Topic: Color changing process midprint and weird happenings
I've been doing a lot of playing around with changing colors on my current project and I seem to have run into an interesting trait of the gcode. When I change colors in the middle of a layer and resume the print there are a few things to note that happen. As a precursor, I'm still using Pronterface and slicing with Slic3r .93 since it has been working for me. RH hasn't loaded onto my Acer netbook I've committed to the printer yet and it's hard to argue or change what's working.
1. Before resuming, the extruder evens out the reverse/extrude balance. For example, I reverse 25 mm twice to remove the filament and then extrude 25 mm twice to get the filament extruding again before resuming. However, if I reverse 35 which is just enough to remove the filament, but it takes 45 usually to get it extruding again, the extruder will reverse 10 mm before starting again. This means when it resumes, it will start off 10 mm behind and leave a very thin layer of oozing filament for a small space. Not a huge deal, but enough to ruin a print that is color dependent. Therefore, I just make sure to reverse and extrude the same amount.
2. Be careful of ooze. As soon as I pause, I lower the bed 10 mm. I have my tweezers ready to remove the little blob of plastic that inevitably rests on the print. I also remove the stringy ooze from the nozzle immediately. I then just lay a little piece of paper under the nozzle on the print to catch the pre-extrude filament. This has worked consistently and is quite straight forward.
3. Z backlash is not the same in both directions. I had to play around and screw up a good bit before I realized that it is difficult to get back to exactly the right z height. Even with the backlash nut inserted, if I lower the bed 10 mm and then raise it back up 10 mm, it comes back to about .07 mm above where it was last at. In my case, that means that I raise it 10mm and .1 mm when I finish changing filaments. It makes the layer ever so slightly smooshed since it has .03 mm less room to work with. Not enough I notice, but it is enough if I don't raise the extra .1 mm.
4. The biggest anomaly is that sometimes the extruder will slow way down for what seems like an arbitrary amount of gcode commands. This is a computer controlling everything, so I know it is not arbitrary, but I can't figure out a consistent pattern of what makes it happen. I've changed colors about a gazillion times, and the things that it does not appear to be related to are; the amount of time into a layer before changing filament, the amount of time remaining in a layer when changing filament, how long it takes to change the filament, whether it is on a perimeter or infill? ( not positive about this one), what speed the extruder is supposed to be running at. This is by far the most annoying anomaly, because a large flat solid layer will take half an hour when this happens sometimes and it ruins the adhesion, the extruder seems less good at flow rates at this speed, and it bothers me. It seems to run at about 2-3 mm/s. I have started digging into the gcode and firmware, but still haven't found the logic argument that causes this to happen.
So anyways, I get some really cool prints and great effects when changing colors, but would really like to make this a more consistent process. It seems like I have mostly everything figured out but the last anomaly. I'd love to hear what process others use and if anyone knows what causes the slow extruding/maneuvering after the change, you could save me a lot of time reading lots of lines of code.



