ply2

Haha it works, it works! Can't wait to mount this onto my UM2. The only "special" parts used are 2x E3D springs for the latch. Once tested and debugged this will be released as CC0 ofc. First manual tests are promising: the huge knurled nuts have an iron grip but leave almost no markings on the filament. The whole thing has a little high inertia because of the combined weight of all the 608 bearings i used. So to the feeder-modding-pros in this thread: What do you think, what are the chances that this assembly will not destroy my printer? Am right that i can divide filament diameter by sqrt(2) to achieve double feeding? Like, instead of 2.85mm i dial in a filament diameter of 2.02mm and then the feeder will double it's action? Or are there (apart from the retraction setting) other magic things the cura->firmware toolchain does that totally depend on 2.85mm (wiping, coasting, blabla) that might ruin my prints? Update: Noooo! Seems the "feeder knurled wheel" (part 1179-B2P-A) and it's small set screw seem to be glued onto the motor shaft? tried to unscrew the set screw with wd40, heatgun, inch-unit-allen key (which is a bit too big and therefore fits extremely tight), it didn't move at all. I thought i tried this before i started but now it seems i didn't. So it looks like i have to replace the feeder motor if i want to try my feeder :-( The UM2 docs do not say anything about the type of motor used as feeder motor. Only the Z motor is specified in detail. Can i use a regular 1.8° nema17 or is some special motor needed? I think i heard the board is reading the motor feedback to decide for skipbacks. (I have a bunch of ACT 17HS5425/max 2.5A here)

Here's another one A couple of 608 bearings, 2x DIN 467 knurled nut, 2 springs, endless hours with freecad and about 1 spool of pla. I changed the transport ratio 1:2 compared to the original system to compensate for the friction of the gears and to make skipbacks more unlikely in general. Is there a preferred method to tell the printer that the feeder has to turn twice as much as normal?

No it didn't work for me. It turned the temp sensor into Schroedingers PT100. (It's OK unless you look at it). The hull of the sensor was ripped off and still sticks inside the heater block. The fun part is that the tiny sensor chip with it's itsy bitsy cables survived this like nothing happened this is why i posted the pics. This is how it should look: https://ultimaker.com/en/support/view/254-disassembly

You could download the "old" cura (15.04.2 or so?) and tinker with the "fix horrible" settings in the "expert settings" dialog (first idea: try turning them all OFF except "extensive stitching"). Alternatively you can try minkowski() with your object and a tiny cube in openscad.

What i learned today: Dont pull the temp sensor out while it's hot! :-( After a lot of WD40, heating, cooling, twisting, pulling and cursing i had the clever idea that it might help to head up the block to ~220°C. A soft pull was enough to get the sensor out. It looks like the PT100 is intact, maybe i can give him a second life as ambient temp sensor in another printer...

when you pour aluminium into a PLA-filled gypsum mold the explosion might KILL you. The mold has to be burned at "it glows"-temperatures (much more then needed to get rid of any plastic) anyway before you can use it for aluminium casting. If you are looking for a material which allows very thin layers and produces almost invisible ripple, XT might be an alternative to PLA. To me it seems that PLA gives a little more detail and less stringing while XT gives a little more surface smoothness and allows smaller layer heights. If your parts dont need the full strength of aluminium, maybe you should try one of the many "silicone-cast then epoxy+metal powder" recipies instead of the suicide stuff. The resulting material looks & feels like metal, can be polished like metal, can rust like metal. Casting aluminium in a quality where the UM2s print quality becomes a relevant factor would require master skills anyway.

I checked the PTFE isolator and it was in pretty bad shape - quite surprising since the machine is almost new. :-/ Also the small silver tube between the isolator and the brass nozzle was not fully screwed down onto the brass. I treated the "elephant foot" of the PTFE coupler by "polishing" the inside with a 3.5mm drill for a while. Now i can PLA-print the famous test tower with 4mm³ at 228°C and 5mm³ at 238°C (Innofil PLA) completely skipback-safe - horray! :-) Here ist what i concluded today: 1. the UM2 with a straight PTFE isolator can print with PLA very nice! 2. the PTFE isolator does not like it hot. 3. the PTFE isolator is also a bit short (when i push down the filament while the print head cools down for an atomic pull, i get softened PLA above the coupler - if this happens at the end of a print you probably have your filament blocked the next time you are starting a print. 4. the spare PTFE isolator you can buy is diamond filled or so. :-( 5. At the Ultimaker factory the hex screws of the print head are tightened by a gorilla. Meanwhile, i brushed thin ABS-Acetone-Slurry on my ABS parts. It seems to be quite failsafe when you need strong and dimensionally stable parts where the surface look is not so important. The slurry makes the shell of the part up to about 0,5-1mm depth very soft (depends on how much slurry you brush on) and fully fuses the material, while the infill and inner layers are somehow not affected and keep the shape of the part exact. On some large "ceilings" a few top layers delaminated and got bumpy but they somehow auto-repaired over night. The downside: Respirator or fume extraction = Mandatory

The main reason i prefer ABS is that printing other materials literally takes forever. Maybe something is wrong with my nozzle, don't know. I can do the speed test tower with PLA at max ~ 2 mm³/s, XT and PET+ at 3 mm³/s, PLA with sewing machine oil almost 3 mm³/s before getting skipbacks. ABS doesnt seem to have a relevant speed limit, 10 mm³/s -> no problem (actually there is a small problem that the build platform starts to vibrate when printing flat roofs too quick. it helps to position the print at the back of the platform). It was like this from the beginning, atomic pulls are impossible. I ordered "the olssen block", once able to use a bigger nozzle I will definitely go for PLA because it has nice strength+rigidity indeed. Another thing i will try is to cast the ABS print in gypsum and heat it in the oven slowly. I have a small old oven here that i can afford to treat with blasting cement bricks

Hi! I am printing big parts that need to have some mechanical strength from ABS. The printed parts have a lot of inner tension, making them want to delaminate, when they come out of the printer they already have tiny cracks in the corners, especially when the part is big & printed with high speed. How can i release this inter-layer tension? I figured out that painting them with ABS-acetone-slurry somewhat bonds the shell quite nice but i would prefer a less handcrafty method if possible. I think maybe cold acetone vapour or boiling the parts in hot water could help but didn't try yet. Update: Boiling ABS in water converts tension into shape :-(