Paul & Sabrina's cheap DIY 144v motor controller

[adamj12b]

Hey everyone,

here is a video showing the new firmware that Fran wrote and the use of the serial configuration. This is very well written code.

Enjoy

http://www.youtube.com/watch?v=23txj4t2eR8

-Adam

[jackbauer]

I like the data streaming feature very much. Wouldn't take much to put together a vb program that would run on windows and allow changes via radio buttons etc.

I'm aiming the igbt setup to be flexible in terms of power capacity. 1000amps at 300v is possible at least in theory. Busy with work but will start doing some research on the gate drive. My 1600v 800amp igbts should be in from Germany next week. I'll be using a 12mm thick aluminium plate as the heatsink and initially one device as the switch as one as the freewheel diode. Later i'll order two more devices and use 2 as the switch and two as the freewheel. With water cooling that should mean anything up to 1000amps. I'll have to look at an upgrade on the lem to handle this range. I think some form of snubbing will be necessary at such currents.

[jyanof]

Originally Posted by jackbauer

The following takes place between 4pm and

In regards to the gate ringing, I wonder what the drain-source voltage waveform looks (or whatever the igbt equivalent is- collector emitter?)

I notice that you have some significant legth of wire between your caps and igbts.. I had a similar setup when I was playing around with a charger and found excessive ringing on the Vds turnoff. When I asked the evtech list about it, they pointed out that the current in that length of wire goes from zero to max with every swithing cycle. Any inductance in that wire will cause voltage ringing, even if it's just the parasitic inductance of the wire itself. They suggested to minimize the lenth connectio between input cap, switcher, and rectifying diode to reduce ringing, and it solved the problem.

So, it might be that the gates are picking up some noise, or maybe the gate wire connnections are long and their parasitic inductance is contributing (they also recommended to keep those short)

Anyway, some things to try if you haven't found a solution..

[jackbauer]

Good points. The layout is a disaster i just threw it together to see if it would work. Soon as i get the new devices i'll get the whole thing better organised.

[jackbauer]

Paul I tested the lem as per the circuit in the datasheet. Same result. Vout = 0.73v. Any ideas?

[MPaulHolmes]

crap! crap crap! Unless the wiring is wrong (which I doubt) the dang lem is broken.

I have another one I can mail to you. makes me mad! stupid digikey! I will punch them right in the head!

[jackbauer]

i'll send the old one back to yourself so at least you can get a replacement. no panic its not a show stopper for now as i'm still on low power testing. Certainly explains the problems i've been having

[MPaulHolmes]

don't worry about mailing it back. I doubt digikey would take it back anyway. I'll just ship you a new one. it's all good. Make sure you punch that lem right in the head for me. Stab it in the eyeball with a needle.

[drbobwoolery]

Paul, check out AD623 for the shunt amplifier. You should have two shunts in the mail in the next few days. Don't be dismayed that the first is so big; that's for testing. The second fits between two pieces of a busbar with 5/16 holes one inch apart. To heck with these finicky Hall devices!

[jackbauer]

Would we not have a concern of lack of isolation between the power and logic sections of the controller using a current shunt?