VTOL Carsh due to EKF3 IMUO MAG0 ground MAG ANOMALY

[xingxing_wong]

Before start, my mail configuration is as follows
FC:speedy bee F405 WING
GPS:BN880
VTOL MOTOR: EXMAX 2807 1300KV
ESC:DLRC 4IN1
REAL MOTOR:JF2814 700kv
FIRMWARE: Plane 4.4.4
I converted a delta-wing into a VTOL,It has flown successfully several times.

The last time I took off from the roof of my car, there was a fault message as the plane took off:
EKF3 IMUO MAG0 ground MAG ANOMALY，YAW-Reakugbed,
I did not pay attention to it at that time and chose to continue flying. During trans to FIX wing mode, the plane suddenly turned over. I switch back to QLOITER mode, but it could not became stabilize, and the plane crashed.

Unfortunately, the plane crashed, but I installed DJI air unit on the plane, and the video has OSD data throughout the flight. And the flight controller also recorded the entire flight process. I hope some one can analyze what is the reason for this crash, and write out the entire process of analysis, not only to find out the reasons for developers to improve, but also as a case to learn from each other.
VIDEO:

VTOL crash because of EKF VARIANCE

log file:https://drive.google.com/file/d/1R4iN4i-KvPauPrdcfT0qclxSHEZwXxIF/view?usp=sharing

I also checked the forum and found that many people have encountered this malfunction, but in the end, it seems that the specific cause has not been determined

EKF3 IMU0 MAG0 ground mag anomaly, yaw re-aligned

Hello! To begin with, I will tell you a little about my aircraft. My frame size is 800x800x500mm. I use Cube Orange, Here3 GNSS, Hobbywing XRotor Pro X6 motors and 2 LiPO 22000mah batteries. I’ve been flying steadily for a few months, but recently I’ve been getting the “EKF3 IMU0 MAG0 ground mag anomaly, yaw re-aligned” error when climbing. When this error occurs, my drone behaves very unstable. After I climb to a height of 10-15 meters, the compass is recalibrated and the drone behaves stabl…

THE GPS MODULE HAS BEEN PLACE FAR AWAY FROM THE POWER WIRE.

[Curly_Fork]

Maybe the magnetometers are confused because of taking off on top of the car.
If this is the case, you should try to power on and take off away from large metal objects

[xingxing_wong]

I looked at the flight video again carefully, maybe it was not the compass problem that led to the crash. From the video, we can see that during the conversion to the fixed wing, the current was too large, reaching a maximum of 132A, which is impossible to be that large normally, and the vertical speed was negative during the climb process. There may be a problem inside the flight controller, or there is a problem with the firmware.The reason why the plane suddenly crashed in the end was because the current was too large, and the ESC overheated protection