I have a MK4S with MMU3, both updated to latest firmware (6.5.3+12780)
I've been getting the MMU3 working nicely on my MK4S recently. I almost only use PETG for functional parts where colour might be used for inbuilt labels. Therefore, a purge tower to put a version number on the top of a model is a big waste of filament.
I have set my "extra load distance" in firmware to correctly specify the distance between the nozzle and the filament sensor,* now my final stage is getting the colour changes to work nicely.
I noticed that the default T0 code suggested by prusaslicer seem much more likely to result in a failed swap, and I think this is because it does not initiate ramming to shape the filament tip as it leaves. Therefore I get very stringy, unpleasant tips which leave material in the extruder gears etc.
When I compare this to M701 and M702, they seem almost flawless, it seems like a missed opportunity to not use them.
This is my filament change G-Code
M118 Change filament Gcode section;
;Debug notes:
;Previous {previous_extruder}
;Current {current_extruder}
;Next extruder{next_extruder}
{if current_extruder != next_extruder}
G60 S0 ;Save current position
G0 Z{layer_z+8} F{travel_speed_z*60} ;Move Z away from the print
G0 X{min(first_layer_print_min[0]-10,0)} Y{max(first_layer_print_min[1]-10,0)} F{travel_speed*60} ; move away from the print
M702 Z1
M701 P{next_extruder} Z1
M118 finished loading using M701;
T{next_extruder}
M109 S{new_filament_temp-5}
M104 S{new_filament_temp}
G61 S0 F{travel_speed*60} ;Return to saved position
{endif}
These are my MMU settings from start G-Code:
;BLOCK_START
M17 ; Enable steppers
M140 S60 ;Start bed heating
M109 S130; Hotend wait for 130
;Adjust Extruder steps per mm
;No idea why such a large adjustment seems to be needed from the default 380, but I calibrated this very carefully
M92 E419;
;;;;;;;;Original prusa code;;;;;;;;;;
; Perform nozzle diameter check
M862.1 P0.4 ; nozzle check
M862.3 P "MK4S" ; Check printer model
M862.5 P2 ; Check G-code level
M862.6 P "Input shaper"; Check firmware has Input shaper
M862.6 P "MMU3" ; Check firmware has MMU3 enabled
M115 U6.5.3+12776 ; Report firmware version
; Configure MMU settings
;MMU registers https://prusa3d.github.io/Prusa-Firmware-Doc/Prusa-Firmware-MMU/group__register__table.html
M708 A0x1e X14 ; Set pulley current to ~260mA
; https://prusa3d.github.io/Prusa-Firmware-Doc/Prusa-Firmware-MMU/group__TMC2130.html
M708 A0x0d X140 ; Set unload feedrate
M708 A0x11 X140 ; Set load feedrate
M302 S200 ;Lower cold extrusion limit
;;;;;;;;;END Original prusa code;;;;;;;;
; Configure MMU settings
M708 A0x14 X6 ; Set slow feedrate 8mm/s is 19cubic mm/s
M708 A0x22 X380 ;bowden length
M708 A0x0b X{extra_loading_move}; Set extra load distance
;BLOCK_END TemplateGCode
*I find there is about 36.5mm between the filament sensor and the nozzle tip.
I have a MK4S with MMU3, both updated to latest firmware (6.5.3+12780)
I've been getting the MMU3 working nicely on my MK4S recently. I almost only use PETG for functional parts where colour might be used for inbuilt labels. Therefore, a purge tower to put a version number on the top of a model is a big waste of filament.
I have set my "extra load distance" in firmware to correctly specify the distance between the nozzle and the filament sensor,* now my final stage is getting the colour changes to work nicely.
I noticed that the default T0 code suggested by prusaslicer seem much more likely to result in a failed swap, and I think this is because it does not initiate ramming to shape the filament tip as it leaves. Therefore I get very stringy, unpleasant tips which leave material in the extruder gears etc.
When I compare this to M701 and M702, they seem almost flawless, it seems like a missed opportunity to not use them.
This is my filament change G-Code
These are my MMU settings from start G-Code:
*I find there is about 36.5mm between the filament sensor and the nozzle tip.