Paul & Sabrina's cheap DIY 144v motor controller

[apowers]

OK so when they are all on they are all on. Do you have 6 , 144 volt or whatever volt , battery packs?

[reissmachinist]

I had been running on 9 12 volt batteries for a week or two and getting to
like this EV thing. I took my dad out for lunch in it twice, He was impressed
with the performance.
Friday, I loaded up a bunch of junk batteries and traded them for 6 good used
batteries with a 30 day guarantee and $54. So now I had 15 batteries and 180,195
volts to the armature. It did have the higher rated mosfets, diodes and capacitors.
I eased out of the driveway, down the alley and into the street. It seemed to have
a little more pep. I gave it a little more throttle and started to accelerate.
It did for a short distance and then seemed to accelerate abruptly, more than I had
commanded and there was a sound like popcorn popping fast. I pushed in the
clutch, hit the main disconnect breaker and coasted to a stop.
Because in this short ride I hadn't stepped on the throttle hard, I think the failure
was probably voltage related instead of too much amperage.
The damage is the worst at one end of the board, burning it away between the B+and B- bus bars.


i dot imgur dot com slash uX6BP dot jpg

i dot imgur dot comslash MxrUK dot jpg

i dot imgur dot com slash MxrUK dot jpg

i dot imgur dot com slash MxrUK dot jpg

Sorry about that, I don't have enough posts to post links.

I took it all apart and found the aluminium bar and insulator tape under the mosfets OK.
I cleaned the control board and powered it up, the LED's both lit up solid.
The mosfets were $7 in January and now they are $12. Digi-Key has 124 of them.
Next build maybe I will stay with 144 volts.
Hot and Humid here!

[Darxus]

Originally Posted by apowers

OK so when they are all on they are all on. Do you have 6 , 144 volt or whatever volt , battery packs?

I don't have anything. I've been planning to use Headway 3.2v 10Ah cells in 72 volt 22 cell packs with a 72v Mars ME0709 motor, on a motorcycle. 6 of those packs, totalling 132 cells, is 4.2 kWh.

And fewer than the 144 cells in this model:

[Greg Fordyce]

Originally Posted by Darxus

I don't have anything. I've been planning to use Headway 3.2v 10Ah cells in 72 volt 22 cell packs with a 72v Mars ME0709 motor, on a motorcycle. 6 of those packs, totalling 132 cells, is 4.2 kWh.

You will need some sort of BMS for the headway cells. In that configuration the BMS is 6 times more expensive. If you parallel the 6 cells first so you have a single 60ah 72 volt pack, you then only need a BMS with 22 nodes instead of 132.

Greg

[voltmatic automaton]

i need some advice on the control board. I was wiring everything up for my install, had the power section in and wired up and went to wire in the control board.
something went ZAP and now the control board won't light up on a 12 volt supply...nothing...at all.
so my question is which components should i start looking at for a failure if 120v somehow zapped the board?

just to clear the confusion, im using the power stage of a logisystems controller. just removed the lem and installed the revolt board. since there wasn't the space i decided to go with the zilla idea of making the power stage a seperate module to the control board stage...kinda like the hairball idea.

but now the control stage is dead...please help!

[sawickm]

Originally Posted by Georgia Tech

hey guys whats the story on the Synchronous controller? The one that can handel 1000 amps..

I found this website this morning.

http://b2600ev.org/prototype-control...ng-part-3.html

Some guy's who modified the Revolt for Synchronous conversion.

- Mark

[apowers]

Originally Posted by voltmatic automaton

i need some advice on the control board. I was wiring everything up for my install, had the power section in and wired up and went to wire in the control board.
something went ZAP and now the control board won't light up on a 12 volt supply...nothing...at all.
so my question is which components should i start looking at for a failure if 120v somehow zapped the board?

just to clear the confusion, im using the power stage of a logisystems controller. just removed the lem and installed the revolt board. since there wasn't the space i decided to go with the zilla idea of making the power stage a seperate module to the control board stage...kinda like the hairball idea.

but now the control stage is dead...please help!

I would start by checking to see if the 12 volt supply leads are reversed. I'm not sure it would go zap even if they were. The positive should go to R1 first.

[jyanof]

Originally Posted by Greg Fordyce

The controller and motor together form a Buck converter.

Thanks for explaining this more clearly!

We could use a separate motor for each phase

This would work!

[voltmatic automaton]

the 12volt leads are not providing power in either orientation. I have hooked them up and can get meter readings thru the big resistor R1 and the first diode D1.
after that i'm not sure where to check the voltage.
but i know i can get power into the board at least passed those components.

any others i should start lookin at?
i am suspicious of the cimcon, there are no voltage readings on any part of the board passed that component.

[jyanof]

Originally Posted by reissmachinist

So now I had 15 batteries and 180,195
volts to the armature. It did have the higher rated mosfets, diodes and capacitors.

Dude, that's too bad! gonna think out loud now...

I'm wondering how overvoltage would play into this. What were the components rated for? There will be a voltage spike when the mosfets turn off - measurements with a scope show something like 15-20V or so at 200A or so using a gate resistor of 20 ohms. I forget, but I posted a while ago on that in this post. It gets worse as the current goes up and as the gate resistor value goes down.

Another thing to consider is that mosfet switching losses increase proportionally to votage squared (according to heat generation formulas for mosfets). So, if you increase the input voltage from 108 to 180, that's a factor of 180/108 = 1.66x. But, the switching losses will increase 1.66^2 = 2.77x! So, your mosfets may have been generating almost 3x as much heat as before! They may have heated up faster than the heat spreader could and before overtemp could kick in.