How to eliminate the "hairy travel"? Using S3D.

[jsc]

You should be using a .40mm extrusion width anyway. Anything from .4 to .5, but .4 is nice because it is a nice round number you can design around. S3D has a minimum infill length option under Infill, mine is set to 5mm. What's yours set to?

[swbluto]

Just tried .4mm, man, the two-walled pieces are now fairly flimsy and there's gaps in the wall. Making me think that the -.3mm restart is a bad idea when using .4mm, should probably try out 0 mm restart w/ .4mm. Still, I don't think that'd resolve the flimsiness? Maybe my extrusion multiplier is too low at 1.00. So many combos to try out.

On the plus side, no hairs. I'm starting to think I won't be doing multi-piece prints on this machine, since that seems to introduce the greatest amount of nasty little hairs and tumors. Still, it'd nice to forget it for 2 hours instead of reloading it every 40 minutes. Oh, the trade offs.

Does anyone have a rough idea of what the extrusion multiplier should be for hatchbox ABS when using .4mm extrusion width on a .35mm nozzle? I think 1.00x might be a little low, but I don't have test data for comparison.

SAT question alert!!! lol

With Pink 1.75mm Hatchbox ABS:

On my makerbot, .4 mm nozzle -> .4 mm extrusion width -> 1.05x extrusion rate produces perfect surfaces.
On my M2, .35mm nozzle -> .35mm extrusion width -> 1.08 extrusion multiplier produces perfect surfaces.
On the M2, .35mm nozzle -> .4mm extrusion width -> X extrusion multiplier produces perfect surfaces. What is X?

Also, my minimum infill length is set to 5mm. Appears to be the default.

[Jules]

1.08. The software is going to make the width correct if you have the correct extrusion multiplier. (Trick question.)

[ednisley]

Maybe my extrusion multiplier is too low at 1.00.

IIRC, you chose the multiplier to produce smooth top surfaces, rather than accurate extrusion widths, so "gaps in the walls" may be the result: the thread comes out narrower than slicer expects.

If you calibrate the multiplier for accurate width, perhaps you could tune the top surface using Top Solid Infill Width (in Slic3r, S3D surely has something similar). Even with the width set to Auto, Hilbert Curve infill suffices for my simple needs:
http://softsolder.com/2015/08/17/clover ... on-holder/



Magenta PETG is impossible (for me, anyhow) to photograph up close; the camera sensor saturates with red.

The top surface comes out dead flat, with a slightly granular texture, and doesn't look at all like the usual hand-knitted 3D printer finish.

[swbluto]

Maybe my extrusion multiplier is too low at 1.00.

IIRC, you chose the multiplier to produce smooth top surfaces, rather than accurate extrusion widths, so "gaps in the walls" may be the result: the thread comes out narrower than slicer expects.

Nah, I just chose 1.00 because I thought it'd compensate for the slightly thicker extrusion width (.4mm vs .35; that is, 1.00x.40 is roughly equal to .35 x 1.08). BUT... I think I might've been thinking wrong about that, as the laid down width is larger, so it really would need the exact same extrusion multiplier of 1.08, I think? I'll try it out next time.

If you calibrate the multiplier for accurate width, perhaps you could tune the top surface using Top Solid Infill Width (in Slic3r, S3D surely has something similar). Even with the width set to Auto, Hilbert Curve infill suffices for my simple needs:
http://softsolder.com/2015/08/17/clover ... on-holder/

I miss some of Slic3r's options (S3D doesn't have anything similar to that). Even though it sucked at parsing non-manifold models, it did print out some superior top solid infills. I think I might just start experimenting with Slic3r with my makerbots, see if I can improve the top layer. S3D does perfectly fine with the top layer on my M2.

[swbluto]

Experimenting with Slic3r. Pain in the ass trying to get it working on my Makerbot Replicator 2X, but after multiple gcode revisions, it's finally printing. It's now printing the first layer and, woah!, that's some really thin lines you got there. My settings are exactly the same in s3d as far as I can tell, but slicer's lines seem to be much thinner. The settings are correct, though, because the bottom layer looks like it's being filled in fine so I don't think the extruder is too close or any of that jazz. I verified the z-heights by comparing the g-code with other g-code that S3D produces, and the z-heights are the same. I guess Slic3r just does much smaller 'real' filament width than S3D given the exact same settings.

We'll see how this turns out. I guess the much thinner widths allows it to get much better detail on the top layer. I hope the intermediary layers aren't this thin, that'd take forever! lol. *Checks slic3rs layer preview window*

Oh, it appears that the lines in Slic3rs are twice as thin on the bottom layer as the rest of the layers. Is 100% first layer width in Slic3r really the equivalent to 50% first layer width in S3D, because the first layer's width looks half that as the width of every other layer in Slic3r. So to get 100% like in S3D, I really need 200% for the first layer width in Slic3r?

*Types in 200%*, now that looks much more like the other layers! Just slightly larger, really, but still much more 'normal' than 100%! I guess Slic3r is kind of squirrely with its first layer widths.

Seems I can't tell Slic3r to do rectilinear for the bottom layer, and then hilbert for the top. Hilbert on the bottom seems to have a tendency to create holes in random places, at least for 100% f0_width. I'll see what it is for 200%.

Edit: Looks like it's still creating holes in the bottom layer at 200%. Can I find a workaround to this? (Edit: First layer's finish, and my lord! Looks like Berlin after the city got demolished by carpet bombing in WW2! lol)

Maybe I'll just splice together S3D gcode for the lower layers and Slic3r gcode for the top layer. I was /really/ hoping I could avoid S3D altogether. I suppose I could create two different gcode files from Slic3r, one based on rectilinear and the other on hilbert, and splice them together to get a rectilinear bottom and hilbert top. If slic3r has a command-line capability, I suppose I /could/ do that programmatically. But, first, let's see it print to see its top layer quality before going that direction. (Hey, looks like slic3r is incredibly easy to use commandline wise so programming that would be a piece of cake.)

Well, the first signs of the top surface appearance of hilbert is coming through: looks and feels like stucco! lol.

Edit: Just tried using the rectilinear pattern in Slic3r and the surface finish is much smoother than S3D. However, S3D looks "shinier" and it seems like Slic3r's has more "fatigued white plastic residue" than S3D. That is, S3D looks better in that the surface doesn't have as much fatigued plastic (Parts of the pink plastic look white) and it's shinier/more-shimmery, but it's definitely rough to the touch which Slic3r is smooth to the touch. Well, this is cool - it turns out I can use my colored sharpies to hide the white marks in the plastic. Now it looks uniform in color and feels smooth, best of all worlds.

Yep, I think I'm going to integrate slic3r into my processes. S3D is such a pain in the butt to automate and it has a rough surface affect on my printer, whereas Slic3r does something special to smooth the top layer.

[ednisley]

given the exact same settings.

Back in the day, different slicers applied the Extrusion Multiplier setting in different ways: some used it as a linear multiplier, some used it as a quadratic factor. Perhaps Slic3r does it one way and S3D the other? At least with Slic3r, you can read the code and figure out what it's doing (and change it to work the way you want, if you prefer).

In any event, if you haven't calibrated the Extrusion Multiplier for the filament / temperature / settings you're using, I'd lay long money that the actual filament width isn't what the slicer expects: Slic3r is doing exactly what you told it to do, just not what you expected it to do.

rant ON

Seems I can't tell Slic3r to do rectilinear for the bottom layer, and then hilbert for the top

You can apply entirely different settings to different volumes, not just different Z levels, by defining and positioning volumes where you want them, then modifying the settings within those volumes. Slic3r gives you very finely detailed control over the settings, at the cost of being very tedious.

I hope the intermediary layers aren't this thin, that'd take forever!

You can (much more easily) control the ratio of interior to exterior layer thickness & thread width, although with a non-obvious constraint on the actual extrusion rate. I haven't bothered figuring that out, because it wouldn't improve my lifestyle in the least, but other folks swear by it.

Slic3r is kind of squirrely with its first layer widths

You can control the first layer width and thickness (by percentage or millimeters), the XY speed, the fan setting, and anything else you want. At the cost of additional fiddling, you can do that for every single layer and any particular volume, all the way throughout the model, from the bottom to the top.

If Slic3r is not doing what you want, then tell it what you want. If you don't understand what it's doing or how to tell it differently, RTFM: Slic3r actually has a manual, it has lots of good info, you need to read it. If that's not enough, read the source code: it's out there, it's not secret, it's not hidden.

In Expert Mode, which is definitely what you want, Slic3r has about ten gazillion settings, all of which interact and all of which require tweaking if you're at all fussy about the results, which you certainly are. But Slic3r can't read your mind, it can't figure this stuff out on its own, it's not magic... but it's very, very capable.

Sit down and learn how the tool works before complaining that it doesn't work.

rant OFF

[swbluto]

given the exact same settings.

Back in the day, different slicers applied the Extrusion Multiplier setting in different ways: some used it as a linear multiplier, some used it as a quadratic factor. Perhaps Slic3r does it one way and S3D the other? At least with Slic3r, you can read the code and figure out what it's doing (and change it to work the way you want, if you prefer).

In any event, if you haven't calibrated the Extrusion Multiplier for the filament / temperature / settings you're using, I'd lay long money that the actual filament width isn't what the slicer expects: Slic3r is doing exactly what you told it to do, just not what you expected it to do.

rant OFF

Just found out that slic3r calculates it as a percentage over the first layer height. So my first layer height was .255mm, so slic3r was pushing .255mm out of a .4mm nozzle at 100% whereas S3D is pushing out .4mm out of a .4mm nozzle. So, I solved it, no problems there. The last test case turned out marvelous.