Extrusion calibration impossible steps/mm. Very rare.
this is really a strange thing...
I am very experimented in 3d print but this subject is impossible understand for me
When i calibrate the Extrusion step/mm for that i move 100mm the extruder move 100mm i must adjust in 643 steps.
With 643 steps i test
-- Move 100mm without hot end-> Yes move 100mm +/-0.25mm
-- Move 100mm with hot end with PLA or PETG-> Yes move 100mm +/-0.25mm
-- Test in diferentes speeds, 1-2-4 and 5mm/s. Always is OK
-- Test with nozzle 0.6 and 0.8. Always is OK
Clarify this i put a print and ALWAYS i have a BIG over extrusion.
How is possible?
If a test a cube with only 1 outline and i measure the wall have +6/7% over extrusion.
When is a 100% infill the result is a weird...
If i am testing to down the steps/mm WHILE is printing i see that with 605-608 seem ok, but not is sure. Is better but not perfect, but of course is really better 605 than 643...
How is possible this? I test with 2 duet3D and happen the same... diferent versions and happen the same....
My extruder is DyzeXtruder GT
The website say 578, if i use this is really under-extrusion!
Somebody have this? What steps/mm are you using?
Some idea about this?
I also want add that i test 4 DyzeXtruder GT and happen the same in the four...
I’ve noticed the same. I’m new to duet but not to 3d printing and this has me a little perplexed. Ended up reducing the extrusion multiplier to 90 and reducing the steps to get the width correct. What slicer are you using? Absolute or relative extruded moves?
NitroFreak last edited by NitroFreak
whats your filament diameter like? It could be that if it´s too big. the linear distance is correct (the 100mm) but since the filament is thicker, it creates volumetric overextrusion. Independent of Duet, Dyzextruder, etc.
@mwolter always use simplify3d (with most update version).
Relatives, i never used absolutes
@nitrofreak i measure a lot times, the filament is from formfutura and have 1.75 and some times 1.74. But never 1.76 or more.
I use simplify3d also and also use relative extruded positioning. I have a feeling the over extrusion is caused by an improper extrusion volume calculation in the slicer. I’ll try using another slicer. Would like to use Prusa Slicer but not sure it supports setting the extruded temperature via G10 commands.
Not sure but, I believe another thing that could contribute to this is if the Z steps are somehow out of calibration. I’ll double check this but I believe this is accurate on my printer.
I've been thinking about this the past couple days and your post prompted me to print a single wall cube and calibrate this again... I might have a little info to help with this.
First, are you using non-linear extrusion? Reviewing my notes when initially configuring my printer and found with non-linear extrusion enabled and calibrated, using a 0.4 nozzle and S3D set at 0.4 extrusion width at 0.2 layer height, single wall extrusions measured 0.52 to 0.54 using calipers. I then disabled non-linear extrusion and extrusion widths were much closer to what was expected (0.43 to 0.44mm).
Second, reading other posts, specifically on the extrusion width setting for S3D, I found that even though you have a 0.4 nozzle, the extrusion width is always a little more. Sometimes up to 1.2x the nozzle width. That's why when you click the auto checkbox next to extrusion width it automatically enters 0.48 when you use a 0.4 nozzle. If I'm understanding this correctly, when your extruder steps/mm is properly configured and you're using 100% extrusion multiplier, the actual extrusion width will always be slightly larger than your nozzle diameter. I tested this by printing two single wall cubes using the same 0.4 nozzle and 100% extrusion multiplier, one with extrusion width at 0.4 and another at 0.43 and both cubes measured 0.43 wall thickness. So as far as I can tell, changing the extrusion width doesn't actually change how much filament is extruded. It only changes the spacing between extrusions.
Please keep in mind that this was only tested with two cubes but I think it's accurate.
@mwolter i do a test few minutes ago, i slicer the same line in S3D, Prusa Slicer and Cura.
A line of 100mm with 0.3 layer, nozzle 0.6 and line of 0.75 have the follow results:
Prusa Slicer: E8.48729
9.28425/8.48729= 9,3% MORE in S3D and Cura. OverExtrusion??
One question independent of this:
---- Is possible that when Duet3D read 9.2843, really move 10 or 9.5?
@mwolter I test a lot parts... and i know that you say about the auto width.
@fbg For me, I had severe under extrusion with Prusa Slicer. I like many of the new features of prusa slicer and was going to look into it's under extrusion issue but since it doesn't properly support changing temp gcode from M104 to G10 for the extruder so I stuck with S3D.
callielli last edited by
@FBG I have a DyzeXtruder on a test rig I have build for PEEK. I'm working through the setup calibrations and noticed peculiar results vs. what I was expecting to see also.
Have you had any success in resolving your issues?
@callielli Hi, not, nothing resolve. The problem continue, i don't know if is problem of S3D or that...
The single wall thickness test is not a good test IMO because a line of filament that is not constrained on either side will spread out. What I do when trying a new filament is to adjust the extrusion factor in DWC to get good top infill. Usually I need between 95 and 100%. Then I always use that extrusion factor with that filament.
i don't test a single wall, i test cubes with 100% of infill.
Are you using S3D?