"Loose" yaw on follow mode
  • Hiya!

    I've got a Nebula 4000 Lite handheld stabiliser (using the 8-bit SimpleBGC), and whilst it's fantastic for the most part, its follow on the yaw axis seems kinda loose. I often have to rotate what feels like 20 or more degrees to persuade the camera to turn, and it doesn't always return to exactly the center point.

    I've tried altering the obvious settings - deadband and speed on the Follow tab - and done my best to tune the gimbal following the SimpleBGC guide, but that doesn't seem to have helped much.

    Any suggestions much appreciated!
  • Hi!
    Which PID values for yaw? Deadband/Expo/Speed in follow mode?
  • I am having exactly the same problem since I changed my 8-bit controller to a 32-bit one (with a second IMU mounted above the yaw axis on my handheld gimbal with a Panasonic GH4).

    Everything works really smoothly in "lock on" mode but the yaw axis is vague and weak when in follow mode. It also moves in a jerky fashion, like it is jumping from pole to pole and, as you say cairmen, it doesn't always return to the centre.

    P:65, I:0.02, D:44 deadband 5º and all other follow settings at default. Raising the PID values just make it lose control of the yaw axis completely and go into oscillation, followed by the roll axis drifting.

    When I first changed the boards (2 days ago), I had a lot of trouble with PID settings in general, even though my rig is nicely balanced and was working great with the 8-bit board. The auto setup came up with PID settings that were so high they were nearly off the scale with the power set at 100 for all axes (a total guess, I don't know where to start with that). The gimbal was a shaking mess after the setup process finished. My manual settings are much lower but I can't even get in the ball park with following the yaw axis.

    I would also greatly appreciate any suggestions.

  • If the yaw motor is weak, you need to move really slowly. Also you could activate acceleration limiter and try some below 100 values on it.
  • My settings:
    Yaw PID: 25 0.02 20 (Power 120, num.poles 23)

    Deadband 3 degrees, expo 80, yaw speed 25

    FYI, the gimbal I'm using is the FilmPower Nebula 4000 - http://www.camotionllc.com/nebula-4000-a-111.html
  • And acceleration limiter setting?
  • No acceleration limitation set.
  • Hi Garag

    When I say the motor is weak, I mean that the response is weak. All of the motors are G5208-200 and the yaw response was strong when I had the 8-bit board. I can't remember how I had it set up and I don't have that board to read with the GUI.

    Can you recommend a good starting point for these motors (Hi-fly frame and a camera lens combination of about 1kg)? I can't find any detailed information on how to choose motor power settings. As I said, when I had them all at 100 the PID auto-tune wan't P and D values in the 200s and the result was a mess. I have double checked everything like pole number settings and motor directions.
  • If the motor feels weak, just put in as much power as needed (starting from 100 and increasing 30 at the time or so), but monitor frequently with hand that it does not get hot to touch. Hand warm is ok. If it gets hot you need to reduce power.

    After setting power it need to be PID tuned. Try auto tuning one axis at the time, hold the gimbal in hand relatively steady. Use good stability settings.

    Note: motor inverted detection may not work correctly if power is too weak. redo it when good power.
  • @cairmen, what about trying it, it probably would solve your problem.
  • I've given it a test - thanks! It does indeed significantly improve the problem.

    Only question: it also seems to mean that the camera now turns much more slowly, which makes tracking moving subjects tricky. I'm assuming that if I push the degrees/sec up that'll help with that problem - at what point is the acceleration limitation going to stop being useful to solve the yaw drift? You mentioned 100 deg/sec above - is that the practical maximum?

    (Sorry for the noob questions!)
  • It is a compromise between skipping motor steps and reaction speed. Put the acceleration as high as the motor can follow, when it starts skipping steps, lower it a bit. Also try different follow speeds, using lower follow speed and bigger expo helps too.

    To get higher speed than that, you need to raise power, or get stronger motor. (but do not raise power so much motor gets hot.)
  • Thanks for your help with this Garug. I have made a bit of an improvement with mine. I can now get it to follow Yaw smoothly but it is very slow (maybe 5-6 seconds to turn 90 degrees). If I try to move it too quickly I lose control of both the yaw and the roll axis. I spent two hours trying different yaw PID settings but can't get it any better than this. I have been very careful with the positioning of both IMUs and done 6 point calibrations on them. These are all things I did before when I set up the 8-bit board. That time though I got everything working very well. One thing that is happening though is that the yaw is more willing to move clockwise. When it breaks synchronisation it always goes clockwise. I don't know what that means about my calibration.

    cairmen, I hope you didn't mind me posting on your thread. I thought we might be looking for the same answers and I didn't want to double post.
  • Hi
    I have 2 different gimbals with GB-85 and Iflight 8108 on "Pan" I was testing also iflight 8017 and the smoothness in movement is also a problem. Tuning PIDs doesn't help increasing power also. I checked also acceleration limiter and don't see any improvement. When I am testing the gimbal holding it in the hand I just feel that this motors are not so smooth ( I feel skipping magnets from time to time). This bigger motor have better power but they are not so smooth like iflight 5208-200t. Do you have any solution for smoother move of big motors ?
  • On follow yaw they should work quite smooth. when panning with joystick I can feel some jitter at certain panning speeds. At slow movements it is very smooth (this gimbal is with GB90 and GB85)

    But at some 20 speed I get kind of small jitter, at fast speed it works smoothly again. I think this is because of the gimbal dynamics and many poles (42).

    It is worth checking though that your motors are ok. I had one motor missing one phase, it was working, but not well. Also I have had motor inverted status set wrong The gimbal can work, but not well. always set the inverted status with the Auto.

    This gimbal is with GB85. It is a year old video, so not the newest FW today ;-) but works good. It has the same at certain speed yaw jitter. Not much jitter at all on pitch, though it has the same GB85 on it.

    This is with my newest gimbal with GB54-2 motors (14 poles) it is indeed very smooth, but its strongest feature is the light weight.