Jump to content

ultiarjan

Ambassador
  • Posts

    3,778
  • Joined

  • Last visited

  • Days Won

    76

Posts posted by ultiarjan

  1. just water and soap... but only after many prints... and I mostly use a super small amount of glue-stick spread with a moist paper towel.

     

    I also sometimes scrape the glass-plate when its dirty from to much glue, while still in the printer, with a glass knive, and then reapply a thin layer of glue, or just use a moist paper towel without adding new glue.

     

    This really is my nr1 tool.

     

     

    image.png.dcb476b7dc446f7b5c82a6da1e9b71e3.png

    • Like 2
  2. 5 hours ago, mkaj2019 said:

    Great mod! How far could you reduce the retraction distance with your mod?

     

    I only used flex material, though direct drive is obviously beneficial for retraction behavior, I feel the added weight of this hack is so bad that it doesn't make any sense to look at other materials. 

     

    5 hours ago, mkaj2019 said:

    Have you tried using the the Z offset Plugin from the Cura Marketplace.

     

    I did not, but I think it's basically another means to achieve the same, weather you reduce flow at the same height, or increase height at the same flow, the result is avoiding to much pressure build up.

     

  3.  

    S5.thumb.jpg.154c9f9d198d1fc71fa335cd61f5b0db.jpg
    I was looking into a way to occasionally print Ninjaflex or other real flexible materials (85A)
    There's no way you'll be doing that reliable with a bowden setup (ok, I've seen people using oil in the bowden but nahhh), so I compared a quick and dirty direct drive conversion on an Ultimaker2 head (tested on my DIY GO) and on the ultimaker S5

     

    S5_inside_electronincs.thumb.jpg.d0c95060d2356b902cef98751c6bd525.jpg

     

         ****
    first a warning, Ultimaker does NOT want you to open the S5, it contains an open style high voltage power supply, so it's DANGEROUS!!!, also the electronics can be damaged and I'm pretty sure you loose your warranty.  

         ****

     

    I used two materials for testing, Extrudr TPU Soft and Ninjaflex, both 85A.

     

    2-pancake.thumb.png.218c07fbd75ce763a4d1a5b8c0f3d0fd.png

    I used the regular feeder from the UM2+ , combined with a pancake stepper motor (LDO-42STH25-1404MAC#190920) this stepper is less than half the weight of the normal stepper. I had to cut some length of the shaft to make it fit. The stepper has a different # steps per rotation and as it's impossible to change the steps setting for the S5 on one feeder it needs compensation in flow (200% flow = 100%)

     

    Obviously the result is a crazy top heavy setup, but if you just want to print an occasional gasket it should be fine, specially since you need to print materials this flexible very slow (15 mm/s) anyhow. On an Ultimaker2 the extra weight is enormous, on the S5 not so much as it's very heavy already anyhow.

     

    3_UM2-S5-direct.thumb.jpg.21fff551508b1eac5cba51441bcd2806.jpg

     

    The nice thing is you can see into the feeder so during testing it's easy to spot when the feeder can't cope if you try to speed it up to much. If something goes wrong it's mainly during the first few layers, so watch them... otherwise you'll spend a few minutes again taking the feeder apart...

    arggghhh.thumb.jpg.dd64d054705667b4afe6919d87ceb391.jpg


    To validate the basic setup I printed regular PLA on both converted machines, to make sure the setup (including changed steps/flow due to different motor) was correct.

     

    On the UM2 head the ninjaflex was pretty easy to get working, i managed to print with constant and reliable extrusion using a 0.4 nozzle.

    The materials prints best around (I tested with 0.15 layers)


    Flow :         110-115 % (with  feeder tension set to lowest point)
    Temperature :    210 c
    Speed : 15 to 20 mm/s

    The easiest way to test temperature and flow is by printing simple cilinders, foir the S5 you can use the "tune" menu to play with values on the fly.

     

    Tune-to-test.thumb.png.d5aafe36716ca2c7ea4e857f757f1e13.png

     

     

    Getting ninjaflex to work on the S5 proved to be more difficult than on the UM2, at the same settings using an 0.4 mm nozzle prints failed all the time, it seems the CORE, is much more of a challenge that a UM2 setup. I blame the longer filament path, and the metal tube of the core. Only when I changed to the 0.8 nozzle and the CC0.6 I was able to get reliable results.

    2vs5.thumb.jpg.c6d4b436955f9a3108e6449b3e3448bc.jpg


    My main S5 tests are with the CC0.6, atm my best results are with the following settings;

     

    The S5 levels pretty close to the bed, making every print fail at normal flow rates.

    Layer height 0,15
    Initial layer height 0,25
    
    Top/bottom pattern & Innitial layer pattern ; concentric (avoid lot's of small lines, head vibration)
    
    Innitial printing temperature/ innitial layer : 230c
    Printing teperature : 215c
    
    Skirt/Brim flow 140% (= realy 70%, due to the different stepper motor)
    Innitial layer flow 150% (= realy 75%, due to the different stepper motor)
    Flow 225%  (= 2x the real flow, due to the different stepper motor)
    
    Print speed 15 mm/s
    Innitial layer speed, Skirt/Brim speed 10mm/s
    
    Cooling at 100%, start with 20% and let it rise over 10 layers)

     

    Turn off the flow sensor (in the settings menu) on the S5, as you're not using that feeder..

     

    Conclusion, even if you eliminate the bowden, printing 85A flex on an Ultimaker S5 is still a bit of a challenge. The used feeder (UM2+UM3 model) and filament path are not ideal. But, usually if you make it past the first few layers, and have relatively simple prints that keep the flow going it's very well possible to print ninjaflex on the S5.

     

    I have not experimented a lot with it jet, but I also managed a simple test print using PLA for support on the 2th extruder.

     

    the hardware hack;

    The feeder plate mounts to 3 screws of the head, so you can leave one and the head will stay assembled during mounting.

    4-3screwMount.thumb.png.afe9381460ee73d7fc78d946ba58c985.png
    I made a little breakout plug for the left stepper motor on the back of the printer, so I can change between direct and bowden in just a few minutes. I just leave the black mounting plate on the head. In needed to make a hole in the bottom plate for the DIY extension cable. I isolated the plug with hot glue. 

    20200216_193702.thumb.jpg.db905153fcc6f95c07c35dd9621b2843.jpg

    20200216_123631.thumb.jpg.23197577acd486dbccc236c17bf1125c.jpg

    some side notes;
    Cura can sometimes create fantastic vibrator g-codes, not nice with this heavy setup, so one thing you can do to prevent this is  to change line fill to cylindrical fill, and check the layer preview.

     


    openRCwheel.thumb.jpg.568133de05a249a72da6b52f18a0ce18.jpg

    ninjaflex openRC wheel, printed on S5

     

    gasket.thumb.png.90efc908d195ba453c5e81b918514b3e.png

    extrudr TPU soft (85A just like ninjaflex) gasket, printed on S5

     

    S5 stl's

    S5_FlexDirect.zip

     

     

     

    • Like 11
  4. There are differences in solubility between different brands of PVA, but more soluble also means harder to print, so probably not the best trade off.

     

    It's better to use "support blocker" in cura to avoid support in places you do want it, don't ask me to explain how as it's not the most intuitive function of cura...

     

  5. Sure , I did both already 🙂 at least a proof of concept, never got to really publishing stuff...

     

     

     

    I did make a better version of the single core head, using a single core bottom plate 3d printed in metal (shapeways), but I never really documented it properly..... always a lack of time and starting to many new projects 🙂

     

     

     

     

     

  6. 22 hours ago, DrCeeVee said:

     

    I'd appreciate it if you could share any information about your drybox, and how it performs when used with PVA. Thanks.

     

     

    I just use a diy plastic box with diy spoolholder, filled with a large bag of silica gel, but considering the low price of the polybox just get that one... (probably get one myself someday..)  For me PVA keeps performing very good for many months.  Between jobs I store the fillament in printdry containers. I assume results will vary a lot depending on when you live in the world...

  7. just to get a feeling for the scale, original UM2 vs Hemera, both without fans

     

    image.thumb.png.6ddc54493550b347b62dddf48f6f90b8.png

     

    I've been looking at this as I would like one of my machines converted to direct drive for flexible materials. The hemera is to big for my taste, a pancake motor version would already be better. 

     

    I'm also looking into a bondtech BMG + pancake + mosquito hotend, this looks a more interesting option imho.

     

     

    • Like 1
  8. For most materials I like to use glue sticks for bed adhesion (I use TESA easy stick) and paper towel, to (a little wet) spread the glue to a very thin layer. Get good quality paper towel that wont loose fibers when getting a bit wet.

     

    I use Ultimaker adhesion sheets for some flexible materials .

     

    Next to a "spachtel" i like to use a glass-knive like this one from stanly in case you can't get get under the print with a "spachtel"

     

    image.png.8edef1283a9a7259986aadedc88a10ec.png

     

    For some materials it helps to cut the end at an angel before loading the material, and bending it straight a biut (specially for older materials, or the last bit of a spool)

     

    image.png.07c3374fef620e71667f16f5dfb7b667.png

     

    • Thanks 1
×
×
  • Create New...