Re: MOT leakage

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Borislav Trifonov" <bdt-at-shaw.ca> 

Thanks for the reply.

The reason I got this MOT was that it was larger than any others I have 
come across, and I'm only drawing about two thirds of its rated power. 
That's why putting two together seems overkill.

What I want to ask is about shielding.  Someone suggested I put it in a 
steel box filled with potting compound.  This should attenuate the leakage 
to the outside and the compound should prevent the box from vibrating too 
much.  Reinforcing the power supply chassis should deal with the remaining 
hum.  However, I'm wondering if heat will dissipate well enough through 
potting compound.  What sort of material is recommended, that is not too 
hard to obtain (I'm also in Canada, which makes getting things like that 
harder)?  Another thing is, I've noticed that most leakage seems to be 
around the exposed parts of the windings rather than the core.  The chassis 
buzzes least when the transformer is set in such a way that the closest 
part to the chassis is an outer side of the core.  So what I'm wondering is 
if it's OK to just shield  the sides of the transformer, so there is better 
cooling.  I have seen many enclosed transformers that only cover the 
windings, though perhaps that is more of a safety feature rather than 
shielding (I'm referring to tube amp transformers which often are mounted 
on top of the chassis).

Another person suggested an aluminum chassis but I need the shielding of a 
steel one.

>It may get slightly quieter under load, as the IR drop in the primary will
>reduce the maximum flux that the core sees. But to get a big improvement you
>might have to build a new transformer with a larger core area or more
>primary turns.
>You might like to try "mating" the E parts of two identical MOTs together
>and connecting the primaries and secondaries in series. This will give the
>same output voltage but will half the peak flux density so it should run
>much quieter and leak less flux.
>Steve C.
>