Trouble with Valve Amp
 

Hi folks

I have built a 30W valve amp from a circuit that is in the amateur's guide to valve selection, page 46. Its a Mullard publication, and for some strange reason after warm up the dreadful feed back howl that comes out of the LS is almost like an oscillation taking place. The only component I could not use was the exact output transformer. I used one made by RS which has an 8 ohm output and 6.5k anode to anode on the primary. Even the feed back resistor is as of the diagram, any idea's?

Ken

Re: Trouble with Valve Amp
 

Have you connected the feedback to the right end of the output winding?

Re: Trouble with Valve Amp
 

Has it got a negative feedback path.
It could be that the speaker winding is the wrong way round.

Re: Trouble with Valve Amp
 

Echoing barrymagrec, I would suggest connecting the feedback to the other end of the o/p transformer secondary winding. What is happening is almost certainly that you are getting positive instead of negative feedback. This will cause the amp. to become an oscillator. The way I was taught it is that NFB 'reduces gain and improves (increases) frequency response. Conversely, PFB, increases gain, and (almost certainly) reduces frequency response. edit; crossed with Refugee's post!

Re: Trouble with Valve Amp
 

Thanks everyone, I will try this and see if it does cure this problem, will let you all know.

Ken

Re: Trouble with Valve Amp
 

As others have said, picking the wrong end of the transformer secondary would certainly induce howling. But it's not out of the question that something else could be causing it and in that case moving the wire from what is in fact the correct end to the wrong end could make the howl so bad that something is damaged. So I'd recommend that to begin with you just disconnect the wire and try the amp without feedback at all. If this doesn't stop the howling then we'll need to have another think about the possible cause. Of course disconnecting the feedback will make the amplifier very much more sensitive. So do take care to feed it only with a small input signal.

If the howling stops when you disconnect the feedback lead then you can power down, connect the wire to the other end of the secondary (and swap the secondary ground wire too) and switch the amp back on. Keep your finger on the power switch so that if howling starts again you can turn it off immediately. Running a 30W amp as a power oscillator can fry the tweeter in your speaker very quickly indeed !

I've not seen a copy of the book you mention. The only online link that Google finds for me is to a Gramophone magazine column announcing its publication in Jan 1953. So that will pre-date the EL34. Are you using EL37s as the power output valves ? Also did the original design have an 8ohm secondary on the output transformer ? If not then you might find that the feedback level is different from what the designers intended.

Cheers,

GJ

Re: Trouble with Valve Amp
 

The reference suggests 1948 so either EL31, EL37 or 6L6 one would presume.

The secondary impedance may be wrong but it would already be a very flaky design if that pushed it over the edge, I would have thought.

http://www.r-type.org/refs/ref4035.htm

Re: Trouble with Valve Amp
 

Some Mullard designs used quite heavy feedback, which will only work well if the output transformer and the circuit construction are good copies of what Mullard intended. Depart too far from the 'script' (especially in OPT quality) and instability is likely.

Re: Trouble with Valve Amp
 

I was curious about this design and so I did a bit more rummaging through my old valve data books and I came across Mullard's Valve & Service Reference Manual (2nd edn 1951). Sure enough, in the back of that is a multi-purpose circuit for a 10W or 30W amplifier with separate component lists for the two powers. I never realised that it was there, so thanks for that ;D.

Attached are scans of the circuit and the component list for the 30W option. Is this the amp we're discussing ? Compared with the level of detail that Mullard go into in their Circuits for Audio Amplifiers just a few years later the description here is very, um, terse. This is literally all there is.

Cheers,

GJ

Re: Trouble with Valve Amp
 

That looks strikingly like the circuit I reverse engineered from the amps I use as a front room stereo.
Look out for how they got more gain:blah:

Re: Trouble with Valve Amp
 

Cured the howling. I should have known really and yes, it's just odd by just swapping the feedback wire and the LS. Mullard circuits are one of the best to work on, all I have to do is get a couple of KT66 valves as spares.

Ken

Re: Trouble with Valve Amp
 

I always used to get the connections sorted by trial and error as the transformers I scrounged back in the day rarely had any discernable markings.

Re: Trouble with Valve Amp
 

A better way is to swap the primary connections. If the secondary is reversed then the earthy connection (0) to the speakers is actually connected to the 16 ohm & feedback connection. Ok it’s not dangerous but could upset the system if someone decides to Earth the speaker neg connection and short out the transformer secondary.

Re: Trouble with Valve Amp
 

Can't see any attempt at tailoring the NFB or even the forward signal path, so presumably the overall amount of feedback must be quite small.

Re: Trouble with Valve Amp
 

In those days (1948) use of negative feedback in audio amplifiers was quite recent and the transformers available not always good enough to support a large amount.

D.T.N. Williamson, Peter Walker and Harold Leak amongst others did much in this area.

Re: Trouble with Valve Amp
 

Hi folks

Swapping the LS connections did cure the unwanted howling, I have put a picture of the finished amp on here so all can see what it looks like. A good friend of mine who has since passed on gave me a sheet of aly of which I knocked up the chassis out of, I will not forget what he said to me and it is this; "It doesn't matter what it looks like so long as it works"

Best wishes

Ken