modulated the wedge-shaped pattern appears; the higher the modulation percentage the wider and more pointed the wedge becomes. At 100% modulation it just makes a point on the axis A at one end and the height PQ at the other end is equal to twice the carrier height YZ.

99 CHAPTER FIVE

n eleach� AmaieuA ilancgooh

Overmodulation in the upward direction is

indicated by increased height over PQ, and

in the downward direction by an extension

along the axis X at the pointed end. The modu-

lation percentage may be found by measur-

ing the modulated and unmodulated carrier

heights in the same way as with the wave

envelope pattern.

Non-symmetrical waveforms -- In voice

waveforms the average maximum amplitude
in one direction from the axis is frequently

greater than in the other direction, although
the average energy on both sides is the same.

iFnorthtehiusp

reason the percentage of modulation direction frequently differs from that

in the down direction, and with agiven voice

and microphone this difference in modulation

percentage is usually always in the same direc-

tion. Since overmodulation in the downward

direction causes more out-of-channel interfer-

ence than overmodulation upward because

of the steeper wavefront (� 6-1), it is advisable

to "phase" the modulation so that the side

of the voice waveform having the larger excur-

sions causes the instantaneous carrier power to

increase and the smaller excursions to cause a

power decrease. This reduces the likelihood

of overmodulation on the down peak. The

direction of the larger excursions can readily be

found by careful observation of the oscillo-

scope pattern. The phase can be reversed by

reversing the connections of one winding of

any transformer in the speech amplifier or

modulator.

Modulation monitoring -- While it is de-

sirable to modulate as fully as possible, 100%

modulation should not be exceeded, particu-

larly in the downward direction, because har-

monic distortion will be introduced and the

channel width increased (� 5-2), thus causing

unnecessary interference to other stations.

The oscilloscope may be used to provide a

continuous check on the modulation, but sim-

pler indicators may be used for the purpose,

once calibrated. A convenient indicator, when

a Class-B modulator (� 5-6) is used, is the

plate milliammeter in the Class-B stage, since

plate current fluctuates with the voice inten-

sity. Using the oscilloscope, determine the

gain-control setting and voice intensity which

gives 100% modulation on voice peaks, and

simultaneously observe the maximum Class-B

plate-milliammeter reading on the peaks.

When this maximum reading is obtained, it

will suffice in regular operation to adjust the

gain so that it is not exceeded.

A sensitive rectifier-type voltmeter (copper

oxide type) can also be used for modulating

monitoring. It should be connected across the

output circuit of an audio driver stage where

the power level is a few watts, and similarly

calibrated against the oscilloscope to determine

the reading which represents 100% modulation.
The plate milliammeter of the modulated r.f. stage may also be used as an indicator of overmodulation. Since the average plate current is constant (� 5-3, 5-4, 5-5) when the amplifier is linear, the reading will be the same with or without modulation. When
the amplifier is overmodulated, especially in the downward direction, the operation is no longer linear and the average plate current will change. A flicker of the pointer may therefore be taken as an indication of overmodula-
tion or non-linearity. However, it is possible that the average plate current will remain constant with considerable overmodulation under some operating conditions, so such an indicator is not wholly reliable unless it has been previously checked against an oscilloscope.
Linearity -- The linearity (� 5-2) of amodulated amplifier may readily be checked with the oscilloscope. The trapezoidal pattern is more easily interpreted than the wave envelope pattern and less auxiliary equipment is required. The connections are the same as for measuring modulation percentage (Fig. 517). If the amplifier is perfectly linear, the sloping sides of the trapezoid will be perfectly straight from the point at the axis up to at least 100% modulation in the upward direction. Nonlinearity will be shown by curvature of the sides. Curvature near the point, extending the point farther along the axis than would occur with straight sides, indicates that the output power does not decrease rapidly enough in this region; it may also be caused by imperfect neutralization (a push-pull amplifier is recommended because better neutralization is possible than with single-ended amplifiers) or r.f. leakage from the exciter through the final stage. The latter condition can be checked by removing the plate voltage from the modulated stage, when the carrier should disapriear and only the beam spot remain on the screen (Fig. 518-F). If a small vertical line remains the amplifier should be re-neutralized to eliminate it; if this does not suffice, r.f. is being picked up from lower-power stages either by coupling through the final tank circuit or through the oscilloscope pickup circuit.
Inward curvature at the large end of the pattern is caused by improper operating conditions of the modulated amplifier, usually improper bias or insufficient excitation, or both, with plate modulation. In grid-bias and cathode-modulated systems, the bias, excitation and plate loading are not correctly preportioned when such curvature occurs, usually because the amplifier has been adjusted to have too-high carrier efficiency without modulation (� 5-4, 5-5).

loo CII APTER

fedideferhony

For the wave-envelope pattern it is necessary to have alinear horizontal sweep circuit in the oscilloscope and a source of sine-wave

OPERATING

1/Cnarrnroideulrafed

audio signal (such as an audio oscillator or signal generator) which can be synchronized with the sweep circuit. The linearity can be

(A) IPO�O; USPF WAORR D
MODULATION AuLSdiwimiiortGsl-roifd

judged by comparing the wave envelope with a

true sine wave. Distortion in the audio cir-

cuits will affect the pattern in this case (such distortion has no effect on the trapezoidal pattern, which shows the modulation characteristic of the r.f. amplifier alone), and it is also readily possible to misjudge the shape of the

&modulated

AMPLIFIER

Carrier

(ED O LIRGHETnLI YTAWTIAODNED

INCORRECT Abuod.uitosGorrid Swin9

modulation envelope, so that the wave enve-

lope is less useful than the trapezoid for check-

ing linearity of the modulated amplifier.
Fig. 519 shows typical patterns of both types. The cause of the distortion is indicated for grid-bias and suppressor modulation. The patterns at A, although not truly linear, are representative of properly-operated grid-bias modulation systems. Better linearity can be obtained with plate modulation of a Class-C amplifier.
Faulty patterns--The drawings of Figs.

annwidaled

OvERnoommioN Can'ei"

CAUSED BY U-

s." CESSIVE AUDIO

GRID VOLTAGE

ends of Audio Grid

Swing

Fig. 519-- Oscilloscope patterns representing proper and improper grid-bias or cathode modu ation. The pattern obtained with a correctly adjusted amplifier is shown at A. The other two drawings indicate nonlinear modulation.

518 and 519 show what is normally to be expected in the way of pattern shapes when the oscilloscope is used to check modulation. If the actual patterns differ considerably from
those shown, it is probable that the pattern is faulty rather than the transmitter. It is important that only r.f. from the modulated stage be coupled to the oscilloscope, and then only to the vertical plates. The effect of stray r.f, from other stages in the transmitter has

characteristic with grid-bias modulation there is normally a slight upward change in plate current of astage so modulated, but this occurs only at high modulation percentages and is
barely detectable under the usual conditions of voice modulation.
With plate modulation, a downward shift in plate current may indicate one or more of the following:

been mentioned in the preceding paragraph. If r.f. is also present on the horizontal plates,

1. Insufficient excitation to the modulated r.f. amplifier.

the pattern will lean to one side instead of being upright. If the oscilloscope cannot be

2. Insufficient grid bias on the modulated stage.

moved to aspot where the unwanted pick-up disappears, a small by-pass condenser (10

3. Wrong load resistance for Class-C r.f. amplifier.

gpfd.) should be connected across the horizon-

4. Insufficient output capacity in filter of

tal plates as close to the cathode-ray tube as

modulated amplifier plate supply.

possible. An r.f. choke (2.5 mh. or smaller)

5. Heavy overloading of Class-C r.f. ampli-

may also be connected in series with the un-

fier tube or tubes.

grounded horizontal plate. "Folded" trapezoidal patterns occur when
the audio sweep voltage is taken from some point in the audio system other than that where the a.f. power is applied to the modulated stage, and are caused by aphase differ-
ence between the sweep voltage and the modulating voltage. The connections should always be as shown in Fig. 517-B.

Any of the following may cause au upward shift in plate current:
1. Overmodulation (excessive audio power, audio gain too great).
2. Incomplete neutralization of the modulated amplifier.
3. Parasitic oscillation in the modulated amplifier.

Plate-current shift-- As mentioned above,
the d.c. plate current of amodulated amplifier will be the same with and without modulation so long as the amplifier operation is perfectly
linear and other conditions remain unchanged. This also assumes that the modulator is working within its capabilities. Because there is usually some curvature of the modulation

When a common plate supply is used for both Class-B (or Class AB) modulator and modulated r.f. amplifier, the plate current of the latter may "kick" downward because of poor power-supply voltage regulation (� 8-1) with the varying additional load of the modulator on the supply. The same effect may occur with high-power transmitters because of poor

CHAPTER FIVE 101

Dfl e Pali� Amaieur'� -11unitool

regulation of the a.c. supply mains, even when a separate power supply unit is used for the

input hum is the likely cause. The various parts of the transmitter may be checked through

Class-B modulator. Either condition may be detected by measuring the plate voltage applied to the modulated stage; in addition, poor
line regulation may also be detected by a downward shift in filament or line voltage.
With grid-bias modulation, any of the following may be the cause of a plate current shift greater than the normal mentioned above:
Downward kick: Too much r.f. excitation;
insufficient operating bias; distortion in modulator or speech amplifier; too-high resistance in
bias supply; insufficient output capacity in plate-supply filter to modulated amplifier; amplifier plate circuit not loaded heavily enough;
plate-circuit efficiency too high under carrier
conditions. Upward kick: Overmodulation (excessive
audio voltage); distortion in audio system; regeneration because of incomplete neutralization; operating grid bias too high.
A downward kick in plate current will ac-
company an oscilloscope pattern like that of Fig. 519-B; the pattern with an upward kick will look like Fig. 519-A with the shaded portion extending farther to the right and
above the carrier, for the "wedge" pattern. Noise and hum on carrier -- These may be
detected by listening to the signal on areceiver sufficiently removed from the transmitter to
avoid overloading. The hum level should be low compared to the voice at 100% modulation. Hum may come either from the speech amplifier and modulator or from the r.f. �ection of the transmitter. Hum from the r.f. section can be detected by completely shutting

as described above. Spurious sidebands -- A superheterodyne
receiver having a crystal filter (� 7-8, 7-11) is needed for checking spurious sidebands
outside the normal communication channel (� 5-2). The r.f. input to the receiver must be kept low enough, by removing the antenna or by adequate separation from the transmitter, to avoid overloading and consequent spurious receiver responses (� 7-8). With the crystal filter in its sharpest position and the beat oscillator turned on, tune through the region outside the normal channel limits (3 to 4kc. each
side of the carrier) while another person talks into the microphone. Spurious side bands will be observed as intermittent beat notes coin-
ciding with voice peaks, or in bad cases of
distortion or overmodulation as "clicks" or crackles well away from the carrier frequency. Sidebands more than 4 kc. from the carrier should be of negligible strength in aproperly modulated 'phone transmitter. The causes are overmodulation or non-linear operation (� 5-3).
R.f. in speech amplifier -- A small amount
of r.f. current in the speech amplifier -- particularly in the first stage, which is most susceptible to such r.f. pick-up -- will cause overloading and distortion in the low-level stages.
Frequently also there is a regenerative effect which causes an audio-frequency oscillation or "howl" to be set up in the audio system. In such cases the gain control cannot be advanced very far before the howl builds up, even though the amplifier may be perfectly stable when the r.f. section of the transmitter

off the modulator; if hum remains when this is done the power-supply filters for one or more of the r.f. stages have insufficient smoothing (� 8-4). With a hum-free carrier, hum introduced by the modulator can be checked by turning on the modulator but leaving the speech amplifier off; power-supply filtering is the likely source of such hum. If carrier
and modulator are both clean, connect the speech amplifier and observe the increase in
hum level. If the hum disappears with the gain control at minimum, the hum is being intro-
duced in the stage or stages preceding the gain control. The microphone may also pick up hum, a condition which can be checked by removing the microphone from the circuit
but leaving the first speech-amplifier grid circuit otherwise unchanged. A good ground on the microphone and speech system is usually

is not turned on. Complete shielding of the microphone,
microphone cord and speech amplifier are
necessary to prevent r.f. pick-up, and aground connection separate from that to which the
transmitter is connected is advisable. Unsymmetrical or capacity coupling to the antenna (single-wire feed, feeders tapped on final tank circuit, etc.) may be responsible in that these systems sometimes cause the transmitter chassis to take an r.f. potentional above ground. Inductive coupling to a two-
wire transmission line is advisable. This antenna effect can be checked by disconnecting the antenna and dissipating the power in a dummy antenna (� 4-9) when it usually will be found that the r.f. feedback disappears. If it does not, the speech amplifier and microphone shielding are at fault.

essential to hum-free operation. Hum can be checked with the oscilloscope,
where it appears as modulation on the carrier in the same way as the normal modulation. Usually the percentage is rather small, but if the carrier shows modulation with no speech

�5-11 FIt EQUENCY MODULATION
Principles --In frequency modulation the carrier amplitude is constant and the output frequency of the transmitter is made to vary about the carrier or mean frequency at arate

102 CHAPTER FIVE

i()achoietephony

corresponding to the audio frequencies of the

speech currents. The extent to which the fre-

quency changes in one direction from the un-

modulated or carrier frequency is called the

frequency deviation. It corresponds to the

change of carrier amplitude in the amplitude-

modulation system (� 5-2). Deviation is us-

ually expressed in kilocycles, and is equal to

the difference between the carrier frequency

and either the highest or lowest frequency

reached by the carrier in its excursions with modulation. There is no modulation percent-

age in the usual sense; with suitable circuit de-

KILOCYCLES

sign the deviation may be made as large as desired without encountering any effec equivalent to overmodulation in the amplitud

Fig. 520 -- Triangular spectrum of noise response in an f.m. receiver compared with amplitude modulation. Deviation ratios of 1and 5are shown.

system.
Deviation ratio--The ratio of the ma mum frequency deviation to the audio fr quency of the modulation is called the devi
lion ratio. Unless otherwise specified, it is
taken as the ratio of the maximum frequency deviation to the highest audio frequency to be transmitted.
Advantages of f.m. -- The chief advantage of frequency modulation over amplitude modulation is noise reduction at the receiver. All electrical noises in the radio spectrum, including those originating in the receiver, are r.f.
oscillations which vary in amplitude, this variation causing the noise response in ampli-
tude-modulation receivers. If the receiver does not respond to amplitude variations but
only to frequency changes, noise can affect it only by causing aphase shift which appears as frequency modulation on the signal. The effect
of such frequency modulation by the noise
can be made small by making the frequency change (deviation) in the signal large.
A second advantage is that the power required for modulation is inconsequential, since there is no power variation in the modulated output.
Triangular spectrum-- The way in which noise is reduced by a large deviation ratio i illustrated by Fig. 520. In this figure the nois is assumed to be evenly distributed over t channel used, an assumption which is alm always true. It is also assumed that au frequencies above 4000 cycles (4 kc.) are n necessary to voice communication, and that t audio system in the receiver has no respons above this frequency. Then if an amplitude modulation receiver is used and its selectivity
is such that there is no attenuation of sidebands (� 5-2) below 4000 cycles, the noise components of all frequencies within the channel will produce equal response when they beat with a carrier centered in the channel. This is shown by the line DC.
In the f.m. receiver the output amplitude is proportional to the frequency deviation, and

noise components in the channel can be considered to frequency-modulate the steady

carrier with a deviation proportional to the

difference between the actual frequency of the

component and the frequency of the carrier,

and also to give an audio-frequency beat of

the same frequency difference. This leads to a

rising response characteristic such as the line

OC, where the noise amplitude is proportional

to the audio beat frequency. The average noise

power loutput is proportional to the square

roo � ,.the sum of the squares of all the ampli-

tudelrvalues (� 2-7), so that the noise power

with f.m. having adeviation ratio of 1is only

that with amplitude modulation, or an improvement of 4.75 db.

'fee deviation ratio is increased to 5, the

noise response is represented by the line OF,

but only frequencies up to 4000 cycles are re-

produced in the output so the audible noise

is confined to the triangle OAB. These relations

only hold when the carrier is strong compared

to the noise. With weak signals the signal-to-

noise ratio is better with adeviation ratio of 1.

Linearity-- A transmitter in which fre-

quency deviation is directly proportional to

the amplitude of the modulating signal is said

to be linear. It is also essential that the carrier

amplitude remain constant under modulation,

which in turn requires that the transmitter

tuned circuits be broad enough to handle with-

out discrimination the range of frequencies

transmitted. This requirement is easily met

under ordinary conditions.

Sidebands -- In frequency modulation there

is a series of sidebands on either side of the

carrier frequency for each audio-frequency

component in the modulation. In addition to

the usual sum and difference frequencies

(� 5-2) there are also beats at harmonics of

the fundamental modulating frequency, even

the latter may be apure tone. This oc-

c 'ecause of the necessity for maintaining

the roper phase relationships between the

carrier and sidebands to keep the power output co; tant. Hence afrequency-mod ulated signal

103 CHAPTER FIVE

Dhe Radio Amateur'3 -naLgooh

inherently occupies a wider channel than an amplitude-modulated signal, and be use of the necessity for conserving space in the usual
communication spectrum the use of f.m. is confined to the ultra-high frequencies.
The number of sidebands for asingle modulating frequency increases with the frequency
deviation. When the deviation ratio is of the order of 5the sidebands beyond the ms,ximum

through the oscillator tank, giving the same effect as though an inductance were connected across the tank (in an inductance the current lags the voltage by 90 degrees -- �2-8). The
frequency is therefore increased in proportion to the lagging plate current of the modulator. This in turn is determined by the a.f. voltage applied to the No. 3grid of the 6L7, hence the oscillator frequency varies with the audio

frequency deviation are usually negligible, so that the channel required is approximately twice the frequency deviation.

Other circuit arrangements to produce the same effect can be used. It is convenient to use atube (such as the 6L7) in which the r.f. and

� 542 METHODS OF FREQUENCY

a.f. voltages can be applied to separate con-

MOM II.ATION

trol grids; however, both voltages may be

Requirements and methods -- At present there are no fixed standards of frequency deviation in amateur work. Since a deviati n ratio
of 5is considered high enough in any cse, the
fmoar xainmuupmpedrevaiuadtiioo-nfrneeqcueesnscaryyliismi1t5oftp2o0w koc.r

applied to the same grid with suitable precautions taken to prevent r.f. from flowing in the external audio circuit and vice versa (� 2-13).
The modulated oscillator is usually operated on a relatively low frequency so that a high order of carrier stability can be secured. Fre-

4000 cycles (� 5-2), or a channel width of 30 quency multipliers are used to raise the freto 40 kc. The permissible deviation is deter- quency to the final frequency desired. The

mined by the receiver (� 7-18), since deviation

beyond the limits of the receiver pass-band

413

causes distortion. If the transmitter is designed

to be linear (� 5-11) with adeviation of about

15 kc. it can be used at lower deviation ratio

simply by reducing the gain in the speech

amplifier, and thereby made to conform to the

requirements of the particular receiver i use.

The several possible methods of fre

Osc. Tank

modulation include mechanical (for ins , ce,

varying condenser plate spacing in accor e nee

AT Input

with voice vibrations), initial phase modula-

tion which is later transformed into frequency modulation, and direct frequency modulation of an oscillator by electrical means. Th latter, in the form of the reactance modnl&orl is the
simplest system. The reactance modulator -- The re ctance
modulator is a vacuum tube amplifier connected to the r.f. tank circuit of an oscillator in such away as to act as avariable inductance or capacity of a value dependent upon the
instantaneous a.f. voltage applied to its grid. Fig. 521 is arepresentative circuit. The control grid circuit of the 6L7 tube is connected across

+B
Fig. 521 -- Reactance modulator circuit using a6L7 tube.
-- Oscillator tank capacity. -- 3-10 pad. (See text.) C2-250 pad. C3 -- 8-pfd. electrolytic (a.f. by-pass) in parallel with
0.01-dd. paper (a. by-pass). C4 -- 0.01 eifd.
-- Oscillator tank inductance. Ri -- 50,000 ohms. 112 -- 0.5 megohm.
R3 -- 30,000 ohms.
R4 - 300 ohms. Re -0.5 megohm.

the small capacity C1,which is in series with the resistor R1 across the oscillator tank cir-

cuit. Any type of oscillator circuit (� 3-7) may

be (�

used. R1 is large compared to 2-8) of CI,so the r.f. current

the reactance
through W I

will be practically in phase (� 2-7) with the r.f.

voltage appearing at the terminals of the

tank circuit. However, the voltage across

will lag the current by 90 degrees

8).

The r.f. current in the plate circuit of the L7

will be in phase with the grid voltage

and consequently is 90 degrees behind the cur-

rent through C1,or 90 degrees behind the r.f tank voltage. This lagging current is drawn

frequency deviation increases with the number of times the initial frequency is multiplied; for instance, if the oscillator is operated on 7 Mc. and the output frequency is to be 112
Mc., an oscillator frequency deviation of 1000 cycles will be raised to 16,000 cycles at the
output frequency. Design considerations -- The sensitivity of
the modulator (frequency change per unit change in grid voltage) increases when Ci is made smaller, for a fixed value of RI, and also increases with an increase in LIC ratio in the oscillator tank circuit. Since the carrier

104 CHAPTER FIVE

leaclidelephony

stability of the oscillator depends on the L/C plate current as the a.f. modulating voltage

ratio (� 3-7) it is desirable to use the highest is increased. The distortion will be within ac-

tank capacity which will permit the desired ceptable limits with the tube and constants

deviation to be secured while keeping within given in Fig. 521 when the plate current does

the limits of linear operation. When the circuit not change more than 5% with signal.

of Fig. 521 is used in connection with a7-Mc.

Non-linearity is accompanied by a shift in

oscillator, a linear deviation of 2000 cycles the carrier frequency, so it can also be checked

above and below the carrier frequency can by means of a selective receiver such as one

be secured when the oscillator tank capacity with a crystal filter (� 7-11). A tone source is

is approximately 200 ppid. A peak a.f. input convenient for the test. Set the receiver for

of two volts is required for full deviation. At high selectivity, switch on the beat oscillator

56 Mc. the maximum deviation would be and tune to the oscillator carrier frequency.

8 X 2000 or 16 ke.

(The check does not need to be made at the

Since achange in any of the voltages on the output frequency, and the oscillator frequency

modulator tube will �ause a change in r.f. usually is more convenient since it will fall

plate current and consequently a frequency within the tuning range of a communications

change, it is advisable to use aregulated plate receiver.) Increase the modulating signal until

supply for both modulator and oscillator. adefinite shift in carrier frequency is observed;

At the low voltages used (250 volts), the this indicates the point at which non-linearity

required stabilization can be secured by means starts. The modulating signal should be kept

of gaseous regulator tubes (� 8-8).

below this level for minimum distortion.

Speech amplification -- The speech ampli-

A selective receiver also can be used to check

fier preceding the modulator follows ordinary frequency deviation, again at the oscillator

design (� 5-9) except that no power is require

frequency. A source of tone of known fre-

from it and the a.f. voltage taken by the modu quency is required, preferably a continuously

lator grid is usually small -- not more tha

variable calibrated audio oscillator or signal

10 or 15 volts even with large modulator tubes. generator. Tune in the carrier as described

Because of these modest requirements only a above, using the beat oscillator and high selec-

few speech-amplifier stages are needed; atwo- tivity, and adjust the modulating signal to the

stage amplifier consisting of apentode followed maximum level at which linear operation is

by a triode, both resistance coupled, will suf- secured. Starting with the lowest frequency

fice for crystal microphones (� 5-8).

� available, slowly raise the tone frequency while

R.f. amplifier stages -- The frequency mull listening closely to the carrier beat note. As the

tiplier and output stages following the modu- tone frequency is raised the beat note will de-

lated oscillator may be designed and adjusted crease in intensity, disappear entirely at adefi-

in accordance with ordinary principles. No nite frequency, then come back and increase

special excitation requirements are imposed, in intensity as the tone frequency is raised still

since the amplitude of the output is constant. more. The frequency at which the beat note

Enough frequency multiplication must be disappears multiplied by 2.4 is the frequency

used to give the desired maximum deviation deviation at that level of modulating signal;

at the final frequency; this depends upon the for example, if the beat note disappears with

maximum linear deviation available from the an 800-cycle tone the deviation is 2.4 X 800,

modulator-oscillator. All stages in the trans- or 1920 cycles. The deviation at the output

mitter should be tuned to resonance, and care- frequency is the oscillator deviation multiplied

ful neutralization of any straight amplifier by the number of times the frequency is multi-

stages (� 4-7) is necessary to prevent r.f. plied; in this example, if the oscillator is on

phase shifts which might cause distortion.

7 Mc. and the output on 56 Mc., the final

Checking operation -- The two quantities deviation is 1920 X 8, or 15.36 kc.

to be checked in the f.m. transmitter are linear-

The output of the transmitter can be

ity and frequency deviation. With amodulator checked for amplitude modulation by observ-

of the type shown in Fig. 521, both the r.f. ing the antenna current. It should not change

and a.f. voltages are small enough to make the from the unmodulated carrier value when the

operation Class A (� 3-4) so that the plate transmitter is modulated. If there is no an-

current of the modulator is constant so long; tenna ammeter in the transmitter, aflashlight

as operation is over the linear portions of th lamp and loop can be coupled to the final tank

No. 1 and No. 3 grid characteristics. Ilene coil to serve as acurrent indicator. The lamp

non-linearity will be indicated by achange in brilliance should not change with modulation.

mu CHAPTER FIVE

CHAPTER SIX

�

KEYING PRINCIPLES AND

CHARACTERISTICS

Requirements -- The keying of a transmitter can be considered satisfactory if the method employed reduces the power output to zero when the key is open, or "up," and permits full power to reach the antenna when the key is closed, or "down." Furthermore, it should do this without causing keying transients or "clicks," which cause interference with other amateur stations and with local broadcast reception, and it should not affect the frequency of the emitted wave.
Back-wave -- From various causes some energy may get through to the antenna during keying spaces. The effect then is as though the dots and dashes were simply louder portions of a continuous carrier; in some cases, in fact, the back-wave, or signal heard during the keying spaces, may seem to be almost as loud as the keyed signal. Under these conditions the keying is hard to read. A pronounced backwave often results when the amplifier stage
feeding the antenna is keyed; it may be present because of incomplete neutralization (� 4-7) of the final stage, allowing some energy to get to the antenna through the grid-plate capacity of the tube, or because of magnetic coupling
between antenna coupling coils and one of the low-power stages.
A back-wave also may be radiated if the key-
ing system does not reduce the input to the keyed stage to zero during keying spaces. This trouble will not occur in keying systems which cut off the plate voltage when the key is open, but may be present in grid-blocking systems (� 6-3) if the blocking voltage is not great enough and, in power supply primary-keyed systems (� 6-3) if only the final stage power supply primary is keyed.
Keying waveform and siclebands-- A c.w. signal can be considered equivalent to any modulated signal (� 5-1) except that instead of being modulated by sinusoidal waves and
their harmonics, it is modulated by a rectangular wave as in Fig. 601-A. If it were modulated by asinusoidal wave of single frequency,
as in Fig. 601-B, the only sidebands would be those equal to the carrier frequency plus and minus the modulation frequency (� 5-2). A keying speed of 50 words per minute, sending sinusoidal dots, would give sidebands only

20 cycles either side of the carrier. However, when harmonics are present in the modulation, the sidebands will extend out on both sides of the signal as far as the frequency of the highest harmonic. The rectangular wave form contains an infinite number of harmonics of the keying frequency, so a carrier modulated by truly rectangular dots would have sidebands covering the entire spectrum. Actually the high-order harmonics are eliminated because of the selectivity of the tuned circuits (� 2-10) in the transmitter, but there is still enough energy in the lower harmonics to extend the sidebands considerably. Considered from another viewpoint, whenever a pulse of current has a steep front (or back) high frequencies are certain to be present. If the pulse can be slowed down, or caused to lag, through a filter the highest-order harmonics are filtered out.
Key clicks--because the high-order harmonics exist only when the keying character is started or ended (when the amplitude is building up or dying down) their effects outside the normal communication channel are observed as pulses of very short duration. These pulses are called key clicks.
Tests have shown that practically all operators prefer to copy a signal which is "solid" on the "make" end of each dot or dash; i.e., one that does not build up too slowly, but just slowly enough to have aslight click when the key is closed. The same tests indicate that the most pleasing and least difficult signal to copy, particularly at high speeds, is one that has a fairly soft "break" characteristic; i.e., one that has practically no click as the key is opened. A signal with heavy clicks on both
A
Fig. 601 -- Extremes of possible keying waveshapes. A, rectangular characters; B, sine-wave characters.

106 CHAPTER SIX

make and break is difficult to copy at high speeds (and also causes considerable interference), but if it is too "soft" the dots and dashes will tend to run together. It is ;elatively
simple to adjust the keying of a transmitter so that for all normal hand speeds (15 to 4. 0 w.p.m.) the readability will be satisfactoty while the keying still will not cause interference to reception of other signals near the frequency of the transmitter.
Break-in keying--Since in code transmission there are definite intervals, between dots and dashes and between words, when no power is being radiated by the transmitter it is poss�ble, with suitable keying methods, to allow the receiver to operate continuously and thus be capable of receiving incoming signals during the keying intervals. This practice facilitates
communication because the receiving operator can signal the transmitting operator, by holding down the key of his transmitter, whenever
he has failed to copy part of the message and thus obtain a repetition of the missing part without loss of time. This is called break-in operation.
Frequency stability -- Keying should have no effect upon the output frequency of aproperly designed and adjusted transmitter. However, in many instances keying will cause a
"chirp," or small frequency change at the instant of closing or opening the key, which makes the signal difficult to read. Multi-stage transmitters keyed in a stage subsequent to the oscillator are usually free from this condi-
og tion unless the keying causes line-voltage
changes which in turn affect the frequency the oscillator. When the oscillator is keyed for break-in operation special care must be taken to insure that the signal does not have keying chirps.
Selecting the stage to key -- It is advantageous from an operating standpoint to design the c.w. transmitter for break-in operation. In ordinary cases this dictates that the oscillator be keyed, since a continuouslyrunning oscillator will create interference in the
receiver and thus prevent break-in operation on or near the transmitter frequency. On the other hand, it is easier to avoid achirpy signal by keying abuffer or amplifier stage. In either
case, the tubes following the keyed stage must be provided with sufficient fixed bias to limit the plate currents to safe values when the key is up and they are not being excited (� 8-9). Complete cut-off reduces the possibility of a back-wave if a stage other than the oscillator is keyed, but the keying waveform is not as well preserved and some clicks can be introduced even though the keyed stage itself produces no clicks. It is agood general rule to bias the
tubes to take a key-up plate current equal to about 5% of the normal key-down value.

Keyed power -- The power broken by the key is an important consideration, both from the standpoint of safety for the operator and arcing at the key contacts. Keying the oscillator or a low-power stage is favorable in both respects. The use of a keying relay is highly recommended when a high-power circuit is keyed. �Se KEYING CIRCUITS
Plate-circuit keying -- Any stage of the transmitter can be keyed by opening and closing the plate power circuit. Two methods are shown in Fig. 602. In A the key is in series with the negative lead from the plate power supply to the keyed stage. It could also be placed in the positive lead, although this is to be avoided whenever possible because the key is necessarily at the plate voltage above ground and there is danger of shock unless a keying relay is used.
sg plate Fig. 602 -- A, plate keying; B, screen-grid keying. Oscillator circuits are shown in both cases, but the keying methods also can be used with amplifiers.
Fig. 602-B shows the key in the screensupply lead of an electron-coupled oscillator. This can be considered to be a variation of plate keying.
The circuits of Figs. 602-A and B respond well to the use of key-click filters, and are particularly suitable for use with crystal and self-controlled oscillators operating at low plate voltage and power input.
Power-supply keying -- A variation of plate keying, in which the keying is introduced in the power supply itself rather than between the power supply and transmitter, is illustrated by the diagrams in Fig. 603.

107 C II Al' E SIX

Dheleach� Amaleur'6 .-liancgooh

25V Trans1f1oVrmer

(A)
Xeenq relay h.v. inseation

Pnrnary not used

(B)

Fig. 603 -- Power-supply keying. Grid-control rectifiers are used in A. Transformer T is asmall multiplesecondary unit of the type used in receiver power supplies, and is used in conjunction with the full-wave rectifier tube to develop bias voltage for the grids of the hi ghvoltage rectifiers. RIlimits the load on the bias supply when the keying relay is closed; 50,000 ohms is asuitable value. CImay be 0.1 pfd. or larger. L and C constitute
the smoothing filter for the high-voltage supply in both
circuits. B shows primary keying.

be sufficient to overcome the r.f. grid vol-

tage, in the case where the bias is applied

+ to the .control grid, and hence must be con-

siderably higher than the nominal cut-off

value for the tube at the operating d.c.

plate voltage. The fundamental circuits are

shown in Fig. 604.

In both circuits the key is connected in

25V

series with aresistor, RI,which limits

Windmq used the current drain on the blocking-

05 P'"nory bias source when the key is closed.

Rei is a resistance-capacity filter

(� 2-11) for controlling the lag on

make and break of the key circuit. The lag

ihcreases as the time constant (� 2-6) of

this circuit is made larger. Since grid current

flows through R2 when the key is closed in

Fig. 604-A, additional operating bias is de-

+ veloped, hence somewhat less bias is

I needed from the regular bias supply. -0* The operating and blocking biases can

be obtained from the same supply, if

desired, by utilizing suitable taps on a

voltage divider (� 8-10) or if no fixed bias is

used R2 can be the regular grid leak (� 3-6)

for the stage.

With blocked-grid keying arelatively small

direct current is broken as compared to other

systems, thus sparking at the key is reduced.

The keying characteristic (lag) readily can be

controlled by choice of values for C1 and R2.

Fig. 603-A shows the use of grid-controlled rectifier tubes (� 3-5) in the power supply. Keying is accomplished by applying suitable bias to the grids to cut off plate current flow when the key is open, and removing the bias when the key is closed. Since this circuit is used only with high-voltage supplies, a wellinsulated keying relay is a necessity. Direct keying of the primary of the plate power transformer for the keyed stage or stages is shown in Fig. 603-B. This and the method at B inherently have akeying lag because of the time constant (� 2-6) of the smoothing filter.
If enough filter is provided to reduce ripple to alow percentage (� 8-4) the lag (� 6-1) is too great to permit crisp keying at speeds above about 25 words per minute, although this type of keying is very effective in eliminat-
ing key clicks. A single-section filter (� 8-6) is about the most that can be used for areasonably-good keying characteristic.
Blocked-grid keying -- Keying may be accomplished by applying sufficient negative bias voltage to acontrol or suppressor grid to cut off plate current flow when the key is open, and by removing this blocking bias when the key is closed. The blocking bias voltage must

(B)

- t B/o�kinq
Bias -'

+BlBoicaki5-n�g

Fig. 604 -- Blocked-grid keying. 111, the current. limiting resistor, should have a value of about 50,000 ohms. Ci may have acapacity of 0.1 to 1/dd., depending
upon the keying characteristics desired. R2 is similarly
variable, values being of the order of 5000 to 10,000 ohms in most cases.

Cathode keying -- Opening the d.c. circuits of both plate and grid simultaneously is called cathode keying, or center-tap keying with a directly-heated filament-type tube, since in the latter case the key is placed in the filamenttransformer center-tap lead. The circuits are shown in Fig. 605.

108 CHAPTER SIX

--Xeyine

Grid Return

gale Return

Grid Return

gale Return

Fig. 605 -- Center-tap and cathode keying. Condenser C is an r.f. by-pass condenser having acapacity of 0.001 to 0.01 �fd.

Cathode keying results in less sparking at the key contacts, for the same plate power, as compared with keying in the plate-supply lead. When used with an oscillator it-does not respond as readily to key-click filtering (� 6-3) as does plate keying, but there is little difference in this respect between the two systems when an amplifier is keyed.
� 6-3 KEY-CLICK REDUCTION
R.f. filters -- A spark at the key contacts, even though minute, will cause a damped oscillation to be set up in the keying circuit which may modulate the transmitter output or may simply be radiated by the wiring to
KLeiynieng

Fig. 606-- R.f. filter for elimi-

nating effects of sparking at key

contacts. Values are discussed in

RFC

RFC the text.

Key
the key. Interference from this source is usually confined to the immediate vicinity of the transmitter, and is similar in nature and effects to the click which is frequently heard in a receiver when an electric light is turned on or off. It can be minimized by isolating the key from the wiring by means of alow-pass filter (� 2-11), which usually consists of an r.f. choke in each key lead, placed as close as possible to the key, by-passed on the keyingline side by a condenser, as shown in Fig. 606. Suitable values must be determined by experiment. Chokes values may range from 2.5 to 80 millihenrys, and condenser capacities from 0.001 to 0.1 ;dd.
This type of filter is required in nearly every keying installation, in addition to the lag circuits discussed in the next paragraph.
Lag circuits -- A filter used to give adesired shape to the keying character to eliminate unnecessary sidebands and consequent interference is called a lag circuit. In one form,

suitable for the circuits of Figs. 602 and 605, it consists of acondenser across the key terminals and an inductance in series with one of the leads. This is shown in Fig. 607. The optimum values of capacity and inductance must be found by experiment, but are not especially critical. If a high-voltage low-current circuit is being keyed, a small condenser and large inductance will be necessary, while if a lowvoltage high-current circuit is keyed, the capacity required will be high and the inductance
small. For example, a 300-volt 6-ma. circuit will require about 30 henrys and 0.05 isfd., while a 300-volt 50-ma, circuit needs about 1 henry and 0.5 jafd. For any given circuit

To Keyed arcuit
I

Fig. 607 -- Lag circuit for shaltin gthe keyin gcharacter. Values are discussed in the text.

From Hey and rffilter
and fixed values of current and voltage, increasing the inductance will reduce, clicks on "make," and increasing the capacity will reduce the clicks on "break."
Blocked-grid keying is adjusted by changing the values of resistors and condensers in the circuit. In Fig. 604, the click on "make" is reduced by increasing the capacity of C1 and the click on break is reduced by increasing and/or R2. The values will vary with the amount of blocking voltage and the grid current. The constants given in Fig. 604 will serve as afirst approximation.
Tube keying -- A tube keyer is aconvenient adjunct to the transmitter because it, allows the keying characteristic to be adjusted easily without necessitating condenser and inductance values which may not be readily available. It uses the plate resistance of a tube (or tubes in parallel) to replace the key in a plate or cathode circuit, the keyer tube (or tubes) being keyed by the blocked-grid method (� 6-2). A typical circuit is shown in Fig. 608. Type 45 tubes are suitable because of their low plate resistance and consequently small voltage drop between plate and cathode. When atube keyer is used to replace the key in aplate or cathode circuit the power output of the stage will be somewhat reduced because of the voltage drop across the keyer tube, but this can be compensated for by a slight increase in the supply voltage. A tube keyer makes the key itself very safe to handle, since the high resistance in series with the key and blocking voltage prevents shock.

CHAPTER SIX 109

..7he Palio AnaleuA --flaniloo1

Fig. 608 -- Vacuum -tube keyer circuit. The voltage
drop across the tubes will be approximately 90 volts with the two type 45 tubes shown when the keyed current is 100 milliamperes. More tubes can be connected in parallel to reduce the drop. Suggested values are as follows: Ci -- 2pfd., 600-volt paper. C2 -- 0.003-pfd. mica. C3 -- 0.005 -pfd. mica. RI-- 0.25 megohm.
1R123,--R450--,0500meohgmosh,ms1.0-watt.
115 -- 0.5 megohm. Swi, Sw2 -- 3-position 1-circuit rotary switch. Ra -- 325 volts each side c.t., with 5-volt and 2.5-volt
windin gs. A wider range of lag adjustment may be obtained by using additional resistors and condensers. Suggested values of capacity, in addition to the above (C2 and Ca), are 0.001 and 0.002 �fd. Resistors in addition 20 R2 could be 2, 2, 3and 5megohms.
� 6-4 CHECKING TRANSMITTER KEYING
Clicks-- Transmitter keying can be checked by listening to the signal on a superheterodyne receiver. The antenna should be disconnected so that the receiver does not overload and, if necessary, the r.f. gain can be reduced as well. Listening with the beat oscillator and a.v.c. off, the keying should be adjusted so that aslight click is heard as the key is closed, but practically none is heard as the key is released. When the keying constants have been adjusted to meet this condition, the clicks will be about optimum for all normal amateur work. If the clicks are too pronounced, they will cause interference with other amateurs and possibly to nearby broadcast receivers.
Chirps--Keying chirps (instability) may be checked by tuning in the signal or one of its harmonics on the highest frequency range of the receiver and listening with the b.f.o. on and the a.v.c. off. The gain should be sufficient to give moderate signal strength but it should be low enough to preclude the possibility of overloading. Adjust the tuning to give a low-frequency beat .note and key the transmitter. Any chirp introduced by the keying adjustment will be readily apparent. By listening to a harmonic, the effect of any

instability is magnified by the order of the harmonic and thus made more perceptible.
Oscillator keying -- The keying of an amplifier
is relatively straightforward and requires no special considerations other
than those mentioned, but afew additional precautions are necessary with oscillator keying. Any oscillator, either selfexcited or crystal, will key well if it will oscillate at low plate voltages (of the order of one or two volts) and if its change in frequency with platevoltage changes is negligible. A crystal oscillator will oscillate at low plate voltages if a regenerative type of circuit such as the Tritet or grid-plate (� 4-5) is used and if an r.f. choke is connected in series with the grid leak to reduce loading on the crystal. Crystal oscillators of this type are generally free from chirp unless there is a relatively large air-gap between the crystal and top plate of the holder, as is the case with a variable-frequency crystal set at the high-frequency end of its
range. Self-controlled oscillators can be made to
meet the same requirements by using a high CIL ratio in the tank circuit, low plate and screen currents, and judicious adjustment of the feedback (� 3-7). A self-controlled oscillator intended to be keyed should be designed for good keying rather than maximum out-
put. Stages following keying-- When akeying
filter is being adjusted, the stages following the keyed tube should be made inoperative by
removing the plate voltage. This facilitates monitoring the keying without the introduction of additional effects. The following stages should then be added one at atime, checking the keying after each addition. An increase in click intensity (for the same carrier strength) indicates that the clicks are being added in the stages following the one which is keyed.
The fixed bias on such stages should be sufficient to reduce the idling plate current (no excitation) to a low value but not to zero. Under these conditions any instability or tendency toward parasitic oscillations, either of which can adversely affect the keying characteristic, usually will evidence itself. It is
particularly necessary that the transmitter be free from parasitic oscillations, since they
can be the cause of key clicks which do not respond to the methods of treatment outlined in the preceding sections.

//0 CHAPTER SIX

CHAPTER SEVEN

Receiver

rtricipie3 and' cOeJign

� 7-1 ELEMENTS OF RECEIVING SYSTEMS
Basic requirements -- The purpose of a radio receiving system is to abstract energy from passing radio waves and convert it into
a form which conveys the intelligence contained in the signal. It must also be able to select adesired signal and eliminate those not wanted. The fundamental processes involved are amplification and detection.
Detection -- The high frequencies used for radio signalling are well beyond the audiofrequency range (� 2-7) and therefore cannot be used directly to actuate aloudspeaker. Neither can they be used to operate other devices, such as relays, by means of which a message
might be transmitted. The process of converting a modulated radio-frequency wave to a usable low frequency, called detection or demodulation, is essentially that of rectification
(� 3-1). The modulated carrier (� 5-1) is thereby converted to a unidirectional current the amplitude of which will vary at the same rate as the modulation. These low-frequency variations are readily applied to a headset, loudspeaker, or other form of electro-mechanical device.
Code signals -- The dots and dashes of code (c.w.) transmissions are rectified as described, but in themselves can produce no audible tone in aheadset or loudspeaker because they are of constant amplitude. For aural reception it is necessary to introduce a second radio frequency, differing from the signal frequency by a suitable audio frequency, into the detector circuit to produce an audible beat (� 2-13). The frequency difference, and hence the beat note, is generally of the order of 500 to 1000 cycles since these tones are within the range of optimum response of both the ear and the headset. If the source of the second radio frequency is a separate oscillator the system is known as heterodyne reception; if the detector itself is made to oscillate and produce the second frequency, it is known as an autodyne detector.
Amplification -- To build up weak signals to usable output level, modern receivers employ considerable amplification -- often of the order of hundreds of thousands of times. Amplifiers are used at the frequency of the incoming signal (r.f. amplifiers), after detection

(a.f. amplifiers) and, in the superheterodyne
receiver, at one or more intermediate radio frequencies (i.f. amplifiers). The r.f. and i.f. amplifiers practically always employ tuned circuits.
Types of receivers -- Receivers may vary in complexity from asimple detector with no amplification to multi-tube arrangements having amplification at several different radio frequencies as well as at audio frequency. A regenerative detector (� 7-14) with or without audio frequency amplification is known as a regenerative receiver; if the detector is preceded by one or more tuned radio-frequency amplifier stages the combination is known as a t.r.f. (tuned radio frequency) receiver. The superheterodyne receiver (� 7-8) employs r.f. amplification at a fixed intermediate frequency as well as at the frequency of the
signal itself, the latter being converted by the heterodyne process to the intermediate frequency.
At ultra-high frequencies the superregenerative detector (� 7-4), usually with audio ampli-
fication, is used in the superregenerative receiver,

.1.

.,,

� 100 4e,..�,.

3.4.

-20

-.0 -5e56Kc).5 �.0 2.5 �20

KILOCYCLES FROM RESONANCE

Fig. 701 -- Selectivity curve of a modern superheterodyne receiver. The relative response is plotted against deviations above and below the resonance frequency. The scale at the left is in terms td voltage ratios; the corresponding decibel steps (� 3-3) are shown at the right.

CHAPTER SEVEN

Dhe Radio -AmaieuA ilancltooh

providing high amplification of weak signals with simple circuit arrangements.
� 7-2 RECEIVER CHARACTERISTICS
Sensitivity -- Sensitivity is defined as the strength of the signal (usually expressed in microvolts) which must be applied to the input terminals of the receiver to produce aspecified audio-frequency power output at the loudspeaker or headset. It is ameasure of the am.plification or gain.
Signal-to-noise ratio -- Every receiver generates some noise of ahiss-like character, and signals weaker than the noise cannot be separated from it no matter how much amplification is used. This relation between noise and a weak signal is expressed by the signal-to-noise ratio. It can be defined in various ways, one simple one being to give it as the ratio of signal power output to noise output from the receiver at aspecified value of modulated carrier voltage applied to the input terminals.
Since the noise is uniformly distributed over the whole spectrum, its effect will depend upon the selectivity of the receiver, greater selectivity giving smaller noise output and hence a higher signal-to-noise ratio.
Selectivity -- Selectivity is the ability of a receiver to discriminate against signals of frequencies differing from that of the desired signal. The overall selectivity will depend upon the selectivity of the individual tuned circuits and the number of such circuits.
The selectivity of areceiver is shown graphically by drawing acurve which gives the ratio of signal strength required at various frequencies off resonance, to the signal strength at resonance, to give constant output. A resonance curve of this type (taken on a typical communications-type superheterodyne receiver) is shown in Fig. 701. The band-width is the width of the resonance curve (in cycles or kilocycles) at aspecified ratio; in Fig. 701, the band-widths are indicated for ratios of 2 and 10.
Selectivity for signals within afew kilocycles of the desired signal frequency is called adjacent-channel selectivity, to distinguish it from the discrimination against signals considerably removed from the desired frequency.
Stability -- Stability of a receiver is its ability to give constant output, over a period of time, from a signal of constant strength and frequency. Primarily, it means the ability to stay tuned to a given signal, although a receiver which at some settings of its controls has a tendency to break into oscillation, or "howl," is said to be unstable.
The stability of a receiver is affected principally by temperature variations, voltage changes, and constructional features of a mechanical nature.

Fidelity -- Fidelity is the relative ability of the receiver to reproduce in its output the modulation (keying, 'phone, etc.) carried by the incoming signal. For exact reproduction, the band-width must be great enough to accommodate the highest modulation frequency, and the relative amplitudes of the various frequency components within the band must not be changed. � 7-3 DETECTORS
Characteristics -- The important characteristics of adetector are its sensitivity, fidelity or linearity, resistance, and signal-handling capability.
Detector sensitivity is the ratio of audiofrequency output to radio-frequency input. Linearity is a measure of the ability of the detector t� reproduce, as an audio frequency, the exact form of the modulation on the incoming signal. The resistance or impedance of the detector is important in circuit design, since a relatively low resistance means that
(A)
(B)
AFOutput
(C)
4 FOutpui
Fig. 702 --Simplified and practical diode detector circuits. A, the elementary half-wave diode detector; B, apractical cricuit, with r.f. filtering and audio output coupling; C, full-wave diode detector, with output coupling indicated. The circuit 1.2Ci is tuned to the signal frequency; typical values for C2 and RIin A and C
are 250 $tttfd. and 250,000 ohms, respectively; in 13,
C2 and Cs are 100 add. each; Ri, 50,000 ohms; and 112,
2al5l0,t0hr0e0eodhimasg.raCm4s.is 0.1 etfd. and 14, 0.5 to 1megotim in

112 CHAPTER SEVEN

Receiver Principte3 and 243iytt

power is consumed in the detector. The signalhandling capability means the ability of the detector to accept signals of aspecified ampli-
tude without overloading. Diode detectors -- The simplest detector is
the diode rectifier. Circuits for both half-wave
and full-wave (� 8-3) diodes are given in Fig. 702. The simplified half-wave circuit at 702-A includes the r.f. tuned circuit L2C1,with a coupling coil L1from which the r.f. energy is fed to L2C1;the diode, D, and the load resistance R1 and by-pass condenser C2. The flow of rectified r.f. current through R1 causes a d.c. voltage to develop across its terminals, and this voltage varies with the modulation on the signal. The -- and + signs show the polarity of the voltage. Variation in amplitude of the r.f. signal with modulation causes corresponding variations in the value of the d.c. voltage across RI.The load resistor, RI,usually has a rather high value so that a fairly large voltage will develop from a small rectified-current flow.
In the circuit at 702-B, R1 and C2 have been divided for the purpose of filtering r.f. from the output circuit (� 2-11); any r.f. voltage in the output may cause overloading of asucceeding amplifier tube. These audio-frequency variations can be transferred to another circuit
through a coupling condenser, C4in Fig. 702, to a load resistor R3,which usually is a "potentiometer" so that the volume can be adjusted to adesired level.
The full-wave diode circuit at 702-C is practically identical in operation to the half-wave circuit, except that both halves of the r.f. cycle are utilized. The full-wave circuit has the advantage that very little r.f. voltage appears across the load resistor, RI,because the mid-
point of L2 is at the same potential as the cathode or "ground" for r.f.
The reactance of C2must be small compared to the resistance of R1 at the radio frequency being rectified, but at audio frequencies must be relatively large compared to R1 (� 2-8, 2-13). This condition is satisfied by the values shown. If the capacity of C2is too large, the response at the higher audio frequencies will be low.
Compared with other detectors, the sensitivity of the diode is low. Since the diode consumes power, the Q of the tuned circuit is reduced, bringing about a reduction in selectivity (� 2-10). The linearity is good, however, and the signal-handling capability is high.
Grid-leak detectors -- The grid-leak deteetor is a combination diode rectifier and audio-frequency amplifier. In the circuit of Fig. 703-A, the grid correspodis to the diode plate, and the rectifying action is exactly the same. The d.c. voltage from rectified current flow through the grid leak, R1biases the grid negatively with respect to cathode, and the

R F input

(A)

A.F
put
4.0

R Input

Fig. 703 -- Grid-leak detector circuits, A, triode; B, pentode. A tetrode may be used in the circuit of B by neglecting the suppressor-grid connection. Transformer coupling may be substituted for resistance coupling in A, or a high-inductance choke may replace the plate resistor in B. 1.4C: is a circuit tuned to the signal frequency. The grid leak, 111, may be connecte4 directly from grid to cathode instead of across the grid condenser as shown. The operation with either connection will be the same. Representative values are:

Component

Circuit A

Circuit B

100 to 250 �dd. 100 to 250 M.

0.001 to 0.002 ;dd. 250 to SOO ierfd.

C.t

0.1 dd.

0.1 �M.

0.5 irfd. or larger.

1to 2megohrns. 1to 5megohms.

50,000 ohms.

100,000 to 250,000 ohms.

50,000 ohms.

20,000 ohms.

Interstate audio

transformer.

500-henry choke.

The plate voltage in A should be about 50 volts for best sensitivity. In B the screen voltage should be about 30 volts, plate voltage from 100 to 250.

audio-frequency variations in voltage across
R1 are amplified through the tube just as in a normal a.f. amplifier. In the plate circuit, R2is the plate load resistance (� 3-3) and C3 a bypass condenser to eliminate r.f. in the output
circuit. C4is the output coupling condenser. With a triode, the load resistor R2 may be replaced by an audio transformer, T, as shown, in which case C4 is not used.
Since audio amplification is added to rectification, the grid-leak detector has considerably greater sensitivity than the diode. The
sensitivity can be further increased by using a screen-grid tube instead of a triode, as at 703-B. The operation is equivalent to that
of the triode circuit. Cg, the screen by-pass condenser, should have low reactance (� 2-8, 2-13) for both radio and audio frequencies. R3
and R4 constitute a voltage divider (� 8-10)

113 CHAPTER SEVEN

5he Radio -Amateur's ilancgooh

from the plate supply to furnish the proper

Circuits for triodes and pentodes are given

d.c. voltage to the screen. In both circuits, Cg in Fig. 704. C3 is the plate by-pass condenser,

must have low r.f. reactance and high a.f. reactance compared to the resistance of RI; the same consideration applies to C3 with respect to R2.
The sensitivity of the grid-leak detector is higher than that of any other type. Like the

R1is the cathode resistor which provides the operating grid bias (� 3-6), and Cg is aby-pass, for both radio and audio frequencies, across R1 (� 2-13). R2is the plate load resistance 3-3) across which a voltage appears as a result of the rectifying action described above. C4is the

diode, it "loads" the tuned circuit and reduces output coupling condenser. In the pentode

its selectivity. The linearity is rather poor, and circuit at B, R3 and R4 form avoltage divider

the signal-handling capability is limited.

to supply the proper potential (about 30 volts)

Plate detectors -- The plate detector is to the screen, and Cg is a by-pass condenser

arranged so that rectification of the r.f. signal between the screen and cathode. Cg must have

takes place in the plate circuit of a triode, low reactance for both radio and audio fre-

tetrode or pentode, as contrasted to the grid � quencies.

rectification just described. Sufficient negative

The plate detector is more sensitive than

bias is applied to the grid to bring the plate the diode, since there is some amplifying action in the tube, but less so than the grid-leak de-

tector. It will handle considerably larger sig-

nals than the grid-leak detector, but is not

quite as tolerant in this respect as the diode.

Linearity, with the self-biased circuits shown,

is good. Up to the overload point, the detector

takes no power from the tuned circuit and

hence does not affect its Q and selectivity

$13

(� 2-10). Infinite impedance detector -- The circuit

of Fig. 705 combines the high signal-handling

RF Input

Fig. 704 -- Circuits for plate detection. A, triode; B, pentode. 14C1 is tuned to the signal frequency. Typical values for other constants are:

Component C2 C3 C.4 C5 RI
R2
113 114

Circuit A 0.5 pfd. or larger.
0.001 to 0.002 dd.
0.1 dd.
10,000 to 20,000
ohms.
50,000 to 100,000
ohms.

Circuit 13

0.5 pfd. or larger.

250 to 500 ppfd.

00..15

pfd. pfd.

larger.

10,000 to 20,000 ohms.

100,000 to 250,000 ohms.

50,000 ohms.
20,000 ohms.

Plate voltages from 100 to 250 volts may be used. Screen voltage in B should be about 30 volts.

current nearly to the cut-off point, so that the application of asignal to the grid circuit causes an increase in average plate current. The average plate current follows the changes in signal amplitude in a fashion similar to the rectified current in adiode detector.
114 CHAPTER SEVEN

OUAT.PFUT

Fig. 705 -- The infinite impedance detector. LaCi is tuned to the signal frequency. Typical values for other constants are: C2 -- 250 pad. Ca -- 0.5 dd. C4 -- 0.1 pfd. Ri --0.15 megohm. R2 -- 25,000 ohms. R3 -- 0.25-megohm volume control.
A tube having amedium amplification factor (about 20) should be used. The plate voltage should be 250 volts.

capabilities of the diode detector with low distortion (good linearity) and, like the plate detector, does not load the tuned circuit to which it is connected. The circuit resembles that of the plate detector except that the load resistance, RI,is connected between cathode and ground and is thus common to both grid and plate circuits, giving negative feedback for the audio frequencies. The cathode resistor is by-passed for r.f. (C1) but not for audio
(� 2-13), while the plate circuit is by-passed

Receiver Principled and 2 0edign

to ground for both audio and radio frequencies.

R2 with C2 forms an RC filter (� 2-11) to iso-

late the plate from the "B" supply at audio

(A)

frequencies. The plate current is very low at no signal,

increasing with signal as in the case of the

plate detector. The voltage drop across R1

similarly increases with signal because of the

increased plate current. Because of this and

the fact that the initial drop across R1 is large,

the grid cannot be driven positive with respect

to the cathode by the signal, hence no grid

current can be drawn.

+ 8

-8

Fig. 706 -- Triode and pentode regenerative detector circuits. The circuit LaCi is tuned to the signal frequency. The grid condenser, Ca, should have a value of about 100 ��fd. in all circuits; the grid leak, Ri, may range in value from 1 to 5 megohms. The tickler coil, La, will ordinarily have from 10% to 25% of the number of turns on La; in C, the cathode tap is about 10% of the number of turns on La above ground. Regeneration control condenser Ca in A should have a maximum capacity of 100 �dd. or more; by-pass condensers C3 in B and C are likewise 100 midd. Ca is ordinarily 1pfd. or more; Ra, 50,000-ohm potentiometer; 113, 50,000 to 100,000 ohms. L4 in B (La in C) is a500-henry inductance, C4 is 0.1 Mfd. in both circuits. Ti in A is aconven.' tional audio transformer for coupling from the plate of atube to afollowing grid. RFC is 2.5 mh.
In A, the plate voltage should be of the order of 50 volts for best sensitivity. The pentode circuits require about 30 volts on the screen; plate voltage may be from 100 to 250 volts.

� 7.4 REGENERATIVE DETECTORS
Circuits-- By providing controllable �r.f. feedback or regeneration (� 3-3) in atriode or pentode detector circuit, the incoming signal can be amplified many times, thereby greatly increasing the sensitivity of the detector. Regeneration also increases the effective Q of the circuit and hence increases the selectivity (� 2-10), by virtue of the fact that the maximum regenerative amplification takes place at only the frequency to which the circuit is tuned. The grid-leak type of detector is most suitable for the purpose. Except for the re-
generative connection, the circuit values are identical with those previously described for this type of detector, and the same considerations apply.
Fig. 706 shows the circuits of regenerative detectors of various types. The circuit of A is for a triode tube, with a variable by-pass condenser, C3,in the plate circuit to control regeneration. When the capacity is small the tube does not regenerate, but as it increases toward maximum its reactance (� 2-8) becomes smaller until a critical value is reached where there is sufficient feed-back to cause oscillation.
If L2 and L3 are wound end to end in the same direction, the plate connection is to the outside of the plate or "tickler" coil, /4 ,when the grid connection is to the outside of the tuned circuit coil, L2.
The circuit of B is for ascreen-grid tube, regeneration being controlled by adjustment of the screen-grid voltage. The tickler, L3,is in the plate circuit. The portion of the control resistor between the rotating contact and ground is by-passed by a large condenser (0.5 dd. or more) to filter out scratching noise when the arm is rotated (� 2-11). The feedback should be adjusted by varying the number of turns on /13 or the coupling (� 2-11) between id2 and L3 so that the tube just goes into oscillation at ascreen voltage of approximately 30 volts.
Circuit C is identical with B in principle of operation, except that the oscillating circuit is
of the Hartley type (� 3-7). Since the screen and plate are in parallel for r.f. in this circuit,

115 CHAPTER SEVEN

Dhe Radio Amaleur'6 iland`ooh

only a small amount of "tickler" -- that is, relatively few turns between cathode tap and ground -- is required for oscillation.
Adjustment for smooth regeneration -- The ideal regeneration control would permit the detector to go into and out of oscillation smoothly, would have no effect on the frequency of oscillation, and would give the same value of regeneration regardless of frequ9ncy and the loading on the circuit. In practice the effects of loading, particularly the loading that occurs when the detector circuit is coupled to
an antenna, are difficult to overcome. Likewise, the regeneration is affected by the frequency to which the grid circuit is tuned.
In all circuits it is best to wind the tickler at the ground or cathode end of the grid coil, and to use as few turns on the tickler as will allow the detector to oscillate easily over the whole tuning range with the proper plate (and screen, if apentode) voltage. Shoukl the tube break into oscillation suddenly, making aclick,
as the regeneration control is advanced, the operation can frequently be made smooth by changing the grid-leak resistance to a higher or lower value. The wrong grid leak plus toohigh plate and screen voltage are the most frequent causes of lack of smoothness in going into oscillation.
Antenna coupling -- If the detector is
coupled to an antenna, slight changes in the antenna constants (as when the wire swings in a breeze) affect the frequency of the oscillations generated, and thereby the beat fre-
quency when c.w. signals are being received. The tighter the antenna coupling is made, the
greater will be the feedback required or the higher will be the voltage necessary to make the detector oscillate. The antenna coupling should be the maximum that will allow the detector to go into oscillation smoothly with the correct voltages on the tube. If capacity coupling (� 2-11) to the grid end of the coil is used, only a very small amount of capacity will be needed to couple to the antenna. Increasing the capacity increases the cou-
pling. At frequencies where the antenna system is
resonant the absorption of energy from the oscillating detector circuit will be greater, with the consequence that more regeneration is needed. In extreme cases it may not be possible to make the detector oscillate with normal voltages, causing so-called "dead spots". The
remedy for this is to loosen the antenna coupling to the point which permits normal oscillation and smooth regeneration control.
Body capacity -- A regenerative detector occasionally shows a tendency to change frequency slightly as the hand is moved near the
dial. This condition (body capacity) can be caused by poor design of the receiver or by

the antenna, if the detector is coupled to an antenna. If the body capacity is present when the antenna is disconnected, it can be eliminated by better shielding, and sometimes by r.f. filtering of the 'phone leads. Body capacity present only when the antenna is
connected is caused by resonance effects in the antenna which tend to cause part of a standing wave (� 2-12) of r.f. voltage to appear on the ground lead and thus raise the
whole detector circuit above ground potential. A good, short ground connection should be made to the receiver and the length of the antenna varied electrically (by adding asmall coil or variable condenser in the antenna lead) until the effect is minimized. Loosening the coupling to the antenna circuit also will help reduce body capacity.
Hum -- Power-supply frequency hum may be present in aregenerative detector, especially when it is used in an oscillating condition for c.w. reception, even though the plate supply is free from ripple (� 8-4). It may result from the use of a.c. for the tube heater, but effects of this type are normally troublesome only when the circuit of Fig. 706-C is used, and then only
at 14 Mc. and higher frequencies. Connecting one side of the heater supply to ground, or grounding the center-tap of the heater transformer winding, is good practice to reduce
hum, and the heater wiring should be kept as far as possible from the r.f. circuits.
House wiring, if of the "open" type, will
have a rather extensive electrostatic field which may cause hum if the detector tube, grid lead, grid condenser and leak are not electrostatically shielded. This type of hum is easily recognizable because of its rather high
pitch, a result of harmonics (� 2-7) in the power-supply system. The hum is caused by a species of grid modulation (� 5-4) because the field causes a small a.c. voltage to develop across the grid leak.
Antenna resonance effects frequently cause a hum, of the same nature as that just de-
scribed, which is most intense at the various resonance points and hence varies with tuning. For this reason it is called tunable hum. It is prone to occur with arectified a.c. plate supply (� 8-1) when a standing wave effect of the
type described in the preceding paragraph occurs, and is associated with the non-linearity of the rectifier tube in the plate supply. Elimination of antenna resonance effects and by-passing the rectifier plates to cathode usually will cure it.
Tuning -- For c.w. reception, the regeneration control is advanced until the detector breaks into a"hiss," which indicates that the detector is oscillating. Further advancing of the regeneration control after the detector starts oscillating will result in aslight decrease

116 CHAPTER SEVEN

Receiver Principl� and 2 1eJign

o 2000 Cycles

o Zero beat
B

woo cycles
Li

12000
10000 e000-

6000 -

4000 -

2000 -�
o

' I '
45

50 DIAL SETTING

1111 SS

Fig. 707 -- As the tuning dial of areceiver is turned past ac.w. signal, the beat note varies from ahigh tone down through "zero beat" (no audible frequency difference) and back up to ahigh tone, as shown at A, B and C. The curve is a graphical representation of the action. The beat exists past 8000 or 10,000 cycles but usually is not heard because of the limitations of the audio system of the receiver.

in the strength of the hiss, indicating that the sensitivity is decreasing.

Superregeneration-- The limit to which
ordinary regenerative amplification can be carried is the point at which oscillations com-
mence, since at that point further amplification ceases. The superregenerative detector over-
comes this limitation by introducing into the detector circuit an alternating voltage of a frequency somewhat above the audible range (of the order of 20 to 100 kilocycles) in such a way as to vary the detector's operating point (� 3-3). As aconsequence of the introduction of this quench or interruption frequency the detector can oscillate only when the varying operating point is in a region suitable for the production of oscillations. Because the oscillations are constantly being interrupted the regeneration can be greatly increased and the signal will build up to relatively tremendous proportions. The superregenerative circuit is suitable only for the reception of modulated signals, and operates best on ultra-high frequencies where it has found considerable application in simple receivers.
A typical superregenerative circuit for ultra-

The proper adjustment for the reception of c.w. signals is just after the detector has started to oscillate, when it will be found that c.w. signals can be tuned in and will give atone

with each signal depending on the setting of the tuning control. As the receiver is tuned

through a signal, the tone will first be heard as a very high pitch, go down through "zero beat" and then disappear at a high pitch

on the other side, as shown in Fig. 707. It will be found that a low-pitched beat-note cannot be obtained with a strong signal because the detector "pulls in" or "blocks," but this condition can be corrected by advancing the regeneration control until the beatnote occurs again. If the regenerative detector is preceded by an r.f. amplifier stage, the blocking can be eliminated by reducing the gain of the r.f. stage. If the detector is coupled to an antenna the blocking condition can be elimi-
nated by advancing the regeneration control or loosening the antenna coupling.
The point just after the receiver starts oscillating is the most sensitive condition for c.w. reception -- further advancing of the regeneration control makes the receiver less prone to blocking by strong signals but less capable of receiving weak signals.
If the receiver is in the oscillating conditionand a'phone signal is tuned in, asteady beatnote will result and, while it is possible to listen to 'phone if the receiver can be tuned to exact zero beat, it is more satisfactory to

Fig. 708 -- Superregenerative detector circuit with separate quench oscillator. LIC: is tuned to the signal frequency. Typical values for other components are as
,ed. follows: -- 100 Ca -- 500 odd. C4 -- 0.1 idd. Ri -- 5megohms. R2.-- 50,000 ohms. Rs -- 50,000-ohm potentiometer. R4 -- 50,000 ohms.
Ti -- Audio transformer, plate-to-grid type.
RFC -- Radio-frequency choke, constants depending upon frequency of operation. Special low. capacity chokes are required for ultra-high frequencies.

reduce the regeneration to the point just before the receiver goes into oscillation. This is the most sensitive operating point for this type of reception.

high frequencies is shown in Fig. 708. The regenerative detector circuit is an ultraudion
oscillator (� 3-7). The quench frequency, obtained from the separate quench oscillator, is

117 CHAPTER SEVEN

Pali� .%

_Amateur:3 -Want!itool

introduced in the plate circuit. Many other circuit arrangements are possible.
If regeneration in an ordinary regenerative circuit is carried sufficiently far, the circuit will break into a low-frequency oscillation simul-
taneously with that at the operating radio frequency. This low-frequency oscillation has
much the same quenching effect as that from a
separate oscillator, hence acircuit so operated is called a self-quenching superregenerative detector. This type of circuit is more successful
at ultra-high than at ordinary communication frequencies. The frequency of the quench oscillation depends upon the feedback and
upon the time constant of the grid leak and condenser, the oscillation being a form of
"blocking" or "squegging" in which the grid accumulates a strong negative charge which cannot leak off rapidly enough through the grid
leak to prevent a relatively slow variation of
the operating point. The superregenerative detector has little
selectivity, but discriminates against noise such as that from automobile ignition systems. It also has marked automatic volume control action, since strong signals are amplified to a much smaller extent than weak signals.
Adjustment of superregenerative detectors --Because of the greater amplification, the hiss when the superregenerative detector goes into oscillation is much stronger than with the ordinary regenerative detector. The most sensitive condition is at the point where
the hiss first becomes marked. When asignal is tuned in the hiss will disappear to a degree which depends upon the signal strength.
Lack of hiss indicates insufficient feedback at the signal frequency or inadequate quench

voltage. Antenna loading effects will cause dead spots similar to those with regenerative detectors and these can be overcome by the same methods. The self-quenching detector may require critical adjustment of the grid leak and grid condenser values for smooth operation, since these determine the frequency and amplitude of the quench voltage.
� 7-5 AUDIO-FREQUENCY AMPLIFIERS
General-- Audio-frequency amplifiers are used after the detector to increase the power to alevel suitable for operating aloud-speaker or, in some cases, a headset. There are seldom more than two stages of a.f. amplification in a receiver, and often only one.
In all except battery-operated receivers, the negative grid bias of audio amplifiers is usually secured from the voltage drop in acathode resistor (� 3-6). The cathode resistor must be bypassed by acondenser having low reactance, at the lowest audio frequency to be amplified, compared to the resistance of the cathode resistor (10% or less) (� 2-8, 2-13). In batteryoperated sets, a separate grid-bias battery generally is used.
Headset and voltage amplifiers --The circuits shown in Fig. 709 are typical of those used for voltage amplification and for providing sufficient power for operation of headphones (� 3-3). Triodes usually are preferred to pentodes because they are better suited to working into an audio transformer or headset, which have impedances of the order of 20,000 ohms.
In these circuits, R2 is the cathode bias resistor and C1 the cathode by-pass condenser. RI, the grid resistor, gives volume control

IAn.p1u1t-1

(A)

Fig. 709 -- Audio amplifier
circuits for voltage amplifica.

.C3

tion and headphone output. The tubes are operated as

A Inpul

Output

Class-A amplifiers ( 3-4).

-6

(c)

(D)

118 CHAPTER SEVEN

Peceiver Principio ancl 2)ettign

-a (3)

Fig. 710 -- Audio poiier output amplifier circuits. Class A or AB (1 3-4) amplification is used.

action (� 5-9). Its value ordinarily is from 0.25 to 1 megohm. Cg is the input coupling condenser, already discussed under detectors; it is, in fact, identical to C4 in Figs. 703 and 704 if the amplifier is coupled to adetector.
Power amplifiers-- A popular type of power amplifier is the single pentode; the
circuit diagram is given in Fig. 710-A. The grid resistor, RI, may be a potentiometer for volume control as shown at R1in Fig. 709. The output transformer T should have aturns ratio (� 2-9) suitable for the speaker used; most of the small speakers now available are furnished complete with output transformer.
When greater volume is needed a pair of pentodes or tetrodes may be connected in
push-pull (� 3-3) as shown in Fig. 710-B. Transformer coupling to the voltage-amplifier stage is the simplest method of obtaining pushpull input for the amplifier grids. The inter-
stage transformer, 7'1, has a center-tapped secondary, with asecondary-to-primary turns ratio of about 2to 1. An output transformer, 7'2,with acenter-tapped primary must be used. No by-pass condenser is needed across the cathode resistor, R, since the al. current does not flow through the resistor as it does in single-tube circuits.
Tone control-- A tone control is a device for changing the frequency response (� 3-3) of an audio amplifier; usually it is simply a method for reducing high-frequency response. This is helpful in reducing hissing and crackling noises without disturbing the intelligibility of the
signal. R4 and C4 together in Fig. 709-D form

an effective tone control of this type. The maximum effect is secured when R4 is entirely out of the circuit, leaving C4connected between grid and ground. R4 should be large enough compared to the reactance of C4 ( 2-8) so that when it is all in circuit the effect of C4 on the frequency response is negligible.
�7-6 RADIO-FREQUENCY AMPLIFIERS Circuits-- Although there are variations in
detail, practically all r.f. amplifiers conform to the basic circuit shown in Fig. 711. A screengrid tube, usually a pentode, is invariably used, since atriode will oscillate when its grid and plate circuits are tuned to the same frequency (� 3-5). The amplifier operates Class A, without grid current (� 3-4). The tuned grid circuit, LiCi, is coupled through L2 to the antenna (or, in some cases, to a preceding stage). R1 and C2 are the cathode bias resistor and cathode by-pass condenser, C3 the screen by-pass condenser, and Rg the screen dropping resistor. L3 is the primary of the output transformer (� 2-11), tightly coupled to L4 which, with C3,constitutes the tuned circuit feeding the detector or a following amplifier tube. LiCI and L4C5 are both tuned to the frequency of the incoming signal.
Shielding --The screen-grid construction prevents feedback (� 3-3) from plate to grid inside the tube, but in addition it is necessary to prevent transfer of energy from the plate circuit to the grid circuit external to the tube. This is accomplished by enclosing the coils in grounded shielding containers, and by keeping the plate and grid leads well separated. With "single-ended" tubes care in laying out the wiring to obtain the maximum possible physical separation between plate and grid leads is necessary to prevent capacity coupling.
The shield around acoil will reduce the inductance and Q of the coil (� 2-11) to an extent which depends upon the shielding material and its distance from the coil. Adjustments to the inductance therefore must be made with the shield in place.
By-passing -- In addition to shielding, good by-passing (� 2-13) is imperative. This is not simply a matter of choosing the proper type
�
Fig. 711--The circuit of a tuned radio-frequency amplifier. Circuit values are discussed in the text.
119 CHAPTER SEVEN

he Radio

ccieur'�

and capacity of by-pass condenser. Short
separate leads from C3 and C4 to cathode or ground are a prime necessity, since at the higher radio frequencies even an inch or two
of wire will have enough inductance to provide feedback coupling, and hence cause oscillation, if the wire happens to be common to both the plate and grid circuits.
Gain control -- The gain of an r.f. amplifier usually is varied by varying the grid bias. This method is applicable only to variable-g type
tubes (� 3-5) hence this type usually is found in r.f. amplifiers. In Fig. 711, R3 and R4 comprise the gain-control circuit. R3 is the control
resistor (� 3-6) and R4 a dropping resistor of such value as to make the voltage across the outside terminals of R3 about 50 volts (� 8-10). The gain is maximum with the variable arm all the way to the left (grounded) on Rg and minimum at the right. R3 could simply be placed in series with RI,omitting R4 entirely, but the range of control is limited when this connection is used.
In amulti-tube receiver, the gain of several stages would be varied simultaneously, asingle
control sufficing for all. In such a case, the lower ends of the several cathode resistors (RI)would be connected together and to the movable contact on Rg.
Circuit values -- The value of the cathode resistor, RI, should be calculated for the minimum recommended bias for the tube used. The capacities of C2,C3 and C4 must be such that the reactance is low at radio frequencies; this condition is easily met by using 0.01-44fd. condensers at communication frequencies, or 0.001 to 0.002 mica units at ultra-high frequencies up to 112 Mc. R2 is found by taking
the difference between the recommended plate and screen voltages, then substituting this and the rated screen current in Ohm's Law (� 2-6).
R3 must be selected on the basis of the number of tubes to be controlled; a resistor must be
chosen which is capable of carrying, at its lowresistance end, the sum of all the tube currents plus the bleeder current. A resistor of suitable current-carrying capacity being found, the bleeder current necessary to produce a drop
through it of about 50 volts can be calculated by Ohm's Law. The same formula will give R4, using the plate voltage less 50 volts for E and the bleeder current just found for I.
The constants of the tuned circuits will de-
pend upon the frequency range, or band, being covered. A fairly high L/C ratio (� 2-10) should be used on each band; this is limited, however, by the irreducible minimum capaci-
ties. An allowance of 10 to 20 peifd. should be made for tube and stray capacity, and the minimum capacity of the tuning condenser also must be added.
If the input circuit of the amplifier is con-

nected to an antenna, the coupling coil L2 should be adjusted to provide critical coupling (� 2-11) between the antenna and grid circuit. This will give maximum energy transfer. The turns ratio L1/L2 will depend upon the frequency, the type of tube used, the Q of the tuned circuit, and the antenna system, and in general is best determined experimentally. The selectivity will increase as the coupling is reduced below this "optimum" value, a consideration which it is well to keep in mind if selectivity is of more importance than maximum gain.
The output circuit coupling depends upon the plate resistance (� 3-2) of the tube, the input resistance of the succeeding stage, and the Q of the tuned circuit L4C2.L3 is usually coupled as closely as possible to L4 (this avoids the necessity for an additional tuning condenser across L3)and the energy transfer is about maximum when L3 has 9. to 5.� as many turns as L4,with ordinary receiving screengrid pentodes.
Tube and circuit noise -- In any conductor electrons will be moving in random directions simultaneously and, as aresult, small irregular voltages are developed across the conductor terminals. The voltage is larger the greater the resistance of the conductor and the higher its temperature. This is known as the thermal agitation effect, and it produces a hiss-like noise voltage distributed uniformly throughout the radio-frequency spectrum. The thermal agitation noise voltage appearing across the terminals of atuned circuit will be the same as in a resistor of a value equal to the parallel impedance of the tuned circuit (� 2-10) even though the actual circuit resistance is low.
Hence the higher the Qof the circuit the greater the thermal agitation noise.
Another component of hiss noise is developed in the tube, because the rain of electrons on the plate is not entirely uniform. Small irregularities caused' by gas in the tube also contribute to the effect. Tube noise varies with the type of tube, and is proportional in a
general way to the inverse ratio of the mutual conductance (� 3-2) of the tube to the square root of the plate current.
To obtain the best signal-to-noise ratio, the signal must be made as large as possible at the grid of the tube, which means that the antenna coupling must be adjusted to that end, and also that the Q of the grid tuned circuit must be high. A tube with low inherent noise obviously should be chosen. In an.amplifier having good signal-to-noise ratio the thermal agitation noise will be greater than the tube noise. This can easily be checked by grounding the grid through a0.01-pfd, condenser and observing whether there is a decrease in noise. If there is no change, the tube noise is greatly

120 CHAPTER SEVEN

Peceiver Princip leo and 2)e�i9n

predominant, indicating a poor signal-tonoise ratio in the stage. The test is valid only if there is no regeneration in the amplifier. The signal-to-noise ratio will decrease as the frequency is raised because it becomes increasingly difficult to obtain atuned circuit of high effective Q (� 7-7).
The first stage of the receiver is the important one from the signal-to-noise ratio standpoint. Noise generated in the second and subsequent stages, while comparable in magnitude to that generated in the first, is masked by the amplified noise and signal from the first stage. After the second stage, further contributions by tubes and circuits to the total noise are inconsequential in any normal receiver.
Tube input resistance -- At high frequencies the tube may consume power from the tuned grid circuit even though the grid is not driven positive by the signal. Above 7 Mc. all tubes load the tuned circuit to an extent which depends upon the type of tube. This effect comes about because the time necessary for electrons to travel from the cathode to the grid becomes comparable to the time of one r.f. cycle, and because of a degenerative effect (� 3-3) of the cathode lead inductance. With certain tube types the input resistance may be as low as a few thousand ohms at 28 Mc. and as low as a few hundred ohms at ultrahigh frequencies.
This input loading effect is in addition to the normal decrease in the Q of the circuit alone at the higher frequencies because of increased losses in the coil and condenser. Thus the selectivity and gain of the circuit are both adversely affected.
Comparison of tubes -- At 7Mc. and lower frequencies, the signal-to-noise ratio, gain and selectivity of an r.f. amplifier stage are suffi-
ciently high with any of the standard receiving tubes. At 14 Mc. and higher, however, this is no longer true, and the choice of atube must be based on several conflicting considerations.
Gain is highest with high mutual-conductance pentodes, the 1851 and 1852 being examples of this type. These tubes also develop less noise than any of the others. The inputloading effect is greatest with them, however, so that selectivity is decreased and the tunedcircuit gain is lowered.
Pentodes such as the 6K7, 6J7 and corresponding types in glass have lesser inputloading effects at high frequencies, moderate gain, and relatively-high inherent noise.
The "acorn" and equivalent miniature pentodes are excellent from the input-loading standpoint, the gain is about the same as with
the standard types, and the inherent noise is somewhat lower.
Where selectivity is paramount, the acorns are best, standard pentodes second, and the

1851-52 types last. On signal-to-noise ratio the 1851-52 tubes are first, acorns second, and standard pentodes third. The same order holds for overall gain.
At 56 Mc. the standard types are usable, but acorns are capable of better performance because of lesser loading. The 954, 956 and the corresponding types 9001 and 9003 are the only usable types for r.f. amplification at 112 Mc. and higher.
� 7=7 TUNING AND BAND-CHANGING METHODS
Band changing --The resonant circuits which are tuned to the frequency of the incoming signal constitute a special problem in the design of amateur receivers since the amateur frequency assignments consist of groups or bands of frequencies at widelyspaced intervals. The same LC combination cannot be used for, say, 14 Mc. to 3.5 Mc., because of the impracticable maximum-minimum capacity ratio required, and also because the tuning would be excessively critical with such a large frequency range. It is necessary, therefore, to provide ameans for changing the circuit constants for various frequency bands. As a matter of convenience, the same tuning condenser usually is retained, but new coils are inserted in the circuit for each band.
There are two favorite methods of changing inductances; one is to use aswitch, having an appropriate number of contacts, which connects the desired coil and disconnects the others. The second is to use coils wound on forms with contacts (usually pins) which can be inserted in and removed from asocket.
Band spreading -- The tuning range of a given coil and variable condenser will depend
(A)

Fig. 712 -- Essentials of band-spread tuning systems.

(B)

121 CHAPTER SEVEN

..7fle Ruche AntuhuA iluncgooL

upon the inductance of the coil and the change in tuning capacity. For ease of tuning it is desirable to adjust the tuning range so that practically the whole dial scale is occupied by the
band in use. This is called band-spreading. Because of the varying widths of the bands, special tuning methods must be devised to give
the correct maximum-minimum capacity ratio
on each. Several of these are shown in Fig. 712. In A, a small band-spread condenser Ci
(15 to 25 ��fd. maximum capacity) is used in
parallel with acondenser, C2,which is usually large enough (140 to 175 ��fd.) to cover a2-to-1 frequency range. The setting of C2 will deter-
mine the minimum capacity of the circuit, and the maximum capacity for band-spread tuning will be the maximum capacity of C1 plus the setting of C2.The inductance of the
coil can be adjusted so that the maximum-
minimum ratio will give adequate bandspread. In practicable circuits it is almost impossible because of the non-harmonic relation of the various bands to get full band-
spread on all bands with the same pair of condensers, especially when the coils are wound
to give continuous frequency coverage on C2, which is variously called the band-setting or main-tuning condenser. C2 must be re-set each time the band is changed.
The method shown at B makes use of condensers in series. The tuning condenser, C1,
may have amaximum capacity of 100 mi2fd. or more. The minimum capacity is determined principally by the setting of C3, which usually has low capacity, and the maximum capacity by the setting of C2, which is of the order of 25 to 50 iipfd. This method is capable of close adjustment to practically any desired degree of band-spread. C2 and C3 must be adjusted for each band or else separate pre-adjusted condensers must be switched in.
The circuit at C also gives complete spread on each band. C1,the band-spread condenser, may have any convenient value of capacity;
50 ilpfd. is satisfactory. C2 may be used for continuous frequency coverage ("general coverage") and as a band-setting condenser. The effective maximum-minimum capacity ratio depends upon the capacity of C2 and the point at which C1 is tapped on the coil. The nearer
the tap to the bottom of the coil, the greater the band-spread, and vice versa. For a given coil and tap, the band-spread will be greater
if C2 is set at larger capacity. C2 may be mounted in the plug-in coil form and pre-set, if desired. This requires a separate condenser for each band, but eliminates the necessity for
re-setting C2 each time the band is changed. Ganged tuning--The tuning condensers
of the several r.f. circuits may be coupled to-
gether mechanically and operated by asingle control. This operating convenience involves

more complicated construction, bath electrically and mechanically. It becomes necessary to make the various circuits track -- that is, tune to the same frequency at each setting of the tuning control.
True tracking can be obtained only when the inductance, tuning condensers, circuit minimum capacity and maximum capacity are identical in all "ganged" stages. A small trimmer or padding condenser is connected across the coil so that variations in minimum capacity can be compensated. The fundamental circuit is shown in Fig. 713, where C1 is the trimmer and C2 the tuning condenser. The use of the trimmer increases the minimum circuit capacity, but is a necessity for satisfactory tracking. Condensers having maximum ca-
Fig. 713--Showing the use of atrimmer condenser across the tuned circuit to set the minimum circuit capacity for ganged tuning.
pacities of 15 to 30 peifd. generally are used for the purpose.
The same methods are applied to bandspread circuits which must be tracked. The circuits are identical with those of Fig. 712, although if both general-coverage and bandspread tuning are to be available, an additional trimmer condenser must be connected across the coil in each circuit shown. If only amateur-band tuning is desired, however, then C3 in Fig. 712-B, and C2 in Fig. 712-C serve as trimmers.
The coil inductance can be adjusted by starting with a larger number of turns than necessary, then removing aturn or fraction of aturn at atime until the circuits track satisfactorily. An alternative method of adjusting inductance, providing it is reasonably close to the correct value initially, is to make the coil so that the last turn is variable with respect to the whole coil, or to use asingle short-circuited turn the position of which can be varied with respect to the coil. These methods are shown in Fig. 714.
U.H.F. circuits--Tube interelectrode capacities are practically constant for a given tube type regardless of the operating frequency, and the same thing is approximately true of stray circuit capacities. Hence at ultra-high frequencies these capacities become an increasingly larger part of the usable tuning capacity and reasonably-high L/C ratios (� 2-10) are more difficult to secure as the frequency is raised. Because of this irreducible minimum capacity, standard types of tubes cannot be tuned to frequencies higher than about 200 Mc., even when the inductance in

122

CHAPTER SEVEN

Peceiver Principie3 and Zedyn

(A)

(B)

Fig. 714 -- Methods of adjusting inductance for ganging. The half turn in A can be moved so that its magnetic field either aids or opposes the field of the coil. The shorted loop in B is not connected to the coil, but operates by induction. It will have no effect on the coil inductance when the plane of the loop is parallel to the axis of the coil, and will give maximum reduction of the coil inductance when perpendicular to the coil axis.

the circuit is simply that of a straight wire between the tube elements.
Along with these capacity effects the input
loading (� 7-6) increases rapidly at ultra-high
frequencies so that ordinary tuned circuits have very low effective Q's when connected to the grid circuit of a tube. The effect is still further aggravated by the fact that losses in the tuned circuit itself are higher, causing a still further reduction in Q. For these reasons the frequency limit at which an r.f. amplifier will give any gain .is in the vicinity of 60 Mc.,

(A)

Concentric Line

InRpFut

(B)

Short-circuited Fig. 715 -- Circuits of improved Q for ultra-high frequencies. A, reducing tube loading by tapping down on the resonant circuit; B, use of aconcentric-line circuit, with the tube similarly tapped down. The line should be a quarter-wave long, electrically; because of the additional shunt capacity represented by the tube the physical length will be somewhat less than given by the formula (It10-5). In general, this reduction in length will be greater the higher the grid tap on the inner conductor. One method of coupling to an antenna or preceding stage is indicated. The coupling turn should be parallel to the axis of the line and insulated from the outer conductor.

with standard tubes, and at higher frequencies there is a loss instead of amplification. The condition can be mitigated somewhat by taking steps to improve the effective Q of the circuit, either by tapping the grid down on the coil as shown in Fig. 715-A or by using a lower L/C ratio (� 2-10). The Q of the tuned circuit alone can be greatly improved by using a linear circuit (� 2-12), which when properly constructed will give Q's much higher than those attainable at lower frequencies with conventional coils and condensers. The concentric type of line, Fig. 715-B, is best both from the standpoint of Q and adaptability to non-symmetrical circuits such as are used in receivers. Since the capacity and resistance loading effect of the tube are still present, the Q of such acircuit will be destroyed if the gridcathode circuit of the tube is connected directly across it, hence tapping down, as shown, is necessary.
Ultra-high frequency amplifiers should employ tubes of the acorn type which have the smallest loading effect as well as low interelectrode capacities. This is because the smaller loading effect means higher input resistance and hence, for a given loaded Q of the tuned circuit, higher voltage developed between grid and cathode. Thus the amplification of the stage is higher.
A concentric circuit may be tuned by varying the length of the inner conductor (usually by using close-fitting tubes, one sliding inside the other) or by connecting an ordinary tuning condenser across the line. Tapping the con:denser down as shown in Fig. 715-B gives a band-spread effect which is advantageous, and in addition helps to keep the Q of the circuit higher than it would be with the condenser connected directly across the open end of the line, since at ultra-high frequencies most condensers have losses which cannot be neglected.
U.h.f. oscillators such as those used in the superregenerative detector usually will work well at frequencies where r.f. amplification is impossible with standard tubes (as in the 112-Mc, band) since tube losses are compensated for by energy taken from the power supply. Ordinary coil and condenser circuits are practicable with such tubes and circuits at 112 Mc., and although not as good as linear circuits are more convenient to construct.
� 7-8 THE SUPERHETERODYNE
Principles-- In the superheterodyne, or superhet, receiver the frequency of the incoming signal is changed to anew radio frequency, the intermediate frequency (i.f.), then amplified, and finally detected. The frequency is changed by means of the heterodyne process (�7-1), the output of an adjustable local oscil-

123 CHAPTER SEVEN

DA,. Path� AmaisuA ilartilooh

lator (h.f. oscillator) being combined with the incoming signal in a mixer or converter stage (first detector) to produce a beat frequency equal to the i.f. Fig. 716 gives the essentials of the superhet in block form. C.w. signals are
made audible by heterodyning the signal at the second detector by an oscillator (the beat frequency oscillator (bl.o.) or beat oscillator), set to differ from the i.f. by a suitable audio
frequency. As anumerical example, assume that an in-

mixer tube. Also, the higher the intermediate frequency the higher the image ratio, since
raising the i.f. increases the frequency separation between signal and image and thus places the latter farther away from the peak of the resonance curve (� 2-10) of the signalfrequency circuits.
Other spurious responses -- In addition to images, other signals to which the receiver is not ostensibly tuned may be heard. Harmonics of the high-frequency oscillator may

\17

R F AMPLIFIER

FREQUENCY CONVERTER

IF AMPLIFIER

SECOND DETECTOR

, AUDIO AMPLIFIER

Fig. 716 -- The basic superheterodyne arrangement.

HF OSCILLATOR

BEAT OSCILLATOR

termediate frequency of 455 kc. is chosen, and that the incoming signal is on 7000 kc. Then the h.f. oscillator frequency may be set to 7455 kc. in order that the beat frequency (7455 minus 7000) will be 455 kc. The h.f. oscillator also could be set to 6545 kc., which will give the same frequency difference. To produce an audible c.w. signal of say 1000 cycles at the second detector, the beat oscillator would be set either to 454 !cc. or 456 kc.
Characteristics-- The frequency-conversion process permits r.f. amplification at a relatively-low frequency where high selectivity can be obtained, and this selectivity is constant regardless of the signal frequency. Higher gain is also possible at the low frequencies used for intermediate amplification. The separate oscillators can be designed for stability, and since the h.f. oscillator is working at a frequency considerably removed from the signal frequency its stability is practically unaf-
fected by the strength of the incoming signal. Images --Each h.f. oscillator frequency
will cause i.f. response at two signal frequencies, one higher and one lower than the oscillator frequency. If the oscillator is set to 7455 kc. to respond to a7000-kc. signal, for example,
it will also respond to a signal on 7910 kc., which likewise gives a 455-kc. beat. The undesired signal of the two is called the image. When the r.f. circuit is tuned to the desired signal frequency, and desired-signal and image voltages of equal magnitude are alternately applied to the circuit, the ratio of desiredsignal to image i.f. output is called the signalto-image ratio, or image ratio.
The image ratio depends upon the selectivity of the r.f. tuned circuits preceding the

beat with signals far removed from the desired frequency to produce output at the intermediate frequency; such spurious responses can be reduced by adequate selectivity before the mixer stage and good shielding to prevent signal pickup by any means other than the antenna. When astrong signal is received, the harmonics (� 2-7) generated by rectification in the second detector may, by stray coupling, be introduced into the r.f. or mixer circuit and be converted to the intermediate frequency to go through the receiver in the same way as an ordinary signal. These "birdies" appear as a heterodyne beat on the desired signal and are principally bothersome when the incoming signal is not very greatly different from the intermediate frequency. They can be prevented by proper circuit isolation and shielding. Harmonics of the beat oscillator also can be converted and amplified through the receiver in similar fashion; these responses can be reduced by shielding the beat oscillator and operating it at low output level.
The double superhet -- At high and ultrahigh frequencies it is difficult to secure an adequate image ratio when the intermediate frequency is of the order of 455 kc. To reduce image response the signal frequently is first converted to a rather high intermediate frequency (1500, 5000, or even 10,000 kc.), and then -- sometimes after further amplification -- reconverted to alower i.f. where higher adjacent-channel selectivity can be obtained. Such areceiver is called adouble superheterodyne.
�7-9 FREQUENCY CONVERTERS
Characteristics -- The first detector or mixer resembles an ordinary detector. A cir-

124 CHAPTER SEVEN

Peceiver Principtu and 21Jign

cuit tuned to the intermediate frequency is placed in the plate circuit of the mixer so that the highest possible i.f. voltage will be developed. The signal- and oscillator-frequency
voltages appearing in the plate circuit are bypassed to ground since they are not wanted in
the output. The i.f. tuned circuit should have low impedance for these frequencies, a condition easily, met if they do not approach the intermediate-frequency.
The conversion efficiency of the mixer is measured by the ratio of i.f. output voltage
from the plate circuit to r.f. signal voltage applied to the grid. High conversion efficiency is obviously desirable. The mixer tube noise
also should be low if a good signal-to-noise ratio is wanted, particularly if the mixer is the first tube in the receiver.

(A)
RF lnpu t

I.F T.

OSC .VOLTA GE

(B)

R3
osc

T
VOLTAJE

Fig. 717 -- Mixer or converter circuits. A, grid injection with a pentode plate detector; B and C, separate injection circuits for converter tubes.
Circuit values are as follows:

Circuit ,4

CI, C2, Ca- 0.01-0.1 pfd.

Ci --

approx. 1add.

Ri --

10,000 ohms.

R2--

0.1 megohm.

Ra --

50,000 ohms.

Circuit B
0.01-0.1 pfd. 50-100 ,pfd. 300 ohms. 50,000 ohms. 50,000 ohms.

Circuit C
0.01-0.1 afd. 50-100 pad. 500 ohms. 15,000 ohms. 50,000 ohms.

Plate voltage should he 250 in all three circuits. If an 1851 or 1852 is used in Circuit A, RIshould he changed lo 500 ohms.

The mixer should not require too much r.f. power from the h.f. oscillator, since it may be
difficult to supply the power and maintain good oscillator stability (� 3-7). Also, the conversion efficiency should not depend too critically on the oscillator voltage (that is, a small change in oscillator output should not change the gain appreciably) since it is diffi-
cult to maintain constant oscillator output over awide frequency range.
A change in oscillator frequency caused by tuning of the mixer grid circuit is called pulling. If the mixer and oscillator could be completely isolated, mixer tuning would have no effect on the oscillator frequency, but in practice this is a difficult condition to attain.
Pulling causes oscillator instability and should be minimized, because the stability of the whole receiver depends critically upon the stability of the h.f. oscillator. The pulling effect decreases with the separation between the signal and h.f. oscillator frequencies, hence
is less with high intermediate frequencies. and greater with low i.f.'s.
Circuits -- Typical frequency-conversion circuits are given in Fig. 717. The variations are chiefly in the way in which the oscillator
voltage is introduced. In 717-A, the screen-
grid pentode functions as aplate detector; the oscillator is capacity-coupled to the grid of the tube, in parallel with the tuned input circuit. Inductive coupling may be used instead. The conversion gain and input selectivity are generally good so long as the sum of the two voltages (signal and oscillator) impressed on the mixer grid does not exceed the grid bias. It is desirable to make the oscillator voltage as high as possible without exceeding this limitation. The oscillator voltage required is small
and the power negligible. A pentagrid-converter tube is used in the
circuit at B. Although intended for combination oscillator-mixer use, this type of tube usually will give more satisfactory performance
when used in conjunction with aseparate oscillator, the output of which is coupled in as
shown. The circuit gives good conversion efficiency, and because of the electron coupling gives desirable isolation between the mixer
and oscillator circuits. A small amount of power is required from the oscillator.
Circuit C is for the 6L7 mixer tube. The value of oscillator voltage can vary over a considerable range without affecting the conversion gain. There are no critical adjustments and the oscillator-mixer isolation is good. The oscillator must supply somewhat more power than in B.
A more stable receiver generally results,
particularly at the higher frequencies, when separate tubes are used for the mixer and oscillator. The same number of circuit coin-

125 CHAPTER SEVEN

Dhe Radio AnuzieuA ilandZooh

ponente is required whether or not a combi-
nation tube is used, so that there is little difference from the cost standpoint.
Tubes for frequency conversion-- Any sharp cut-off pentode may be used in the circuit of Fig. 717-A. The 1851 or 1852 give very high conversion gain and an excellent signal-to-noise ratio -- comparable, in fact, to the gain and signal-to-noise ratio obtainable with r.f. amplifiers, and in these respects far superior to any other tubes used as mixers. HoWever, this type of tube loads the circuit more (� 7-6) and thus decreases the selectivity.
The 6K8 is a good tube for the circuit at B; its oscillator plate connection may be ig-
nored. The 6SA7 also is excellent in this circuit, although it has no anode grid (No. 2grid in the diagram). In addition to these two types, any pentagrid converter tube may be
used.

which often is a cause of hum modulation of the oscillator output at 14 Mc. and higher frequencies when 6.3-volt heater tubes are used. Hum is usually not bothersome with 2.5-volt tubes, nor, of course, with tubes which are heated by direct current. The circuit of 718-C overcomes hum with 6.3-volt tubes since the cathode is grounded. The two coils are advantageous in construction, since the feedback adjustment (number of turns on L2) is simple mechanically.
Besides the use of afairly high CIL ratio in the tuned circuit, it is necessary to adjust the
(A)

� 7-10 THE HIGH-FREQUENCY OSCILLATOR
Design considerations -- Stability of the receiver (� 7-2) is chiefly dependent upon the stability of the h.f. oscillator, and particular care should be given this part of the receiver. The frequency of oscillation should be insensitive to changes in voltage, loading, and mechanical shock. Thermal effects (slow change in frequency because of tube or circuit heating) should be minimized. These ends can be attained by the use of good insulating materials
and good-quality circuit components, by suitable electrical design, and by careful mechanical construction.
In addition, the oscillator must be capable
of furnishing sufficient r.f. voltage and power to the particular mixer circuit chosen, at all frequencies within the range of the receiver, and its harmonic output should be as low as possible to reduce spurious response (� 7-8).
It is desirable to make the L/C ratio in the oscillator tuned circuit as low as possible (high-C) since this results in increased stability (� 3-7). Particular care should be taken to insure that no part of the oscillator circuit will vibrate mechanically. This calls for short leads and very "solid" mechanical construction. The chassis and panel material should be heavy and rigid enough so that pressure on the tuning dial will not cause torsion and a shift in the frequency. Care in mechanical construction is well repaid by increased frequency sta-
bility. Circuits--Several oscillator circuits are
shown in Fig. 718. The point at which output voltage is taken for the mixer is indicated by the "X" or "Y" in each case. A and B will give about the same results, and require only one coil. However, in these two circuits the cathode is above ground potential for r.f.,

(E.)

(c)

Fig. 718 -- High-frequency oscillator circuits. A, screen-grid grounded-plate oscillator; B, triode groundedplate oscillator; C, triode, tickler circuit. Coupling to mixer may be taken from points X and Y. In A and B, coupling from Y will reduce pulling effects, but gives less voltage than from X; this type of coupling is therefore best adapted to those mixer circuits with small oscillator-voltage requirements.
Typical values are as follows:

Circuit A

Circuit B

Circuit C

CI--
Ci --
Ca -- Ili -- Et2 --

100 �pfd. 0.1 odd. 0.1 dd. 50,000 ohms. 50,000 ohms.

100 �dd. 0.1 dd. 50,000 ohms. 10,000 to 25,000 ohms.

The "B" supply voltage should be 250 volts. In circuits B and C, R2 is for the purpose of dropping the supply voltage to 100-150 volts; it may be omitted if this voltage is taken from avoltage divider in the power supply (1 8-10).

126 CHAPTER SEVEN

Receiver Principles ana.' 24319rn

feedback to obtain optimum results. Too much
feedback will cause the oscillator to "squeg," or operate at several frequencies simultaneously (� 7-4); too little feedback will cause the output to be low. In the tapped-coil circuits (A, B) the feedback is increased by moving the

Pa,e/e/ Ihmmer
Pane/el rrienner, Sets ej17.Cap.

Tuai

ezbed

erond Oct. Tube

Cathode

SenkeeekingoeReeiny-..regs ffix.C..at,
rTouobsec_

Fig. 719 --Converter circuit tracking methods. Ap-
proximate circuit values for 450- to 465-kc. intermediates with tuning ranges of approximately 2.15-to-I, CIand Cs having amaximum of 140 gpfd. and the total minimum capacitance, including Ca or C4, being 30 to 35 pad.

Tuning Range

L7-4 Mc. 3.7-7.5 Mc. 7-15 Mc. 14-30 Mc.

50 ph. 14 ph. 3.5 ph. 0.8 ph.

40 ph. 12.2 ph. 3 ph. 0.78 �h.

0.0013 pfd. 0.0022 pfd. 0.0045 pfd. None used

Approximate values for 450- to 465-kc. i.f. wth a 2.5-to-1 tuning range, CIand C2 being 350-oad. maximum, minimum capacitance including C3 and C4 being 40 to 50 gpfd.

tap toward the grid end of the coil; in C, by increasing the number of turns on L2 or by moving L2 closer to LI.
The oscillator plate voltage should be as low as is consistent with adequate output. Low plate voltage will reduce tube heating and thereby reduce frequency drift. The oscillator and mixer circuits should be well isolated, preferably by shielding, since coupling other than by the means intended will often result in pulling.
To avoid changes in plate voltage which may cause the oscillator frequency to change,
it is good practice to regulate the plate supply by means of a gaseous voltage regulator tube
(� 8-8). Tracking -- For ganged tuning there must
be a constant difference in frequency between the oscillator and mixer circuits. This difference is equal to the intermediate frequency
7-8). Tracking methods for covering a wide frequency range, suitable for general-coverage receivers, are shown in Fig. 719. The tracking capacity C6 commonly consists of two condensers in parallel, a fixed one of somewhat less capacity than the value needed and asmaller variable in parallel to allow for adjustment to the exact proper value. In practice, the trimmer C4 is first set for the highfrequency end of the tuning range and then the tracking condenser is set for the lowfrequency end. The tracking capacity becomes larger as the percentage difference between the oscillator and signal frequencies becomes smaller (that is, as the signal frequency becomes higher). Typical circuit values are given in the accompanying table. In amateur-band receivers tracking is simplified by choosing aband-spread circuit which gives practically straight-line-frequency tun-
ing (equal frequency change for each dial division) and then adjusting the oscillator and mixer tuned circuits so that both cover the same total number of kilocycles. For example, if the i.f. is 455 kc. and the mixer circuit tunes
from 7000 to 7300 kc. between two given points on the dial, then the oscillator must tune from
7455 to 7755 kc. between the same two dial readings. With the band-spread arrangement of Fig. 712-C the tuning will be practically straight-line frequency if the capacity actually in use at C2 is not too small; the same is true of 712-A if C1 is small compared to C2.

Tuning Range

0.5-1.5 Mc. 1.5-4 Mc. 4-10 Mc. 10-25 Mc.

240 gh. 32 �h. 4.5 ph. 0.8 ph.

14
130 ph. 25 ph. 4 �h. 0.75 ph.

C6
425 ppfd. 0.00115 pfd. 0.0028 pfd. None used

� 7-II THE INTERMEDIATE FREQUENCY AMPLIFIER
Choice of frequency -- The selection of an intermediate frequency is a compromise between various conflicting factors. The lower the i.f., the higher the selectivity and gain, but a low i.f. brings the image nearer the desired

127 CHAPTER SEVEN

Dhe Palio .AmaieuA idandtooh

signal and hence decreases the image ratio (I 7-8). A low i.f. also increases pulling of the oscillator frequency (� 7-9). On the other hand, ahigh i.f. is beneficial to both image ratio and pulling, but the selectivity and gain are lowered. The difference in gain is-least important.
An i.f. of the order of 455 kc. gives good selectivity and is satisfactory from the standpoint of image ratio and oscillator pulling at frequencies up to 7 Mc. The image ratio is poor at 14 Me. when the mixer is connected to the antenna, but adequate when there is a tuned r.f. amplifier between antenna and mixer. At 28 Me. and the ultra-high frequencies the image ratio is very poor unless several r.f. stages are used. Above 14 Mc. pulling is likely to be bad unless very loose coupling can be used between mixer and oscillator.
With an i.f. of about 1600 kc., satisfactory image ratios can be secured on 14, 28 and 56 Mc., and pulling can be reduced to negligible proportions. However, the i.f. selectivity is considerably lower, so that more tuned circuits must be used to increase the selectivity. For ultra-high frequencies, including 28 Mc., the best solution is to use a double superhet (� 7-8), choosing one i.f. for image reduction (5 and 10 Mc. are frequently used) and the second for gain and selectivity.
In choosing an i.f. it is wise to avoid frequencies on which there is considerable activity by the various radio services, since such signals may be picked up directly on the i.f. wiring. The frequencies mentioned are fairly free of such interference.
Circuits-1.f. ampifiers usually consist of one or two stages. Two stages at 455 kc. will give all the gain usable, in view of the minimum receiver noise level, and also give suitable selectivity for good-quality 'phone reception (� 7-2).
A typical circuit arrangement is shown in Fig. 720. A second stage would simply duplicate the circuit of the first. In principle, the i.f. amplifier is the same as the tuned r.f. amplifier (� 7-6). However, since a fixed frequency is used the primary as well as the secondary of the coupling transformer is tuned, giving higher selectivity than is obtainable with aclosely-coupled untuned primary. The cathode resistor, RI,is connected to again control
prPeslctaeatdegieonfg

circuit of the type previously described (� 7-6); usually both stages, if two are used, are controlled by a single variable resistor. The decoupling resistor, R3 ( 2-11), helps isolate the amplifier and thus prevent stray feedback. C2 and R4 are part of the automatic volume control circuit (� 7-13); if no a.v.c. is used the lower end of the i.f. transformer secondary is simply connected to ground.
In atwo-stage amplifier the screen grids of both stages may be fed from acommon supply, either through a resistor (R2)as shown, the screens being connected in parallel, or from a voltage divider (� 8-10) across the plate supply. Separate screen dropping resistors are preferable for preventing undesired coupling between stages.
When two stages are used the high gain will tend to cause instability and oscillation, so that good shielding, by-passing and careful circuit arrangement to prevent stray coupling, with exposed r.f. leads well separated, is necessary.
I.F. transformers-- The tuned circuits of i.f. amplifiers are built up as transformer units consisting of ashielding container in which the coils and tuning condensers are mounted. Both air-core and powdered-iron-core universal-wound coils are used, the latter having somewhat higher Q's and, hence, greater selectivity and gain per unit.
Variable tuning condensers are of the midget type, air-dielectric condensers being preferable because their capacity is practically unaffected by changes in temperature and humidity. Iron-core transformers may be tuned by varying the inductance (permeability tuning) in which case stability comparable to that of variable air-condenser tuning can be obtained
by use of high-stability fixed mica condensers. Such stability is of great importance, since a circuit whose frequency "drifts" with time will eventually be tuned to a different frequency than the other circuits and thereby reduce the gain and selectivity of the amplifier.
Besides the type of i.f. transformer shown in Fig. 720, special units to give desired selectivity characteristics are available. For higher than ordinary adjacent channel selectivity (� 7-2) triple-tuned transformers, with a third tuned circuit inserted between the input and output
Fig. 720 -- Intermediate-frequency amplifier circuit. Typical
vCaIl--ues0.a1repafds.foaltlo4w5s5: kc.; 0.01 pfd.
at 1600 kc. and higher. Cl -- 0.01 �fd.
C3, CS, Cs-- 0.1 pfd. at 455 kc.;
0.01 pfd. at 1600 kc. and higher. Ri -- 300 ohms. 112 -- 0.1 megohm. 113 -- 2000 ohms. R4 -- 0.25 megohm.

128 CHAPTER SEVEN

Peceiver Principi� and 216iyin

windings, are used. The energy is transferred

from the input to the output windings via this tertiary winding; thus adding its selectivity to the overall selectivity of the transformer. Variable-selectivity transformers also can be obtained, these usually being provided with a

third (untuned) winding which can be connected to aresistor, thereby loading the tuned circuits and decreasing the Q and selectivity

(� 2:-10) to broaden the selectivity curve. The variation in selectivity is brought about by switching the resistor in and out of the circuit. Another method is to vary the coupling be-

tween primary and secondary, overcoupling being used to broaden the selectivity curve, undercoupling to sharpen it (� 2-11).
Selectivity -- The overall selectivity of the i.f. amplifier will depend on the frequency and

the number of stages. The following figures are indicative of the band-widths to be expected with good-quality transformers, with construction in which regeneration is kept to a minimum:

Intermediate Frequency

Band Width, kc.

2 times 10 times 100 times

down

One stage 455 kc. (air core)

8.7

One stage 455 kc. (iron core)

4.3

Two stage 455 Ice. (iron core) . 2.9

Two stage 1600 kc

11.0

Two stage 5000 kc

25.8

17.8 10.3
6.4 16.6
46.0

32.3 20.4
10.8 27.4
100.0

Tubesfor I.F. amplifiers -- Variable-g pentodes (� 3-5) are almost invariably used in i.f.
amplifier stages, since grid-bias gain control (� 7-6) is practically always applied to the i.f. amplifier. Tubes with high plate resistance will have least effect on the selectivity of the amplifier, and those with high mutual conductance will give greatest gain. The choice of i.f. tubes will have practically no effect on the signal-tonoise ratio, since this will have been determined by the preceding mixer and r.f. amplifier (if used).
If single-ended tubes are used, care should be taken to keep the plate and grid leads well separated. With these tubes it is advisable to mount the screen by-pass condenser directly on the bottom of the socket crosswise between the plate and grid pins to provide additional shielding, making sure that the outside foil of the condenser is connected to ground.
Single-signal effect -- In heterodyne c.w. reception with a superhet receiver the beat oscillator is set to give a suitable audio-frequency beat note when the incoming signal is converted to the intermediate frequency. For example, the beat oscillator may be set to 456 kc. (the i.f. being 455 kc.) to give a 1000-cycle
beat note. Now if an interfering signal appears at 457 kc., it also will be heterodyned by the beat oscillator to produce a 1000-cycle beat. This audio-frequency image corresponds to the high-frequency images already discussed

(� 7-8) and can be redueed by providing enough selectivity since the image signal is off
the peak of the i.f. resonance curve. When this is done, tuning through a given
signal will show a strong response at the de-
sired beat tone on one side of zero beat only,
instead of the two beat notes on either side of zero beat which are characteristic of less selective reception, hence the name "single signal" reception.
The necessary selectivity is difficult to obtain with non-regenerative amplifiers employing ordinary tuned circuits unless avery low intermediate frequency or alarge number of circuits is used. In practice it is secured either by regenerative amplification or by the use of a crystal filter.
Regeneration -- Regeneration can be used to give a pronounced single-signal effect, particularly when the i.f. is 455 kc. or lower. The resonance curve of an i.f. stage at critical regeneration (just below the oscillating point) is extremely sharp, aband width of 1kc. at 10 times down and 5kc. at 100 times down being
readily obtainable in one stage. The audiofrequency image of agiven signal can thus be reduced by a factor of nearly 100 for a 1000cycle beat note (image 2000 cycles from resonance).
Regeneration is easily introduced in an i.f. amplifier by providing a small amount of
capacity coupling between grid and plate (bringing ashort length of wire, connected to the grid, into the vicinity of the plate lead, usually will suffice) and may be controlled by the regular cathode-resistor gain control. When the i.f. is regenerative, it is usually preferable
to operate the tube at reduced gain (high bias) and depend upon the regeneration to bring the signal strength back to normal. This prevents overloading on strong signals and thereby
increases the effective selectivity. The higher selectivity with regeneration re-
duces the response to noise generated in the
earlier stages of the receiver, just as in the case of high selectivity produced by other means, and therefore improves the signal-to-noise
ratio. The disadvantage is that the regenerative gain varies with the signal strength, being less on strong signals, and the selectivity varies accordingly.
Crystal filters -- The most satisfactory
method of obtaining high selectivity is by the use of apiezo-electric quartz crystal as aselective filter in the i.f. amplifier (� 2-10). Com-
pared to agood tuned circuit, the Q of such a crystal is extremely high. The dimensions of the crystal are made such that it is resonant at
the desired intermediate frequency, and it is then used as a selective coupler between i.f. stages.
Fig. 721 gives a typical crystal-filter reso-

129 CHAPTER SEVEN

Dn eleaclioAmaieur'� -Wand:got:Ph

1000 rawriezwerz

500

111111111�1111111.1�111111111M.
�1111111111111111111MM

�

1.1111.111111.11.111

. 111111111e�

20 11.1MIIMMIMII.

s Z.: :KIWI
IIIINIMM111111111

O
40 �..111ma

parallel-tuned resonant circuit at a frequency
slightly higher than its series-resonant frequency. Signals at the parallel-resonant frequency are thus prevented from reaching the output circuit. The phasing control, by varying the effect of the holder capacity, permits shifting the parallel-resonant frequency over aconsiderable range, thus providing adjustable rejection of interfering signals. The effect of rejection is illustrated in Fig. 721, where the audio image is reduced far below the value that would be expected if the resonance curve were symmetrical.
Variable selectivity-- In circuits such as A and B, Fig. 722, variable selectivity is obtained by adjustment of the variable input impedance, which is effectively in series with the crystal resonator. This is accomplished by

eeee eee ee? 60 KILOCYCLF-S
Fig. 721 -- Graphical representation of single-signal selectivity. The shaded area indicates the region in which response is obtainable.

nance curve. For single-signal reception the audio-frequency image can be reduced by a factor of 1000 or more. Besides practically
eliminating the a.f. image, the high selectivity of the crystal filter provides great discrimination against signals very close to the desired signal in frequency, and, by reducing the band width, reduces the response of the receiver to noise both from sources external to the receiver and in the r.f. stages of the receiver itself.
Crystal-filter circuits; phasing-- Several crystal-filter circuits are shown in Fig. 722. 'Those at A and B are practically identical in performance, although differing in details. The crystal is connected in abridge circuit (� 2-11) with the secondary side of 7'1,the input transformer, balanced to ground either through a pair of condensers, C-C, (A) or by acenter-tap on the secondary, L2 (B). The bridge is completed by the crystal X, and the phasing condenser, C2,which has a maximum capacity somewhat higher than the capacity of the crystal in its holder. When C2 is set to balance the crystal-holder capacity the resonance curve of the crystal circuit is practically symmetrical; the crystal acts as aseries-resonant circuit of very high Q and thus allows signals of the desired frequency to be fed through Cs to L3L4, the output transformer. Without C2 the holder capacity (with the crystal acting as adielectric) would by-pass signals of undesired frequencies to the output circuit.
The phasing control has an additional function besides neutralization of the crystal-holder capacity. The holder capacity becomes apart of the crystal circuit and causes it to act as a

(C)
Fig. 722 -- Crystal filter circuits of three types. All give variable band-width, with C having the greatest range of selectivity. Their operation is discussed in the text. Suitable circuit values are as follows: Circuit A, Ti, special If. input transformer with high-inductance primary, 1,1, closely coupled to tuned secondary, La; C1, 50-pfd. variable; C, each 100-pmfd. fixed (mica); Ca, 10- to 15-gpfd. (max.) variable; Ca, 50-ggfcl. trimmer; L3C4, i.f. tuned circuit, with L3 tapped to match crystalcircuit impedance. In Circuit B, Ti is the same as in Circuit A except that the secondary is center-tapped; CIis 100-umfd. variable; Ca, C3 and C4 same as for Circuit A; L3L4 is atransformer with primary, L4, corresponding to tap on La in A. In Circuit C, Ti is a special id. input transformer with tuned primary and low-impedance secondary; C, each 100-ttpfd. fixed (mica); Ca, opposed-stator phasing condenser, app. 8 mad maximum capacity each side; L3Ca, high-Q i.f. tuned circuit; R, 0to 3000 ohms (selectivity control).

130 CHAPTER SEVEN

Receiver Princip lei anti 21Jign

varying C1 (the selectivity control) which tunes the balanced secondary circuit of T1.When the secondary is tuned to i.f. resonance, the parallel impedance of the L2Ci combination is maximum and is purely resistive (� 2-10). Since the secondary circuit is center-tapped, approximately one-fourth of this resistive impedance is in series with the crystal through C3 and L4. This lowers the Q of the crystal circuit and makes its selectivity minimum. At the same time, the voltage applied to the crystal circuit is maximum.
When the input circuit is detuned from the crystal resonant frequency, the resistance component of the input impedance decreases, and so does the total parallel impedance. Accordingly, the selectivity of the crystal circuit becomes higher and the applied voltage falls off. At first the resistance decreases faster than the applied voltage, with the result that at first the c.w. output from the filter increases as the selectivity is increased. The output then falls off gradually as the input circuit is detuned farther from resonance and the selectivity becomes still higher.
In the circuits of A and B, Fig. 722, the minimum selectivity is still much greater than that of a normal two-stage 455-kc. amplifier, and it is desirable to provide a wider range of selectivity, particularly for 'phone reception. A circuit which does this is shown at Fig. 722-C. The principle of operation is similar, but a much higher value of resistance can be introduced in the crystal circuit to reduce the selectivity. The output tuned circuit L3C3 must have high Q. A compensated condenser is used at C2 (phasing) to maintain circuit balance, so that the phasing control does not affect the resonant frequency. The output circuit functions as a voltage divider in such a way that the amplitude of the carrier delivered to the next grid does not vary appreciably with the selectivity setting. The variable resistor,
may consist of a series of separate fixed resistors selected by atap switch.
� 7-12 THE SECOND DETECTOR AND BEAT OSCILLATOR
Detector circuits-- The second detector of asuperhet receiver performs the same function as the detector in the simple receiver, but usually operates at ahigher input level because of the relatively great r.f. amplification. Therefore the ability to handle large signals without distortion is preferable to high sensitivity. Plate detection is used to some extent, but the diode detector is most popular. It is especially adapted to furnishing automatic gain or volume control (� 7-13), which gives it an additional advantage. The basic circuits are as described in �7-3, although in many cases the diode elements are incorporated in a multi-

purpose tube which also has an amplifier section.
The beat oscillator -- Any standard oscillator circuit (� 3-7) may be used for the beat oscillator. Special beat-oscillator transformers are available, usually consisting of a tapped coil with adjustable tuning; these are most conveniently used with circuits such as those shown at Fig. 718-A and -B, with the output taken from "Y." A variable condenser of about 25 liafd. capacity often is connected between cathode and ground to provide fine adjustment of the beat frequency. The beat oscillator usually is coupled to the second detector tuned circuit through a fixed condenser of a few aefd. capacity.
The beat oscillator should be well shielded to prevent coupling to any part of the circuit except the second detector, and to prevent its harmonics from getting into the front end of the receiver and being amplified like regular signals. To this end, the plate voltage should be as low as is consistent with sufficient audio output. If the beat-oscillator output is too low, strong signals will not give a proportionately strong audio response.
A regenerative second detector may be used
to give the audio beat note, but since the detector must be detuned from the i.f. the selectivity and signal strength are reduced, while blocking (� 7-4) is pronounced because of the high signal level at the second detector.
� '7-13 AUTOMATIC VOLUME CONTROL
Principles-- Automatic regulation of the gain of the receiver in inverse proportion to the signal strength is agreat advantage, especially in 'phone reception, since it tends to keep the output level of the receiver constant regardless of input signal strength. It is readily accomplished in the superheterodyne by using the average rectified d.c. voltage developed by the received signal across aresistance in adetector circuit (� 7-3) to vary the bias on the r.f. and i.f. amplifier tubes. Since this voltage is proportional to the average amplitude of the signal, the gain is reduced as the signal strength is greater. The control will be more complete as the number of stages to which the a.v.c. bias is applied is greater. Control of at least two stages is advisable.
Circuits-- A typical circuit of adiode-triode type tube used as a combined a.v.c. rectifier, detector and first audio amplifier is shown in Fig. 723. One plate of the diode section of the tube is used for signal detection and the other for a.v.c. rectification. The a.v.c. diode plate is fed from the detector diode through the small coupling condenser, C3. Negative bias resulting from the flow of rectified carrier current is developed across R4, the diode

131 CHAPTER SEVEN

DX.. Radio .Amateur's -llancliooh

Fig. 723 --Second-detector and first audio circuit

with a.v.c., using duo-diode-triode tube.

Ri -- 0.25 megohm.

R2 -- 50,000 to 250,000 ohms.

Ra -- 2000 ohms.

R4 -- 2to 5megohms. Rs -- 0.5 to 1megohm. Rs, R7, 119 -- 0.25 megohm. Ro -- 0.25 megohm. Rio -- 0.5-megohm volume con-

EGrid

ee/1.sEtG.rid

trol. Ci, Ca, Ca -- 100 pfd.

-- 0.1 pfd.

Ca, Ce, C7 -- 0.01 pfd.

CCoI,�

C. -- --5.

0.01 90 0.1 pfd. to 10-pfd. electrolytic.

Cli -- 250 ppfd.

E 2nd 1FGrid Lme

4.8 CoTuoplBi.Fn.90Coro,
R9 11C-1-.A.FOutput
i.F.T

R1
; riirreW
c2Z"

C10 CgT

load resistor. This negative bias is applied to the grids of the controlled stages through
the filtering resistors (� 2-11) R5, Rg, R7 and Rg.
It does not matter which of the two diode plates is selected for audio and which for a.v.c. Frequently the two plates are connected together and used as a combined detector-a.v.c. rectifier. This could be done in Fig. 723. The
a.v.c. filter and line would connect to the junction of R2 and C2,while C3 and R4 would be omitted from the circuit.
When S1is closed the a.v.c. line is grounded, thereby removing the a.v.c bias from the
amplifier stages. Delayed a.v.c. -- In Fig. 723 the audio diode
return is made directly to the cathode and the a.v.c. diode return to ground. This places negative bias on the a.v.c. diode equal to the d.c. drop through the cathode resistor (a volt or two) and thus delays the application of a.v.c. voltage to the amplifier grids, since no
rectification takes place in the a.v.c. diode circuit until the carrier amplitude is large enough to overcome the bias. Without this delay, the a.v.c. would start working even with a very small signal. This is undesirable because the full amplification of the receiver then cannot be realized on weak signals. In the audio diode circuit this fixed bias would cause distortion and must be avoided, hence the return is made directly to the cathode.
Time constant -- The time constant (� 2-6) of the resistor-condenser combinations in the a.v.c. circuit is an important part of the system. It must be high enough so that the modulation on the signal is completely filtered from the d.c. output, leaving only an average d.c. component which follows the relatively slow carrier variations with fading; audio-frequency
variations in the a.v.c. voltage applied to the amplifier grids would reduce the percentage of modulation on the incoming signal and in the practical case would cause frequency distortion. On the other hand, the time constant should not be too great since the a.v.c. then

would be unable to follow rapid fading. The values indicated are satisfactory for high-
frequency reception. Signal strength and tuning indicators --
A useful accessory to the receiver is an indicator which will show relative signal strength. Not only is it an aid in giving reports, but it also is helpful in aligning the receiver circuits, in conjunction with a test oscillator or other steady signal.
Three types of indicators are shown in Fig. 724. That at A uses an electron-ray tube, several types of which are available. The grid of the triode section is usually connected to the a.v.c. line. The particular type of tube to use will depend upon the voltage available for its grid; where the a.v.c. voltage is relatively large, aremote-cutoff type tube such as the 6G5 or 6N5 should be used in preference to the sharpcutoff type (6E5).
In B, a milliammeter is connected in series with the d.c. plate lead to one or more r.f. and i.f. tubes whose grids are controlled by a.v.c. Since the plate current of such tubes varies
with the strength of the incoming signal, the meter will indicate relative signal intensity and may be calibrated in "S" points. The scale range of the meter should be chosen to fit the number of tubes in use; the maximum plate current of the average remote-cutoff r.f. pentode is from 7 to 10 milliamperes. The shunt resistor R enables setting the plate current to the full-scale value ("zero adjustment"). With this system the ordinary meter reads downwards from full scale with increasing signal strength, which is the reverse of normal pointer movement (clockwise with increasing reading). Special instruments with the zero-current position of the pointer at the
right-hand side of the scale are used in com-
mercial receivers. The system at C uses a0-1 milliammeter in
a bridge circuit arranged so that the meter reading and signal strength increase together. The current through the branch containing R1
should be approximately equal to the current

132 CHAPTER SEVEN

leeceiver Principles and 2 0eityn

through that containing R2. In some manufactured receivers this is brought about by draining the screen voltage-divider current and the current to the screens of three r.f. pentodes (r.f.

is generally called apreselector, its purpose, in part at least, being to discriminate in favor of the signal against the image. The preselector
may consist of one or more r.f. amplifier stages. When its tuning is ganged with that of the

/Meg

mixer and oscillator, its circuits must track with the mixer circuit.

A VC LINE

oneq.
o

The circuit is the same as discussed earlier (� 7-6). An external preselector stage may be used with receivers having inadequate image ratios, in which case it is built as a separate

+ 250 V.

unit, often with a tuned output circuit which gives a further improvement in selectivity.

The output circuit usually is link-coupled

(� 2-11) to the receiver.

R f

IF.

Signal/noise ratio-- An r.f. amplifier will

have abetter signal-to-noise ratio (�7-2) than a

mixer because the gain is higher and because

the mixer electrode arrangement results in

higher internal tube noise than does the ordi-

nary pentode structure. Hence apreselector is

advantageous in increasing the signal-to-noise

ratio over that obtainable when the mixer is fed directly from the antenna.

Image suppression-- The image ratios

(� 7-8) obtainable at frequencies up to and

including 7Mc. with asingle preselector stage

are high enough, when the intermediate fre-

R F oA

quency is 455 kc., so that for all practical purposes there is no appreciable image response.

Average image ratios on 14 Mc. and 28 Mc. are 50-75 and 10-15, respectively. This is the overall selectivity of the r.f. and mixer tuned cir-

cuits. A second preselector stage, adding one

SCREENS AND VOLTAGE
DIVIDER
(c) +.
Fig. 724 -- Tuning indicator or "S"-meter circuits for superhet receivers. A, electron-ray indicator; It, plate-current meter for tubes on a.v.c.; C, bridge circuit for a.v.c. controlled tube. In B, resistor It should have a maximum resistance several times that of the milliammeter. In C, representative values are: RI, 250 ohms; 112, 350 ohms; R3, 1000-ohm variable.
and i.f. stages) through R2, the sum of these currents being about equal to the maximum plate current of one a.v.c. controlled tube. Typical values for this type of circuit are given. The sensitivity can be increased by making RI, R2 and R3 larger. The initial setting is made with the manual gain control set near maximum, when R3 should be adjusted to make the meter reading zero with no signal.
�7-14 PRESELECTION Purpose-- Preselection is added signal-fre-
quency selectivity before the mixer stage is reached. An r.f. amplifier preceding the mixer

more tuned circuit, will increase the ratios to several hundred at 14 Mc. and to 30 or 40 at 28 Mc.
On ultra-high frequencies it is impracticable to attempt to secure agood image ratio with a 455-kc. i.f. Good performance in this respect can be secured only by using ahigh-frequency i.f. or by using adouble superhet (� 7-8) with a high-frequency first i.f.
Regeneration-- Regeneration may be used in apreselector stage to increase both gain and selectivity. Since this makes tuning more critical and increases ganging problems, regeneration is seldom used except at 14 Mc. and above where adequate image suppression is difficult to obtain with non-regenerative circuits. The same disadvantages exist as in the
case of a regenerative i.f. amplifier (� 7-11). The effect of regeneration is roughly equivalent to the addition of another non-regenerative preselector stage.
The regeneration may be introduced by the same method used in regenerative i.f. amplifiers (� 7-11). The manual gain control of the stage will serve as avolume control.
Regeneration does not improve the signalto-noise ratio, since the tube noise is fed back to the grid circuit along with the signal to add

133 CHAPTER SEVEN

Dh R h e ac oAntateur'� --ilandlool,

noise reduction in code

Phones

reception can be accomplished by amplitude lim-

iting arrangements ap-

plied to the audio output

circuit of areceiver. Such

limiters also maintain the

signal output nearly con-

(A )

+ 220 V.

(B)

+220 v

stant with fading. Diagrams of typical output

Fig. 725 -- Audio output limiting circuits.

limiter circuits are shown

:1-- 0.25 �M. C2-- 0.01 dd. Ca -- 5 dd. Ri -- 0.5 megohm.

B2 -- 2000 ohms. B3 -- 50,000-ohm potentiometer. T -- Output transformer. Li -- 15-henry choke

in Fig. 725. Circuit A employs a triode tube operated at reduced plate voltage (approximately 10

to the thermal agitation noise originally present. The latter noise also is amplified.

volts) so that it saturates at a low signal level. The arrangement of B has better limiting characteristics. A pentode audio tube is

� 7-15 NOISE REDUCTION

operated at reduced screen voltage (35 volts or

Types of noise--In addition to tube and circuit noise (� 7-6) much of the noise interference experienced in reception of amateur signals is caused by domestic electrical equipment and automobile ignition systems. The
interference is of two types in its effects. The first is of the "hiss" type consisting of overlapping pulses, similar in nature to the receiver noise. It is largely reduced by high selectivity in the receiver, especially for code reception. The second is the "pistol shot" or "machine gun" type, consisting of separated impulses of high amplitude. The "hiss" type of interference is usually caused by commutator sparking in d.c. and series a.c. motors, while the "shot" type results from separated spark discharges
(a.c. power leaks, switch and key clicks, ignition sparks, and the like).
Impulse noise-- Impulse noise, because of the extremely short duration of the pulses as
compared to the time between them, must have high pulse amplitude to give much average energy. Hence noise of this type strong enough
to cause much interference generally has an instantaneous amplitude much higher than that of the signal being received. The general principle of devices intended to reduce such noise is that of allowing the signal amplitude to pass through the receiver unaffected, but making the receiver inoperative for amplitudes greater than that of the signal. The greater the amplitude of the pulse compared to its time of duration the more successful the noise-reducing device, since more of the energy in the pulse
can be suppressed. In passing through selective receiver circuits
the time duration of the impulses is increased because of the Q or flywheel effect (� 2-10) of the circuits. Hence the greater the selectivity ahead of the noise-reducing device the more difficult it becomes to secure good noise
suppression. Audio limiting-- A considerable degree of

so) so that the output power remains practically constant over agrid excitation voltage range of more than 100 to 1. These output limiter systems are simple and adaptable to nearly all receivers. However, they cannot prevent noise peaks from overloading previous
circuits. Second detector circuits-- The circuit of
Fig. 726 "chops" noise peaks at the second detector of asuperhet receiver by means of a biased diode which becomes non-conducting above apredetermined signal level. The audio output of the detector must pass through the diode to the grid of the amplifier tube. The diode would normally be non-conducting with the connections shown were it not for the fact that it is given positive bias from a 30-volt source through the adjustable potentiometer R3. Resistors R1 and R2 must be fairly large in value to prevent loss of audio signal.
The audio signal from the detector can be considered to modulate (� 5-1) the steady diode current, and conduction will take place
so long as the diode plate is positive with respect to the cathode. When the signal is sufficiently large to swing the cathode positive with respect to the plate, however, conduction ceases and that portion of the signal is cut off from the audio amplifier. The point at which
cut-off occurs can be selected by adjustment of R3. By setting R3 so that the signal just-passes through the "valve," noise pulses higher in amplitude than the signal will be cut off. The circuit of Fig. 726-A, using an infinite-impedance detector (� 7-3) gives a positive voltage on rectification. When the rectified voltage is negative, as from the usual diode detector (� 7-3) a different circuit arrangement, shown
in Fig. 726-B, is required. An audio signal of about ten volts is required
for good limiting action. When abeat oscillator
is used for c.w. reception the b.f.o. voltage should be small so that incoming noise will not

134 CHAPTER SEVEN

Receiver Principte3 and 2)eiiyn

have astrong carrier to beat against and produce large audio output.
A second-detector noise limiting circuit which automatically adjusts itself to the received carrier level is shown in Fig. 727. The diode load circuit (� 7-3) consists of R6,R7,118 (shunted by the high-resistance audio volume control, R4)and R8 in series. The cathode of the 6N7 noise-limiter is tapped on the load resistor at apoint such that the average rectified carrier voltage (negative) at its grid is approximately twice the negative voltage at the cathode, both measured with reference to ground. A filter network, RIC', is inserted in the grid circuit so that the audio modulation on the carrier does not reach the grid, hence the grid potential is maintained at substantially the rectified carrier voltage alone. The cathode, however, is free to follow the modulation, and when the modulation is 100% the peak cathode voltage will just equal the steady grid voltage.
At all modulation percentages below 100% the grid is negative with respect to cathode and current cannot flow in the 6N7 platecathode circuit. A noise pulse exceeding the peak voltage which represents 100% modula-
2ND DETECTOR
I. F.
.C3 745 audio amp/ear

(A)

30 V.

Fig. 726 -- The series-valve noise-limiter circuit. A, with an infinite-impedance detector; B, with diode detector. Values are as follows:
RI-- 0.25 megohm.
Ra -- 50,000 ohms. R3 -- 10,000-ohm potentiometer. R4 -- 20,000 to 50,000 ohms.
-- 250 lapfd. Ca, Ca -- 0.1 pfd.
Diode circuit constants in B are conventional.

Fig. 727 -- Automatic noise limiting circuit for super. het receivers. T -- 1.f. transformer with balanced secondary for
working into diode rectifier. Ri, 112, Re -- 1megohm. 114 -- 1-megohm volume control. Rs -- 250,000 ohms. Re, Re -- 100,000 ohms. 117 -- 25,000 ohms. Ci -- 0.1-pfd. paper. Ca, Ca -- 0.05-pfd. paper. Cs, Ce -- 50-ppfd. mica. Co -- 0.001-pfd. mica (for r.f. filtering, if needed). Sw -- S.p.s.t. toggle (on-off switch).
The switch should be mounted close to the circuit elements and controlled by an extension shaft if necessary.
tion will, however, make the grid positive with respect to cathode and the relatively-low plate-cathode resistance of the 6N7 shunts the high-resistance audio output circuit, effectively short-circuiting it so that there is practically no response for the duration of the noise peak over the 100% modulation limit,
R6 is used to make the noise-limiting tube more sensitive, by applying to the plate an audio voltage out of phase with the cathode voltage so that at the instant the grid goes positive with respect to cathode, the highest positive potential also is applied to the plate, thus further lowering the effective platecathode resistance.
I.F. noise silencer-- In the circuit shown in Fig. 728 noise pulses are made to decrease the gain of an i.f. stage momentarily and thus silence the receiver for the duration of the pulse. Noise voltage in excess of the desired signal's maximum i.f. voltage is taken off at the grid of the i.f. amplifier, amplified by the noise amplifier stage and rectified by the fullwave diode noise rectifier. The noise circuits are tuned to the i.f. The rectified noise voltage is applied as apulse of negative bias to the No. 3 grid of the 6L7 used as an i.f. amplifier, wholly or partially disabling this stage for the duration of the individual noise pulse, depending on the amplitude of the noise voltage. The noise amplifier-rectifier circuit is biased so that rectification will not start until noise voltage

135 CHAPTER SEVEN

DZ.leach. -Amateur ilandlooh

exceeds the desired-signal amplitude, by means of the "threshold control," R2. For reception with automatic volume control, the a.v.c.
voltage can be applied to the grid of the noise amplifier to augment this threshold bias. This system of noise silencing gives signal-noise ratio improvement of the order of 30 db. (power ratio of 1000) with heavy ignition interference, raising the signal-noise ratio from --10 db. without
the silencer to +20 db. in atypical instance. Circuit values are normal for i.f. amplifiers
(� 7-11) except as indicated. The noise rectifier
transformer T1 has an untuned secondary closely-coupled to the primary, center-tapped for full-wave rectification. The center-tap
rectifier (� 8-3) is used to reduce the possibility of r.f. feedback into the i.f. amplifier (noise silencer) stage. The time constant (� 2-6) of the noise rectifier load circuit, RiC1C2, must be

ent from the intermediate frequency (� 7-8).
This adjustment may be made by tuning in a moderately-weak steady carrier, with the beat oscillator turned off, for maximum signal
strength as determined by maximum hiss, then turning on the beat oscillator and adjusting its frequency (leaving the receiver tuning alone) to give asuitable beat note. Subsequently the
beat oscillator need not be touched except for occasional checking to make certain the frequency has not drifted from the initial setting. The b.f.o. may be set on either the high- or
low-frequency side of zero beat. The use of a.v.c. (� 7-13) is not generally
satisfactory in c.w. reception because the receiver gain rises in the spaces between dots and dashes, giving an increase in noise in the same intervals, and also because the rectified
beat oscillator voltage in the second detec-

IF Irans.

+6 CI NOISE RECT.
6H6

Fig. 728-- ILL noise-silencing circuit. The "B" supply should be 250 volts.
-- 50-250 pea.; use smallest value possible without r.f. feedback.
C2 -- 50 ,dd. Ca -- 0.1 pfd. RI-- 0.1 megohm. R2 -- 5000-ohm volume control. R3 -- 20,000 ohms. R4, 1115 --0.1 megohm. Ti -- Special i.f. transformer for noise
rectifier.

small to prevent disabling the noise silencer stage for alonger period than the duration of the noise pulse. The radio-frequency choke, RFC, must be effective at the intermediate frequency.
Adequate shielding and isolation of the noise amplifier and rectifier circuits from the noise silencer stage must be provided to prevent possible self-oscillation and instability. This 'circuit is preferably applied to the first i.f. stage of the receiver before the high-selectivity circuits are reached, and is most effective when the signal and noise levels are fairly high (one or two r.f. stages before the mixer) since several volts must be obtained from the noise rectifier for good silencing.
� 7-16 OPERATING SUPERHET RECEIVERS
Cay. reception -- Proper adjustment of the beat oscillator is to afrequency slightly differ-

tor circuit also works the a.v.c. circuit. This gives a constant reduction in gain and prevents utilization of the full gain of the receiver. Hence the gain is preferably manually ad-
justed to give suitable audio-frequency output. To avoid overloading in the i.f. circuits it is
usually best to control the i.f. and r.f. gain and keep the audio gain at a fixed value, rather than to use the a.f. gain control as a volume control and permit the r.f. gain to stay fixed
at its highest level. Tuning with the crystal filter -- If the re-
ceiver is equipped with a crystal filter the tuning instructions in the preceding paragraph still apply, but more care must be used both in initial adjustment of the beat oscillator and in tuning. The beat oscillator is set as described above, but with the crystal filter in operation
and adjusted to its sharpest position, if variable selectivity is available. This initial adjust-
ment should be made with the phasing control

136 CHAPTER SEVEN

Receiv.er Principle3 and 2 ,edign

(� 7-11) in the intermediate position, and after it is completed the beat oscillator should be left set and the receiver tuned to the other side
of zero beat (audio-frequency image) on the same carrier to give a beat note of the same tone. This beat will be considerably weaker than the first, and may be "phased out" almost completely by careful adjustment of the phasing control. This is the adjustment for normal operation, and it will be found that one side of zero beat has practically disappeared, leaving the receiver with maximum response on the desired side.
An interfering signal having a beat note differing from that of the a.f. image can similarly be phased out, provided its carrier frequency is not too near the desired carrier.
Depending upon the filter design, maximum selectivity may cause the dots and dashes to lengthen out so that they seem to "run to-
gether." This, plus the fact that the tuning is quite critical with extremely high selectivity, may make it desirable to use somewhat less selectivity in regular operating. However, it
must be emphasized that to realize the benefits
of the crystal filter in reducing interference it is necessary to do all tuning with it in the circuit. The selectivity is so high that it is almost impossible to find the desired station quickly should the filter be switched in only when interference is present.
'Phone reception-- In reception of 'phone signals the normal procedure is to set the r.f. and i.f. gain at maximum, switch the a.v.c. on, and use the audio gain control for setting the volume. This insures maximum effectiveness of
the a.v.c. system in compensating for fading or maintaining constant audio output when either
strong or weak signals are tuned in. On occasions astrong signal close to the frequency of a weaker desired station may take control of the a.v.c., in which case the weaker station will practically disappear because of the reduced gain. In such asituation better reception may result if the a.v.c. is switched off, using the manual r.f. gain control to set the gain at a point which prevents "blocking" by the stronger signal.
A crystal filter will do much toward reducing interference in 'phone bands. Although the high selectivity cuts sidebands and thereby reduces the audio output, especially at the higher audio frequencies, it is possible to use quite high selectivity without destroying intelligibility even though the "quality" of the
transmission suffers. As in the case of c.w. reception, it is advisable to do all tuning with the filter in circuit when interference is likely to
occur. Variable-selectivity filters permit a choice of selectivity which give varying degrees
of sideband cutting to suit conditions. An undesired carrier close in frequency to a

desired carrier will heterodyne with it to produce a beat note equal to the frequency difference. Such aheterodyne can be reduced by adjustment of the phasing control when the crystal filter is used. It cannot be prevented in the "straight" superhet having no crystal filter.
A tone control often will be of help in reducing the effects of high-pitched heterodynes, sideband splatter (� 5-2) and noise, by cutting off the higher audio frequencies. This, like sideband cutting with high selectivity, causes some reduction in naturalness.
Spurious responses--Spurious responses can be recognized without a great deal of difficulty. It is often possible to identify an image by the type of station transmitting, knowing the frequency assignments applying to the frequency to which the receiver is tuned. However, an image also can be recognized by its behavior with tuning. If the signal causes a heterodyne beat note with the desired signal and is actually on the same frequency, the beat note will not change as the receiver is tuned through the signal, but if the interfering signal is an image the beat will vary in pitch as the receiver is tuned. The beat oscillator in the receiver must be off for this test. Using a crystal filter with the beat oscillator on, the image will peak on the opposite side of zero beat to that on which the desired signal peaks.
Harmonic response can be recognized by the "tuning rate," or movement of the tuning dial required to give aspecified change in beat note. Signals getting into the i.f. via highfrequency oscillator harmonics will tune more rapidly (less dial movement) through agiven change in beat note than signals received by normal means.
Harmonics of the beat oscillator can be recognized by the tuning rate of the beat oscillator pitch control. A smaller movement of the control will suffice for agiven change in beat note than is necessary with legitimate signals.
� 7-17 SERVICING SUPERHET RECEIVERS
Il.alignment-- A calibrated signal generator or test oscillator is apractical necessity for initial alignment of an i.f. amplifier. Some means for measuring the output of the receiver also is needed. If the receiver has a tuning meter, its indications will serve for this purpose. Alternatively, if the signal generator is of the modulated type an a.c. output meter (high-resistance voltmeter with copper-oxide rectifier) can be connected across the primary of the output transformer feeding the loudspeaker, or from the plate of the last audio amplifier through a0.1-�fd. blocking condenser (� 2-13) to the receiver chassis. The intensity of sound from the loud-speaker can also be judged by ear (with the modulated test oscilla-

137 CHAPTER SEVEN

5he Radio Amaieur'� --ilanitooh

tor) if no output meter is available, although formers may be aligned by ear, using a weak

this method is not as accurate as those using unmodulated signal adjusted to the crystal

instruments. The procedure is as follows: The test oscil-

peak. Switch on the beat oscillator, adjust to a suitable tone, and align the transformers for

lator is adjusted to the desired intermediate maximum audio output. �

frequency and the "hot" or ungrounded out-

An amplifier which is only slightly out of

put lead is clipped on the grid lead of the last i.f. amplifier tube. The grounded lead is connected to the receiver chassis. The trimmer

alignment as aresult of normal drift from temperature, humidity or aging effects can be realigned by using any steady signal, such as a

condensers of the transformer feeding the sec- local broadcasting signal, in lieu of the test

ond detector are then adjusted for maximum oscillator. Allow the receiver to warm up

signal output. The hot lead from the generator is next clipped on the grid of the next

thoroughly (an hour or so), tune in the signal as usual and "touch up" the i.f. trimmers for

to the last i.f. tube and the second from last
i.f. transformer brought into alignment by adjusting its trimmers for maximum output. This process is continued, working back from the second detector, until all of the i.f. transformers have been aligned. It will be necessary to reduce the output of the signal generator as

maximum output. R.f. alignment-- The object of alignaient
of the r.f. circuits in agang-tuned receiver is to secure adequate tracking over each tuning
range. The adjustment may be carried out with atest oscillator of suitable frequency range or even on noise or such signal as may be heard.

more of the i.f. amplifier is brought into use because the increased gain is likely to cause overloading and consequent inaccurate readings. It is desirable in all cases to use the minimum signal strength which gives useful output readings. The i.f. transformer in the plate cir-
cuit of the mixer is aligned with the signalgenerator lead connected to the mixer grid.

Set the tuning dial at the high-frequency end of the range in use and adjust the h.f. oscillator trimmer condenser for maximum hiss. Next
adjust the mixer trimmer condenser for maximum hiss or signal, then the r.f. trimmers. Reset the tuning dial to the low-frequency end of the range and repeat; if the circuits are �
properly designed no change in trimmer set-

Since the tuned circuit feeding the mixer grid may, because it id tuned to a considerably higher frequency, effectively short-circuit the signal-generator output, it may be necessary to disconnect this circuit. With tubes having
a top grid connection this can be done by removing the grid cap.
If the tuning indicator is used as an output meter the a.v.c. should be switched on; if the audio output method is used the ii.v.c. should be off. The beat oscillator should be off in
either case. If the i.f. amplifier has a crystal filter, the

tings should be necessary. Should it be necessary to increase the trimmer capacity in any circuit, more inductance is needed; if less capacity resonates the circuit, less inductance is required. In the oscillator circuit, the proper frequency range may be secured by adjustment of the tracking condenser capacity (� 7-10)
as well as by inductance adjustment. Tracking is seldom perfect throughout a
tuning range, so that acheck of alignment at intermediate points in the range may show it
to be slightly off. Normally the gain variation from this cause will be small, however, and it

filter should first be switched out and the alignment carried out as above, setting the signal

will suffice to bring the circuits into line at both ends of the range. If most reception is in a

generator as closely as possible to the frequency of the crystal. When completed, the crystal should be switched in and the oscillator frequency varied back and forth over asmall

particular part of the range, such as an amateur band, the circuits may be aligned at that frequency to insure maximum performance, even though the ends of the whole frequency range

range either side of the crystal frequency to may be slightly out of alignment.

find the exact frequency, which will be indicated by a sharp rise in output. Leaving the

Oscillation of r.f. or i.f. amplifiers-- Oscillation in high-frequency amplifier and mixer

generator set on the crystal peak, the i.f. trimmers may be realigned for maximum output. The necessary readjustment should be small. The oscillator frequency should be checked frequently to make sure it has not drifted from the crystal peak.
A modulated signal is not of much value for aligning acrystal-filter i.f. amplifier, since the

circuits is evidenced by squeals or "birdies" as the tuning is varied. It can be caused by poor connections in the common ground cir-
cuits, especially to the tuning condenser rotors.
Inadequate or defective by-pass condensers in cathode, plate and screen-grid circuits also can cause such oscillation. In some cases it may he advisable to provide ashield between

high selectivity cuts sidebands and the results the stators of pre-r.f. amplifier and first-de-

may be inaccurate if the audio output of the receiver is used as a criterion of alignment. Lacking the a.v.c. tuning meter, the trans-

tector ganged tuning condensers, in addition to the usual tube and inter-stage shielding. A metal tube with an ungrounded shell will cause

138 CHAPTER SEVEN

leeceiver Principte� and 2ieJign

this trouble. Improper screen-grid voltage, which might result from a shorted or too-low
screen-grid series resistor, also could be responsible.
Oscillation in the i.f. circuits is independent of high-frequency tuning and is indicated by a continuous squeal which appears when the gain is advanced with the c.w. beat oscillator on. It can result from similar defects in i.f. amplifier circuits. Inadequate cathode resistor by-pass capacitance is acommon cause of such oscillation. Additional by-pass capacitance, of 0.1 to 0.25 j.ifd. usually will remedy it. Similar treatment can be applied to screen-grid and plate by-passes of i.f. tubes.
Instability-- "Birdies" or amushy hiss occurring with tuning of the high-frequency oscillator may indicate that the oscillator is "squegging" or oscillating simultaneously at
high and low frequencies (� 7-4). This may be caused by adefective tube, too-high oscillator plate or screen-grid voltage, excessive feed back in the oscillator circuit or too-high grid-
leak resistance. A varying beat note in c.w. reception indi-
cates instability in either the h.f. oscillator or
beat oscillator, usually the former. The stability of the beat oscillator can be checked by introducing asignal of intermediate frequency (from a test oscillator) into the i.f. amplifier; if the beat note is unstable, the trouble is in the beat oscillator. Poor connections or defective parts are the likely cause. Instability in the high-frequency oscillator may be the result of poor circuit design (� 7-10), loose connections, defective tubes or circuit components, or poor voltage regulation in the oscillator plate and/or
screen supply circuits. Mixer pulling of the oscillator circuit (� 7-9) also will cause the
beat-note to chirp on strong c.w. signals because the oscillator load changes slightly under these conditions.
In 'phone reception with a.v.c., a peculiar
type of instability (" motorboating") may appear if the h.f. oscillator frequency is sensitive to changes in plate voltage. As the a.v.c. voltage rises the electrode currents of the controlled tubes decrease, decreasing the load on
the power supply and causing the plate voltage on the oscillator to rise. The oscillator frequency changes correspondingly, detuning the
circuit and reducing the a.v.c. voltage, thus tending to restore the original conditions. The
process then repeats itself at arate determined by the signal strength and the time constant of
the power supply circuits. It is more pronounced with high selectivity, as when a crystal filter is used, and can be cured by de-
signing the oscillator circuit to be relatively insensitive to plate voltage changes and by
regulating the voltage applied to the oscillator
(� 7-10).

� 748 RECEPTION OF FREQUENCYLATER SIGNALS

Fan. receivers-- A frequency-modulation

receiver differs in circuit design from one

designed for amplitude modulation chiefly in

the arrangement used for detecting the signal.

Detectors for amplitude-modulated signals do

not respond to frequency modulation. It is also

necessary, for full realization of the noise-re-

ducing benefits of the f.m. system, that the

signal applied to the detector be completely

free from amplitude modulation. In practice

this is attained by preventing the signal from

rising above agiven amplitude by means of a

limiter (� 7-15). Since the weakest signal must

be amplitude-limited, high gain must be pro-

vided ahead of the limiter; the superheterodyne

type of circuit is almost invariably used to

provide the necessary gain.

The r.f. and i.f. stages in such asuperhet are

identical in circuit with those in an a.m. re-

ceiver. Since the use of f.m. is confined to the

ultra-high frequencies (above 28 Mc.) a high

intermediate frequency is employed, usually

between 4 and 5 Mc. This not only reduces

image response but also gives the greater

band-width necessary to accommodate wide-

band f.m. signals.

Receiver requirements-- The primary re-

quirements are sufficient r.f. and i.f. gain to

"saturate" the limiter even with a weak

signal, sufficient band-width (� 7-2) to ac-

commodate the full frequency deviation either

side of the carrier frequency without undue

attenuation at the edges of the band, alimiter

circuit which functions properly on both rapid

and slow variations in amplitude, and adetec-

tor which gives a linear relationship between

frequency deviation and amplitude output.

The audio circuits are the same as in other

receivers (� 7-5) except that it is desirable to

cut off the upper audio range by means of a

low-pass filter (� 2-11) because the higher-

frequency noise components have the greatest

amplitude in an f.m. receiver.

The limiter-- Limiter circuits are generally

of the plate saturation type (� 7-15) where low

plate and screen voltage are used to limit the

plate current flow at high signal amplitudes.

Fig. 729-A is atypical circuit. The tube is self.

biased (� 3-6) by agrid leak, RI,and condenser,

C1.R2,R3 and R4 form a voltage divider

(� 8-8) which puts the desired voltages on the

screen and plate. The lower the voltages the

lower the signal level at which limiting occurs,

but the r.f. output voltage of the limiter also

is lower. C2and C3 are the plate and screen

bthye-paisnsteccroomnnedddeeinnassteeerrss,

conventional value for used. The time

constant (� 2-6) of RICI determines the be-

havior of the limiter with respect to rapid and

slow amplitude variations. For best operation

�

139 CHAPTER SEVEN

.h.e leach.

leur 'i ..flanAoo

.-,_,. , , F LF rrons.

+13

on impulse noise (� 7-15) the time constant should be small, but a small time constant limits the range of signal strengths which the limiter can handle without departing from the constant-output condition. A larger time constant is better in the latter respect but is not so effective for rapid variations, hence acompromise set of constants must be used.
The cascade limiter, Fig. 729-B, overcomes this by making the time constant in the first grid circuit suitable for effective operation on impulse noise and that in the second grid (C4R6) optimum for a wide range of input signal strengths. This results, in addition, in more constant output over avery wide range of input signal amplitudes because the voltage at the grid of the second stage is already par-
tially amplitude-limited, thus giving the second stage less work to do. Resistance coupling (R6C4R6)between stages is used in preference to transformer coupling for simplicity and to prevent unwanted regeneration, additional gain at this point being unnecessary.
The rectified voltage developed across R1 in either circuit may be used for a.v.c. (� 7-13).
Discriminator circuits and operation -- The f.m. detector is commonly called a discriminator, because of its ability to discriminate between frequency deviations above and those below the carrier frequency. The circuit generally used is shown in Fig. 730-A. A special i.f. coupling transformer is used between the limiter and detector. Its secondary, LI, is centertapped and is connected back to the plate side of the primary circuit, which is otherwise conventional. C4 is the tuning condenser. The load circuits of the two diode rectifiers (RIC', R2C2) are connected in series; the

Fig. 729 -- P.m. limiter circuits. A, single-tube plate.

saturation limiter; 13, cascade limiter. Typical values

are as follows:

Ci, cuit A

Circuit E

Cu --

100 add.

100 pad.

C2, --

0.1 Al.

0.1 pfd.

250 add.

0.1 megohm.

50,000 ohms.

112

2000 ohms.

Ra --

50,000 ohms.

114 --

0 to 50,000 ohms. 0to 50,000 ohms.

4000 ohms.

He --

0.2 megohm.

Plate supply voltage should be 250 volts in each

circuit.

constants used are of the same order as in ordinary diode detector circuits (� 7-3). The audio output is taken from across the two load resistances.
The primary and secondary circuits are both adjusted to resonance in the center of the i.f. pass-band. The voltage applied to the rectifiers consists of two components, that induced in the secondary by the inductive coupling, and that fed to the center of the secondary through C2. The phase relations between the two are such that at resonance the rectified load currents are equal in amplitude but flow in opposite directions through RI and R2,hence the net voltage across the terminals marked "audio output" is zero. When the carrier deviates from resonance, the induced secondary current either lags or leads, depending upon whether
the deviation is to the high- or low-frequency side, and this phase shift causes the induced current to combine with that fed through C2 in such away that one diode gets more voltage than the other when the frequency is below resonance, while the second diode gets the larger voltage when the frequency is higher

14 0 CHAPTER SEVEN

Receiver Principled and �Oediyn

than resonance. The voltage appearing across
the output terminals is the difference between
the two diode voltages, hence a characteristic
like that of Fig. 731 results, where the net rectified output voltage has opposite polarity
for frequencies on either side of resonance, and up to acertain point becomes greater in amplitude as the frequency deviation is greater. The straight-line portion of the curve is the useful detector characteristic. The separation between the peaks which mark the ends of the linear portion of the curve depends upon the Q's of the primary and secondary circuits and the degree of coupling. The separation becomes
greater with low Q's and close coupling. It is ordinarily set so that the peaks fall just outside
the limits of the pass-band, thus utilizing most of the straight portion of the curve. Since the
audio output is proportional to the change in d.c. voltage with deviation, it is advantageous
from this standpoint to have the peak separation the minimum necessary for alinear charac-
teristic. A second type of discriminator circuit is
shown in Fig. 730-B. Two secondary circuits Si and 82 are used, one tuned above the center
frequency of the i.f. pass-band, the other below. They are coupled equally to the primary,
which is tuned to the center frequency. As the carrier frequency �deviates, the voltages induced in the secondaries will change in ampli-
tude, with the larger voltage appearing across the secondary nearer resonance with the in-
stantaneous frequency. The detection characteristic is similar to that of the first type of discriminator. The peak separation is deter-
mined by the Q's of the circuits, the coefficient of coupling, and the tuning of the two secondaries. high Q's and loose coupling are necessary
for close peak separation.
F.m. receiver alignment-- Alignment of f.m. receivers up to the limiter is carried out as

described in �7-17. For output measurement, a 0-1 milliammeter or 0-500 microammeter should be connected in series with the. limiter
grid resistor (RIin Fig. 729) at the grounded end; or, if the voltage drop across R1 is used for a.v.c. and the receiver is provided with a tuning meter (� 7-13), the tuning meter may be used as an output meter. An accurately calibrated signal generator or test oscillator is desirable, silice the i.f. should be aligned to be as symmetrical as possible; that is, the-output reading should be the same for any two test oscillator settings the same number of kilo-
cycles above or below resonance. It is not necessary to have uniform response over the whole band to be received, although the output at the edges of the band (limit of deviation �5-11) of the transmitted signals) should not be too
low -- not less than 25% of the voltage at resonance. In communications work a bandwidth of 30 kc. or less (15 kc. or less deviation) is commonly used.
Output readings should be taken with the test oscillator set at intervals of afew kilocycles either side of resonance until the band limits are reached, and the i.f. trimmers adjitsted to give as symmetrical acurve as possible.
After the i.f. (and front end) are aligned the limiter operation should be checked. This can be done by temporarily disconnecting C3, if the discriminator circuit of Fig. 730-A is used, disconnecting Riand Ci from the upper diode's cathode in the same diagram, and inserting the milliammeter or microammeter in series with R2 at the grounded end. This converts the discriminator to an ordinary diode rectifier. Varying the signal generator frequency over the channel, with the discriminator transformer adjusted to resonance, should show no change in output (at the band-widths used for common i ti oils purposes) as indicated by the rectdi, ,.I ,�iirrent read by the meter. At this

Fig. 730 -- F.m. discriminator circuits. In both circuits typical values for Ci and C2 are 100 �dd. each; RIand R2, 0.1 megohm each. C3 in A is .`oproximately 50 ufd., depending upon the intermediate frequency; RFC should be of the type designed for the i.f. in use (2.5 mh. is satisfactory for i.f.'s of 4 to 5megacycles). The special three-winding transformer in B is described in the text.
In either circuit the ground may be removed from the lower end of C2 and moved to the junction of Ci and C2 for push-pull audio output.

Limiter Plate

RFC
C1 Audio Output (b) 2
CHAPTER SEVEN 14 1

5he Radio -AmaleuA iiandlooh

Fig. 731 -- Characteristic of a typica fan. detector. The vertical axis represents the voltage developed across the load resistor as the frequency varies from the exact resonance frequency.
A detector with this characteristic would handle f.m. signals up to a band-width of about 150 kc. over the linear portion of the curve.
point various plate and screen voltages can be tried on the limiter tube or tubes to determine the set of conditions which gives maximum output with adequate limiting (no change in rectified current).
When the limiter has been checked the discriminator connections can be restored, leaving the meter connected in series with Ri. Provision should be made for reversing the connections to the meter terminals to take care of the reversal in polarity of the net rectified current. Set the signal generator to the center frequency of the band and adjust the discriminator transformer trimmer condensers to resonance, which will be indicated by zero rectified current. Then set the test oscillator at

the deviation limit (� 5-11) on one side of the
center frequency and note the meter reading. Reverse the meter terminals and set the test
oscillator at the deviation limit on the other side. The two readings should be the same. If
they are not, they can be made so by aslight adjustment of the primary trimmer. This will necessitate re-checking the response at resonance to make sure it is still zero. Generally speaking, the secondary trimmer will chiefly
affect the zero-response frequency, while the primary trimmer will have most effect on the symmetry of the discriminator peaks. A detector curve having satisfactory linearity can be obtained by cut-and-try adjustment of both
trimmers. Tuning and operation -- An f.m. receiver
gives greatest noise reduction when the carrier is tuned exactly to the center of the receiver pass-band and to the point of zero response in
the discriminator. Because of the decrease in noise, this point is readily recognized. Aside from this no special tuning instructions are necessary. The effectiveness of the receiver
will depend almost wholly on how accurately it is aligned.
When an amplitude-modulated signal is tuned in, its modulation practically disappears at exact resonance, only those nonsymmetrical modulation components which may be present being detected. If the signal is to one side or the other of resonance, however, it will be heard and is capable of causing interference to an f.m. signal.

142

(: II A l' 'I' E II SEVEN

CHAPTER EIGHT

Power Supp4

�8-1 POWER SUPPLY REQUIREMENTS
Filament supply -- Except for tubes designed for battery operation, the filaments or heaters of vacuum tubes used in both transmitters and receivers are universally operated on alternating current obtained from the power line through a step-down transformer (� 2-9)
delivering a secondary voltage equal to the
rated voltage of the tubes used. The transformer should be designed to carry the current taken by the number of tubes which may be connected in parallel (� 2-6) across it. The filament or heater transformer is generally
center-tapped to provide a balanced circuit for eliminating hum (� 3-6).
For medium- and high-power r.f. stages of
transmitters, and for high-power audio stages, it is desirable to use aseparate filament trans-
former for each section of the transmitter, installing the transformer near the tube sockets. This avoids the necessity for abnorm-
ally large wires to carry the total filament current for all stages without appreciable filament voltage drop. Maintenance of rated filament voltage is highly important, especially with thoriated-filament tubes, since underor over-voltage may reduce filament life.
Plate supply -- Direct current must be used for the plates of tubes, since any variation in plate current arising from power supply causes will be super-imposed on the signal being received or transmitted, giving an undesirable
type of modulation (� 5-1) if the variations occur at an audio-frequency (� 2-7) rate. Unvarying direct current is commonly called pure
d.c. to distinguish it from current which may be unidirectional but of pulsating character. The use of pure direct current on transmitting tubes is required by FCC regulations on frequencies below 60 megacycles.
Sources of plaie power -- D.c. plate power is usually obtained from rectified and filtered alternating current, but in low-power and portable installations may be secured from batteries. Dry batteries may be used for very low-power portable equipment, but in many cases astorage battery is used as the primary source of power, in conjunction with an inter-
rupter to give pulsating d.c. which is applied to the primary of a step-up transformer (� 8-10).

Rectified a.c. supplies --Since the power line voltage is ordinarily 115 or 230 volts, a step-up transformer (� 2-9) must be used to obtain the desired voltage for the plates of the tubes in the equipment. The alternating secondary current is changed to unidirectional cur-
rent by means of diode rectifier tubes (� 3-1), then passed through an inductance-capacity filter (� 2-11) to the load circuit. The load resistance in ohms is equal to the d.c. output
voltage of the power supply divided by the current in amperes (Ohm's Law, �2-6).
Voltage regulation -- Since there is always
some resistance in power supply circuits, and since the filter normally depends to aconsider-
able extent upon the energy storage of inductance and capacity (� 2-3, 2-5) the output voltage will depend upon the current drain on the supply. The change in output voltage with
change in load current is called the voltage regulation of the supply. Expressed as a percentage,

% Regulation

100 (E1 -- Es) E2

where E1 is the no-load voltage (no current in the load circuit) and E2 the full-load voltage (rated current in load circuit).

�8-2 RECTIFIERS
�Purpose and ratings -- A rectifier is a device which will conduct current in only one direction. The diode tube (� 3-1) is used almost exclusively for the purpose in d.c. power supplies used with radio equipment. The important characteristics of tubes used as power supply rectifiers are the voltage drop between plate and cathode at rated current, the maxi-
mum permissible inverse peak voltage, and the permissible peak plate current.
Voltage drop -- Tube voltage drop depends upon the type of tube. In vacuum rectifiers it increases with the current flowing because of space-charge effect (� 3-1), but can be minimized by using very small spacing between plate and cathode as is done in some rectifiers for receiver power supplies. Mercury-vapor rectifiers (� 3-5) llave aconstant drop of about 15 volts regardless of current. This is much smaller than the voltage drops encountered in vacuum rectifiers.

CHAPTER EIGHT 143

7he Palio _Ain afenr's -.11anidooi

Inverse peak voltage -- This is the maximum voltage developed between plate and cathode of the rectifier when the tube is not conducting; i.e., when the plate is negative with respect to the cathode.
Peak plate current -- This is the maximum instantaneous current flowing through the rectifier. It can never be smaller than the load current in ordinary circuits, and may be several times higher.
Operation of mercury-vapor rectifiers -- Because of its constant voltage drop, the mercury vapor rectifier is more susceptible to damage than the vacuum type. With the latter, the increase in voltage drop tends to limit current flow on heavy overloads, but the mercury-vapor rectifier does not have this limiting action and the cathode may be damaged under similar conditions.
In mercury-vapor rectifiers a phenomenon known as "arc-back," or breakdown of the mercury vapor and conduction in the opposite direction to normal, occurs at high inverse peak voltages, hence such tubes always should be operated within their inverse-peak voltage ratings. Arc-back also may occur if the cathode temperature is below normal, therefore the heater or filament voltage should be checked to make sure that the rated voltage is applied. This check should be made at the tube socket
4.
o
(A)
(B)
(C)
Fig. 801-- Fundamental rectifier circuits.

to avoid errors caused by drop in the leads from the filament transformer to the tube. For the same reason the cathode should be allowed to come up to its final temperature before plate voltage is applied; the time required for this is of the order of 15 to 30 seconds. When atube is first installed or is put into service after a long period of idleness, the cathode should be heated for aperiod of 10 minutes or so before applivation of plate voltage.
�U-3 RECTIFIER CIRCUITS
Half-wave rectifiers -- The simple diode rectifier (� 3-1) is called ahalf-wave rectifier because it can pass only half of each cycle of alternating current. It is shown in Fig. 801-A. At the top of the figure is arepresentation of the applied a.c. voltage, with positive and negative alternations (� 2-7) marked. When the plate is positive with respect to cathode, plate current flows through the load as indicated in the drawing at the right, but when the plate is negative with respect to cathode no current flows. This is indicated by the gaps in the output drawing. The output current is unidirectional, but pulsating.
In this circuit the inverse peak voltage is equal to the maximum transformer voltage, which in the case of asine wave is 1.41 times the r.m.s. voltage (� 2-7).
Full-wave center-tap rectifier -- Fig. 801B shows the "full-wave center-tap" rectifier circuit, so called because both halves of the a.e. cycle are rectified and because the transformer secondary winding must consist of two equal parts with aconnection brought out from the center. When the upper end of the winding is positive, current can flow through rectifier No. 1to the load; this current cannot pass through rectifier No. 2 because its cathode is positive with respect to its plate. The circuit is completed through the transformer center-tap. When the polarity reverses, the upper end of the winding is negative and no current can flow through rectifier No. 1, but the lower end is positive and therefore rectifier No. 2passes current to the load, the return connection again being the center-tap. The resulting wave shape is shown at the right.
Since the two rectifiers are working alternately in this circuit, each half of the transformer secondary must be wound to deliver the full load voltage, hence the total voltage across the transformer terminals is twice that required with the half-wave rectifier. Assuming negligible voltage drop in the particular rectifier which may be conducting at any instant, the inverse peak voltage on the other rectifier is equal to the maximum voltage between the outside terminals of the transformer. In the case of asine wave this is 1.41 times the total secondary r.m.s. voltage (� 2-7).

144 CHAPTER EIGHT

Power Supply

Because energy is delivered to the load at twice the average rate as in the case of ahalfwave rectifier, each tube carries only half the load current.
The bridge rectifier -- The "bridge" type of full-wave rectifier is shown in Fig. 801-C. Its operation is as follows: When the upper end of the winding is positive, current can flow through No. 2 to the load, but not through No. 1. On the return circuit, current flows through No. 3by way of the lower end of the transformer winding. When the polarity reverses and the lower end of the winding becomes positive, current flows through No. 4 and the load and through No. 1by way of the upper side of the transformer. The output wave shape is shown at the right.
The inverse peak voltage is equal to the maximum transformer voltage, or 1.41 times the r.m.s. secondary voltage in the case of a

The ripple frequency depends upon the line frequency and the type of rectifier. In general, it consists of a fundamental plus a series of
harmonics (� 2-7), the latter being relatively unimportant since the fundamental is hardest
to smooth out. With a half-wave rectifier the fundamental is equal to the line frequency; with a full-wave rectifier the fundamental is equal to twice the line frequency, or 120 cycles in the case of a60-cycle supply.
Types of filters-- Inductance-capacity filters are of the low-pass type (� 2-11), using series inductances and shunt capacitances. Practical filters are identified as condenserinput and choke-input, depending upon whether a capacity or inductance is used as the first element in the filter. Resistance-capacity filters (� 2-11) are occasionally used ira applications, particularly in receivers and speech amplifiers, where the current is very low and

sine wave (� 2-7). Energy is delivered to the the voltage drop in the resistor can be tolerload at the same average rate as in the case of ated.

the full-wave center-tap rectifier, so that each

Bleeder resistance --Since the condensers

pair of tubes in series carries half the load in afilter will retain their charge for aconsid-

current.

erable time after power is removed (provided

�

FILTERS

Purpose offilter -- As shown in Fig. 801, the output of a rectifier is pulsating d.c., which would be unsuitable for most vacuum-tube applications (� 8-1). A filter is used to smooth out the pulsations so that practically unvarying direct current flows through the load cir-

the load circuit is open at the time) it is good practice to connect aresistor across the output of the filter to discharge the condensers when the power supply is not in use. The resistance is usually high enough so that only arelatively small percentage of the total output current is consumed in it during normal operation of the supply.

cuit. The filter utilizes the energy-storage [weenies of inductance and capacity (� 2-3, 2-5) by virtue of which energy stored in elec-

Components-- Filter condensers are made in several different types. Electrolytic condensers are available for voltages up to about

tromagnetic and electrostatic fields when the voltage and current are rising is restored to
the circuit when the voltage and current fall, thus filling in the "gaps" or "valleys" in the rectified output.

800, and combine high capacity with small size, since the dielectric is an extremely thin film of oxide on aluminum foil. Condensers for higher voltages are usually made with a dielectric of thin paper impregnated with oil.

Ripple voltage and frequency-- The pulsations in the output of the rectifier can be considered to be caused by an alternating

The working voltage rating of a condenser is the voltage which it will withstand continuously.

current superimposed on a steady direct current (� 2-13). Viewed from this standpoint, the filter may be considered to consist of bypass condensers which short-circuit the a.c. while not interfering with the flow of d.c., and chokes or inductances which permit d.c. to flow through them h.ut which have high reactance for the a.c. (� 2-13). The alternating component is called the ripple. The effective-
ness of the filter may be measured by the percent ripple, which is the r.m.s. value of the a.c. ripple voltage expressed as a percentage of the d.c. output voltage. With an effective

Filter chokes or inductances are wound on
iron cores, with asmall gap in the core to prevent magnetic saturation of the iron at high currents. When the iron becomes saturated its permeability (� 2-5) decreases, consequently the inductance also decreases. Despite the airgap, the inductance of a choke usually varies to some extent with the direct current flowing in the winding, hence it is necessary to specify the inductance at the current which the choke is intended to carry. Its inductance with little
or no direct current flowing in the winding may be considerably higher than the load value.

filter the ripple percentage will be low. Five percent ripple is considered satisfactory for c. w. �8-5 CONDENSER-INPUT FILTERS

transmitters, but lower values (of the order of 0.25%) are necessary for hum-free speech transmission and receiver plate supplies.

Ripple voltage --The conventional condenser-input filter is shown in Fig. 802-A. No simple formulas are available for computing

CHAPTER EIGHT 145

5he Radio AniuieuA -llattcllooh

the ripple voltage, but it will be smaller as both
capacity and inductance are made larger. Adequate smoothing for transmitting purposes can be secured by using 4to 8Add. at CIand C2, and 20 to 30 henrys at Li with 120-cycle ripple (� 8-4). A higher ratio of inductance to capacity may be used at higher load resistances
(� 8-1). For receivers, an additional choke, L2, and
condenser, C3, of the same approximate values,

may decrease to the average value of secondary
voltage, or about 90% of the r.m.s. voltage or even less. Because of this wide range of output voltage with load current the voltage regulation (� 8-1) of the condenser-input filter
is inherently poor. The output voltage obtainable from agiven
supply cannot readily be calculated, since it depends critically upon the load current and filter constants. Under average conditions it will be approximately equal to or somewhat

less than the r.m.s. voltage between center-

From (A )Rectifier
SINGLE-SECTION

Output

tap and one end of the secondary in the fullwave center-tap rectifier circuit (� 8-3).
Ratings of components -- Because the output voltage may rise to the peak transformer
voltage at light loads, the condensers should

have a working-voltage rating (� 8-4) at least

this high and preferably somewhat higher as a

03)Fmn

1-1

TC RextifterTL t

TWO-SECTION 0

L2
2

DC. Output
o

safety factor. Thus in the case of acenter-tap rectifier having a transformer delivering 550 volts each side of the center-tap, the minimum
safe condenser voltage rating will be 550 X 1.41, or 775 volts. An 800-volt or preferably a

Fig. 802 -- Condenser-input filters.

1000-volt condenser should be used. Filter

chokes should have the inductance specified

as shown in Fig. 802-B, are used to give additional smoothing. In such supplies the three condensers are generally 8mfd. each, although

at full-load current, and should have insulation between winding and core adequate to withstand the maximum output voltage.

the input condenser, CI,is sometimes reduced to 4mfd. Inductances of 10 to 20 henrys each will give satisfactory filtering with these capacity values.
For ripple frequencies other than 120 cycles,
the inductance and capacity values should be multiplied by the ratio 120/F,where F is the

� 8-� CHOKE-INPUT FILTERS
Ripple voltage --The circuit of a singlesection choke-input filter is shown in Fig. 803-A. For 120-cycle ripple aclose approximation of the ripple to be expected at the output of the filter is given by the formula:

actual ripple frequency. The bleeder resistance R should be chosen to
draw 10% or less of the rated output current of the supply. Its value is equal to 10,000E11,

Single Section Filter

100 % Ripple = LC

where E is the output voltage and 1 the load where L is in henrys and C in pfd. The product

current in milliamperes. Rectifier peak current -- The ratio of
rectifier peak current to average load current is high with a condenser-input filter. Small rectifier tubes designed for low-voltage supplies (type 80, etc.) generally carry load-cur-
rent ratings based on the use of condenser-

LC must be equal to or greater than 20 to reduce the ripple to 5per cent or less. This figure represents, in most cases, the economical limit for the single-section filter. Smaller percent ages of ripple are usually more economically obtained with the two-section filter of Fig.

input filters. With rectifiers for higher power,

such as the 866/866A, the load current should

not exceed about 25% of the rated peak plate current of one tube when a full-wave rectifier is used, or 3 the rating with half-wave rectification.
Output voltage -- The d.c. output voltage

(A) From

1-1

Rectifier

0.C. Output

SINGLE-SECTION

from acondenser-input supply will, with light

loads or no load, approach the peak trans-

former voltage. This is 1.41 times the r.m.s.

voltage (� 2-7) of the transformer secondary

in the case of Figs. 801-A and C, or 1.41 times the voltage from center-tap to one end of the secondary in Fig. 801-B. At heavy loads it

803 Fig.

-- Choke-input filters.

146 CHAPTER EIGHT

803-B. The ripple percentage (120-cycle ripple) with this arrangement is given by the formula:

Two Section Filter

% Ripple

650 LiL2 (C1 -I- C2) 2

For aripple of 0.25 per cent or less, the denominator should be 2600 or greater.
The formulas can be used for other ripple frequencies by multiplying each inductance and capacity value in the filter by the ratio 120/F, where F is the actual ripple frequency.
The distribution of inductance and capacity in the filter will be determined by the value of input-choke inductance required (next paragraph), and the permissible a.c. output impedance. If the supply is intended for use with an audio-frequency amplifier the reactance (� 2-8) of the last filter condenser should be small (20% or less) compared to the other a.f. resistance or impedance in the circuit, usually
the tube plate resistance and load resistance (� 3-2, 3-3). On the basis of a lower a.f. limit of 100 cycles for speech amplification (� 5-9), this condition is usually satisfied when the output capacity (last filter capacity) of the filter is 4to 8�dd., the higher values being used for the lower tube and load resistances.
The input choke -- The rectifier peak current and the supply voltage regulation depend almost entirely upon the inductance of the input choke in relation to the load resistance (� 8-1). The function of the choke is to raise the ratio of average to peak current (by its energy storage) and to prevent the d.c. output voltage from rising above the average
value (� 2-7) of the a.c. voltage applied to the rectifier. For both purposes its impedance (� 2-8) to the flow of the a.c. component (� 8-4) must be high.
The value of input choke inductance which prevents the d.c. output voltage from rising above the average of the rectified a.c. wave is called the critical inductance, and for 120cycle ripple frequency is given by the approximate formula:

Loeit.-- Load

resistance 1000

(ohms)

For other ripple frequencies, the inductance required will be the above value multiplied by the ratio of 120 to the actual ripple frequency.
With inductance values less than critical the d.c. output voltage will rise because the
filter tends to act as a condenser-input filter (� 8-5). With critical inductance the peak plate current of one tube in acenter-tap recti-
fier will be approximately 10% higher than the d.c. load current taken from the supply.
An inductance of twice the critical value is called the optimum value. It gives afurther re-
duction in the ratio of peak to average plate

Power SuPP4

current, and represents the point at which further increase in inductance does not give a corresponding return in improved operating characteristics.
Swinging chokes --The formula for critical inductance indicates that the inductance required varies widely with the load resistance. In the case where there is no load except the bleeder (� 8-4) on the power supply the critical inductance required is highest; much lower values are satisfactory when the full-load current is being delivered. Since the inductance of a choke tends to rise as the direct current flowing through it is decreased (� 8-4) it is possible to effect an economy in materials by designing the choke to have a"swinging" characteristic such that it has the required critical
inductance value with the bleeder load only, and about the optimum inductance value at full load. Thus in the case where the bleeder resistance is 20,000 ohms and the full-load resistance (including the bleeder) 2500 ohms, a choke which swings from 20 henrys to 5 henrys over the full output-current range will fulfill the requirements.
Resonance --Resonance effects in the series circuit across the output of the rectifier formed by the first choke (Li) and first filter condenser (C1) must be avoided, since the ripple voltage would build up to large values (� 2-10). This is not only the opposite action to that for which the filter is intended, but also may cause excessive rectifier peak currents and abnormally high inverse-peak voltages. For full-wave rectification the ripple frequency will be 120 cycles for a60-cycle supply (� 8-4) and resonance will occur when the
product of choke inductance in henrys times condenser capacity in microfarads is equal to 1.77. The corresponding figure for 50-cycle supply (100-cycle ripple frequency) is 2.53 and for 25-cycle supply (50-cycle ripple frequency) 13.5. At least twice these products should be used to ensure that no resonance effects will be present.
Output voltage -- Provided the inputchoke inductance is at least the critical value, the output voltage may be calculated quite closely by the equation:

E. = 0.9E8 -- (L,

(Ri 1000

Rs)

where E. is the output voltage; Eeis the r.m.s. voltage applied to the rectifier (r.m.s. voltage between center-tap and one end of the secondary in the case of the center-tap rectifier); lb and //, are the bleeder and load currents, respectively, in milliamperes; R1 and R2 are the resistances of the first and second filter chokes; and Er is the drop between rectifier
plate and cathode (� 8-2). These voltage drops are shown in Fig. 804.

147 CHAPTER EIGIIT

.7he Pacho Amaieur'i -llutitgooi

�-h--E,

of the filter chokes, and E, is the voltage drop in the rectifier. Ei is the full load r.m.s. (� 2-7)

secondary voltage; the open-circuit voltage

usually will be 5% to 10% higher.

Load

Volt-ampere rating-- The volt-ampere rating (� 2-8) of the transformer depends upon

the type of filter (condenser or choke input).

With acondenser-input filter the heating effect

Bleeder

in the secondary is higher because of the high ratio of peak to average current, consequently

Fig. �tal --

drops in the power supply circuit.

the volt-amperes consumed by the transformer may be several times the watts delivered to

At no load IL is zero, hence the no-load
voltage may be calculated on the basis of bleeder current only. The voltage regulation may be determined from the no-load and fu!!-

the load. With a choke-input filter, provided the input choke has at least the critical inductance (� 8-6), the secondary volt-amperes can be calculated quite closely by the equation:

load voltages (� 8-1).

Sec. V.A. = 0.00075 El

Ratings of components-- Because of better voltage regulation, filter condensers are
subjected to smaller variations in d.c. voltage than in the condenser-input filter (� 8-5). However, it is advisable to use condensers rated for the peak transformer voltage in case the bleeder resistor should burn out when there is no external load on the power supply, since in this

where E is the total r.m.s. voltage of the secondary (between the outside ends in the case of a center-tapped winding) and /is the d.c., output current in milliamperes (load current plus bleeder current). The primary volt-
amperes will be 10% to 20% higher because of
transformer losses.

case the voltage will rise to the same maximum value as with acondenser-input filter.
The input choke may be of the swinging type, the required no-load and full-load inductance values being calculated as described above. The second choke (smoothing choke) should have constant inductance with varying d.c. load currents. Values of 10 to 20 henrys are ordinarily used. Since chokes are usually placed in the positive leads, the negative being grounded, the windings should be insulated
from the core to withstand the full d.c. output voltage of the supply.

� 8-8 VOLTAGE STABILIZATION
Gaseous regulator tubes--There is frequent need for maintaining the voltage applied to alow-voltage, low-current circuit (such as the oscillator in asuperhet receiver or the frequency-controlling oscillator in atransmitter) at a practically constant value regardless of the voltage regulation of the power supply or variations in load current. In such applications gaseous regulator tubes (VR105-30, VR150-30, etc.) can be used to good advantage. The voltage drop across such tubes is constant over a moderately-wide current

� 8-7 THE PLATE TRANSFORMER
Output voltage--The output voltage of the plate transformer depends upon the required d.c. load voltage and the type of rectifier circuit. With condenser-input filters the r.m.s. secondary voltage is usually made equal to or slightly more than the d.c. output voltage, allowing for voltage drops in the rectifier tubes and filter chokes as well as in the transformer itself. The full-wave center-tap rectifier requires a transformer giving this voltage each side of the secondary center-tap (� 8-3).
With achoke-input filter the required r.m.s. secondary voltage (each side of center-tap for a center-tap rectifier) can be calculated by the equation:

range. The first number in the tube designation indicates the terminal voltage, the second
the maximum permissible tube current. The fundamental circuit for agaseous regu-
lator is shown in Fig. 805-A. The tube is con-
nected in series with a limiting resistor, RI, across a source of voltage which must be
higher than the starting voltage, or voltage required for ionization of the gas in the tube.
The starting voltage is about 30% higher than the operating voltage. The load is connected in parallel with the tube. For stable operation a minimum tube current of 5 to 10 milliamperes is required. The maximum permissible current with most types is 30 milliamperes,
consequently the load current cannot exceed 20 to 25 milliamperes if the voltage is to be

E, = 1.1 [E. � 1(R110+00R2) Eal

stabilized over arange from zero to maximum load current.

The value of the limiting resistor must lie

where E. is the required d.c. output voltage, /is the load current (including bleeder current) in milliamperes, R1 and R2 are the resistances

between that which just permits minimum tube current to flow and that which just passes the maximum permissible tube current when

148 CHAPTER EIGHT

Power suppty

there is no load current.. The latter value is generally used. It is given by the equation
R -- 1000 (E.-- E,.)
Where R is the limiting resistance in ohms, E. the voltage of the source across which tube and resistor are connected, E,. is the rated voltage drop across the regulator tube, and I is the maximum tube current in milliamperes (usually 30 ma.).
Fig. 805-B shows how two tubes may be used in series to give a higher regulated voltage than is obtainable with one, and also to give two values of regulated voltage. The lim-

A -

regFuilga.to8r05tu--besV.oltage stabilizing circuits using gaseous

iting resistor may be calculated as above, using
the sum of the voltage drops across the two tubes for E,.. Since the upper tube must carry
more current than the lower, the load connected to the low-voltage tap must take small current. The total current taken by the loads on both the high and low taps should not exceed 20 to 25 milliamperes.
Voltage regulation of the order of 1% can be obtained with tubes of this type.
Electronic voltage regulation-- A voltage regulator circuit suitable for higher voltages and currents than the gas tubes, and also hav-
ing the feature that the output voltage can be
varied over a rather wide range, is shown in Fig. 806. A high-gain voltage amplifier tube (� 3-3), usually asharp-cutoff pentode (� 3-5) is connected in such away that asmall change in the output voltage of the power supply causes achange in grid bias and thereby acor-
responding change in plate current. Its plate current flows through a resistor (R5)the voltage drop across which is used to bias asecond tube -- the "regulator" tube -- whose platecathode circuit is connected in series with the
load circuit. The regulator tube therefore functions as an automatically-variable series resistor. Should the output voltage increase
slightly, the bias on the control tube becomes
more positive, causing the plate current of the control-tube to increase and the drop across
R6to increase correspondingly. The bias on the regulator tube therefore becomes more negative and the effective resistance of the regulator tube increases, causing the terminal

voltage to drop. A decrease in output voltage causes the reverse action. The time lag in the action of the system is negligible and with proper circuit constants, the output voltage can be held within a fraction of a per cent of the desired value throughout the useful range of load currents and over awide range of supply voltages.
An essential in the system is the use of a constant-voltage bias source for the control tube. The voltage change which appears at the grid of the tube is the difference between a fixed negative bias and a positive voltage which is taken from the voltage divider across the output. To get the most effective control, the negative bias must not vary with plate current. The most satisfactory type of bias is adry battery of 45 to 90 volts, but agaseous regulator tube (VR75-30) or aneon bulb of the type without the resistor in the base may be used instead. This is indicated in the diagram. If the gas tube or neon bulb is used, anegativeresistance type of oscillation (� 3-7) may take place at audio frequencies or above, in which ease acondenser of 0.1 dd. or more should be connected across it. A similar condenser between the control tube grid and cathode is also frequently helpful in this respect.
The variable resistor R3 is used to adjust the bias on the control tube to the proper operating value. It also serves as an output voltage control, setting the value of regulated voltage within the existing operating limits.
The maximum output voltage obtainable is equal to the power supply voltage minus the minimum drop through the regulator tube. This drop is of the order of 50 volts with the
4SPuFoPpwpOelTry
Fig. 806 --Electronic voltage regulator. The regulator tube is ordinarily a2A3 or anumber of them in parallel, the control tube a6SJ7 or similar type. The filament transformer for the regulator tube must be insulated for the plate voltage, and cannot supply current to other tubes when afilament-type regulator tube is used. Typical circuit values are as follows: RI, 10,000 ohms; R2, 25,000 ohms; Rs, 10,000-ohm potentiometer; R4, 5000 ohms; RS,0.5 megohm.

149 CHAPTER EIGHT

Dne Radio Amaleur'� ilancliooh

tubes ordinarily used (power triodes having low plate resistance, such as the 2A3). The maximum current is also limited by the regulator tube; 100 milliamperes is a safe value for a 2A3. Two or more regulator tubes may be connected in parallel to increase the currentcarrying capacity, no other changes in the circuit being required.
� 8-9 BIAS SUPPLIES
Requirements -- A bias supply is not called upon to deliver current to a load circuit but simply to furnish afixed grid voltage to set the operating point of atube (� 3-3). However, in most applications it is nevertheless true that current flows through the bias supply, because such supplies are chiefly used in connection with power amplifiers of the Class-B and Class-C type where grid-current flow is a feature of operation (� 3-4). In circuit design abias supply resembles the rectified a.c. plate supply (� 8-1), having a transformer-rectifierfilter system employing similar circuits. Bias supplies may be classified in two types, those furnishing only protective bias, intended to prevent excessive plate current flow in a power tube in case of loss of grid leak bias (� 3-6) from excitation failure, and those which furnish the actual operating bias for the tubes. In the former type voltage regulation (� 8-1) is relatively unimportant; in the latter it may be of considerable importance.
In general, a bias supply should have wellfiltered d.c. output, especially if it furnishes the operating bias for the stage, since ripple voltage may modulate the signal on the grid
of the amplifier tube (� 5-1). Condenser-input filters are generally used, since the regulation of the supply is not a function of the filter. The constants discussed in �8-5 are appli-
cable. Voltage regulation-- A bias supply must
always have a bleeder resistance (� 8-4) connected across its output terminals to provide a d.c. path from grid to cathode of the tube being biased. Although the grid circuit takes no current from the supply, grid current flows
through the bleeder resistor and the voltage across the resistor therefore varies with grid current. This variation in voltage is practically independent of the design of the bias supply unless special voltage-regulating means are
used. Protective bias -- This type of bias supply
is designed to give an output voltage sufficient
to bias the tube to which it is applied to or near the plate-current cut-off point (� 3-2). A typical circuit is given in Fig. 807. The resistance R1 is the grid-leak resistor (� 3-6) for the amplifier tube with which the supply is used, and the normal operating bias is developed by the flow of grid current through this

fC
Fig. 807 -- Supply for furnishing protective bias to a power amplifier. The transformer T should furnish a peak voltage at least equal to the protective bias required. Other constants are discussed in the text.

resistor. R2is connected in series with R1across the output of the supply to reduce the voltage across RI,when there is no grid-current flow,
to the cut-off value for the tube being biased. R2 is given by the formula

Eg -- Ec Eo

where Eg is the output voltage of the supply with R2 and R1 in series as a load, .L is the cut-off bias for the tube with which the supply is used, and R1 is as described above.
When such asupply is used with a Class-C
amplifier, the voltage across R1 from gridcurrent flow will normally be higher than that from the bias supply itself, since the latter is adjusted to cut-off while the operating bias will be twice cut-off or higher (� 3-4). In some cases the grid-leak voltage may even exceed the peak output voltage of the transformer
(1.41 times half the total secondary voltage, in the circuit shown). The filter condensers in such abias supply must therefore be rated to stand the maximum operating bias voltage on the Class-C amplifier, if this voltage exceeds the nominal output voltage of the supply.
Voltage stabilization--When the bias supply furnishes operating rather than simply protective bias, the value of bias voltage should be as constant as possible even when the grid current of the biased tube varies. A simple method of improving bias voltage regulation is to make the bleeder resistance
low enough so that the current through it from the supply is several times the maximum grid current to be expected. By this means the percentage variation in current is reduced. This method, however, requires that aconsiderable amount of power be dissipated in the bleeder, which in turn calls for arelatively large power
transformer and filter choke. Bias voltage variation may also be reduced
by means of aregulator tube, as shown in Fig. 808. The regulator tube is usually a triode having aplate-current rating adequate to carry
the expected grid current. It is cathode-biased (� 3-6) by the resistor RI,which is of the order of several hundred thousand ohms or a few
megohms so that with no grid current the tube

150 CHAPTER EIGHT

Power Supp4

is biased practically to cut-off. Because of this high resistance, the grid current will flow through the plate resistance of the regulator tube, which is comparatively low, rather than through R1 and R2,hence the voltage from the supply across R1 and the cathode-plate circuit of the regulator tube in series can be
Fig. 808 -- Automatic voltage regulator for bias supplies. For best operation the tube used should be one of high mutual conductance (� 3-2).
considered constant. The bias voltage is equal to the voltage across the tube alone. When grid current flows the voltage across the tube will tend to increase, hence the drop across R1 decreases, lowering the bias on the regulator and reducing its plate resistance. This in turn reduces the tube voltage drop, and the bias voltage tends to remain constant over afairly wide range of grid current values.
At low bias voltages it may be necessary to use anumber of tubes in parallel to get sufficient variation of plate resistance for good regulating action. The bias supply must furnish the required bias voltage plus the voltage required to bias the regulator tube to cut-off, considering the output bias voltage as the plate voltage applied to the regulator. The current taken from the bias supply is negligible. R2 may be tapped to provide arange of bias voltages to meet different tube requirements.
Multi-stage bias supplies-- When several power amplifier tubes are to be biased from a

From Bias Supply

-c

Fig. 809 --Isolating circuit for multiple-stage bias supply.

single supply, the various bias circuits must be isolated by some means. If the grid currents of all stages should flow through a single bleeder resistor a variation in grid current in one stage would change the bias on all, a condition which would interfere with effective adjustment and operation of the transmitter.
When protective bias is to be furnished several stages, the circuit arrangement of Fig. 809, using rectifier tubes to isolate the individual grieleaks of the various stages, may be employed. In the diagram two type 80 rectifiers are used to furnish bias to four stages. Each pair of resistors (RiR2) constitutes a separate bleeder across the bias supply. R1 is the grid-leak for the biased stage; R2 is adropping resistor to adjust the voltage across R1 to the cut-off value (without grid-current flow) for the biased tube. The values of RI and R2 may be calculated as described in the paragraph on protective bias. In this case the bias supply should be designed to have inherently good voltage regulation; i.e., a choke-input filter with appropriate filter and bleeder con-
Fig. 810 -- Use of gaseous regulator tubes to stabilize bias voltage.
stants (� 8-6) should be used, the bleeder being separate from those associated with the rectifier tubes. When the voltage across R1R2 rises because of grid-current flow through R1, the load on the supply will vary (hence the necessity for good voltage regulation in the supply) but there is no interaction of grid currents in the separate bleeders because the rectifiers can pass current in only one direction.
When asingle supply is to furnish operating bias for several stages, a separate regulator tube circuit (Fig. 808) may be used for each one. Individual voltages for the various stages may be obtained by appropriate taps on R2.
Well-regulated bias for several stages may be obtained by the use of gaseous regulator tubes when the voltage and current ratings of the tubes permit their use. This is shown in Fig. 810. A single tube or two or more in series can be used to give the desired bias voltage drop; the bias supply voltage must be high enough to provide starting voltage for the tubes in series. R1 is the protective resistance (� 8-8); its value should be calculated for mini-

151 CHAPTER EIGHT

Dh R f e a ioAntaieuA -.11andiood

mum stable tube current. The maximum grid current that can be handled is 20 to 25 milliamperes with available regulator tubes.
�8-10 MISCELLANEOUS POWER SUPPLY CIRCUITS
Voltage dividers -- A voltage divider is a resistance connected across asource of voltage and tapped at appropriate points from which voltages lower than the terminal voltage may be taken (� 2-6). Since the voltage at any tap depends upon the current drawn from the tap, the voltage regulation (� 8-1) of such adivider is inherently poor. Hence a voltage divider is best suited to applications where the currents drawn are constant, or where separate voltageregulating circuits (� 8-8) are used to compensate for voltage variations at the taps.
A typical voltage divider arrangement is shown in Fig. 811 The terminal voltage is E, and two taps are provided to give lower voltages El and E2 at currents Ii and /2 respectively. The smaller the resistance between taps in proportion to the total resistance, the smaller the voltage between the taps. In addition to the load currents /1 and /2 there is also the bleeder current, /b. The voltage divider may be the bleeder for the power supply. For convenience, the voltage divider in the figure is considered to be made up of separate resistances, RI,R2, Ra, between taps. RI carries only the bleeder current, Ib. R2 carries /1 in addition to /b; R3carries /2, Ii

From Power Supply
16,

+E E2 E,

Fig. 811 -- Typical voltage-divider circuit.

and lb. For the purpose of calculating the resistances required, ableeder current Ib must be assumed; generally it .is low compared to the total load current (10% or so). Then

RI =
I b

11.2

E -- E2
-- lb + 11 � 12

the currents being expressed in amperes. The method may be extended to any de-
sired number of taps, each resistance section being calculated by Ohm's Law (� 2-6) using the voltage drop across it and the total current through it. The power dissipated by each sec-

tion may be calculated by multiplying the
same quantities together. In case it is desired to have the bleeder re-
sistance total to a predetermined value, the
same method of calculation may be followed, but different values of bleeder current should be tried until the correct result is found.
Transformerless plate supplies -- It is pos-
sible to rectify the line voltage directly, without using a step-up power transformer, for certain applications (such as some types of receivers) where the low voltage so obtained is satisfactory. A simple power supply system of this type, using a half-wave rectifier, is shown in Fig. 812. Tubes for this purpose are provided with heaters operating at relatively high voltages (25, 35, 70, or 115 volts) which can be connected across the line in series with other tube filaments and/or a resistor R of suitable value to limit the current to the rated
value for the tube heater. The rectifier is often

115 v

output It0000,1-0+
c L cTI D.C.

o
Fig. 812 -- Transforruerless plate supply with half. wave rectifier. �

incorporated in the same tube envelope with an audio power amplifier tube.
The half-wave circuit shown has a fundamental ripple frequency equal to the line frequency (� 8-4) and hence requires more inductance and capacity in the filter for agiven
ripple percentage (� 8-5) than the full-wave rectifier. A condenser-input filter is generally
used, frequently with a second choke and third condenser (� 8-5) to provide the necessary smoothing.
A disadvantage of the transformerless cir-
cuit is that no ground connection can be used on the power supply unless care is used to insure that the grounded side of the power line is connected to the grounded side of the supply. Receivers using this type of supply are generally grounded through a low capacity (0.05 eifd.) condenser to avoid short-circuiting the line should the line plug be inserted in the socket the wrong way. The input condenser should be at least 16 and preferably 32 dd. to keep the output voltage high and to improve voltage regulation.
Voltage-doubling circuits -- The circuit arrangement of Fig. 813, frequently used in transformerless plate supplies, gives full-wave rectification combined with doubling of the output voltage. This is accomplished by using a double-diode rectifier, one section of which charges CI when the line polarity between its
plate and cathode is positive while the other

152 CHAPTER EIGHT

Power supply

Fig. 813 -- Full-wave voltage-doubling transformer. Iras plate supply circuit.
section charges C2 when the line polarity reverses. Each condenser is thus charged separately to the same d.c. voltage, and they discharge in series into the load circuit. For effective operation of this circuit the capacities of C1 and C2 must be at least 16 mfd. each and preferably higher.
The ripple frequency with this circuit is twice the line frequency, since it is afull-wave circuit (� 8-4). The voltage regulation is in-
herently poor and depends critically upon the capacities of C1 and C2, being better as these capacities are made larger. A typical supply with 16 dd. each at CIand C2 will have an output voltage of approximately 300 at light loads, dropping to about 210 volts at the rated current of 75 milliamperes.
No direct ground can be used on this supply or on the equipment with which it is used. If an r.f. ground is made through a condenser, the condenser capacity should be small (about 0.05 nfd.) since it is in shunt from plate to cathode of one rectifier. A large capacity (low reactance) would by-pass the rectifier and thereby nullify its operation.
Duplex plate supplies -- In some cases it may be advantageous economically to obtain two plate supply voltages from asingle power
supply, making one or more of the components serve a double purpose. Two circuits of this type are shown in Figs. 814 and 815.
In Fig. 814 a bridge rectifier is used to obtain the full transformer voltage, while aconnection is also brought out from the center tap

to obtain a second voltage corresponding to half the total transformer secondary voltage. The sum of the currents drawn from the two taps should not exceed the d.c. ratings of the rectifier tubes and transformer. Filter values for each tap should be computed separately (� 8-6).
Fig. 815 shows how a transformer with multiple secondary taps may be used to obtain both high and low voltages simultaneously. A separate full-wave rectifier is used at each tap. The filter chokes are placed in the common negative lead, but separate filter condensers are required. The sum of the currents drawn from each tap must not exceed the transformer rating and the chokes must be rated to carry the total load current. Each bleeder resistance should have avalue in ohms of 1000 times the maximum rated inductance in henrys of the swinging choke, LI, for best regulation (� 8-6).
Rectifiers in parallel -- Vacuum-type rectifiers may be connected in parallel (plate to plate and cathode to cathode) for higher cur-
�d.V.
Fig. 815--Power supply circuit in which a single transformer and set of chokes serve for two different voltages.
rent carrying capacity. No circuit changes are required.
When mercury vapor rectifiers are connected in parallel, slight differences in tube characteristics may make one ionize at aslightly lower voltage than the other. Since the ignition voltage is higher than the operating voltage, this means that the first tube to ionize carries the whole load, since the voltage drop is then too low to ignite the second tube. This condition can be prevented by connecting resistors of 50 to 100 ohms in series with each plate as shown in Fig. 816, thereby insuring that a high-enough voltage for ignition will always be available.

+ H V

Fig. 814--Combination

bridge and center-tap recti-

�

fier to deliver two output

voltages with good regula-

tion.

+L.V.
CIIAPTER EIGHT

153

..7he Radio AmaleuA ilandtooh

the upper half of the transformer primary.

The magnet coil is again energized and the

cycle repeats itself, usually at a rate about

equivalent to a60-cycle supply frequency.

The synchronous circuit of Fig. 817-B is

provided with an extra pair of contacts which

rectify the secondary output of the trans-

former, thus eliminating the need for a sepa-

rate rectifier tube. The secondary center-tap

furnishes the positive output terminal when

the relative polarities of primary and second-

ary windings are correct. The proper connec-

Fig. 816 -- Operating mercury-vapor rectifiers in parallel. Resistors marked R should have values between 50 and 100 ohms.

tions may be determined by experiment, reversing the secondary connections if the first trial is wrong.
The buffer condenser, Cg, across the trans-

Vibrator power supplies --For portable former secondary is used to absorb surges

or mobile work the most common source of which would occur on breaking the current,

power for both filaments and plates is the 6- when the magnetic field collapses practically

volt automobile-type storage battery. Fila- instantaneously and hence causes avery high

ments may be heated directly from the battery, voltage to be induced in the secondary (� 2-5).

while plate power is obtained by passing cur- Its value is usually between 0.005 and 0.03

rent from the battery through the primary of ;dd. and for 250-300 volt supplies should be

asuitable transformer, interrupting it at regu- rated at 1500 to 2000 volts d.c. The proper

lar intervals to give the changing magnetic field required for inducing avoltage in the secondary (� 2-5), and rectifying the secondary out-

value is rather critical and should be determined experimentally, the optimum value being that which results in least battery cur-

put. The rectified output is pulsating d.c. which rent for agiven rectified d.c. output from the

may be filtered by ordinary means (� 8-5). Fig. 817 shows two types of circuits used,
both with vibrating-reed interrupters (vibrators). At A is shown the non-synchronous type of vibrator. When the battery circuit is open the reed is midway between the two contacts, touching neither. On closing the battery circuit
the magnet coil pulls the reed into contact with the lower point, causing current to flow

supply. Sparking at the vibrator contacts causes r.f.
interference ("hash") when such a supply is used with areceiver. This can be minimized by installing hash filters, consisting of RFC1 and C1 in the battery circuit, .and RFC2 with C3 in the d.c. output circuit. C1 is usually from 0.5 to 1etfd., a50-volt rating being adequate. RFC2 consists of about 50 turns wound to

through the lower half of the transformer primary winding. Simultaneously the magnet coil is short-circuited and the reed swings back,

about half-inch diameter, No. 12 or No. 14 wire being required to carry the rather heavy battery current without undue loss of voltage.

and is carried by inertia into contact with the C3 may be of the order of 0.01 to 0.1 pfd.,

upper point, causing current to flow through and RFC2 a2.5-millihenry choke of ordinary design. Equally as important as

the hash filter is thorough shield-

(A)

ing of the power supply and its connecting leads, since even a

small piece of wire or metal will

radiate enough hash to cause in-

terference in asensitive receiver.

Line-voltage adjustment --

In some localities the line voltage

may vary considerably from the nomi-

nal 115 volts as the load on the power

Vibratin9 reed

(B)

RFC,

RFC a

system changes. Since it is desirable to operate tube equipment, particularly filaments and heaters, at constant voltage for maximum life, a means of ad-

Smoothin9 Filler justing the line voltage to the rated

value is desirable. It can be accom-

plished by the circuit shown in Fig. 818, utiliz-

Fig 817 -- Vibrator power supply circuits. Constants ing a step-down transformer with a tapped

and operation are discussed in the text,

secondary connected as an auto-transformer

154 CHAPTER EIGHT

Poem- supply

(� 2-9). The secondary should preferably be tapped in steps of two or three volts, and should have sufficient total voltage to corn-
Fig. 818 -- Line-voltage compensation by means of tapped step-down transformer.

pensate for the widest variations encountered. Depending upon the end of the secondary
to which the line is connected, the voltage to the load can be made either higher or lower than the line voltage. A secondary winding capable of carrying five amperes or so will be
adequate for loads up to 500 volt-amperes on a 115-volt line.

155 CHAPTER EIGHT

CHAPTER NINE
Wave Propagation

�94 RADIO WAVES iNniiire of radio wares --Radio waves are
electromagnetic waves, consisting of traveling electrostatic and electromagnetic fields so related to each other that the energy is evenly divided between the two, and with the lines of force in the two fields at right angles to each other in a plane perpendicular to the direction of propagation as shown in Fig. 901. Except for the difference in order of wavelength, they have the same nature as light waves, travel with the same speed (300,000,000 meters per second in space), and, similarly to light, can be reflected, refracted and diffracted.
Polarization -- The polarization of a radio wave is taken as the direction of the lines of force in the electrostatic field. If the direction of the electrostatic component is perpendicular to the earth, the wave is said to be vertically polarized, while if the electrostatic component is parallel to the earth the wave is horizontally polarized. The electromagnetic component, being at right-angles to the electrostatic, therefore has its lines of force vertical when the wave is horizontally polarized, and horizontal when the wave is vertically polarized.
Reflection -- Radio waves may be reflected from any sharply-defined discontinuity, of suitable characteristics and dimensions, in the medium in which they are propagated. Any good conductor meets this requirement provided its dimensions are at least comparable
Electrostatic lines ofForce
Magnetic lines of Force
r////? / Fig. 901-- Representation of eectrostatic and electromagnetic lines of force in aradio wave. Arrows indicate instantaneous directions of the fields for awave traveling out of the page toward the reader. Reversing the direction of one get of lines would reverse the direction of travel.

with the wavelength. The surface of the earth also forms such adiscontinuity, and waves are readily reflected from the earth.
Refraction --Refraction of radio waves is similar to the refraction of light; that is, the wave is bent when moving obliquely into aregion having a different refractive index from that of the region it leaves. This bending results because the velocity of propagation differs in the two regions, so that the part of the wavefront which enters first travels faster or slower than the part which enters the new region last, causing the wavefront to turn.
Diffraction -- When awave grazes the edge of an object in passing it is bent around the object. This bending is called diffraction.
Ground and sky waves --Two types of waves occur, one traveling along the surface of the ground, the other traveling through the atmosphere and having no contact with the ground along most of its path. The former is called the ground wave, the latter the sky wave. The ground wave dies out rather rapidly, but the sky wave call travel to great distances, especially on high frequencies (short wavelengths).
Field strength -- The intensity of the electrostatic field of the wave is called the field strength at the point of measurement. It is usually expressed in microvolts per meter, and is equivalent to the voltage induced in a wire one meter long placed with its axis parallel to the direction of polarization.
09-2 THE GROUND WAVE
Description -- The ground wave is continuously in contact with the surface of the earth and, in cases where the distance of transmission makes the curvature of the earth important, is propagated by means of diffraction, with refraction in the lower atmosphere also having some effect. The ground wave is practically independent of seasonal and day and night effects at high frequencies (above 1500 kc.).
Polarization -- A ground wave must be vertically polarized because the electrostatic field of ahorizontally-polarized wave would be short-circuited by the ground, which acts as a
conductor at the frequencies for which the ground wave is of most interest.
Ground characteristics and losses -- The

156 CHAPTER NINE

Wave Propagation

wave induces acurrent in the ground in traveling along its surface. If the ground were aperfect conductor there would be no loss of energy,
but actual ground has appreciable resistance so that the current flow causes some energy dissipation. This loss must be supplied by the wave, which is correspondingly weakened. Hence the transmitting range depends upon the ground characteristics. Because sea water is a good conductor, the range will be greater over the ocean than over land. The losses increase with frequency, so that the ground wave is rapidly attenuated at high frequencies and above about 2 megacycles is of little importance except in purely local communication.
Range of ground wave -- At frequencies in the vicinity of 2megacycles the ground wave
range is of the order of 200 miles over average land and perhaps two or three times as far over sea water, for amedium-power transmitter (500 watts or so) using agood antenna. At higher frequencies the range drops off rapidly, and above 4megacycles the ground wave is useful only for work over quite short distances.

and the wavelength. Unusually high ionization may cause complete absorption of the wave energy, especially when the ionization is high in the lower regions of the ionosphere and below the lowest normally-useful layer. When the wave is absorbed in the ionosphere it is no more useful for communication than if it had passed through without sufficient bending to bring it back to earth.
In addition to refraction, reflection may take place at the lower boundary of alayer, if that boundary is well-defined; i.e., if there is an appreciable change in ionization within arelatively short interval of distance. For waves
approaching the layer at or near the perpendicular, the change in ionization must take place within adifference in height comparable to the wavelength, hence reflection is more apt to occur at longer wavelengths (lower frequencies).
Criticalfrequency -- When the frequency is low enough, awave sent vertically upward to the ionosphere will be bent sufficiently to return to the transmitting point. The highest frequency at which this occurs, for a given
state of the ionosphere, is called the critical

� 9-3 THE IONOSPHERE Description -- Since asky wave leaving the
transmitting antenna has to travel upward

frequency. It serves as an index for transmission conditions, although it is not the highest useful frequency since waves which enter the ionosphere at smaller angles than 90 degrees

with respect to the earth's surface, it would simply continue out into space if its path were not bent sufficiently to bring it back to the earth. The medium which causes such bending is the ionosphere, aregion in the upper atmosphere where free ions and electrons exist in sufficient quantity to cause achange in the refractive index. Ultraviolet radiation from the
sun is considered to be responsible for the ionization. The ionosphere is not a single region but consists of aseries of "layers" which occur at different heights, each layer consisting of a central region of ionization which tapers off in intensity both above and below.
Refraction, absorption, reflection--For a

(vertical) will be bent sufficiently to return to earth. The maximum usable frequency, for
waves leaving the earth at very small angles to the horizontal, is in the vicinity of three times the critical frequency.
Besides being directly observable, the critical
frequency is of more practical interest than ionization density because it includes the effects of absorption as well as refraction.
Virtual height -- Although alayer is a region of considerable depth, it is convenient to assign to it adefinite height, called the virtual height. The virtual height is the height from

4ciaal

�

given intensity of ionization the amount of re-

Neieee'444101(tfeete �

fraction becomes less as the frequency of the

wave becomes higher (shorter wavelength).

The bending is therefore smaller at high than at

low frequencies, and if the frequency is raised

to ahigh-enough value the bending eventually

will become too slight to bring the wave back

to earth, even when it enters the ionosphere at

avery small angle to the "edge" of the ionized

zone. At this and higher frequencies long-distance communication becomes impossible.

Fig. 902 -- Bending in the ionosphere and method of determining virtual height.

The greater the intensity of ionization the greater the bending on agiven frequency. Thus an increase in ionization increases the maximum frequency which can be bent sufficiently
for long-distance communication. The wave loses some energy in the ionosphere, and this energy loss increases with ionization density

which a pure reflection would give the same effect as the refraction which actually takes place. This is illustrated in Fig. 902. The wave
traveling upward is bent back over apath having appreciable radius of turning, and ameas-
urable interval of time is consumed in the turning process. The virtual height is then the

157 CHAPTER NINE

5heRadio -Amaieur'� -11tritcgooh

height of the triangle formed as shown, having equal sides of atotal length equivalent to the actual time taken for the wave to travel from T to R.
The E layer --The lowest normally-useful
layer is called the E layer. Its average height (maximum ionization) is about 70 miles. The ionization density is greatest around local noon, and the layer is only weakly ionized at night when the radiation from the sun is not
present. This is because the air at this height is sufficiently dense so that free ions and electrons very quickly meet and recombine.
The F, F1and fq layers-- The second principal layer is the F, which is at aheight of about 175 miles at night. In this region the air is so thin that recombination of ions and electrons takes place very slowly, since the particles can travel relatively great distances before meeting. The ionization decreases after sundown, reaching aminimum just before sunrise. In the daytime the F layer splits into two layers, the F1 and F2, at average virtual heights of about 140 miles for the F1and around 200 miles for the F2. These are most highly ionized at about local. noon, and merge again at sunset into the F
layer. Seasonal effects--In addition to day and
night variations, there are also seasonal changes in the ionosphere as the quantity of radiation
received from the sun changes. Thus the E layer has higher critical frequencies in the summer (about 4 Mc., average, in daytime) than in the winter, when the critical frequency is near 3 Mc. The F layer shows little variation, the critical frequency being of the order of 4to
5Mc. in the evening. The F1layer, which has a critical frequency in the neighborhood of 5
Mc. in summer, usually disappears in winter. The critical frequencies are highest in the F2 layer in winter (11 to 12 Mc.) and lowest in summer (around 7Mc.). The virtual height of
the Fi layer is also leas in winter (around 185 miles) than in summer (average 250 miles).
In the spring and fall a transition period occurs, and conditions in the ionosphere are more variable at these times of the year.
Sunspot cycles--The critical frequencies mentioned in the preceding paragraph are

mean values, since the ionization also varies with the 11-year sunspot cycle, being higher during times of greatest sunspot activity. Critical frequencies are highest during sunspot maxima and lowest during sunspot minima. The E critical frequency does not change greatly, but the .1? and F2 critical frequencies change in aratio of about 2to 1.
Magnetic storms and other disturbances -- Unusual disturbances in the earth's magnetic field (magnetic storms) usually are accompanied by disturbances in the ionosphere, when the layers apparently break up and expand. There is usually also an increase in ab� sorption during such aperiod. Radio transmission is poor and there is adrop in critical frequencies so that lower frequencies must be used for communication. A storm may last for several days.
Unusually high ionization in the region of the atmosphere below the normal ionosphere may increase absorption to such an extent that sky-wave transmission becomes impossible on high frequencies. The length of such adisturbance may be several hours, with agradual falling off of transmission conditions at the beginning and an equally gradual building up at the end of the period. Fadeouts, similar to the above in effect, are caused by sudden disturbances on the sun. They are characterized by very rapid ionization, with sky-wave transmission disappearing almost instantly, occur only in daylight, and do not last as long as the first type of absorption.
� 9-4 TIlE SKV WAVE
Wave angle (angle of radiation)-- The smaller the angle at which the wave leaves the earth, the smaller the bending required in the ionosphere to bring it back and, in general, the greater the distance between the point where it leaves the earth and that at which it returns (� 9-3). This is shown in Fig. 903. The vertical angle which the wave makes with a tangent to the earth is called the wave angle, or angle of radiation, the latter term being used more in connection with transmitting than receiving.
Skip distance--Since more bending is re-

355)

i'..? .ee

4 /

/4 ,

/el'% /.�

4/ I

, �

�

� %

� S%

� �.% le.

� ,

�

158 CHAPTER NINE

Fig. 903 --Refraction of sky waves, showing critical wave angle and skip zone

Wave Propagation

quired to return the wave to earth when the wave angle is high, it is found that at high frequencies the refraction frequently is not great
enough to give the required bending unless the wave angle is smaller than acertain angle called the critical angle. This is shown in Fig. 903, where wave angles A and lower give useful signals, but waves sent at higher angles travel through the layer and do not return. The distance between T and R1 is therefore the shortest possible distance over which sky-wave communication can be carried on. The area between the end of the useful ground wave and the beginning of sky-wave reception is called the skip zone. The skip distance depends upon the frequency and the state of the ionosphere and is greater the higher the transmitting frequency and the lower the critical frequency
(� 9-3). It also depends upon the height of the layer in which the refraction takes place, the higher layers giving longer distances for the
same wave angle. The wave angles at the transmitting and receiving points are usually, although not necessarily, approximately the same for agiven wave path.
It is readily possible for the sky wave to pass through the E layer and be refracted back to earth from the F, F1 or F2 layers. This is because the critical frequencies are higher
in the latter layers, so that asignal too high in frequency to be returned by the E layer can still come back from the F1,F2 or F, depending upon the time of day and the conditions existing. Depending upon the wave angle and fre-
quency, it is also possible to have communications via either the E or F1-F2 layers on the same frequency.
Multi-hop transmission-- On returning to earth the wave can be reflected (� 9-1) upward and travel again to the ionosphere where
refraction once more takes place, again with bending back to the earth. This process, which can be repeated several times, is necessary for
transmission over great distances because of the limited heights of the layers and the curvature of the earth, since at the lowest useful wave angles (of the order of a few degrees, waves at smaller angles generally being absorbed rapidly at high frequencies by being in
contact with the earth) the maximum one-hop distance is about 1250 miles with refraction

from the E layer, and around 2500 miles from the F2 layer. Ground losses absorb some of the energy from the wave on reflection, the amount of loss varying with the type of ground and being least for reflection from sea water. When the distance permits, it is better to have one hop rather than several, since the multiple reflections introduce losses which are higher than those caused by the ionosphere.
Multi-hop transmission is shown in Fig. 904, two- and three-hop paths being indicated.
Fading --Two or more parts of the wave may follow slightly different paths in traveling to the receiving point, in which case the difference in path lengths will cause a phase difference to exist between the wave components at the receiving antenna. The field strength may therefore have any value between the numerical sum of the components (when they are all in phase) and zero (when there are only two components and they are exactly out of phase). Since the paths change from time to time this causes a variation in signal strength called fading. Fading can also result from the combination of single-hop and multi-hop waves, or the combination of a ground wave and sky wave. The latter condition gives rise to an area of severe fading near the limiting distance of the ground wave, better reception being obtained at both shorter and longer distances where one component or the other is considerably stronger. Fading may be rapid or slow, the former type usually resulting from rapidly changing conditions in the ionosphere, the latter occurring when transmission conditions are relatively stable.
�9-5 ULTRA-HIGR-FREQUENCV PROPAGATION
Direct ray--In the ultra-high frequency part of the spectrum (above 30 megacycles) the bending of the waves in the normal ionosphere layers is so slight that the sky wave (� 9-4) does not ordinarily play any part in communication. The ground-wave (* 9-2) range also is extremely limited because of high absorption in the ground at these frequencies. Normal u.h.f. transmission is by means of a direct ray, or wave traveling directly from the transmitter to the receiver through the atmosphere. Since the energy lost in ground absorption by awave

Fig. 904--Multi-hop transmission.

159 CII iPTER NINE

DA, Radio AsnaleuA fiancliooi

traveling close to the ground decreases very rapidly with its height in wavelengths above ground, an ultra-high-frequency wave can be relatively close (in physical height) to the ground without suffering the absorption effects which would occur at the same physical heights on longer wavelengths.
Since the wave travels practically in a straight line, the maximum signal strength can be obtained only when there is an unobstructed atmospheric path between the transmitter and receiver. This means that the transmitting and receiving points should be sufficiently high to provide such a path, and on long paths the curvature of the earth must be taken into account as well as the intervening terrain.
Reflected ray--In addition to the direct ray, part of the wave strikes the ground between the transmitter and receiver and is reflected upward at aslight angle to produce areflectedray component at the receiver. This is shown in Fig. 905. The reflected ray is more or less out of phase with the direct ray, hence the net field strength at the receiving point is less than that of the direct ray alone. The canceling effect of the reflected ray depends upon the heights of the transmitter and receiver above the point of reflection, the ground losses when reflection takes place, and the frequency,
Fig. 905 -- Direct and reflected waves in u.h.f. transmission.
decreasing with an increase in any of these factors.
Atmospheric refraction -- There is normally some change in the refractive index of the air with height above ground, its nature being such as to cause the waves to bend slightly towards the ground. Where curvature of the earth must be considered, this has the effect of lengthening the distance over which it is possible to transmit a direct ray. It is convenient to consider the effect of this "normal" refraction as equivalent to an increase in the earth's radius in determining the transmitting and receiving heights necessary to provide aclear path for the wave. The equivalent radius, taking refraction into account, is 4/3 the actual radius. �
Range vs. height -- The height required to provide aclear path ("line of sight") over level ground from an elevated transmitting point to. areceiving point on the surface, not including the effect of refraction, is
h=1- .d521

20,000
1 . 1111111111111010613 2%

10,000
MIM�11��111.111�MMI.IMUMM��������11..10.�...ff����
,0 IMIIMIMMIMIIMIIIMMIMMIM11.1fflMMM�����r/1.1.1M
7 00 11MIMMIMERSignirniiMMIlifflifflgirnirgralM1
5,000

3,000

2000 I IMM11I1U11M1M 11i1l 11i11l1i11M1I11HMM MITlM la1il1d11111.�111. 1111

Ill

1,000
700

ir.z.n.emrregrcec:ihrrdwmearc.mr.,T...re.A.E:F:mLr=.=

SOO IMIIIIMIIIIII�11111111111M1/1111111.1111M1111

300 miliMiggiannglinsuguismiggiinunn
2 200

't 100

�

ammmeullairmemaCiiniiimi�i�revmAmrommimeafifl'f"flZuZm:mi:g:��== VARIMMIRIMMIIIIIMum

� SO

as
20

I1I1I1M1M1I1M.I1I1II1I1I1I1M.E1I�I1II3MM1M1I1I1M1I1I11I1I1I1I1M1I1Z1Z1�I1M1I1I1I1M1MMIM

10
7 mrartinnumm:ra::::gram

2 3 5 7 10

20 30 50 70 00 200

LINE-OF-SIGHT DISTANCE IN MILES

Fig. 906 -- Chart for determining line-of-sight dia. tance for u.h.f. transmission. Solid line includes effect of
refraction, dotted line is the optical distance.

where his the height in feet and dthe distance in miles. Conversely, the line of sight distance in miles for a given height in feet is equal to 1.23%/Ii. Taking refraction into account, the latter equation becomes 1.41Vh. The graph of Fig. 906 gives the answer directly when one quantity is known.
When transmitter and receiver are both elevated the maximum direct ray distance to
ground level as given by the formulas can be determined separately for each. Adding the two distances so obtained together will give the maximum distance by which they can be separated for direct-ray communication. This is shown in Fig. 907.
Diffraction -- At distances beyond the direct-ray path, the wave is diffracted around the curvature of the earth. The diffracted wave is attenuated very rapidly, so that beyond the maximum direct-ray distance the signal strength decreases considerably faster

Fig. 907 -- Method of determitt itt gtotal line-of-sight distance when both transmitter and receiver are ele Nated. Since only earth curvature is taken into account in Fig. 906, irregularities in the ground between the transmitting and receiving points must be considered for each actual path.

160

(: IL% �'TER NINE

Wave Propa yalion

with distance than it does within the directray path.

pirea.R.ay.

� 9-6 TROPOSPHERE REFRACTION

Temperature inversions -- The refractive

index of the lower atmosphere depends prin-

'NORMAL" AIR CONDITIONS

cipally upon the temperature, moisture con-

tent and pressure. Of the three, only tempera-

ture differences cause a large enough change in refractive index to refract ultra-high fre-

Direct Raq Refracted Downward

quency waves in such a way as to extend the distance range beyond the normal direct-ray

, ..

and diffracted-wave ranges discussed in the

preceding section. This occurs when there is a

"temperature inversion," or alayer of warm air over cooler air near the ground. Tempera-

TEMPERATURE INVERSION

ture inversions are relatively frequent in the

Fig. 908 -- Effect of a temperature inversion in ex-

summer, and usually occur at heights from a tending the range of u.h.f. signals.

few thousand feet to two miles or so above the ground.
Lower atmosphere bending -- When there is a sufficiently marked temperature version; i.e., a rapid rise of temperature with height, a wave is refracted back to earth in much the same way as in the ionosphere, although the cause of the change in refractive index is different. The amount of bending is small compared to the bending in the ionosphere. Consequently the wave angle (� 9-4) must be quite low (zero or nearly so), but since the bending takes place at alow altitude it is possible to extend the range of u.h.f. signals to several hundred miles when both transmitter and receiver are well below the line of sight.
Fig. 908 illustrates the conditions existing when the air is "normal" and when atemperature inversion is present. Since the bending is relatively small, it is advantageous to have as much height as possible at both the receiving and transmitting points, even though these heights may be considerably less than those necessary for "line of sight" transmission.
Frequency effects -- The amount of bending is greater at longer wavelengths (lower frequencies) but is not usually observed at frequencies much below 28 Me., partly because it
is masked by other effects. The upper limit of frequency at which useful bending ceases is not known, but transmission by this means is frequent on 56 and 112 Mc.

hence communication by means of it is restricted to transmitting and receiving localities so situated with respect to the cloud and to each other that a refracted wave path is possible.
The abnormal ionization usually disappears in the course of afew hours. Sporadic E ionization is more frequent in the summer than winter, and may occur at any time of the day or night.
Transmission characteristics -- Sporadic E refraction may take place at all frequencies up to the region of 60 megacycles. At the present time there are no known cases of such refraction on 112 Mc. When sporadic E ionization is present, skip distance is greatly reduced (when awavepath via the cloud is possible to agiven receiving location) on the frequencies where transmission is normally by means of the F, F1 and F2 layers; that is, from about 3.5 to 30 megacycles at night. The skip zone may in fact disappear entirely over most of the high-frequency spectrum, since the critical frequencies may rise to as high as 12 Mc. for sporadic E.
At ultra-high frequencies the bending is relatively small compared to lower frequencies, and only wave angles of the order of 5 degrees and less are useful in most cases. The transmitting and receiving points thus must be
sufficiently distant from the cloud to enable awave leaving the transmitter at such angles

� 1147 SPORADIC-E IONIZATION
Description -- Under certain conditions small regions or "patches" of unusually dense ionization may appear in the E layer of the ionosphere, for reasons not yet clearly understood. This is known as sporadic E ionization, and the change in refractive index in such a patch or cloud is frequently great enough to cause waves having frequencies as high as 60 Mc. to be bent back to earth. The dimensions of asporadic E cloud are relatively small,

to strike it, and the cloud should be approximately on, and near the center of, the line joining the transmitter and receiver. Unless the ionization is extremely high, the minimum distance of transmission on 56 Mc. is of the order of 800 miles and the maximum distance about 1250 miles.
Multi-hop transmission by means of two sporadic E clouds properly situated with respect to atransmitter and receiver is possible, but rather rare. Distances up to 2500 miles or so have been attained on 56 Mc. by this means.

161 CHAPTER NINE

CHAPTER TEN
Artlerma

lem�

� 10-1 ANTENNA PROPERTIES
Wave propagation and antenna design --
For most effective transmission, the propagation characteristics of the frequency under consideration must be given due consideration in selecting the type of antenna to use. These have been discussed in Chapter 9. On some frequencies the angle of radiation and polarization may be of relatively little importance; on others they may be all-important. On agiven frequency, the type of antenna best suited for long-distance transmission may not be as good as a different type for shorter-range work.
The important properties of an antenna or antenna system are its polarization, angle of radiation, impedance, and directivity.
Polarization -- The polarization of a straight-wire antenna is its position with respect to the earth. That is, a vertical wire transmits vertically polarized waves and a horizontal antenna generates horizontally polarized waves (� 9-1). The wave from an antenna in a slanting position contains both vertical and horizontal components.
Angle of radiation -- The wave angle (� 9-4) at which an antenna radiates best is
determined by its polarization, height above ground, and the nature of the ground. Radiation is not all at one well-defined angle, but rather is dispersed over a more or less large angular region, depending upon the type of
antenna. The angle is measured in a vertical plane with respect to a tangent to the earth at the transmitting point.
Impedance-- The impedance (� 2-8) of the antenna at any point is the ratio of voltage to current at that point. It is important in connection with feeding power to the antenna, since it constitutes the load resistance represented by the antenna. At high frequencies it consists chiefly of radiation resistance (� 2-12). It is understood to be measured at a current loop (� 2-12) unless otherwise specified.
Directivity -- All antennas radiate more power in certain directions than in others. This characteristic, called directivity, must be considered in three dimensions, since direc-
tivity exists in the vertical plane as well as in the horizontal plane. Thus the directivity of the antenna will affect the wave angle as well

as the actual compass directions in which maximum transmission takes place.
Current --The field strength produced by an antenna is proportional to the current flowing in it. Since standing waves are generally present on an antenna, the parts of the wire carrying the higher current therefore have the greatest radiating effect.
Power gain-- The ratio of power required to produce agiven field strength with a"comparison" antenna, to the power required to
produce the same field strength with aspecified type of antenna is called the power gain of the latter antenna. It is used in connection with antennas intentionally designed to have direc-
tivity, and is measured in the optimum direction of the antenna under test. The comparison antenna is almost always ahalf-wave antenna having the same polarization as the antenna under consideration. Power gain is usually expressed in decibels (� 3-3).

� 10-2 HALF-WAVE ANTENNA
Physical and electrical length-- The fundamental form of antenna is a single wire whose length is approximately equal to half the transmitting wave-length. It is the unit from which many more complex forms of antennas are constructed. It is sometimes known as aHertz or doublet antenna.
The length of ahalf wave in space is

length (feet)

492 Freq. (Mc.)

(1)

The actual length of a half-wave antenna will not be exactly equal to the half wavelength in space but is usually about 5% less, because of capacitance at the ends of the wire (end effect). The reduction factor increases slightly as the frequency is increased. Under average conditions, the following formula will give the length of ahalf-wave antenna to sufficient accuracy, for frequencies up to 30 Mc.

Length of half-wave antenna (feet) =

492 X 0.95 Freq. (Mc.)

468 Freq. (Mc.)

(2)

At 56 Mc. and higher frequencies the somewhat larger end effects cause aslightly greater reduction in length, so that for these frequen-
cies,

162 (II AP't'ER TEN

Anienna Sy�lems

Length of half-wave antenna (feet) =

492 X 0.94 _ 462

Freq. (Mc.) Freq. (Mc.)

(3)

or length

(inches)

5540 Freq. (Mc.)

(4)

Current and voltage distribution -- When power is fed to such an antenna the current and voltage vary along its length (� 2-12). The distribution, which is practically asine curve,
is shown in Fig. 1001. The current is maximum at the center and nearly zero at the ends, while the opposite is true of the r.f. voltage.

--rattle

Current

over a wider frequency range. The effect is greater as the diameter/length ratio is in-
creased, and is aproperty of some importance at ultra-high frequencies where the wavelength is small.
Radiation characteristics-- The radia-
tion from a half-wave antenna is not uniform in all directions but varies with the angle with respect to the axis of the wire. It is most intense in directions at right-angles to the wire, and zero along the direction of the wire itself, with intermediate values at intermediate
angles. This is shown by the sketch of Fig. 1002, which represents the radiation pattern in free space. The relative intensity of radiation is proportional to the length of aline drawn from the

center of the figure to the perimeter. if the an-

tenna is vertical, as shown in the figure, then

the field strength (� 9-1) will be uniform in all

horizontal directions; if the antenna is hori-

Fig. 1001 -- Current and voltage distribution on a half-wave antenna.
The current does not actually reach zero at the current nodes, (� 2-12) because of the end

zontal, the relative field strength will depend upon the direction of the receiving point with respect to the direction of the antenna wire. e 10-3 GROUND EFFECTS

effect; similarly, the voltage is not zero at its node because of the resistance of the antenna, which consists of both the r.f. resistance of the wire (ohmic resistance) and the radiation resistance (� 2-12). Usually the ohmic resistance of a half-wave antenna is small enough, in comparison with the radiation resistance, to be neglected for all practical purposes.
Impedance -- The radiation resistance of a half-wave antenna in free space -- that is, sufficiently removed from surrounding objects so that they do not affect the antenna's charac-

Reflection -- When the antenna is near the ground the free-space pattern of Fig. 1002 is modified by reflection of radiated waves
from the ground, so that the actual pattern is the resultant of the free-space pattern and ground reflections. This resultant is dependent upon the height of the antenna, its position or orientation with respect to the surface of the ground, and the electrical characteristics of the ground. The reflected waves may be in such phase relationship to the directly-radiated

teristics -- is 73 ohms, approximately. The

value under practical conditions will vary with

the height of the antenna, but is commonly

taken to be in the neighborhood of 70 ohms. It is pure resistance, and is measured at the center of the antenna. The impedance is minimum at

the center, where it is equal to the radiation resistance, and increases toward the ends (� 10-1). The end value will depend on anum-
ber of factors such as the height, physical construction, and the position with respect to ground.
Conductor size --The impedance of the antenna also depends upon the diameter of the conductor in relation to its length. The figures above are for wires of practicable sizes. If the diameter of the conductor is made large, of the order of 1% or more of the length, the impedance at the center will be raised and the impedance at the ends decreased. This increase in
center impedance (of the order of 50% for a diameter/length ratio of 0.025) is accompanied by a decrease in the Q (� 2-10, 2-12) of the antenna, so that the resonance curve is less sharp. Hence the antenna is capable of working

Fig. 1002 -- Free-space radiation pattern of half. wave antenna. The antenna is shown in the vertical position. This is across-section of the solid pattern described by the figure when rotated on its axis (the antenna). The "doughnut" form of the solid pattern can easily be visualized by imagining the drawing glued to cardboard, with a short length of wire fastened on to represent the antenna. Then twirling the wire will give a visual representation of the solid pattern.
waves that the two completely reinforce each other, or the phase relationship may be such that complete cancellation takes place. All intermediate values also are possible. Thus the effect of a perfectly-reflecting ground is such that the original free-space field strength may be multiplied by afactor which has amaximum value of 2, for complete reinforcement, and having all intermediate values to zero, for complete cancellation. Since waves are always reflected upward from the ground (assuming

163 CHAPTER TEN

5hePalio ...Amateur's ilanitooh

that the surface is fairly level) these reflections only affect the radiation pattern in the vertical
plane -- that is, in directions upward from the
earth's surface -- and not in the horizontal plane, or the usual geographical directions.
Fig. 1003 shows how the multiplying factor varies with the vertical angle for several representative heights for horizontal antennas.
As the height is increased the angle at which complete reinforcement takes place is lowered until it occurs at avertical angle of 15 degrees for a height equal to one wavelength. At still greater heights not shown on the chart the first maximum will occur at still smaller angles.

the higher radiation angles are effective so that again a reasonable antenna height is not difficult of attainment. Heights between 35 and 70 feet are suitable for all bands, the higher figures generally being preferable if
circumstances permit their use. Imperfect ground -- Fig. 1003 is based on
aground. having perfect conductivity, which is not met with in practice. The principal effect of
actual ground is to make the curves inaccurate at the lowest angles; appreciable high-frequency radiation at angles smaller than afew degrees is practically impossible to obtain at heights of less than several wavelengths. Above 15 degrees, however, the curves are accurate enough

2.0

1.8

e'L'ef\''
7/

.,. /

4.-.-. H= 3/4 .1

1.6 CC C-3 l. 4
47 12

ifi .1 I

il
.if l

'
,
eri

1
t
I i

h
. \,...,,

for all practical purposes, and may be taken as indicative of the sort of result to be expected at angles between 5and 15 degrees.
The effective ground plane -- that is, the
plane from which ground reflections can be considered to take place -- seldom is the actual surface of the ground but is afew feet below it, depending upon the character of the soil.

t.D 10 ill

i li

Impedance -- Waves which are reflected directly upward from the ground induce a

08 .iI
k 0.6

i
i11 II

\
\ \

k 0.4
02
o

N= "=4-A. 'ii1I1\itI e Vy 1

\ \

current in the antenna in passing and, depending on the antenna height, the phase relationship of this induced current to the original current may be such as either to increase or decrease the total current in the antenna. For the same power input to the antenna, an increase in current is equivalent to adecrease in

10 � 20 � 30 � 40 � 50� 60 � 70� 80� 90� VERT/CAL ANGLE

impedance, and vice versa. Hence the impedance of the antenna varies with height.

Fig. 1003 -- Effect of ground on radiation at vertical The theoretical curve of variation of radiation

angles for four antenna heights. This chart applies only to horizontal antennas, and is based on perfectly conducting ground.

resistance for an antenna above perfectlyreflecting ground is shown in Fig. 1004. The impedance approaches the free-space value as

When the half-wave antenna is vertical the maximum and minimum points in the curves of Fig. 1003 exchange positions, so that the
nulls become maxima, and vice versa. In this case, the height is taken as the distance from

the height becomes large, but at low heights may differ considerably from it.
Choice of polarization-- Polarization of the transmitting antenna is generally unimportant on frequencies between 3.5 and 30 Mc.

ground to the center of the antenna. Radiation angle -- The vertical angle, or

100

angle of radiation, is of primary importance,

especially at the higher frequencies (� 9-4, 9-5).

It is therefore advantageous to erect the an-

tenna at aheight which will take advantage of

, 70

ground reflection in such away as to reinforce

the space radiation at the most desirable angle.

$:.! 50

Since low radiation angles usually are desirable, this generally means that the antenna should

tt 40

be high; at least AI wavelength at 14 Mc. and

030

preferably 3 % or 1wavelength; at least 1wave-
length and preferably higher at 28 Mc. and the ultra-high frequencies. The physical height decreases as the frequency is increased so that good heights are not impracticable; a half

``. 14 10
o

1 /4 '/ 3/4 1.0 1/4 1V2 13'4 20 HEIGHT ABOVE GROUND

wavelength at 14 Mc. is only 35 feet, approximately, and the same height represents afull

Fig. 1004 -- Radiation resistance of ahalf-wave horizontal antenna as afunction of height above perfectly-

wavelength at 28 Mc. At 7 Mc. and lower, reflecting ground.

164 CHAPTER TEN

Anienna

denu

However, the question of whether the antenna should be installed in a horizontal or vertical
position deserves consideration on other counts. A vertical half-wave antenna will radiate equally well in all horizontal directions, so that it is substantially non-directional in the

despite the fact that this line runs in the same geographical direction as OA. At some higher angle OC the radiation, still in the same geographical direction, is still more intense. The effective radiation pattern therefore depends

usual sense of the word. If installed horizon-

tally, however, the antenna will tend to show directional effects, and will radiate best in the

direction at right-angles, or broadside, to the

wire. The radiation in such acase will be least

in the direction toward which the wire points.

This can be seen readily by imagining that Fig. 1002 is lying on the ground and that the pattern is looked at from above.

The vertical angle of radiation also will be affected by the position of the antenna. If it

were not for ground losses at high frequencies, the vertical half-wave antenna would be pre-

ferred because it would concentrate the radia-

tion horizontally. Practically, this theoretical advantage over the horizontal antenna is of little or no consequence.

At 1.75 Mc. vertical polarization will give more low-angle radiation, and hence is better

for long-distance transmission; at this fre-

quency the ground wave also is useful and must
be vertically polarized. On ultra-high frequencies, direct-ray and lower troposphere trans-
mission require the same type of polarization at both receiver and transmitter, since the waves suffer no appreciable change in polarization in transmission (� 9-5, 9-6). Either vertical or horizontal polarization may be used, the latter being slightly better for longer distances.
Effective radiation patterns -- In deter-

Fig. 1006 -- Horizontal pattern of ahorizontal half. wave antenna at three vertical radiation angles. Solid line is relative radiation at 15 degrees. Dotted lines show deviation from the 15-degree pattern, for angles of 9 and 30 degrees. The patterns are useful for shape only, since the amplitude will depend upon the height of the antenna above ground and the vertical angle considered. The patterns for all three angles have been proportioned to the same scale, but this does not mean that the maximum amplitudes necessarily are the same. The arrow indicates the direction of the antenna wire.

mining the radiation pattern it is necessary to consider radiation in both the horizontal and

upon the angle of radiation most useful and for long-distance transmission is dependent upon the conditions existing in the ionosphere.

These conditions may vary not only from day

Fig. 1005 -- Illustrating the

importance of vertical angle of

radiation in determining antenna

directional effects. Ground re-

flection is neglected in this

drawing.

�

to day and hour to hour, but even from minute to minute. Obviously, then, the effective directivity of the antenna will change along with transmission conditions.
At ultra-high frequencies, where only ex-

tremely low angles are useful for any but

vertical planes. When the half-wave antenna is vertical, the vertical angle of radiation chosen does not affect the shape of the horizontal pattern, but only its relative amplitude. When the antenna is horizontal, however, both the shape and amplitude are dependent upon the angle of radiation chosen.
Fig. 1005 illustrates this point. The "freespace" pattern of the horizontal antenna shown is a section cut vertically through the solid pattern. In the direction OA, horizontally along the wire axis, the radiation is zero. At some vertical angle represented by the .line OB, however, the radiation is appreciable,

sporadic-E transmission (� 9-7) the effective radiation pattern of the antenna approaches the free-space pattern. A horizontal antenna
therefore shows more marked directive effects than it does at lower frequencies, on which high radiation angles are effective.
Theoretical horizontal-directivity patterns for half-wave horizontal antennas at vertical angles of 9, 15, and 30 degrees (representing average useful angles at 28, 14 and 7 Mc. respectively) are given in Fig. 1006. At inter-
mediate angles the values in the affected regions also will be intermediate. Relative field strengths are plotted on adecibel scale (� 3-3) so that they represent as nearly as possible the actual aural effect at the receiving station.

165 CHAPTER TEN

54, Radio Amcdeur '� -11anclioo

�10-4 APPLYING POWER TO THE ANTENNA
Direct excitation-- When power is transferred directly from the source to the radiating antenna, the antenna is said to be directly excited. While most of the coupling methods (� 2-11) may be used, the more common ones are shown in Fig. 1007. Power is usually fed to the antenna at either a current or voltage loop (� 10-2). If at acurrent loop, the coupling is called current feed; if at avoltage loop, it is
called voltage feed. Current feed--This is shown in Fig.
1007-A. The antenna is cut at the center and a small coil coupled to the output tank circuit of the transmitter, with adjustable coupling

Disadvantages of direct excitation--Direct excitation is seldom used except on the lowest amateur frequencies because it involves bringing the antenna proper into the operating room and hence into close relationship with the house and electric wiring. This usually means that some of the power is wasted in heating poor conductors in the field of the antenna. Also, it usually means that the shape of the antenna must be distorted so that the expected directional effects are not realized, and likewise means that the height is limited. For these reasons, in high-frequency work practically all amateurs use transmission lines or feeder systems which permit putting the an-
tenna in adesirable location.

so that the transmitter loading can be controlled. Since the addition of the coil "loads"
the antenna, or increases its effective length because of the additional inductance, the series condensers CI and C2 are used to provide electrical means for reducing the length to its original unloaded value; in other words, to cancel the effect of the inductive reactance
(� 2-10). Voltage feed--In Fig. 1007 at B and
the power is introduced into the antenna at a point of high voltage. In B the end of the antenna is coupled to the output tank circuit
through a small condenser; in C a separate tank, connected directly to the antenna, is used. This tank is tuned to the transmitter frequency and should be grounded at one end or at the center of the coil, as shown.
Adjustment of coupling--Methods of tuning and adjustment correspond to those used with transmission lines and are discussed
in �10-6.

�10-5 TRANSMISSION LINES
Requirements-- A transmission line is used to transfer power, with a minimum of loss, from its source to the device in which the power is to be usefully expended. At radio frequencies, where every wire carrying r.f. current tends to radiate energy in the form of electromagnetic waves, special design is necessary to minimize radiation and thus cause as much as possible of the power to -be delivered to the
receiving end of the line. Radiation can be minimized by using aline
in which the current is low, and by using two conductors carrying currents of equal magnitudes but opposite phase so that the fields about the conductors cancel each other. For good cancellation of radiation the two conductors should be parallel and quite close to each
other. Types-- The most common form of trans-
mission line consists of two parallel wires, maintained at a fixed spacing of two to six

inches by insulating spacers or spreaders at

C, (--,Q0L9 Q,41 c2
xmi-R TANK

suitable intervals (open-wire line). A second type consists of rubber-insulated wires twisted together to form aflexible line without spacers (twisted-pair line). A third uses a wire inside

(A)

and coaxial with a tubing outer conductor,

separated from the outer conductor by insu-

XMTR TANK c

lating spacers or "beads" at regular intervals (coaxial or concentric line). A variation of this

Short connecbon

type uses solid rubber insulation between the

(B)

inner and outer conductors, the latter usually

being made of metal braid rather than solid

KIKTR

x/a

TANK

tubing so that the line will be flexible. Still another type of line uses a single wire alone, without asecond conductor (single-wire feeder);

in this case radiation is minimized by keeping

the line current low. Spacing of two-wire lines--The spacing

between wires of an open-wire line should be

Fig. 1007 -- Methods of direct feed to the half-wave antenna. A, current feed, series tuning; B, voltage feed, capacity coupling; C, voltage feed with inductively coupled antenna tank. In A, the coupling apparatus is not included in the antenna length.

small in comparison to the operating wavelength to prevent appreciable radiation. At the same time it is impracticable to make the spacing too small because when the wires

166 CHAPTER TEN

Antenna SyrItem�

swing with respect to each other in awind the line constants (� 2-12) will vary and thus cause a variation in tuning or loading on the transmitter. It is also desirable to use as few insulating spacers as possible to keep the weight of the line to a minimum. In practice aspacing of about six inches is used for 14 Mc. and lower frequencies, with four and two inch spacing being common on the ultra-high frequencies.
Balance to ground -- For maximum cancellation of the fields about the two wires it is necessary that the currents be equal in amplitude and opposite in phase. Should the capacity or inductance per unit length in one wire differ from that in the other this condition cannot be fulfilled. Insofar as the line itself is concerned, the two wires will have identical characteristics only when the two have exactly the same physical relationships to ground and
to other objects in the vicinity. Thus the line should be symmetrically constructed and the two wires should be at the same height. Line unbalance can be minimized by keeping the line as far above ground and as far from other objects as possible.
To overcome unbalance the line is sometimes transposed, which means that the positions of the wires are interchanged at regular intervals
(Fig. 1008). This is more helpful on long than

Fig. 1008 -- Method of transposing atwo-wire open transmission line to preserve balance to ground and nearby objects.

on short lines and usually need not be re-

sorted to for lines less than a wavelength or two long.

Characteristic impedance -- The square

root of the ratio of inductance to capacity

per unit length of the line is called the characteristic or surge impedance. It is the impedance

which along line would present to an electrical

impulse induced in the line, and is important

in determining the operation of the line in

conjunction with the apparatus to which it is connected.

The characteristic impedance of air-insulated transmission lines may be calculated from

the following formulas:

Parallel-conductor line

Z = 276 log -b a

(5)

where Z is the surge impedance, bthe spacing, center to center, and a the radius of the conductor. The quantities band amust be meas-

ured in the same units (inches, etc.). Surge impedance as a function of spacing for lines

using conductors of different size is plotted in chart form in Fig. 1009. Coaxial or concentric line

Z = 138 log -ab

(6)

where Z again is the surge impedance. In this case b is the inside diameter (not radius) of the outer conductor and ais the outside diameter of the inner conductor. The formula is true for air dielectric, and approximately so for a line having ceramic insulators so spaced that the major proportion of the insulation is air.
When asolid insulating material is used between the conductors the impedance decreases, because of the increase in line capacity, by the
factor 1R/Tc, where kis the dielectric constant of the insulating material.
The impedance of asingle-wire transmission line varies with the size of the conductor, its

LO 800

SPACING, INCHES 2 3 4 5678910

700

u) 600 k 0 V
%.1
tkx SOO 0

..... u 400

i>.

Lk'0u. c!1le,)

300

! k

%., 200

}WORE SIZE

A q'

TUBING DIR.

ioo1o2

3 4 5 6 7 8910

CENTER-TO-CENTER SPACING (INCHES)

Fig. 1009 -- Characteristic impedances of typical spaced-conductor transmission lines. Use outside diameter of tubing.

height above ground, and orientation with respect to ground. An average figure is about 500 ohms.
Electrical length -- The electrical length of aline is not exactly the same as its physical length for reasons corresponding to the end effects in antennas. (� 10-2). Spacers used to
separate the conductors have dielectric constants larger than that of air, so that the waves
do not travel quite as fast along aline as they would in air. The lengths of electrical quarter waves of various types of lines can be calculated from the formula

167 CHAPTER TEN

-7he /each� Amateur's -Mac/Loh

Length

(feet)

246 Freq.

X V (Mc.)

(7)

where V depends upon the type of line. For lines of ordinary construction, V is as follows:

sistance of a value equal to the characteristic impedance of the line.
Reactance, resistance impedance--The input end of aline may show reactance as well as resistance, and the values of these quantities

Parallel wire line

V = 0.975

will depend upon the nature of the load at the

Parallel tubing line

V = 0.95

output end, the electrical length of the line,

Concentric line (air-insulated) V = 0.85

and the line characteristic impedance. The

Concentric line (rubber-insu-

reactance and resistance are important in

lated)

V = 0.56-0.65 determining the method of coupling to the

Twisted pair

source of power. Assuming that the load at the

Input and output ends-- The input end of a line is that connected to the source of power; the output end is that connected to the
power-absorbing device. When aline connects a transmitter to an antenna, the input end is at the transmitter and the output end at the antenna; with the same line and antenna connected to areceiver, however, the energy flows from the antenna to the receiver, hence the input end of the line is at the antenna and the
output end at the receiver. Standing-wave ratio-- The lengths of
transmission lines used at radio frequencies are of the same order as the operating wave
lengths and therefore standing waves of current and voltage may appear on the line (� 2-12). The ratio of current (or voltage) at a loop to the value at a node (standing-wave ratio) depends upon the ratio of the resistance of the load connected to the output end of the line, or termination, to the characteristic impedance of the line itself. That is,

Z. Z.

Standing-wave ratio = -- or --

(8)

output end of the line is purely resistive, which is essentially the case since the load circuit is usually tuned to resonance, aline less than a
quarter wavelength long electrically will show inductive reactance at its input terminals when the output termination is less than the charac-
teristic impedance, and capacitive reactance when the termination is higher than the
characteristic impedance. If the line is more than aquarter wave but less than a halfwave long the reverse conditions exist. With still longer lengths the reactance characteristics
reverse in each succeeding quarter wavelength. The input impedance is purely resistive if the
line is an exact multiple of a quarter wave in length. The reactance at intermediate lengths is higher the greater the standing wave ratio, being zero for aratio of 1.
Impedance transformation-- Regardless of the standing-wave ratio, the input impedance of aline ahalf-wave long electrically will be equal to the impedance connected at its output end. Such a line (the same thing is true of a line any integral multiple of a halfwave in length) can be considered to be aoneto-one transformer. However, if the line is a

where Z. is the characteristic impedance of the line and Ztis the terminating resistance, Zt is

quarter-wave (or an odd multiple of aquarter wave) long the input impedance will be equal to

generally called an impedance, although it must be non-reactive and therefore corre-
spond to apure resistance for the line to operate as described. This means that the load or termination, when an antenna, must be reso-

Z82
z; =-
Z.
where Z. is the characteristic impedance of the

nant at the operating frequency.

line and Ztthe impedance connected to the out-

The formula is given in two ways because it put end. A quarter-wave line therefore can be

is customary to put the larger number in the used as an impedance transformer, and by

numerator so that the ratio will not be frac- suitable selection of constants a wide range tional. As an example, a 600-ohm line termi- of input impedance values can be obtained.

nated in a resistance of 70 ohms will have a Furthermore, the impedance measured be-

standing wave ratio of 600/70, or 8.57. The ratio on a 70-ohm line terminated in a resistance of 600 ohms would be the same. This means that if the current as measured at a node is 0.1 amp., the current at aloop will be
0.857 amp. A line terminated in aresistance equal to its

tween the two conductors anywhere along the line will vary between the two end values, so that any intermediate impedance value can be selected. This is aparticularly useful property since a quarter-wave line may be shortcircuited at one end (� 2-12) and used as a
linear transformer with adjustable impedance

characteristic impedance is equivalent to an ratio.

infinitely long line, consequently there is no reflection and no standing waves appear. The standing wave ratio therefore is 1. The

Losses-- Air-insulated lines operate at quite high efficiency. Parallel-conductor lines average 0.12 to 0.15 db. (� 3-3) loss per wave-

input end of such aline appears as apure re- length of line. These figures hold only if the

168 CHAPTER TEN

Anienna Sefem�

standing wave ratio is 1. The losses increase with the standing-wave ratio, rather slowly up to aratio of 15 to 1, but rapidly thereafter. For standing-wave ratios of 10 or 15 to 1the increase is inconsequential provided the line is well balanced.

proper value (impedance matching) and for tuning out any reactive component that may be present (2-9, 2-10, 2-11). The resistance and reactance considered are those present at the input end of the line, and hence have nothing to do with the antenna itself except insofar as

Concentric lines with air insulation are excellent when dry, but losses increase if there is moisture in the line. Provision therefore
should be made for making such lines airtight, and they should be thoroughly dry when assembled. This type of line has the least radiation loss. The small lines (%-inch outer conductor) should not be used at high voltages, hence it is desirable to keep the standingwave ratio down.
Good quality rubber insulated lines, both

the antenna load may affect the operation of the line (� 10-5).
Untuned coil -- A simple system, shown in Fig. 1010-A, uses acoil of afew turns tightly coupled to the plate tank coil. Since no provision is made for tuning, this system is suitable only for non-resonant lines which show practically no ceactance at the input end. Loading on the transmitter may be varied by varying the coupling between the tank inductance and pickup coil, as it is frequently called, or by

twisted pair and coaxial, average about 1db. loss per wavelength of line. At the higher frequencies, therefore, such lines should be used only in short lengths if losses are important. These lines have the advantages of compactness, ease of installation, and flexibility. Ordinary lampcord has aloss of approximately 1.4 db. per wavelength, when dry, but its losses become excessive when wet. The parallel
m�ulded-rubber type is best from the standpoint of withstanding wet weather. The characteristic impedance of lampcord is between 120 and 140 ohms.
The loss in db. is directly proportional to the length of the line. Thus aline which has aloss of 1db. per wavelength will have an actual loss of 3db. if the line is three wavelengths long. In

changing the number of turns on the pickup coil. A slight amount of reactance is coupled into the tank circuit by the pickup coil, since the flux leakage (� 2-11) is high, so that slight retuning of the plate tank condenser may be necessary when the load is connected.
Taps on tank circuit -- A method suitable for use with open-wire lines is shown in Fig. 1010-B, where the line is tapped on abalanced tank circuit with taps equidistant from the center or ground point. This symmetry is necessary to maintain line balance to ground (� 10-5). Loading is increased by moving the taps outward from the center. Any reactance present may be tuned out by readjustment of the plate tank condenser, but this method is not suitable for large values of reactance and

the case of line losses, the length is not ex- therefore direct tapping is best confined to use pressed in terms of electrical length but in with non-resonant lines.

physical length; that is, a wavelength of line, in feet, is equal to 984/Freq. (Mc.) for computing loss. This permits a direct comparison

Adjustment of untuned systems -- Adjustment of either of the above systems is quite simple. Starting with loose coupling, apply

of lines having the same physical length. The electrical lengths, of- course, may differ considerably.
Resonant and non-resonant lines -- Lines

power to the transmitter, and adjust the plate tank condenser to minimum plate current. If the current is less than the desired load value, increase the coupling and again resonate

are classified as resonant or non-resonant de- the plate condenser. Continue until the desired pending upon the standing-wave ratio. If the plate current is obtained, always keeping the

ratio is near 1, the line is said to be non- plate tank condenser at the setting which gives resonant. Reactive effects will be small and minimum current.

consequently no special tuning provisions need be made for cancelling them (� 2-10) even when the line length is not an exact multiple of aquarter wavelength. If the standing-wave ratio is fairly large, the input reactance must be cancelled or "tuned out" unless the line is

Pi-section coupling -- A coupling system which is electrically equivalent to tapping on the tank circuit, but using a capacity voltage divider in the plate tank circuit for the purpose, is shown in Fig. 1010-C. Since one side of the condenser across which the line is con-

a multiple of a quarter wavelength, and the nected is grounded, some unbalance will be

line is said to be resonant.

introduced into the transmission line. The

�10-6 COUPLING TO TRANSMISSION LINES

method is used chiefly with low-power portable sets because it is readily adjustable to meet a fairly wide range of impedance values. A

Requirements -- The coupling system be- -single-ended amplifier, either screen-grid or a tween a transmitter and the input end of a grid-neutralized triode (� 4-7) is required,

transmission line must provide means for since the plate tank circuit is not balanced.

adjusting the load on the transmitter to the Coupling is adjusted by varying C1, re-resonat-

169 CHAPTER TEN

Dfle

Amaleur'd --flancidooh

ing the circuit each time by means of C2, until the desired amplifier plate current is obtained. In general, the coupling will increase as C1 is made smaller with respect to C2. Relatively large-capacity condensers are required to give a suitable impedance-matching range while
maintaining resonance. Pi-section filter -- The coupling circuit
shown in Fig. 1010-D is alow-pass filter capable of coupling between afairly wide range of impedances. The method of adjustment is as follows: First, with the filter disconnected from the transmitter tank, tune the transmitter tank to resonance, as evidenced by minimum plate current. Then, with trial settings of the clips

on L1 and L2 (few turns for high frequencies, more for lower) tap the input clips on the final tank coil at points equidistant from the center
so that about half the coil is included between them. A balanced tank circuit must be used.
Set C2 at about half scale, apply power, and rapidly rotate CI until the plate current drops to minimum. If this minimum is not the desired full-load plate current, try anew setting
of C2 and repeat. If, for all settings of C2, the plate current is too high or too low, try new settings of the taps on L1 and L2, and also on the transmitter tank. Do not touch the tank condenser during these adjustments. When, finally, the desired plate current is obtained,

Tight
Cduphrly
Trans Line (A)
I-0 Trans Line

(H)

RFC

Trans [me
�

Trans
tme

f---0

Trans line

c 2

(E)

RFC

Fig. 1010 -- Methods of coupling the transmitter to the transmission line. Circuit values and adjustment are discussed in the text. Condensers marked "C" are fixed blocking condensers to isolate the transmitter plate voltage from the antenna; their capacity is not critical, 500 pad. to 0.002 jifd. being satisfactory values. Voltage rating should at least equal the plate voltage on the final stage.
170 CHAPTER TEN

-Antenna .-S7ydienu

set C1 carefully to the exact minimum plate- for by tuning of C1 and adjustment, if neces-

current point. This adjustment is important in sary, of L1 by means of taps. Parallel tuning

minimizing harmonic output.

is suited to resonant lines when avoltage loop

With some lengths of resonant lines, particu- appears at the input end.

larly those not exact multiples of a quarter

The tuning procedure is quite similar to

wavelength, it may be difficult to get proper that with series tuning. Find the value of

loading with the pi-section coupler. Usually, coupling between L1and L2 which will bring

these lengths also will be difficult to feed with the plate current to the desired value as CI is

other systems of coupling. In such cases, the tuned through resonance. Again aslight read-

proper loading often can be obtained by vary- justment of the amplifier tank condenser may

ing the L/C ratio of the filter over aconsider- be necessary to compensate for the effect of

ably wider range than is used for normal loads. coupled reactance.

Series tuning -- When the input impedance

Link coupling -- Where tuning of the cir-

of the line is low, the coupling method shown cuit connected to the line is necessary or

in Fig. 1010-E may be used. This system, desirable, it is possible to separate physically

known as series tuning, places the coupling the line-tuning apparatus and the plate tank

coil, tuning condensers, and load all in series, circuit by means of link coupling (� 2-11).

and is particularly suitable for use with reso- This is often convenient from aconstructional

nant lines when a current loop appears at the standpoint, and has the advantage that with

input end. Two tuning condensers, as shown, proper construction there will be somewhat

usually are used to keep the line balanced to less harmonic transfer to the antenna since

ground, but one will suffice, the other end of the stray capacity coupling is lessened with the

line being connected directly to the end of LI. smaller link coils.

The tuning procedure with series tuning is as

Figs. 1010-G and H show a method which

follows: With C1 and C2 at minimum capacity, can be considered to be a variation of Fig.

couple the antenna coil L1loosely to the trans- 1010-B. The first (G) is suitable for use with a

mitter output tank coil and observe the plate single-ended plate tank, the second (H) for a

current. Then increase C1 and C2simultane- balanced tank. The auxiliary tank on which

ously, until a setting is reached which gives the transmission line is tapped may have ad-

maximum plate current, indicating that the justable inductance as well as capacity to pro-

antenna system is in resonance with the trans- vide a wide range of reactance variation for

mitting frequency. Readjust the plate tank compensating for line reactance. The center of

condenser to minimum plate current. This is the auxiliary tank inductance may be grounded

necessary because tuning the antenna circuit if desired. The link windings should be placed

will have some effect on the tuning of the plate at the grounded parts of the coils to reduce

tank. The new minimum plate current will be capacity coupling and consequent harmonic

higher than with the antenna system detuned, transfer. With this inductively-coupled system but should still be well below the rated value the loading on the auxiliary tank circuit in-

for the tube or tubes. Increase the coupling be- creases as the taps are moved outward from the

tween L1and L2 by asmall amount, readjust center, but since this decreases the Q of the

C1 and C2for maximum plate current, and circuit the coupling to the plate tank simul-

again set the plate tank condenser to mini- taneously decreases (� 2-11), hence acompro-

mum. Continue this process until the mini- mise adjustment giving proper loading must

mum plate current is equal to the rated plate be found in practice. Loading also may be

current for the amplifier. Always use the de- varied by changing the coupling between one

gree of coupling between L1 and L2 which will link winding and its associated tank coil;

just bring the amplifier plate current to rated either tank may be used for this purpose. When

value when C1 and C2 pass through resonance. The r.f. ammeters should indicate maximum feeder current at the resonance setting; these

the auxiliary tank is properly tuned to compensate for line reactance the plate tank tuning will be practically the same as with no load on

meters are not strictly necessary, but are useful the circuit, hence the plate tank condenser

in indicating the relative power output from need only be readjusted slightly to compensate

the transmitter.

for the small reactance introduced by the link.

Parallel tuning -- When the line has high

Link coupling also may be used with series

input impedance, parallel tuning, as shown in and parallel tuning, as shown in Figs. 10104

Fig. 1010-F, is required. Here the coupling and J. The coupling between one link and its

coil, tuning condenser and line are all in paral- associated coil may be made variable to give

lel, the load represented by the line being di- the same effect as changing the coupling be-

rectly across the tuned coupling circuit. If the tween the plate tank and antenna coils in the

line is non-reactive, the coupling circuit will ordinary system. The tuning procedure is the

be tuned independently to the transmitter same as described above for series and parallel

frequency; line reactance can be compensated tuning. In the ease of single-ended tank cir-

171 CHAPTER TEN

..7he Pail� Amaieur'� --llandhooh

cuits, the input link would be coupled to the grounded end of the tank coil, similarly to the method in Fig. 1010-G.
Circuit values -- The values of inductance and capacity to use in the antenna coupling system will depend upon the transmitting frequency, but are not particularly critical. With
series tuning (Fig. 1010-E, I) the coil may consist of afew turns of the same construction as is used in the final tank; average values will run from one or two turns at ultra-high frequencies to perhaps 10 or 12 at 1.75 Mc. The number of turns preferably should be adjustable so that the inductance can be changed should it not be possible to reach resonance with the condensers used. The series condensers should have amaximum capacity of 250 or 350 pad. at the lower frequencies; the same values will serve even at 28 Mc., although 100 �1.4fd. will be ample for this and the 14-Mc. band. Still smaller condensers can be used at ultra-high frequencies. Since series tuning is used at alow-
voltage point in the feeder system, the plate spacing of the condensers does not have to be
large. Ordinary receiving-type condensers are large enough for plate voltages up to 1000, and the smaller transmitting condensers have high-enough voltage ratings for higher-power applications. With high-power 'phone it may be necessary to use condensers having aplate spacing of approximately 0.15 to 0.2 inch.
In parallel-tuned circuits (F, G, H, J) the antenna coil and condenser should be approximately the same as those used in the final tank

circuit. The antenna tank circuit must be capa-
ble of being tuned independently to the transmitting frequency, and if possible provision should be made for tapping the coil so that the L/C ratio can be varied to the optimum value (� 2-11) as determined experimentally.
In Fig. 1010-D, C1 and C2 may be 100 to 250 �Leif d. each, the higher-capacity values being used for lower-frequency operation (3.5 and 1.75 Mc.). Plate spacing should in general be
at least half that of the final amplifier tank condenser. For operation from 1.75 to 14 Mc., L1 and 1./2 each should be 15 turns 2 inches
in diameter, spaced to occupy 3inches length, and tapped every three turns. Approximate settings are 15 turns for 1.75 Mc., 9 turns for 3.5 Mc., 6turns for 7 Mc., and 3turns for 14 Mc. The coils may be wound with No. 14
or No. 12 wire. This method of coupling is very seldom used at ultra-high frequencies.
Harmonic reduction-- It is important to prevent, insofar as possible, harmonics in the output of the transmitter from being transferred to the antenna system. Untuned (Fig.
1010-A) and directly-coupled (Figs. 1010-B) systems do not discriminate against harmonics, and hence are more likely to cause harmonic radiation than the inductively-coupled tuned systems. Low-pass filter arrangements such as those at C and D, Fig. 1010, do discriminate against harmonics, but the direct coupling frequently is asource of trouble in this respect.
In inductively-coupled systems, care must
be taken to prevent capacity coupling between coils. Link coils should always be coupled at a point of ground potential (� 2-13) on the plate tank coil, and so should series and paralleltuned coils (E and F) when possible. Capacity coupling can be practically eliminated by the
use of a Faraday shield (� 4-9) between the
two coils.

'Yz
Fig. 1011 -- Half-wave antennas fed from resonant
lines. A and B, end feed with quarter- and half-wave
lines; C and D, center feed. The current distribution is shown for all four cases. Arrows indicate instantaneous direction of current flow.

� 10-7 RESONANT LINES
Two-wire lines -- Because of its simplicity of adjustment and flexibility with respect to the frequency range over which an antenna system will operate, the resonant line is widely used with simple antenna systems. Constructionally, the spaced or "open" two-wire line is best suited to resonant operation; rubber-insulated lines such as twisted pair will have excessive losses when operated with standing waves.
Connection to antenna -- A resonant line is usually -- in fact, practically always -- connected to the antenna at either a current or voltage loop. This is advantageous, especially when the antenna is to be operated at harmonic frequencies, since it simplifies the problem of determining the coupling system to be used at the input end of the transmission line.
Half-wave antenna with resonant line -- It is often helpful to look upon the resonant

172 CHAPTER TEN

line simply as an antenna folded back on itself. Such a line may be any whole-number multiple of a quarter-wave in length; in other words any total wire length which will accommodate awhole number of standing waves.
(The "length," however, of a two-wire line is always taken as the length of one of the wires.)
Quarter- and half-wave resonant lines feeding half-wave antennas are shown in Fig. 1011. The current distribution on both antenna and line is indicated. It will be noted that the quarter-wave line has maximum current at one
end and minimum current at the other, determined by the point of connection to the antenna. The half-wave line, however, has the same current (and voltage) values at both ends.
If aquarter-wave line is connected to the end of an antenna as shown in Fig. 1011-A, then at the transmitter end of the line the current is
high and the voltage low (low impedance) so that series tuning (� 10-6) can be used. Should the line be a half-wave long, as at 1011-B, current will be minimum and voltage maximum (high impedance) at the transmitter
end of the line, just as it is at the end of the antenna. Parallel tuning therefore is required (� 10-6). The line could be coupled to a bal-
anced final tank through small condensers,

-Antenna Sy.31ein3
as in Fig. 1010-B, but the inductively-coupled
circuit is preferable. An end-fed antenna with resonant feeders, as in 1011-A and B, is known as the "Zeppelin," or "Zepp," antenna.
The line also may be inserted at the center of the antenna at the maximum-current point. Quarter- and half-wave lines used in this way are shown at Fig. 1011-C and D. In C, the antenna end of the line is at ahigh-current, lowvoltage point (� 10-2), hence at the transmitter end the current is low and the voltage high. Parallel tuning therefore is used. The halfwave line at D has high current and low voltage at both ends, so that series tuning is used at the transmitter end.
The four arrangements shown in Fig. 1011 are thoroughly useful antenna systems, and are shown in more practical form in Fig. 1012. In each case the antenna is a half wavelength long, the exact length being calculated from Equations 2, 3or 4 (� 10-2) or taken from the charts of Fig. 1015. The line length should be an integral multiple of a quarter wavelength, and may be calculated from Equation 5(� 10-5) the result being multiplied by any whole number which gives a total length convenient for reaching from the antenna to the transmitter. If there is an odd number of quarter waves on the line in the case of the end-fed antenna,

(A )

C, TOOLDAOTA TaA'itie 11".1 1-2

XMTR ML 2 � TANK

XMTR

TANK

Fig. 1012 -- Practical half-wave antenna systems using resonant-line feed. In the center-feed systems., the antenna length "X" does not include the length of the insulator at the center. Line length is measured from the antenna to the tuning appartus; leads in the latter should be short enough to be neglected. The two meters shown are helpful for balancing feeder currents; however, one is sufficient for tuning for maximum output, and may be transferred from one feeder to the other, if desired. The systems at (4) and (C) are for feeders an odd number of quarter-waves in length; (B) and (D) for feeders amultiple of ahalf wavelength. The drawings correspond electrically to those of Fig. 1011.
173 CHAPTER TEN

Dhe ledio Amuieur'i ilandhooh

series tuning will be used at the transmitter end; if an even number of quarter waves, then parallel tuning is used. With the center-fed antenna the reverse is true.
Practical line lengths--In general, it is best to use line lengths that are integral multiples of a quarter wavelength. Intermediate
lengths will give intermediate impedance values and will show reactance as well (� 10-5). The tuning apparatus is capable of compensat-
ing for reactance but it may be difficult to get suitable transmitter loading because simple series and parallel tuning are suitable for only
low and high impedances, respectively, and neither will perform well with impedances of the order of afew hundred ohms. Such values of impedance may reduce the Q of the coupling
circuit to such apoint that adequate coupling cannot be obtained (� 2-11). However, some departure from the ideal length is possible --
even as much as 25% of a quarter wave in many cases -- without undue difficulty in tuning and coupling. In such cases the type of
tuning to use, series or parallel, will depend on whether the feeder length is nearer an odd number of quarter waves or nearer an even number, as well as on the point at which the
feeder is connected to the antenna. Line current -- The feeder current as read
by the r.f. ammeters is useful for tuning purposes only; the absolute value is of little importance. When series tuning is used the current will be high, but very little current will be indicated in a parallel-tuned system. This is because of the current distribution on the feeders as shown by Fig. 1011. With a given antenna and tuning system, of course, the
greatest power will be delivered to the antenna when the readings are highest. However,
should the feeder length be changed no useful conclusions can be drawn from comparison between the new and old readings. For this

reason any indicator which registers the relative intensity of r.f. current can be used for tuning purposes. Many amateurs, in fact, use
flashlight or dial lamps for this purpose instead of meters. They are inexpensive, and when shunted by short lengths of wire so that considerable current can be passed without burnout will serve very well even with high-power transmitters.
Antenna length and line operation -- Insofar as the operation of the antenna itself is concerned, departures of afew per cent from the exact length for resonance are of negligible consequence. Such inaccuracies may influence the behavior of the feeder system, however, and as aresult may have an adverse effect on the operation of the system as a whole. This is true of the end-fed antennas such as are shown in Fig. I012-A and -B.
For example, Fig. 1013-A shows the current distribution on the half-wave antenna and quarter-wave feeder when the antenna length is correct. At the junction of the "live" feeder
and the antenna the current is minimum so that the currents in the two feeder wires are equal at all corresponding points along their length. When the antenna is too long, as in B, the current minimum occurs at apoint on the antenna proper, so that at the top of the live feeder there is already appreciable current flowing, whereas at the top of the "dead" feeder the current must be zero. As aresult, the feeder currents are not balanced and some power will be radiated from the line. In C the antenna is too short, bringing the current minimum to apoint on the live feeder, so that again the currents are unbalanced. The more serious the unbalance the greater the radiation from the line.
Strictly speaking, a line having an unbalanced connection such as the one-way termination at the end of an antenna cannot be truly

sm(oCp)T

Fig. 1013 -- Effect on feeder balance of incorrect antenna length. With center feed, incorrect antenna length does not unbalance the transmission line, as it does with end feed.

174

CHAPTER TEN

Anienna Syi Fern �

balanced even though the antenna length is correct. This is because of the difference in loading on the two sides. The effect is fairly small, however, when the currents are balanced.
If the antenna is fed at the center the undesirable effects of incorrect antenna length balance out so that the line operates properly under all conditions. This is shown in Fig. 1013 at D, E and F. So long as the two halves of the antenna are of equal length, the distribution of current on the feeders will be symmetrical so that no unbalance exists, even for antenna lengths considerably removed from the correct value.
� 10-8 NON-RESONANT LINES
Requirements -- The advantages of nonresonant transmission lines -- minimum losses, and elimination of the necessity for tuning -- make this type of operation attractive. The chief disadvantage of the non-resonant line, aside from the necessity for more care in initial adjustment, is that when "matched" to the ordinary antenna it is matched only for one frequency, or at most for a small band of frequencies on either side of the frequency for which the matching is done. Except for afew special systems, this means that the antenna is unsuitable for work on more than one amateur band.
Adjustment of anon-resonant line is simply that of adjusting the terminating resistance to match the characteristic impedance of the line. To accomplish this, the antenna itself must be resonant at the selected frequency, and the line must then be connected to it in such away that the antenna impedance as looked at by the line is the right value. The matching may be done by connecting the line at the proper spot along the antenna, by inserting an impedance transforming device between the antenna and line, or by using a line having an impedance equal to the center impedance of the antenna.
In the following discussion of ways in which different types of lines may be matched to the antenna, ahalf-wave antenna is used as an example. Other types of antennas may be treated by the same methods, making due allowance for the order of impedance that appears at the end of the line with more elaborate systems.
Single-wire feed-- In the single-wire-feed system the return circuit is through the ground. There will be no standing waves on the feeder when its characteristic impedance is matched by the impedance of the antenna at the connection point. The principal dimensions are the length of the antenna L, Fig. 1014, and the distance D from the exact center of the antenna to the point at which the feeder is attached. Approximate dimensions

oantytluenneitd4kte
Fig. 1014--Single-wire-feed system. The length L (one-half wavelength) and D are determined from the chart, Fig. 1015.
can be obtained from Fig. 1015 for an antenna system having afundamental frequency in the most used amateur bands.
In constructing an antenna system of this type the feeder must run straight away from the antenna (at aright angle) for adistance of at least one-third the length of the antenna. Otherwise the field of the antenna will affect the feeder and cause faulty operation. There should be no sharp bends in the feeder wire at any point.
With the coupling system shown in Fig. 1016-A, adjustment is as follows: Starting at the ground end of the tank coil, the tap is moved towards the plate end until the amplifier draws the rated amount of plate current. The plate tank condenser should be readjusted each time the tap is changed, to bring the plate current to minimum. The amplifier is loaded properly when this "minimum" is the rated current. The condenser in the feeder is for the purpose of insulating the antenna system from the high-voltage plate supply when series plate feed is used. It should have a voltage rating somewhat above that of the plate supply. Almost any capacity greater than 500 gpfd. will be satisfactory. The condenser is unnecessary, of course, if parallel plate feed is used.
Inductive coupling to the output circuit is shown in Fig. 1016-B. The antenna tank circuit should tune to resonance at the operating frequency and the loading is adjusted by varying the coupling between the two tanks, both being kept tuned to resonance.
Regardless of the type of coupling, a good ground connection is essential with this system. Single-wire feed works best over moist ground, and poorly over rock and sand.
Twisted-pair feed-- A two-wire line composed of twisted rubber-covered wire can be constructed to have an impedance approximately equal to that at the center of the antenna itself, thus permitting the method of connecting the line to the antenna shown in Fig. 1017. Any discrepancy which may exist between line and antenna impedance can be compensated for by aslight fanning of the line

175 CHAPTER TEN

R h ..7he ac o Adnaleur'� ...flantllooh

1750

1800 c 1850
iz 1900
1950 cy 2000 rt 4- 2050
220 16 3500 d 3600

�
e
I �

230 17

240 18

38
�
250 19

260

270

(3 3700
�
3800

k+ 3900 cc 4000
114

e
118

700 0

122 9'3"

126 9'4'

130

134

9'6"

ku 7050

7100

liso
4, 7200 C:r 4.1 7250 ltte.
7300 62
tu
14.1

e I

48'

14.2

14.3 C3/ 144 4' 328"
2'2' 28.0 Z1 28.4

324"

3217'

2'3'

33'0' 2'4'

334"

33 8'

2'5"

28.8

29.2 29.6

4CC1

300 15

'0"

e e e'

e
15'4"

158'

06.e.
(.3 57.0 k 58.0

1'1.5'

fe t. 4.

lee

16'8"

1'25'

k 59.0

e600

7'6'

7' a�

8'2'

1Y4'

176 CHAPTER TEN

where it connects to the two halves of the antenna, as shown at B in Fig. 1017.
The twisted line is aconvenient type to use, since it is easy to install and the r.f. voltage on it is low because of the low impedance. This makes insulation an easy matter. Special twisted line for transmitting purposes, having lower losses than ordinary rubber-covered wire, is available.
The antenna should be one-half wavelength long for the frequency of operation, as determined by charts of Fig. 1015 or the formulas (� 10-2). The amount of "fanning" (dimension B) will depend upon the kind of cable used; the right value usually will be found between 6 and 18 inches. It may be checked by inserting ammeters in each antenna leg at the junction of the feeder and antenna; the value of B which gives the largest current is correct. Alternatively, the system may be operated continuously for a time with fairly high r.f. power input, after which the feeder may be inspected (by touch) for hot spots. These
indicate the presence of standing waves, and the fanning should be adjusted until they are eliminated or minimized. Each leg of the feeder forming the triangle at the antenna should be equal in length to dimension B.
Coupling between transmitter and transmission line is ordinarily by the untuned coil method shown in Fig. 1010-A (� 10-6).
Concentric-line feed -- A concentric transmission line readily can be constructed to have a surge impedance equal to the 70-ohm impedance at the center of a half-wave antenna.
Such a line, therefore, can be connected directly to the center of the antenna, forming the system shown in Fig. 1018.
Solving Equation (6) (� 10-5) for an air-
insulated concentric line shows that, for 70ohm surge impedance, the inside diameter of the outer conductor should be approximately 3.2 times the outside diameter of the inner conductor. This condition can be fulfilled by using standard 5 -inch (outside-diameter) copper tubing for the outer conductor and No. 14 wire for the inner. Ceramic insulating spacers are available commercially for this combination. Rubber-insulated concentric line having the requisite impedance for connection to the center of the antenna also is available.
The operation of such an antenna system is similar to that of the twisted-pair system just described, and the saine transmitter-coupling arrangements may be used (� 10-6).
The outer conductor of the line may be grounded if desired. The feeder system is
Fie. 1015 -- Charts for determining the length of half-wave antennas for use on various amateur bands. Solid lines indicate antenna length (lower scale); dotted lines point of connection for single-wire feeder (upper scale) measured from center of antenna.

Fig. 1016-- Methods of coupling the sinfle-wire feeder to the transmitter. Circuits are shown for both singleended and balanced tank circuits.

qle wire

feeder

_Antenna.

y�terni

slightly unbalanced because the inner and outer conductors do not have the same capacity to ground. There should be no radiation, however, from a line having a correct surge impedance.
Delta matching transformer -- Because of the extremely close spacing required, it is impracticable to construct an open-wire transmission line which will have a surge impedance low enough to work directly into the center of a half-wave antenna. Such wire lines usually have impedances between 400 and 700 ohms, 600 ohms being a widely-used

must be designed for exact impedance values as well as frequency values and the dimensions are therefore fairly critical.
The length of the antenna is figured from the formula (� 10-2) or taken from Fig. 1015.
The length of section C is computed by the formula:

C (feet) --

123

Freq. (Mc.)

The feeder clearance E is found from the equation:

E (feet)

148 Freq. (Mc.)

The above equations are for feeders having a characteristic impedance of 600 ohms and will not apply to feeders of any other impedance. The proper feeder spacing for a600-ohm trans-
mission line is computed to asufficiently close approximation by the following formula:

D = 75 X d

Fig. 1017--Half-wave antenna center-fed by a twisted pair line.
value. It is therefore necessary to use other means for matching the line to the antenna.
One method of matching is illustrated by the antenna system of Fig 1019. The section E is "fanned" to have a gradually increasing impedance so that its impedance at the antenna end will be equal to the impedance of the antenna section C, while the impedance at the lower end matches that of apracticable transmission line.
The antenna length L, the feeder clearance E, the spacing between centers of the feeder wires D, and the coupling length C are the important dimensions of this system. The system

where D is the distance between the centers of the feeder wires and d is the diameter of the wire. If the wire diameter is in inches the spacing will be in inches and if the wire diameter is in millimeters the spacing will be in millimeters.

�oo-Olon line
ony length

Speeder:

Outer Tub/n Optional ground be outer Conductor

Inner wire
O, Tubmg Concentric Line
(Any Length)

To Coup/inf
Fig. 1018--Half-wave antenna with concentric transmission line.

Fig. 1019 -- Delta-matched antenna system. The dimensions C, D, and E are given in the text. It is important that the matching section, E, come straight away from the antenna without any bends.
Methods of coupling to the transmitter are discussed in �10-6, Figs. 1010-C, D, G and H being suitable.
"Q"-section transformer -- The impedance of atwo-wire line of ordinary construction (400 to 600 ohms) can be matched to the impedance of the center of a half-wave antenna
177 CHAPTER TEN

5L9 Pacho

ilancllooh

by utilizing the impedance-transforming prop-
erties of a quarter-waveline (� 10-5). The matching section must have low surge impedance and therefore is commonly constructed of large-diameter conductors such as aluminum or copper tubing, with fairly close spacing. This system is known as the "Q"
antenna. It is shown in Fig. 1020. The important dimensions are the length of the antenna, the length of the matching section, B, the spacing between the two conductors of the matchinisection, C, and the impedance of the untuned transmission line connected to the lower end of the matching section.
The required surge impedance for the match-
ing section is

Z. = VriTz

(9)

where Z1 is the input impedance and Z2 the output impedance. A quarter-wave section matching a 600-ohm line to the center of a half-wave antenna (72 ohms) should have a surge impedance of 208 ohms. The spacings

Speadenr
en�anec,bne any kyny,th
Shore'',hhk Fig. 1021-- Half-wave antenna systems with quarterwave open wire matching transformers.

Fig. 1020 -- The "Q" antenna with quarter-wave matching section using spaced tubing.

between conductors of various sizes of tubing and wire for different surge impedances are given in graphical form in Fig. 1009. With half-inch tubing, for example, the spacing should be 1.5 inches for an impedance of 208
ohms. The length, B, of the matching section
should be equal to a quarter wavelength, and
is given by

Length of

_ 234

wave line (feet) Freq. (Mc.)

The length of the antenna can be calculated from the formula (� 10-2) or taken from the
charts of Fig. 1015. This system has the advantage of the sim-
plicity of adjustment of the twisted pair feeder system and at the same time the superior insulation of an open-wire system. Figs. 1010-B, D, G (� 10-6) and H represent suitable methods of coupling to the transmitter.
Linear transformers --Fig. 1021 shows two methods of coupling a non-resonant line to a half-wave antenna through a quarter-

wave linear transformer (� 10-5) or matching section. In the case of the center-fed antenna
the free end of the matching section, B, is open (high impedance) since the other end is connected to a low-impedance point on the antenna. With the end-fed antenna the free end
of the matching section is closed through a shorting bar or link; this end has low impedance since the other end is connected to a high-impedance point on the antenna (�10-7).
In the center-fed system, the antenna and matching section should be cut to lengths found from the equations in �10-2 and �10-5.
Any necessary on-the-ground adjustment can be made by adding to or clipping off the open ends of the matching section. The matching
section in the end-fed system can be adjusted by making the line alittle longer than necessary and adjusting the system to resonance by moving the shorting link up and down.
Resonance can be obtained by exciting the antenna at the proper frequency from a temporary antenna nearby and measuring the current in the shorting bar by alow-range
r.f. ammeter or galvanometer. The position of the bar should be adjusted for maximum current reading. This should be done before the transmission line is attached to the match-
ing section. The position of the line taps must be deter-
mined experimentally, since it will depend upon the impedance of the line as well as the antenna impedance at the point of connection. The procedure is to take a trial point, apply power to the transmitter, and check the transmission line for standing waves. This can be done by measuring the current in the
wires, using adevice of the type pictured in

178

CHAPTER TEN

A ntenna Sy.itemi

Fig. 1022. The hooks (which should be sharp enough to cut through insulation, if any, of the

4';17'jteoe351e

Let wooden spacer

wires) are placed on one of the wires, the spac-

ing between them being adjusted to give a

suitable reading on the meter. At any one position along the line the currents in the two

Heavy sle wins

wires should be identical. Readings taken at

intervals of aquarter wavelength will indicate whether or not standing waves are present.'

8com.to support mi.�

It will not usually be possible to obtain complete elimination of standing waves when
the matching stub is exactly resonant. The line taps should be adjusted for the smallest obtainable standing-wave ratio. Then afurther

Ihermo mdbommele.r o� 4so or 0- 250 .stale
Fig. 1022--Line-current measuring device for adjustment of untuned transmission lines.

"touching up" of the matching stub tuning will eliminate the remaining standing wave, provided the adjustments are made carefully. The stub must be readjusted because when
resonant it exhibits some reactance as well as resistance at all points except at the ends, and the slight lengthening or shortening of the
stub is necessary to tune out this reactance. The required readjustment is quite small, however.
When the connection between matching section and antenna is unbalanced, as in the end-fed system, it is important that the antenna be the right length for the operating frequency if a good match is to be obtained (� 10-7). The balanced center-fed system is less critical in this respect. The shorting-bar method of tuning the center-fed system to resonance may be used if the matching section is extended to a half wavelength, bringing a current loop at the free end.
An impedance mismatch of several per cent is of little consequence so far as power trans-
fer to the antenna is concerned. It is relatively easy to get the standing wave ratio down to 2or 3to 1, aperfectly satisfactory condition in practice. Of considerably greater importance
is the necessity for getting the currents in the
two wires balanced both as to amplitude and phase. If the currents are not the same at corresponding points on adja�ent wires, and the loops and nodes do not also occur at corresponding points, there will be considerable radiation loss. This balance can only be brought about by perfect symmetry in the line, particularly with respect to ground. This symmetry should extend to the coupling apparatus at the transmitter. An electrostatic shield between the line and the transmitter coupling coils often will be of value in pre-
venting capacity unbalance, and at the same time will reduce harmonic radiation.

integral multiple of a half-wavelength. When the antenna is more than ahalf-wave long, it is usually called a long-wire antenna, or a harmonic antenna.
Current and voltage distribution--Fig. 1023 shows the current and voltage distribution along awire operating at its fundamental
frequency (where its length is equal fo a. half wavelength) and at its second, third and fourth harmonics. For example, if the fundamental frequency of the antenna is 7 Mc.,
the current and voltage distribution will be as shown at A. The same antenna excited at 14 Mc. would have current and voltage distribution as shown at B. At 21 Mc., the third harmonic of 7 Mc., the current and voltage distribution would be as in C; and at 28 Mc., the fourth harmonic, as in D. The number of the harmonic is the number of half-waves contained in the antenna at the particular operating frequency.
The polarity of current or voltage in each standing wave is opposite to that in the adjacent standing waves. This is shown in the
figure by drawing the current and voltage curves successively above and below the antenna (taken as a zero reference line) to indicate that the polarity reverses when the current or voltage goes through zero. Currents flowing in the same direction are in phase; in opposite directions, out of phase.
It is evident that one antenna may be used for harmonically related frequencies, such as the various amateur bands. The long-wire or harmonic antenna is the basis of multi-band operation with one antenna.
Physical lengths-- The length of a longwire antenna is not an exact multiple of that of a half-wave antenna because the end effects
(� 10-2) operate only on the end sections of the antenna; in other parts of the wire these effects are absent and the wire length is ap-

�10-9 LONG-WIRE ANTENNAS

proximately that of an equivalent portion of

Definition-- An antenna will be resonant if an integral number of standing waves of

the wave in space. The formula for the length of along-wire antenna therefore is:

current and voltage can exist along its length; in other words, so long as its length is some

Length (feet) -- 492 (N-0.05) Freq. (Me.)

(10)

CHAPTER TEN 179

Dhe Radio AnalettA ..._11anclloJ

ous angles with the wire. In general, as the

length of the wire is increased the direction of

maximum radiation tends to approach the line of the antenna itself.

Directional characteristics for antennas one

FUNDAMENTAL (HALF-WAVE)

wavelength, three half-wavelengths, and two

wavelengths long are given in Figs. 1025, 1026

and 1027, for three vertical angles of radiation.

Note that as the wire length increases the

radiation along the line of the antenna becomes

more pronounced. Still longer antennas can

2Ho HARMONIC (FULL -WAVE)

be considered to be practically "end-on" radiators, even at the lower radiation angles.

Methods of feeding--In a long-wire an-

tenna the currents in adjacent half-wave sec-

tions must be out of phase, as shown in Fig.

1028 and Fig. 1023. The feeder system must

3RD HARMONIC (3/2.-WAVE)

not upset this phase relationship. This requirement is met by feeding the antenna at

either end or at any current loop. A two-wire feeder cannot be inserted at a current node,

however, because this invariably brings the

currents in two adjacent half-wave sections in

phase; if the phase in one section could be re-

4114 HARMONIC (2-WAvE)

versed then the currents in the feeders would

Fig. 1023--Current and voltage distribution along
an antenna operated at various harmonics of its funda. mental resonant frequency.

be in phase and the feeder radiation would not
be cancelled out. Either resonant or non-resonant feeders may
be used. With the latter, the systems employ-

where N is the number of half-waves on the antenna. From this it is apparent that an antenna cut as a half-wave for a given frequency will be slightly off resonance at exactly

ing a matching section (� 10-8) are best. The non-resonant line may be tapped on the matching section as in Fig. 1021 or a "Q" type section, Fig. 1020, may be employed.

twice that frequency (on the second harmonic) because of the diffeient behavior of end effects

pix.

when there is more than one standing wave on

the antenna. For instance, if the antenna is

160

cut to exact fundamental resonance on the

second harmonic (full wave) it should be 2.6% longer, and on the fourth harmonic

140 .

A���\ /e./..........

(two-wave), 4% longer. The effect is not very

important except for a possible unbalance in

120

the feeder system (� 10-7) which may result

in some radiation from the feeder in end-fed

systems.

,100

Impedance and power gain-- The radiation resistance as measured at a current loop

:,80 (

becomes larger as the antenna length is increased. Also, a long-wire antenna radiates

r !

more power in its most favorable direction

than does a half-wave antenna in its most i

favorable direction. This power gain is secured

at the expense of radiation in other directions.

Fig. 1024 shows how the radiation resistance and power in the lobe of maximum radiation

vary with the antenna length.

Directional characteristics-- As the wire is made longer, in terms of the number of half

234 AN5 TEN6 NA7 LE6 NGTH -9A

.- .

wavelengths, the directional effects change. Instead of the "doughnut" pattern of the half-wave antenna, the directionai characteristic splits up into "lobes" which make vari-

Fig. 1024-- Curve A, variation in radiation resistance
with antenna length. Curve B, power in the lobes of maximum radiation for long-v, ire antennas as aratio to
the maximum of ahalf-wave antenna.

180 CHAPTER TEN

60 7

�

130.

�0 1

10 I

20 I

90 o
25 .... �
4 4
0

25 10 30
90

% 2,0'

,
� .
�

1 10
I
I

I 20, I
30
JO

60 i:a

Antenna Syilem�
� 10-10 MULTI-RAND ANTENNAS
Principles-- As suggested in the preceding section, the same antenna may be used for several bands by operating it on harmonics. When this is done, it is necessary to use resonant feeders, since the impedance matching for non-resonant feeder operation can be accomplished only at one frequency unless means are provided for changing the length of amatching section and shifting the point at which the feeder is attached to it. A matching section which is a quarter-wavelength long on one frequency will be a half-wavelength long on twice that frequency, and so on, and changing

Fig. 1025 -- Horizontal patterns of radiation from a full-wave antenna. Solid line, vertical angle 15 degrees; dotted lines, deviation from 15-degree pattern at 9and 30 degrees.
All three patterns are drawn to the same relative scale; actual amplitudes will depend upon the height of the antenna.

In such case, Fig. 1029 gives the required surge impedance for the matching section. It can
also be calculated from Equation 9 (� 10-8) and the radiation resistance data in Fig. 1024;
Methods of coupling the line to the transmitter are the same as described in �10-6 for the particular type of line used.

7
60 50

80 9 I9
k25

20
I 10
I
0 a

� 9* .

�.%

e.
i,, 5

�

20 �� .

.. .., ���

60 7 80

a � I

���
SO 60 70

20 30

Fig. 1026 -- Horizontal patterns of radiation from an antenna three half-wavelengths long. Solid line, vertical angle 15 degrees; dotted lines, deviation from 15-degree pattern at 9 and 30 degrees. The minor lobes coincide for all three angles.

Fig. 1027 -- Horizontal patterns of radiation from an antenna two wavelengths long. Solid line, vertical angle 15 degrees; dotted lines, deviation from 15-degree pattern at 9 and 30 degrees. The minor lobes coincide for all three angles.
the length of the wires, even by switching, is inconvenient.
Also, the current loops shift to a new position on the antenna when it is operated on harmonics, further complicating the feed situation. It is for this reason that half-wave antennas center-fed by rubber-insulated lines are practically useless for harmonic operation; on all even harmonics there is avoltage maximum at the feed point and the impedance mismatch is so bad that there is a large standing-wave ratio and consequently high losses in the rubber dielectric.
When the same antenna is used for work in several bands, it must be realized that the directional characteristic will depend on the band in use.
Simple systems -- Any of the antenna arrangements shown in �10-7 may he used for

CHAPTER TEN 181

n eRack� ...Amateur's ilaneMooh
(A)
(B)
es..
(C)

not effective radiators -- it is simply that the directional characteristic will not be that of a long-wire having the same overall length. Rather it will resemble the characteristic of one side of the antenna, although this is not exact.
Antennas with afew other types of feed systems may be operated on harmonics for the higher-frequency bands, although their performance is somewhat impaired. The single-

Fig. 1028 -- Current distribution and feed points for long-wire antennas. A 3/2-wave antenna is used as an illustration. With two-wire feed, the line may be con nected at the end of the antenna or at any current loop (not at acurrent node).

X14 TR

For operation on 3 S1.14 and
28 me

multi-band operation by making the antenna a
half wave long at the lowest frequency to be used. The feeders should be aquarter wave, or some multiple of a quarter wave, long at the same frequency. Typical examples, with the type of tuning to be used, are given in Table I. The figures given represent acompromise to give satisfactory operation on all the bands considered, taking into account the change in required length as the order of the harmonic
goes up. A center-fed half-wave antenna will not
operate as along wire on harmonics because of the phase reversal at the feeders previously mentioned (� 10-9). On the second harmonic, the two antenna sections are each ahalf wave long, and since the currents are in phase the directional characteristic is different from that of afull-wave antenna even though the overall length is the same. On the fourth harmonic, each section is afull wave long and again because of the direction of current flow the system will not operate as a two-wave antenna. It should not be assumed that these systems are

fTruenqeude'nlcoyIagsoIeronsindier

OPeTR

For operation
on /75 ,77C

Fig. 1030 -- A simple antenna system for five amateur bands. The antenna is voltage fed on 3.5, 7, 14 and 28 Mc., working on the fundamental, second, fourth and eighth harmonics, respectively. For 1.75 Mc. the system is a quarter-wave grounded antenna, in which case series tuning must be used. The antenna wire should be kept well in the clear and should he as high as possible.
If the length of the antenna is approximately 260 feet, voltage feed can be used on all five bands.
wire fed antenna (� 10-8) may be used in this way; the feeder and antenna will not be matched exactly on harmonics with the result that standing waves will appear on the feeder, hut the system as a whole will radiate. The

sno
i260
240
zno

'es
182

ASPEN/VA LENGTH, IN WAVELENGTHS

CHAPTER TEN

Fig. 1029 -- Required surge impedance of quarter-wave matching sections for radiators of various lengths. Curve A is for atransmission line impedance of 440 ohms, Curve B for 470 ohms, Curve C for 580 ohms and Curve 1) for 600 ohms.

_Antenna

y3temi

Fig. 1031-- Current distribution on

antennas too short for the fundamental

(A)

frequency. These systems may be used

when space for afull half-wave antenna

is not available. The current distribution

on the second harmonic also is shown to

the right of each figure. In A and C, the total length around the system is ahalf-

wavelength at the fundamental fre-

quency. Arrows show instantaneous

direction of current flow.

(C)

Tee

TABLE I
MULTI-BAND RESONANT-LINE FED ANTENNAS

Antenna Length (ft.)

Feeder Length
(ft.)

Band

Type of Tuning

With end feed: 243

120 1.75-Mc. 'phone series

4-Mc. 'phone parallel

14 Mc.

parallel

28 Mc.

parallel

136

67 3.5-Mc, c.w. series

7 Mc.

parallel

14 Mc.

parallel

28 Mc.

parallel

134

67 3.5-Mc. c.w. series

7Mc.

parallel

7 Mc.

series

14 Mc.

parallel

28 Mc.

parallel

With center feed:

272

135

137

1.75 Mc. 3.5 Mc. 7 Mc. 14 Mc. 28 Mc. 3.5 Mc. 7 Mc. 14 Mc. 28 Mc.

parallel parallel parallel parallel parallel parallel parallel parallel parallel

67.5

7 Mc.

14 Mc.

28 Mc.

parallel parallel parallel

The antenna lengths given represent compromises for harmonic operation because of different end effects on different bands. The 136-foot end-fed antenna is slightly long for 3.5 Mc., but will work well in the region (3500-3600 kc.) which quadruples into the 14-Mc. band. Bands not shown are not recommended for the particular antenna. The center-fed systems are less critical as to length; the 272-foot antenna may, for instance, be used for both c.w. and 'phone on either 1.75 or 4Mc. without loss of efficiency.
On harmonics, the end-fed and center-fed antennas will not have the same directional characteristics, as explained in the text.

same is true of the delta-matched antenna. The "Q" antenna (� 10-8) also can be operated on harmonics, but the line cannot operate non-resonant except at the fundamental frequency of the antenna. For harmonic operation the line must be tuned and, therefore, the feeder length is important. The tuning system will depend upon the number of quarter waves on the line, including the "Q" bars. The concentric-line fed antenna (� 10-8) may be used on harmonics if the concentric line is air-insulated. Its operation on harmonics is similar to that of the "Q." This antenna is not recommended for multi-band operation with a rubber-insulated line, however.
A simple antenna system, without feeders, for operation in five bands is shown in Fig. 1030. On all bands from 3.5 Mc. upward it operates as an end-fed antenna -- half-wave on 3.5 Mc., long wire on the other bands. On 1.75 Mc. it is only aquarter-wave in length and must be worked against ground, which in effect replaces the missing half of the antenna. Since on this band it is fed at a high-current point, series tuning (� 10-6) must be used.
Antennas for restricted space -- If the space available for the antenna is not large enough to accommodate the length necessary for ahalf-wave at the lowest frequency to be used, quite satisfactory operation can be secured by using ashorter antenna and making up the missing length in the feeder system. The antenna itself may be as short as aquarter wavelength and still radiate fairly well, although of course it will not be as effective as one a half-wave long. Nevertheless such a system is useful where operation on the desired band otherwise would be impossible.
Resonant feeders are a practical necessity with such an antenna system, and acenter-fed antenna will give best all-around performance. With end feed the feeder currents become badly unbalanced and, since lengths midway
between those requiring series or parallel tuning ordinarily must be used to bring the
entire system to resonance, coupling to the transmitter often becomes difficult.
183 CHAPTER TEN

Dhe Radio AnzateuA

With center feed, practically any convenient length of antenna can be used if the feeder length is adjusted to accommodate at
least one half-wave around the whole system. Typical cases are shown in Fig. 1031, one for
an antenna having a length of one-quarter wave (A) and the other for an antenna somewhat longer (C) but still not ahalf-wave long. Current distribution is shown for both fundamental and second harmonic. From the points
marked X resonant feeders any convenient number of quarter waves in length may be
extended to the operating room. The sum of the distances on each wire from X to the antenna end must equal a half-wave. It is suffi-
ciently accurate to use Equation 2 (� 10-2) in calculating this length. Note that X-X is a high-current point on these shortened antennas, corresponding to the center of ahalf-wave antenna. It is also apparent that the antenna at A is ahalf-wave antenna on the next higher-
frequency band (B). The practical antenna can be made as in
Fig. �1032. Table II gives afew recommended lengths. Remembering the preceding discussion, however, the antenna can be made any convenient length provided the feeder is considered to "begin" at X-X, and the line length adjusted accordingly.

TA BLE II ANTENNA AND FEEDER LENGTHS FOR SHORT
M ULTI- BAND ANTENNAS, CENTER-FED

Antenna Length (ft.)
137
100

Feeder Length
us.)
68
38

Band
1.75 M c. 3.5 Mc. 7 Mc. 14 M c. 28 M c. 3.5 M c. 7M c. 14 M c. 28 Mc.

TTuynpienogf series parallel parallel parallel parallel parallel series series series or parallel

A
Timing Appaueus Fig. 1032 -- Practical arrangement of a shortened antenna. The total length A + B B � A, should be ahalf-wavelength for the lowest-frequency band, usually 3.5 M c. See Table II for lengths and tuning data.
Bent antennas-- Since the field strength at adistance is proportional to the current in the antenna, the high-current part of ahalf-wave antenna (the center quarter-wave, approximately) does most of the radiating (� 10-1). Advantage can be taken of this fact when the space available does not permit erecting an antenna a half-wave long. To accomplish it, the ends may be bent, either horizontally or vertically, so that the total length equals a half wave, even though the straightaway horizontal length may be as short as aquarter wave. The operation is illustrated in Fig. 1033. Such an antenna will be a somewhat better radiator than the arrangement of Fig. 1031-A on the lowest frequency, but is not as desirable for multi-band operation because the ends play an increasingly important part as the frequency is raised. The performance of the system in such a ease is difficult to predict, especially if the ends are vertical (the most convenient arrangement) because of the combination of horizontal and vertical polarization as well as dissimilar directional characteristics.
� 10-11 LONG-WIRE DIRECTIVE ARRAYS
The "V" antenna--It has been emphasized that as the antenna length is increased the lobe of maximum radiation makes a more acute angle with the wire (� 10-9). Two such

67.5

3.5 M c.

series

7 Mc.

parallel

14 Mc.

parallel

28 Mc.

parallel

7 Mc.

parallel

14 Mc.

parallel

28 Mc.

parallel

7 Mc.

parallel

Fig. 1033-- Folded arrangement for shortened an-

14 Mc.

parallel

tennas. The total length is a half-wave, not including

28 Mc.

parallel

the feeders. The horizontal part is made as long as

convenient and the ends dropped down to make up the

7Mc.

parallel

required length. The ends may be bent back on them-

14 Mc.

series

selves in feeder fashion to cancel radiation partially.

28 Mc.

parallel

The horizontal section should be at least aquarter-wave

long.

184 CHAPTER TEN

An(enfla Syibm�

wires may be combined in the form of a oo

horizontal "V" so that the main lobes

tos

from each wire will reinforce along aline

bisecting the angle between the wires.

This increases both gain and directivity,

since the lobes in directions other than

DESIGN CHART
FOR HORIZONTAL "V
(,ir maXikurn ouffud)

along the bisector cancel to a greater or

lesser extent. The horizontal "V" antenna therefore transmits best in either

direction (is bi-directional) along the line ".N 75

bisecting the "V" made by the two wires.

The power gain depends upon the length

of the wires. Provided the necessary space is available, the "V" is asimple antenna

to build and operate, and can be used

readily on harmonics so that it is suitable

for multi-band work. The "V" antenna

is shown in Fig. 1034. Fig. 1035 shows the dimensions that
should be followed for an optimum design to obtain maximum power gain for differ-

1(311

(aLrE) NGT3H2(31L)--4(W24A.V)ELE5NG.5T6H)S 6(t6')

7(or)

804"

toques 47parent/6mi nersairto,q6n6/4,4& 40,04406,ten'Might engofooehothwiefc,qa

ent-sized "V" antennas. The longer systems give good performance in multi-band op-

Fig. 1035 -- Design chart for horizontal "V" antennas. Enclosed angle between wires versus length of sides.

eration. Angle ais approximately equal to twice

the angle of maximum radiation for a single wire equal in length to one side of the "V."
The wave angle referred to in Fig. 1035 is the vertical angle of maximum radiation (� 10-1). Tilting the whole horizontal plane of the "V" will tend to increase the low-angle radiation
off the low end and decrease it off the high end. The gain increases with the length of the
wires, but is not exactly twice the gain for a single long wire as given in Fig. 1024. In the longer lengths, the gain will be somewhat increased because of mutual coupling between the wires. A "V" eight wavelengths on aleg, for instance, will have a gain of about 12 db. over a half-wave antenna, whereas twice the gain of a single 8-wavelength wire would be approximately 9db.
The two wires of the "V" must be fed out of phase for correct operation. A resonant line may simply be attached to the ends as shown in Fig. 1034. -Alternatively, a quarter-wave matching section may be employed and the
antenna fed through a non-resonant line
(� 10-8). If the antenna wires are made multiples of a half-wave in length (use Equation 10, �10-9, for computing the length) the

matching section will be closed at the free end. The rhombic antenna -- The horizontal
rhombic or "diamond" antenna is shown in
Fig. 1036. Like the "V," it requires agood deal of space for erection, but it is capable of giving excellent gain and directivity. It can also be used for multi-band operation. In the terminated form shown in Fig. 1036 it operates, like anon-resonant transmission line, without standing waves, and is uni-directional. It may also be used without the terminating resistor, in which case there are standing waves on the
wires and the antenna is bi-directional. The important quantities influencing the
design of the rhombic antenna are shown in Fig. 1036. While several design methods may be used, the one most applicable to the condi-
tions existing in amateur work is the so-called "compromise" method. The chart of Fig. 1037 gives design information based on a given length and wave angle to determine the remaining optimum dimensions for best operation. Curves for values of length of 2, 3and 4 wavelengths are shown, and intermediate values may be interpolated.
With all other dimensions correct, an increase

in length causes an increase in power gain and

a slight reduction in wave angle. An increase

in height also causes areduction in wave angle

/i,e rranflIVISIOn

�re Direction

and an increase in power gain but not to the same extent as a proportionate increase in length.

For multi-band work, it is satisfactory to

AB.CD Fig. 1034--The "V" antenna, made by combining two long wires in such a way that each reinforces the other's radiation. The important quantities are the length of each leg and the angle between legs.

design the rhombic antenna on the basis of 14-Mc, operation, which will permit work on
the 7-and 28-Mc, bands as well. A value of 800 ohms is correct for the termi-
nating resistor for any properly constructed

185 CHAPTER TEN

.5he Radio ArnctieuA -Wanellooh

pedance will be matched by an 800-

ohm line, which may be constructed

from No. 16 wire spaced 20 inches

krmeity

or from No. 18 wire spaced 16 inches.

WRECIiveY

The 800-ohm line is somewhat un-

TOP VIEW

RECEIVING

gainly to install, however, and may

be replaced by an ordinary 600-ohm

re44�

line with only anegligible mismatch.

A t

,,,!1461.Pe e.

.4 4p 1

C
L

Alternatively, a matching section may be installed between the antenna terminals and alow-impedance line. However, when such an arrangement

a/�-4� 4,e7///71

is used it will be necessary to change the matching section constants for

SI DE ELEVATION '

0 MOLE OF TILT (DEGREES) =WAVE Awgie (DEGREES)

L. LENGTH OF ORE SIDE (WAVELENGTHS)
Na HEIGHT (WAVELENGTHS)

each different band of operation. The same design details apply to
the unterminated rhombic as to the

Fig. 1036 -- The horizontal rhombic or diamond an-

terminated type. Resonant feeders

tenna, terminated.

are preferable for the unterminated rhombic.

A non-resonant line may be used by incorpo-

rhombic, and the system behaves as apure resistive load under this condition. This terminating resistor must be capable of safely

rating amatching section at the antenna, but is not readily adaptable to multi-band work.
Rhombic antennas will give apower gain of

dissipating one-half the power output (to eliminate the rear pattern) and should be non-inductive. Such aresistor may be made up

10 db. or more when constructed according to the charts given. In general, the larger the antenna the greater the power gain.

from a carbon or graphite rod or from along 800-ohm transmission line using resistance wire. If the carbon rod or a similar form of

� 10-12 DIRECTIVE ARRAYS WITH DRIVEN ELEMENTS

lumped resistance is used the device should be

Principles -- By combining individual half-

suitably protected from weather effects, i.e., covered with good asphaltic compound and

wave antennas into an array with suitable spacing between antennas (called elements)

sealed in a small light-weight box or fibre and feeding power to them simultaneously, it

tube. Suitable resistors are available commercially.
For feeding the antenna, the antenna im-

is possible to make the radiated fields from the individual elements add in afavored direction, thus increasing the field strength in that

1.5
ut \ 1.3 \
1.2

COMPROMISE DESIGN
(area Lana'either0ord to find otherenentigns)

A1 LO .9

.6 .- L=4 Wee/myths

L.3 Wavelengths

' 4

.2 �

L.2 Wavelengths

WAVE ANGLE (k)-DEGREES

Fig. 1037 -- Compromise method design chart for various leg lengths and wave angles. The following examples illustrate the use of the Chart: (1) Given: Length (L) = 2wave-lengths.

Desired wave angle

(�) = 20�. To Find: H,

Method:

Draw vertical line through point "a" (L = 2 wavelengths) and point "b" on abscissa (A -= 20�.)

74%

Read angle of tilt (4) for

I I

point "a" and height (H) from intersection of line

I I l

702

ab at point "c" on curve H.

(6.t.e1k'.

Result: � = 60.5�. H = 0.73 wavelength.

(2) Given:

I
Iif

Length (L) = 3 wavelengths.
Angle of tilt (s)= 78�. To Find: 11, A. Method:
Draw vertical line from point "d" on curve L = 3 wavelengths at 4, = 78�. Read intersection of this line on curve H (point

"e" and intersection at

point "f" on the abscissa for A.

Result:

H = 0.56 wavelength.

A = 26.6�.

186 CHAPTER TEN

Anienna

tisfeins

direction as compared to that produced by one antenna element alone. In other directions the fields will more or less oppose each other, giving a reduction in field strength. Thus the power gain in the desired direction is secured at the expense of apower reduction in other directions.
Besides spacing between elements, the instantaneous direction of current flow (phase) in individual elements determines the directivity and power gain. There are several methods of arranging the elements. If they are strung end to end so that all lie on the same straight
line, the elements are said to be collinear. If they are parallel and all lying in the same plane, the elements are said to be broadside when the phase of the current is the same in all, and end-fire when the currents are not
in phase. Elements which receive power from the transmitter through the transmission line are called driven elements.
The power gain of a directive system increases with the number of elements. The proportionality between gain and number of elements is not simple, however, but depends upon the effect of the spacing and phasing upon the radiation resistance of the elements, as well as upon their number.
Collinear arrays -- Simple forms of collinear arrays, with the current distribution, are shown in Fig. 1038. The two-element array at A is popularly known as "two half-waves
in phase." It will be recognized as simply a center-fed antenna operated at its second harmonic. The way in which the number of elements may be extended for increased directivity and gain is shown in Fig. 1038-B. Note that quarter-wave transmission lines are used between each element; these give the reversal in phase necessary to make the currents in individual antenna elements all flow in the same direction at the same instant. Another way of looking at it is to consider that the whole system is a long wire with alternate half-wave sections folded so that they do not radiate. Any phase-reversing section may be
used as a quarter-wave matching section for attaching a non-resoffant feeder (� 10-8), or a resonant transmission line may be substituted for any of the quarter-wave sections. Also, the antenna may be end-fed by any of the systems
previously described (� 10-7, 10-8) or any

Feeder

Fig. 1039 -- The broadside array using half-wave

elements. Arrows indicate direction of current flow. The transposition in feeders is necessary to bring the an-

tenna currents in phase. Any reasonable number of

elements may be used. The array is hi-directional per-

pendicular to the plane of the antenna; i.e., perpendicu-

larly through this page.

�

element may be center-fed. It is best to feed at the center of the array so that the energy will be distributed as uniformly as possible among the elements.
The gain and directivity depend upon the number of elements and their spacing, centerto-center. This is shown by Table III. Al-
though 3%-wave spacing gives greater gain, it is difficult to construct a suitable phasereversing system when the ends of the antenna elements are widely separated. For this reason the half-wave spacing is generally used.

TABLE III THEORETICAL GAIN OF COLLINEAR HAI-F-WAVE
ANTENNAS

Spacing Between Centers of Adjacent
Half Waves

a. Number ofHalf Waves
in Array vs. Gain in

$ 6

3. Wave
% Wave

1.8 3.3 4.5 5.3 6.2 3.2 4.8 6.0 7.0 7.8

Collinear arrays may be mounted either horizontally or vertically. Horizontal mounting gives horizontal directivity, with vertical directivity the same as for asingle element at
the same height. Vertical mounting gives the same horizontal pattern as a single element, but concentrates the radiation at low angles. It is seldom possible to use more than two elements vertically at frequencies below 14 Mc. because of the height required.

Fig. 1038 -- Collinear half-wave antennas in phase. The system at A is generally known as "two-half-waves in phase." B is an extension of
the system; in theory it may be carried on indefinitely, but practical considerations usually limit the number of elements to four.

187 CHAPTER TEN

DheRadio -AmateuA -ilancgooh

Broadside arrays -- Parallel antenna elements with currents in phase may be combined as shown in Fig. 1039 to form abroadside array, so named because the direction of maximum radiation is broadside to the plane containing the antennas. Again the gain and directivity depend upon the number of elements and the spacing, the gain for different spacings being shown in Fig. 1040. Half-wave spacing is generally used, since it simplifies feeding when the array has more than two
elements. Table IV gives theoretical gain as a
function of the number of elements. Broadside arrays may be suspended either
with the elements all vertical or with them horizontal and one above the other (stacked). In the former case the horizontal pattern is quite sharp while the vertical pattern is that of one element alone. If the array is suspended horizontally the horizontal pattern is that of one element while the vertical pattern is sharp, giving low-angle radiation.
Broadside arrays may be fed either by resonant transmission lines (� 10-7) or through quarter-wave matching sections and nonresonant lines (� 10-8). In Fig. 1039, note the "crossing over" of the feeder, necessary to bring the elements in proper phase relationship.
Combined broadside and collinear arrays -- Broadside and collinear arrays may be combined to give both horizontal and vertical directivity, as well as additional gain. The general plan of constructing such antennas is shown in Fig. 1041. The lower angle of radiation resulting from stacking elements in the vertical plane is desirable at the higher frequencies. In general, doubling the number of elements in an array by stacking will raise the gain 2 to 4 db., depending upon whether vertical or horizontal elements are used -- that is, whether the stacked elements are broadside or collinear.
The arrays in Fig. 1041 are shown fed from
5
l,,,.a--rO(Euntdo-FfiPrhea)se
4
gfirnoaPdhstaislee)
3
2

(A)
1 /2
Fig. 1041-- Combination broadside and collinear arrays. A, with vertical elements; B, with horizontal elements. Both arrays give low-angle radiation. Two or more sections may be used.
The gain in db. will be equal, approximately, to the sum of the gain for one set of broadside elements (Table IV) plus the gain of one set of collinear elements (Table III). For example, in A each broadside set has four elements (gain 7db.) and each collinear set two elements (gain 1.8 db.) giving a total gain of 8.8 db. In B each broadside set has two elements (gain 4 db.) and each collinear set three elements (gain 3.3 db.) making the total gain 7.3 db. The result is not strictly accurate because of mutual coupling between elements, but is good enough for practical purposes.
one end, but this is not especially desirable in the case of large arrays. Better distribution of energy between elements, and hence better all-around performance, will result when the feeders are attached as nearly as possible to the center of the array. Thus in the 8-element array at A the feeders could be introduced at the middle of the transmission line between the second and third set of elements, in which case the connecting line would not be transposed. Or the antenna could be constructed with the transpositions as shown and the feeder connected between the adjacent ends of either the second or third pair of collinear elements.
A four element array of the general type shown at B is frequently used. It is shown, with the feed point indicated, in Fig. 1042.
End-fire arrays -- Fig. 1043 shows a pair of parallel half-wave elements with currents out of phase. This is known as an end-fire array
TABLE IV
THEORETICAL GAIN VS. NUMBER OF BROADSIDE ELEMENTS (HALF-W AVE SPACING)

1/8

1 / 4

3/8

3/4

ELEMENT SPACING

Fig. 1040 -- Gain vs. spacing for two parallel half. wave elements.

188 CHAPTER TEN

No. of Elements 2 3 4 5 6

Gain 4db.
5.5 db. 7db. 8 db. 9 db.

Antenna Seterno

because it radiates best along the line of the antennas, as shown.
The end-fire array may be used vertically or horizontally (elements at the same height) and is well adapted to amateur work because it gives maximum gain with relatively close element spacing. Fig. 1040 shows how the gain varies with spacing. End-fire elements may be combined with additional collinear and broadside elements further to increase the gain and directiv ity.
Either resonant or non-resonant lines may be used with this type of array, the latter being

...�������

�����

(A)

(B)

Fig. 1043 --End-fire arrays. They are shown with half-wave spacing to illustrate feeder connections. In practice, closer spacings are desirable, as shown by Fig. 1040. Direction of maximum radiation is shown by the large arrows.

lines between parallel elements can be calculated from the formula

Length of half- 492 X 0.975

480

wave line (feet) Freq. (Mc.) Freq. (Mc.)

feed
Fig. 1042 -- A four-element combination broadside. collinear popularly known as the "lazy H" antenna. A closed quarter-wave stub may be used at the feed point to match into a600-ohm line, or resonant feeders may be attached at the point shown. The gain over ahalf-wave antenna is 5to 6db.
preferably matched to the antenna through a quarter-wave matching section (� 10-8).
Checking phasing-- Figs. 1041 and 1043 illustrate a point in connection with feeding a phased antenna system which sometimes is confusing. In Fig. 1043 when the transmission line is connected as at A there is no crossover in the line connecting the two antennas, but when the transmission line is connected to the center of the connecting line the crossover becomes necessary (B). This is because in B the two halves of the connecting line are simply branches of the same line. In other words, even though the connecting line in B is a half-wave in length, it is not actually a
half-wave line but two quarter-wave lines in
parallel. The same thing is true of the untransposed line of Fig. 1041. Note that under these conditions the antenna elements are in phase when the line is not transposed, and out of phase when the transposition is made. The opposite is the ease when the half-wave line simply joins two antenna elements and does not have the feed line connected to its center, as in Fig. 1039.
Adjustment of arrays--With arrays of the types just described, using half-wave spacing between elements, it will usually suffice to make the length of each element that given by the equation for a half-wave antenna in �10-2, while the half-wave phasing

The spacing between elements can be made equal to the length of the phasing line. No special adjustments are needed provided the formulas are followed carefully.
With collinear arrays of the type shown in Fig. 1038-B the same formula may be used for the element length, while the quarter-wave phasing section can be calculated from Equation 7(� 10-5). If the array is fed at its center it will not be necessary to make any particular adjustments, although if desired the whole system may be resonated by connecting an r.f. ammeter in the shorting link on each phasing section and moving the link back and forth to find the maximum current position. This refinement is hardly necessary in practice so long as all elements are the same length and the system is symmetrical.
Simple arrays--Several simple directive antenna systems using driven elements are in rather wide use among amateurs. They are
shown in Fig. 1044. Tuned feeders are assumed in all cases; however, a matching section (� 10-8) readily can be substituted if a nonresonant transmission line is preferred. Dimensions given are in terms of wavelength; actual lengths readily can be calculated from the equations in �10-2 for the antenna and Equation 7 (� 10-5) for the resonant transmission line or matching section. In cases where the transmission line proper connects to the midpoint of aphasing line, only half the length of the latter is added to the line to find the quarter-wave point.
At A and B are two-element end-fire arrangements using close spacing. They are electrically equivalent; the only difference is in the method
of connecting the feeders. B may also be used as a four-element array on the second harmonic, although the spacing is not quite opti-
mum (Fig. 1040) in that case. A close-spaced four-element array is shown at C. It will give

18 9 CHAPTER TEN

.7h, leach. -Amateur Jiandtooh

about 2 db. more gain than the two-element array. The antenna at D is designed to take advantage of the greater gain possible with collinear antennas having greater than halfwave center-to-center spacing, but without introducing feed complications. The elements are made longer than ahalf wave to bring this
Yz

/4a

064A
(D)

0.64A. Y--r-- 0 II A

X-1
Fig. 1044 -- Simple directive sy stems. A, two-element end-fire array; B, same with center feed, which permits use of the array on the second harmonic, where it becomes afour-element array with quarter-wave spacing
C, four-element end-fire array with %-wave spacing.
D, extended in-phase antennas ("extended doubleZepp"). The gain of A and B is slightly over 4 db. On the second harmonic, B will give about 5db. gain. With
C, the gain is approximately 6db., and with D, approxi-
mately 3db. In the first three, the phasing line contributes about
1/16th wavelength to the transmission line; when B is used on the second harmonic this contribution is Yg wavelength. Alternatively, the antenna ends may be bent to meet the transmission line, in which case each feeder is simply connected to one antenna. In D, points 1C -Y indicate a quarter-wave point (high current) and X-X ahalf-wave point (high voltage). The line may be extended in multiples of quarter-waves, if resonant feeders are to be used.
A, B, and C may be suspended on wooden spreaders.
The plane containing the wires should be parallel to the ground.
about. The gain is 3db. over asingle half-wave antenna, and the broadside directivity is quite sharp.
The antennas of A and B may be mounted either horizontally or vertically; horizontal suspension (with the two elements in a plane parallel to the ground) is recommended, since this tends to give low-angle radiation without an unduly sharp horizontal pattern. Thus these systems are useful for coverage over a wide horizontal angle. The system at C, when

mounted horizontally, will have a sharper horizontal pattern than the two-element arrays.
� 10-13 DIRECTIVE ARRAYS WITH P Olt ASITIC ELEMENTS
Parasi tic excitation -- The antenna arrays described in �10-12 are bi-directional; that is, they will radiate both to the "front" and the "back" of the antenna system. If radiation is wanted in only one direction (for instance, north only, instead of north-south) it is necessary to use different element arrangements. In most of these thiadditional elements receive power by induction or radiation from the driven element, generally called the "antenna," and reradiate it in the proper phase relationship to achieve the desired effect. They are called parasitic elements, as contrasted to driven elements which receive power directly from the transmitter through the transmission line.
The parasitic element is called a director when it reinforces radiation on a line pointing to it from the antenna, and is called areflector when
the reverse is the case. Whether the parasitic element is adirector or reflector depends upon the parasitic element tuning (which usually is adjusted by changing its length) and, particularly when the element is self-resonant, upon the spacing between it and the antenna.
Gain vs. spacing -- The gain of an antenna-

A)
...../

L 1
A4ee4 8
..,.. ,..........

*......._....,..�

tion Resistance 20

-25

I 1 l

0 0.05 0.1 0 5 02 0.25 0.3 0.35 0.4

ELEMEN rSPACING -WAVELENGTH

Fig. 1045 -- Gain vs. element spacing for an antenna and one parasitic element. Zero db. is the field strength from ahalf-wave antenna alone. Greatest gain is in the direction A at spacings less than 0.14 wavelength; in direction B at greater spacings. Front-to-back ratio is
the difference in db. between curves A and B. Variation
in radiation resistance of the driven element also is shown. These curves are for self-resonant parasitic element. At most spacings the gain as areflector can be increased by slight lengthening of the parasitic element; as
a director, by shortening. This likewise improves the front-to-back ratio.

190 CHAPTER TEN

_Antenna Sy�lem�

reflector or antenna-director combination varies chiefly with the spacing between elements. The way in which gain varies with spacing is shown in Fig. 1045, for the special case of self-resonant parasitic elements. This
chart also shows how the attenuation to the "rear" varies with spacing. The same spacing
does not necessarily give both maximum forward gain and maximum backward attenuation. Backward attenuation is desirable when the antenna is used for receiving, since it greatly reduces interference coming from the
opposite direction to the desired signal. Element lengths -- The antenna length is
given by the formulas in �10-2. The director and reflector lengths must be determined experimentally for maximum performance. The preferable method is to aim the antenna at a receiver a mile or more distant and have an observer check the signal strength (on the "S" meter) while the reflector or director is
adjusted afew inches at atime, until the length which gives maximum signal is found. The attenuation may be similarly checked, the length being adjusted for minimum signal. In
general, the length of adirector will be about 4% less than that of the antenna, for best
front-to-back ratio. The reflector will be about 5% longer than the antenna.
Simple systems -- the rotary beam -- Practical combinations of antenna, reflector and director are shown in Fig. 1046. Spacings for maximum gain or maximum front-to-back ratio (ratio of power radiated in the desired
direction to power radiated in the opposite direction) may be taken from Fig. 1045. In the chart, the front-to-back ratio in db. will be the sum of gain and attenuation at the same spacing.
Systems of this type are popular for rotary beam antennas, where the whole antenna is rotated to permit its gain and directivity to be utilized for any compass direction. They may be mounted either horizontally (plane containing the elements parallel to the earth) or vertically.
Arrays using more than one parasitic ele-
ment, such as those shown at C and D in Fig. 1046, will give more gain and directivity than
is indicated for the single reflector and director by the curves of Fig. 1045. The gain with a properly adjusted three-element array (antenna, director and reflector) will be 5 to 7 db. over ahalf-wave antenna, while somewhat higher gain still can be secured by adding a second director to make afour-element array. The front-to-back ratio is correspondingly improved as the number of elements is increased.
The elements in close-spaced (less than onequarter wavelength element spacing) arrays preferably should be made of tubing of half-

Di Ft ANT
ANT

0 IX

0 ,5�

D1R t ANT
REF

Ole'

Cl or RISA

CAR

ERR

NIA

t ANT REF

O IA

4 4

Olar015�

Fig. 1046 -- Half-wave antennas with parasitic elements. A, with reflector; B, with director; C, with both director and reflector; D, two directors and one reflector. Gain is approximately as showtrby Fig. 1045 in the first two cases and depends upon the spacing and length of the parasitic element. In the three- and four-element arrays a reflector spacing of 0.15 wavelength will give slightly more gain than 0.1-wavelength spacing. Arrows show direction of maximum radiation. The array should be mounted horizontally (these are top views).

to one-inch diameter both to reduce the ohmic resistance (� 10-2) and to secure mechanical rigidity. If the elements are free to move with
respect to each other the array will show detuning effects in awind.
Feeding close-spaced arrays -- While any of the usual methods of feed may be applied to the driven element of aparasitic array, the fact that with close spacing the radiation resistance as measured at the center of the driven
element drops to avery low value makes some systems more desirable than others. The preferred methods are shown in Fig. 1047. Reso-
nant feeders are not recommended for lengths greater than ahalf wavelength.
The quarter- or half-wave matching stubs shown at A and B in Fig. 1047 preferably
should be constructed of tubing with rather close spacing, in the manner of the "Q" section. This lowers the impedance of the match-
ing section and makes the position of the line taps somewhat less difficult to determine
accurately. This line adjustment should be made only with the parasitic elements in
place, and after the correct element lengths have been determined should be checked to compensate for changes likely to occur because of element tuning. The procedure is the same as that described in �10-8.

191 CHAPTER TEN

Wm.
./he leach� Am ateur IIJiandiooh

The concentric-line matching section at will work with fair accuracy into aclose-spaced parasitic array of 2, 3 or 4 elements without necessity for adjustment. The line is used as an impedance inverting transformer, and if its characteristic impedance is 70 ohms will give an exact match to a 600-ohm line when the resistance at the termination is about 8.5 ohms. Over a range of 5 to 15 ohms the mismatch, and therefore the standing-wave ratio, will be less than 2 to 1. The length of the quarterwave section should be calculated from Equation 7 (� 10-5).
The delta matching transformer shown at D is an excellent arrangement for parasitic arrays, and probably is easier to install, mechanically, than any of the others. The positions of the taps (dimension a) must be determined experimentally, along with the length b, by checking the standing-wave ratio on the line as adjustments are made. Dimension bshould be about 15% longer than a.
60o-ohm A
---..-.s.t.Suhbenaqdrjoudsstfmoernt
/me
6n0n0 e-ohm
Fig. 1047 -- Recommended methods of feeding the driven antenna element in close-spaced parasitic arrays. The parasitic elements are not shown. A, quarter-wave open stub; B, half-wave closed stub; C, concentric-line quarter-wave matching section; D, delta matching transformer.

Sharpness of resonance-- Peak perform-
ance of amulti-element directive array depends upon proper phasing or tuning of the elements, which in all but the simplest systems can be exact for one frequency only. However, there is some latitude, and most arrays will work well over arelatively narrow band such as 14 Mc. If frequencies in all parts of the band are to be used, the antenna system should be designed for the mid-frequency; on the other hand, if only one frequency in the band will be used the greater portion of the time the antenna might be designed for that frequency and some degree of misadjustment tolerated on the occasionallyused spare frequencies.
When reflectors or directors are used the tolerance is usually less than in the case of driven elements, partly because the parasitic-element lengths are fixed and the operation may change appreciably as the frequency passes from one side of resonance to the other, and partly because the close spacing ordinarily used results in a sharp-tuning system. With parasitic elements operation should be confined to asmall region about the frequency for which the antenna is adjusted, if peak performance is to be secured.
Combination arrays -- It is possible to combine parasitic elements with driven elements to form arrays composed of collinear driven and parasitic elements and combination broadside-collinear-parasitic elements. Thus two or more collinear elements might be provided with acollinear reflector or director set, one parasitic element to each driven element. Or both directors and reflectors might be used. A broadside-collinear array could be treated in the same fashion.
When combination arrays are built up, a rough approximation of the gain to be expected may be obtained by adding the gains for each type of combination. Thus the gain of two broadside sets of four collinear arrays with aset of reflectors, one behind each element, at quarter-wave spacing for the parasitic elements, would be estimated as follows: From
Table III, the gain of four collinear elements is 4.5 db. with half-wave spacing; from Fig. 1040
or Table IV, the gain of two broadside elements at half-wave spacing is 4.0 db.; from Fig. 1045 the gain of aparasitic reflector at quarter-wave spacing is 4.5 db.; the total gain is then the sum, or 13 db. for the sixteen elements. Note that using two sets of elements in broadside is equivalent to using two elements, so far as gain is concerned, similarly with sets of reflectors as against one antenna and one reflector. The actual gain of the combination array will depend, in practice, upon the way in which the power is distributed between the various elements, and upon the effect of mutual coupling between elements upon the

192 CHAPTER TEN

-Antenna Seient�

radiation resistance of the array, and may be somewhat higher or lower than the estimate.
A great many directive antenna combinations can be worked out by combining elements according to these principles.

good ground connection because of sandy soil
or other considerations, it is preferable to use a counterpoise or capacity ground instead of
an actual ground connection. The counterpoise consists of a system of wires insulated from

� 10-14 MISCELLANEOUS ANTENNA SYSTEMS
Grounded antenna -- The grounded antenna is used almost exclusively for 1.75-Mc. work, where the length required for a halfwave antenna would be excessive for most lo-

ground running horizontally above the earth beneath the antenna. The counterpoise should
have asufficient number of wires of sufficient length to cover well the area immediately under the antenna. The wires may be formed into any convenient shape, i.e., they may be spread

cations. An antenna worked "against ground" need be only a quarter-wave long, approxi-

Fig. 1048--Typical grounded antenna for 1.75 Mc., consist

mately, because the earth acts as an electrical "mirror" which supplies the missing quarter wave. The current at the ground connection with aquarter-wave antenna is maximum, just

ing of a vertical section and horizontal section having a total length (including the ground lead if the latter is more than afew feet long) of

as it first the center of a half-wave antenna. On 1.75 Mc. the most useful radiation is
from the vertical part of the antenna, since

one-quarter wavelength. Coil L should have about 20 turns of No. 12 on athree-inch di-
ameter form, tapped every

vertically-polarized waves are characteristic of ground-wave transmission. It is therefore desirable to make the down-lead as nearly vertical as possible, and also as high as possible.

two or three turns for adjustment. C is 250 to 500 ,dd. variable. The inductive coupling between L and the final tank coil should be variable.

This gives low-angle sky-wave transmission

which is most useful for long-distance work at night, in addition to agood ground wave for local work. The horizontal portion contributes to high-angle sky-wave transmission, which is
useful for covering short distances on this band at night.
Fig. 1048 shows a grounded antenna with the top folded to make the length equal to a quarter wave. The antenna coupling apparatus

out fan-shape, in a radial pattern, or three or
more parallel wires separated a few feet running beneath the antenna may be used. The counterpoise should be elevated six or seven feet above the ground so it will not interfere with persons walking under it. Connection is made between the usual ground terminal of the transmitter and each of the wires in the counterpoise.

consists of the coil L, tuned by the series condenser C, with L inductively coupled to the transmitter tank circuit (� 10-4, 10-6).
For computation purposes, the overall length of agrounded system is given by

"J" antenna -- This antenna, frequently used on ultra-high frequencies when vertical polarization is desired, is simply a half-wave radiator fed through aquarter-wave matching
section, (� 10-8), the whole being mounted

L (feet) =236
f(Mc.)
This is the total length from the far end of the antenna to the ground connection. The length is not critical, since departures of the order of 10% to 20% can be compensated by the tuning apparatus.
The ground should preferably be one with conductors buried deep enough to reach natural moisture. In urban locations, good grounds can be made to water mains where they enter the house; the pipe should be scraped clean and a low-resistance connection made with a tightly-fastened ground clamp. If no waterpipes are available several pipes, six to eight feet long, may be driven into the ground at intervals of six or eight feet, all being connected together. The transmitter should be located so as to make the ground lead as short as possible.
In locations where it is impossible to secure a

vertically as shown in Fig. 1049. Adjustment
and tuning are as described in �10-8. The bottom of the matching section, being practically at zero r.f. potential, can be grounded through a metallic conductor for lightning protection.
Coaxial antenna-- With the "J" antenna there is likely to be some radiation from the
matching section and transmission line which tends to combine with the radiation from the antenna in such away as to raise the angle of radiation. As this is undesirable on ultra-high frequencies where the lowest possible radiation angle is essential, the coaxial antenna shown in Fig. 1050 was developed to eliminate feeder radiation. The center conductor of a 70-ohm concentric transmission line is extended one quarter wave beyond the end of the line to act as the upper half of a half-wave antenna, the lower half being supplied by the quarterwave sleeve, the upper end of which is connected to the outer conductor of the eoncen-

193 CHAPTER TEN

Radio Amaieur '3 ilandlooh

trio line. The sleeve acts as ashield about the transmission line and very little current is
induced on the outside of the line by the antenna field. The line is non-resonant, since its characteristic impedance is the same as the center impedance of the half-wave antenna

TLriannes.

Fig. 1049 --The "J" antenna. It is usually constructed of metal tubing; frequently with the O%-wave vertical section shown an emxetteanlsimoanst.ofThae sgtruobunmdaeyd be adjusted by a sliding shorting bar.

\- Maybe grounded here

(� 10-5) is shown in Fig. 1051. Essentially, it consists of acenter-fed half-wave antenna with another half-wave element connected directly between its ends. The spacing between the two sections should be quite close -- not more than a few per cent of the wavelength. As used at ultra-high frequencies, the spacing is of the order of an inch or two with elements
constructed of metal tubing. The impedance at the terminals of the an-
tenna is four times that of ahalf-wave antenna, or nearly 300 ohms, when the antenna conductors are all the same diameter. A 300-ohm line will therefore be non-resonant when the antenna is connected to its output end (� 10-5), while the standing-wave ratio with a600-ohm
line will be only 2to 1. The total length around the loop formed by
the antenna may be calculated by Equation
10 (� 10-9) for atotal length of one wavelength. Corner reflector antenna-- A type of an-
tenna system particularly well-suited to the u.h.f. ranges above 56 Mc., is the "corner" reflector shown in Fig. 1052. It consists of
two plane surfaces set at an angle of 90�, with the antenna set on aline bisecting this angle.

(� 10-2). The sleeve may be made of copper or
brass tubing of suitable diameter to clear the transmission line. The coaxial antenna is somewhat difficult to construct, but is superior to simpler systems at low radiation angles.
Folded dipole -- An arrangement which combines the radiation characteristics of a half-wave antenna with the impedance-trans-
forming properties of a quarter-wave line

TronLsirnne.ssion Fig.1051-- Folded dipole for increasing the value of impedance at the feed point.

TiMReotdal /nsulator ocCtfoooclnnooidnnnuuecetcceetntroterrdic SMleeteav/e
7co0n-coehntmric //ne

Fig. 1050--Coaxial anotfenntah.e T7h0e-oihnmnerlinceonidsucctoonrnected to the quarter-wave metal rod which forms the upper half of the antenna.

The distance of the antenna from the vertex
should be 0.5 wavelength, but some compromise designs can be built with closer spac-
ings (see Table V). The plane surfaces do not need to be solid, and can most easily be made of spines spaced about 0.1 wavelength apart. The spines do not have to be connected
together electrically. The resistance of the antenna is raised when
a corner reflector is used. The transmission line should be run out at the rear of the reflector to keep the system as symmetrical as
possible and thus avoid any unbalance. Two simple antennas which can be used with the corner reflector are shown in Fig. 1053.
The corner reflector can be used with the antenna either horizontal or vertical, and the plane of polarization will be the plane of the antenna. The relative positions of the antenna and reflector must remain the same,
however, which means that asupport for both horizontal and vertical polarization would require a means for rotating the reflector
about its horizontal axis. Receiving antennas -- Because of the high
sensitivity of modern receivers alarge antenna

194 CHAPTER TEN

is not necessary for picking up signals at good strength. Often it will be found that an indoor wire only 15 or 20 feet long will give quite good results, although alonger wire outdoors is better.
The use of atuned antenna greatly improves the operation of the receiver because the signal

Spacing of driven dipole
to vertex (2'2'1

Reflector elements

his reflector element at
vertex

Feed line
/Page (opt/one
Side member of mod or metal

Sdpeacmienngtest

reflector (r)

4 Onven �dement or dipole Crossmember
Medea .'" side member

Re/lector element �-lie wire

feed bne to tronseutter.._
Driven element or
dipole

Shhemember wood (ornsetal)
ellenegmtehntelsr<e2f.l-e7c)tor
Standoff insulator

_Antenna Sice.item3

TABLE V

Frequency Band

Length Lon th of

of Side

tor
Elements

Number Re i:enct Elemen�4

SRpadcielengoorf

Spacing of Driven

Elements toE rporleon

224-230 Mc.

(1% meter) 4' 2"

4' 7"

112-116 Mc.
(2% meter)

8,4,,

5,2,,

5"
10"

2' 2"
4' 4"

112-116 Me.' AtRrt (2h meter) " -

Ai 9/, - -

Dye

y 6"

5(trejeotUo re). 16' 8" 10' 4"

1' 8" 8' 8"

56-60 Mc. (5 meter)

13' 4"

10' 4"

1' 8" 6' 11"

Table V.-- Dimensions of square-corner reflector for the 224-, 112-, and 56-Mc. bands. Alternative design. are listed for the 112- and 56-Mc. bands. These designs marked (5), have fewer reflector elements and shorter aides, but the effectiveness is only slightly reduced There is no reflector element at the vertex in any of the designs.

the receiver to the transmitter while the transmitter is on the air. The directive effects and power gain of directive transmitting antennas are the same for receiving as for transmitting, and should be utilized for best reception.

Fig. 1052-- A corner reflector antenna system with grid-type reflector. The reflector elements are stiff wire or tubing. The dimensions are for 224 Mc., and should � be doubled for 112 Mc. (See Table V.) The gain of the system is close to 10 db.

strength is greater in proportion to the stray noises picked up by the antenna than is the case with the antenna of random length. Since the transmitting antenna is usually given the best location, it can be used to great advantage for receiving, especially when a directive antenna is used. A change-over switch or relay connected in the antenna leads can be used to transfer the connection from

Two wire bee tuned at transmitter

450 (0 Ohm

h5a0g0/

OW /2 wfre,

spaced 2"I

s A
Fig. 1053 -- Dipoles suitable for use with the corner reflector antenna system. The len gth L is 25 inches for 224 Mc., s��� 1inch for the same band.

195 CHAPTER TEN

CHAPTER ELEVEN

Radio equipment

� A ONE-TUBE REGENERATIVE RECEIV ER

T HE SIMPLEST receiver capable of

giving at all satisfactory results in everyday

operation is one consisting of a regenerative

detector followed by an audio amplifier. This

type of receiver is sufficient for headphone re-

ception, and is quite easy to build and adjust.

A dual tube may be used for both stages,

thereby reducing cost. Figs. 1101 to 1105 show such a receiver,

using a 6C8G twin-triode tube, one triode

section being the regenerative detector and

the other the audio amplifier. The circuit diagram is given in Fig. 1103. The grid coil, Li,

el is tuned to the frequency of the incoming
signal by means of condensers Ci and Ca,

being the bandsetting or general coverage

condenser and C3 ,the bandspread condenser.

Regeneration is supplied by means of the

tickler coil L2;the variable plate by-pass con-
denser, C2, is the regeneration control. The receiver is coupled to the antenna through Cr,, alow-capacity trimmer condenser. Ri and C4 are the grid leak and grid condenser.
The audio amplifier section of the tube is coupled to the detector by the audio transformer T1. Bias for the audio stage is supplied by a midget flashlight cell, this type of bias being quite convenient as well as cheaper than other methods. The choke, RFC, is necessary

Fig. 1102 -- A rear view of the one-tube receiver. The grid condenser and grid leak are supported by their wire leads between the stator plates of the tuning condenser and the grid cap on the tube.
to prevent r.f. current from flowing in the primary winding of the audio transformer; without the choke the regeneration control condenser Cr, may be ineffective. A switch, Sr, is provided for turning off the "B" supply when transmitting.
The construction of the re-

ceiver is shown in the photo-

graphs. The chassis measures

5% by 9% by 1% inches. The

three variable condensers are

mounted on the panel three

inches from the bottom edge,

with C3 in the center, C1 at the

right and Cr, at the left. All

ground connections may be

made directly to the chassis,

making sure that the paint is

scraped away and that good con-

tact is secured. The headphone connections

are made by means of tip jacks

mounted on the rear edge of

the chassis. Filament and plate

Fig. 1101 -- A one-tube regenerative receiver, using adouble triode as a regenerative detector and audio amplifier. Plug-in coils for different frequency ranges are shown in front of the receiver.

power are brought in through a four-wire cable which enters the chassis through the rear edge.

196 CHAPTER ELEVEN

Fig. 1103 -- Bottom of chassis view of the one-tube regenerative receiver. The construction and wiring are extremely simple. The power supply cable and the headphone tip jacks are brought out at the rear of the chassis, and the antenna and ground terminals are mounted on one side.

Radio equipmenl

The coils are made as shown in Fig. 1105 and the coil table. Both windings should be in
the same direction. L1for the B, C and D coils should have its turns evenly spaced to occupy the specified length; the wire may be held in
place when the coil is finished by running some Duco cement along the ridges of the coil forms.
The heater supply for the receiver may be either a 6.3-volt filament transformer (the 1-ampere size will be ample) or a 6-volt battery. A 45-volt "B" battery should be used for the plate supply. The "B" current drain is only a few milliamperes, and a medium- or small-size "B" battery will give excellent service.
After the wiring has been checked, the heater and plate supplies, headphones and antenna and ground can be connected to the receiver, and the C coil plugged in. Turn the regeneration condenser, C2, starting from mini-
mum capacity (plates all out) until the set

late, the coil L2 must be moved nearer to L1 or, in extreme cases, a turn or two must be added to L2. This is best done by rewinding with more turns rather than by trying to add aturn or two to the already-wound coil. For any given band of frequencies, adjust C6 so that the detector oscillates over the whole range, using as much capacity at C5 as is possible. This will give the best compromise between dead spots and signal strength. It will be found that less advancing of the regeneration control, C2, is required at the high-frequency end of a coil range (C1 at or near minimum capacity) than at the low-frequency end. The best adjustment of the antenna condenser, C5, and the feedback coil, L2, is that which requires almost amaximum setting of the regeneration control at the low-frequency end (maximum capacity of C1) of any coil range.
ONO. ANt

goes into oscillation. This phenomenon is

easily recognizable by adistinct click, thud or

1.7

hissing sound. The point where oscillation just begins is the most sensitive operating point at that particular dial setting.

The tuning dial may now be slowly turned,

the regeneration control knob being varied simultaneously (if necessary) to keep the set

just oscillating. A number of stations will prob-

ably be heard. A little practice will make tuning easy.

If the set refuses to oscillate, the sensitivity

will be poor and no code signals will be heard on

the frequencies at which such signals should be expected, It should oscillate easily, however, if the coils are made exactly as shown. It some-
times happens that the antenna takes so much energy from the set that it cannot oscillate,
this usually resulting in "holes" in the range where no signals can be picked up (and where the hissing sound cannot be obtained). This can be cured by reducing the capacity of C5 (unscrewing the adjusting screw) until the detector again oscillates. If it still refuses to oscil-

Fig. 1104 -- Circuit diagram of the one-tube regenerative receiver.
C1 -- 100-,pfd. band-set variable. C2 -- 100-ourfd. regeneration control variable. C3 - 15-ererfd. band-spread variable. C4 - 100-irerfd. mica grid condenser. C3 - 3-30-iridd. adjustable mica antenna coupling. RI-- 1-megohm, j�-watt grid leak. Li, 1,2 -- Grid coil and tickler coil. See coil table for
dimensions. Tt-- Interstage audio transformer, 3:1 ratio. Si -- S.p.s.t. B+ toggle switch. RFC -- 2.5-mh. detector plate r.f. choke.

197 CHAPTER ELEVEN

Dh e leadio Amateur '� ilancido�

ONE-TUBE REGENERATIVE RECEIVER COIL DATA

Coil

Grid Winding (Li)

Tickler (La)

56 turns No. 22 enamelled

15 turns No. 24 enamelled
5" " "

Ali coils wound on 1%-inch diameter forms (Hamrnarlund SWF-4> Grid windings on coil. B, C and D spaced to occupy alength of 11/2
incites; grid winding on coi! A close-wound. Tickler coils all closewound, spaced 1/8 inch from bottom of grid winding. See Fig. 1105.

Frequency range Coil: A -- 1700 to 3200 kc. B -- 3000 to 5700 kc. C -5400 to 10,000 kc.
1) -9500 to 18,000 kc.

Coil A misses the high-frequency end of the

BOTTOM OF SOCKET OR COIL FORM

broadcast band, but it is possible to hear police stations and other services. The band is most easily located by listening at night (when there is the most activity), setting C3 at maxi-

Fig. 1105 -- Method of winding coils for the one-tube regenerative receiver. Pin 1connects to ground, pin 2to the plate of the detector, pin 3 to RFC and the stator plates of C2, and pin 4to.the stator plates of CIand Ca.

mum and slowly tuning with C1 until some of the police stations are heard. These stations

heating radiator or water piping is usually

operate on 1712 kc., so that once found they good.

become "markers" for the low-frequency end of the band. Further tuning then should be done with the main tuning dial.
Locating the other ranges is done in exactly the same manner, by searching carefully with C1 and looking for marker stations of known frequencies. Activity will be greater above 12 Mc. during the day and below that frequency
at night. A suitable antenna for the receiver would be
50 to 75 feet long, and as high and clear of surrounding objects as possible. The ground lead should preferably be short; a ground to a

� A REGENERATIVE SINGLE-SIGNAL
RECEIVER
An inexpensive amateur-band receiver using i.f. regeneration for single-signal reception, is shown in Fig. 1106. Fig. 1108 gives the circuit diagram. Regeneration also is used in the mixer circuit to improve the signal-to-image ratio and to give added gain. This receiver is designed to give the maximum of performance, in the hands of acapable operator, at minimum cost. Selectivity, stability and sensitivity are primary considerations.
The mixer, a6SA7, is coupled to the

antenna and is separately excited by

a6J5 oscillator. There is asingle 460-

ke. i.f. stage, using a 6SK7 and per-

meability-tuned transformers. The second detector and first audio amplifier is a 6SQ7 and the audio output tube for loud-speaker operation is a

6F6. The separate beat oscillator circuit uses a6C5. A VR-105 voltage regulator tube is used to stabilize the plate voltage on the oscillators and the screen voltage on the mixer and

i.f. tubes. To make construction easy and to
avoid the necessity for additional

trimmer condensers on each coil, the mixer and high-frequency oscillator

Fig. 1106 -- A 7-tube superhet using regeneration to give singlesignal reception and an improved image ratio. The dial is of atype that can be directly calibrated for each amateur band. Two plug-in coils are used for each range. The chassis is 11 by 7by 2inches and the panel 7by 12 inches.

circuits are separately tuned. Main tuning is by the oscillator bandspread condenser, C3, which is operated by the calibrated dial. C2 is the oscillator band-setting condenser. The mixer circuit is tuned by CI, and regeneration in this circuit is

198 CHAPTER ELEVEN

Fig. 1101 -- Top view of the 7-tube superhet without coils in place. The band-spread tuning condenser, Ca, is at the front center; at the left is the mixer tuning condenser and at the right the oscillator band-set condenser. The oscillator tube is in the center, with the mixer tube to the left on the other side of a baffle shield. This shield, measuring 4% by 4.1;4 inches, is used to prevent coupling between oscillator and mixer. The mixer coil socket is at the left edge of the chassis, and the oscillator coil socket is next to C3.
The If. and audio sections are along the rear edge of the chassis. The transformer in the rear left corner is Ti; next to it is the If. tube, then Ta. Next in line is the 6SQ7, followed by the 6C5 beat oscillator, the h.o. transformer, T3, and finally the 6F6. The VR-105 is just in front of Ta.

6SA7

Cis 6SK7

Palio equipment

6SQ7

6F6

RFC S IC13
A/I I/Irs

vR -105 =

-250 �

Fig. 1108 -- Circuit diagram of the regenerative superhet.

Ci -- 5O-pfd. mixer tuning. Ca -- 50-ppfd. oscillator band-set. C3-- 35-ppfd. oscillator band.
spread. C4-- 50-pufd. mica oscillator cou-
pling. Ca, Cs -- 0.1-pfd. cathode by-pass. Co -- 0.1-pfd. mixer screen by-pass. C7-- 0.1-pfd. mixer plate by-pass. Ca -- 0.01-pfd. b.f.o. plate by-pass.
Cil -- 0.01-pfd. audio coupling. Cla -- 0.01-�fd. i.f. amplifier grid
by-pass. CI3 -- 0.005-pfd oscillator plate
by-pass. CI4 -- 0.005-pfd. mixer grid by-
pass. CI5 -- 3-30-ppfd. adjustable If. feed-
back. CHI -- 250-ppfd. diode by-pass. CI7 -- 100-pfd. rd. filter. CI5 -- 100-pfd. oscillator grid. CI0, C20 -- 25-pfd. electrolytic cath-

Czi -- 25-ppfd. b.f.o. tuning. C22 -- 100-ppfd. mica diode block-
ing. RI-- 200-ohm, %2-watt cathode
bias. R2 -- 20,000-ohm, 3/2-watt injec-
tion grid leak. R3 -- 50,000-ohm, 3/2-watt oscil-
lator grid leak. Rs -- 50,000-ohm, %-watt b.f.o.
plate dropping. Rb -- 50,000-ohm, ,1/-watt diode
r.f. filter. R5 -- 300-ohm, 1/2-watt cathode
bias. R7 -- 0.2-megobm, %-watt diode
load. RR-- 2000-ohm, 34-watt cathode
bias. Rs -- LO-megohtn, 2-watt a.v.c.
filter. Rio -- 0.1-megohm, j�-watt plate

R; ;-- 0.5-megohm,

att grid.

R12 -- 450-ohm, 1-watt cathode

bias.

Rua -- 75,000-ohm, 1-watt bleeder.

Ria -- 5000-ohm, 10-watt voltage

droppin g.

Ris -- 10,000-ohm mixer regenera-

tion control.

Rio -- 25,000-ohm If. gain control.

R17 -- 2-megohm audio volume

control.

R18 2-megohm, 3/2-watt a.v.e.

load.

Ti, Ta -- 465-kc. IL transformer,

permeability tuned.

T3 -- 465-kc. b.f.o. unit.

RFC -- 2.5-mh. r.f. choke.

J-- Closed-circuit jack.

Si -- S.p.s.t. B+ toggle switch.

Li-La, inc. -- See coil table.

X indicates jumper inside VR-10.;

ode by-pass.

load.

base.

199 CHAPTER ELEVEN

.74e Radio -AonaleuA -1,/ancl4004

�
Fig. 1109 -- The below-chassis Wing is shown in this view of the 7-tube superhet. The controls along the bottom edge of the panel are, from left to right, the mixer regeneration control, Ills, the i.f. gain control, Rio, the audio volume control, 1117, and the beat-oscillator tuning condenser, C21. l'he latter has one corner of one rotary plate bent over so that when the condenser plates are fully meshed the condenser is short-circuited, thus stopping oscillation.

controlled by R76, connected across the mixer tickler coil, L3.
R76 is the i.f. amplifier gain control, which also serves as an i.f. regeneration control when this stage is made regenerative. Cm is the regeneration condenser; it is adjusted to feed back a small amount of i.f. energy from the plate to the grid of the 6SK7 and thus produce regeneration. If the high selectivity afforded by i.f. regeneration is not wanted, C76 may be
omitted. Diode rectification is used in the second de-
tector circuit. One of the two diode plates in the 6SQ7 is used for developing a.v.c. voltage, being coupled through C22 to the detector diode. The detector load resistor consists of R6 and R7 in series, the tap being used for r.f. filtering of the audio output to the triode section of the tube. R78 is the a.v.c. load resistor; �R9,C14 and C12 constitute the a.v.c. filter circuit. 82 cuts the a.v.c. out of circuit by grounding the rectifier output. The headphones connect in the plate circuit of the triode section of the 6SQ7. R17is the audio volume control.
The tube heaters are all in parallel, one side of each being grounded right at the tube socket. Only one filament wire need be run from tube
to tube. In wiring the i.f. amplifier, keep the grid and
plate leads from the i.f. transformers fairly close to the chassis and well separated. Without C76, the i.f. stage should be perfectly stable and should show no tendency to oscillate at full gain.
The method of winding the coils is shown in Fig. 1110 and complete specifications are given
in the coil table. The i.f. amplifier can be aligned most con-
veniently with the aid of amodulated test oscillator. First alignment should be made with C76 disconnected so that the performance of the

amplifier non-regenerative can be checked. A headset or loud speaker can be used as an � output indicator. The mixer and oscillator coils should be out of their sockets, and R15 should be set at zero resistance.
After the i.f. amplifier is aligned, plug in a set of coils for some range in which there is likely to be a good deal of activity. Set the oscillator padding condenser, C2, at almost
maximum capacity and the mixer regeneration control, R76,for minimum regeneration -- no resistance in circuit. Connect an antenna. Switch on the beat oscillator by turning C21 out of the maximum position, and adjust the screw on T3 urtil the characteristic hiss is
heard. As C1is tuned over its range, there should be
two points where there will be a definite increase in noise and in the strength of any signals which may be heard. The peak on the low-
capacity side corresponds to the image fre-
quency, and the mixer condenser should always be tuned to the peak which occurs at the higher
capacity setting. After the signal peak on C1 has been identi-
fied, tune C3 over its whole range, following with C1 to keep the mixer circuit in tune, to see how the band fits the dial. With C2 properly set, the band edges should fall the same number of main dial divisions from 0and 100; if the band runs off the low-frequency edge, less capacity is needed at C2,while the converse is true if the band runs off the high edge. Once the band is properly centered on the dial, the panel may be marked at the appropriate
point so that C2 may be reset readily when changing bands.
To check the operation of the mixer regener-
ation, tune in a signal on C3,adjust C1 for maximum volume, and slowly advance the regeneration control, R76. As the resistance

200 CHAPTER ELEVEN

lea dio tiuipoteril

MIXER

OSCILLATOR

ANT.
TOP OF SOCKET VIEWS Fig. 1110 -- Coil and socket connections for the 7-tube superhet. The small coil inside the mixer coil must be oriented properly for regeneration to take place.
increases retune CI to maximum, since the regeneration control will have some effect on the mixer tuning. As regeneration is increased, signals and noise will both become louder and C1 will tune more sharply. Finally the mixer circuit will break into oscillation when, with C1 right at resonance, aloud carrier will be heard since the oscillations generated will go through the receiver in exactly the same way as an incoming signal. As stated before, oscillation should occur with R16 set at half to threequarters full scale. In practice, always work with the mixer somewhat below the critical regeneration point and never permit it actually to oscillate. On the lower frequencies, where images are less serious, the tuning is less critical if the mixer is non-regenerative. In this case, always set R16 at zero since there will be a range on the resistor where, without definite regeneration, the signal strength will be less than it is with zero resistance.
After the preceding adjustments have been completed the i.f. regeneration may be added. Install C15, taking out the adjusting screw and bending the movable plate to make an angle of about 45 degrees with the fixed plate. Realign the i.f. As the circuits are tuned to resonance the amplifier will oscillate, and each time this happens the gain control, R16, should be backed off until oscillations cease. Adjust the trimmers to give maximum output with the lowest setting of R16. At peak regeneration the signal strength should be about the same, de-

COIL DATA FOR 7-TURE SUPERRET

Band

Wire Coil Size Turns

Length

Tap

1.75 Mc. 3.5 Mc. 7 Me. 14 Mc. 28 Mc.

1.2

70 15 15 42 15 35 9 12 25 10 20 5 9 14 8 10 5 7 7 4 4 4 1.5
3.6 2.4

Close-wound

.11

1inch Close-wound
1inch Close-wound
1inch Close-wound
1inch Close-wound
1inch Close-wound
1inch Close-wound
1inch Close-wound

Top 18 6 2.4 1.4

All coila except L3 are 13,6 inches in diameter, wound with enamelled wire on Hammarlund SWF Forras. Spacing between Li and 14, and between Li and 14 ,approximately
inch. Band-spread taps are measured from bottom (ground) end of L4.
L3 for 28 Me. is interwound with L1 at the bottom end. L3 for all other coils is self-supporting, scramble-wound to a diameter of 54. inch, mounted inside the coil form near the bottom of Li.

spite reduced gain in the amplifier, as without regeneration at full gain. Too much gain with regeneration will have an adverse effect on selectivity.
For single-signal c.w. reception, set the beat oscillator so that when R16 is advanced to make the i.f. just go into oscillation the resulting tone is the desired beat-note frequency. Then back off on R16 to obtain the desired degree of selectivity. Maximum selectivity will be secured with the i.f. just below the oscillating point. The "other side of zero beat" will be very much weaker than the desired side.
Power supply requirements for the receiver are 2.2 amp. at 6.3 volts for the heaters and 80 ma. at 250 volts for the plates. Without the pentode output stage asupply giving 6.3 volts at 1.5 amp. and 250 volts at 40 ma. will be sufficient.
� COMMERCIAL SUPERRETEROVITNE RECEIVERS
Although some very advanced types of superheterodyne receivers have been built at home, most of the superheterodynes used by amateurs are of the manufactured type. They range from simple, inexpensive ones with a minimum of tubes to large receivers with many

201 CHAPTER ELEVEN

Dhe Radio Amaieur'� Jlancgooh

features. Two of the latter type are shown in
Figs. 1111-1116. The National HRO, shown in Fig. 1111,
1112 and 1113, is ahigh-gain superheterodyne
using plug-in coil gangs for the various frequency ranges. Coils are available for bands of frequencies between 50 kc. and 30 Mc., arange made possible only by the use of plug-in coil gangs. Standard equipment is four sets of coils for 1.7 to 30 Mc. Provision is made on the coil gangs in the high-frequency range for either general coverage or amateur bandspread, depending upon the coil connections used. Two

is set to the end of the scale. A tuning meter mounted on the panel can be switched in
during 'phone reception for comparative measurements of carrier strength. The main tuning dial is an ingenious device which must make 10 complete revolutions for 180-degree rotation of the tuning condenser gang. The numbers,
visible through small windows on the dial, change every revolution to give consecutive numbering by tens from 0to 500.
A high order of frequency and alignment stability is obtained by the use of air-dielectric trimmer and padder condensers.

stages of r.f. amplification are used

ahead of the mixer, and a separate

high-frequency oscillator is used for

greater stability than is possible with

some of the combination converter

types of tnbes. The mixer is followed

by acrystal filter and two stages of i.f.

amplification. The crystal filter is re-

sponsible for ahigh order of selectiv-

ity and is used for "single signal"

reception of c.w. signals. A double

diode,

pentode is used for the

second detector and first audio stage,

and is followed by a pentode audio

output tube. A beat-frequency oscil-

lator is loosely coupled to the diode

second detector. The coil gang plugs in at the
lower center of the panel, as can

Fig. 1111--The National HRO 9-tube superheterodyne. Thr gang of four coils is plugged in under the tuning dial.

be seen in Fig. 1111. The controls on the panel, aside from the main tuning dial, are audio and r.f. gain, beat-frequencyoscillator frequency control, and selectivity and phasing controls for the crystal filter. An "on-off" switch is included for turning off the receiver during transmission periods, and an-
other switch turns on the a.v.c. when radiophone signals are being received. A third switch, mounted on the b.f.o. control, turns off the b.f.o. for 'phone reception when the control

The Hammerlund "Super Pro," shown in Figs. 1114, 1115 and 1116, is a 16-tube superheterodyne using coil-switching for band changing. Several models are available -- the two common ones cover arange of 0.54 to 20 Mc. and 1.25 to 40 Mc. There are two stages of r.f. amplification ahead of the mixer, and a separate high-frequency oscillator is used. The
mixer output feeds into a wide-range crystal filter followed by three stages of i.f. amplifies-

Fig. 1112 -- A top view of the HRO receiver. The ganged tuning condensers can be seen either side of the dial mechanism.

202 C:EIAPTER ELEVEN

tion. The crystal filter is adjustable in steps from a selectivity wide enough to admit 'phone signals to sharp enough for singlesignal c.w. reception. To increase further the selectivity range of the i.f. amplifier, the i.f. transformen, are given avariable-selectivity characteristic by making the coupling between coils adjustable from the panel. Thus, by turning the i.f. selectivity knob on the panel, the i.f. characteristic can be made broad enough for highfidelity 'phone reception or sharp enough for some-
thing approaching single-signal perfur malice.
The i.f. amplifier is followed by a diode second detector, an audio noise limiter (useful for removing interference of the type caused by automobile ignition and similar sources), and two stages of audio amplification which drive apush-pull pentode output stage. The a.v.c. voltage is obtained from aseparate diode rectifier which is driven by an a.v.c. amplifier tapped into the output of the i.f. amplifier. This use of "amplified" a. v.v. results in a wide range of control. .\ b.f.o. is coupled into the plate circuit of the last i.f. amplifier stage.
The power supply of the Super Pro is aseparate unit connected to the receiver through aflexible cable. It is a heavy duty affair with aseparate rectifier for the C-bias supply. This rectifier obtains its voltage from atap on the power transformer. The field

kill 1;0

I(*pill ['Pi

)1� Fig. 1113 -- This under-chassis view of the HRO shows the coil compartment and the wiring of the receiver.
4� Fig. 1114 -- The Hammarlund Super � Pro 16-tube superheterodyne. The band. change switch is located between the two dial windows, just below the S meter. Fig. 1115 --- A top view of the Super Pro, showing the shielding of the coils and tuning condensers (center).

203 l'TE It ELEVEN

..7.he Radio ...Amateur's iiandllooh

parative carrier strengths, and the sensitivity of the metering system can be adjusted by a control at the rear of the chassis so that the
meter will indicate S9 on any signal from 10 to 10,000 microvolts, depending upon the requirements of the operator.

�AN ULTIR A-1116HFlt Eel E% Y SUPER� REGENERATIVE WECEIVER

The superregenerative receiver

shown in Figs. 1117, 1118 and

1119 has excellent sensitivity in

both the 112- and 224-Mc. ama-

teur bands, but it is not entirely

free from radiation as would be a

similar receiver with an r.f. ampli-

Fig. 1116 -- The tuning unit of the Super Pro with the shield cover removed. Each coil is mounted on an Isolantite base along with its associated trimmer condenser.

fier between the antenna and the detector. However, such areceiver will permit good reception in

the u.h.f. range with a minimum

of the loud speaker serves as one of the two fil- of expense, and the radiation is not great

ter chokes in the plate supply filter system.

enough to bother receivers more than a half

Two separate gangs of tuning condensers are used for tuning the Super Pro. A "band set"
gang is used to set the receiver in the portion of the spectrum one wishes to operate in, and the

mile or so away. A 9002 u.h.f. triode is used in the detector
circuit, followed by two stages of audio amplification. Only one stage of audio is used for head-

"band spread" gang is then used to tune slowly over this range. When the band spread dial, which is directly calibrated in frequency,

phone reception, and a jack is provided for plugging in the 'phones. The detector circuit is aform of the ultraudion circuit often used for

is set to the high-frequency end of an amateur band, the band spread dial will read frequen-
cies in that band. The manufacturer's tolerance for such calibration is less than 0.5% of the highest frequency in each range.

u.h.f. detectors -- the grid leak is returned to the positive plate supply rather than to the cathode, which results in slightly smoother
superregeneration. The tuning condenser, CI, is shunted by amica trimmer, C2, which is used

An "S" meter is included for reading com- as a band-set condenser and which allows

Fig. 1117 -- Left -- The panel of the two-band superregenerative receiver measures 7inches square. The knob in the upper right-hand control adjusts the antenna coupling and the knob below the tuning dial controls the regen eration. Right -- Aview of the back of the two-band superregenerative receiver shows the variable antenna coupling and the placement of parts. Note the 224-Mc, coil in the foreground -- the 112-Mc, coil is in the coil socket.
204 CHAPTER ELEVEN

Radio equipment

greater bandspread to be obtained with a reasonable value for C1. Variable antenna coupling is used to aid in matching to different antennas and to make the use of plug-in coils easier. Regeneration can be increased by ad-

The polystyrene tube socket for the 9002 is mounted on a metal bracket which is placed close enough to the tuning condenser to allow avery short lead from the tuning condenser to the plate connection and just enough room be-

An?o

RFC, RFC 2
Ic4 3

6F6

Speaker

Heaters

it v wov.+

Fig. 1118 --Wiring diagram of the two-band superregenerative u.h.f. receiver.

Ci -- Two-plate band-spread tuning.

-- 3-30.4t4cfd. adjustable mica band-set. Cs -- 50-pmfd. grid condenser.

C4 - 0.003-pfd. by-pass.

interruption-frequency

Ca, C7 -- 10-pfd. electrolytic cathode by-

pass.

C2 -- 0.01-pfd. audio coupling.

RI-- 10-megohm, 3'2-watt grid leak.

R2 - 50,000-ohm regeneration control.

Ra -- 0.1-megohm, 1-watt voltage drop-

ping.

R4 2500-ohm, .4-watt cathode bias.

Ra 0.1-megohm, 3/2-watt fringe-howl suppressor.

R6 -- 0.1-megohm, /-watt plate load.

R7 -- 0.1-megohm, 3/2-watt grid.

�

R8 -- 500-ohm, 1-watt cathode bias.

J-- Closed-circuit 'phone jack.

S-- S.p.s.t. B+ toggle switch.

-- Interstage audio transformer, 3:1 ratio.

RFC1 -- U.h.f. rd. choke.

RFC2 -- 8-mh. interruption frequency r.f.

choke.

Li -- Antenna coil -- 1turn No. 14 enam. wire, %-inch inside diam.

L2 -- Detector coil -- 112 Mc.: 3 turns No. 18 enam., winding length. Tap 1% turns from plate end.

diam., spaced to . 34,-inch

224 Mc.: 2 turns No. 18 enam., %-inch diam., spaced to 3/2-inch winding length. Tapped at center.

vancing the regeneration control, R2, or by
loosening the coupling between L1 and La. The receiver is built on a7- by 7- by 2-inch
chassis. The dial is mounted in the center of the panel and is connected to the tuning condenser by a bakelite flexible coupling. The condenser is mounted on a metal bracket cut out in the shape of a"U" to clear the stator connections of the condenser.
The socket for the plug-in coils is made from the contacts taken from a miniature tube socket. They are obtained by squeezing the socket in avise until the bakelite cracks, after which they can be easily removed. One of these contacts is soldered to each of the tuning con-
denser connections and athird is soldered to a lug supported by one of the extra holes of the Isolantite base of the tuning condenser. The only care necessary in mounting the contacts is to see that they are all the same height, so that the plug-in coil will seat well on them. The band-set condenser is mounted by soldering short strips of wire to the ends and then solder-
ing these wires to the tuning condenser terminals.

tween the rotor of the condenser and the grid connection of the tube for the grid condenser to fit. The heater and cathode leads are brought down to the underside of the chassis through a rubber grommet.
The variable antenna coupling coil is mounted on apolystyrene rod supported by a shaft bearing. The rod is prevented from moving axially in the bearing by cementing afiber washer to the shaft and tightening the knob on the other side so that the shaft does not move
too freely. The antenna coupling loop should be adjusted so that it will just clear the coils when they are plugged in the socket.
The coils are mounted on small strips of 3inch polystyrene which have three small holes drilled in them corresponding exactly to the tops of the coil sockets. The coil is cemented to the strip with Duco cement at the points where the wire passes through the strip. The No. 18 wire used for the coils will fit snugly in the sockets if the sockets are pinched slightly. A coil socket of this type allows very short leads to be used. The coils are trimmed to the bands by spreading the turns slightly. The

205 CHAPTER ELEVEN

Dhe Radio

lateUr

e
�

JicinidooL

band-set condenser gives some further range of adjustment and, in the receiver as described,

receiver of the superheterodyne type, with provision for the reception of either amplitude-

it is screwed down fairly tightly for the 112-Mc. band and loosened about four revolutions for
224 Mc. Two things will be found to influence the

or frequency-modulated signals. Since ordinary receiving tubes are of little
value above 60 Mc., special u.h.f. tubes of the "acorn" type are used for r.f. amplifier, mixer

sensitivity of the receiver, the value of C4 and the degree of antenna coupling. It is recommended that values of C4 from 0.001 to 0.005 Mfd. be tried. The antenna coupling will, of course, vary greatly with the setting of the coil and with the type of antenna that is used, and it is well worth while to tune the antenna circuit and then vary the coupling with the panel control. Tight coupling will usually give better results than loose coupling, and the coupling can be increased almost up to the point where it is impossible to make the detector oscillate
with no ill effects except increased radiation and QRM for other receivers in the vicinity.
No audio volume control was included in this
receiver because the parts were held down to a minimum, but one could easily be added. In

and high-frequency oscillator. A front-ofpanel antenna trimmer control allows the input circuit to be matched closely to the antenna in use, for maximum gain. The mixer is followed by atwo-stage 5.25-Mc. i.f. amplifier which has two degrees of selectivity. A
switch on the panel allows the selectivity of the i.f. amplifier to be set at "sharp" for a.m. or narrow-band f.m. reception, or to "broad" for wide-band f.m. reception. Depending upon whether a.m. or f.m. reception is required, the output of the i.f amplifier is fed into a diode second detector or a limiter and discriminator (the latter for f.m. rectification). A b.f.o. is included for c.w. reception, and an audio noise limiter can be switched in for quieter reception of a.m. signals. The two

this receiver, the value of R7 was adjusted until normal loud-speaker output was ob-

detection methods are followed by audio amplifier stages ending in a push-pull output

tained, and it can be varied to meet anyone's stage.

particular requirements.

A dual-purpose "S" meter serves as a carrier-level meter for a.m. signals and as a

� A COMMERCIAL EAU?. RECEIVER
The Hallicrafter S-27 u.h.f. receiver shown in Figs. 1120, 1121 and 1122 affords continuous coverage of 27 to 145 Mc. in three ranges of 27-46, 45-84 and 81-145 Mc. It is a 14-tube

tuning meter for f.m. signals. More So than in any other type of receiver, accurate tuning is
a requisite of distortionless f.m. reception. Several degrees of response can be obtained
from the audio amplifier through the incor-

Fig. 1119 -- 1.eft -- A close-up view of the tuning assembly shows how the leads from tuning condenser to tube socket have been kept short and how the coil socket is mounted on the tuning condenser. Hidden by the grid condenser (the 50-aafd. condenser so prominent in the picture), the plate terminal of the tube socket goes to alug that has been added to the stator of the tuning condenser. Hight -- The arrangement of parts under the chassis can be seen in this photograph. The 6J5 socket is on the left and the 6F6 socket is on the right, near the speaker terminals.

206 CHAPTER ELEVEN

Fig. 1120 -- The Hallicrafter S-27 uhf. receiver can be used for either a.m. or f.m. reception. The scale at the left indicates the frequency -- the center scale is a vernier reading for accurate logging of signals.

/each� equipment

geared much higher and serves as a vernier indicator of the main dial setting.
Panel controls, other than the tuning, are band switch, a.f. gain, a.m.-f.m.switch, b.f.o. pitch control, tone selector switch, a.v.c. switch, noise limiter on-off switch, b.f.o. switch, send-receive switch and antenna trimmer control.
�RECEIVER POWER
SUPPLIES
Power supplies of most minufactured radio receivers are bu.lt into
<--�
Fig. 1122 -- A close-up view of the coil and condenser assembly of the S-27 with the shield cover removed. The acorn tube oscillator can be seen clearly -- the mixer and r.f. stage acorn tubes are mounted on the interstage shields.

poration of audio filters in the receiv-
er. Setting the tone-control switch at "low" attenuates the higher audio frequencies in the manner of the ordinary tone control. In the "normal" setting, a response similar to the ordinary broadcast receiver is obtained,
while in the "high fidelity" position the high-frequency range is extended to include the higher frequencies
transmitted by f.m. broadcast stations. A fourth "bass boost" position brings up the lower frequencies below 100 cycles slightly more than
they would normally be, to compensate for loud speaker and other acoustical shortcomings.
Only one tuning control is used, in conjunction with two dials. One
dial indicates the frequency in megacycles and rotates about 340 degrees
for 180-degree rotation of the tuning condenser. The other dial is

Fig. 1121 -- A top view of the S-27, showing the shield covering the coil and condenser assembly and the gears of the dial mechanism.

207 CHAPTER ELEVEN

-% Radio Amateur'� ilandtoole

Fig. 1123 -- This small receiver power supply will deliver 300 volts at 130 ma. with achoke-input filter and about 430 volts with acondenser-input filter. The chassis measures 7by 9by 2inches. The circuit diagram is shown in Fig. 1124.
the receiver cabinet, thus making the receiver and power supply a single unit. On the other hand, it is often more convenient for the amateur who builds his receivers to make the power supply aseparate unit so that it may be used with different receivers that may be built from time to time. In either case, the power supply requirements are substantially the same: correct voltage (within 10% or so), an adequate power capability, low hum and noise level and good regulation. The first two requirements are fulfilled by proper selection of the power transformer and current ratings of the filter chokes, the third depends upon the use of sufficient filter and, in some cases, avoiding the use of mercury-vapor rectifiers, and the fourth can be satisfied by using low-resistance filter chokes and, where necessary, voltage regulator tubes. It is not necessary, however, to use regulator tubes except where the voltage must be held quite close, as in the case of critical biases and the plate voltage of highfrequency and beat-frequency oscillators. For this reason, the regulator tubes are usually used to control only these voltages.

A typical heavy-duty amateur power supply is shown in Figs. 1123 and 1124. It is built upon a small metal chassis, with all of the wiring under the chassis. Heater and highvoltage connections are brought to a fourprong socket, so that the power cable from the receiver can be plugged in, and the connection to the 110-volt line is made through a male plug mounted on the side of the chassis. A switch mounted on one side of the chassis turns on the 110-volt supply to the primary of the transformer. As shown in the wiring diagram, the output voltage will run about 450, too high for most receiver use, but by removing the input condenser, C1,the output voltage will drop to about 300, correct for most receiver work. Further reduction of voltage can of course be obtained by the use of adropping resistor (but with an adverse effect on the regulation) or by the use of a lower-voltage transformer.
� A SIMPLE TETIIIIDE OSCILLATOR TIIANSAIITTER
The unit shown in the photograph of Fig. 1125 represents one of the simplest types of amateur transmitters. The various components are assembled on aplain wooden baseboard.
A simple tetrode crystal-oscillator circuit (� 4-4) is used and is shown in Fig. 1126. Parallel feed (� 3-7) is used in both plate and grid circuits so that the only exposed highvoltage points are the plate-circuit r.f, choke and the high-voltage power terminal. Parallel plate feed also permits mounting the plate tank condenser, C1,directly on the baseboard without insulation. Voltage for the screen is reduced to proper value by means of the dropping resistor, R2. Bias is obtained entirely from the voltage drop across the cathode resistance, R1 (� 3-6). The r.f. chokes are placed so as to be out of the direct field of the plate tank coil. By-pass condensers (� 2-13) are located close to the points to be by-passed. A common grounding point (� 2-13) is provided by awire running the length of the baseboard to which all ground connections shown in the circuit diagram are made.
Since this circuit is not designed for frequency doubling, a separate crystal is re-

To 110 VAC.
208

CHAPTER ELEVEN

Fig. 1124-- Circuit diagram of a typical receiver power supply. Ci -- 4-Add. electrolytic input. C -- 8- or 16-mfd. electrolytic out-
put. L -- 10-by., 150-ma, low-resistance
filter choke. R -- 15,000-ohm, 25-watt bleeder. T -- Type 80 rectifier. Tr -- Power transformer: 400 volts
each side of center tap at 140 ma., 6.3 volts at 6 amperes and 5volts at 3ampeees.

Radio eiripmeni

quired for each frequency at which it is desired to operate.
Simple direct . coupling to the antenna (� 10-4) is shown in the diagram. Coupling is adjusted by moving the tap up or down on the plate tank coil. As indicated by the dotted

807 as an amplifier-doubler. The circuit diagram is shown in Fig. 1129.
The Tri-tet oscillator circuit (� 4-5) is chosen because of its ability to supply output at harmonic frequencies of the crystal, as well as at the fundamental. Sufficient output may

be obtained at both the second and fourth

harmonics (� 2-7) to drive the 807 amplifier.

The amplifier stage is capacitively coupled

(� 4-6) to the oscillator through the coupling condenser, Cg. The 807 is a screened tube,

therefore no neutralizing circuit is required

(� 4-7). A link, Lg, is provided at the output for coupling to the grid tank circuit of a

following amplifier or an antenna tuner (� 4-7; �10-6).

Series plate feed (� 3-7) is used in both

stages, while parallel grid feed is used in the

amplifier. The resistances R4, Rg, R g and R9

Fig. 1125 -- A simple breadboard oscillator transmitter. The crystal is plugged in the tube socket to the left, while the socket to the right holds the plate tank coil, Li. The 6L6 oscillator tube is near the center. The grid r.f, choke is between the crystal and 6L6, while, the plate r.f. choke is to the right of the 61.6. The cathode and screen resistors are to the rear of the 6L6. The blocking condenser, C2, is between the tube and the tank condenser, CI, to the left of Li.

form a voltage divider (� 2-6) to provide suitable voltages for the oscillator screen and plate. Two separate resistors, R8 and R9 are used instead of a single resistor in this case because their physical sizes permit suitable
space inside the small chassis to be found. A second voltage divider for the amplifier screen is composed of the resistances Rg and R7.
Safety bias for the oscillator tube, in case

lines, link coupling to aconventional antenna tuner (� 10-6) may be used, if preferred, by

the oscillator circuit ceases to function, is provided by the cathode resistance (� 3-6), Rg.

adding a suitable link winding at the bottom

of the plate tank-coil form.

Connections to apower unit, such as the one

shown in Fig. 1123, may be made by aplug and

cable connected to the terminal strip at the

rear of the baseboard. Plate voltages up to 450

may be employed when using the type 6L6,

while the same arrangement, without change,

may be used with the type 6V6 at lower plate voltages.

A meter with ascale of 100 or 200 ma. may

be connected in series with the key for check-

ing, for tuning purposes, combined plate and

screen currents which flow through the cathode circuit

With a6L6 tube and aplate supply deliver-

ing 400 volts, the screen voltage will run about

250. The tube will draw about 75 ma. non-

oscillating, dipping to about 50 ma. at

resonance (� 4-4) with the antenna discon-

nected. It should be possible to load up the circuit until the tube draws about 80 ma. at resonance. Under these conditions, the power output on each band should be 15 to 20 watts.
�A TWO-TUBE PLUG-IN COIL EXCITER In the two-tube exciter or low-power trans-
mitter shown in the photographs of Figs. 1127 and 1128, a 6L6 oscillator is used to drive an

oscFiilgl.ato1r1t2r6an--sCmiirttceuri.t diagram of the simple tetrode Ci -- 250-pgfd. plate tank condenser. C2 -- 0.001-pfd. mica plate blocking condenser. Ca, C4, C5 -- 0.01-pfd. paper cathode, screen and plate
by-pass condensers. RI-- 200-ohm, 2-watt cathode biasing resistor. R2 -- 15s,i0st0o0r-.ohm, 2-watt screen voltage-dropping reRFC -- 2.5-mh. r.f. chokes. Li -- Plate-tank inductance, 38 phy., 12 phy., and 5
shy., respectively, for 1.75, 3.5 and 7Mc. Power connections are made through the plug and
cable indicated.

209 CHAPTER ELEVEN

..7he Radio AmaleuA -ilattiL�

Fie. 1127 -- A two-tube plug-in coil exciter, built to conserve space in a relay rack. The crystal socket to the
left is submounted in the panel. The dials control the plate tank condensers, CIand Ci. The knob to the right is
for the meter switch, Sw2, while the toggle switch, Sty:, is to the left.

Additional operating bias (� 4-8) is obtained
from the grid-leak resistance (� 3-6), R1.In the amplifier circuit, protective bias is furnished from an external 45-volt supply, such as a "B" battery. Additional bias is obtained from the grid leak, R2,when the circuit is in
operation. A switch is provided so that the milliamme-
ter may be shifted to read either oscillator plate current or amplifier plate current for tuning purposes. With the keying system shown, both stages are keyed simultaneously in the common cathode lead (� 6-2).
The unit is designed to operate from asingle high-voltage power supply delivering 750
volts at 250 ma. Although the output stage may be operated
as a frequency doubler (� 4-1; �4-11) when it is necessary to obtain output at the eighth harmonic of the crystal frequency, greater power output may be obtained, without exceeding the dissipation rating of the tube, if it is operated as astraight amplifier.
The unit is constructed in amanner to conserve panel space in arelay rack. The crystal socket is mounted on the panel so that crystals

may be conveniently changed. The cathode tank coil, the oscillator tube and the oscillator plate tank coil are grouped closely to permit short connecting wires. L1 and L2are placed
so that their axes are at right angles to reduce inductive coupling between the two (� 2-11). C1 is mounted inside the chassis so that short leads may be passed through clearance holes in the chassis between its terminals and those
of L2.The 807 tube is mounted horizontally fr�m asmall metal panel which, together with the cylindrical can, provides a measure of shielding against external stray capacitive coupling between the amplifier input and output circuits which, if allowed to exist, might cause self oscillation in the amplifier (� 7-6).
The horizontal mounting also permits ashort plate lead to the output tank circuit. The output tank coil is placed at the extreme end of the chassis to reduce inductive coupling be-
tween the two tank circuits. In preparing to place the unit in operation, a
crystal must be chosen whose fundamental frequency or whose second-, fourth- or eighthharmonic frequency falls at the output frequency desired. For output at the fundamental

Fig. 1128-- Underneath view of the two-tube plug-in coil exciter. The components mounted along the rear edge of the chassis from left to right are: The output tank coil, Ls, the 807 amplifier tube, the oscillator plate tank coil, L2, the 6L6G oscillator tube and the cathode coil, Li. Inside the chassis are the two tank con-
densers, CIand Cz, which must
be insulated from the chassis, and the various resistors. Insulating couplings are used between the tank-condenser shafts and the two dials.

210 CHAPTER ELEVEN

e cho (...quipmen1

frequency or the second- or fourth-harmonic
frequencies of the crystal, coils must be selected for L2 and L3 which will resonate at the desired output frequency. If the output frequency desired is the eighth-harmonic frequency of the crystal, Ci-L2 should tune to the fourth harmonic, while the C2-L3 circuit should be tuned to the desired output frequency. The coil used for L1 should always correspond to
the frequency of the crystal in use. When
Cr-L2 is tuned to the fundamental frequency of the crystal, Stv 1 should be closed. The cir-
cuit is then that of a simple tetrode crystal oscillator (� 4-4). The purpose of this is to prevent excessive crystal r.f. currents which may damage the crystal when the oscillator tube is operated at high-power input at the crystal fundamental frequency (� 4-3).
When the unit is operated at 750 volts, the power output obtainable should run between 40 and 55 watts, depending upon the output frequency, if the output stage is operated as a straight amplifier. When the output stage is used as afrequency doubler, the input must be reduced to prevent excessive plate dissipation

and, therefore, the power output obtainable in practice will be reduced to between 18 and 25 watts.
� A 450-WATT PUSII-PULL AMPLIFIER
Figs. 1130 and 1131 show two views of a push-pull r.f. amplifier designed for a pair of tubes of the 1500-volt, 150-ma, class, such as the types T40, T55, 812, 8005, R K51, HF100, etc. A similar arrangement, with a plate tank condenser of less plate spacing (� 4-8), is also suitable for tubes of the 1000-volt, 100-ma. class.
The circuit, which is conventional, is shown in the diagram of Fig. 1132. A tuned tank circuit is provided for the grid circuit, as well as the plate circuit, to permit link coupling to an exciter. A system of plug-in coils is used to shift operating frequency from band to band. To provide sufficient tank-circuit capacity (� 4-8) without the use of variable condensers of excessively-large physical dimensions, provision has been made for plugging in a fixed air padding condenser in parallel with each of

6L66

807

Fig. 1129 -- Circuit diagra m of the two-tube exciter.

Ci -- 140-ssfd. oscillator plate tank condenser. Ca 150-ssfd. amplifier plate tank condenser. Ca -- 100-ssfd. mica cathode tank condenser. C4 -- 20-ssfd. mica coupling condenser. Cs, C0 -- 0.01-sfd., 600-volt, paper cathode by -pass
condensers. Cee, C10 -- 0.01-iifd., 600-volt, paper screen by-pa..
condensers.

-- 10,000 ohms, 25-watt (see text). Rs -- 3500 ohms, 25-watt (see text).
Re, R7 -- 15,000 ohms, 25-watt (see text). Rs, Rs -- 1250 ohms, 50-watt (see text. Rio, Rn -- 10-ohm, 1-watt meter-shunting resistances. RFC -- 2.5-mh, r.f. choke. Swi -- S.p.s.t. toggle switch (see text).

C7 -- 0.01-Add., 600-volt, paper oscillator plate by-pass Siva -- D.p.d.t. rotary meter switch.

condenser.

Li -- Oscillator cathode inductance -- 35 shy., 3.2 shy.,

Cs -- 0.01-sfd., 600-volt, paper grid-circuit, by-pass

and L75 shy., respectively, for 1.75-, 3.5. and

condenser.

. 7-Mc. crystals.

CH --0.01-14fd., 1000-volt, paper amplifier plate by- La -- Oscillator plate tank inductance -- 54 shy., 15

pass condenser. RI-- 20,000-ohm, 1-watt oscillator grid leak.

shy., 4.2 shy., 1.25 shy. and 0.5 shy., respectively, for 1.75, 3.5, 7, 14 and 28 Mc.

82 -- 25,000-ohm, 2-watt amplifier grid leak. R -- 200-ohm, 2-watt oscillator cathode biasing re-
sistance.

La -- Amplifier plate tank inductance -- 52 shy., 16 shy., 5.7 shy. 1.5 shy. and 0.7 shy., respectively, for 1.7d, 3.5, 7, 14 and 28 Mc.

211 CHAPTER ELEVEN

a.Radio -AmaleuA -ilamilooh

the tank condensers for 1.75-Mc. operation. The rotor of the split-stator plate tank con-
denser, C2, is insulated from ground for d.c., while C10 provides a ground path for r.f. currents. This permits ahigh d.c.-voltage connection to be made to the rotor of C2, thereby removing the difference of d.c. potential between rotor and stator plates of the tank condenser. This connection permits the use of atank condenser with less plate spacing, since the peak r.f. voltage is the maximum to appear
between the condenser plates (� 4-4). The tuned circuits 1.43-C6 and L4-05 are trap
circuits tuned to trap out ultra-high-frequency parasitic oscillations which are usually encountered in an amplifier of this type (� 4-10).
The 100-ma. meter may be switched to read
either d.c. grid current or total cathode current. When switched to read cathode current, the meter is shunted by a low resistance, R2, which is of the correct value to give a meter-

scale multiplication of five (see Chapter 12). The shunting resistance, RI,is of sufficiently high value to have no practical effect upon the reading of the meter when it is switched to
read grid current. The purpose of the disposition of components
shown in the photographs is to arrive at an arrangement which will permit both short r.f. connecting leads and good isolation between grid and plate circuits. The plate tank condenser and coil are placed on either side of a vertical partition ,with clearance holes for the connecting leads. The tubes are placed on the chassis with their plates close to the stator terminals of the plate tank condenser and their sockets submounted so that their grid terminals are close to the stator terminals of the grid tank condenser underneath the chassis. The grid tank coil is mounted on the vertical partition to the left with its axis at right angles to that of the plate tank coil. The chassis

Fig. 1130 -- A 450 -% at t

ptuasnhk-piunldlucatmapnlcief,ie1r4.,Tihs eto

grid the

left and the plate tank in-

ductance, L2, to the right.

The two neutralizing con-

densers are staggered be-

tween the two tubes on the

chassis. The plate tank con-

denser, C2, is mounted on the

right-hand partition. The

parasitic trap tank circuits,

La--Cs and IA--Cs, are in the

plate leads to the tubes.

�

�
Fig. 1131 -- Bottom view of the 450-watt push-pull amplifier, showing the position of the grid tank condenser between the two sub. mounted tube sockets and the two air padding condensers in place for 1.75-Mc. operation.

212 CHAPTER ELEVEN

e cho equipment

No grid leak is shown in the diagram, since

it is assumed that biasing voltage will be

obtained from one of the simple units con-

taining the required grid-

leak resistance discussed in

the chapter on power sup-

Output ply (� 8-8).

For maximum plate effi-

ciency with plate modula-

tion, an exciter having an output rating

of not less than 25 watts is required. The

exciter unit described in the previous section should be suitable.

� A GRID-STABILIZED.815 112-MC. TRANSMITTER

--BIAS

C.7. 7.5V.

+H.V. +BIAS

ampFliigf.ie1r1.32 -- Circuit diagram of the 450-watt push-pull

Ci -- 100-spfd. per-section grid tank condenser, 0M3inch spacing.

C2 --

100-pidd. per-section plate tank condenser, 0.07inch spacing.

Ca, C4 -- Micrometer-type neutralizing condensers. C5, Ca -- 30-ppfd. mica-trimmer parasitic-trap con-
densers.

C7 -- 0.0c1o-npdfedn.s,er6.C/0-volt, paper grid-circuit by-pass

C8, C9 -- 0.01-mfd., 600-volt, paper filament by-pass condensers.
Cio -- 0.001-pfd., 7500-volt, mica plate-circuit by-pass condenser.

Cii -- 50-plifd., air grid-tank padding condenser for 1.75 Mc., 0.05-inch spacing.
C12 -- 50-ppfd. air plate-tank padding condenser for 1.75 Mc., 0.125-inch spacing.
Ri -- 25-ohm, 1-watt meter-shunting resistance. Ra -- Meter-multiplier resistance for 5-times multi-
plication.

RFC -- 1-mh. r.f. choke. MA -- D.c. milliammeter, 100-ma. scale.
-- Grid tank inductance -- 70 ,by., 38 shy., 13 phy.,

4.5 phy. and 0.8 shy., respectively, for 1.75, 3.5, 7, 14 and 28 Mc., 3-turn links. -- Plate tank inductance -- 70 phy., 35 phy., 14

phy., 3phy. and 1;illy., respectively, for 1.75, 3.5, 7, 14 and 28 Mc., 2-turn links. 1,3,L4-- Parasitic trap coils -- 4 turns No. 12 wire, 3/2-inch diameter, 4-inch long.

The transmitter shown in Figs. 1133 and 1134 uses an 815 double beam tube in agridstabilized oscillator circuit and will run at an input of 60 watts with good efficiency. The

and double vertical partitions provide shielding against undesirable couplings (� 7-6) between input and output circuits, which might otherwise be sufficient to make complete neutralization (� 4-7) impossible. The neutralizing condensers are placed in aposition where short, direct connecting leads are possible. The parasitic-trap components are soldered directly in the leads from the tube plate caps to the stators of C2.
For operation at maximum input, a platevoltage supply delivering 1500 volts at 300 ma. is required. Filament supply, depending upon the tubes selected, will also be required.

Fig. 1133 -- The grid-stabilized 112-Mc, transmitter is mounted on a3- by 4- by 5-inch metal box, and the box houses the filament transformer and the various fixed condensers, resistors and the r.f. choke. The frequency is changed by adjusting the length of the grid lines by sliding the inner tubes in and out. The power supply cable plugs on the plug mounted on the side of the box. Wires for feedback control run from the plate caps of the 815 close to the grid lines.

213 CHAPTER ELEVEN

5Le Palio Antaieur'� -llancidoofl

circuit is similar to the tuned-grid tuned-plate except that it uses alinear circuit instead of a coil and condenser in the grid circuit. By tapping the grids down on the line the line is loaded lightly and consequently retains its high Q. The 815 does not have ahigh-enough
grid-plate capacity to give all of the necessary feedback, and some additional capacity must be added from plate to grid of both sections of the tube. This is easily done by two short

of the transmitter outside of the adjustment of the feedback condensers. This can best be done with adummy load such as a25-watt electric lamp connected to the antenna terminals. The lead from the grid leak, RI, to ground should be opened and a 0-10 millammeter connected in the circuit. Plate voltage can be applied and the plate tuning condenser rotated for maxi-
mum output as indicated by the brilliancy of the lamp. The grid current should be between 3.5 and 5 ma. at this point -- if it is higher

815
r, -

there is too much feedback and the feedback

capacity should be reduced by trimming off a

short length of the wire or by moving

it away from the grid lines. It is not

L3 Ant. too critical asetting but it should be

done before the transmitter is put on

the air. After the proper feedback ad-

RFC

justment is found, the antenna can

be coupled to the transmitter and modulation

applied. The frequency can be checked by means of Lecher wires or a wave-meter. The

antenna coupling is tightened until the plate

current is 150 ma. and the grid current should

A�.

_ 400V.

Fig. 1134 -- Wiring diagram of the grid-stabilized

be between 3.5 and 5 ma. under these conditions.
The power supply is required to deliver

2%-meter oscillator.

slightly over 165 ma. at 400 volts, and the

Ci --15.��fd. per section dual plate tuning. C2 -- 0.002-mfd. mica screen by-pasq. Cr -- Feedback condenser. See text and Fig. 1133.

modulator must give at least 30 watts to modulate fully the oscillator.

Ri -- 15,000-ohm, 1-watt grid leak. R2 -- 25,000-ohm, 10-watt screen dropping. Li -- Grid lines: j,�-inch diem. copper tubing 23 inches
long. Spaced 1inch on centers; grids tapped 2% inches from shorted end. 1,2- Plate inductance: 2 turns No. 12 enam., 1-inch diam., turns spaced % inch. -- Antenna coupling coil: 2 turns No. 12 enam., i%-inch diam., turns spaced % inch.

* A TIRANSMFITER FOR 224 MC.
As one operates on frequencies higher than 116 Mc. he finds considerable difficulty in getting good performance with tubes other than

RFC -- U.h.f. ri. plate choke. Ti -- 6.3-volt filament transformer.

lengths of wire running from the plate terminals to points near the grid lines.
The grid line is made of half-inch copper tubing and is supported a half inch from the box by three feed-through insulators which also serve as convenient connectors to the grids
and to the grid leak. The open ends of the parallel tubings take 3-inch lengths of 3/8-inch diameter tubing which can be moved in and out to adjust the frequency of the oscillator. They are held securely in place by set screws
through the half-inch tubing. The plate condenser is supported by a3-inch
steatite pillar which also acts as a guide for the sliding variable antenna coupling. Two large 866-type plate caps are slid over the pillar and the antenna binding-post assembly is fastened to them by short lengths of No. 12 wire. By sliding this assembly up and down the antenna coupling can be set to any value de-
sired. There is nothing unusual about the tuning

Fig. 1135 -- A 224-Mc. transmitter using the IlY75. A rectangular hole in the top of the Preedwood chassis allows the tuning condenser to be placed for shortest leads. The tunine condenser is adjusted by an insulated
screwdriver.

214 CHAPTER ELEVEN

Radio equipment

HY 75

( L2
L1

oAnt C,

RFC2 =

RFC3

RFC,

6.3 V.

Filament Transfomer Center tap

400 +

Fig. 1136 -- Wiring diagram of the 224-Mc. oscillator.
Ci -- 100-�pid. midget variable tuning. Ri -- 5000-ohm, 10-watt wirewound grid leak. Li -- Series-tuned tank circuit: 5'-inch copper tubing
33i inches long, spaced 32 inch on centers. L2 Antenna coupling loop: 2-inch loop No. 16 bare
wire. RFC1 -- U.h.f. r.f. plate choke. RFC2, RFC3 -- Filament chokes: 10 turns No. 18
enam. closewound on34-inch diam., selfsupporting.

the tightness of coupling and the setting of C1. It will be found that the output is alittle better towards the maximum-capacity end of the range of C1.The frequency coverage of the transmitter should now be checked, by Lecher wires or a wavemeter, to make sure that it will cover the range. The coverage can be adjusted slightly by changing the separation of the copper tubes, but if this is not enough the tubes will have to be made shorter or longer.
The transmitter requires a power supply capable of furnishing 60 ma. at 400 volts, and the modulator should be capable of delivering 12 watts of audio.
Because of its small size, a transmitter of this type can be built right into arotatable antenna for the 224-Mc, band if desired. It is desirable not to run afeed line for any great distance at this frequency because of the chances for loss in the line.
� A 112-MC. "WALKIETALKIE" TRANSMITTER4tECEIVE111

those designed expressly for u.h.f. operation. However, there are several inexpensive tubes available to amateurs that will perform well on 224 Mc., and the transmitter shown in Figs. 1135 and 1136 shows how the HY75 can be used.
The transmitter is built on a 3M- by 6Minch strip of 3'-inch Presdwood supported by two strips of 1- by 2-inch wood. A rectangular hole is cut in the center of the Presdwood to accommodate the tuning condenser which is
supported by two metal pillars at one end. The tuned circuit consists of two pieces of 3d-inch copper tubing supported at one end by two feed-through insulators. The screws of the feed-through insulators are sweated into the ends of the tubing, and the tuning condenser connects to two lugs right at this point. Connec-

Many battery-operated lightweight 112-Mc. stations are of the "transceiver" variety, which usually consists of two tubes, one serving as the transmitter oscillator or receiver de-
tector and the other as modulator or audio amplifier, depending upon the position of the "send-receive" switch. A transceiver of this
type has the disadvantage that too many compromises must be made -- in the "receive" position the radiation is almost as great as in
the "send" position, and the transmitted and received signals must be of the same frequency because only one tuned circuit is used. Further, it is practically impossible to obtain good trans-

tion from the tubing to the grid and plate leads of the tube is made through M inch of flexible braid. Filament chokes, the plate r.f. choke and the grid leak are mounted under the chassis.

The antenna coupling consists of a loop of wire parallel to the copper tubing and terminating in the antenna binding posts. The coupling is varied by moving the loop nearer to or far-
ther away from the copper tubing. The transmitter should first be
tested with adummy load, and a 10watt electric lamp is excellent for the purpose. The load is connected to the antenna posts and the power supply is then turned on. If everything is connected properly the lamp will
light, its brilliancy depending upon

Fig. 1137 -- A 112-Me, pack set ready to go. The station is bull into aknapsack which contains two complete sets of batteries.

215 CHAPTER ELEVEN

..7Le Radio Amateur'i ilandhooh

mitter stability and receiver sensitivity at the
same time in such an arrangement. The transmitter-receiver shown in Figs.
1137, 1138 and 1139 is a big improvement over a transceiver in that it uses only two tubes but has separate tuned circuits for re-
ceiving and transmitting, thereby reducing considerably the compromises in performance. In the "receive" position alow-C tank circuit is switched in. This has been adjusted for best sensitivity. In the "send" position a

high-C tank circuit which gives afair order of
frequency stability is used. The circuits can be tuned to any desired frequency in the band, and thus no operating limitations are imposed. Optimum antenna coupling to each circuit is assured by individual transmitter and receiver antenna-coupling condensers. The carrying case is large enough to hold two complete sets of batteries; aswitch cuts in anew set in the event that batteries begin to fail during operation in the field. However, the batteries are
not expensive, and more than 50 hours of operation can be expected from a single

set. A 1Q5GT, with plate and screen grid con-

nected together to give atriode characteristic,

is used for the r.f. tube. A two-section twoposition rotary switch is used for changing

from "send" to "receive" and back again. One section of the switch (Si) has Isolantite insulation and is used to switch the r.f. circuits: the

grid and plate of the 1Q5GT and the antenna. The other section of the switch (82) is bakelite-

insulated and controls the d.c. and audio circuits. In the "receive" position, 82 connects in the interruption-frequency feedback condenser, C7 (removed during transmission because it by-passes too much of the audio from the microphone transformer), opens the microphone circuit to reduce battery drain, closes the receiver regeneration-control supply

circuit and switches the output of the audio tube (also a 1Q5GT) from the modulation

choke, L, to the headphones. In the "receive" position, the superre-

Fig. 1138 -- Wiring diagram of the 112-Mc, pack set.
Ci -- 10-anfd. midget receiver tuning.
-- 35-mafd. midget transmitter tuning. Ca -- 3-30-anfd. adjustable mica receiver antenna
coupling. C4 3-30-anfd. adjustable mica transmitter antenna
coupling. Ca -- 100-add. receiver grid.
-- 100-nnfd. transmitter grid. C7 -- 0.004-pfd. interruption-frequency by-pass. Cs -- 0.01-pfd. audio coupling. Ri --0.5-megohm, Y2-watt receiver grid leak. R2 -- 15,000-ohm, %-watt transmitter grid leak. 113 -- 0.25-megohm, %-watt audio grid. R4 -- 50,000-ohm regeneration control. RFC' -- U.h.f. receiver plate choke. RFC2 -- U.h.f. transmitter plate choke. RFC3 -- 80-mh. interruption frequency filter choke.

generative detector is impedance-coupled to the audio amplifier tube by the secondary of the microphone transformer, T (acting as a coupling impedance), coupling condenser Cs and grid resistor R3. Regeneration is controlled by the setting of R4. Bias for the audio tube is obtained from asmall 4.5-volt battery. When transmitting, the plate circuit of the audio tube is coupled to the oscillator across L.
A toggle switch, 83, turns on the station, and another toggle switch, Sa, is used to switch to the spare supply of batteries.
The station is built on a plywood panel which is fastened to a plywood box fitted snugly in the knapsack. The box extends 2 inches above the panel, thus protecting the
controls from accidental movement. The dimensions of the box will, of course, be de-

J-- Open-circuit microphone jack. Si -- 3-circuit 2-position Isolantite r.f. switch. S2 4-circuit 2-potation bakelite d.c. and audio switch.
Si and Si are ganged. Sa -- D.p.s.t. on-off toggle. S4 D.p.d.t. power supply selection toggle. T -- S.b. microphone-to-grid transformer. L-- 15-henry, 40-ma, modulation choke. Li -- Receiver inductance: 3 turns No. 14 enam., 3/e.
inch inside diam., 7/16-inch long. -- Transmitter inductance: 1 turn No. 12 enam.,
%-inch diam.

termined by the size of the knapsack. The box does not extend to the bottom of the knapsack. Corner posts, of -inch square wood, are provided for fastening the panel in place. Glue and brads are used to hold the box and corner posts together. The box is prevented from slipping into the knapsack by tacking strips of 3d-inch quarter-round trim around the top
edges of the box.

216 CHAPTER ELEVEN

Fig. 1139 -- A view of the back of the 112-Mc. pack-set panel. The receiver tuning condenser can be seen to the left of the "send-receive" switch. The two tubes are mounted on brackets on the switch. The microphone transformer, microphone jack, headphone tip jacks and regeneration control can be seen at the left-hand edge. The modulation transformer, L, is directly behind the two tubes.

k),,,/I. � / 10 --f.i il ip/n ett

The arrangement of the parts in the r.f. portion of the station is centered around the "send-receive" switch because of the necessity for short r.f. leads. The two tube sockets are supported by metal brackets mounted on the
switch, thus placing the r.f. tube socket directly over the Isolantite switch section and the audio tube over the bakelite switch section. The brackets can be made of 3/2-inch strips of 3-le-inch thick aluminum or copper. The plywood panel is backed up by a 3 by 5-inch plate of aluminum to reduce hand-capacity
effects. The tuning condensers, C1 and C2, are mounted on 1-inch pillars as close to the Isolantite switch as possible.
Considerable care must be exercised in wiring the r.f. portions of the station because additional inductance in the leads will prevent the obtaining of proper L-C ratios in the tank circuits. Short leads of No. 12 or 14 wire should be used. The d.c. wiring is relatively unimportant and can be run wherever convenient. The lead from RFC3 to the audio transformer had best be shielded to prevent r.f. pickup along its length.
The antenna is a38-inch length of automobile antenna. It is supported by aporcelain feedthrough insulator. The antenna is tapped to take a 6-32 screw, and a long 6-32 screw is passed through the insulator. The antenna can be unscrewed from the assembly when not in use.
To test the station, connect the batteries, antenna, headphones and microphone and throw SI-82 to the "receive" position. Set the antenna condensers, C3 and C4,to the minimum-capacity position. When S3 is closed and the tubes heat up (which takes only an instant), the usual superregenerative hiss should be heard as R4 is advanced. If no hiss is heard, it may be necessary to try different
values of C7 or RI,providing, of course, that the wiring has been thoroughly checked and

found to be correct. The receiver inductance, LI,can now be adjusted by squeezing the turns together or pulling them apart until the tuning range covered by C1 includes the 112-Me. amateur band. Then close C3 until it is necessary to advance R4 well towards the end of its range to obtain regeneration.
Upon switching to "send," the inductance L2 should be trimmed until the transmitter tunes to 112 Mc. with C2 almost completely meshed. This will give the best stability obtainable with this rig. It is probable that C4 can be closed up to very near full capacity without any tendency for the transmitter to go out of oscillation. If full capacity does prevent oscillation, C4 can be opened abit.
Since the range of the "walkie-talkie" depends to a great extent upon the location of the station, it is highly desirable to operate the ,unit from as high apoint as possible. A change in height of only a few feet may increase the signal strength enough to make communication possible where it would not be sufficient from alower position. In acrowded city, the range may be only a few blocks, but in open terrain and between elevated points, the range will run up to one-half mile or more.
� A 40-WATT SPEECH AMPLIFIER OR MODULATOR
Fig. 1140 is the photograph of high-gain audio amplifier which may be used to plate modulate a low-power final amplifier or to drive a high-power Class-B modulator. The circuit diagram of this unit will be found in Fig. 1141.
The microphone output is fed to the grid circuit of a high-gain voltage-amplifier stage using a 6J7 pentode. Resistance coupling (� 5-9) is used between the output of this stage and the input of a second stage with a 6J5 triode. The 6J5 is transformer coupled to the input of a push-pull amplifier which fur-

217 CHAPTER ELEVEN

..7ne Pail� Amaleur'3 -ilani`oo1

Fig. 1140 -- A 40-watt speech amplifier or modulator of inexpensive construction. The 6J7 and first 6J5 are at the front, near the microphone connector and volume control, respectively. Ti is just behind them, and the push-pull 6J5s are at the rear of the chassis behind Ti. T2, the 6L6s, and T8 follow in order to the right.
�

nishes sufficient power to drive apair of 6L6s
operating as Class AB2 power amplifiers. Gain is controlled by the potentiometer, lith which adjusts the signal voltage delivered to the grid
of the 6J5. Operating bias is obtained from the voltage drop across cathode resistors (� 3-6) in all stages except the output stage. Bias for
this stage must be obtained from a steadyvoltage, low-resistance source, such as a battery, to prevent instantaneous changes in biasing voltage when the grids are driven positive and grid current flows (� 5-6).
No by-pass condenser is required across the cathode resistor of the push-pull stage, 14, because, with a circuit of this type, the sum of
the instantaneous cathode currents of both

tubes does not change and, therefore, degeneration cannot take place.
T1 is of the type commonly known as an interstage audio transformer with aturns ratio which provides a step up of signal voltage
from the plate of the single 6J5 to the grids of the push-pull stage. T2 is astep-down trans-
former designed especially to provide the required driving voltage for the 6L6s with good regulation over the excitation cycle (� 5-6). The output transformer, 7%, is chosen with a turns ratio to provide the correct operating load resistance for the 6L6s, depending upon the value of the load connected across the secondary of the transformer.
Rio and R11 comprise a voltage divider

6J5

6L6

Output
Car
6L6
1. 270

��22.5

3b0

Fig. 1141 -- Circuit diagram of the 40-watt speech amplifier modulator.

Ci --

screen by-pass condenser.

C8 -- 0.01-pfd. coupling condenser. Ci -- 20-pfd., 50-volt electrolytic cathode by-pass con-

denser.

C4, Ca, Ce -- 8-pfd., 450-volt electrolytic audio-filter

condensers.

Ri -- 5-megohm, 3/2-watt grid resistance. 118 -- 1300-ohm, 34-watt cathode biasing resistance.

R5 -- 50,000-ohm, 34-watt plate decoupling resistance. Re -- 1-megohm volume control. R7 -- 1500-ohm, 1-watt cathode biasing resistance. R8 -- 750-ohm, 1-watt cathode biasing resistance. 1:12 -- 12,000-ohm, 1-watt decoupling resistance. Rio -- 20,000 ohms, 25-watt (see text). Ru -- 1500 ohms, 10-watt (see text). Ii -- Interstage audio transformer, single plate to p.p.

Rs -- 1.5-megohm, 34watt screen voltage-dropping

grids, 3:1 ratio.

resistance. Ro -- 0.25-megohm, 34-watt plate load resistance.

T2 -- Driver transformer, p.p. 6J5s to 6L6s, Class AB2. T8 -- Output transformer, multi-tap.

218 CHAPTER E-LEVEN

e cho equipment

�CLASS-B MODULATORS

Class-B modulator circuits are practically identical, no matter what the power output of the modulator. The diagrams of Fig. 1142, therefore, will serve for almost any modulator of this type. The circuit for triodes is shown at

A, while that for tetrodes is shown at B. When small tubes with indirectly-heated cathodes are used, the cathodes should be connected to ground.

An output transformer should be chosen

I/5V

-H V

MOO +14 V To Had Amp. Plate

which will permit matching (� 5-3) the rated modulator load impedance to the modulating
impedance of the r.f. amplifier and, similarly, adriver transformer selected which will couple the driver stage properly to the Class-B grids.

Driver plates or /,re

The input transformer, 7'1, may couple

-r

directly between the driver tube and the modulator grids or may be designed to work from a

low-impedance (200- or 500-ohm) liae. In the

latter case, atube-to-line output transformer

for Mod Amp

must be used at the driver stage. This type of coupling is recommended only when the driver

Fd Trans'
115 V --NV

+5.6. 500+ 5V.

must be at a considerable distance from the modulator, because the second transformer not only introduces additional losses but also further impairs the voltage regulation.

Fig. 1142 -- Class-B modulator circuit diagrams. The circuit for triodes is shown at A, the circuit for tetrodes at B.
(� 2-6) for obtaining lower plate voltage for the tubes of the first three stages. R5 and Rg and Cg, C6 and Cg form a filtering system for prevention of coupling between stages via power circuits (� 5-9).
The transformers and tubes are arranged on the chassis so as to permit reasonably short connecting leads. The input circuit of the 6J7 is shielded as completely as possible to prevent the picking up of hum (� 5-9).

Condenser C1 in these diagrams will give a "tone-control" (� 7-5) effect and filter off highfrequency side-bands (splatter) caused by distortion in the modulator or preceding speech-amplifier stages. Values in the neighborhood of 0.002 to 0.005 �M. are suitable. The voltage rating should be adequate for the peak voltage across the transformer secondary. The plate by-pass condenser in the modulated amplifier will serve the same purpose.
The plate power supply for the modulator should have good voltage regulation and

The transformers are arranged with their cores at right angles to reduce

the possibility of feed-back coupling. Resistors and condensers are mounted underneath the chassis.

The voltage gain provided by the unit is sufficient to operate the output stage at a rated power output of 40

watts from the input signal from a crystal microphone.
Besides heater voltage for the six tubes, the power unit supplying this amplifier should deliver 360 volts with good regulation over the current range of approximately 140 ma. to 265 ma.
The power output obtainable is

sufficient to plate modulate an input

of 80 watts to a Class-C r.f. amplifier. It may also be used to drive ahigh-power Class-B modulator.

Fig. 1143 -- A conventional chassis arrangement far low and medium power Class-B modulators. The layout follows the circuit diagram.

CHAPTER ELEVEN 219

DneRadio Amaleur'� ilanclhooh

should be well filtered. It is particularly important, in the case of atetrode Class-B stage, that the screen supply have excellent regulation to prevent distortion. The screen voltage should be set as exactly as possible to the recommended value.
When "C" bias is required for the modulator, the bias source must have very low resistance. Batteries are the most suitable source of bias. In cases where the voltage values are right, regulator tubes such as the VR-75, VR105, etc., may be connected across atap on an a.c. bias supply (� 8-9) and will hold the bias
5V,,A.0 Fig. 1144 -- Circuit diagram of the combination 1000
and 400-volt power supply.
C1, CI-- 2 �M., 1000-volt. C3 -- 4 dd. elect rot% tit', 600-volt working. C4 -- 8mfd. electroly ti,, 600-volt working. Li, L3 -- 5/20 hy. swinging choke, 150-ma. 122, L4 -- 12 hy. smoothing choke, 150-ma. Bi -- 20,000 ohms, 75-watt. R2-- 20,000 ohms, 25-watt. Ti -- High-voltage transformer, 1075 and 500 volts
r.m.s. each side of center, 125- and 150-ma. simultaneous current rating. T2-- Filament transformer, 2.5 volts, 5-amp. 1.2-- Filament transformer, 5volts, 4-amp.

voltage steady under grid-current conditions. Generally, however, zero-bias modulator tubes are preferable not only because no bias supply is required but also because the loading on the driver stage is less variable and driver distortion is consequently reduced.
An example of modulator construction is shown in Fig. 1143.
� POWER SUPPLY UNITS
An example of medium-power plate-supply (see �8-1 to 8-7) construction is shown in Fig. 1145. The unit pictured varies slightly
from the types most frequently en. countered in that it is a duplex sup-
ply from which two independent voltages may be obtained. Either section, however, by itself is typical of the form in use in the great majority of installations.
The circuit diagram is shown in Fig. 1144. Operating from the low-voltage taps of the secondary of the high-voltage transformer, T1,is atype-83 full-wave rectifier, the output of which is fed into the double-section, chokeinput smoothing filter, consisting of L3-C 3 and L4-C4.The filament voltage for the rectifier tube is supplied by the filament transformer, T3.A pair of type-866 jr. half-wave rectifiers, whose filaments are supplied from T2,operates from the high-voltage taps of T1 and the output is fed through asecond similar filter system consisting of L1-Ci and L2-C2. The resistances R1 and R2 are the bleeder resistances designed to assist in voltage regulation of the output and to discharge the filter condensers when the power supply is turned off.

Fig. 1145 -- This power supply makes use of a combination transformer and dual filter s` stem delivering 1000 volts at 125 ma. and 400 volts at 150 ma. simultaneously. The circuit diagram is shown in Fig. 1144. The 1000-volt bleeder resistance is mounted on the rear edge of the chassis with aprotective guard made of screening to provide ventilation. Safety terminals are used for the two high-voltage
terminals. Ceramic sockets should be used for the 866 jrs.

220

CHAPTER ELEVEN

The arrangement of power-supply com-
ponents is seldom critical. The important points of construction in the unit shown are the use of 'ceramic sockets for the high-voltage
rectifiers (where the insulation of bakelite or fibre sockets would be insufficient to insure against voltage break-down to the chassis), the placing of the high-voltage bleeder resistance, RI, where it may be adequately ventilated and the use of specially-insulated posi-

knit� equipment
SI
C2

RFC, C3

RFC,

Ailiot." A'gero-ure 114.

Fig. 1147 --Wiring diagram of the VP-552 vibrator supply.

Ci --

50-volt vibrator hash suppression.

-- 0.007-lifd., 1600-volt buffer. C3 -- 0.02-�fd., 1000-volt r.f. filter.

Iti -- 5,000-ohm, 1-watt buffer.

RFC' -- Low-resistance "A" hash filter.

RFC2 -- "B" hash filter choke.

Si -- Two-pole 4-position voltage selector.

T -- Transformer.

V -- Vibrator.

Fig. 1146 -- The Mallory VP-552 vibrator supply operates from a6-volt d.c. source and delivers 300 volts at 100 ma. maximum. Lower values of voltage and current can be obtained by proper setting of the switch on the side of the chassis. This particular unit includes no output filter but does contain all of the necessary hash filters.
tive high-voltage terminals and a protective screen around R1 to reduce the chances of accidental contact by the operator.
The plugs showing in the rear edge of the chassis are for the 115-volt power connections for the plate transformer and the two rectifier filament transformers which are mounted underneath the chassis.
� EMERGENCY POWER SUPPLIES
For emergency and field operation, gasolineengine-driven generators are almost ,universally used when the power demand is above 100 watts. However, for low-powered operation above 5 watts (dry batteries are generally used when the power demand is less than 5

watts) the most universally acceptable selfcontained power source is the storage battery. It has high initial capacity and can be recharged, so that its effective life is practically infinite. It can be used to provide filament or heater power directly, and plate power through
associated devices such as vibrator-transformers, dynamotors and genemotors, and a.c. converters. For emergency work, a storage battery is a particularly successful power
source, since practically no matter what the circumstances, such batteries are available. In a serious emergency it would be passible to obtain 6-volt storage batteries as long as there were automobiles to borrow them from. For this reason, the 6-volt storag& battery makes an excellent unit around which to design the low-powered portable or emergency station.
For maximum efficiency and usefulness, the power drain on the storage battery should be
limited to 15 or 20 amperes from the ordinary 100- or 120-ampere-hour, 6-volt battery. This should provide acarrier power when transmitting of 20 to 30 watts, which is usually adequate. In connecting the battery, heavy leads of the automotive-cable type should be used
to minimize the voltage drop; ordinary carreceiver leads are definitely not satisfactory. Similarly, heavy-duty low-resistance switches are required.
Vibrator Power supplies--The vibrator power supply consists of a specially-designed transformer combined with a vibrating interrupter. When the unit is connected to astorage battery, the circuit is made and reversed rap-

221 CHAPTER ELEVEN

ifiLe leach� AmaleuA ilandiooh

idly by the vibrator contacts and the squarewave d.c. which flows in the primary of the transformer causes an alternating voltage to be developed in the secondary. This highvoltage a.c. is in turn rectified, either by a vacuum-tube rectifier or by an additional synchronized pair of vibrator contacts, and
filtered, providing outputs as high as 400 volts at 200 ma. Tube rectifiers are ordinarily used
only when the negative side of the circuit cannot be grounded, arequirement with the self-
rectifying type. The high-voltage filter circuit is usually identical with that of an equivalent power source operating from the a.c. line. Noise-suppression equipment, serving to minimize r.f. disturbances, is incorporated in
the manufactured units. Some of the commercial units include ahum
filter and some do not, but the design of this filter is, for the most part, conventional. A typical commercial unit of the self-rectifying type is shown in Figs. 1146 and 1147. The vibrator supplies used with automobile receivers are satisfactory for receiver application but usually are not desirable for use with a transmitter except where the power requirements are slight. The efficiency of vibrator packs runs from 60% to 75%. Vibrator supplies are not intended to withstand much overloading, but fusing of the battery cable will

Fig. 1149 -- Rear view of the 150-watt transmitter. Immediately below the antenna-tuner unit at the top is aunit containing asingle-tube final amplifier, which is driven by the multi-tube all-band exciter unit below. The speech-amplifier and modulator units may be seen below the exciter, with one of the power-supply units showing at the lower edge of the photograph.

Fig. 1148 -- A complete 150-watt rack transmitter for 'phone and c.w.

eliminate any danger of failure through over-
loading. Dynamotors and Genemotors-- A dyna-
motor is adouble-armature high-voltage generator, the additional winding operating as adriving motor. It is usually operated from a6-, 12- or 32-volt battery, and may deliver voltages from 300 to 1000 or more. Dynamotors have been widely used in military work and many of those in amateur use derive from such ori-
gins. The genemotor is arefinement of the dyna-
motor designed especially for automobile receiver, sound truck and similar applications. It has found wide acceptance among amateurs as a source of transmitting power, having good regulation and efficiency combined with economy of operation. It is also used in connection with portable receiver installations, although a rather high inherent noise level limits this application in sensitive amateur high-frequency
receivers. Genemotors are made to fill almost every
need. Their cost, at amateur net prices, runs from about eight to twenty-four dollars. Stand-

222 CHAPTER FI EVEN

Pail� C9uipment

ard models range from 135 volts at 30 ma. to 300 volts at 200 ma. or 500 volts at 200 ma. Parallel and series operation of identical units to provide higher capacity is entirely practical. The normal efficiency averages around 50%, increasing to better than 60% in the higherpower units. The regulation is comparable to
Fig. 1150 -- The Flallierafter IIT6 transmitter. well-designed a.c. supplies; it is largely dependent upon external IR drops.
Successful operation of dynamotors and genemotors implies heavy, direct leads, mechanical isolation to reduce vibration, and thorough r.f. and ripple filtration (the purchase of manufactured filter units is recommended). The shafts and bearings should be thoroughly "run in" before regular operation is attempted, and the tension of the bearings should be checked occasionally.

watts to the output amplifier is the maximum allowed amateur stations by law. The output of the transmitter may be keyed for telegraphic
communication or modulated, by one of several different systems, for voice or, in the case of broadcasting stations, for voice and musical transmissions.
One of the simplest forms of transmitter for covering more than very short distances is the simple oscillator transmitter shown in Fig. 1125. Since the operation of crystals ground to frequencies lower than 10 Mc. is more reliable than that of higher-frequency crystals, it is common practice to use the lower-frequency crystals in conjunction with frequency-multiplying stages in the transmitter when output at the higher frequencies is desired. This practice is also often followed in a transmitter employing a variable-frequency self-excited oscillator, because better frequency stability may be obtained with the oscillator operating at alower frequency, and in amateur transmitters, which operate in harmonically-related bands, to obtain output in any one of several bands from a single crystal. Such frequency multiplication is often done at low-power levels with a series of relatively-small tubes in aunit commonly referred to as an exciter. An example of such an exciter is shown in Fig. 1127.
When greater output than that obtainable from the output stage of the exciter is desired, an amplifier, such as the one shown in Fig.
1130, operating at the output frequency of the

� COMPLETE Tit
TEO!.
A complete radio transmitter may range in size and power from a simple receiving-tube oscillator of almost pocket size coupled directly to the antenna and operating from low-voltage battery supply to large multistage installations feeding elaborate antenna systems and operating with a power input of several hundred kilowatts, although an input of 1000

Fig. 1151 -- Wiring underneath the 1116.

223 CHAPTER ELEVEN

5heRadio -AmaleuA ilanihooh

exciter may be added to step up the power de-
livered to the antenna. When modulation of the output wave is
desired, audio-frequency units, such as those shown in Figs. 1140 and 1143, are added, which, together with the power-supply units,
complete the transmitter. Transmitters may be found built in many
different forms, depending upon the service for which they are designed, but all types will be found to follow the same general plan here outlined. One commonly-used arrangement is the rack-and-panel system pictured in the photographs of Figs. 1148 and 1149. In this system, the transmitter is made up of several units of standard dimensions which fit into
a pre-drilled frame or rack also of standard dimensions. The arrangement permits the removal or change of certain units without disturbing other units of the transmitter.
A commercially-built transmitter, of which a wide variety is normally available for various services is shown in the photographs of Figs. 1150 and 1151. It is completely self-contained with modulator and power supply and has an output rating of 25 watts. Provision is made for obtaining output in any amateur band from 1.75 Mc. to 56 Mc. with suitable crystals and coils.
The r.f. circuit employed in this unit is somewhat similar to that of Fig. 1129. A type
6L6 tube is used in the oscillator to drive an 807 output amplifier. Grid and cathode connections are brought out from the oscillator to allow avariety of circuits. On the four lower amateur-frequency bands, the 6L6 is operated as a simple crystal oscillator with output at the crystal frequency. If desired, the crystal may be replaced by a special plug-in unit which converts the circuit to that of an elec-

tron-coupled, variable-frequency oscillator. For 28-Mc. output, the oscillator circuit is
converted to the Tri-tet type by means of a special plug-in unit, so that oscillator output at the second harmonic of a 14-Mc, crystal may
be obtained. For 56-Mc. output, the 6L6 is used as a
frequency doubler, while a special unit consisting of a separate crystal oscillator is
plugged in. At all frequencies the 807 is operated as a
straight amplifier. No tuning controls appear on the front
panel. The various plug-in tank circuits are broadly tuned to permit satisfactory output
to be obtained at any frequency within the specified band without the necessity for accurate tuning for that frequency. Provision is made so that pre-tuned units for any three bands may be plugged in simultaneously and output in any one of these bands may be selected by aswitch on the front panel.
The audio section consists of a two-stage triode speech amplifier, with sufficient gain for a crystal microphone, and a modulator using apair of 6L6s as Class AB power ampli-
fiers. The plate and screen of the 807 are modulated simultaneously.
The cabinet contains two power units, one supplying the r.f. section, while the other sup-
plies the audio section. A milliammeter is provided which may be switched to read the plate current of the oscillator, r.f. amplifier or modulator, or grid current of the r.f. amplifier.
The output-coupling system is suitable for use with resonant antenna systems or systems employing "fiat" transmission lines. Apparatus must be supplied externally when the unit is to be used with antenna systems requiriag
tuning.

224 CHAPTER ELEVEN

CHAPTER TWELVE
Meaittremenli an] Mectiuring
equipment

�12-1 MEASUREMENT TECHNIQUE
Taz NATURE of radio and electronic equipment is such that visual inspection is of little value in determining satisfactory performance. A wide variety of test and measuring apparatus, based on the same principles as are used in the equipment itself, has been devised to enable accurate evaluation of the performance of experimental equipment as well as to facilitate production testing and adjustment and, ultimately, efficient maintenance in the field.
Because of the variety of such measuring and test equipment and its widespread application in all branches of the art, it is essential for the radio technician to be familiar with its principles and use.
Fundamentally, the process of measurement is that of comparing aquantity with areference standard. Measuring equipment divides into two types: (1) fixed standards giving areference point of known accuracy, with associated
equipment for making comparisons, and (2) direct-reading instruments or meters calibrated
in terms of the quantity being measured. The basic quantities to be measured are: (1)
frequency; (2) the primary electrical quantities: current, voltage and power; (3) circuit constants: resistance, impedance, capacity, inductance and Q; and (4) waveform. This chapter will describe commonly-used methods of measuring these quantities, and the application of these methods to the testing of specific types
of equipment: receivers, transmitters, antennas, tubes.

source, either matching it directly by varying a calibrated source (heterodyne frequency meter), or measuring the difference between it and a fixed source (frequency standard), the frequency of which is known with high precision, by interpolation.
Calibrated Receiver --In the absence of
more elaborate frequency-measuring equipment, acalibrated receiver may be used to indicate the approximate frequency of an oscillator. If the receiver is well-made and has good inherent stability, aband-spread dial calibration can be relied on to within perhaps 0.2 per cent. Some manufactured models having factory calibration may be used to even closer limits. For most accurate measurement the oscillator should be unmodulated and maximum response in the receiver indicated by acarrier-operated tuning indicator (� 7-13), the receiver beatoscillator being turned off.
In checking transmitting frequency the receiving antenna should be disconnected. If the signal is too strong and blocks the receiver, the transmitter frequency may be checked by listening to the oscillator, with power amplifier turned off.
Absorption frequency meters -- The simplest type of frequency meter consists of acoil and condenser, tunable over the frequency range desired. A frequency meter of this type, when tuned to the frequency of the transmitter and loosely coupled to the tank coil, will extract asmall amount of energy from the tank. The energy thus extracted can be used to light a small flashlight lamp. Maximum current will

� 12-2 FREQUENCY MEASUREMENT
Frequency (� 2-7) is measured by counting the number of cycles or oscillations per secemd. Since this cannot be done directly, except at very low frequencies, in practice the measurement is made (a) by noting the response of a selective resonant device, such as a tuned circuit (absorption frequency meter, Wien bridge, etc.) or mechanical resonator (tuning fork, vibrating reed, etc.) previously calibrated in terms of frequency, or (b) comparing the unkne-vn with aknown frequency from aseparate

flow in the lamp when the frequency meter is tuned exactly to the transmitter frequency, hence the brightness of the lamp indicates resonance. A more accurate indication may be
obtained by substitution of a thermo-galvanometer or vacuum-tube volt-meter as the indicator. A crystal detector can also be used.
Although this type of frequency meter is not well adapted to precise measurement of frequency, it is useful for checking (1) the fundamental frequency of an oscillating circuit, (2) presence and order of amplitude of harmonics, (3) frequency of parasitic oscillations, (4) neu-

225 CHAPTER TWELVE

..7ne /each� Amateur's Jjanclbooh

Fig. 1201 -- A simple absorption frequency meter circuit is shown at left. It is used in transmitter checking with link line coupling to the circuit being checked. Circuit at right has bulb indicator loosely coupled to tuned circuit, giving asharper resonance point.
B -- 1.4-volt 50-ma. dial light. C -- 150--ggfd. variable. L-- Coils covering high-frequency spectrum with overlapping ranges, wound on 1%-inch dia. forms.

Freq. Range 1.1-3.5 Mc. 2.5-8.0 Mc. 4.5-14 Mc. 7.5-25 Mc. 22-70 Mc. 40-120 Mc.

Wire Size No. 28 e. No. 24 t. No. 20 t. No. 16 t. No. 16 t. No. 16 t.

No. of Turns 81% 37% 17%
8%

Length of Winding
174" 1%1" 1142" 1%"

Link' 17 t. 11 t. 6 t. 4 t. 2 t. t.

1Closewound, No. 30 d.s.c., i`-inch from bottom end of primary winding.

Fig. 1202 -- A sensitive absorption frequency meter with a crystal detector-rectifier and d.c. milliammeter indicating circuit. Individual calibration charts mounted on each coil form make the meter direct-reading. A toggle switch connects a10-ma, shunt across the 0-1 ma. meter; the 10-ma. range is used for preliminary readings, to avoid burning out meter or crystal. The meter gives indications several feet from alow-power oscillator.

tralization of an amplifier, (5) field strength on
aqualitative basis, (6) presence of r.f. in undesired places such as power wiring, or any other application where detection of asmall amount of r.f. and measurement of its frequency provides useful information.
Calibration of the absorption frequency
meter is most easily accomplished with a receiver of the regenerative type to which the coil in the meter can be coupled. With the detector oscillating weakly, the frequency meter should be brought near the detector coil and tuned over its range until a setting is found which causes the detector to stop oscillating. The coupling between meter and receiver
should then be loosened until the stoppage of oscillations occurs at only one spot on the
meter tuning dial. The meter is then tuned to the frequency at which the receiver is set. If
the receiver is set on several stations of known frequency, anumber of points for acalibration curve can be obtained for each coil.
The same method may be used with asuperheterodyne receiver, but it is necessary to remember that the oscillator frequency differs from the signal frequency by the intermediate frequency. For instance, if the receiver dial reads 6500 kc. and the receiver i.f. is 456, the oscillator frequency will be 6956 kc., which is the frequency to be marked on the meter calibration scale. It is necessary to know if the
oscillator is on the high or low side of the incoming signal; in most receivers the high side is

used throughout, but some receivers shift to the low side on the high-frequency ranges.
If the oscillator coils in the receiver are not accessible, the frequency meter may be capacity coupled through afew turns of insulated wire wrapped around the frequency-meter coil with one end of the wire placed near the stator plates of the oscillator condenser.
For transmitter frequency checking, aflashlight lamp or other indicator is not entirely necessary, since resonance will be indicated by achange in the plate current of the stage being checked as the meter is tuned through resonance. However, for locating parasitic oscillations, determining the relative amplitude of harmonics, checking neutralization, locating
D
Fig. 1203 -- Absorption frequency meter with crystal' detector indicator. CI 140.riafd. variable. CR 0.001-pfd. mica. D -- Fixed crystal detector (Philmore). Li, La -- Same as in Fig. 1201.* M -- 0-1 ma. d.c. milliammeter. Ri -- 3-ohm shunt; see general data on meter shunts. S-- S.p.s.t. toggle switch.
Crystal polarity must be determined by experiment; if meter reads backwards, reverse crystal connections.
*Experiment with number of turns on Lx for maximum current indication is necessary to compensate for variations in impedance of individual crystal detectors.

226 CHAPTER TWELVE

ril and ea�uremenb

Meadurinq equipment

stray r.f. fields, etc., a sensitive indicator is indispensable.
The inherent errors in the absorption-type frequency meter ordinarily limit its useful accuracy to about 1%.
Lecher wires -- At ultrahigh frequencies it is possible to determine frequency by actually measuring the length of the waves generated. The measurement is made by observing standing waves on a two-wire transmission line or Lecher-wire system. Such a line shows pronounced resonance effects, and it is possible to determine quite accurately the current loops (points of maximum current) as shown in Fig.
1204. The distance between two consecutive

lamp shows a sharp loss in brightness. This point should be marked and the shorting bar moved out until asecond dip is obtained. The distance between the two points will be equal to half the wavelength. If the measurement is made in inches, the frequency will be

Fmc.

5906 length (inches)

If the length is measured in meters

Fme .--

150 length (meters)

In checking asuperregenerative receiver, the Lecher wires may be similarly coupled to the

Fig. 1204 -- Lecher wire system for measuring wavelength at u.h.f. Typical standing-wave distribution is shown, with positions of the shorting bar at current loops indicated. The distance "X" equals one-half wavelength.

current loops is equal to one-half wavelength. Thus the wavelength can be read off directly in meters (inches x39.37 if ayardstick is used) or centimeters for the very short wavelengths.
The line should be at least awavelength long at the lowest frequency to be measured and entirely air-insulated except where it is supported at the ends. The wires may be stretched tightly between any two convenient supports, using a spacing of 1to 1% inches. The positions of the current loops are found by means of a"shorting bar" which is slid along the line to vary its effective length (Fig. 1205).
Resonance indications may be obtained in several ways. A convenient and fairly sensitive indicator can be made by soldering the ends of aone-turn loop of wire to aflashlight bulb, then coupling the loop to the tank coil until the bulb glows moderately brightly. A similar coupling loop should be connected to the ends of the Lecher wires and brought near the tank coil.
Then the shorting bar should be slid along the wires outward from the transmitter until the

receiver coil. As the bar is slid along the wires aspot will be found where the receiver goes out of oscillation. The distance between two such spots is equal to ahalf wavelength.
Lecher wire measurements may easily be
made to an accuracy of 1% or better. If sufficient care is used, measurements accurate to
0.1% at 112 Mc. are possible, representing a linear distance of about 1 millimeter. This is accomplished by loosening the coupling for the
final adjustment until indications are just discernible. It is helpful to use ahighly sensitive indicator. The crystal-detector absorption frequency meter previously described will enable closer measurements when used as aresonance indicator than will the flashlight bulb indicator, for example.
Heterodyne frequency meters --For more accurate measurement of transmitter frequency, aheterodyne frequency meter should be used. This is asmall oscillator with aprecise
frequency calibration covering the lowest fre�quency band used, completely shielded. It

Fig. 1205 -- One end of a typical Lecher wire system, showing turnbuckles for maintaining tension. Wires are No. 16 bare copper. The shorting bar is of brass with asharp edge for better contact and precise indication; the slider and side-guides keep the bar at right angles to the wire. A horizontal strip of bakelite at the back keeps the wires right against the bar. A transparent celluloid centimeter rule is cemented to the front of the slider. The beam is marked off in decimeter (10 cm.) units. The sum of the reading of the slider and the lowest adjacent numeral on the beam gives the wavelength.

227 CHAPTER TWELVE

Dhe Radio AmaleuA -Jiancgoon

R=D. R=.
ec x

65.17

RFC

SIGNAL INPUT
110

65A7

fl.43
Phone.S

Fig. 1206 -- Electroncoupled heterodyne frequency meter circuit diagram.
-- 350-pmfd. zero temp.

Cs -- 40-pmfd. negative temp.

Ca -- 100-apfd. midget.

W1C6 C 1CC <='

C4 -- 50-gpfd. trimmer. Ca -- 100-add. silv�r-mica.

X 1

`I!.L-'

Co, Cs -- 0.005-pfd. mica. C7, Co, CIO -- 0.002-Add. mica.

Cii -- 25-pfd., 25-volt electrolytic.

-- 0.01-�fd., 400-volt paper.

Cis --

400-volt paper.

Cis -- Dual 8-dd., 450-volt electro-

L2
00000
--T vR 'so

lytic. RI, Rs --1 megohm, 3/2-watt. Ra -- 20,000 ohms, 1/2-watt. R4 -- 150 ohms, %-watt. Rs -- 5000 ohms, 1-watt.

Re -- 50,000 ohms, 2-watt. R7 -- 3500-ohm, 10-watt wire-wound.

Rs -- 2500-ohm, 5-watt wire-wound.

Li -- 60 turns No. 24 d.c.c., close-wound on 1%-in. dia, form tapped at 12 turns.

-- 10-henry 40-ma. filter choke. RFC -- 65 turns No. 28 e. close-wound, 9,16-in. dia.

T -- 300-volt 50-ma, power transformer.

must be so designed and constructed that it can be accurately calibrated and will retain its calibration over long periods of time.
The signal from this oscillator (or a harmonic thereof) is fed into areceiver or simple detector together with the signal to be measured, and the two frequencies are heterodyned. When the frequency meter oscillator is tuned to zero beat with the signal, its frequency or the harmonic multiple is the same as the unknown.
The oscillator used in the frequency meter must be very stable. Mechanical considerations are most important in its construction. No matter how good the instrument may be electrically, its accuracy cannot be depended upon if it is flimsily built. Inherent frequency stability can be improved by avoiding the use of phenolic compounds and plastics (bakelite, polystyrene, etc.) in the oscillator circuit, employing only high-grade ceramics for insulation. Plug-in coils or switches are not ordinarily used; instead, a solidly-built and firmlymounted tuned circuit is permanently installed and the oscillator panel and chassis reinforced
for rigidity. To obtain high accuracy the frequency meter
must have adial that can be read precisely to at least one part in 500; ordinary dials like those used on transmitters and inexpensive receivers are not capable of such precision without the addition of vernier scales. The dial should have fine lines for division marks, and an indicator set close to the dial scale so that the readings will not appear different because of parallax when the dial is viewed from different angles.
A stable oscillator suitable for use in the frequency meter is one using the electron-
228 CHAPTER TWELVE

elea.lurement.4. and Meaiuring

att.pntent

number of points have been established, they
may be marked on graph paper and acalibration curve drawn. For maximum convenience, adirect-reading dial scale can be constructed.
If no frequency standard is available, calibration points may be obtained from other
sources of known frequency, such as the transmitter crystal oscillator, harmonics of local broadcasting stations, or even checks by other amateurs on the air. As many such points as possible should be secured, so that inaccuracies will average out.
In use, the signal from the frequency meter can be fed into the receiver by connecting a wire from the plate of the oscillator through a very small capacity to the input of the receiver. The signal to be measured is then tuned in in the usual way and the frequency meter adjusted to zero-beat.
For convenience in checking the frequency
of the transmitter or other local oscillators which generate sufficiently strong signals, it is desirable to incorporate a detector in the frequency meter which will combine the signals
and deliver the audio beat-note output to headphones or a visual zero-beat indicator. A frequency converter tube such as the 6L7 or 6SA7 is especially suited for this purpose.

With astable oscillator, a precision dial and frequent and careful calibration, an overall ac-
curacy of 0.05 to 0.1% may be expected of the heterodyne frequency meter. The principal limiting factors are the precision with which the calibrated dial can be read and the "reset" stability of the tuned circuit.
Frequency standards -- To make more precise frequency measurements, particularly of amateur-band limits, asecondary frequency standard is required. This is a highly stable low-frequency oscillator, usually operated at 50 or 100 kc., the harmonics of which provide reference points every 50 or 100 kc. throughout the spectrum. A 1000-kc. frequency is often added to facilitate preliminary identification of frequency ranges, especially on u.h.f.
An electron-coupled oscillator built according to the principles previously outlined for frequency meters, equipped with a tuned circuit for 50 or 100 kc., will serve as asimple and inexpensive standard. A standard of this type is inherently more accurate than aheterodyne frequency meter because (a) the low-frequency
oscillator has better inherent stability and (b) the frequency setting once made is not thereafter changed, eliminating the re-set and calibration errors.

6SJ 7 C2

65 C7

j CSGC

Ri2

Ria �-Wner.000,

Cil

vF 105

63V 4// litnr Fig. 1207 -- C rcuit diagram of aprecision crystal-controlled 100-kc. frequency standard. LCCCCCCC2oaISSSTI,,S,,------h--CCCCei7So7PS5D0,l---c----0.iuluhprCp-0nigeaayia0pn-n0spl1i0r2ckfirt.e0,.e-tnday-0r30cpiCil0,.1.6esfbii1-ic5dpr4-pomav-i.foip0fsedpid--ladfr-jp.md..ddamfuig,.ctmdasd0eaui.tB,.4gtmsndl1e0eivtei-fgdm0atdilpde-irltgfetttvcimydyeeooaatip.rtmblo.ecSg4ilct.amie0chfO.iv,0paorcCe-a.kecavpeo-q.doe.r1uelrn0est.-0inrcey(doIsR1IRCRCCitsc1tas1iino4oiiZ,2a,,lg,------lun--RRR--asaCt7aseli0o1.1.sr31------5s3-m4,tO--c3er00muo0025eg0it-e-.00l8n0o5Ap,dg,.g-ph00pudotimoff00mthn.hedhd00,m..mgv,oo,oaYpthh4rh2orcpmbmmsA5i-mtiioesisa0-awreim,b-c,,wmacnnmlsvrutaetteeu1o1iAtoti.s-brlt-.-htomwst.wwe.wmXaotaaineetiut-lltrttttnle.Xteyu.tcr.tt.i.ecuniodnngie)lft.oth(roeNFto1oTSRRRdRRh.R0riiiii1eiF,0naaa23o--0cCSgc0----h-- --a--rrk,ey--acPdcsm..S2521o80tkc.105.5a0w..i11,,l10c0ne--58W.000g.rmu00oionme0o0hStt1inh.ghmrh0toopm.soa,0.hh1sn,s0hCmmAcr.s,-mi.stsof-k3f.,,,.onc/1in2r.n0tnce-1m31oe-chw/a--egwcpaoh2wgrwroatt-klaa,itteweefttmn.tad.2i.totta..5srtbx,.0mei)tvam.wmu,e1ame45,yn00

229 CHAPTER TWELVE

Dh e Palio Antaieuris Jiancitooh

For highest accuracy, the most suitable in-

SIGNAL

strument is a crystal-controlled standard,

(I)

provided with a 10-kc. multivibrator (� 3-7) for frequency division. Such a standard will

mark 10-kc. intervals throughout the com-

munications spectrum. The frequency of a

(2)

L SIGNAL

signal can then be checked by noting its location with respect to two adjacent 10-kc. points on the dial of a calibrated receiver or heterodyne frequency meter and estimating the exact

RECEIVER
r IH=ef
METER
SIGNAL
(3)

frequency by interpolation. In the adjustment of the frequency standard
at least a15-minute warm-up period should be allowed. For initial adjustment, its output may be coupled into areceiver operating on the broadcast band and the oscillator tuned to zero beat with a broadcasting station on a frequency that is a multiple of 100 kc. (800
kc., 900 kc., 1000 kc., etc.) If the oscillator is self-excited, a second station 100 kc. away

HETERODYNE FREQUENCY M ET ER

should be checked, to make certain that the
oscillator is working on 50 or 100 kc. rather than another frequency which gives an odd harmonic. Since broadcasting stations are re-

10 KC. I IMULTIVIBRATOR1

quired to stay within 20 cycles of assigned frequency, the maximum error of such a source will be less than 30 parts in one

100 KC. STANDARD

million. For highest precision, the standard may be
calibrated on the National Bureau of Standards WWV standard frequency transmissions, which are accurate to better than 1part in 10

(4)

SIGNAL RECEIVER

million. These transmissions may be tuned in on areceiver operating on 5Mc. (receiver beat oscillator off) and the standard adjusted until its harmonic is exactly at zero-beat with WWV. The calibration should be rechecked whenever

precise measurements are to be made.

10 KC. MULTIVIBRATOR

AF INTERPOLATION
OSCILLATOR

In adjusting the multivibrator, two adjacent 100-kc. points are first noted on the dial of a calibrated receiver. The multivibrator is then

turned on, and its frequency control (R8 in

100 KC. STANDARD

Fig. 1207) set at half scale. The number of separate audio beats between the two marked
100-kc. points is then counted. If it is anumber other than nine (indicating 10-kc. intervals),

Fig. 1208 -- Frequency measurement methods. A frequency meter (with built-in detector) used alone is the simplest arrangement for checking the frequency of local oscillators (1). With areceiver (2) incoming received signals can be measured as well. A heterodyne
frequency meter can also be used as alinear interpolation oscillator in conjunction with a 100-kc. standard (3), with or without a 10-kc. multivibrator. The standard
provides accurate check points on the frequency meter scale. Alternatively, areceiver (if adequately calibrated) may be substituted for the frequency meter. For greatest
precision, method (4) is used with an interpolation audio oscillator having alinear scale.
With careful design and construction, high precision can be attained with methods (3) and (4). Using (3), the

readjust R8 until nine beats are observed. Mark this point. Note also the points on the R8 scale where 8 and 10 beats occur, indicating approximately 11- and 9-kc. separation. The
odd frequencies are occasionally useful in checking frequencies very close to the 10-kc. harmonics where the low beat-frequency makes it difficult to secure zero-beat, particularly when an interpolation oscillator is used. Mathematical calculation is required to de-
termine the exact frequency. In practice the 100-ke. points can usually be
identified as being louder than the 10-kc.

accuracy can be 0.01% (100 parts in amillion). Method (4) is accurate to 10 parts in a million with ordinary equipment; precision laboratory apparatus is reliable.
to better than 1part in amillion.

multivibrator harmonics. This identification
process can be facilitated by applying audio modulation to the 100-kc, signal only, causing

230 CIIAPTER TWELVE'

Meaiuremenb and Illeaduriny equipment

apoints to stand out because of lye tone. Aation -- When measuring exact fre,ds with the aid of afrequency standard multivibrator providing equi-spaced harJnic points, it is necessary to determine the exact location of the unknown frequency by interpolation between adjacent standard harmonics. This can be done (a) by use of acalibrated receiver or heterodyne frequency meter with a scale that is linear with frequency, or (b) by comparison of the audio beat frequency with acalibrated audio oscillator. In method (a), the points at which the unknown frequency and the nearest lower and
higher harmonics appear on the dial of the receiver or frequency meter are noted, as shown
in Fig. 1209. Knowing the exact frequencies of the harmonic points fi and /2,the unknown
frequency, h, can be determined as follows

S0.2 -- o (./.2 - fi)

where Si is the dial setting for fi ,S2 for f2 and S. for 1"..
Method (b) consists of beating the standard and unknown frequencies in a detector and measuring the resulting audio frequency by zero-beating with acalibrated audio oscillator having a linear frequency range covering half the difference between adjacent harmonics (0-5000 cycles with a 10-kc. multivibrator), as shown in Fig. 1209. The measured frequency is then equal to the reading of the audio oscillator added to or subtracted from the nearest

standard harmonic, as the case may be. To determine whether to add or subtract this audio difference it is necessary that the frequency be known to better than 5 kc. from the receiver (or auxiliary heterodyne frequency meter) calibration.
In addition to the beat note resulting from the nearest adjacent harmonic, fi,there will also be another higher beat from fi. However, by tuning the receiver midway between fi and
h, its adjacent-channel selectivity will dis-
criminate against fi and reduce the higher beat note to anegligible level.
The interpolation audio oscillator should have ascale that reads linearly with frequency (as opposed to the logarithmic scale commonly found in laboratory oscillators). A beat-frequency oscillator (Fig. 1228) with a straightline capacity tuning condenser in series with
the correct value of fixed capacity will have a nearly linear scale. Some forms of resistance-
capacity oscillators can also be made to have such ascale.
A suitable detector is apentagrid converter
(� 7-9) with some form of zero-beat indicator in the plate circuit. The interpolation audio oscillator is connected to the oscillator grid, the audio beat note from the receiver being applied to the signal grid.
Zero-Beat Indicators -- Use of the hetero-
dyne method of frequency comparison requires a means for determining when the known and unknown frequencies are synchronized; i.e., when they are at zero beat. The point at which zero-beat occurs can be determined approxi-

fsL..fhio +10
22.1oo Kc =1000 Kc

f2 1 1 I I

+20 +30

+40

+50 +60

+70

+90

MU LTIVI BRATOR 7-1AR MON iCS

rh It

ZERO SEA?'

CYCLES

RECEIVER OR FREQUENCY METER
SCALE

INTERPOLATION OSCILLATOR

o o o

SCALE

Fig. 1209 -- Use of interpolation methods in measuring frequencies between standard harmonics. At top is shown the relative location of the frequency-standard fundamental and harmonics in the spectrum, together with the multivibrator harmonics, as related to the unknown frequency under measurement (J.). At the left is shown asmall segment of this spectrum as it appears on the dial of acalibrated receiver or heterod ne frequency meter, and at right the appearance of the audio oscillator dial when using the comparison andiu beat-note method.

231 CHAPTER TWELVE

DfleRadio Amateur's Jiancgooh

Fig. 1210 -- Wien bridge for audio frequency measurements. Ri -- 2000 ohms, %-watt. R2 -- 1000 ohms, 3.4-watt. Ra, R4 -- 10,000-ohm wire-wound variable. RB -- 25-ohm potentiometer. CI, C2 -- 0.01-pfd. for 2,000-20,000 cycles
0.1-pfd. for 200-2,000 cycles 1.0-mfd. for 20-200 cycles The input transformer should be of the balanced-toground type with an electrostatic shield between primary and secondary.
mately by listening to the output of the receiver or detector with headphones or loud speaker. For greatest accuracy some form of auxiliary visual zero-beat indicator is desirable, however. This may be a rectifier-type a.f. voltmeter with copper-oxide or diode rectifier (� 2-3), a neon tube "flasher" or an electronray tube (� 7-13) with its triode grid connected to the receiver output. The headphones will still be required for preliminary adjustments since the visual indicator usually responds only to frequencies under about 25 cycles.
Audio-Frequency-Measurement -- The measurement of unknown audio frequencies can also be accomplished by either direct or comparison methods. Direct measurements are beet made with an a.c. bridge (� 2-11). The most suitable form is the Wien bridge as shown in Fig. 1210. The a.c. voltage to be measured is applied across the input terminals and the variable resistors Ra and R4 varied simultaneously until balance is obtained as shown by zero response in the indicator. Then

sound through a power amplifier a. � ' speaker and measured by aural comp.. with a properly-tuned piano, rememberi,
that middle C is 256 cycles and each octave above or below doubles or halves the frequency. Intermediate points can be obtained by multiplying each successive half-note above C in
any octave by 1.05946 (e.g., if C is 1, C# equals 1.05946, D equals 1.1225, etc.).
The cathode-ray oscilloscope (� 3-9) is extremely useful in measuring frequencies by the comparison method when a reliable standard

AB
/0

180 �

135� OR 225 �

CDE
0 0\

90 � OR 270 �

45* OR 315 �

.1 REPLHAATSIEONS

2 D

HORIZONTAL TIMING AXIS VERTICAL TIMING AXIS

6:1

9:2

41/2 :I

f --

1 21- R C

where R is the resistance effectively in the cir-
cuit at R3 and C is equal to the value of Ci. Ri and R2 may be ganged on one shaft and provided with a scale calibrated in terms of fre-
quency for each value of C (the ranges being selected by a 2-gang switch). RB is a small balancing potentiometer used to compensate for minor variations in the two arms. The indicator may be a pair of headphones as shown, although a sensitive visual indicator will give greater range and higher accuracy.
Where acalibrated audio oscillator is available, measurements may be made by comparison as previously described in this chapter. If no electrical frequency standard is available, the audio frequency can be converted into

16.3

51/3:1

Fig. 1211 -- Lissajou's figures as used in measuring audio frequencies by comparison with a known source on acathode-ray oscilloscope. Figures A through E illus-
trate the pattern produced by different phase relationships when the two voltages have a1:1 frequency ratio. Figures F through Jshow the same phase relationships with a2:1 frequency ratio, the higher frequency being applied to the vertical plates. The next figure shows a ratio of 6:1, determined by counting the peaks of the waves in the horizontal plane (in this case the higher frequency is applied to the horizontal plates). Complex ratios are identified by one or more crossovers, as indicated by the arrows opposite the 9:2 and 16:3 figures. In principle, frequency ratios are determined by counting both horizontal and vertical peaks (number of cross-overs plus 1). Care must be taken not to confuse the back lines (return trace shown by light line in 6:1 figure) in counting cross-overs. This can be done by counting only the peaks travelling in the same
direction across the screen when the frequency is adjusted so that the pattern rotates slowly.

232 CHAPTER TWELVE

Illeadurement� and MeaJurinq equipment

source is available. Applying voltages from
the unknown and the standard to the opposite pairs of cathode-ray tube deflecting plates results in patterns of varying form termed Lissajou's figures. By proper interpretation of these figures, as shown in Fig. 1211, fre-

plier, it is necessary to know the resistance of the meter. If it is desired to extend the range of avoltmeter, the value of resistance which must be added in series is given by the formula:

R =

(n -- 1)

quency ratios up to 10 to 1 can be obtained conveniently. Thus with a1000-cycle oscillator calibration points between 100 and 10,000 cycles are possible. The 60-cycle a.c. supply can be used as a calibration source up to 600 cycles or so.

where R is the multiplier resistance, R. the resistance of the voltmeter, and n the scale
+���--
<>Multiplier

Frequency Monitoring -- Contrasted with the problem of frequency measurement as a

Shunt

single operation is that of continuously monitoring the frequency of an oscillator, particularly in broadcasting and other single-frequency
transmitters. This requires an automatic device -continuously reading frequency deviation. Common practice is to provide ahighly-stable temperature-controlled standard having a frequency slightly different (usually 1 kc.) from the assigned carrier frequency. Both carrier and standard frequencies are fed into adetector and the resulting beat measured on
a direct-indicating audio-frequency meter. Variations in the audio beat then show devia-

Fig. 1212 -- How voltmeter multipliers and milliammeter shunts are connected,
multiplication factor. For example, if the range of a 10-volt meter is to be extended to 1000 volts, n is equal to 1000/10 or 100.
If a milliammeter is t� be used as a voltmeter, the value of series resistance can be found by Ohm's law (� 2-6)
R -- 1000 E

tions of the carrier frequency in either direction. Alternatively, selective circuits and vacuum-tube relays controlling panel lamps may be used instead of the meter, the lamps flashing to indicate deviations beyond pre-set upper and lower limits.
� 12-3 MEASUREMENT OF CURRENT, VOLTAGE AND POWER
D.c. instruments -- Instruments for measuring direct current (� 2-6) are based on the d'Arsonval moving-coil principle, comprising an indicating pointer moving across acalibrated scale, actuated by the flow of current through acoil located in aconstant magnetic field.
Ammeters and voltmeters are basically identical instruments, the difference being in the method of connection. An ammeter is connected in series with the circuit and measures the current flow. A voltmeter is amilliammeter (ammeter reading one-thousandth of an ampere) which measures the current through a high resistance connected across the source to be measured; its calibration is in terms of voltage drop in the resistance, or multiplier.
The ranges of both voltmeters and ammeters can be extended by the use of external resistors, connected in series with the instrument in the case of avoltmeter or in shunt in the ease of an ammeter. Fig. 1212 shows at (A) the manner in which a shunt is connected to extend the range of an ammeter and at (B) the connection of avoltmeter multiplier.
To calculate the value of ashunt or multi-

where E is the desired full-scale voltage and I the full-scale current reading of the instrument in milliamperes.
To increase the current range of amilliammeter, the resistance of the shunt can be found from the formula:
-- n-- I
where R. is the meter resistance as before. A portable combination milliammeter-volt-
meter having several ranges is extremely useful for experimental purposes and for troubleshooting in receivers and transmitters. As a voltmeter such an instrument should have high resistance so that very little current will be drawn in making voltage measurements. A low-resistance voltmeter will give inaccurate readings when connected across a high-resistance circuit. A resistance of 1000 ohms per volt is satisfactory for most uses; a 0-1 milliammeter or 0-500 microammeter (0-0.5 ma.) is the basis of most multi-range meters of this type. Microammeters having a sensitivity of 0-50 ea., giving a voltmeter resistance of 20,000 ohms per volt, are found in units available at reasonable cost. Multipliers for the various ranges are selected by switches.
The various current ranges on amulti-range instrument are also obtained by using anumber of shunts individually switched in parallel with the meter. Particular care must be taken to minimize contact resistance.
When d.e. voltage and current are accurately

233 CHAPTER TWELVE

r eacho AmateuA ilanihooi

rectifier, which converts a low-resistance 0-1

d.c. milliam meter into ahigh-resistance 0-0.909

a.c. milliammeter, making possible the con-

struction of a.c. voltmeters having asensitivity

of 1000 ohms per volt and an accuracy of

about 5%. The design of multipliers for such a

voltmeter must allow for the fact that the rec-

tifier resistance varies with current. Two scales

are usually provided, one for use above 50

volts and one below. The frequency error is ap-

proximately 0.5% per 1000 cycles.

A.c. power measurements are more complex

than for d.c., the simple multiplication of cur-

rent and voltage being in error unless the load

is purely resistive. If the current and a.c. im-

pedance are known, the power is PZ. For or-

dinary amateur power calculations, such as the

input to a power transformer, the product of

a.c. voltage and current can be considered

sufficiently accurate. Commercial power measurements are made on awattmeter, which is a

Fig. 1213 -- A typical inexpenbive multi-range combination volt-ohm-milliammeter, showing meter with 1-ma, basic range and knobs for range switch and zero re-setting resistor. Flexible test leads with insulatedhandle test probes are used to make connection to the

complex instrument relating average current and voltage in terms of asingle reading.
Rt. instruments-- The measurement of very high-frequency a.c. or r.f. quantities involves special problems. Practical instru-

circuit under test.

o EHXTI.-OBAHTM.

Fig. 1214 -- Circuit of the low-cost V-O-M.

0+ scroov.

Ri -- 2000-ohm wire-wound variable. R2 -- 3000 ohms, %-watt.

~Re2V

0+
o
7

112 -- 10-ma. shunt, 3.6 ohms. R4 -- 100-ma. shunt, 0.33 ohms. Rs -- 40,000 ohms, 3/2-watt. Re -- 4megohms, 4-watt (four 1-meg.,
1-watt resistors in series). R7 -- 0.75 megohm' 1-watt (0.5
megohm and 0.25 megohm, 34-watt in series). Rs --0.2 megohm, 3/2-watt. R9 40,000 ohms, %-watt. Rio -- 10,000 ohms, l/2-watt. SW -- 9-point 2-pole switch (Mallory-Yaxley 3109).

known, the power can be stated by simple ap-

B -- 4.5 volts (Burgess 5360).

plication of Ohm's law: P = EI (� 2-6).

A.c. instruments D.c. meters will not

respond on alternating current, and it is there-

fore necessary either to rectify the a.c. and

measure the resulting d.c. or to use special instruments that will indicate on ac. (� 2-8).
A.c. ammeters and voltmeters utilize the moving iron-vane principle. Since the maxi-

� D.C.
VOLTS

mum sensitivity is 15 to 25 ma., making the

MILLIAMPERES

ohms-per-volt 40 to 67, iron-vane voltmeters

consume substantial power. They are suitable

for measuring filament and line voltages, but

cannot be used in circuits which are unable to sustain a measuring load. Iron-vane meters

are not accurate above afew hundred cycles.

For measurements where iron-vane meters are not suitable, special devices enabling the use of d.c. movements are employed. The most

Fig. 1215 -- Seale of the typical multi-purpose meter. It is read in accordance with the range in use. In this meter the ranges are "high" (0-250,000) and "low"

common of these for the power and audio frequency range is the full-wave copper-oxide

(0-500) ohms. 100, 10 and 1uta., and 10, 50, 250, 1000 and 5000 volts.

234 CHAPTER TWELVE

Meoitiremenb an,/ MeaJaring

ments read in terms of d.c. from aconversion device.
R.f. current is usually measured by means of a thermoammeter. This is a sensitive d.c.
microammeter connected to a thermocouple associated with aheater made of ashort piece
of resistance wire. Thermoammeters have been made with asensitivity of 1ma., but the ranges used by amateurs for measuring antenna current, etc., are from 0-0.5 amperes up.
The most suitable r.f. voltmeter is a peakreading vacuum-tube voltmeter (Fig. 1216). When properly designed, its accuracy is limited at r.f. only by the variation of the input resistance with frequency. The peak diode voltmeter has little error even at 60 Mc. The same is true of the self-biased and slide-back types if tubes having low input capacity are used. The oscilloscope can also be used as an r.f. voltmeter for potentials of several volts or more.
R.f. power measurements can be made by measuring the current through a resistor or reactance of known value (� 4-9). Approximate power measurements may be made by using ordinary 115-volt light bulbs as asubstitution or "dummy" load, connected either singly or in series-parallel to provide the required resistance and power rating. The approximate resistance of the bulb can be computed from its wattage rating at 60 cycles. Special non-inductive resistance units, enclosed in vacuum bulbs mounted on standard tube bases, with resistances of 73 and 600 ohms at power ratings up to 100 watts, are available for this purpose. For higher power the units may be connected in series-parallel.
Where the substitution load method is impractical, r.f. power can be measured by multi-
plying the current through a thermoammeter in the circuit by the r.f. voltage across the circuit as indicated by an r.m.s. meter (or 70.7% of the reading on apeak voltmeter).
Another method of meas�ring r.f. power is the photometric method. A calibrated lightsensitive cell (a photographer's exposure meter is suitable) is used to measure the relative brilliance of an electric light bulb as asubstitution load and its normal brilliance on 115-volt 60-cycle supply.
Vacuum -tube voltmeters -- The most useful instrument for the measurement of both d.c. and a.c. voltages is the vacuum-tube volt-
meter. Its chief advantages are (a) negligible power taken from the circuit under measure-
ment and (b) accuracy over a wide frequency range including r.f.
The v.t.v.m. operates by virtue of the
change in plate current caused by achange in grid voltage (� 3-2). In the measurement of a
d.c. voltage, the voltage--termed the "signal"
-- represents simply a change in grid bias. In

(A)
(8)

Ca14.6rittiort: � o RI

Re.) � .2.22 RI
(a, a.c ono' r�. to 10

Ro0.,-,ornesroher Rinput OR

DIODE RECTIFIER

Calibration: =Rf
E,,,,, oo.00,45

PEAK DICOODOE RECT IFIE R. -SHU NT TYPE

(C)

Calz�ratien �

(ma

Typical Calve`r.f.E.

� 3 L2

C�am -o.siff,1

R 4enephens

GRID RECTIFI ER

04 .4
-0 02 At W. 00 .0 0 404.
1040 - Ei-45.122

Calihration:
Ed, Typicalcurue Axle.

(D) E

C �aim -tee �d

R � i-10

Pc . 2000

nregole...

C'� reMireeelelegea. Cp. 51 wife/a:U.44. ""

PLATE RECTIFIER (SELF -BIASED)

IW
00 -- 7,
10
o 0.4 40 L., Li 20 O 2 4 6 0 .0
E 6J5-Eb R,4 2500 ,-for vot(ede R,s5RooR. for ice bolt mile

(E)

RI-R3 RIOR2 R2> 'note, ,rsistanes BALANCING CIRCUITS

(F)

E, E (max) ez'
...weeny on b, when R, is balanced far initial
-SD l00frIbA1ac. 2i.c.500dui fordC.

- E,

RIa20012 R2� zoo",

per per

volt volt

ooffEE,,

SUDE -BACK PEAK VOLTMETER

Fig. 1216-- Vacuum-tube voltmeter circuits.

the case of a.c., the tube acts as a rectifier and the measurement is in terms of rectified d.c.
Representative vacuum-tube voltmeter circuits are shown in Fig. 1216. The simple diode rectifier (A) can be almost any vacuum tube; in a triode or multi-grid type, all electrodes except the control grid are connected to cathode (or negative filament). A Type 30 or 1G4G tube with a flashlight cell for filament supply makes aconvenient portable unit. A tube with low input capacity (1N5G, 6T7G, 954) should

235 CHAPTER TWELVE

5he Radio �naleur'i ilandhooh

be used for high frequencies. The frequency (or section of the voltage divider, in an a.c.

range is limited by the tube input capacity shunting the load resistance. Calibration is linear above 2 or 3 volts provided the load

power supply) is connected back to the meter through a variable resistor, providing a controllable opposite current flow which can be

resistance exceeds 0.1 megohm. The meter M should be asensitive mieroammeter (0-100 or

made to equal exactly the residual plate current of the tube. When used with (C), this

200 p.a.); a 0-1 ma. meter can be used with balance circuit allows the meter terminals to

reduced sensitivity. The peak diode voltmeter at (B), shunt-

be reversed, thereby making it read forward instead of backward with signal. The resistor

connected to eliminate the need for a d.c. re- R should be not less than ten times the internal

turn in the measured circuit, reads peak a.c. voltage. The input resistance is comparable to

resistance of the meter. At the right in (E) an automatic balancing

that of the simple diode for equivalent sensitivity, but the high-frequency error is less. The time constant of the RC circuit should be at

circuit is shown wherein a duplicate triode (usually the second section of a twin triode) takes the place of the adjustable resistor R.

least 100 for the lowest frequency to be measured (RCF>100). Typical values are 0.5

Current flow through R1and R2being equal and opposite with no signal, there is no poten-

megohm and 0.5 pfd. for audio, and 0.1 megohm and 0.05 mfd. (mica) for r.f. and i.f.
The grid rectification circuit shown at (C)

tial across the meter and consequently no current flow. When avoltage is placed on the grid of the voltmeter triode this balance is dis-

can be considered equivalent to the diode rectifier of (B) followed by a zero-bias triode amplifier. The sensitivity is greatly increased over the ordinary diode. The input resistance

turbed, however, and the meter registers current flow. A small zero-setting resistor, R3, is provided to correct for any discrepancies in tubes or resistors. The values of R1 and R2 will

is low with small inputs (0.1 to 1.0 megohm)

because of grid current. The plate current is at

maximum when idling and decreases with

signal. This circuit is useful chiefly because it can be used with inexpensive meters. The in-

strument can be calibrated from a known

60-cycle source; the scale is square-law for

small signals, becoming linear with increasing

input. The value of R is non-critical. C should have areactance small compared with R at the operating frequency, i.e., 0.01 dd. mica from 1

kc. up, 0.1 pfd. paper for low a.f. For d.c. C is,

of course, omitted. A high-11 tube is preferable,

to reduce the idling or no-signal plate current. The self-biased plate-rectification or reflex

voltmeter at (D) has a very high input resistance and fair sensitivity. It is normally

connected directly across the circuit to be measured; if this does not provide ad.c. return

a coupling circuit must be added as shown by

dotted lines (C -- 0.01 dd., R -- 10 megohms

or more). A

tube is preferable, to min-

imize contact potential and grid current. The

cathode resistance Rc controls sensitivity; the

higher it is the more nearly linear and stable

will be the calibration. A range switch can be

provided, connecting in various values of

cathode resistance from 2000 ohms to 0.5

megohm to give full-scale ranges from about

2.5 to 250 volts. The plate and cathode by-

passes may be 0.001-pfd. mica condensers, the

cathode being shunted by a10-pfd. electrolytic

for 60-cycle calibration and low a.f. measure-

ments. The no-signal plate current present in the

circuits of (C) and (D) can be balanced out by

bridge or bucking circuits, typical forms of

depend on the plate supply voltage available; the higher they are, the better the sensitivity

and stability. The minimum value is several

times the meter resistance.

The "slide-back" voltmeter at (F) is acom-

parison instrument in which the peak value of

an a.c. or r.f. voltage is read in terms of d.c.

substitution voltage; the voltmeter tube and

milliammeter, M ,merely indicate when the

two are equal. With the input terminals

shorted and R1 set so that V reads zero, the

tube is biased nearly to cut-off by adjustment

of R2.The residual plate current is the refer-

ence current (/re.) or "false zero." If an a.c. voltage, E,is now applied across the input

terminals, plate rectification of the positive

peaks will cause the plate current to rise. By

adjusting RI,additional bias voltage may be

introduced to balance out the a.c. voltage. The

additional bias required to bring the plate

current back to the reference value

equal to the peak value of the signal being

measured. In operation R1 should be set (after setting /re.) so that all of E1 is in the circuit,

to avoid burning out the milliammeter when

the signal is applied. After the unknown volt-

age has been connected the bias is reduced by

R1 until the reference current is reached. The

slide-back voltmeter is capable of high accu-

racy and has the advantage of requiring no

a.c. calibration.

Oscilloscopes-- Perhaps the most useful of

all measuring devices is the cathode-ray oscil-

loscope. (� 3-9) Although relatively expensive,

its applications are so numerous that it can

replace a number of other less satisfactory

types of measuring equipment. It can be used

which are shown at (E). An auxiliary battery on d.c., a.c. and r.f., and is particularly suited

236 CHAPTER TWELVE

VERTICAL
o 0

AMPLIFIER MFO
o-- e
AreoLerIER

1.- 2

2306

R ��

00000 s--VezMw--..--vmmme,--

.00poo

a...boon

0 miGH

MOR It ON TA L
0 0

:orr J- 2 0-- AMPLIFIER

AMPLIFIE R GC G
<

gArlaAll HaldVHD

POVVER TRANOFORMER

6C 6 6 66 mCATER6
HY RECTIFIER 80

R -6 WNW os000e.
o

8
e

CATHODE RAY ec,E.
D

110 -120 50-0010 10.1 A
a-5 POWER

elA^ffle

L.V. RECTIFIER 80

e
8
0

�le

i000
- 0000,
L-1 REACTOR

VERTICAL 0-23
C.000 A ORI2Oref
R-22
280,000 a

TCR1RG

oyrerroyar.6 TarnsroreTr...1 Et, 3,0

TIMING

R.,

884

o -,0

*-�

�0.000 �L

2200

TVO

22 1_

0 Orb�
* RANGE -<-6 5.50, 000.0351.c1.0,,0.

C.11 .8 0400

820

T -`o7 879

Fig. 1217 -- Circuit diagram of a commercially-built cathode-ray oscilloscope with sweep circuit and voltage amplifiers. (Courtesy R.C.A.)

Dheleach� -Amateur ilancitooh

teur equipment -- resistors, condensers, coils, etc. -- both as ameans of identification and in checking accuracy. The advanced amateur will also be interested in measuring impedances and the characteristics of devices of his own construction or under other than rated conditions.
Resistance -- The volt-ammeter, ohmmeter and Wheatstone bridge methods are commonly used in measuring resistance. In the volt-ammeter method, the resistance is determined from Ohm's Law by measuring the current through the resistor when a known d.c. voltage is applied. The resistance can be determined with a voltmeter alone if its internal
resistance is known
R = e--RL" -- R.

Fig. 1218 -- A representative oscilloscope of the ty pe widely used in amateur, service and commercial applications, using a906 3-in, cathode-ray tube. The circuit is shown in Fig. 1217. All necessary controls are located on the front panel, including intensity, focus, timing circuit synchronization, frequency and range, and centering, gain and input selection controls for both vertical and horizontal axes. (Courtesy R.C.A.)
to a.f. and r.f. measurements because of the high input resistance and small frequency error.
The oscilloscope is in effect acomplex voltmeter capable of measuring any two voltages simultaneously by the deflection of a weightless electron-beam pointer. Moreover, because this pointer projects its indication on aretentive luminous screen, the measurements include the additional factor of time. It is possible therefore to see the actual form of one or more repetitive cycles of an a.c. voltage by means of the oscilloscope, and to measure thereby not only its amplitude but also its fre-
quency and waveform. When used as asimple voltmeter the signal
is applied to the vertical plates and its amplitude measured in terms of the height of the resulting trace. Approximate measurements can be made by calibrating the sensitivity of the cathode-ray in volts per inch. This varies with the anode voltage and type of tube; typical figures for small tubes are 25 to 75 volts
per inch, peak-to-peak. The initial calibration can be made with a variable d.c. voltage and comparison voltmeter.

where R is the resistance under measurement, E is the voltage read on the meter,
eis the series voltage applied, and ll,,, is the internal resistance of the meter
(full-scale reading in volts x ohms-per-
volt). The ohmmeter is a practical application of this method, with alow-current d.c. voltmeter and asource of voltage (usually dry cells), connected in series with the unknown resistance. If the meter reads full-scale with the connecting leads shorted, insertion of the resistance under measurement will cause the reading to decrease in proportion to the amount of resistance inserted. The scale can therefore be calibrated in ohms. In Fig. 1219-A, the series resistance is adjusted until the milliammeter reads full-scale
when the test leads are shorted. When the meter reading changes as the battery ages, this resistance is reduced, compensating for the change. In (B) the series resistance is kept constant but the sensitivity of the meter is varied to compensate for the changing voltage. The circuit of (C) is useful for measuring resistances below afew hundred ohms. The unknown resistance is connected as a shunt across the meter, reducing the current reading. Values of
afraction of an ohm can be read in this way.

(A)

(B)

� 12-4 IL 2,C, L AND re 111EASUUE1HENTS
It is frequently necessary to measure the components used in the construction of ama-

Fig. 1219 with series

--comOphemnsmaettieorn.cir(cBu)itsS.er(iAe)s

Series ohmmeter ohmmeter with

shunt compensation. (C) Shunt ohmmeter for measuring

low resistances.

238 CHAPTER TWELVE

7neuiuremenb ana! Meuiurinq equipment

The ratio of resistances which can be measured on a single ohmmeter range averages about 100 to 1, or from one-tenth to ten times the center-scale value.
Only approximate measurements can be
made with an ohmmeter. For greater accuracy the unknown resistor may be compared with a standard resistance of known accuracy by means of a Wheatstone bridge (� 2-11). If resistance measurements only are to be made, the bridge may be powered by abattery and a milliammeter used for the balance indicator. If reactances are also to be measured, an a.c. source is required (Fig. 1221).
Capacity and inductance -- The capacity of condensers and the inductance of coils can be measured (a) in terms of their reactances, (b) by comparison with astandard, and (c) by substitution methods.
The reactance method is simplest but least accurate. The method is similar to the d.c. ohmmeter, except that impedance is measured instead of resistance. In Fig. 1220, at (A) the unknown reactance is placed in series with an a.c. rectifier-type voltmeter across the 115-volt a.c. line. With a 1000-ohms-per-volt meter, capacities can be identified from approximately 0.001-dd. to 0.1-12fd. At (B) the reactance is confiected in series with a1000-ohm resistance; the proportionate voltage drop across this
resistance indicates the reactance of condensers from 0.1 dd. to 10 dd. and inductances from 0.5 henry to 50 henries, when Q is greater than 10. Because the lower end of the scale of a rectifier-type meter is somewhat crowded, abetter reading can be had by using
the connection at (C) for large reactances. Approximate calibrations for each connection may be made by checking typical condensers and coils of known values and drawing calibration curves for the voltmeter in use.
The reactance method at best gives only approximate indications of inductance and capacity. For accurate measurements, an a.c. bridge must be used.
A simple bridge for the measurement of R, C and L is shown in Fig. 1221. Its accuracy will depend on the precision of the standards, the sensitivity of the detector or balance indicator, the voltage and frequency of the a.c. source, and the ratio of the unknown value to the standard. The signal source can be a 1000cycle audio oscillator with low harmonic content and the detector a pair of headphones or an electron-ray tube.
For maximum accuracy the ratio of the un-
known to the standard should be kept small, so that R is read near the center of its scale. The ratio can be as high as 10 to 1in either direction with good accuracy, and an indication can be had even at 100 to 1. Additional standards can be included for other ranges if desired.

(B)

(C)

Fig. 1220 -- Reactance-measurement circuits for checking capacity and inductance values.

The potentiometer R must be calibrated as accurately as possible in terms of the ratio of resistance on either side of its mid-point, which may be arbitrarily marked 10. If the potentiometer is next set at 500 ohms, the ratio of resistances is 1 to 10 and the scale may be
marked 1. The corresponding point on the other end of the scale is marled 100. Intermediate points are similarly marked according to the resistance ratios. These ratios will then correspond with the ratio of the unknown resistance, inductance or capacity to the standard in use, when the bridge has been balanced for anull indication on the detector.
Since direct current flowing through a coil changes its inductance, allowance must be made for this in measuring choke coils and transformers carrying d.c.
Condensers should be checked for leakage as well as capacity. This check must be made
with the rated d.c. voltage applied, amicroammeter being connected in series with the high
voltage source. The resistance of good paper condensers should be above 50 megohms per microfarad, that of mica above 100.
The condition of electrolytic condensers can be checked roughly with an ohmmeter. With the positive terminal of the condenser connected to the positive of the ohmmeter bat-

ToA.C.Source

RI R2
Fig. 1221 -- Simple a.c. bridge for measuring resistance, inductance and capacity. Ci -- 0.01-pfd. mica. Cs -- 1.0-pfd. paper. 11 -- 10,000-ohm linear wire-wound potentiometer. RI-- 100 ohms, wire-wound (1% accuracy). R2 -- 10,000 ohms, wire-wound (1%). Li -- 125-millihenry iron-core r.f. choke.
-- 12-henry iron-core choke (Thordarson T-49C91).

CHAPTER TWELVE 239

DheRadio AntaieuA

To A.C.Source
�.7 Fig. 1222 -- Substitution-type capacity bridge.
Ci -- 1000-ggfd. straight-line-capacity condenser (may
be two-ghng 500-egfd. with sections in parallel). C2 -- 900-gpfd. silver mica. Ca -- 100-��fd. variable trimmer.
RI, R2 -- 5000-ohm wire-wound (1%).
Rs -- 1000-ohm wire-wound potentiometer.
tery, high-voltage electrolytics should show a resistance of 0.5-megohm or so; low-voltage cathode by-pass condensers should be over 0.1 megohm. Electrolytics can also be checked by measuring the leakage current when the rated d.c. polarizing voltage is applied. It should read about 0.1 ma. per dd. The maximum for a useful unit is about 0.5 ma. per afd. Low leakage current also indicates a faulty unit. Electrolytic condensers which have lain idle on the shelf will show leakage currents as high as 2ma. per pfd. per 100 volts. After "aging" for a few minutes at rated d.c. voltage they should return to normal, however.
The measurement of small capacities under 0.001 pfd. is not possible with abridge of the type previously described because stray reactances cause errors. A more accurate bridge for measurement of small capacities is shown in Fig. 1222. It is of the substitution type with a calibrated air condenser, C1,for the variable arm. Cg is a fixed reference capacity. C3 is used to balance out stray capacity including that of the leads to C.. The bridge is first balanced by adjusting C3,with C1at maximum capacity and the leads to c. in place. C. is then connected and the bridge again balanced by adjusting C1.The difference in capacity (AC) of C1 between its new setting and its maximum capacity represents the capacity of C..
It is impossible to get a zero null indication from the detector unless the resistances as well as the capacities of the two condensers being compared are equal. R3 is therefore included to aid in achieving aresistive balance. Generally speaking, R3 will be in the Cg leg when measuring a mica condenser and in the C1 leg for

an air condenser. The bridge is brought into balance by alternately varying the standard
capacity C1 and equalizing the power factors by R3until zero indication is obtained.
The bridge can be made direct-reading in �pfd. by using adial with 100 divisions and a 10-division vernier installed so that 0 on the
dial corresponds to maximum capacity on C1. Then as the capacity of C1is decreased to compensate for the addition of C., �C is numerically equal to the dial reading times 10. The true capacity of C1 will depart from linearity with the dial setting as it nears zero, but the
percentage error remains small up to at least 90 on the dial (C. < 900 ��fd). The overall accuracy can be made better than 1%.
L, C and Q measurements at r.f.-- The
low-frequency a.c. bridge method of measuring inductance is of value only with highinductance coils f9r use at power and audio frequencies. I.f. and r.f. coils must be measured
at frequencies corresponding to those at which
they are used. The method commonly employed is to de-
termine the frequency at which the coil reso-
nates when connected across a capacity of known value. This may be done (a) by connecting the coil-condenser combination in a negative-resistance oscillator (� 3-7) and observing the resulting oscillation frequency on a
calibrated receiver, or (b) by connecting the coil to a calibrated condenser, supplying the circuit with r.f. power from asuitable oscillator, and tuning the condenser until resonance is indicated by maximum indication on a vacuum-tube yoltmeter (Fig. 1223). With the
capacity known in i.infd. and the resonant frequency in kc., the apparent inductance of the
coil in microhenries can be computed

L -- (159, 160y 1

12 C

The apparent inductance thus computed is in error, however, in that it also includes the distributed capacity of the coil. This will be discovered if asimilar measurement is made at another frequency (for example, the harmonic of h), for it will be found that adifferent value of inductance results. However, by combining the two measurements the true inductance can be found

1 842212

1 -- C2

when 12 is the second harmonic of fi, C2 is the
capacity required to tune to fi and C2 to ./.2. A convenient source of r.f. power for the two-
frequency method of inductance measurement is the transmitter exciter unit, provided it has good second harmonic output. The oscillator output and link circuit (shown inside dashed lines in Fig. 1223) should either be shielded or

240 CHAPTER TWELVE

Meaiurements and Mecourinq equipment

sufficiently remote from the measuring circuit

so that the vacuum-tube voltmeter shows no

indication when there is no coil in the circuit.

The calibrated condenser must, of course, have

sufficient capacity to tune over a 2-to-1 fre-

quency range. It can be calibrated by abridge such as the substitution-type capacity bridge

of Fig. 1222.

The resonance method can also be used for

accurate measurement of capacity. A standard

coil of suitable inductance must be provided;

the exact value is not important. The standard

condenser C1 is first tuned to resonance with the oscillator frequency. The unknown capac-

ity, C., is then added in parallel and the capacity of CI reduced until the circuit again resonates at the oscillator frequency. The difference between the two settings (aC) represents the capacity of C..
The arrangement of Fig. 1223 is additionally useful in that it can be used as a Q meter, measuring r.f. resistance and impedance (� 2-10).
As is shown by Fig. 1224, resistance in a tuned circuit broadens the resonance curve.

Fig. 1223-- (A) Circuit used in measuring inductance, capacity and Q at r.f. The calibrated variable. frequency oscillator should have atuning range in excess of 2-to-1. (B) Circuit for calibrating the v.t.v.m. for Q
measurements from 60-cycle a.c. REIC is 70.7% of RAC.
With the switch in position A, Ris is adjusted to give a voltmeter deflection near the upper part of its scale; this is the peak-deflection reference point. The switch is then turned to position B, and the new reading noted. By making a number of measurements with different initial input levels, agraph can be plotted showing peak and 70.7% readings for awide range of inputs.

By measuring the frequency difference between the two points at which the output voltage equals 70.7% of the peak voltage (where the resistance in the circuit equals its reactance), the Q of coils and condensers can be computed.
There are two methods of determining these points. One involves the use of a calibrated
variable frequency oscillator to determine the bandwidth in terms of frequency change and

When the frequency of the oscillator is fixed and acalibrated variable condenser is used, the capacity at resonance (Cr)is noted, as well as that on either side at which the voltmeter reads 70.7% of maximum. Then

2 Cr

the other a calibrated variable condenser to measure the capacity change.
When the calibrated variable oscillator and v.t.v.m. are used, the frequency and r.f. voltage at resonance are first noted. The oscillator frequency is then varied on either side of resonance until the v.t.v.m. reads 70.7% of its initial value: Then Q is equal to the frequency divided by the bandwidth, or
Q
af where Af is the difference between the frequencies fi and /2.

The foregoing applies when measuring the Q
of coils. Actually, the figure of Q thus derived is not that for the coil alone but for the tuned circuit as a whole, including the condenser. The Qof the standard condenser must therefore be kept high. An efficient air condenser with steatite or mycalyx insulation is required; it should be operated near maximum capacity. Short, heavy leads must be used and the stray capacities kept as low as possible.
The Q of other air condensers and of mica condensers can be determined by first measuring the Q of the circuit with astandard coil in place, then connecting C. in parallel with C and

Fig. 1224--Resonant curves of (A) a high-Q tuned circuit and (B) a low-Q circuit. Q is determined directly from the ratio of the frequency at owfhimcahxriemsupmon.se is maximum to that at which it is 70%

FREftQwUfErtN.CY A

21-

f).

fIL

FR EQ UENCY

CHAPTER TWELVE 24 1

Dhe Pack� Amaleur '.3 ,ilaticlhooh

again measuring the Q. The Q of the unknown condenser is

(CI-- C2) Qi Q2 C1 (Qi -- Q2)

Low-Q mica and paper condensers (Q<

1000) can be measured by inserting the un-

known in series with L and C. Q1 is measured with a shorting bar across the unknown; the

bar is then removed and Q2 determined. Then

- (c2 -- Qi Q2

�

CiQi - C2 Q 2

If C2 is larger than C1,the reactance is inductive rather than capacitive; i.e., the "con-

denser" is actually an inductance at the meas-

urement frequency.

The r.f. resistance, reactance and impedance

of other components can be measured by the

same methods. If an external r.f. impedance

(such as an antenna or transmission line or an r.f. choke) is inserted in a coil-condenser cir-

cuit, it will both detune the circuit and broaden

its resonance curve. By observing the capacity required to bring the circuit back to resonance

and measuring the additional resistance introduced by re-measuring its Q, the reactive and

resistive components of the external impedance

can be computed.

5% THIRD HARMONIC

s% sumo moo�

le 'Hun) HARMONIC

10% MONO HARMONIC

7% be PLUS 5% 7.0

Fig. 1225-- Waveforms as viewed on the cathode-ray oosfcvialrloisocuospet,ypsehsoowfinhgartmhoenidcisdtiisntgouritisohni.ngThchearwaacvteerisshtaipces is symmetrical with odd-harmonic distortion, asymmetrical with even.

Using astandard coil and condenser suitable for the operating frequency, connect the unknown quantity across C1 (for high resistances) or in series with L and C (for low resistances), and proceed as previously outlined. If C1 must be increased to restore resonance, the reactance of the unknown is inductive; if it must be decreased, the reactance is capacitive.
�12-4 WAVEFORM
Wave Analyzers -- In working with alternating currents it is often necessary to know not only the peak or average values of voltage and current, but also something of the waveform (� 2-7). By the use of awave analyzer it is possible to measure the amplitude and frequency of the fundamental and each of the important harmonic components of acomplex electrical waveform.
The most widely-used instrument for waveform analysis is the cathode-ray oscilloscope equipped with alinear sweep circuit. As shown in �3-9, the pattern traced on the screen represents the actual electrical shape of the wave under study. By study of this shape the character of the wave can be determined. Fig. 1225 shows typical patterns shapes resulting from different kinds of harmonic content.
The oscilloscope does not readily enable analysis of separate small harmonic components and in laboratory work special wave analyzers are used for this purpose. In the heterodyne type of wave analyzer the voltage under measurement is heterodyned with the signal from acalibrated variable oscillator and the resultant applied to asquare-law detector or balanced modulator which delivers its output to ahighly-selective amplifier using piezoelectric crystal filters or sharply-tuned circuits. The frequency of the component under.measurement is indicated by the variable oscillator setting, while the amplitude is read on a vacuum-tube voltmeter. Since it first magnifies the relative separation between the harmonic components by heterodyning, and then uses a highly-selective amplifier with a pass-band of but afew cycles, the individual components of acomplex voltage can be measured even when the fundamental lies in the lowest audio range.
Simpler wave analyzers of the resonant type are satisfactory on higher frequencies. R.f. harmonics may be measured qualitatively with an ordinary tuned circuit equipped with asensitive resonance indicator. An absorption frequency meter or field-intensity meter may be used for the purpose (� 12-2, �12-8). By tuning the instrument to the fundamental and each harmonic in turn, keeping the pick-up constant, a useful indication of relative amplitudes can be obtained.
Modulation Percentage -- Determination of the percentage of modulation in amodulated

242 CHAPTER TWELVE

Mea�urement� and Meadurinq equipment

wave (� 5-2) requires analysis of the wave
envelope. Use of the oscilloscope for this purpose is described in �5-10. Another form of modulation percentage indicator is shown in Fig. 1230-B. Essentially, it constitutes apeak diode voltmeter which reads the r.f. carrier steady-state level when the plug is in the lefthand jack, while in the right-hand position the meter reads the average rectified component of the audio modulation. The relative amplitudes of the reading indicate the modulation percentage.
e12-6 RECEIVER CHARACTERISTICS
Measurements in connection with receiving

Testing and Alignment--The measurement of receiver performance requires precision laboratory equipment beyond the scope of the average amateur. Sufficient apparatus for servicing receivers should be available in every amateur station, however. This may be as little as amulti-range volt-ohm-milliammeter, atestsignal source of some description, and a few spare tubes.
For the alignment of tuned circuits asimple test oscillator is required, preferably one that can be modulated by a400-cycle audio oscillator. A rectifier-type voltmeter can be used for the output meter.

equipment come under two heads: (1) overall

performance, and (2) servicing and alignment. Performance Measurement -- Laboratory

measurements of overall performance require

(1) astandard-signal generator, which is atest

oscillator covering the frequency range in use,

capable of being modulated 30% at 400 cycles

and equipped with an output attenuator ac-

curately calibrated in microvolts; (2) astand-

ard dummy antenna approximating an ordinary receiving antenna, consisting of a 20 ph.

inductance, shunted by 400 ppfd. and 400

ohms, in series with a200 ppfd. condenser, used

to couple the signal generator to the receiver

antenna input; (3) an output-power measuring

device consisting of an a.f. voltmeter connected across aresistance representing the output load of the receiver; and (4) abeat-frequency audio oscillator continuously variable from 30 to 15,000 cycles. A wave analyzer is also desirable for measuring audio distortion.
With this equipment the receiver characteristics listed in �7-2 can be measured and
charted. Sensitivity is stated in terms of the input in microvolts of a30 %-modulated signal required to give an audio output of 0.5 watt (0.05 watt for battery receivers, etc.) at various frequencies throughout the range. Selectivity curves (Fig. 701) are taken by varying the signal-generator frequency either side of resonance in increments of 1to 10 }Lc., increas-

Fig. 1226 -- I.f. test signal generator circuit.

CI-- 100-ppfd. variable with 200-iimfd. fixed in parallel.

-- 100-pmfd. midget mica.

C3, C4 -- 250-Add. midget mica.

C6 -- 0.005-pfd. mica.

C6 0.1-pfd., 400-volt paper.

C7 500-Add. midget mica.

RI-- 50,000 ohms, Y2-watt.

R2 -- 2000 ohms, -watt.

R3 -- 20,000 ohms, 1-watt.

R4 -- 20,000 ohms, 2-watt.

Rs -- 500-ohm carbon potentiometer.
L-- 440-510 kc.: 140 turns No. 30 enamel close-wound

on 1%-inch plug-in form. Cathode tap 35 turns

from ground end.

1400-1550 kc.: 42 t. No. 20 d.s.c., tap at 10 t.

4500-5500 kc.: 11 t. No. 18 spaced, tap at 3t.

RFC1 --2.5-mh. r.f. choke.

RFC.2

r.f. choke.

ing the signal input each time to keep the output constant. Fidelity curves are made by modulating the signal generator at 30 % with a
continuously-variable beat-frequency oscillator and recording the audio output of the receiver relative to the normal response at 400
cycles. The signal-to-noise ratio is determined by measuring the noise output at maximum gain with no signal, and then applying asignal sufficient to double the output. Frequency stability is measured by methods described in �12-2, zero-beating the receiver oscillator against a stable standard and observing the drift (a) during the warm-up period of the receiver, (b) with variation in power supply voltage, and (c) with various input signal levels.

The frequency meter is a suitable signal source for r.f. alignment provided the harmonic amplitude on the higher-frequency bands is great enough. A harmonic amplifier and outp�t attenuator are useful in this application.
The i.f. test oscillator circuit shown in Fig. 1226 consists of a simple e.c.o. with plug-in coils. The output level is controlled by a potentiometer so connected as to present aconstant input resistance to the receiver. The oscillator should be shielded so that direct pick-up is minimized. It is helpful to make all ground returns to aheavy copper strap which is connected to the cabinet at only one point --
the output ground terminal. The plug-in-coil should be separately shielded.

CHAPTER TWELVE 243

Dire leach� Arna leur ivianiho�

60, 6A 0 7B3

Fig. 1227 -- Negative-resistance audio-frequency oscillator.

h o�

C I-- 0.15-mfd., 400-volt paper.

C'veeue 1-- R3
+250 V.

2 ' C80--.25tifd-efdierPttqlseirp.er. RI, 112 -- 50,0(10 ohms, 1-watt. Rs -- 50,000-obus volume control. 14 -- 1.2-henry choke (Thordarson T-14C61 with iron
T-- Outcporuet rtermaonvsefdo)r.mer (interstage audio, 1:3 ratio).

The test oscillator may be suppressor-grid modulated by applying approximately 10 volts
of audio (for 50% modulation), as shown in the diagram. The suppressor-grid is biased 10-volts negative for modulated use; if an
unmodulated signal is desired, the upper terminal can be grounded as indicated. This will increase the output from the oscillator. Conversely, if the output potentiometer does not attenuate the signal sufficiently, additional d.c, negative bias can be applied between the modulation terminals.
Ordinarily there is no requirement for precise calibration of the test oscillator. In i.f. alignment most communications receivers are equipped with acrystal filter and the oscillator
frequency is set to correspond with the crystal response. If the receiver contains no crystal
filter, the oscillator should be set at the design i.f. as closely as its calibration will permit.
With an unmodulated test signal the output indicator can be the " "-meter in the receiver, a microammeter in the detector or
a.v.c. circuit, or avacuum-tube voltmeter. It

is not advisable to use the receiver beat oscilla-
tor to generate an audible note for output indications. When a modulated test signal is used, the output indicator can be a copper-
oxide rectifier-type voltmeter which reads the a.f. voltage across the rated output load resistance. Power output can be computed as
previously described. The a.f. modulating source for the test oscil-
lator can be any audio oscillator capable of delivering 10 to 20 volts at the standard receiver checking frequency of 400 cycles.
Audio Oscillators -- A simple oscillator of the negative-resistance or two-terminal variety using apentagrid tube is shown in Fig. 1227.
A frequency of approximately 400 cycles is generated with the tuned circuit shown.
Fig. 1228 shows a beat-frequency type of audio oscillator with a continuously-variable range from 0to 15,000 cycles, of the kind used in making fidelity measurements and for other purposes requiring a wide-range variable a.f.
source. The electron-coupled oscillators are initially tuned to an identical frequency in the

R. 6J7

6N7

Cri

6J5

7 75/-1ie-1.1 -
111, RS, R9 -- 50,000 ohms, Y2-watt. R2, RS-- 20,000 ohnts, 1-watt. Rs -- 2000 ohms, z'-w att. Rs, R2, R8 -- 40,000 ohms, ,1/2-watt. Rio, Ru --O.! megohm, 1112, 1113 -- 25,000-ohm potentiometer. RI4 -- 500 ohms, %-watt. RFC -- 125-mh. iron-core r.f. choke.
244 CHAPTER TWELVE

--A0Q_O RFC

+250

Fig. 1228 -- Beat-frequency audio oscillator.

Ci -- 500-nisfd. variable. C2, C7-- 100-gpfd. variable trimmer.

Ca, Cs -- 0.002-nnfd. mica padding condenser.

C4, Cs -- 500-unfd. mica.

Cs, Cis --

600-volt paper.

Cs, C11 -- 0.01-mfd., 600-volt paper.

C12 -- 50-nfd. 50-volt electrolytic.

C13, C14, CIS, C18 -- 250-gpfd. mica.

C17, C18 -- 1.0-pfd., 600-volt paper.

T -- High-fidelity push-pull interstage or tube-to-line transformer, according to application.
14-La -- 1-mh. shielded coil with cathode tap (may be 456-kc. receiver heat-oscillator coil, with selfcontained trimmer used as C2).
L2 -- 100 turns No. 32 wire close-wound over Li. -- 20 turns uound over 14 (separated by electrostatic shield).

IlleaMtremen13 and Mea�leriny

Fig. 1229-- C.w. and 'phone monitor.

Ci -- 50-juifd. midget variable.

Ct -- 0.002-dd. midget mica.

C3 -- 100-mild. midget mica.

Ri -- 1megohm, 34-watt.

Si-S2 -- 2-section 4-position rotary band switch.

S3 -- S.p.d.t. low-capacity switch.

S4 -- Toggle switch.

Band

1750 3500 7000 14000

90 t. No. 30 e. 50 t. No. 30 d.c.c. 30 t. No. 22 d.c.c. 10 t. No. 22 d.c.c.

24 t. 16 t. 10 t. 6 t.

All coils close wound on 1-inch diameter forms, grouped around Si-St with adjacent coils at right angles. L and L' approximately h-in, apart.

L2' = L.g

L.;

IG4G, IH4G, 30. etc
1

vicinity of 100 kc., the variable tuning con- be constructed solidly so that it can be moved

denser C1 being set at zero. The oscillator out- around the station without necessity for re-

puts are coupled to the balanced modulator tuning when listening to asignal.

which passes the beat frequency along to the

The circuit of a simple monitor with band-

push-pull audio amplifier. In construction the switching covering four amateur bands is

two oscillator circuits should be made as nearly shown in Fig. 1229. Any 1.5- or 2-volt filament

identical as possible, so that any tendency triode can be used, as well as any batteries of

toward drift will be equal in both. The tuned asize that will fit into the container selected.

circuits should be isolated from tube heat and A plate-tickler switch (S3)is provided to make

made mechanically stable. An electrostatic the monitor non-oscillating when checking

shield (� 4-9) on the fixed-oscillator output aids 'phone signals. If desired, aregeneration con-

in reducing pulling at very low frequencies. trol could be incorporated (� 7-4).

A voltage-regulated power supply is desirable.

Any type of simple detector with a means

� 12-7 TRANSMITTER CHAILACTERISTICS

for picking up asmall amount of r.f. from the transmitter can be used as a 'phone monitor. A satisfactory monitor can be constructed

The transmitter characteristics requiring using a diode rectifier and untuned pick-up

measurement are carrier frequency, d.c., a.c. coil, as shown in Fig. 1230-A. Headphones are

and r.f, voltage, current and power, keying used for listening checks. The monitor can

and modulation quality, modulation percent- also be employed as an over-modulation indi-

age distortion, carrier noise and spurious radia- cator by use of the 0-1 milliammeter. The

tions. Instruments for the measurement of most of these quantities have been discussed under the appropriate headings.

pick-up coil is loosely coupled to the tank circuit of the final r.f. amplifier until the milliammeter reads approximately 0.9 ma. The speech

Keying and modulation checks may be amplifier is supplied with a400-cycle sine-wave

made by several methods; the two commonly tone from an audio oscillator such as that

used by amateurs are aural checks with moni- shown in Fig. 1227, and the gain control turned

tors, and visual checks with the oscilloscope. up until the monitor meter starts to rise, indi-

For data on the use of the oscilloscope in check- cating overmodulation.

ing modulation using "wave envelope" and

The circuit at 1230-B indicates the percent-

"trapezoid" methods, see �5-10.

age of modulation directly. The a.c. milliam-

Monitors -- A monitor is a miniature re- meter is first plugged into the left-hand jack

ceiver, usually having only a single tube, en- and the pick-up coupling adjusted to give closed with its batteries in some sort of metal afull-scale meter reading on the unmodulated

box which serves as ashield. The requirements carrier. Then the meter is plugged into the

for a satisfactory monitor for checking c.w. right-hand jack and the transmitter modulated

signals are not difficult to satisfy. It should by atone or speech signal. The modulation per-

oscillate steadily over the bands on which centage will be 140 times the reading of the the station is to be active; the tuning should meter; e.g.,, for 100% modulation the meter

not be excessively critical, although the degree will read approximately 0.7 ma. In measuring

of band-spread ordinarily considered desirable the percentage of modulation with speech the

for receivers is not essential; the shielding inertia of meter will cause it to undershoot

should be complete enough to permit the on peaks; the maximum swing should there-

monitor to be placed near the transmitter and fore be limited to something less than 0.7 ma.

still give agood beat note when tuned to the

Power -- The power input to aradio trans-

fundamental frequency of the transmitter mitter is rated as the total d.c. plate input

(this is often impossible with the receiver (E xI) to the final stage. The power output is because the pick-up is too great); and it should rated as the actual a.c. power measured at the

CHAPTER TWELVE 245

2:7ne leach,' _Amateur's -ilancgooL

(A)
(3)

6H6

RFC

Fig. 1230 -- (A) Simple diode 'phone monitor and overmodulation indicator.
Ci -- 0.005-pfd. midget mica. C2 -- 0.01-pfd. paper. Ri -- 0.15 megohm, 3-watt. Li -- Pick-up coil (enough turns of hook-up wire to
give 1-ma, deflection on meter when loose!) coupled to final tank, connected to unit through twisted-pair line). M --0-1 ma. d.c. milliammeter.
(B) Modulation percentage indicator.
CI, RIand Li same as above. Cs -- 0.005-mfd. midget mica. C4 -- 1.0-gfd. paper. R2 -- 0.25 megohm, 3/2-watt. (Should equal Ri plus
impedance of L2 at modulation test frequency.) L2 -- 30-50 henry iron-core choke. M -- 0-1 ma. a.c. milliammeter (d.c. meter with copper-oxide rectifier).

teristics, chiefly involving impedance and

resistance. Elaborate field-strength measuring sets

calibrated in microvolts-per-meter or similar

output terminals of the transmitter unit proper
when connected to its normal load circuit or equivalent. A c.w. transmitter is rated in terms of its power capability under keying conditions, defined as intermittent operation with the key
down 50% of the time. An amplitude-modulated transmitter is rated in terms of the average power under modulation when being modulated to its specified capability (the modulation capability and distortion at the test frequency being stated).
Instruments and methods of measuring both d.c. and r.f. power have been described in �12-3. Additional methods used to measure transmitter power are the calorimeter and anode-dissipation methods. In the calorimeter

units have been devised for use in commercial work, incorporating standard-signal generators and standard-gain amplifiers for measuring field intensity by the comparison method. Ordinary antenna checks and adjustments can be made by asimple field-intensity meter, however.
The instruments described for r.f. measurement (thermocouple ammeter, vacuum-tube voltmeter, L, C and Q meter) are all applicable to antenna measurement.
Antenna resistance (� 10-1) can be determined by inserting anon-inductive resistor of known value and a thermoammeter in series in the antenna, preferably at a current loop. Then the antenna resistance in ohms is

method a non-inductive resistor carrying the

r.f. power is cooled by water or other liquid surrounding it. The power dissipated is then
calculated from the temperature rise of the water. The anode-dissipation method is used in connection with large water-cooled tubes. The total power delivered to the filament grid and plate circuits is first measured. The power
dissipated in the cooling fluid is then measured in terms of its temperature rise and the differ-

I -- Ik
where Ik = antenna current with the resistance Rk in the circuit and I = the current with the resistance short-circuited.
With the resistance and current known, the antenna power can be calculated:
P. = I2R.
as The effective antenna reactance (as well

ence between the total input power and that the resistance) can be measured by the substi-

dissipated in the water is considered to be the tution method outlined under �12-4. The

useful power output delivered to the load.

measurement must be made at the operating

� 1248 ANTENNA MEASIMEMENTS

frequency or band of frequencies. Antenna impedance can also be measured by

Antenna measurements are made for the feeding power to the antenna from asource of

purposes (a) of securing maximum transfer of known impedance through auniform low-loss

power to the antenna from the transmitter, quarter-wave line (� 10-5). The current or

and (b) of adjusting directional antennas to voltage is measured at both ends, whereupon

conform with design conditions. Measurements are therefore made of the current (power) in the antenna, voltage and current relationships, resistance, and radiated field intensity. Related to measurement of the antenna proper is
the measurement of transmission line chame-

z _ Zi

Zi E,, Ei

where the subscripts iand oindicate the input (oscillator) and output (antenna) ends of the
line.

246 CHAPTER TWELVE

Meaiurement� ani Meaiuring equipmenl

Field-intensity meters--In adjusting an-

tenna systems for maximum radiation and in

exploring radiation patterns use is made of a

field intensity meter. Basically afield-intensity

meter is a vacuum-tube voltmeter provided

with atuned input circuit. It is used to indicate

the relative intensity of the radiation field

under actual radiating conditions. It is par-

ticularly useful on the ultra-high frequencies and in adjusting directional antennas. Checks

-I +

with afield-intensity meter should be made at points not less than several wavelengths distant from the, antenna and at heights corresponding with the desired angle of radiation. Standard comparative tests are made at a distance of one mile with one watt input to the antenna.

Fig. 1231 -- Sensitive field-intensity meter. Ci -- 50-aafd. midget variable. C2 -- 250-pmfd. midget mica. C3 -- 0.002-pfd. midget mica. Ri -- 1megohm, 2-watt. L-- Coil to tune to frequency in use in conjunction
with CI. Diode tap in center of coil. M -- 0-1.5 milliamperes.

The absorption frequency meter shown in

Figs. 1202-03 may be used as field strength amplifier tube. Because of the logarithmic

meter if it is provided with apick-up antenna. grid voltage-plate current characteristic of

This can be ashort length of brass rod or an this tube, a 0-1 ma. meter in its plate circuit

automotive-type antenna mounted on astand- can be calibrated arbitrarily with an approxi-

off insulator and connected to the stator of the mate linear db. scale as shown.

tuning condenser through a small trimmer.

The scale covers approximately 25 db. and is

The crystal detector is not linear, so that a linear over a range of about 20 db. At very

given increase in current does not indicate a small signals it departs from linearity, and 0

directly proportional increase in field strength. db. is therefore placed at 90% of the scale. A

A more sensitive field-intensity meter of use variable meter shunt compensates for varia-

in examining the field-strength patterns of tions in tubes and battery voltages. In use

lower-frequency antenna systems is shown in this resistor is initially adjusted to give afull-

Fig. 1231. It consists of a diode rectifier and scale reading of 1 ma. The signal pick-up is

d.c. amplifier in the same envelope. The initial then made such as to cause the meter to indi-

plate current reading is in the neighborhood cate 0 db. Alternatively, the initial reading

of 1.4 ma.; with signal input, the current dips can be set arbitrarily at 10 db.; adjustments

downward. The scale reading is linear with will then be indicated as losses or gains in re-

signal voltage, a characteristic that is advan- lation to that figure.

tageous in making comparative measurements. Radiated power variations will be as

� I2-9 TUBE CHARACTERISTICS

the square of the field-voltage indication.

Accurate measurement of the operating

Power gain in antenna systems is usually ex- characteristics of vacuum tubes (� 3-2) under

pressed in terms of decibels. A field intensity dynamic conditions requires the use of a

meter that reads directly in db. is shown in vacuum-tube bridge, in which the tube co-

Fig. 1232. It consists of a self-biased linear efficients constitute the "unknown" arm.

triode voltmeter followed by avariable-p d.c. Such a bridge is a complex and costly affair.

III ,,'
- 225V+

Islillil
- 45V. +

Fig. 1232 -- Logarithmic field-intensity meter with db. calibration using miniature dry-cell type tubes.

CI-- 3-30-aufd. mica trimmer. Ca -- 50-add. midget variable, Cs, Ca --500-apfd. midget mica. Ft, -- 10 megohms, %-watt.

RR-- 1000-ohm wire-wound vanable.
L -- Coil to tune to frequency in use.

SI, SR-- Toggle switches or d.p.s.t. switch.
M -- 0-1 ma. d.c. milliammeter.

247 CHAPTER TWELVE

5neleadio Amateur'�

'Fortunately, however, the need for measuring vacuum-tube characteristics as such is rarely encountered in ordinary radio work, since complete data concerning standard types is published by the manufacturers.
The principal need for checking a tube lies in determining when, through aging or other defect, it departs from the original characteristics. The best method of checking areceiving or transmitting tube is by direct comparison in
its own socket with anew tube of known quality under actual operating conditions. Any other test falls short of an actual performance
test. A number of commercial tube checkers of
the type used by servicemen are on the market. In purchasing one the following qualifications should be sought: (1) complete facilities for checking shorts between any pair of electrodes; (2) atransconductance rather than an "emis-
sion" test (the emission of a tube may vary widely with no effect itin its performance, while genuinely faulty tubes may show rated emission); (3) provision for checking plate and
screen currents under typical conditions (at rated voltages); (4) gas and noise tests.
Commercial tube-checkers are elaborate assemblies bearing all standard socket types and switching arrangements to supply correct element potentials for all types of tubes. In
the absence of such a unit, for an occasional need the amateur can assemble acircuit using an existing power source in accordance with Fig. 1233 to make a reasonably accurate standard transconductance test. A pentode tube is shown; for other types omit or add electrode connections as required. The voltages
applied should correspond with those listed in the published tube tables within 5% (ex-
pecially grid voltage, plate voltage for triodes and screen voltage for pentodes). With the

Fig. 1233 -- Circuit for measuring vacuum-tube transconductance.
grid switch in No. 2 position, the plate and screen currents should read near the rated values; wide variations from normal indicate adefective tube.
To make the transconductance test, note the plate current with the grid switch alternately on positions 3 and 1, changing the bias from exactly 0.5 volt less than rated bias to exactly 0.5 volt more. The resulting plate current change multiplied by 1000 equals the transconductance in micromhos. This value can be checked against the tables. Tubes will usually operate satisfactorily until the transconductance falls to 70% of the rated value.
Pentagrid and heptode frequency converters may be checked by this method if the rated d.c. electrode voltages are applied. The oscillator section can be checked separately by noting the oscillator-anode current change.
Diodes can be checked by applying 50 volts of 60-cycle a.c. between plate and cathode, in series with a 0.25-megohm load shunted by a
2-pfd. condenser, and reading the rectified current on a 0-1 ma. d.c. meter. A reading of 0.2 and 0.25 ma. indicates a satisfactory
tube.

24 8 CHAPTER TWELVE

CHAPTER THIRTEEN
Worhihop Practice

� TOOLS

W HILE THE GREATER the variety of tools available, the easier and, perhaps, the better the job may be done, with a little thought and care it is possible to turn out a fine piece of equipment with comparatively few common hand tools. A list of tools which will be found indispensable in the construction
of equipment will be found on this page. With these tools it should be possible to perform any of the required operations in preparing panels and metal chassis for assembly and wiring. A few additional tools will make certain operations easier.
The following list will be found helpful in making aselection:

Bench vise, 4-in. jaws

Tin shears, 10-in., for cutting thin sheet metal

Taper reamer, holes

for enlarging small

Taper reamer, 1-in., for enlarging holes Countersink for brace

INDISPENSABLE TOOLS

Long-nose pliers, 6-in.

Diagonal cutting pliers, 6-in. Screwdriver, 8- to 7-in., h-in. blade Screwdriver, 4- to 5-in., h-in. blade

Scratch awl or ice pick for marking lines Combination square, 12-in, for laying out work

Hand drill, h-in, chuck or larger, 2-speed type preferable
Electric soldering iron, 100 watts Hacksaw, 12-in. blades Center punch for marking hole centers Hammer, ball peen, 1-lb. head Heavy knife

Yardstick or other straight edge

Carpenter's brace with adjustable hole cutter or

socket-hole punches (see text)

l'air small "C" clamps for holding work

Large coarse flat file
Large round or rat-tail file, 34-in, diameter

Three or four small and medium files, flat, round,

half-round, triangular

Drills, particularly

and Nos. 18, 28, 33, 42

and 50

Combination oil stone for sharpening tools

Solder and soldering paste (non-corroding

Medium-weight machine oil

Carpenter's plane, 8- to 12-in., for woodworking

Carpenter's saw, cross-cut Motor-driven emery wheel for grinding

The soldering iron may be kept in good condition by keeping the tip well tinned with

Long-shank screwdriver with screw-holding solder and not allowing it to run at full voltage

clip for tight places

for long periods when it is not being used. After

Set of socket wrenches for hex nuts Wood chisel, 34-in. Cold chisel, 34-in. Wing dividers, 8-in, for scribing circles

each period of use, the tip should be removed and cleaned of any scale, which may have accumulated. An oxidized tip may be cleaned by dipping in sal ammoniac while hot and wiping

Set machine-screw taps and dies Folding rule, 6-ft. Dusting brush

clean with arag. Should the tip become pitted, it should be filed until smooth and then tinned by dipping it in solder. _

Several of the pieces of light woodworking machinery, often sold in hardware stores and

All tools should be wiped occasionally with an oily cloth to prevent rust.

mail-order retail stores, are ideal for amateur radio work, especially the drill press, grinding head, band and circular saws and joiner.

� USEFUL MATERIALS
Small stocks of various miscellaneous materials will be required from time to time. Most

� CARE OF TOOLS
A few minutes with the oil stone or emery wheel now and then will maintain the fine

of them may be purchased from hardware or radio-supply stores. A representative list follows:

cutting edges of knives, drills, chisels, etc. Drills should be sharpened at frequent inter-
vals so that grinding is kept at aminimum each time. This makes it easier to maintain the rather critical surface angles for best cutting with least wear.

34-in, by 1/16-in, brass strip for brackets, etc. (half-hard for bending)
square brass rod or 34-in, by 34-in. by 1/16-in, angle brass for corner joints %-in.-diam, round brass rod for shaft extensions

249 CHAPTER THIRTEEN

5ne leach� AmaieuA ilanitooh

Machine screws: Round-head, flat-head with nuts to fit. Most useful sizes, 4-36, 6-32, and 8-32 in lengths from 4-in. to 13�in. (Nickeled iron will be found satisfactory except in strong r.f. fields where brass should be used.)
Plain washers and lock washers for screws Bakelite and hard rubber scraps Soldering lugs, panel bearings, rubber
grommets, lug terminal strips, cambric tubing
Machine screws, nuts, washers, soldering lugs, etc., are most reasonably purchased in

for leads under i.f. transformers, etc., as well
as holes for wiring leads. By means of the square, lines indicating ac-
curately the centers of shafts should be extended to the front of the chassis and marked on the panel at the chassis line by fastening the panel temporarily. The hole centers may now be punched in the chassis with the centerpunch. After drilling, the parts which require mounting underneath may be located and the
mounting holes drilled, making sure by trial that no interferences exist with parts mounted on top. Mounting holes along the front edge of the chassis should be transferred to the

quantities of agross.

panel by once again fastening the panel to

� CHASSIS CONSTRUCIION
With a few essential tools and proper procedure, it will be found that building radio gear on a metal chassis is no more of a chore than building with wood and a more satisfac-

the chassis and marking from the rear. Next mount on the chassis the condensers
and any other parts with shafts extending to the panel, and measure accurately the height of the center of each shaft above the chassis as illustrated in Fig. 1301. The horizontal dis-

tory job results.

placement of shafts having already been

Much trouble and energy can be saved by marked on the chassis line on the panel, the spending plenty of time on the planning end vertical displacement may now be measured

of the job. When all details have been worked from this line and the shaft centers marked on out, the actual construction will be greatly the back of the panel and the holes drilled.

simplified. Cover the top of the chassis with apiece of
wrapping paper, or preferably cross-section paper, folding the edges down over the sides of the chassis and fastening with adhesive tape. Next, assemble parts to be mounted on top of

Holes for any other panel equipment coming above the chassis line may now be marked and drilled and the remainder of the apparatus mounted.
� DRILLING AND CUTTING HOLES

the chassis and move them about until asatis-

In drilling holes in metal with the hand drill,

factory arrangement has been found, keeping it is important that the centers be well located

in mind any parts which are to be mounted with the center punch so that the drill point underneath so that interferences in mountings will not "walk" away from the center when

will be avoided. Place condensers and other starting the hole. The drill should be held at parts with shafts extending to the panel first right angles to the surface being drilled. Care

and arrange so that the controls will form the should be used to prevent too much pressure

desired pattern on the panel. Be sure to line up the shafts square with the chassis front. Locate any partition shields and panel brack-

with small drills which bend or break easily. When the drill starts to break through, special care should be used and it is often an advantage

ets next and then sockets with their shields, to shift a two-speed drill to low gear at this

if used, and other parts, marking accurately point. Holes near %-in, in diameter may be the mounting-hole centers of each on the paper. started with a smaller drill and reamed out

Watch out for condensers whose shafts do not with alarger drill.

line up with the mounting holes. Do not forget

The chuck of the usual type of hand drill is

to mark the centers of socket holes and holes limited to '4-in, drills. Although it is rather

tedious, the 14-in, hole may be filed out to

larger diameters with round files. Another pos-

sible method with limited tools is to drill a

series of small holes with the 'hand drill along

the inside of the diameter of the large hole,

placing the holes as close together as possible.

The center may then be knocked out with a

cold chisel and the edges smoothed up with a

file. Taper reamers which fit in the carpenter's

brace make the job much easier. A large rat-

tail file may be clamped in the brace. This

Fig. 1301 -- Method of measuring heights of shafts.
If the square is adjustable, the end of the scale should he set flush with the face of the head.

makes avery good reamer for holes up to the diameter of the file if the file is revolved counter-clockwise.

250 CHAPTER THIRTEEN

l/VorLhop Practice

Number

NUMBERED DRILL SIZES

Diameter (mils)

Will Clear Screw

Drilled for Tapping Iron, Steel or Braes*

1 2 3 4
5 6 7 8 9
10 11 12 13
14 15 16
17 18
19
20
21 22 23 24
25
26
27 28
20 30
31 32 33
34 $6 36 37
38 39
40 41 42
43 44 45
46 47
48
49 60
51
52 53
54

228.0 221.0 213.0 209.0
205.0 204.0 201.0 199.0 196.0
193.5 191.0 189.0 185.0
182.0 180.0 177.0
173.0 100.6
166.0
161.0
159.0 157.0 154.0 152.0
149.5
147.0
144.0 140.6
1$4.0 128.5
120.0 116.0 1111.0
111.0 110.0 106.5 104.0
101.5 099.5
098.0 096.0 0113.6
089.0 * 086.0 082.0
081.0 078.5
076.0
073.0 070.0
067.0
063.5 059.5
055.0

12-24 12-20 10-32 10-24
4-$0 4-40 3-48 2-56

14--24
- -
12-- 24 12-20 10-32 10-24 8--$2
8-32
4-30- 4-40 3-48 2--56

*Use one size larger drill for tapping bakelite and hard rubber.

For socket holes and other large round holes, an adjustable cutter designed for the purpose may be used in the brace. When the cutter is
well sharpened, it makes the job easy. Occasional application of machine oil in the cutting groove usually helps. The cutter should first be tried out on ablock of wood to make sure that it is set for the correct diameter. The best de-
vice of all for cutting socket holes is the sockethole punch. The preferred type works by pressure applied by turning ascrew with awrench.
Square or rectangular holes may be cut out
by using the series of small holes previously

described, but more easily by drilling a )/2-in. hole inside each corner, as illustrated in Fig. 1302, and using these holes for starting and turning the hacksaw. The socket-hole punches may also be of considerable assistance in cutting out large rectangular openings.
The burrs or rough edges, which usually remain after drilling or cutting holes, may be removed with a file, or sometimes more conveniently with asharp knife or chisel. It is a good idea to keep an old wood chisel sharpened up for this purpose.
� CUTTING THREADS
Brass rod may be threaded or the damaged threads of ascrew repaired by the use of dies. Holes of suitable size (see drill chart) may be threaded for screws by means of taps. Either are obtainable in any standard machine-screw
A
Fig. 1302 - Cutting rectangular holes in a chassis. If the corner holes are filed out as shown in the shaded portion of B, it will be possible to start the hacksaw blade along the cutting line. A shows a single-ended handle for a hacksaw blade.
size. A set usually consists of taps and dies for 4-36, 6-32, 8-32, 10-32 and 14-20 sizes with a suitable holder. The die may be started easily by filing ataper or bevel on the end of the rod. In tapping ahole, extreme care should be used to prevent breaking the tap. The tap should be kept at right angles to the surface of the material and rotation should be reversed arevolution or two whenever the tap starts to turn hard. With care, holes may be tapped rapidly by clamping the tap in the chuck of the hand drill and using slow speed. Machine oil applied to the tap usually makes cutting easier and sticking less troublesome.
� CUTTING AND BENDING SHEET METAL
If a sheet of metal is too large to be conveniently cut with a hacksaw, it may be marked with scratches as deep as possible along the line of the cut on both sides of the

251 CHAPTER THIRTEEN

e cho Ainaleur'� ilandlooL

sheet and then clamped in avise and worked back and forth until the sheet breaks at the line. Do not carry the bending too far until the break begins to weaken, otherwise, the edge of the sheet may become bent. A pair of iron bars or pieces of heavy angle stock, as long or longer than the width of the sheet, used in the vise will make the job easier. "C" clamps may be used to keep the bars from spreading at the ends. The rough edges may be smoothed up with afile or by placing alarge piece of emery cloth or sandpaper on a flat surface and running the edge of the metal back and forth over the sheet.
Bends are made similarly. The sheet should be scratched on both sides, but not too deeply.
0:CLEANING AND FINISHING METAL
Parts made of aluminum may be cleaned up and given a satin finish, after all holes have been drilled, by placing them in asolution of lye for half to three-quarters of an hour. Three or four tablespoonfuls of lye should be used to each gallon of water. If more than one piece is treated in the same bath, each piece should be separated from the others so as to expose all surfaces to the solution. Overlapping of pieces may result in spots or stains.
� CRACKLE FINISH
Wood or metal parts may be given acrackle finish by applying one coat of clear Duco or Tri-Seal and allowing it to dry over night. A coat of Kem Art Metal Finish is then sprayed or put on thickly with a brush, taking care that the brush marks do not show. This should be allowed to dry for two or three hours and the part should then be baked in ahousehold oven at 225 degrees for one and one-half hours. This will produce a regular commercial job. This finish comes in several different colors and should be obtainable through any dealers handling Sherwin-Williams paint products.
� HOOK-UP WIRE
A popular type of wire for receivers and low-power transmitters is that known as "push-back" wire which comes in sizes of No. 18 or 20 which is sufficiently large for all power circuits except filament. The insulating covering, which is sufficient for circuits where voltages do not exceed 300 or 400, may be pushed back a few inches at the end making cutting of the insulation unnecessary when making aconnection. Filament wires should be of sufficiently large conductor to carry the current without appreciable voltage drop. Rubber-covered house-wire, sizes No. 14 to No. 10, is suitable for heavy-current transmitting tubes, while No. 18 to No. 14 flexible wire is satisfactory for receivers and low-drain trans-

mitting tubes where the total length of wire is not excessive.
Stiff bare wire, sometimes called bus-wire, is most favored for the high r.f.-potential wiring of transmitters and, where practicable, in receivers. It comes in sizes No. 14 and No. 12 and is usually tin-dipped. Soft-drawn antenna wire may also be used. Kinks or bends may be removed by stretching 10 or 15 feet of the wire and then cutting into small usable lengths.
The insulation covering power wiring which will carry high transmitter voltages should be appropriate for the voltage involved. Wire with rubber and varnished cambric covering, similar to ignition cable, is usually available at radio dealers. Smaller sizes have sufficient insulation to be safe at 1000 to 1500 volts, while the more heavily insulated types should be used for voltages above 1500.
� WIRING TRANSMITTERS AND RECEIV ERS
It is usually advisable to do the power-supply wiring first. The leads should be bunched together in cable form as much as possible and kept down close to the surface of the chassis. Chassis holes for wires should be lined with rubber grommets to fit the hole to prevent chafing of the insulation. In cases where powersupply leads have several branches, it is often convenient to use fibre terminal strips as an-
chorages. These strips also form handy mountings for wire-terminal resistors, etc. When any particular unit is provided with a nut or thumb-screw terminal, solder-lug wire terminals to fit are useful.
High-potential r.f. wiring should be well spaced from the chassis cit. other grounded metal surfaces and should run as directly as possible between the points to be connected without fancy bends. When wiring balanced or
push-pull circuits, care should be taken to make the r.f. wiring on each side of the circuit as symmetrical as possible. When it is necessary to pass r.f. wiring through the chassis, a feed-through insulator of low-loss material should be used, or the hole in the chassis should be of sufficient size to provide plenty of air space around the wire. Large-diameter rubber grommets may be used to prevent accidental short-circuit to the chassis.
By-pass condensers should be connected directly to the point to be by-passed and grounded immediately at the nearest available mounting screw, making certain that the screw makes good electrical contact with the chassis. In using tubular paper by-pass condensers, care should be taken to connect the side marked "ground" or "outside foil" to
ground. Blocking and coupling condensers should be
mounted well spaced from the chassis.

252 CHAPTER THIRTEEN

WorLhop Practice

High-voltage wiring should be done in such a manner that exposed points are kept at a minimum and so that those which cannot be
avoided are rendered as inaccessible as possible to accidental contact.

purpose. Never use panel bearings of the nonmetal type unless the condenser shaft is grounded. The metal bearing should be connected to the chassis with a wire or grounding strip. This prevents any possible danger.

� SOLDERING
The secret of good soldering is in making certain that all members of the joint and the tip of the iron are clean and in allowing time for the joint, as well as the solder, to attain sufficient temperature. Sufficient heat should be applied so that the solder will melt when it comes in contact with the wire forming the joint without the necessity for touching the solder to the iron. Soldering paste, if the noncorroding type, is extremely useful when used correctly. In general, it should not be used for radio work except when it is necessary to make the soldered joint with one hand. In this case, the joint should first be warmed slightly and the soldering paste applied with a piece of wire. Only the soldering paste which melts from the warmth of the joint should be used. If the soldering iron is clean, it will be possible to pick up adrop of solder on the tip of the iron which can be applied to the joint with one hand, while the other is used to hold the connecting wires together. The use of excessive soldering paste causes the paste to spread over the surface of adjacent insulation causing leakage or breakdown of the insulation. Except where absolutely necessary, solder should never be depended upon for the mechanical strength of the joint; the wire should be wrapped around the terminals or clamped with soldering terminals.
� CONSTRUCTION NOTES
Lockwashers should be used under nuts to prevent loosening with use, particularly when mounting tube sockets or plug-in coil receptacles subject to frequent strain.
If a control shaft must be extended or insulated, aflexible shaft coupling with adequate insulation must be used. Satisfactory support for the shaft extension may be provided by means of a metal panel bearing made for the

� COIL WINDING
Dimensions for coils for various inductances or frequencies may be determined from the data given in Chapters Four and Fifteen.
The number of turns specified should be spaced out to fill the specified length on the form. The length specified should be marked on the form and holes drilled opposite the pines to which the ends of the winding are to connect. Scrape one end of the wire and pass it through the lower hole in the form to the pin
to which the bottom end of the winding is to connect and solder this end fast. Unroll an amount of wire approximately sufficient to make the winding and clamp the spool in a vise so it will not turn. The wire should be pulled out straight and the winding started by turning the form in the hands and walking up toward the vise. A fair tension should be
kept on the wire at all times. The spacing can be judged by eye. lf, as the winding progresses,
it becomes evident that the spacing is going to be incorrect to fill the required length, the winding may be started over again with a different spacing. If the spacing is only slightly off, the winding may be finished, the top end fastened and the spacing corrected by pushing each turn. When complete, the turns should be fastened permanently in place with coil cement. After alittle practice, the job of determining the correct spacing will not be difficult.
Sometimes it becomes necessary to adjust the number of turns on a coil experimentally to fit a particular job. The easiest way to do this is to bring awire out from one of the pins, extending through the hole in the form for a half-inch or so. The end of the winding may then be soldered to this extension, rather than to the pin itself, and the nuisance of repeatedly fishing the wire through to the pin avoided until the correct size of the winding has been determined.

253 CHAPTER TIIIill'EEN

CHAPTER EOUHTEEN

,Learning Me Cole

You are about to learn a new language -- the language of code. In learning you
should consider it a language of sound only, for the sole form in which it is written is its
English equivalent. By far the best way to learn this new lan-
guage is to secure the help of a competent operator skilled in code. If that is possible, you should never have to look at an alphabetic table of code equivalents; you should be taught by sound. Since everyone cannot get tutoring, however, it is necessary for us to teach this sound language as best we can by means of the printed page.

You don't think of the spoken letter "u," for
example, as being composed of two separate and distinct sounds -- yet actually it is made
up of the pure sounds "ee" and "oo," spoken in rapid succession. You learned the spoken letter "u" as asound unit itself. Similarly, you should learn code letters as sounds themselves,
and not as combinations of other sounds. A skilled operator does not think of dits and dahs
when copying, but actually hears the head-
phones speak words to him. He has mastered this new language. The sound "dab didididit
dit" is just as familiar to him as the sound of the spoken word "the," for which it stands.

The Continental (International Morse) code is the type used in all radio work. It consists of various combinations of two different-length sounds properly interspersed with spaces, forming letters of the alphabet, numbers, etc. In

�LEARN BY SOUND
Do not use the "didah" tables below for memorization, for if you do you will have a visual concept of code, instead of asound con-

nearly all radio work, this sound takes the form of ahigh-pitched audio beat or oscillation, usually ranging from 300 to 1200 cycles, depending upon the receiving operator's individual preference. Lacking an audio oscillator

cept; when you hear asound you will have to convert it first into "didah" language, and then into letters. If you learn the sounds directly with their letters, however, without using visual reference, you will eliminate this

itself, the best way to simulate the high-pitched extra step while learning.

sound of code in headphones is to whistle. Since
awhistle cannot be very well shown in print, terms have been coined which closely duplicate the actual letter-sounds in code. We speak of the short sound as "dit," of the longer one (three times as long, actually) as "dah."
For purposes of explanation to you and to

So, while we must print atable of code equivalents, memorize them in sound. If possible, get someone who knows the code to start you off. A member of your family or afriend (pref-
erably not one wishing to learn code) will suffice if he will study these paragraphs and practice the voice exercises ashort time. Then

whomever might assist you, let us say that the sound "dit" is pronounced as "it" with a "d" before it. The sound "dah" is pronounced with "ah" as in "father." The sound "dah" is
always stressed or accented -- not in a different tone of voice, but slightly drawn out and

ask him to pronounce the sounds to you, identifying them with the letters for which they stand. Take afew letters at atime, such as in the groups suggested below. As your "instructor" sings out the sounds to you, you should be
able to call out the letters, or vice versa. This is

the least bit louder. The sound "dit" is pro- excellent practice. It will come harder at first,

nounced as rapidly and sharply as possible; for purposes of easy combination, as aprefix, it is often shortened to "di." When combinations of the sounds appear as one letter say them

possibly, than other ways, but you are set for a successful code career if you learn by sound. In other words, after you read this chapter do not go back and memorize the code tables. Don't

smoothly but rapidly, remembering to make the sound "di" staccato, and allowing equal stress to fall on every dah. There should never

even take another look. Get someone to call out the sounds to you, either by "didah" language or whistling, and preferably acombina-

be aspace or hesitation between dits and dahs
of the same letter. These are simply convenient vocal terms to
use in duplicating the sound of code. You should never think of a letter as being composed of sounds, but rather as asound itself.

tion of both. Before taking up letter-sounds, let us ob-
serve some of the element sounds. Practice saying to yourself the sound "dididididi ..
(dits in rapid succession). It should sound like ablast from amachine-gun: staccato, evenly-

254 CHAPTER FOURTEEN

eeearnin9 the Cole

spaced, precise. To make certain you get correct timing, start tapping the top of a table

dididididit

continuously in smooth, even sequence, like clockwork; if ametronome is available, it will

dandandandandah

serve admirably. The tapping (or metronome ticks) should be at about 100 per minute, or a

dit

bit less than two per second. Synchronize your tonguing of dits at four per metronome beat.

dah

Be careful not to say "didididi ...didididi " thereby leaving a space after every

didah

fourth dit; they should be as even and regular as the metronome beat, but four times as fast.

didahdit

If it will help, slightly accent the dit which

coincides with the metronome beat. When you

can do this easily, begin repeating "dandahdandandah ..." so that the beginning of each second dah synchronizes with a metronome tick (or table rap). Let the dahs run
smoothly and make them of equal length. Do

The "5" should merely be five staccato dits in the same sequence and speed you practiced above; the zero should similarly be five dahs. "A" is our first character using both sounds. Make that di very short, the dah the usual

not let them become choppy; your voice should be almost continuous, broken only for that short instant your tongue cuts off the tone to make the "d" sound.
Now alternate, repeating dits for amoment and then switching to dahs, without stopping to take abreath. This will show you the proper proportion of dits to dahs at this speed, or any speed to which the metronome is set; that is, four dits or two dahs per metronome tick. You should now practice alternating this voice simulation with whistling. Again, make your whistled dits very short and staccato, your whistled dahs smooth and full.
Another excellent practice exercise is the alternating of single dits and dahs. Practice repeating (and then whistling) "didandidahdidandidah .. "and then "dandidandidahdidandi ...." Here again the tone should be almost continuous, the dit as short as possible. An hour, in several shorter periods, is not too much time for these exercises. If you can master them, you will have no difficulty in forming
the various letter and number sounds.

length, properly stressed; it should have the same metric swing as does the word "to-day,"
rapidly spoken and strongly accenting "day." Similarly, the "didandit" of "r" should have the same metric swing as "reni_ted," the second syllable strongly stressed, the final one not accented and as short as you can make it.
Spend at least ahalf-hour on this group; and preferably more, though your practice should then be split up into shorter periods. Take your time and learn the sounds of these letters thoroughly. Repeat them in "didah" language, and then whistle the characters.
You should immediately begin practicing copying down the characters your "instructor" calls out to you in sound. Copy them simul-
taneously speaking the letter, if you like. For the first copying practice it is well to print; in the transmission of non-English text, which you often will be getting for letter-practice, longhand letters are sometimes confused with each other. For example, "1" is mistaken for "e," an uncrossed "t" for "1," etc. The U. S.

� LETTER SOUNDS

Signal Corps requires printing ability by its field operators, since much of the text handled

When you have mastered the timing and is in code groups and errors would cause much

rhythm exercises explained above, you can be- difficulty in reading the actual message. Print

gin to combine these groups and form letter until you are able to receive about ten words

sounds. In doing so remember that this is a per minute, above which speed you will want new type of spoken language you are learning, to resort to longhand to keep up with the text.

not acollection of short and long code symbols taken from a printed page. In fact, it might be better to forget that you are learning

Learn to print the characters rapidly and without conscious effort; your mental effort must be on reading the code sounds and not

"code" and think of it instead as the "didah" language.
Let us take a few letters and numbers to start with, as shown in the following table.
Remember that you should look at this table only briefly while reading this text; when you

on your finger movements. Never look back over your copy when receiving, nor try to
guess what word is coming; copy what is sent. If you do not immediately recognize acharacter, skip it and devote your full attention to the next one; if you try to remember it, you

come to study, have someone hold this book doubtless will miss several letters in a row.

and coach you in memorizing the code-sound By checking your written copy with the trans-

of each letter.

mitted text, you can determine what letters

255 CHAPTER FOURTEEN

Dne Radio AmaleuA -llandhooh

are giving you trouble and give special attention to them.
Here is another group of characters. Before beginning their study, have your "instructor" review you on the first group. Intersperse learned sounds with new ones when studying this next group:
dididit
didahdidit
dididah
dahdahdidah
didahdahdah

and dahs; practice that again momentarily, and you should have no trouble.
When learning the sound of "v," practice "h" and then make the final dit adah; practice the two letters intermittently as you did "u" and "s," repeating in your mind what letter is concerned with each sound.
Here is afourth group:
didit
dahdididit
dahdidahdah
didahdahdit

The first letter obviously is simply three staccato dits. "L" is more difficult; its metric swing is like that of "fraternity"; again, make
dits rapidly, particularly the final ones, and stress the dah. "U" is similar to "s" with the final dit changed to a dah; practice the two letters interchangeably to get the rhythm. "Q" and "j" should be smooth; stress each
dah equally, make the dit short. Study these in the same manner as the pre-
vious group. You have nearly twice as many letters to remember now, so your progress will be a bit slower. Don't rush; learn each sound thoroughly before proceeding to the next. Speed will come later, and it will come rapidly if you learn by sound. There are quite a number of words which can be made from
the letters so far studied, and examples are given for each successive lesson on alater page. Practice saying the sounds to yourself, particularly between study sessions. Occasionally you should have an operator who knows code check you on your progress.
Here is a third group, to be taken up only after you have learned the previous eleven
characters well:

By now you should have developed sufficient timing sense and code consciousness that you can pronounce new sounds without difficulty. Simply remember to keep characters smooth, without spaces.
In between practice periods, when convenient, notice street signs and posters containing these letters; when you see one you know, call out its code-sound. Keep this practice principle always in mind and use it agreat deal; you will find agoodly amount of idle time you can put to good use -- riding back and forth to work, or walking to the corner drugstore. Or even if you are not in the vicinity of signs, there are a number of short words you can practice saying to yourself in sound language. Make your character formation snappy, leave a recognizable space between letters, plenty of space between words. Use both whistle and voice technique. We want to get you thinking subconsciously of code whenever you see letters and words.
Now, another group:
di- dah- dah
dahdidah

didididit

dandandah

����1

dandit

dandidandit

dahdahdidit
dahdah
When these have been thoroughly learned, you may proceed to the final group:

didididah

dandidit

The first two letters should give no trouble. Be careful not to have aspace in the sound for "n"; nor to make the dit anything but short. "C" can best be simulated by remembering our earlier exercise of alternating single dits

dandididah

dididandit

dandandit

256 CHAPTER FOURTEEN

epeearning the Code

You already have learned the two simplest numerals: five and zero. The others are:

fore attempting to use the key. A reliable rule is that there should be avertical movement of

didan-dandandah

about one-sixteenth inch at the key knob. This is measured from the top surface of the

dididandandah

knob. It is set by the rear screw adjustment. When the knob's top surface goes down about

didididandah

'one-sixteenth inch upon pressing the key, you have the approximate "average" spacing

dididididah

between the key contacts. In making any key adjustment, be sure to loosen the lock nuts

first, so that you do not strip the threads.

dandidididit

Tighten all lock nuts when you complete the

adjustment. The contact points should be per-

dandandididit

fectly aligned by means of the side screws. There should be a very slight degree of side

dandandandidit

play, between the two side screws. These screws should be tightened, then loosened just

dandandandandit

abit so that the key moves freely and does not bind. Recheck the contacts after this adjust-

From here, your progress will be principally a matter of practice. Listen in on the shortwave bands for commercial automatic-tape stations sending slowly. While the schedule varies, you will usually find suitable stations in the daytime between 7,000 and 20,000 kilocycles, and at night between 4,000 and
10,000 kilocycles. Do not be surprised if you seem to copy "solid" but your text does not make sense; much commercial transmission is in coded groups or foreign language. Tape
sending is precision itself; the more you listen the more will become fixed in your mind the sound of code perfectly sent and the more easily will you be able to simulate it.

ment to be sure they are true, making any slight readjustment necessary.
Although the "one-sixteenth inch" rule is agood one to follow for first adjustment, the amount of vertical movement can later be
changed to suit your particular fist. The spring tension, likewise, must be set for the individual operator. Some prefer a heavier spring than others. The primary consideration is to send good code; how you have your key adjusted to do this depends on what you find best for you. However, it should be remembered that too heavy a spring tends to make your sending "choppy," causing you to "clip" your dahs
and dits as well as being tiring for long periods of sending. Similarly, too light a spring tends

�LEARNING TO SEND
It is important that you should learn the correct sound of code letters thoroughly before ever touching a telegraph key. If you do not know how a code letter should sound, no amount of playing with akey will teach you. When you are at the point where you can unhesitatingly call out each letter as your "instructor" pronounces or whistles the sound, you are ready to learn how to handle akey. First, however, you need some device to produce a tone. A buzzer set will serve the purpose, but amuch better tone source is the vacuum-tube oscillator since it duplicates the audio beat note a radib operator copies. Construction data on both types are given near the end of this chapter.
A telegraph key is simply an on-off switch in convenient form for rapid manipulation. Pressing the key knob closes the electrical contact and produces the "mark" or sound; releasing it allows the contact to break, producing the' space or no-signal period. Correct key adjustment is that adjustment which fits your particular touch, and it is important that you arrive at the correct adjustment be-

to cause you to run characters together, there being insufficient control of the key. Remember, you are making the characters, the key
isn't! Generally speaking, a somewhat heavy spring allows better control, particularly of dots. With a spring of "feather-weight" tension, the dits are likely to run away from you and you will find yourself slurring them. No treatise on key adjustment ever can solve
the individual's problems in this line. Only by personal use can you find the correct adjustment for yourself.
There is a definite sending posture which should be observed. Sit upright in your chair, square with the operating table, with your arm on aline with the key. The key knob should be about eighteen inches from the front edge of the operating table, allowing room for the elbow to rest on the table. The muscle of the forearm should support the weight, and the wrist should be off the table. A table about thirty inches in height is best. The key may be fastened by means of wood screws directly to the table if one is available for permanent use. Otherwise, it may be fastened to arectangular piece of thin board such as three-ply veneer, about six inches wide and two feet long.

257 CHAPTER FOURTEEN

Dhe leadio Amateur'� -ilanihoolz

The manner of grasping akey knob is also the choice of the individual operator. When learning, place the thumb against the left edge
of the key knob, the first finger on top of the knob at the rear and lapping over the rear

level at which you make aminimum of errors. Accuracy and perfection come first -- speed will come with practice. Here again, it is well to have someone skilled in code check your prog-
ress occasionally.

edge just abit; and the second finger against the right edge of the knob, about in the center or slightly to the rear of center. In no event should the grip be tense. The first and second

� ADVANCED PRACTICE
To become expert in transmitting good code, after you have thoroughly learned each letter and numeral and can both send and copy let-

ters without hesitation, your best practice is to

listen to commercial automatic tape stations.

Perfectly-sent code can be accomplished only

by amachine, and you want to get fixed in your

mind, indelibly, the correct formation of each

and every code character and in particular tilt

associated spaces. Notice the formation of each

letter, the spaces left between letters and

Fig. 1401 -- Illustrating the correct positon of the hand and fingers in using atelegraph key.

words, and the proportion in length of dits to dahs.
One of the best methods for deriving this

fingers should be slightly arched, not held out straight. The third and fourth fingers should be permitted to curl naturally toward the palm
of the hand, but they should not be tightly clenched. Keep the fingers, hand and wrist
relaxed at all times. Now that you are all set, you can begin
sending practice. Before taking up lettersounds, it is best to achieve smoothness and facility with element sounds, as we did in learning by voice. Begin by making aseries of dits -- ten or more in a row. Make them evenly spaced, precise, a bit slower than the speed at which you learned the voice-sounds. Work on both dits and dahs, and then alternate, just as you did in voice work. Here again
ametronome, or asubstitute such as someone rapping on the table for you, will be avaluable tinting guide. Remember to synchronize your
sending at four dits, or two dahs, per beat. Make the beats about 60 per minute -- one per second -- as compared to the 100 in receiving practice. Synchronize your sending with the spoken sounds, if you wish. Keep your wrist flexible; allow it to bob up and down with your sending. If you find that your fingers or whole arm are doing the work, stop, and start over again: Be particularly careful to make dahs smooth and full; the tone should be almost
continuous, broken only for that tiny instant the key contacts open.
When you have achieved smoothness in these exercises, begin on letter-sounds in the same sequence as you learned the voice, starting out with the five and the zero. Aim at perfect rhythm in duplicating just what you have learned by voice. In particular, be careful not to leave spaces between elements of letters. Practice sending material from a book, and
also non-English text such as the cipher groups printed in this chapter. Keep your speed at the

association is to find a commercial or other tape station sending at about your maximum receiving speed. Listen to the transmissions as you would at amusical concert, concentrating on assimilating every detail.
The spaces between words may seem exaggerated, simply because you have probably been running yours together. A score of other details where the automatic transmission is dif-
erent than yours will very likely show up in the same text. From all this you will learn where your own faults lie and be able to correct them.
If you can locate a tape station sending double (repeating each word) you can get
excellent rhythm practice. Set up an audio oscillator alongside your short-wave receiver so you can hear both simultaneously. As each word comes through the receiver fix it in your
mind; then, as the tape repeats it, send the same word on your oscillator simultaneously with the tape, as closely to perfect synchronism as possible. Perhaps you will find yourself leaving too much or too little space between characters, or making certain dahs too long --
these are the most common errors. Remember that all inaccuracies are yours, and profit accordingly. By such constant practice you will learn the proper rhythm and precision of perfect code. It's bound to work itself, sub-
consciously, into your sending. Probably the most important single factor
in sending ability is this sense of proper spacing. While it can be much more easily obtained
by listening to tape than by visual study, the student should understand the mechanical relation of the various marks and spaces, as shown in the accompanying chart.
In transmitting text there of course is aneed for code symbols for punctuation marks, and some special procedure signals to facilitate rapid transmission. The important ones are:

258

CHAPTER FOURTEEN

ofearning the Code

period comma question mark double dash end of message end of work wait invitation
to transmit

didahdidahdidah d--ahd--ahdidi- dah- dah dididahdahdidit dahdidididah didahdidahdit didididahdidah
-- -- didahdididit
dandidah

Others which you might encounter in future code work, and which you therefore should learn after you become reasonably proficient with the previous characters, are:

hyphen

dandididididah

parentheses

dandidandandidah

colon semicolon quotes error

- danda- ndandididit dandidandid--andit di--dandidid--andit didididididididit

apostrophe

di-dand--and- and- andit

fraction bar

dandididandit

�HIGH-SPEED OPERATION
Perfect copy at high speeds should be the eventual objective of every operator. Ability to read rapid code "in your head" means little;

what counts is what you can transcribe to paper correctly. Since the limit of writing ability is about 30 words per minute, one must resort to atypewriter for copy at higher speeds. The first essential is touch-typing ability; no "two-fingered" typist ever became a really good operator. If you have not been schooled in touch typing, therefore, try to find some class giving the necessary instruction; often you can find an evening public high school offering such courses. Remember there should be amargin of about twenty words per minute between your straight typing speed and the code speed which you can expect to obtain on the "mill."
Assuming you have touch typing ability, then, you simply substitute the typewriter for your pencil. Use standard letter-size paper, and
write double-space. Do not try to capitalize any letters at first; all lower case type will suffice until you become proficient. Using some source of copy at about your normal receiving speed, such as acode machine or acommercial short-wave station if you can locate one, practice copying smoothly and evenly, preferably a letter or two behind the transmitted text. Do not listen and then type ferociously
for a second ...and listen ...and type hurriedly again. Your typing must be dissociated, consciously, from your code reception. After you are able to handle this first speed,
pick stations sending a bit faster (or step up the code machine), so that you get about 90 per cent of the text; when you copy solid, again step up the speed.
But all the ability is not in typing; you must be able to read the code as well. In high-speed copying a new principle is involved. It is one you will reach automatically if you progress sufficiently far in your practice, but it is worth explanation here.
When one first learns the code by sound, he learns letters first. (In some cases, students
studying visually learn parts of letters first, but we have tried to obviate that error by teaching code sounds in letter-units.) With a

dit dah. element .space character space word space

Nan

31 Time duration ratio, relative to "dit" as
7

III !Ili T I IL

IiI Ii

ItIi iIi ttI I I11 11

MBAR

CHAPTER FOURTEEN 259

-5h.e Radio Atnaleur'J ilanddooh

good deal of practice, one may slowly increase his speed of copying until he reaches acertain point -- differing with various individuals -- which is the maximum speed at which he can copy individual letters without having them seem to run together or blur in his mental
thought. The average is around 28 words per
minute. Progress beyond that point must be on a
new principle of copying, then. It is simply the process of copying by word-sounds instead of letter-sounds. An operator capable of receiving,
say, 25 words per minute, can listen to 35 or 40 w.p.m. text and easily pick out the shorter, more common words such as "the," "and," "but," and so on. He can do so only because he is copying word-sounds and not letter-
sounds. A skilled operator does not hear letters,
but actual syllables and words. The code sound "didit dandit dandandit" (the familiar suffix
"ing"), for example, is mentally heard by the skilled operator as a complete sound and not as three different letters; again, it is as if someone had pronounced the syllable to him.
There is no secret to the attainment of this ability, except continued practice. You can

motion. Relieved of the fatigue of tense motions involved in making rapid dits, the fingers are able to tap out code of much higher
speed. The position of the "bug" should be similar
to that of the straight key. The arm again should be relaxed, and the right side of the hand should rest on the table immediately in front of the key. To the left of the control paddle is the thumb, which when moved to the
right trips the dit vibrator; to the right of the paddle are the index and second fingers (some operators prefer to use only one), which when moved to the left operate the dab contact, similar to astraight key set on edge.
Operating motions consist of an easy roll of the wrist and hand from left to right, and return. With the key connected to atone source, practice sending aseries of dits, then of dahs,
then of alternating single sounds -- just as we did in learning voice sounds. Take lettersounds in that sequence, as well. Before attempting to use a "bug" in actual operating, you should have become proficient with it through practice with an audio oscillator and
headphones.

help yourself, though, by having someone send to you rapidly (about 10 w.p.m. above your normal receiving speed) the common words and syllables such as and, of, the, to, a, in, is, it, for, ing, ion, of, on, end, that, and so on. Nothing can equal the practice obtainable by

�TONE SOURCES FOR PRACTICE
A buzzer set, connected as shown in the diagram of Figure 1402, will serve the purpose of a signal source to be keyed for code practice.

copying commercial tape transmissions, however.
It should be pointed out that one important

ro Pho,es

liro Dry Cehi/17 series cor,ected

prerequisite to high-speed copying is ability to spell. Since you do not hear letters, but entire

word-sounds, spelling of a word on the type-

writer is up to you. True, it was sent only one

way, but since you did not hear individual let-

ters you do not know exactly which were sent;

you know only what the word sounded like to

you. You must know the spelling of words be-

fore you can recognize them from their code

sounds and be able to transcribe them cor-

rectly to the typewritten page. The complement of high-speed receiving is
of course high-speed sending. You should never let your sending speed outstrip your receiving speed, however; if that does happen, it simply means that your conception of the code is mechanical rather than in terms of sound. A semi-automatic key, or "bug," can be used for speeds higher than obtainable on a"straight" key, but in no event should astudent attempt to handle a "bug" until he has mastered the regular key. The semi-automatic key is a mechanical device which produces a series of dits (when the proper lever is tripped) by
means of a vibrating contact. Dahs are made in the usual manner, although the operating knobs work in ahorizontal instead of avertical

Fig. 1402 -- Circuit of abuzzer code practice set. The headphones are connected across the coils of the buzzer with a condenser in series. The size of this condenser determines the strength of the signal in the 'phones. If the value shown gives an excessively loud signal, it may be reduced to 500 pad. or even 250 liedd.
A much better tone, however, since it is identical to the audio beat note a radio operator reads, is that produced by a vacuum-tube audio oscillator. Beside a pair of headphones and key, the parts required are an old audio transformer, grid resistor and condenser, tube and socket. For battery operation, the tube may be any 1%- or 2-volt filament type, such as a 1G4G, 1LE3, '30, etc., power being sup-

260 CHAPTER FOURTEEN

1046

cleearniny Me Code
throwing a few switches, facing pairs of students may join in two-way conversations, and each such pair may be monitored in turn by
6F8G

Fig. 1403 -- iring diagram of a-simple audio oscillator for use as acode practice set.

plied by a No. 6 dry cell and a 22%-volt B battery. The parts may be mounted on abaseboard, as shown in the layout drawing; or they
may be enclosed in aportable carrying-case in which event flashlight cells and ten or more volts of B battery will suffice for power. If nothing is heard in the 'phones when the key is depressed, reverse the leads going to either transformer winding; reversing both sets of
leads will have no effect. If desired, an oscillator powered from the 110-volt line may be used; the hookup of a simple one is shown in Figure 1405.

Fig. 1405 -- Wiring diagram of the a.c. oscillator. Ci -- 250-mdfd. mica. C2 -- 25-pfd. electrolytic, 25 volts. J1, Ja -- Small closed-circuit jacks.
RI---0.6.13'-5fv-om1l9etF,g8o0h).1m-sa,mp%e-rweatt.transformer (Thordarson
-- Small audio transformer (Thordarson T-13.434).

�A CODEANSTBUCTION TABLE
The preceding material has been written primarily for the individual who must learn his code with little or no outside help. If anumber of persons wish to learn the code as agroup, or to develop speed after learning, and if some competent instructor can be located, the best method of instruction is by r.neans of a code table.
Any such table should be so wired as to permit the instructor to send to the whole class, but by alittle special wiring many other things are possible. In the one shown here, for instance, each student can practice sending, independently and to himself; yet the instructor, by means of the selector switch, can listen in on each student in turn, can break in on him and correct his errors. Moreover, simply by

Fig. 1406 -- The a.c. oscillator ma, be built on a home-made chassis, of which this is abottom view, consisting of 1x2white pine supporting strips and abaseboard of 34-inch plywood. The 6F8G is the only component mounted on top the chassis.

audFiiog-.os1c4i0ll4at--or Lcaoydoeutpraocfttichee set. All parts can be screwed onto awooden baseboard approximately 5x7inches.

*A' BAT { I.5 VOLTS "B" BAT. { 22 5VOLTS

'PHONES

261 CHAPTER FOURTEEN

..7he Radio Ainalertrii

headsets (and keys) may be paralleled by

Example practice words, paragraphed in groups corresponding to the sequence of letters learned: ate rat tear 50 era rate art 05 tare

closing the inter-connecting switches, each switch being associated with the position of the
same number. A tap from each position is taken off to the multi-point switch (S) at the

sell request jar lure ruse stare suet squeal jet queer slate jute quart lesser quell just tesla sales

instructor's position so that the latter can also place his headset and key in parallel with those

coverlet alone cancel vocal hover the collect never that shone lathe sheet these conquest there neutral severe tenet runner enclose reluctant jocular love helen lunar conclave recluse quiver
bottle corncob yesteryear honey this poppy jitters honey battery bay sissy reconciliate council bor. phone join supper pave ship pay capon nylon coin

at any of the ten positions. When the instructor sends to the entire class
all the small switches are closed, connecting all positions in parallel. With all switches open each student can practice sending by himself, and the instructor can listen to any student by

pyrites copy boil pebble cavity vicious isotope

setting the selector switch Sto the appropriate

work buzzer zero slow hark warmer kind jerk Su-

position. This also puts the instructor in

zanne bump quirk make wink simmer map skim milk tomorrow wholesome jam qualms lowly wharves pompom war mark womanish causeway know ersatz mink waltz
dock kind finger dolores dexterous kidder fixtures goodly golf jigger foggy jinx stuffing dog flight draught fling fox faddist god guffaw dagger mexican doddering textile exchange paradox xylograph

parallel with the student at that position so that the instructor can "work "the student.
Suppose now that students are to practice together in groups of two. Closing Sw2 connects Positions 1and 2; closing Sw4 connects 3 and 4; Sw6 connects 5 and 6, and so on. The odd-numbered switches would be open in this

case. The instructor can listen in on any group by setting the selector switch to either position

the instructor -- all on the one table and from the common oscillator. Bigger "nets" can be switched at will, while other students at the same table continue practice either singly or in pairs, all under control and supervision.
The instructor should sit at one end of the
table, facing down its length, the students arranged along each side. A center partition and cross-partitions make asort of "private office" of each operating position, each with its �phones, key and switch. The oscillator is at, or handy to, the instructor's end of the table, his controls on alow panel in front of him, running
across the table. At each student's position, akey and head-
set in series are bridged across the line carrying the continuously-running audio tone. The

in the group, and again can break in on the work. Larger groups can easily be formed; for instance, closing switches Nos. 2, 3and 4will connect positions 1, 2, 3and 4together; closing Sw6 and Sw 7 will connect positions 5, 6and 7
together, and so on. The number of positions can be extended
indefinitely by following the same wiring system. Separate tables, grouping perhaps ten men to atable, can be used if the room is too small to accommodate asingle table for alarge class. In such a'case the instructor could have a central position with a separate selector
switch for each table. The code-practice oscillator described earlier
in this chapter is suitable only for two or three sets of 'phones; it is, therefore, necessary to

10 � tie
�

sw,

swz �sw,

swa sw,

sw,

sw.8-

INSTRUCTOR >)

STUDENTS' POSITIONS

Fig. 1407 -- Wiring diagram of the code-instruction table.

262 CHAPTER FOURTEEN

Zearning de eel,

PRACTICE CIPHER GROUPS

Cipher groups make better practice material than plain English because you can't foresee the next letter. The groups below, taken from the operator's manual of the Signal Corps, are representative both of the practice material used in service schools and of ciphered messages actually used in the services. Their breakdown into 5-letter units makes it easy for you to determine your approximate speed, figured on the
basis of 5characters to aword. With someone to send toyou, this material

of course provides the best possible copying practice. Check back for your errors, con-
centrate on the letters that are proving difficult for you. If you find yourself memorizing some of the combinations or their order, use each group backwards or start from the bottom of the page. There are
plenty of practice possibilities in the following list, even for the experienced operator who is seeking higher speed on the typewriter.

OKICQ
2468 0 CKTOG AXBTR

CEMID JXEFY
EFMEY 4 9 2 8 5

0 3759 CWXCK DKAKX
WQYFZ

JNABD JBCYD
PGZUN CTILL

OMIWG
HSPCQ XGLDT 80701

THQVI
HMOUS TVCPT CPTKO

RKMOZ EKJCD DJHNG HAHGJ

20 184 VNFBH DIEUY JCUWH

LDOSJ 12310
BAYQT QNB AV

8 3 7 4 6 LKDOP
WEQFQ 1 7 8 6 3

2 5 8 9 4 JXHGQ 980 15 MBUDF

MVHVD APLKM NDHBH JTIOE

AEIOU 13579
CJRNO YOUMK

MNLFG
YOUMK VOIUM 127 0 9

9720 1 EFXNK
LCKQH OCYAA

ALCRJ DLZIK
PHVTX HIMAD

ZLJBI 8470 7
KDKNG FHFUX

CPNZI JBOAH
GYEVZ ARJZU

NTVMG
WOHZR IPAZQ FNZQJ

DXAQN
�11 90 2 0 6143 CPTKO

PGMAE 10 273 DXNZE 9 8 2 3 4

IRAGO PZOKA OQUIW MCNUY

JEHDY MCNCB
PAOKU MDHNH

BCNBX ZAOIZ
30 21Q TQRYW

UEYTA QYEUR QWZXN MNOPA
KQZAX LARDO MQECV BEXZB

ZMNZB NDBGY KDIOE CVQAR
NASTF CFGIL CADXA TTLOR

MAIDP JRNOL 76321 EWSKM 70 365 HZGNA OHWIM SFMCG

VTNEE HELKF
47382 MRFXE

FMEMI ZDLYS
KTLES
WBPCM

4 0 367 SPEJN OBRDK VIODO PVFKQ CUBEZ RBEFP AXTRX

KDIEY MDIDH 0 9165 QZASU

KCNYQ EIURY
MCNBS IHDJI

CVFZA ZLKAM
JDMNF EIOUW

SCQZW ZJHFG DKUER MXNBG

ZXVGX UYRHI DYAFH 12754

OIEUQ
LKCCN LPWCA 0E VAN

PONDL
NPQRT SNKAD KOBDZ

FAWZX
QWERT CUTYR
RGLBH

263 CHAPTER FOURTEEN

..7.ne echo Ametteur'd ilanihooh
OUTPUT 100 5020
70 L7GT

Fig. 1408 -- Circuit of code-practice oscillator suitable for large groups.
Ci -- 0.05-pfd. paper (for tone of approximately 650 cycles with Li = 1henry); substitute lower circuit for variable tone.
Ci -- 25-pfd. electrolytic, 25-volt. C3, C4 -- 0.01-pfd. paper. C3, C6 30-pfd. electrolytic, 150-volt. RI-- 2000 ohms, 1-watt. R2 -- 0.25 megohm, i/2-watt. 113 -- 0.25-megohm volume control. R4 -- 150 ohms, 1-watt. Rs -- 3000 ohms, 1-watt. 114, R7 -- 10,000 ohms, 1-watt. Li -- 1henry, approx. (see text). Lz -- 8-henry 55-ma. filter choke. Ti -- Universal output transformer, tube
to voice coil (set for matching Bohm v.c. to 2000-ohm plate load). Si, Sz -- S.p.s.t. toggle. S3 -- 2-pole 6-position wafer switch. Line Cord -- 220-ohm.

build a special type of oscillator for use with the code table. A circuit diagram suitable for handling up to thirty or forty headsets is
shown in the diagram herewith. Operating directly from the 115-volt line, it consists of a 12SJ7 oscillator followed by a 70L7GT power amplifier and rectifier. The pitch of the tone is determined by the constants of the tuned circuit Lei. The inductance should be of the order of 1 or 2 henrys for use with readily-

available paper condensers. In the unit shown in the photographs this inductance is an ordinary small filter choke (Thordarson T-14C61) with the straight section of the core removed
and apiece of wood of the same size and shape substituted so that the mounting clamp can be replaced. This gives a choke of approxi-
mately 1henry inductance. A choice of tones should be available to
avoid monotony. Variable tone is secured by

T241

4 -
24'

4 -i-

-36 --.4.-

Ire

---->1

c�.

G 2

'V-I Z -

36---.1-- 36--+--- 36- --4-- 36=-4.-24'-��

Fig. 1409 -- A suggested form of construction for the code-instruction table. Dimensions may be varied to suit individual needs. Legs may be made of 4by 4stock; top and side rails of 1-inch boards. Appropriate bracing should be provided underneath. A coat or two of shellac after the table is finished will prevent shrinkage.
264 CHAPTER FOURTEEN

earning the Lode

means of the switching arrange-

ment shown below the main

wiring diagram. With three

condensers of the values indi-

cated, an assortment of six tones

ranging from about 600 to

1600 cycles can be obtained.

If only one tone is needed, however, simply use the con-

stants shown in the main diagram.

The output switch S2 en-

ables the instructor to cut off

the tone from the entire class,

and thus serves as a simple

means of attracting attention

when group work or individual

sending practice is being car-

ried on.

No specific construction is

indicated for the table, since

bracing and other details will

depend upon the size, and the

builder's preferences. The im-

portant thing is to be sure

each position provides enough

room for the student to work

comfortably. In particular there

should be sufficient depth so

that the key can be placed the

proper distance -- about 18

inches -- from the edge of the

table.

The interconnecting switches, dw2, etc., can be ten-cent

ti. 1LW

l'he codeliractice oscillator circuit of fig. 1108 built into a

3by 4by 5box. All wiring is insulated from the case.

store s.p.s.t. knife switches and

the instructor's selector switch a wafer-type
unit. The latter can be obtained in the singlepole type with as many as 23 contacts. Since no special precautions need be taken with
respect to insulation or voltage drop, probably

the most economical wiring job can be done by using ordinary bell wire. In cases where the students bring their own headsets it would be
advisable to mount atip-jack assembly at each position to facilitate connection.

265 CHAPTER FOURTEEN

_7n., Radio _Atnafeur'.3 --11anclhooh

Scales Used in Expressing Signal Strength and Readability

Strength

(See QRK and QSA in the Q Code)

Readability

Q QSSAA 1 2.......

Barely perceptible.
...... ......... Weak.

Q QSSAA 3 4..............

..................... .....................

Fairly Good.

good.

QSA 5......

................ Very good.

dable.

QRK2. ..... .................Readable now and then.

QRK3.

..Readable with difficulty.

QRK4.. ...... ...............Readable.

QRK 5........ ...... .........Perfeetly readable.

�"Q" CODE
IN THE REGULATIONS accompanying the existing International Radiotelegraph Convention there iq avery useful internationallyagreed code designed to meet major needs in

international radio communication. This code follows. The abbreviations themselves have the meanings shown in the "answer" column. When an abbreviation is followed by an interrogation mark (?) it assumes the meaning
shown in the "question" column.

Abbresiation
QRA QRB
QRC
QRD QRG QRH QRI QRJ QRK QRL QRM QRN QRO QRP QRQ QRS QRT QRU QRV QR W
QR X
QRY QRZ QSA QSB QSD QSG
QSJ QSK
QSL QSM QS0
QSP QSR
QSU QSV

Question

Answer

What is the name of your station? How far approximately are you from my station?

The name of my station is

The approximate distance between our stations

nautical miles (or

kilometers).

What company (or Government Administration) settles the accounts for your station?

The accounts for my station are settled by the company (or by the Government Ad-

ministration of

Where are you bound and where are you from? Will you tell me my exact frequency (wave-length)
in ko/s (or m)? Does my frequency (wave-length) vary? Is my note.good? Do you receive me badly? Are my signals weak? What is the legibility of my signals (1 to 5)? Are you busy?

1am bound for

from

Your exact frequency (wave-length) is

Itchi

(or

m).

Your frequency (wave-length) varies.

Your note varies. Icannot receive you. Your signals are too weak.

The legibility of your signals is .... (1 to 5).

Iam busy (or Iam busy with

). Please do

not interfere.

Are you being interfered with? Are you troubled by atmospherics?

Shall Iincrease power?

Shall Idecrease power?

Shall Isend faster?

Shall Isend more slowly?

Shall Istop sending?

Have you anything for me?

Are you ready?

Shall I tell

that you are calling him on

kc/e (or

m)?

Shall Iwait? When will you call me again?

Iam being interfered with. Iam troubled by atmospherics.

Increase power. Decrease power.

Send faster (

words per minute).

Send more slowly (

words per minute).

Stop sending.

Ihave nothing for you. Iam ready.

Please tell

that Iam calling him on

kc/s (or

m).

Wait (or wait until Ihave finished communicating

with

)Iwill call you at

o'clock

(or immediately).

What is my turn?

Your turn is No

(or according to any other

method of arranging it).

Who is calling me? What is the strength of my signals (1 to 5)? Does the strength of my signals vary? Is my keying correct; are my signals distinct?

Shall Isend

telegrams (or one telegram)

at atime? What is the charge per word for

including

your internal telegraph charge? Shall I continue with the transmission of all my

traffic, Ican hear you through my signals?

Can you give me acknowledgment of receipt?

Shall Irepeat the last telegram Isent you?

Can you communicate with

direct (or

through the medium of Will you retransmit to

)? free of charge?

Has the distress call received from

been

cleared? Shall Isend (or reply) on

Ws (or m) and/

or on waves of Type Al, A2, A3, or BP

Shall Isend aseries of VVV

You are being called by

The strength of your signals is

(1 to 5).

The strength of your signals varies.

Your keying is incorrect; your signals are bad.

Send

telegrams (or one telegram) at a

time.

The charge per word for

francs,

including my internal telegraph charge.

Continue with the transmission of all your traffic, I

will interrupt you if necessary.

Igive you acknowledgment of receipt.

Repeat the last telegram you have sent me.

Ican communicate with

direct (or through

the medium of

Iwill retransmit to

free of charge.

The distress call received from

has been

cleared by

Send (or reply) on

kc/s (or

and/or on waves of Type Al, A2, A3, or B.

Send aseries of VVV

266

CHAPTER FOURTEEN

me code

Abbreviation

Question

Answer

QSW
QSX QSY
QSZ QTA QTB
QTC QTE
QTF
QTO
QTfl QTI QT.; QTM
QTO QTP QTQ QTR QTU QUA QUB
QUC QUD QUF QUG QUH QGJ
QUK QUL QUM

Will you send on

kc/e (or

and/or on waves of Type Al. A2, A3, or B?

Will you listen for

(call sign) or

Ws (or

m)?

Shall Ichange to transmission on

kes (or

m) without changing the type of wave? or

Shall Ichange to transmission on another wave?

Shall Isend each word or group twice?

Shall I cancel telegram No.

as if it had

not been sent?

Do you agree with my number of words?

Iam going to send (or Iwill send) on

kcia

(or A3, or B.

m) and/or on waves of Type Al, A2,

Iam listening for

(call sign) on

Ws (or

m).

Change to transmission on

Ws (or

m) without changing the type of wave

Change to transmission on another wave.

Send each word or group twice.

Cancel telegram No. sent.

as if it had not been

Ido not agree with your number of words: Iwill re-

How many telegrams have you to send?

What is my true bearing in relation to you?

What is my true bearing in relation sign)?

(call

What is the true bearing of

(call sign) in

relation to

(call sign)?

Will you give me the position of my station according to the bearings taken by the direction-finding stations which you control?

Will you send your call sign for fifty seconds fol-

lowed by adash of ten seconds on

kc/s

(or becring?

m) in order that Imay take your

peat the first letter of each word and the first figure of each number.

Ihave

telegrams for you (or for

Your true bearing in relation to me is .... degrees or

Your true bearing in relation to

frail sign)

degrees at

The true bearing of

to (time).

(call sign) is

(time)

(call sign) in relation

degrees at

The position of your station according to the bearings taken by the direction-finding stations which Icon-

trol is

latitude

longitude.

Iwill send my call sign for fifty seconds followed by a

dash of ten seconds on

kc/s (or

m) in order that you may take my bearing.

What is your position in latitude and longitude (or by any other way of showing it)?
What is your true course? What is your speed?
Send radioelectric signals and submarine sound sig. nabs to enable me to fix my bearing and my distance.
Have you left dock (or port)? Are you going to enter dock (or port)? Can you communicate with my station by means
of the International Code of Signals? What is the exact time? What are the hours during which your station is
open?

My position is

latitude

longitude

(or by any other way of showing it).

My true course is

degrees.

My speed is per hour.

knots (or

kilometers)

Iwill send radioelectric signals and submarine sound

signals to enable you to fix your bearing and your distance. .

Ihave just left dock (or port).

Iam going to enter dock (or port).

I am going to communicate with your station by

means of the International Code of Signals. The exact time is

My station is open from

Have you news of

(call sign of the mobile

station)?

Can you give me in this order, information concern-

ing: visibility, height of clouds, ground wind for (place of observation)?

Here is news of

(call sign of the mobile sta-

tion).

Here is the information requested

What is the last message received by you from
(call sign of the mobile station)? Have you received the urgency signal sent by
(call sign of the mobile station)? Have you received the distress signal sent by
(call sign of the mobile station)?
Are you being forced to alight in the sea (or to land)? Will you indicate the present barometric pressure
at sea level? Will you indicate the true course for me to follow,
with no wind, to make for you?

The last message received by me from

(call

sign of the mobile station) is

Ihave received the urgency signal sent by (call sign of the mobile station) at
Ihave received the distress signal sent by

(time).

(call sign of the mobile station) at Iam forced to alight (or land) at

(time). (place).

The present barometric pressure at sea level is (units).

The true course for you to follow, with no wind, to

make for me is

degrees at

Can you tell me the condition of the sea observed

(place or coOrdinates)?

Can you tell me the swell observed at

(place or coOrdinates)?

Is the distress traffic ended?

(time). The sea at .-...

(place or coordinates) is

The swell at

(place or coOrdinates) is

The distress traffic is ended.

Special abbreviations adopted by the American Radio Relay League. QST General call preceding a message addressed to all amateurs and ARRL Members. This is in effect "CQ AREL." QRR Official ARRL "land SOS." A distress call for use by stations in emergency zones only.

267 CIIAPTER FOURTEEN

CHAPTER FIFTEEN

ifiliceltneotti la la

THIS CHAPTER represents a compilation of miscellaneous data and reference information intended to illustrate and supplement the basic material throughout the remainder of this Handbook.
Inductance (L)
The formula for computing the inductance of air-core radio coils is:

where: A --= area of one side of plate (sq. cm.) n = total number of plates d --= separation of plates (cm.) k = dielectric constant of dielectric.
When A is the area of one side of one plate in square inches and dis the separation of the plate in inches,
C = 0.2235 --kdA (n -- 1)

L --

0.2 A 2N2

3A -I- 9B -F 10C

where: L is the inauctance in microhenrys A is the mean diameter of the coil in

inches B is the length of winding in inches C is the radial depth of winding in

inches N is the number of turns. The quantity C may be neglected if the coil

is a single-layer solenoid, as is nearly always

the case with coils for high frequencies. For example, assume acoil having 35 turns
of No. 30 d.s.c. wire on a receiving coil form

having a diameter of 1.5 inches. Consulting the wire table, we find that 35 turns of No. 30

d.s.c. will occupy a length of one-half inch.

Therefore,

A --= 1.5

B = .5

N = 35

and

L -- 0.2 X (1.5)2 X (35)2

(3 X 1.5) -I- (9 X .5)

or 61.25 microhenrys. To calculate the number of turns of asingle-
layer coil for arequired value of inductance:

\ i3A + 9B 0.2A 2

X L.

More rapid and convenient calculations in designing coils can be made with the ARRL
Lightning Radio Calculator (Type A).

Condenser Capacity (C) The formula for the capacitance of a condenser is:
C 0.088 --kdA (n-1) ��fd.

The dielectric constant determines the quantity of charge which a given separation and area of plates will accumulate for a given applied voltage. "k" is the ratio of the capacitance of a condenser with a given dielectric
to its capacitance with air dielectric.

Table of Dielectric Constants

Dielectric

Power Factor 1

Air (normal pressure) ... 1.0

Asphalts.

...

2.7-3.1

Bakelite -- See Phenol

Beeswax

2.9-3.2

Casein plastics 4.. . .. 6.1-6.4

Caster oil

4.3-4.7

Celluloid.. ........... 4-16

Cellulose Acetate .... 6-8

Cellulose Nitrate ....... 4-7

Ceresin wax... ........ 2.5-2.6

0.2-0.5 2.3 3
5.2-6 7
5-10 3-6 2.8-5 0.12-0.21

Enamel (wire)... ......

Fibre....

..... . 5-7.5

4.5-5

Glass:

Common window.. ... 7.6-8

Crown

6.2-7

Electrical ......

4-5 �

Flint ...... ......... 7-10

0.7 1.4
13
0.5 0.4 0.25

Photographic ..

Plate

Pyrex

Gutta Percha

Lucite.... ..... .....

Mica

Mica (clear India)...

Mycalex

7.5 6.8-7.6
4.5 2.5-4.9
2.5-3 2.5-8
6.4-7.3 68

0.8-1 0.6-0.8
0.7
0.01-0.06 0.01-0.02
0.2-0.3

Paper Paraffin wax (solid)..... 1.9-2.6

0.1-0.3

Phenol: 5

Pure

Asbestos base....

7.5

Black molded

5-5.5

Fabric base

5-6.5

Mica-filled .....

5-6

Paper base...

3.8-5.5

Yellow

5.3-5.4

Polystyrene .....

.. 2.4-2.9

Porcelain (dry process).. 6.2-7.5

15 3.5
3.5-10 0.8-1 2.5-4 0.36-0.7
0.02 0.7-15

Puncture Voltage 2 19.8-22.8
25-30
165 380
600 300
500-750 150-180
-200-250
500 2000
335 200-500
-600-1500
250
300
-90-150 400-500 150-500 475-600 650-750
500 500 40-100

268 cH %PT ER FIFTEEN

el�cellaneou3 20ata

Porcelain (wet process).. 6.5-7

Presaboard (untreated).. 2.9-4.5

Preasboard (oiled) .. .. ..

Quartz (fused). .... .. .. 4.2-5.1

Rubber (hard) 7.... .. .. 2-3.5

Shellac ........ ..,, ... 2.5-4

Steatite 2

. 6.1

Titanium Dioxide 2.. ... 90-170 Urea Formaldehyde res-

ins " ...... ...... . .. 5-7

Varnished cloth (black).

Varnished cloth (yel-

low) 11

2.5

Vinyl resins

Vitrolex ......... ... Wood (dry oak)

6.4 2.5-6.8

Wood (paraffined maple) 4.1

0.6 -- -0.03 0.5-1 0.09 0.06-0.2 0.1
2.4 2
3 1.4-1.7
0.3 3.85 --

150 125-300
750 200 450 900 150-315
300-400 550 440
400-500
115

1At 1Mc. 2 In kilovolts per inch. Most data applies to relatively thin sections and cannot be multiplied directly to give breakdown for thicker sections without added safety factor. 3At 1kc.
4 Includes such products as Aladdinite, Galalith, Erinoid, Lactoid, etc.
3 Phenolaldehyde products include Acrolite, Bakelite, Celeron, Dielecto, Durez, Durite, Formica, Micarta, Synthane, Textolite, etc. Yellow bakelite is so-called "lowloss" bakelite.
Includes Amphenol 912A, QuartzQ, Styron, Trolitul, Victron, etc.
7 Also known EIE Ebonite. Soapstone -- Alberene, Alsimag, leolantite, Lava, etc. Rutile. Used in low-temperature-coefficient fixed con-
densers.
1�Includes Aldur, Beetle, Plaskon, Pollopas, Prystal, etc. 11 Includes Empire cloth.

BMA RADIO COLOR CODES Standard color codes have been adopted by the Radio
Manufacturers Association for the identification of the values and connections of standard components.
Resistors and Condensers:

Condenser voltage ratings are indicated by asupplementary dot or band as follows:

Brown .. Red ..... . Orange Yellow Green Blue

100 volts 200 volts 300 volts 400 volts 500 volts 600 volts

Violet

700 volts

Grey

800 volts

White

900 volte

Gold

1000 volts

Silver .

2000 volts

None ...... .. 500 volte

In some instances condensers are identified by a second (lower) row of three dots or a series of three narrow color bands to the right of the wide capacitance banda. In these cases the first color of the second row or series indicates the number of ciphers following the capacitance digits, the second color indicates tolerance and the third voltage rating.

I. F. Transformers:
Blue-- plate lead. Red -- B + lead. Green -- grid (or diode) lead. Black -- grid (or diode) return.
Nam: If the secondary of the i.f.t. is center-tapped, the second diode plate lead is green-and-black striped, and black is used for the center-tap lead.

A.F. Transformers:
Blue-- plate (finish) lead of primary. Red -- B lead (this applies whether the primary is plain
or center-tapped). Brown -- plate (start) lead on center-tapped primaries.
(Blue may be used for this lead if polarity is not important.) Green -- grid (finish) lead to secondary. Black -- grid return (this applies whether the secondary is plain or center-tapped). Yellow-- grid (start) lead on center-tapped secondaries. (Green may be used for this lead if polarity is not important.)
NOTE: These markings apply also to line-to-grid, and tubeto-line transformers.

For identification of resistance and capacitance values of small carbon-type resistors and midget mica condensers, numbers are represented by the following colors:

0-- Black 1-- Brown 2-- Red 3-- Orange 4-- Yellow

5-- Green 6-- Blue 7-- Violet 8-- Gray 9-- White

Three colors are used on each resistor to identify its value. The body color represents the first figure of the resistance value; one end or tip is colored to represent the second figure; acolored band or dot near the center of the resistor gives the number of zeros following the first two figures. A 25,000-ohm resistor, for example, would be marked as follows: body, red (2); tip, green (5); band, orange (3 zeros).
Small mica condensers usually are marked with three colored dots, with an arrow or other symbol indicating the sequence of colors. Readings are in micromicrofarads (oofd.),
with the color code as above. For example, a 0.00025-ofd. (250-oofd.) condenser would be marked as follows: red (2), green (5). brown (1 zero).
An auxiliary color code has been established covering the tolerances of resistors and condensers and the voltage ratings of condensers. Tolerances are indicated by the following colors, which appear as a fourth dot or band:

Red Orange Yellow Green
Blue

I% . 2%
3% 4% 5%

Violet Gray White Gold Silver.

None ... ....... ..

*Commonly used on resistors.

7% 8% 9% 5%* 10%*
20%*

Loudspeaker Voice Coils:
Green-- finish. Black -- start.

Field Coils:
Black and Red -- start. Yellow and Red -- finish. Slate and Red -- tap (if any).

Power Transformers:

1. Primary Leads If tapped:

. . ..... ... .. .... -Black

Common

.Black

Tap ......... .. ...... ...Black and Yelbno Striped

Finish 2. High-Voltage Plate Winding.

.Black and Red Striped Red

Center-Tap ... .. .. .. .. .. ..Red and Yellow Striped

3. Rectifier Fil. Winding

.Yellow

Center-Tap ..... ....... .. Yellow and Blue Striped

4. Fil. Winding No. 1

.Green

Center-Tap

Green and Yellow Striped

Center-Tap 6. Fil. Winding No. 3
Center-Tap ........

.Brown and Yellow Striped . .Slate
.... Slate and Yellow Striped

LC Constants The product of the inductance and capacity required to resonate at agiven frequency is known as the LC constant. If any value

269 CHAPTER FIFTEEN

.5he Radio Antaieur'� ilanelhooh

of inductance (or capacity) is divided into that constant the quotient is the capacity (or inductance) required for resonance at that frequency. The following table gives LC constants for amateur and intermediate frequencies in terms of microhenries and micro-
microfarads:

Frequency Kc.
100 455 1000 1600 1750 1900 2000 3500 4000 5000

LC Constant ph. X �yid
2,533,030. 122,355. 25,330.3 9,894.64 8,271.12 7,016.68 6,332.57 2,067.78 1,583 .14 1,013.21

Frequency Mc.
7.0 7.3 14.0 14.4 28.0 30.0 56.0 60.0 112.0 116.0

LC Constant uh. X �dd.
516.944 475.399 129.236 122.185
32.3090 28.1448
8.07726 7.03620 2.01931 1.88245

Grid conductance Grid resistance Grid bias voltage Plate potential Plate current Plate conductance Plate resistance
Plate supply voltage Emission current Mutual conductance Amplification factor Filament terminal voltage Filament current
Grid-plate capacity Grid-cathode capacity Plate-cathode capacity
Grid capacity (input)
Plate capacity (output)

go r,, E, Ep,
gp r,, E I.
Ej
I
COP
e re Cpk Cg Cp

For other frequencies: LC - 25330.3 (Freq.m c.) 2
(The above data are contributed by Henry R. Hesse, W 2ERY.)

NOTE - Small letters refer to instantaneous values.

Units of Length

English

Metric

Symbols for Electrical Quantities

Admittance Angular velocity (2wf)
Capacitance Conductance Conductivity Current Difference of potential Dielectric constant
Energy Frequency Impedance
Inductance Magnetic intensity Magnetic flux Magnetic flux density Magnetomotive force Mutual inductance Number of conductors or turns
Permeability Phase displacement Power Quantity of electricity Reactance Reactance, Capacitive Reactance, Inductive Resistance Resistivity Susceptance Speed of rotation
Voltage Work

Y, Y
G, g I, i E, e K or e
z, z
H
4,
N O/� or 41 P,
QX,, qx
L R, r
E, e

Letter Symbols for Vacuum Tube Notation

Grid potential Grid current

E0,e,
jo,i,,,

mil = 0.001 inch

1millinieter= 39.37 mils

= 0.0254 millimeter inch = 2.54 centimeters

1centimeter = 0.3937 inch = 0.0328 foot

foot = 30.48 centimeters
yard = 0.9144 meter mile = 1.6093 kilometers

1meter = 3.28 feet = 1.094 yards
1 kilometer = 0.6214 mile

1micron = 10-6 meter = 0.0001 centimeter

= 10,000 Angstrom units (A.)

1Angstrom = 10-10 meter = 10-8 centimeter

= 0.0001 micron

Relative Electrical Conductivity of Metals at Ordinary Temperatures
(Based on Copper as 100)

Aluminum (28: Pure) Aluminum (alloys):
Soft-annealed. Heat-treated

Cadmium

Chromium

Climax

Cobalt. .....

...

Constantin...

Copper (hard drawn).

Copper (annealed).

Everdur

German Silver (18%)

Gold Iron (pure).

59 Iron (cast) .

. 2-12

Iron (wrought) . ... 11.4

45-50 Lead

30-45 Manganin

. 3.7

Mercury ..........1.66

Molybdenum... ... 33.2

Monel

1.83 Nichrome

4 1.45

16.3 Nickel .

3.24 Phosphor Bronze

89.5 Platinum

100

Silver

6 Steel

5.3 Tin

Tungsten

17.7 Zin

12-16 36 15
106
3-15
13 28.9 28.2

Approximate relations: An increase of 1in A.W.G. or B. & S. wire size increases
resistance 25%. An increase of 2 increases resistance 60%. An increase of 3 increases resistance 100%:
An increase of 10 increases resistance 10 times.

270 CHAPTER FIFTEEN

Mi� ce bectneott� 2 0ctia

Current Capacity of Power Wiring
The National Board of Fire Underwriters has established the following as maximum current densities for commonly-used sizes of copper wire in electrical power circuits:

Gauge No. Circular B. & S. Mil Area

Amperes

Rubber

Other

Insulation Insulation

1 2 4
8 6
10 12
14
16
18

83690 66370 41740 26250 16510
10380
6530 4107 2583
1624

100
90 70 50 35 25 20
15
6 3

150
125 90 70 50 30
25 20 10
6

Greek Alphabet
Since Greek letters are used to stand for many electrical and radio quantities, the names and symbols of the Greek alphabet with the equivalent English characters are given.

Greek Letter Greek Name

English Equivalent

A a

Alpha

Is 0 .

Beta

I` y

Gamma

A �

Delta

E e
Z e-

Epsilon Zeta

e z

11 t1

Eta

�

e O

Theta

I t

Iota

Kappa

A X

Lambda

M pi

N v

;.;.e. k

0 o

Omicron

II 7r

Rho

l lf

Sigma

T 7

Tau

T u

Upsilon

4, ip

Phi

X x

Chi

`1, e

Psi

0 1.0

Omega

Multiples and Sub-Multiples

Ampere Ampere

= 1,000,000 microamperes -= 1,000 milliamperes

Cycle Cycle

= 0.000,001 megacycle = 0.001 kilocycle

Farad

= 1,000,000,000,000 micro-

mierof arads

Farad Farad

= 1,000,000 microfarads = 1,000 millifarads

Henry

=- 1,000,000 microhenrys

Henry Kilocycle Kilovolt

= 1,000 millihenrys = 1,000 cycles = 1,000 volts

Kilowatt

= 1,000 watts

Megacycle

= 1,000,000 cycles

Megohm

= 1,000,000 ohms

Mho Mho

= 1,000,000 micromhos = 1,000 millimhos

Microampere

= 0.000,001 ampere

Microfarad Microhenry

= 0.000,001 farad = 0.000,001 henry

Micromho

= 0.000,001 mho

Micro-ohm

= 0.000,001 ohm

Microvolt

= 0.000,001 volt

Microwatt

= 0.000,001 watt

Micromicrofarad = 0.000,000,000,001 farad

Micromicro-ohm = 0.000,000,000,001 ohm

Milliampere

= 0.001 ampere

Millihenry

= 0.001 henry

Millimho

= 0.001 mho

Milliohm

= 0.001 ohm

Millivolt

= 0.001 volt

Milliwatt

= 0.001 watt

Volt
Volt Watt Watt

= 1,000,000 microvolts
= 1,000 millivolts = 1,000,000 microwatts = 1,000 milliwatts

Watt

= 0.001 kilowatt

Metric Prefixes

1 1,000,000

One-millionth

micro-

1,000
1
100
d

100

1,000 10,000 M 1,000,000

One-thousandth milli-

One-hundredth

centi-

One-tenth

One

Ten

�

One hundred

One thousand Ten thousand One million

deci-
uni-
deka-
hekto-
kilomyriamega-

271 CHAPTER FIFTEEN

..7fl, Radio Antaleur'd ilancgoon

VOLTAGE,CURRENT OR POWER RAT/OS

9 a
6 5 4 3
2

t I t

I I
I
i

1 r I

I
i-N. I , I

I
.,,1, 1 , I

;o--f;z) 1i,1 .1 / 4 1 +

i I t I
,. II /IIi -1 / 4 I

t t
.:14)1-1-,,7- )o I I I

1 1 i
04.,I .., 1 I ,

1
t
I
.0.00,1,--,,.
I , ,

II I i
(o`eI4l- t I , I

oo4-i
1
I

t
t
I i
oo2�,1
i ,

190 8
6 5

e,-- _
e eee

.-
e e e e

4
3

e e e

e I

e e V e

e I

e e

1
o

60
12

7 14

9_0_

1.1-,0

DECIBELS (0 )

POWER

VOLTAGE OR CURRENT

The chart aboye is direct-reading in terms of decibels for all power, voltage or current ratios. The top scale goee from Oto 100 db. and is useful for very large ratios; the lower scale permits closer reading between O and 20 db., or one cycle of the extended scale. Solid lines show voltage or current ratios; dotted lines, power ratios. To find db. gain, divide output power by corresponding input power and read db. value for this ratio, using the appropriate curve (i.e., " X 1" for ratios from 1 to 10, " X 10" for ratios from 10 to 100, " X 100" for ratios from 100 to 1000, and so on). To find db. loss, as when,output is less than input, (livide input value by output value. Current and voltage ratios in db. can be found similarly, provided the input and output impedances are the same. Power, voltage and
current values must be in the same units (watts, millivolts, microamperes, etc.).

ABBREVIATIONS FOR ELECTRICAL AND RADIO TERMS

Alternating current

a.c.

Ampere (amperes) Antenna

a. ant.

Audio frequency

a.f.

Centimeter

cm.

Continuous waves Cycles per second

e. w. c.p.s.

Decibel

db.

Direct current Electromotive force Frequency Ground

d.c.
e.m.f.
f. gnd.

Henry

High frequency

h.f.

IntermediAte frequency

if.

Interrupted continuous waves i.c.w.

Kilocycles (per second)

kc.

Kilowatt

kw.

Megacycle (per second)

Mc.

Megolim Meter Microfarad
Microlienry Micromicrofarad
Microvolt Microvolt per meter Microwatt Milliampere Millivolt Milliwatt Modulated continuous waves
Ohm Power Power factor Radio frequency Ultra-high frequency
Volt (volts)
Watt (watts)

IsU2 m. mfd.
�h. ;44fd.
mv. mv/m. biw� ma. mv. mw. m.c.w.
I? P. p.f. r.f. u.h.f.
v.
w.

272 CHAPTER FIFTEEN

..z.1
e
fi

iiiiile.r..

����� MIMIM,111M10.01MMI ��� 1, MM01. 1����FAIMMUMMI.M�110,0

0, IMIM.������IMIal

111,0MM

iiimummmiummumminv......�....m..........m......mmi. Ismos.........1i11s0s��u:m�1.1.1F.A.WI.I.I.MM.I.M.MI.M.Pi.e.M.IMwP.UmIIuMsIsBU.Z.II.I.P.AL.M.I.II.I.I

IIIIIMIESMIWA IIIMPd1�1111'..M

111111111M.M11111,231101IMSIMOIMilleillIMIIMINIES11.511MMI ,.

e: ; E�78' e.
eg -5 "e'

1oo gi

Ii9111.0111inallinellUMIMIIIIIInligalinni ���...- g91111 1111 10 C .1�O ��M..M .� I ..iE .���N ���: ���M ���I � .E ..4M ���I �0:C �� I m...I ....Z ..mrO .mmeI ....I ,...I ..r.I m.�.M ....E �.. n08

'.....-- ,er. -- P.=:.::etre..�r..- --.1:,--.--, - 71-- r--

s1:s:i:o:n:mZoUmiMmImiCn:g�rreuneia-.s:sm.u=ss1:taPiUmMmeUmMknIeLiIstnatnss-.r...Um1a:g1e6m9s1t:inziaweaUsmi Ls kb

1t1 2�l'il

el9.'`g� E

1.2 e 1

tin; **. t ie.I.
��. E 2..42a2. 7Pe2.,

4 ,t.1 ..b1 .�*.
.;P2,i.e20 3l>i.

�se IlliinifiNginFIZIMMeiliiinakiMIK".ilinlipMEnilignine�

-..,.....

linnumemomeanimmeerinindminlIIIMIM

i :l::i `..g n.ekl'..7.a ...n � d:i ::.�t�-�.�i .�.�..r...g.e"1g.:l1i:ZgetierVedr.i=e::ee7l9:i1g11i."a =:r =1:g ::i �.:lZqW U W 0"

e, -i, ilsiungngimnmeemmeenugtmuBmueimeegimmeimlsIyMsMeIm1e1a1m1g1mMis1u1g1a0r1.1mMu1m1u1s3g1s1mWri.lmlmeoln
lirialilleillinfillile1151111111Kellie.11111M011011111111M
I:li:i.i:l:l=,ialal=l"4=r1i :::lLiTeg�l='1 ..m0..1...04-.1...1.1.5kr M=ienilltigeIidgirneMiinEEeD74016.1r. e1o%=

1109
005 00Z
001

4 : t u � E.E = o re -.;1t..- �:'15,1f�
s Ei4 8 gli

U2 ; ...ge..;:iie''.'

e"t.4".:1
8;1 ier

e..e..ea Ei tle, Wiltlelsakr"A3%C\W=O'M�IjI1I1I4M4I5M1UvMiInI1IIMaI'IaInIiIlIiIlIiInIiIeKdWrIiMMMrIaIiInMeMeI'UUMrM\ImCMInIIIII:MPOaI.MlUtMeIrIfIaI=M1

o0I. s.

: .g

z e

AiiiniMliagliiin.M.V1iiiniiihMedireiniMIEWiiiininging

J 14j et

e0t,

4 61 127

It" ."; Si

IM11111111MOMIIIMINIIIIIIMM111111111n1g1119111eA OE

elligaitillIIIIMEMMICENIUMIII:liel

� 01

e e34: f,emi---L ..3 e�r11 ''',iers'111

.0 e .1i

, :i

Ussens:ieg.a"t..s1m:m--o,.g"e.rCi;:gdgisasZu.1e"g.u:m=IunseelaiseggUmra.mbrs,i�smtBeiIgiulrdArihsmiiggUim=elnigggsrsismuien"gassg:m= L s
r1:111111M11I1r1e1e1i_tMelWlAiIIgO1n1a1l1l1b1M1IMn11101151W1011111111W11M111111012511111111K1.1M1IM1IFPAIMIIINMMIIMI.IN,Is

0�0�.m0m1mim0.mrA.�mi����LorIamN./ mmmur..,r.d.00I1m0 am�

���

mum. �m04:

.�

������

.11111����,1M1Mee

0001

6�10131.11/11�MM/.110�11110M) IMIIM'4110061�01,11M1M- MIIMIIM10�11101M,1�01.11001.10661M
OOL � IIIIMMYOMMV/411111�1111161MMI.M.M/40 00. 0 .0 .1114010M 1rA:MMMEMII.WM1WWIMIIMIIIWIIIMIIM`MM

gio1e11ng11a1m11m1ui1n1i11u1m11w81iR8Or8CoO8MnEu8idt81g1�1e10m1�1i1ng11u21m11o1u61lm1u 10.11N1m 11_oo0so0sr5 1di>

SPIN� N/ 3.,.NV.I.,le321

.; -8@�'e ll: Ij

.rt.0=- .- =" ,.;
e.e. e'tc

.e. tri
E 2 11

...,.. , e 6. 00 61 .1.) 61
MU

ter mej

...

_z�a.m.l9l.,e-�-5,'1rO4i.�41�l;),

X>,It�Tiocev,.1eE.:-Ei,.-=.gli^LoI .. ti-ZIEi 3E1 fis

CHAPTER FIFTEEN

oo .-1 CCOV 0C. bb.. .01 ts b. LO

,eII4

II
ice

II
iN

II
lee

.4 c. 0 1
2"i �il
t'l II

-, ...,

II el'l cie: X

0 CV CO 0.1 0 u,

)1O

C, VI) I
C\I

07)

0 --

CD 00 CD 0005CC nC e�CC N

���c--otet, C)Ci-ep�eot..0114000�10

�

CO 00

0 0

ce.

II E

II II II II II II II II II k . k`1 .1.1kikis�U
`.T. I �
e

�

5Le Radio AinaleuA Jiandtooh

Gauge No.

Standard Metal Gauges

American or B. cfb .S.1

U. S. Standard 2

Birmingham or Stubs 3

Effect of Coil Shields on Inductance
It is well known that enclosing a coil in a shield decreases the inductance of the coil. An easily-applied graphical method of determining

.2893

.28125

..300

the extent of the decrease has been worked out

2 3
4 5 8
7 8 9 10 11 12 13 14 15 16
17 18 19 20
21 22 23 24
25 28
27 28 29 30 31 32 33

.2578 .2294 .2043 .1819 .1620 .1443 .1285 .1144 .1019 .09074 .08081 .07196 .06408 .05707 .05082 .04526 .04030 .03589
.03196 .02846 .02535 .02257 .02010 .01790 .01594
.01264 .01126 .01003 .008928 .007950 .007080

.265625 .25 .234375 .21875
.203125 .1875 .171875 .15625 .140625 .125 .109375 .09375
.078125 .0703125 .0625 .05625 .05 .04375 .0375 .034375 .03125 .028125 .025 .021875 .61875 :0171875 .015625 .0140625 .0125
.0109375 .01015625 .009375

.284 .259 .238 .220 .203 .180 .165 .148 .134 .120
.109 .095
.083 .072
.065 .058 .049 .042 .035
.032 .028 .025 .022
.020 .018 .016 .014 .013 .012 .010 .009 .008

by the Radiotron Division of RCA Manufacturing Company and published as a tube application note.'
Considering the shield as asingle turn having low resistance compared to its reactance, the following formula for inductance of the coil within the shield can be worked out:
L = L.(1 -K2)
where L is the desired inductance, L. is the inductance of the coil outside the shield, and K2 is a factor depending upon the geometric dimensions of the coil and shield. Values of K2 have been plotted as afamily of curves in the
chart reproduced on the opposite page. The notations are as follows:
b- length of winding of coil a- radius of coil A - radius of shield
The curves are sufficiently accurate for all practical purposes throughout the range shown when the length of the shield is greater than that of the coil by at least the radius of the coil. If the shield can is square instead of circular, A may be taken as 0.6 times the width of one side. The reduction factor, K2,is plotted against b/2a (ratio of length to diameter of coil), for a

.006350

.00859375

.007

series of values of a/A, the ratio of coil radius

35 36 37
38 39

.005615 .005000 .004453
.003965 .003531

.0078125
.00703125 .006640625
.00625 .....

.005 .004
����

to shield radius (or coil diameter to shield
diameter). The following example will illustrate the use
of the chart. Assume an r.f. coil 11 /2 inches long

.003145

and 3 % inch in diameter to be used in ashield

'Used for aluminum, copper, brass and non-ferrous alloy sheets, wire and rods.
2 Used for iron, steel, nickel and ferrous alloy sheets,

1% inches in diameter. The inductance-reducing effect of the shield is to be calculated. The
values are:

wire and rods. 3 Used for seamless tubes; also by some manufacturers for
copper and brass.

b = 1.5 a = 0.375 A = 0.625

Decimal Equivalents of Fractions

b/2a = 1.5/0.75 = 2 a/A = 0.375/0.625 = 0.6

2 1/16 3/32 1/8 5/32 3/18 7/32 1/4 9/32
5/18 .11/32 � 3/8 13/32
7/16 15/32
1/2

03125 06
.09375 .125
.15625 .1875 .21875
.25 .28125 .3125 .34375 .375 .40625 .4375 .46875 .5

17/32 .. .. .... .... .53125

.......... 5625

19/32 ..... ...... . .59375

5/8 ............625

21/32

.65625

11/18

.6875

23/32 .............71875

3/4. .

.75

25/32 .............781 25 13/18 ..........8125

27/32 ..... ....... .84375

7/8 ...... �.... .875

29/32 ... ...... ... .90625

15/16

.9375

31/32 .. .. .. .. .. . .96875

1...........

1.0

From the curves, K2 is 0.28; the inductance of the coil is therefore reduced 28% by the shield, or conversely, the inductance of the shield coil is 72% of its unshielded value.
The reduction in inductance does not become serious with coils of b/2a ratios of 2 or less, until the shield diameter becomes less than twice the coil diameter. With an a/A ratio of 0.5, the reduction in inductance will be of the
order of 15%.
Application Note No. 48, Copyright, 1935, RCA Manu-
facturing Co..

274 CIHAPTER FIFTEBN

,st

MfifU!9I icre. mr.Line.

1111 0111111:i1iHi igiirnimm

rim ,,.....,,..........,..,, ing
IUlIlIiIIiiiIIU n ,..
111111 111111E11E1 BUBB MI 11111 Igi 11111 111111111 111111111111111 111111 1111 111111inai imMLIImalrlIM
11111i11n1111i1111r111i11u11im11111I111l111u1111n111i1M1M2111a11Bt1E1Iu1Mi1111r11i111p01011101 `1'1111M1111I111M1111I1111Hr11,111I11l1h1u11n0i I1l1l111111111111111I11111111111111111111a1S:i0iLPnlA'thiaittrijena.

o1M 1111 110 III II SIN

1'11111 11111111,1112 1 ill:111 I:MIMI Hillf hlfilhhhIll 111'11111111111 id IILl'il'i'�M 0

4J HD 1 -MUM' IME in I0! 1131 IBEMEIMIECUMMIC]higiffrhiiiM11:BghihniMiliii_iing�P�ten41�-.7-- : .1'' I'1�� Ira MEW1 191 gRl irilllip 1iF,11'ni ll- 3.1=ir '�.e:

utd-le" I� UU
,1an1iaih 11111..m..,,. ...1,1=zzemIlinoln.. .11r:eilmiiuttinfeiF nu'

-1m1i1n1u1mEi1ir.1e1nl1mif1ail1kli1il1n41t-11i1zg1i3il1tu1usEne1e1ie m1an 1n1ue1r1wo1na 1i1l g1g� 1iOgmRutiiflliilnnveeFtUili.riel'.l-a. 1
Znili4-PliIiELpI.1Mg:erz.e.-,-Z1-'2S1f,,1f1ilakiilvdINMIiiIlIlIlMitIpel.lmIed.li17ilii-ibibri!e!!i"P:1 .11.11.11 cihil ia li...k, !up lale

,, M.�
UUU...MMMMIIMI MIUMMI�gulm

-a lCCYI

M 1M I1I1II 1!1!f1g 1V 01W1I 1I1EM 111,111Mn0.1,01I1M111I1V.1.I1MI11g1M1Ii1Mi1IIII1iIILL1iI'.1I..1iI..1rI.1.Mi0..I1.u.1LIn"11IL1uI1M1MoU:1ImMU:M1Mm.IV1Iua1aI1u1:l1Im1l:11:m1ir10ulMi1Lln.�LIf1IilI1iIn'1IClauIM0rI1im1I1.li1I1tr1It1Mi11in111ldi1qm1i1nuf1n1on11im11u11lm11nr11ii11i1,i11.n11l111rl1;1ii11i1tn,1l1:k1n1d,l1u911i111mk1I;11InI111I1l11i1M1l11Ik1I11a1:i111i111l11111:i1111n11i111m0M11i1c11i1i1m1i11k1i.1.11..1,1..1.11I11.1.1..t.1l1u.1.i1m"1.z0..m1i1lr"1amee1ilm1rr11ioe1i2Zv11�x11o1l.2m1nW

e'
-C' IIV

_iiiiiiiiminim 1111111111MIlin 11111111110111 1111111k 111111111111.111110011111111.11011k1 11111 1111110 11 11111puthounni iminammuminummomma Iu111 11111011111 Illk 1111ill111111111111.141111 NV1111111111 1110q111R11111111111111111 MIIMil 11111111111INIIMMUI _

co 91' '', ''', ''''' kn"rrillitilieTe-NEINI.:"1:1111,1 71' HI lieEetahff+..1=`- -%P.

,elm�M

01:1 11 .� �
i lr�11.���11 ,,.

MIIIMM��MM���I�

mAave 1....�MrMMaMM e.lzcz 0

.F ii ii ,, 71,

1,A,iouuiiutn.n.rot.o,amunmr"aon"mlu:rkum9:miv:vw:me:hux1yi1nri0li1:1im{nu1n11i1rpi;t mila�1m11 �i.I1,.

umisretwalelamm

11 11 ,,, 111

41111!F11111111111011�111.1011 1111 1110

11 h

-J-21,3

11 4: '14 i;i 1

III

guinionza "{MUM l11o1m11ffi11v1aEnZItErRKaMmE
!Imo disieniz 11101111111111r1ap 111 1o 1Utw el.sl

PieeMe@

N331.41.4 )13IdVII9

t\D

COPPER WIRE TABLE

Gauge
No. B.& S.

Diane. in
Mast

Circular hid Area

Turns perlanearineh 2

Emma

S.C.C.

D.S.C. or
See.

D.C.C.

TurnsperSquarelneh 2

S.C.C.

emote! S.C.C.

D.C.C.

FeetperM.

Bare

D.C.C.

Ohm
per molt. 26�C.

Current Carrying
Capacity 1600aCi.111.
per Arne

Diam. in nun.

Nearea British S.W.G.
No.

289.3

83690

257.6

66370

229.4

52640

204.3

41740

181.9

33100

182.0

26250

144.3

20820

128.5

16510

7.6

114.4

13090

8.6

101.9

10380

9.6

90.74

8234

10.7

80.81

6530

12.0

71.96

5178

13.5

64.08

4107

15.0

57.07

3257

18.8

50.82

2583

18.9

45.26

2048

21.2

40.30

1624

23.6

35.89

1288

26.4

31.96

1022

29.4

28.46

810.1

33.1

25.35

642.4

37.0

22.57

509.5

41.3

20.10

404.0

46.3

17.90

320.4

51.7

15.94

254.1

58.0

14.20

201.5

64.9

12.64

159.8

72.7

11.26

126.7

81.6

10.03

100.5

90.5

8.928

79.70

101

7.950

63.21

113

7.080

50.13

127

6.305

39.75

143

5.615

31.52

158

5.000

25.00

175

4.453

19.83

198

3.965

15.72 224

3.531

12.47

248

3.145

9.88 282

-----
-----
------
--18.9 21.2
23.6 26.4 29.4 32.7
36.5 40.6 35.3 50.4
55.6 61.5
68.6 74.8 83.3 92.0 101 110 120 132 143 154
166 181 194

----
-----
7.4
8.2 9.3 10.3 11.5 12.8 14.2 15.8 17.9 19.9 22.0 24.4 27.0 29.8 34.1 37.6 41.5 45.6 50.2 55.0 60.2
65.4 71.5 77.5 83.6 90.3 97.0 104 111 118 126 133 140

----
----7.1
7.8 8.9 9.8 10.9 12.0 13.8 14.7 18.4 18.1
19.8 21.8 23.8 26.0 30.0
31.6 35.6
38.6 41.8 45.0 48.5
51.8 55.5
59.2 62.6 66.3
70.0 73.5 77.0 80.3
83.6 86.6 89.7

----
-----
87.5 110 138 170 211 262 321 397 493 592 775 940 1150 1400 1700 2060 2500 3030 3870 4300 5040 5920 7060 8120 9600 10900 12200 -----

------
-84.8
105 131 162 198 250 306 372 454 553 725 895 1070 1300 1570 1910 2300 2780 3350 3900 4860 5280 8260 7360 8310 8700 10700 -----

--------
-80.0 97.5
121 150 183 223 271 329 399 479 625 754 910 Imo 1260 1510 1750 2020 2310 2700 3020 -----------

3.947 4.977 8.276 7.914 9.980 12.58 15.87 20.01
25.23 31.82
40.12 50.59 63.80 80.44 101.4
127.9 161.3 203.4 256.5 323.4 407.8 514.2 648.4
817.7 1031 1300
1639 2067
2607 3287 4145
5227 6591 8310 10480 13210
16660 21010 26500 33410

--
----
----
19.6
24.6 30.9 38.8 48.9 61.5 77.3 97.3 119
150 188 237 298 370 461 584 745 903 1118 1422
1759 2207 2534 2768 3137 4697 egg
6737 7877 9309
10666 11907
14222

.1264 .1593
.2009 .2533 .3195 .4028 .5080 .6405
.8077 1.018 1.284 1.819 2.042 2.575 3.247 4.094
5.183 8.510 8.210 10.35 13.05 16.46 20.76 26.17 33.00 41.62 52.48 66.17 83.44 105.2 132.7 167.3 211.0 268.0 335.0 423.0 533.4
672.6 848.1 1069

55.7 44.1 35.0 27.7 22.0 17.5 13.8 11.0
8.7 6.9 5.5
4.4 3.5 2.7 2.2 1.7 1.3 1.1
.86 .68
.54 .43 .34 .27 .21 .17 .13 .11 .084
.087
.053 .042 .033 .026 .021
.017 .013
.010 .008
.006

7.348
6.544 5.827 5.189 4.821
4.115 3.665 3.264
2.906 2.588 2.305 2.053 1.828 1.828 1.450 1.291 1.150 1.024
.9118
.8118 .7230 .6438
.5733 .5108 .4547 .4049
.3606 .3211 .2859 .2546 .2288 .2019
.1798 .1801 .1426
.1270 .1131 .1007
.0897 .0799

1 3
4 5 7
8 9 10
11 12
13 14 15 18 17 18 18
19 20 21 22 23 24 25 26 27 29 30 31 33 34
36 37 38 38-* 39-4 41 42 43 44

IA mil s1/1000 (one thousandth) of an inch. 2 The figures given are approximate only, since the thickness of the insulation varies with different manufacturers.
The current-carrying capacity at 1000 C.M. per ampere is equal to the circular-mil area (Column 3) divided by 1000.

Mi�ceitaneous 2 0aia

25
15
10 9 8 6
4 3

3 100 90
80 4
TO
5 60

6 SC (1/
1
40 �.)

rzs)

Is (i)

cc
/-Ui 20

o>" 20

L1.1 25

100 90 80 TO 60 50 40
30
20
15

9 35

RELATION BETWEEN INDUCTANCE, CAPACITY AND FREQUENCY
With this chart and astraight-edge any of the above quantities can be determined if the other two are known. For example, if acondenser has aminimum capacity of 15 ��fd. and amaximum capacity of 50 ppfd., and it is to be used with acoil of 10 ph. inductance, what frequency range will be covered? The straight-edge is connected between 10 on the left-hand scale and 15 on the right, giving 13 inc. as the high-frequency limit. Keeping the -traight-edge at 10 on the left-hand scale, the other end is swung to 50 on the right-hand scale, giving alow-frequency limit of 7.1 mc. The tuning range would, therefore, be from 7.1 mc. to 13 mc., or 7100 kc. to 13,000 kc. The center scale also serves to convert frequency to wavelength.
The range of the chart can be extended by multiplying each of the scales by 0.1 or 10. In the example above, if otrhe0.c7aptaoci1t.i3esMcar.e A1l5t0erannatdiv5e0l0y,pp1f.d5. taon5d ptphfed.inadnudct1anpch.e w1i0l0l pghi.v,eth7e1 rtaon1g3e0bMecco.mes approximately 231 to 422 meters

277 CHAPTER FIFTEEN

..7he Radio Arnaleur'� -ilandho�

�USE OF LOG TABLES
The accompanying four-place table will give results which are accurate to three significant figures without interpolation and to four figures with interpolation. The use of log tables has been discussed in �1-6. Only the mantissas are given in the tables; the decimal point (omitted in the tables for convenience) should be placed to the left of the mantissas in every case. The characteristic of any log can be found readily if the number is factored by a power of

ten (� 1-5) wIth the significant factor assigned a value between 1and 10. For example:
258,000,000 = 2.56 X 108 log = 8.4082
0.000000256 = 2.56 X 10-7 log i.4082
If the number has only two digits it will be found in the "N" column and the corresponding logarithm will be beside it in the " "column. For example, the mantissa of any number having only two significant figures, such as 5400, is 0.7324; in this example the characteristic is 3 so the corn-

1 2 3

FOTIB-PLACE

56 78 9

Proportional Farts 1234 56 7 89

110 11 12 13
14

0000 0043 0086 0128 0170 0414 0453 0492 0531 0569 0792 0828 0864 0899 0934 1139 1173 1206 1239 1271 1461 1492 1523 1553 1584

15 1761 1790 1818 1847 1875 18 2041 2068 2095 2122 2148 17 2304 2330 2355 2380 2405 18 2553 2577 2601 2625 2648 /5/ 2788 2810 2833 2856 2878

20 3010 3032 3054 3075 3096 21 3222 3243 3263 3284 3304 22 3424 3444 3464 3483 3502 23 3617 36363655 3674 3692 24 3802 3820 3838 3856 3874

25 3979 3997 4014 4031 4048 28 4150 4166 4183 4200 4216 27 4314 4330 4346 4362 4378 28 4472 4487 4502 4518 4533 29 4624 4639 4654 4669 4683

20 4771 4786 4800 4814 4829 31 4914 4928 4942 4955 4969 32 5051 5065 5079 5092 5105 33 5185 5198 5211 5224 5237 34 5315 5328 5340 5353 5366

35 5441 5453 5465 5478 5490 38 5563 5575 5587 5599 5611 37 5682 5694 5705 5717 5729 38 5798 5809 5821 5832 5843 39 5911 5922 5933 5944 5955
40 6021 6031 6042 6053 6064 41 6128 6138 6149 6160 6170 42 6232 6243 6253 6263 6274 43 6335 6345 6355 6365 6375 44 6435 6444 6454 6464 6474

45 6532 6542 6551 6561 6571 46 6628 6637 6646 8656 6665 47 6721 6730 6739 8749 6758 48 6812 6821 6830 6839 6848 49 6902 6911 6920 8928 6937

50 6990 6998 7007 7016 7024 51 7076 7084 7093 7101 7110 52 7160 7168 7177 7185 7193 53 7243 7251 7259 7287 7275 54 7324 7332 7340 7348 7356

0212 0253 0294 0334 0374 0607 0645 0682 0719 0755 0969 1004 1038 1072 1106 1303 1335 1367 1399 1430 1614 1644 1673 1703 1732

*4 8 12 17 21 25 29 33 37 4 8 11 15 19 23 26 30 34 3 7 10 14 17 21 24 28 31 3 6 10 13 16 19 23 26 29 3 6 9 12 15 18 21 24 27

1903 1931 1959 1987 2014 2175 2201 2227 2253 2279 2430 2455 2480 2504 2529 2672 2695 2718 2742 2765 2900 2923 2945 2967 2989

*3 6 8 11 14 17 20 22 25 3 5 8 11 13 16 18 21 24 2 5 7 10 12 15 17 20 22 2 5 7 9 12 14 16 19 21 2 4 7 9 11 13 16 18 20

3118 3139 3160 3181 3201 3324 3345 3365 3385 3404 3522 3541 3560 3579 3598 3711 3729 3747 3766 3784 3892 3909 3927 3945 3962

2 4 6 811 13 15 17 19 2 4 6 8 10 12 14 16 18 2 4 6 8 10 12 14 15 17 2 4 6 7 9 11 13 15 17 2 4 5 7 9 11 12 14 16

4065 4082 4099 4116 4133 4232 4249 4265 4281 4298 4393 4409 4425 4440 4456 4548 4564 4579 4594 4609 4698 4713 4728 4742 4757

2 3 5 7 9 10 12 14 15 2 3 5 7 8 10 11 13 15 2 3 5 6 8 9 11 13 14 2 3 5 6 8 9 11 12 14 1 3 4 6 7 9 10 12 13

4843 4857 4871 4886 4900 4683 4997 5011 5024 5038 5119 5132 5145 5159 5172 5250 5263 5276 5289 5302 5378 5391 5403 5416 5428

1 3 4 6 7 9 10 11 13 1 3 4 8 7 8 10 11 12 1 3 4 5 7 8 9 11 12 1 3 4 5 6 8 9 10 12 1 3 4 5 6 8 9 10 11

5502 5514 5527 5539 5551 5623 5635 5647 5658 5670 5740 5752 5763 5775 5786 5855 5866 5877 5888 5899 5966 5977 5988 5999 8010

1 2 4 5 6 7 9 10 11 1 2 4 5 6 7 8 10 11 1 2 3 5 6 7 8 9 10 1 2 3 5 8 7 8 9 10 1 2 3 4 5 7 8 9 10

6075 8085 6096 6107 6117 6180 6191 6201 6212 6222 6284 6294 8304 6314 6325 6385 6395 6405 6415 6425 6484 6493 6503 6513 6522

1 2 3 4 5 6 8 9 10 123 4 56 789 1234 56789 1234 56 789 123 4 567 89

6580 6590 6599 8609 6618 6675 6684 6693 6702 6712 6767 6776 6785 6794 8803 6857 6866 6875 6884 6893 6946 6955 6964 6972 6281

1234 56789 1234 56778 123 45 56 78 123 44 56 78 123 4 45678

7033 7042 7050 7059 7067 7118 7128 7135 7143 7152 7202 7210 7218 7226 7235 7284 7292 7300 7308 7316
7364 7372 7380 7388 7396

123 34 5678 1 2 3 3 4 5 '6 7 8 122 3 4 56 77 12 234 56 87 1223 45867

5 6 7 8 9

*Interpolation in this section of the table is inaccurate.
278 CHAPTER FIFTEEN

1234 56789

plete log is 3.7324. For numbers having three significant figures, locate the first two in the "N" column and move to the right, reading the mantissa under the appropriate digit; e.g., log 5430 3.7348, the characteristic being found as before. For numbers having four significant figures, proceed as before and add to the mantissa the figure given on the same line in the "Proportional Parts" column under the appropriate digit. For example, log 5434 =- 3.7351.
In determining numbers whose logarithms are given (such
numbers are called anabgarithms and are written log-I) the procedure given above is reversed. For example, given 1.9304 as the logarithm of anumber to be found, inspection of the

7llacebeaneou� 2Itia
table shows that the mantissa 0.9304 corresponds to the number 852; the characteristic is 1so there are two integral places, hence the number is 85.2. If the given log had been 1.9306, there is no exactly corresponding mantissa in the table, but the next lower one is 0.9304. The difference is 2 (0.0002, actually), and on referring to the columns of proportional parts on the same line it is found that adifference of two may indicate either 3or 4as the fourth digit in the antilogarithm. Hence the antilogarithm of 1.9306 is either 85.23 or 85.24. This is the maximum uncertainty which will be encountered with four-place tables; in moat cases the fourth figure will have no uncertainty.

LOGARITHMS

0 1234

5 6 7 8 9

Proportional Parta 123 4 56 789

55 7404 7412 7419 7427 7435

56 7482 7490 7497 7505 7513

57 7559 7566 7574 7582 7589

58 59

7634 7642 7649 7657 7664 7709 7716 7723 7731 7738

SO 7782 7789 7796 7803 7810 61 7853 7860 7868 7875 7882 62 7924 7931 7938 7945 7952 63 7993 8000 8007 8014 8021 64 8062 8069 8075 8082 8089

65 8129 8136 8142 8149 8156 66 8195 8202 8209 8215 8222 87 8261 8267 8274 8280 8287 68 8325 8331 8338 8344 8351 69 8388 8395 8401 8407 8414

76 8451 8457 8483 8470 8476 71 8513 8519 8525 8531 8537 72 8573 8579 8585 8591 8597 73 8633 8639 8645 8651 8657 74 8692 8698 8704 8710 8716

75 8751 8756 8762 8788 8774 78 8808 8814 8820 8825 8831 77 8865 8871 8876 8882 8887 78 8921 8927 8932 8938 8943 79 8976 8982 8987 8993 8998

86 9031 9036 9042 9047 9053 81 9085 9090 9096 9101 9106 82 9138 9143 9149 9154 9159 83 9191 9196 9201 9206 9212 84 9243 9248 9253 9258 9263

85 9294 9299 9304 9309 9315 86 9345 9350 9355 9360 9365 87 9395 9400 9405 9410 9415 88 9445 9450 9455 9460 9465 89 9494 9499 9504 9509 9513

40 9542 9547 9552 9557 9562 91 9590 9595 9600 9605 9609 92 9638 9643 9647 9652 9657 93 9685 9689 9694 9699 9703 94 9731 9738 9741 9745 9750

95 9777 9782 9786 9791 9795 96 9823 9827 9832 9836 9841 97 9868 9872 9877 9881 9886 98 9912 9917 9921 9926 9930 99 9956 9961 9965 9969 9974

7443 7451 7459 7466 7474 7520 7528 7536 7543 7551 7597 7604 7612 7619 7627 7672 7679 7686 7694 7701 7745 7752 7760 7767 7774

12 234 5 56 12 23 45 567 122 34 55 67 1 123 44 56 1123 4 4 567

7818 7825 7832 7839 7846 7889 7896 7903 7910 7917 7959 7966 7973 7980 7987
8028 8035 8041 8048 8055 8096 8102 8109 8116 8122

112 34 4 566 1 123 4 456 6 112 3 34 566
1123 34 5 56 1123 3 4 556

8162 8169 8176 8182 8189 8228 8235 8241 8248 8254
8293 8299 8306 8312 8319 8357 8363 8370 8376 8382 8420 8426 8432 8439 8445

11233 4 556 1123 3 45 56 11233 4 556
1123 3 44 56 112 23 44 58

8482 8488 8494 8500 8506 8543 8549 8555 8561 8567
8603 8609 8615 8621 8627
8663 8669 8675 8681 8686 8722 8727 8733 8739 8745

112 23 44 56
1 12 234 45 5 1 1223 4 455 112 23 44 55 1 122 3 4455

8779 8785 8791 8797 8802
8837 8842 8848 8854 8859 8893 8899 8904 8910 8915
8949 8954 8960 8965'8971 9004 9009 9015 9020 9025

112 23 3 455 1 122 33 4 55 112 23 3 445 I 1 2 �2 3 3 4 4 5
1 1223 3 445

9058 9063 9069 9074 9079 9112 9117 9122 9128 9133 9165 9170 9175 9180 9186
9217 9222 9227 9232 9238 9269 9274 9279 9284 9289

112 23 34 45 1122 334 4 5 112 233 4 45 1 12 233 4 45 112 23 3 445

9320 9325 9330 9335 9340 9370 9375 9380 9385 9390
9420 9425 9430 9435 9440
9469 9474 9479 9484 9489 9518 9523 9528 9533 9538

1 122 33 445
112 23 3 445 O 112 23 3 44 O 1 12 23 344
O 112 23 3 44

9566 9571 9576 9581 9586 9614 9619 9624 9628 9633 9661 9666 9671 9675 9680 9708 9713 9717 9722 9727 9754 9759 9763 9768 9773

O 11223344
O 1 12 23 3 44
O 11223344 O 11223344 O 1 12 23 34 4

9800 9805 9809 9814 9818 9845 9850 9854 9859 9863 9890 9894 9899 9903 9908 9934 9939 9943 9948 9952
9978 9983 9987 9991 9996

O 11223344 O 1 12 233 44 O 11223344 O 1 12 23 344 O 11223334

0 123 4

56 78 9

12 3 4 56 7 89

279 CHAPTER FIFTEEN

GOOD BOOKS

Every amateur should maintain a carefully selected bookshelf; a few good books, consistently read and consulted, will add immeasurably to the interest and knowledge of the owner. We suggest aselection among the following works, all of which have been carefully chosen and are recommended in their various fields.
Fundamentals:
Audel's New Radioman's Guide, by E. P. Anderson. For one who wants to get a working knowledge of radio. Not an engineering text, but filled with useful information for the practicing radioman. 765 pages, 519 illustrations. 1940. Audel, $4.00.
Fundamentals of Radio, by F. E. Terman. An elementary version of the author's "Radio Engineering" with simplified treatment intended for readers of limited mathematical ability. 458 pages, 278 illustrations. 1938. McGraw-Hill,
$3.75. Getting Acquainted With Radio, by Alfred Morgan. Gives
the neophyte in radio the basic principles of the science in popular language. 285 pages, 130 illustrations. 1940. Century. $2.50.
Principles of Radio, by Keith Henney. This popular book covers the range from the fundamentals of electricity to modern concepts of modulation and detection. 495 pages, 311 illustrations. 3rd edition, 1938. McGraw-Hill, $3.50.
Theory and Engineering:
Electrical Communication, by Arthur L. Albert. General treatment of the whole field of electrical communications, both wire (telegraphy and telephony) and wireless (radio). 534 pages, 397 illustrations. 2nd edition, 1940. Wiley, $5.00.
Electrical Engineers Handbook: Communication and Electronics, by Pender and Mcllwain. Engineering reference book covering all phases of communication and electronics. 1022 pages, 981 illustrations. 1936. Wiley, $5.00.
Radio Engineering, by F. E. Terman. A comprehensive treatment covering all phases of radio. The recognised authority in its field. 813 pages, 475 illustrations. 2nd edition, 1937. McGraw-Hill, $5.50.
Radio Engineering Handbook, Keith Henney, Editor. An authoritative handbook for radio engineers, with technical data on all aspects of radio. 948 pages, 837 illustrations. 1941. McGraw-Hill, $5.00.
Principles of Electron Tubes, by H. J. Reich. The theory, characteristics and applications of electron tubes and their circuits. Includes data not elsewhere available. 397 pages, illustrated. 1941. McGraw-Hill, $3.50.
Experiments and Measurements:
Measurements in Radio Engineering, by F. E. Terman. A comprehensive discussion of measurement problems encountered in engineering practice, with emphasis on basic principles. 400 pages, 208 illustrations. 1935. McGraw-Hill, $4.00.

Radio Frequency Electrical Measurements, by H. A. Brown. A laboratory course in r.f. measurements for communications students. Contains practical information on methods. 384 pages, 177 illustrations. hid edition, 1938. McGraw-Hill, $4.00.
The Cathode-Ray Tube at Work, by John F. Rider. Cathodray tube theory, sweep circuita, a.c. wave patterns and oscilloscopes, including actual photographs of screen patterns. 322 pages, 444 illustrations. 1935. Rider, $2.50.
Commercial Operating and Equipment:
Radio Operating Questions and Answers, .by Nilson and Hornung. Gives answers to questions in the FCC study guide covering all six elements of the commercial examinations. 415 pages, 87 illustrations. 7th edition, 1940. McGrawHill, $2.50.
The Radio Manual, by G. E. Sterling. A practical handbook, especially valuable to the commercial and broadcast operator, covering principles, methods and apparatus. Illustrated. 1120 pages, 1938. D. Van Nostrand, $6.00.
Books Dealing with Specialized Radio Topics:
Aeronautic Radio, by Myron F. Eddy. Supplies the information needed by students, pilots, mechanics, operators and executives. Prepares for aviation radio license exams. 502 pages, 198 illustrations. 1939. Ronald, $4.50.
Frequency Modulation, by John F. Rider. Practical exposition of the subject from the serviceman's standpoint, but including worthwhile theoretical background. 136 pages, illustrated. 1940. Rider, $1.00.
Mathematics for Electricians and Radiomen, by Nelson M. Cooke. Furnishes the student with a sound mathematical foundation and shows how to apply this knowledge to practical problems. 604 pages, illustrated. 1942. McGrawHill, $4.00.
Principles of Television Engineering, by D. G. Fink. Covers the television system from the camera to the viewing screen in the receiver, with descriptions of modern equipment. 541 pages, 313 illustrations. 1940. McGrawHill, $5.00.
Radio as a Career, by J. L. Hornung. A realistic discussion of the opportunities to be found in the various radio fields and the relationship of the radio amateur to these fields. 212 pages. 1940. Funk & Wagnalls, $1.50.
Servicing Superheterodynes, by John F. Rider. Theory and practice of superheterodynes, with adjustment and troubleshooting data. 278 pages. Rider, ELM
Publishers addresses:
D. Appleton-Century Co., 35 West 32nd St., New York. D. Van Nostrand Co., 250 Fourth Ave., New York. Funk & Wagnalla Co.. 354 Fourth Ave., New York. John F. Rider, 404 Fourth Ave., New York. McGraw-Hill Book Co., 330 West 42nd St., New York. Ronald Press, 15 East 26th St., New York. Theo. F. Audel & Co., 49 West 23rd St., New York. John F. Wiley & Sons, 440 Fourth Ave.. New York.

280 CHAPTER FIFTEEN

index

A Battery � �

.. .. .. ...

PAG4E2

2830 A.C.-D.C. Power Supplies. .. .... .... ....... ...... 152 AT-Cut Crystals. .. .. .. .... ..... .............. 61

Abbreviations, Radio and Electrical... .... .. .... .. 272

Abscissae

Absorption .. ........ .... .............. ... ...... 157

Absorption Frequency Meters

22..5-227

Acorn Tubes ....... ...... ..... ..... ....... ..... 121

Adjacent-Channel Selectivity.. ....... .. .. .. .. .. 112 Air-Gap Crystal Holder .. .......... .. .. ..........61-62

Alignment, Receiver

.137-138, 141-142, 243-244

Alternating Current ... .. ....... ...... ...... ....28-30

Alternating Current, Ohm's Law for. .. .. .. .. .. .. ..30-31

Alternating Current Spectrum.

Ammeters Ampere.

23, 32, 233, 235 23

Ampere Turns

Amplication Factor

Amplification, Vacuum Tube Amplifier Classifications

.43, 44-46 46-48

Amplifiers (see "Receivers," "Transmitters" and "Radiotelephony")

Amplitude Modulation

Analyzers, Wave242-243

Angle, Amplifier Operating.. ..................71, 79-80

Angle of Radiation .. .. ... ................153, 162, 164

Angular Velocity

5node

Anode-Dissipation Power Measurement. ..... .. .... 246

Antenna Systems:

Angle of Radiation .. .. .. ..... ... .. .. -158, 162, 164

Beam Antennas .... ... .................. _184-194

Bent Antennaq

184

Broadside Arrays

.. 187, 188

Coaxial Vertical Radiator

-193-194

Collinear Arrays

........ ........ .....187-188

Compact Antennas. .. ................. -183-184

Conductor Size

163

"Corner" Reflector Antennas

194

Coupling to Receiver .

116

Coupling to Transmitter ... .. ........ .. .. ..169-172

Current Distribution

163, 179

Directive Antennas Directivity.

18.'4-194 162

Director .

190

Doublet

162

Dunund Antennas

.. ..... ....... 76

End End-Fire Arrays188-189

162 .1ffi,

Feed Systems:

Balance to Ground .

167

Characteristic Impedance ........... . . 167

Concentric Lines.

1611, 1i6-177

Current .166

Delta Matching Transformer... .. .. .... .. .. 177

Harmonic Resonance.... ... ........ .....

Height Impedance

... ........... .........1fi3-164 40, 167, 168

Length

167-168, 189

Losses

.168-169

Linear Transformers Non-Resonant Lines

..168, 178-179 .....169, 175-179

Open-Wire Line "Q" Antenna

.166. 172 177-178

Reactance

168

Resistance

168

Resonant Lines... ..................169, 172-175

Single-Wire Feed

.166, 175

Standing-Wave Ratio

168

Stub Matching ..... ....................178-179

Tuned Transmission Lines

.172-175

UTwnitsutneedd-PTariarnLsimnies.si..on..L.i.n.e.s.�1�7�1�-�1�7�9���166, 175- 176

Volta

166

Folded Dipole ..... ............... .. .. ..... 194

Ground Effect .

.18;4-165

Grounded Antennas... ..... ...... . . ..... 193

Hertz Antenna

162

Impedance

.162, 163, 164, 181

J Antenna

162

Length

162-163, 174-175, 179, 191, 193

Long-Wire Antennas

179-182, 184-186

Marconi (Grounded) Antennas Measurements . Multi-Band Antennas

PAGE
193, .246-248
182-183

Parasitic Elements Patterns. Radiation ...... ..............

190-193 165

Phased

186-190

Polarization ... ...... ........... .. iM, 162, 164-165

Power Gain

.162, 181

Radiation Characteristics163

Radiation Resistance.

Receiving..

40 .194-195

Reflection, Ground .. .................... -1133-164

Reflector

190

Rhombic or Diamond Antennas

.185-186

Rotary-Beam Antennas.

191

Standing Waves ....... ... .. .......... .. ..

Transmission Lines ...................... _166-169

V Antennas

..I84-185

Voltage Distribution

.163, 179

Anti-Nodes

Arc-Back ....... ...... ............ ...........

144

Arrays, Antenna .

186-195

Atoms ... .. ... ..... ... ... ..... ..... .... .. .. 21

Atmospheric ........

160, 161

Attraction ........ ...... .......... ............. 22

Audio Frequencies

....... ...... ... _ 30

Audio Frequency Measurement .. ....... ...... -232-233

Audio Image ...... ... .................... ...... 129

Audio Limiting ..... ... .. .. .. .. .. .. .....97, 134

Audio Oscillators ......... ........245, 26n-261, 2E14-265

Autodyne Reception

111

Automatic Volume Control. ....... .............131-133

Auto-Transformer Average Current Value Axes, Cathode-Ray Tube

33 55-5269

B Battery ......... ... ............. ........

Back-Wave

106

Baffle Shield

Balanced Transmission Line ..................... 167

Balancing Circuits, Meter.

236

Band Changing..

....... ......

. 121

Band-Pass Filters

Band Spread

.121-122

Band Width

112

Bands, Antenna Lengths for Base (of Logarithm)

176, 183, 184, 195 15

Batteries

Beam Antennas (see "Antenna Systems")

Beam Tubes ......... ..... .... .. ... ��.......... 49

Beat Frequencies .. ... _.. .. ....................40-41

Beat Frequency Audio Oscillator

243, 245

Beat Note

Ill

Beat Oscillator

124, 131

Bent Antennas

184

Bias Calculation Bias Modulation.

.50, 71-72 .88-89

Bias Supplies.. .. ................ ...... .... _ .150-152 Birdies. ... ......... .......... .......... .124, 138-139

Bleeder ..

145

Blocked-Grid Keying.. ....... ...... ............. 108

Blocking.

118

Blocking Condensers ....... ........ .....

Body Capacity

.67, 116

Books .280

Brackets .

.11-12

Break-In

107

Bridge Circuits

.38-39, 232, 239-240

Bridge Rectifiera .

145

Broadside Arrays Buffer Amplifier .

.188 58

Buffer Condenser

154

Bus Wire

252

Button, Microphone.

Buzzer Code Practice Set

260

By-Pass Condensers .. .. ... ....... ..........41, 119-120

.35,53,72-73

C. W. Recepiion.. .... ...... :... -11.1, 114-117, 136-137

Calorimeter Method of Power Measurement ...... .. 246

Cancelling Out Terms ...... . .. .. .. .. . ... ..... 12 Capacitive Coupling. .. .. . .. .. .. .... ........ ..36, 66

Capacitive Reactance

281

Jnclex

PAGE

Capacity.

. .

22-23

Capacity, Distributed

31-32

Capacity, Inductance and Frequency Chart

277

Capacity, Interelectrode

Capacity Measurement ..... ... .. ..........239-240, 241

Capacity of Condenser, Computing

268

Capacity-Resistance Time Constant .. .............27-28

Carbon Microphones

Carrier

58,84

Cascade Amplifiers.

Cathode

.23, 24, 42, 49-50

Cathode Bias.

Cathode By-Pass Cathode Keying

50 .108-109

Cathode Modulation

86, 89-90

Cathode-Ray Oscilloscopes. .... ... 55-57, 98-101, 236-238

Code Learning: Audio Oscillators Buzzer Practice Set Cipher Groups Code Table High-Speed Operation Practice Cipher Groups Practice Words Receiving Sending Sounds Tone Sources
Code Table Code, Q Code Practice Seta Coefficient Coefficient of Coupling

PAGE
260-261, 264-265 260 263
261-265 259-260
263 262 254-257 257-259 254-257 260-265 261-265 268-267 260-265
11 36

Cell ........_. ,... ........... ...... ............. Center-Tap, Filament

5240

Center-Tap Full-Wave Rectifier

.144-145

Center-Tap Keying

108-109

Center-Tap Modulation

.89-91

Channel

.85-86

Characteristic Curves

Characteristic Impedance

167

Characteristic (of Logarithm) ......... ........... 16

Charges, Electrical... ....... ....................21-23 Charts and Tables:
Abbreviations for Electrical and Radio Terms... 272 Antenna and Feeder Lengths ... ....176, 183, 184, 195

Antenna Gain ..... .. .. ... .. .. .. .. .. ..187, 188, 190

CBaatnhd-oWdiedMtohd,uTlyaptiicoanl PIe. rFformance Curves.. .. .. 19209

Characteristic Impedance .. - ..............167, 180 Coil Shields, Effect on Inductance ..... .....274-275

Coil-Winding Data. ..

. 74

Color Code for Radio l'aria. :.:.:.. .. ::: :: :: :: :. 269

CCoonntdiuncetnitvailtyCoofdeMetals ...

.. ��...... ...2.5. 5-225770

Conversion Factors

271

Current Capacity of Power Wiring ... ...... ... 271

Decibel Chart

272

Decimal Equivalents

274

Dielectric Constants

268-269

Drill Sizes ... ... ... ........... ��.. .. .. .. ... 251

Electrical Symbols Gain of Directive Antennas

270 187, 188, 190

Gauges, Metal

274

Greek Alphabet

271

Inductance, Capacity and Frequency Chart

277

LC Constants

270

Length, Unite of

270

Line-of-Sight U. H. F. Range

160

Log Tables

278-279

Metric Prefixes

271

Multiples and Sub-Multiples

271

Numerical Values

273

Q Code

268-267

Radiation Angles Radiation Patterns Radiation Resistance

164, 165 182
164, 181

Reactance Chart

273

Resistance-Coupled Amplifier Data

Rhombic Antenna Design Schematic Symbols

186 Frontispiece

Surge Impedance Tank Circuit Capacity

167, 180 73

Tap Sizes

249

Tolerances (Color Code)

269

Tools

249

V-Antenna Design Chart

185

Vacuum Tube Symbols

270

Voltage Ratings (Color Code)

269

Wire Table

269

Chassis Layout

250

Chirp, Keying

107, 110

Choke Coil

Choke-Coupled Modulator

87-88

Choke, Filter Choke-Input Filter

147 145, 146-148

Cipher Groups Circuit Diagram Symbols

263 Frontispiece

Circuit, Electric

Circuits, Receiving (see "Receivers")

Circuits, Transmitting (see "Transmitters")

Circuits, Tuned Class A, AB, B, and C Amplifiers Class B Modulators

33-35 46-48 91-92, 219-220

Coefficient, Temperature

Coil Shields, Effect on Inductance

274

Coils (see "Inductance")

Coils, Winding

253

Collecting Terms

Collinear Arrays

187-188

Color Codes, RMA

269

Colpitts Circuit Combined A.C. and D.C. Complex Waves Concentric Line Circuits

51
40
40 40, 81-82

Concentric Transmission Line

166, 176-177

Condenser

22-23

Condenser Capacitance (Computing)

268

Condenser Color Code Condenser-Input Filter

269 145-146

Condenser Microphones

Condenser Series and Parallel Connections

Condenser Reactance Condensers, Electrolytic Condensers, Testing

31 145 239-240

Condensers, Voltage Rating Conductance Conduction Conduction, Electrolytic

74, 145 43 23 24

Conduction, Gaseous

Conduction, Thermionic

Conductivity of Metals

270

Conductors

Constant, Time

27-28

Continental Code

255-257

Control Grid

Controlled-Rectifier Keying

108

Conversion Efficiency Conversion Factors Converters, Frequency

125 271 124-126

Coordinates

Copper-Oxide Rectifier Meter

234

Core

Corner Reflector Antenna

194

Coupled Circuits Coupling

35-39 35-37

Coupling, Antenna to Receiver Coupling, Antenna to Transmitter

116 169-172

Coupling Condenser

Coupling, Interstage

Crackle 'Finish

252

Critical Angle

159

Critical Coupling Critical Frequency

38-37 157

Critical Inductance

147

Cross-Modulation

Crystal Detector

225

CCrryyssttaall, FFialitlerusre to Oscillate Crystal Microphones Crystal Oscillators Crystal-Controlled Transmitter

129-131, 136-16337 93-94
52, 62-65 Construction (see

"Transmitters" and "Ultra-High Frequencies")

Crystals, Piezo-Electric

35, 61-62

Current, Alternating Current Capacity of Wiring (Table)

28-30 271

Current, Direct

23-24

Current Feed for Antennas Current Flow

166 21, 23-24

Current in A.C. Circuits

CCuurrrreenntt MLeaagsaunrdemLeenatd

28, 32, 233-23305

Cut-Off Frequency

Cut-Off, Plate Current

Cycle

Clicks, Keying

106-107, 110

CClooasxeidalCiLricnueit Coaxial Vertical Radiators

166, 176-12717 193-194

D'Arsonval Movement Damping

233 34

282

index

Dead Spots Decibel Decimals Decoupling Decrement Deflection Plates Deflection Sensitivity

PAGE
116 18, 46
7-9 96 34 5544

Degeneration

45-46

Delayed A. V. C

132

Delta Matching Transformer

177

Demodulation Detection Deviation Ratio Deviation-Reading Frequency Meter Diagrams, Schematic Symbols for DDiiealemcotnridc Antenna Dielectric Constant

111 Ill, 112-118, 131
103 Frontispi2e3c3e
18252-,1228326

Dielectric Constants (Table)

268-269

Dies Diffraction Diode Detectors Diode Voltmeters Diodes Dipole, Folded Direct Coupling Direct Current

251 156, 160-161
113 235-236
42 194 3236

Direct Feed for Antennae

166

Direct-Ray Transmission Direction of Current Flow

159-160 24

Directional Antennas (see "Antenna Systems") Director, Antenna Disassociation Discriminator Dissipation, Plate Distortion DDiivsitdreirbsu,teVdolItnadguectance and Capacitance
Doubler. Frequency Double Resonance

140-11429130
44 44 31-32 58, 79-2870 77

Double Superheterodyne

124

Doublet Antenna

162

DDDrroiiwfltln,wSiFazrreesdq(uMTeoandbcluyel)ation DDrriilvleinngEHloelmeesnts

101-102
53, 61, 225218 250-251

Driver DDrroivpi,nVgoPlotawgeer
Dry Cell

66,19817 45, 72,279,1-1924423

Dummy Antennas

76 243

Duplex-Diode Triodes and Pentodes

Duplex Power Supplies

153

Dynamic Characteristics

Dynamic Instability Dynamic Microphones

52-53 94

Dynamotors

222

EEE-dLdayyeCrurrent Effective Current Value Efficiency, Amplifier Efficiency, Transformer Electric. Charges Electric Circuit Electrical Length EElleeccttrriiccailtyUnits and Symbols
Electricity, Static Electrolytic Condensers Electrolytic Condensers, Testing Electrolytic Conduction Electromagnetic Deflection Electromagnetic Radiation Electromagnetism Electromotive Force (E.M.F.) Electrons EElleeccttrroonn-FCloouwpled Oscillator
Electron Gun Electron-Ray Tubes Electronic Conduction Electrostatic Coupling Electrostatic Deflection Electrostatic Field Electrostatic Shield Element, Antenna Emission, Electron End Effect End-Fire Arrays

26 158, 1236917
71 32 21-23 21 167-168
2702,1-2227132 45
239-240 23-24 54 40 24-26 22 21
59-60, 526413 132 24 36 2524 76
186-187 24
162 187, 188-189

Envelope, Modulation Excitation Exciter Exciter Units (see "Transmitters") Exponential Functions Exponente

PAG8E4 44, 45, 52, 74-75
58
9, 1147--1158

F-Layers Factor Fadeouts

158 1598

Fading

159

Farad

Faraday Shield

Feed, Series and Parallel

Feed-Through Insulators

252

Feedback

45-46, 51-52

Feeders and Feed Systems (see "Antenna Systems")

Fidelity

112

Fidelity, Measurement of

243

FFiieelldd, IEnlteecntsriotsytaMteitcers

247-22428

Field, Magnetic Field, Strength Field-Strength Measuring Sets

24-26 156 246

FFiillaammeenntt Supply

24, 49-15403

Filter, Crystal FFiilltteerrss, R.F Final Amplifier First Detector

129-131 38, 145-148
124-11250689

Five Meters (see "Tjltra High Frequencies")

Flat Response

Flexible Couplings

253

Flow, Current

Fluorescent Screen

Flux Density, Magnetic FFllyyw-hBeaeclk Effect

5266
34

Focussing Electrode

FFoorlcdee,d EDliepctolreomotive

194 22

FFoorrcmeu,laLsi:nes of

22, 24-25

A.C. Average, Effective and Peak Values

Amplification Factor

Antenna Impedance Measurement

247

Antenna Length

162, 163, 179, 193

Antenna Power Measuremont

247

Antenna Resistance Measurement

247

Bias

Bias Supply Bleeder

150

Bridge Balance Capacitive Reactance

38, 39
31

Capacity of Condenser

268

Cathode Bias

Characteristic Impedance

167

Coefficient of Coupling

Combined A.C. and D.0

Complex Wave

Coupling Transformer Turns Ratio Critical Inductance

91, 92
147

CDreycsitbaell Frequency-Thickness Ratio Delta Matching Transformer Design

61
18, 46
177

Filter Design

147-148

GFrriedquIemnpceydance Grid Leak Bias Impedance

33
a371o2

Impedance Matching Impedance Ratios

33, 86, 168, 178 32

Impedance Transformation

168

Inductance, Apparent

240

Inductance Calculation

268

Inductance, True

241

Inductive Reactance

Input Filter Choke

147

Interpolation, Frequency

231

Lecher Wires

227

Linear Matching Sections Modulation Impedance Modulation Percentage

168 84, 9896

Modulator Transformers, Turna Ratio Multipliers (Meter) Mutual Conductance

91, 92 233 43

Ohmmeter (from voltmeter)

238

Ohm's Law (A.C)

Ohm's Law (D.C.)

Parallel Impedance

Piano Frequencies

232

Plate Resistance

283

index

PAGE

Power

Power Factor

Power Supply Output Voltage

147

PQower Supply Transformer Voltage

35, 241, 214428

Q Antenna

178

R/C Time Constant

Reactance

Regulation

143

Resonance

Resonant Impedance

Rhombic Antenna

186

Ripple

146, 147

Series, Parallel and Series-Parallel Capacities.... 27

Series, Parallel and Series-Parallel Inductance... 27

Series, Parallel and Series-Parallel Resistances. .. 27

Shunts, Meter

233

Standing-Wave Ratio Surge Impedance Time Constant

168 -167 27-28

Transformer Volt-Ampere Rating Transmission Line Length

148 168, 178, 189

Transmission Line Spacing Turns Ratio U.H.F. Range Voltage Dividers

167 32 160
27, 152

Voltage Regulation

143

Voltage Regulator Limiting Resistor

149

Wavelength

Wavelength-Frequency Conversion

Wien Bridge

232

Formulas, Explanation Frequency

10-14 28-29, 39

Frequency Converters

124-126

Frequency Deviation

103

Frequency Distertion Frequency Divider Frequency Drift

44 53 53, 61, 228

Frequency, Inductance and Capacity Chart

277

Frequency Measurement: Absorption Frequency Meters

225-227

Audio Frequencies Frequency Standards Heterodyne Frequency Meters

232-233 229-231 227-229

Interpolation

231

Lecher Wires

227

Monitoring

233

Zero-Beat Indicators Frequency Modulation
Deviation Ratio Discriminator Limiter

231-232 84
103 140-141 139-140

Methods

104-105

Principles

102-104

Reactance Modulator

104

Reception

139-142

Frequency Multipliers

58, 79-80

Frequency Response Frequency Spectrum

44, 93 30

Frequency Stability Frequency Standards Frequency-Wavelength Conversion

52-53, 107, 228 229-231 39

Full-Wave Rectifiers Functions (Mathematical)

144-145 16-18

Fundamental Frequency

29, 39

Fundamental Radio System

Gain Control Galvanometers Ganged Tuning Gaseous Conduction Gaseous Regulator Tubes Genemotors General Coverage Tuning Graphs Grid Grid Bias Grid-Bias Modulation Grid-Control Rectifiers Grid Current Grid Driving Power Grid Emission Grid Excitation GGrriidd KIemypiendgance Grid Leak Grid Leak Detectors Grid Neutralization Grid-Plate Oscillator Grid-Plate Transconductance Grid Swing Ground Effect

96-97, 120 28 122 23
148-149 222 122
18-20 42
43, 44, 50, 52, 71-72 86, 88-89 49, 108 43 45 77 44 45 17028 50 113-114 67, 68 64 43 46 163-165

Ground Potential Ground Wave Grounded Antennas Grounds Gun, Electron

PAGE
41 156-157
193 41 54

Hairpin Coupling Loop

Half-Wave Half-Wave Antenna

39 162-163

Half-Wave Rectifiers

144

Hand Capacity Harmonic Distortion Harmonic Frequencies

67 44 29, 39

Harmonic Generation

79-80

Harmonic Operation of Antennas Harmonic Resonance

39, 179 39

Harmonic Suppression

76-77, 172

Hartley Circuit

51, 59

Hash

154

Headphone Impedance

118

Heater

Heating Effect

Henry

Hertz Antenna

162

Heterodyne Frequency Meters

227-229

Heterodyne Reception

111

High-Frequency (Oscillator (Receiver).... 123-124, 126-127

High-Pass Filters High Speed Code Operation

38 259-260

High-Vacuum Rectifiers

143

Hiss Noise

120, 134

Holes, Drilling and Cutting

250-251

Hook-Up Wire Hum

252 98, 102, 116

Image, A.F. Image, R.F Image Ratio

129 124 124, 133

Impedance Impedance, Antenna Impedance, Grid Impedance Matching Impedance Ratio, Transformer Impedance, Tank

31, 34-35 162, 163. 164, 181
72 33, 168
32-33 60, 71, 72-73

Impulse Noise Index (of a Root) Indicators, Tuning

134 9
132-133

Inductance Inductance Calculation IInndduuccttaannccee,, cDaipsatcriitbyutaendd Frequency Chart

25 268 312-7372

Inductance Measurement Inductance Winding Charts

239-241 74

Induction Inductive Coupling Inductive Neutralization Inductive Reactance Infinite Impedance Detector Input Capacity, Tube Input Choke

2356 68-69
31 114-115 ' 43, 45
147

Input Resistance, Tube Instability

121 52-53

Instantaneous Plate Current

144

InstrAu.m0ents:

32, 234-235

Ammeters Audio Oscillators Capacity Bridges D.0

28, 233-235 245
39, 239-240 28, 233-234

Field-Intensity Meters

247-248

Frequency Meters Frequency Standards

225-229 229-231

I.F. Teat Oscillator

244-245

Inductance Bridge L, C and Q Meter

239 240-242

Multi-Range V-O-M Ohmmeters Oscilloscopes

234 238-239 55-57, 98-101, 236-238

Reactance Meter Resistance Bridges Vacuum-Tube Voltmeters

239 239 235-236

Voltmeters Wave Analyzers

28, 233-234 242-243

Wheatstone Bridge

230

Wien Bridge

232

Insulators Integer Interelectrode Capacitances Intermediate Frequency Intermediate Frequency Amplifiers Intermediate Frequency Test Oscillator

21 8
43 123, 127-128 111, 124, 127-131
244-245

284

index

Interpolation (Frequency Measurement) Interpolation (Graphs) Interruption Frequency Interstage Coupling Interstage Transformers. II' Inverse Peak Voltage Inversion, Temperature Inverted Amplifier Inverter, Phase Ionisation Ionosphere Ions Iron-Vane Meters

PAGE
231 20
11 7-118 66
128-129 144 161 71 97 23
157-158 23243

Brackets Cancelling Out Characteristic Coefficient Collecting Terms Coordinates Decibels Decimals Exponential Functions Exponents Formulas GFurnacpthisons

J Antenna

Index

193

Interpolation

1P1A-G1E2
12 16 11 12 18 18 7-9
17- 18 9 14 -15
10-14
1186--21908
20

Key Chirps Key Clicks Key, flow to Use Keying:

107,110 106-107,110
257-258

Log Paper Logarithmic Functions Logarithms Mantissa

19 17-18
15-16
16

Back Wave Break-In Key-Click Reduction Methods:
Cathode or Center-Tap Blocked Grid Grid-Controlled Rectifier Plate Primary Screen Grid Waveform Suppressor Grid Oscillator Keying Parasitic Clicks Kilocycle Kilohm Kilovolt Kilowatt Kilowatt-Hour

106

ONrodtiantaitoens

107

,Plotting Graphs

109

�Polar Cofirdinates

.Power Function

110088

,.PRaodwiecrasl Sign

108

Roots

107

Scales

108

Signs

107

Square Root

106

Transposition

108

Variables

110 Maximum Current Value

110 Measurements:

2296

CAanptaecnintay

2266

CFurrerqeunetncy

Inductance

7,10 18 18 20 17
9,11,15 9
9-10,11,15 18-20 10-11 9-10 12-13 16-17
29
246-248 239-241 28,32,223235--223353 239-241

LC Constants (Table)

L-C Ratios

LD-Cut Crystals

L-Section Filters

Lag Circuits

Lag, Current

Lag, Keying

Layer Height

Lead, Current Leakage, Checking Condenser

Leakage Reactance

Lecher Wires

Limiter

Limiting, Audio

Limiting (Noise) Circuits

Line-of-Sight Transmission

Line Voltage Adjustment

Linear Amplifiers

Linear Antenna Transformers

Linear Circuits

Linear Time Base

Linearity

111.

Lines of Force

Lines, Resonant Link Coupling

Lissajou's Figures

Load

Load Impedance

Local Oscillator (Receiver)

Locking

Log Paper

Log Tables

Logarithmic Functions

Logarithms

Long-Wire Antennas Loops

Loose Coupling

Low-Level Modulators

Low-Pass Filters

Lumped-Constant Circuits

Magic-Eye (Electron-Ray) Tubes Magnetic Field Magnetic Storms Magnetism Magnetomotive Force Mantissa Matching Impedances Materials Mathematics:
Abscissae Base

35,53,72-27730 61
13089 13006 157-158 30 240 32 139-212470 97,134 134-136 160,161 154-155 168,178-17497 39-40,80-82 56,363 85,86,90,100,103 22,24-25 40 36,66,171-172 232-233 21,33,43,44,143 123-124,1264-41,7217 53 19 278-279 17-18 15-16 179-182 39 36 92 38 82-83
132 24-25
158 2254,-2266
16 33 249-250
18 15

Modulation ............ .. .... .98-102,243,245-246

QPower

28,234,235,246 241-242

Receivers

243-245

TReescihsntiaqnucee Transmitters Tubes

238-239,224452,-222442675
248

MMeeagsaVucoyrlcitlnaeggeInstruments Megohm Memorising the Code Mercury Vapor Rectifiers Metals, Relative Conductivity of Meters (see "Instruments") Metric Prefixes Mho

28,233-238 28,32,225-22493
26
254
49,144 270
271
43

Microammeters MMiiccrroofaamrpaedreand Micromicrofarad Microhenry Micromhos MMiiccrroovpohlotnes Microwatt

28233 , 26
22
25
43 84,92-94
26 26

Milliammeters

Milliam pere

Millihenry

Millivolt

MMiilxleirwsatt Modulation:
Adjustments and Testing Amplitude Modulation Capability Cathode Modulation Characteristic Envelope Frequency Modulation Grid-Bias Modulation Impedance Measurements Methods Percentage PPloawteerModulation Modulators Molecular Friction Molecules Monitoring Frequencies Monitors Motorboating

26
124-126
98-102 84-86 85 89-91 85 84
102-105
88-8as9
98-102,245-246 84 84
91-92,21876-2-288068 26 21
233 100 98, 139

285

inc/ex

Mounting Crystals Moving Coil Meters Moving Iron Vane Meters MMuulta-Band Antennas Multi-Element Tubes Multi-Hop Transmission Multi-Purpose Tubes Multipliers, Frequency Multipliers, Voltmeter Multivibrator Mush Mutual Conductance Mutual Inductance
Negative Charges Negative Resistance Negative Resistance Oscillators Neutralization Neutralizing Bridge Circuits Neutralizing Condensers NNooidsees Noise Ratio Noise Reduction Non-Resonant Lines Non-Synchronous Vibrators Notation Nucleus Numerical Values

PAGE
61-62 233 234
182-18443 48-49 159 49
58, 79-80 28, 233 52, 230 139 43 36
22 53 53 67-70 39 ' 6399 97-98, 102, 120-121, 134 112, 120-121, 133 134-136 175-179 154 7, 10 21 273

Piezo-Electricity Piezo-Electric Crystals Piezo-Electric Microphone
Plate

Plate Current Plate Current Shift

Plate Detectors

PPllaattee

Dissipation
Efficiency

PPllaattee LKeoyaidnRgesistance

Plate Modulation

Plate Neutralization

Plate Resistance

Plate Supply

Plotting Graphs

Point, Decimal

Polar Coordinates

PPoollaarriiztyat ion

�

Pole

Positive Charges Pot Oscillator PPootteennttiiaolmeDtieffresrence Power Power Amplification Power Amplifiers, Receiver Power Factor Power Functions Power Gain, Antenna

PAGE
35 61-62
2943,-9442
42 101-102
114
45, 4542 14037
86-88 67-68
43 143
18 7
20
156, 162, 164-1262225 22
82-83
2227 28 45 119
31
17 162

Ohm

2266

PPoowweerr IMnepausturement

234, 235, 24465

Ohm's Law OOhhmm'msetLearws for A.0 Open Circuit Open-Wire Line Operating Angle, Amplifier Operating Point Ordinates Oscillating Detector Oscillation Oscillator Keying Oscillators Oscilloscopes Output Capacity, Tube Output Limiting Output Transformer Overmodulation Overmodulation Indicators
P Padding Condenser Parallel Amplifiers Parallel Antenna Tuning Parallel Capacities Parallel Circuits Parallel Feed Parallel Impedance Parallel Inductances Parallel Resistances Parallel Resonance Parasitic Elements PPaartatseirtnisc, OOsscciillllaotsocrospe Patterns, Radiation Peak Current Value Peak Diode Voltmeter PPeeaakk VPloaltteagCeurRraetnitng Pentagrid Converters Pentode Amplifiers Pentode Crystal Oscillators Pentodes
Period
Permeability PPehrassiestence, Screen

31
238-239 21
166 71, 79-80
44 18 115-117 50-51 110 50-53 55-57, 98-101, 236-238 43 9474 85
243, 245-246
26
122 46
171
27 33-34
52 33, 34-35
27
28-27 33-34
190
7575--7596 165
29
236 73-17444
125 66-67 62-65
25 4298 -26
29-3540

Phase Angle Phase Difference PPhhaassee MInovdeurlsiaotnion Phase Relations, Amplifiers Phased Antennas Phasing Control P'hPohtoonmee(tsreiec"RPaodwieorteMleeapshuorneym ")en ts
Piano Frequencies Pickup Coil Pi-Section Antenna Couplers Pi-Section Filters Pierce Oscillator

30 30
9874
45
186-190
130
235 225 169
169-171
38
63

Power Output Power Sensitivity Power Supply Keying

PoweCroSnusptprluyc:tional

Dry Batteries
Dynamotors

Emergency

Genemotors

Principles

Receiver Power Supplies

Storage Batteries

Vibrators

Vibrator Power Supplies

Powers (of aNumber)

Practice Cipher Groups

Practice Code Words

Prefixes, Metric

PPrreismealrecytion

Primary Keying

Propagation

Pulling

Pulsating Current

Pure D.0

Push-Back Wire

Push-Pull Amplifiers

Push-Pull Neutralization

QPush-Push Doubler

Q Antenna

Q Code

Q Measurement

Quarter Wave

Quartz Crystals (see "Crystals ')

Quench Frequency

52, 75- 76, 246 45 107
220-221 22224 221 222
143-155 208 24 154
221-222 9 11, 15
263 262 271 321,3335--13364 108 156-161 125
40 143 252
46 69 34, 35, 37, 60, 72, 7890 177-178 286-267 241-242 39
117-118

R R.M.S. Current Value Radian Radiation Radiation Angle Radiation Resistance Radiator (see "Antennas") Radical Sign Radio Frequency Amplifiers (Receivers) Radio Frequency Choke Coil Radio-Frequency Resistance

26 29 29 40 158, 162 40
9 119-121
41 34

Radiotelephony: CAldajsuss-tBmeMnotdsulaantdorTsesting

98-105 91-92, 219-220

Con4s0t-rwuactttioSnpaele:ch Amplifier-Modulator 112 Mc. "Walkie-Talkie"

217-219 215-217

Class-B modulators Microphones

219-220 92-94

Modulation Monitore

84-91 245-246

Principles

84-105

286

Jaclex

Reception Speech Amplifiers Range, U.H.F. Ratio Arms Ray, Direct and Reflected Reactance Reactance, Capacitive Reactance Chart Reactance, Inductive Reactance Meter Reactance Modulator Receivers:

PAGE
137 94-98
160 39
159-160 30-31 31 272 31 239
104-105

Antennas for Constructional:

194-195

1-Tube 2-Stage Regenerative 7-Tube Regenerative Single-Signal 9-Tube National HRO 16-Tube Hammarlund Super-Pro Fre(Sqeueenalcsyo MeUalsturar-eHmiegnhtFwrietqhuencies -- Measurements Power Supplies Principles and Design Testing and Alignment

196-198 198-202 202-203 Receive2r0s3"-)204
225 243-245
208 111-142 243-245

Receiving Code

2M-257

Reflex Voltmeter Regenerative I.F. Rectification RReeccttiiffiieerrsCircuits

236 129 42 42,49,143-144

RReefelde,ctVeidbrIamtpiendgance Reflected Waves

144-145 2323935

Self-Oscillation Self-Quenching Sending Code Sensitivity, Measurement of Sensitivity, Microphone Sensitivity, Receiver Series Antenna Tuning Series Capacities

5P0A-G5E1
118
257-259 243
92-93
112
171 27

Series Circuits

Series Feed

Series Inductances

Series Resistances ..

-26-27

Series Resonance Servicing Receivers

33 137- 139

Sharpness of Resonance

Sheet Metal Working

251-2 52

Shielding

37-38,78-77,102,119

Shielding, Effect on Inductance

274

Shields, Electrostatic

76-77

Short Wave Receivers (see "Receivers ")

Short Noise Shunts, Meter Sidebands SSiiggnnaall Generators

134 28,233 85-88,103-104,10484 243-245

Signal-Handling Capability Signal-Noise Ratio Signal-Noise Ratio, Measurement Signal-Strength Indicators Signs

113
112 ,120-121 ,133 243
132-133
10-11

SSSiiilnneeencW Ceua rsrv,veeNoise

134- 136 28 28

Reflection of Radio Waves

156,157-160,163 Single-Signal Reception

129-131

Reflector

190

RReefgreancetriaotnioonf Radio Waves Regeneration Control Regenerative Detectors Regenerative Receivers Regulation, Voltage Regulator, Voltage Relaxation Oscillator RReellauyc,taKnecyeing

157 45 115-116 115 Ill 148-114493 53 107

RRRRRRRRRRRRRRRRRRRRRRReeeeeeeeeeeeeeeeeeeeeeevssssssssssssssssssssspeiiiipiiiooooioioiiioiurssssssossnsnnnssssnnnlsttttttnttttttataaaaaasaoaaoaasaaaaaaanlnnninnnnnrnrnennnnnn,ntotttcccctcc,cccccCcceeeneeeeeePeeeToeCFL,,,IsF-h-i-rli-rCmCMrCrRCRNiCaanCoeuopeecnaa.oaeensorqureaquppFudguseuSpvdiausCCap.ipameeltaeicceluaolitrsinriinenteirsidcncttdciecdsveeugyyysoeyiemiAAnotaTFeumnsininpdldlmtitLieeioPfrnaiACeremsoarpnlllsDietaflaitneatrs(Chart2)38-239,321674,32222992-36,57511--22-1--.444732633243143992592322545035569273047366788626

RRihbobmobnicMiAcnrtoepnhnoanes Ripple Rochelle-Salt Crystals Root-Mean-Square Value Roots Rotary Antennas

185-186 94
145 93 29 9-10,11,15 191

Single-Wire Feed Skin Effect Skip Distance Sky Wave Slide-Back Voltmeter Smoothing Choke SSoolladrerCiyncgle Sounds, Code Space Charfe Spectrum, .C. Speech Amplifiers Sporadic-E Ionizati on Square Root Squegging Stability, Frequency Stability, Receiver SSttaabgieliGzaatiinon, Voltage Standard Signal Genera tor Standards, Frequency Standing Waves Standing Wave Ratio Starting Voltage Static Characteristics Static Charge Static Electricity Storage Batteries Stub Antenna Matching Substitution Measurements Sunspot Cycle Super-Control Tubes Superheterodyne Superimposed Currents Superregeneration Superregenerative Detectors Suppressor Grid Suppressor-Grid Keying Suppressor-Grid Modulation Suppressor, Noise Burge Impedance

188,175

158-15394 156.158-159
236

147
158
253
254-257

4320 94-98,217-219
181
9-10
118,139
52-53 112,139 148-150
95

243
229-231 39
169 148 43 21-22 21-23 24 178-179 240 158 49 111 ,1 23- 124 40-4 1

117-118

111-112117--118

�

108

134-136

167

SMeters Saturation Point Sawt,00th Wave Scales Schematic Symbols Screen Grid

132-133 42
55-56 Fronlis1p8i-e2e0e
48

Sweep Sweep Circuits Swing, Grid Swinging Choke SSyymmbboollss,, ESlcehcetmraictailc Sync hron ization

55 56 46 147 Front ispi2e7c0e
56-57

Screen-Grid Keying Second Detector Secondary Secondary Emission Selectivity Selectivity Control Selectivity, Measurement of Self-Bias Self-Controlled Oscillators Self-Induction

124,131,134-113057 32,35-36 48
34,112,129 131 243 50
59-81 25

Synchronous Vibrators

154

Tables and Charts (see "Charts and Tables")

Talking Level, Microphone

Tank Circuits

52,60,72-73

Taps

249

Tapped Circuits

Temperature Coefficient of Crystals

Temperature Inversion

161

Termination, Line

168

287

.-9/14111X

Test Equipment Test Oscil ator Tetrode Amplifiers Tetrode Crystal Oscillators Tetrodee Textbooks Thermal Agitation Thermionic Conduction Thermionic Emission Thermocouple Ammeters Thickness of Crystals Threads, Cutting Threshold Control Thump Filters Tickler Tight Coupling Time Constant Tolerances, Color Code Tone Control Tone Sources Tools Trace, Cathode-Ray Tracking Transceivers Transconductance, Grid-Plate Transformer Color Codes Transformer Coupling Transformer Design Transformerless Power Supplies Transformers
Transformers, I.F Transmission Linee Transmitters:
Antenna Coupling

PAGIO
233-248 244-245
66-67 62-63
48 280 120
24 24 32,235 61 251 136 109 51 36-37 27-28 269 119 260-265 249 54,56 127 215 43 269 36, 86, 96 148 152 32-33 128 129 166-179 169-172

Ultra-High Frequencies: Constructional: Rec2e2i4ve&rs:112 Mc. Superregenerative Hallicraf ter S-27 FM-AM Superhet Tra1n1s2m-iMtct.erB-aRtetceeriyve"rWs:alkie-Talkie" Tra1n1s2m-iMtet.er8s1:5 Linear Oscillator 224-Me. HY75 Oscillator
Upward Modulation
V Antennas V-Cut Crystals Vacuum Tube Voltmeters Vacuum Tubes Vacuum Tube Testing Vacuum Tube Symbols VVaarriiaabbllee-pSeTleucbteivsity Velocity, Angular Velocity Microphone Velocity of Radio Waves Vibrating Reed Vibrator Power Supplies Vibrators Virtual Height Volt Volt-Amperes Voltage Amplification Voltage Dividers Voltage Drop Voltage-Doubling Circuits VVoollttaaggee FMeeeadsufroremAennttennas

PAOE
204-206 207-208 215-217 213-214 214-215 101-102 184-185
61 235-236 24, 42-57
248 27409 129, 130-131
29 94 39 225 221-222 154 157-158 22 31, 148 45 ,.27, 152 27 152-153 28, 233-213686

ConCsotmrpulcteitoenaTlr:ansmitter Assemblies Hallicrafter HT-6 25-watt
Exciters or Low-Power Transmitters: 15-Watt Output 6L6 Bread-Board 50-Watt Output 6L6-807 Plug-In Coil.
Power Amplifiers: 450-Watt Input Push-Pull

223-224 224
209
.209-212
212-213

(See also "Ultra-Iligh Frequencies -- Measurements Principles and Design Transposition Trapezoidal Pattern Triangular Spectrum Triggering Voltage Trimmer Condenser Triodes Triode Crystal Oscillators Triple-Tuned Transformers Tri-Tet Oscillator Troposphere Refraction Trouble Shooting (Receivers) Trough Line Oscillator TTuubbee CKheercrking Tube Noiee

Transmitt2er4s5"-)246 58-83 12-13
98-100 103 56 122
42-44 62
128-129 63-65 161
137-139, 243, 244 82
109, 224685 120-121

Tubes (see "Vacuum Tubes") Tuned Circuits Tuned-Grid Tuned-Plate Circu it

33-35 52

Tuned R.F. Receiver TTuunniinngg FInodrikcators

111
132-212353

Voltage-Ratings (Color Code) Voltage Ratio, Transformer Voltage Regulation Voltage Rise VVoollttamgeet-eSrtsabilized Power Supplies Volume Control

269 32
143 34 28, 233-234, 213458--215308 96 -97, 131-133

WWV Standard Frequency Transmissions Watt WWaattttm-eHtoeurr Wave Analyzers Wave Angle Wave Envelope Pattern WWaavveefPorrompagation Wavelength Wavelength-Frequency Conversion Wavemeter (see "Frequency Meters") Wheatstone Bridge Wien Bridge Winding Coils Wire Table Wiring Apparatus Wiring, Current Capacity Wiring Diagrams Working Voltage Workshop Practice

230 26
28, 22364 242-214538
98-100 29, 44, 110506,-116016
39 39
38-39, 239 232 253 276
252-253 271
Frontispiece 145
249-253

X-Cut Crystals

Tuning Receivers Tuning Transmitters TTwuirsntseRdaPtaiior Feed
Two-Conductor Line

116-117, 136-137, 142
60-61, 83, 64-65, 74-71,72 166, 175-176
80-52

Y-Cut Crystals Z Zero Beat ZZeerpop-BFeeaetd IfnodricAanttoernsnas

61 31 117 231-213723

Ultraudion Circuit

4288

THE RADIO AMATEUR&#39;S HANDBOOK

Full PDF Document

Extracted Text

THE RADIO AMATEUR'S HANDBOOK