[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Sold state IGBT disruptive coil spark gap idea

Original poster: Vardan <vardan01@xxxxxxxxxxxxxxxxxxxxxxx>

Hi Again,

First, all the current readings in today's scope pictures are off by a factor of ten. Everywhere where the scope says 200A/div is really 20A/div. I set the calibration number wrong on the current probe. I put a note in the file too now.

I took some more readings and ran some numbers and all the times, energies, and power levels work out so "everything is under control" :o))

I thought this would happen... I almost put in the diode originally but I was trying to get away with it :o) The SIDACs are staying on very long (many microseconds) even with a current reversal. They stay shorted just like a "real" spark gap... That shorts out the capacitor and takes all the power out of the gate drive circuit. The circuit keeps turning off every cycle and is burning up all the energy thus the poor efficiency...

So, it should just need a say 1N5918 diode between the transorb and the capacitor to trap the voltage in the gate drive circuit regardless of what the SIDACs are doing. Just a 12 cent part so no big deal at all...

I will make the change and retest tomorrow.




I got the parts today and built up the circuit. I tested it out and the charging circuit for the gate drive is not working right. Does not look like anything serious, just have to figure it out. I put my notes here:


Sorry, it turned out at 1.8MB with all the pictures...



Hi All,

I was up late last night working on an idea for an IGBT replacement for a conventional coil spark gap. I woke right up at 2PM this afternoon (semi-retired people can do that :O))) and started getting some parts and wrote a paper up on it here:


It is not super cheap, but not insanely expensive either. Might be great for lower voltage MOT systems with E-bay big IGBTs...

There has been ZERO real testing so don't get over excited ;-))

It basically uses IGBTs fired by strings of SIDACs. Each section should be fully independent and they can be put in series as needed for any voltage. One could "tap" the string for various power levels. Power losses seem very pleasant.

It will take some real world testing to see if and how well it works, but the things seems sound. It is an easy and simple circuit that hopefully not much can go wrong with :o))

Parts on order...