Paul & Sabrina's cheap DIY 144v motor controller

[MPaulHolmes]

The overtemperature starts at 75 degC, and limits maximum motor amps in 10 steps all the way down to zero amps at 85 degC. You probably need to mount the controller on a heatsink if it's getting that hot.

[SEBART PL]

Originally Posted by MPaulHolmes

The overtemperature starts at 75 degC, and limits maximum motor amps in 10 steps all the way down to zero amps at 85 degC. You probably need to mount the controller on a heatsink if it's getting that hot.

The heatsink it is not a point, I drove for 75 miles with no problem, I did overload and system have not a chance to minimalize current, I have put wrong gear to see what happend and I know that I can't give car to my wife or other man. If only PWM could stop as quick as possible mosfets could get more chance to survive.

Other people experiance - Duncan, Adam who blown up mosfets proved that system is not quick enough when temperature is too high and they haven't got another chance to put bigger heatsink- in my opinion.

Paul, is it problem to put new temperature command or possible to overtemperature start quicker 65 C and limit motor amps i 2 steps half and 0at 75? I wish to have controll this parametrs.

[adamj12b]

Originally Posted by SEBART PL

Other people experiance - Duncan, Adam who blown up mosfets proved that system is not quick enough when temperature is too high and they haven't got another chance to put bigger heatsink- in my opinion.

Just so this is clear,

My controller did not blow up because of temperature. It blew up because of 2 reasons. The main reason is the case came lose and shorted the controller out stressing the fets and taking out the weakest ones. After that, I decided to not save it and finish off the rest of the fets.

It was actually pretty low temperature when this occurred.

-Adam

[MPaulHolmes]

I see what you are saying. It would be easy for it to come on at 65c instead, but maybe you would need to mount the temperature sensor closer to a mosfet to get a quicker response? I remember Ian down in Australia was using a Curtis 500amp controller with his direct drive conversion and it blew up after a short time because it was the equivalent of driving around in a high gear I think with a large motor. I'm not sure how fast of a response you can get without putting the sensor actually embedded in the back metal case of a mosfet which Otmar (who makes the zilla) has actually done.

[SEBART PL]

Ok, thank you all for your anserws, I think cut off contactors should be a standard couse in some cases like overload, shortcircuit, fets sticiking, diodes shorts --- >higway controller should react ultra fast, ideal software could check throtle position and current as well for 0 throtle PWM=0 if not true contactor off...

...anyway I am happy what I have and I have decided to put few thermostats 60 degree C opened on mosfets conected in series on contactor line and driving with laptop to see what temeprature I can have on heatsink maybe it is easiest way to read current, temperature etc.?

I hope so I will not kill any one on road if some above cases will occur to me, lucky I have clutch.

[duncan]

Originally Posted by SEBART PL

Other people experiance - Duncan, Adam who blown up mosfets proved that system is not quick enough when temperature is too high and they haven't got another chance to put bigger heatsink- in my opinion.

My controller was not hot - it blew up on initialization - nothing to do with overheat

The initial blow-up was probably due to dirt or a poor solder joint,
and it happened the first time I switched it on.
This blew one MOSFET, I actually drove the car for about 40Km with the blown MOSFET

The second blow up was due to my not replacing damaged parts
Also I was draining the caps by operating the throttle with the main contactor open every time I shut down
I think this contributed to a failure Paul had in one of his prototypes

[Roche]

Would I be right in saying that this is less about sudden, large, temperature rise and more about sudden (and possibly sustained), large current drain?

[SEBART PL]

Originally Posted by Roche

Would I be right in saying that this is less about sudden, large, temperature rise and more about sudden (and possibly sustained), large current drain?

I think it is important on 50/50 current and temperature three actions:

1sudden overload=current rise=temperature rise (series motor take huge current even with slow motion temperature on mosfets rise very quickly especially if you have only ten mosfets)

2,then/or
sudden blown up mosfets cause shortcircuit/sticking drain and source=current rise out of control=motor run out of control (only main contactor can stop motor)

or
3environment temperature rise = boiling mosfets (very sunny day, traffic jam often starts and stops)
then 2

[Roche]

So in your case 1, the existing current/temperature monitoring should take care of any destructive conditions.

Case 2, you might not see maximum current, but equally you possibly won't see a huge temperature rise, just an increase in speed without motor control - I think that's taken care of in software somewhere.

Case 3 - slow temperature rise, the existing temperature sensor should pick it up...

In short, I think the bases are pretty well covered.

[jyanof]

Originally Posted by SEBART PL

I think it is important on 50/50 current and temperature three actions:

1sudden overload=current rise=temperature rise (series motor take huge current even with slow motion temperature on mosfets rise very quickly especially if you have only ten mosfets)

2,then/or
sudden blown up mosfets cause shortcircuit/sticking drain and source=current rise out of control=motor run out of control (only main contactor can stop motor)

or
3environment temperature rise = boiling mosfets (very sunny day, traffic jam often starts and stops)
then 2

Just to review some history of this design and its many applications, there are many instances of the controller's thermal cutback algorithms working correctly and saving the controller from failure. I won't search through the past posts, but there are several cases that come to mind where the user didn't initially realized the importance of a heatsink. The controller would overheat and power would be reduced as expected. I recall another case where in addition to limited heatsinking, the motor was undersized forcing the controller to output max current (500A) for long periods of time.

In any event, the original thermal design of this controller should be adequate for normal use. I myself use the controller in my EV during the hot Arizona summers where ambient temperature can be 50C (especially in the sun). Actually, just this year, I got the a/c working and currently all the hot air from the condensor gets dumped into the engine bay under the hood.

A thermal sensor under the hood indicates that the ambient air temp under the hood can get up to 60C with the a/c on and the controller still works fine! (knock on wood).

I see the pictures previously posted of your controller design in this post. The power section does not appear to be similar to the original Revolt design.

The Open Revolt software and hardware are designed to work together. While changing the software to work better with the current hardware you have may work, you might consider changing the hardware to match the original design intent. There are many more things to consider in the power board layout other than thermal performance - changing to a proven design may end up being more effective than trying to develop software mods which may continue to result in blown up hardware.

Good luck in your efforts!