UM3 Idea - Abort before grinding occurs

[drofnas 15]

Not sure if this idea has already been suggested, but after a bunch of failed prints lately (ton of 16+ hour ones), I wish there some sort sort of sensor right above the feeder that can see if the filament is still moving when the stepper motors says it's suppose to be. If after some appropriate amount of time (maybe a setting in Cura) it can abort the print, then show the layer it was last printing on the screen.

This way someone could come back, clean the nozzle, ensure there isn't an issue with snagged filament, etc. And maybe resume where it left off at.

Am I off my rocker thinking this would be an awesome upgrade?

P.S. Right now I'm building the Mostly Printed CNC machine which requires pieces to have 50-80% infill, so they are rigid enough to handle the forces of moving around a spindle. I'm sure my failures are related to to-many retractions, but unsure, and I still have 100 hours worth of prints to do... shrug.

Edited February 5, 2016 by Guest

[neotko 1,417]

https://ultimaker.com/en/community/17860-ultimaker2marlin-and-filament-runout-sensor

[drofnas 15]

Thanks, I'll take a look at their solution. It would still be an awesome addon/kit for a UM2 (or the eventual UM3).

[neotko 1,417]

I would prefer a more complex, but solid solution, like a filament size reader (there's a few projects for marlin that do that) so if the filament size changes it adjust the flow to compensate. Ofc a simple 'running out of filament' pause, move to a corner and wait for the user to put new filament to continue, that could be really nice for less hardcore users.

[drofnas 15]

yes, but size is only subset of the issues.

Reading the forums it sounds like in my case I might have to much retractions during my 16+ hour prints, which then heats up the knurled feeder and eventually causes grinding to occur.

I want the printer to know if the filament is still moving or not, so it can abort before it gets to an "Air Print" scenario and a bunch of grind'ed up filament all over my feeder.

[mastory 44]

Sounds like a work-around for bad mechanics. Sure, programming can compensate for bad mechanisms, but it is really a band aid for the real problem. It would add unnecessary complexity and thereby more stuff to break/malfunction. Why not just install a proven feeder design that doesn't grind the filament. There are proven designs available.

[Labern 775]

I think if it was going to be implemented properly i don't think it would add to much complexity. A lot of production machines have monitor encoders. we have encoders that are over 20years old on our machines with no faults.

I agree that for the sake of grinding that its not the best solution, But for the sake of filament tangles, empty roll runout or blocked nozzle (mainly for filaments like woodfil) then this can be a very helpful solution. It could even measure amount of feeder slippage and bring up a warning if its to high. This could be an good early warning system for worn things like PTFE couplers or partially blocked nozzle.

[drofnas 15]

Why not just install a proven feeder design that doesn't grind the filament.  There are proven designs available.

 

If you got an idea of one I can install, I'd love to hear it.

Right now I'm already running with IRobertI Alternative Feeder, IRobertI Low Friction Spool Holder, 35W Heater upgrade, and an Olsson Block (0.4 nozzle). So I thought I was doing all the stuff to reduce these issues, but I'm still very much a nub.

 

I agree that for the sake of grinding that its not the best solution

 

Yea, your reasons are much better than mine. My specific reason is a little short sited just because of my current issue trying to print these parts.

Edited February 6, 2016 by Guest

[IRobertI 521]

Sounds like a work-around for bad mechanics.  Sure, programming can compensate for bad mechanisms, but it is really a band aid for the real problem.  It would add unnecessary complexity and thereby more stuff to break/malfunction.  Why not just install a proven feeder design that doesn't grind the filament.  There are proven designs available.

 

Bad or good mechanics isn't going to make a difference for tangled filament or dirt in the nozzle for example.

[nallath 1,124]

Why not just install a proven feeder design that doesn't grind the filament.  There are proven designs available.

 

There is no such thing as a "proven" feeder design, not when you take into account that an Ultimaker can be fed with pretty much everything. I do agree that hardware problems should be solved with hardware as much as possible, but in this specific case it's good to also have some (software based) feedback. A more closed loop system should greatly improve the quality of the machine.