Just built up a custom spider/deadcat-style quadcopter and had two flyaways on the first day trying it out. Talk about two exciting maiden flights! Miraculously, I recovered the quad after both flyaways with no damage on the first and only minor damage on the second. I've known and loved the APM on a previous Tricopter build - any help with the log files would be greatly appreciated.
The first flyaway occurred in Loiter mode. After about 30sec in Loiter mode, the quad took off, straight up at 100% throttle. No response to lowering the throttle to minimum. Quad hit about 650ft relative altitude and was getting hard to see. I then panicked and completely turned off the Tx hoping to invoke the failsafe Landing mode. That seemed to work (see below...don't believe it actually worked, must have been coincidence) - quad came straight down at a slow, controlled speed and all was well - no damage.
The second flyaway was much more exciting - this time it occurred in Altitude Hold mode. After about 2 min of relatively gentle flying, the quad once again railed the throttle to 100% and took off. I lowered the throttle to minimum - no response. Since it seemed to work previously, I then turned off the Tx... no response - still climbing! At this point all I could do was wait and hope to see/hear it come down. Quad hit ~1,800ft - well out of sight and I thought it was gone for good. By some miraculous stroke of luck, the wind somehow blew the quad back within sight (see the GPS track in Google Earth, it's unbelievable) as it was coming down about 2 min later - I saw it drop below a ridge about 200 yards away. Took 1.5 hrs to find it but I've got it back relatively unscathed and with the log file to show for it.
I haven't determined what caused the 100% throttle flyaway nor what caused the decent in either case. (In the second case, the Tx was completely off for 60-90 seconds before the quad started decending.)
Any ideas determining what caused these flyaways would be very helpful. I've looked through the logs and don't see anything obvious. The vibration levels at the APM are higher than desired but could that really be the root cause? Thanks for looking.
PS - I've since read that you should switch back to stabilize mode during a flyaway - one of many lessons learned from this experience.
Replies
You may want to check the power connections to the RX and do a range test. Are you running telemetry? I am on a mobile so I do not have a way to run the logs as of now
Thanks Tearig - I'll double-check the range test of the Rx, but the short-range mentioned above was because I actually turned off the Tx hoping to put the copter in fail-safe land mode. I'm not running telemetry, just have the log files at the moment.
Agreed - I've read that switching back to stabalize (I actually took off in stabalize mode both times) would have allowed me to regain control. Do you think the vibration caused the flyaway?
Yes, vibrations i Z axis are a common reason to gaining height, (it responds to "falling")
Also, Alt-Hold mode does not care about GPS, so the only difference between stabilize and Alt-Hold, is the throttle control.
Excellent - thanks for the advice!
In both cases the quadcopter lost RC link and failed to to into failsafe.
I would suggest you double-check the failsafe on the ground whilst the copter is disarmed.
Also check the range of the RC transmitter too. It seems awfully short range.
Thanks Stephen! You're absolutely correct, I actually turned off the Tx on purpose (hence the short range) and the copter did not go into failsafe - I'll take your suggestion and look into getting that working on the ground with it disarmed.
Do you see any reason it may have taken off like that in the first place?