Consistent yaw problem ?

[RoboticTundra]

I purchased an MPP 2 weeks ago. I've been experimenting with both GO4 and Litchi and am having consistent yaw problems, mostly to left, occasionally to right. Took a bit of time to determine that it was likely not just user error (though may yet be).

First, am I correct that the term 'heading' is the direction that the Mavic is facing, not the direction that it is flying? What then is the term for the direction that it is flying?

I created a mission in both GO4 and Litchi that flies directly north (0°) from waypoint 1 to waypoint 2 through a garden that has fences and paths running 0° north<>south and 90° east<>west. It appears to fly correctly between the waypoints but is nearly always aimed about -10° (so heading of 350°? But flying 0°). Same behavior in GO4 and Litchi.

In one instance it flew the path correctly but had a heading of about +10°. This was the first run after it had turned CCW to begin the run vs others where it turned CW. So it's possible that it is frequently turning less, in whatever direction, than it should?

Some prior discussion here: Correcting Yaw w/ Litchi ?

Path:



Images from flight to follow.

 

[RoboticTundra]

Images from flights. All are GO4.

1st flight. Yawed about +10°:

2nd flight yawed about -10°:

3rd flight yawed about -10°

 

[RoboticTundra]

Flight from Airdata:

 

[Former Member]

What then is the term for the direction that it is flying?

Course = Flight Path
Heading = Orientation

Try recalibrating IMU, compass and controller.

 

[RoboticTundra]

Thanks.

I did IMU and compass last week. I can maybe find a park or something a bit more open to do the compass again. Would controller calibration have an impact when it's flying waypoints from Go4 or Litchi?

 

[RoboticTundra]

Did IMU, Compass and Controller calibrations. Definitely better but still not there. It seems much more stable and when free flying it no longer slowly rotates left as it did previously. Is that IMU (gyros?) controlling that and happier now or likely RC sticks?

I did 9 runs of the short test flight with Go4. It never nailed it. When yawing left it is on average not as far off as before. A couple of runs were off by as much as before but others were closer so definite improvement.

It definitely differs based on the direction of turn prior to starting the run. If it approaches the start point from the east and rotates CW then it is always off towards the west (E.G., it is not rotating far enough).

If it approaches from the west and so rotates CCW it is close (and possibly as close as it can be?) though there was some variability in these three runs. On one run it was slightly off to the east (right), on the other two it started the run slightly off to the left and then slowly moved to being off to the right. This may be correct based on the waypoints though.

It also missed the two precision landing attempts by a bit. Both times I'd taken it up to 11m and let it sit for a bit. It started from a 30" orange pad w/ an H in the middle. One time it had two legs in and two out but gimbal in the grass and the second time it had one leg on the edge and three legs out.

BTW, MPP, Go4 and Litchi are all latest firmware.