DC power

The recent posting from Wes B. (copied below) is very interesting.  Is there
anyone out there running a large Tesla coil (in the 2 kva range) on DC power?

"Subject: Capacitor Charge Rate


I'm glad my note on the subject made some sense. I wrote it in a hurry and
gushed out a lot of different items, and was concerned after sending it that
it might have seemed a bit incoherent. If you were able to get out the
information I was trying to put in, then that's just great.

As far as using rectifiers to always charge the cap in the same direction
goes, I don't think that the cost of the rectifiers should be a problem.
Commercial high voltage rectifiers tend to be made by stacking lower voltage
rectifiers in series, with each rectifier shunted with a very large
resistor. The idea is to make the resistors small enough so that the
inverse voltage is distributed equally among each diode, but large enough
so that the reverse leakage is acceptable. With 1KV 1A rectifiers selling
for ten cents each, it shouldn't have to cost a fortune to roll our own. The
problem as I see it is controlling the kickback from the tank circuit.

While the power transformers seem to be tough enough to handle a lot of
kickback-related punishment, it only takes one nearly-instantaneous glitch
of overvoltage to waste a semiconductor device. This means that kickback
prevention really will need to be designed rather than estimated. I think
that this problem is quite solvable, and it's one of the items I'm working
on, but it's going to take a lot of time at the rate I'm going. Still,
adequate kickback prevention can be based entirely on passive filter design,
which has been around a long time. It should be theoretically possible to
limit the kickback voltage seen by the transformer to whatever level you
want, and the difference between the line frequency and the frequency of the
kickback waveform is so great that it should be practical to be able to
limit the kickback into insignificance. After this is done, rectifiers will
be more practical in the circuit, and a lot more efficiency should be

Wes B."

Thanks, Ed Sonderman