I have just been thinking about the design of the brushes on a variable transformer and how they might affect the efficiency of the device.

My understanding is that the width of the brush must be sufficiently great to bridge at least two adjacent turns of the transformer to permit continuous current flow whilst adjusting the output voltage. But bridging the turns is shorting them and that is not good! Presumably to avoid high short circuit currents the brush would need to be resistive and I assume as one turn of the winding is likely to be three or four volts, the brush would need to present at least a couple of ohms between the turns. But if that were the case and even if the brush was made with a material with a good negative resistance coefficient it is likely to run hot as well as presenting a poor output regulation.

I have an eight amp Variac (type 100R) and this doesn’t appear to be the case. So how have the manufacturers solved the problem and achieved such a good result?

I am sure one of you will explain how this is achieved and I will probably kick myself for not seeing the solution sooner.

 

There is no problem to solve.

When the brush bridges two adjacent turns, all it does is remove one turn from the circuit.
Taking out one turn makes very little difference when there is a large number of turns.

 

Well on my Variac, the brush narrows right down, and when in operation barely spans two adjacent windings.'

 

Here's an article that discusses that.
It states the contact brush has a non-linear resistance that goes down with current.

 

There is no problem to solve.

When the brush bridges two adjacent turns, all it does is remove one turn from the circuit.
Taking out one turn makes very little difference when there is a large number of turns.

No sir. One or Two turns is/are shorted and there is a circulating current. May be small, but there is a shorted turn. The copper conductor is designed to handle that current without damage.

 

Here's an article that discusses that.
It states the contact brush has a non-linear resistance that goes down with current.

Thanks very much, Crutschow, for referring me to Rod Elliott’s article on the Variac. It answered all the questions and explains the variable transformer very clearly.

It is interesting the graphite brush is anisotropic having a different resistance in one direction compared to another. I’m not a physicist but somehow find this understandable. However in a latter part of his article he said that the brush resistance when measured with a meter may read more than 20 ohms but is less than one ohm when passing a current. This I find harder to understand and it suggests the material acts more like a (low voltage) suppressor rather than a non-linear resistor.

Thanks, Crutschow for the excellent solution.

 

One serious caution about both Variacs and Powerstats is that it is important to keep those contacts on the coils clean of both dust and oxide. A poor contact can lead to overheating as the rated current is approached.