Nozzle expansion offset gets reset by an order of magnitude at the start of a print. VC4/500
I've gone through all the Beacon calibration and it comes back with a sane number. (e.g. T0 expansion coefficient 0.069375) and then when starting a print the Z offset will be wildly off either + or - by millimeters.
So the prime blob will start in mid air, or in this case, it drove the nozzle into the textured PEI and the extruder was clicking because it was trying to extrude inside the PEI surface.
I haven't tried setting the expansion to false yet.
4:09 p.m. RatOS: Printing... 4:09 p.m. RatOS | BEACON: Nozzle expansion offset of -1.231562mm applied to T0 4:09 p.m. RatOS | Priming: Bridging with 40% fan speed.. 4:09 p.m. SET_GCODE_OFFSET Z=0 MOVE=1 <--- This is me clicking the CLEAR button to remove the -1.2mm Z offset while the hotend is being driven into the PEI and the extruder clicks 4:09 p.m. RatOS | Priming: Starting prime blob.. 4:09 p.m. RatOS | Priming: Moving to 495.0, 5 along the edge of the print area.. 4:09 p.m. RatOS | Priming: Lifting Z to 50.0.. 4:09 p.m. RatOS | BEACON: Nozzle expansion offset of -1.231562mm applied to T0 4:09 p.m. RatOS | Priming: Priming nozzle with prime blob.. 4:09 p.m. RatOS | TEMPERATURE_WAIT: Temperature for toolhead T0 reached. 4:08 p.m. RatOS: please wait... 4:08 p.m. RatOS | TEMPERATURE_WAIT: Waiting for sensor: extruder, MINIMUM: 250, MAXIMUM: 255 4:08 p.m. RatOS: Heating Extruder... 4:08 p.m. RatOS | Adaptive Mesh: Saving offset adjustment of 0.06343749968021539 in probe_for_priming_end_result 4:08 p.m. Last reading: 5474223.85Hz, 41.99C, -infmm 4:08 p.m. Result is z=0.063437 4:08 p.m. probe at 470.000,52.500 is z=0.063437 4:08 p.m. RatOS | Adaptive Mesh: Probing the end of the prime location at 470.0, 52.5
4:09 p.m. RatOS: Printing... 4:09 p.m. RatOS | BEACON: Nozzle expansion offset of -1.231562mm applied to T0 4:09 p.m. RatOS | Priming: Bridging with 40% fan speed.. 4:09 p.m. SET_GCODE_OFFSET Z=0 MOVE=1 <--- This is me clicking the CLEAR button to remove the -1.2mm Z offset while the hotend is being driven into the PEI and the extruder clicks 4:09 p.m. RatOS | Priming: Starting prime blob.. 4:09 p.m. RatOS | Priming: Moving to 495.0, 5 along the edge of the print area.. 4:09 p.m. RatOS | Priming: Lifting Z to 50.0.. 4:09 p.m. RatOS | BEACON: Nozzle expansion offset of -1.231562mm applied to T0 4:09 p.m. RatOS | Priming: Priming nozzle with prime blob.. 4:09 p.m. RatOS | TEMPERATURE_WAIT: Temperature for toolhead T0 reached. 4:08 p.m. RatOS: please wait... 4:08 p.m. RatOS | TEMPERATURE_WAIT: Waiting for sensor: extruder, MINIMUM: 250, MAXIMUM: 255 4:08 p.m. RatOS: Heating Extruder... 4:08 p.m. RatOS | Adaptive Mesh: Saving offset adjustment of 0.06343749968021539 in probe_for_priming_end_result 4:08 p.m. Last reading: 5474223.85Hz, 41.99C, -infmm 4:08 p.m. Result is z=0.063437 4:08 p.m. probe at 470.000,52.500 is z=0.063437 4:08 p.m. RatOS | Adaptive Mesh: Probing the end of the prime location at 470.0, 52.5
Solution:Jump to solution
Doing the double "SAVE_Z_OFFSET" in the console while the print was going actually seems to have worked.
14 Replies
Please, share ratos-debug.zip
You can find it under the configurator.
Will do tomorrow. I've turned the printer off for the night.
I turned expansion calibration off and got a fairly good result.
Very next test has different Z offsets again:
2:55 p.m. RatOS | Adaptive Mesh: Saving offset adjustment of -0.02625000031971303 in probe_for_priming_end_result 2:55 p.m. Last reading: 5477968.69Hz, 43.80C, -infmm 2:55 p.m. Result is z=-0.026250 2:55 p.m. probe at 470.000,52.500 is z=-0.026250 2:55 p.m. RatOS | Adaptive Mesh: Probing the end of the prime location at 470.0, 52.5 2:55 p.m. RatOS | Adaptive Mesh: Saving offset adjustment of -0.06750000032012404 in probe_for_priming_result 2:55 p.m. Last reading: 5479410.89Hz, 43.79C, -infmm 2:55 p.m. Result is z=-0.067500 2:55 p.m. probe at 470.000,7.500 is z=-0.067500 2:55 p.m. RatOS | Adaptive Mesh: Probing the start of the prime location at 470.0, 7.5And I've tightened the nozzle while it's hot, and the screws on the Beacon are also very tight.
Bed/frame isn't perfect. But 1/3rd of a mm should be easily compensated.
Stripped down the hotend and found just about every single bolt and screw was loose.
Even on the carriages.
Lock tight is your friend
Well, things are definitely better, but still not working right.
Is it possible to heat the bed/Beacon for too long? Trying to print PETG. I let the bed heat to 80c for 20-30 minutes, run the first layer test print. Get the Z offset sorted out. Keep the bed heating at 80 after it's done and run another first layer test to confirm it looks good. Which it has been pretty much. Maybe a slight bit of Z babystepping needed. Bed has been at 80c for ~45 to 60 minutes by now when that print finishes. And then any further prints I try and do, the Z offset that just worked a few minutes ago is suddenly not good.
Maybe repeat the beacon complete calibration now that you have solid hardware. It will do the thermal expansion tests
I've done that. Also enabled the bed compensation with a contact mesh. New input shaper tuning too. Though I can't remember if I checked the belt tension graphs I think I must have.
I'll check the directory and post the graphs.
I pretty much did the whole initial calibration over again.
Somewhere during this whole process I did a full reinstall of rc2, and then a one by one update starting with the configurator.
Finally got back around to working on this.
Solution
Doing the double "SAVE_Z_OFFSET" in the console while the print was going actually seems to have worked.