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

    Tramming bed multiple times

    Scheduled Pinned Locked Moved
    Tuning and tweaking
    5
    19
    765
    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.
    • JamesMundefined
      JamesM @deckingman
      last edited by JamesM

      @deckingman
      My leadscrews are actually one in the middle left and two on the right side. I have the railcore II cast bed from 713maker on my BLV cube printers.

      I was always under the impression that I was supposed to do my bed.g with probing by the leadscrews but if that's not the case then I'll change that. I assumed that is how it was supposed to be set up since the Duet documentation lists this example:

      G28 ; home
      G30 P0 X20 Y190 Z-99999 ; probe point 0 near a leadscrew
      G30 P1 X180 Y190 Z-99999 ; probe point 1 near a leadscrew
      G30 P2 X100 Y10 Z-99999 S3 ; probe point 2 near a leadscrew and calibrate 3 motors
      

      Should I still probe by the leadscrews and then add additional locations such the left rear, left front and also the right center and then rear center and front center?

      Something like this

      M562 						;clear bed transform
      G28 						;home all
      M401 P0 					;pickup Euclid probe
      G30 P0 X310 Y32.5 Z-99999 	; right front leadscrew
      G30 P1 X310 Y179 Z-99999 	;right center
      G30 P2 X310 Y287.5 Z-99999 	; right rear leadscrew
      G30 P3 X40 Y290 Z-99999 	;left rear
      G30 P4 X40 Y159.5 Z-99999 	;left center leadscrew
      G30 P5 X40 Y30 Z-99999 		;left front
      G30 P6 X138 Y30 Z-9999 		;front middle
      G30 P7 X138 Y290 Z-99999 S3 ;rear center
      G1 X138 Y179 F9000 			;go to center of bed in advance of probe that point
      G30 						;probe center to re-establish Z
      M402 P0						;Return Euclid probe
      
      JamesMundefined o_lampeundefined 2 Replies Last reply Reply Quote 0
      • JamesMundefined
        JamesM @JamesM
        last edited by

        Gave me a pretty nice heightmap

        6e99dee8-ba6d-4a41-b025-dac87a7abc08-image.png

        deckingmanundefined 1 Reply Last reply Reply Quote 0
        • deckingmanundefined
          deckingman @JamesM
          last edited by

          @JamesM Did you watch my YT video? If not, I suggest you do. To summarise (again), if you probe close to the lead screws, you'll get a plain that is level to a plain described by the bottom fixing positions of the lead screws. However, with a cuboid shaped printer frame, the XY gantry will run on rails fixed to the top of the frame. If the XY rails aren't absolutely parallel with the lower fixing positions of the lead screws, then they won't be tram with your (levelled) bed. So regardless of the actual positions of the lead screws, after levelling the bed, it's a good idea to check the diametrically opposite positions.
          If you are absolutely sure that your frame is perfectly square (within say 0.1mm) and that the frame will stay that way, then you can omit that check.
          I made my comments based on your title which refers to "tramming" as opposed to "levelling". A build plate might be level, but if it isn't level with respect to the XY axes of the print head, then it won't be tram.

          Ian
          https://somei3deas.wordpress.com/
          https://www.youtube.com/@deckingman

          moth4017undefined JamesMundefined 2 Replies Last reply Reply Quote 1
          • moth4017undefined
            moth4017 @deckingman
            last edited by

            @deckingman hi great video, so it got me thinking if you have a good probe ( good repeatability ) you could apply the same measures technique as you used with the dial gauge to do an auto tram to see if the frame is twisted if your using mutiple motors , side note my core xy i build 6 years a go has three lead screws and 1 motor and a belt and i used the same technique to alter the height ..

            <

            1 Reply Last reply Reply Quote 0
            • o_lampeundefined
              o_lampe @JamesM
              last edited by

              @JamesM This new bed.g would trigger the bed tramming, since it has more probe points than leadscrews.
              But it doesn't repeat the probing if deviation is bigger than 0.x.
              You'd have to wrap the code between pickup euclid probe and return euclid probe in a while loop, as shown before.

              1 Reply Last reply Reply Quote 0
              • JamesMundefined
                JamesM @deckingman
                last edited by

                @deckingman
                Yes, I did watch your video and I think I had actually watched it in the past.

                I do seem to be getting better prints with the additional probing now but I do think I'll need to tweak my bed mesh manually since there are still a few spots when printing where I'll get the filament ripple on the first few layers from when it is too close.

                deckingmanundefined 1 Reply Last reply Reply Quote 0
                • deckingmanundefined
                  deckingman @JamesM
                  last edited by deckingman

                  @JamesM I don't use any form of flatness or levelling compensation, so I can't comment on your mesh results. But I have found in the past that a slight change to the overall Z offset will fix first layer rippling if it occurs, even if that ripple is only across part of the print. On my machine at least, it seems that if one bead of filament lifts slightly due to a spec of dust or for some other reason, then subsequent touching beads will also lift at the same spot. Generally reducing the offset by 0.05 or 0.1mm is enough to cure the problem for me.
                  Edit. To be clear, if I find that the bed is level but not tram with print head, I fix the problem with the frame, rather than using mesh compensation.

                  Ian
                  https://somei3deas.wordpress.com/
                  https://www.youtube.com/@deckingman

                  o_lampeundefined 1 Reply Last reply Reply Quote 0
                  • o_lampeundefined
                    o_lampe @deckingman
                    last edited by

                    @deckingman said in Tramming bed multiple times:

                    I fix the problem with the frame, rather than using mesh compensation.

                    Many people underestimate gantry sag. They spend $$$ for an extra flat bed, but it still doesn't work. (don't ask, how I know 😉 )

                    deckingmanundefined 1 Reply Last reply Reply Quote 0
                    • deckingmanundefined
                      deckingman @o_lampe
                      last edited by

                      @o_lampe said in Tramming bed multiple times:

                      ......... Many people underestimate gantry sag. They spend $$$ for an extra flat bed, but it still doesn't work. (don't ask, how I know 😉 )

                      That too can be a problem. Personally I use two parallel rails with my (heavy) hot end fixed between them. No sagging but it does take a up a bit of potential travel in the Y direction compared to a single rail.

                      Ian
                      https://somei3deas.wordpress.com/
                      https://www.youtube.com/@deckingman

                      JamesMundefined 1 Reply Last reply Reply Quote 0
                      • JamesMundefined
                        JamesM @deckingman
                        last edited by

                        @deckingman & @o_lampe
                        On my big ratrig v-core 3.1 I suspect it has some gantry sag..it's a big printer...500x500. I am re-doing a few things on that right now, replaced the gantry extrusion with a quality one from misumi, re-printed most of the printed parts, new y-linear rails since one that came with it was less than ideal.

                        I'm really curious to see how it is after all the replacements I'm doing on it. I even had two small custom extrusions cut from misumi to use as alignment spacers for the extrusions the y-rails mount to so they would be precisely lined up.

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