Reasons for excessive FET temps?

[HK12K]

I've got a Max-E pushing a QS205 50H 4t and have noticed some strange behavior lately.

Last week I pointed the bike up a steep hill, nailed the throttle, and around half way up limiting started kicking in. When I checked motor temps were well within range but the fet temps had climbed towards overheating. Pulled over and off throttle the temps quickly dropped from 70 to 40 but as soon as I leaned on the throttle again the fet temps quickly shot back up.

Since then this seems to have become a more common occurrence, and it doesn't necessarily have to be uphill. Riding around town on reasonably flat ground the fet temps will suddenly decide to climb and after that point any real throttle input will have them shooting from 40 towards 70. Even throwing it into eco mode with battery and phase amps set low enough to walk beside it at WOT doesn't stop this behavior at that point, but if I keep my throttle inputs to a minimum it seems that I can continue to limp it along.

It doesn't seem to be a progressive type of thing either, or at least not linear. That is to say it'll slowly warm from 30-40 but once this starts happening it's an extremely rapid rise from 40 to 70 even under very little load, then immediately back down to 40 again off throttle.

The controller IS mounted inside the frame, but it was in there during the summer as well without this being an issue so I'm not sure if that's directly related. Hadn't recently made any changes within the Adaptto that could account for this, however I did change a flat not too long before this happened. Positive the wires are all in the same orientation when it went back together, but not positive there isn't some unseen insulation damage hidden somewhere. Motor is running quiet and cool, but wondering if emf leakage could account for this or what?

 

[amberwolf]

If the FETs are already damaged (from whatever cause) their resistance tends to go up, which causes greater heating, which causes resistance to increase, etc. Thermal runaway and smoke can follow if it gets bad enough.

Gate drive being insufficient can also cause this, so whatever powers the gates inside the controller might be low in voltage, or unable to supply enough current to the gates. (meaning the FETs aren't turning fully on fast enough so they heat up more than they should).

Poor connections might cause it if there's arcing (RF) in the connection somewhere.

Wrong phase/hall order (if it's sensored) can definitely cause it, but usually causes excessive current draw, motor heating, phase wire heating, rough behavior under some conditions, poor motor torque, etc. So you would likely have other problems too, if this was the case.

 

[HK12K]

Thanks AW, you've given me some things to consider and explore.

I'll break out my fine toothed comb.

 

[HK12K]

So, a little update. I disassembled the bike with the intent on relocating the controller to the exterior but as it turned out I would have needed to rewire a number of things to more appropriate lengths and I wasn't really prepared to do it all on that day, so it went back together with the controller mounted internally. I did however change the orientation of the controller so it was on end with as much surface area as possible touching the metal side plates of the Raptor to more easily facilitate heat transfer.

After reassembly I took the bike for a rip through 1.5' of snow in boost mode, throttle pinned all the while, and the fet temps stayed cool as a cucumber. Put it away for a few days and went for an extended ride earlier today. Ended up doing the speed limit on a 50kph road going up the side of a mountain in Coquitlam without a hiccup, temps all good. Until I pulled onto a multi use path and had to drop my speed. Suddenly my fet temps started spiking going from 40 to 70 in the blink of an eye. Stop and let it rest, temps drop to the high 30's, ease on throttle and bam back to 70c with less than a 20 amp load per the display. Finally I gave the side cover a hard open handed slap and the problem immediately subsided for the rest of the day.

Gonna go out on a limb and say it's an internal controller issue.

Does anyone know if Adaptto is still doing repairs in any capacity, or know of anyone who can service them?

Guess I should just go ahead and order that Nuc 24f I've had my eye on. (If only they'd finalize and release their BMS already)

 

[HK12K]

Giving this a bit more thought...

Does anyone know how exactly the controller determines mosfet temp?

I'm familiar with the concept of measuring temperature using changes in wire resistance and suspect the controller is doing something along those lines to determine temps, but that is a wild guess at best. Practically speaking I have no idea how this is really being accomplished under the cover. Are the fets themselves being monitored for resistance changes or is there an external heatsink with a sense wire attached or something else entirely?

 

[billvon]

Does anyone know how exactly the controller determines mosfet temp?

It's almost certainly a thermistor. They are cheap, small and simple. There are ways to use specialized FETs to measure internal die temperature, but then you are locked into a specific (and expensive) FET.

 

[HK12K]

Does anyone know how exactly the controller determines mosfet temp?

It's almost certainly a thermistor. They are cheap, small and simple. There are ways to use specialized FETs to measure internal die temperature, but then you are locked into a specific (and expensive) FET.

Beautiful, that's exactly what I wanted to hear. If that's ultimately my issue it sounds possible to fix locally.

Anyone happen to know exactly how and where it's implemented in the Max-E? If I can lay eyes on it I may be able to discern why it's acting up, or at least point at the right part when showing it to a local generic repair service, if I can find one.

 

[amberwolf]

usually a thermistor is just a tiny few-mm blob on the end of a pair of very thin wires. they're usually either glued or bonded in some way to the device they are measuring, so most likely stuck to the heatsink for the fets, or one of the fet bodies.

otherwise, post some pics of the internals if you have it apart; maybe a "trained eye" can spot it.

this links to various images of differetnt ypes of thermistor
https://www.google.com/search?q=thermistor&source=lnms&tbm=isch
such as this

 

[HK12K]

I don't have it open at the moment but will see if I can find some high res photos of the internals. I did however find the following video and around 1:13 it looks like there might be one nestled there in the middle between two fets. Hard to say for certain though at the moment.

https://www.youtube.com/watch?v=WoA0uC53djE