The SCRs wont cut off until they reach 0 volts, I control the pulse by controling the inductance/resistance of the coil. IGBT and FETs are a great choice, but not as affordable at the same amperage.

I dont power the SCRs directly from the Micro, since they require more current than the micro can deliver. I have a transistor/mosfet combo that actuates the SCR.

But I have read about reverse biasing SCRs to get them to shut off early.

I havnt had any time to work on it recently, but ill get you some better schematics as soon as i can.

Hmm. I was hoping for more detail about how you're driving the SCR gates. It looks like you put them on the high side of the cap but I don't see a high-side driver? Most that I've worked with use 5V logic. The little details are important when trying to figure out what happened. Also please add the part numbers so I can check the datasheet for that part. ;)

Also I've been thinking, those freewheel/clamp diodes are really just too puny for what you're doing. You should rate them with the expectation of seeing the full current spike.. and as you may know diodes don't share current very well, the junction voltage drop actually reduces with temperature. Not sure how much this is a factor when dealing with just a spike.. but either way I would suggest fewer and larger parts there.

Im sorry I messed up the schematic, the SCRs are on the low side.

The SCRs being used were 2n3890 manufactured by Westinghouse, they are actuated by an arduino and pull power via the 5v rail of my power supply(max current at 5v is 60A via a transistor and a mosfet). I will upload some better detailed schematics tonight.

The freewheel diodes are being replaced with a diode module that can withstand 900 amps continuous, and a 20ka spike. pn powerex LS412460.

The SCRs are going to be replaced by 5x in parallel of these: Eupec tt56 SCR modules. Im going to wire them as follows, terminal 1 to ground, terminal 2 from coil 1, terminal 3 from coil 2. To fire the first coil all pin 4s will be brought high. To fire Second coil all of pins 4 and pins 6 will be brought high.

And please excuse my electronics experience I learn as I go, I never pick projects that I dont have to learn something.

Ah cool that should help. :)

I plan on making a gate driver circuit for mine to make sure it switches on and off as fast as possible, not worrying about current limits of the micro's outputs. Plus adding a degree of separation between those power parts and the logic. series R will let the internal clamps work should a spike kick back thru..
Do the SCR's turn off when you shut off the gate? I've used some that wanted to stay stuck on before, wouldn't turn off until zero crossing on the anode+cathode. I have it in my head that IGBT or FET are the path I want because fine tuning the pulse duration will be important. SCR may be ok though, will have to research this.. a lot of high-power options out there.

posted a new project update, if you'd like to look at it.