Paul & Sabrina's cheap DIY 144v motor controller

[MPaulHolmes]

That'll do, Oprah. That'll do. (Babe, slightly modified)

Hi Ben! I got the driver board done for the possible motor link that you sent to me. I also got one done that would be for IGBTs and super high voltage. I've found some big gosh dang igbts in a TO-264 package. Rated for like 200-250 amps each and 600 volts, and they only cost around $12 or $14 each. The same sort of thing that the 1000amp controller uses can be done with them. Just solder their backs to copper plates, and clamp the plates against the heatsink. I think a 300-400amp per phase AC controller would be pretty dang cheap. It would require a chassis mount film capacitor or 2 though rather than the crap electrolytics. And the 6 isolated drivers are only a couple dollars each since I have a super fancy way to do it!

[rayjay]

Also add a non-conductive adhesive to the mating surface of the lexan to the pcb for strength

[isaac_alaska]

might use an adhesive from the lexan to the other surface, as long as you're sure it never has to come off. but screws and stand-offs should be more than plenty strong to hold the pcb to the lexan. unless you're mounting low frequency (in other words big and heavy) transformers to the pcb o.O

[Anaerin]

Alternatively, if you want to link your ground planes, you could always just use aluminium L-brackets to support the board at 90 degrees.

[isaac_alaska]

paul, i'm using the little melexis board you sent me to measure current for the peter perkins bms i'm putting together. the bus bar i tested it on is 1 inch wide instead of 3/4 inch, and gives an output of 30 mV/20 amps, for a full range of +/- 1500 amps. i think i may mill the busbar a little narrower (down to 3/4 inch) where the current sensor mounts to get a little higher resolution, but for now i've just changed the software and it works nicely

(do you have any more of the little pcb's made up? i might need to buy a couple more)

[MPaulHolmes]

Heck yes I do... I gots me friggen fiddy of dem I think for just such emergencies. $3 each would friggen cover it fo sheezy feta cheesy alfredo.

Hey! I just did a little video!

http://www.youtube.com/watch?v=St7HZ...ature=youtu.be

[isaac_alaska]

sweet! i should order a couple of them at least. maybe 3 and we'll call it ten dollars. have you got a paypal adress listed somewhere that i can send moneys to? (:

also if you have them already built up with the melexis and the caps i'll pay for them that way so i don't have to build them (:

loved the video. and the song! it's good

[jackbauer]

Happy new year everyone

[MPaulHolmes]

Uh Oh! Irish people and New Years Eve! You stay out of that EV, Mister!!! haha

[MPaulHolmes]

I bought 3 used batteries today, so here's a test at 60 volts. I got it up to around 115 motor amps. There's no noticeable heating of the base plate or bus bars. Gate to source turn on looks pretty nasty, like the driver has fits barreling through the miller knee, but it's turning on fast, so maybe it doesn't matter. Maybe I need a new driver? The total turn on capacitance per driver is like 85000pF, and it's rated for 60000pF, but it works fine with the 500amp controller, which is driving over 100000pF. Maybe not all picofarads are created equal:

http://www.youtube.com/watch?v=mdwPm...ature=youtu.be

EDIT: This paper suggests that the gate oscillation is the result of the switching from drain to source. I could reduce the oscillation if I slow down the switching, but I've already seen similar oscillation (but more minor I think) with 15 Ohm gate resistors.
http://ieeexplore.ieee.org/xpl/freea...number=1713572

The way I see it, what do I care if there's gate ringing as long as the mosfet is getting turned on and off fast, and the gate turns on fast too, and the gate to source voltage is clamped by a transient voltage suppressor. The TVS diodes respond in like 5 picoseconds.