Re: center tapped transformers

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: Jim Lux <jimlux@xxxxxxxxxxxxx>

At 12:51 PM 11/2/2005, Tesla list wrote:
> Original poster: "Qndre Qndre" <qndre_encrypt@xxxxxxxxxxx>
>
> There is a question I asked myself over and over again since the 
> first time I've seen a schematic for a Tesla coil on the net: Why do 
> coilers ground the center tap of the secondary of the high voltage 
> transformer (NST or OBIT) instead of just doing the opposite 
> (insulating the case from ground very carefully or cut the 
> connection of the secondary to the case)?

You're not the first to wonder..
It's really the centertap of the winding connected to the core, which 
is then connected to the case, and it's done so that you can use less 
insulation on the HV windings (since the inside of the winding is 
close to core potential, and each successive layer is farther out, 
and at higher voltage.

And, because they KNOW the core is going to be at low (earth) 
potential, they don't put a heck of a lot of insulation on the 120V 
winding (just enough to pass the hi-pot test, if any, at manufacture time).

There's also a more subtle reason, and it has to do with the parastic 
capacitive coupling from the windings to the core.  Keeping the 
windings symmetrical with respect to the core (i.e. connecting the 
center tap) helps reduce the leakage current to "outside the box".

If you want an isolated secondary, you need to use a 120V:120V 
isolation transformer with 15kV insulation, which would cost more 
than just getting a transformer designed for 120V:15kV with an 
isolated secondary.

As always, the surplus market can change the economics of the 
situation.  I found a couple big (as in many kVA and 100lbs of iron) 
isolation transformers for 30kV+, designed to float the 
cathode/heater power supply in a RF power amplifier.  These would 
make it a trivial matter to stack multiple NSTs in series, if I were 
so inclined.


> As far as I know this would be safer since touching one of the 
> output wires (or being accidently connected to one by a streamer) 
> wouldn't electrocute you in this case since the output terminals 
> weren't hot in respect to ground.

Except that even though they're not "wired" to the case, they are 
still capacitively coupled to the case, and you could still get a shock.

Jim