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

Solved Calibration / Calculating Steps

Discussion in 'Troubleshooting' started by izumi5188, Mar 14, 2018.

  1. izumi5188

    izumi5188 Member

    Joined:
    Feb 12, 2017
    Messages:
    59
    Likes Received:
    20
    So, I'm confused. I don't know if it is tolerance on my machine, my digital caliper, or PLA shrinking.

    I've printed this object to calibrate after doing several 20mm cubes, https://www.thingiverse.com/thing:195604

    I did it to calibrate for 100mm. I calibrated to be pretty close in the X and Y, 100.02 and 100.18. But now when I print the small cube I get 20.8 all around. Should calibrating for a larger number be more accurate?

    When I first calibrated for the 20mm cube and printed the 100mm, I was short, around 95mm.

    Am I going about this wrong?

    Thanks ahead of time.
     
  2. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    What are you changing to perform your adjustments?
     
  3. izumi5188

    izumi5188 Member

    Joined:
    Feb 12, 2017
    Messages:
    59
    Likes Received:
    20
    I am changing my step values.
     
  4. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    Why would you do that? Steps are a function that is fixed by the mathematical relationship of the parts chosen. they don't change.
     
  5. izumi5188

    izumi5188 Member

    Joined:
    Feb 12, 2017
    Messages:
    59
    Likes Received:
    20
    I changed it because all the information I was finding about honing in my accuracy suggested that. If that isn't the right approach, what should be adjusting?
     
  6. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    Well if you need to have sub mm accuracy for sizes you also have to take into account shrink rates of various plastics, some are extreme. FDM printing is not intended to make precision parts without taking that into account. Example; the largest user of ABS in the world purposely builds molds with incorrect dimensions so that the cooled parts are precisely the same size. They also control the chemical make-up of that plastic so they know that one part made 20 years ago fits together with a part made yesterday. We don't have that luxury. If you set up your printer to move incorrectly to account for a single model then even a change of suppliers would through all that work out the window. The only real way to do this is to model your parts oversized based on the output required, allowing for the specific shrink rate.

    Unfortunately you cannot use most of the shrink rate charts on the web as they are specifically for molded parts. Which means you have to guess at worst, or iteratively make multiple prints of a single part to get things correctly dimensioned.

    There is more to consider but I don't have time to discuss all the issues here in a single post.

    My suggestion for the time being, put the steps back to the mathematically correct numbers and over extrude to account for the shrinkage, then redesign the model and reprint until you get it the size you want.
     
  7. izumi5188

    izumi5188 Member

    Joined:
    Feb 12, 2017
    Messages:
    59
    Likes Received:
    20
    I can definitely do that. This means there is a lot of bad information on the web about calibration. I figured the material had something to do with, but wanted to make sure there wasn't a mechanical solution that would help.

    Thanks.
     
  8. Geof

    Geof Volunteer Moderator
    Staff Member

    Joined:
    Nov 9, 2015
    Messages:
    6,757
    Likes Received:
    2,339
    It doens't just stop at calibration :D and everyone does it differently. You want your esteps to be driving 100mm of raw filament. (Good info in this link)

    Normally if I have to have EXACT fit with FDM (across all machines not just Robo) I either post finish to size or adjust the model for the "rough dimension" it is off. If your X Y and Z are all off .2 (not terrible) then adjust then +.2 or -.2 in the model so after printing they come out "nominal" to your drawing. Kind of like if you make 2 parts that press into eachother you'll notice a .2 difference in nominal size from Model 1 to Model 2 so they press togather tight (on a well tuned slicer profile). Thats about standard.

    If your not creating your own parts then it really doesn't matter, you'll get the result you get. As far as shrinkage- PLA doesn't really do that all to bad. But for ABS you can scale your part uniformly by the amount you shrink. Say you shrink 3%, print at 103%. Its not exact, but neither are the printeres. (for what its worth- I dont do that with my machines-if it shrinks, it still going to do its job. If it needs super tight tolerances, I machine then in after the fact)
     
  9. izumi5188

    izumi5188 Member

    Joined:
    Feb 12, 2017
    Messages:
    59
    Likes Received:
    20
    Hi Geof,
    If the initial printing was just .2 off, I would take that in a heart beat. I set my steps back and on a 40mm object I'm get under 38mm of print. This didn't bother me until I was trying to print an object that need to fit a very specific piece of equipment on my CNC router. My object there was like 4mm under for size. So I printed a 100mm object and it was 5mm under size. So I looked into calibrating my machine to be more accurate.

    I'll look into my filament extrusion, maybe that is where all my problems are arising from. I'll report back after I try to calculate those steps and reprint my 40mm object.

    Thanks.
     
    WheresWaldo and Geof like this.
  10. izumi5188

    izumi5188 Member

    Joined:
    Feb 12, 2017
    Messages:
    59
    Likes Received:
    20
    Hey Geof,
    So I marked and ran 100mm. And it only came about a 1mm off. So I don't think that would be my cause for a 40mm print to come up over 2mm shy with PLA filament.

    I'm running .8mm walls. The particular roll of filament I thought 210* printed best. And I'm printing at 60mm for speed. Could that speed be to fast, causing the print to be that far off?

    I might just be back to figuring out what scale I need make my prints to get a more accurate print.
     
  11. Geof

    Geof Volunteer Moderator
    Staff Member

    Joined:
    Nov 9, 2015
    Messages:
    6,757
    Likes Received:
    2,339
    can you send me your gcode and the file? I'm guessing a slicer issue. 1mm of filament off can cause some underextrusion but not 2mm worth. Please also upload a picture of your print 2 mm off ? If its actually printing it to scale and slipping up by 2mm then it should be one ugly cube.
     
  12. izumi5188

    izumi5188 Member

    Joined:
    Feb 12, 2017
    Messages:
    59
    Likes Received:
    20
    Yeah, let me jump on the computer that has the files of the object I'm currently running. It's isn't cube. But for the most part it appears to be printing "square". Give me about 10min.
     
    Geof likes this.
  13. Geof

    Geof Volunteer Moderator
    Staff Member

    Joined:
    Nov 9, 2015
    Messages:
    6,757
    Likes Received:
    2,339
    No problem, will get back to you asap.
     
  14. izumi5188

    izumi5188 Member

    Joined:
    Feb 12, 2017
    Messages:
    59
    Likes Received:
    20
    Well, I'm feeling a little silly. That way I pulled in the part it was rotated that my points were measuring 40mm. And actual flat edges should be 38.04 according to my cad soft ware. Now I don't feel so bad about it's current calibration of this object. My flat sides and coming in at 37.4X to 38.3X . I think it looks a little lop sided but not 2 bad.

    Every flat side should be 38.04 from flat edge to flat edge.

    But here are the files and a picture from underneath the object. And yes I know my file names are not great. But when I'm fine tuning a model, saving it every time before it's finalized is obnoxious.

    The print took me, I think 30min but I already don't remember.

    This model isn't the model I was trying to get the best precision out of. But that model takes like 5 hours to print, a little hard to use for calibration.
     

    Attached Files:

    #14 izumi5188, Mar 15, 2018
    Last edited: Mar 15, 2018
  15. izumi5188

    izumi5188 Member

    Joined:
    Feb 12, 2017
    Messages:
    59
    Likes Received:
    20
    Geof,
    I don't know if you are planning on printing out my model. I don't think it is necessary at this point. It is there if you'd like to take a look at it.

    I think my next set of plans is to try and print the calibration object about and see about calculating an X, Y, and Z multiplayer and see how that works for like 20, 50, and 100mm. Hopefully that will allow for better accuracy for objects that I need it for.

    Most things I don't care to much, but when I'm trying to print an object to fit around a manufactures specific piece, then I care more.

    Thanks for your current insights.
     
  16. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    If you are using a 0.40 nozzle I think this is a large contributor to your problem. If left on Auto I am pretty sure the default extrusion width is 0.48 mm It would be hard to push out enough plastic to have a 0.80 mm extrusion width with that size nozzle.
     
    Geof likes this.
  17. izumi5188

    izumi5188 Member

    Joined:
    Feb 12, 2017
    Messages:
    59
    Likes Received:
    20
    Hi WheresWaldo,
    I can't say I've measured the individual wall perimeters. I do know if I set it to .4, .8, 1.2, I get 1,2,3 perimeters respectfully.
     
  18. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    Try the same calibration model at a more normal extrusion width of say 0.5 or the default 0.48 and see if you are closer on the cube measurements. If you want something that wide, you would usually increase the nozzle diameter like running a 0.6 or 0.8 mm nozzle.
     
  19. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    When I ran my R1 everyday, the 20 mm calibration cube using PETG was always around 19.95 mm so way closer than you are getting. But I never changed the extrusion width to try to push out anything more than 0.5 mm extrude.
     
    Geof likes this.
  20. Geof

    Geof Volunteer Moderator
    Staff Member

    Joined:
    Nov 9, 2015
    Messages:
    6,757
    Likes Received:
    2,339
    ditto, I never go much over .46 ext on a .4 nozzle. smaller to get fill in areas on small details sometimes, but larger will cause some grief.
     

Share This Page