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

Solved *Read pg 8 if you use a Geckotek bed!* Changed power supply, now y axis waves

Discussion in 'Troubleshooting' started by Rat_Patrol, Jun 10, 2017.

Thread Status:
Not open for further replies.
  1. OutsourcedGuru

    OutsourcedGuru Active Member

    Joined:
    Jun 3, 2017
    Messages:
    752
    Likes Received:
    141
    I'd agree: if the amplitude of the wave is consistent then it sounds like something's loose. The amount of deviation is directly related to the amount of slop in something, you'd think.

    The varying "frequency" of the wave could be explained if you're changing the travel speed of the extruder. If not, then something unexpected like ambient temperature/humidity might be the cause.

    Don't forget that it's not just the deep-Y that's making this worse. It looks like the problem is triggered with some mid-Z, right? Is that a common distance from bed or is that variable?
     
  2. Rat_Patrol

    Rat_Patrol Member

    Joined:
    Dec 14, 2016
    Messages:
    178
    Likes Received:
    24
    It starts at probably zero, but too little to detect. Visually, I can see it starting at about 20mm of height. The wobble gets worse and worse the higher Z travels. Frequency stays the same on any given model, but amplitude increases with the Z height.
     
  3. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    Try moving the extruder carriage/platform (with the power off) along the Y (push/pull it) at different heights.
    I know you checked the solid rods, but the mounts that hold them to the bed plate are small and easily cracked.
    (There are alternatives on thingiverse)

    If one of those was missing a screw or got a crack in it I could see that causing Y motion at varying heights.
     
  4. Ryan TeGantvoort

    Ryan TeGantvoort Active Member

    Joined:
    Mar 11, 2016
    Messages:
    343
    Likes Received:
    172
    Shot in the dark, but have you adjusted your acceleration and jerk settings in the firmware?
     
  5. OutsourcedGuru

    OutsourcedGuru Active Member

    Joined:
    Jun 3, 2017
    Messages:
    752
    Likes Received:
    141
    If the amplitude increases with Z then this isn't a constant amplitude, then. This might not be looseness/slop, at least not the standard variety.

    But it could be getting sloppier the hotter it gets. But then the part printed in the front of the bed you'd think would still misbehave at the same rate. You have replaced the fan and it blows like crazy, you indicated.

    Just stating what I've heard:
    * problem (Y-deviation) increases directly with Y, albeit with a periodicity
    * problem (Y-deviation) increases directly with Z, albeit with a periodicity.. and yet
    * Y=FrontOfBed with increasing Z seems to be fine
    * "frequency" of wave is unique to the print job but consistent throughout that job
    * printing a single tower doesn't have the problem, printing multiple towers does
     
  6. Rat_Patrol

    Rat_Patrol Member

    Joined:
    Dec 14, 2016
    Messages:
    178
    Likes Received:
    24
    Yup, and I am printing 3 towers right next to each other toward the very rear of the build plate, where the last print was at its worst. Its still going, but the print is actually looking very much better.

    In that case, the variable left that triggers the issue is long moves from Y0 to Y(extreme end), but NOT vise versa??
     
  7. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    OK, that ... look closely at the belt and gears that drives the bed (Y axis) and watch as you slide it back and forth full-stop. Is the belt centered and stays centered on the capstans in the center as you move it or does the belt try to ride up/down?
     
  8. Rat_Patrol

    Rat_Patrol Member

    Joined:
    Dec 14, 2016
    Messages:
    178
    Likes Received:
    24
    So after finishing the test print, here are the results:

    The model in the back of the Y plate was on top.

    You can see a line where the entire model shifted on the bottom model, that is where I paused the print and started pushing and pulling on stuff (everything was solid).
     

    Attached Files:

  9. OutsourcedGuru

    OutsourcedGuru Active Member

    Joined:
    Jun 3, 2017
    Messages:
    752
    Likes Received:
    141
  10. Rat_Patrol

    Rat_Patrol Member

    Joined:
    Dec 14, 2016
    Messages:
    178
    Likes Received:
    24
    I had at one point turned my jerk down to 5, acceleration down to 200 or 300 IIRC, made no difference. I have done a LOT since then, may be worth trying again <walks away, comes back> just turned jerk/accel/speed down for Y to about 50% of what it was. We will see.
     
  11. Ryan TeGantvoort

    Ryan TeGantvoort Active Member

    Joined:
    Mar 11, 2016
    Messages:
    343
    Likes Received:
    172
    Jerk should be about 25-30% your print speed. Acceleration should be around 600, default setting is like 9000 or something ridiculous like that.
     
  12. OutsourcedGuru

    OutsourcedGuru Active Member

    Joined:
    Jun 3, 2017
    Messages:
    752
    Likes Received:
    141
    If lowering the jerk/accel down fixes this, then...

    * the original PSU change gave everything extra mojo which then introduced the original jerk problem
    * the intra-tower distance plus the infill direction changes would be related to your mysterious "wave frequency"
     
  13. Rat_Patrol

    Rat_Patrol Member

    Joined:
    Dec 14, 2016
    Messages:
    178
    Likes Received:
    24
    Lowering the jerk/accel did nothing.

    As bad as this shifting is, one would expect to find something noticeably loose. Everything is rock, short of the right size Z axis rod bearing, which is still better than stock.
     
  14. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    There are so many little interrelated things ... :)

    At this point i am betting it is something simple.
    So simple that your eyes don't even see it, they just zoom right on by because "that is fine".
     
  15. OutsourcedGuru

    OutsourcedGuru Active Member

    Joined:
    Jun 3, 2017
    Messages:
    752
    Likes Received:
    141
    So what happened with your test-it-while-it's-still-hot test?
     
  16. Rat_Patrol

    Rat_Patrol Member

    Joined:
    Dec 14, 2016
    Messages:
    178
    Likes Received:
    24
    Everything was fine, no slop anywhere.

    Ran out of time today, won't get to it this weekend either. I will pick this project back up on Monday/Tuesday with new bearings installed.
     
    Geof likes this.
  17. Rat_Patrol

    Rat_Patrol Member

    Joined:
    Dec 14, 2016
    Messages:
    178
    Likes Received:
    24
    OK, Monday/Tuesday turned into Thursday... Life happens..

    Anyway, put in new Z rod bearings. I even sprung for the 45mm long bearings instead of 2 short bearings (Misumi specifically states that 2 bearings should not be used side-by-side like that, and 1 bearing should be used) to no avail.

    I don't like using the term desperate, but I find it being more and more appropriate.

    I agree is must have something to do with the Z axis, as the shifting changed with the new rods to much longer, deeper waves with the new TR8*8 rods/motors. If i start pushing hard on stuff, I can get things to move, but nothing is outright loose. This machine is STILL tighter than it was when new with the undersized/cheap rods/bearings.
     
  18. OutsourcedGuru

    OutsourcedGuru Active Member

    Joined:
    Jun 3, 2017
    Messages:
    752
    Likes Received:
    141
    I think Waldo's advice back on the other (acceleration/jerk) thread was to tighten up the belts, for what it's worth.

    If it were me, I think I would walk up there an hour into the print and do a pause, wait then minutes, (clip the extruded filament) and resume and repeat this every hour. If the pauses break the pattern then it might be heat-related and you could learn something (or rule out a variable).
     
  19. Rat_Patrol

    Rat_Patrol Member

    Joined:
    Dec 14, 2016
    Messages:
    178
    Likes Received:
    24
    I will do that heat testing, just to try. The only way it will work is if I release the steppers, as they stay hot if held. I may get some shifts, but I can see if the wave pattern keeps going or stops.
     
  20. OutsourcedGuru

    OutsourcedGuru Active Member

    Joined:
    Jun 3, 2017
    Messages:
    752
    Likes Received:
    141
    Or... stand there with a hair dryer (which has a no-heat mode) and if it starts to wave, blast the hotend or the part with some cool air. If the effect is positive or negative, again you might know something about whether or not this is heat-related.

    Knowing something could knock two of these off the table:
    * heat-related z problem
    * slack-related z problem
    * heat-related y problem
    * slack-related y problem
     
Thread Status:
Not open for further replies.

Share This Page