No Retraction When Leaving Prime Tower

[Benjamin4456 0]

Hey all, I've been sorting out a Chimera Ender 3 setup and have managed to get most of the kinks worked out, but I just ran into this odd problem.  When leaving the prime tower the extruder is not retracting.

 

The prime tower looks normal since linear advance is taking care of the residual pressure, but upon arriving at the part enough filament has oozed that it deposits a few blobs/strands.  The nozzle change g-code works and looks fine, but rather it's when it is printing a consistent material and building up the prime tower that this issue occurs.  This is confirmed by looking at the g-code files where sure enough there is no retraction on these moves.  I was running 4.12.1 and now I'm on 4.13.1 and the issue persists.  It could be that I'm missing a setting, although if so I certainly am having trouble finding it.

 

The prime tower is located at (45,20).  Attached are the relevant files I can think of.  Let me know if anything else is needed.

E3PROC_Cable_Strain_Relief_Cover_v19gcodetest.gcode E3PROC_Cable_Strain_Relief_Cover_v19gcodetest.3mf

[Benjamin4456 0]

Well I've managed to figure out a way around the problem, but not the problem itself.  Using the exact same profile if I have the layer height set to 0.24mm the prime tower retraction issue presents itself.  If I drop it to 0.20mm the retraction before leaving the prime tower gets added back in.

 

I have no clue why this is happening and it'd be great if someone could take a look here and see if they notice anything weird.  Below I'll attach the two gcode files and the profile I was using for folks to take a look at.  I'll also zip the machine definition files below, but note that the issue persists even with the default Cura custom machine.  Let me know if there's anything else I can provide.

E3PROC_Cable_Strain_Relief_Cover_v19gcodetest (0.24).gcode E3PROC_Cable_Strain_Relief_Cover_v19gcodetest (0.20).gcode ASA - DualE Base Profile.curaprofile Ender 3 Pro Chimera.zip

[Benjamin4456 0]

Ok.... New part model and now the layer height work around doesn't work.  Genuinely confused here.  Am I nuts and there's a setting somewhere that I've missed, or is this a bug?  I'd really love to know because getting strings and then areas of slight under extrusion at the beginning of a layer is not something I'm particularly fond of.

 

Attached is the model where the issue pops up again.

E3PROC_Chimera Carriage for E3P v8_ASA.gcode

[GregValiant 1,410]

I received an "Unknown Printer" warning and I could only open your 3mf as a model file.  Your settings didn't come across.

 

This is the travel to the first layer prime tower in your gcode file.  There is no retraction.

G1 X93.578 Y93.426 E49.76969
G0 F3600 X37.862 Y24.192
;TYPE:PRIME-TOWER
G1 F1800 X37.037 Y24.634 E49.81451

............Then leaving the prime tower

G1 X35.403 Y20.521 E59.10517
G0 F3600 X102.462 Y108.599
;TYPE:SUPPORT
G1 F1800 X104.96 Y108.601 E59.22481

 

This from a gcode file I produced from your model:

G1 X147.409 Y142.321 E69.7767
G1 F2100 E64.7767   Retract
;MESH:E3PROC_Cable_Strain_Relief_Cover_v19gcodetest.3mf
G0 F9000 X197.225 Y194.381
G0 X201.07 Y199.07
;TYPE:PRIME-TOWER
G1 F2100 E69.7767   Prime
G1 F4500 X202.559 Y197.85 E69.8463

........Leaving the prime tower

G1 X207.891 Y205.891 E97.98674
G1 F2100 E92.98674    Retract
G0 F9000 X197.959 Y193.779
G0 X135.434 Y111.483
G0 X134.728 Y110.662
G0 X134.692 Y110.453
;TYPE:WALL-INNER
G1 F2100 E97.98674    Prime
G1 F4500 X134.513 Y110.525 E97.99372

 

There were 152 retractions in your gcode file and 354 in the file I generated.  But without knowing the differences in the settings, we can't really tell what's going on.

 

Here is a 3mf file using your model.  I used an Ender 3 Pro configured as a 2-in-one-out dual extruder.  Extruder2 is used for the support and prime tower is enabled.  Basic default settings everywhere else.

 

GV PrimeTowerTest.3mf

[Benjamin4456 0]

Thanks for taking a look.  I tried downloading your 3mf but ran into the same "printer does not exist" issue, and I can't set an Ender 3 Pro up a dual extruder unless I go modify the definition files to allow more than one extruder which just puts us back to square one with a custom machine again.

 

I included the definition files for my printer in the second post - they're in the zip folder since the forum doesn't allow the upload of json files.  That paired with my material profile which was also attach should allow you to open it up.

 

I'll attach the original stl file below so that we don't have to worry about Cura not recognizing the name of the printer in the 3mf.

 

Thanks again for the help.

 

 

Edit:  Even without matching machines, shouldn't the material profile still be able to be imported?  That would then give us equivalent slicer settings aside from machine parameters.

Chimera_Carriage_for_E3P_v8.stl

Edited February 19, 2022 by Benjamin4456

[Benjamin4456 0]

Alrighty, for the sake of completeness I went ahead and made a new quality preset/profile from scratch, only adjusting settings that I needed and... still the exact same issue.  I've gone ahead and put all the relevant files that I can think of into the zip folder attached below.  I hope that it's enough that someone might be able to replicate this issue on their end.

 

Realistically this isn't an end of the world sort of problem, especially since supports are always printed first after leaving the prime tower.  However it's still a nuisance and if anybody happens to figure out what's going on that would be awesome.  I'll keep toying around with things on my end, but I'm just about out of ideas.

 

Thanks.

Dual Extruder Problem Files.zip

[GregValiant 1,410]

I installed the printer.  I think the problem is in that "ASA_Dual E..." profile.

 

This is from the gcode using your ASA profile.  It is the nozzle moving from the prime tower to the model and there is no retraction.  Moving to the prime tower there wasn't a retraction either.

G1 X0.853 Y21.184 E15.41702
G0 F7200 X20.2 Y20.2
G0 X126.25 Y99.299
;TYPE:SUPPORT-INTERFACE
G1 F2400 X125.948 Y99.501 E15.42886

 

This is with the Fine profile (I picked "fine" at random) with Extruder#2 doing the Support Interface:

G1 X169.691 Y207.691 E18.56354   Last extrusion at prime tower
G0 F7200 X195.2 Y213.2
G1 F1500 E12.06354                          Retract
G0 F7200 X53.39 Y99.839
G0 X7.576 Y99.839
G0 X4.033 Y103.5
;TYPE:SUPPORT-INTERFACE
G1 F1500 E18.56354                          Prime
G1 F2400 X2.733 Y103.5 E18.58516

 

There doesn't seem to be a problem moving to the prime tower either.

My suggestion would be to start from scratch using a generic profile, customizing it to your needs, and see how it goes.

 

[Benjamin4456 0]

Ok, I had just typed out my conclusion and then figured I should test a couple other files.  Unfortunately doing so threw my findings right out the window (good thing I checked).

What I had first thought was the issue was altering the location of the prime tower.  Moving it anywhere from the default location on the first model would make the retraction issue appear.  If it stayed put, all the layer heights I tested worked fine.

 

Then I tried a completely different model with no profile changes from the above and now it's back to the same issue.  This was a profile I made completely from scratch (second time now) while testing every profile setting change by slicing it each time and inspecting the gcode.  Guess I get to do it over again here and see if anything different rears itself... Yippee....

[GregValiant 1,410]

It occurs to me that within your definition file you directly inherit "fdmprinter" and bypass the "Creality_Base" definition.

 

Maybe there are some over-rides within the Creality_base file that you needed?  When I set up my Ender 3 Pro for multi-extruder - I made the changes to the Extruder Trains within the Creality_Base definition file rather than creating an entirely new definition.

 

That may give you something else to look at when you get tired of staring at Gcode.

[Benjamin4456 0]

Well after getting tired of staring gcode, I took your advice Greg and tried messing with the original Creality definition.  I certainly don't understand it, but it seems to be working again, though I have noticed a couple things.

 

For one, the time estimates between the two (with identical material profiles) were radically different.  I don't recall how accurate the custom machine was, but the Creality based config is pretty accurate while being nearly double the time of the custom machine.  Perhaps this is because I did not define the accel and jerk values in the custom definition.  I did edit the Creality definition to match my firmware settings though.

 

The fans/extruder swap also do not work as perfectly with the Creality setup vs the custom one.  I noticed that the first time the material swaps, the second extruder stays at the Initial temp rather than ramping up fully to Print temp.  After the second switch it acts as expected.  The fans are the opposite in that they swap back and forth the first extruder change, but from then on they both stay on full time.  Not really sure what is going on there, but the print finished successfully and the slight amount of extra cooling isn't the worst thing in the world when the build chamber is already 50C+.

 

For now I'm not going to try and figure out what the heck caused these variations (I've already spent hours going back and forth trying to do so).  Looking at the Creality_Base definition I did not see anything that would appear mission critical that I was missing in my custom definition, but who knows.  Mostly it was just material settings that I didn't care about having defaults for anyway.

 

I'll fiddle with the setup again at some point here, but I've had enough headaches getting dual extrusion to work properly for the time being.  I'm sure I'll throw some more info in here if I discover something (or if something else breaks...).

 

Thanks for the help Greg.

[GregValiant 1,410]

Check the fan numbering in the Machine Settings.  I think Extruder#1 should be fan 0 and Extruder #2 should be fan 1.  If the gcode is trying to operate fans 1 & 2 then there will likely be a problem when internally the printer has them numbered 0 & 1.

[Benjamin4456 0]

I know from when I configured the firmware that the fans are 0 and 1.  Fan 2 is the hotend cooling fan.  This also matches what I can control either directly on the printer or though the terminal.