Yes, I missed the context change to model 430.

My 429-MA is getting curious.

First, there is a feed back arrangement to the output valve, with one side of the output transformer secondary grounded and the other going directly to the output valve cathode. And second, the output valve is a 6AQ5. (The original N78 cathode would be directly grounded.)

I don't think that the substitution is original, looking at the soldered connections, but the feedback looks original.

So I have to decide whether to return this radio to original, or go with the 6AQ5. They are not direct substitutes looking on line.

Is a 6AQ5 a reasonable substitute for an N78 with circuit changes? Would such a substitute account for the feedback arrangement?

The 430-MA, with the same official back-end valve line-up (6AR7GT, N78), also has a feedback arrangement; they might have decided to apply it to the 429-MA as well. In case it matters, the other, front end, valve is a 6BE6 for the 429 (BC) and a X61 for the 430 (dual wave).

I am about to track the circuit around the 6AQ5 to see what other differences there are to the official circuit.

You are describing the circuit arrangement on the amended version that uses the 6AQ5.

This amended schematic can be found in AORSM Vol. 10, page 73.

If you can't find this schematic I'll post one here.

AWA 429MA Circuit Diagram

OK, so there was an amendment. I don't have that schematic so if you could send would be great. I'll open my email.

Mine has a 7000 ohm output tranny: 6AQ5 looks like it should be 5000 ohm. Is that the case?

Darned couriers! thread answers a lot of questions about output valve variations in the 429-MA.

Hello MonochromeTV - just realised that you said you would post the schematic here - even better for discussion.
So I will re-hide my email.

7000 ohms will be fine for the 6AQ5. It will run cooler and produce more output. (yes I found that out by testing - the ohms specified for an output valve should be considered a minimum rather than optimum).

Robbert is 100% correct. Valves are always happier pushing a higher impedance load, within reason.

Decided to study negative feedback to see if I could figure out why they decided to introduce it with the 6AQ5 on this radio.

Scroggie has a good explanation in Ch 20 on Audio-frequency amplification - a bit too technical really but managed to get the following.

Pentodes are more prone to distortion and negative feed back acts to reduce this for 6AQ5 and the 6AR7GT since the feedback goes from the speaker secondary back to the 6AR7GT. It can also do this with no reduction in output power. The feedback being out of phase with the distortion, counteracts it. Because the feedback is from the output transformer secondary, transformer distortion is also reduced.

Speakers have a tendency to ring at various frequencies, accentuating and prolonging them. A low resistance across the speaker damps this ringing, and the effect of negative feedback is to reduce the output resistance of the pentode, which has a high output resistance compared to a triode which naturally damps ringing. This emphasises Ian's comment above that "Valves are always happier pushing a higher impedance load".

This begs the question as to whether the 6AQ5 is more prone to distortion than the alternatives, or whether the speaker used was prone to it, requiring the introduction of negative feedback.

Has anyone noticed sound issues with this radio? If not it would seem that in introducing negative feedback AWA were anticipating a problem with the valve change, or maybe a speaker change. If there is a problem with sound, say due to the speaker, maybe they were reacting to it.

Feedback can be tricky and may cause inaudible low frequency or high frequency oscillation. So best not to change the layout.

Most radios with miniature output valves had some form of negative feedback, some obvious, some not so. There's nothing stopping the use of negative feedback on older radios, but I suspect the designers were simply trying to get the maximum output rather than worrying about fidelity.

The 6AQ5 was (more or less) a derated version of the 6BW6 which in its turn was a derated version of the 6V6GT. By derated, I mean it works the same, but due to its smaller size it could dissipate less heat, and so was rated at less maximum power.

Output valves of the era were often quoted as having 10% distortion, so the introduction of negative feedback reduced it to a much lower figure.

Paper caps will unlikely be good now and that is why I have put my question to SC. Even if it is a "Love job". Reliability can only be of the highest likelihood if the job is done properly. There is no difference in Standard here, between a personal fixer and a Commercial one.

The whole reason for testing the cap in the Roberts was to provide me with yet another example of a Wax paper cap passed its use by date. I did in a similar exercise with HRSA "Radio Waves", leakage test all of the "Chenex" & some "Ducon" wax paper caps (same company) from radios from & in "the line". Horror story: Could not expect any radio to run well with that lot.

As far as I am concerned, I can replace any of those leaking caps with one that isn't leaking. So if the NP cap is leaking its a goner.

On that I would again point out that a "Peak 200H" meter came my way. It has a cap for measuring OP, when calibrating. A lot of Astor's had a plug just for that. When tested the cap was found wanting & replaced. It is not the only meter / apparatus (not only mine) with old caps where they have failed.

Thanks for the amended schematic- marking this up now and comparing to the original.
One thing -C23 is not there on the circuit or in the parts list. But I found it on my radio - as bold as brass on the top of the oscillator side of the tuning gang, in parallel with C7 & C8.
It is not in the top-side photo with the original circuit.

Re replacement of paper caps, my recollection is that the philosophy in the '80s in Electronics Australia vintage radio articles was to replace paper caps in the high impedance AGC circuit; and in interstage signal paths to grids, where DC leakage was damaging to performance and valves. Also where a HT leakage could conceivably find its way onto the aerial or any other terminal.

Leaky cathode bypass caps could stay in place as long as they were not very bad; after all they were usually paralelled by a bypass resistance.

I suppose that philosophy then applied to collectors who operated radios under their supervision. They might have felt differently when servicing radios leaving their care.

My attitude is fairly ruthless, as the majority of my work on valve stuff is commercial: Rework is something I do not need and is not good for business.

As noted many times Waxed paper & Oil filled caps in radios etc.cannot be trusted. If in doubt I test them. That is in 99.9% of the time a waste of time as the mentioned fail & Mica and Ceramic are rare failures. No resistor should be left unchecked.

Electrolytics are supposed to leak a bit: They are chemical and need the leakage to sustain their insulation. That's why they die if left unused. Cathode bypass caps commonly loose "form" (polarity) which is why I advocate change all electrolytic caps. If they present as a short, there goes the bias & if open, there goes audio quality.

As for coupling caps: Never forget that as applied in most of the valve radios which we see, the grid current is insignificant, so a cap with a resistance below 200meg is going to seriously impinge on bias. Do consider the effect of the majority I see at around 100K, bleeding positive from a Plate onto a control grid. OR One on the AGC loading down the AGC; not only dragging down its voltage, but also dragging down the signal on say the signal grid of a 6A8.

Food for thought.

I agree with all you say Marcc - radios must be megged at 1000V as well.

The amended circuit changes amount to the following:

One side of the output transformer grounded, the other side connected to 6AQ5 cathode; cathode grounding used for original N78 removed. Connection from cathode of 6AQ5 to C18 of volume control tap components; ground for C18 used with original circuit removed. (No feedback connection to previous stage (6AR7) as described by Scroggie -so maybe feedback effect would not be strong).

Other change is to the values of grid biasing train of resistors used for 6AQ5 - 75 ohm at power transformer secondary centre tap changed to 175 ohm, 0.5 Mohm changed to 1 Mohm- 50 Kohm unchanged.

Only other change is addition of 9uμF padder previously mentioned.

A potentially deadly fault on my radio which was only discovered when replacing C20, one of the 24μF filter caps. The positive terminal end of this cap is located very close to the back of the tone control pot which has the two pole mains switch incorporated. This switch switches both mains conductors of the twin lead which connect to the common and 230-260V terminals of the power transformer. The switch terminal on the tone pot connected to the common lead was jammed through the brown cardboard case of the capacitor, shorting the mains primary to the capacitor body and radio chassis.

This probably would not have killed the radio, but might have killed the operator if contacting the case screws.

Otherwise the radio started up well after recapping. The high value resistors had gone high, but only one, the 2.5 Mohm in the AGC circuit was over 20% high (3.7 Mohm). I'll replace some of these next.