I'm getting some significant snapping and crackling in certain spots when I adjust the variac setting in a piece of equipment I've been working on. I've cleaned the tracking off of the gold surface and adjusted the brush tension but the issue persists. It only occurs under load. Am I right in suspecting that the carbon has simply worn down too far and is starting to bridge certain windings?

Judging by the charring on the sides of the carbon I would venture a guess that the arc energy is high enough to be pitting the gold. Not something that I can just sweep under the rug.

I'm asking for a second opinion because replacement brushes are spendy little bastards - nearly more than I paid for the equipment itself. Alternatively, would it be worth trying to file down the taper to shrink the contact surface back into tolerance?

Thoughts?

 

Has the tip of the brush worn so it is too wide shorting several turns of the variac together ?

Les.

 

Unplug it and spray all contacts with WD40. Wait 30 minutes before powering it up again. This worked for me.

 

It looks like the wires themselves could do with a little bit of a clean up (to remover the black muck that is on them), which might help the connection. But you should maybe take the brush out and see how it looks, just a bit of a spring clean on it could work wonders.

 

I removed the brush and brush carrier for closer inspection. I found some wear marks in the carbon. Significant but not terrible. The pattern tells me that there are some raised spots on the windings from pitting. I managed to uncover some data on the brush for this particular transformer. The spec for it is 920 thou and mine measured in at 960. That's minimal wear for 50 years old. Both of these observations lead me to believe that this transformer was subjected to multiple overloads during it's life. I also noticed that the guide on the aluminum brush carrier was oxidized from contact with the brass brush cartridge.

I dressed the contact surface of the carbon and then ground the taper down to 870 thou using some 3000 grit carbide on a flat surface. I also cleaned up the brush cartridge and guide then applied a thin film of oil to both surfaces.

The brush now glides freely like it's supposed to and the arcing has been noticeably reduced but still persists. I turned off the lights and ran the brush up and down with a load on it and noticed repeatable arcing when transitioning between certain pairs of windings. This tells me that their surfaces are no longer true.

So ultimately this transformer has finite life left in it. The arcing will continue to get worse until the thing becomes unusable. The only way to true up the track would be to surface-grind it flat and then plate it with fresh gold... unless I dare try sanding it. The former is too rich for my purposes and the latter is far too likely to ruin the gold.

Moral of the story: DON'T overload your variacs! You'll distort the track.

 

Moral of the story: DON'T overload your variacs!

Mine has a built in fuse at the rated current for the device.
Max.

 

Mine has a built in fuse at the rated current for the device.
Max.

Unfortunately the previous owners of mine had no such foresight. What a shame to see such a finely made transformer on it's last long mile.

 

The part that appears to be a plain carbon brush on a variac brush is a special type of material -anisotropic graphite- that if goes bad; would need replacement, I doubt sanding or reforming can fix it. It conducts mostly in one axis to avoid shorting windings that are simoultaneously contacted and should be oriented properly.
The price of those brushes tell they are not just any carbon brush ---->http://www.carbonbrush.com/variacbrush.htm
And some explanation: ----> https://sound-au.com/articles/variac.htm#s1

 

You're suggesting that the arcing could be due to a failed anisotropic matrix within the carbon? What could be the cause of such a molecular restructuring? Heat?