How to fix a buzzing transformer

[mikeydee77]

I have a trace elliot super tramp where the transformer is making a very loud buzzing sound. I can see that the thinner wires are a little loose in places so assuming that a varnish treatment might do the job. Many years ago I used to work in a place repairing VDU and we would spray the line output transformer with hair lacquer. Not sure how effective it was long term but it definitely was a short term fix.

I read about dunking in varnish and then drying in the oven. Can anyone confirm this is a sound approach and provide a recipe?


M

[Jon Snell]

It is not the windings that make a noise it is either the lamminations or the lack of fibre washers causing a 'shorted turn' in the fixing bolts, in turn causing Eddy Currents that dissipate into vibration.

[mikeydee77]

Thanks Jon, didn't realise that.

There are two securing bolts but six available holes. The two bolts have fibre washers on one end only.

I guess I could try more bolts to compress the laminates.

Do you think fibre washer at one end is OK? Surely both ends better.

[mikeydee77]

Here is further pic

[Helmholtz]

Eddy currents in mounting bolts are insignificant even if not insulated.
Transformer buzz is either caused by loose laminations or loose wire turns.
Reason is magnetic forces.
These are repelling only between laminations and attracting only between wire turns. Also there are magnetic forces between windings and core.
In both cases line frequency is doubled.
In my job we had a case where a ferrite core choke was buzzing solely due to the windings. Problem was cured by potting the bobbin.

[Jon Snell]

Here is further pic

Most new transformers have fibre washers at both ends, infact some have sleeves over the threads of the fixing bolts. This eliminates Eddy currents that are a cause of rattling.
The current losses can be measured and tightening/slackening the fixing bolts with no fibre insulation washers can be measure with a current meter and recorded as loss.

Many years ago and the last run of colour TVs with tubes, there were many tuned circuit chokes in the line and frame power stages that used to whistle and whine. These were cured by applying nail varnish.
Hitachi and sets made by Vestel were the worst offenders.

[J M Fahey]

If varnish dunking, it must reach deep.

Outside spraying is only cosmetic.

Best is to get dedicated transformer insulation rated varnish, but if not, a good substitute is good quality "marine" or "outside deck" varnish.

Not silly "water based" or "ecological" type, water in a winding is your enemy, just the old fashioned solvent based type, which gets thinned with some turpentine, (natural or synthetic) say adding 20% or so, so as to make it more liquid.

You put transformer in any suitable container, fully submerge it, and leave it at least a couple hours so it reaches *deep* , no more bubbles exit.

Then you pull it by the wires which you have bundled together and hang it from a broomstick over 2 chairs over the varnish can until excess drips out.

Leave it there overnight, then a couple days extra so it fully dries on its own.

No need to setup an oven for "just one", do not use the Kitchen one because it will stink forever, Wifey will kill you for good reason.

If OCD, you *can* get real transformer varnish, buy the air dry type, to avoid the oven.

https://mgchemicals.com/distributors/usa/

[mikeydee77]

Thanks Juan
That makes sense and double thanks for the oven tip no need to rock the boat.

In UK it is about 3 degrees so may need to adjust drying times accordingly. In summer I would put in our glasshouse with tomatoes to get the nice varnish smell out of the house.

[Mick Bailey]

I've never been wholly convinced that eddy current losses are of any consequence in a guitar amp. Peavey used to weld the core to a baseplate right across the laminations on both corners and there are plenty of amps where there are no insulating bushes or washers.

My own experience of trying to fix buzzing transformers has been mixed. The best penetration of varnish is under vacuum, but even this doesn't guarantee results because solvent accounts for most of the varnish. If the transformer is really poorly made, once thoroughly cured the buzz can still come back, but usually not as bad. I gave up with proper transformer varnish because it's expensive and the whole can goes off quite quickly once opened.

A good soak though in polyurethane varnish is probably the best you're going to do unless you go down the route of making a vacuum pot - all too involved.

[mikeydee77]

Thanks Mick
This transformer looks in good condition on the laminates with no visible rust so I will give it the varnish treatment and report back in a few days on the results.

I also saw a few YT videos illustrating using clamps to identify if it is a lam problem or wiring problem so I might experiment with that first just to create an illusion of scientific rigour even just for my own peace of mind.

[mikeydee77]

Eddy currents in mounting bolts are insignificant even if not insulated.
Transformer buzz is either caused by loose laminations or loose wire turns.
Reason is magnetic forces.
These are repelling only between laminations and attracting only between wire turns. Also there are magnetic forces between windings and core.
In both cases line frequency is doubled.
In my job we had a case where a ferrite core choke was buzzing solely due to the windings. Problem was cured by potting the bobbin.

Thanks for the very well explained info. I will measure the frequency and let you know as this buzz is a horrible frequency, definitely not hum

[Helmholtz]

The current losses can be measured and tightening/slackening the fixing bolts with no fibre insulation washers can be measure with a current meter and recorded as loss.

Nice theory, but could you really measure a difference?
I tried and couldn't.
No measurable difference in inductance and losses between no screws, insulated and non-insulated screws.
I think the reason for the sleeves and the fiber washers is mechanical damping of screw vibrations and reduced vibration transfer to the chassis.

[Helmholtz]

I will measure the frequency and let you know as this buzz is a horrible frequency, definitely not hum

As the magnetic forces are non-linear the buzz will also contain higher harmonics of 100Hz.
What frequencies you're actually hearing depends on the resonances of the main radiating parts, including things like end bells, mounting frames and last not least the chassis.
For such tests I like to use a stethoscope.

[J M Fahey]

Magnetic lines travel through the shortest possible path, so at corners and top/bottom of center leg flux is minimal.

Not zero but so low as to become irrelevant.

That´s why factories punch holes *there*.

[mikeydee77]

As the magnetic forces are non-linear the buzz will also contain higher harmonics of 100Hz.
What frequencies you're actually hearing depends on the resonances of the main radiating parts, including things like end bells, mounting frames and last not least the chassis.
For such tests I like to use a stethoscope.

Yep I had an old metal stehoscope that was for listening to engine pistons... was very effectinve in picking up the sound. I forgot to measure the noise before I took to dunking it. It is drying now so will report back by Monday, fingers crossed.