Home built HV transformers (fwd)

[<sroys@xxxxxxxxxxxxxxxxx>]

---------- Forwarded message ----------
Date: Wed, 10 Nov 2004 22:35:48 -0600
From: Steve Ward <steve.ward@xxxxxxxxx>
To: High Voltage list <hvlist@xxxxxxxxxx>
Subject: Home built HV transformers

Ive been working on a high power C-W voltage multiplier recently.  The
plan is to use 15kvRMS input at something over 10khz (preferably
50-150khz) and over 1000w input power.  My efficiency goal for the
actual multiplier is 60%.

A side question before i get down to business.  For my HV diodes i am
using strings of SF 1600 diodes, they are 1600V, 1A, 75ns Trr,
Avalanche rated diodes.  Im using strings of 40 diodes, so that is a
64kv rating, there will be of course about 40kv peak across them.  Ive
heard of many people going for 2X the peak voltage across the diode,
but they did not mention using avalanche rated diodes.  My question
is, should my diode strings be sufficient with their 1.6X rating?

Ok, transformers...  I had 2 different ideas as to how to generate the
15kv.  First was a simple ferrite based transformer (i have a few
1.5kva C cores).  I calculated that i would need 30 primary turns at
170VDC input... that gave me something like 3500 turns needed on the
secondary.  I do not plan to put any gaps in the core like flybacks
use (for energy storage in the inductance of the coils).  So im
curious, will i really need 3500 turns if i perhaps operate the
transformer at the secondaries resonance (like a TC... well more like
a SSTC) ?  Also, does anyone have any suggestions for winding such a
transformer?  What material to use between layers?  How many turns are
acceptable per layer before you start to have too much self C?

Second idea was to build a small SSTC, but to load down its output
with a string of doorknob capacitors (something around 100pf on the
output).  This would lower the impedance of the secondary, as well as
lower the voltage to around the 15kv.  I tried some Pspice
simulations, but i was getting an absurd 2A RMS in the secondary coil
at 15kvAC!!  I was then also considering perhaps not using the caps to
lower the Fr, but rather use more inductance and wind a coil as a
stack of "pancake" coils with plastic dividers every 1/2" or so.  Any
thoughts on this?

So i guess im just looking for any suggestions as to how to generate
15kvAC at a high frequency (50-150khz) and at over 1kw (maybe up to
1500W).

Thanks,

Steve Ward