danilius

I was thinking along those lines. I fully appreciate that this is non-trivial, but it is nonetheless jolly useful. This would of course could break Marlin in a big way, but perhaps not so drastically. As it is, the whole gcode thing is a bit long in the tooth for 3D printing, but assuming that we want to stick with it, one could add some very simple features into the firmware to handle this. For example, bracketing each layer of each object with a start/end marker in comments as @ahoeben suggests. Standard firmware will simply ignore these comments, updated firmware will understand what they represent and could, upon receiving a custom gcode, simply skip to the next object on that layer. Of course each object will need label, which could be set and/or autogenerated in Cura. Maybe this is something for @Tinkergnome to play with?

No idea if anyone has suggested this already, but I would love to have some way to be able to top printing one or more objects in a multi-object print if they have failed somehow during the print.

Has anyone got Cura 2.1 working with Octoprint, and if yes, care to share the details?

It looks pretty and amazing as well as pretty amazing, but as you can see in the "use cases" section, it's an all-round compromise. It does everything OK. Which is fine, really. If you are after precision or quality surface finish then you will have a hard time with that machine, though. I need both; precision and surface finish. So far I have not found a 3D printer that has made me think, "Oh I wish I had bought that one instead of the Ultimaker". The day I do think that that, I will buy it.

Brilliant win - by far the most outstanding entry. Superb function with great form without re-inventing the wheel. Or fork, in this case. Oh, and a nice choice of colours, although you should go all the way and get some nicer-looking rubber bands This is *exactly* what desktop printing makes sense for. Fulfil a need that is not being serviced properly by existing offerings, and make it oh-so-much-better with some careful thought.

Seriously cool!

Closed build chamber is not all that necessary for ABS. I buy really cheap ABS off ebay, and print some large things (+15 hours prints, longest one 36 hours). In order to reduce warping, I add some geometry to hold things in place, like mouse ears and bridges, and use 20 line brims . This reduces warping considerably if not entirely mitigating it. The greatest importance, for me anyway, is dual-head printing for supports. This will really be a game changer for me, because then I can print the kind of parts I want to design. With ABS I get near to injection quality at 0.1mm layer heights. The next really important upgrade would be a more powerful processor that would allow for faster printing. The current one struggles with circles at speeds above 60mm/sec. Wifi or ethernet would be nice :-)

OK, an official entry. A tool for measuring e-steps easily. [print=3552][/print]

Wurl, this does not technically count as an entry, but a solution to something that was really bugging me, namely the bowden tube distorting horribly. In order to calculate how many rubber bands I needed and what tension would be required, I used a complex method with essentially involves grabbing several at random and agonising over the colour choice. Works a charm.

The results your tests produce appear to indicate that there is a condition in which the pressure gradually builds up and then is suddenly released. Increasing the temperature alleviates this condition by reducing the viscosity of the plastic. I had this issue and was always printing PLA at 210C for 0.1mm layers at 30mm/sec. Just recently I bought a "pimp" 40 watt heater from 3Dsolex and what a difference that has made! It leaves me to believe that the previous heater was woefully underpowered, and no doubt Ultimaker genuinely believed that it was a 25 watt heater, but it was probably a lot less than that.

Just in case you need sheer numbers to convince you: the only technology that can pull this off at the moment is powder printing - aka SLS. Trying to do this using FFF/FDM (same thing, just means filament-based printing) is as close to impossible as you can possibly imagine, or certainly so painful it's not even worth bothering. In fact UM should highlight these images as canonical "what you really don't want to try printing with FDM technology" examples. Until we get dual head printing, of course, then you are on to a winner.

Cool, there has to be a badge for dumping that much g-code.

Of course allowing IGES and STL files as the very lowest level is great, and provided the IGES files are converted into STL for those users who only model with polygons, it should be a reasonable base solution. But it's only a partial solution, because I frequently find that I need the original files to make rapid changes without too much of a headache. For example, it might be easier to modify a sketch rather than try to add some more geometry on to something or reverse-engineer a shape. Likewise in Blender I make heavy usage of booleans which makes updating dimensions and other changes much easier and will want all the meshes used for the boolean operations present. There is no perfect solution, I am simply advocated for something that will stand the test of time. FreeCAD and Blender are open source, hence it will always be possible to open any files generated with them provided the older sources are made available or are backed up somewhere.

Regarding tools, if you plan on using onshape or Autodesk you are reducing the available pool of contributors or their respective terms and conditions changing suddenly making the software unaffordable to the tinkerers. If you used FreeCAD and Blender as the modelling tools, you have the best of both worlds (traditional CAD and polygonal modelling), no risk of them becoming unaffordable, and they both run Windows, Mac and Linux. I use both of them in my daily work for designing products and prototypes. While Blender is far more mature and has an insane amount of modelling tools, FreeCAD is growing by leaps and bounds and is eminently useful.

Did you have to fiddle around with the limit switch at all? There isn't all that much clearance under the platform that I can see on my UM2. If you are shy of opening the case, then an epoxy glue like araldite will do the trick. Just clamp the beams in place while drying or wrap some painters tape around it it all until dry.

@cloakfiend, where do you get colorfabb filament that cheaply? Does it involve propitiating demons and the sacrifice of maidens of virtue?

Will do. Thanks.

I don't suppose anyone could direct me me to the beta.....

I have designed my own extruder that will not fit into the available space, so will have to mount it externally. I want to test it without removing the extruder currently installed, so want to buy a separate motor to test it with. So, what spec motor do I need to buy for this? Secondly, can I connect it to the electronics without having to unplug the existing motor, which will allow me to run tests with my new design and then get back to printing with the existing one without having to constantly unplug the things?

I should have made myself clearer, apologies for that. Does the new extruder have more torque, and does that mean one can push more plastic through at a given temperature, assuming that only the extruder has been replaced and not the installed sensor? @Labern, what exactly were you printing at 130C?

Maybe the platform could be raised using something like this (sketchfab is doing something weird with the animation, but you get the idea), and two or more could be linked together as units.

Oh, if you insist. It's ABSolutely making me fume.

Does the feeder have more torque to enable printing at a lower temperature?

Aside from the gear on the new feeder motor, and that it has a lower resolution than the older one, what else is different? Is it more powerful than the old one?