The light is coming! I was finally able to understand how to make the MMU2 perfectly working.
At the moment I think (cross fingers) I started to understand the "whole" MMU2<->MK3S word.
In facts in those days I was able to improve the reliability from 0 to 99%, working in different fronts.
I started from the Hardware. This is the working situation:
EXTRUDER:
- Original PRUSA MK3S small diameter 1.85 PTFE
- Replaced the PRUSA heat break (the one with diameter change) with the original E3D heat break (with fixed diameter)
- Replaced PTFE tube to the MMU2 with a new one with an inside diameter of 4mm
BUFFER:
- Replaced the PTFE tubes with new ones, with an inside diameter of 4mm.
MMU2:
- Added five 50mm PTFE tubes with an inside diameter of 2mm and with an hydraulic joint at the exit of the MMU2. This will be linked to the tubes from the buffer, for an easy plug/unplug in case of need.
Then was the turn of the software.
I understood that was PrusaSliced to drive everything, so I started to work on it. I was finally able to improve PrusaSlicer in a way that allowed me to do multicolour prints with all the reliability I was expecting from my MMU2. 😍 I spent some days to understand how to set up the development environment and how the PrusaSlicer internal code was. Then I started the coding of some ideas I had to manipulate the filament to get perfect shape. I spent the remaining time to code, and print tests, changing the code based on the results, until I was ready. I called this clever handle "filament dribbling".
Yesterday, I downloaded the latest PrusaSlicer 2.1.0-alpha1 version and apply all my changes to produce almost perfect filament in extraction. In addition I fixed the annoying bug that prevents to print in multicolour with different filament diameters.
Today I launched a multicolour print with a lot of changes, to see if all my successfully tests, were confirmed by a real long print, and... YES... the print went perfect and smooth, without any intervention for all the colour changes.
The real secret was to use a clever slicer to produce a very good filament on extraction . 😊
I learnt that in MMU2 mode, almost all is controlled by PrusaSlicer, EVEN the filament eject and load.
I was convinced that the Slicer used special gcode to load and unload (e.g. single instruction). I was wrong. The Slicer when it is time to unload the filament calculates exactly where the filament is and created gcode instructions to moves the E axe (the filament) just above the gears. Then it triggers the MMU2 command to retract the filament, but it should be OUTSIDE from the extruder gears.
It is crucial NOT TO CHANGE some parameters in the Printer Settings tab, even if they are not really correct. All the PrusaSlicer unload logic is based on those parameters. I understood that some issues I got in unload filament, were due to some arbitrary changes to that parameters that I had done (on suggestion to fix another issue). So I reverted back all to the default values, since that should stay as they are.
Another thing I realized is that I will need to apply my "dribbling" filament handling also to the MMU2 single mode, and to the firmware itself, since a lot filament ejects, done directly by the firmware, cause jams and terrible heads. 😲
There is really very few documentation on "how it works" and why. This will be a long journey.
There are also several bug in the firmware either MMU2 either MK3S, and I plan to work also on that in the near future.
Regards
RE: The light is coming! I was finally able to understand how to make the MMU2 perfectly working.
Hi,
I'm glad we have another happy MMU2 user, welcome on board. Just finished 26 hours, 450 tool changes print without any intervention. It was my second PLA+PETG print and I will post details later. Using my still stock MK3s+MMU2s, 1.6-rc2 mmu2 firmware and 3.8.0-rc1 firmware with the longer waiting time patch ( https://github.com/prusa3d/Prusa-Firmware/pull/2087).
Just wandering if your software/settings are matched to the E3D heat break instead of the MMU2 heat break. So it will not be really usable for stock MMU2 heat break. As I'm tweaking only temperature and keeping all other settings default, I never really had an issue with bad filament tips. The ones which are causing load/unload issues. They might not look perfectly shaped, but MK3s+MMU2s is not really dependent on that.
Looking forward to see details of your tweaks. If you're creating pull request, feel free to link it here. So we can have a discussion here instead of GitHub.
Often linked posts:
Going small with MMU2
Real Multi Material
My prints on Instagram
RE: The light is coming! I was finally able to understand how to make the MMU2 perfectly working.
Hello Nikolai,
Just wandering if your software/settings are matched to the E3D heat break instead of the MMU2 heat break. So it will not be really usable for stock MMU2 heat break. As I'm tweaking only temperature and keeping all other settings default, I never really had an issue with bad filament tips.
The settings should work either with the original. The intent is to get good filament tips once extracted. As you can see in my picture, this is the quality I was able to obtain: 😊
One thing to say, is that I always try to keep temperature low, I print PLA at low temperature 190-198 max (fist layer 205), because then it starts to strings.
I am in doubt if to implement back the temperature change in dribbling, since changing the temperature takes time during print, and the print could take some hours more than without. 🙄
I replaced the heat break, because the 2.2 internal diameter creates filament tip of 2.2 diameter, that bare fit, or does not fit at all, in the PTFE tube of the cooler, the PTFE tube between the extruder and MMU, AND all the MMU2 PTFE tubes. You as me, can also mitigate the effect by replacing the PTFE tubes with a larger inside diameters. However with the original heath break you will never get those FLAT 1.94 mm diameter filaments. 😀
I will do a ZIP file with the excutable so that who is interested in, can test Dribbling, but before I want to check some other things when it is used in MMU2 Single mode, to see if that could be improved also.
Regards
RE: The light is coming! I was finally able to understand how to make the MMU2 perfectly working.
Hello Nikolai,
Just wandering if your software/settings are matched to the E3D heat break instead of the MMU2 heat break. So it will not be really usable for stock MMU2 heat break. As I'm tweaking only temperature and keeping all other settings default, I never really had an issue with bad filament tips.
The settings should work either with the original. The intent is to get good filament tips once extracted. As you can see in my picture, this is the quality I was able to obtain: 😊
One thing to say, is that I always try to keep temperature low, I print PLA at low temperature 190-198 max (fist layer 205), because then it starts to strings.
I am in doubt if to implement back the temperature change in dribbling, since changing the temperature takes time during print, and the print could take some hours more than without. 🙄I replaced the heat break, because the 2.2 internal diameter creates filament tip of 2.2 diameter, that bare fit, or does not fit at all, in the PTFE tube of the cooler, the PTFE tube between the extruder and MMU, AND all the MMU2 PTFE tubes. You as me, can also mitigate the effect by replacing the PTFE tubes with a larger inside diameters. However with the original heath break you will never get those FLAT 1.94 mm diameter filaments. 😀
I will do a ZIP file with the excutable so that who is interested in, can test Dribbling, but before I want to check some other things when it is used in MMU2 Single mode, to see if that could be improved also.
Regards
Are you saying to use the Prusa heat break or the E3D one, and what ID should be used for the PTFE Tube in it?
RE: The light is coming! I was finally able to understand how to make the MMU2 perfectly working.
Seriously impressed by your efforts. Do you have any issues with print delamination by printing at lower temperatures?
RE: The light is coming! I was finally able to understand how to make the MMU2 perfectly working.
Seriously impressed by your efforts. Do you have any issues with print delamination by printing at lower temperatures?
If I exaggerate, yes. 😆 😆 😆
Today for wrong I printed a multi color PET-G object, setting the wrong filaments in PrusaSlicer, assigning the corresponding colours to PLA colour spools instead of PETG colours !!!! 😖
During the print I heard the sound of delamination, and look at the printer I noticed a lot of delamination and plate detachments...
Then my eyes went on the LCD that said... 198°/60°.... 😱
The astonish thing was that my SUNLU PETG that has a low temperature print range from 200 to 220, at around 198 was still able to be extruded and did not clogged the nozzle..... Of course I aborted everything and re-sliced everything with the correct materials and reprinted it.
I got a very nice object. No issue with colour change. But the object was sliced with my PrusaSlicer 2.1-rc-DRIBBLING version.