brunoosti

In my case it does nothing (rotated to the part so as one of the walls would be parallel to the lines). I'm using arachne engine.

Sure, I understand what you mean: for precision and not having the plugin escaping the intended shape. But sometimes it's evident the part would be round if not for the low poly count (besides skipping treating the model part ). How does the plugin calculate the curved path? Tangent to the original lines?

Now, how do I convert a low poly shape into a curved one? (like this shape attached) It seams like tolerance and resolution should do the trick. But if I configure resolution over 0.07mm, cura won't slice, whatever the tolerance. Overhang_Test_Solid.stl

Hey! After changing that file line, I think Arc Welder indeed "fixed" Junction Deviation. Making some tests right now, but the part that always gave me zits before is completely smooth. No rattling during curved moventes and great corners. The only thing a bit different is that the extruder is more aggressive on the corners with linear advance. Give me some curved surface to test with that always gave you problems using JD.

Solution for slowing too much before some curves (anycubic i3 mega-s with that firmware in the link, in my case): https://github.com/knutwurst/Marlin-2-0-x-Anycubic-i3-MEGA-S/issues/80 henniero commented 8 hours ago I had the same issue. I found this thread in the reprap forum https://reprap.org/forum/read.php?415,869738,869785 I changed line 219 in Marlin>src>gcode>motion>G2_G3.cpp if (!planner.buffer_line(raw, scaled_fr_mm_s, active_extruder, 0 // "0" was "seg_length" in original version I replaced "seg_length" with "0" at the end. I built a new firmware with my printer-configuration from the platform.io envelope choice. Issue fixed for me with that version.

Before it even showed the gcode file size savings percentage. Still didn't get any luck with my anycubic i3 mega-s. Could it be the minimal segments is set to 24? You can really notice the issue when printing a Benchy. The printer sloths at each lettering under the boat, not extruding (maybe because the printer is so slow that the extruder cant "keep up") or melding what it had extruded.

@FormerLurker, I'm running a slightly modified version of Knutwurst's i3 MEGA (M/S/P/X) Firmware, based on Marlin 2.0.5, it seams: https://github.com/knutwurst/Marlin-2-0-x-Anycubic-i3-MEGA-S My board is also 8bit. Could that interfere?

Can someone confirm this? when printing the first layer of a benchy (lots of curves on the letters) the printer slows down so much that the letters get melted, or the extruder simply is not that precise enough and cant extrude so little filament, leaving an under extruded first layer. It has no relation to linear advance, as I've tested with and without. Maybe it's my printer's firmware arc settings? // G2/G3 Arc Support // #define ARC_SUPPORT // Disable this feature to save ~3226 bytes #if ENABLED(ARC_SUPPORT) #define MM_PER_ARC_SEGMENT 1 // (mm) Length (or minimum length) of each arc segment (default: 1) //#define ARC_SEGMENTS_PER_R 1 // Max segment length, MM_PER = Min #define MIN_ARC_SEGMENTS 24 // Minimum number of segments in a complete circle (default: 24) //#define ARC_SEGMENTS_PER_SEC 50 // Use feedrate to choose segment length (with MM_PER_ARC_SEGMENT as the minimum) #define N_ARC_CORRECTION 25 // Number of interpolated segments between corrections #define ARC_P_CIRCLES // Enable the 'P' parameter to specify complete circles #define CNC_WORKSPACE_PLANES // Allow G2/G3 to operate in XY, ZX, or YZ planes #endif

In addition to my tests, it seams Arc Welder aid some Linear Advance calculation issue where the printer would rattle a bit in inside walls (or maybe overlapping ones - I'm printing a 2mm curved shell), trying to compensate extrusion vs path, with not that smooth movements. With Arc Welder the printer must input the long arc for Linear Advance to calculate extrusion and movement is always continuous and smooth. That is such an improvement : )). Sometimes I still see the printer almost stopping at corners right before a curve, extruding absolutely no filament (LA in the works) but leaving that melted dot as described before.

Deleted. Sorry, had a hard time trying to quote my previews message. intended message below.

I'm running a slightly modified version of Knutwurst's i3 MEGA (M/S/P/X) Firmware, based on Marlin 2.0.x: https://github.com/knutwurst/Marlin-2-0-x-Anycubic-i3-MEGA-S I don't think it has any new implementation of JD as on a recent update he disabled it.

Testing junction deviation with linear advance and arc welder: First test: linear advance + junction deviation + arc welder + "G90 influences extruder" on. This caused the extruder to vibrate a lot on corners, with the XY movement almost completely stopping there. Gave up the print on the first layer afraid it could damage de extruder. Curved motion was smooth on the XY axis. Second test: linear advance + junction deviation + arc welder + "G90 influences extruder" off. Same. Maybe LA is causing the issue. Third test: junction deviation + arc welder + "G90 influences extruder" on. No vibration on the extruder. Overall smooth movement, but the printer still stops for a second on some corners leaving a melted spot (not a blob of filament). Forth test: classic jerk + linear advance + arc welder + "G90 influences extruder" on No vibration on the extruder. Overall smooth movement. No stopping on corners. It seams the problem was JD all along. Unfortunately arc welder does not help "fixing" junction deviation.

Cool! Hope so : ) (the outermost is a skirt and the rest would be a 2mm cylinder) I'm preparing everything and finishing a print right now. Will come back with junction deviation results.

There you go. Maybe it's just the graphical representation,? But this is what i see in NC viewer: normal.gcode with AW.gcode

Oh sorry! That should be a circle. See the gcode with no WA. The one with WA shows more straight lines in the arc (if that makes any sense)