Brief Power Failure at ~5 second intervals, near crash

Log Analysis -- deprecated Copter Logs -- deprecated
1

Hi there, I’m looking for analysis of my log file. I am training a University group to use their new X8 from 3DR, and the craft nearly crashed catastrophically today. We are using a brand new copter, with all the calibrations complete and preliminary flight testing complete.

Yesterday we got the better part of an hour in of stick time, with no ill effects. And earlier today, we ran a short test mission in auto and again the copter performed silky smooth.

Cue the latest mission. It was a longer test mission in auto. The first half went fine: ascent to 80 meters went well and the copter started on its route. However about halfway through, it began to regularly cut its thrust completely for about half a second, about every five seconds. Happened very regularly. I have been using Arducopter systems for about three years and have never seen this problem. The copter is extremely difficult to fly because of the power shutting off like clockwork, then coming back on and rapidly correcting. In loiter I bring it down from 80 meters to as near the ground as I dare. When I change to stabilize, the copter is no longer controllable. I luck out and the copter makes a soft landing in a tree, where I then recover it with just one broken prop.

My log is attached. Any help at all is appreciated, so that I can get this university group’s equipment back in the air. Thank you!

2

Well, the pixhawk certainly isn’t trying to shut off the motors, and the pixhawk isn’t losing power. I’d say you have some really bizarro hardware issue. It’s also possible that your battery was really low if your voltage sensor isn’t calibrated. It says 3.7v/cell when the problem starts, but maybe it’s really 3.5v/cell. Could you have started with a low battery? Seems like the ESCs wouldn’t have all hit cutoff at the same time, though.

3

Ok, I suspected it was mechanical, but I wanted to eliminate the pixhawk as the source of the problem.

Definitely not battery: I managed to duplicate the error (safely) across two different batteries. I was of course sure that they were charged, and I also use these to monitor my battery in flight because I don’t trust the power module.

Of course now I am unable to duplicate the error. The only thing that has changed between working and not working is the humidity. All the times when it hasn’t had this problem it has been sunny and relatively dry. Both times when I encountered this error it was sopping wet outside, had just stopped raining.