So after sitting for at least 30, it seems stable. EKF was almost at the bottom and the compass was locked.
I went outside and moved it around, rotated, etc. and put it back on the ground. Watching the screen now, after about 10 minutes of it not moving, ekf is staying around .3 - .4 (green) and yaw is slowly rotating counterclockwise.
I have the exact same problem and its really frustrating.
Mine started happening after I tried plugging in the HK digital airspeed, I lost gps and compass, unplugged it and realised that the HK digital airspeed has the pins opposite to the 3DR one. So I cut and soldered the 4 wires, so they’re correct and recalibrated everything, but get the same problem. I’ll try to flash the new beta code tonight and let you know how that goes, the new beta code also has a new version of EKF, however that version requires a stable compass.
To be able to fly the last couple of weeks I’ve just disabled compass and done a manual take off.
Well, my compass “drift” seems to be under control, but now I’m a little concerned about the high ekf numbers. It seems like the slightest bit of movement sends it in to the yellow zone and it will stay there almost indefinitely.
Any chance of some dataflash log files (from the SDCard) when you see this issue? Matching tlog’s fromteh ground control station would be ideal as well. That will help us analyse the problem.
When you say goes into the yellow zone which bar in the graph are you talking about? A screen shot would be great.
Thanks, Grant.
One EKF status screen, the bar for the compass will go yellow (or red) and stay there…sometimes for many minutes, sometimes indefinitely.
I still haven’t thrown this thing up in the air, although I do have a little more confidence that my initial problem was mostly resolved after I messed with the declination. Just need to find some time now.
I’ll capture the logs when I do.
I’m still not sure if the “yellow zone” thing and my phantom yawing are related or not.
Hi Everyone,
I appear to have the same issue and I have just spent 10 hours looking at it.
I “think” I know what’s going on.
Here is what I found.
Both external compass readings appear on compass 1 channel (looking at the MP calibration 3 axis twirly bird).
No compass 3 channel readings in calibration.
Can’t calibrate compass 1 external
compass 2 internal looks good in calibration
Heading output is porridge, EKF red.
Calibration works and looks good on both external and internal. (The dots for both compass follow what I’m doing with the plane in calibration)
Heading output is porridge when setting for external primary and only use external. (Compass 1)
Heading output is OK EKF triggers when setting for internal primary Compass 2 and use external Compass 1
Everything stabilizes when disabling the external compass.
compass 2 primary, enable compass 2, disable compass 1.
(You need to reboot to get the EKF to clear)
To be clear on my board the internal compass gets moved to the compass 2 designation when an external compass is plugged in.
I wish for CLI access with each compass channel output available as a txt stream. All the sensors would be good.
I’m “thinking” plane 3.4 only works with an internal compass and I’m fairly sure it won’t work with 3.
Please let me know if my observations are consistent with your setups.
Thanks
Paul