Currently when a large print job is executing and one nozzle is not being used for a longer period, a print job could fail if the filament needs to be unloaded later on when another filament is required, but the dormant filament broke in the PTFE tube, causing unload failures, due to a part of filament being stuck in the tool head.
To prevent this, it would be good that the filament is unloaded after use or after staying dormant for a predefined period.
To my opinion, retracting the filament, when it is foreseen that it wont be used for a longer time, can improve the stability of the process.
I had this issue regularly at the end of a print job, because the printer wanted to unload both filaments but one was broken.
At the end of the print, not so much of an issue, but during a print, i can stop the print.
It would be nice if this is conditional and based on the flexibility/sensitivity to break of the filament.
External spools should be excluded from this behavior.
I am considering to make a workaround in the G-code sections for “Change Filament” and “Change Layer”
Basically a filament change would reset a counter for the left or right nozzle.
On every layer change the counter is increased for the inactive nozzle.
If the left or right counter reaches a certain threshold, the filament can be unloaded for that specific nozzle.
I need to check if there is then specific action required when the same or other filament is required again.
I presume that a change to a new filament is properly handled by the slicer, but when it want the use the same filament (of which the the slicer assumes that it is still loaded), a full load of the filament needs to be triggered again.
The threshold could be time or a number of layers.
Time would be the best, but layers is probably easier implemented.
By the way, I even have PLA broken when dormant for 10 hrs.
Just a thought here, but… Could the unload the filament by backing it away from the nozzle a couple of inches would cure this and if the same filament is again needed then it could move back in quite rapidly. By doing this at every change they could reload quickly or retract the filament back to the AMS if a color change is needed.
It might be a consideration but it needs to pull back at least 200mm to het out of the danger zone.
And I wouldn’t want to do this, if the same filament is required again 2 layers later. Then you lose more time than having a full retraction only when really needed.
I also think that for stability, you might want to prevent that there is a 3rd state between loaded and unloaded, where in theory you don’t exactly know where the end of the filament is.
Agreed, it might as well pull it back all the way to the normal position. I think it would much simpler to keep track of where a given filament is. IE: Either ready to be loaded, or not loaded at all.
I also would like to see this as an added feature. I’ve been lucky so far but I’m sure my time is coming… Maybe allow a setting that turns it on and even how may layers to kick in the retract.
I will forward it to our team for their consideration.
If possible, please provide a few use cases where this has happened to you, and what filament broke in the PTFE tube, to better asses the situation and find the best solution.
I had it happen with even just normal PLA, when printing a large part. Especially if the tool head is moving in a way that the ptfe tube has been repeatedly bent but no extrusion happens (so the bending strain would repeatedly damage the same spot for the filament in the left nozzle)
A theoretical case is, a 30cm x 30cm x 30cm solid cube, with ONLY the first and last layer using the left nozzle, and all other layers using just right nozzle. With something like this the yellow filament which stayed in the other extruder path is definitely gonna break
In real world it doesn’t have to be a solid cube, but as long as there’s repeated bending without (at least some length of) extrusion in the left nozzle, then it would eventually break.
This is exactly the case.
However in your use case, it might not go wrong, because the same filament is used again on the top. Big chance that the AMS will push the broken bits into the extruder and all goes well.
There are at least 2 scenarios possible:
There is no 2nd color change at the top.
The printer tries to unload the filament at the end of the print job. the broken bit is not removed, and the printers stops at the end of the print job during the unload with an error.
Result: print job complete, but error at unload.
The top layers require a different filament.
In that scenario the AMS also tries to retract the filament, but now the print job cannot be completed because of the same error.
Result: print job incomplete, and possible quality issues due to cooling or other effect due to the stall of the print.
This happened to me already 5 times in a project with a large model and was in this case also PLA Basic.
The large model (lots of XY motion of the tool head) and the PTFE tubes hitting the right side of the printer are amplifying this effect.
I also had this issue a few times with a long print and PLA+ loaded in the right nozzle. Because the right ptfe tube is constantly pushed against the side wall if the print head is on the right side of the build plate, the filament will eventually break in the ptfe tube, which causes an unload error.
I hope Bambu Lab will implement this feature request, because this is needed!
For the time being, playing you model as far to the left on the build plate will help (if there is room).
We have this problem also with large objects.
There are imprints on the bottom where filament of one nozzle is only used for some layers. However it stays inside the tube all the time and when the print is finished it is broken. Even with very fresh and dried PLA.
