Re: I'm going to need help with a 34kV power supply

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

Original poster: "Jim Lux by way of Terry Fritz <twftesla-at-qwest-dot-net>" <jimlux-at-earthlink-dot-net>




>    That said, I'm now itching to use that pretty gold box to power a larger
> tesla coil.  I realize that the extremely high voltage from this power
> supply is going to cause some major design problems with the primary
> circuit.  I have enough pulse caps to hold off 140kV and give me 7.5nf.  The
> caps will obviously have to be immersed in oil, the connections and wiring
> between all of the components will all have to be heavily insulated and
> probably coated with an insulator to keep corona losses down, but I'm not
> sure about the construction of the primary coil.  I'll most likely use 3/8"
> copper tubing, but how far apart should the turns be spaced?.

An insulator like tubing won't help much with corona losses, because if
there is ANY air gap between copper tubing and insulator, that's where the
corona will start. The same applies for small surface imperfections (saw or
file dings, dust, etc.).  However, if the radius of your conductor is > 1
cm for every 30 kV, you won't get much corona, insulated or not.

Here are some things to contemplate: 
1) The voltage between turns is pretty low (assuming you have, say, 10
turns, it's only a few kV) so corona inception/breakdown probably isn't a
problem.  At the ends of the coil, though....

2) You could make the primary with somewhat lower inductance than required,
pot the entire thing in silicone, and then use a second tuning inductor in
series. The tuning inductor could be bare, and on the "cold" side of the
primary winding, so the voltage across the tuning inductor never gets above
a few kV, relative to ground.

3) You won't have to immerse your caps in oil at only 30-40 kV.  Just pay
attention to radius of curvature and your connections.  A bit of "corona
dope" will also help (silicone globs work well on things like screws, etc.)

4) A bit of corona loss isn't a real problem (heck, the secondary is going
to have corona all over it...), but unintended flashovers in the primary
would be.


>    The spark gap is another problem.  I'd like to use an async rotary with a
> good bit of radial velocity (3450rpm, instead of 1725rpm), but what would be
> good to use for the stationary electrodes so that they won't glow with
> corona and ionize the air around the gap excessively?  Would a nice fan take
> care of that, or should something like a small doorknob be used?

A fan won't change whether corona occurs, but would make it less noticeable
(maybe).  Whether you get corona is a function of the electric field
strength and the dielectric (air). This is entirely a function of radius of
curvature, surroundings, and air pressure.
>    If I go with a static gap, I worry about corona losses at the edges of
> the copper pipes, if copper pipes are the best choice for this gap.

You could use pipe caps on the ends to get a nice rounded surface.  Or,
turn down some bar stock to give it hemispherical ends.
>    How do you feel about a triggered gap?  The trigger electrode would have
> to be powered with a significant amount of voltage, given that that tank
> circuit will be powered with a 34kV power supply.

I'd certainly consider a triggered gap. The main electrodes can be spheres
of suitable diameter (or hemispheres brazed/soldered onto a larger flat
plate with rounded edges). The trigger doesn't need signficant voltage,
just because you're switching 30 kV.  Say you have your gap set so that it
self breaks down at 40 kV.  The trigger electrode (assuming a midplane, or
centered three ball)  only nees to break down half the gap (swinging
cascade form), or 20 kV.  Once it breaks down, that half effectively is a
short circuit with a few hundred volts across is, so the entire 34 kV is
across the other half gap, which will readily break down.  
If you are using a trigatron type gap, the small spark from the trigger
only needs to bridge the gap between the trigger pin and the main
electrode.  The ions and UV from the trigger spark move into the main gap
to reduce the breakdown.
If it is a field distortion gap, it's sort of like the swinging cascade,
except that the trigger just "pushes" the E field around to increase the
field in part of the gap beyond breakdown. A streamer forms, propagating
into the breakdown area, effectively shortening the gap, and resulting in
total breakdown.

In any case, a run of the mill HEI coil generates 40 kV.


>    Any advice on how to properly design and build such a high-voltage beast
> would be greatly appreciated.  I won't mind comments about safety, either,
> as I don't want to miss any issues on safety just because I hadn't
> remembered them at the time.

By and large, a clearance of 1 inch per 10 kV is a good guideline.
> 
> Thanks,
> Joe Duva