Power transformer magnetic coupling issue. Solution: toroid?

[soapfoot]

Put together a little build with UTC A-20s on the outputs and a small Hammond 16V power transformer.

It turns out they're a bit too close together (see pics) and there's some hum coupling going on... worse in the right channel (power transformer is a bit closer to that side).

It's not extreme, and it borders on "usable," but I'd like to improve it as much as possible.

Rotating the power transformer along its axis (orienting vertically instead of horizontally) didn't seem to help.

It just occurred to me that I haven't tried rotating the A-20s yet, so I'll try that next.

In any case, would it be advisable to go to a toroidal power transformer in this case? Would it be likely to afford a substantial enough reduction in electromagnetic interference to be worth the trouble? My experience with toroidal power transformers is very limited.

Thanks for any insight!

 

[mjrippe]

Did you listen while rotating the transformer?  Sometimes there is a spot more or less than 90 degrees where hum cancels.

Alternately, if there is room you can move the audio (or power!) transformers inside the case.  There are also magnetic shields for the A series called A-33, but they are not cheap.

 

[abbey road d enfer]

I see you have space inside the case, so you may want to move the xfmr in.
Using a toroidal is not a guaranteed improvement. Actually some winders are taking advantage of their shape to increase the magnetizing current, which is what produces the magnetic field, because it allows them to increase the nominal power.
You could use a xfmr with a 230V (2x115) primary and a 2x18V secondary. It would run at low magnetizing current.
You would use a model with nominal power 1.5 times the existing.

 

[soapfoot]

Did you listen while rotating the transformer?  Sometimes there is a spot more or less than 90 degrees where hum cancels.

Alternately, if there is room you can move the audio (or power!) transformers inside the case.  There are also magnetic shields for the A series called A-33, but they are not cheap.

Thanks everyone

I did try rotating it while listening... it didn’t get significantly better in any mountable orientation (maybe slightly better when stood up on end)

I do think I could find room to mount it inboard, which might be the answer. Then I could mount it on the chassis floor in the quieter orientation (and it would be further from the I/O transformers). Only concern there is whether it then gets too close to low level audio, but I can experiment.

OTOH, those A-33 cans would sure be an easy solution. If they truly attenuate by the 20dB claimed, that would be plenty, and I could keep my chassis nice looking.

Not too many for sale right now, though.

But seems like either course would be better than a toroid?

 

[mjrippe]

Thanks everyone
...

But seems like either course would be better than a toroid?

Well, if you have a torroid on hand it is certainly worth trying but I wouldn't order one thinking it is a guaranteed fix. 

 

[soapfoot]

Okay! I happened to run across a pair of those A-33 shielding cans for a good price, so I'll wait for those to arrive, slip them on, and re-evaluate. That might well be the path of least resistance.

Otherwise, I'll move the PT inboard and re-evaluate. I should probably try one variable at a time just so I have an honest sense of what I'm doing.

 

[soapfoot]

Update:

Moving the power transformer inboard and central (rotated such that it's mounted on the floor) has completely solved the issue. There's only a very slight amount of magnetic coupling remaining; it's only evident with all controls at minimum, no input connected, and headphones completely floored. I can't get it to even meter in my DAW.

As soon as any control is turned up or any input device connected, any hum is totally swamped by the inherent noise floor of the circuit.

I still have the A-33 cans on the way, and will attach them when they arrive to see if they improve matters even further, but I consider the current state to be more than acceptable; quite good even.

 

[emrr]

Great.  My general observation has been you can have a power transformer closer to ANYTHING ELSE other than an audio transformer with lesser results.  Seems to have been true here too. 

 

[soapfoot]

The inverse-square law never ceases to amaze me. Just barely doubling the distance and changing the orientation moved me from "obvious problem" to "zero problem."

 

[mjrippe]

Excellent!  And the layout really looks like it was meant to be that way  ;D

 

[mjrippe]

Quick follow-up.  I was about to mount some open frame Jensens inside a piece with some huge coils and magnets and I thought about this thread.  Fortunately I have this little device that came with my tape head degausser, a magnetometer.  It clearly showed that the space for the transformers was inside the magnetic field of the other components, so I mounted them externally.  ;D

Edited: Initially I thought inside was OK, but then I turned the magnetometer on a different axis!

 

[soapfoot]

Ah! Very cool