Grundig Mandello radiogram fading away...

[brymac]

Having made good progress on getting this radiogram back into the land of the living, I am stuck on a fading/distortion problem with occurs once the two ECL86s get hot.

With reference to the extract from the schematic attached, I have replaced the usual suspect capacitors feeding the grids of the push pull output valves (C61 and C62) as well as the elderly capacitors around the stereo balance control, which (and this may be a clue) appears to have no effect...

When initially switched on, I am rewarded with the usual excellent Grundig sound, but after about 4/5 minutes, it starts fading and becomes distorted...

What have I missed pelase?!!!!

Bryan

[Nuvistor]

If the balance pot has a open circuit the output stage will ‘runway’ get very hot and distort the sound.

[majex45]

It sounds like a reduction (failure?) of the power rails.
I would suspect something in the power supply, perhaps one of the diodes becomes open circuit when it gets hot?
It that a thermistor (S1) in the primary circuit?

[brymac]

Thanks both - the balance pot is a devil to get at, which is why I've worked around it! And I am wondering about the solid state rectifier - I need to see what's happening with the HT I guess..... No - S1 is a fuse....

[Nuvistor]

You can check o/c control by measuring G1 of the pentode section of each ECL86, it should be about 1Mohm to chassis, if both read ok then it’s off to check other items.
The HT could well be falling through a faulty PSU but if one or both the output valves draw excess current the HT will fall.

[Edward Huggins]

Originally Posted by brymac

With reference to the extract from the schematic attached, I have replaced the usual suspect capacitors feeding the grids of the push pull output valves (C61 and C62) as well as the elderly capacitors around the stereo balance control, which (and this may be a clue) appears to have no effect...

Bryan

er, not Push-Pull which would have been even nicer.....

[vidjoman]

The pot on the pentode grids is surely a tone control - the pot in the cathodes of the triode sections looks like a balance pot to me. How would the balance control cause a runaway of the output valves? The tone control could.

[Nuvistor]

Originally Posted by vidjoman

The pot on the pentode grids is surely a tone control - the pot in the cathodes of the triode sections looks like a balance pot to me. How would the balance control cause a runaway of the output valves? The tone control could.

Yes my mistake, yes the tone control.
Must try harder as my teachers told me

[jjl]

If both channels are affected then there is most likely to be a power supply fault.
I'd imagine that this set is fitted with a contact cooled selenium bridge rectifier - Grundig were fond of these. These rectifiers often fail, with the forward voltage drop of one or more diodes increasing and serving to gradually reduce the HT voltage.
Monitoring the HT voltage and observing whether it reduces when the fault appears should tell you whether the rectifier is suspect.

John

[brymac]

Thank you all - as usual with this world, life has intervened and I've had to spend time with various children and grandchildren. When will they learn that a part restored bit of vintage radio with a fault is much more important....?!!

I'll work through the power supply and I too have my eye on that selenium rectifier, much loved of 1960s German audio manufacturers....

I'll keep you posted!

Bryan

[brymac]

Well - after a long and rather frustrating day on this, I have managed to exclude some of the usual suspects; here goes (with reference to the original schematic at the top of this thread)

1.G1 of the pentode section of each ECL86is indeed about 1Mohm to chassis

2. Voltage from rectifier to C70 is a healthy 247V

3. Voltage at C58 and C59 is 198V.

4. Although that indicates that the 3 in 1 smoothing cap is OK, it was running hot, so I took it out of circuit temporarily and replace it with the "lash up " you see in the photo attached. That failed to resolve fading after 4 mins on both channels with distortion. Pressing the RH tone control (see photo of front panel) causes a slight increase in volume but with more distortion. Bear in mind that when switched on cold, this thing performs very well with superb sound....!

5. C67 replaced, although there appeared to be nothing wrong with it.

6. R64 tested spot on - but as I write this I realise I didn't check that it was still spot on when the problem occurred and it would be hot. I'll do that tomorrow, just in case....

7. When problem occurs, no change in HT (see photos)

8. C59 out and tested - OK so I put it back

9. Balance control still has no effect, even in the first 4 minutes before the volume decreases.

10. Both ECL86 Valve voltages steady at 200V (L&R) (pin3) 244R/189L pin 6, 79V (R) 109(L)
- they seem low and may give a clue??

I think that's it so far - both valves test OK on my Mullard valve tester so I haven't rushed to replace them - an in any case this fault affects BOTH channels!

Grateful for some more wisdom please - this is one of those challenges that you end up having nightmares about!!

Many thanks,

Bryan

[G6Tanuki]

What is the voltage across the common cathode-resistor R64?

Does this go high when the fading-out happens?

[Michael Maurice]

One thing you could check is the voltage on the grids of the pentode section of the ECL86’s. They should be around 0V. If they rise, it could indicate that they are going into grid current.

[brymac]

Yes - as indicated, failing to check what was going on with R64 when it started fading may have been a mistake - I'll be on that in the morning....

Bryan

[jjl]

Hmm, I didn't notice the common cathode resistor at first.
With this arrangement, one output valve misbehaving by taking more or less cathode current than nominal will affect the other as the bias voltage for both output valves will shift about. This is a case where a fault on one channel of a stereo amplifier could affect the other, blameless channel.

John

[brymac]

OK - I'll watch what is going on with the common cathode resistor first before investing in a replacement valve ; but both tested OK on my Mullard Valve tester....!

[Michael Maurice]

One thing you could do is to remove the link between the two cathodes and replace the cathode resistors and capacitors with 180 ohm or 220 ohm and 22uF capacitors for each channel. That way you can see if one valve was affecting the other

[snowman_al]

Sorry to butt in, but:-

Can you please measure the common cathode voltage pin 7 of either ECL86 (see post #12)
And what voltage do you measure on the input grids of the pentode sections pin 8 on both valves (post #13)

The very low anode voltages on the ECL86 pentode sections with 247 volts on the transformer primary is a concern. You are loosing 150 volts in the output transformer primaries

[Techman]

Originally Posted by brymac

but both tested OK on my Mullard Valve tester....!

They will do. Testers won't show up faults like this, particularly that Mullard one. They'll give a 'go/no go' answer and are not testing the valve at the parameters operated at in the real circuit. I've said this many times, there's only one good test for a valve and that is in the circuit it was designed to operate in. Testers are generally a waste of time unless you're a valve dealer and need to provide some figures for your customers.

The ECL86 is trouble once it gets old and has had a hard life with leaky capacitors, and replacing the coupling capacitors is often still too late for the valve. The EL84 can go the same way, but it's nowhere near as bad. The last EL84 I had with this symptom was a Russian version in a Rigonda radiogram where it would fade into distortion (with grid emission) after several hours of running.

If you've been keeping up with things on the forum then you'll have been reading the concurrent thread that's running at the moment regarding the ECL86 valve. I've posted a link to it below and you'll see my comments a couple of days ago where I was referring to this thread in post #8. We may be wrong, but it's looking very much like one or both of the valves are 'running away' and are faulty.
https://www.vintage-radio.net/forum/...d.php?t=182529

[brymac]

Latest -

Got as far as looking at voltage across R64 (combined cathode resistor) - it went from 0V (working OK) to gradually up to 8V after 4 mins as fading/distortion occurred. Have replacement, but not time to fit it, as I’ve had to pause for family stuff - just about to go to Bristol to visit grandchildren for the weekend......