Tuning dimensional accuracy

[bdejong11129]

As part of my update project for one of my Voron's I have been struggling to get it to product dimensionally accurate parts.  It has spent the past few years printing brackets and kids toys so it didn't need to be perfect.   

I did start with the basics of the Ellis tuning guide to get that all reset.  I also tuned in bridging and the hole dimensions so that they are printed correctly.  Now my issue is poor results from Califlower for X and Y.   My last run had a correction of 0.16% for X and 0.35% for Y.  Since these two axis are tied together I am at a loss as to how to adjust them separately so that I can get the prints to come out correctly.   Help an old man out here. Please...

[mvdveer]

29 minutes ago, bdejong11129 said:

As part of my update project for one of my Voron's I have been struggling to get it to product dimensionally accurate parts.  It has spent the past few years printing brackets and kids toys so it didn't need to be perfect.   

I did start with the basics of the Ellis tuning guide to get that all reset.  I also tuned in bridging and the hole dimensions so that they are printed correctly.  Now my issue is poor results from Califlower for X and Y.   My last run had a correction of 0.16% for X and 0.35% for Y.  Since these two axis are tied together I am at a loss as to how to adjust them separately so that I can get the prints to come out correctly.   Help an old man out here. Please...

I have solved most of my dimensional inaccuracies with flow calibration/Extrusion multiplier and skew correction. If you have followed the Ellis guide, then you would have flow/EM  sorted. Skew calibration with Califlower is well worth the money you pay for it. It makes a difference. 

[bdejong11129]

I have all of the califlower versions, even the latest Calilantern.  The skew was easily corrected, the hole dimensions were also easily resolved.  The main issue I am having now is that certain parts just don't fit.  The main culprit is the [a]_guidler_a and [a]guidler_b parts for the Stealth Burner.  These two parts should just snap together but I have to almost press them in place after some shaving. 

 

All versions of Califlower and Calilantern show a difference between X and Y. However, there is not individual adjustment for x vs y.  

[mvdveer]

The only other thing I can think off that may affect it is shrinkage, in the filament settings of the slicer. You can adjust XY seperate from Z. These are normally set at 100%. Mine for e-Sun ABS is set at 101% and for ASA is at 99.6%. Took a bit of filling and lots of printing dimensional test pieces from a very old file on thingiverse. I am sure there are better and newer ones out there.

[Penatr8tor]

12 hours ago, bdejong11129 said:

All versions of Califlower and Calilantern show a difference between X and Y. However, there is not individual adjustment for x vs y.  

Neither your printer nor any one of ours nor anyone else's will print exact dimensions every time. That leads me to the next question... how much of a difference? There is an acceptable range... generally .01-.02mm so if you're numbers are in that range you're good. Even if they vary by a .02-.04 you're still not bad. As far as parts fitting too tightly... dropping the extrusion multiplier 0.1-0.2 might help. FWIW... I did the Calilantern skew correction thing a while back. Since then, I've just put things together as best as I could and run them. Although there's nothing wrong with being thorough, I haven't found that particular calibration to be a big game changer really and that's why it's not on many peoples' radars. If I print a Voron cube and I'm within .01-.02mm of the 30mm dimension... I call that good.

[Mr.Shaky]

I've also done the calilantern hokey-pokey....and it's just another tool for working with a prototyping device.  It'll get you in the ballpark along with some manual adjustments, but don't expect extremely high accuracy or reproducibility.

Amusingly, there's a (new) Maker's Muse video on the broader accuracy/quality topic; food for thought: