flashp0int

When I changed the PTFE orientation (180°), it delayed retraction problems and soon after the retraction problems I started to have serial-cloggs. The new PTFE solved that and it printed better than on day 1. But after just a month, I had to get a new PTFE, as 30 hours printing Colorfabb XT already ruined the new one ( a very expensive print ). My intend is to talk about this with Ultimaker in Paris. If this is something you need to change monthly, it should be much much cheaper.

Any idea where we can get Vespel ? It looks like a very specialized material. I have three burnt PTFE I could cut in order to beta-test.

That looks really promising, The teflon is really a weak point of the UM2 if you printing something different than PLA. In the mean time, Ultimaker should sell them at a lower price imho.

FYI, I got a new Teflon piece from Ultimaker, that solved out the retraction issue totally. Now I can print the 9 tubes with retraction and no underextrusion.

This looks really great! And there is a Rostock version in the pipe, that's even better as this is my second printer

Ok, then there might still be hope. It is about a week ago. I am not used to so long delays. in my business, we have 4 to 24 hours SLAs based on what type of device. So a week seemed very long to me. Thanks.

How do you contact UM2 support ? I have tickets open there since over a week with no answers.

Thanks @MartijnvG Retraction speed is 25 mm/s (default I guess), the 30mm/s are the print speed (sorry for the confusion). I'll definitely look into playing with that setting. I have to print slower, as the tubes move and vibrate if I print at 50mm/s. Any thoughts on the Teflon piece and the hot-end ? Btw, would be great if we could tell Cura to Densify infill close to the shell (somewhat possible on Slic3r) Change layer thinness and speed at given height (even better if automatic) Best regards

Would really appreciate some expert insight on that one. Thanks in advance !

I dismantled the head in order to check the Teflon. Turns out the Teflon has some sort of dent on one side and not on the other. Could this be the reason that prevents proper retraction and flow? It seems to be on the side where bowden tube might pressure the material into the teflon piece. (check picture below). Another surprise was that the heat cartride and sensor seem to be clued or stick to the heat piece. I did not dare to pull with anything else than my fingers. Is this normal ? Unless there was a clog I never pushed over 230°C.

On the original one, if I nearly fully loose the screw, it works with PLA and FlexPLA. But PET and Nylon did not. With the 0.8 mm longer yoke, i sort of fixed that. The difference is really small, but I could not loosen the screw more at that stage. I'll measure exactly the wheel.

Thanks @3Dmaker4U The temperature is 16-19°C, that should not be too hot. and the bottom of the UM2 is quite well ventilated. I guess I am going to take the head appart and check the teflon to be sure. Enabling retractions on prints with way fewer retractions does not cause under-extrusion. This really starts happening over time. So retractions every 2-3 seconds seem to cause problems over time. What you cannot see here is that the first 5mm the tube walls are thick. I saw under-extrusion at 6-7mm and paused the printer in order to make sure material was properly inserted. As you can see, the tube get better for a couple of centimeters (thin walls), then it starts again. But it does not deteriorate completely to the point of not printing anymore (like under-extrusion I had with the stock feeder). Btw: Temperature is at 210°C

Looks like your GCode steps up fans with layers. That's why even if you set manually, once you hit the next layer, your code provides a new fan speed instruction. I guess it scales up with layers, let me know if it is not the case. Can you print something high ? I use the same settings, only start cooling after a few layers and progressive ( like after 5 layers, 5% fan speed and increments 5% per layer until 100%).

Well, in order to extrude the right amount of filament E-Step need to be 300. Really tested this extensively with Pronterface. But I agreed that the value difference is weird. Therefore I am wondering if my MK-? has different specs. I recall I ordered mine from a RepRap shop in Denmark, would need to check the invoice to be sure.

@EldRick: I designed a slightly longer Yoke (0.8 mm). Happy to send you the STL if you want it. The original was working, except with PET and Nylon. That's why designed it. It is pretty much the only change. I might create one that is maybe 0.4-0.5 mm longer, because the new one results in inconsistent experience with NinjaFlex or FlexPLA (the original Yoke was better). I have the spool under the table, but the filament moves fine through the feeder. But after a lot of checks, my E-Step value is definitely 300. So starting to wonder if our MK-8 are really the same :???: