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

Unresolved Failed Overhang

Discussion in 'Troubleshooting' started by spankurmonkey, Feb 16, 2017.

  1. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    So colder was fine?
    I assume you started with one below 190...
    Some of the temperature tests go down as far as 170 or 180... you don't need to be that cold.
     
  2. spankurmonkey

    spankurmonkey Member

    Joined:
    Dec 26, 2016
    Messages:
    102
    Likes Received:
    3
    I ran this at 190 with these two fans at 100%

    Some areas the print looks like it should and other, especially on the right side, have this stringing.. but I am not sure that is really stringing per say but some other term but I don't know what it is called.

    [​IMG] [​IMG]


    image upload no size limit
     
  3. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    That last print really looks like there are bigger issues than cooling.
     
  4. spankurmonkey

    spankurmonkey Member

    Joined:
    Dec 26, 2016
    Messages:
    102
    Likes Received:
    3
    Such as ??? I am having a hard time troubleshooting this..
     
  5. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    Under-extrusion, maybe layers are too big, it is hard to say with just one image, what about all the other failed prints?
     
  6. spankurmonkey

    spankurmonkey Member

    Joined:
    Dec 26, 2016
    Messages:
    102
    Likes Received:
    3
    Is it possible that in S3D I messed something up...

    I.e. Nozzle Diameter, Extrusion Multiplier, Extrusion width, etc??

    I am baffled .. I printed this the other day on my old fan set up... https://www.thingiverse.com/thing:1624412

    Thought the layer height was .10 and it came out perfect... the old fan is just a radial on this -- http://www.thingiverse.com/thing:2049748

    Yet that radial will not print those MPCNC parts nor that planter from today.
     
  7. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    Pictures of the other fails or it didn't happen! :eek:

    Unless all your mechanical settings are pretty close to perfect, I find that using single processes for multiple layer heights is less than optimal. S3D has too many common parameters that work against you if you have one of those processes where you can choose fast/medium/high.

    Also, why are you printing everything at 0.1 mm, is it necessary or do you just want it to be harder and slower to print?

    First make sure you are mechanically sound, did you modify or change firmware in any way from Robo's default?
    Did you calibrate your extruder to make sure your printer knows that 100 mm of filament is really 100 mm long?

    With just descriptions and very little evidence (pictures) it makes the forum members job a lot harder.
     
    #67 WheresWaldo, Feb 26, 2017
    Last edited: Feb 26, 2017
  8. spankurmonkey

    spankurmonkey Member

    Joined:
    Dec 26, 2016
    Messages:
    102
    Likes Received:
    3
    someone on here told me to start at .10 - When I first got the printer only messing with the firmware was to accept the lcd controller that's it.

    I am not sure how to calibrate the extruder, so no I have not.

    I started at the .10 per the other post when the overhang tests were failing. I started this thread because some of the parts for the MPCNC which had overhangs failed miserably.

    Like I said the Dragon and a Benchy printed just fine at the .10 but when I try something different or layer heights of .26 or .30 I get that filament run over like a bunch of stranded hairs and that is what I am trying to figure out the cause of and how to correct it.
     

    Attached Files:

    • 006.JPG
      006.JPG
      File size:
      84.6 KB
      Views:
      4
  9. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
  10. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    .3 is a little high for a 0.4 nozzle. Your best results for any nozzle with this style FDM machine will be in the ball-park of 50% nozzle size. Now, can you go higher or lower? Sure, with tweaking and testing you can actually use 30%->70% of the nozzle diameter for a layer height*, but it will not be at the "default" settings. For new users just getting into printing I always suggest you stay at the sweet spot. If you need 0.1mm layer heights go with a 0.2mm nozzle (and slow down the print speed by 50% or more).



    *you can push it to real extremes -- we have seen 0.05 layers with a 0.4 nozzle, but this is not easy nor best practice :) I refer you to @tesseract one of our first users to really push things:

    http://community.robo3d.com/index.php?threads/highest-quality-print.5578/#post-47891

    He has done some amazingly tiny prints with stock stuff. It is just not easy nor simple.
     
    #70 mark tomlinson, Feb 26, 2017
    Last edited: Feb 26, 2017
  11. spankurmonkey

    spankurmonkey Member

    Joined:
    Dec 26, 2016
    Messages:
    102
    Likes Received:
    3
    I will run the extrusion tests today.

    When I printed the MPCNC part where the overhang failed I was printing at .26 layer height.

    Just so I am clear you are saying the sweet spot is 50% of the nozzle, thus .20 ?
     
    mark tomlinson likes this.
  12. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    Somewhere near there will be the best spot at all of the default settings and hardware (0.4mm nozzle).

    If you go with that you are safe to use all of the default profile settings. Also, when you are making adjustments, keep it to one thing at a time. Make one change, then test. Otherwise you will never sort out what affected your print since too many of the settings interact with each other.

    Another really helpful rule is to reduce print speed as you reduce layer height (i.e. increase resolution). I use a linear relationship and that works well (half the layer height -> half the speed). Same is true for nozzle sizes. If you go down in size, go down in speed.

    That relationship is why people use the volcano adapter -- larger bore nozzle with a huge meltzone means that you can push the printer to the max speed possible and still get acceptable prints. Even if you do NOT push the speeds the nozzle diameter does not have the sort of quality impact you might imagine. A 0.8 nozzle does not make hideous prints and assuming all things are equal (in terms of the mechanical bits working correctly) the quality is not hugely different visually. Same is true for most models as you go down in layer height. Smaller, highly details models will certainly show a difference, but usually on a printer bed this size you are not printing 50mm squre or less prints :)
     
  13. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    OK, that print on the left. That is not stringing really, that is filament being dropped in place (rather than pressed).
    Not having analyzed your printer :) I would suggest that this is either a mechanical issue (loose drive allowing sloppy layers to be put down) or printing too fast (which means drive lash is the real issue).

    From what I can see of the top of that print there appears to be some spacing between layers as well and that can also be from the filament not being pressed flat. Start with default settings and then do a couple of small tests. I have attached the first layer test (circle bed level) and the DePrime test. Let's start small :)

    Also if you have updated the firmware at any point on this printer let me know exactly which firmware you used. Also refresh us on exactly which printer model you have and ANY upgrades (I think the answer is "just fans" but we need to be certain)
     

    Attached Files:

    #73 mark tomlinson, Feb 26, 2017
    Last edited: Feb 26, 2017
  14. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    I was the one that told you to print at 0.1 mm but that was just a test! I don't recall which printer you have and don't see it here in this thread. But as @mark tomlinson pointed out there are limits to what you can do without changing hardware. For me, as a rule of thumb I try to stay between 25% and 75% of the orifice diameter for layer height. That means for a 0.4 mm nozzle I don't print below 0.1 mm or above 0.3 mm layer heights. At the same time I had to create different profiles for each layer height I typically use, the Fast/Medium/High choice in an S3D profile are too limiting especially since there are too many shared parameters (changing it at one layer height changes all layer heights). For example, at 0.12 layer height I print the first layer at 200% height, that way I am guaranteed an amount of extrusion that will properly build a first layer. I don't need that extra at 0.2 or 0.28 layer heights. I rarely print over 0.28 mm. You will also notice I am not printing at 0.1, 0.2, 0.3. This is because I want all my layers to have an even number of steps and no fractional steps. You can use the Prusa calculator to find the ideal number of steps per layer at http://www.prusaprinters.org/calculator/), making sure all your chosen layer heights are in full steps.

    But all this said, the very first thing you need to do is to follow the video that @mark tomlinson posted.
     
    #74 WheresWaldo, Feb 26, 2017
    Last edited: Feb 26, 2017
    mark tomlinson likes this.
  15. spankurmonkey

    spankurmonkey Member

    Joined:
    Dec 26, 2016
    Messages:
    102
    Likes Received:
    3
    In regards to changes to the firmware.. I had the smaller LCD which required a firmware change... then I got the XXL so I switch it back. No other firmware changes. In regards to firmware version I am not sure... remind me how to check so I can report back.

    the model is the Robo 3D R1 Plus - Yes, I have messed with many fans and ducts. right now the underside is the printed part for the Black Widow fan set up allowing a fan on both side in the default orientation as the stock one with no duct. The only other upgrade is the cable chain.

    In regards to speed perhaps that could be part of the issue the MPCNC part I was printing at 45mm/s at the .26 layer height. I think the stuff I was printing yesterday... yes I should have mentioned this but I think I increased the flow rate up.

    So for today I will do the extruder calibration test and then I will do the other two tests that Mark provided.
     
  16. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    If you look at the firmware files you downloaded, in Configuration.h line 79 there is a CUSTOM_MENDEL_NAME, what is it?
    line 445 should be DEFAULT_AXIS_STEPS_PER_UNIT what are those numbers?
     
  17. spankurmonkey

    spankurmonkey Member

    Joined:
    Dec 26, 2016
    Messages:
    102
    Likes Received:
    3
    // Define this to set a custom name for your generic Mendel,
    #define CUSTOM_MENDEL_NAME "ROBOR1PLUSV1" //robo


    #define DEFAULT_AXIS_STEPS_PER_UNIT {80,80,800,723.38} // default steps per unit for RoBo 3D R1
     
  18. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    Okay, I can tell you with 100% certainty that you are under-extruding. Follow the video and the 4th parameter on line 445 will be much closer to 800 than 723.38 once you correct that you can then tackle the other issues one at a time.
     
    mark tomlinson likes this.
  19. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    One of mine is slightly over even. Dead certain it will be more than 723 :)
     
  20. spankurmonkey

    spankurmonkey Member

    Joined:
    Dec 26, 2016
    Messages:
    102
    Likes Received:
    3
    First, I ran both print tests - untouched with standard settings in S3D.

    To me , both prints passed the test.

    [​IMG]

    [​IMG]


    I then ran the Extrusion test.

    1. Loaded white filament.
    2. Measured up 6" from where the filament meets the machine. So in MM terms that is 152.4 MM
    3. Heated extruder to 212c
    4. Using Octoprint terminal I sent a G92E0 command
    5. Using Octoprint terminal I sent a G1E100F90 command.
    6. I turned off the hot end after the extruder finished running.
    7. I measure up to my mark on the filament which read approximately 58.35MM

    So doing the math... 152.4 - 58.35 = 94.05 MM so not quite 100MM

    Following the video....

    1. Take old steps per MM ( 723.38 ) x 100 = 72338
    2. Take the Sum of #1 divided by the extruded length ( 72338 / 94.05 )
    3. New steps per MM = 769.14

    Since I have never done this please let me know if you agree with the outcome...... I can easily run another test after making the change to the firmware.
     

Share This Page