How to cure Magnetostriction?

[Mr. Lime]

Hi guys,

in my 18W amp with Trainwreck kind of preamp and 6V6 tubes I think I have the problem of magnetostricion.
The output transformer seems to sing along with the speaker and I wonder what's responsible for that?

I also think that this causes this annoying high end fizz that can only be removed with a conjunction filter (cap between inverting and none-inverting outputs of the PI).
With the filter the amp gets to dark to cut through in a band.
The OT I use is a Hammond 1750PA rated with 18W so I thought it might be fine with the 6V6?

Is the PI, the 6V6s or the OT driven to hard?

Maybe someone has experience and can give some guesses?

Thanks a lot!

[sluckey]

The OT lamentations may be loose. Try tightening the screws. That OT should be OK for 6V6s.

[Stevem]

All output transformers sing audibly like that at a certain output level, if your getting interstage reactions then it's due to the lenght and or position of the transformers output and plate input leads.

On a few 1950s Gibson amps where the plate leads ran along near the complete lenght of a 18" chassis they went so far as to sheild those 3 wires even with the low gain that those amps had!

I have fixed other people's amp builds that had a oscillation issue caused by the mounting location of the Marshall type impeadance selector switch at half volume .

To prove out that is not a preamp section that's making your fizz issue take a 5 pf to 500 pf cap of the needed voltage rating and solder it in from the plate to inout grid on each preamp to and see/ hear if it makes a difference.

[Mr. Lime]

Thanks for the replies!

Well I thought I read somewhere that this happens when specific parts are undersized or the bass frequencies too dominating.
As long as it won't hurt the amp I'm fine with the OT singing..

Yes, the OT wiring could be shorter, it is my first DIY amp and there's definitely room for improvements for the next one.

Amp OT Placing.jpg

Oscillation isn't a problem at all.

Concerning the fizz, I have tried the power amp with other preamps inserted at a FX loop return feeding the PI and they are way too bright as well.
I don't think it's a preamp problem.
Here's my power amp, what's missing is a bootstrapped double MV on the PI instead of R14 and R15 which helps to tame the highs but of course I would like to have it working with the MV full on which is basically what the schematic shows. At the moment the nfb wire is removed, I like the whole amp more without it.

PA schematic.png

Maybe it's because of PI imbalance?

Thanks a lot!

[R.G.]

Magnetostriction is a property of the transformer iron itself. It's not curable. It's worse at high flux densities, hence the technobabble about not overdriving things.

Loose laminations are vibrated by the changing magnetic field. Varnish is supposed to help with this, but if they're loose, they will sing. Tighten bolts.

Copper coils move and vibrate around due to Lenz' Law (force on a conductor in a magnetic field, even a field caused by the conductors). This is another thing that impregnation with varnish is intended to help with. Well impregnated coils mostly don't sing.

Panels resonate. Magnetostricting iron can vibrate chassis sections. Just like in speaker cabs, the solution to a vibrating panel is to make it stiffer (driving any resonances way high in frequency) and damp with soft, sound-sorptive stuff.

[Mr. Lime]

Thanks for the insight!

I have another 20W OT with more iron now coming to me the next week and I will try if replacing the OT makes the situation better.

[Stevem]

You need a osciliscope to trace out in what stage the issue is stemming from , not new OT iron!