Filament ejects after a Z height around 120mm with MK3S MMU2S
Hi everyone!
I'm having a problem with tall prints on my MK3S & MMU2S. The filament is ejected (like it runs out of filament, but that's not the case), wants me to remove the filament and after "fixing" the issue (I simply cut the tip and pull it back in) it loads the filament, extrudes a bit, asks me if the color changed correctly and resumes the print.
After a print reaches a height around 120mm, the problem occures randomly several times until the print is finished.
This is a problem I'm facing with ALL bigger prints so far, no matter if MM mode or single mode, no matter which filament brand, no matter which filament slot and for Firmware 3_7_1 / 1_0_5 and 3_7_2 / 1_0_5
Once it even knocked over my purge block, because it caused a blob on it and I had to cancel the whole print (40 hours in a 60 hour print).
I checked the FINDA (tightened/loosened it), The filament sensor (calibrated several times), did run the wizzard again and again, reinstalled the firmware (printer and MMU) many times, used different SD cards, different GCODES, cleaned everything to make sure nothing is stuck, checked all the PTFE tubes, screws, connections and so on.
I'm really out of ideas right now.
It's weird, because smaller prints don't have that issue at all.
Any ideas what might cause this?
Best Answer by A.Dorn:
Ideas are plenty, but I'm also more or less shooting into the dark when facing random loading issues.
My current bet is on Mechanical problems around the IR-Sensor. Sometimes I tightened the screw that holds door too tight - resulting in errors after about 2 hours. If the door is not happy and it gets hot around door/spring maybe the IR-Sensor is more prone to errors. During calibration for me everything always looked fine.
Next bet is on the filament forming a fat tip (that has noticeable resistance in the tube). These might trigger the IR-Sensor differently. Theories about why sometimes the tip gets fat are plenty... (heatbreak, model, filament, temperature...). At least these are very easy to fix by simply cutting the filament.
Next bet is on physical resistance somewhere in the path... can be anywhere... maybe only a filament with a strange tip/bend will be able to see the problem.
The firmware could be a lot more helpful by telling why it rejected a filament. (IR, Finda, ?!)
Also during fixing the problem the firmware walks a bit into the twilight zone. Doing a "load to nozzle" of filament 1 once fixed my problems with filament 2 - but if there's an error while fixing the error the firmware most likely hits buggy territory - I've seen the LCD go completely blank, saw it trying to extrude with a cold nozzle, ... Basically anything other than simply cutting the tip and pushing a button to continue feels really unsafe.
I hope you find out more about your problem...
RE: Filament ejects after a Z height around 120mm with MK3S MMU2S
Ideas are plenty, but I'm also more or less shooting into the dark when facing random loading issues.
My current bet is on Mechanical problems around the IR-Sensor. Sometimes I tightened the screw that holds door too tight - resulting in errors after about 2 hours. If the door is not happy and it gets hot around door/spring maybe the IR-Sensor is more prone to errors. During calibration for me everything always looked fine.
Next bet is on the filament forming a fat tip (that has noticeable resistance in the tube). These might trigger the IR-Sensor differently. Theories about why sometimes the tip gets fat are plenty... (heatbreak, model, filament, temperature...). At least these are very easy to fix by simply cutting the filament.
Next bet is on physical resistance somewhere in the path... can be anywhere... maybe only a filament with a strange tip/bend will be able to see the problem.
The firmware could be a lot more helpful by telling why it rejected a filament. (IR, Finda, ?!)
Also during fixing the problem the firmware walks a bit into the twilight zone. Doing a "load to nozzle" of filament 1 once fixed my problems with filament 2 - but if there's an error while fixing the error the firmware most likely hits buggy territory - I've seen the LCD go completely blank, saw it trying to extrude with a cold nozzle, ... Basically anything other than simply cutting the tip and pushing a button to continue feels really unsafe.
I hope you find out more about your problem...
RE: Filament ejects after a Z height around 120mm with MK3S MMU2S
Based on your description it sounds like filament runout is being tripped. My guess is that things aren't perfect with the IR or FINDA sensors and the way the filament is getting jostled around due to the bend in the PTFE tube is just enough to have one or the other trip.
An interesting test of this would be to load filament to the extruder and then run some gcode that extrudes while moving back and forth on the X axis while raising the Z axis. While the gcode is running, watch the values in Support -> Sensor info. My bet is that as you approach that 120mm mark that you'll start seeing erroneous sensor values.
While obviously more annoying to troubleshoot, I expect the fix is just the normal path once you determine which sensor is the issue. My guess is that it's the FINDA as I believe it is responsible for runout detection.
RE: Filament ejects after a Z height around 120mm with MK3S MMU2S
does sound like the filament in the ptfe tube going to the extruder might flex just enough to move the ball in the finda. maybe lower finda just a tad.
I have a Prusa,therefore I research.
RE: Filament ejects after a Z height around 120mm with MK3S MMU2S
Thank you @andreas-d10, my IR sensor door was too tight as well. Loosening did the trick.
Thank you very much! 👍
RE: Filament ejects after a Z height around 120mm with MK3S MMU2S
so I bet the filament was flexing around the IR probe.
you can run a print with the sensor info showing, it gets odd when the printer goes into fault, but it is nice. almost wish they had 3 boxes to show on the screen
I have a Prusa,therefore I research.
RE: Filament ejects after a Z height around 120mm with MK3S MMU2S
Very well possible. It was pretty hot lately, so the PLA got soft. This might also be something to consider.
As the PTFE tube to the extruder is getting bent more in higher Z regions, this might have an impact on softer filament as well.