EEVblog Electronics Community Forum

General => General Chat => Topic started by: cdev on December 20, 2016, 05:13:16 pm

Title: Curing transformer hum by use of methylated spirits/shellac?
Post by: cdev on December 20, 2016, 05:13:16 pm
While looking for info on transformer hum I have read that some old timers faced with that situation (and having eliminated other potential causes of problems) either using an eye dropper or a soak, to some degree re-impregnate transformers with some mixture of methylated spirits and shellac-

Example

https://www.flipclockfans.com/forum/forum/clock-show-and-tell/clock-restoration-and-repair/1346-fix-for-noisy-transformers (https://www.flipclockfans.com/forum/forum/clock-show-and-tell/clock-restoration-and-repair/1346-fix-for-noisy-transformers)

Does anybody have any experience with this?


Hummmm.. just found this old post (https://www.eevblog.com/forum/beginners/philips-pe1542-transformer-hum/) here..   Also, re-mounting the transformer may help.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: aargee on December 21, 2016, 02:42:41 am
We used to put the transformer in a pot of varnish/shellac/etc and stick it in a bell jar attached to a vacuum pump - just one of the venturi types that attach to a tap (faucet).
Keep the vacuum going until no bubbles rise to the top, every little nook and cranny will be full of the stuff.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: BradC on December 21, 2016, 03:07:20 am
We used to put the transformer in a pot of varnish/shellac/etc and stick it in a bell jar attached to a vacuum pump - just one of the venturi types that attach to a tap (faucet).
Keep the vacuum going until no bubbles rise to the top, every little nook and cranny will be full of the stuff.

Huh. I do the same thing with Vodka and Bananas.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: Sceadwian on December 21, 2016, 04:26:34 am
You need to keep it really thin so that it gets into the coils but still can evaporate to cure, not sure if it's possible to get really ultra low viscosity epoxy? You'll never 'cure' the hum but you'l probably reduce it.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: aargee on December 21, 2016, 04:39:43 am
We used to put the transformer in a pot of varnish/shellac/etc and stick it in a bell jar attached to a vacuum pump - just one of the venturi types that attach to a tap (faucet).
Keep the vacuum going until no bubbles rise to the top, every little nook and cranny will be full of the stuff.

Huh. I do the same thing with Vodka and Bananas.

Does it stop them humming? (Are you thinking what I'm thinking B1? I think I am B2)
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: cdev on December 21, 2016, 04:50:20 am
One of the posts I saw seemed to say that one could do a much more limited treatment with just a few drops of varnish/methylated spirits when old transformers developed a hum, and often after the transformer dried out it would be found that would cure acquired transformer hum.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: tautech on December 21, 2016, 05:01:42 am
You need to keep it really thin so that it gets into the coils but still can evaporate to cure, not sure if it's possible to get really ultra low viscosity epoxy? You'll never 'cure' the hum but you'l probably reduce it.
Thermosetting epoxies and phenolics are used for transformers to encapsulate them. They're specialty products.

http://www.globalspec.com/learnmore/materials_chemicals_adhesives/electrical_optical_specialty_materials/insulating_varnishes_impregnating_resins (http://www.globalspec.com/learnmore/materials_chemicals_adhesives/electrical_optical_specialty_materials/insulating_varnishes_impregnating_resins)

Practical transformer winding:
https://ludens.cl/Electron/trafos/trafos.html (https://ludens.cl/Electron/trafos/trafos.html)
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: BradC on December 21, 2016, 07:16:15 am
We used to put the transformer in a pot of varnish/shellac/etc and stick it in a bell jar attached to a vacuum pump - just one of the venturi types that attach to a tap (faucet).
Keep the vacuum going until no bubbles rise to the top, every little nook and cranny will be full of the stuff.

Huh. I do the same thing with Vodka and Bananas.

Does it stop them humming? (Are you thinking what I'm thinking B1? I think I am B2)

Dunno, after the first Banana I hum myself, and yes B2 I think I'm thinking what you're thinking.

On topic though, vacuum impregnation with wax is another method that works pretty well. As with all of the vacuum impregnation techniques you do have to manage the vac level so you don't boil your encapsulant. I've done some pretty bad damage to epoxy and 2 part elastomers by getting careless with the vac and boiling it. So much so I built a vacuum controller so I can dial in the required vacuum and a set of solenoids makes sure it stays where it is supposed to. Helps a lot for degassing, vacuum moulding and vacuum filtering sensitive liquids. (plus bananas & watermelons). Vacuum is also a wicked way to get a great steak marinade done in minutes rather than hours, and a great way to cheap red wine (at room temperature) into quite a drinkable brandy / lawnmower fuel.

Don't use the same pot you use for epoxy/wax to do the food or grog, and don't use a fragile container. A vacuum implosion can do a *lot* of damage, make a lot of noise and upset the missus.

RS sell nice little air/venturi vacuum devices for a couple of 10's of $$. So you don't need a machanical vac pump to do it either.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: calexanian on December 22, 2016, 01:24:16 am
This sort of thing works for magnetic mechanical hum. Induced hum can be cause by too much leakage inductance. This is a function of poor transformer design. More exciting current than necessary under light load via low primary inductance in usually the culprit. It can induct hum right onto adjacent ferrous things such as the chassis.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: David Hess on December 23, 2016, 06:52:28 am
As with all of the vacuum impregnation techniques you do have to manage the vac level so you don't boil your encapsulant. I've done some pretty bad damage to epoxy and 2 part elastomers by getting careless with the vac and boiling it.

Use the vacuum chamber to remove the gasses from the encapsulant as a separate step before using it on the transformer.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: BradC on December 23, 2016, 07:43:18 am
As with all of the vacuum impregnation techniques you do have to manage the vac level so you don't boil your encapsulant. I've done some pretty bad damage to epoxy and 2 part elastomers by getting careless with the vac and boiling it.

Use the vacuum chamber to remove the gasses from the encapsulant as a separate step before using it on the transformer.

Yes, but at higher vaccum levels you actually boil components out of the encapsuant, rendering it "not quite as good as it should have been". Most vacuum aparatus will pull a deep enough vacuum to do damage so you need to temper that. It took me ages to figure out that the "gas" escaping from the epoxy wasn't just very well trapped gas but components boiling (I was using a 2 stage oil sealed vane pump though). Probably not that important for transformers, but when making replacement elastomer bits for printers damaging the rubber can have significant long term effects.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: calexanian on December 24, 2016, 02:25:13 am
yeah, a two stage rotary pump is a little too much. typically only a few inches of vacuum is needed. Diaphragm pumps are typically used.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: cdev on December 24, 2016, 03:10:13 am
The end one (the one on the left if you are looking at the front) is the transformer whose hum I'm trying to cure. Its part of a Sorensen triple bench supply.  (the unit by itself would be an XTS7-6) It works, it stays cool as a cucumber while delivering its rated 7 volts at six amps right into a current shunt and meter.. it doesn't even get hot..

 But it hums.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: rrinker on December 24, 2016, 05:01:14 am
 Maybe it just doesn't know the words?

Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: BradC on December 24, 2016, 05:10:18 am
First thing I'd do is nip the screws up. If that fails, put a bit of tube in your ear and use that like a stethoscope to narrow down the source of the hum to the laminations, windings or something nearby and go from there.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: SeanB on December 24, 2016, 06:17:26 am
Likely leakage field inducing motion into the steel backplate. Simplest cure is to mill some air cooling slots that run parallel to the laminates, so the current path is very long.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: calexanian on December 24, 2016, 06:33:02 pm
Or try wedging some rubber sheet or foam in there. Sounds like the transformer is perfectly fine and you just have, as said before, some induced or mechanical hum. Physical dampening would be best. Just push stuff with a rubber eraser until it quotes down and then figure out what to do from there.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: cdev on December 24, 2016, 07:31:27 pm
 SeanB Thank you for the tube suggestion. Its a useful technique.

Its clear the voltage regulator chips are getting hot. Whether that is abnormal? Maybe not.

Kind of interesting..  I brought out a liquid crystal sheet I have which shows temperature.  (remember "mood rings" ? Like that.. Which shows that visually. (Photos below)  This is basically "poor mans thermography" and in some ways its even more useful than with a $500 camera..

And then I had to clear off the kitchen table for now. So am reading up on voltage and current regulation (https://web.archive.org/web/20060203095838/http://www.national.com/rap/Application/0,1570,24,00.html) ..

The LC sheet is a sample from LCR Hallcrest -  http://www.liquid-crystals.com (http://www.liquid-crystals.com)


Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: tautech on December 24, 2016, 08:01:10 pm
Does the hum change with load ?

As you've a IR gun check a few more components for heating....caps drying out ?
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: cdev on December 24, 2016, 08:14:56 pm
No, not at all but then again,  is the internal shunt on my meter a good load? it would seem to be..   The meter is rated for 10A current measurement but - probe wires were getting warm. Its an old Radio Shack pre-digital VOM.  I also have a UT61e

What is the best kind of load to use?

maybe a light bulb or a longer piece of copper wire?   The current meter also works. (the voltage side is pinned at 9.99 volts, but the voltage does adjust as its supposed to.. and there is a linear "row of 10 LEDs" volt and current meter which works.)

It also has "sense" wire connections. Its a very nice power supply. Definitely worth the $60 (plus almost $40 shipping) I paid..

the caps look as old as the unit (1986!) but they are not getting hot or bulging, as far as I can tell.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: tautech on December 24, 2016, 08:30:06 pm
Whether it's related to the hum or not I'd be concerned at the temps those 317's are running at.  :scared:
Something tells me they shouldn't be running that hot.........what's drawing the current ?  :-//
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: cdev on December 24, 2016, 08:50:00 pm
Regardless of ESR status, perhaps those 40 year old Sprague caps all should be replaced.  I think I will send off for replacements, might as well.  I might as well get it out of the way.

The 7V 6A unit has four, the other two units each have two.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: tautech on December 24, 2016, 08:58:11 pm
Regardless of ESR status, perhaps those 40 year old Sprague caps all should be replaced.  I think I will send off for replacements, might as well.  I might as well get it out of the way.

The 7V 6A unit has four, the other two units each have two.
As they're in a linear PSU ESR is not as important as real value and leakage.
They were a good brand decades ago, I think I'd just test them then make the call.
They should hold a charge for an hour or so but to get some idea, leakage specs are in datasheets of any of the better cap manufacturers and you can do a few simple calcs to find out roughly what it should be.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: Zero999 on December 25, 2016, 03:55:57 pm
As with all of the vacuum impregnation techniques you do have to manage the vac level so you don't boil your encapsulant. I've done some pretty bad damage to epoxy and 2 part elastomers by getting careless with the vac and boiling it.

Use the vacuum chamber to remove the gasses from the encapsulant as a separate step before using it on the transformer.

Yes, but at higher vaccum levels you actually boil components out of the encapsuant, rendering it "not quite as good as it should have been". Most vacuum aparatus will pull a deep enough vacuum to do damage so you need to temper that. It took me ages to figure out that the "gas" escaping from the epoxy wasn't just very well trapped gas but components boiling (I was using a 2 stage oil sealed vane pump though). Probably not that important for transformers, but when making replacement elastomer bits for printers damaging the rubber can have significant long term effects.
How about reducing the temperature for the vacuuming stage? If you put it in the freezer, the liquid will boil at a lower pressure. I0m, aware that it won't cure at low temperatures but once it's been vacummed, the curing can take places at higher temperatures.
Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: cdev on December 25, 2016, 08:31:39 pm
So, was scanning the old posts on sorensen power supplies and found this old post (https://www.eevblog.com/forum/chat/latest-acquisition/msg39636/#msg39636) by tekfan in a 2011 thread about the same power supply as I have, which describes preregulation.

Edit: reading the theory of operation in an online PDF (https://doc.xdevs.com/doc/SORENSEN/XANTREX%20XT%2060%20WATT%20SERIES.pdf) of the 2001 version of the manual (missing schematics) it doesn't mention pre-regulation as far as I can tell.

:(


Title: Re: Curing transformer hum by use of methylated spirits/shellac?
Post by: johansen on December 26, 2016, 06:30:32 am
So, was scanning the old posts on sorensen power supplies and found this old post (https://www.eevblog.com/forum/chat/latest-acquisition/msg39636/#msg39636) by tekfan in a 2011 thread about the same power supply as I have, which describes preregulation.

Edit: reading the theory of operation in an online PDF (https://doc.xdevs.com/doc/SORENSEN/XANTREX%20XT%2060%20WATT%20SERIES.pdf) of the 2001 version of the manual (missing schematics) it doesn't mention pre-regulation as far as I can tell.

:(

the preregistration discussed there is using an SCR as a switch (like in a buck converter).. and you need an inductor for energy storage. its going to make a lot more noise than a transformer and such systems should be easily identified compared to traditional linear power supplies. for example, the pre regulator and the linear regulator should be mechanically and electrically isolated from each other, and a large inductor, approximately the size of the main supply transformer is required. wouldn't supprise me if you find two additional transformers to provide power for the control circuitry, one for the pre regulator and one for the linear regulator.