Notifications
Clear all

Most dimensionally accurate FDM prints under 1k?  

  RSS
jerry.m7
(@jerry-m7)
New Member
Most dimensionally accurate FDM prints under 1k?

hello,
Howdy!

I've been out of the 3d printing game for a couple of years now (minus ordering stuff on shapeways) and I've had a bunch of ideas stewing that I want to try out.

I picked up a Monoprice Mini Delta a month ago but I (and many others) have been having a ton of trouble getting accurate-enough prints out of it. It's been great for cosmetic prints, but not functional.

A lot of things that I want to print are adapters between different products (ex: an adapter for a water filter for hiking) that require pretty precise tolerances.

I feel like I have two choices, and would love your help deciding on the right one.

Buy a new printer that is more accurate. From what I read, the Prusa i3mk3 is my best bet, but maybe there's something better? Seems like some of their auto-leveling/screw correction could result in really accurate prints?

Use some sort of software to calibrate my existing printer. I feel like I should be able to print some sort of calibration model, measure it with calipers and then feed those measurements into my slicer. Then, it would manipulate the model to print 'accurately' without the printer itself having to do anything different.

Any suggestions or help would be greatly appreciated! I'm still doing my best to make the MP Mini Delta work for me since I love the idea of such an affordable and compact printer, but a lot of people more talented than I are struggling with this issue so I don't have high hopes for this particular version.

Tutuapp ApkTextNow ApkPhotomath Apk
Napsal : 10/07/2018 3:38 pm
dimprov
(@dimprov)
Noble Member
Re: Most dimensionally accurate FDM prints under 1k?

Interesting question!

Maybe try printing this tolerance test on your existing printer and then see how it compares to what others report for their printers?

If there's a better test, maybe someone can post what it would be?

or, maybe this?
https://www.thingiverse.com/thing:533472

Please do let us know what you come up with.

Napsal : 11/07/2018 12:11 am
arthur.c2
(@arthur-c2)
Trusted Member
Re: Most dimensionally accurate FDM prints under 1k?

I made test parts on my MK3 to correct settings and now my parts are precise to 0.01 mm.

Stock MK3 is precise to 0.1mm or less but I wanted to reach the limits.

I don't have a delta printer but they are not known for their precision, especially a cheap one.

Whatever try some test with your printer to see what precision you can reach (there are a lot on thingiverse).

Personally I didn't like a test that I found so I made my own test part and an excel file to determine corrections settings, it works well, I should post it on thingiverse

Napsal : 11/07/2018 11:35 am
dimprov
(@dimprov)
Noble Member
Re: Most dimensionally accurate FDM prints under 1k?


I should post it on thingiverse

^This, and a link here to it. 🙂

Napsal : 11/07/2018 2:59 pm
arthur.c2
(@arthur-c2)
Trusted Member
Re: Most dimensionally accurate FDM prints under 1k?

I will try to do it this evening 😉

Napsal : 11/07/2018 3:37 pm
bobstro
(@bobstro)
Illustrious Member
Re: Most dimensionally accurate FDM prints under 1k?


[...] Buy a new printer that is more accurate. From what I read, the Prusa i3mk3 is my best bet, but maybe there's something better? Seems like some of their auto-leveling/screw correction could result in really accurate prints?
The Prusa i3 Mk3 is very accurate once calibrated. You can expect to print parts with 0.2mm clearance and have them function. I've printed the Maker's Muse calibration test, various screws and bearings, and generally things printed with 0.2mm clearance will turn. Considering I just got into the game, that's impressive. Some of the 0.15mm parts almost turn, but you have to really work on them with a knife afterwards to get parts turning. Dimensionally, you should be able to get easily within 0.1mm, but you have to calibrate for the specific filament you're using if you need it extremely precise.

Of course, you need to generate the models to print appropriately. Fusion 360 has some great tools for producing usable threads with a minimum of fuss. These have worked well for me when exported to STL and printed on my Mk3.

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Napsal : 11/07/2018 3:52 pm
dimprov
(@dimprov)
Noble Member
Re: Most dimensionally accurate FDM prints under 1k?

That's good, but 0.01mm is a whole different level. Even the Maker's Muse guy didn't get close to that.
I'm really curious as to how Arthur.c2 did it.

Napsal : 11/07/2018 4:27 pm
dimprov
(@dimprov)
Noble Member
Re: Most dimensionally accurate FDM prints under 1k?

As a follow-up, I printed this tolerance test: https://www.thingiverse.com/thing:2375342
and with no effort at all, it worked on all of them. 😀 It's tightest tolerance was 0.1mm, so I presume my I3 MK3 might do even better than that. I'll have to search for some other test to find out what its actual limits are.

I printed it in PETG at a speed of 50mm/sec, 0.48mm width, 0.1mm layer height, 20% infill.

[Edit: Then I printed this tolerance test:
https://www.thingiverse.com/thing:2750861
and it could do the 0.1mm test, but failed the 0.05mm test. So, with no real effort, the tolerance would appear to be somewhere between 0.05mm and 0.1mm.]

Napsal : 12/07/2018 8:03 pm
dimprov
(@dimprov)
Noble Member
Re: Most dimensionally accurate FDM prints under 1k?


I made test parts on my MK3 to correct settings and now my parts are precise to 0.01 mm.

Stock MK3 is precise to 0.1mm or less but I wanted to reach the limits.

I don't see how 0.01mm accuracy is even possible. When I print a calibration cube and measure the distance from one side to the other, I get measurements which vary by more than that amount on the same cube, just depending on where on the surface face I measure it at: https://shop.prusa3d.com/forum/general-discussion-announcements-and-releases-f61/what-are-the-expected-tolerances-for-x-y-and-z-len-t23009.html

I'm not saying you're wrong. I'm just saying I don't see how you can be right.

Napsal : 13/07/2018 7:46 pm
stoofer
(@stoofer)
Estimable Member
Re: Most dimensionally accurate FDM prints under 1k?

Delta's are notoriously hard to calibrate for dimensional accuracy as, due to allmovements being a combination of three axis, the whole build volume really needs to be calibrated, rather than a small central spot - often moving the test cube to be very near one of the three struts will result in quite a different print - whereas on a cartesian printer like the Prusia, generally what the X Y and Z are doing for a small area of the overall volume will extrapolate out for the whole area.

You can made a Delta accurate within it's entire volume, but it's hard.

Napsal : 13/07/2018 8:04 pm
bobstro
(@bobstro)
Illustrious Member
Re: Most dimensionally accurate FDM prints under 1k?


As a follow-up, I printed this tolerance test: https://www.thingiverse.com/thing:2375342
and with no effort at all, it worked on all of them. 😀 It's tightest tolerance was 0.1mm, so I presume my I3 MK3 might do even better than that. I'll have to search for some other test to find out what its actual limits are.
I can't get past 0.2mm, although I did print at 0.2mm layer height. I haven't spent much time tuning for it though, so I'm not overly worried.

Just remember that it doesn't count if you're using Cura to slice!

I printed it in PETG at a speed of 50mm/sec, 0.48mm width, 0.1mm layer height, 20% infill.I'll give 'er a go with those settings.

My notes and disclaimers on 3D printing

and miscellaneous other tech projects
He is intelligent, but not experienced. His pattern indicates two dimensional thinking. -- Spock in Star Trek: The Wrath of Khan

Napsal : 13/07/2018 8:50 pm
arthur.c2
(@arthur-c2)
Trusted Member
Re: Most dimensionally accurate FDM prints under 1k?


As a follow-up, I printed this tolerance test: https://www.thingiverse.com/thing:2375342
and with no effort at all, it worked on all of them. 😀 It's tightest tolerance was 0.1mm, so I presume my I3 MK3 might do even better than that. I'll have to search for some other test to find out what its actual limits are.

This test is not a tolerance test for me but an assembly printing test, to see what clearance you should let in your models to print functional assembly. And that's a good test for that.

When printing the deposit is hot and "liquid" so without enough clearance, parts can be bond, but when getting cold there is rectractation and final dimensions are reached. I always let my parts get cold before measuring.

If your printer is well calibrate you should be able to assemble males and females parts printed separately with a cleareance of 0.05.

Moreover I always add a radius (don't know the english world) on male parts for the angles because printed sharp corners are not sharp and false your measurement of the surfaces.



I don't see how 0.01mm accuracy is even possible. When I print a calibration cube and measure the distance from one side to the other, I get measurements which vary by more than that amount on the same cube, just depending on where on the surface face I measure it at: https://shop.prusa3d.com/forum/general-discussion-announcements-and-releases-f61/what-are-the-expected-tolerances-for-x-y-and-z-len-t23009.html

I agree with those variations, saying that I'm precise to 0.01 is a little optimist. What sounds honest is to say that I can reach +/- 0.02mm, which is a really appreciable tolerance for me with a desktop machine.

I tried my calibrating process again because I did it when receiving my printer months ago. My caliper is tired and bug for dimensions over 80mm so I will buy a good one and do the test again, then I will post the process.

Whatever it is possible to set your softawre and printer to assemble parts printed in nominal dimensions (without adding clearance on your model).

Napsal : 13/07/2018 10:22 pm
Share: