Hey Folks,
thx to the improved weather I found some time today to play with the Position Hold and RTH, but I'm not getting along well with them with my M4 ladybird. I actually have two issues, but one is more my different expectation, so I'll here focus first on the first issue, but come back to the second one later.
As advised by the wiki I first did the RTH tests. I spend three batteries on it, but could not achieve or observe consistent nor satisfying behavior. Probably I'm just doing something obvious wrong, but I don't get it.
The craft didn't really stayed in position, often it would describe kind of circles with ca 1m radius (toilet bowl like), mostly it just "floated" around within a square of ca 1x1m^2, often it would slowly decent after switching to PH, sometimes it would even drift fastly away because the craft isn't leveled but no counter action is taken.
As regards RTH it of course starts moving when flipping the RTH switch, and the movement is roughly in direction of the home position, but in no case it returned to a position I would identify as its home, sometimes it would just move few meters, although it would have to return ca. 5-10m. The altitude remained totally unclear to me (the wiki is quite confusing on this point).
Since it was cold outside, I had the ladybird sitting on the balcony for more than 1/2 hour, before I went to the tests. Since it's not clear to me what values are reset when the M4 is disarmed, I always switched it off, waited for the blue led to come up, and then started. I also did a couple of tests, where I waited another min or so after the blue led went solid before I lifted the ladybird (to see if the GPS might be a bit better then). This had no effect as much as I could observe. The RTH were done by lifting the ladybird to about 2.5-3m, hover there, switching to PH, and centering the thr stick. The hover thr was just a tiny mark below the center position. The craft flew perfectly in manual mode, exactly as indoors. I then flew around, and with the craft typically 5-10 m away triggered the RTH. The calibrations were done indoors. For practical reasons I didn't had telemetry.
The RT calibration of course suggests itself as the root of the issues, especially I could imagine that the toilet bowl could be due to a improper mag calibration. This, however raises the question of how accurate it needs to be. Please don't answer now with "accurate" . Should the determined south be within 1°, 5°, ???. Could the temperature effect be strong enough? Also, the acc calibration might be wrong and induce off-level and hence the fast drifting in one direction. But in manual mode it would fly nicely. Anyways, the observed inconsistent behavior during the same afternoon runs, just by repeating the procedure many times, doesn't suggest a calibration thing as main issue, since it should have been equally inaccurate during all the test runs.
At this point my reasoning stops. What am I doing wrong?
Thx!
Olli