Pixhawk Overheat?

ArduCopter
1

While shooting video of a nearby butte, I momentarily lost sight of my Tarot quad as it flew around the side and crashed before I hit RTL to bring it back (Click here). Fortunately, I had equipped it with a RID module that broadcast it’s position and I was able to locate and recover the aircraft.

Upon examination, the only apparent damage was to the power distribution board and two chipped propeller blades. I replaced the PDB, soldered BEC connections for the R/C receiver and controllers and was able to command the landing gear to extend to the landing position and reassembled the aircraft. After completing a radio and compass calibration, I decided it was ready for a test flight.

Takeoff was okay and it seemed to fly okay in STABILIZE mode, but when I switched to ALT HOLD, it started to climb and fly away. I immediately switched to RTL, and it responded as expected - climbed to its assigned altitude, flew over the launch point and descended. The descent was very unstable with significant wobble. I had a GoPro Hero 4 camera in the gimbal and it fell off due to the wobbling. Despite the shift in C/G, it landed okay and I immediately disarmed the motors. Then I noticed a faint whisp of smoke coming from around the center of the aircraft and when I touched the flight controller mount, I almost burned my fingers.

The Pixhawk mount is raised up a bit using nylon spacers in order to mount the landing gear controller underneath. When I removed the mount, the front left and right rear spacers were partially melted and the bolt heads were scorched.

DSCF3873
DSCF38731920×1440 615 KB

You can see heat damage to the mount where the spacers made contact. Examination of the flight controller casing showed no heat damage whatsoever and later, I was able to plug it into my laptop running Mission Planner and it appears to be running okay.

I removed the top deck, exposing the connections between the ESCs and the wires going through the arms to the motors - no visible indication of heat damage.

Middle Deck
Middle Deck1920×1440 543 KB

The long bolts that secure the flight controller mount go into the threaded holes at the ends of the arms - again no visible sign of heat damage. At this point, I am mystified as to what the heat source was.

I will have to strip away the heat shrink around the bullet connectors in order to disconnect the ESCs from the motor wires so that I can remove the middle deck and expose the power distribution and ESC layout. BTW, the PDB is functional as I was able to apply power to it.

2

Check for frayed wiring everywhere. I imagine those screws served as a path to ground for some short, maybe out in the arms by the motors. Carbon fiber is conductive, and splinters can pierce wiring shielding with ease. That is what I would do first.

1 Like
3

I think you are correct! I was totally mystified by the lack of visible heat damage anywhere else. So, after a close visual inspection of the wires coming out of the arms, I found one on the left front and one on the right rear where the insulation was compromised. I suspect that as I screwed the hex bolts in to secure the controller mount, they pierced the insulation, and the carbon fiber plate completed the circuit. I plan to cover the spots with some heat shrink and carefully put it back together. I don’t know if the associated ESCs are okay - I will be optimistic and see what happens on the next test flight.

I have built another Tarot quad - also waiting on propellers from China; I will inspect it in order to make sure that I didn’t repeat the mistake.

Thanks for your suggestion!

1 Like
4

If you can see the motor coils, and observe them between flights, you can often catch a motor that is starting to cook. Regarding ESCs, there’s no way to tell, but if the short isn’t through the ESC then you might be fine. You’d have to think through the potential short paths and inspect everything involved.

Edit: The other failure mode is that hot wires start to melt insulation, and then something metal is absorbed through the melted insulation and makes contact. It might be hard to tell the difference, but check all your wiring to see if that is happening in other places.

5

I did a close visual inspection of the motors - they all appear to be okay. I have applied some heat shrink over one compromised area and I’ve decided to glue nylon spacers to protect the wires when I screw the hex bolts in during reassembly:

Spacer
Spacer885×684 171 KB

After I have the aircraft put back together, I plan to run the motor/compass test, installing the propellers upside down to produce negative thrust. I will be able to throttle up the motors for an extended time and see how they behave and to make sure that there is no heat issues. If successful, then a brief flight test - hover just out of ground effect for a minute or so. will be in order.

6

Update:

I have repaired the damaged motor wires and reassembled the aircraft, but there is some heat damage to the bottom of the flight controller shock absorber mount:

Heat Damage
Heat Damage1920×1440 636 KB

I have ordered a replacement which should arrive tomorrow (Saturday). After I mount and hook up the Pixhawk, I will conduct a radio and compass calibration and wait for favorable weather to try a test flight.

1 Like