EN CN
high frequency vibrations mainly on PITCH Motor
  • Hi!
    I try to get my 32bit board to work.
    Before I used the 8bit 3.5 board and the gimbal worked quite good.

    But now, with the 32bit version, I can't set up the PID and Power value good enough.
    I use three GB85-1 motors with 42 poles.

    Roll and Yaw works quite nice but the Pitch is impossible to set up.
    If I use P=3 I=0.1 D=2 Power=70 the gimbal already vibrates high frequency.
    And also tilt movement is not very smooth.

    I also tried to connect the motors and math it in the software in another way to test if the Pitch channel has an defect but the result was the same.

    Anyone an idea? Pleas help, thanks.
  • GB85-1 is strong and really demanding for the frame stiffness. Your PID is way off for GB85-1.

    What battery do you use? I would not suggest more than 4S. What camera/lens do you have? where is your IMU located?

    But in any case try Power 130 (for 4S and Canon 5D sized camera ) P 20, I 0.05, D 20 and start rising P keeping D as low as possible after done rise 'I'

    You could also try Gyro high sensitivity, works good for me. Here is my settings for 3xGB85-1, 5D http://www.rcgroups.com/forums/showpost.php?p=28031121&postcount=9436

    And alternative way of tuning http://www.rcgroups.com/forums/showpost.php?p=28045237&postcount=9447

    I am using 8 bit, but I think it is really the FW 2.4 that is relevant here.

  • Hi! Thanks for your answer.
    I use a 4s Lipo and a 600d for testing, later I also want to use bigger sized cameras.

    I had similar values with the 3.5 armbgc board (8bit), there it worked quite good.
    But the 32bit board is much more difficult to calibrate or as Garug mentioned the 2.4 firmware.
    All values over P=2 I=0.1 D=2 Power=60 cause high frequency vibrations mainly on PITCH Motor.

    Any ideas? Thanks
  • 32 BIT BOARDS THE PID seem to be different than 8 bit boards.
  • 'I' is wery much different on 2.4 FW. 32 bit has 2.4.

    Try setting 'I' to 0.01 and then raising P and D as needed. I would also recommend trying gyro high sensitivity.
  • I tried to do, but it changed nothing. gyro high sensitivity does not work, the hole gimbal vibrates.

    Do you think it is because of the hardware (Board, Gimbal, Motors) or software?

    It is strange that it worked with the 8-bit board.
  • many seem to have problems when changing from 2.3 to 2.4, the PID tuning really is different, and the 2.4 maybe is more demanding for frame stiffness.
  • The new blog entry http://www.basecamelectronics.com/company/blog/show/?newsid=69 helped me to config the gimbal.
    The best way to eliminate the high vibrations was to set the GYRO LPF to 2 smoothing the gyro.

    Now the balance is quite good, but I have one last problem with the pitch.
    The camera does not balance very smooth in Pitch Axis. It is jittering and with high focal length the stabilizing results are not good.

    You will see it on the Video:


    Someone an idea? Thanks!
  • and here a Testshot (50mm with 600d = 75mm):
  • If the camera is well balanced also in the pitch axis and the motor seems to be OK, then it's something about SimpleBGC.

    Set 0 all PID/power and then set the correct power for the pitch axis. I guess you know how it's done.

    Then start with the PIDs. Just set I to 0.01 at the beginning and raise P and D as it's explained in the guides until the stabilization is good enough.

    Gyro high sensitivity could be worth trying, but you must consider that P and D will be in fact doubled, so a retuning would be necessary.

    If nothing works, try high Gyro LPF, like 100 or something.

    Comment what happens! Best regards.
  • It changes nothing, same effect on pitch. Gyro LPF max is 15 I think.

    May it is due to the 42 poles GB85-1 motor in combination with the 32bit board?
  • Oh, sorry, I confused Gyro LPF with Gyro trust... ¬¬

    Anyway, Gyro LPF works only in some cases, and it's harder to get the right PIDs.

    I don't think the number of poles and board version has nothing to do, although I'm not 100% sure...

    I think it could be something in the settings, or maybe something physical...

    It's just a matter of patience and research, as always.

    I'm also struggling to get rid of vibrations on my roll axis...
  • Hi!
    I tried to change something in the construction on pitch.
    So I was able to raise up the P and D. But the pitch stabilisation isn't smooth jet.

    http://www7.pic-upload.de/16.04.14/3k99in27mqj4.jpg
  • You are on the right track. Really the more solid you get it the better it is.

    Also worth trying different IMU positions. On my gimbal on flash shoe was the best position until I made a gage that connected the camera both from bottom screw and flash shoe. After this modification at bottom of the camera was the best place for IMU. This is how it works now with the 8 bit board.

    http://youtu.be/hV7hpWdBzhA

    BTW this is Canon 5D with 24-105 4.0 F and I can use it with all the zoom range though it is balanced at 24 mm (and I have lens hood and ND filter connected.) At 105 mm it is really out of balance, but works good for stationary shots. When walking at 105 mm I get a bit similar vibrations as on your video, due it is so much out of balance (and at 105 mm all vibrations show much more clearly).

    That actually might be one reason for your problems, 50 mm on a small sensor (what camera is that) 24 mm on full sized sensor is much more forgiving. Also I have never much liked the construction of the canon 50 mm F 1.4 lens, on mine the lens barrel moves a lot when shaken.
  • I use a Canon 600D with 50mm (=75mm) for testing, to see the stabilisation better.
    I tried also to use this standard mount, but there the high vibrations occur very fast.
    (only the mount, with my gimbal)
    http://i00.i.aliimg.com/wsphoto/v2/903692102/-In-Stock-BG001-Famoushobby-DSLR-3-axis-Brushless-font-b-Gimbal-b-font-handle-font.jpg

    I hope the 8bit board will arrive in the next days to test it by comparison.
    I allready used it with the 8bit and it worked fine.
    I am shure, the 32bit is much more diffuclt to setup.


  • Today I also tried the 8bit board, but with the new firmware I have similar problems with the pitch axis.

    So I tried to change somethin on the construction. I added two rods. It's not the best solution but it works quit good.
    http://www7.pic-upload.de/17.04.14/hpmm6s564ow3.jpg
    But with these rods it is no so easy to mechanically balance the system, beause there is much more weight on one side.
    Does someone have an idea to fix it better or know a webshop with such components? I do not find any solution in online shops.
  • You could add balancing weight to roll axis (opposite the pitch motor), but really I would try to rebuild the camera attachment all together. Simplest thing to try would be adding the bottom tube. That would add rotational stiffness. Second to attach the camera directly to the bottom plate.

    If the gimbal manufacturer does not provide additional parts for this gimbal, it probably will be very difficult to find parts that fit directly, but hopefully you find someone locally that could manufacture those, or someone who has already done a modification to this gimbal.

    If you could find/fit something like on this gimbal http://www.foxtechfpv.com/eagle-eye-hhg5d-dslr-bl-handled-gimbalwith-controller-p-1278.html it would probably be good. Attaching the camera from bottom and top makes a very solid camera installation.
  • do not know what your trying to accomplish with all the structural mods but your wasting your time! please complete the PID tuning . your tilt is shaking not the gimbals fault it your fault! you are wrong this gimbals is fully upgrade with parts sold by me by a pro aluminum camera tray for $200+. eagle eye gimbal DYS gimbal are hobby not pro gimbals not in the same class as movixx gimbal aka famous hobby hiy fly gimbal 4mm thick carbon plate 25mm CF tube 2mm carbon fiber. If you got a fake model fiber glass copy you get what you pay for. I KNOW I SELL THEM ALL!! 32bit board are easier to set up one sensor mode!! meany people with this gimbal it was released first on the market by me proven design. to say DYS eagle eye is even close in preposterous.

    http://mymobilemms.com/OFFTHEGRIDWATER.CA/index.php?main_page=index&cPath=20_22&sort=0&page=2

  • Aopen3434, I do not follow what you are saying. My experience is that attaching camera from bottom and top is much better than just bottom and that was the only thing I was refering to with the link provided.

    I do not know either gimpal but for me, just looking the gimbal it is clear how to improve it for the camera theflamez is using. If you can better help him, please do so, just saying that "not the gimbal fault, it is your fault " is not helping anyone.
  • if you looking for an upgraded camera a they? your wrong I have put together meany of these rigs no modification is needed. you need to properly tune the gimbal making these mods are waste of time. thousands of people have gone threw this removing pitch vibration is a PID thing not structural issue.
    spend your time on PiDS.
  • As I see it, most gimbals can be PID tunend to work, maybe even satisfactory, but a well working gimbal starts from the constructio and well working condtruction is easy to PID tune.

    I also think manufactures providing good gimbas should be able to provide good starting point for PID tuning, i.e. Excambles for different cameras and lenses, not only PID, but all the GUI settings. ofcourse some PID tunig will always be needed, but good starting values will help most users a great deal.

    So I think a bit more support than "spend your time on PIDs" would be apropiate, don't you?
  • PID are camera , balance , specific never the same even with the same camera setup PID differ.
    auto tuning would be nice too. no substitution for hard work but if you do not read the documentation and understand what the valued do my be issue in tuning. sorry you really can not copy some one ealses settings outright
  • start with default settings!
  • Sorry about my posts today, I was bored and got irritated for no reason at all. I stand behind what I wrote, but I had no reason to write what I wrote, and there was no point. Sorry.
  • Hi!

    I think the temporary solution with the two rods is ok.
    But the PID settings for Pitch are not optimal.
    You will see it on the video:


    May someone has an idea?
    I use the GB85-1 with the 32-bit ArmBGC board.
    Thanks
  • I wish there were short cuts too. some times it not PID could be bad wiring bad cable too
  • lower your power setting
  • Set to 0 all PIDs for pitch.

    Set the power at the exact amount it needs to hold the camera in all positions, no more, no less.

    Then try rising a little bit P. Does it still shakes?

    Try to reset PID values looking at the realtime data tab and the reaction of the pitch when moving the gimbal or touching the pitch.

    You need to raise P and D as needed, no more, no less. I mean, they shouldn't be too low or too high, higher isn't always better.

    The I value could be just 0.01, or whatever works for you if P and D are good.
  • You guys are using lenses that are way too long. On a cropped sensor a 75, its like 100mm basically. mechanical ie jib, cran slider, steady cam or whatever are made to use wide angles for the most part. One, it reduces the appearance of movement or shake. two it has for ground. when you use a long lens you are pulling the background towards you. so there is no foreground for dynamic movement anymore. First thing is too stop with the long lenses. go 16mm or wider. 24mm ok on full frame but not cropped IMO.
  • Hi! Thanks for your answers. I will try as Octovisuals says.

    To OST_RC, I use a 50mm with cropfactor it is 75mm.
    And BeSteady is also able to use long lenses, like a 100mm. Here you see.


    I primary will use my 50mm for testing to balance it exactly.
  • Yeha they did a great job with that. it looks great but it is a test, not practical IMO. You can achieve most of the shots they did with just a tripod. Some shots are really cool but mostly mechanicals will call for a wide lens. I would set up with a wide lens then take it from there if you want to tune up. Whatever floats your boat. But chasing something or someone around with 100mm lens is not easy. If you want a versatile set up go 24mm and lower.
  • I'm also struggling with slightly roll vibrations/shakes/fast bouncing in extreme situations, and tried many different settings...

    I'm still wondering if it's something physical or that even more tuning is needed... I'm still testing with power and the PIDs. It's really a complex matter, very time-demanding on many gimbals.

    It's not easy to get the actual correct PID/power settings for every gimbal and every situation. There's just too many variables. But it's worth trying since one learn very much and can achieve the best stabilization (sort of...).
  • Hello I have a problem with Alexmos 32 bit two IMU 3 axis gimbal (second IMU below YAW). I'm using it with my DSLR Canon (1,2kg). When I walk the image is shaking but I don’t know on which axis (pitch, roll?). This is more visible when I run. There is my PID settings (Basecam 2.41b):
    ROLL:
    P: 22, I: 0.06, D: 28 - POWER = 170
    PITCH:
    P: 5, I: 0.07, D: 14 - POWER = 150
    YAW:
    P: 49, I: 0.05, D: 86 - POWER = 180
    FOLLOW MODE ON.
    (when you compare to old PID setting from earlier version than 2.41b you have to divide P and D by 4 and I by 2).
    Sample video:
    Can anyone please give me some help with this or any advice about the settings?
    Thanks,
    Lukas
  • Your Pitch values look really low compared to other axis? Are the axis well balanced?
  • Yes, they are.
  • Is the gimbal strong, is the gimbal flexing when walking?

    Anyway, totally impossible to say with the available information what parameters you should change and often the problems are mechanical.

    my guess would be: try to get Pitch values higher, lowering power if needed. Same for other axis. 'I' on all axis is low.
  • Thank you I will try that. I have cheap Chineese gimbal http://www.ebay.com/itm/181350900225?redirect=mobile with english controller from desirerc. I think its quite flexible. Can I do something to eliminate that flexibility?
  • Shortening the tubes as much as possible would likely help, it provides less pending momentum to the motor axis. The motor axis are probably the biggest problem. It is difficult to see from the images, but is there anything else than the motor axle supporting the Yaw and Roll axis?

    Camera installation is also often problem and on this gimbal it looks it might be flexing and also many cameras flex if installed from the bottom only. See if there is anything you can do to improve the camera installation.

    Where is your camera IMU installed, try different locations. On this gimbal it might work best if installed on camera, hotshot etc.

    There is also other parameters than PID and Power. Try setting on Advanced tab 'Gyro LPF' to 1 or 2 and retune. It should help getting PID parameters higher. (LPF, Low Pas Filter)
  • What else can support Yaw and Roll axis except for the motor?
    First IMU is installed on camera hotspot, second is below Yaw, perpendicularly to Roll axis.
  • There could be additional bearings outside the motor.
  • Thanks! I made a new gimbal, it works amazing for two weeks, but now I have problem. Every time YAW is affected by ROLL and vice versa. Here is example (follow mode off, secound IMU below YAW):


    Is it controller problem, or IMU (i have two IMU, 32 bit controller).
    Please, help!
  • Hello imago10
    That Gimbal is no good too much flex! so when the gimbal vibrates it looses its shape over time that why you have this issue. re balance the gimbal make sure the IMU are fat and straight re do ACC , gyro callab.
  • Hi AOPEN3434

    I solved problem! I accidentally hit calib.acc on GUI for Frame IMU. I repeat 6point calib. and everything works fine! Also I have bad YAW balance.

    Thanks!