Oh look I'm sure you can tell I have no formal electronics knowledge or training so I will leave the "hows and whys" of operation to someone who does.

I can observe that the internals are original though and that the valve sockets contain the valves the designer intended (because they are labelled), so logic tells me that either this thing never worked, or it did, and does.

I will post some better images of the chassis shortly and you can see what you think. I will also label the caps in the image with their values,

Back to the caps. I too would have expected them to be very common but I seem to be getting shot down at every turn.
Evatco, for example, clearly show the items on their website, yet send me this response;

"Hi Ash, an unusual order to say the least.. not to many of those values are carried but information below..
regards
Kev"

The information showed they could provide about half what I needed..

In terms of the ratings, again I am simply working from the original values and assume that if I go higher, I cant go wrong.

Originals as below.

0.1μF - 400/600 mixed
0.25μF - 400
0.01μF - 600
0.02μF - 600
0.05μF - 600

16μF - 525
8μF - 600
25μF - 40

Ash

OK, here is a labelled image of the internals - http://1drv.ms/1NxjOCV

G'day Ash.

With the photo of the valves, is the bottom one a 6U7G or 6G8G?

That's the 6U7, the text rubbed off as I removed it. Still legible in person, just not the image.

IMO, the supplier of the best range of high voltage capacitors -- especially axial types -- is Justradios in Canada: http://www.justradios.com/cart.html

... but the US$20 order minimum can be a problem, plus there's the shipping. I usually order resistor packs to boost the order, if necessary.

Yes, the rule of thumb is to go higher (i.e. to the next step in the range) with capacitance and voltage, never lower.

When you say "0.1μF - 400/600 mixed" I guess that's a multi-section capacitor in a can? If so -- and assuming a direct replacement isn't available -- your choices are to re-stuff the can with modern caps, or disconnect the existing can and solder new caps below it under the chassis. (My own preference is to re-stuff when possible.)

You are approaching this in the Correct way. At first I did not know that μF and mf were the same so I was going batty lol. Yeah the 8 μF is now a 10, the 16 is a 20 and 25 is still available if not a little higher is good. These guys are all very experienced in this field so just listen and every one has a different approach. I reccomend you get the kits from justradio and make sure you order any ones you need that are not in the kits. Their postage costs are not real expensive but its up to you.it took about a week and a half to recieve the parcel which is pretty good.

What may be confusing is that capacitors are now in a standardised range of values & voltages.

The caps to replace all wax paper types are 630VDC
Some values are now not the same.

0.1mfd (104) is what it was 0.01 is what is was (103)
0.25mfd becomes 0.22mfd (224) and 0.02mfd becomes .022mfd (223)
0.05mfd becomes 0.047mfd (473)

The only Non polarised one that may be different is the Plate bypass on the 6V6 some use 1KV or more.

Thanks a lot everyone, really do appreciate the help.
It's definitely been confusing but a good learning exercise too, which is always a good thing.

I'll take the advice and order from Just Radios, I'll give an update when the parts arrive.

I'd still go to Jaycar for all those parts, faster and probably cheaper. The main filters electros may need higher-than normal voltage though, some of those older radios have a HT of up to 400V.

Any caps that are cathode bypasses or agc filter, or volume centre terminal to grid, and maybe some around the coils can be ordinary 50V greencaps, very cheap price for those. But as you're probably unaware of what cap does what, it's safer to get the high voltage ones. The giant 25μF 40V electro can be replaced with an ordinary (and very tiny) 22μF electro that costs just a few cents.

For some reason I can't see a link to view the valves.

Yeah the 600v requirement is what caused the problem, Jaycar had basically nothing.
Why would the original designer use 600V items if they were not required? (presumably they were expensive in the day)

Some radios did require them. On some HMV receivers from the 1940s the HT on the speaker field coil is around 750 volts. I measured this once after getting a belt off it. I was whacked straight through the paper 'insulation'. In some sets the field coil doubles as a choke to filter mains hum.

The primary reason I recommend high working voltages is simply because the higher the rating, the better the component can handle the load placed on it.

Up-speccing, also known as over-engineering, usually equates to more reliability. Take note of what Marc said above regarding codes on some types of condensers. Your receiver may likely need re-alignment once all the busted components have been replaced.

Some people are known to not go ahead with a refurb if they have to spend more than a fiver on replacement parts. I'd happily spend more than $100 if needed. Getting a radio going again is worth it in the long run.

Yeah the 600v requirement is what caused the problem, Jaycar had basically nothing.

Strange, the Jaycar I use (at Campbelltown) has 630V-rated caps, 10nF, 22nF, 47nF, which covers most things. It's only the electrolytics where you either get 10μF 450V fairly cheaply, or spend lots of money on capacitors meant for motors and so on, and which probably won't fit in the radio.

Strange, the Jaycar I use (at Campbelltown) has 630V-rated caps,

It would be interesting to subject Jaycar's capacitors to some HV stress testing.

The filter cap problem lies with the "B" voltage & the rectifier. The most common B Voltage is 250V on load.

As I have said before, some sets exceed that, but if you have an 80 / 5Y3 Filament rectifier, or similar and Silicon Diodes, feeding Heater tubes, you will get a surge that can reach twice EMF, prior to the heaters becoming effective and the valves start to conduct. So 250V suddenly becomes 500V, hence the 525 Surge volt Red Ducons of old.

Logically if "B" voltage is higher, in the above situation, the surge is greater & so must be the filter cap. Catch 22: We do not know so much as the secondary voltage of your transformer, to even predict "B" voltage. If there is a 600v one it may be the first cap in a choke input filter, as that is the one that cops the highest voltage all of the time?

I have used non polarised motor run (note run) caps in a 1929 set that was built pre-electrolytic cap. There are some non polarised film caps that some Americans are using as filters, as these do not deteriorate like Electrolytic's. The 1929 set was running close to 400V on the OP valve plate & over -54 Volts on its Grid so there were two separate rectifiers.

Hi Marc,

Would you recommend a low value (x100Ω) high wattage resistor, prior to the first filter cap, to soften the surge voltage??? If so, what specs??

The 'trade-off', of course, is a voltage drop under load ....... but this may be tolerable.

Thanks,
Ian