Filament Feed - feeder issues with UM3s

[FalmouthLouis 3]

Every so often I have a long print fail because the drivers at the back of my UM3s stop feeding the filament through - instead they just eat into the filament so that all forward movement stops.

 

This has just happened to me about 75% of the way through a 2 day print. This was despite the fact that I had been checking every couple of hours that the filament was hanging loosely on the spools at the back. I put prints on hold over night, and that was not an issue. There was no sign of a filament tangle. Since I didn't spot that something was going wrong until several hours after the breakdown (breakaway scaffolding was still printing happily), I can't totally guarantee that the PLA reel had not been offering too much resistance at the point of the failure.

 

I have a sense that having two spools running simultaneously at the back of UM3s causes problems because (i) there's a fair amount of resistance as the spools rotate and (ii) the far spool is being pulled forward as well as rotated.

 

So, if I want to improve the reliability of long prints, will I need to think about one of those systems whereby the filament is fed upwards from below? If so, what is the most cost effective recommendation>

 

 

[Smithy 1,145]

I don't think it is caused by the two spools, this is normally working fine without problems. But it can have several reasons and you should try to find out if the problem is before your feeder or after it. As a third option, it can be the feeder itself, the knurled wheel can also wear out, especially if you print a lot of abrasive materials.

 

Check also if your nozzle is clean and not clogged, maybe try if it is the same with another print core. Additionally, check your Bowden tube. The tube should be changed every year to keep friction low. You can open the Bowden tube at the feeder side, heat up your print core and feed manually filament through it. Then you can feel it there is something strange or not. 

 

Another reason could be that some materials are more prone to tangling than normal PLA for example, especially if it comes to the end of the spool and the filament is winded with a smaller diameter.

[gr5 2,224]

The feeders can pull/push with more than 10 pounds or 5kg of force.  A little friction of the spools touching is too small to matter.  I'm going to split your problem into 3 categories of problems (although there are 20 possibilities):

1) Feeder

2) print core

3) too many retractions

 

#3 is simplest to fix and diagnose.  Some layers may have hundreds of retractions.  By default Cura lets you do 30 or 40 retractions on the same spot of filament and that can grind it up.  10 retractions on the exact same spot is safe.  20 is pushing it.  40 almost guaranteed to fail.  So look at the failing layer in cura and count up retractions to get a feel for this.  You can change retraction settings if necessary.

 

It's probably more likely #2 than #1 but #1 is easy to test.  You need this thing anyway so print one of these to help with loading, unloading filament and testing things (to physically feel yourself what it feels like when working properly):

https://www.youmagine.com/designs/wedgebot-for-ultimaker2

 

Now feed the filament half way down the bowden and do "move material" on the menu to energize the extruder and then pull down on the filament hard.  You should be able to pull at least 10 pounds without the filament slipping.  Tension on feeder should be in the center position of the indicator.

 

#2 - try changing cores.  Obviously if that fixes the issue then the problem was with the core.  3dsolex cores are nice because you can change the nozzle.  A bad core can usually be cleaned with hot and cold pulls (see this feature on the UM3 menu system).  You may have a "bad" core.  It's kind of a mystery to me what makes a core bad but it's common.  I think it's related to the teflon insdert somehow.  Even 3dsolex cores can be problem cores (but most are fine). (I sell 3dsolex cores so I'm biased but I do believe they are better).

 

Once you get a tangle in filament it will not go away unless you remove the tangle.  The tangle just travels along forever if you don't fix it.  If you never lose the end of the filament (hard to do) it will never tangle.  It is possible to remove a tangle by passing the entire spool "through the knot".  Hard to explain but easy once you figure it out.

 

[Whiteknuckle 1]

I have UM3 with .25 core. When I load filament I get a good stream through the core but when I print, it does not make a puddle but extrudes a small amount and drags it to the print, the filament never gets to proper flow and collects around the hot end.

I have looked a the filament...it has no bad grind spots. I would appreciate suggestions. JP

[FalmouthLouis 3]

Thanks GR5. What you say about retraction intensity and print core reliability is very interesting. I'll do some thinking and checking.

[Smithy 1,145]

On 1/30/2019 at 7:12 PM, Whiteknuckle said:

I have UM3 with .25 core. When I load filament I get a good stream through the core but when I print, it does not make a puddle but extrudes a small amount and drags it to the print, the filament never gets to proper flow and collects around the hot end.

I have looked a the filament...it has no bad grind spots. I would appreciate suggestions. JP

Your print core is probably a little bit clogged, so you should do some hot/cold pulls to clean the nozzle. Additionally, if priming is ok but during the print not, you print maybe too cold or too fast.

[Whiteknuckle 1]

Thank you for the reply, I did hot/cold pulls and did get a bit of junk out of the core. I am going back to basics on the settings.

 

I am attempting to do more than my experience allows. I am printing nylon open ended capsules 16mm long and 9mm in diameter with a .5mm thick wall and port holes around its periphery . I have achieved good prints but I am now trying to enlarge ports in the capsule wall that result in overhang printing....tough.

 

Thanks again.