So I tried using Slic3r’s “Wipe before rectract” to address the issue shown in this photo (the gunk between the two towers), and I also tried using Cura, which solved this problem but created others.
So, instead of continuing to try and find a “work-around”, I’m interested in understanding the cause of the problem so I can fix it and keep using Slic3r, since in every other way it produces the best results I’ve seen with my designs and my printer.
Given that it goes away when I use Cura, that would indicate that it’s something that can be solved on the software/configuration side (as the hardware remained constant during these tests). What’s not clear to me is what the root cause is; is it over-extrusion, under-retraction, etc.?
Seems like others have experienced this and found work-arounds, but I’m interested in getting to the bottom of the issue and putting it to bed once and for all If that means extensive testing, experimentation, etc. that’s great (had to do the same thing to get to the bottom of my “banding” issues awhile back :).
Thanks again to everyone in this community for your help, it’s amazing how much my printing has improved since joining.
I did the “keep lowering the temp until extrusion stops” test, but maybe there’s something more sophisticated?
I found that the lowest temp I could consistently extrude (PLA) at was measured as 180C on my setup, but layer bonding was weak so I stepped it up gradually until settling on 200C. This seems a bit high, but my hot-end setup is a bit “custom” so it might not be directly comparable to other setups.
That said, I’ll try this print again at 180C, and see if that eliminates the gunk, or at least changes it.
Try turning cooling off under filament settings in slic3r and see what effect that has. I don’t expect it to fix everything (and will probably cause other issues), but it will be informative.
Dropping the temp to 180 (after the first layer) seems to have helped @Anderson_Ta , in the sense that it delayed the onset of the gunk until the last 1/3 of the print. In other words, instead of having it strung from the base of the towers to the top, it only starts to appear above the midway point.
@Whosa_whatsis I’ll try disabling cooling, although I’m a little concerned about turning it off outright as I’m using a jhead and I don’t want the heat to creep too far up. I have two fans, one aimed squarely at the cooling fins of the jhead, and a second pointed at the output area, so what I could safely do is electrically disable the output area fan, but do you think that would have the effect you desire?
The fan aimed at the cooling fins on your extruder should be wired directly to 12V, not FET-controlled. The point was to get rid of the slowdown due to minimum layer time. You could also just set the minimum layer time to zero and leave the other cooling settings unchanged.
The point of this test is that I think you’re traveling too slowly. No matter how much you retract, if the nozzle is hanging in open air long enough between towers, plastic will ooze out.
Is your Slic3r up to date? I know it used to slow down travel speed with the cool setting, but I thought that it had been fixed so that travel speed was left alone and only printing speed was slowed to ensure minimum layer time. To minimize ooze, your travel speed should be set as high as possible without causing skipping.
Nice, I never considered the travel time… My Slic3r should be up to date (as of last Wednesday at least but I have been experimenting with adjusting speeds and may have slowed travel down inadvertently.
So it sounds like the root problem is ooze, would you agree? And perhaps the reason it gets worse at higher layers is due to the decrease in surface area, which in turn causes Slic3r to slow things down to maintain minimum layer time?
I’ll also mention that this latest print seemed to finish much faster, which would make sense in light of removing the minimum layer time requirement.
So, if I can reduce ooze by lowering temp (as in the first suggestion), and speed-up travel (so the nozzle spends less time hanging out in the open air), that should reduce the gunk.
Additionally, would increasing retract help with ooze? Right now I’m using 1mm and probably a speed of 40mm/s (if I remember correctly).
I get this a lot, and it is ooze, in fact watching closely I can see the ooze as the head travels then it gets caught on these little towers and pulls off. Travel time is definitely an issue as the printer I see this on is very slow max travel is 25mm/sec. I found retract did not help much with ooze in this context. Funnily enough my older Jhead Mk4 with 0.5 nozzle does not ooze at all, the new Mk5 with 0.35 nozzle oozes like the devil. Brian mentioned adding a nipple to the nozzle may help. but that is a special order
Yeah @Wolfmanjm , what’s strange to me is that I don’t have this problem with models sliced by Cura, so there must be a software way to address it, but I haven’t found it yet…
I did another print with travel cranked up to 200mm/s, retract cranked up to 4mm and temp at 180C. It was a little better, but had the same sort of artifacts, just shaped a bit different (I can post a photo if it helps).
I might try going through the Gcode and compare the Slic3r code and the Cura code and see if there are any clues to the difference in there.
Yeah other than this one problem, I get great results with Slic3r. There are things I like about Cura, and I could probably dial it in to output as well as Slic3r, but there’s a couple bugs (for me at least) that are showstoppers.
@John_Ridley Slic3r is being actively developed and they will release another version. The average time between slic3r releases is less than the time elapsed since 0.9.10b. I don’t know why there are problems compiling, and I recently attempted to compile it from Git for the first time without success too.
the problem comes partially from under retraction speed, usually due to motor stalling on retract. i eliminated this by tuning my Acceleration constant very carefully to get maximum retract speed w/o stalling.
you need the extrudate to suck back into the nozzle, fast retract helps that.
I didn’t think about stalling during retract @vince_geisler , I’ll take a look at that and make sure I’m not loosing steps when the drive is backing up.
If that means extensive testing, experimentation, etc. that’s great (had to do the same thing to get to the bottom of my “banding” issues awhile back :).
but I have been experimenting with adjusting speeds and may have slowed travel down inadvertently.