Recently had a interesting event with a large X8 style copter that I would like some help with. During a test flight there was a sudden drastic departure of controlled flight that occurred once. The aircraft recovered and was landed safely, but determining the cause has been difficult.
The aircraft was in forward flight and reached a peak of around 6.5 m/s when the front right arm suddenly ascended. The copter pitched up to 32deg and rolled left to 45 deg (angle_max was set to 25deg). During the pitch/roll excursion the copter gained some altitude while CTUN.ThO dropped to 0.
Right now it is believed to be a mechanical or electrical fault. The first thought that occurred was that a motor or arm briefly failed in flight. However, the RCOU data does not show any of the expected peaks in output from trying to increase thrust on a dead motor/arm. Some of the outputs actually drop to idle during the event. None go above ~60%. Additionally, after a tear down and component testing we can not locate any potential cause.
If anyone has any time and can look at this log it would be greatly appreciated.
Log File Link: https://drive.google.com/file/d/1Mk0hJrPRwp0ulAFBaOsAQJRfiZPhMKgt/view?usp=sharing
Looking at the log, my guess is one of the two motors/escs on the opposite arm (back left) had a brief failure and recovered. It’s had to picture a scenario where motors increase power out without a command to do so, so it’s far more likely an opposite motor had some kind of hiccup dropping power.
It appears ArduCopter responded rapidly and appropriately. It shows the uncommanded pitch up and roll left. Motors 1 and 5 went to minimum, and motors 2 and 6 went almost to minimum. This is consistent with responding to the uncommanded pitch up, while maintaining yaw control. The rear motors had a similar opposite response, increasing thrust to bring to the pitch down and correct the roll.
It also appears that this copter is extremely overpowered. This entire flight, the throttle output only ranged from 12-18%, even in a takeoff climb. That actually leaves very little room from the motors to vary thrust to compensate for problems. I’m not sure it contributed to anything about this, since I think the problem was mechanical. But this could be a problem for you down the road. Or, it could be that this is a heavy lift bird that just didn’t have any payload on it. You may want to strap a rock to it or something to keep the power more balanced.
Thank you for the quick response. You are absolutely correct that this is an unloaded heavy lift platform (It is an older copter that is being refurbished). It makes sense that the vehicle would not command full throttle in that case. Especially considering the ascent that followed. I don’t have a complete explanation for that, but maybe because the vehicle was in forward flight and rapidly pitched up it entered a state similar to auto-rotation where it gained lift even with the drastically reduced throttle.
We have replacement motors and ESCs ready to be fitted and I have a calibrated rock and some scotch tape on standby by for the next flight. We will continue to examine the old parts for the precise cause of the failure. I will try to update when the issue is determined.
@DroneWrangler - Maybe we should add rock calibration to the Wiki lol!
Just an update, we think the problem was a damaged power connector leading to the ESC. The connector was slightly crushed and likely not making full contact. There were no signs of arching or other electrical damage on the connector. Our current theory is that it was connected enough to supply low current, but was unable to supply larger amounts of current or it was vibrating in a way that allowed it to momentarily disconnect. We have replaced the connector and completed a few more test flights. The vehicle needs some tuning, but is otherwise airworthy.
Thank you all for your help