This is only a preview of the January 2012 issue of Silicon Chip. You can view 26 of the 104 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. Items relevant to "A Stereo Audio Compressor":
Items relevant to "Build A Simple AM Radio":
Articles in this series:
Items relevant to "3-Input Stereo Audio Switcher":
Items relevant to "Playing USB-Stick & SD/MMC Card Music Without A PC":
Purchase a printed copy of this issue for $10.00. |
Vintage Radio
By Rodney Champness, VK3UG
The AWA PA1002 50W
PA Amplifier
Boasting a power output
of 50W RMS, the AWA
PA1002 is a typical valve
PA amplifier from the
mid-1960s. It’s a rugged,
well-made unit with many
useful features and is
easy to troubleshoot and
restore.
T
HE VERY FIRST PA systems
used a straight tapered funnel arrangement, referred to as a bull horn,
megaphone or loudhailer, without any
form of electronic amplification. They
certainly “reinforced” the user’s voice
to some extent but they were quite
directional.
When valves later became readily
available, PA amplifiers capable of a
few watts were quickly developed.
Those early units were used with carbon microphones, while the speakers
were larger, more powerful versions
of the horn speakers used with early
receivers. Horn speakers were very
efficient compared to modern loudspeakers but their frequency response
was quite restricted and the distortion
from these units was relatively high.
As a result, early PA systems were
only suitable for voice communication.
PA amplifiers of various output
powers were later used during World
War II, some so big that they used
radio transmitter valves in their output
stages to feed the loudspeakers. By
contrast, the commercial PA amplifiers
90 Silicon Chip
used for the general public following
the war were quite modest, with output powers ranging from around 5W
up to about 25W.
As an aside, when I first worked in
the radio service industry in the late
1950s, my employers hired out a PA
system. This used a 25W amplifier
(similar to the unit described here)
and was teamed with a dynamic microphone, a record-playing turntable
and up to four reflex-horn speakers.
The speakers could be located some
distance from the amplifier, as the
output was fed to a 100V (medium
impedance) speaker line.
The connecting cable consisted of a
length of twin-lighting flex which had
low losses at the speaker impedances
used. Standing under the speakers
when the unit was operating at full
power was a deafening experience
and it could be heard up to 3km away
(depending on wind direction).
AWA PA1002 PA amplifier
My vintage radio collection includes
several valve PA amplifiers, the largest
of which is the AWA PA1002, a 50W
unit from the mid-1960s. PA amplifiers
improved considerably after World
War 2 and top-end 50-100W valve PA
amplifiers quickly reached the pinnacle of their development.
Unfortunately, I had no luck in
obtaining a circuit diagram for the
PA1002 although I do have circuits of
other AWA PA amplifiers of the era.
AWA obviously experimented with
many different circuit designs because
none of the circuits I have are anything
like the PA1002. As I result, I eventually traced out the output and driver
circuits of the amplifier and this revealed a push-pull output stage that’s
quite different to what’s normally
expected. In addition, the chassis
has holes punched for two additional
output valves, so that a 100W version
could be manufactured using as much
common circuitry as practical.
The amplifier itself weighs a hefty
14kg without any accessories and it
measures 430 x 230 x 230mm (W x D
x H), not allowing for the knobs and
mounting feet. A feature of the unit
is that there is plenty of room on top
of the chassis for accessories, such as
siliconchip.com.au
a monitor speaker and a radio tuner.
These accessories are not fitted as
standard but could be easily added as
their connection sockets are already
wired.
As can be seen from the photos, access to the parts for service (or restoration) is relatively easy. To gain access
to the valves, it is only necessary to
remove four screws from the ends of
the chassis that hold the U-shaped perforated steel cover in place. It is then
quite practical to carry the amplifier by
one or both of the metal rods located
at the top of the cabinet that attach the
front and back panel to the chassis.
The circuitry under the chassis is
accessed by turning the unit upside
down and then removing the four
self-tapping screws holding the rubber
buffers and the bottom shield in place,
followed by the baseplate itself. It’s
then quite easy to access most of the
parts although some parts are tucked
in under a ledge at the back.
The bottom of the cabinet is fitted
with rubber anti-scuff buffers and
there are also ventilation holes in the
bottom cover sheet.
The patch terminals on the rear panel allow the amplifier’s output to be
matched to many different impedances for both 100V and 75V lines and
are connected according to the table. In addition, there are two line output
terminals plus antenna and earth terminals for the optional tuner.
screwdriver adjustable preset controls:
(1) signal limiter adjustment; (2) bias
adjustment for the V4 output valve; (3)
bias adjustment for V5; and (5) an AC
balance control to null any residual
hum in the audio output.
Finally, at extreme right is another
screwdriver adjustable control. This
is simply a 2-position switch which
is labelled “Audio Bass Full/Cut”. It
very effectively removes any residual
hum in the output.
Inputs & controls
The PA1002 has inputs for two microphones and these are plugged into
sockets on the lefthand end panel of
the chassis. Another socket adjacent
to these connects to the record player.
The various controls are conveniently located along the bottom edge of
the front panel and (from left) include
a volume control for each microphone
plus another volume control for a
record player or the optional radio
tuner. These latter two units are selected by a phono/radio switch.
To the right of this is a tone control
for the phono/radio channel, a small
pilot light (for power) and an optional
mains on/off switch, although the latter is not fitted to my unit (the hole
location is “blanked off”). The final
control is the monitor on/off switch
but in my unit, someone in the past
has rewired it to serve as the mains
on/off switch.
In fact, none of the available accessories are fitted to my particular
amplifier. However, the top section of
the front panel has provision for their
controls. At the left is a blanked-off
hole that’s intended to accommodate
the dial for the radio tuner but the
purpose of the blanked-off hole in the
centre of the panel is a mystery. The
siliconchip.com.au
Circuit details
An old exponential horn speaker as
used in many PA systems during the
1960s. Similar speakers are still used
in modern PA systems.
perforations on the right are for the
optional monitor speaker.
Rear panel facilities
A patch panel on the rear allows
the amplifier’s output to be matched
to many different audio line impedances. There are 12 patch terminals in
all and these are connected in various
combinations according to a table.
There are also a couple of terminals
and a couple of sockets to which the
speaker lines are connected. Another
two terminals are provided for an antenna and an earth for the tuner.
The bottom edge of the rear panel
carries a number of controls and other
features. At the left is the mains power
lead, followed by a 1A mains fuse and
then a 500mA high-tension fuse. An
adjacent covered panel conceals four
Fig.1 shows the output section of
the circuit. It’s quite conventional in
many areas but as mentioned above,
the push-pull output stage is different
to other circuits.
The two microphone outputs are fed
to each half of a 12AU7 twin triode
which amplifies these signals. Its
outputs in turn drive the input of an
EF86 signal limiter via level control
potentiometers and mixing resistors.
The signal from the radio tuner or the
record player is also fed to this stage,
again via a potentiometer and mixing
resistor.
The EF86 signal limiter operates
as follows. First, a small amount of
the audio output is used to drive a
small lamp and this illuminates an
LDR (light dependent resistor) sealed
inside a small lightproof tube. The LDR
is connected across the grid resistor
of the EF86 (see Fig.1), so the actual
grid resistance decreases as the audio
input increases.
Basically, the EF86’s grid circuit
has the LDR and its grid resistor in
January 2012 91
3
1
4
7027A (V5)
4 F
470
8
4.7k
5
6
33k
220k
100nF
33k
–65V
BIAS
5k
2 F
4.7k
4.7k
33k
5k
2 F
220k
7
2
3
8
1
6
1/2 12AU7
470
8
5
6
4.7k
100nF
4.7k
220k
220k
4 F
+525V
HT
4 F
4 F
220k
220k
3.3
44
37
SPEAKER
LINES
44
37
OUTPUT
TRANSFORMER
4.7k
1
4
3
7027A (V4)
33k
Output transformer
100nF
1M
3
ELECTRONIC OUTPUT LIMITER
AUDIO
INPUT
LDR
EF86
68k
1.2M
2
1
10nF
92 Silicon Chip
2.2k
HT
1M
1/2 12AX7
100nF
2.7k
1/2 12AU7
Fig.1: the output stage of the
AWA PA1002 PA amplifier.
The 12AX7 serves as a
phase splitter and this
drives a twin-triode 12AU7
amplifier stage. The 12AU7
in turn drives a push-pull
output stage based on
7027A beam power tetrodes
and an output transformer.
The circuit segment at
bottom left shows the signal
limiter.
series with the resistors in the inputcombining network. As a result, the
LDR-resistor combination acts like an
electronic volume control and even if
By contrast with this arrangement,
some PA amplifier circuits use an
AGC voltage to control the EF86, even
though it is a sharp cut-off valve. This
involves altering the operating conditions for the EF86.
The next stage consists of a twintriode 12AX7, with the first section
used as an audio amplifier and the
second section as a phase splitter. A
twin-triode 12AU7 amplifier stage
follows, with one section driving
valve V4 and the other driving valve
V5, both 7027A beam power tetrodes
which form the push-pull output pair.
Most amplifiers of this type use the
phase splitter to directly feed the two
output valves in push-pull. However,
in this amplifier, a higher input level is
required for the output stage than such
an arrangement could provide, as the
output valves have quite a high negative bias applied to their grids (about
-56V each). Hence the inclusion of the
12AU7 twin-triode stage.
The 7027A valve is an upgraded
version of the older 6L6G and when
used as a pure class-A single-ended
amplifier, its operating conditions are
almost identical. In this circuit, the
bias for each valve is obtained separately from the bias supply so that it
can be adjusted for performance variations between the individual valves.
someone bellows into a microphone,
the amplifier will not be overloaded
(assuming that the signal limiter is set
up correctly).
The output transformer has a total
of eight windings: four to match the
output to the speaker line impedance
plus another four in the plate/screen
and cathode circuits. Each cathode
winding has a resistance of around
44Ω, while the plate windings are
about 37Ω.
It appears that the circuit has been
designed with considerable negative
feedback to stabilise it against the
wide impedance variations reflected
back into it from the long speaker
lines. It’s certainly an unusual circuit
but it works very well. However, I do
wonder why they went to so much
trouble when a more conventional
circuit would have done the job, as
with many of AWA’s other designs.
The power supply (Fig.2) uses a
solid-state voltage-doubler circuit
and this has an output of +525V DC
with the high mains voltage (250V) at
my place. The bias is obtained from
a separate half-wave solid-state rectifier, which gives a maximum of -65V
DC. As well as the windings shown in
siliconchip.com.au
Fig.2, the transformer also has a 6.3V
heater winding with two 47Ω resistors
across it to give an artificial centre-tap
to earth, ie, one end of each resistor
goes to the chassis.
Finally, the unit includes a circuit
that artificially injects hum into the
amplifying chain to cancel out and
minimise hum in the output.
Restoration
Three badly overheated resistors
were the most obvious fault when I
removed the bottom cover from the
amplifier. Two were on the screens
of the 7027A output valves, so it was
quite possible that these had also been
damaged after such abuse.
This sort of problem often occurs if
the speaker transformer primary winding goes open circuit, which removes
the high tension (HT) from the plate of
the output valve. The screen then tries
to function as the plate and the valve
is then often ruined due to excessive
power dissipation in the screen circuit.
Most of the inter-coupling capacitors are polyester types so I didn’t
expect any problems with these. In
fact, it was unlikely that there would
be any major problems in the early
stages at least, as nothing looked or
tested faulty with the power off.
The output stage was a different
story though, due to the badly-burnt
resistors in the output valve screen
circuits. To troubleshoot this problem,
I decided that it would be best to remove the output valves and check all
the voltages and components around
that stage with power applied.
Insulation checks
Before applying power, I first needed
to confirm that it was safe to do so. I
also needed to check and reform the
electrolytic filter capacitors on the HT
line. These checks involved using a
high-voltage insulation tester to check
for leakage between the mains wiring
and chassis and between the HT line
and chassis.
These checks proved satisfactory.
The mains-to-chassis resistance exceeded 200MΩ, while the HT-to-chassis resistance was greater than 50kΩ.
With these checks complete (and
the output valves removed), I applied
power for a second or two then turned
it off for a few seconds. This allowed
me to check how quickly the voltage
built up and decayed on the HT line.
This step was then continuously
siliconchip.com.au
These above-chassis views show the simplicity of the layout. The two large
vacant holes are for additional output valves, as used in the 100W version.
500mA
FUSE
+525V HT
NO LOAD
230V
MAINS
–65V
BIAS
Fig.2: the HT rail is
produced by a solidstate voltage-doubler
and this can produce
up to about 525V DC
(if the high mains
voltage is high). The
-65V bias supply
is derived via a
separate winding on
the transformer and
a half-wave rectifier.
January 2012 93
This view shows
the two badly overheated resistors
on the screens of
the 7027A output
valves. Fortunately,
the cathode resistors had gone open
circuit and saved
the output valves
from damage.
Silicon Chip
Binders
REAL
VALUE
AT
$14.95
PLUS P
&
P
These binders will protect your
copies of S ILICON CHIP. They
feature heavy-board covers & are
made from a dis
tinctive 2-tone
green vinyl. They hold 12 issues &
will look great on your bookshelf.
H 80mm internal width
H SILICON CHIP logo printed in
gold-coloured lettering on spine
& cover
H Buy five and get them postage
free!
Price: $A14.95 plus $A10.00 p&p
per order. Available only in Aust.
Silicon Chip Publications
PO Box 139
Collaroy Beach 2097
Or call (02) 9939 3295; or fax (02)
9939 2648 & quote your credit
card number.
Use this handy form
Enclosed is my cheque/money order for
$________ or please debit my
Visa Mastercard
Card No:
_________________________________
Card Expiry Date ____/____
Signature ________________________
Name ____________________________
Address__________________________
__________________ P/code_______
repeated for a couple of minutes to
reform the electrolytic filter capacitors
on the HT line.
If these capacitors charge quickly
and then slowly discharge over a few
seconds, it indicates that they have
reformed correctly. In this case, it
all worked out as expected so I reapplied power and checked the cathode
voltages on the output valve sockets.
They were both over +100V instead
of the correct 0V, which wasn’t good.
With the unit now disconnected
from power, I checked the cathode
resistors and found that they were
both open circuit. In addition, both
of the 2µF capacitors in the grid-tocathode circuits were leaky, as were
the 100nF coupling capacitors from
the two 12AU7 valve plates to the
respective grids.
Surprisingly, the screen resistors
were still intact and measured 4.7kΩ
and 5.6kΩ. In the end, I replaced all
these resistors and capacitors. The
original capacitors were paper and
metallised-paper types and it was
known that they eventually became
leaky. In fact, it would have been better if AWA had used the more reliable
polyester or polystyrene capacitors in
this section of the amplifier, just as
they had done in the front-end.
The other burnt resistor was in the
feedback network and this was also
replaced. The set was then wired on
the speaker patch panel for a 12Ω line,
to match my triple-cone test speaker.
That done, I applied power to the set
and adjusted the bias potentiometers
to give about -50V at each grid pin.
The smoke test
It was now time to find out if the
94 Silicon Chip
output valves had survived the rough
treatment meted out by the faulty components. As a result, these valves were
reinstalled and the power applied. The
bias voltages remained roughly the
same as before and both valves were
drawing some current, determined by
checking the voltage drop between
each cathode and chassis.
At this point, I adjusted the bias pots
until I had 16V between the cathodes
and chassis. The fixed bias was now
around -37V, which meant that the
total bias was about -53V with respect
to chassis.
The current drawn by each valve
under these conditions is 32mA,
which means that the dissipation in
each valve is around 16W. This is well
within the 35W plate dissipation rating for the 7027A valves. I don’t know
what cathode current AWA intended
these valves to draw with no signal
but 32mA and 16W dissipation seem
to be safe values and the distortion in
the output was low.
It’s strange that the bias adjustments
are hidden behind a plate on the rear.
After all, it’s also necessary to remove
the cover under the chassis to gain
access to the valve cathodes in order
to measure the bias voltages. Surely it
would have been more convenient to
mount the bias adjustment pots adjacent to each cathode circuit?
Getting back to the restoration, the
pots were all given a spray with Inox
contact cleaner/lubricant to eliminate noise. The valve sockets were
all in good order and did not require
lubrication.
Once all that work had been completed, the old PA amplifier worked as
normal. Some of the valves are a little
siliconchip.com.au
noisy and there is a little induced hum
but considering the gain of the amplifier, it really is quite acceptable. In this
unit, the grid circuits are high impedance and these can be susceptible to
induced hum from mains voltages.
Fortunately, it appears that the output valves suffered little in the way of
damage and there’s no point replacing
them. They were probably saved from
destruction by the 470Ω cathode resistors going open circuit.
It’s probable that quite a few of these
AWA PA1002 PA amplifiers were affected by leaky capacitors around the
output stages. One website that I found
has an under-chassis view of one of
these amplifiers and the same resistors
were burnt out in that unit as well!
Cleaning up
Like most PA amplifiers of the era,
the PA1002 is housed in a steel cabinet
painted a grey hammer-tone colour.
The U-shaped cover on my unit had
a number of marks on it, while the
front and rear panels were relatively
unmarked. The chassis itself is also
far from pristine, with discolouration
largely due to the ravages of oxidation
on the plating.
I tried cleaning the chassis with
a small paintbrush and a kerosenesoaked rag but there was little improvement in its appearance. It’s quite
typical of the deterioration that occurs
with plating over time, especially in a
slightly hostile environment.
The valves were relatively clean but
I like them to look like new if possible.
To achieve this, I first washed all the
miniature valves in warm soapy water,
taking care not to rub the type numbers
off them. By contrast, the octal valves
need to be cleaned more carefully;
they cannot be immersed completely,
otherwise water may get into the base
and create electrical leakage paths.
For this reason, I only immerse the
glass envelopes and not the bases of
octal valves in the water and gently
rub the envelopes with soapy water
using my fingers. The valves are then
rinsed in clean water and left to dry by
either supporting them upside down
(so water doesn’t run into the bases) or
by laying them on their sides.
By contrast, the miniature types
are left to dry by standing them on
their pins.
The U-shaped cover was restored
by first sanding it down to remove
any rust and discolouration in the
paintwork. It was then given several
light coats of a grey hammer-tone paint
(called “Galmet”) on the top and sides.
It ended up looking as good as new.
The front and rear panels were
tidied up by spraying them with Inox
and then wiping away the excess. This
removed most of the minor blemishes
and these panels now look quite acceptable. I also sprayed and wiped
the knobs and they came up looking
like new.
The third overheated resistor was in
the feedback network and was easy
to spot.
Summary
The PA1002 is a well-built PA amplifier and is still capable of providing
years of reliable service. The output
stage is rather unusual but both it and
the signal limiter stage work well. As
expected, the cabinet is rather utilitarian but it’s rugged and provides good
protection for the internal parts.
My tests indicate that the PA1002
gives its rated output with little hum.
The only thing that I can criticise is
the use of those paper and metallisedpaper capacitors around the output
stage. As stated above, these eventually become leaky and cause serious
All the faulty parts were in the pushpull audio output stage. They included
five resistors and four capacitors.
problems. Of course, it’s possible that
polyester capacitors were unavailable
in the values needed at that time,
which is why the less-reliable paper
types were used.
In my case, I was lucky that the
cathode resistors had gone open circuit and saved the output valves from
destruction. That said, the unit was
easy to troubleshoot and restore and
SC
is a good performer.
Major Vintage 1920’s-1960’s Radio Auction: Sunday 26 February 2012 at 12 Noon
Under instructions from Garfield Wells, past secretary of the NSW HRSA: an entire collection spanning 50 years – 200 rare
Bakelite, Coloured, Wooden, Mantel and Cathedral radios including Art Deco and “Empire State” models. Original and in
working order. Also rare valves, collection of vintage horn speakers including AR15’s, RCA’s and Brown models.
Absentee, telephone and “live” internet bidding available. On view: Saturday 18 February 9am - 12 noon; Saturday 25
February 9am - 12 noon; Sunday 26 February from 9am or by appointment. Catalogue on line from Saturday 18 February
Enquiries: Phillip Thomas 9552 1899
Raffan Kelaher &Thomas Pty Ltd – FINE ART & GENERAL AUCTIONEERS & VALUERS
Member of the Auctioneers & Valuers Association of Australia
42-48 John Street, Leichhardt www.rkta.com.au
Note: above pictures are for illustration only and do not necessarly depict items for sale
siliconchip.com.au
January 2012 95
|