User avatar
david.f42
Posts: 537
Joined: Sat Feb 17, 2018 4:53 am
Topics answered: 4
Contact: Send private message

What are the expected tolerances for x,y, and z lengths?

Fri Jul 13, 2018 2:56 pm

I printed a 20mm test cube: https://www.thingiverse.com/thing:1278865
and then measured it with a micrometer.

X measures 19.81mm
Y measures 19.78mm
Z meaures 20.09

That last digit varies depending upon where on the cube face that I measure it.

Is this the expected tolerance?

I printed the test cube in PETG with speed 50mm/sec, 230C, layer height 0.1mm, calibrated extrusion width 0.48mm, using a 0.4mm nozzle.

Given that, I don't see how I can do any better on Z. I'm not sure how much shrinkage plays into this for X and Y. Does a slicer typically try to compensate for shrinkage automatically, based on the material, or is that something I'm expected to do myself?

According to Matter Hackers, "PETG has a shrink ratio (or shrink rate) of less than 0.004 in/in", so both X and Y seem to measure out a bit short regardless of shrinkage.
Last edited by david.f42 on Fri Jul 13, 2018 4:25 pm, edited 1 time in total.

bobstro
Posts: 367
Joined: Sun Dec 31, 2017 2:28 am
Topics answered: 3
Contact: Send private message

Re: Should I attempt to calibrate x,y, and z lengths?

Fri Jul 13, 2018 4:24 pm

Using Mad Maker's PLA with a 1.0 extrusion factor (and the M221 code removed from Slic3r gcode), I get:

X 20.11 (top) 20.00 (center) 20.03 (bottom)
Y 20.11 (top) 20.14 (center) 20.12 (bottom)
Z 20.11 (left) 20.13 (center) 20.13 (right)

I think these are pretty typical ranges. I have not tried calibrating to 0.1mm accuracy yet, but I know you need to measure your filament and possibly tweak extrusion. Slic3r does have Print Settings->Advanced->Other->X-Y Size Compensation, although that will applied to all filaments.

User avatar
david.f42
Posts: 537
Joined: Sat Feb 17, 2018 4:53 am
Topics answered: 4
Contact: Send private message

Re: Should I attempt to calibrate x,y, and z lengths?

Fri Jul 13, 2018 4:49 pm

bobstro wrote:
Fri Jul 13, 2018 4:24 pm
Using Mad Maker's PLA with a 1.0 extrusion factor (and the M221 code removed from Slic3r gcode), I get:

X 20.11 (top) 20.00 (center) 20.03 (bottom)
Y 20.11 (top) 20.14 (center) 20.12 (bottom)
Z 20.11 (left) 20.13 (center) 20.13 (right)

I think these are pretty typical ranges. I have not tried calibrating to 0.1mm accuracy yet, but I know you need to measure your filament and possibly tweak extrusion. Slic3r does have Print Settings->Advanced->Other->X-Y Size Compensation, although that will applied to all filaments.
What speed did you print at?

bobstro
Posts: 367
Joined: Sun Dec 31, 2017 2:28 am
Topics answered: 3
Contact: Send private message

Re: Should I attempt to calibrate x,y, and z lengths?

Fri Jul 13, 2018 4:55 pm

david.f42 wrote:
Fri Jul 13, 2018 4:49 pm
What speed did you print at?
Slic3rPE defaults with 0.2mm layer height.

User avatar
david.f42
Posts: 537
Joined: Sat Feb 17, 2018 4:53 am
Topics answered: 4
Contact: Send private message

Re: What are the expected tolerances for x,y, and z lengths?

Fri Jul 13, 2018 9:44 pm

At present I'm using Simplify3D, which doesn't know about linear advance, so for now it's turned off.

The test cube I reported above had only 5% infill and 0% overlap between the infill and the outer perimeter. As a result, I think the test cube may have slightly imploded.

I made some adjustments that are improving the results. The changes I made were:
100% infill
99% overlap between infill and outer perimeter (that's as high as it goes). I'm not sure what value slic3rPE uses for this, as it does't seem to be exposed.

Doing that, I printed a new cube and took 5 measurements for each dimension and averaged:
X: 19.97, 19.92, 19.98; 19.96, and 19.96. Average: 19.96
Y: 19.98, 19.97,19.92,19.98, and 19.99. Average: 19.97
Z: 20.14, 20.09, 20.11, 20.09, 20.09. Average: 20.10

I can't really complain about these numbers. The X and Y do show some shrinkage, which is consistent with what matterhackers said about PETG. Z is going to be affected by how well I do the live Z, and I'm not sure how to be exact about that.

So, yeah, tolerance of plus or minus 0.1mm sounds about right.

Would the expected tolerance be worse for a larger test cube? That I don't know. Is tolerance best expressed as an absolute maximum error in millimeters, or as a percentage?

User avatar
RHDreambox
Posts: 329
Joined: Mon Oct 23, 2017 8:38 am
Location: Sweden
Contact: Send private message

Re: What are the expected tolerances for x,y, and z lengths?

Sat Jul 14, 2018 12:15 pm

Slic3r PE with default settings for PETG 0.2 mm layer height.
It is difficult to measure exactly as there are different measurement values each time (100th of a mm)
But I am very pleased with the tolerances. They are better than I had expected from a "hobby printer".

X = 19.98 in the middle (20.04 average)
Y = 20.04 in the middle (20.02 average)
Z = 20.03 in the middle (20.04 average)
Happy owner of a Prusa MK3

Return to “General discussion, announcements and releases”

Who is online

Users browsing this forum: corey.d3, Dreide, ismaelab, richard.l and 13 guests