This is only a preview of the May 2005 issue of Silicon Chip. You can view 38 of the 112 pages in the full issue, including the advertisments. For full access, purchase the issue for $10.00 or subscribe for access to the latest issues. Articles in this series:
Items relevant to "Voice Recorder":
Items relevant to "MIDI Theremin, Pt II":
Items relevant to "What’s this? Free PC Boards for Schools?":
Articles in this series:
Purchase a printed copy of this issue for $10.00. |
Vintage Radio
By RODNEY CHAMPNESS, VK3UG
The AWA Radiola B29 8-transistor
radio with battery eliminator
Manufactured in the 1960s, the AWA B29 is an
interesting 8-transistor radio that could be run
from batteries or an external power adapter.
T
RANSISTOR RADIOS came in
three basic sizes during the period
that Australian manufacture still existed: small (little bigger than a large
pack of cigarettes), medium (about
the size of a house brick) and large
(about the size of a medium-sized
valve portable).
During this period, a lot of work
was done to reduce the size of the
sets while still maintaining good
sensitivity and audio output. For
example, in medium-sized receivers,
200mm x 12mm-diameter ferrite rods
were used to ensure good sensitivity,
while the batteries were kept small to
keep the size and weight to a reasonable level.
Unfortunately, using small batteries
also meant that they had to be replaced
frequently at some cost. Many of these
portables spent more time in the
kitchen than outside, although they
were also often used for entertainment
at the beach.
Saving batteries
In order to conserve the batteries,
manufacturers had to look at ways
of minimising or even eliminating
battery drain in some circumstances
– eg, when the radio was used in the
Although still functional, the old AWA Radiola B29 was somewhat worse
for wear. It’s a fairly conventional 8-transistor set from the 1960s.
98 Silicon Chip
kitchen. The answer was to provide
a small external power supply that
would allow the set to run off the
mains. A switching contact on the set’s
power socket isolated the battery when
the external supply was plugged in.
The AWA B29 8-transistor radio is
one such set that can be used with an
external power supply (or “battery
eliminator”). It was a medium-sized
receiver weighing 1.6kg without a
battery, or about 1.85kg with its 2364
battery fitted. By contrast, the larger
AWA B32 8-transistor receiver (which
has an RF stage) weighs 3.2kg without
a battery and just under 4kg with its
quite sizable battery fitted.
The two sets draw around the same
current. However, the battery in the
B32 is more than three times heavier
than the B29’s 2364 battery (800g versus 250g) and it provides nearly four
times the operating life.
Although hardly a lightweight at
4kg, the B32 weighed much less than
the mains-powered valve portables
from the 1950s and 1960s. These
weighed as much as 8kg, which made
them rather heavy to move around.
Generally, a high current drain
relative to battery size means a short
operational life. For example, some
small sets used a 216 battery (weighing less than 40g), which meant that
operating times were down to just a
few hours. By contrast, those sets using
four AA cells (52g) did have a longer
operational life.
There were other options, however.
For example, the Kriesler 41-27 used a
286 battery which was reputed to give
around 1000 hours of operation. It is a
weighty portable (approximately 3.6kg
with the battery fitted), although it was
lighter than the AWA B32.
siliconchip.com.au
The B29 is more conveniently sized
and is an effective portable receiver
when using its internal battery. However, it’s still best to use an external
mains adapter to power the set if it is
to be used in the home.
The A&R PS82
Probably the first manufacturer
to provide a battery eliminator
was Philips with their circular
shaped unit. However, other manufacturers soon climbed aboard the
bandwagon.
One such company was A & R
Transformers and their PS82 battery
eliminator was very popular at the
time. This unit was double-insulated
and featured switchable 6V and 9V
(100mA) DC outputs.
One advantage of the PS82 is that
it is easily dismantled for service, requiring just one screw to be removed
to split the case in half. As shown in
Fig.1, the circuit is really quite simple and uses diode D1 as a half-wave
rectifier. The 6V and 9V settings are
selected by switching taps on the
transformer secondary, while two electrolytic capacitors and a 10W resistor
filter the output from D1.
Fitting four diodes in a bridge rectifier circuit would have achieved better
voltage regulation under load, so why
wasn’t this done? The reason is probably to do with cost. Silicon diodes
were much more expensive then than
they are now and so this would have
sharply increased the cost.
The A&R PS82 battery
eliminator featured
selectable 6V and 9V
DC outputs.
Fig.1: there’s not much in the A&R
PS82 battery eliminator – just a
transformer, a diode, a couple
of electrolytic capacitors and a
switch to select the transformer
taps.
AWA B29 circuit details
The circuit of the AWA B29 is conventional for the era – see Fig.2. First,
there is a large ferrite antenna rod and
this has three windings: (1) a tuned
winding for RF signal pick-up; (2) a
transistor base feed winding; and (3)
a winding that can be interfaced to an
external antenna. Coil L1 connects to
this third winding and acts as a loading coil to boost the receiver’s performance with a relatively short antenna
(eg, a car radio antenna).
The autodyne converter stage uses
a 2N1639 or a 2N1636 germanium
transistor. Following this is a 2-stage
neutralised intermediate frequency
(IF) amplifier using either 2N1638 or
2N1634 transistors. The neutralising
capacitors are C11 and C17. Note that
the IF transformers (TR3-TR5) each
have only one tuned winding, their
low impedance coils coupling to either
siliconchip.com.au
the next transistor base or, in the case
of TR5, to the detector diode (MR1).
The automatic gain control (AGC)
voltage from the detector diode is
applied to the base of the first IF transistor (VT2). For medium strength
signals, AGC control only occurs in
the first IF stage. As the AGC voltage
increases, the transistor draws more
current rather than less as in a valve
circuit.
This also means that as the AGC
voltage rises, the voltage on VT2’s collector drops and this changes the bias
on VT3 (2N406) such that it begins to
draw current. As a result, VT3 functions a variable shunt across the tuned
circuit in TR3, which reduces the
signal applied to the first IF amplifier.
It’s no exaggeration to say that this
type of AGC circuit is foreign to most
who have only been involved with
valve AGC circuits. Note too that the
circuit is drawn with a positive earth,
which makes it just that much harder
to follow.
The detector stage is followed by
a conventional 4-transistor amplifier
circuit, with transformer TR6 driving a
push-pull class B output stage (VT7 &
VT8). There is no output transformer,
the output transistors directly driving
a centre-tapped 80-ohm loudspeaker.
Because the speaker is so unusual, you
need to keep your fingers crossed that
it never needs replacement!
By the way, germanium transistors
are rather sensitive to temperature
and voltage variations. If these factors increase, the output stage draws
May 2005 99
more current, which increases the
temperature, which leads to a further
increase in current and so on. Unless
precautions are taken, this can lead to
a condition known as thermal runaway
and result in the destruction of the
output transistors.
In this circuit, the current through
the output stage is stabilised by a
network consisting of R25, R26, R28,
R29 and thermistor TH1 (AS2). In
operation, TH1 monitors the temperature of the output transistors. As the
temperature goes up, the thermistor’s
resistance goes down and this reduces
the forward bias on the output transistors. As a result, they draw less standing current and so the current through
them is kept to a safe limit.
Finally, note that because the set
has a positive earth, the plug from the
external power source must have its
centre pin (tip) as the negative output.
Most external supplies (including
plugpacks) have the centre pin wired
as the positive rail, so take care here.
Repairing the cabinet
Fig.2: the B29 receiver uses a fairly conventional superheterodyne circuit
with eight transistors. The audio output stage is a little unusual though,
as it drives a centre-tapped loudspeaker coil – see text.
100 Silicon Chip
The B29’s case measures 235mm
x 120mm x 70mm and is made from
brown leather, cardboard and plastic,
with a metal front. The front of the set
is similar to the B79 described in the
December 2004 issue.
As can be seen in one of the photographs, the carrying handle had come
adrift from the lugs on the top of the
case. This method of attaching the
handle was obviously inadequate, so I
looked at making some improvements
here. Unfortunately, this involved
dismantling the set and that wasn’t
all that easy.
First, the handspan tuning knob
and the on-off-volume control knob
were removed. However, I couldn’t
persuade the dial pointer to come off
– it is a tight friction fit to the tuning
gang shaft.
Next, working from the back of
the set, the five screws that secure
the PC board in place were removed.
That done, the loopstick antenna was
eased out of its clips and the four nuts
(one at each corner of the cabinet
at the front) were removed using a
small socket wrench (spintite) on an
extender shaft
By this stage, everything was “flopping” around inside the case and so the
dial-scale was now worked backwards
and forwards while I tried to remove
the PC board assembly. It eventually
siliconchip.com.au
Removing the “works” from the cabinet isn’t all that easy, the entire process
taking around 20 minutes. Note the large loopstick antenna which contributes
to the set’s good sensitivity.
came free and I was able to extract
the PC board, the chassis and the rod
antenna from of the case.
This gave access to the four bifurcated rivets that hold the handle in
place. These rivets (along with the
handle) were then removed.
That done, I cut some light-gauge
galvanised flashing (try your local
hardware store) into two small rectangles. These were then bent into
semi-circular straps, after which two
holes were drilled in each strap to
match the holes in the leather handle
assembly.
The idea behind these metal straps
is that they would take all the pressure off the leather handle. The accompanying photographs shows the
basic scheme.
Re-attaching the handle assembly
to the case took a little while but the
scheme worked well. It may not look
100% due to the damage that has been
done to the various parts over the years
but it’s a lot better (and stronger) than
it was.
The case was also looking a bit shabby, so the next step was to give it a good
clean. First, the leather was washed
with soapy water on a cloth and then
put aside until it was completely dry. It
was then polished using silicone wax
and came up rather nicely.
siliconchip.com.au
The front panel of the set was also
cleaned using soapy water, this time
with the aid of a nail brush. However,
this job had to be approached with a
good deal of caution, to avoid splashing water onto the paper dial scale.
The front panel was then polished, by
which time the receiver was starting
to look rather good, even in its disassembled state.
The knob and handspan dial were
also given a wash, ready for the final
assembly. However, that would have
to wait until the circuit had been
checked out.
Getting it going
It was time for the smoke test. After
making sure that the PC board was
properly isolated, I connected a power
supply and switched the set on. It immediately showed signs of life but the
volume control was very noisy.
A quick spray with a suitable
cleaner solved that particular problem. I then twisted the tuning control
shaft with my fingers and a number
of stations were heard. I closed the
tuning gang and found to my amazement that the set was tuning down to
around 495kHz. And with the gang
fully open, it tuned all the way up to
about 1700kHz.
However, the set was never really
designed to cover this tuning range.
Someone in the past had adjusted the
various coils and trimmers so that it
covered this range and then flooded
the coil cores with beeswax. I was able
to readjust the IF coils but the oscilla-
VALVES
AUDIO HI-FI
AMATEUR RADIO
GUITAR AMPS
INDUSTRIAL
VINTAGE RADIO
We can supply your valve needs,
including high voltage capacitors,
Hammond transformers, chassis,
sockets and valve books.
WE BUY, SELL and TRADE
SSAE DL size for CATALOGUE
ELECTRONIC
VALVE & TUBE
COMPANY
PO Box 487 Drysdale, Vic 3222
76 Bluff Rd, St Leonards, 3223
Tel: (03) 5257 2297; Fax: (03) 5257 1773
Email: evatco<at>pacific.net.au
www.evatco.com.au
May 2005 101
The next problem was that the audio sounded rather distorted at high
volume, so I starting checking out the
audio amplifier. I soon found that the
supply rail dropped with increasing
volume, getting down as low as 7V.
This is lower than I would have liked to
have seen and indicates that a full-wave
bridge rectifier in the power supply
would have been a good idea.
As an experiment, I tested the set
with a small regulated supply set to
9V. This gave greater output and less
The handle assembly on the old B29
was repaired and strengthened using
two small metal brackets (above and
right). The restored receiver looks good
and gives reasonable performance.
tor coil is well and truly sealed and so
there is no way of adjusting the bottom
end of the tuning range. When I obtain
a surplus coil from another similar set,
I’ll replace it.
At the other end of the scale, the
top frequency was adjusted to around
1640kHz with the tuning gang fully
open. I then peaked the antenna coil
near the bottom end of the dial (at
about 600kHz) by sliding the tuned
winding along the ferrite rod for
best reception, after which the set
was tuned to about 1500kHz and the
antenna trimmer adjusted for best
reception.
The set was now performing quite
satisfactorily although like most sets of
that era, the RF transistor isn’t exactly
quiet. As a result, the set is a bit “hissy”
on the weaker stations.
distortion but I still felt that there was
something wrong with the output
stage.
Next, I checked the voltages on the
collectors of the 2N217S and they
were nearly identical (ie, with the
set delivering a reasonable volume).
I then fed a 1kHz modulated signal
from a signal generator into the set
and checked the voltages across R28
and R29. They varied equally with
increased volume but there was still
noticeable distortion.
The old CRO
The original 2364 battery type used in the B29 is no longer available but
six AA cells in a holder can be used instead.
102 Silicon Chip
It was time to bring out the heavy
artillery, so I fired up my CRO to track
down the source of this distortion.
This showed that the audio was a beautiful sinewave up to the bases of the
2N217S output transistors but when
I put the probe onto either collector, I
got a horrible looking waveform.
I had expected to see a half sinewave
but that wasn’t the case. I then tried
disabling one half of the output stage
siliconchip.com.au
by getting rid of the signal alternatively
to each base but it still looked terrible
and I was getting nowhere fast!
My next step was to use the dual
inputs of the CRO. By placing a probe
onto each section of the output, I could
observe the phasing of the output stage.
It still looked terrible but at least the
two waveforms showed that the problem occurred in each transistor output
on each half wave. I then decided to
use the add function and what did I
see? – a beautiful sinewave.
I’d been trapped by the rather
unusual nature of the circuit. If the
speaker had been fed from a centretapped transformer, the CRO would
have shown a sinewave on the voice
coil. However, the system used in this
set is different, as the output goes to a
centre tapped choke (ie, the speaker)
and the waveforms are different.
So what was causing the distortion?
If I’d thought about it, I would have
realised what the answer was – this
is a personal portable with low-output
transistors feeding the speaker so I
was really expecting too much of the
audio stage.
Putting it back together
Having solved that problem, it was
now time to reassemble the receiver.
I won’t go into all the problems I ran
into but it took me no less than 20
minutes to do the job.
By contrast, I remember when the
local film processing laboratory was
giving a radio away with each film developed. These radios must have cost
them next to nothing but they were still
rather nice little 5-transistor sets. And
they could be dismantled and reassembled in about one minute!
That said, the fully assembled B29
now looks good and performs quite
credibly. And with an external antenna and earth, it really works very
well indeed.
Substitute battery
Type 2364 batteries are now unobtainable, so I decided to see if six Ccells could be fitted into the receiver’s
case. Unfortunately, the holder was
just a little too large but a pack of six
AA-cells in a holder can be fitted and
will give reasonable life. It is only
necessary to solder a battery snap lead
to the 2364 battery plug and the B29
is ready to use as a portable.
To prevent any shorts, a couple of
pieces of plastic tubing were placed
siliconchip.com.au
Photo Gallery: Precedent 5-Valve Receiver
Designed and manufactured by Firth Brothers, Melbourne, the
Precedent c1933 was a stylish table model receiver that tuned the
medium-wave band. The valves fitted were as follows: 57 autodyne
mixer; 58 IF amplifier; 57 anode bend detector; 2A5 audio output;
and 80 rectifier. Photo: Historical Radio Society of Australia, Inc.
over the exposed lugs on the 2364 battery plug. The 6-cell pack fits snugly
inside the case, as can be seen in the
photograph.
Summary
The B29 is quite a nice little set,
rather typical of many sets of the same
era. Its performance is good (although
the audio suffers as the battery voltage
drops) and the handspan dial system
works well.
In summary, it is a pleasant little
transistor set which I’m quite pleased
SC
to have in my collection.
A Few Gremlins
A few gremlins crept into the April
2005 column, as follows:
(1) Page 80, third column, end of
paragraph four should read: “This
feeds one section of a 6SN7-GT as
the second stage, while the second
section acts as a phase splitter . . .”
(2) The AGC bypass capacitor at
the extreme left of the circuit (Fig.1)
should be 47nF not 47µF.
(3) The asterisks marking the com-
ponents replaced were omitted. The
components replaced were: all the
electrolytic capacitors, the AGC and
audio coupling capacitors (except the
coupler to the grid of V7), the screen
dropping resistor feeding the 6U7-G
valves and the accompanying screen
bypass capacitor.
(4) V4’s cathode should be earthed
and not connected to the 25kW resistor, volume control or the 100pF
capacitor. The same error is on the
amended AGC circuit (Fig.3).
May 2005 103
|