EN CN
Roll and yaw affected by pitch
  • nwilhem
    Hi everyone, I've seen some topics about the same subject but no good answers for me.
    When I use the pitch in RC, the roll is "rolling" ;-) and there's some very little moves on the yaw.
    I precise that my roll is disable in the GUI.

    Thank you in advance.

    Nicolas
  • Garug
    It could be that your IMU is not perfectly aligned.
  • OctovisualsOctovisuals
    Something pretty similar happens to me, and nobody seems to know anything:

    http://forum.basecamelectronics.com/index.php?p=/discussion/864/roll-affects-yaw-and-vice-versa#Item_1

    I already perform the 6-point accelerometer calibration and my IMU is perfectly parallel to all axes.

    Thanks.
  • Garug
    But how do you know it is perfectly parallel? The IC on the board could easily be 1 degree of the parallel and that would cause already 1.7% error on other axis. See http://www.invensense.com/mems/gyro/documents/PS-MPU-6000A-00v3.4.pdf page 44. Could be good to try to change the IMU heading and see how that affects.
  • Marian
    I had to mount the IMU pcb off parallel to the axis of rotation to awoid the same issue. It could be that the IC is not sitting square on the board. It takes some experimenting with IMU pisitioning to to minimize the interference of movement from one axis to the others.
  • OctovisualsOctovisuals
    I can't know for sure, but if I recall correctly, at the beginning there was any interference.

    It started since I updated to 2.40 b7, or something, I don't know exactly when.

    Nothing to worry about, though. Just curiosity.

    I'll keep testing.
  • nwilhem
    For me, I did the calibration a few times on different flat surfaces and with different technic but still the same problem. When I move my gimbal with my hands it's seems to be all right but when I use the RC the pitch affects yaw and roll.

    Does the problem come from my RC?
  • Marian
    No. I believe that the problem is caused by the gimbal movements. The axis of rotation must be perfectly perpendicular to each other. It is very hard to achieve on these kit home made gimbals as the construction allows for the impetfection to be built in and that causes the interaction of movement between axis. The position of IMU to the axis of rotstion is crucial as well. When any of the above mentioned conditions are not met 100% the IMU is moving in both axis hence the undesirable movements of roll when you move the pitch ftom Tx. It is a hard one to eliminate completely...
  • nwilhem
    Thank you Marian, it seems to be the problem. I'll try to check all my angles but it sounds right.
  • Marian
    On most of these home assembled gimbals the lenght of arms from the pivoting points on both ends of the pitch assembly can have different lengths which invariably causes the IMU to move in two planes instead of one even when only pitch input is applied from RC. All these things are going to cause interaction of movements between axis. The arm lengths must be equal on both sides of the gimbal. For pitch from the motor shaft and on the other side from middle of the supporting bearing to the bottom bar which holds the camera. Can be hard to measure and even harder to make perfect.
Add a Comment