I did mention electrolytics used as coupling caps, go dry & decrease coupling. A reason to change them, or at least test for correct capacity.

Just completed the voltage tests on the transistors. Q3 Q4 and Q5 all had base 0.6v more than emitter so that checks out as per Ians' advice for silicon. But interestingly the last two transistors Q7 and Q8 germanium had under .2 volts with Q6 testing 131mv and 9mv on base and emitter and Q7 141mv and 9mv. I checked the supply voltage and my lashed up 3xAA battery supply was only outputting 3.8 volts. So I made up a fresh one and tested again. Actually I did't get a chance to take voltage readings. As I was hooking up the wires I felt Q6 and Q7 getting very hot. The output transformer was getting hot also. So I shut down.
Am I on the right track thinking that initially there was not enough voltage difference to turn the output transistors on? Then when I used fresh batteries the output transistors turned on but due to a fault either in the transformer or the transistors themselves they overheated? Or is this symptomatic of something else?

There is no real evidence that the audio transformer is good. As per elsewhere, in respect of a thermistor. It is there as Germanium transistors have a bad habit of getting thermal runaway. They may have now, self destructed.

That transformer need testing & more likely binning. The transistor like those, I believe need bias to cause them to run, unlike tubes if they loose bias, or it goes positive, they become diodes.

Thanks Marcc. I will look for a transformer. I think the LT700 might be a suitable replacement. And I did change out that polarised cap. It tested 1.5μF when I removed it so I pit a 1μF electro in it's place.

Document uploaded to Post 1.

If that was the pastel blue one, there appears to be a10 visible? Is that 10μF, or 10V? If it was labeled 10μF it was a dud.

I think you will find one at Jaycar as well as a Transformer.

No Marcc it was the green one next to the inner end of the ferrite rod. That big blue cap was the first one I changed. It was 100μF. The tall green one (1uf) was the last of the polarised caps. I didn't change it originally as it looked different to the others and I was unsure if an electro would work as a replacement. So now all of the electrolytics/polarised caps have been changed.
Thanks to Brad for posting the link to my schematic on page 1. Taking a pic worked really well and it came out quite clear. Now at least anyone interested in helping me will be able to see what I am banging on about. There are a few question marks where I could not make out the component values on the original and had to take my best guess.
On the transformer. I did do a resistance test on it (not under load) a while back and there were no shorts. The resistances looked logical to me also (they are documented in an earlier post). Are there any other tests I can do to work out if it is causing the overheating? I hope it is OK as the LT700 I mentioned earlier is hard to get in Australia. There are a few on the auction sites but postage is like $25. Jaycar has one that is 1k:8ohm center tapped and it is under $10. Would that work in this set? Also I am resigned to the fact that I have probably damaged the two output transistors. I have read that if you replace germanium transistors with silicon you have to change the base resistors to get the .6 voltage differential needed to switch them on. This is what is stopping me from just changing them out and seeing what happens. Hey Ian. Now that you can see the schematic is there any chance you can advise me on what transistors (maybe BC 547 or 8) would be a good swap and which resistors if any I would need to change? Thanks to all for your help so far.

Finally got the site to work & I could post this: Frustrating.

I suspect that the iron core of the transformer has been sufficiently compromised, that it is no longer viable.

The Germanium transistor's, unlike the Silicon ones had a voltage divider on the base. That would need reworking, to give the correct bias for the Silicon Transistors, which do not suffer thermal runaway. If the output is push pull, that usually requires an NPN and a PNP transistor. As previously noted.

Marcc, your describing a complementary symmetry output pair.
This radio uses a B class push pull output where both transistors are NPN.
JJ

I went ahead and tested both output transistors starting with Q7 as it was the one getting really hot. It was faulty. Tester says it is a diode and not a transistor. Q6 tested OK. I replaced both Q6 and Q7 with BC548's. As per Ian's tip on base/emitter bias I changed out the 1.3K resistor in the supply to both bases to bring the voltage difference up to .6v as the new transistors are silicon. After a bit of experimenting reducing the resistance gradually I found that a 180 ohm resistor gave the following: Base .840v Emitter .165v for a difference of .675v. Not sure how exact this voltage difference needs to be but I am hoping someone will tell me if it's OK or not. I am using a digital multimeter and do not know exactly how accurate it is. In any case the radio still did not work. Yes the howl was gone but no life from the speaker.
Back to the drawing board. I decided to take a deep breath and try to think logically. I must have missed something. So a probe through the circuit starting from the volume pot revealed the following. Clear signal at the base of Q4 which is amplified but becomes a bit distorted at the base of Q5. The signal stops there. Q5 emitter and collector have a very weak barely audible signal even with the tracer at full gain. The collector of Q5 had just a few millivolts on it. Looking at the schematic I would expect to see almost 4volts. So I checked the input transformer winding and found 4volts on one side of the primary ( I think it would be the primary) but just a few millivolts on the other end. I think the winding is open circuit and this supplies the voltage to Q5 collector.
Hopefully one of you guys will have time to look at the schematic and advise me on what I think is going on. Am I right? Should there be something like 4 volts on the collector of Q5? By the way the resistor at the top of the supply rail is 68ohms and not 220 ohms as shown in my drawing.

Yes, Q5 should have several volts on the collector.
Primary of drive transformer could be open circuit, or shorted Q5.
Must be getting close to some results now.
JJ

That should have been: Generalising NPN "or" PNP; Similar to transformer coupled tube PP. However, still think the divider would need reworking and that transformer has issues and needs to go.

Its a pity there is no oscilloscope, albeit a probe might do? To actually see where any signal existing, or fed into the audio is getting lost.

Albeit that my technique takes out some of the fun. I have found in commercial fixing, over decades, that RF, AF, & Function Generators, coupled with an Oscilloscope, a formidable combination for finding lost signal, & distortion.

Thanks for the confirmation Johnny. It gave me the confidence to pull the transformer and sure enough it was open on the primary winding. In a much earlier post I mentioned that I bought a cheap kit radio purely as a source for components. I have already raided it for the volume pot and switch. Lucky for me that kit had a push pull amplifier and so had an input transformer. It was a bit smaller than mine so I had to extend the legs which was a pain but I eventually got it mounted. After I don't know how many hours spent hunched over that circuit board I was rewarded with some noise from the speaker. I tuned a station and I got good reception with no sign of the distortion from earlier on. The set worked much better than I expected. Hugely satisfying. Just a bit of tidying up before I can move on to making some battery terminals from scratch and cleaning up the case. I will also need to find a 70mm speaker for it but at least for now I can get my life back and take my time finishing this one.
Just out of interest I re fitted the original1.3k bias resistor for Q6 and Q7 to see if it had been necessary. The set worked but only just. The audio was horribly distorted and not very loud. So it really did need re biasing. Just as Ian said. I refitted the 180ohm resistor and everything returned to normal. Thanks again to everyone that took an interest. As usual your help was a big factor in my small success.

I assume you replaced the germanium output transistors with PNP, i.e. BC557?

No Ian I used the BC548s. I bought a half dozen a week or two back for I think about 30 cents each. The germaniums were both NPN so I changed like for like. Except I used your advice to re bias the base to emitter voltage and they worked just fine.