na5n
lvl.4
United States
Offline
|
One of my motors failed on my 16th flight just as it was landing (new version). No crash, just a mild tip-over with a distinct puff of white smoke coming from the arm and smelling like burned PCB.
To DJI and others:
I am an electronic engineer by trade. Took the Phantom apart this evening. The ESC on the bad motor arm is completely fried. Two of the mosfet ICs are completely blown - split wide open typical of when the die gets blown out of the package (catastrophic substrate failure). All three wires are melted. The insulation on the red wire got so hot, the insulation peeled off (split) and the wire melted into the black wire causing the black and red to be fused and shorted together. That's dumping nearly the full 5A of the battery into the short. No doubt what blew the mosfet drivers. Yellow wire melted into the two blown ICs.
What is a bit disturbing is the motor wires on 2 of the 3 remaining arms are also partially melted and clearly showing signs of severe heat stress from the ESC board all the way to the motors. Obviously, the new motors with 25% more power are drawing far more current than the wires (AWG#22 ?, generally good for 1A) can handle. I suspect the heating of the wiring caused my black and red to fuse and short together, blowing the mosfet drivers. The good motor/ESC wiring doesn't look far behind - the insulation is warped and distorted.
Disconnecting the bad motor and one good motor, the stator winding resistance is about the same, hopefully indicating the motor is OK.
I recommend going to #20 or #18 wire for the motor drives. Rework the ESC so the motor wires terminate on the motor end of the board (not the side) - this way, the wires would be shorter and arranged so they do not touch, and reduce the inductance for the reverse voltage fly-back from the stepper motors. The mosfet drivers are rated at 6-8A which should be sufficient. Disconnecting the bad ESC, I ran the other motors, which looks like you're using a 400Hz 3-phase stepper drive, PWM modulated for speed control. What is the maximum motor current draw at maximum PWM duty cycle? I didn't want to rev up the motors without the load of the props to measure the current.
For others who open up your new Phantom, I suggest rearranging the 3 motor wires (black, yellow and red) to ensure they are not touching to avoid a short when the insulation gets so hot, it starts getting soft. If the ESC fails in flight, it's coming down like a rock.
I love my Phantom and there is plenty of clever, well thought-out engineering in the craft. However, it does appear the new Ver. 2 ESC falls a bit short with the higher current draw of the new motors.
Paul
I have photos of the failed ESC if anyone at DJI wants to see them.
|
|