Print head is back together after the self induced meltdown. I added the two changes I mentioned (better seal fit, and a backup seal with a path for leaked filament) to the existing print head. Welcome to version 9.1.
I immediately used the leak port as I didn't tighten my set screw and therefore failed to spring load my seal properly the first time. Cleanup, another new seal (as I damaged it on removal), and another assembly and I was back in business.
Testing got serious today with a single wall test print. Printed about 30 of these.
These are the last 9, about when I started to get everything dialed in as best I could.
These are at 0.2mm layer height, printing one color per layer, using cyan, yellow and magenta only. White just made it lighter, and black made it look like mud at 0.2mm. At 0.35mm, black was nicely contrasted. Black takes just about the whole layer on one of these to purge as the saturation is so much higher than the CYM I have.
I started to double the bottom layer for better adhesion as my bed wasn't exactly level, and so I didn't have to keep changing my tape.
At that layer height, the only time you can really see the color is if a filament drive skips. You have to look really close, but every layer is a new color. I still haven't changed my damaged stepper drivers, and as a side effect they run pretty hot so sometimes I have to give those a push.
I still get a little bubble where I change layers, with or without a color change. As all extruding is controlled by the velocity extrusion calculations, and not E values, I'm looking at the code for clues.
Retracting while feeding the old, changing the color code, and then un-retracting the new resulted in the fastest color changes. It also made the layer change bubble bigger. More retraction doesn't fix it. I'm assuming the idle filament continues to expand slightly while retracted. I can *probably* fix that..