Another question about tilted horizon..

[jenius]

Sorry for starting another topic about tilted horizon..

Mine would tile just by rotating the aircraft. The attached photo shown is when the aircraft is stationary, but rotated 90 degrees to the right. If I rotate through 360 degrees, it seems to go back to normal at some point. There was little wind..

I am not convinced that this has to do with the wind speed since I was not moving the aircraft other than rotating.

I have done IMU calibration, Gimbal calibration. Nothing seem to help. Manually calibrating the gimbal in flight doesn't help, because as soon as I rotate to another direction, it will be not level again.

This is a replacement DJI unit that I have from DJI after repair. I don't seem to notice this problem in my last Mavic 2 Pro units.

What are your thoughts?

 

[Former Member]

Some gimbals do this but most dont. The gimbal is gyro stabilized, meaning that it does not immediately respond the the "level" command given to it, but takes its time and does it smoothly.

The best way to insure a level gimbal is to fly slowly, yaw slowly, and not make sudden movements.

If your gimbal has always done this, and continues to do this, you can manually offset the "roll" by keeping the "adjust gimbal" window open and leveling it when its off. Not too much trouble. If you cant live with that, then send the drone back.

One last thing. If you need to "adjust gimbal" roll to more than 3.0 in either direction, then something in your gimbal is bent, and thats what is causing the tilt.

 

[webvan]

Huh, I thought this was a MP1 problem only and that it had been fixed on the M2 ?

 

[Former Member]

Huh, I thought this was a MP1 problem only and that it had been fixed on the M2 ?

Nope. A tilted horizon can happen on any drone.

It really depends on a number of things. Crash damage, worn gimbal motors, leaving your gimbal clamp on too many times can ruin any gimbal, then it will need to be adjusted often.

 

[webvan]

Right, I can understand some "misuse" requiring resets from time to time but I should have clarified that I meant the "drunken tilt" where the horizon tilts as you yaw and slowly corrects itself. That doesn't look like something that is due to "misuse" ?

 

[Former Member]

Right, I can understand some "misuse" requiring resets from time to time but I should have clarified that I meant the "drunken tilt" where the horizon tilts as you yaw and slowly corrects itself. That doesn't look like something that is due to "misuse" ?

Nope that is normal gyro-stabilized gimbal movement. The gyros dampen the harsh movement of the gimbal, and sometimes it does take a moment or two for the gimbal to find level again.

Thats why I urge people not to spin their drone too fast around the yaw axis. The yaw motor is the weakest of the 3 motors, and continued extreme motion might cause it to prematurely fail.

 

[AMann]

Nope that is normal gyro-stabilized gimbal movement. The gyros dampen the harsh movement of the gimbal, and sometimes it does take a moment or two for the gimbal to find level again.

Thats why I urge people not to spin their drone too fast around the yaw axis. The yaw motor is the weakest of the 3 motors, and continued extreme motion might cause it to prematurely fail.

Curious- is it an electronic gyro (accelerometer) or a electromechanical gyro?

 

[Former Member]

Curious- is it an electronic gyro (accelerometer) or a electromechanical gyro?

The former (im too lazy to type it lol)

I remember when I first started into RC helis, some people still used the latter. Took them a while to spool up, and were never as accurate as the former.

These were used in the old days:

 

[AMann]

Thanks! I like the video. My father taught mechanical engineering and had some from various missiles in his office- loved powering them up and playing with them. Too bad they belonged to the school, it would have been cool to keep one.

 

[Paul2660]

I see this on both MP2's when shooting video. One seems worse than the other, and the faster the yaw, the tilt increases slightly. On the 1st Mp2, I did make a manual adjustment at times since the tilt did not correct it self after the yawing. I finally did a full gimbal adjustment and re calibrated the IMUs which seemed to help it stay stabilized and level. However on a windy day, like my last flight, I just gave up as the drone was fighting a strong wind all day long.

Just flew this a.m. and in calmer conditions, no problems, as long as I keep the yaw to a steady speed.

Another note, is that when I switch to photo mode the gimbal pops right back into level, and stays as such until I try a video pan.

The off level amount is not as bad as what was shown in OP's picture.

Paul C

 

[webvan]

Isn't the problem due more to the roll axis than to the yaw ? I mean the Spark has no yaw mechanical stabilization and it does not exhibit any of the "drunken tilt" of the MP1, at least the one I had, and many others based on feedback like this one :

 

[Former Member]

Isn't the problem due more to the roll axis than to the yaw ? I mean the Spark has no yaw mechanical stabilization and it does not exhibit any of the "drunken tilt" of the MP1, at least the one I had, and many others based on feedback like this one :

Good observation. Its a combo of both roll and yaw. While yawing the drone, the roll has to maintain the horizon best it can. The faster you yaw the drone, the harder it is for the gimbal roll motor to keep up. Finally when the drone stops yawing, the roll can stabilize. The yaw motor has limited range, whereas the roll has alot more and responds alot faster.

 

[Former Member]

By the way, by yawing the drone too fast, the yaw motor can get "dizzy" and disoriented. As it tries to keep up with the drones yaw movement, it is being pushed to its limits. Once the drone stops, the gyro tries to recenter the yaw axis, and it might take a few moments for it to get reoriented. Same happens when a pilot applies hard braking and causes the drone to rear up like a King Cobra. The pitch motor is pushed to its limits, and might take a while to stabilize.

 

[webvan]

Makes sense thanks, in your experience isn't it less visible on the MA and the M2 though ? I can't really find any videos that show any tilting to that extent.

 

[Former Member]

Makes sense thanks, in your experience isn't it less visible on the MA and the M2 though ? I can't really find any videos that show any tilting to that extent.

Yes it was alot more pervasive in the MP and MPP than on the MA, M2P or Z. Still, after contact or damage, any gimbal will do it. Worn motors, or that have been subjected to stress, like being held in place with the gimbal lock on, will more likely do this as well.

The M2 gimbal has its own peccadilloes. On the M2Z, if the gimbal gets tapped by contact while its powered on, the pitch motor could spin around 180° and get stuck there. If an owner is lucky, it can be forced back into position without requiring disassembly.

On the M2P, If you land too hard even one time, the left side of the gimbal pitch arm can pop out of the assembly and then it will need to be disassembled to be reset into place.

 

[webvan]

So I sent back the Mavic Alpine and got an Onyx Air and...there is none of that "horizon tilting" as I had hoped for. The horizon stays level when you yaw (even too fast, just for test purposes) and turn. Too bad it doesn't have the signal stability of Occusync when flying through trees and its battery life !