Z-axis height way off

Discussion in 'Calibration, Help, and Troubleshooting' started by richgain, Apr 21, 2017.

  1. richgain

    richgain Well-Known Member

    Joined:
    Dec 18, 2015
    Messages:
    73
    Likes Received:
    35
    Are you still using Marlin in your printers, @Alex9779 ? Which version?
     
  2. Alex9779

    Alex9779 Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    2,411
    Likes Received:
    735
    @richgain you might have missed it but I changed to a Duet a while ago... I was annoyed by the maintaining policy and just the thing you are talking about was one of those things: actually they told me they just fixed a bug and it should have always been that the head moves when switching tools. I was able to get that extended command to be able to keep the heads in position when switching but I wanted the "old" behaviour because IMHO it is always better to do no move on a tool change and you can choose to if you want, not the way they implemented it...
     
  3. richgain

    richgain Well-Known Member

    Joined:
    Dec 18, 2015
    Messages:
    73
    Likes Received:
    35
    Finally got Marlin version 1.1.1 running and tried a 'Both Extruders' print of two 20 mm cubes,
    and ...

    it didn't fix the problem!
    X = 20.04
    Y = 20.03
    Z = 22.65

    Damn!
     
  4. Greg Holloway

    Greg Holloway Administrator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    1,022
    Likes Received:
    631
    How many layers are there in the print? is there any indication of the error per layer?
     
  5. richgain

    richgain Well-Known Member

    Joined:
    Dec 18, 2015
    Messages:
    73
    Likes Received:
    35
    S3D calls it 200 layers of 0.2 mm with the final layer being at 20.00 mm. This seems to be because of double counting the layers as the tool change is labelled as a new layer even though layer 2 is the same height as layer 1.
    Code:
    M117 Layer 2, Z=0.2
    M104 S190 T0
    T1
    ; tool H0.200 W0.480
    ; inner perimeter
    On inspection the layers all seem to be very even, with no obvious discrepancy anywhere.
     
  6. Stefan

    Stefan Well-Known Member

    Joined:
    Feb 17, 2016
    Messages:
    324
    Likes Received:
    43
    Any chance you can share your 1.1 config, which branch exactly you use and your 20 mm test cubes g-code files?
    I would like to try and print to see if this happens to me too.
    Off to install the 2nd hot end...
     
  7. richgain

    richgain Well-Known Member

    Joined:
    Dec 18, 2015
    Messages:
    73
    Likes Received:
    35
    Hi Stefan
    I posted the cube STL and code files in post #7 above.
    The Marlin version I currently have installed is the 1.1.1 which I downloaded 10 days ago from https://github.com/MarlinFirmware/Marlin
    Richard
     
  8. Stefan

    Stefan Well-Known Member

    Joined:
    Feb 17, 2016
    Messages:
    324
    Likes Received:
    43
    OK I'm almost there... just have to align the nozzles... thats the part I hate and what stopped me from dual until now, now I have a reason to do it :)
    Any way can you post your config.h an config_adv.h? so I have your setup, and save some time since I'm running my modded rc7 or something.
     
  9. TimV

    TimV Well-Known Member

    Joined:
    Mar 13, 2016
    Messages:
    151
    Likes Received:
    26
    mmm strange, this is exactly what I am seeing with my E1.
    115% infill is the only way to get my extrusion spot on, meanwhile, when running the 'calibrate 10cm' it's spot on 10mm, without the 115% multiplier.
    Perhaps share your 20mm cube, I'll print it on mine and come back with the results. (To ensure we start from the same base)
    Also, I checked my nozzle size, with Carburator reamers, and they are spot on 0.4mm.
    I also swapped the stepperdriver(with a brand new one) and it's does exactly the same thing.
     
  10. richgain

    richgain Well-Known Member

    Joined:
    Dec 18, 2015
    Messages:
    73
    Likes Received:
    35
    @TimV - the critical point here Tim is that when printing from a single nozzle all the dimensions of the test print are perfect; but as soon as I try and print anything using both nozzles the Z dimension increases by 20% from both nozzles. The test cube was shared in post #7 above.

    @Stefan - here you go. There were three changes to files that I found in my Dropbox; I can't tell you what they were since I'm away on holiday at the moment and working on my iPad.

    https://www.dropbox.com/s/p7uv8jqi8mard1w/Marlin_main.cpp?dl=0
    https://www.dropbox.com/s/j086jyedrfwim5a/Configuration.h?dl=0
    https://www.dropbox.com/s/qn5yogih41go8ku/Configuration_adv.h?dl=0
     
  11. Spoon Unit

    Spoon Unit Well-Known Member

    Joined:
    Sep 6, 2015
    Messages:
    1,560
    Likes Received:
    483
    I went through the problem at least twice now that, with z-hop turned on, I could, for seemingly no reason, end up with z raising far more than it should during a single layer print. Probably cause, z-hop up was not equal to z-hop down. Your problem sounds like the same sort of result. I couldn't imagine where to look, but Gregs advice at the time was to unplug and re-plug everything - which actually worked. The problem did return, but then went away again of it's own accord.

    My other Z issue was something more recent, where I noticed that moving up, was not moving as far as moving down even when given the same command. This was when using the LCD to manual drive the motion system. There was a strange whining sound when moving up. I fixed that one by upping the voltage to Z, but it left me wondering whether my X, Y, and Z have the right voltage (at some point I moved to TMC drivers, so this is also in the mix for me). In any case, this all left me wondering about the whole quality of motion control from the RUMBA, but maybe it's something I could fix by just adding more voltage.

    The advice for voltage seems to be - enough to do the job and no more. But how can I really asses truly whether enough is enough?

    I'm guessing Rich that your problem would be "solved" by disabling all z-hops. I realize this will also cause a mess, but something that might be possible to workaround with either an outer shell or prime column.

    At some point, I ended up writing a single GCODE script that was just dozens of up/down motions to see if the head stayed in the same spot of moved. However, that always seemed to result in the head moving correctly (i.e. staying at the same height). So maybe the issue only occurs when you couple certain kinds of movement together.

    Sorry - I'm rambling.
     
  12. Paul Winter

    Paul Winter Well-Known Member

    Joined:
    Jun 11, 2016
    Messages:
    50
    Likes Received:
    9
    I cannot remember if you mentioned this, but if you implement a prime pillar or tower when doing a dual extruder print, does it reflect the same changes that the model itself is experiencing?
    I only ask to find out further indicators into what might be happening.
     
  13. Stefan

    Stefan Well-Known Member

    Joined:
    Feb 17, 2016
    Messages:
    324
    Likes Received:
    43
    yeah I managed to glogged the one heat break with petg.... so not dual tests until I cleaned it some how or need a new one :(
     
  14. richgain

    richgain Well-Known Member

    Joined:
    Dec 18, 2015
    Messages:
    73
    Likes Received:
    35
    Hi @Spoon Unit
    Thanks for taking the time to consider my issues and share your experiences.
    I haven't tried that yet - I might give it a go.

    I took the Z voltage up to 0.8V which didn't seem to make any difference. It's now back at 0.6V but with seemingly identical results.

    I did try printing without Z-hops but still had the problem of increased Z height.
    I need to get back to doing some printing now to start clearing a backlog of orders, but hopefully will find the time to return to tinkering soon.
     
  15. richgain

    richgain Well-Known Member

    Joined:
    Dec 18, 2015
    Messages:
    73
    Likes Received:
    35
    Yes, @Paul Winter.
    The Ooze Shield was exactly the same (incorrect) height as the printed parts.
     
  16. Kick2box

    Kick2box Well-Known Member

    Joined:
    Mar 15, 2016
    Messages:
    198
    Likes Received:
    27
  17. richgain

    richgain Well-Known Member

    Joined:
    Dec 18, 2015
    Messages:
    73
    Likes Received:
    35
    Hi @Kick2box
    Thanks for taking the time to think about this issue and comment.
    Yes, I did see it when I was researching the problem but I discounted it for a couple of reasons.
    1. I think that a Switching extruder is a different kind of hardware that needs specific code to tilt the two heads from side to side.
    2. The original problem (https://github.com/MarlinFirmware/Marlin/issues/5966) appears to have been introduced after the date of the version that I am using.
    Even so, it is a very good match for the kind of problem that I am still having, so I will try and look into it a bit further.
    Eventually though, I may be forced into baling out and installing the Duet like @Alex9779 did.
     
  18. gonzalo

    gonzalo Well-Known Member

    Joined:
    Mar 17, 2016
    Messages:
    255
    Likes Received:
    146
    Did you ever solve this?
    I just started having the same problem!
     
  19. richgain

    richgain Well-Known Member

    Joined:
    Dec 18, 2015
    Messages:
    73
    Likes Received:
    35

Share This Page