Tested Runout function - didn't go well
Just for grins, I decided to test the filament runout/recovery function of my MK3S. I cut the filament a few millimeters above the printhead and sure enough, after a few moments, the sensor triggered, the head moved away from the print and it asked me to remove the filament. Unfortunately, when I pressed the knob, the filament did not eject from the top of the print head. When I indicated the filament did not eject properly, it said I had to remove the filament manually and the printer moved the print head a little to the left. When I opened the side door, it was clear that the filament was too low to be grasped by the gears and I had a devil of a time getting the rest of the filament out.
When I tried to autoload the filament, it would not work, but there was a manual mode to load the filament. The good news is that after inserting new filament, the printer continued the print as expected. However, that doesn't seem to be the way this should work. Any thoughts would be appreciated!
I'm using the latest firmware and printing through the latest build of Octoprint. The filament sensor seems to work as it should as autoload has worked perfectly and the menu system shows the sensor status updating correctly.
RE: Tested Runout function - didn't go well
I experienced the same issue on my Mk3S, no Octoprint. I have an email into support questioning it.
RE: Tested Runout function - didn't go well
are you using the stock MK3s IR sensor setup or the MK#S/MMU2s chimney setup?
RE: Tested Runout function - didn't go well
Sounds like the flag is not adjusted correctly (it is a really poor design). So that instead of triggering filament out as soon as the filament passed the ball, the flag arm was slow to move and block the IR path; letting the printer use more filament than it should. That or if the filament was past the gear, then the sensor never triggered at all.
Unload filament, then check for a 1:0 state change using the LCD status display and a loose piece of filament.
RE: Tested Runout function - didn't go well
are you using the stock MK3s IR sensor setup or the MK#S/MMU2s chimney setup?
I'm using the stock MK3S setup.
RE: Tested Runout function - didn't go well
Sounds like the flag is not adjusted correctly (it is a really poor design). So that instead of triggering filament out as soon as the filament passed the ball, the flag arm was slow to move and block the IR path; letting the printer use more filament than it should. That or if the filament was past the gear, then the sensor never triggered at all.
Unload filament, then check for a 1:0 state change using the LCD status display and a loose piece of filament.
Thanks. Testing the IR sensor shows it responds quickly to filament in/out. Not sure what else to check at this point. The printer clearly continued to extrude even after the sensor tripped. Of course, it is an N of 1 at this point. The part I was printing at the time had some pretty large straight lines and I'm wondering if the printer (or Octoprint) had to finish a long extrusion gcode line before it could interrupt the print. Any thought on that? I may try testing on a smaller part.
RE: Tested Runout function - didn't go well
I'm still thinking it was a sensor failure. It simply isn't that reliable (in my opinion). That, or it failed when the filament is moving slowly - and you needed it to work, and works fine when quickly moving filament in and out. A screw they use as an axle has been known to have too much friction if tightened a bit too far.
RE: Tested Runout function - didn't go well
I tried this on a second, smaller print. The operation happened flawlessly with the filament ejected correctly. In addition to the smaller object, I had viewed a video in which Josef Prusa demonstrated the function and said something to the effect that the cut had to be straight and not at an angle or it won't work correctly. So, I made sure the cut was straight. I don't recall paying any special attention to the cut I made the first time. I now have an N of 2 and am batting 50% 🙂