Simplify3D User Forum

I read an article in a forum that I can't recall or find, that may be the answer to this issue.

First, this problem is more evident on horizontally curved sections and is absent on horizontally straight sections, such as the side of a cube.

The article had to do with computing the speed at which material is fed to the extruder (I'll call this feed rate, and differentiate it from the rate at which the head travels). It stated that because there is less material needed on the inner part of the trajectory, the feed rate must be reduced. (This theory doesn't explain why the excess on the inside isn't offset by the additional material needed on the outside of the curve. The details of the computation would be needed to understand.) This jibes with the observation that blobs are worse on smaller radii of curvature. The question is if the slicers compensate for this effect, or if there is some issue with the coding.

So what's the mechanism ... good question. On a curved section, does an excess of material build up and the SLICER compensates by periodically creating a blob in the tool path? This implies that the slicer "knows about it" but is unable to reduce the feed rate. Reducing the extrusion factor may help, but causes underfill on planar vertical surfaces of the same object.

My assessment is that this is a deep and subtle issue as evidenced by the fact that it is exhibited by several slicers. But, I wish someone would solve it, because it really destroys the surface finish. I'd be happy to help, just tell me what you would like to do. My main printer is an M2.

Rich Thall

I have seen this problem before and for me the issue was simply that the computer could not keep up with the gcode message passing (i.e., CPU or bandwidth limitations) for curved surfaces. So what happens is the printer pauses every once in a while to wait for more commands. This causes a blob at that point. Decreasing gcode resolution (such as oversample resolution in KISSlicer) fixes it, or decreasing STL res, or getting a faster computer and/or minimizing gcode preview during prints (in Pronterface this works for me). Both KISSlicer and Slic3r have a way to limit the gcode resolution, but I'm not aware of a way to do this in S3D?

jd_3d wrote:I have seen this problem before and for me the issue was simply that the computer could not keep up with the gcode message passing (i.e., CPU or bandwidth limitations) for curved surfaces. So what happens is the printer pauses every once in a while to wait for more commands. This causes a blob at that point. Decreasing gcode resolution (such as oversample resolution in KISSlicer) fixes it, or decreasing STL res, or getting a faster computer and/or minimizing gcode preview during prints (in Pronterface this works for me). Both KISSlicer and Slic3r have a way to limit the gcode resolution, but I'm not aware of a way to do this in S3D?

The speed of you computer will not cause the printer to pause, unless your computer cannot send data over the USB at the requires speed, which no computer 99% of the people are using are that bad. I have a dual quad core machine with 10gb of ram running a super fast SSD as the HD and I get blobs, so machine speed is not it. It is a problem with the slicing method and extrusion amount. I have gotten blobs on a 20mm test cube, so no high mess or resolution. It was sliced at 0.4mm. Slic3r and KISSlicer didn't blob that same model, so it's the math.

My trial ran out so I cannot test that out, but if you set your extrusion multiplier down some I bet the blobs will go away or get less pronounced.

Update:

Ok, after some exhaustive testing and twiddling, I've come to the conclusion it's actually my printer at issue and not the software, itself. I think I can chop this up to a timing coincidence.

After the trial ended, I reverted back to my Repitier software to do some work and was experiencing the exact same results with globbing. It wasn't doing it before I switched over and first print with the S3D software was ok, but every one after had the issue and now with the old software (Repitier).

I Re-calibrated everything (for the umpteenth time) and still no luck. Changed out the filament, etc. I even went as far as moving the printer to another computer. No joy.

TBH, the Solidoodle 3 I bought has been dicey, since day one. Long story, but it's never printed a "quality" piece, to date. I thought perhaps (hoped, actually) it was the Repitier software. The support staff at Solidoodle haven't been very helpful, either. Actually, they've been pretty bad. I can't even get them to give me exact measurments on what the tension on the belts should be, other than "finger tight". Ridiculous.

So.... I'm ditching this crap-tastic Solidoodle 3 and will be getting an M2 or a LeapFrog Creatr, most likely. Lesson learned: spend a little more and get a better printer.

I want to thank Clayton for his continued assistance, even while he was on vacation. A lot can be said about getting support from someone, even when it impedes on their own personal time. I'll be giving the software another go, once my new printer arrives.

Well I have done the same test as you, but with Slic3r and KISSlicer printer prints perfect, but with S3D it blobs

wrex wrote:Update:

Ok, after some exhaustive testing and twiddling, I've come to the conclusion it's actually my printer at issue and not the software, itself. I think I can chop this up to a timing coincidence.

After the trial ended, I reverted back to my Repitier software to do some work and was experiencing the exact same results with globbing. It wasn't doing it before I switched over and first print with the S3D software was ok, but every one after had the issue and now with the old software (Repitier).

I Re-calibrated everything (for the umpteenth time) and still no luck. Changed out the filament, etc. I even went as far as moving the printer to another computer. No joy.

TBH, the Solidoodle 3 I bought has been dicey, since day one. Long story, but it's never printed a "quality" piece, to date. I thought perhaps (hoped, actually) it was the Repitier software. The support staff at Solidoodle haven't been very helpful, either. Actually, they've been pretty bad. I can't even get them to give me exact measurments on what the tension on the belts should be, other than "finger tight". Ridiculous.

So.... I'm ditching this crap-tastic Solidoodle 3 and will be getting an M2 or a LeapFrog Creatr, most likely. Lesson learned: spend a little more and get a better printer.

I want to thank Clayton for his continued assistance, even while he was on vacation. A lot can be said about getting support from someone, even when it impedes on their own personal time. I'll be giving the software another go, once my new printer arrives.

All,

I don't know if this helps explain it or not. I set up a print with Z-lift and retraction (both at 1.0 mm IIRC). It seemed like the "sequence" of these operations was handled incorrectly. I expected:

1. Retract and/or Lift Z (together or sequentially)
2. Do a fast move to a new place on the part
3. Drop Z back where it was and/or return filament where it was
4. Start next extrusion from there

What I saw happen (and I'm quite certain of this because of the length of the retraction and Z lift moves --- I can take a movie of it if needed) was this:

1. Retract and Lift Z
2. Drop Z back where it was and return filament (!)
3. Do the fast move (at the original Z, not the lifted Z) to the next location on the part
4. Start next extrusion

When the fast move occurs after the Z drop, the nozzle can "clunk" into imperfections or blobs in the layer. The whole point of the Z lift, I thought, was to "jump" over imperfections. As it works on my M2, the Z lift doesn't do much but pull up a blob/whisker, moosh it, then move quick across the part, with AUDIBLE clunks into the part. I had a round (screw) hole in the part caught by the nozzle on a move and knocked out-of-round this way.

I sent my .factory file to support at simplify3d in a separate e-mail. Just thought I'd see if others can confirm this behavior of Z-lift and/or retraction. Using Creator 1.1.4 on Linux Mint 13 / Cinnamon, on MakerGear M2 with black MG PLA (kind of "damp", not fresh dry PLA, sorry).

Can anyone else confirm this sequence of moves with retraction and/or Z lift?

Dale

I can confirm that the Z-lift / retraction / fast move sequence was fixed in 1.1.5.

I wonder what else (possibly not listed in the release notes?) was fixed from 1.1.4 to 1.1.5. Is it just my settings, or did the blobs go away? (See attached photo... note the short outer face in the center of the picture. There is an "opposite" face off-photo to the left that has "blobs" on the layers for which this face does not have blobs.)

"M2 Light Bar" printed with MG black PLA 1.75 mm, 0.35 mm nozzle, .2 mm layers, 195C/70C/50% speed first layer, 190C/65C remaining layers, 0.6 mm retraction, 0.5 mm Z-lift. (The bottom one has the super-bright white LEDs in place, with the leads folded out to hold the LEDs in place for the moment.)

Dale