I’m new to these forums but I tried APM 1-2 years ago. gave up on it and now I’m back with pixhawk and arducopter 3.1. everything is great so far and I’ll soon throw away / sell all my Aeroquad and DJI stuff
Everything went perfectly so far, no crashes. I’m using an old DJI F450 frame with their motors 8x4.5 props with 4S lipo. pixhawk is setup correctly as far i can tell and see from the manuals, 3DR gps / mag mounted on 10-15cm mast. stable is autotuned and very stable, responsive and not oscillating. loiter is also nearly perfect. I setup with the mission planner 2.0.7 and I installed the latest 3.1 version from mission planner on the pixhawk.
But, today I tried auto and guided mode. both work but I see strange twitches on the motor 1-2 axis (on a X frame front right to back left). as I can visually see in flight is that motor 2 tilts down about 20-40deg and then quickly recovers with a bit of an overshoot tilting down motor 1. this mostly happens at the waypoints, sometimes but less often during the flight from one to next waypoint. sometimes i looks like motor 1 is going down first, not sure about that.
attached is the bin log from the pixhawk. if you graph RCOUT you see channel 1 and 2 (the motor axis i see twichting) going up and down way too far as I can tell.
problem is, I have no idea what could cause this. NTUN DPos and Pos looks good for me.
There does appear to be a problem causing spikes in demanded roll and pitch.
By the way, you’ve set a couple parameters in the wrong unit.
You set RATE_YAW_IMAX to 8. It should be 800.
You set LOITER_LAT_IMAX to 4. It should be 400.
Seems like a bug to me. We’re in the middle of completely redoing pretty much everything at the moment, so I would just wait on the next release. Your velocity tracks desired velocity pretty reasonably, and there are no spikes in desired velocity.
thanks for the answer. about LOITER_LAT_IMAX and RATE_YAW_IMAX, that was a bit confusing as there seemed to be change in mission planner and scaling in the GUI. i tried to set this to 4 and 8 instead of 400 and 800, just for testing. didn’t make any difference at all, so i forget to reset it. i’ll try with 400 and 800 again, but if I got you right, this will not help with the roll / pitch command spikes?
the strange thing for me, if its a bug in roll / pitch demand why is it on the motor axis and not the copter axis (on a X frame, from motor 1 to motor 2)?
Yeah, shouldn’t make any difference… Just something I noticed.
We’ve started avoiding unit differences between the GCS-displayed values and the actual parameter values. That’s why it used to be 4 or 8 and now it is 400 or 800.
It isn’t. There are single-sample spikes in demand on varying axes.
Thank you. I’m not that good in interpreting the logs / graphs, for me it looked like its only the motor 1 to motor 2 “axis”. anyway, does this mean that everyone is having the same issue on 3.1.2 in auto / guided?
ok a short update. I reset everything yesterday as described in my last post in (viewtopic.php?f=21&t=6313) and gave it another try today. and yes, it worked, no strange twitches in auto / guided anymore. I’m pretty sure that somewhere in the parameters i found something saying, that the quad is set up as + (plus frame) and not X. but I did not find it again. Anyway it works now and I’ll mark this thread as solved. for anyone having similar issues, I’m pretty sure it has something todo with upgrading APM Planner 2 version or APM:Copter version. I really need to read about updating APM, maybe I missed something.
