Help troubleshooting motor controller fire

hat_tr1ck

1 µW
Joined
May 7, 2019
Messages
4
Few weeks back while running a 50KW BLDC purchased off Alibaba with a custom 80kV winding by Dongguan Freerchobby Model Co., the Flier ESC F-120V-500A-A caught on fire after the motor starting up (jerky).

Here is a video of the setup, the motor acting before the incident, the incident itself, and the post mortem of the motor controller plus the power schematic.

The Flier ESC is rated to 120V and 500A as per their websitte. Not that I take their website as the absolute truth given they are not a highly reputable supplier, but I don't think over voltage was the issue. The controller was pre-charged using the schematic shown in my video, although Flier insists that the controller does not require precharge (even with the 120V battery input).
Here is the motor's Alibaba page. It's a sensor less motor.


My two standing theories are:
1. Controller and motor lost sync, leading to a current spike - Unlikely because we were running low throttle / low speed
2. Motor winding is short - Could be, supplier assures me they tested at the factory and it was fine. Motor had barely been ran before the incident, and pos incident i do not spot any burnt windings. At the same time, winding resistance to ground is all over the place (some in the MOhm range, some kOhm. My understanding is that they should have 'infinite' resistance)

Has anyone dealt with something like this before? Depending on the most likely cause, I may send the motor back to the supplier for 'repairs', or just look elsewhere for a different supplier. Motor is for an eVTOL project.
 
The theory about the current spike is probably correct.
Looks like the prop was too big and the ESC couldn't run it because it's expecting it spin up fast.
Look at Helicopter ESC's vs Quadcopter ESC's. The Helicopter ESC has a longer ramp up time from what I understand.
I've had first hand experience with some big prop motors going out of sync constantly.

IMO you should install some Hall sensors, opens up more Controller options.
 
For context, we did run this motor-propeller combo before with no problem, so it's definitely not the load of the propeller. Plus we started slow and always kept small throttle increments. Also, the prop is super light (below 1kg). That being the cause sounds unlikely to me.

We had a hall sensor but it did not talk to the ESC.
 
You should be able to test the motor. If any of the windings are shorted, there will be a lot of resistance to spinning it.

I seem to recall the same model of controller catching fire for somebody else but can't find it. There are many reasons that could cause this, but hard to find out after it's burned. Personally, I would look for another brand.
 
I'd lean towards it being a sync issue on the basis I've seen that exact behavior on startup with a IGBT drive that had a blown phase. I suppose you could at least verify your phase wiring was OK after the fire?
 
Something appears to be wrong with the phase wiring as I state on my post just based on resistance to ground being all over the place.

Don't have another ESC in hand to try running the motor again. Are there any tests that I could do without powering it? Quick visual inspection of the window looks OK, might take it apart to get a closer look. I have heard of motors losing sync and blowing up, but at higher speeds.
 
Can you borrow a megohmmeter? This is the correct way to check the integrity of the winding insulation (phase-to-phase and phase-to-ground). You are right that a regular multimeter should show near infinite resistance, but it won't pick up weak insulation as the test voltage is very low (<5V).
 
Loading on the video setup is near zero-load at low RPM.
Load progresses to max relative to RPM.
Not a spin-up issue.

There are a dozen reasons a controller can smoke out on first start.
My understanding is that this was the first run of that controller?

... These days most controllers have algorithms to help them not self destruct... but ... if you bounce against the limit repeatedly (as it appears is happening in the video) you could no doubt overheat.

What would be most useful is the profile you had set up on the controller.
For first run on unknown you generally want to set everything to 10% or 20%

Max Current 10%
Max Power 10%
Ramp Rates Low

Set everything as soggy as possible. . . just enough to overcome freewheel power
(so like 50W or a couple hundred watts)

-methods
 
Back
Top