We had an unusual failure and need help analyzing the logs. We were flying our 3DR Quad Frame with an APM 2.0 with MediaTek GPS and ArduCopter version 2.7.1. After taking off in stabilize mode, we lost control from the RC (Spektrum DX6i & AR6115e) and the quad started to drift away. After drifting over a parking lot and not regaining any RC control, we issued commands through Mission Planner to change the mode to RTL, then Auto, then RTL again. We're were in a bit of a panic which is why we didn't just leave it in RTL. The quad flew back near where we took off under Auto since we had an auto mission loaded and then once we switched it to RTL again it entered Loiter (expected since we didn't have RTL set to autoland). Once in loiter it started to drift away again (we think because the RC inputs were still stuck and in loiter you can affect direction). It finally hit the side of a building and fell about 3 ft down to the roof where we recovered it. Luckily no major damage, just a new set of props needed. But we were pretty suprised by this kind of failure. We did extensive testing of the RC system and ruled it out as a cause mainly because a loss of RC link during stabilize would have led to an immediate crash since the throttle value drops to 0 in that case. It appears that the APM inputs are either going bad, or part of the software that reads those inputs hung. But that's about as far as we were able to figure out. Here are the log files from this flight, any help is greatly appreciated!
You need to be a member of diydrones to add comments!
Replies
Possible scenarios:
Question - what/how did you program your receiver fail-safe? If they resemble the numbers above you may have simply lost signal and they locked to their programed values.
That was a tough break - even if you got it to come back through Mission Planner, (quick thinking) I'm not sure how you would have got it down. Maybe a couple of lost propellers was the best ending in this case.
Did you ever find the cause for loss of RC inputs?
I had something somewhat similar occur to me. My sequence of steps was something like this:
I initalized APM2.0; allowed it to acquire GPS, etc.
While on the ground, with plane in my hands, I switched into Stabilize mode and moved the plane around to observe/confirm APM attempts to Stablize the plane.
I also provided some manual RC inputs (move flaps a few times).
Turned OFF Stabalize mode.
I launched plane in Manual mode.
Gained some altitude.
Deliberately took a nose dive with motor OFF, then turned Stablize ON and allowed APM to automatically stablize the plane.
Then turned Stabilize OFF.
Then turned motor ON.
Gained some altitude.
Deliberately took an inverted nose dive with motor OFF, then flipped Stabilize ON for couple seconds (I am not sure if it actually succeeded at Stablizing the plane in this instance).
Then turned motor ON.
Stabilize OFF.
Gained some altitude.
Deliberately took another nose dive with motor OFF, then flipped Stablize ON.
For a second it seemed like APM may have been trying to stabalize the plane, I do not know for sure.
At this point in time the plane appeared to have not been responding to my RC Inputs.
I think during this time, in an attempt to save the plane, I probably switched once or twice between Manual/Stabilize mode.
Then plane finally hit the ground.
Fortunately no electronics appeared to have made contact with ground, just damage to the front of the plane.
I have not played with APM since the crash, been busy (rebuilding the front of the plane, installing more powerful prop motor, testing new motor, attending Horizon Air Meet 2012, etc). I plan to check if I have any logs (I don't remember if log feature turns on automatically, hopefully yes).
I actually have a video of this incident:
https://www.youtube.com/watch?v=Sxamei2wJds
Damn, this is scary.