Consistent Feeder Grinding Issues

[fbrc8-erin 272]

7 hours ago, whoneyc said:

No harm in a little shameless promotion.  The parts sent from fbrc8 were paid for by the NA support folks.  I haven't heard back from them, but I'm guessing they aren't too into doing a full upgrade.

 

I'm just a little curious about the Bondtech feeders you have as a side note.  I see a two-part feeder gear, and that is the only real difference.  Is the rest of the kit really necessary if I can't print my own custom lever?  Do the 1.75mm feeders work well on the UM3?  That would be a nice side-grade to consider in the future.

 

For support, we send out whatever parts are needed under the warranty (so all OEM parts, no bondtech). It looks like my technichian responded on your support thread after this last post and has some other things to try with you that he wanted to check out. 

[whoneyc 10]

1 hour ago, fbrc8-erin said:

 

For support, we send out whatever parts are needed under the warranty (so all OEM parts, no bondtech). It looks like my technichian responded on your support thread after this last post and has some other things to try with you that he wanted to check out. 

 

Oh, I know I'm not getting anything besides OEM.  My sarcasm doesn't always come across in text.

 

Thanks for saying that S responded with some stuff to try.  His email got bumped into spam, so I might not have noticed it.

[gr5 2,167]

I don't recommend a 1.75 conversion unless you have at least $400 or 400 euros of 1.75 filament that you don't know what to do with.

[gr5 2,167]

You can I believe buy only the bondtech gears with the splined feeder part and design your own feeder around that.  The UM3 feeder works quite well though in my experience.

[whoneyc 10]

10 hours ago, gr5 said:

I don't recommend a 1.75 conversion unless you have at least $400 or 400 euros of 1.75 filament that you don't know what to do with.

 

Only about $200 worth of filament left over from when I had a Makerbot.  The plan now is to shred it all and make thicker filament on the Filastruder.

[kmanstudios 1,120]

5 minutes ago, whoneyc said:

 

Only about $200 worth of filament left over from when I had a Makerbot.  The plan now is to shred it all and make thicker filament on the Filastruder.

I would like to know how that goes :)

[whoneyc 10]

14 minutes ago, kmanstudios said:

I would like to know how that goes ?

 

I will be sure to post information for you here as well as on the Filastruder forums.  I have probably 8kg of unusable filament now, be it 1.75 mm old stuff or the massive bucket of failure from this blasted issue.  I also have 5 kg of virgin PLA to mix in.  I know that PLA isn't friendly for heat recycling, so I had a crazy idea.  I bought some methylene chloride and alizarin dye.  The proposed primary degradation pathway for PLA during repeated extrusions is free radical damage.  One method to stabilize these radicals is to add a quinone-based additive, which preferentially absorbs the free radicals before they can cause damage to the polymer.  My plan is to make some masterbatch from the virgin PLA with a crazy high loading of alizarin, which is a quinone (also cheap and accessible to the public).  My hypothesis is that I will end up with way more orange-red polymer than I know what to do with, but the material will be stable enough for recycling.  If that works, I will publish it on this forum, the Filastruder forum, and an appropriate academic journal.

[kmanstudios 1,120]

25 minutes ago, whoneyc said:

 

I will be sure to post information for you here as well as on the Filastruder forums.  I have probably 8kg of unusable filament now, be it 1.75 mm old stuff or the massive bucket of failure from this blasted issue.  I also have 5 kg of virgin PLA to mix in.  I know that PLA isn't friendly for heat recycling, so I had a crazy idea.  I bought some methylene chloride and alizarin dye.  The proposed primary degradation pathway for PLA during repeated extrusions is free radical damage.  One method to stabilize these radicals is to add a quinone-based additive, which preferentially absorbs the free radicals before they can cause damage to the polymer.  My plan is to make some masterbatch from the virgin PLA with a crazy high loading of alizarin, which is a quinone (also cheap and accessible to the public).  My hypothesis is that I will end up with way more orange-red polymer than I know what to do with, but the material will be stable enough for recycling.  If that works, I will publish it on this forum, the Filastruder forum, and an appropriate academic journal.

Way cool :) thanks!!

[dmboston 2]

Curious to know how your printer is working, and if you have any updates on your recycling project. I had the same issue, and noticed the same wear marks on the tension lever, and also the top of the feeder housing before the filament goes into the bowden tube.

 

I finally contacted the NA support, fortunately the printer was under warranty, and replaced the feeder. No problems now! (Thanks to the folks at the NA customer support for helping me get my printer running super quickly)

[whoneyc 10]

34 minutes ago, dmboston said:

Curious to know how your printer is working, and if you have any updates on your recycling project. I had the same issue, and noticed the same wear marks on the tension lever, and also the top of the feeder housing before the filament goes into the bowden tube.

 

I finally contacted the NA support, fortunately the printer was under warranty, and replaced the feeder. No problems now! (Thanks to the folks at the NA customer support for helping me get my printer running super quickly)

 

I have not forgotten about this thread, and I had no intention of leaving without reporting a final solution.  I have printed successfully since changing the lever and resetting the flow to 100%.  However, it started acting up again with the same underextrusion patterns on my latest print.  Sadly, this error came up on the same week as my back went out and I was driving my cat around to different emergency vets, so it got pushed to the back-burner.  I have not taken apart the extruder yet to check for damage to the lever, but I will do that as soon as possible.

[whoneyc 10]

As promised,here is an update.  I'm tagging the folks that have been helpful and seem to be running some support shops: @gr5 @fbrc8-erin @kmanstudios

 

The fix presented by the Ultimaker support folks was a replacement lever and setting the material flow to 100% from a previous test.  I ran this off and on for about two weeks without issue.  I produced 3 moderately sized statues during this time using PLA and Breakaway filament in the extruders.  I did this for more than just a desire to collect a pantheon of archaic gods, but because these were designs with complex surfaces and required a lot of material switching (aka lots of retractions).  These were all successful.  I had one hiccup with a failed print that made my heart skip, but that was just a simple filament tangle.  No big deal.  I was telling any friends and family that feigned concern for my printer woes that the issue was fixed.  

 

And so I believed it was, until recently.  I printed a very simple design with few, if any retractions, for my filawinder build.  This print failed with the mesh-like underextrusion I have grown to loathe with a passion.

 

 

So there is a still a problem.  Naturally, since the lever was recently replaced, I looked to it first as the source of problems.  And as I suspected, the lever did have a problem.  The two halves of the lever which are held in place by a friction lock as designed by the manufacturer, had partially separated.

 

 

You can see in the image that the axis holding the bearing in place has separated, making it such that the bearing is not aligned.  I corrected this issue by pressing the halves back together, but something about the tactile feedback in my fingers after I had pressed it together suggested that it is not permanently affixed.  I believe it will happen again.

 

I doubt this was a spontaneous split.  I believe that the torsion of the filament in the feeder is applying a force to the bearing or plastic slot such that the part splits.  I will be continuing to print to see if this continues to occur.

 

Until more information is brought to our attention on this, let's do a little speculation.  What is causing this to occur?  Is there something wrong with the alignment in my machine which could cause this?  Is there a fundamental issue with the design?  We know that I am not the only one to have lever problems, as the Ultimaker Support guys shared a picture which had the grooved lever like mine was.  Are there more users such as @dmboston who are having the same failed part?  What can we do to remedy this as users?  Do we need to switch over to a different material in the lever construction to solve this?  I mean, a lever machined from Al sounds awesome, albeit expensive.

 

I should note that I have made the following modifications to my printer since the last time I posted here:

• Added clawed feet which lift the printer up about an inch
• Added a 120mm fan in the space made by the clawed feet to lower temperatures of motor controllers
• Added 5V to 12V buck boost between fan and USB header

None of these mods should be affecting the feeder.

 

 

[dmboston 2]

So, out of curiosity, how do you load your filament? Do you pull up on the lever and pass material in, then use the motor power to get it to the head? Or do you use the motor power to feed it in from the start using the "load" option in material settings?

 

Given the wear pattern is on the side of the lever, I have a suspicion that, at least on our machine, it may have worn during loading/unloading when someone inserted the filament not aligned with the bearing. Hard to say for sure in my case because the printer is in an academic lab and sees a lot of hands constantly loading/unloading different filaments, but it sounds like you're the sole operator of your printer?

 

I agree, an Al lever seems like it might be better, but I'd go with die-cast, much cheaper ?

[whoneyc 10]

1 hour ago, dmboston said:

So, out of curiosity, how do you load your filament? Do you pull up on the lever and pass material in, then use the motor power to get it to the head? Or do you use the motor power to feed it in from the start using the "load" option in material settings?

 

Given the wear pattern is on the side of the lever, I have a suspicion that, at least on our machine, it may have worn during loading/unloading when someone inserted the filament not aligned with the bearing. Hard to say for sure in my case because the printer is in an academic lab and sees a lot of hands constantly loading/unloading different filaments, but it sounds like you're the sole operator of your printer?

 

I agree, an Al lever seems like it might be better, but I'd go with die-cast, much cheaper ?

 

I have used both the auto load and cold load (where I lift the lever and insert to the end before using the "move" function to put the filament in the correct spot). The cold load happens most often as a matter of convenience when the damaged filament can be pulled out from the Bowden tube, snipped, then reinserted easily.

 

Which extruder are you having trouble with? If it is the left side, the filament spool is not perfectly in line with the extruder.  It might be an issue of the filament coming in at a weird angle.  There was a suggestion earlier in this topic by @gr5 that spools on the floor work best.  Perhaps this is another solution.  Limited access can complicate things, especially when there is that one person in the lab that does things in an odd manner (there is always that guy).  You are correct that I am the sole operator of this printer.  

 

Die cast might be an option if we have enough people that express desire.  If this is a thing, I have experience running group buys for mechanical keyboards, so I would not be above setting one up if the community becomes interested.  

[fbrc8-erin 272]

The lever splitting like that definitely looks unusual; while I have seen the groove issue, I haven't seen one split like this. The levers are press fit with specialized tools. It's always possible the tolerances was off on one of the plastic pieces; respond back on your ticket on the support system and we'll get another lever out on Monday.

[whoneyc 10]

29 minutes ago, fbrc8-erin said:

The lever splitting like that definitely looks unusual; while I have seen the groove issue, I haven't seen one split like this. The levers are press fit with specialized tools. It's always possible the tolerances was off on one of the plastic pieces; respond back on your ticket on the support system and we'll get another lever out on Monday.

 

Thanks, but I'm honestly not that worried about the splitting.  That can be easily remedied with glue and a clamp.  I'm more concerned about what causes the split.  If it is a press-fit thing that can be fixed with glue, that is dandy.  If the split occurred due to some mis-alignment, then there are bigger problems.

[fbrc8-erin 272]

It is a press-fit. The bearings on the housing parts are also press-fit. If your motor feeder motor is screwed in tight and none of the bearings have popped out, I don't think there should anything to worry about alignment-wise.

[whoneyc 10]

Necro-follow up.

 

So it has been a few months since I posted this issue, and I am happy to report it is fixed.  

 

The grinding issue continued for a few months after this thread died.  The Ultimaker support folks at fbrc8 (shout out to Spencer P. who has the patience of a saint) were helping me work on it.  We went through several iterations of tests and fixes.  During which we found that:

• One of my bowden tubes was installed with the beveled edge on the wrong end.
• The extruder lever began to break apart with the stress of the grinding,
• The breaking extruder lever was easily fixed with carefully applied superglue.

In the end, the fix came due to weather changes.  I was working on my printer before and after the first big cold front of the autumn came through.  Once my shop got cold, things got better.  While my sample size is just n=1, I think that it is very likely that the weather was causing the problems.  I have been running my printer since then with no issues related to this grinding problem.  

 

So what was this weather like? There were several hiatuses during the email chain with Spencer P. in which it was literally inhospitable in my shop.  The shop is a converted garage with poor insulation and no climate controls.  The temperature was usually above 90F/32C and humid.  I was only running automated stuff like the printer, server, etc. in the shop because I could not comfortably work very long.  Tiny tools and electronics are difficult to handle when covered in sweat.

 

But with the cold front change in the weather, the humidity dropped.  My shop became usable, and the printer started working.  I think it is likely that it was a combination of +90F heat and +90% humidity that caused issues.  Remember how I replaced one of the Bowden tubes because I could no longer fit filament through it? Once the humidity was down, the filament could go through the old tube.  I think there may have been a combination of mushy filament and Bowden tube friction that caused grinding failures.

 

So, I have a solution.  It isn't the best thing I could hope for.  I have lost:

• Time spent working on this problem
• Time in the year I am able to use my printer
• Pay from a contract job I was unable to do
• Literally 5gal of filament in wasted prints and grinding damages.

 

Pictured: a handful of 2' long unusable filament lengths.

 

But I don't hold Ultimaker to blame.  Sure, this could have gone much better, but I have access to a strong support forum and support persons through them.  However, I would recommend that they be more upfront about this with consumers.  Perhaps it would be wise to have a page on where to put your printer.  This would have been useful to me.  My previous printer had no issues with weather, likely because it was not a Bowden tube design.

 

 

TL;DR A printer designed in a cool climate cannot handle Oklahoma summers.  If you have grinding troubles, think about your printing environment.

[kmanstudios 1,120]

Environment is something that I find to be a very important factor in printing.

 

I am very happy that you got your issues sussed out and moving along.

 

And, yeah, FBRC8 folk are really good.

[alnavasa 5]

On 5/25/2018 at 6:08 AM, gr5 said:

6) feeder spring issues - too tight, too loose.  You want the tension such that you can clearly see the diamond pattern biting into the filament.  You want to see at least 2 columns of diamonds.  4 columns is too much.  You usually want the tension in the center.

 

 

On The Ultimaker original, does this apply too? I find my self struggling on how much is too much and how little is too little pressure, The knurled bolt on the original Ultimaker original has a different pattern than on the UM2 (I think), so I guess this apply differently right? how much is the right tension? I can't find any post explaining this, 

Only found the Ultimaker website, where "no marks" is too little and "material grinded" is too much

[SilverDynamic 1]

Hi,

 

I wanted to update this thread because I've recently had the exact same problem and came by a solution that allowed me to continue printing.

 

To cut a long story short it was the same issue - suddenly all my prints failed shortly after starting with grinding. Although later I did establish that prints with a much smaller X/Y travel worked fine.

 

I re calibrated everything, tried different high quality filaments, replaced Bowden tubes and nozzles and used both extruders. (UM3+).

 

I came by this thread and was intrigued by the proposed route cause - high temperatures and high humidity.

I live in Scotland and while the summer has been good so far it CAN NEVER BE TOO HOT IN SCOTLAND!

 

However the temperature was 20+ degrees Celsius.

 

After pursuing a mechanical issue as the cause I hit a dead end then some things came together:

1. I then thought back to the ambient temperature issues in this thread. Yesterday when I replaced the Bowden tubes it was ~25 degrees and very humid. The hottest day of the year so far.
    
2. This reminded me also of a comment UM support had made about pooling melted fluid in the nozzle. When this problem first occurred I noticed immediately that the ends pulled from the nozzle were unusually thick, basically the same dimensions as the metal tube. And then I realised that this was because the filament had fully melted.
    
3. I'd been surprised that I was having the same issue on both extruders so it couldn't really be the Bowden tube as ext 2 has hardly been used.

So I then reduced the print temperature from 200 deg to 185 deg and the problem was solved. I've since upped the build plate from 60 to 65 degrees as there was some minor adhesion issues (first layer bubbling). But since then it's remained hot outside (relatively speaking) here and I've done 3 major prints with no issues.

But I'm still not sure why my UM2+ in the same location is merrily printing away at an unchanged 220 degrees...

 

Thanks for all the contributors to this thread.

TL;DR I solved this issue by reducing the print temperature from 200 to 185 degrees C.

 

 

Pictures: 

1. Failed print.

2. Jammed in tube guide for extruder 2.

3. Thickened end after melting.

Edited May 29, 2020 by SilverDynamic

[whoneyc 10]

11 hours ago, SilverDynamic said:

TL;DR I solved this issue by reducing the print temperature from 200 to 185 degrees C.

 

 

Thanks for the detailed report @SilverDynamic.  

 

As a follow up: I have had no problems related to my OP since moving houses to a location where the printer could live in an air conditioned room.  Temperature appears to be an important issue with UM3.

[SilverDynamic 1]

Time for another update!

This issue recurred again but it's September in Scotland and a cool steady 17 degrees C in the house, plenty of printing being done successfully but once again on large flat bottomed prints the same issue.

OK so after a few Cura updates the print temp had reset to 200 degrees C but in the normal course of printing this was not a problem.

Then a project with a large bottom surface area came up and the same again.

This time when I tried to extract the filament it got stuck in the Bowden tube because the piece from the hot end had got excessively thickened after melting. 

 

If it's only 200 deg, why is it melting?

 

So I checked in the advanced settings and found TOP/BOTTOM SPEED which was set at 38mm/s as opposed to 100 mm/s for normal printing.

Which gives the filament time to heat more, melt and block off the nozzle.

I punched the Top/Bottom Speed up to 100 and tried again at 200deg.

Of course this will only be an issue if your bottom layer is a large area which gives the PLA time to fully melt as it's sitting in the extruder for a long period.

I tried the new setting of 100mm/s Top/Bottom Speed and a perfect print.

-------------

This never happens on my UM2+ machines so I checked the setting for Top/Bottom Speed and the default here is 50mm/s.

 

TL;DR - the real problem is the Top/Bottom Speed setting. It's too slow and allows the filament to fully melt and jam off the nozzle.

Edited September 11, 2020 by SilverDynamic