EN CN
Mos 32bit 3axis 2.41 b5 - All stable until RC input - then pitch oscillates
  • Stolman
    RCTimer gimbal operates normally with the RC input disabled. Once the receiver is powered up with the XMTR on, the pitch begins to oscillate. Both roll and yaw work flawlessly with or without receiver & XMTR power.

    Sony A6000 camera.

    Pitch has a VERY slow (1 x per 5 sec) oscillation. Power on and XMTR / Receiver (Spektrum DX9 / AR9020) on.

    The motors are set to 110 (120 and above they start to get pretty warm) and the PID settings are:

    P-I-D
    Pitch: 1-.01-0 (INV) Oscillation only gets more intense with any increase in P or I or D settings, regardless of permutation.
    Roll: 3-.05-1
    Yaw: 3-.05-1

    No matter what settings I use for Pitch, I cannot eliminate the slow oscillation. When I increase D to the oscillation is significantly exacerbated. Same with increasing I and P. I went all the way up to I=.05 and at each .01 step I increased first P from 1 to 9 and then D from 1 to 9.

    GYRO Trust: 80
    PWM Freq: High
    RC Control: Angle (for Pitch and Yaw)

    FRAME IMU: Installed but disabled (disabling made no difference)

    I have very little hair left after tearing it out on this.
    Any helpful suggestions would be greatly appreciated, especially if you have had/seen this kind of problem.

    Thanks in advance.
  • Garug
    Your PID settings are very low.

    Make sure that your basic settings are correctly performed. (Mechanical, Balancing, IMU orientation, Motor poles/inverted....) Perform the setup only with Camera IMU. When everything is working, then add the Frame IMU.

    Not knowing your gimbal, but generally P and D should be well above 10. Does your gimbal need the maximum power? if not reduce power. Too much power is not good. Also as your motors are getting hot already at Power 110, you could use lower voltage.

    However, this is strange "Pitch has a VERY slow (1 x per 5 sec) oscillation. Power on and XMTR / Receiver (Spektrum DX9 / AR9020) on." I do not know why PID settings would cause this. What does the realtime tab show for the Pitch RC input, is it oscillating?
  • Stolman
    Garug,

    Thank you for the prompt reply!

    I started over from scratch. When I did I had an interesting observation.

    With Frame IMU disconnected and disabled, I secured the camera IMU on a wooden block, sitting on table with drone.
    The initial position of the camera IMU is the same as when mounted on gimbal (connector points to rear - away from camera & "X" points Right)

    With GUI Sensor Orientation of Top= "-Z" and right ="X" - I hit Calibrate Accelerometer.

    The small LED on the controller flashes quickly about 10 times then goes to a continuous steady ON state with all Yaw/Pitch/Roll indicators on GUI showing perfect and correct alignment.
    BUT
    Before continuing with the advanced mode calibration, if I simply observe the Yaw indicator on the GUI, it begins to precess to the right at a rate of about .75 (3/4) degrees per minute. If I wait long enough, it stops at almost 180 degrees out. (there are no magnets or metal objects near the IMU)

    Now, if I do not pause and continue with the advanced mode calibration, the remaining 5 positions appear to be successfully calibrated.
    But by the time I remove the IMU from my wood calibration block and reinstall it on the gimbal, the Yaw precession is almost 180 degrees out.

    NOTE: Receiver is not connected during the above calibration and the Beeworks B32 gimbal controller is powered by my 6S. If I connect the receiver prior to starting this entire process, then the precession becomes 3 degrees per minute.

    I am hardly an expert but could this be an IMU issue?
  • Garug
    The IMU has no absolute bearing. When it is powered of it looses the bearing and uses the start up position as new bearing. The ACC calibration does not affect bearing at all. Gyro calibration corrects the bearing drifting.
  • Stolman
    Understand. Thanks.

    The manufacturer has been quite supportive but it still will not calibrate Pitch regardless of power or PID settings. If I leave pitch out and calibrate Yaw and Roll. They are easy and work just fine.

    At this point I have spent at least 30 hours in the last 10 days just trying to get the pitch working. (My wife is livid). My projects for income parts of my work are significantly delayed and I have limited time to devote to this in the next few days. For now, I will just tape the pitch axis in place and do my video shoot. When I next come up for air I will start all over again.

    Thanks very much for your suggestions. If you think of anything else please let me know.
  • Zeb
    Just had a quick read of your problem, the 180 degree change sounds like the axis is 'looking' for the nearest pole on the motor and then moving to it, not sure why it would not stay aligned in the first place though, it should not need to do this if the gyro is calibrated for both IMU's?
  • Stolman
    Hi Zeb,
    According to the manufacturer, the Frame IMU should reduce the precession. Manufacturer advised that even with the Frame IMU there will be some precession. I understand this but it was the rate that concerned me. Thanks for your post.
  • Garug
    To get it working, perform only one point calibration. There is no need to perform 6 point calibration to get it working perfectly. I have never performed more than 1 point calibration, but when you have it working well, 6 point calibration can of course be performed to see if it improves anything. And as mentioned earlier, set it working first only with Camera IMU.

    You should test the Pitch motor, it could be broken.

    Also make absolute sure that you set first the IMU orientation correct, then perform the Motor configuration Auto to get the Inverted status right and finally set num.poles as per the actual number of poles on your motor. (The gimbal can fork surprisingly well even with wrong Inverted etc. but with plenty of problems)

    Also, make sure you have connected Pitch and Roll motor wires as per the 32bit board picture, not the labelling on your board. some boards have wrongly labeled Pitch and Roll. Though this is probably ok in your case.

    A video out side the gimbal and your settings would help analysing it further.

  • Stolman
    Thanks Garug, Do you have a suggested "test" for the pitch motor?

    I tried one point calibration with Camera only IMU. Results are the same.
    However I performed the motor configuration Auto again and this time none of the motors are inverted (previously pitch was inverted).

    I flew the drone with the pitch motor at zero power and taped the pitch at the desired angle. It worked just fine.

    However, as this my first time with a gimbal, I did not realize that I need to now set the FOLLOW parms because as I rotate the drone to the right or left, the Yaw remains solidly facing the original direction.

    If the motor "tests" as defective and I am unable to get it to work with a replacement, I will make an external video that includes the images of the drone, the gimbal and the GUI screen(s) so you can see exactly what is going on. OR, I can do a skype live video with you. Whatever is most convenient for you.

    Thanks.
  • Garug
    You could measure the motor phases, the resistance should be the same on each 3 phases (i.e. between any motor wire) also make sure that there is no connection between motor fires and motor metal parts. Also move the cables a bit when measuring, sometimes there is loose connections.

    Other way to test is to make sure that the motor can rotate freely (separate from frame if needed) and power the system and move IMU so that motor starts spinning. It should spin smoothly.

    "I tried one point calibration with Camera only IMU. Results are the same.
    However I performed the motor configuration Auto again and this time none of the motors are inverted (previously pitch was inverted)." Are you sure the IMU directions are correctly configured on GUI, the inverted status should not change.
  • Stolman
    Garug,

    Some results:

    1) Resistance is .4 ohms on each wire. Note: a continuity tester shows continuity between each of the three terminals on the PITCH motor (i.e., they are all 3 electrically connected). The same holds true for the ROLL motor.
    2) There are no shorts in the wiring to the PITCH motor.
    3) The PITCH motor does not rotate freely nor does it spin smoothly at low PID settings.
    4) The ROLL motor exhibits the same oscillations when connected to the PITCH connector on the controller board.
    5) The AUTO-TUNE software on GUI 2_42.b6 yields "motor is missing steps" when all three motors are selected for auto-tune.

    Looks like I need to order another PITCH motor, agree?
  • Zeb
    All motors should rotate freely with very slight 'notching' but essentially free.
  • Garug
    "1) Resistance is .4 ohms on each wire. Note: a continuity tester shows continuity between each of the three terminals on the PITCH motor (i.e., they are all 3 electrically connected). The same holds true for the ROLL motor."

    Is this just the wire resistance? The resistance between motor phases should be somewhere in between 8 to 20 ohms. i.e. measure from any of the 3 motor pins to any other of the motor pins and it should always give the about same resistance (maybe +-1 ohm deviation max.)
  • Stolman
    Garug, Sorry, I misunderstood your message. Yes, it is just the wire resistance. Between the motor pins the readings are different. Approx 24 ohms between both outer pins, 24 ohms between the middle and one side but 48 ohms between the middle and the other side.
  • Garug
    That indicates that there is a problem with the motor, they should all be the same. looks like you have one phase missing.

    If you have other similar (unbroken) motor it measures likely around 16 ohms between each pin.
  • Stolman
    Garug, You are so right! The ROLL motor is the same model (GBM4108-130T) and it indeed reflects about 16 ohms between each pin.

    Please accept my sincerest thanks for your time and advice. This challenge has been exacerbated by my lack of knowledge (ignorance) of exactly how the system components SHOULD be responding. Since the PITCH motor was faulty and since the GUI instructions say to start with the PITCH motor, I was left trying everything else.

    The GUI instructions are quite simple and easy to follow but I have read so much on the web about those having chronic problems that I figured my issue must be my own user error.

    Thanks again for helping restore my sanity!

  • Garug
    I am glad I could help. I have been in the same situation and spending days before figuring out my Yaw motor was missing phase.

    Now days if something does not work I usually quite soon verify the motors. There is to kind of problems motors can have, missing phase or the phases touching metallic parts of the motor. Both are bad and need to be fixed.

    Just yesterday I was tuning a gimbal that worked just perfect on Pitch and Roll, but not very well on Yaw, and it took me some time to figure it out that Pitch motor bearings had developed play and causing the problems, so the mechanical problems are not always where one would easily think them to be.

    Boards and IMUs can fail too, but not that often and when they do it is usually clear that they are broken but not always. I ones had a Broken IMU where only Accelerometers where broken, it took long time to figure that out. Often the problem is not perfectly performed basic setup cause problems, it just takes one motor inverted status or IMU orientation to be wrong to make it impossible to get it working perfect. Surprisingly sometimes it works quite good though. Most often it is simply a mechanical problem.

    If everything else is ok, setting Power levels right and PID tuning takes some time to learn, but is actually quite easy to do. If PID tuning does not seem to work, then the problem is likely elsewhere.

Add a Comment