RTL Land At Wrong Location After Battery FS

I have posted this in a couple of other areas and I cannot seem to find an answer. Thought I would try here to see if someone might be able to see something in my logs that I missed?

I loaded Copter 3.4.6 on my Pixhawk this weekend to do some flying in auto mission. I wanted to test out some of the FS options. I experienced a behavior I cannot explain with battery failsafe.

I purposely left the battery failsafe voltage higher than I normally would so the event would trigger in the middle of the auto mission. I have battery failsafe set to RTL. As I had planned, about 3-4 minutes into my auto mission the battery failsafe triggered and the pixhawk quad entered failsafe and went into RTL mode. (All expected behaviors at this point).

The quad turned and started to fly straight towards home. It flew for about 30 seconds along its course toward home until it stopped, and started to descend and land (unfortunately it was still above a woods). It was not a crash but a controlled land by the quad. The quad was still about 1240 ft from home location when it decided to land. (The quad did end up crashing as the land happened above a woods and into trees.)

I double checked the FS parameters and looked at the data flash logs. The FS parameters for the battery look correct and are set to RTL. I cannot find any additional errors in the data flash logs.

I downloaded the data flash file and posted it here as I did not know if it was too large to post otherwise:
https://drive.google.com/drive/folders/0B8-ATMPoKFG6bTluNUMwSWwyX0E?usp=sharing

Does anyone have an idea of what and where I should look next to understand what might have caused this behavior?

You appear to have partially lost power on the corner corresponding to motor 4.

You can see that that motor saturated earlier in the flight too - and it also failed to achieve its desired height during that time.

Suggest looking at physical effects on that motor - most likely a loose bullet connector, bad bearings, that sort of thing.

Peter

Peter - Thank you for that analysis. I am checking my solder joints to see if I have anything loose. What I don’t understand is why did the copter go into a controlled landing? I would have expected it to crash trying to get home instead of stopping its forward velocity and go into a landing.

Do you or anyone else have any idea or is this expected behavior given the analysis you found?

Thanks.

That wasn’t land mode - that was RTL :slight_smile:

I believe it was prioritising stability over altitude, and managed it, it seems.

Peter

Thanks for the help. That makes sense. Happy flying.