MMU2 - Real Multi Material  

Re: MMU2 - Real Multi Material

NICE!!!!

Thanks again,
Eric

Re: MMU2 - Real Multi Material

I've been testing this out today and there is something weird going on with the first change. I'm going from PLA with a 215 first layer and a 205 other layers. when it does the first change it'll heat up to 215 before unloading the PLA, which causes strings. All the rest of the changes seem to work fine.

Re: MMU2 - Real Multi Material

I've been testing this out today and there is something weird going on with the first change. I'm going from PLA with a 215 first layer and a 205 other layers. when it does the first change it'll heat up to 215 before unloading the PLA, which causes strings. All the rest of the changes seem to work fine.

Hi Shaun,

It would be great if you could open up an issue ticket. https://github.com/workinghard/mmuGcodeParser
Please provide the original, the parsed gcode and the version of the parser you've used. I need to reproduce it in order to provide some more inside.

Thanks,
Nikolai

RE: MMU2 - Real Multi Material

As there is still no traction towards PrusaSlicer fixing this issue, I'm bumping up this thread for new users.

RE: MMU2 - Real Multi Material

its really sad to see no response from prusa on this even on github, ill give your script a try in the next few days as testing multi material stuff is exactly what i wanted to do with the MMU.

gonna do some testing regarding adhesion of different filament on each other.

RE: MMU2 - Real Multi Material

Can I ask what kind of purge volumes you use for PETG and PLA together? I'm interested in trying this out with PETG support material for PLA prints, but would like to get an idea of how much to purge the PETG to avoid having it mix and weaken the PLA print. (I've noticed in the past that after changing to PLA from PETG my first layer perimiters can delaminate easily from PETG contamination)

RE: MMU2 - Real Multi Material

I've used twice the default setting to be on the safe side. That worked well for me but I'm pretty sure it can be tweaked to the better setting.

RE: MMU2 - Real Multi Material

In regards to purge volume. I've printed now couple times PLA objects with PETG support. It works like a charm. The only thing I would like to improve is to make PETG more brittle. So it's easier to remove. My idea so far (untested) is to lower the PETG temp to somewhere 230-240C and use less purge for the transition PLA->PETG. This way the support should be stable enough during print and easy to remove after it. 

RE: MMU2 - Real Multi Material

Interesting. I'm already using a low-temp PETG to begin with. Makes sense you could get away with less purge on the support transition.

RE: MMU2 - Real Multi Material

Since PR developers still don't consider to merge the provided tiny pull request into main branch, I still have to recompile it every time.

In case someone have similar issues during filament unload, here is a compiled firmware version with this patch applied. It helps a lot in case you don't have perfect filament tips (which is very common by mixing different materials).

RE: MMU2 - Real Multi Material

Hi Nikolai,

as you may have read in the forum I created the DRIBBLING version of PrusaSlicer 2.1 for MMU2, that basically allows to have perfect tip during filament changes and avoid any manual intervention. I never though to mix different filament types, but it sounds interesting, and from your experiments it proves to be realistic and that it works.

I could implement the temperature change logic directly inside my PrusaSlicer 2.1 "Dribbling", so that it could support  filament swaps with perfect tips and temperature changes without any needs for people to worry about external scripts, since all is done inside PrusaSlicer.

You may find PrusaSlicer 2.1 "Dribbling" Version here if you never tried it:
https://github.com/antimix/PrusaSlicer/releases/tag/Dribbling_2.1.0

Regards

RE: MMU2 - Real Multi Material

@antimix

That would be really great. It would also improve the gcode creation time.

All the puzzle pieces to improve MMU2 experience drastically are there. But I don't see a commitment from PR to invest in it. So be prepared to maintain your version for a long time. PR is developing a lot but it looks to me like you need some kind of VIP pass or big shit storm in order to put community work into the main branch nowadays.

RE: MMU2 - Real Multi Material

Hi All,

I might be fighting a losing battle here but I'm trying to get old PVA (bubbling during extrusion 😖 ) and PET to work seamlessly together. For the most part I have won using this threads help with custom branches Dribbling & Skinnydip & Slic3r++ etc. Out of frustration I wrote a g-code 'script' that doesnt print anything but simply cycles through whatever combination of 'tool' changes implementing any tuning setting. The idea works and the outcome is that I get perfect filament tips every time with very little tuning actualy required. The only tuning is simply extract the filament when its cold enough not to string and would rather break kind of like a non-newtonian type fluid. I have found that ramming is irrelevant and the cooling moves can be reduced to 1 or 2 fairly rapid moves. The downside is heating and cooling the nozzle which takes time but at this stage thats the only disadvantage on an already realllyy long print...

I am now at the point where if the nozzle sits on the wipe tower during cooling, blobs form which then build up over time and have caused crashes and its too complex for me to implement wipe movements ina  post processing script so I'm kinda stuck. For anyone wanting to try this I can upload the gcode, its long so not sure if I can upload here...

@antimix I would love to try your latest version which I assume is beta3, need to figure out how to compile unless you upload a release anytime soon?

RE: MMU2 - Real Multi Material

Posted by: @duncan-s 

I am now at the point where if the nozzle sits on the wipe tower during cooling, blobs form which then build up over time and have caused crashes and its too complex for me to implement wipe movements ina  post processing script so I'm kinda stuck.

If you're using the script then the Hot->Cold transition is happening during the wipe movement. In most cases I had,  nozzle was already near cold state and the waiting time was short. If your nozzle is waiting long on the wipe tower then you can improve it by simple make the wipe tower bigger (extrude more). The default value 70 is way to small anyway for material swap. Go way up to 120, 150, 200. Otherwise PVA will mix into the PETG in your object. 

You can also modify the script and remove the gcode to wait for the target temperature.

RE: MMU2 - Real Multi Material

Hi Nikolai,

I am using your script and still need to really test it in depth. One thing I have noticed is that tool change temperature reduction (M109 S245 ; MMUGCODEPARSER v0.2) only works for the first round of filament changes, thereafter the temperature reduction only works when there is a different filament. IE if I have 4x PET's and 1 x PVA then the temp drop only works consistently between PVA->PET or PET->PVA and NOT between PET->PET after the first 5 changes. 

I hunt for your "M109 S245 ; MMUGCODEPARSER v0.2" comments in NP++ and replace them with my bold G91->G90 code below.

Also, for some reason the first 5x changes are all M109, thereafter all M104? If M104 is used then the nozzle temp cant drop quick enough to actually make use of dropping the temperature to prevent stringing before extraction. I use quite drastic temperature drops as you can see below and I wait for the nozzle temp to be reached and its that that causes blobs. I tried to work around it by reducing nozzle pressure & zhopping but still need to test this further.

I want to try fiddle with the dribbling effect (@antimix) too, perhaps its possible to get away with no temp drop at all just by doing a series of rapid extended rams from cooling tube right into nozzle a number of times but I think this will have a higher chance of grinding.

This is a portion of my gcode tip tuning script is is really my first attempt at gcoding directly to excuse the excessive commenting...

;-------------------------------------------------------------------------------------------------------------------------------
; CP TOOLCHANGE START
; toolchange #1
; material : PET -> PVA
;--------------------

; CP TOOLCHANGE UNLOAD

G91 ;SET POSITIONING TO RELATIVE FOR SMALL Z HOP
G4 S0 ; Wait for buffer to complete and printer to catchup all mvoes before continuing - same as M400
M106 S255 ; Part fan on to cool hotend
G1 E-3 F1000 ;SMALL EXTRACTION TO REDUCE PRESSURE
G1 Z5 ;ADD A SMALL HOP TO AVOID BULGES ON WIPE TOWER
M109 S200 ; TOOLCHANGE WAIT FOR TEMP NORMAL MODE
G4 S2 ;ADD A SMALL PAUSE TO ALLOW TEMPERATURE TO SETTLE
G1 Z-5 ;ADD A SMALL HOP TO AVOID BULGES ON WIPE TOWER
M106 S0 ; Fan off
G90 ;SET POSITIONING TO ABSOLUTE TO CONTINUE PRINT

;G91\nG4 S0\nM106 S255\nG1 E-3 F1000\nG1 Z5\nM109 S200\nG4 S2\nG1 Z-5\nM106 S0\nG90

; BEEP WHEN EXTRACTING
M300 S2000 P100

; Extract from nozzle
G4 S0
G1 E-15.0000 F5000
G1 E-24.5000 F1200
G1 E-7.0000 F600
G1 E-3.5000 F360

; This is the cooling moves
G4 S0
G1 E20.0000 F600
G1 E-20.0000
G1 E20.0000
G1 E-20.0000

; This is the filament extraction from extruder prior to handover to MMU2
G1 E-35.0000 F2000

; RESTORE PRE-TOOLCHANGE TEMP NORMAL MODE TO PURGE NOZZLE
G4 S0
M104 S245

; Selects extruder 5
T4

; Sets current extruder position to zero
G92 E0

; RESET THE FILAMENT PRINT TEMEPERATURE FOR NEXT FILAMENT
; THIS IS THE ONLY PIECE OF CODE WHICH NEEDS TO BE TUNED FOR THE NEXT FILAMENT
G4 S0
M104 S205

; EXTRUDE SOME MATERIAL TO PURGE NOZZLE - THIS WOULD NORMALLY BE DONE IN WIPE TOWER
G1 E150 F250

; CP TOOLCHANGE END
;****************************************************************************************************************************

;-------------------------------------------------------------------------------------------------------------------------------
; CP TOOLCHANGE START
; toolchange #2
; material : PVA -> PET
;--------------------

; CP TOOLCHANGE UNLOAD

G91 ;SET POSITIONING TO RELATIVE FOR SMALL Z HOP
G4 S0 ; Wait for buffer to complete and printer to catchup all mvoes before continuing - same as M400
M106 S255 ; Part fan on to cool hotend
G1 E-3 F1000 ;SMALL EXTRACTION TO REDUCE PRESSURE
G1 Z5 ;ADD A SMALL HOP TO AVOID BULGES ON WIPE TOWER
M109 S165 ; TOOLCHANGE WAIT FOR TEMP NORMAL MODE
G4 S2 ;ADD A SMALL PAUSE TO ALLOW TEMPERATURE TO SETTLE
G1 Z-5 ;ADD A SMALL HOP TO AVOID BULGES ON WIPE TOWER
M106 S0 ; Fan off
G90 ;SET POSITIONING TO ABSOLUTE TO CONTINUE PRINT
;G91\nG4 S0\nM106 S255\nG1 E-3 F1000\nG1 Z5\nM109 S165\nG4 S2\nG1 Z-5\nM106 S0\nG90

; BEEP WHEN EXTRACTING
M300 S2000 P100

; Extract from nozzle
G4 S0
G1 E-15.0000 F5000
G1 E-24.5000 F1200
G1 E-7.0000 F600
G1 E-3.5000 F360

; This is the cooling moves
G4 S0
G1 E20.0000 F600
G1 E-20.0000
G1 E20.0000
G1 E-20.0000

; This is the filament extraction from extruder
G1 E-35.0000 F2000

; RESTORE PRE-TOOLCHANGE TEMP NORMAL MODE TO PURGE NOZZLE
G4 S0
M104 S205

; Selects extruder 1
T0

;RESET EXTRUDER OFFSET
G92 E0

; THIS IS THE ONLY PIECE OF CODE WHICH NEEDS TO BE TUNED FOR THE NEXT FILAMENT
; RESET THE FILAMENT PRINT TEMEPERATURE
G4 S0
M104 S245

; EXTRUDE SOME MATERIAL TO PURGE NOZZLE - THIS WOULD NORMALLY BE DONE IN WIPE TOWER
G1 E150 F250

; CP TOOLCHANGE END
;*************************************************************************************************************************

;-------------------------------------------------------------------------------------------------------------------------------
; CP TOOLCHANGE START
; toolchange #3
; material : PET -> PET
;--------------------

; CP TOOLCHANGE UNLOAD

G91 ;SET POSITIONING TO RELATIVE FOR SMALL Z HOP
G4 S0 ; Wait for buffer to complete and printer to catchup all mvoes before continuing - same as M400
M106 S255 ; Part fan on to cool hotend
G1 E-3 F1000 ;SMALL EXTRACTION TO REDUCE PRESSURE
G1 Z5 ;ADD A SMALL HOP TO AVOID BULGES ON WIPE TOWER
M109 S200 ; TOOLCHANGE WAIT FOR TEMP NORMAL MODE
G4 S2 ;ADD A SMALL PAUSE TO ALLOW TEMPERATURE TO SETTLE
G1 Z-5 ;ADD A SMALL HOP TO AVOID BULGES ON WIPE TOWER
M106 S0 ; Fan off
G90 ;SET POSITIONING TO ABSOLUTE TO CONTINUE PRINT

; BEEP WHEN EXTRACTING
M300 S2000 P100

; Extract from nozzle
G4 S0
G1 E-15.0000 F5000
G1 E-24.5000 F1200
G1 E-7.0000 F600
G1 E-3.5000 F360

; This is the cooling moves
G4 S0
G1 E20.0000 F600
G1 E-20.0000
G1 E20.0000
G1 E-20.0000

; This is the filament extraction from extruder
G1 E-35.0000 F2000

; RESTORE PRE-TOOLCHANGE TEMP NORMAL MODE TO PURGE NOZZLE
G4 S0
M104 S245

; Selects extruder 2
T2

;RESET EXTRUDER OFFSET
G92 E0

; THIS IS THE ONLY PIECE OF CODE WHICH NEEDS TO BE TUNED FOR THE NEXT FILAMENT
; RESET THE FILAMENT PRINT TEMEPERATURE
G4 S0
M104 S245

; EXTRUDE SOME MATERIAL TO PURGE NOZZLE - THIS WOULD NORMALLY BE DONE IN WIPE TOWER
G1 E150 F250

; CP TOOLCHANGE END
;****************************************************************************************************************************

RE: MMU2 - Real Multi Material

Wouldn't it be easier to edit the python script instead? Looking for M109 after the processing is not accurate. Depending on the transition low->high, high->low, same->same and different parameters it's doing different things.

For example PET->PET transition with the same base temperature is doing nothing if the parameter ram_temp_diff is set to zero.

If it's possible, please provide your project file, so I can look what's happening there. 

RE: MMU2 - Real Multi Material

It would absolutely be easier to edit the script and I have been trying to get a .py to .exe converter working and havent managed to figure out how to run .py directly from Slic3r...

Attached the file I generated this morning.

RE: MMU2 - Real Multi Material

It wont allow to attach gcode or txt files so not sure how to send the code?

RE: MMU2 - Real Multi Material

Perhaps a zip will work. This is just the raw output of your code run through your script. I havent added my 2cents yet.

RE: MMU2 - Real Multi Material

I managed to get the .py script executing in slic3r which is cool so yeah now its easier to modify the .py accordingly but will take me some time to understand how to speak python...

I attached 3 files, the original gcode from slic3r, the adjusted gcode & the script as I have it now.

the issue is that even if I set:

# Drop the temperature by 10C during the ramming process. Checking if it might help
ram_temp_diff = 30
# Set this to True if you want to drop the temperature even for the same filament
ram_temp_diff_wait_for_stabilize = True

The resulting code is the following which clearly misses out the temperature drop even for same filaments.

;--------------------
; CP TOOLCHANGE START
; toolchange #94
; material : PET -> PET
;--------------------
M220 B
M220 S100
; CP TOOLCHANGE UNLOAD
G1 X170.750 Y140.510
M900 K0
G1 E-15.0000 F9000
G1 E-24.5000 F2400
G1 E-7.0000 F1200
G1 E-3.5000 F720
G1 Y141.350
G1 X220.750 E20.0000 F3000
G1 X170.750 E-20.0000 F2500
M73 Q20 S530
M73 P20 R526
G1 X229.250 E20.0000 F2340
G1 X170.750 E-20.0000 F1755
G1 E-35.0000 F2000
G1 Y141.180 F2400
G4 S0
; Filament-specific end gcode
; RESTORE PRE-TOOLCHANGE TEMP NORMAL MODE TO PURGE NOZZLE
;G4 S0
;M104 S245