Thats a good question.
Notice that the skirt printed oddly too, it printed layer 1 and then printed layer 2 about 50% off of position.
You may be right, the yellow is not shifted at all, only the red, and the red is shifted twice. Perhaps it is a firmware problem after all. The yellow got nozzled up somehow and all melted on the top, but is otherwise oriented correctly.
I have printed a true multicolor print, this crazy cat thing directly after those cubes. It came out okay, alot of extra extrusion spaghetti all over the place, but it didn’t experience the same problem. So if it was a firmware situation, why would something as complicated as this come out (somewhat) correctly but a simple pair of wall calibration cubes go so nuts?
This machine is such a weird oddity.
Currently running PETG. Which I will say, has been coming out reasonably well considering the length of time I spent on tuning the machine in. I always had a lot of problems with petg, but i barely did anything but esteps and it came out like this, even without my glass bed on.
I need to better understand the pre and post tool change gcode to properly retract. I don’t quite understand how I should be doing this, but in my head I am thinking it should be like this:
If the new tool is tool 1, set tool 0 active, retract, then set tool 1 active.
if new tool is tool 0, set tool 1 active, retract, then set tool 0 active.
But I am not sure if that is the proper approach, whether i need to re-assign the new tool back to active tool, etc.
or inother words, i am not clear if post tool changes occur the line before the change or the line after the change, and whether or not i need to assist with the tool change itself.