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

Unresolved 3d Systems Cube3 Quality > My ROBO R1

Discussion in 'Troubleshooting' started by TheGlutton, Jul 28, 2016.

  1. TheGlutton

    TheGlutton New Member

    Joined:
    Jul 2, 2016
    Messages:
    26
    Likes Received:
    9
    Hey Everyone!

    I just want to start by saying thank you to all those who helped get me up and running through my last thread! :D

    Ok so I think I have my z-offset issues dialed, calibrated my E-Steps and I am now getting much more reliable first layer quality.

    My friend purchased a 3d Systems Cube 3 printer at the same time that I got my ROBO R1 printer (I turned him onto it, closeout for $220) and in addition to his experience being nearly plug-and play the quality of his first Benchy (unfinished) blows the best I have managed out of the water. I am totally impressed with his Cube3 and feel like there is LOTS of room for improvement on my own print. I am not entirely sure where to start with troubleshooting and was hoping you folks might at least name the faults in my print so I can get on the right track with google.

    I can't let the Cube3 beat my ROBO! (at least not this badly!)

    I printed this in CURA 15.04.6 at:

    BASIC SETTINGS:
    Layer Height = .07mm
    Shell Thickness = .8

    Bottom/Top Thicnkess = 1.2mm
    Fill Density= 15%

    Print Speed = 30mm/s
    Extruder Temp= 190C
    Bed Temp= 50C

    Brim = Yes

    ADVANCED SETTINGS:
    Retraction Speed = 40mm/s
    Distance = .5mm

    Initial Layer Thickness= .3mm
    Initial Layer Line width = 125%
    Travel Speed = 150mm/s
    Bottom Layer Speed = 30mm/s
    Infil Speed = 50mm/s
    Top/Bottom Speed = 40mm/s
    Outer Shell Speed = 30mm/s
    Inner Shell Speed = 40mm/s

    Minimal Layer Time = 7 sec.

    1607_untitled_11153.jpg 1607_untitled_11152.jpg 1607_untitled_11151.jpg

    The Cube3 print was cancelled partway through because my friend was going to bed and concerned about leaving the machine running overnight. He is on latest firmware and software for the machine and chose the best quality option, not sure what settings were used beyond .07mm layer height.
    1607_untitled_11154.jpg 1607_untitled_11155.jpg

    I know it is silly to get into a competition over prints made at a quality that is too slow to be practical, but when I saw what he made with nearly 0 effort on his machine I realized that I should be able to accomplish the same with some effort, right?

    Thank you!
     
  2. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    A couple of observations.
    1) Print Speed is the one big knob you can tweak to adjust quality.
    2) Layer height... there is a lot to read on that (for example: http://forums.reprap.org/read.php?1,125511,125511) however as a rule of thumb I try to keep my layer height 30-60% of the nozzle diameter for best results. Assuming the stock 0.4 then: 0.12 -> 0.24 mm (0.1 mm is usually OK too so lets say 0.1 -> 0.25). 0.07? That, sir is pushing it ... drop to a 0.2 nozzle.

    Resolution (layer height/nozzle diameter) all affect max speed too so as you get to finer resolutions you really need to make overall print speed decrease linear with resolution increase.
     
    TheGlutton and Ryan TeGantvoort like this.
  3. daniel871

    daniel871 Well-Known Member

    Joined:
    Apr 18, 2015
    Messages:
    1,322
    Likes Received:
    510
    To be perfectly honest, both examples look like hot garbage to me.

    I'd go back to printing calibration cubes and fiddling with settings until I didn't see all those bumps and whatnot everywhere.
     
    TheGlutton and Ryan TeGantvoort like this.
  4. Ryan TeGantvoort

    Ryan TeGantvoort Active Member

    Joined:
    Mar 11, 2016
    Messages:
    343
    Likes Received:
    172
    So I looked through one of your previous threads dealing with the Calibration Cube height. Trying to see what you have done and what printer you have.

    Initial Changes I noticed:
    #1 - Lead Screw Upgrade
    #2 - Lower Acceleration Settings
    #3 - Calibrate Extruder (if not previously done)

    Plus what @mark tomlinson said.
     
    TheGlutton likes this.
  5. TheGlutton

    TheGlutton New Member

    Joined:
    Jul 2, 2016
    Messages:
    26
    Likes Received:
    9
    I just finished a 5hr 20min Benchy, and it looks a little better, but MAN was that painfully slow...

    @mark tomlinson Speed definitely kills, but I have to find a balance between acceptable quality and waiting days for prints to finish. Thanks for that article, makes sense I will shift back to .1mm as my max resolution.

    @daniel871 That is actually very comforting to hear, because I have been pretty disappointed with my results thus-far. I am totally at a loss for that is causing the ugly globs and such on what appears to be the rear-facing vertical surfaces. What settings would you start to tweak? I am printing another 20mm cube now.

    @Ryan TeGantvoort Thank you for the suggestions! I am tight on cash (arent we all) and so the Lead Screw Upgrade will have to wait, there has to be a way for me to get acceptable prints with this machine out of the box. I did calibrate the Extruders E-Steps in my firmware :) But have read mention of manually twisting the Lead screws to level? I cant find where I saw it mentioned now, but was there a step I should have performed?

    The Cube just finished here were my settings:
    Screen Shot 2016-07-28 at 2.49.13 PM.png Screen Shot 2016-07-28 at 2.49.24 PM.png Screen Shot 2016-07-28 at 2.49.33 PM.png

    And here was the output:
    1607_untitled_11158.jpg 1607_untitled_11157.jpg 1607_untitled_11156.jpg

    I changed acceleration from 1300 to 700, tweaked the probe movement speed during bed-leveling to speed it up, and also lowered my z-offset further because my brim lines were not touching each other. I am printing another cube now.
     
  6. Ryan TeGantvoort

    Ryan TeGantvoort Active Member

    Joined:
    Mar 11, 2016
    Messages:
    343
    Likes Received:
    172
    When you changed the 1300 to 700, you should also change these:

    #define DEFAULT_MAX_ACCELERATION {600,600,100,10000}

    I think you still have them set at 9000, which is the default setting within Marlin for some reason? :D As you can see I lowered mine to 600. Your XY Jerk setting should be about 20-30% your print speed as well (a good starting point anyways). Adjusting these setting should give you better results, especially at corners.

    Manually turning the Lead Screws is typically used for manually leveling the X-axis. Most people call it Paper Level. Take a piece of paper and place on the bed jog the nozzle down until it just grabs paper. Power everything off, slide the Carriage back and forth and adjust the Lead Screws until they are even.
     
    TheGlutton and Geof like this.
  7. Geof

    Geof Volunteer Moderator
    Staff Member

    Joined:
    Nov 9, 2015
    Messages:
    6,757
    Likes Received:
    2,339
    I'd be happy to slice your model in s3d if you'd like. The benchy was insanely hard for me with curs. If interested send me a pm and we can work on it
     
    TheGlutton likes this.
  8. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    MAX speed will affect your overall quality a lot more than resolution will. You should take a small but complicated test model (like the DePrime) and play with speeds. Then you can quickly get a good test and make an informed decision :) You do not have to run it in creep mode, but find the spot above which quality noticeably falls off.

    Insanity. That is a reason.

    600 is a good number.
     
    Ryan TeGantvoort, TheGlutton and Geof like this.
  9. TheGlutton

    TheGlutton New Member

    Joined:
    Jul 2, 2016
    Messages:
    26
    Likes Received:
    9
    @Ryan TeGantvoort Thank you, I have just updated the XY_Jerk to 12, and the acceleration settings to {600, 600, 100, 10000}, added a belt tensioner to my X axis belt, and printed the following cube:
    1607_untitled_11159.jpg 1607_untitled_11162.jpg 1607_untitled_11160.jpg 1607_untitled_11161.jpg

    These were my settings:
    Screen Shot 2016-07-28 at 4.54.17 PM.png Screen Shot 2016-07-28 at 4.54.30 PM.png Screen Shot 2016-07-28 at 4.54.35 PM.png

    I lifted the print bed and checked my Y-Axis belt, and it is pretty loose, is it supposed to be?
     
    Ryan TeGantvoort likes this.
  10. TheGlutton

    TheGlutton New Member

    Joined:
    Jul 2, 2016
    Messages:
    26
    Likes Received:
    9
    @Geof Thank you for the offer, I may take you up on it in the future :)
     
    Geof likes this.
  11. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    Geof and TheGlutton like this.
  12. TheGlutton

    TheGlutton New Member

    Joined:
    Jul 2, 2016
    Messages:
    26
    Likes Received:
    9
    Time for an update.

    Made big strides, but still having some issues. I used the Paper Leveling method and found that the right side of the carriage was quite a bit higher than the left. I have also tweaked the XY-Jerk to 12 in the firmware. Using Hatchbox Yellow PLA.

    My latest and best Benchy printed at .2mm layer heights, and 50mm/s speed, took about an hour and a half, on a cool night with a window fan blowing on the printer at max speed:
    1607_untitled_11178.jpg 1607_untitled_11179.jpg 1607_untitled_11180.jpg 1607_untitled_11181.jpg

    I watched another earlier Benchy print and noticed that the print was staying too soft, and was squishy under the print head as it printed each layer. This is my best so far and was printed at "poorer" settings than the one I shared above at .7mm layer height and 20mm/s speed (over 6 hours to print).

    The print is far from perfect, but is MUCH better. I printed this: http://www.thingiverse.com/thing:967881 and will see if I can get it installed tonight to see if that will get me similar results.

    However, there are still some issues, the main issues I can identify are (name anything you see that I didn't catch):

    EDIT: ALL OF THE EXAMPLES BELOW WE PRINTED AT .2MM AND 50MM/S SPEED

    1) Z-banding or ribbing, which I understand is a limitation of the threaded rods my R1 uses. I will live with it until I can afford the lead screw upgrade.

    2) This strange gapping:
    1607_untitled_11164.jpg 1607_untitled_11166.jpg

    3) The top layer of some parts of this print did not close?
    1607_untitled_11176.jpg

    4) Corners seem to be protruding, especially when the printhead changes direction 90 degrees from moving along the y-axis to an x-axis direction. This always ends up being the front-left and back-right corners of my test cubes.
    1607_untitled_11156.jpg

    5) There seems to be one point where the printer "catches" or stutters/hiccups, on any shape, circles seem to show it best. It seems to be at the 7:30 position if the print were on a clock, again, right about where the print-head movement changes from primarily Y-axis to X-axis. This is maybe the same issues as #4 above?
    1607_untitled_11182.jpg

    If anyone has any ideas about how I might fix issues 2-5 of if you have any other pointers for improving my print quality I would greatly appreciate them!
     
    #12 TheGlutton, Jul 31, 2016
    Last edited: Jul 31, 2016
  13. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    If your description is correct then some of the issues lie in the fact that you physically cannot print a layer of .7 mm and expect anything but print failures. Typical values range between 25% to 75% of the nozzle diameter. If you are using the default 0.4 mm nozzle, that means between .1 mm and .3 mm layer heights. Just that alone can result in the gapping you see in picture 2. Retraction settings could contribute to the rounded corners in picture 4 along with over-extrusion that would be necessary with a layer height that is simply too tall.

    I have nothing else to add at the moment.
     
    TheGlutton likes this.
  14. TheGlutton

    TheGlutton New Member

    Joined:
    Jul 2, 2016
    Messages:
    26
    Likes Received:
    9
    @WheresWaldo sorry for the mixup! I edited my post above to say that all of the example photos in that post were a .2mm printed at 50mm/s speed. Sorry!
     
  15. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    Excellent! Moving on :)

    There is a lot more than 'gapping' going on there. Look at the thickness of the layer as you move across the bed.
    It is wildly uneven from side to side. The gapping is likely because the nozzle is too close to the bed and can't extrude enough filament. you have some first layer leveling to do if you want to improve this.

    http://printedsolid.com/firstlayer

    Auto Bed Leveing you say? I'd suggest a lot more reading :) It will not account for too large of a variance across some axis (i.e. the X needs to be pretty level from the start or the ABL mechanism will not work reliably). It has to assume some given values are 'good' in the current incarnation.

    We would need to see the model, but I suspect the gap is due to the nozzle size/layer width/layer thickness versus the model. When the slicer did its job it found some 'impossible' sections and just omitted them. You can make walls too thin for a given nozzle configuration and how small you defined the resolution in the slicer settings can affect it as well (like @WheresWaldo commented above)

    Wild guess, but either the extruder need to be calibrated to get the extrusion rate set correctly () or there is a little slop in one or more of the drive belt systems. That does not mean the belt is loose (although that would affect it too) you could have a loose drive gear on a stepper motor for an axis...they are famous for not being setup correctly (i.e with loctite) and they come loose and move a slight amount.

    If it is a loose drive part then yes...

    Good luck.
     
    TheGlutton and Ryan TeGantvoort like this.

Share This Page