Unless anyone can find an issue in the DAT file the drone is in the box and being sent to DJI Europe tomorrow. After speaking to DJI live tech support this morning and showing them the videos, they have admitted that something is not right and it needs to go in for repair. Refurbished drone coming my way I guess!
Crossing my fingers and hoping @sar104 finds something.
Okay - there is a serious problem with something on this aircraft. If you look at the comparison of magnetic yaw with IMU yaw the problem is quite obvious:
At -70 seconds the FC computes the declination and inclination as follows:
-60.921 : 1365 [L-NS][AHRS] wmm dec: -1.839030
-60.921 : 1365 [L-NS][AHRS] wmm inc: 65.357033
The difference between the magnetic yaw and initialized IMU yaw values should therefore be 1.84°, but in fact the difference is 23°. Worse than that, the difference varies with yaw:
Both IMUs are showing a similar problem, which probably rules out fault rate gyros etc. In any case, that characteristic sinusoidal variation with a period of 360° usually indicates that the compass is not properly calibrated to subtract the magnetic field of the aircraft, but that doesn't explain the incorrect initialization. If careful compass calibration doesn't fix it then it will need to be replaced.
@BudWalker ?