This doesn’t seem to be an issue with purging black from the white. More like there is black filament being deposited unintentionally or ousing out of the nozzle after it has made that specific area. I have had issues with purged filament getting stuck and ending up on the plate but these are small areas and there is no surface defect, just the small color "bleed’.
You can turn on Quality>Avoid crossing walls, but not sure it will help here.
I would imagine the black has moisture and is oozing at temp, you can verify by manually heating the nozzle to 220+ and then manually extrude some filament, clip it off with tweezers quickly then see if it continues to creep out. If it does it suggest moisture and drying might help.
Well yes moisture in the filament is one of the major root causes of nozzle oozing and stringing.
Multi color prints with high color contrast differences are always difficult and need perfect settings to minimize defects.
Recommendations from my side:
Calibrate flow and K values for each of the filament colors since the different pigments can influence these characteristics even if the underlying base material is the same.
Also do stringing tests for each and fine tune the retraction and nozzle z-hop and wipe settings.
Try with Arachne instead of Classic Wall Generator as this can sometimes reduce the amount of z-hops and travels over the different colored areas. Note using “Avoid crossing walls” won’t help in such multi color areas since each color area is seen as a different entity by the slicer so will still cross over them.
Manually painted areas usually go several layers deep and depth is not uniform across the part and there is no control over that (as far as I know). This significantly increases the amount of nozzle crossings of the different colors and with it the chances of having unwanted strings of the black material deposited into the white areas. Better would have been a design where the black color is its own object as an inlay to the white main part with a height of only one or two layers to the black colored inset.
quality and appearance would be far better if the colored insets are done flush on the first layer(s) of the build plate (face down) but with this part design that would impact the appearance of the outer bevel area as well as screw hole recesses the latter of which would need supports if printed in this face down orientation while removing the need for supports on the connector recesses on the rear with the part orientation you currently have.
Thank you for all of the tips, multicolor 3d printing is an extra depth of 3d-printing I don’t (yet) have much experience with.
Thank you for the detailed steps on what could improve the result and I will try to apply all of your suggestions to my future designs.
One quesion does remain for me, for your fourth suggestion you shared that a seperate object of the black text would be a better way to design the black areas instead of painting them in the slicer. What is the benefit to design seperate object for it instead of painting? Is it becauce of the recessed area that I used to be able to more easily paint the text or is there a different more substancial benefit in the slicer by using a seperate inlay object?
The main reason is that you can design to a specific depth so that it will print a defined number of layers in that color. Manual painting usually causes varying depth in the painted areas without you being able to control it. You will see that with your connector panel model in Bambu Studio when you go into review and scroll trough the layers, there are black areas that are 2 layers deep and others that are up to 5 layers deep.
Every layer with color change increases the chance that some unwanted black gets deposited onto the white area causing a visual defect. Black on White can read through several layers since White is rather translucent. Black on the other hand is opaque thus one print layer of it is enough.
This also reduces the amount of wasted material due to color changes and reduces print times (your 3mf has 11 color changes that could be reduced down to 2)
