GE SSB Jr. User manual

HAM NEWS
SS , Jr.
Presenting a 3 -Tube 5-Watt SSB Transmitter with Superior Performance
rtr"Pir-._ .11 ..
° -; ,;_.D .i-
.- ° & " 43 1-.
- _ . . i . :Y. /`
Fig. 1. Front panel view of the SSB Jr. For
single-frequency operation none of the con-
trols need be adjusted (except the audio gain
control). Front-panel mounting of the controls
permits a compact physical layout to be
obtained.
wr '
FEATURES -
Simple to construct
Uses inexpensive parts
Has sideband-reversing control
Usable as emergency, portable
or home transmitter
CONTENTS
SSB Jr. pages 1-9
Designer's Corner (Notes on the design of the SSB Jr. rig) page 10
Sweeping the Spectrum page 11
Technical Information (6A07) page 12

SSB, Jr.
ELECTRICAL CIRCUIT
AUDIO
INPUT I2AU7GREEN
R2 3
1R3
t //
PHASE SHIFT NETWORK
I q0B
R15 i A' C7 ¡ I
98G1G"
BUJE J
1. ;O' I 12AT7
lo;l
Q l
RED
R4
R5
BLUE
T2
\TEL.:a
/+ G2A
i'
XTAL OR V. F.O.
O
RED C3
a WH.
117(
o Cl2
RED
RI!
18
BLUE RED
YEL.E'RE1-4/
Suj
13
A.F.
BALANCE
02B
R12
2 12ÁU7
6
R14
m.C4
~G5~I
C13T
*L5
TWIST
OUTPUT
RF 0 1
RFC 019
*MOUNTING END
C18 6AG7 x
6aL4
14
5
G2C YC2D
GI7
K 1
RFC
L3 Y
g G20-
021T
G16
RFC
GI
G
014>m
G2 R16
a
G3
G15
C+, B-
10.5 V.
- 300 V.
Fig. 2. Circuit diagram of the SSB Jr.
041110004X10G6 -
L21,
1,80e818e081,
RFC
GI,
Circuit Constants
(All resistors and capacitors *20% tolerance unless specified otherwise)
C, 0.5 mf 200 volt paper
Cz 20-20-20-20 mf 450 volt electrolytic
C,, C,, Cie, C,2, C,e 1000 mmf mica or ceramic
Cs, C,,, Cl- 250 mmf mica X10%
C,: May not be necessary, see text
Ls
L,Ls
16 turns No. 19 enamel wire spaced to
fill Millen No. 69046 coil form. Tap at
8 turns. Link of 1 turn at center.
Same as L, except no link used
28 turns No. 19 enamel wire. Link on
Cr 2430 mmf (0.002 mf mica *5%a with 170-780
mmf trimmer in parallel) open end to match load. (4 turn link
matches 72 ohm load.)
C,
C,
4860 mmf (0.0043 mf mica 5% with 170-780
mmf trimmer in parallel)
1215 mmf (0.001 mf mica *5% with 50-380
mmf trimmer in parallel)
RFC
R,RºRs
Radio-frequency choke 0.5 millihenry
1 megohm potentiometer
10,000 ohm, % watt
750 ohm, % watt
C,,, 607,5 mmf (S00 mmf mica X10% with 9-180-
mmf trimmer in Parallel) R,Rs 430 ohm, 3,1 watt (*5%)
100 ohm potentiometer
Cu,n C,, 0 005 mf mica or ceramic R, 1600 ohm, A watt (*5%
CC 350 mmf 600 volt mica X10% (250 mmf in
parallel with 100 mmf)
0 01 mf mica or ceramic
R,, R,,
Rs, R,
R 133,300 ohm, % watt (*1%0
100,000 ohm, % watt ( 1%)
510 ohm, A watt (*5%)
C,,, C2, 0 002 mf mica X10% R,2 500 ohm potentiometer
G,, 02, 0,, G,... 1N52 germanium diode or equivalent,
see text R,,R 330 ohm, 3.4 watt
47,000 ohm, 34 watt
J Open circuit jack R,,, R,, 20,000 ohm, 1 watt
L,, L2 33 turns No. 21 enamel wire close wound R,6, R,; 1000 ohm potentiometer
on Millen No. 69046 iron core adjust- S DPDT toggle switch
able slug coil form. Link of 6 turns of
hookup wire wound on open end. T,Ts, T, Stancor A -53C transformer
UTC R-38A transformer
The SSB Jr. is a complete single-sideband trans-
mitter-just add microphone and antenna and you
are on the air. No longer must amateurs feel that
single-sideband equipment ís too complex to under-
2
stand or too complicated to build. The SSB Jr, rig
is no more difficult to build or adjust than any
modern 3-tube transmitter. This rig should bring
SSB within the reach of anyone that is interested.

Further, any amateur can build the SSB Jr. rig and
be assured that his single-sideband signal will be
second to none in quality. Performance has not been
sacrificed in the interest of simplification.
The peak power output is 5 watts and the total
power input, not including filament power, is 18 watts
(300 volts at 60 ma.). The SSB Jr. rig features a self-
contained crystal oscillator (or buffer for VFO
operation), 40 db. sideband suppression, and me-
chanical and electrical ruggedness that make it ideally
suited as a complete portable, mobile, emergency
transmitter, or as an exciter for a home transmitter.
The system used in the generation of the single-
sideband signal is a simplified phasing method that
is daringly direct and effective. Inexpensive and easily-
available components are used throughout.
All of the information necessary to construct and
adjust the SSB Jr. rig appears in this article. Tech-
nical details on the new phase-shift network and the
new modulator design are explained in the Designer's
Corner section of this issue.
With reference to the circuit diagram, Fig. 2, the
first tube, a 12AU7, is a twin-triode, combination
speech amplifier oscillator. A 12AT7 serves as a twin-
channel amplifier in the output of the phase-shift
network, and the final is a 6AG7 pentode.
Starting with the audio circuit, an input gain con-
trol potentiometer feeds the grid of the self-biased
input tube, which is one-half of the 12AU7 miniature
tube. The output of this tube is coupled into a
newly designed audio phase-shift network by means
of transformer T1. The outputs of the phase-
shift network feed separate triode sections of the
12AT7 miniature tube. These two tube sections are
transformer coupled to two balanced modulators
each of which employs a pair of germanium crystal
diodes.
The balanced modulators are also supplied by r-f
signals from the crystal oscillator, which is the other
half of the 12AU7. These r -f signals are picked up by
separate link windings on L1 and L2, which comprise
portions of a 90 degree r-f phase-shift network in the
plate circuit of the oscillator. The balanced modulators
work into a balanced load circuit (L:,, C21, C21) which
is link coupled to the grid circuit (L4, C17) of the class
AB' linear power amplifier tube, a 6AG7.
This power amplifier works into a conventional
tank circuit (LU, Cm) that is link coupled to the load.
All circuit tuning is accomplished by adjustable
slug-tuned coils wound on Millen No. 69046 powdered -
iron coil forms.
Sideband switching is accomplished by the reversal
of audio polarity in one of the audio channels (switch
S). Provision is made for equalization of gain in the
audio channels, this equalization being necessary in
order to achieve normal sideband cancellation. In
addition, a semi-fixed control (R7) is provided for
phase-shift network adjustment. Use of this control
eliminates the need for a special transformer, or the
need for two non-standard precision resistors. Stable
modulator balance is achieved by the balance/buffer
resistors R16 and R17 in conjunction with the ger-
manium diodes.
The audio characteristic of the SSB Jr. is designed
to emphasize the intelligence-bearing frequencies
from 300 to 3000 cycles per second. This feature is
obtained jointly by the action of C1 and the audio
transformer T1. Low differential phase-shift is main-
tamed in audio circuits following the phase-shift
network by means of lightly loaded output trans-
formers which are shunt-fed to reduce harmonic
distortion caused by direct current in their windings.
A 5 by 7 by 2 inch chassis provides ample space,
with good access, for all component parts. A cabinet,
as shown, may be used, although this is not essential.
It is recommended that parts layout shown in the
sketches and the photographs be followed exactly.
Obviously other layouts will work, but the layout
shown has been carefully made and many layout
problems have been eliminated.
Before starting work ón the main chassis it is
advisable to make the audio phase-shift network
board. This is diagrammed in Fig. 3. The base ma-
terial may be thin bakelite or any insulating material.
The dimensions are 4 inches by 2% inches. Note that
one corner is cut off to permit access to the 12AU7
tube. This board uses four fixed mica condensers
which are padded with four adjustable mica trimmers,
and four precision resistors (Continental Nobleloy
X-3, plus or minus 1% tolerance). In the unit
shown R, and R9 are as specified, that is, they are
Continental Nobleloy 100,000 ohm resistors. How-
ever, the 133,300 ohm resistors were made by taking
two 150,000 ohm precision Continental Nobleloy
resistors and paralleling each of them with a one-half
watt 1.2 megohm (plus or minus 10% tolerance)
resistor. Careful selection of the 1.2 megohm units
will permit close adjustment to the desired value of
133,300 ohms. A convenient way to mount the 1.2
znegohm resistors is to slip them inside the hollow
body of the precision 150,000 ohm resistors.
The phase-shift network sub-assembly is mounted
on three half-inch long spacers under the chassis
directly below transformers T1 and T2. It is best to
dress the leads from these transformers flat against
the chassis to clear the phase-shift network. Time
will be saved by installing the network sub-assembly
as the last step in the construction.
Mount the phase-shift network elements as shown
in Figs. 3A and 3B. The dashed connections should
be omitted initially, since the detailed alignment
procedure described later presumes that these con-
nections will be made at the proper time only.
21x4K4" TExT .ITE THICK
C7 - 0.002 MICAWITH 170-780 MME
TRIMMER. (2430 MME TOTAL)
8 - 0.0043 MICA WITH 170.780 MME
TRIMMER. (4860 MME TOTAL)
9- 0.001 MICA WITH 50-380 MME
TRIMMER. (121S MME TOTAL)
C10- 500 NNW MICA WITH 9 T0(80
MME TRIMMER. (607.5 MME
TOTAL)
A
I7: pa . 5 _`(
- +7 '; y
B .
Fig. 3. (A) Mechanical arrangement of the audio phase-
shift network. (B) Detail view of the audio phase -shift
network
3

A word of caution about the coils. Make sure that
the hot and cold ends are as specified on the circuit
diagram-the asterisk indicates the end which is
the mounting end, that is, the end with the long
tuning screw.
The links on the coils are wound over the cold end,
as indicated in Fig. 11. As a suggestion, wind the
links with solid insulated hookup wire. This type
of wire is convenient, holds on well, and makes a
nice looking job. Twist the wires together when
running from one coil to another coil, or to another
connection point. A small terminal strip may be
placed under L, to serve as a convenient junction
point for the links coming from L, and L2 and going
to the balanced modulators.
The small fixed mica tuning condensers that con-
nect across L1, L2 and L3 are mounted on the coil
form terminals. The coupling capacitor between L1
and L2 (C6) is shown dotted in the circuit diagram,
since the amount of actual capacitance needed at
this point will depend on stray coupling effects in
the particular unit you build. More information
will be given on this later.
Note that the grid connection of the 6AG7 is
above the panel from the hot end of La through a
hole in the chassis right next to pin number 4 (the
grid terminal) of the 6AG7 socket. Direct strapping
of terminals 1, 3 and 5 of this socket to the chassis is
desirable to ensure stable amplifier operation. Note
also that a 2 by 2)4 inch brass or aluminum shield
is placed between coils L2 and L; below deck.
The unused transformer leads may be cut off close
to the winding and forgotten. The secondary wind-
ings of T2 and Ta have several intermediate taps that
are not used. All leads from the three transformers
are fed through small rubber grommets in the chassis
to circuits on the underside. All, that is, except the
secondary leads from T3 which remain above chassis.
Twist these leads together before running them to the
sideband reversing switch on the front panel.
Do not ground either heater lead in the chassis,
as you may wish to use an a-c heater power supply
or perhaps run your automobile engine while trans-
mitting if the rig is used for mobile work.
Ample mounting space for C1 and R3 will be found
near C2, the four-section electrolytic condenser. With
reference to C2, one 20 mf section is C2A, another is
C2n, etc. The heater leads that run from the 12AU7
may be cabled together with the other leads from
T1, T2, T3.
The germanium diodes deserve special care in
handling. Do not bend the leads close to the diode
unit itself. The diodes are mounted by means of their
leads between the coil terminals of L3 and the ap-
propriate ends of 121, and R77. Protect the germanium
diodes from heat while soldering by holding the lead
with cold pliers between the diode itself and the end
where the soldering is taking place. Further, use
only as much heat as is necessary to make a good
joint.A four-wire shielded cable brings power from the
power supply to the exciter. The shield serves as
the negative plate supply lead and should be con-
nected to chassis ground. A male plug at the other
end of the cable makes a convenient connection to the
power supply.
The power supply is not unusual in any respect.
Any source of power supplying 300 volts and 60 mils
or more may be used. It is not necessary to use
electronic bias either, and a standard battery sup-
plying 10.5 volts may be used for bias.
4
The power supply used with the SSB Jr. rig pic-
tured is shown in Fig. 7 and the circuit diagram
given in Fig. 6. A 5V4 -G rectifier tube feeds a single-
section filter to supply 300 volts, and a 6H6 tube
acts as a bias rectifier to supply 10.5 volts. Resistor
R1 adjusts the bias voltage obtainable.
The main a-c switch is S1, and the stand-by switch
is S2. Note that resistor R2 acts as a low resistance
bleeder to drop the positive voltage to zero quickly
when the rig is turned off. A double-pole switch is
employed with the switch arms tied together, as this
arrangement gives the effect of a double break con-
tact.There is nothing critical about the power supply
layout, and any arrangement may be used to suit
your convenience.
The SSB Jr. rig as designed requires that a high-
output microphone circuit be used. A single-button
carbon microphone, connected as shown in Fig. 8B
is quite adequate, even desirable, if mobile operation
is contemplated.
On the other hand, low-level microphones, such
as the usual type of crystal or dynamic microphone,
may be used if a one-tube preamplifier is provided.
A suggested circuit is shown in Fig. 8A. This pre-
amplifier may be built as a separate unit or incor-
porated into the SSB Jr. rig. Either the preámplifier
shown or the single-button carbon mike circuit will
provide in excess of the 2 volt (RMS) signal level
required as a minimum input signal to the SSB Jr.
As is true with many transmitter designs, there
are some component parts used in the SSB Jr. rig
that must be chosen carefully. Obviously, the pre-
cision resistors specified are important. If precision
resistors are not available-although you should
try to get them if at all possible-you may use non-
precision resistors which have been checked on a
good resistance bridge. You may find that these
resistors will change value after they have been
used for a while, and that is why it is desirable to
use precision resistors initially.
The adjustable mica trimmers used in the audio
phase-shift network may be any good grade of mica
trimmer. Those actually used are El-Menco mica
trimmers-T52910 for the 170 to 780 mmf range;
T52510 for the 50 to 380 mmf range; and T52310 for
the 9 to 180 mmf range.
Resistors R4, Rs and R11 are specified as plus or
minus 5% tolerance. This is because the values stated
are required, and these values only come in the 5%
tolerance series.
The germanium diodes are specified as 1N52
diodes. Other types, such as 1N48, 1N51 and 1N63
may be used instead. If possible, select four diodes
which have about the same forward resistance. The
forward resistance is the low resistance as checked
on an ohmmeter. To determine approximately what
it is, measure the resistance in one direction, then
reverse the leads to the diode and make a second
I measurement. The two readings should be quite
different. The lower resistance is the one of interest.
I Make this measurement on the four diodes you
intend to use to make sure that the forward resistance
of any one of the diodes is within ten per cent of the
average resistance of the group.
The diodes used in the rig shown measured ap-
proximately 250 ohms on a Weston 772 analyzer
when the analyzer was set to the RX10 scale. (Dif-

_1_
R1 AUDIO GAIN
111, .í .9 .
N1
f.
-1' 1
-----__ -1---.\
Fig. 4. Rear view of the 55B Jr.
3"
PHASE SHIFT NETWORK
(BELOW ON I/2" SPACERS)
Ti STANCOR
e No3c
Sill ,//
L----____ _, .T I
---- i' xi,
r-
1 I '\I2AU7 ¡ t
i i r.-'-1.1 i 1/2"
L___ ----J; n'-4--'-'-ii + .1
XTAL FRONT
SOCKET LI
(BELOW)
R5
(USE SHAFT LOCK)
r-----,
II -----J
Ii12AT7 i
C2 a,b,c,d
CABLE OUTLET
-c+
5"X 7"X2"
CHASSIS
i`// \
AG7 KEEE
\ /
\\ /
L2 L3 L4 (ABOVE)
(BELOW) (ABOVE) L5 (BELOW)
Fig. 5. Chassis layout for the 55B Jr. (top view)
411.
5

T
ELECTRICAL CIRCUIT
Co 5.0V5V4 -G
L,
-O J 300V.
-300V.
10.5V.
-10.5V.
6.3 V.
6.3V.
Fig. 6. Circuit diagram of the SSB Jr. power supply
Circuit Constants
(All resistor, and capacitors *20% tolerance unless specified otherwise)
C, 40 mf 450 volt electrolytic Rx .. 1000 ohm 1 watt
C:, C, 50 mf 50 volt electrolytic S, SPST toggle switch
J Closed circuit jack or terminal posts to S. DPDT toggle switch
L permit metering with 0-100 mil meter 7 henry choke, 160 mils (UTC R-20) T Power transformer, 350-0-350 at 75 mils,
6.3 volts at 3 amperes, 5.0 volts at 3
R, 100 ohm potentiometer amperes (UTC R-11) l
ferent ohmmeters may give different readings, since
the diodes are non-linear in nature.)
It is recommended that transformers T1, T2 and
T3 be as specified. Do not make any substitution un-
less you wish to duplicate a long series of tests to
determine if the substitutes are suitable. The types
indicated are standard parts, inexpensive, and easy
to procure. Observe that the connections are indicated
on the circuit diagram by their color code.
It is further recommended that you use Millen
No. 69046 coil forms as specified. While the coils are
not critical, they must have a certain inductance
and distributed capacitance, and ifyou adhere to the
specifications given you should encounter absolutely
no difficulty coil-wise.
The adjustment of the audio phase-shift network
circuits is most easily done with the phase-shift sub-
assembly out of the chassis. The resistors R7 and R.
(and R10 and Ro) should bear the ratio of 133,333 to
6
It
-.
Fig. 7. SSB Jr. Power supply
100,000, that is, 4 to 3, as closely as can be deter-
mined. If in doubt as to the ratio of the resistors
you used, double-check their value on an accurate
bridge. The adjustment of the phase-shift network
now consists only of setting the four capacitors (C7
through C10) to their proper values. Several methods
can be used. The most accurate will be described.
An audio oscillator capable of operation from 225
to 2750 cycles per second (with good waveform)
is required, plus an oscilloscope. The oscillator should
be carefully calibrated by the method described
later. Connect the output of the audio oscillator
through a step-down transformer (the Stancor A-53C
will serve nicely) to a 1000 ohm or 2000 ohm poten-
tiometer with the arm grounded.
Adjust the arm position so that equal (but op-
posite) voltages appear on each half of the poten-
tiometer. A steady audio frequency signal of any
convenient frequency may be used with an oscilloscope
acting as a convenient voltmeter for this job. Swing
the vertical deflection lead from one end of the
potentiometer to the other and adjust the arm to
obtain equal voltages (a true center tap). Set up a
temporary double cathode-follower circuit using a
12AT7 with 500 ohms from each cathode to ground
and connect as shown in Fig. 9. (It will be convenient
to provide leads M, N, and 1 and 2 with clips at the
ends to facilitate checking.) One may use the 12AT7
in the rig as the double cathode follower by temporar-
ily short circuiting the plate of each tube to its
respective center tap of the UTC R-38A transformers.
Be sure to remove the 12AU7 and the 6AG7 at this
time, and of course supply operating voltages for the
12AT7. Pins 3 and 8 should connect to the H and V
deflection amplifiers in the oscilloscope, and the oscil-
loscope common connection should be made to the
chassis.
First connect lead M to terminal A on the phase-
shift unit, and lead N to terminal A'. Connect leads
1 and 2 to terminal M. (Note that the dashed con-
nections are missing at this stage of adjustment.)
I
4

ELECTRICAL CIRCUIT
R2M 12AT1
HIGH
IMPEDANCE
MIKE 2
S. B. CARBON
MIKE
B 4.5T06V.
DC
T
5
AUD0
OUTPUT
300
TC5
6.3V.
,(C4 AUDIO
OUTPUT
Fig. 8. Suggested microphone circuits for use with the SSB Jr.
Circuit Constants
(All resistors and capacitors *20% tolerance unless specified otherwise)
Cs, Cz, Cs 0.05 mf 400 volt paper R, 470 ohm, 3 watt
Ca, Cs 8 mf 450 volt electrolytic R7 47,000 ohm, M watt
Rs 2 megohm, 3 watt Re 33,000 ohm, M watt
R2 0 1 megohm, 1 watt T Microphone to grid transformer
R3 0 5 megohm, Mj watt
Adjust the horizontal and vertical gains on the
oscilloscope to produce a line about WI inches long
slanted at 45 degrees when the oscillator is set to a
frequency of 490 CPS (an exact method of setting
frequency will be described later). If the oscilloscope
has negligible internal phase shift the display will be
a straight line instead of a narrow slanting ellipse.
If the latter display appears it is necessary to correct
the oscilloscope phase shift externally by using an
adjustable series resistance (a 50,000 ohm poten-
tiometer) mounted at either the vertical or horizontal
input terminal, depending on what correction is
necessary.
At any rate, the objective here is to get a single
straight line at 490 CPS. In some cases a series
capacitor may be needed to provide the necessary
correction. Try values from 0.05 to 0.0005 mf. Now
shift lead 1 from terminal A to terminal B on the phase
shifter. Adjust the trimmer of C7 to obtain a circle
on the oscilloscope. It will be noted that as this ad-
justment is made the display will shift from an ellipse
"leaning" to one side through a circle or ellipse (with
axes parallel to the deflection axes) to an ellipse
which leans the other way. If desired or necessary,
the appropriate gain control on the oscilloscope may
be changed so that a circle instead of a "right"
ellipse is obtained at the point of correct adjustment.
After changing the gain control on the oscilloscope,
check (and correct, if necessary) the phase shift in
the oscilloscope by moving lead 1 back to terminal
A, and then repeat the setting of C, with lead 1 back
on terminal B.
In general, always make certain that the oscillo-
scope is used in a phase-corrected manner. As a
double-check (if the deflection plates in the oscillo-
scope are skewed, for instance) connect lead 2 to
terminal A'. If the circle changes to a slanting ellipse,
readjust C7 to produce an ellipse "half-way" between
the ellipse (obtained by switching lead 2) and a
circle. Changing lead 2 from A' to A and back again
should give equal and opposite skew to the display
when C7 is set correctly. Failure to get symmetrical
ellipses (egg-shaped, or other display) is due to dis-
tortion, either in the oscilloscope, the oscillator, the
transformer, or the cathode follower. Conduct the
test at as low a signal level as possible to avoid dis-
tortion.
Next connect leads M and N to terminals E and E',
respectively. Connect leads 1 and 2 to E, set the oscil-
lator frequency to 1960 CPS, correct oscilloscope phase
shift as before, and move lead 1 to terminal G. Adjust
C10 for a circle as was done for C7, using the precautions
outlined for that case.
Now connect lead M to terminal D, and lead N
to terminal F. Connect leads 1 and 2 to terminal D,
set the oscillator frequency at 1307 CPS, correct
oscilloscope phase shift as before, and move lead 1
to the junction of R9 and C,. Adjust C9 for a circle
on the oscilloscope, as before.
Repeat the above procedure for the remaining
R-C pair, Rs and C,. Use terminals D and C this time
and set the oscillator for 326.7 CPS. This completes
except for a final check the adjustment of the phase-
shift network. Connect A to A', E to E', B to C, F to
G, and A to E. Be certain to remove the temporary
short circuiting connections between the 12AT7
plates and T2, T3.
If the oscilloscope did not require changes in ex-
ternal compensation over the four frequencies used
an over-all frequency check can now be made easily
on the phase-shift network. To do this, connect lead
1 to point B, C, lead 2 to point F, G, lead M to point
A, A', E, E', and lead N to point D. Now shift
the arm of the potentiometer toward M until a circle
appears on the oscilloscope screen at a frequency of
250 CPS. Then, as the oscillator frequency is varied
from 250 CPS to 2500 CPS, this circle will wobble a
little from one side to the other, passing through a
perfect circular display at 440, 1225 and 2500 CPS.
The audio band over which the wobble indicates
a plus or minus 1.3 degree deviation from 90 degrees
is 225 to 2750 CPS, or 12 to 1 in range. This means
that when other circuits are properly adjusted, a
7

AUDIO
IWO
Om POT
N
PASSE
SHIFT
NET-
WORK
The frequency ratios (just as the resistance ratio
previously mentioned) are far more important than
the actual values of frequency (or resistance) used.
Fig. 9. Audio phase-shift network test layout
sideband suppression ratio of 39 db is possible at the
worst points within this range. The average suppres-
sion ratio will be about 45 db. Proper phase-shift
network operation is necessary to obtain this class of
performance, so the adjustment procedures have been
explained in great detail as an aid toward this goal.
The phase shift network should never require read-
justment, so that when you are satisfied with the
adjustment you may seal the trimmers with cement.
r
It will be noted that the frequency ratios are such
that the 12th harmonic of 326.7 CPS, the 8th har-
monic of 490 CPS and the 3rd harmonic of 1306.7
CPS are all the same as the 2nd harmonic of 1963
CPS, namely, 3920 CPS. Thus, if a stable source of
3920 CPS frequency (such as a thoroughly warm
audio oscillator) be used as a reference, the frequency
of the test oscillator can be set very closely to one-
half, one-third, etc., of this reference frequency if
both oscillators feed an oscilloscope and the resulting
Lissajous figures observed.
Use of a calibrating frequency in this manner
assures that the frequency ratios used are correct,
even though the exact frequencies used are unknown.
R12 AUD. BAL.
+ ¡t/ DPDT
SWITCH
Iá (SB-REV.)
A.F. INPUT
JACK
Install the phase-shift network in the chassis,
remove the 6AG7 output tube, plug in a crystal
(3850 to 4000 KC) or supply a signal to the crystal
socket from a VFO at not less than a 10 volt (RMS)
level, set L1 and L2 for minimum inductance (slug
out, counterclockwise) and apply power. The current
drain should be about 35 to 40 MA at 300 volts
under this condition with the oscillator operating. If
the current drain is over 45 MA, turn off the B+
power, adjust LI, reapply power, etc., until the crystal
oscillates. This may be checked by means of a re-
ceiver tuned to the crystal frequency. Continue to
advance the slug in LI with the crystal operating until
oscillation ceases. Then back the slug out a few turns
to assure stable crystal operation. For VFO input
simply adjust L1 for minimum total current.
Apply an audio signal of 1225 CPS to the input
jack of the exciter and connect the horizontal de-
flection of the oscilloscope to a cathode (pin 3) of the
12AT7, and the vertical deflection to the other cathode
(pin 8) after making certain that the oscilloscope is
phase-compensated at the frequency of 1225 CPS.
Adjust RS to produce a circle on the screen. Adjust
R12 to about mid-range. This test should be made at a
reasonably low audio signal level (in general, the
lower the better).
Now plug in the 6AG7, after checking to see that a
bias of about 101..f, volts is supplied. Connect the out-
put link on L; to the vertical plates of the oscilloscope
(no amplifier used). Deliberately unbalance one of the
/ \RI6
7 MOD. I \
' BAL. 1
3"
38
/ MOD.
/ LOAD
/ I
1---i-
\L3R17 7"
+ ` - 116
MOD.II
3"\ BAL 5x7"X2"
4 , CHASSIS
P. A.
--+ /L4
2 8 1r12AU7 I SHIELD
I I SOCKET 12"X2!"
L-+- R. F. I1,7 2
/ RI GAIN\\ I'Í OSC.
IPLATE
PHASE
+-IO\ 2
i l I + 1 4
L
8
I
I
.L.,/ " /l ) XTALOR VFO 4 1
Am,/ SOCKET
I" I" I" I" --i
4 16 18--I8 I16 ~ IÍ6
Fig. 10. SSB Jr. panel layout (front view)
P.A.
PLATE
3 15"
12
1"
1
1

modulators by setting R16 appreciably off-center.
Adjust L6 for maximum vertical deflection at any
convenient sweep speed. This deflection may be
small at first since other circuits are not yet tuned.
Adjust L3 for further increase of deflection (maximize),
and then finally tune L, for maximum output. As
this tuning is done it may be necessary to reduce the
modulator unbalance to keep from overloading the
output stage. Check the tuning again on L5, L3, and
L4, in that order. Next remove all audio input by
turning R1 to zero, and, by successive alternate
adjustments of R16 and 1247, balance the modulators
for zero output as seen on the oscilloscope. It will be
noted that as the correct points are reached the mini-
mum point becomes successively sharper on each
control.
Next apply some 1225 CPS audio tone to the ex-
citer by advancing R1. Undoubtedly some RF en-
velope will be seen. Adjust L2 (the RF phase control)
in such a direction as to reduce the "modulation"
appearing on the output. Remove the tone, check
modulator balance (R16 and R17), and repeat the
adjustment of L2. The crystal (if used) may stop
oscillation during this operation due to interaction
between L2 and L1 tuning. If so, back out the slug on
L1 until stable crystal operation is obtained. With
the 1225 CPS audio signal still applied continue to
adjust L2 for minimum "modulation" or ripple on
the envelope, checking modulator balance periodi-
cally. When a minimum point is reached, adjust R12 to
still further reduce this ripple, then adjust L2 for more
reduction, etc. until a substantially ripple-free dis-
play is seen.
With L2 tuned it is now time to check the r -f volt-
ages applied to the modulators. Temporarily remove
the audio tone and connect the vertical deflection
plate of the oscilloscope to the arm of R16. Always
keep the common connection of the oscilloscope
grounded to chassis. Note the deflection and then
check the voltage on the arm of R17 in a similar
manner. If this is appreciably lower than the first
voltage (on the arm of R16) more coupling capacity
(C6) is necessary between L1 and L2.
Usually very little capacity is required, and this
can be provided conveniently by making a condenser
of two pieces of insulated wire twisted together for
half an inch or so. Adjust the amount of capacity by
clipping off a little bit at a time to approximately
equalize the RF signals appearing on the arms of
R16 and R17. (Note: check both voltages after each
adjustment of capacity, since both voltages will
change.) Connect the oscilloscope to read r-f output
from L6 as before, and then check modulator balance.
Apply the 1225 CPS tone and make whatever slight
adjustment is necessary in L2 tuning to obtain the
ripple-free display obtained before the coupling ca-
pacitor (if necessary) was installed. Remove the audio
tone and check modulator balance (Ri6 and R1 ).
This completes the adjustment of SSB, Jr. A dummy
load may now be connected, or the output used to
drive a high power linear amplifier.
Note that when changing frequency, Ll, L2, L3, L4
and L5 should be readjusted, since these circuits
constitute the tuning adjustments of the rig. The
principal effect of mistuning L3, L4, and L; will be
lower output or efficiency. The principal effect of mis-
tuning L2 will be degraded sideband suppression.
It is quite important, therefore, to adjust L2 very
carefully. It may be noticed that when large audio
signals are applied, the envelope develops some ripple.
There are two possible causes for this action. The
first is carrier unbalance (carrier shift), and the other
is harmonic distortion in the audio circuits. (It is
assumed that a pure sine wave of 1225 CPS is used
_- o
Z.® ;0. .
4 .
..i'r ..
?..
'11._
r r
.. ..-
Fig 11. Under chassis view of the SSB Jr.
as the input signal.) One may isolate these two
effects by setting carrier balance at high-level audio
operation (where these effects generally are most
pronounced) to reduce the "ripple." With the carrier
ripple (which is easily identified when the carrier
balance controls R16 and R17 are moved) balanced
out, adjust L2 slightly (in conjunction with RIO for
minimum envelope ripple. The remaining ripple
should be less than 5% of the display and is most
probably caused by audio distortion, either in the
audio source or in the audio system of the transmitter.
In observing ripple, the oscilloscope should be syn-
chronized from the 1225 CPS audio signal at a fre-
quency of about 122.5 CPS to show ten cycles or so
of carrier ripple. Unwanted sideband ripple will show
twice as many peaks, and so will second harmonic
audio distortion. Third harmonic audio distortion
will show three times as many peaks, etc. Of course,
all these distortions (and maladjustments) may occur
simultaneously, so a little care and thought is ad-
vised. In the sample SSB Jrs. tested, third harmonic
audio distortion is the principal component, and is
easily identified at high levels.
When feeding a load the total input current will
rise to about 80 MA at full level with a single tone
input. With speech input the current will rise syl-
labically from a resting value of about 60 MA to
around 70 MA, depending on the waveform. Always
use an oscilloscope to determine maximum operating
levels. Overload will cause degradation of the side-
band suppression, and so is to be avoided. Sideband
cancellation adjustments performed at about half
peak level are probably the most reliable ones. Carrier
balance is best made with little or no audio input.
Peak level is the audio signal level which causes
flattening of the peaks due to amplifier overload. A
higher input level can be used when working into a
load, but the overload condition should be avoided
while making adjustments and later, too, when operat-
ing the rig.
The sideband selector switch is used to control
which sideband (upper or lower) is generated. Find
out which switch position corresponds to upper
sideband by tuning the exciter output signal on a
receiver with its BFO supplying carrier. Conduct a
talk test and tune the receiver for normal speech
output. Then tune the receiver to a slightly lower
frequency. If the voice pitch rises, the upper side -
band is being generated. Identify switch positions
accordingly.
It takes about 15 minutes from a "cold" start to
make all the adjustments described here after a little
experience is gained. Do not be frightened away from
single-sideband because of a lengthy description of
the adjustment procedure, since the adjustments are
simple to do, and you will find that the description is
actually very detailed and complete. Another reason
for not being frightened away from single-sideband
is that extremely modest equipment affords the most
reliable 'phone communication yet developed.
9

NOTES ON THE DESIGN OF
THE SSB, JR. RIG
Because the SSB Jr. rig design is made possible
by a new type of phase-shift network, and a new
style modulator, it seemed desirable to have the
designer, W2KUJ, explain these units in further
detail for the benefit of the technically minded
readers of Ham News.-Lighthouse Larry
The SSB Jr. is a superbly simple rig. Such things
just don't happen by accident, however. Throughout
the design many new ideas were employed to save
space and reduce complication while not sacrificing
performance in any way. Easy adjustment for
optimum performance was a foremost point of design.
The phase-shift network is an example of simplifica-
tion of this sort. Literally hundreds of laborious cal-
culations were made along the way to the final solu-
tion. The result is a better performing network that
has only eight parts and is really very easy to adjust
properly. Two methods of adjustment are possible.
The first (and preferred one) has already been ex-
plained in detail. The other one is obvious. Merely
put in accurately measured values and call the job
done. The problem here is to obtain the accuracy
needed (absolute accuracy) since standards of re-
sistance and capacity are obviously of a different
nature. By making adjustments which involve both
resistance and capacitance values simultaneously in
conjunction with a single reference frequency, almost
all sources of error are eliminated. And that is why
the preferred method is preferred. All this accuracy is
wasted, however, if the components used are not
stable enough to hold their values after selection. This
is why precision resistors are specified, and why only
a small range of adjustment is provided by the trim-
mer capacitors, since the trimmers are the most likely
circuit elements to change. In this way good stability
is obtained.
A word about operating conditions necessary for
the phase-shift networks. The outputs must feed very
high impedance circuits. The effective source imped-
ance should be low, and the voltage supplied to A,E
must be minus 0.2857 times the voltage supplied to
D. Incidentally, the voltage output of each section is
equal to the voltage at A,E from zero frequency to a
matter of megacycles. The design center frequency for
the two networks (yes, there are actually two) is 800
CPS. The differential phase-shift versus frequency
curve is symmetrical about this point and holds to
within 1.3 degrees from 225 CPS to 2750 CPS, as
indicated in Fig 12. A slight error in setting the refer-
USEFUL ~WC
200 300 500 E00 1250 2000 3000
rnEOUENCT M O.P.S.
Fig. 12. Audio phase -shift network performance
10
ence frequency (3960 CPS) will result only in shifting
this band up or down by the same percentage. The
operating band is adequate-even desirable-for
voice communication. One need not fear reports of
poor quality when using this rig.
Another simplification which deserves comment is
the balanced modulator used in SSB Jr. Let's take a
few moments to consider what takes place in the cir-
cuit. Fig. 13 shows just one modulator consisting of
two germanium diodes, G1 and G2 with associated cir-
cuits. First, suppose a high frequency signal of a few
volts is applied at point R. On the positive crest of
signal, current passes through G2 into the center
tapped resonant circuit and tends to pull point S in the
same direction. Point T naturally tends to go negative
because of the phase inverting properties of the res-
onant circuit, but, of course, no current flows through
G1. One half cycle later current passes through G, from
the source, tending to pull point T in the negative
direction. But at this time point T would be at a
positive potential because of the "inertia" of the
resonant circuit. The net result of the battle between
G, and G2 to cause current to flow in the resonant cir-
cuit is a draw. No net voltage appears across this
circuit at the source frequency and energy is dissipated
in the balancing resistor and in G1 and G2. Thus far,
we have currents in the resonant circuit, but none at
its operating frequency. This seems like a long way to
go to get nothing, but wait.
Now, let us imagine a bias applied at U. If the
voltage at U is positive, G2 will pass more current into
the resonant circuit, and G1 will pass less current.
This, in effect, unbalances the circuit and a radio
frequency voltage will appear across the resonant cir-
cuit, with point S in phase with the voltage at R. If
the bias voltage at U is negative, G1 passes more
current than G2, and the circuit is unbalanced in the
other direction. Under this condition the voltage at T
will be in phase with that at R. Obviously, if the
voltage at U is an audio frequency voltage, the cir-
cuit is unbalanced in one direction or the other (at an
audio frequency rate) and the resulting radio fre-
quency voltage across the resonant circuit is actually
two sets of sidebands with no carrier. When another
pair of diodes (such as G3 and G, ofFig. 2) is connected
to feed currents into the resonant circuit from related
audio frequency and radio frequency sources respec-
tively 90° out of phase with the first, sideband cur-
rents caused by these signals flow through the res-
onant circuit in such a manner as to reinforce one set
of sidebands and to cancel the other set. The result is
a single-sideband suppressed carrier signal. In the
case of SSB Jr., it is a really high grade one.
The function of the balancing resistors (R,; and
1217 of Fig. 2) is to equalize minor differences in the
characteristics of the diodes and to balance out stray
couplings. Thus, any one balanced modulator is not
necessarily perfectly balanced, but the action of two
such modulators fed with polyphase signals allows a
complete composite balance.
What about operating SSB Jr. in other amateur
bands or at other frequencies, in general? As described,
the radio frequency circuit design is for the 75 meter
(Continued on page 12)
GI
Fig. 13. SSB Jr. modulator circuit

s. .! tr.
Well, the problem my boss had about the bound
volume of Ham News is well on its way to being
solved. At the time this column was written over two
hundred of you had sent in "yes" votes, and when I
presented this evidence to the boss he agreed that
maybe a bound volume was a good idea. So, if every-
thing goes along as expected I'll be notifying you soon
that the bound volume is available. Those of you who
sent in "yes" votes will be notified personally.
I recently had the pleasure of addressing the
Evansville-Owensboro Section of the Institute of
Radio Engineers. My subject was the SSB Jr. rig
described in this issue of the Ham News. D. E.
Norgaard, W2KUJ, had also been invited to talk to
this group, but he was unable to appear, so I made a
wire recording of Don's talk and took it along with me.
Whenever Don or I give talks on single-sideband
we like to demonstrate inverted speech, because it is
so easy to produce with SSB equipment. As you know
inverted speech is that strange sounding stuff that you
hear on the short-wave bands on transoceanic com-
munication systems. At least, inverted speech used to
be used a great deal, although now more complicated
systems of scrambling are employed.
At any rate, you produce inverted speech by taking
an upper sideband, let us say, and placing it on the
low frequency side of a carrier. This can be done on a
receiver by tuning it on the high frequency side of a
so-called upper sideband. The effect is to make low
pitched sounds high in pitch and vice versa. You
should hear the wolf-whistle coming through on in-
verted speech! I can guarantee that you would never
recognize it.
In fact, until you become familiar with inverted
speech it is practically impossible to recognize any-
thing. For example, if you say "General Electric
Company" into an inverted speech system, what
comes out sounds like "Gwunree Oyucktruck Krin-
kino." Conversely, if you say the latter phrase into an
inverted speech system, what comes out sounds like
"General Electric Company."
In other words, you can form a new language, and
if you speak this new language into an inverted speech
system, what comes out is understandable English.
As an example, "metz pee wee" means "nuts to you"
and "eee wye" says "oh yeah." But you can go even
further, as Don and I did. We decided that it would
be nice to be able to recite the poem Mary Had A
F
Little Lamb in inverted speech, and after an hour of
intense concentration we succeeded in the decoding
job.We thought you would like to see this poem in
"Sweeping the Spectrum," so here it is:
Naarow hod O yutty yarng,
Uts feeious yiz yelt uz snee,
Arnd I view hair bop naarow yump,
No yarng yiz sla pee bay.
A word of caution. When practising this poem in
inverted speech language, make sure that you are
alone. People have strange enough ideas of amateurs
as it is!
It doesn't take an editor of a magazine long to
realize that he has a bunch of sharp-eyed readers.
Even though I do know this, every so often something
happens that makes me realize that the Ham News
readers are product -conscious. For example, VE1WM
had a question in the September-October, 1950 Ham
News (page 5) regarding radio interference from
fluorescent lamps. I answered the question at some
length, and referred to a home-made filter which
might be made, consisting of three 0.07 mf condensers
connected in delta.
Just the other day one of my readers wrote me, and
pointed out that two manufacturing concerns make
just such a special condenser, that is, a single unit
which contains three 0.07 mf delta-connected con-
densers. One such concern is Sprague Electric Com-
pany, and their interference filter has the number
IF-37.
This same person continues, and points out, that
the other company making such filters is the General
Electric Company! Oh well, looks like I'll have to
surround myself with more G-E catalogs. The G-E
unit, by the way, carries the number 25F214.
The reason this issue of the G-E Ham News feels
thicker or heavier is that it contains twelve pages.
This is not going to be the standard size of the Ham
News from now on, but in this case, because of the
SSB Jr. rig, I deemed it desirable to give as much in-
formation as possible on this newest of amateur de-
velopments. It is possible that one or two issues a year
may be twelve pages long, if the material warrants
it. --L'iyh toserle Jc 44,
1 1

TECHNICAL INFORMATION
6AG7
GENERAL DESCRIPTION
Principal Application: The 6AG7 is a metal high-
vacuum type power amplifier pentode designed
for use in the output stage of television video
amplifiers.
Cathode: Coated Unipotential
Heater Voltage (A -C or D-C) 6 3 Volts
Heater Current 0 65 Ampere
Envelope: Metal Shell, MT-8
Base: B8-21 Small Wafer Octal 8 -Pin
Base Material: Phenolic
PHYSICAL DIMENSIONS
MAX
NIA 8-6
HAM NEWS
FROM:
The tube is capable of operating at high plate cur-
rent levels and features a high transconductance.
Direct Interelectrode Capacitances:
Grid to Plate 0.06 µµf
Input 13 µµf
Output 7.5 µµf
Grid to Screen (Approx) 5.8 µµf
Grid to Cathode (Approx) 5.2 µµf
Heater to Cathode (Approx) 10.7 µµf
DESIGNER'S CORNER (Cont.)
band, 3850 to 4000 KC. There is no reason, however,
to think that equally successful performance would
not be obtained on 20 or 10, or even on what is left of
160. It's simply a matter of coil design.
The unit pictured in this issue of Ham News was
the second one ever built. Ten minutes after the last
solder joint had cooled down, the rig was perfectly
adjusted and was delivering 5 watts peak power to a
75 ohm dummy load-and I followed the adjustment
procedures described in the article. Maybe it will take
some people a little longer to read the instructions
than it did for me (after all, I wrote them), but.1, 2, 3
procedure really does the job. I didn't peek ahead in
the instructions, either.
If you get one-tenth the fun out of building and
operating SSB Jr. as I did in designing, building and
using it, you are in for the most enjoyment you have
ever had in ham radio.-W2KUJ
1