I’ve had a Creality CFS upgrade for my K1 Max sitting on my workbench for probably a month waiting for me to install it. Part of that delay is that I knew it would take a while to install, and I am glad I waited.
I finally got around to doing that yesterday, and I have thoughts… First off, the Creality installation documentation is ok, but not great. I would have been in a lot more pain without Thinkering with Jerd’s tutorial video. So thank you to Jerd, whoever you may be. I also agree with Jerd that the extensive use of hot glue on connectors is super annoying. While I suspect it makes sense in terms of ensuring that devices work when they arrive to a customer, it makes upgrades super annoying and surely there is a better way. Overall I think I spent at least six hours on the install.
But what about multicolor printing now that I’ve finished the upgrade? Well, my initial observations align mostly with my expectations in that:
Printing is slow if there are color changes. This is especially true if you include the time taken by filament jams and failures to feed that I didn’t notice immediately. My first non-trivial test print will likely take 16 hours to complete. So, you can’t just set a complicated print going and go to work. It will be unlikely to have finished by the time you get home because it will jam somewhere along the way. This is especially annoying given that the way I’ve been clearing the jams is simply to press “retry”, which seems to resolve the issues. Could the machine not do that for me?
Multicolor printing is also a bit wasteful, even when using “flush to infill”. I was expecting this, but yeah its definitely a thing.
These are of course only true for multicolor prints, if you treat the thing as a loader which can deliver whatever single color you need for this specific print then its pretty cool out of the box.
To address the multicolor print issues, I have learned you should minimize the numbers of colors per layer to reduce time and waste, as well as ordering some unobtainium PFTE tube replacement which alleges it will jam less by being “more slippery”. We’ll see how that goes I suppose.
Finally, I was also surprised by how the filament change worked. I assumed that anything already loaded into the feeder tube would be flushed, and therefore tube length should be minimized. This is in fact not true, the filament gets retracted — the change of filament waste is driven by flushing the extruder and having to clean the extruder between colors. So, you can in fact have the CFS further away from the printer than I expected, although I had to consume PFTE tube to learn that. I do expect that the longer the feed tube the more filament failures to load, but then again the tighter the bends in the tube the more of those too, so there is a balance here somewhere.