PWM DC motor controller for EV

I am toying around with the idea of making a high powered DC motor controller for an electric car. I want it to be on par with the available controllers (curtis, zilla, et.al I have read about) as far as power handling. They have models that will switch 300+V at 1200+A. I am wondering what kind of technology they use to switch those high currents. My first thought was that they probably use IGBTs, so I went to newark and the highest powered IGBT I can find is rated @ 890A which in itself might be sufficient for my needs, but assuming I want to make a drag strip car, I would want more than that. Can I parallel IGBTs to double the current capacity?
I looked at MOSFETS also and they have even lower AMP capacity. Can those be paralleled? I'm thinking that if either IGBTs or MOSFETS were used, there would have to be some form of load sharing implemented or else one IGBT/MOSFET would switch the bulk of the load until it blew, and then the next one would blow, and so on in a domino effect. How would one set up load sharing like that? OR am I barking up the wrong tree with paralelled components? is there a higher powered MOSFET or IGBT out there?
Thanks

I've been toying with a similar idea except mine is a small electric car. I think the choice depends on the motors and battery voltages you select. If you are going to stay below 100 volts (say 72) then MOSFETs are probably easiest because they are easy to parallel because their resistance goes up with temperature so you can just put them in parallel with there own gate resistor. If you are going to use 300 or 400 volts which is what it sounds like you are thinking then IGBTs are the better choice, but they are hard to parallel unless you are a manufacture and can match them. If not they need a small resistance in the emitter that will drop .2 volts or so at rated current. These turn out to be large resistors. The modules are made with matched dies, but you pay dearly for it.
Do you have a choice for motors and batteries yet?

No I haven't chosen batteries, motor, donor car, anything. I'm still only learning about it. I want to approach the idea with research from all angles before any purchasing action. Whatever the end game plan, I want to cut costs (but not corners) anywhere possible and it seems that designing my own controller would be a good (if not the only) place to start cutting costs, so I'm looking into that first. Can't very well design my own motor and batteries. My thinking was that if I design a controller that is on par with the big dogs, then any motor/battery combo I choose later on will work.

The modules are made with matched dies, but you pay dearly for it.

Click to expand...

so are you saying that modules (also referred to as arrays?, like these) then they are matched by the manufacturer and shouldn't need the large resistors you mentioned?

I think if you were going to parallel modules you would have the same problem. If your going for a Zilla type spec. I think I would just buy the Zilla. Don't think you can build a one of as cheap or as reliable. Think about how you would circulate the water and all that stuff. The cost of one ahh s..t (If you survive the explosion) would make it a steal.

ronv I have been reading through threads on the DIY EV forum of guys who made/tried to make/are making homebrew controllers and I gained some perspective. I totally agree with you at this point. I think trying to reinvent the wheel here would cost me more in the long run because I'd probably fail 10X before I got anything to work. And these are not cheap failures.