Two Variacs - Claude Lyons 8A, Slidup 5A

[mark_in_manc]

Well, two dead Variacs made it onto the bench today.

The Lyons one used to work, but then started arcing and only worked over a narrow range of outputs in mid-range.

At first I had a vague memory of dropping it, and thought the coil had shifted downwards relative to the brushes - the coil sits in a kind of cradle and can be shimmed up towards the brushes (see pic). Moving it up a bit restored operation over the whole range, but with lots of arcing and smoke - I switched it off quick!

It turned out that what I thought were brushes were in fact metallic brush holders. The brushes themselves are tiny, and were scrubbed away to nothing. So I made two more from some dead angle grinder brushes - and now it all works nicely.

[mark_in_manc]

The 'Slidup' 5a variac came with the big pile of stuff which I've been selling for charity - and seemed unhappy. The knob went round and round, and it sounded like there were loose wires inside.

Having taken the top off, it seems both wires leaving the coil have snapped, though with long ends which can be re-attached. Also the end-stop which should be there, isn't - the thing is made of a kind of black Bakelite and is all rather brittle.

So I've made a new stop from Milliput epoxy putty, and we'll see what happens when it goes off.

[Lucien Nunes]

Did you investigate the anistropy of resistance of the carbon at all? I have always understood that variac brushes were made of material with an especially high resistance along the axis tangential to the core, to minimise circulating current in the shorted turns. How true or important this is I have never got around to investigating by experiment.

[mark_in_manc]

The remnants were so small, I didn't attempt it. The brush dimensions seem to have been chosen to address the problem you mention - they are only about as thick as the wire used for the winding - so at most you'd have two adjacent shorted turns.

One of the 'new ones' snapped almost straight away - the carbon is pretty soft. Luckily the bit that broke off was long enough to re-install, and being shorter there is less of a bending moment acting on it as one rotates the wheel. I would guess it's not worth making these with more than a couple of mm poking out of the holder.

I have a big lump of what looks like graphite rod - 2" dia by 8" long. I wonder if this would work for home-made brushes.

[mark_in_manc]

Well, the 5A Japanese variac works too. It's nowhere near as nicely engineered as the Lyons unit, which appears designed to go into battle - on this one the bakelite tabs which 'hold' the coil have snapped, and it seems free to vibrate - it certainly makes a fair row at 100Hz when plugged in. But, it works. I must now read up a little on auto-transformers and convince myself I understand _how_ it works

[TrevorG3VLF]

I thought that a Variac brush consists of two pieces of carbon glued together to give the anisotropy mentioned. I made a brush for a 1A Variac from a 6B pencil, it seems to work OK at very low currents.

I understood that the winding is machined flat on the conducting face and then hard gold plated. Not something to be damaged if it can be avoided.

[dseymo1]

Originally Posted by Lucien Nunes

Did you investigate the anistropy of resistance of the carbon at all? I have always understood that variac brushes were made of material with an especially high resistance along the axis tangential to the core, to minimise circulating current in the shorted turns. How true or important this is I have never got around to investigating by experiment.

I suspect that I'm being denser than usual, but I can't see how it would help. For simplicity, consider a brush as being composed of two insulated 'slices', each contacting one of two adjacent turns. Certainly, the turns would not be directly shorted, but surely the short-circuit would simply 'move up' to the common connection (at the brush-holder or wherever).
What am I missing here?

[mark_in_manc]

That seems to make sense to me. Which is good, in the absence of a counter argument, since it makes my bodge feel less bodgy!

[Lucien Nunes]

the turns would not be directly shorted, but surely the short-circuit would simply 'move up' to the common connection

Relative to typical load resistances, the resistance along the length of the brush is trivial, but against the EMF of only the bridged turns it limits the current significantly. The laminated or anisotropic brush construction avoids a very short effective path through the brush little more than the turn pitch, that would otherwise occur. I.e. it makes the short longer!

[backflipper]

Does anyone have any advice on reterminating the winding if it has gone open to the terminal? I have one that is open but the windings cannot be soldered on to the terminal.

I think they welded them in..

[ex 2 Base]

Could you Silver solder an new connection.

[mark_in_manc]

Can you adapt the terminal so that the connection to the end of the winding is mechanical - clamp or screw it on, in other words.