I have a #Printrbot Metal Simple with the Z-probe. At the start of a print, I run a G29 command to probe the bed level and determine the bed plane. This reports values in the form.
< 11:34:56 AM: Bed x: 10 y: 142.40 z: 0.42
< 11:35:04 AM: Bed x: 10 y: 10 z: 1.63
< 11:35:12 AM: Bed x: 142.40 y: 10 z: 1.72
These values will vary +/- 0.05 mm between calibration runs.
This works well enough for most prints but are not quite repeatable enough for print in place designs with very tight tolerances or low layer height prints where a 0.1mm height difference between corners may cause the print to fail.
I would like to manually set the z plane by setting the bed plane in Gcode at the start of the print. ie. Manually set the z value for 3 points as above. Any hints on what gcode to use to do this?
Is it possible to have the bed probed at more points? I noticed that the Afinia leveling scripts tests 9 points. Code suggestions?
The very newest Marlin supports the G30 gcode, which allows you to probe a single spot on the bed. Though I’m not sure if it then uses that information to calculate the surface plane.
The decision whether to probe three spots or many more in a square grid is made in the firmware’s “auto bed leveling” configuration, which i explain in detail here: https://www.youtube.com/watch?v=EcGFLwj0pnA Grid vs. three-point at 5:54.
So to change the mode your firmware chooses, you’ll have to edit and reupload your printer’s firmware.
-Also, what slicer are you using? Every slicer except cura makes a fatter bottom layer which helps with adhesion and helps alleviate minor imperfections causing your entire print to fail.-
@ThantiK , Cura has had a setting for first layer extrusion width since July 3rd (version 14.07)… The feature request was submitted by @Whosa_whatsis .
I’m using Cura. This isn’t really an issue with normal prints but if I want to start playing with smaller nozzle sizes and lower layer height, I would like to be able to set the bed plane values manually rather than rely on z-probe hitting same values each time. Someone at Maker Faire NY said they took this approach but couldn’t remember the gcode commands.
When I record z heights for 3 leveling points on multiple runs I get these sorts of values.
Run 1: 0.35 , 1.56 , 1.61
Run 2: 0.44, 1.66, 1.69
Run 3: 0.42, 1.63, 1.72
Run 4: 0.43, 1.62, 1.71
(I have gone through and tightened as much as I can to reduce any wobble in the bed.)
These fluctuations on 0.1 mm are sometimes enough to cause one of the sides of the print to be either just a little too high or a little too low. I don’t know if this amount of error is typical for these probes.
Would like to manually feed values that I know work well.
@Matthew_Satterlee , https://github.com/daid/Cura/issues/769 – Last comment was that you had to manually go in and edit util/sliceEngine.py in order to actually enable this? I haven’t seen it in the UI as of late…am I just dumb and blind?
Stupid question: do you know for sure that the variation is in the probe? You’re asking the bed bearings and the cantilevered Z assembly to have slop of less than 2 thicknesses of a human hair. Even things like the weight of filament above the extruder could have measurable effects on that order.
