Strange Jamb

[printbus]

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.

[CalifDan]

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.

[Mjolinor]

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.

[jtice]

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.

[CalifDan]

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.

[AbuMaia]

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.

[Mjolinor]

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.

[AbuMaia]

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

[CalifDan]

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.

[theAut0]

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