gauthier

I switched back to my old regular 0.4 mm nozzle, but still see similar problems. I made this model to check for wiggle along my X: The result looks good on both sides, so I don't suspect carriage wiggle anymore: 3mf of this test model attached (xy-wiggl.3mf) Yet, this rib is solid in my model, but has a gap in my print: Here is how it looks like in PrusaSlicer with Arachne (I switched to PrusaSlicer to see if it would perform better in that regard): 3mf for that also attached (strip-top.3mf) I am at a loss. What is the problem? I'm starting to think that Arachne might be too optimistic with regard to extrusion consistency, which the classic perimeter algorithm covered with a fat gap line. xy-wiggle.3mf strip-top.3mf

This is the same problem as in this thread. The issue does not seem related to filament, it happens with another as well. It surprises me that the lines seem to always be grouped in pairs. It might point to the nozzle itself begin the issue? If there's wiggle room, maybe it drags after on the X direction?

This is mostly noticeable in spiral mode, but also in regular prints. I switched nozzle to 0.6 mm hardened steel, re-calibrated the build plate. Slicing in Cura 5.1, tried both Marlin and Ultimaker gcode flavors. I put z offset to -0.1 mm, it wouldn't stick otherwise. What are the parameters I should look into? Cubes in spiral mode (wall thickness is correct, maybe even a little fat): In Cura: And a regular model:

As is typical, I finally found someone else with that issue, right after posting here. For now I will just lift the extruder lever and push manually, although it would be nice to understand the amount of retraction that I should compensate for, so that I could input that in the script.

I'm trying to get a color change at a given layer. The post-processing script "Filament change" did not seem to do anything, and I've seen the recommendation to use "Pause at height", and manually change the filament in the printer interface. I do that, let the extruded filament change color, then resume the print. However, as soon as I resume, no filament gets out of the nozzle. Here is the part of the gcode that does the pause, I cannot find anything wrong with that: G0 F6750 X100.669 Y111.02 ;MESH:NONMESH G0 F600 X100.669 Y111.02 Z0.67 G0 F6750 X101.288 Y111.377 ;TIME_ELAPSED:22.753103 ;TYPE:CUSTOM ;added code by post processing ;script: PauseAtHeight.py ;current layer: 2 M83 ; switch to relative E values for any needed retraction G1 F300 Z1.67 ; move up a millimeter to get out of the way G1 F9000 X190 Y190 G1 F300 Z15 ; too close to bed--move to at least 15mm M0 ; Do the actual pause G1 F300 Z0.67 G1 F9000 X101.288 Y111.377 G1 F300 Z0.67 ; move back down to resume height G1 F1862.7 ; restore extrusion feedrate M82 ; switch back to absolute E values G92 E61.71386 ;LAYER:2 ;TYPE:WALL-OUTER ;MESH:Part Studio 2 - Part 1.stl G1 F2400 X106.288 Y102.717 E62.61384 My best guess is that the filament change procedure in the Ultimaker FW does a retraction after the change. I have no retraction in the Pause script setup, but I guess I could put a negative value there, if I knew how much the FW retracts after purging the nozzle.

I also want M600, and it's to change filament in the *middle* of a layer. This would enable me to add virtual extruders in the printer profile of my slicer. I would then run M600 on tool change: instead of changing the tool, I'd just change filament. See here how it can be used on Marlin FW.

@GregValiant Ultimaker 2+

This guide refers to paths such as "resources/definitions/ultimaker_original_dual.def.json". I believe these are for the source code of Cura. How can I add a virtual extra extruder, without rebuilding Cura? On Linux, I have found ~/.local/share/cura/4.12/definitions (empty) and ~/.config/cura/4.12/cura.cfg. But from there I'm not sure what to do.

Sure! Only printed PLA on it, but got better adhesion than PLA on my Prusa stock textured sheet. Many other factors could be playing in of course. I've had paper glue on it once for a print that took most of the area, otherwise bare. You can choose to put the stick magnetic sheet on the glass bed like @K3Dds and I. The advantage is that you don't butcher the aluminium plate, and apparently you can temporarily flip the glass bed, would you ever want to print on glass again (in that case you need to use other clamps because of added thickness). I use the side of the glass that had a little damage (caused by me putting it in the freezer to free a stuck print). Someone here told me that the magnetic sheet could be unglued a couple of times, I don't remember how. Since I want to avoid replacing the aluminum plate, I went the glass route. It works. Here is a thread that describes how I installed it.

PrimaCreator FlexPlate-Powder Coated PEI 220 x 220 mm

I am using a Prima Creator plate. Here are some pictures of the measuring and mounting. I made small squares, which I placed in Cura on the very edges of the print area, to see where to place the magnetic stick sheet. I was worried that I'd lose much print area, but as you can see, there quite much of the original glass that wasn't actually accessible to print on.

Temporarily solved by making the bridges two lines wide instead of one. At least they don't mush together on Y. On the other hand I have to be there to watch the actual start of the lines, because the second line covers the first one. So I'm still wondering if the single bridge wall line being printed twice is by design, and if there is a way to get rid of it.

I created a model to calibrate the speed and flow of bridge walls. The goal is to print a single line bridge, and make it as good as it gets before going to bridge skin. I expected the printer to print a line then go on. Instead, it prints the line a second time at the same position as the first line. It is problematic, because the second take goes over the first one, and since it isn't hardened yet, it often knocks it down. The gcode looks like this (my added comments after #): ;BRIDGE G1 F900 X99.325 Y92.107 E143.50126 ;TYPE:WALL-OUTER G1 F1200 X98.325 Y92.107 E143.51689 G1 F1500 X97.325 Y92.107 E143.5294 G1 F1725 X96.675 Y92.107 E143.53647 ;TYPE:WALL-OUTER ;BRIDGE # First take: X96 to X76 G1 F900 X76.325 Y92.107 E143.69557 ;TYPE:WALL-OUTER # Some kind of anchor in the holder G1 F1200 X75.325 Y92.107 E143.71121 G1 F1500 X74.325 Y92.107 E143.72372 G1 F1725 X73.7 Y92.107 E143.73052 G1 F2400 X73.7 Y92.057 E143.73091 G1 X74.763 Y92.057 E143.73209 G1 X76.325 Y92.057 ;TYPE:WALL-OUTER ;BRIDGE # Back to X96! Same Z, Y is very close to the first take, 0.05 mm. G1 F900 X96.675 Y92.057 E143.75482 ;TYPE:WALL-OUTER G1 F1200 X97.675 Y92.057 E143.75706 G1 F1500 X98.675 Y92.057 E143.75884 G1 F1725 X99.325 Y92.057 E143.75985 Is this the expected behavior for bridge walls, or is there something else that is wrong? I've tried to change the width of the bridge lines in the stl (source in onshape), but even with the lowe st value that still appears in the slice, I get this double take (0.40 mm width is not sliced at all, 0.41 mm slices with double take). UM2_bridge test.3mf UM2_bridge test.3mf

You were right! I solved it that way: And got 104.5 mm. Thanks!

Right! There is a flow rate field just below the extrude control in the octoprint dialog. I tried to set it to 20 %, but it somehow wouldn't bite. On the other hand I didn't want to believe in that, because I would expect the extruder to skip if that was the case. It felt smooth and silent. Was that a wrong assumption?