Duet3D Logo Duet3D
    • Tags
    • Documentation
    • Order
    • Register
    • Login

    Input shaping cant work with short move

    Scheduled Pinned Locked Moved
    Firmware developers
    4
    8
    644
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • leoyuundefined
      leoyu
      last edited by

      Hello, i am trying the input shaping feature with RRF 3.4. I used to have a good result with klipper but failed with RRF 3.4. The cause i found is the move is very short (like 2mm with 10000 acc)will not shaping the command. Could anyone have idea to fix this issue?

      dc42undefined 1 Reply Last reply Reply Quote 0
      • dc42undefined
        dc42 administrators @leoyu
        last edited by dc42

        @leoyu input shaping cannot work if the relevant acceleration and/or deceleration times are less than N times the input shaping period (i.e. reciprocal of frequency), where N varies from about 2 to 4 depending on the type of shaping selected. This often requires acceleration to be reduced. The M593 L parameter specifies the minimum that RRF will reduce acceleration to. If it sold need to be reduced more than that, RRF will not apply input shaping to the move. I am not aware of Klipper having a similar parameter.

        That said, the move you describe is very likely an accelerate-decelerate move with no steady speed phase. It may be the case that Klipper applies IS to such moves when RRF will not. If you can't resolve this issue using the M593 L parameter, please provide details of the move parameters and your M593, M201, M203 and M566 parameters so that I can investigate further.

        Duet WiFi hardware designer and firmware engineer
        Please do not ask me for Duet support via PM or email, use the forum
        http://www.escher3d.com, https://miscsolutions.wordpress.com

        1 Reply Last reply Reply Quote 0
        • leoyuundefined
          leoyu
          last edited by

          I tried L parameter but not work still get high vibration
          This is my config:
          M203 X30000 Y30000 Z300 E3000:3000
          M201 X10000 Y10000 Z100 E10000:10000
          M203 P10000 T10000
          M566 X300 Y300 Z60 E300:300
          M593 P"mzv" L100
          the following is my test gcode:
          G28
          G1 Z10 F300
          G0 X200 Y200 F12000
          G0 X198 Y200 F12000
          G0 X200 Y200 F12000
          G0 X198 Y200 F12000

          with klipper it can move smoothly, but RRF3.4 it get very high vibration

          1 Reply Last reply Reply Quote 0
          • leoyuundefined
            leoyu
            last edited by

            After reading the klipper document, the klipper would reduce the top speed to make sure there must be enough steady phase.
            f288cb51-d1cb-423a-b402-0b8f76999f1f-image.png

            1 Reply Last reply Reply Quote 1
            • CCS86undefined
              CCS86
              last edited by

              Is this short move limitation also why IS isn't applied when the print head does very small (segmented) arcs, like this?

              e5d8a013-0639-4673-b982-bc88e6e05c6e-image.png

              This is a really big limitation of IS at the moment.

              botundefined leoyuundefined 2 Replies Last reply Reply Quote 1
              • botundefined
                bot @CCS86
                last edited by

                @ccs86 sadly this is why I've not yet tried IS. All my prints are highly segmented to the point that IS would seldom play a role.

                *not actually a robot

                1 Reply Last reply Reply Quote 1
                • leoyuundefined
                  leoyu @CCS86
                  last edited by

                  @ccs86 Yes, such short move with high acceleration would cause big vibration

                  1 Reply Last reply Reply Quote 1
                  • CCS86undefined
                    CCS86
                    last edited by

                    @bot said in Input shaping cant work with short move:

                    @ccs86 sadly this is why I've not yet tried IS. All my prints are highly segmented to the point that IS would seldom play a role.

                    Glad I'm not alone there. I really try to avoid sharp outside corners in most of my designs, so the current IS implementation really falls short for me.

                    I'm hopeful that there is a solution to the OP's problem and ours all in one.

                    1 Reply Last reply Reply Quote 0
                    • First post
                      Last post
                    Unless otherwise noted, all forum content is licensed under CC-BY-SA