Please help me getting the Gantry perpendicular

New Home Forum Mostly Printed CNC – MPCNC Troubleshooting – MPCNC Please help me getting the Gantry perpendicular

This topic contains 22 replies, has 7 voices, and was last updated by  Hans-Georg 4 months, 3 weeks ago.

Viewing 23 posts - 1 through 23 (of 23 total)
  • Author
    Posts
  • #97473

    Hans-Georg
    Participant

    Very important first: Please tell everybody that you have to adjust your 3d-printer, before you print any parts for the MPCNC!

    Al my mpcnc-parts are printed very clean with a layer high of only 0.06 mm

    Now my trouble: Whatever I do, the gantry does not get perpendicular.

    check if mpcnc gantry is perpendicular

     

    I checked my 3d-printer as follows and it prints perfecly perpendicular!

    checking if the 3d-printer prints perpendicular

    Also it prints dimensionally stable in both directions, the red corners should bee 200 mm apart.

    ckeck if the 3d-printer pints dimensionally stable

    What can I try?

    The 25 mm conduits run very smooth and positioning with 12000 mm/min without any loss of steps seems to be no problem.

     

    #97484

    MegaDad3D
    Participant

    Al my mpcnc-parts are printed very clean with a layer high of only 0.06 mm

    Are you sure that this is what you’re using for layer height? I have a .4mm extruder nozzle and typically print about .2mm layer height. That’s an order of magnitude different, and would result in 16.66 layers/mm and take FOREVER to print.

    Moving on, I’d look at the ‘Tension Bolts C’ to make sure that they’re not over-tightened.

    “Tension bolts “C” should have the locknuts and the bolt heads tightened so they are just touching the plastic. Equal light tension keeps the Z axis perpendicular to the work surface and locks the assembly together.”

    https://www.v1engineering.com/assembly/middle/

    #97487

    Ryan
    Keymaster

    You can not adjust anything without all six rails in place.

    #97488

    Ryan
    Keymaster

    Your printer test is only checking XY, ZX and ZY are more important and more often the problem.

    1 user thanked author for this post.
    #97495

    Jamie
    Participant

    You can not adjust anything without all six rails in place.

    This got me thinking, do you think it would be possible or reasonable to shim the gantry rails at the junction with side rails where they are clamped by the roller and motor mount? There are fewer moving (flexing) parts and maybe there could be a procedure thats more deterministic.

    I may give this a shot. My gantry is slightly out of square and I force it into square when I energize the steppers.

    #97496

    Ryan
    Keymaster

    The rollers should already clamp them square, or more square. That is why the center should be adjusted before the rollers come into play.

    #97498

    Jamie
    Participant

    Ok you are talking about the missing z axis then. That makes sense because they contribute to the overall flexing, yes?

    #97501

    Hans-Georg
    Participant

    Thank you MegaDad3D,

    yes, it takes more than a while to print with that layer high, but it makes a visible difference in accuracy.

    coparison between 0.2 mm and 0.06 mm layer high

    Here is a link to a bigger view: http://metallmodellbau.de/pictures/mpcnc/IMG_1910.JPG

    But the big parts of the gantry, the XY_F_Burly take more than 43 hours each, so you need a lot of patience.

    Of course I followed the instructions for the middle assembly and tried out all possible combinations of tightening all screws A, B and C, but thank for your advice.

     

    Thank you Ryan,

    as mentioned before, I followed the instructions, but you are right, I didn’t check the XZ and YZ – orientation of my 3d-printer.

    After I checked this, I’ll be back.

    #97502

    Hans-Georg
    Participant

    OK, forgott to click submit before.

    Ryan, you were absolutly right, the Z-axis of my printer is not (yet) perpendicular.

    I gues this should be straight up.

    How about, I expand the hole for the long screw to the right on both parts?

    To print those parts again is not funny.

    #97503

    Ryan
    Keymaster

    Try putting in the Z axis, your rails will flex the correct way to fix your issue.

    #97507

    Hans-Georg
    Participant

    The Z-axis was in place, when measuring. Not on the picture above, but before.

    #97508

    Ryan
    Keymaster

    From the picture above that is generally what it looks like without a Z (looks like it could work to me), but if you checked otherwise, reprint is my suggestion. In terms of strength thicker layers are better so you might want to crank it up for these parts the second time around, after you fix your printer.

    #98143

    Hans-Georg
    Participant

    My printer was fixed and prints perfect perpendicular now in YZ and has 0.5 mm difference in XZ at a high of 100 mm.

    After I printed all four parts of the gantry again, this time with 0.2 mm layer high, the difference is still 4 mm in a distance off 400 mm. Tightening the long screw B has no positive effect, also the other screws.

     

    As you can imagine, I am very unhappy with the mpcnc. What can I do?

    #98150

    Ryan
    Keymaster

    I am very unhappy with the mpcnc.

    1-Have you used it yet? I can see your frustration with it not being absolutely perfect but this is in the FAQ’s some people’s builds need to be broken in for some reason.

    2-Can you use it for a while for non mission critical learning exercises as recommended in the FAQ’s for this issue? Imagine if you got it perfect and then drew a few crowns broke a few endmills to find your build settled and is now not perfect. Try using it, draw a crown, mill some wood. then double check you angles.

    3-Did you buy everything else from me?

    4-Are you using Dual endstops?

    5-How far off is it when fully assembled?

    6-Are any of the bolts touching things they should not be, recently we had the 4 1.25″ bolts hitting the plastic near the 2.5″ bolts (you can either file off the 1mm,  or turn them around). one built had the 1.25″ bolts hitting near the very center. I believe this is from over tightening the 1.25″ bolts.

    7-Loosening the rest of the tension bolts (keeping the 2.5″ just touching) what happens when you tighten the 5″? In regards to gantry rail angle there are only 3 adjustments to make they all should be changed in sets 5″, 1.5″, 2.5″. As an example just to be clear, never tighten just one 2.5″ all should be equal.

    8-0.2mm layers are not as strong as thicker layers, this has nothing to do with your current dilemma.

    #98151

    Ryan
    Keymaster

    Tightening the long screw B has no positive effect, also the other screws.

    If nothing moves when messing with these bolts you have bigger issues. These will and do have an effect.

    #98170

    Hans-Georg
    Participant

    Hi Ryan,

    thanks for your prompt reply.

    1 Initially I followed your advice in the instructions, not to care to much about perfect adjustments and milled some things successfully. Than I milled an L 250 mm x 400 mm to see, if the mpcnc mills perpendicular, the difference was around 3 mm on the long side. I am milling and turning steel a lot on real machines (Arboga and Weiler LZ280) and used to work slightly precise. The MPCNC will be used for RC-Planeparts and at least, it should mill perpendicular and not only diamonds.

    2 I drawed your crown already and milled sucessfully some easter bunnys without breaking a 2 mm 2 flute endmil, but I am expirienced enough to know, that the problem I have, will not disappear by milling around. I guess the parts for the F-version I use for the 25 mm pipes are not OK. The fault seems to be systematically.

    3 I bought nothing from you. The screws and the pipes are straight and on the gantry are no other parts.

    4 Until now, I dont use any endstops, but how should they affect the angle of the gantry itself?

    5 As mentioned before, 3 mm at a length of 400 mm was my best milling result.

    6 No!

    7 I followed your instructions, but the angle error stays, whatever I do!

    8 The last time you proposed to use 0.2 mm layer high and I did for the new prints. To me the prints with 0.06 mm layer hight seem to be far more compact and stronger than thicker ones, but I may be wrong.

    Is it possible, that the F-version has a systematically fault and nobody remarked this until now?

     

     

    #98171

    Ryan
    Keymaster

    real machines

    Ouch…

    Joking.

     

    1 Initially I followed your advice in the instructions

    Have you done it with the new parts…reason to follow.

     

    I guess the parts for the F-version I use for the 25 mm pipes are not OK. The fault seems to be systematically.

    I assure you there are a large number of successful builds out there.

     

    3 I bought nothing from you. The screws and the pipes are straight and on the gantry are no other parts.

    Maybe your bolts are tight, 8mm? As previously suggested a break in period could be required. Your prints could have tighter than usual ID’s. You have none of my proven parts so I am trying to help you understand why I suggest things. I build a machine that works everywhere in the world but with every part sourced/made by you I can only do so much.

     

    4 Until now, I dont use any endstops, but how should they affect the angle of the gantry itself?

    See the dual end stops, auto squaring page.

     

    6 No!

    Appologies if that was somehow offensive, it is the largest issue with this problem since the Burly update.

     

    Is it possible, that the F-version has a systematically fault and nobody remarked this until now?

    I think you are underestimating the number of builds. Is it possible, of course; likely, no.

     

    Well, I am not sure if you are willing to keep trying or not, perhaps you want to have a closer look at your Z axis. Twisted, parallel rails? Rail to rail OD measured in between the clamps vs the bottom?

    #98226

    Hans-Georg
    Participant

    ‘Well, I am not sure if you are willing to keep trying or not…’ Good joke!

    What else can I do? Give up, milling diamonds? Never!

    But the good news is: It’s done, the gantry is now perpendicular in X and Y.

    After I dis- and reassambled it without setting in the Z-rails, it looked for the first time a little bit more than perpendicular. More means the angle was slightly to smal insted of way to big.

    Why? Nobody knows, but this was a new startpoint for me. After I put in single Z-rails, the angle was OK. So I had to work on the tool mount and the Z-yokes.

    Some sanding paper glued on a 25 mm tube in the lathe helped me to decrease the distance between the the Z-rails at the end.

    What a horror trip!

    Now I hope that the Z-axis is perpendicular to X and Y.

    #98237

    Jan Walter
    Participant

    Hey I have a similar problem with my mpcnc. I don’t get the gantry perpendicular. 🙁 I already reassembled it. If I cut squares, the are not rectangular. It is not drastic on smaller cuts. But on bigger cuts it gets worse. Is there maybe a way to tune everything a little via software?

    #98238

    Ryan
    Keymaster

    Jan it is best to start a new thread with details of your build so we can help, each issue at this point, tends to be different.

    1 user thanked author for this post.
    #98441

    GTKplusplus
    Participant

    Hans, just a small primer on 3d printing:

    Bigger layer heights are much stronger. The weak point in any print are the layer lines, the less of them you have, the stronger your print will be. Also, if your printer isn’t perfectly tuned, and doesn’t have a reduction on the extruder (like the titan or bondtech bmg do) the precision of your extrusion is lower, and it can have difficulties handling small layer heights.

     

    Also, you should consider the so called magic numbers when using really small layers: you’d rather avoid positioning on a microstep position, since they are less precise then a full step (on some drivers, like the drv8825 the difference is quite dramatic, microsteps 17 has an almost half step jump, and the remaining steps are much smaller). A magic number is pretty much the minimum movement you can have with your machine if moving at full steps. With a normal motor and 2mm pitch leadscrew, it is 0.01mm and allows for pretty much all layer heights, but with an 8mm pitch you have far more issues. Technically, you should print at layers multiples of 0.08mm. Actually, though, you could use halfsteps for the calculation without much of an issue, they are really accurate too. So if you had a 8mm pitch screw, you should avoid layers with heights not multiple of 0.04.

     

    One last thing, 0.5mm over 100 of error on Z is quite dramatic, and shouldn’t happen on a well tuned printer. So there is that too that may cause an error.

    1 user thanked author for this post.
    #98711

    kd2018
    Participant

    6-Are any of the bolts touching things they should not be, recently we had the 4 1.25″ bolts hitting the plastic near the 2.5″ bolts (you can either file off the 1mm, or turn them around). one built had the 1.25″ bolts hitting near the very center. I believe this is from over tightening the 1.25″ bolts.

    I think this was me? For the record I promise they weren’t overtightened. Everything was tightened to spec using a calibrated torque screw driver I use for rifle scopes, and then actually backed off a bit. Tension bolts were kept finger tight until adjusting.

    I had the same problem as op. Dual endstop firmware is what squares my center up. (Also, my outer rails are perfectly squared with a trusted  precision machined fab square)

    One commonality is OP and I both have layer heights of 0.2mm instead of 0.3mm…

    #98783

    Hans-Georg
    Participant

    Hello GTKplusplus,

    thank you for your hints, I shall consider all you wrote from now on.

Viewing 23 posts - 1 through 23 (of 23 total)

You must be logged in to reply to this topic.