Odd hum problem with newly built stereo 3-3

[Aerodyne]

I built this 3-3 stereo amp exactly to the Mullard spec, i.e. two 3-3 amplifiers but with common controls and a single power supply.

Now here's the problem. There's a low level hum immediately upon switch-on. Remove the valves including the rectifier... and the hum remains. Disconnect the secondary HT and low voltage supplies from the mains transformer... and the hum remains. Although there's a full screen box over the twin output transformers and the mains transformer, additional mains transformer screening has no effect upon the hum. Disconnect the transformer lamination chassis grounding... and the hum remains.

Otherwise, and when working (playing a disc or MP3) the hum cannot be heard and the amp works as well as I'd expect, i.e. it sounds excellent.

What if anything have I missed? I'll admit to being out of ideas.
Just for clarity: O/P transformers are placed at right angles to the mains transformer. Full wave valve rectification is used - but as said above, even without any valves inserted, the faint hum remains.

Tony

[G8HQP Dave]

Clearly there is magnetic coupling between the mains and output transformers. This could be either through 'air' or via a steel chassis (aluminium will not do this). 'Screening' of transformers does not always work very well, and is often included for cosmetic rather than functional reasons.

Unfortunately the solutions all require some reworking:
1. change orientation
2. move further apart
3. swap steel chassis for aluminium

There is one faint hope of remedy without major work: it is just possible that if you used a steel chassis and put the two wires from each OPT secondary through separate chassis holes (i.e. signal and return in separate holes) then this could be the route of magnetic coupling. The two wires from each OPT secondary should go though the same hole to avoid coupling to the mains transformer; the two wires from the other channel secondary should go through a different hole to avoid inter-channel crosstalk.

[Radio Wrangler]

There is also the issue of large currents coming from stack-clamping bolts. This can put large currents in the chassis and can couple the reverse way into output transformers.

David

[Edward Huggins]

There should be virtually no hum with this circuit. Don't just let the audio mask it. Some experimentation may be needed WRT the orientation of the 2 x Output TX to each other and in turn to the Mains TX.

[Refugee]

I would be tempted to mount the mains transformer on insulators to isolate the stack clamps from the chassis.

[Aerodyne]

Many thanks for the replies, gents. I did think this had to be some kind of field problem but couldn't think what to try next! FYI the chassis is 16 gauge aluminium. Output transformer leads are fed through grommeted holes, each TRX with its leads together but each channel well separated.

I will experiment with isolating the stack and also check loudspeaker termination wiring for positioning. Orientation should be fine - the transformers are well apart from each other and at right angles to the mains transformer - that said, I take your point and will try the effect of repositioning them.

I'd really like to sort this out. I've built literally dozens of amplifiers over many years and never before has this been a problem.
Once again, much obliged for your help.

Regards
Tony

My 3-3 had a weeny bit of hum like this too, merrily moving a transformer 1/16th of an inch worked.

[Aerodyne]

Merlin, clearly you live up to your name! I've always considered designing and building amplifiers to be more art than science.

Tony

[Aerodyne]

Hello again

A quick update on the hum problem. I tried everything mentioned in the very helpful threads I received. Finally found the only way I could minimise the hum was to turn the output transformers onto their backs! See attached photo. I had little room to spare but managed it.
Now there is virtually no hum whatsoever, an excellent result.

Thanks again to all who offered advice.

Best wishes

Tony

[GrimJosef]

If it's any consolation, this sort of problem can trouble commercial amps too. A while ago I worked on a pair of large mono amps which had an interstage transformer coupling the driver to the output valves and also doing the phase-splitting. This small transformer was mounted inside the chassis, offset not very far from the mains transformer. The amp hummed. I remounted the interstage transformer on a gimbal (red-ringed in the picture) and tilted it, dropping the hum by 7-10dB at a stroke.

Cheers,

GJ

[Aerodyne]

Yes, definitely more art than science, GJ, but what matters in the end is the result and your result with the interstage TRX proves it. Great idea, using a gimbal to locate the 'null' hum point. I just held one of my output transformers in my hand (taking care to avoid the hot terminals) and moved it around until I found the sweet spot!

Regards,

Tony

[GrimJosef]

I confess I waved the transformer around by hand too. I hadn't expected the null orientation to be as angled as it was (the offset from the mains transformer must have accounted for the local magnetic field being tilted). But once I did locate it it was easier to build a simple gimbal than to try to machine a bracket at just the right angle.

I know what you mean about trying not to forget the high voltage. In my case the output valves were 211s which run with a 1050V HT. The silver cylinders just to the left of the transformer are the reservoir capacitors for it. Accidentally bumping a hand-held metal object into the terminals on those might well be the last thing you do.

Cheers,

GJ

[Ed_Dinning]

Hi Gents, some of the earlier Ferrograph tape recorders had the mains transformer mounted on circular slots so it could be rotated to minimise hum.

Ed

[Edward Huggins]

Yes, as did one of the small, very well made, late 1950s Dulci amps based on the 3-3. Seeing the very substantial looking chassis in the OP's Post #9, I note the EF86s are not screened and am not quite sure if they should be?

[G6Tanuki]

One thing to be aware of is the issue of inadvertent induction of currents into the chassis because of incorrect power-transformer wiring.

I've come across this in the past, where either the supply to the secondary or the two wires from the LT winding on the power-transformer were fed through separate-but-adjacent holes in the chassis. When you've got 3 or 4 amps of heater-current flowing, feeding the 2 wires through different chassis-holes produces a sort-of 'single-turn-secondary' effect and can lead to significant circulating AC currents in the chassis, which can then induce themselves into all sorts of inconvenient places.

[Aerodyne]

GJ, that is truly shocking... If I'd been shocked in this case it would have been a mere 300V or so. I must confess that even after all these many years I still hate getting shocks and do my level best to avoid it - but I do not always succeed...

Ed, the circular slot mounting idea seems to be a really good one. Even so in this case I doubt it would have suggested to me that the transformers should be turned over!

Edward, good point. I did try dropping grounded screening cans over the EF86s because I wondered the same thing but it made no obvious difference. Now I've got the hum tamed, it might be worth trying again just to see if the very slight 'presence' which is evident could be tamed. Maybe I'm being a perfectionist (again).

G6, that is a good point and one I'd not given a lot of consideration, having never built an amp power supply wired in such a way. I can see how a problem could occur through poor wiring methods.

Again, my thanks to you all. Always good to gain fresh insight.

Tony

[G8HQP Dave]

Glad to hear the original problem has been solved. I suspect proximity was part of the problem - the picture in post 9 shows OPT right next to the mains transformer. There is nothing like a bit of space to reduce magnetic coupling.

Given the choice, put the valves near each other and the transformers well apart rather than the converse.

[Aerodyne]

If I build yet another amplifier - a possibility, though I'm getting long in the tooth for protracted metalwork and electronic construction - I'll bear all the points made in this thread in mind and certainly will heed your 'proximity' point, Dave.

Actually, I originally did plan for reasonable separation between the transformers. I built the chassis 300mm wide but I hadn't reckoned on changing the output transformers for much larger, better ones. I now can see exactly where I went wrong. Isn't hindsight frustrating!

Tony

[Sideband]

With output transformers that size, I'll bet it sounds good! I was amazed at just how good these 3-3 amps were when I built one into a record player back in the 70's. The original internal speaker was hopelessly inadequate and I had to replace it with something much more substantial. I also mounted some external speaker sockets at the rear of the cabinet so a larger speaker could be used when it was necessary to turn up the wick during parties!

[G8HQP Dave]

All my projects would be much better if only I built them twice, with the benefit of debugging the first build. The snag is that the only thing I have built twice (stereo speakers) I found the second one was very boring to build because there was little new to discover and anyway they had to be identical.