Gimbal - Maintaining the image level with Horizon during flight - An observation

[RadioFlyerMan]

I've replaced my gimbal this year (on an MPP), and switched out up to five lens barrels to improve the focus in the entire field of view. During this whole process over the last 4 months, I've noticed that the gimbal had lost it's ability to fully compensate for the sideways roll of the AC when flying sideways or fighting a strong wind. There was some compensation, for example rotating the AC sideways when hand held, but not enough when flying. Naturally at each change I recalibrated the gimbal which did not fix this problem.

However, on my last barrel change yesterday, I did an IMU calibration and then a gimbal calibration. I then tested its ability to maintain a level image by hovering in a strong side wind (which buffeted the AC significantly), and by flying sideways into a 15+ mph wind. The results: the camera is now solid as a rock, with only a very very slight mismatch to horizontal in a worst case scenario.

So my assumptions are that:

1. The AC maintains the image at horizontal through connectivity between the gimbal and the IMU.
2. The problem was fixed through calibration and had nothing to do with installing a new lens barrel. In other words, there is nothing within the electronics of the lens barrel itself that serves to keep it level.

Do I have this correct? Comments, experiences, and first hand knowledge welcome.

 

[brett8883]

I've replaced my gimbal this year (on an MPP), and switched out up to five lens barrels to improve the focus in the entire field of view. During this whole process over the last 4 months, I've noticed that the gimbal had lost it's ability to fully compensate for the sideways roll of the AC when flying sideways or fighting a strong wind. There was some compensation, for example rotating the AC sideways when hand held, but not enough when flying. Naturally at each change I recalibrated the gimbal which did not fix this problem.

However, on my last barrel change yesterday, I did an IMU calibration and then a gimbal calibration. I then tested its ability to maintain a level image by hovering in a strong side wind (which buffeted the AC significantly), and by flying sideways into a 15+ mph wind. The results: the camera is now solid as a rock, with only a very very slight mismatch to horizontal in a worst case scenario.

So my assumptions are that:

1. The AC maintains the image at horizontal through connectivity between the gimbal and the IMU.
2. The problem was fixed through calibration and had nothing to do with installing a new lens barrel. In other words, there is nothing within the electronics of the lens barrel itself that serves to keep it level.

Do I have this correct? Comments, experiences, and first hand knowledge welcome.

I do t remember if it was from experience or explicit instruction but that’s how I’ve always done it.

Cold IMU calibration
Restart
Gimbal Calibration
Restart
Compass calibration
Restart

It makes sense that they are intertwined because while the gimbal calibration is calibrating the horizonal position based on the resting position of the Mavic on its landing gear with a calibrated offset determined by the aircraft IMU. Even If the gimbal is correctly calibrated and the IMU calibration gets out of whack I’d presume eventually it had to determine the gimbal attitude as an offset from the frame attitude. If the IMU is incorrectly reporting frame attitude due to a bad calibration you’d expect the gimbal to be crooked.

To be clear I don’t know if this is how it does it but I’m just saying it would make sense if it was. I don’t know all the intricacies of how the gimbal attitude is determined but I concur with your observations.

 

[DanMan32]

IMU to gimbal, I'll be rolling 7 degrees clockwise in 5, 4, 3, 2, 1, ROLL!

 

[RadioFlyerMan]

I do t remember if it was from experience or explicit instruction but that’s how I’ve always done it.

Cold IMU calibration
Restart
Gimbal Calibration
Restart
Compass calibration
Restart

It makes sense that they are intertwined because while the gimbal calibration is calibrating the horizonal position based on the resting position of the Mavic on its landing gear with a calibrated offset determined by the aircraft IMU. Even If the gimbal is correctly calibrated and the IMU calibration gets out of whack I’d presume eventually it had to determine the gimbal attitude as an offset from the frame attitude. If the IMU is incorrectly reporting frame attitude due to a bad calibration you’d expect the gimbal to be crooked.

To be clear I don’t know if this is how it does it but I’m just saying it would make sense if it was. I don’t know all the intricacies of how the gimbal attitude is determined but I concur with your observations.

I have definitely learned it makes a difference as to the procedure one uses. I stumbled on to the IMU calibration simply on an observation that the green line of the accelerometer was getting longer, and not because of the gimbal issue. The only thing I did not do was the compass calibration, which I will do prior to my next flight.