Transformer noise

[IvoS]

I am using BTA24-800 triac and MOC3041 optocoupler to turn toroidal transformer ON/OFF. When unloaded, there is really loud noise coming from the transformer. When loaded, the noise goes partially away. If I power the transformer directly from 120V AC there is no noise. So the triac is the problem. Is it a DC that is being generated by the triac a problem? Or what is going on?

[GoneTomorrow]

Probably the triac is driving the transformer with a square(er) wave. Mains will be a near-perfect sine wave. A square wave will create more noise due to the much shorter rise time.

You can attempt to mitigate noise by damping the transformer, eg heatshrink, epoxy, hot snot, etc. (keeping operating temps in mind)

[Mr.B]

You don't mention if you are using the snubberless version - BTA24...W
I thought triacs switching inductive loads needed snubbers.

[Richard Head]

Use a relay to switch the transformer. Triacs will struggle with the varying phase angle of the load and the holding current and the possibility of flux walking due to possible asymmetry of excitation voltage. Toroidals are sensitive to this as they have tape wound cores that have no airgap at all resulting in a high remanent flux.

[IanB]

I am using BTA24-800 triac and MOC3041 optocoupler to turn toroidal transformer ON/OFF. When unloaded, there is really loud noise coming from the transformer. When loaded, the noise goes partially away. If I power the transformer directly from 120V AC there is no noise. So the triac is the problem. Is it a DC that is being generated by the triac a problem? Or what is going on?

Are you trying to run the transformer from a phase angle dimming circuit? Noise is usual from such circuits at low power settings. Even incandescent bulbs make a humming noise when you do that.

But "dimming" an inductive load like a transformer is probably not a good idea.

[GoneTomorrow]

Oh, you mean turning on the transformer as in like a switch to the mains? For  some reason I had in my head that you were "switching" the transformer as in switchmode PSU type switching 

In that case (or many really), triacs aren't ideal for switching on inductive loads. As above, try a relay instead. You've already proven this will solve the problem.

[Andy Watson]

Is it a DC that is being generated by the triac a problem? Or what is going on?

I think DC is the problem. I think the zero-crossing opto is compounding the problem because (my best guess): When the transformer is unloaded it presents a largely inductive load for the triac. Say that triac is conducting for the positive half-cycle. When the supply voltage crosses into the negative half-cycle the triac is still conducting - because the current is maintained due to the inductance of the transformer. At some point in the negative cycle the current decays to zero and the triac turns off - if the voltage at this point is greater than the detection voltage of the opto, it (correctly) holds off triggering until the next zero-crossing at the beginning of the positive cycle. And repeat. So the transformer is effectively being feed with only positive half-cycles.

[IvoS]

Yes, I forgot to mention that it is a snubberless triac. Although, I don't think it makes any difference. I think this triac is just more robust and can handle inductive load. So, is it the optocoupler being fooled by the inductive load that screwes up the correct "zero-cross" detection and therefore it doesn't turn on-off for positive and negative cycle properly....??
No phase shifting circuitry is used for dimming or anything like that, straight rock solid ON and OFF only.

There is a suggestion for snubber in MOC3041 datasheet on page 7
http://www.mouser.com/ds/2/149/MOC3041M-195885.pdf
would that help?
Thanks.

[IvoS]

OK, so maybe non-zero cross opto isolator would solve the problem.