This is the schematic for the Terry style filter for NSTs. I'm not buying 16 metal-oxide varistors just to make a filter. I'm not sure why the person drew the varistors as current sources, but whatever.
This is the physical setup of the prototype of my filter. The concept, put simply, is that high inductances block high frequencies and pass low frequencies linearly, and small capacitances pass high frequency and block low frequency and DC. A simple filter to pass low frequency 60Hz and block high frequency RF would have high inductance in series and low capacitance to ground in parallel. I might put a safety gap in also, but safety gaps are stupid so probably not anytime soon.
This is my a shot of the filtered 500kHz versus the unfiltered 500kHz. The unfiltered is the odd-looking
sine wave (yes, the university's function generator is a junk heap, my good
one was at home) and the filtered signal is the nearly-zero ripple. I think it
works, and all the components are good for the voltages. I'm looking forward
to the full-power test tomorrow.
|
This is the schematic for my filter. I am not sure what the inductances total to, I am going to solder and measure tomorrow.
I mounted all the inductors on a piece of plywood that I had cut to an appropriate size, which I had intended on mounting on the front of the Tesla cart vertically, but I am seriously reconsidering now, given the approximate 40 pound weight to it. The red doorknob capacitors are above on the lab bench, out of the photo.
This is version 2.0 of the filter, since at higher powers the filter tended to burst into flames. These spark coils (hopefully) at worst will only have the mineral oil boil over and make a little bit of a mess. That's the plan, anyhow.
|