"Singing" output transformers

[tigro]

I replaced the output transformer on my power amplifier Antique Sound Lab
AQ1008DT. I wanted to test its performance with a scope and signal generator. I connected a 5ohm resistor as load and at 40V P-P output signal I could hear it loud and clear coming from the output transformer
starting at about 2.2Khz and going up as far as I could hear.
I checked the original transformer and the phenomenon ocurred again at a slightly different frequency.
Is this a mechanical resonance of the transformer?
Otherwise the amplifier performs very well though I am not sure if there is any difference between the original transformer and the new one.
Comments on the issue are welcome.


Regards

Tigro

[Bandersnatch]

hey-Hey!!!,
That is entorely normal. Change the magnetic flux on the steel core and it changes its shape/volume. The effect is called magnetostriction. Takes about 7% Si to get that effect at a minimum and that alloy is not so friendly either mechanically or magnetically compared to 3.5% Si.
cheers,
Douglas

[kavermei]

It can also be the core moving back and forth in the magnetic field generated by the windings, especially if the bobbin is not fixed to the core with lacquer or wax.

[tigro]

Hello

Thank you very much for your answer.To my best knowledge the whole thing is potted
but there must be some freedom because I hear the signal, thus something vibrates.

Regards

Tigro

[rknize]

Square waves will really make them sing.

[SGregory]

So will clipping a sine wave, at least on the amp I am working on.

[Michael Koster]

It's not resonance. Resonance occurs at one or more particular frequencies. It's the bobbin moving on the core or the core lams vibrating against one another, etc. due to electromagnetic forces. It could also be the core changing dimension due to magnetostriction I guess.

Potting is usually the cure for a singing inductor. Either yours is not well potted or is being overdriven or maybe that's just what you get. Most of my non-potted OPTs sing at least a little, but the potted ones are quiet (havent tried the 300W ones yet though...)

[jrenkin]

Originally Posted by tigro

5ohm resistor as load and at 40V P-P output signal

Tigro

I may not have this right, but the amp at 100W peak output power and a 5R load would at max be putting out what, 20-25 V p-p. Isn't 40V p-p pushing those OPTs a little hard?

[Michael Koster]

Originally Posted by jrenkin

I may not have this right, but the amp at 100W peak output power and a 5R load would at max be putting out what, 20-25 V p-p. Isn't 40V p-p pushing those OPTs a little hard?

I think if you convert 50V pk-pk to RMS (17.7VRMS)and use VRMS^2/Zload you get about 62 watts.

[kevinkr]

Originally Posted by jrenkin

I may not have this right, but the amp at 100W peak output power and a 5R load would at max be putting out what, 20-25 V p-p. Isn't 40V p-p pushing those OPTs a little hard?

Not quite 40Wrms.. (based on OPs original 40Vpp)

Incidentally mathematically equivalent - you can square Vp, then divide by 2, and finally divide by the load impedance and that gives you watts rms..