Uploaded May 2024 | Updated September 2026, 2 weeks ago
I bought 100 10nf capacitors I plan to wire up in something like 5S 10P for 20nF or add another S for around 16nF.
I added a generic phase lead on the schmitt trigger to see how I could get it to run without PLL feedback but decided I was better off with the PLL given how much the arc loading probably changes the phase shift. Tuning the phase lead at max voltage is probably the way to do it but I can't hit that with the primary streaking anyway. Going to eventually convert this over for 120VAC line input. I'll just use a couple transformers to isolate my control voltages and double the 120VAC for about 340VDC into the bus caps. I'll probably have to use a basic precharge circuit for 20,000uF. You will trip a breaker trying to slam 170VDC pk onto that so I have a 3p contactor on the input that closes the line input to the bridge rectifier when control power is on. For precharge I just need to bypass that contactor with some resistance and only have it close (bypassing the resistor) when the bus reaches a certain voltage threshold. To run it in any kind of mode all I'd have to do is bypass the buck circuit after the caps for smoothed or switch out the buck entirely for unsmoothed.
The smaller 4s battery powered setup looks like it could hit 3ft arcs with a little higher ramp voltage. I tried winding the ZVS transformer so that it could charge about 6000uF to around 200VDC tops pretty quickly from a 16V supply. If I ever find a perfect enclosure for it I'll rework it using smaller caps and it will be pretty awesome. I think a lot of these switches are able to handle 1-2ms or so of some incredible peak power and since the arcs are fairly slow in the BPS and still satisfying, you can probably get away with a whole lot with some small ~60-100A switches. The next one I want to build will probably be real small and use 200V mosfets for about 150VDC bus voltage and maybe 100V ramps tops. That one can also be 4s powered and I'm thinking somewhere around 1-2mhz.
I bought 100 10nf capacitors I plan to wire up in something like 5S 10P for 20nF or add another S for around 16nF.
I added a generic phase lead on the schmitt trigger to see how I could get it to run without PLL feedback but decided I was better off with the PLL given how much the arc loading probably changes the phase shift. Tuning the phase lead at max voltage is probably the way to do it but I can't hit that with the primary streaking anyway. Going to eventually convert this over for 120VAC line input. I'll just use a couple transformers to isolate my control voltages and double the 120VAC for about 340VDC into the bus caps. I'll probably have to use a basic precharge circuit for 20,000uF. You will trip a breaker trying to slam 170VDC pk onto that so I have a 3p contactor on the input that closes the line input to the bridge rectifier when control power is on. For precharge I just need to bypass that contactor with some resistance and only have it close (bypassing the resistor) when the bus reaches a certain voltage threshold. To run it in any kind of mode all I'd have to do is bypass the buck circuit after the caps for smoothed or switch out the buck entirely for unsmoothed.
The smaller 4s battery powered setup looks like it could hit 3ft arcs with a little higher ramp voltage. I tried winding the ZVS transformer so that it could charge about 6000uF to around 200VDC tops pretty quickly from a 16V supply. If I ever find a perfect enclosure for it I'll rework it using smaller caps and it will be pretty awesome. I think a lot of these switches are able to handle 1-2ms or so of some incredible peak power and since the arcs are fairly slow in the BPS and still satisfying, you can probably get away with a whole lot with some small ~60-100A switches. The next one I want to build will probably be real small and use 200V mosfets for about 150VDC bus voltage and maybe 100V ramps tops. That one can also be 4s powered and I'm thinking somewhere around 1-2mhz.
