Can I get help from someone with reviewing this issue?
Basically the craft misbehaves in at least 2 ways when it switches from STAB to RTL due to a failsafe:
- it first tilts pretty violently (sometimes 45 degree angle or more)
- then it fights my inputs in a way that it just shakes and stutters
omersiar recommended I post here for more help after saying “When I look at your log file i can tell you that there is similarity in graph lines between COMPASS(MAG) and MOTOR OUTPUT (RCOUT). When APM hits RTL it tries to understand its heading and in your situation it leads shaking problem. Someone who expert on this can help you more”
I have a log dump and video showing the problem in the bug
Your compasses show a lot of divergence as the throttle ramps up for rtl, it doesn’t look like you have done the compass/motor calibration in mission planner.
Try it and see if it improves things. The pitch and roll outputs look like they match the requested pitch and roll requested from the Pixhawk so it looks like the Pixhawk got confused from the compass issue.
thanks for the reply.
this doesn’t quite explain how the X8 would pitch 45 degrees for a second or two (not turn, just pitch) when swtching from STAB to RTL, does it?
This X8 was tuned by 3DR which i’m pretty sure did the compass/motor calibration. Unfortunately when I upgraded from 3.1.x to 3.2, all the settings were lost by the upgrade, and indeed I did the autotune but not the compass/motor calibration. Do you know if 3DR did that before shipping it to me?
If so, it had the same exact problem with 3.1.x calibrated than it then did with 3.2
unfortunately the X8 did get destroyed on landing after the time I lost in the air during those switches that cause momentary loss of control, and the battery dying at 4300mAh used when it was a fully charged 6000mAh battery, so I can’t do further testing.
I can provide logs of similar issues with 3.1.x if that helps, and assuming it was calibrated then.