Tranny failure

[Diabolical Artificer]

The 1070M is nothing like the 1070. I had trouble finding anything on it and could only find a post asking for info etc. After some detective work and lateral thinking I found that the PM250 and PM400 use the same board, chassis etc. AFAIK the 1070M is a PM400 in all but name, see PDF. I use the clearer PM250 schematic for reference.

There is also a video here - https://www.youtube.com/watch?v=a98bYbZ4wdA which will give some idea of some the issues I found and part progress I've made.

Because I'm relatively new to all this, and my understanding of all aspects of the amplifiers circuit is a bit hazy, my fault finding approach is a bit all over the shop at times.

I'll get it on the bench tomorrow and try and give some better data and address some of the points raised.

Andy.

[Mark Hennessy]

Hi Andy,

Thanks for the updated service data. Actually, it's not all that different in overall topology, which isn't all that surprising. Audio amps tend to follow a common theme... As a result, I think that most or perhaps all of my previous questions are still relevant.

Nothing wrong with being a beginner! Everyone has a different approach anyway. I tend to be a bit "theoretical", and like to know how the circuit should be functioning before doing anything else - that approach is admittedly slower than someone else who might jump in and so the "common sense" stuff like swapping caps and sweeping for dry joints, but my approach does have a higher success rate, and is more intellectually satisfying. But it might not put bread on the table - in the trade, time is money!

Also, I like to take the time to see how it works because I can learn new techniques or tricks that help me become a better designer. I think of myself as a designer first, repairer second, but in practice, that balance changes all the time. Before kids, I had much more spare time to design stuff - sometimes simple, sometimes quite complex - but now, I find myself doing more repairs because that fits in easier around family life and work (and it pays a little bit too) - the design stuff tends to be simpler, more self-contained projects that fit a need rather than being driven by interest or research. As the kids get older, the balance will no-doubt shift again...

Anyway, I see this amp has loudspeaker switching, unlike the 1070 IIRC. Stating the obvious, I know, but it might be as daft as switchcleaner

Cheers,

Mark

[Diabolical Artificer]

Managed to do a bit of testing this morning, the first thing I found when I scoped the left channel OP was a 700- ish khz @ 500mV P-P, that is according to my counter. I went and tested it anyway to see what happens to the DC rails on full volume, they went down a bit as expected. However whilst testing I was getting 20 odd volts RMS at the speaker OP. I switched off to unload some video's I'd taken of my exploits, switched on to measure what happens to the input signal and only got 6v RMS. Input sig (120mV) goes down a tad in both cases at full bore.

One other issue I've found is if I go to measure the voltage on the emmiter of Q705 which is the next Q after the LTP/diiff input, with the amp powered through a LL, the lamp glows brightly. The same is true if I touch it's base. All components in that region EG C708/RH06/Q719/R711/R707 are ok, as are the DC voltages on Q705 AFAICT.

Saw TerryKC yesterday, he said to look at the effected channel's FB network. Not knowing a great deal about NFB, apart from the fact that we feed back a bit of the OP signal, I'm not sure what to look for apart from checking C705/R709. I'll look at the zobel components too, C717/R727.

Apart from that, i'm stumped as to why both channels have normal OP power one minute, the next not. I am going to go and resolder every joint though.

Andy.

[Refugee]

That gives me a big clue as to what is going on.
It will be a duff decoupling capacitor that has gone open circuit allowing the amplifier to become an AM transmitter.
One big long wave transmitter. Check it out with a long wave radio and you should be able to tune into it from the end of the street.

[Murphyv310]

That gives me a big clue as to what is going on.
It will be a duff decoupling capacitor that has gone open circuit allowing the amplifier to become an AM transmitter.
One big long wave transmitter. Check it out with a long wave radio and you should be able to tune into it from the end of the street.

Ha Ha.
That could be fun, with an audio input I'd expect the oscillation to vary so it could be a Long Wave FM transmitter

Could even be parasitic oscillation.

[Terry]

If Andy's frequency measurement is accurate it is just below the bottom of the MW ~430m, so a conventional MW/LW receiver won't pick it up.

If there's a good 2nd harmonic, 1400kHz on MW might be worth checking ...

What I actually said to Andy yesterday was that his amplifier was unstable, which means unwanted (positive) feedback - although a fault in the NFB loop might raise the gain to a suitable level.

[Mark Hennessy]

It's possible that an open-circuit capacitor in the lower "limb" of the NFB network could cause this. The gain would fall to unity, and depending on the phase margin, it could easily become an oscillator.

That'll be C703, perhaps - I can't read the scan. It's 100uF, 10V, just to the right of the LTP. The other channel is even less readable.

Given the age, it might be worth changing it anyway.

[Murphyv310]

Hi.
If it was a fault in the NFB then I'd expect the gain to be significantly higher and there could be a big change in frequency response.
Don't discount parasitic oscillation if the amp has been "serviced" in the past.

[Mark Hennessy]

Not necessarily.

High gain would be caused by the resistor between the output and the inverting input going high or O/C. But, as I said, high resistance in the lower "limb" - the resistance between the inverting input and ground - would reduce the gain; all the way to unity if O/C. Now, resistors are generally pretty reliable, but as well as that resistance, there is the series capacitor I mentioned, who's job is to reduce the gain to unity at DC. Well, we all know what electrolytic capacitors do

I've no idea if this is the problem, but it fits in with some bits of the description (low gain, oscillation), and is certainly something I've seen before. Most solid-state power amps are meant to run at 20-30dB, so become oscillators if you turn down the gain. And the reported frequency of 700kHz suggests Nyquist loop oscillation rather than localised parasitic HF/VHF oscillation that emitter followers love to do.

In addition to all that, there is plenty else to check. But it's an easy one to eliminate before going further - simply change those 2 caps. Though frankly, I'd change them all, as there's no real point trying to apply logic until you've replaced all the components that are known to wear out. Some of those Japanese capacitors can be amazingly reliable, but other types aren't.

[Murphyv310]

Not necessarily.

High gain would be caused by the resistor between the output and the inverting input going high or O/C. But, as I said, high resistance in the lower "limb" - the resistance between the inverting input and ground - would reduce the gain; all the way to unity if O/C. Now, resistors are generally pretty reliable, but as well as that resistance, there is the series capacitor I mentioned, who's job is to reduce the gain to unity at DC. Well, we all know what electrolytic capacitors do

I've no idea if this is the problem, but it fits in with some bits of the description (low gain, oscillation), and is certainly something I've seen before. Most solid-state power amps are meant to run at 20-30dB, so become oscillators if you turn down the gain. And the reported frequency of 700kHz suggests Nyquist loop oscillation rather than localised parasitic HF/VHF oscillation that emitter followers love to do.

In addition to all that, there is plenty else to check. But it's an easy one to eliminate before going further - simply change those 2 caps. Though frankly, I'd change them all, as there's no real point trying to apply logic until you've replaced all the components that are known to wear out. Some of those Japanese capacitors can be amazingly reliable, but other types aren't.

Yes I'd agree on the reliability of the Jap caps. If the gain did drop surely the OP would have been able to hear the difference. TBH I've not looked at the circuit and don't know the amp in question, so my remarks are generic.