Coaxial octorotor crash analysis

We had a crash today. It is the second flight of this platform. The first test was done in the lab, stabilized mode worked ok. Now on the second test it was flying on the position mode and it lead to crash. It took off ok, then some disturbance came and it started to increasingly oscillate. I couldn’t find anything from the logs.

The flight seems to be on Altitude mode rather than GPS \ Position.
It also looks like you’ve got some vibrations (before what seems to be loss of control).
Make sure you mount the flight controller as per suggested by CubePilot and setup filters and or dampening as needed.

1 Like


Hi,
According to the log file analysis, the current and motor output were overloaded before the loss of control. Further analysis showed that the EKF environmental data showed that the error was too large, so the motor output force was constantly corrected, which eventually caused the current and motor output to be overloaded.

Hi,

Thanks for the interpretation. Yes at some point motors were clipping, but I don’t think that is the reason, that is the result of some issue. There are excessive vibrations when the drone take-offs from the ground. I think mainly the vibrations are the reason for the estimator to not work properly

Is this drone quite heavy?(15kg up?) During takeoff, the motor output is already very unbalanced. Combined with the large difference between the pitch/roll setpoint and estimate (which could also be due to accelerometer issues), it’s possible that the estimate is overreacting compared to the setpoint, causing the correction to destabilize and crash the drone.

If this is indeed caused by the PID settings, I suggest first lowering the PID gains and making sure that in stabilized mode, even under aggressive stick inputs, the estimated pitch/roll rates stay within expected bounds compared to the setpoints.

(This is based on the assumption that PID is the main issue. From the log, it seems you haven’t adjusted the PID settings. If the drone is heavy and the motors are powerful, such behavior is more likely to occur.)

Thanks for your feedback. drone is a bit heavy, it was around 10kg in this case but designed to take 10kg of payload. Well PID looks like an issue but I am still not sure. Because there are a few cases with the same behavior and it seems like caused by the excessive vibration.