First of all, thanks to the forum members who helped me bring this project to fruition.

Here is the first Wanted link:

https://vintage-radio.com.au/default.asp?f=3&th=1020

With help from forum members I acquired a Tasma 360 chassis and period correct ED speaker.

It was a rusty basket case and I did a bare chassis strip, wirebrush, paint and rebuild. Replacement mains transformer, a NOS 80 in ST style glass. Replica wet electros using paint-stripped Berocca vitamin cans and large bolts. They really look the part.

The paper caps were too weathered to survive a melt-out rebuild with new caps and hotmelt, which is my usual method. The all-open-circuit candohm was repainted and resistors hidden underneath it.

Couldn't match the original peephole dial so I fitted the Tasma's round dial and bakelite surround. I hate to say it but it looks better than the original peephole dial. Can't go back now anyway, the timber's been cut.

Did you know you can source round, domed dial glass from scientific lab suppliers? They call it a watch glass. Glued into the bakelite surround and looks just like new.

Anyway, it was an uneventful resto, Everything went back together and the radio worked first go without needing IF realignment, which is just as well because the IF transformers are potted in the cans with bitumen. Impossible to dismantle and trimmers frozen.

There is a small amount of stagger tuning of the IFs which gives a better frequency response and the radio sounds quite good. The 360 was Tasma's first chassis with a diode detector and AGC. The only change I made was to add grid leak bias to the audio - the 360 circuit has the grid of the first audio connected straight to the volume pot wiper. This is unwise because the bias of the audio shifts with the AGC and results in distorted audio on strong signals and higher volume pot settings.

BUT.....there is an intermittent fault that stubbornly refuses to appear when the chassis is out of the cabinet. It only happens after the radio has been switched off for several hours or more and generally only for about 60 seconds after power on.

The volume drops about 8db and then, with a snap, comes good. Stays good from that time on, until the next day.

The AGC (measured at the pickup terminals) when the radio is good is around -2.2 volts (702 Sydney received on the northern beaches) and drops to -0.9 volts when bad. So it's in the mixer-oscillator or the IF amp., not the audio section. It's not responsive to tapping etc.

The timing and consistency of the fault makes it sound like a thermally sensitive weld in a valve - either the 6A7 or the 6D6 - but I have neither to try.

Anyone struck this fault before?

Anyone struck this fault before?

I've heard of roughly similar symptoms being caused by capacitors within the IF cans.

Seen it heard it: Most likely a Mica cap in the AGC, if the others have been changed. Check if it is actually holding frequency as a Mica in the oscillator circuit may throw it off, as will a large suicidal Arachnid on top of the oscillator terminals = Loss of signal strength loss of AGC voltage.

A leaking Mica on Delayed AGC where the source is a plate: Bleeds positive into the AGC. Also check the top caps. Dry joint.

Micas are rare failures in IF's here, only had two with mechanical failure.

Marc

Have something similar on my STC 830. On turning on it warms up to a second or so of sound and then goes dead for a minute or so an then comes good with a bit of crackling. This was tracked down to the 6a7 mixer valve by tapping lightly around the valves. Then swapped in different 6A7s to the same result. So I assume it is a thermal effect in the valve socket. Have had a spare sitting in the back of the radio until I get around to finding the right rivets and tools.

The uninitiated listening are more impressed with the crackling than with the final good performance.

Yeah thanks guys.

It's not a frequency drift because there's absolutely no change in the audio quality. And it's consistent, it will never happen with the chassis on it's side I've since discovered. So next I might bring some test points out on clip leads, but it's not going to fail again today so it's packed away for now.

Yes I have already tried cleaning and resoldering the top cap wires, seen that before!

Marcc, the AGC is simpler than that, no micas, just comes straight off the diode load, i.e. volume pot. A very simple circuit is the 360.

You might have a point about an IF tuning cap. Can't get to it though. But what I will try is a fire-up on no signal - it could be an OC IF tuning cap coming good one some AC has been applied for about 60 seconds....

Ian. There is a mica across the volume control, or there should be. Labelled "8" on the circuit diagram. An intermittent in that component could cause the problem you describe. I have often found problems in this component, sometimes find it to be missing completely, causing loss of gain and loss of AGC voltage in this case.

Just a historical note; the 360 was not Tasmas first radio to have diode detection and AGC, not by a long way. The 170, which appeared in1933, had these features, and delayed AGC as well.

Cheers, Stuart

Yes Stuart it could be that mica. Although, why does it never fail when the chassis is on its side?? Weird! There's also absolutely NO tell-tale crackling or frying noises, ever.... At least I can now be certain it's before the volume control.

Thank for the historical correction, it was a view I'd formed after scanning through the Tasma circuits looking for something that matched. They really loved their anode bend detectors in those days - horrible things! And then I found the number "360" written underneath the chassis in wax crayon!

Sometime this week I'll pull the chassis from the cabinet, attach some strategic clip leads, and put it back. Then connect half a dozen meters and a scope to it. On 2nd thoughts, I'll put the scope into logging mode using two channels. It's not going to escape!!

Sometimes it takes more than a couple of days of not being powered for the fault to show up so you are never quite sure if you've found it!. I really do need to catch it in the act to be confident it's fixed.

I'll also try not connecting an aerial for about 5 minutes. And putting the chassis in the fridge for a while.

If it does not have the fault when sideways that could be debris as well as a dry joint. I have seen that sort of carry on with a Midwest as the scanting left a big (16Valves) chassis that flexed. Flexing equals stresses.

I have found solder floating inside cans and the tops of top trimmers can get staved in. Have a good look at the valve fly leads. Never unusual with a bit of insulation shrinkage & loss of the sleeving for the bare wire or the terminal to hit a valve can.

Marc

Yeah Marc I tried banging and tapping in all positions with no effect whatsoever. I wish I could strip the IF's down but that would involve melting the pitch out of the cans - last resort stuff and I'd probably try to swap the IFTs out first.

This chassis has a solid earth bus that runs around the chassis and is grounded in multiple places. I was very careful when rebuilding it to ensure no paint and bright metal under the earth lugs. Pretty much everything has been resoldered and contacts cleaned as part of the rebuild process. I think it's sound in this respect given you can give the chassis a hefty whack when it's running and not even a crackle is heard!

I think I was able to confirm your O/C IFT tuning cap theory last night by firing it up without an antenna, letting it run for about 5 mins and then connecting the antenna. The fault was there and lasted about 30 seconds after the antenna was connected. But I haven't been able to repeat that behaviour since. I'll try again tonight.

Interestingly, the AGC line goes positive by about 1.5v while the valves are warming up, then jumps to either -0.8v or -2.2v depending on the fault condition.. I can't see the mechanism for this behaviour, there is no obvious path for a positive voltage under any component failure condition. Except maybe some internal leakage from the triode anode to one of the diodes inside the 75. It's very unlikely to be socket leakage because they are on opposite sides of the pinout. The 75 is a NOS valve dating from the 1930s.

When the heaters are in play and conducting, all grids will go positive if the cathode is open.

There is a large voltage surge where an 80 / 5Y3 is feeding heater valves. That surge approaching twice load volts may be enough to push positive into a bigger AGC cap & charge it So a cap from B+ (delayed AGC), or a faulty cathode resistor / joint on any valve hooked to AGC is on the cards.

Rare cases badly soldered pins on the valve.

Marc

Yes Marc thought of that too.

Common cathode resistor for the 6A7 and the 6D6 is a new part, the old green WW resistor was too far gone with corrosion.

75 cathode is earthed and the joint checked.

This +1.5V appears as soon as the 80 warms up and B+ appears, i.e. within a couple of seconds and well before the rest of the cathodes heat up. As soon as they do, sanity is restored.

Next, I'll check what happens with the 6A7, 6D6 and 75 out of their sockets.

OK, I reckon I've found it!

I pulled the 6A7 and 6D6 out. With them out, there is a 200 megohm leakage, measured at 350v B+, between primary and secondary of the 2nd IFT.

At 200v B+, (dim bulb tester in circuit) it just goes away. Flip the chassis up and test with the DVM. It passes a normal leakage test no problems.

Hmmmm.......! Here's my theory:

It just takes a little while for the front-end valves to reach full emission and pull the B+ down. And they have to be off for a while to go back to their "weak" status (they are 80+ years old after all).

I'm glad I didn't just replace the 6A7 and/or the 6D6 because doing so would probably have "fixed" the fault ..for now...

IFT is potted in bitumen. No tags, just fly leads emerging from the bitumen.

I have an HMV battery portable that had a S/C in the same place so I think I can guess what would be lurking inside this one.

I'll need to rustle up a replacement late 30s square can IFT. I may get time to start rustling on the weekend.

Only question is - why no crackles?? IFT faults are usually very chatty. This fault transition is sharp, it happens with a pop, as you would expect from a DC shift on a B+ or AGC line. No more than that.

This fault transition is sharp, it happens with a pop,

Similar symptoms to the case I referred to in post #2 above.

Yes GTC you got pretty close!

I confirmed the 2nd IFT this morning before it performed its usual trick of coming good.

Leakage between primary and secondary of I.F.T. and R.F. coils is something I have found quite often, but not intermittent like that. Cured sometimes by putting the coil in the oven for a while. An I.F.T. fault involving a fracture in one of the windings will often generate crackling but this fault is different, so not surprising it doesn't crackle.

Cheers.