We had an unexpected crash on our 25 kg MTOW quadcopter prototype. It has over 5 hours of flight time and was working fine up until this crash. I’ll summarize the incident and add the logs and system information.
Summary:
We were trying a flight without RC. Commands were only given from MissionPlanner
We performed an “arm and takeoff”. It took off in Guided Mode. Reached to an altitude of 10 meters. Then pilot switched the mode to “Land” from MissionPlanner.
While it was descending normally, right landing leg touched the ground and it had a 180 degrees flip.
When we checked the logs to find the reason of crash, we saw that the log abruptly cut at the moment of touchdown. We don’t know if that’s a mechanical power failure to the FC or only SD card vibration/contact issues upon impact (i.e. FC was still intact).
Other electronics were working fine even after the crash. We were able to retrieve SiYi camera feed after the crash.
We powered up only the flight controller from USB and created another log file after the incident. There were no crash dump file, so we assume it was not a software problem.
Log file and the video from the downward facing camera is here:
System Information:
Ardupilot Firmware: Arducopter V4.6.2 (1EBD4D99)
Frame: Quad X
FC: Holybro Pix32 v6
Propulsion system: Hobbywing XRotor X6 Plus
Do you have any suggestions how to move forward and find the root cause? Since we don’t have the log, we are a bit lost here. We want to fix the problematic part for the next build.
While the logging rate isn’t high enough to determine exactly when touchdown occurs, it is clear that something goes wrong just before the log cuts out. Assuming from the video that the video is facing craft forward, the IMU thinks it is rolling to the left (negative roll is to the left) so I’m sure the controller decides to counter with a right roll. Also, Z acceleration magnitude increases, which to the controller might be seen as “we’re ascending”, but how much this was playing into aircraft behavior would take a deeper dive.
When I’ve seen IMUs think that something happened that doesn’t look real (you weren’t rolling to the left) my first thought is to make sure that the flight controller didn’t come loose somehow, or maybe some cable or part started vibrating against the cube housing or something right at touchdown.
I looked at the tlog, and telemetry cuts out 10cm before touchdown, assuming that the rangefinder sits 52cm above ground while the landing gear is on the ground. Telemetry is restored about 12 seconds later, AP re-initializes 8 seconds after telemetry drops.
You only need one battery monitor configured for ESC data. It automatically sums the current from all ESCs and averages the voltages. You can disable the BATT2/3/4 monitors entirely.
The problem is the flight controller loses power and all logging stops, as does all control.
Check if your power brick is overheating, and that you dont have two power sources fighting each other. Maybe show a couple of pictures of your wiring and power brick connected to the flight controller.
