Jump to content

Getting Better Dimensional Accuracy - Calibration


Recommended Posts

Posted · Getting Better Dimensional Accuracy - Calibration

Oh dammit, @LaserBrain, my comments were based on your test cube but if your are printing circular geometries then the inner diameter will have a tolerance close to what your stating , i.e. smaller; of course you may already know that ! but I just wanted to be clear.

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration
    12 hours ago, Jakeddesign said:

    When I started at this company they told me they didnt have a use for a printer...then I convinced them to buy and 6 months later this thing NEVER stops printing.  Its crazy....my full time Engineer gig is now more than 50% printer technician!

    That sounds good! What is it used for mostly, if you don't mind me asking? 

     

    9 hours ago, yellowshark said:

    When I say measured I did this by manually extruding 10mm of filament and cutting it off and measuring the diameter.

    Do you mean you extruded the filament not printing on a bed, but just catching it from the air? Wouldn't it stretch because it is still soft / warm and it just hangs there?

     

    9 hours ago, yellowshark said:

    I have no idea if they test each new machine and recalibrate that setting if required.

    We do! I don't think we calibrate stepper motors, as they should all be good and consistent. But we do check them, and if they don't pass our calibration test and quality assurance, it is not going out until it does. 

     

    @ghostkeeper, can you say a little bit more about the post processing script; post stretch script? I can imagine there also lies a partial solution here re the ID of holes. Do we know of a speed / temperature / layer height which is ideal for an accurate ID? I could imagine, a small layer height, at a low temperature and not very fast so the amount of filament you extrude is small and therefor cools fast and is minimally stretched along the nozzle tool path? @Jakeddesign , did you try this script as well? (it can be found under extensions > post processing > modify g-code. 

     

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration
    40 minutes ago, SandervG said:

     

     

    10 hours ago, yellowshark said:

    When I say measured I did this by manually extruding 10mm of filament and cutting it off and measuring the diameter.

    Do you mean you extruded the filament not printing on a bed, but just catching it from the air? Wouldn't it stretch because it is still soft / warm and it just hangs there?

    Hi @SandervG, that is a fair point. What I do first is to extrude 30-40mm to ensure all is extruding well and as it should be, which does mean that on the first cut it is not totally hanging as it is laying on the bed before the cut is made. BUT that does not apply to the subsequent cuts (although as I said above the measurements were cconsistent). I do not think that measuring it  printed on the bed is reliable as that will depend on how squished you have the filament set to via the bed to nozzle distance. Visually it looks right, i.e. once the first tens of mm have been extruded the width of the extruding filament, from the nozzle tip to the bottom looks consistent BUT that is not the same as using a vernier! I guess the good thing for me is that this "stretched" filament was fatter than the specified nozzle diameter and using .45 was better than using .40. Your point though does make me think that I will have another play and maybe try 0.47 or even 0.5 and see what happens.?‍?

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration
    6 hours ago, SandervG said:
    18 hours ago, Jakeddesign said:

    When I started at this company they told me they didnt have a use for a printer...then I convinced them to buy and 6 months later this thing NEVER stops printing.  Its crazy....my full time Engineer gig is now more than 50% printer technician!

    That sounds good! What is it used for mostly, if you don't mind me asking? 

     

    I work for a prosthetics company, we focus mostly on lower limb, but are getting into upper limb prosthetics as well.  Any plastic components were order through Statasys, which is a great service, but quite expensive and long leadtime (unless you pay even MORE money).  We have one product that I have printed 4-5 versions using the UM3 - something we would not have done if we were ordering prints from a vendor.  Saved a lot of money on mold change costs too, since even printing such complex parts on an FDM printer we were able to identify some pretty serious overlooks.

     

    The printer runs for at least 40 hours a week, and if I time prints just right, I can hit 100 hours.

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration
    On 5/8/2018 at 12:18 AM, yellowshark said:

    Oh dammit, @LaserBrain, my comments were based on your test cube but if your are printing circular geometries then the inner diameter will have a tolerance close to what your stating , i.e. smaller; of course you may already know that ! but I just wanted to be clear.

    No Problem @yellowshark, I've been super busy so not had a chance to reply. However, thanks for your info. Round parts are a struggle, especially when they have to fit, I've had to slightly overscale a couple of items to get them to fit, otherwise they are too snug.
    As for the calibration cube, I'll try some of the suggestions and see where it leads too. The hardware looks all good, gave it a once over today, belts etc. 
    Cheers,

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    I found this topic while searching for a solution to the same problem. I have to admit I am really disappointed. I'm new to Ultimaker, I used Prusa for some time before buying the S5. I wanted the S5 for the dual extruder. Without any calibration on Prusa all the parts (in PLA or PETG) are unbelievably accurate (+/- a few microns). If you make 2 parts with the same dimensions (one on the outside and one on the inside) they will simply fit together. Buying a printer that is ~ 7 times more expensive I assumed even better quality. Well... unfortunately I cannot return it... but I probably never use it - I simply need very accurate parts.

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    I have a Prusa, clone, while not the real thing, I have found that it's way less accurate than my Ultimaker. Out of the box, yes I had to adjust the ultimaker slightly and I achieved a +/- 10 Microns outcome. IMHO any FDM printer is going to struggle below this, due to the nature of the print technology variables in temp, movement, filament etc. Each printer is going to differ as it translates any mechanical inaccuracies of the build. I have an Ultimaker 3, for dual extrusion, I looked at the 5S but couldn't justify the expense for just what appears to be size, extrusion sensors and potentially stronger filaments that I'm unlikely to use. I have mine dialled in pretty well and I'm happy with my results. I have multi-part prints that fit so I have my internal and external dimensions sorted now. just took  a little experimentation and tweaks in Cura. I'm sure I could probably get it better but it's good enough for what I need and I'm not on a mission to get +/- 1 micron, I don't have anything that measures that tolerance anyway. Hope you get to where you need.

     

     

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    Thank you for your answer 🙂 My experience in robotics tells me that it is not about mechanical inaccuracies. Mechanical inaccuracies will give you a bad first layer (with all the consequences), but not more. The dimension is given by the steps of the motor, diameter of the wheel (number of teeth) or the lead screw pitch. These are accurately machined parts. I'm sure the printer is much more accurate than what it prints. If you make a 10mm part and it is 10.2 you cannot assume a 0.2 mm inaccuracy in 1cm of belt (actually drive wheel). The only real mechanical problem with the printer is the filament drive mechanism with loooong tubes, but this also cannot explain all the errors. Probably the only problem is the fact that the Cura configuration files are not refined as they should be. The whole printer gives you the impression of a ready to use, push-one-button tool. You have a huge display that shows NO technical details while printing (this is really frustrating), there are long periods of time for "preparing" or other operations while you don't know what happens or if anything happens at all or you should simply restart the thing... You would expect everything is then refined to the last detail, you just press "print" in Cura and it works. Well, no! You have to tune a lot, it doesn't simply work. Again you would expect that from a DIY toy, like Prusa, but not from a "professional" printer. For what I paid I should have this work already done by someone else for me and I should not have to waste days of my time to finish someone else's work.

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    OK @mitu, well you have more experience than me in the mechanical side of life. My Prusa clone is a long way off in terms of parts quality compared to the ultimaker, so my experience differs from yours in that respect.
    I solved the inaccuracies by using the horizontal Expansion settings in Cura which applies settings to all the polygons that make up the model. I can only assume that having these settings is to mittigate the potnetial issues that I had and that you are also seeing. 

    I started out in this thread with the following issue...

    "Having printed a 20mm calibration cube, I get the following, Z=20.01 X=20.32 and Y=20.20. The X & Y being that much over causes parts in the model not to fit."

     

    So yes out of the box my X & Y were not what I wanted either, whether this is by design or an outcome of the manufacturing process, I don't know. However, the 'quality' of the print far exceeded what I could get from my Prusa clone. Only when I started to print parts that needed to fit (internal & external sizing) did I find the problem. It took me just a couple of hours of work to resolve it once I was pointed to the right place in Cura. My method was to create a 20mm cube and a 20mm Cylinder (open one end with 1mm walls). The Cylinder was done to be able to check external, internal and wall measurements. I printed them separately, measured, changed settings and printed again. Then repeated the process until I got my desired result. Finally I combined the cube and cylinder into a model where the cylinder overlapped a corner to create a more complex model. Dimensions were upheld when compared to the final test objects.
    Then I printed the multi part items and bingo, they fit just perfectly, little bit of interference but in this case that was fine as they needed to be held nicely and not loose.  I don't have part sizing issues any more.

    It can be done if you're willing to spend a couple of hours, if you still feel this is not the experience you were expecting then I suggest you take that up with Ultimaker or whoever supplies support for them in your region.

     

     

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration
    9 hours ago, LaserBrain said:

    OK @mitu, well you have more experience than me in the mechanical side of life. My Prusa clone is a long way off in terms of parts quality compared to the ultimaker, so my experience differs from yours in that respect.
    I solved the inaccuracies by using the horizontal Expansion settings in Cura which applies settings to all the polygons that make up the model. I can only assume that having these settings is to mittigate the potnetial issues that I had and that you are also seeing. 

    I started out in this thread with the following issue...

    "Having printed a 20mm calibration cube, I get the following, Z=20.01 X=20.32 and Y=20.20. The X & Y being that much over causes parts in the model not to fit."

     

    So yes out of the box my X & Y were not what I wanted either, whether this is by design or an outcome of the manufacturing process, I don't know. However, the 'quality' of the print far exceeded what I could get from my Prusa clone. Only when I started to print parts that needed to fit (internal & external sizing) did I find the problem. It took me just a couple of hours of work to resolve it once I was pointed to the right place in Cura. My method was to create a 20mm cube and a 20mm Cylinder (open one end with 1mm walls). The Cylinder was done to be able to check external, internal and wall measurements. I printed them separately, measured, changed settings and printed again. Then repeated the process until I got my desired result. Finally I combined the cube and cylinder into a model where the cylinder overlapped a corner to create a more complex model. Dimensions were upheld when compared to the final test objects.
    Then I printed the multi part items and bingo, they fit just perfectly, little bit of interference but in this case that was fine as they needed to be held nicely and not loose.  I don't have part sizing issues any more.

    It can be done if you're willing to spend a couple of hours, if you still feel this is not the experience you were expecting then I suggest you take that up with Ultimaker or whoever supplies support for them in your region.

     

     

    Hi @LaserBrain, I am seeing very similar issues.  My calibration cube measurements are almost identical to yours.  I am willing to put in the time to get the calibration correct.  Originally I was looking for a way to modify the stepper steps/mm as that is what I have done on other models but I don't see a way to send Gcode directly from Cura to the S5.  

    Can you tell me what settings in Cura you adjusted to improve the accuracy of the print?  I am not looking for +/- microns but 0.2 mm is not good enough for this level of machine. 

    Thanks!

  • Link to post
    Share on other sites

    Posted (edited) · Getting Better Dimensional Accuracy - Calibration

    Hi @mzcukier, sure the settings you need are in the 'shell' section of Cura and are called Horizontal Expansion. There are two settings, one for the first layer, this helps with 'elephant foot' problems, the second is for the rest of the model.

    Mine are set to -0.16 in both. A negative setting reduces the overall X/Y dimensions. Also attached are the STL files I created so you can check the outcome. the cube is 20mm all round. The Cylinder is 20mm, print with the base on the bottom, the walls should be 1mm thick giving an internal diameter of 18mm. When you have got your settings right, print the combined Cylinder and cube to see if something more complex upholds all the measurements. If so, you're done, go print a multi-part model and see if the parts fit.

     

    Enjoy!

     

    20mm_Cube.stl

    Combined_Cyl_Cube.stl

    Cylinder.stl

    Edited by LaserBrain
    • Thanks 1
    Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    link to my outcome from testing and setting changes... 
     

     

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    Hello.

    I am getting consistent Y axis bigger than X. I get at the calibr. cube 20.1 on X and 20.5 on Y. I've made a testing tool with holes (m3, m4, 22mm, 20mm square and a 8mm cylinder inside a 60mm on X 40mm on Y rectangle) and in every form de Y alignment gets bigger than in X.

    Deviations are:

    60 on X I get 60.15mm

    40 on Y I get 40.5

    The 20mm square on X 19.56mm and on Y 19.75

    The 22mm circle I get 21.40 mm diameter on X and 21.63mm on Y.

    The 8mm cylinder has a X diameter of 8.1mm and Y 8.4mm.

     

    When I try to print a bearing with a higher tolerance it gets visually perceived a oval form.

     

    I managed to put a needle in the print head and used a good scale, all steps seem perfect all the way in the axis distances.

     

    I think there are 2 controllable variables in play: horizontal expansion and one axis printing bigger than other, but I need to make one of them right so I can approach the other.

    Can you help me?

    Maybe a fixed % compensation in one axis and then start to handle horizontal expansion?

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    What printer is that?

  • Link to post
    Share on other sites

    Posted (edited) · Getting Better Dimensional Accuracy - Calibration

    UM2E

    bondtech DDG feeder, 3D solex heatblock, 0.4 nozzle

    Basic .2 draft set on Cura, nozzle temp 200. All basic settings.

    Edited by my3DBr
  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    @Torgeir, what do you think?  

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    Hi there.

     

    I would check that the two 6 mm shafts is absolute parallel with the two rotating 8 mm shafts it is moving in between.

    I.E. The shaft going across (if you look into the printer) is the X shaft as head go along this shaft. 

     

    There is a tool you can print (we'll need two of them), that can be "clamped" to the (8 & 6)mm shafts in order to fine adjust the pulleye position by unscrew the two setscrews.

     

    To just have an idea of how much off it is, use a caliper and measure between the shafts, -or any fixed object that can be used as a metering device to see that the relative distance is the same. There is 8 places to check..

     

    Here's the Ultimaker site for this problem.

     

    https://support.ultimaker.com/hc/en-us/articles/360017142879-How-to-align-the-axles-on-your-Ultimaker-printer

     

    Thanks

    Torgeir

     

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    Hello @Torgeir. I’ve already done the axis alignment  with those tools, I thought could be it. Nothing changed.

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    This printer was someone’s else before. I did changed pulleys, belts, heat block, feeder.

    Belts are from robotdigg, pulleys I could not check proof of quality.

    I’ve just bought pulleys and belts again from one single supplier from UK (Technobots).

    But if there was to be any problem related to it, the steps should present any kind of deviation, which they don’t.

    I done all of it with help from a more experienced user, using the officials manuals (which present some errors by the way).

    If I could travel through time, I would not change the original pulleys and belts, even with hundreds of printing time. I kept them stored and looking at them nowadays they don’t seem in bad shape.

    The problem was a rusted y axis rusted stepper motor, which I opened and treated.

    I’ve bought new ones from Pololu but did not changed them - lack of personal energy. It is a difficult printer to work on. And the stepping is not altered.

    Well, That’s it. 
    Maybe I can just search for a % percentage os compensation on y axis….

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration
    5 hours ago, my3DBr said:

    Basic .2 draft set on Cura

    Thats not the best profile to test dimension accuracy, please use the normal one.

    It also helps when you equalize all the 7 speed settings (inner, outer, wall, infill, ...) to the same value. Use a speed of 40mm/sec for your tests.

     

     

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration
    1 hour ago, Smithy said:

    Thats not the best profile to test dimension accuracy, please use the normal one.

    It also helps when you equalize all the 7 speed settings (inner, outer, wall, infill, ...) to the same value. Use a speed of 40mm/sec for your tests.

     

     

    Ok, I’ve read before your exact text about the accuracy issue, the “engineer profile” which seems to solve a lot of people’s problems.

    I think it is something I must try - but; there is a discussion I would like to bring: every time I need some sort of accuracy depend on a longer printing time is the best solution? Or, will it solve the difference between Y axis and the other two?
    Let’s say I am using LouBan to edit a mesh of a lumbar spine, and therefore it being a natural element it brings a form that is not “friendly printing format”. I have to cut it in half, the LuBan application automatically does it, inserting a dove for posterior reunion of both pieces.

    I’m having a very troublesome time fighting to it. I never needed accuracy before, never thought this difference between the Y axis and the other two leave me in a not comfortable situation.

    But yes. I should do as you said. And thank you a lot for your participation. 

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    I'll agree with @Smithy as there is better setting for better precision, but I'll think there is more mechanical to check, -cause this is a little to much of an error -IMO.

    So I would like to know how the two sliding bearing inside the head is, -or simply do you have ANY play here?

    You can check this easily without power up the printer. Try to move the head fore and aft (Y axis) without moving the head, if there is any play you will easily feel it. Do the same with the X axis.

    The ting is there can be play that only occur at some places on the bed -but feels OK elsewhere.

    Also, are the two 6 mm axis absolute strait?

    When try to tune the printer up to specs, the above is for sure a must. Been there done that.

     

    Thanks

    Torgeir

     

    There is more, but what about this?

     

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    I greatly agree with you @Torgeir. Too much of an error to be only printing profile (assuming we leave dimensions equal on Cura, off course).

    Both sliding  bearings are original, although only recently I’ve read on a official paper that there is a ideal position for them - the trails os the bearings positioned as an “X” but I don’t think this is the problem.

    I checked and rechecked for any play that would explain it, and could not find any. Not in the print head from the crossed axles nor in belts and pulleys. Sincerely I did not checked the printhead in various positions. 
    Both 6mm print head axis are new, tested and checked for straight accuracy. Bought at Robotdigg. Even more, the difference between axis sizes in printings are reproducible in any place on the bed.

    I am getting close to affirm that this is a mechanical issue and dismounting and replacement a bunch of pieces not knowing what is happening. This is terrible. Or I can try to find a % to reduce from Y axis every time I print.

     

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    I also agree that there is probably a mechanical reason for the difference, but I would still try to print a small part with another profile. Then we could see how big the difference is between x and y. 

    The draft profile is really bad for such things and only useable for rough "drafts" 🙂 

     

     

  • Link to post
    Share on other sites

    Posted · Getting Better Dimensional Accuracy - Calibration

    On to it, tonight! (Brazil time).

  • Link to post
    Share on other sites

    Create an account or sign in to comment

    You need to be a member in order to leave a comment

    Create an account

    Sign up for a new account in our community. It's easy!

    Register a new account

    Sign in

    Already have an account? Sign in here.

    Sign In Now

    • Our picks

      • Introducing Universal Cura Projects in the UltiMaker Cura 5.7 beta
        Strap in for the first Cura release of 2024! This 5.7 beta release brings new material profiles as well as cloud printing for Method series printers, and introduces a powerful new way of sharing print settings using printer-agnostic project files! Also, if you want to download the cute dinosaur card holder featured below, it was specially designed for this release and can be found on Thingiverse! 
          • Like
        • 10 replies
      • S-Line Firmware 8.3.0 was released Nov. 20th on the "Latest" firmware branch.
        (Sorry, was out of office when this released)

        This update is for...
        All UltiMaker S series  
        New features
         
        Temperature status. During print preparation, the temperatures of the print cores and build plate will be shown on the display. This gives a better indication of the progress and remaining wait time. Save log files in paused state. It is now possible to save the printer's log files to USB if the currently active print job is paused. Previously, the Dump logs to USB option was only enabled if the printer was in idle state. Confirm print removal via Digital Factory. If the printer is connected to the Digital Factory, it is now possible to confirm the removal of a previous print job via the Digital Factory interface. This is useful in situations where the build plate is clear, but the operator forgot to select Confirm removal on the printer’s display. Visit this page for more information about this feature.
          • Like
        • 0 replies
    ×
    ×
    • Create New...