These cylinder shapes have proved difficult for me to print on a Taz 6 with nGen Blue. It doesn't matter if I randomize the start points, so I lined them up here.
I've tried an exhaustive combination of temperatures, retraction, calibration, extrusion multipliers, coast, negative extra restart, and wiping. The "fix" has been to change Outline Direction from Inside-Out to Ouitside-In and disable Minimum travel for retraction.
I'm not particularly satisfied with this solution. The reason Inside-Out creates zits is because of gcode and tool path logic. Here is an exaggerated Inside-Out illustration.
The zits happen because slight imperfections (e.g. extra filament) in the adjacent interior perimeter bump out the final outside perimeter. When the inner perimeter finishes (2), regardless of retraction, the movement (3) deposits a small amount of filament. The next perimeter exaggerates the bump.
I see a few ways to fix this with additional (new) settings in my humble order of preference:
• Allow "Perimeter Width" in % - This would allow experimentation in printing thick walls.
• Make the perimeter loops continuous when possible (see below)
• Randomized start points for all perimeter loops (is a loop what I think it is?) so perimeters don't share a common start/end
Here is what a continuous perimeter loop could look like:
Here, the end of the inner perimeter loop (2) keeps extruding as it meets the start of itself (1), makes a left and then right turn (3) and is now printing the start of the outside perimeter (4). At the end of the outside perimeter (5) the slicer knows that there is an extra extrusion width and stops early to accommodate.
Thoughts?
