1. Got a question or need help troubleshooting? Post to the troubleshooting forum or Search the forums!

Solved Strange new fail. Can't find any like it. Z loses its position

Discussion in 'Troubleshooting' started by Ben R, Nov 22, 2014.

  1. Ben R

    Ben R Active Member

    Joined:
    Nov 21, 2014
    Messages:
    651
    Likes Received:
    89
    EDIT just got this ROBO 3d R1. Has been acting the same way no matter what I do ever since print 1

    I've checked everything I know how,
    Found that the e Step and Z steps were a bit out. Adjusted them and get perfect, PERFECT 20x20x10 boxes I mean down to about .01mm
    but... try to print anything else and bam .. it loses Z.

    So what happens is, it starts printing great... then somewhere between the 2nd layer and done, it either stops going up and on to the next layer (just the z axis) or it jumps a couple layers and continues (just the z axis). The extruder still backs moves in and out like it was in the right place.
    If it fails to move z upwards, it usually does it in the first couple mm of the print. If it fails by jumping a couple mm it usually happens a bit higher.

    Larger the print, the lower it happens. I am assuming a communication error or a bad motor.
    I've adjusted the extruder and z , jerk and acceleration down about 20%. Speed doesn't seem to matter it'll fail slow or fast.

    Pic below is of a print that was going fine until it just decided to move up 2 mm and continue printing. Pretty typical of what I'm getting.
     

    Attached Files:

    • fail.JPG
      fail.JPG
      File size:
      475.6 KB
      Views:
      35
    #1 Ben R, Nov 22, 2014
    Last edited by a moderator: Nov 22, 2014
  2. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,915
    Likes Received:
    7,338
    Did you check the Z drive stepper and gears? Make sure there are no loose components (even a pulley on a drive shaft can be loose).
     
  3. Ben R

    Ben R Active Member

    Joined:
    Nov 21, 2014
    Messages:
    651
    Likes Received:
    89
    I'll figure out how to get to the Z axis drive but from the platform up everything seems tip top.
    Reason I would doubt a loose belts/gears is,

    the bed leveling continues to perform as usual (that is, as the head traverses the platform it moves to keep the print level regardless of its position in the z axis being correct. The z axis works fine to keep the bed level and retract the head. Just sometimes, that's all it does.

    Even if I pause, move the z axis manually, it returns to the wrong location at the next level.

    I never hear about anyone else adjusting their steps... as a matter of fact, i've seen techs telling people to return their z step to 2560. Mine was printing 12 mm high for every 10mm. e step was off 1mm in ten as well.
     
  4. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,915
    Likes Received:
    7,338
    I had a broken coupler (that thing the threaded rods are bolted into). When it moved, sometimes the rod--didn't.
     
  5. Ben R

    Ben R Active Member

    Joined:
    Nov 21, 2014
    Messages:
    651
    Likes Received:
    89
    Good advice, but they are solidly in there. Ive tried to wiggle and force thme as hard as i can...and i'm a big guy. They are assembled well.

    I did just take off the stepper gears for the extruder and taped up the nut and replaced it. I'm open to teying anything... But it really seems like a communication problem or ... Maybe computer?
    Tuned up y belt.

    Runs great on small projects... Like it was built for calibration boxes

     
  6. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,915
    Likes Received:
    7,338
    Well, you can try slowing it down by say 50% and see if it improves (I know you said you already tried that).
    If not, then it is not likely a drive issue.
    Looking at the fail example you have (sorry, didn't think to look at it earlier) it is clear that you are good for a number of layers. If it is not the mechanical side of things that leaves electrical hardware or software or the cables in between. I'd normally rule out the last two because it seems to work initially, but a different cable is an easy thing to try. Other than that the connections on the ramps board and the firmware on the arduino are all that is left to check.

    Have you tried running the Z up a long way and back down again (using the manual control features of whichever software package--like Repetier--you are using)? That will let you make sure it is smoothly moving and functional (not some sort of mechanical bind or -- maybe -- a bad stepper).
     
  7. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,915
    Likes Received:
    7,338
    One thing you might check, it may or may not matter since you have an R1, but the nuts that run on the threaded rod.. make sure when at home/zero that they are both firmly seated in the crossbar assembly. If they were not you could potentially get one of them binding up some as it goes up. On the beta it is more obvious since you have to manually level the bed anyway and part of that is making sure the X axis is level across the bed (and in my case if one of the nuts is out of place then it will not be perfectly level across the bed).
     
  8. Ben R

    Ben R Active Member

    Joined:
    Nov 21, 2014
    Messages:
    651
    Likes Received:
    89
    I just tried another print at 50% speed. Took a while... but failed in what looks to be about the same place. Which makes you think... gotta be something wrong with a rod. But there isn't. Runs smoothly all the way up and down. I've tried 3 different firmware counting the v1 it came with and it hasn't changed it. Well, actually, with my current settings, it hasn't failed to move up. just gives up after 2-4 mm and runs too high.

    All my adjusting hasn't helped in anything other than making my small things look good .. If I print out anything nearing the size of the common
    "spocktopus" it fails. I can print the whistle and key chain that came with the mater control software.
    Is there a chance it could be... somehow, my computer? I wish I could plug it into my mac... but some baud issue prevents me.

    Its so strange that it just fails on larger pieces.

    The nuts are in there... "firmly? No.. they rest in there on gravity and there is a little x-y play. I'll gusset it with some paper towel or something...
     

    Attached Files:

  9. Ben R

    Ben R Active Member

    Joined:
    Nov 21, 2014
    Messages:
    651
    Likes Received:
    89
    thanks for the help though.. keep it coming if you have any ideas.

    Changed the cable... if that doesn't do anything.. I think I may be packing it up and exchanging it at amazon. Can't hurt to start anew. At least I would "know" it wasn't my printer.. it was something else. The trojan horse in my computer or something.
     
    #9 Ben R, Nov 22, 2014
    Last edited by a moderator: Nov 22, 2014
  10. Ben R

    Ben R Active Member

    Joined:
    Nov 21, 2014
    Messages:
    651
    Likes Received:
    89
    Swapped cables and computer...
    Just completed a very large and completely successful print.
    Only thing special about it... it did not auto-level before hand. Happy with that, I went on to print something else. This time it went through its auto leveling process. Failed pretty quickly in the same way. Raised up too high and started printing a birds nest.
    Looks like switching out the computer had little to nothing to do with it. I'm currently trying another model that has failed every time.

    When I say it didn't auto-level, I mean, it did not sample the 9 spots on the platform* AND* it didn't move the z axis to compensate for any previous idea of how to level against the platform during printing. Clicked print, it immediately started plotting. The Z axis leveling was completely static through the whole thing. Previous examinations of the leveling process show that the bed is less than 1 mm out across the whole platform. I think it went +0.6 to -0.2 at opposite corners. So, not a horrible amount depending on how big your model is.

    So far, my second model is printing fine, with the z axis not trying to level. Like I said, tried different firmware... old version of software, new version of matter control... no difference in failing. Seems like bed leveling is interfering somehow. Messing up the code. Or matter controller is doing it.
     
  11. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,915
    Likes Received:
    7,338
    There could be any number of settings in play here (or even a mechanical issue with the switches).
    It sort-of sounds like you have accomplished manual leveling of the bed :)
    If the auto-level is at fault then it is a likely a configuration setting with that and I will need to let someone else help (I have a beta, not an R1).
     
  12. Ben R

    Ben R Active Member

    Joined:
    Nov 21, 2014
    Messages:
    651
    Likes Received:
    89
    So, 12 hours and going of good printing without auto-leveling. Fails with auto-leveling. THe R1 cannot be manually leveled. Just has magnets no screws. My numbers come from the auto-leveling log.

    If I start a model, it auto levels. But if I cancel that, then start again, it won't try to auto level. The z axis strays still on x-y movements. The bed is level enough to work. But I "paid" for auto leveling and with no screws I need it. I could go in and program in adjustments I figure, but that;s not the right thing to do.
    So, the fault is probably in either the driver, the firmware, matter control (which I otherwise love) , or windows (which kind of reads "driver). I suppose if there's some sort of feedback from the stepper, that might be at fault too.. but I know nothing of stepper motors.

    exciting. I have a 1 on 1 scheduled for tomorrow. Maybe then.. maybe then I'l get an answer.

    I mean, I've got customer support to answer me... but hours or more between communication make it very slow going.
     
  13. Ben R

    Ben R Active Member

    Joined:
    Nov 21, 2014
    Messages:
    651
    Likes Received:
    89
    daft punktopus says hi.
     

    Attached Files:

    2 people like this.
  14. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,915
    Likes Received:
    7,338
    This is an order of magnitude (maybe two orders of magnitude) better than in the past.
     
  15. Ben R

    Ben R Active Member

    Joined:
    Nov 21, 2014
    Messages:
    651
    Likes Received:
    89
    So i've read. Can't expect much of a startup... but they're hitting their stride now. Costco contract and all..
     
  16. Ben R

    Ben R Active Member

    Joined:
    Nov 21, 2014
    Messages:
    651
    Likes Received:
    89
    Just got done with a 1 on 1 with brian. he set me up. Thing is better than new.
    I just have to use repetier... Only thing stopping me was... I'm new to this and didn't know how to set it up.
    Now that i've been troubleshooting for 5 days... I'm on it. Workin like a dream.
    Now why matter control can't send simple g code... that's a mystery.

    thanks
     
    2 people like this.
  17. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,915
    Likes Received:
    7,338
    Nah, just MC. :)
     
    2 people like this.
  18. Rwalker

    Rwalker Member

    Joined:
    Feb 13, 2015
    Messages:
    56
    Likes Received:
    3
    Hi @Ben R ,

    What ended up happening here?

    I am having a very similar issue with prints.

    Thanks,

    Rian
     
  19. Ben R

    Ben R Active Member

    Joined:
    Nov 21, 2014
    Messages:
    651
    Likes Received:
    89
    I returned the unit to Amazon and got a new one
     
  20. Ben R

    Ben R Active Member

    Joined:
    Nov 21, 2014
    Messages:
    651
    Likes Received:
    89
    But it was possibly a communication error in retrospect
     

Share This Page