AWA: Model PA1003 25w PA Valve Amplifier

AWA was one of the big names in Australian electronics for decades.  Formed as Amalgamated Wireless (Australasia) in 1913, it went on to build everything from radios and communications gear to TVs and commercial audio equipment, so by the time this PA1003 was made in the 1960s, AWA was already a very familiar name in Australian schools, churches, halls and public buildings. 

The PA1003 is very much from that era - a practical Australian-made valve PA, built to do a job rather than win any beauty contests.

AWA MODEL PA1003 25 WATT VALVE PA AMPLIFIER     

Typical of AWA gear from this period, the PA1003 was built for practical public address duties.  It has a high-impedance microphone input, a P.U. input for record players, a tone control for the P.U. channel, and a rear bass cut intended for use with speaker systems that could not cope with much low end (possibly the old-school horn type that was common back then). 

Output is around 25 watts, push-pull arrangement using RCA 7027A valves - these are a higher spec 6L6GC.

This series of AWA PA amps could be ordered with various optional extras. Apparently that even included the power switch.  This one has the labelled hole on the front panel, neatly covered over, but no switch fitted. I guess it was meant to be installed in a place where it would be plugged in and left on for extended periods of time?  

It also missed out on the optional 6 x 4 inch monitor speaker, although oddly enough the monitor speaker switch is still there. The optional radio tuner, on the other hand, is one accessory I am quite happy was never installed.

OUTPUT TRANSFORMER

It has a number of options available for speakers - a few high-impedance 100v lines and the option for a single 12 - 16 ohm speaker.

Output transformer test measurements below.  

Primary

Test voltage: 15.2 VAC

Secondary

Tap span	Measured voltage (VAC)	Turns ratio (Pri:Sec)	Impedance ratio (N²)	Reflected Zp @ 4 Ω	Reflected Zp @ 8 Ω	Reflected Zp @ 16 Ω
Black – Green	0.308	49.35 : 1	2 435	9.7 kΩ	19.5 kΩ	39 kΩ
Black – Yellow	0.666	22.82 : 1	521	2.1 kΩ	4.2 kΩ	8.3 kΩ
Black – Orange	0.687	22.13 : 1	490	2.0 kΩ	3.9 kΩ	7.8 kΩ

The 7027A data sheet lists the following conditions:   An 8 ohm speaker on the black and yellow wires, or black and orange wires, is nearest to these conditions, given the B+ voltage is close to 400V and that the screen voltage is only slightly lower than the B+

DC 12AX7 HEATERS & 7027A CATHODE BIAS

This amp does have one unusual feature: the heaters of the two 12AX7s are used as part of the cathode bias arrangement for the 7027A output valves.  In practice, you can expect around 24V at the 7027A cathodes, with that divided across the two 12AX7 heaters wired in series, so each sees roughly 12V DC.

It is not completely unique.  Some Silvertone amps use a similar arrangement, and it also turns up in certain Fisher hi-fi designs.  One advantage is that the preamp heaters run on DC rather than the more usual AC supply, which in theory can help reduce hum.

There are a few trade-offs, though.  Warm-up tends to be slower, and it also means the 12AX7s are part of the output stage bias network rather than just ordinary heater loads.  In other words, these are not valves you want to hot-swap with the amp running, because removing one changes the bias conditions of the 7027As as well.  This type of setup also limits output valve choice, since the whole arrangement depends on the current draw being in the right ballpark.

Side-mounted inputs for a high-impedance microphone and phono.  Noting that this amp didn't even have a power on/off switch, they probably would have left the mic and turntable permanently plugged in.  Switches on microphones were common back then.

The power transformer is most likely a Ferguson based on looks - they were commonly used in Australian amplifiers.   I can't find it anywhere in their catalogue, so it may have been a special order by AWA, or maybe it's not Ferguson?

I’m sure the output transformer would have been from the same supplier, as above, I can’t find it in any catalogues. 

The cardboard tube capacitor towards the rear did not measure well, and I was sure that it would need to be replaced - after bringing it up slowly on the variac and giving some time to reform, it's fine now.  The Ducon in front measures right on spec.  

The power supply uses a voltage doubler for B+  the cardboard encapsulated capacitor is for insulation purposes (note schematic below).  

Most of the additional holes in the chassis are for mounting optional extras, in this case, the holes are for a radio receiver.  Some holes are for ventilation.

RCA 7027A output pentodes and a Brimar 12AX7.  There's also another pre-amp valve, an ECC83 that the manufacturer's name has worn off (it's in a slip in metal shield).

Plenty of room inside, and it's encouraging to see it full of mustard caps, which are working fine, no leakage or cracks, so they can stay.

 A lot of components are on tag strips, and well spaced out, which is very handy.   The resistors are mostly carbon film and look to be in good shape - can’t see any signs of heat stress.  

The only part that I can see that is broken is the switch for the bass cut, and given that it’s in the negative feedback loop, which I have removed anyway.

AWA PA1003 SCHEMATIC   

The schematic below is very similar to the PA1003 that I have, but not identical.  I actually spent a good deal of time tracing it before I found the schematic below…    anyways, tracing is always a good learning experience.

Mine doesn't have the mic transformer, and there are a few minor differences in values around the mic gain stage.

Further Reading

There's a wealth of information available here, care of Tim Robbins - I believe the schematic above is also by Tim, but I'm not certain of this.   The schematics included are of the PA1003 with his modifications.  

Side note: Tim also has a guide to modding PA heads, Australian ones in particular; the theory probably applies to any weird old vintage PA amp.

AWA PA1003 ADVERTISEMENT

Rebuild Notes

I rebuilt the amp using a typical Fender preamp, with treble, middle and bass controls.  Somewhat similar to a Princeton AA1164 (without the reverb or trem)

• TMB,  T 250k,  M 10k,  Bass 250k.  100k,  250pf, 100n, 22n
• Cathodyne phase inverter
• 10W 250ohm cathode resistor on 7027A, bypassed
• Changed heater wiring to the more regular type, and elevated it from the 7027 cathodes
• Power filtering:  1k / 22u to screens, 18k / 22u to inverter, 18k / 22u to preamp
• Screen: 1W 470ohm resistors added 

In this configuration, the amp has more to give in terms of - the power stage is running conservatively.