[FIXED] Prime pillar being skipped

[TN Yankee]

Perhaps I'm not using the right search terms, but I haven't come up with other posts about this.

(Running 3.1.1) I've noticed some flaky behavior with using 2 extruders to one hot-end (SeeMeCNC Y adapter). Though I usually get a proper filament purge when changing tools, I spotted this yesterday during a print, and went back & found it in the code. The issue seems to be that while a prime pillar is being built with 2 parts (unique process for each part), a prime doesn't occur when a tool change coincides with changing from one part (or process) to another.

PVA support filament is in T1, but when a T0 is issued, no purging/priming takes place - it just jumps into printing a perimeter with the PLA in T0. Other places in the file purge properly, when the tool change is within the same part.

Desired operation:

Code: Select all

G1 X-76.071 Y-21.249 F2700
G1 E0.0000 F600
G1 Z6.350 F12000
T1
; tool H0.250 W0.720
; prime pillar
G1 X-6.640 Y81.200 F13500
G1 Z2.750 F12000
G1 E0.0000 F600
G1 X6.640 Y81.200 E0.8745 F2160

Undesired operation:

Code: Select all

G1 X-42.579 Y-6.845 F2700
G1 Z6.350 F12000
; layer 22, Z = 2.75
T0
; tool H0.250 W0.720
; outer perimeter
G1 X-60.099 Y-25.117 F13500
G1 Z2.750 F12000
G1 E0.0000 F600
G1 X-60.726 Y-24.647 E0.0510 F2700

Am I not understanding something correctly? I've read about not being able to use a prime pillar with processes starting at different Z, but both of these parts start on the bed (secondary part is "low quality" parameters for PLA to fill most of the need for support vertically, then switch to PVA for the final few layers). I've seen the teaser about this very technique being much more automated in the upcoming 4.0, but that won't work for me in this situation anyway due to other reasons.

Any ideas (or links to threads that I missed)? Thanks!

[brian442]

It would probably help if you posted your factory file and then also mentioned the exact layers where the pillar is being skipped. That way others can check your settings.

[hmflash]

I am having the same issue in v4.0 as well. My test part is the lulzbot keychain using the stock Ultimaker 3 configuration.

http://devel.lulzbot.com/TAZ/accessorie ... n_text.stl
http://devel.lulzbot.com/TAZ/accessorie ... n_body.stl

The part starts off with both colors in use and the prime pillar is fine. The middle of the part only uses one color, and again the transition from layers with 2 colors to layers with 1 color is fine. The issue is after printing layers with only 1 color, when the 2 color layer happens, the prime pillar is skipped.

I can reproduce this with the default "Ultimaker 3" settings, and have attached the factory file to be sure.

This is the gcode when for the correct tool changes:

Code: Select all

G92 E0
G1 E-12.0000 F900
; process Color1
; layer 44, Z = 3.360
M218 T1 X0.0 Y0.0 ; reset extruder offset
T1
M218 T1 X-18.0 Y0.0 ; set extruder offset
; uncomment section below to use idle temperatures
;M104 S170 T0 ; idle left extruder
;M104 S195 T1 ; heat right extruder
; prime pillar
G1 X161.557 Y118.374 F15000
G1 E0.0000 F900
G92 E0

This is the gcode that happens at layer 44 when we try to change to tool T1 after printing a couple of layers with just tool T0:

Code: Select all

G92 E0
G1 E-12.0000 F900
; process Color1
; layer 40, Z = 3.060
M218 T1 X0.0 Y0.0 ; reset extruder offset
T1
M218 T1 X-18.0 Y0.0 ; set extruder offset
; uncomment section below to use idle temperatures
;M104 S170 T0 ; idle left extruder
;M104 S195 T1 ; heat right extruder
; tool H0.150 W0.400
; inner perimeter
G1 X74.142 Y115.574 F15000
G1 E0.0000 F900
G92 E0

Here is an image of the bug. Notice S3D is printing the perimeter with T1 (green) even though the top of the prime pillar was skipped and is from what was printed on the previous layer with T0 (blue).



I can attach the full gcode if it is helpful.

[brian442]

I can reproduce this with the default "Ultimaker 3" settings, and have attached the factory file to be sure.

It looks like you posted the fff file. It would help if you posted the actual factory file (File > Save Factory File)

[hmflash]

Oops. Lets try that again.

[S3D-Jason]

This issue has now been fixed with Version 4.0.1. The factory file that was provided now slices correctly in this version with the correct prime pillar sequence for the multi-color layers.

Thank you for your help reporting the issue!