transformer isolation

[dmitriy167]

Hi,

I've been building a p3a (80w) for some time now, and finally got to the boint that it works. Almost. The toroidal transformer (Avel Lindberg Y23 35-0-35 250VA) is giving off some nasty RFI which the input picks up, and as a result, one channel hums when the input wire is in the position of where I want it to be. I moved the wire around and found a position where there is little noise, but I'd rather not have any at all. Does anyone know of a good way to shield the transformer so thre is no noise?

I can post pics if needed.

 

[DonoMan]

I'd say try caging the transformer, put some metal around it... or you could get a new transformer with less bandwidth... You can also try a small non-polarized capacitor on the mains (like .1uF or something) - just make sure the voltage rating is really high (I'd say like 1000v, just so it won't easily burn even if your amp gets a spike)

 

[dmitriy167]

I have some metal foil tape, wondering if that will be fine. Do I need to connect the tape to ground?

 

[DonoMan]

I'm not sure if it will be enough... Can't hurt to try it, though. And go for it on the ground, can't hurt either, as long as you don't short anything with it

 

[dmitriy167]

Will try that and the .1uf cap. Hope it makes a difference; shielding the inputs certainly didn't.

 

[richie00boy]

Is your problem RFI or hum? These are two totally different things.

Have you isolated your signal input sockets from the case? Are you running screened leads from the signal input sockets to the PCB?

Aluminium foil/tape will do nothing for magnetically induced noises.

 

[DonoMan]

I assume he is sure that it's the transformer, since he said moving the wires around changes it...

 

[dmitriy167]

It definitely is the transformer. The input wire has a position where it picks up almost no noise. My signal input sockets are isolated and the leads are coaxial.

 

[Mr Evil]

If the transformer is radiating noise, then there must be noise entering it. Thus I would suggest that your best bet is filtering the incoming mains.

 

[dmitriy167]

I put a .1uf cap on the mains, parallel to the trafo. No change; noise is still being radiated.

 

[Chris Jennings]

Most of the radiated field from a power toriod comes from the area around where the leadout wires exit the winding, often pickup can be greatly reduced by rotating the transformer such that this area is facing away from the input.

Chris

 

[dmitriy167]

I pulled the input leads as far away from the trafo wires as I could, and whaddya know, almost no noise. I better fix the wires in that position and rethink my input/output layout. But is there a way of actually getting rid of the radiated field?

 

[DonoMan]

Like I said before, a capacitor should help. But other than that, all you can do is try to keep the field contained.

 

[EchoWars]

What is it with these Avel Lindberg transformers? On my DoZ amp, nothing I tried worked, including $50 some-odd dollars for TI shield and some mu-metal. I probably built 3 or 4 different styled shields for it using suggestions from here and elsewhere, and I could attenuate it, but not eliminate it.

Transformer is now mounted in its own enclosure about 18" away from the rest of the amp. Nothing else worked.....

 

[dmitriy167]

I think they have a QC problem with their cores on which the wires are wound. They probably assemble them from two halves, since they can't wind the wire on a circular one with a machine. If the core is cracked, the trafo might leak a lot of magnetic energy, but still do its job. Maybe they screwed up a whole batch of the trafos this way, though I hope they didn't.

 

[EchoWars]

If that were the case, seems like you would expect a lot of mechanical noise. However, the transformer is silent...it's just emitting enough EMI to jam a Russian fighter jet.

 

[mcshaggy]

uh oh... im building a p3a and have an avel lindberg 25-0-25 250va transformer... i hope im not in for a nasty treat too (just waiting for the pcbs to come in)

 

[Jean]

I have avel transformers 160va mounted about 5" inches away from the p3a boards and don't have any problems with noise at all It might be grounding problem more than RFI.

 

[AR2]

Hello Dmitriy,
I like your avatar - Elektronska Industrija.

 

[sam9]

Depending on the layout, it might be easier to slip a length of tubular copper mesh over the input wiring. Then connect one end of it to ground. You can definitely get it from Michael Percy but probably from other sources like digikey.

I've always made a point of shielding the input for a number of reasons, not just isolation from the transformer, and so fasr I've not suffered a problem attributible to input wire picking up errant signals of any kind. Maybe I've just benn lucky but my experience is what it is . . . .