I'm trying to adjust calibration cubes to get as close as possible to exact measurements.

my latest print (should be a 20mm cube) gives me:

X: 20.1168
Y: 20.0152
Z: 19.7866

What is acceptable?  Should I be shooting for 1/10 mm, 1/100 mm, or 1/1000 mm?  What's realistic?

Using a modded Ender 3.  At this point the only thing original is the power supply, frame, wheels and pulleys.

Just read an article suggesting 100 microns is the default assumption for accuracy.

I'm a LONG way from 100 microns... lol.

I guess it's time to default all the settings, run a new calibration cube, and start adjusting from there.  I was hoping I could fix it with some minor scaling in the slicer rather than diving in, but that's not to be (that's how I ended up with this latest one, which is slightly better than the last, but not much).

This hobby takes more patience than reloading, I swear it.

You're not using some cheap calipers are you?

Well your using a calibration cube, that is part of your problem. There are much better options available for calibration now.

It really depends on your printer and settings.

Like my Prusa I3MK3 can do under .001" in the XY. Mind you XY is not just dependent on XY travel but also your extrusion amounts. Too much extrusion it will be bigger, not enough it will be smaller. Z is linked to your layer height and first layer settings. So if you have like a .3mm first layer and a .2 layer height and your trying to make a 10mm cube, you will only be able to make a 9.9mm or 10.1mm cube nominally, most slicers underside so most likely the 9.9mm.

should be fine as long as hes not measuring a Glock 10MM chamber

Here's the deal.

They aren't going to be super duper precise.

If you have your extruder calibrated and the flow rate dialed in you'll make parts just fine. (don't measure walls to adjust flow)

At minimum run through the built in calibrations on OrcaSlicer.

It's based on this tuning guide https://ellis3dp.com/Print-Tuning-Guide/

FPNI.

I printed a third (scaled) cube tonight, same file as the second cube so I could observe any variation.

I learned two things:
1. I was using the calipers wrong.  When I tried to place calipers across the middle of the part, it was getting inconsistent readings.  But if I place them edge to edge instead, the measurement is highly repeatable.
2. once I figured out to measure the edges rather than placing them across the middle, I realized the first part I printed is between 0.787" and 0.788" for each axis.  Which is as close as you'll ever get with .001 resolution calipers to 20.000 mm.

This is probably the first time I've used the printer in like 6 or 8 months, turns out I already had it dialed in really well.

That makes me want to shell out the 5 squiggly Ls to get that .stl

I have to grab / buy that stl once I get my next printer up and going.
Thanks.

https://www.printables.com/model/242018-corexy-calibration

https://www.printables.com/model/246935-100mm-calibration

thanks, I'm considering checking that model out.

my print test yesterday went really well, so I'm on the fence whether to attempt any further calibrations.  Beyond 60 degrees, it gets a bit stringy on the bottom, but that's fine, I'll use supports at 65 and I think it will be good.

Really pleased with this filament.  Polymaker PLA PRO

Attached File

Attached File

$7 to find out my printer is 90.02° for skew and within 0.02% XY, which is probably more related to material shrinkage than issues with the printer.

I'll take it.

(For anyone wondering, it's a pretty slick part that comes with an Excel calculator that spits out correction info, and a nice PDF that details what it all means and how to implement it in your printer - worth $7 to me.)