PrusaSlicer 2.7.4 has created an invalid GCODE file

[HTCPCP418]

Description of the bug

I was slicing an STL file for the Prusa MK3+. When I tried to print it, the first few layers went o.k.. Then the stepper motor moved the print head so far to the x- and y-axis, that it hit the end, and the subsequent print layers were offset for about 20 millimeter in both x and y direction.

Project file & How to reproduce

Box_fixed.zip

Checklist of files included above

• [X] Project file
• [ ] Screenshot

Version of PrusaSlicer

2.7.4

Operating system

Windows 11

Printer model

Prusa MK3+

[HTCPCP418]

I am printing the STL sliced with Ultimaker Cura just fine, thus I guess the STL itself is without errors.

[u89djt]

Here's the gcode I got from slicing in your 3mf file Box_fixed_0.15mm_PETG_MK3S_19h18m.zip It looks fine when it's loaded up separately into the viewer, but I haven't run it on my MK3S or XL Maybe compare it to the gcode you're getting? With the gcode looking ok so far, it begins to suggest you might have had a hardware glitch or the - something I've experienced - the tool being held back by jamming against a curled extrusion and gunge around it? I guess you'd have to run the print again and watch closely when it approaches that bit. Maybe video it with your phone or something.

[HTCPCP418]

Yes, that's how it looks in my PrusaSlicer too. But when I print it (I started it twice), when it starts to switch from infill to top layer at the bottom of the box, it runs into the misalignment.

[u89djt]

OK, I think I might have edited my post after you replied, sorry. OK, so it's at the switch from infill to the top layer at the bottom of the box. Lemme look around there out of interest - I'm just another user, so I don't know if I'll be able to see anything.

[HTCPCP418]

This are my two attempts. You can see that both have fizzled out ends to the top and the side, and you can see the layer printed off-set to the previous one. You can see, that the bottom one is off-set just about 10 mm, the first one far more. I guess that the jump in the transport mechanism was different, when the hotend hit the end of the rail.

[HTCPCP418]

(Sorry about them being bent, I just put them in the bin first.)

[u89djt]

That looks like the nozzle caught on those burnt bits? Have you got the cura file there?

[HTCPCP418]

The cura file is printing right now.

[u89djt]

Yep, if you can copy it here I'll have a look. It looks like the infill on your prusaslicer slice has ended up glitchy. If that was my print, I would usually assume the filament was wet. Wondering what cura infill looks like in this case.

[HTCPCP418]

Cura file printing right now. You see some stringing, and yes, the filament might be somewhat wet (It was in a cabinet for a few month. But the other item I printed with it went finde, so I gave it a try.

[HTCPCP418]

What I was printing previously with the same filament from a PrusaSlicer file.

[u89djt]

Aye, I mean the configuration you used in cura - the 3mf file equivalent if it can output one. I'm just downloading cura now out of interest. Those burned spots on the failed print are where the head stayed there longer than it should, trying to get away along the path, so I would assume for myself that it explains the offset in this case. If you want to post the leaf 3mf, we can look at what the infill might have done differently. Yeah, there are wisps on the upper surface - those would happen internally too, and if the infill is dense, you could end up with a pileup I guess. For the main task: if cura is creating gyroid infill at the same density, then there could be information about how to cope with wet filament in this case, so I'm interested in seeing the comparison. It could be of interest to the devs, too :)

[HTCPCP418]

In Cura, I was using the cubic infill.

[u89djt]

OK, so it may be that cubic infill prints more successfully with wet filament. I guess it's simpler than gyroid - straight lines everywhere sitting on top of straight extrusions offset a constant distance. Maybe cubic infill at the same density as you used in cura would succeed via prusaslicer without creating buildups to catch the nozzle, burn and cause loss of stepper motor tracking. To complete the comparison, you could try all the same settings with gyroid in cura.

[HTCPCP418]

PI3MK3M_Box_fixed.zip This is the UltiMaker Cura project file.

[u89djt]

Right, that does look like a more straightforward extrusion task. No stops and starts for the filament to ooze around. OK, this user thinks your issue is a filament quality matter worked around by switching to simpler, less technical infill. I could be wrong, of course, but your experience is well explained by oozing filament and a stuck tool for me.

[HTCPCP418]

Makes sense. I'll try with less infill again.

[u89djt]

Yup, and something with straight lines. You used 20% grid in cura - not cubic - and that's available in prusaslicer.