Trouble with feeder and PLA

[mastheraldo 0]

Hi guys

I've been having problems with some prints. I'm using cura 12.07 and since V0.3 this has been a problem.

With some large prints, there is a point in which the filament just wear so much that the motor isn't able to pull it to the nozzle and the prints just stops.

As you can see on the image.

Then I started to play with the settings, but there is always the same problem in the same point regardless the settings.

Someone has found out a way to solve it?

There are some feeders on thingverse with bearings or springs to avoid a lot of friction on those parts but i haven't had luck with those (because when I printed there is always so much friction on the walls or or the parts just don't match with the design).

At this point I'm trying to print a large print with ABS but what a really want it is to print with PLA.

Could you help me please?

Thanks in advance

[edgar-meyer 0]

Usually towards the end of a long print job the extruder fails - and the problem seems to be that the feed mechanism is unable to release the tension built up in the filament, so the feed-bolt just chews away without advancing the filament.

Might it be that the spiral thread of the modified bolt causes the twist-torsion problem?

EM

[jcosmo 5]

I've seen comments that the UM stock silver can be problematic, I've a roll I've only just started so can't be sure. It's don a couple of 1.5 hour prints without problems though.

[edgar-meyer 0]

As the photo shows, the hobbed bolt makes a slight imprint on the PLA filament as it advances.

As the photo also shows, the imprint is spiral, indicating a spiral twist to the PLA filament - until a tension point is reached, where the PLA filament returns to rest position and the process repeats itself, about every 20mm or so - and here's the rub, as best I can tell - the PLA filament is scarred in transition and the residual PLA begins to till the grooves of the hobbed bolt.

Ultimately (i.e., usually shortly before the end of a print session!) the grooves and gnarled surface of the hobbed bolt are filled and the bolt can no longer grab the filament but continuing pressure causes it to take a bite out of the filament without advancing it, as shown in your photo.

I may be mistaken, but it appears that the fatal flaw is the use of a threaded bolt at the site where it is made hobbed. The way to test this is to use a bolt with blank stock at the crucial place - in fact, Ultimaker should supply such bolts or else refute my hypothesis, that the hobbed bolt is poorly designed and should be replaced with one made from blank stock.

Counter evidence is invited.

EM

[jcosmo 5]

Now you've got me going back to look at the markings on my filament in more detail

[joergen 2]

As the photo shows, the hobbed bolt makes a slight imprint on the PLA filament as it advances.

As the photo also shows, the imprint is spiral, indicating a spiral twist to the PLA filament - until a tension point is reached, where the PLA filament returns to rest position and the process repeats itself, about every 20mm or so - and here's the rub, as best I can tell - the PLA filament is scarred in transition and the residual PLA begins to till the grooves of the hobbed bolt.

Ultimately (i.e., usually shortly before the end of a print session!) the grooves and gnarled surface of the hobbed bolt are filled and the bolt can no longer grab the filament but continuing pressure causes it to take a bite out of the filament without advancing it, as shown in your photo.

I may be mistaken, but it appears that the fatal flaw is the use of a threaded bolt at the site where it is made hobbed. The way to test this is to use a bolt with blank stock at the crucial place - in fact, Ultimaker should supply such bolts or else refute my hypothesis, that the hobbed bolt is poorly designed and should be replaced with one made from blank stock.

Counter evidence is invited.

EM

the filament isn't twisted in the process... the V2 bolt just goes around and leaves the tooth marks on the same side of the filament. no extra force on the filament. so the threaded bolt isn't a fundamental flaw, it works perfectly: I had mine since last September, and no issues (I think I cleaned the bolt 2-3 times since them, print many kilos with heavy retraction). sure V3 might be an improvement, but V2 is already pretty good. V3 is machined, and it wasn't a bolt in it's previous life.

[alaris2 1]

I'm afraid my 6 months of printing experience sides with Joergen on this. the V2 bolt is actually pretty good, rarely needs cleaning and I've found no reason to use the V3 bolt. the 'faults' lie mostly with the hotend.

[edgar-meyer 0]

Glad it works for you. But, by holding the filament lightly, I have felt a 'jerk' which I presumed was the filament releasing tension.

So please tell me, what parameters have you varied and optimized at the hot end: temperature? nozzle size? filament size? (affecting extrusion volume), filament feed rate? or am I missing something?

thanks,

EM

[joergen 2]

Glad it works for you. But, by holding the filament lightly, I have felt a 'jerk' which I presumed was the filament releasing tension.

So please tell me, what parameters have you varied and optimized at the hot end: temperature? nozzle size? filament size? (affecting extrusion volume), filament feed rate? or am I missing something?

The most critical items are:

1. enough tension on the filament knob to hold the filament, but not deform it

2. once you've set the proper tension, check the e steps per mm, since it greatly depends on the tension set in #1

3. measure your filament diameter, and set it accordingly in the slicer

4. test by hand (cold hot-end, nozzle off, clean brass tube) if the filament can actually pass through the whole setup without any impediment: see if the bowden tube clamp on the cold end is fine, and especially if the screws on the hot end are too tight, and cause constrictions at the bowden-PEEK interface

5. make sure the Bowden tube can't pop out: clean the metal teeth in the white clip, or print the Owen clamp, or use Joris's ingenious M7 nut fix

if those 5 items are fine, you can print with a wide range of print parameters. each plastic spool has it's own temp setting, so start slow (50mm/s) and warm, and reduce the temp slowly from hot&runny until you feel the filament grinding and the extrusion becomes matte, not shiny. dial the temp 5C up, and you got a good printing temp. (set the nozzle diameter to what it is in reality for now: 0.4mm)

[edgar-meyer 0]

Thanks

step 2 seems especially relative - I am using Cura 12-08 - so I get the number of steps from the value for the nozzle size or from the filament diameter - or elsewhere?

EM

[edgar-meyer 0]

step 2 - continued...

shat is a 'ball park' number for the number of steps per mm?

thanks,

EM

[joergen 2]

step 2 - continued...

shat is a 'ball park' number for the number of steps per mm?

 

no, that is a very clear and well defined number, almost scientific:

you remove the bowden from the cold end, manually align the end of the filament with the top of the grey plastic tube quick release, and run the initial calibration routine of Cura (which moves 100mm of filament through). then you measure how much filament was actually pushed through (maybe 102 or 98mm, or something like that), and enter that value into dura, or manually calculate the new e steps/mm for the FW settings permanently: 100mm/measured mm * previous e step per mm (865.888).

most likely, you will see 100mm extruded, and can leave the default value of 865.888

[zorfor 0]

Hi, I posted my feeder experiences in another thread..

viewtopic.php?f=7&t=601&p=7744#p7744

cheers