1. Got a question or need help troubleshooting? Post to the troubleshooting forum or Search the forums!

Simplify3D settings

Discussion in 'Software' started by drbanks, Mar 15, 2018.

  1. drbanks

    drbanks Active Member

    Joined:
    Dec 29, 2017
    Messages:
    141
    Likes Received:
    40
    I have been spending A LOT of time recently fighting with zits/blobbing on the exterior surface of my prints, especially when it's a curved surface. I've upped my retraction, told it to avoid retracting on layer changes, upped the coasting, and even reduced the multiplier to 90%. Very little effect.

    I use Simplify3D because I really can't stand Cura. I took their default settings for the Robo R2, and have slowly tweaked them to my liking, except for the zits.

    On a lark, I tried running my test print (just a 3" high, 1" diameter tube with threaded cap) through Cura, at maximum fine (.06mm) settings. Whenever I try that with Simplify3D, the finished product has so many zits as to be unusable. With Cura, the result isn't perfect but the zits are so sparse that you really have to look for them.

    So, I've been side-by-siding the successful settings in Cura against the not-so-successful settings in S3D, and aside from having to deal with slightly different names for some settings, and the fact that S3D sprinkles the relevant retraction settings across at least three different tabs, I have yet to find success.

    Really bothers me that the high priced spread has such inferior results.

    I finally think that maybe where we're going wrong with S3D is that I'd set it to "outside in" and the zits are the start of the extrusion on the new layer. Which would be just a fine hypothesis, except that this would imply that the inside of the Cura print (which was set to "inside out") would be zitfull, but it's not. And I haven't been able to test this because I didn't figure that out until after I'd started a day-long print.

    So, my question is: Has anyone had these issues with S3D and yes or no, do you have any PLA settings profiles for S3D that you recommend?

    (Using Hatchbox PLA, 60 degree bed, 190 degree extruder. Interestingly, I've run some multi-temperature bridging tests and I'm hard pressed to see any differences in the 185-200 degree range. Still thinking I might want to back down to 185, although I do note that the good print out of Cura was at 195.)

    advTHANKSance
     
  2. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    For S3D and zits you really are only assessing the following two parameters Coasting Distance and Restart Distance. That would depend on where the zits are, at the end of the line or at the beginning of the line. It should have little to do with Inside-out or Outside-in. It has been said that Outside-in results in slightly more accurate overall measurements.

    Even though I use S3D about 95% of the time, I have never felt the $150 USD price was justified, really if it were $50 then I could recommend it to nearly everyone. And they really need to strip the DRM, its like they are accusing me of being a thief even after paying for the software. So I usually suggest everyone exhaust the free options before spending a penny, Slic3r/Prusa, Cura, MatterControl in that order.

    I have tweaked my PLA settings a lot in S3D but I also use a 0.60 mm nozzle. I am not having those zits like you although I have had them in the past. Too many parameter changes to say with 100% certainty which ones solved the problem. I included it here so you can see what I did.

    Note, a negative number in Extra Restart Distance should result in slightly less extrude at the start of the line. Make adjustments in small increments something like -0.02 mm at a time.

    An alternative to negative Extra Restart Distance is using Wipe Nozzle and Wipe Distance with proper retraction settings.
     

    Attached Files:

    #2 WheresWaldo, Mar 15, 2018
    Last edited: Mar 15, 2018
    TheOneWhoWhelms and drbanks like this.
  3. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    Yea, no issues with any nozzle size we have used (0.25 -> 1.2)
     
  4. drbanks

    drbanks Active Member

    Joined:
    Dec 29, 2017
    Messages:
    141
    Likes Received:
    40
    So, my experiment of changing from outside in to inside out has resulted in a print with a smooth exterior but zits all over the inside surface. Would seem to imply that the problem is in the first shell of a new layer, and I presume the start of the shell.

    So, that would imply restart distance? Increase or decrease?
     
  5. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    Try a small increase in retraction and add a small bit of wipe.
     
  6. drbanks

    drbanks Active Member

    Joined:
    Dec 29, 2017
    Messages:
    141
    Likes Received:
    40
    tl;dr: in this case, not wipe, not retraction, but something entirely stupid

    I have spent far more time on this than I would have wanted. I've had to get jiggy with print profiles to an extent that I'd otherwise actively avoid. Googling the problem shows that people have had this issue (zits with S3D but not Cura or Slicer) on almost every printer S3D supports, with the same suggested fix (increase retraction, turn on coasting) and with the same frustrating lack of satisfaction.

    I spent some very tedious time comparing the GCode produced by Cura vs. S3D, and nothing popped out at me other than how much "learn to sight-read gcode" was actively not one of my life goals.

    One nice feature of S3D is being able to assign multiple print profiles to one print job. So, I was able to pick a small item that demonstrated the problem, make six copies of it, and assign a different profile to each one, so I could see a progression of changes from my starting point, with parameters being adjusted on each following one.

    I have a nice rack of six of these where the main thing I altered was increasing the retraction amount and/or coasting distance. What I found was quite conclusively that increasing retraction was actually making it worse, to the point where the zits looked more like smallpox.

    Going the opposite direction (reducing retraction and coasting) yielded the expected descent into stringing. At no point are we zitless.

    I'm blaming this ALL on the default Robo R2 configuration that their configuration assistant supplies.

    One point, that seems to matter a little, but not much, is the extrusion width. The default, generic printer profile specifies a extruder diameter of .35mm and a manual extrusion width of .40 (it can calculate automatically, but this was a hard override). In creating the default R2 profile, they increased the nozzle diameter to .40, but did not make corresponding changes to the extrusion width.

    I tried various settings of Z Hop, avoiding passing the extruder over already printed areas, wiping, forcing retraction at layer changes, etc. No Bueno.

    I finally noticed one thing that'd been bugging me all along:

    Infill extrusion Width: 130%
    (Infill) Outline overlap: 40%

    I noticed that the default printer profile had those at 100/20. I noticed that Cura has the second of those at 8%. So, I backed them down to 100/20 and not only did the zits disappear, but so did the stringing from my ludicrously low retraction settings.

    Seems that my zits were actually the infill poking through the sides.

    I really want to kill something.
     
    mark tomlinson likes this.
  7. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    Interesting, you didn't look at the profile I uploaded, they have Outline overlap at 18% and Infill Extrusion width at 99%. I have never used the S3D Robo settings, are they the same as in their GITHUB?
     
  8. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    :)

    Excellent detective work.
     
  9. drbanks

    drbanks Active Member

    Joined:
    Dec 29, 2017
    Messages:
    141
    Likes Received:
    40
    I did look at it. Just didn't think to look at that part. I took what I thought was relevant and copied to my profile. Mea culpa
     
  10. WheresWaldo

    WheresWaldo Volunteer ( ͠° ͟ʖ ͡°)
    Staff Member

    Joined:
    Feb 18, 2015
    Messages:
    5,905
    Likes Received:
    3,593
    Some background on how I do my profiles. Not wanting to use what others think are perfect settings, there is a lot of things done in 3D printing that are kludges to make stuff work. I set up my printer so it is as mechanically accurate as possible. If I ask for 100 mm of filament I want 100 mm of filament to extrude, usually most printers under-extrude out of the box (The R2 is pretty close though). I set offsets so that if I want layer 1 to be 0.20 mm, it actually is 0.20 mm not 0.16 mm because I need it to squish the first layer. Then I take a small model and set the slicer to auto everything and stuff that won't auto to 100% or 0% as appropriate. then print something with 100% infill. It will usually look horrible. I increase extrusion width and overlaps until 100% infill looks like it is solid and watertight. My thought process is that if 100% infill is perfect then everything less will be perfect too.

    On the first layer, common practice is to lower the first layer until it squishes the correct amount, but if my model is 20 mm tall I want it to be 20 mm when I'm finished not 19.9 mm because I needed better bed adhesion. All slicers allow you to address layer one differently than subsequent layers, so I adjust extrusion width or other multipliers until I get the proper bed adhesion without forcing the nozzle to be lower than what I asked it to be.

    Once that is all done, I now have a profile for that type of material with a given nozzle size. I don't need to adjust anything but layer heights for FINE, MEDIUM or FAST quality settings. I generally don't need to change for other manufacturers as long as they are close to the same.

    I usually run through this procedure for each type of filament (not each spool nor each manufacturer) so I end up with a PLA, ABS, PETG, NYLON, etc profile named accordingly followed by the nozzle size. (example PLA - 0.60). If I switch to a different size nozzle all I do is switch nozzle diameter and extrusion width and that gets me to a good starting point that may not need a single other parameter tweaked.

    Of course it takes longer to do things this way than just using someone else's idea of what works, but then I also know what small tweak I might need to make when I switch something out, and at this point they are all small minor tweaks because everything is so close already.

    This way won't work for everyone, you have to have a lot of patience and print a lot of stuff that will end up in the recycle bin, but it works.
     
    Geof, mark tomlinson and drbanks like this.
  11. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    Like Waldo, I have a lot of different profiles, all custom to my machines.
    The only thing I can add to that excellent description is that I don't do any configuration in the startup scripts for the slicer (or ending scripts) that is done in the instance of OctoPrint each printer is using. So regardless of which printer I am going to ultimately print on the slice is what it is. Anything specific to that machine behavior (Z offset or whatever) is done on the printer.
     
    Geof and WheresWaldo like this.
  12. Eric Viglotti

    Eric Viglotti Member

    Joined:
    Jan 18, 2017
    Messages:
    116
    Likes Received:
    22
    Thank you @drbanks for posting this and thank you to @mark tomlinson and @WheresWaldo for all of the notes. I refuse to give up on S3D as it creates generally such better output and so much more customization than Cura. I do believe that Robo just hasn't put enough effort (yet) into really pounding away with S3D and their printers, probably because they know that it's too expensive to bundle with their printers. Though I have been to their office and seen people using/testing with S3D. It definitely is said that @drbanks had to do this much trial and error to basically "fix" the S3D/Robo profile.

    I wanted to chime in here that I had a similar issue on my R1+ with zits and after working through with S3D multiple times, the best I could ever do was move the retraction point to 0,0 xy which worked for me because my curved surface wasn't a total cylinder so I could "buy" the retraction marks on an edge. But that always seemed like a hack that wasn't going to work on all models, so thank you for finding this solution.

    Thanks.
     
    mark tomlinson likes this.
  13. drbanks

    drbanks Active Member

    Joined:
    Dec 29, 2017
    Messages:
    141
    Likes Received:
    40
    In this case, it wasn't really Robo3D's fault. Do they even supply a S3D profile for the R2?

    The broken profile I got is what S3D installed when I ran it the first time and it asked me what kind of printer I had.

    Further research reveals to me that their handling of overlap changed significantly from v2.x to v3.x. I'm betting that they didn't update their library of default printer profiles.
     
  14. Eric Viglotti

    Eric Viglotti Member

    Joined:
    Jan 18, 2017
    Messages:
    116
    Likes Received:
    22
    Yeah, I suppose it's really S3D's issue as they are the one with the profiles. Per your note, they do supply a profile for the reasons you state that you can select it as a printer on that screen :)

    I see this as a bit of a shared thing as S3D had a few R2 printers to pound away at to know that they were producing software that would work with a given printer. However, I think Robo knows people use S3D, so they could be testing it out and suggesting/posting a new profile. That's my $0.02 at least...
     
  15. Eric Viglotti

    Eric Viglotti Member

    Joined:
    Jan 18, 2017
    Messages:
    116
    Likes Received:
    22
    Hello @drbanks and everyone. Good news, on my issue on the "innie", I was able to resolve this after many hours by adjusting the outside perimeters from 2 to 3 (http://community.robo3d.com/index.php?threads/r2-dimples-i-think.20762/#post-126760)

    However, I have a part with a large curved surface where I'm still getting the "outies" described in this post. I changed the settings to the following as noted above:

    Infill extrusion Width: 130%
    (Infill) Outline overlap: 40%

    But unfortunately on my curved surface I still get what you see in the attached photo. But this isn't terribly surprising as the preview in the "default layer.png" screenshot shows that it is going to happen and indeed they do, right in those exact spots.

    I have been able to get around this with changing the default layer start to be as close to 0,0 xy as possible as noted in the "layer" tab. And you can see in that "S3D - 0,0 xy start point layer" screenshot there are no "outies" because all of the "outies" are near a straight line near the 0,0 xy coordinates I have elsewhere in my part. However, for other parts I wouldn't have that luxury.

    What's weird is that I printed a simple cylinder yesterday and there are no outies there, so maybe it's more of of an issue with a gentle long smooth curve than a small circle's curve. Again, I can't help but think this is the infill "sticking out" over the curved surface and this doesn't happen in Cura. I really thought the suggestion above would help but it didn't.

    Anything else anyone can think of?

    Thanks!
     

    Attached Files:

  16. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    You really should open a ticket with Simplify as well (while not as fast they do have support) :)
    Or try their forums as well -- just to cover all the bases.
     
  17. drbanks

    drbanks Active Member

    Joined:
    Dec 29, 2017
    Messages:
    141
    Likes Received:
    40
    130/40 were the bad settings. I changed to 100/20. The 130/40 are the (broken) S3D defaults
     
  18. drbanks

    drbanks Active Member

    Joined:
    Dec 29, 2017
    Messages:
    141
    Likes Received:
    40
    To be clear, my "outies" were a result of the infill poking out the sides. I addressed it by reducing the infill outline overlap percentage to 20. Cura's is 8.

    But when all the zits are lined up along a line like that, it makes me suspect the extrusion start or stop along the point where the layer starts/stops. My problem was complicated (by me) that I was using the random start. But yours are so obviously along a line, it does point back at retraction.

    I took a known bad example that was pretty small, put an array of 6 of the on the bed and assigned each a separate print process, where the difference between each print process was a change in one of retraction distance, restart distance or coasting difference. If nothing else, it gave me a decent enough idea of how each would affect the print.
     
  19. Eric Viglotti

    Eric Viglotti Member

    Joined:
    Jan 18, 2017
    Messages:
    116
    Likes Received:
    22
    Thanks much @drbanks for correcting my misread on this post. Turns out I actually had it correct the first time, saw those marks, then moved it to the incorrect config and saw them again, so as it turns out, either way these are still happening.

    Yeah, to @mark tomlinson point, I have contacted S3D support, many times and actually find them to have outstanding support. Everything we tried didn't work and I always wonder whether the hardware people point to software and software point to hardware, so that's why I figured I would try here :)

    Anyway, I'll keep playing with it, and for now, I'm fine with that 0,0 xy start on this model, so I'll check others to see. There is no question it is somehow related to retraction as that is most certainly where retraction is shown on the preview as these "outies" are not random at all, they line up dead perfectly with what is in S3D's preview.

    Thanks all as always.
     
  20. mark tomlinson

    mark tomlinson ༼ つ ◕_ ◕ ༽つ
    Staff Member

    Joined:
    Feb 21, 2013
    Messages:
    23,912
    Likes Received:
    7,338
    Of course that happens, but remember, while this is a hardware focused forum we are all printer owner/users just like you and we are going to try to solve the problem if we can. Just want to make sure you are using the best resources. Bottom line if this is working from one slicer then failure from another is a slicer issue. Most of the slicers (but S3D more than other current ones) offer a lot of ways to tweak what the slicer is doing at a low level and that is powerful, but also takes a lot of experience to get right :)

    We never give up.
     

Share This Page