Close



Page 2 of 5 FirstFirst 1234 ... LastLast
Results 11 to 20 of 50
  1. #11
    Technologist
    Join Date
    Oct 2016
    Posts
    184
    Ok back at it after some relaxation away from 3D printers . So I noticed that in filaswitch, I had omitted extruder prime step when using purge bucket so the initial ramming length was too short to get a consistent good tip. Fixed the bug and did a rerun of the same benchy gcode as last time.
    This time the print finished without a catastrophic tool change fail, although there a some evidently missed or at least partly missed layers seen. Also I ran out of blue filament right at the end so the roof of the benchy is incomplete.
    IMG_20200814_184134.jpgIMG_20200814_184204.jpg

    So there's still tuning to be done. Currently I'm experimenting with PTFE-lined 2x-feeder part. The idea is that PTFE-lining has less friction obviously so that would make the filament changes more reliable. Still the design requires cutting PTFE tubes and creating junctions which have some edges that might ruin the whole idea. But if this happends to work, I'll be chaining multiple of these to get to 6x material. This will also at least partially fix the problem with current big feeder part that's bulky and causes some stress on the X-carriage when all the PTFE tubes are attached to it, especially when the head is on the edges.
    IMG_20200814_185624.jpg

    Purge bucket: I really need to design a better way to store the waste material than just extrude it on the floor. If I would just throw it away, it wouldn't be that big of a problem, but I'm thinking about recycling all the waste parts I have so I want to save the waste for later. And having it on a long strings is definitely not the way to go because it needs to be cut to small pieces etc... for multi-material printing I'll probably just use purge tower for now, at least all the waste is in compact form and easy to store.

  2. #12
    Technologist
    Join Date
    Oct 2016
    Posts
    184
    Got the first prototype of the PTFE lined 2x feedert built and tested. Some details need to be tuned, but so far the results are promising. The filament path is a bit tight, I'm using Capricorn tube and the M4-nuts compress the tube also, even after opening the tubes with 2mm drill bit. But doesn't seem to matter, the 2x color Benchy had no issues with tool changes, they were seemed to work very smooth, no grinding of the gears or anything.
    IMG_20200818_195754.jpgIMG_20200819_185359.jpgIMG_20200819_185529.jpgIMG_20200820_184015.jpgIMG_20200820_184128.jpg
    The print has some blobs, smaller and bigger. The smaller are probably some retraction or filament related. The larger ones are happening from the head moves from the purge tower and continues printing the model and is realted to the tower retraction handling that filaswitch does. MY guess is that 2 things are happening:
    1. the outer wall of the tower is printed quite fast up until the move back to model, causing bigger pressure on the nozzle and the default retraction isn't enough in this case.
    2. S3D does retractions partly wiping during the retraction so the effective retraction is probably more than just a normal retraction. Filaswitch doesn't do that and I think I'm not going to implement that as it involves quite a lot of gcode ananlysis. However, this wipe-retract might also show as a discrepancy in retraction in model vs. the tower.

    I updated the filaswitch code to slow down printing of the wall. I'll also add option add extra length to the tower retract so it can be fine-tuned better. Currently printing another benchy and it already seems better even with only the wall speed tuning done.

    Oh and my MMU2S is finally shipped, nice

  3. #13
    Staff Engineer
    Join Date
    Jul 2016
    Location
    South Florida, USA
    Posts
    1,248
    Add AutoWiz on Facebook
    Quote Originally Posted by spegelius View Post
    IMG_20200820_184015.jpgIMG_20200820_184128.jpg


    Oh and my MMU2S is finally shipped, nice

    That is Awesome. And that is Awesome.

  4. #14
    Staff Engineer
    Join Date
    Jul 2016
    Location
    South Florida, USA
    Posts
    1,248
    Add AutoWiz on Facebook
    I want to do something to help you buddy. Because you are so cool. I have been studying your multi material benchy and it's blobs and I think I know something that Will help. Linear advance or lin_advance is a way to tune for pressure in the nozzle. In theory a perfectly calibrated lin_advance would need little to no retraction. With long PTFE tubes and in line switches/merge points between the driven gear of the extruder and the nozzle your machine would most certainly benefit from enabling and calibrating this setting. I follow this tutorial for that.. https://www.youtube.com/watch?v=n3yK0lJ8TWM&t=

  5. #15
    Technologist
    Join Date
    Oct 2016
    Posts
    184
    Thanks, I do use Linear Advance with Marlin machines and Pressure Advance with RRF usually, but with current S3D profile I'm using a setting in S3D called 'Wipe while retracting' (or something) that kinda emulates the LA/PA functionality. I've noticed that having this on messes with the LA/PA so for now I'll go without.

    I've gotten the bigger blobs better, there was a small bug in filaswitch code that I fixed. Also tuned the purge tower retraction and as the purge tower is printed with ~200% extrusion, I added some wipe movements to even out the pressure. For the smaller blobs I changed S3D to do the seam on corner so that should hide most of them. Will post pictures later.

  6. #16
    Technologist
    Join Date
    Oct 2016
    Posts
    184
    Well I got the settings tuned quite well, after quite a lot of benchies with varying settings. I actually suffered heatcreep along the way so that probably also affected the results a bit. Especially this last color, red PLA seemed to suffer from it, the small columns were crap (you can just see the uneven extrusion on the second pic). After adding a filament oiler with a bit of olive oil the extrusion stabilized.

    IMG_20200824_204421.jpgIMG_20200824_204428.jpg

    There's still some artifacting from extruder wipe and the post-tower retract isn't 100% correct, but I think I'm gonna call it good for now. And maybe print something else, my armada is getting big enough already. Seems filament changes are working quite stable so far.
    IMG_20200824_204512.jpg

  7. #17
    Staff Engineer
    Join Date
    Jul 2016
    Location
    South Florida, USA
    Posts
    1,248
    Add AutoWiz on Facebook
    I am so jealous of your navy. Hell I am jelli of the purge towers behind them. Man you got your stuff working good.

  8. #18
    Technologist
    Join Date
    Oct 2016
    Posts
    184
    You'll get there too. It took me quite a lot of time to get things tuned so that the quality and realiability is decent. And there's still rooms for improvement, but I'm quite happy how things work at the moment.
    Also the filaswitch development has taken quite a lot more time than I anticipated. I mean, how hard it is to do a post-processing script that just adds few lines of gcode to existing gcode-file? Well now the project has been under development since 2017, has 434 commits, few thousand lines of Python code and I still find random bugs and things to improve now and then..

    Printed a dual color rose as a gift. The stem is printed separately and glued to the petals. Model: https://www.thingiverse.com/thing:2354319. Used FrankenCube for a change, I haven't done any prints on it for a while, nice to see that it's still working as good as it was. I dried the filament properly and might be that there's bit less stringing vs the last time I printed PETG with this printer.
    I kept the purge tower/purge amount quite small as the color contamination is actually a good thing in this model, the clear PETG has some red on it.
    IMG_20200828_183317.jpgIMG_20200828_184210.jpg

    Assembled the 6x feeder with multiple 2x feeder modules. Don't look like much but after short testing it seems to work. In future, it's best not to use capricorn tubing for the feeders, it's simply too tight when clamped inside the feeder parts. Had to do some fine tuning and now it's not too restrictive, but regular PTFE tube seems to work better in this case.
    I did only one test print with two colors, a dual color fidget widget https://www.thingiverse.com/thing:1775648. Currently don't have enough different color PLA, so need to order more colors. But 6-color prints are on the backlog...
    IMG_20200826_191510.jpgIMG_20200828_183326.jpg

    Ordered one set of genuine Bondtech gears. Readl deal on the right. More shiny than the chinese clones. Better? Don't know yet.
    IMG_20200828_184446.jpg

  9. #19
    Technologist
    Join Date
    Oct 2016
    Posts
    184
    Well I had enough filament left for one 6-color print. Didn't finish fully intact, but did finish. Copule of the filaments started jamming at some point and there's few holes here and there. Probably need to open up the feeders made from capricorn tubing. Also the points where the M4 nuts are seem to be a bit tight so maybe a 2.2 mm drill bit would be good...
    IMG_20200831_195508.jpg

    Of and my Haribo arrived, with complementary MMMU2S, how nice. Hope to get this assembled some time this week.
    IMG_20200831_113426.jpg

  10. #20
    Staff Engineer
    Join Date
    Jul 2016
    Location
    South Florida, USA
    Posts
    1,248
    Add AutoWiz on Facebook
    YES! Oh man. Follow them instructions and ration them Haribo's as instructed by the instructions. Congrats! And don't forget this awesome video that is a good introduction to important setup calibrations. https://www.youtube.com/watch?v=XH-F8h7Y1tM&t=

Page 2 of 5 FirstFirst 1234 ... LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •