Some weeks ago I had a crash with my tricopter (3.2RC9), using Pixhawk.
Basically I wrote a small mission flying through an “8” shape and then repeat it with DO_JUMP.
During the second round, once copter started to accelerate, I heard some sound signals from Pixhawk and I then lost any control. (Log, mission, parameters are in the folder below).
Today I was flying with my newly built quad (3.2RC14), made a sample mission that was quite similar to the one in the tricopter (“0” shape from 8 spline waypoints and DO_JUMP at the end). During the second round I had exactly the same issue, Pixhawk made some sound signal, copter started to accelerate and I couln’t control anymore till hit the ground (tried to switch back to stabilize or loiter w/o success). Unfortunatelly I didn’t have any logs now because battery was disconnected during crash. (Parameters, mission file are in the folder below)
So it quite looks like spline waypoints and DO_JUMP together causes total Pixhawk breakdown.
I’ll not try spline WPs again until this BUG or explanation are found… hope you understand.
only the “tricopter crash” had any useful data (.bin) , no log for the other.
The tricopter was clearly unable to maintain roll - plot ATT.Roll va Att.DesRoll , and you will see a big difference - that gave the speed increase.
A typical sign of mechanical failure, ESC, or anything that could result in less thrust than needed, or in this case, even a servo error. - you can see yaw is also going outside desired area.
All I am unsure of, is the PWM and servo output - must assume it’s normal, on a more-than-tricopter I would expect one motor go much higher/max out before deviating that far from desired attitude, but the method is cannot be the same on a tricopter, as it’s power is not symmetric in all directions/axes.
Take that with a grain of salt, I am not really familiar with tricopters, as it’s not exactly a reliable design.
What’s clear: is that it failed to maintain attitude.