Controller shorts out at high power

[420b]

I have recently been struggling with my ebike's controller cutting out when the power is running at max, for instance when I am traveling up a hill at full throttle. My controller is one Chinese 72V silver box controller from ebay. After a few seconds of full power, it cuts out. The type of cut out seems to be a battery short, because it trips my BMS and shuts down my battery to 0V until I unplug. This controller previously worked without issue at this amperage and with the same battery and motor. I tried to decrease the amperage by removing some shunts (I brought it down from 40A to 30A to 20A) and it still has the same cut out issue. There are not any problems when running the bike downhill or at half throttle for a long period of time.

I wonder if anyone has had a similar issue. Maybe a mosfet is failing and causing a short at high amperage? I am lost about what to do because I have visually inspected the controller and there are no damaged components, and no shorts. I will likely buy a new identical controller because I think this one is dead.

 

[Sunder]

Shorting out at high throttle would seem unusual. Its much more likely you have a weak cell or string of cells, and its tripping the BMS because one cell is below LVC when the rest are still strong.

Are you able to build yourself an artificial load generator, good for 20A @ 4v, (A 100w resistor should be good enough for this) and watch the voltage droop for each string of cells?

If one sags more than the rest, that will be your culprit.

 

[amberwolf]

If the controller had a FET problem, it would be all the time (or it would be an overheating issue that would probably cause the FET to fail, or on a more advanced conroller cause a thermal shutdown, not a short), and if a FET was shorted ti would keep the mtoor from turning easily even by hand (and it would not turn the motor itself at all). If it was shorted in a way that caused the BMS to shutdown, it would be permanent, and probably have released smoke, too.

The problem you're having is a battery unable to supply the current being demanded of it. (a cell group dropping in voltage enough to trigger LVC, or the current sensing of the BMS tripping the shutdown to prevent overload).

I would open it up and check each cell group's voltage and note them all down here in this thread, starting from the most negative cell. Then we can go thru them and see which might be problematic.

If they are all about the same, then you'd have to do the same check but with it under a load high enough to cause the issue, but not high enough to shutdown the BMS (which would take the load off). There are a number of threads about different kinds of testing loads, and many threads about battery troulbehsooting that may have more details if you prefer.

 

[420b]

Shorting out at high throttle would seem unusual. Its much more likely you have a weak cell or string of cells, and its tripping the BMS because one cell is below LVC when the rest are still strong.

Are you able to build yourself an artificial load generator, good for 20A @ 4v, (A 100w resistor should be good enough for this) and watch the voltage droop for each string of cells?

If one sags more than the rest, that will be your culprit.

Thank you! Very good thinking, it seems very likely. I will use the motor as a load with the brake slightly applied and test my cells. Just canceled my new controller purchase.

 

[420b]

If the controller had a FET problem, it would be all the time (or it would be an overheating issue that would probably cause the FET to fail, or on a more advanced conroller cause a thermal shutdown, not a short), and if a FET was shorted ti would keep the mtoor from turning easily even by hand (and it would not turn the motor itself at all). If it was shorted in a way that caused the BMS to shutdown, it would be permanent, and probably have released smoke, too.

The problem you're having is a battery unable to supply the current being demanded of it. (a cell group dropping in voltage enough to trigger LVC, or the current sensing of the BMS tripping the shutdown to prevent overload).

I would open it up and check each cell group's voltage and note them all down here in this thread, starting from the most negative cell. Then we can go thru them and see which might be problematic.

If they are all about the same, then you'd have to do the same check but with it under a load high enough to cause the issue, but not high enough to shutdown the BMS (which would take the load off). There are a number of threads about different kinds of testing loads, and many threads about battery troulbehsooting that may have more details if you prefer.

I appreciate your help always. I will check my cells ASAP under load of the motor with brake applied (I have checked them before and they looked good when resting). Very good thinking, this is almost certainly the issue.

 

[420b]

Yep, the cells are definitely out of whack after my first ride with them. I got them all from hoverboard packs but they must be at different age levels because they are draining at different speeds. Here are some pictures of the monstrosity:




This was a sort of learning experience because I figured myself clever in that I could make my own battery and save a few bucks. Months later, I end up with this fireball waiting to happen and it seems I should have just bought a prebuilt battery from the get go.