Strange phenomenon from Homebrew audio amp

[Sideband]

Originally Posted by G8HQP Dave

Did you mean 0.1uF for coupling?

No I meant 0.1uF decoupling...insert the resistors in the HT supply as suggested and then decouple them to chassis with 0.1uF. With experimentation it might be possible to reduce them to .047uF but in amps of this type, 0.1uF seems about right.


Rich

[Radio Wrangler]

The 6BW6 has ben used s a low power HF TX often enough.

How about HF oscillation squegging at 2Hz?

Try a stopper R in the 6BW6 g1.

David

[MrElectronicman]

Thanks for all the suggestions.

First of all I redrew the diagram and corrected all the mistakes, I hope. The original diagram was on a couple bits of paper and in my head, so I had drawn the compete circuit up rather quickly.

Noted the use, or lack thereof, of HT decoupling. I will add a couple of 3K3 or 2K2 resistors in the HT lines, and I happen to have some 16uF 400 Volts capacitors so will try them first.

Last night I had changed the coupling capacitors in the tone control from 0.47uF to 0.1uF and it was working. Today I tried swapping them to 0.47uF on the input and 0.1uF on the output, which also appeared to work, but when I touched the first valve I saw signs of motorboating again.

I will put back the original 0.47uF caps, and add the decoupling and see what happens.

Attached a revised diagram.

Thanks again for all the advice.

[MrElectronicman]

Originally Posted by Sideband

No I meant 0.1uF decoupling...i


Rich

Will that work with such a value? I can try it anyway

[ukcol]

Originally Posted by Sideband

No I meant 0.1uF decoupling...insert the resistors in the HT supply as suggested and then decouple them to chassis with 0.1uF. With experimentation it might be possible to reduce them to .047uF but in amps of this type, 0.1uF seems about right.


Rich

Hi Rich

Surely 0.1uF is too small for decoupling at audio frequencies? I have just looked at four Mullard amplifier circuits and the decoupling, at the particular part of the HT line under discussion, was in each case =>10K ohms and =>8uF.

Colin M

[G8HQP Dave]

Yes, 0.1uF is a suitable value for coupling to a grid with a 100k+ resistor.

Hopelessly useless for decoupling audio supply rails (although OK for decoupling in an IF strip). 10uF would be better, or even more. Remember, the supply decoupler has to work down to subsonic frequencies. A little maths will demonstrate.

[Sideband]

Brain strain......sorry. For some reason I was thinking I.F.....

Yes of course 10uF will be better. I doubt you'll need to go higher than that. The values of resistors will depend on the HT you have available but you probably won't need to go higher than 4.7K. You just want to get rid of any audio that's feeding back via the HT rail.


Hopefully that will solve your problems.



Rich.

[MrElectronicman]

Thanks again for all your comments,

The HT supply I am using is a a bit underpowered for a stereo amp, so I am planning to get a bigger transformer in the near future. The HT is running at about 40mA per channel or 80mA total at about 225V DC, and the transformer is rated at 250V DC at 75mA and thus runs a little warm.

However, I have found another problem, one channel is running at a lower level than than other and it appears to be due to the active filter network for the tone control.
Using my scope and DMM I compared both channels and suspecting a wiring error, I checked it twice, measured the value of the resistors, verify the values of the capacitors and all appears well. Of course I swapped the valves but it made no difference. I can however see a difference in signal levels within the network.

My next step is to made a new temporary network for the active part of the control and substitute it in each channel. I do not know whether one channel is actually lower or the other higher, so I hope my substitution will help me to find out.

I has been quite a while since I did any fault finding, and even longer when using valves, but I must say this is a bit of fun.

[G8HQP Dave]

The thing to keep in mind when chasing a difficult fault is that there is always a physical reason for the wrong circuit behaviour, and often the clues are there in slightly puzzling measurements. Despite some appearances to the contrary, radios and amplifiers don't contain mischievous demons who are bent on making fun of us. I'm not so sure about old TVs!

[Herald1360]

Originally Posted by G8HQP Dave

Despite some appearances to the contrary, radios and amplifiers don't contain mischievous demons who are bent on making fun of us.

Maybe not. I think they may have been employed as designers by Philips, though.......

[MrElectronicman]

Well I found the problem, bad socket in one of the valve holders. I bought some new ceramic valve holders which I thought would be somewhat better than the usual phenolic, but I am not at all impressed with them. Maybe Philips had a hand in their production.

I am now going to put the HT decoupling in, I did not want to change anything until I had resolved the signal problem, high chance of having to work on two problems at the same time.

I must say though the amp sounds very nice, if you don't turn up the base too much.

[Herald1360]

I'm going to get pedantic here and request "turn up the bass" rather than "turn up the base"; the latter conjures up a picture of someone manically folding metal for a bottom cover

[MrElectronicman]

Originally Posted by Herald1360

"turn up the base"; the latter conjures up a picture of someone manically folding metal for a bottom cover

That may be true!