Strange Jamb

Maybe someone can help me with this. I have certainly had a history of jambs with the printer, but this is a new one. I switched the i3v over to 3mm and have not had any jamb problems since. Until yesterday.

I am printing a fairly large part. It is a replacement for the extruder support to allow the installation of a Bulldog extruder and incorporate auto bed leveling. Once I get the work all done and working, I will share the effort.

I prototyped the part about a week ago. No problems with that print. Yesterday I tried to print the part again as adjusted from the prototype. I used the same print settings and the same filament. About half way through it quit feeding filament. The filament was eroded by the extruder wheel and stopped feeding. This happened twice and not in the same spot. Simply resetting the feed allowed me to start over. There was no jamb in the extruder itself. I could easily feed filament through by hand.

I tried several prints today and the same problem, except that today, several of the attempts did not seem to be related to the extruder wheel. The wear when the jamb occurred did not seem deep enough to dissuade the feed. However, on the last one, tightening the extruder idler wheel did allow the filament to restart feeding.

I have tried changing temperature. I did the original prototype starting at 200 first layer and 190 after. It turned out a great looking part. I have adjusted the temps up as high as 210 and I have adjusted the feed rate down thinking that maybe it was trying to push too much through.

I also have the retract set to 3mm and it is set to retract when crossing perimeters. I wonder if that back and forth is causing enough wear to cause the problem. I have tried looser and tighter on the idler.

I took the extruder apart and made sure it was clear. Frankly, each one of these "jambs" has not prevented me from extruding a nice stream through the tip by hand without any changes other than stopping the print, releasing the servos and opening the extruder and pushing the filament.

I am open to any ideas.

CalifDan

How easily does the filament feed off the spool? Could it be getting tangled and snagged there, and the extruder can't pull it?

Quote:

---

Originally Posted by AbuMaia

How easily does the filament feed off the spool? Could it be getting tangled and snagged there, and the extruder can't pull it?

---

I will look at that. I have not noticed that as a problem with this filament, but it might not take much to cause the problem.

Edit: Did the print and it appears that the feed is the problem. I need to make some changes to how that works. Thanks AbiMaia!!

I want to throw this back up to the top. I do not think this is a filament problem caused by tension from the spool.

These failures, when they occur are at areas of a part that are small. I have retraction set to retract 3mm and I am printing PLA through a .4 head (Hexagon) at 190C. The failure occurs at the feed bolt and it wears into the filament enough to stall the feed.

I am wondering if the problem is associated with the back and forth of the retraction? Maybe in conjunction with a lot of filament (3mm) feeding down to a very small print area. This does not allow the filament to advance much and any wear would be exaggerated with multiple retractions and advances.

What I am wondering is how to solve the problem without messing up the prints. Turning retraction on greatly improved the quality of the prints. I am curious what the community thinks. Would turning the temperature up reduce the back pressure enough to prevent the retraction - extension process from wearing into the filament? I am going to experiment, but sure would like to solve this issue.

I would bet you are running too cold. I run PLA at 202 and it runs great, even at fast speed. It starts to click and jam around 195 if my heater can;t keep up during 60mm/s runs.

Thanks. Changing the temperature is something I will try. My concern is that the lower temperature has been working fine from a quality standpoint. I get less stringing and the layers seem to look better, particularly at the joined perimeter spot. Maybe the next question is should the higher temperature also include a speed increase? And if so, how does that reduce the back pressure which I think is the issue. Or, should I be reducing the flow rate?

Like I said, all ideas welcome. I will have three days to mess with it starting on Tuesday, so hope I can figure out a solution that prevents the erosion of the feed bolt and keeps the quality up.

What did you end up using for a slicer? What testing was done to conclude you needed 3mm of retraction?

Printbus,

I have used slicer .9 and 1.6. The part I am having trouble with is using .9 version as recommended by Makerfarm. I was using 4mm of retraction, but varied it a bit based thinking that it did not need to go that far. The point of retraction is to eliminate the ooze when moving between print areas (at least that is what I understand). Frankly, mine stops oozing as soon as the retraction starts, so I thought it would save print time to bring it down a bit. I think that you recommended a retraction rate of 10mm/sec and that is what I am using.

Like I said, I think the wearing and ultimate failure of the filament at the feed bolt is because of the back and forth of the retraction process when working on very small areas (frequent retractions). I think I need a way or combination of settings which will allow the retraction or eliminate the need for it and still maintain quality in the print and perimeter areas.

I also have tried tighter and looser on the feed tension idler. I can't yet consistently tell if that is making much difference.

Quote:

---

Originally Posted by CalifDan

Like I said, I think the wearing and ultimate failure of the filament at the feed bolt is because of the back and forth of the retraction process when working on very small areas (frequent retractions). I think I need a way or combination of settings which will allow the retraction or eliminate the need for it and still maintain quality in the print and perimeter areas. I also have tried tighter and looser on the feed tension idler. I can't yet consistently tell if that is making much difference.

---

I have heard of others having this same problem. I think it was fixed (or at least mitigated) by easing up on the idler tension, and by increasing retraction distance so you're not running the same small bit of filament over the bolt, but a larger area.

AbuMaia,

I will give that a try. Thanks.

I agree that it could be retractions that are causing your problem. I know I have ran into that. I can't speak from exhaustive personal testing, but I had read that direct extruders shouldn't usually require more than 1 or 1.5mm retraction so that might be something to try. Since you're using slic3r, you could also try increasing the setting for minimum travel after retraction to reduce the number of retractions.

FOLLOWUP: If you're using 3mm filament, I would think this is especially true. A 3mm length of 3mm diameter filament seems like an large volume to retract. To evaluate this for yourself, raise the nozzle off the bed and extrude 3mm of filament. Observe how much came out the nozzle. Do you think you should have to pull that much volume back out of the hot end when you're lifting the nozzle for a retraction?

When you mentioned 3mm retraction, I thought maybe you were using Cura since IIRC it defaults to 3mm. I believe it does this since the Ultimaker that Cura is mostly developed for is a Bowden extruder and considerably more retraction is necessary due to hysteresis in the Bowden tube. Cura also has a setting for minimum extrusion before retract that can help reduce retracting the same section of filament. I currently run Cura with retraction distance set to 1.2mm and the minimum extrusion before retract to 0.5mm. This has been a helpful combination, but I don't think Slic3r has the equivalent of the minimum extrusion before retract setting.

I'm not trying to influence you, but you asked for input and you haven't received much. One of the things I disliked about Slic3r (especially on versions newer than 0.9.9) was the way it seemed overzealous in going back over things to touch them up with just dabs or short extrusions of filament. Too many of these in a row and you're setting up for chewing into the filament. I was amazed at the difference in the number of these hops when I started comparing Cura and Slic3r results in a gcode viewer, and dumped Slic3r shortly after.

That said, even with Cura I've learned to look at the layer-by-layer print moves in a gcode viewer for anything with the potential for small amounts of extrusion. If necessary, I'll go back and reslice with retraction turned off in order to keep the hobbed bolt from repeatedly turning on the same length of filament on these prints. The avoid crossing perimeters setting helps minimize issues with stringing. This is another area where I personally think Cura is better at managing perimeter crossing than slic3r.

-----
I have also had filament stop extruding if the nozzle temperature gets too low. This has normally happened when I'm just experimenting with a lower print temperature or messing around with a print cooler and end up blasting the nozzle with more cold air than the heater control loop can deal with. With the nozzle off the bed and motors disabled, you should be able to get a feeling for how easy it is to extrude a particular filament by manually rotating the large gear or releasing the guidler and just feeding the filament manually. As you lower the temperature and start to feel an increase in resistance, up the temperature by 5-10 degrees as the setting for that filament. At least that's what I've read as far as one way to determine the proper temperature to use for a particular spool of filament.

Printbus,

Thanks for the information. I also think Cura does a better job, but, having so many problems with jams over the last few months, changing from 1.75 to 3mm and several different filament purchases, sometimes it is hard to isolate a variable. I will try your suggestions, including a Cura slice of the part. I am also coming to the conclusion that different parts need different considerations. I have found that going back and making minor changes in design has helped me get a better print.

I stopped using slic3r because of that stupid way it went over things. I had a 20mm circle with a knurled edge so like a many pointed star. Slic3r thought it best to draw a circle and then at 60 blobs around the circumference of the circle, irrecoverably brain damaged in my opinion.

I also had some issues on my Prusa i3v 10" 1.75mm.
I was either having filament slip, or SNAP! in the extruder.
With alot of input from here I think it was down to two things, dialing in just the right amount of tension on the extruder, and lowering the retraction settings in Cura.

It has been a long couple of days. With the intent of getting to the bottom of these jams, I have had a couple of real busy days. I have tried various setting in Cura and Slicer. I varied temperatures, I messed with various retraction settings including eliminating retraction all together. Sometimes the part printed, but at least 4 times out of 5 it continued to jam at some point in the print. I even went to the music store and got the guitar string and cleaned out the print head even though it did not appear to need it. The jam was always associated with the grinding away of the filament at the feed bolt.

I think I was concentrating on the wrong side of the feed bolt. As I have said before, it seems to me that this is a bigger problem in the last third or so of a roll of filament. I tried stretching it out and keeping slack in the feed. That seemed to help. but it was not perfect. There were still failures. Often enough to make me think that was not the problem.

Today, I cut off 7 feet of filament (enough for a part) and wrapped the filament loosely around the spool. The part printed without any issues.

I am coming to the conclusion (need more successes) that this issue is related to the tension in pulling tightly wound filament off the roll. I have the standard MakerFarm feed setup on top of the printer. I do have the little printed part that slides over the cross member to allow it to turn easier. Even so, I think that the combination of the friction of the spool on the Makerfarm setup (even modified) tends to create enough friction that the feed will not pull it. Probably the combination of all the resistance forces, but as I get down into the spool, the added effort apparently breaks the bank.

I will need to find a way to modify the spool feed to turn much easier. At this point, I am thinking bearings. I want to remove as much friction as I possibly can.

I will keep you all posted.

One thing you might try in reducing spool friction, get a roll of PTFE tape from Amazon (http://www.amazon.com/CS-Hyde-PTFE-M.../dp/B003DZ16X4). It's thin, slick stuff. I got a half-inch wide roll, and wrap it around my spool holder. So far the only problem I've had is the filament coming off the spool and getting snagged on the holder. If you're willing to fiddle with it, you could put some on the inside of the spool hole as well, to really reduce friction.

Lots of people will probably baulk at this suggestion but I poured rape seed oil down my filament feed tubes on my Stratasys (then blew them out with air) and it made a world of difference without seeming to affect the prints. I think the secret is careful choice of oil and I read about all of them before choosing rape seed.

No responsibilities accepted. :)

I hear some vegetable oils are good to add to PLA, but not so much for ABS.

OK, I'm pretty sure the problem has been identified. I have had six successful prints in a row. For each one, I cut the necessary filament (plus a little) and wound it loosely around an empty spool, still sitting on the Makerfarm spool rack. Essentially, the feed screw is pulling against little or no resistance as the loosely wound filament comes off easily without even turning the reel. No jams.

What I believe was going on was that when the roll is first used the larger diameter of the roll helps lower the resistance of the spool reel to roll around the Makerfarm spool rack. That lower resistance allowed the first part of the reel to work pretty well. Then, as the filament is used up, the extruder is having to pull harder to turn the reel because the diameter it is working against is lower (basically, the rotation lever arm is reduced). Also, the filament is more tightly wound and I believe that this also adds to the turning resistance of the spool.

I have ordered bearings and have designed a setup which should make the spool turn much easier. I will work on that once the parts come and do some testing. If the solution helps the problem, I will post the results and design here. In the mean time, all parts are being made with precut lengths of filament.

I'll make a mention of this, simply because as I skimmed through, I didn't see it, and I ran into this problem today, myself.
Warning, I'm a newb, so this may not be/probably isn't your problem:
For some reason that I have yet to find, after having a bowed bed (in the opposite direction) the center of my glass is HIGHER than the corners. Not one, but all 4. Soooo, lining up the z-endstop meant that when the print begins, obviously in the center, it is putting a lot of pressure on the glass, and the extruder never makes the magic happen. So, the nozzle just goes around the glass, shooting blanks. I would think it were caused by something else, but I hastily, mid-print, lowered a corner of the bed by a substantial amount, and the nozzle started printing only in the areas near that corner. After about 4 cancels in prints, I ended up lowering all 4, also bumped the z-endstop up, and voila, I extrude again. Albeit, there's still a lot of pressure on it. Anyhow, just thought I'd mention it. Don't take it too seriously. I'm brand-spanking new to 3d printing. 3 weeks in, give or take.
I confirmed this by using the typical 4-corners be leveling, then bumping the Z up enough to see light between the nozzle and the bed. All 4 CORNERS had the same height. Move the nozzle to the center on X AND Y? It's touching the glass.

-Aut0