Paul & Sabrina's cheap DIY 144v motor controller

[jyanof]

Originally Posted by Greg Fordyce

Another thing I noticed is the controller gets warm. up to 60-70 Celcius. I just have the supplied steel base plate and was hoping I could get away with that in Scotland. But ebay to the rescue and have just got a nice aluminium heatsink the right size to replace it, so I will do the spacer mods on the bolts at the same time so I don't have the same problem that Ben and Joe had.

What's the ambient temp in Scotland? I don't think my controller got that hot in the hot phoenix summer. When you get your heatsink, you can compare its performance to my setup where I was seeing about an 18-20C rise above ambient as measured from the controller serial data stream with an average current output of about 200A.

Here's an example run

[Greg Fordyce]

Wow Greg, sounds like you have a SAWEET setup!

Range so far is a confirmed 54 miles, but I should be able to get better than 80 miles. The 50 watt bulb is a bit slow to precharge, I may start using 2 in parallel. I've stuck a couple cpu heatsinks on that were lying around, hasn't helped much, but they are pretty wimpy. My plan is to drill and tap the 150mm x 300mm x 40mm heatsink I got for 6mm bolts. This will mean no nuts on the bottom, just screw straight into the heatsink. I've found some bolt insulators, just have to find long enough bolts.

What's the ambient temp in Scotland?

I'm originally from California and when we are having a nice summer day here I like to joke "This is like a fine winters day in LA!" hahaha. Thanks for the link, mine is getting warmer than that, so I have been taking it easy till I get it sorted.

[voltmatic automaton]

Hey guys, here comes the weekend and I plan to build my controller. So my question relates to the bus bars.

My original control board is made for all three outputs coming out the same end.
Is there a way to make all three bus bars come out the same way?
Space is at a premium under the hood, so if anyone can help that would be great.

I also have a question about the epoxy for the heat spreader, would soldering the thing to the board work?
I do have a heat gun, has anyone had any experience in this?

and finally, about the board for mounting the control board to the power stage. where can i get a board for this?, does the control board have to be mounted to something conductive?
The build instructions were a little vague.
and one more, the steel strip for the underneath of the heat spreader, where did that come from? steel only or copper, aluminum or what?

thanks for your continued assitance

[MPaulHolmes]

For the mounting board (board between control board and power board), I just use a piece of FR4 with no copper on either side. See pages 57 through 63 for a step by step of how to mount the control board to it, adn then how to mount it to the power board.

http://www.paulandsabrinasevstuff.co...il,%202011.pdf

I use 8" x 6" unclad FR4. Ebay sells it, or you can get something like it at radio shack. 8"x6" is a pretty common size. Just put the control board on it and mark the mounting holes, adn then drill them out. Then drill the 10 gate holes and 2 ground wire holes. Precision isn't important at all with it.

Make sure you don't connect the aluminum heat spreader directly to the power board. It needs to be electrically isolated from the power board.

The steel bar is just for re-inforcing the bolts, so you can clamp it down nice and tight. You can use copper or aluminum or anything you want.

Good luck!

[jackbauer]

Its alive!! Muhahahahahaha!
I must say the vla module gives a very clean gate drive. Now all i gotta do is ramp up the voltage and current and hope nothing explodes

http://www.youtube.com/watch?v=p_I_Ny7tiYg

[jackbauer]

Right! Anyone got any smart ideas about how to integrate the fault output signal from the igbt driver board into the rev c board?

Option 1: Annoy Paul to do yet another code mod for me to use pb3 on the icsp pin as a shutdown input. Also possibly usefull for a brake input. ie if it goes low then command pwm off.

Option 2: Connect into the uv fault signal.

Option 3: Ignore it.

[DawidvC]

My personal feeling is that it would be best to use pb3 as a shutdown signal. The fault may not be serious, but if it is (igbt gone faulty) it is better to have a defined shutdown cycle that leave everything as safe as possible. This should include forcing a reset to clear the fault.

The real question is - who gets the job of annoying Paul? I got a kid that is ideal for the job, but we are a fair bit away from him...

Regards
Dawid

[MPaulHolmes]

I'm not sure, but I think the desat fault signal would imply that setting PWM to zero would have no affect. If the igbt has failed on, I think your only course of action would be to open the main contactor in the code. But in that case, you could probably just use the hardware overcurrent notification input to the microcontroller. Note the time, and if the current doesn't drop after "x" units of time, then open main contactor. 1/4 of a second is an eternity for a microcontroller, but a car careening out of control for 1/4 of a second means moving forward like 2 inches. haha.

[jackbauer]

Well we already have some of that implemented in the "bauer-code" already. If the device came out of saturation isnt that almost the opposite of falling short circuit? ie the load is so large (short) that it cant get into a fully conducting state? Not a big deal in any event. I would like to look at a brake input sometime. I'll implement the hardware as i have the controller on the bench. Would be a neat feature and would eliminate the stuck acccelerator / failed potbox issue.

If brake signal high then ignore throttle and command pwm to zero. Check current. If more than 10amps open contactor.

[MPaulHolmes]

Oh you are right. The point of the desat detection is to save the igbt, not to become aware of the failure. I need to look up desat detection before commenting on it. hahaha. But I suppose it's not that big of a deal to add a low pass filter after the output of the fault signal, and then monitor it. But I suppose the powerex people might be assuming that you don't have hardware overcurrent protection already. Maybe it's functionally the same thing.