Paul & Sabrina's cheap DIY 144v motor controller

[MPaulHolmes]

If it gets a reading below about 2.0v (which would correspond to like -200amp or something, it causes a fault and forces the power to be turned off. Otherwise, if it reads a negative current, it just treats it as if the current was zero. 2.5v is the zero current point. When the controller first turns on, it waits about 0.1 or 0.2 seconds, and then makes 16 reads of the current sensor. Then it takes the average of those reads and calls it Vref. Usually Vref is around 2.47 volts up to like 2.53 volts. From that point on, Vref is used as the zero baseline for the current read from the microcontroller. Any reading below Vref is assumed to be zero amps.

So, you lifted the foot off the throttle and didn't cycle power, and it was back to normal all day? What happened with the LEM on the charger?

[jackbauer]

Paul , yeh when i headbuted the wheel i came full off the accelerator pedal and slammed on the brakes and reached for the key thinking i'd lost the controller. car stopped. all currents zero. Tried the pedal gently and car behaved as normal. Been talking with Adam and he suggests the LM393 i'm using is rated to 0C. Might be the culprit.

[dave koller]

LM393. 0˚C to +70˚C. LM293. −25˚C to +85˚C

usually the specs are enough that 0 C should still work.. but a 293 would give it more leeway

.

[jackbauer]

thanks Dave. Its looking like that's the culprit. I use a bmw e46 hall effect throttle so i'd hope thats not the issue.

[DJBecker]

Originally Posted by dave koller

LM393. 0˚C to +70˚C. LM293. −25˚C to +85˚C

usually the specs are enough that 0 C should still work.. but a 293 would give it more leeway

.

I very much doubt that a cold weather malfunction is caused by the opamp. They are pretty stable, and an old design like the '393 far exceeds its original specs when moved to more modern processes.

The usual culprits are oscillators and capacitors. An crystal oscillator that fails to start up will just result in the controller not working, and being a little off-frequency won't even be noticed. Capacitors can cause more subtle problems. but there aren't any with critical values in the main circuit.

My speculation: the DC-DC converter didn't start up until it warmed slightly.

[dave koller]

Is that still the EC4A02H DC-DC converter with operating Ambient Temperature Range -25℃ to +71℃? To show a bad component one could use a can of "Freeze". Jack has an open configuration to get to parts and cool them way down to see. LOL use caution !!!

[DJBecker]

Originally Posted by dave koller

Is that still the EC4A02H DC-DC converter with operating Ambient Temperature Range -25℃ to +71℃? To show a bad component one could use a can of "Freeze". Jack has an open configuration to get to parts and cool them way down to see. LOL use caution !!!

It's going to be difficult to reproduce -- it probably will never happen on the bench.

Once you have a few faults such as this, you redesign your software to track anomalous events and log the conditions.

[jackbauer]

The DC-DC is a Vicor VI-202-CX. Rated to -20C. Gonna fire her up in a few minutes. Nice and cold. Wish me luck

[MPaulHolmes]

Good luck! Don't die or anything.

[jackbauer]

Made it! Temp this morning was 5C on the heatsink at switch on. No problems.