Power Transformer leakage

[audio_moksha]

Hello folks,

i just completed my EL34 SET and it sounds great but I am getting a low level hum that is induced by a magnetic flux leakage from the power transformers to the OPT. I know this for sure because when i moved the OPT off the chassis a few inches away, the hum disappeared. Right now I have a few aluminium panels between the transformers and its helping a bit.

So what causes this leakage? Is the transformer saturated? The power transformer is also noisy when turned on and the buzz can be heard a foot away from the amp when all is quiet. I am a bit concerned about the safety of the amp...Someone suggested suing some mu-metal between the transformers. do you think this could eliminate the hum completely?

The transformer is a Hammond 274BX 375-0-375 @180mA. OPT's are hammond 1628SEA 5Kohms impedance. The amp consists of one 6SL7 drive stage and two EL34 power stages for about 6W of SE power.

Anhy advice would be appreciated. Thanks.

[Tweeker]

Mu metal is spendy and fragile. Try steel instead of or in addition to the aluminum, at powerline frequencies it should be more effective.

Power transformer is 90 degrees to output transformer?

[Miles Prower]

Originally posted by audio_moksha
Hello folks,

i just completed my EL34 SET and it sounds great but I am getting a low level hum that is induced by a magnetic flux leakage from the power transformers to the OPT. I know this for sure because when i moved the OPT off the chassis a few inches away, the hum disappeared. Right now I have a few aluminium panels between the transformers and its helping a bit.

So what causes this leakage? Is the transformer saturated? The power transformer is also noisy when turned on and the buzz can be heard a foot away from the amp when all is quiet.

It's possible that the core is saturating. If that happens, the ability of the core to offer magnetic shielding is compromised. Noisey xfmrs are often in saturation, however, it could also be lams that aren't tight enough. It could also be that the coils aren't oriented properly. These should be oriented at 90 deg to minimize stray coupling.

I am a bit concerned about the safety of the amp...Someone suggested suing some mu-metal between the transformers. do you think this could eliminate the hum completely?

That shouldn't be necessary, except for low level stages. How hot does the power xfmr run? Are you certain that you don't have some sort of overload situation, due to something miswired? Shorted secondary you're not using?

[audio_moksha]

Hey guys, thanks for the input.

Tweeker, the xformers are at 90 degrees to each other.

Miles Prower: My grounding, wiring etc looks fine. I havn't checked the xformer leads for continuity I'll do that today after I get home from class and see what it looks like.

[audio_moksha]

Checked the transformers for continuity and it looks like the filaments are well shielded from each other. Te power transformer is still noisy and the leakage continues on!

[BudP]

All, or most anyway, commercial power transformers are made with a very sharp eye on cost rather than low EMF.

They will have flux emissions in the following strength relationships, Measured with a gauss coil at the center of the long core side the emitted field will be 8 times stronger than the field from the short side, where you can see what look like, and are, interruptions in the core sheet edges. The coil faces are 8 times, less noisy than this short side of the core and the joint between coil and core, which you probably cannot see is eight times less noisy than the coil face. All of this means that you can aim the emitted field, through the air, more or less like a searchlight.

A complete decoupling and least application of emitted field into a steel chassis has the power transformer mounted coil down, through a rectangular hole in the chassis. The output transformer is mounted in an upright fashion, with the long side of the core against the chassis and making a T top to the long orientation of the core from the power transformer. You can actually put the two right next to each other without any hum being transmitted from power to output.

I am going to assume that you have both transformers oriented in the same direction and both have their long sides of their cores flush against the chassis. If the chassis is steel you are going to get a direct flux transfer, carried on the steel surface from the power to the output and a field through the air too. You may be able to replace the mounting end bells with ones that allow a rotation of the output or power 90 degrees so that the short core end is flush with the chassis, then rotate one of the units 90 degrees, this is a two decouple orientation and may solve all of your problems.

The removing of the output proves that it is acting as an antenna, but the lead wires can be the problem and the emitter may actually be the chassis itself. If it is cold rolled steel then the EMF applied to the chassis can actually flow down all of the chassis bends. You have to heat treat the chassis at osteonite temperatures, 1300 deg F, to close these cracks back up, When that is done the flux just makes a pool around the power transformer and does not go anywhere, on the chassis. You can still have a through the air coupling.

Are all of your power transformer wires twisted with like wires. Means the two AC mains wires are twisted together, the two or three B+ wires are twisted together and most importantly, the filament wires are twisted together, for all lengths of filament wiring, even as short as an inch. These twists need to be at least one twist per 3 inches but not more than two twists per inch. Then you should twist your output wires together on the primary side only and not more than 1 twist per three inches. Also, where your wires are running is important. I usually have all ground and signal return wires against the chassis and all high current AC wires out on the middle, away from the chassis. This includes filament wires and mains connection wires. Neat dressing here can eliminate a lot of induced noise problems.

Another trick used by many is to mount the output up on 1/4: aluminum standoffs to decouple it from the chassis. If none of the simple stuff works and you want to explore changing the mounting end bells, look in your phone book for a local transformer company and if there is one nearby call and ask if they can aid you. A guitar amplifier repair and mod shop can be another very useful place to turn to for help. If no one is helpful email me and I will drag you through the measurement process, so we can get you what you need and discuss how to do this work safely and neatly.

Bud

[EC8010]

Originally posted by BudP
These twists need to be at least one twist per 3 inches but not more than two twists per inch. Then you should twist your output wires together on the primary side only and not more than 1 twist per three inches.

Why do you suggest a maximum number of twists?

[ulibub]

Originally posted by audio_moksha
Hello folks,

...Is the transformer saturated? The power transformer is also noisy when turned on and the buzz can be heard a foot away from the amp when all is quiet. I am a bit concerned about the safety of the amp...

Besides the electric hum in the speakers - the buzz you hear "a foot away" - do I understand you right - is that a mechanical buzz from the transformer directly?

If so, you should look if the screws that hold the core together are fixed as tightly as possible. If the hum comes form the coils, then a short-circuit between several windings of one coil might be possible. Then, the xformer should quickly get hot, usually the coil first.

Anyway, loud mechanical hum of a xfmr is always a sign of poor quality, if there's no electrical reason - overload or short-circuit. Modern mains transformers, usually vacuum-impregnated, in the sizes we use them in our amps, shouldn't emit almost no audible hum.

Sometimes the chassis can give a resonance and amplify mechanical hum. I that case, some rubber below the transformer helps damping this resonance.

Uli

[audio_moksha]

Thanks for all the input gentlemen. The OPTs to the side have their bell ends flush to the chassis sides (east to west). The power transformer bell go from north to south and it is in the centre with the OPT's on either side. Also, I am using an aluminium 2mm thick chassis. I have decoupled all transformers from the chassis by using strong fibreglass. Whereever possible i have used a Belden shielded twisted pair but i have not twisted the actual transformer leads which I will do when I get done with classes today. As far as the choke is concerned it is mounted directly below the power transformer and it cannot be seen when I set the chassis on the wooden box.

i'm also going to try rotating the power transformer so that its long end does not face the OPT's. hopefully this should solve the problem.Ill get back to you after trying this. Right now its going to be hard to twist the B+ leads because as of now, the leads barely reach the terminal strips to meet the diodes; I might have to elongate the leads some in order to do this.

BudP, it funny that you mention the 3 turns per inch theoratical ideal because both Terry Cain (Cain & Cain) and Doc B (Bottlehead) have both expressed the same opinion.

ulibub, the hum sounds like it might be 60Hz or 120Hz hum its not really buzzy and I can only hear it when I have my ear against the speaker but the perfectionist in me wants the amp to be completely quiet. Also the power transformer is mechanically buzzy but it does not get hot at all.

[SY]

BudP, it funny that you mention the 3 turns per inch theoratical ideal because both Terry Cain (Cain & Cain) and Doc B (Bottlehead) have both expressed the same opinion.

Bud said 1 turn per 3 inches (IOW, 0.33 turns per inch), about ten times lower.