Filament Box
Solution: Pay more attention to filament tangles :-).
I've been trying to print the digit chains for my silly clock using the newly installed MMU3 to print the two color chains.
My first problem was that with all the extra moves back and forth to load black and white filaments, it was easy to knock over the chain links since they have very little bed surface.
That led me to create a custom "brim" to embed in the print to give the chains a better chance to stand up. I designed them with little "sprues" I can easily cut off rather than the normal slicer generated brim which would be a nightmare to remove.
This was printing OK, till the PTFE tube for the white filament apparently got sucked up into the filament bin somehow. That led to the white filament jamming and the print failing.
I cannot for the life of me envision any way force could be applied to the tube to make it want to climb up into the filament bin. The filament is always pulled down, not up. The filament retraction by the MMU doesn't reach the bin, the excess filament merely unwinds into the buffer.
I got everything straightened out and tried printing again, only to notice that a bit of tube was once again sucked up into the box. Why? What is going on?
For now, I have an alarm set every 20 minutes reminding me to check the tube and drag it back down if it has crept up again. Maybe I'll get through the print this time.
If I can't figure out what is going on, maybe I'll have to 3D print some sort of collet to grab the PTFE from the outside and prevent it from rising up into the box. Perhaps some bracket to hold more PC4 fittings in the reverse direction under the box.
I've been using this box for months before I got the MMU with zero weird problems like this.
...OK, it printed the entire thing with no errors. I hope I don't have to always check every print every 20 minutes in order to avoid errors :-).
The only theory I can come up with is that the spool was wound with the filament too tightly wedged in at some point and it wouldn't unwind, resulting on the printer pulling all the slack out of the filament which forced the end of the PTFE tube up into the box. Maybe it got past that and it no longer has any problems. I'll see if I have more problems in the future...
I now have a bit of confirmation. I heard noises from the filament box while printing with the same filament, and observing it, I see the filament, instead of taking a nice smooth path into the PTFE tube, sometimes gets bent down like it is stuck to the next layer of filament, then with a pop it will come loose and spring back up into the curved path you'd expect. Looks like this spool of filament may be problematic (though it hasn't yet got stuck enough to jam again).
Update: While I was writing the above paragraph, the spool got pulled off the rollers and I got a stuck filament detected error :-).
Here's what I found inside the filament box:
Here's what a closer examination of the spool showed:
Apparently that noise I was hearing was the filament somehow being able to unwind and advance the knot every so often. I'd swear everything was perfect when I put the filament in the box after setting up the MMU3, but maybe it somehow managed to unwind under the knot for hours and hours before this failure.
Anyway, I used the Gordian technique and just cut the filament above the knot and reloaded back to the MMU from scratch. While I was at it, I brushed any filament dust out of the hob gear. Maybe this problem is fixed?
At least the new stuck filament detection worked this time. I guess it wasn't stuck enough for the first error I got.
I've been printing with this filament quite a lot since I fixed the tangle, and there have been no more problems, so I guess it is indeed fixed.