Titan is going things that I don't understand and can't explain

Discussion in 'Titan' started by Yurii Haiovyi, Dec 28, 2018.

  1. Yurii Haiovyi

    Yurii Haiovyi Member

    Joined:
    Dec 27, 2018
    Messages:
    9
    Likes Received:
    0
    I've been fighting this issue alone for too long, it's time to give up and ask for community help.

    So I have e3d genuine titan, (genuine) v6, and (genuine) pancake motor all connected in direct extruder (tube size is perfectly cut as per instruction, "a tiny bit shorter rather than longer".)

    I use Marlin, my x, y, z, e steps are calibrated using 0.02mm precise caliper, PID auto tuned, x, y, z are fine tuned using caliper as well, bed UBL auto leveled, belts are tight, recommended 0.48mm line with, 0.25mm line height (doesn't really matter, I've tried various combinations) there should not be any issues, and there's no with walls, they look awesome... but 100% infill and top layer look like at least 2x over extruded. How a hell it's even possible???

    If I try to work around it with flow multiplier, I got gaps between wall lines. Again, without any tweaking prints sizes are not more than 0.2% different from intended. And walls seem absolutely normal as for me.

    So at some point I started blaming slicer. Apparently I was wrong, again.

    So Cura with my settings - the same issue, with default settings - the same, Slic3r - the same, IdeaMaker - the same, different filaments - the same, abs vs pla - the same, different temperatures - the same, disassemble and assemble - the same, different speeds - the same (10-80mm/s), accelerations down to 1500mm/s^2 - the same, different nozzle (confirmed to print well) - the same, linear advance K=0 - the same, any non-zero - the same.

    Bottomline, for $150 setup, it's doing very bad so far, it's so frustrating.

    I'd appreciate any ideas, otherwise I don't have any choice but dispose it, and never buy from e3d again.

    Thanks in advance.

    PS: I measured filament, it's 1.75 straight.

    20181228_004951 (1).jpg
     
  2. mhe

    mhe Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    279
    Likes Received:
    77
    I'd start messing around a bit with the extrusion width overall or with Simplify3D's infill extrusion width (which you can tweak separately). Also make sure your nozzle is clean and free of burrs, the smudges on top look as if there is a tiny burr on the nozzle maybe.
     
  3. Yurii Haiovyi

    Yurii Haiovyi Member

    Joined:
    Dec 27, 2018
    Messages:
    9
    Likes Received:
    0
    I have just installed a brand new original hardened steel nozzle instead of standard one, another $23 wasted.
     
    #3 Yurii Haiovyi, Dec 31, 2018
    Last edited: Dec 31, 2018
  4. mhe

    mhe Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    279
    Likes Received:
    77
    Be that as it may, have you tried tweaking the infill extrusion width?
     
    Yurii Haiovyi likes this.
  5. Yurii Haiovyi

    Yurii Haiovyi Member

    Joined:
    Dec 27, 2018
    Messages:
    9
    Likes Received:
    0
    As a workaround it could work, but c'mon, walls look awesome, and if I set walls thickness to infinity I get no over extrusion, which means flow rate is fine. It fills like the first bottom layer looks fine, and every next one is getting worse and worse. But even the first one looks uneven.
     
  6. mhe

    mhe Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    279
    Likes Received:
    77
    Well the walls look awesome because they only have to look good in one dimension (the exposed side), whereas the top layers get exposed in 2 dimensions - in other words, the inside of your wall could be just as messy, you just don't see it currently.

    If I understood you correctly, the quality of a top layer gets worse and worse the taller a part is?

    You can get quite decent top layers out of a Titan.
    E.g. https://www.dropbox.com/s/b1706dxrdkjsk94/20180826_230047.jpg?dl=0
     
  7. Yurii Haiovyi

    Yurii Haiovyi Member

    Joined:
    Dec 27, 2018
    Messages:
    9
    Likes Received:
    0
    I found couple of issues by now, but fixing them didn't fix entire issue yet, looking for other issues.
    When `S_CURVE_ACCELERATION` you want to turn off `jerk` feature in slicer, and never over tight extruder.
     
  8. Yurii Haiovyi

    Yurii Haiovyi Member

    Joined:
    Dec 27, 2018
    Messages:
    9
    Likes Received:
    0
    I truly can't understand what gives those pulsations on the surface near to edges, it eventually screws up entire print, because every layer adds to the height error. For now I turned off almost every marlin feature I had enabled. Didn't help at all.
     

    Attached Files:

  9. mhe

    mhe Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    279
    Likes Received:
    77
    Looks to me the slowdown towards the turning point doesn't slow down the flow rate synchronously. So the extruder seems to ramp down in steps whereas the x/y axes slow down gradually (or vice versa). How high is extruder jerk in your firmware?

    Jerk is a bit confusing of a term since what it refers to in proper physics is 3rd order motion or rate of change in acceleration. However, in firmwares jerk is usually a speed that is assumed to require no ramp up (or down) so there is no acceleration/deceleration curve calculated in the firmware for it.

    I'd try to lower extruder jerk value in your firmware a bit and see what happens. But I guess this is the value that will need tweaking to arrive at the result you want.
     
  10. Yurii Haiovyi

    Yurii Haiovyi Member

    Joined:
    Dec 27, 2018
    Messages:
    9
    Likes Received:
    0
    #define DEFAULT_XJERK 10.0
    #define DEFAULT_YJERK 10.0
    #define DEFAULT_ZJERK 0.3
    #define DEFAULT_EJERK 5.0
    #define S_CURVE_ACCELERATION

    Jerk is unchecked in Cura.
    But seems like without S_CURVE it's the same anyway.
     
  11. Yurii Haiovyi

    Yurii Haiovyi Member

    Joined:
    Dec 27, 2018
    Messages:
    9
    Likes Received:
    0
    Increasing EJERK to 10.0 made surface better
     
  12. mhe

    mhe Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    279
    Likes Received:
    77
    Interesting that an increase helped. Keep tweaking, will be interesting to see how well you can dial it in!
     
  13. Yurii Haiovyi

    Yurii Haiovyi Member

    Joined:
    Dec 27, 2018
    Messages:
    9
    Likes Received:
    0
    This is madness, so I can print shell with very high speeds, but top/bottom layers and infill should be printed with low jerk and speed, that way I get 'em perfect. I don't think that's normal anyway.
     
  14. Yurii Haiovyi

    Yurii Haiovyi Member

    Joined:
    Dec 27, 2018
    Messages:
    9
    Likes Received:
    0
    Very high jerk, partially works too.
     
  15. mhe

    mhe Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    279
    Likes Received:
    77
    Well, there we go, problem solved I hope? :)
     

Share This Page