Daid

If you print is skewed, then the most common cause is too little tension in belts, or a pulley (on the axes or motor) which isn't secured tightly. If the temperature sensor disconnects from the board on top of the printer head, then you have a major problem, as it will continue heating and your PEEK part will melt. (which smells really bad) The head shouldn't be tilted, so the 3th picture I found odd...

The next version will be a bugfix release. So nothing new I'm doing a "features" -> "bugfix" -> "features" release cycle. This forces me to concentrate on bugs first and features later.

What OS are you using? They might not turn red in MacOS (which has some less capabilities in the GUI toolkit) I've also updated the code, so this problem won't happen in the next version

Same for where I'm from, but it's a common "bug" in software. However, I think the entry box should have turned red to indicate an problem...

890 vs 1400? That sounds like the filament or the extruder is slipping somewhere, as cannot imagine you need almost twice as much rotation to get the same amount of extrusion.

Most likely you used a comma instead of a dot in the number of the "wall thickness", so 1.2 should work, 1,2 will give this error.

I have 2 good rolls from Ultimaker, red and black. (from before their supplier extruder broke down and the bad rolls of grey arrived) 1 roll of flexible black PLA from Ultimaker. The rest is all faberdashery. (And a few scraps I collected, not sure from which sources)

I think (not 100% sure) Ultimaker will replace your bad silver filament if you contact support. The whole machine depends on quality filament, if you don't have quality filament you won't get accurate prints. As for the steps_per_e, the current method is the easiest to do. But you are free to try it with a hot nozzle to see if it makes a difference.

The GCode looks good (thank god). As you can see it pulls a straight line over the window top. So it must be some plastic property that is causing this. You could try lowering the printing speed and the temperature a bit. 210C at 50mm/s for example. Do you have the model shared somewhere? and are you allowed to share it?

Odd, could be some strange Win7 64bit PyPy error, or a problem with the memory of your PC (do you have other programs that crash strangely?)

The bowden tube slides into the white piece, the top part has a ID of the bowden tube, the bottom part an ID of ~3mm. The advantage of this part is that it is ridiculously easy to upgrade current machines with it. I do not object to a new design, but there are also a lot of machines out there that could benefit from upgrades. (I was at UM HQ yesterday, and I've seen more stuff then I can remember. Also a bunch of new nozzles designs. But the reason for my visit will be announced later)

Both the long and the short belts? People tend to forget the short belts.

Nope. What you see there is about 1000 lines of OpenSCAD code.

Maybe there is something wrong with the fan, causing it to draw too much current and thus blowing the darlington?

I tried that, and my result was worse then before. Most likely you need some more belt tension. Make sure the short belts have enough tension by unscrewing the motors a bit and pushing them down. Screw them REALLY tight. For the long belts, you can use some belt tensioners. http://www.thingiverse.com/thing:19892 this one works very well for me.

That should be right. Except for the packing density, that should be 1.00 for PLA. What software version are you using? And did you upgrade firmware or not?

Most likely the "dimension" setting is not enabled, which causes no plastic to be extruded at all. You could try Cura: http://daid.github.com/Cura/ which is simpler to configure and use.

That sounds more like a problem with pypy then the actual python code. Could you try renaming the pypy directory, this should make Cura switch back to python for slicing.

I've made some progress on the design, I have designed a Z stage now, which I hope is stable enough. It was more troublesome to fit in a small area then I had expected. There are still some parts missing, like the winch and the top parts of the Z axes. Also there is still room behind the Z stage, so the Z motor and extruder will fit there. (I hope)

Cool. As the bug happens in "skirt.py", in the function "parseLine". Could you add a debug print statement there? A simple "print line" on line 285 should do it. You need proper indentation (2 tabs in front) or else it won't work and give even more errors. This will produce a shitload of output, but the final print line before the error might give some light on this problem.

For feature requests and real bugs, it's better to use the github issue tracker: https://github.com/daid/Cura/issues That makes it easier to track everything for me. As for the main feature you want. That is not going to happen. Not because I don't see the use, but because it would be to hard to add. Even the current skin option is sub-optimal IMHO, as it just duplicates the outer-wall lines, it doesn't adjust those to get the correct path for that real height. So it works fine for square boxes, but bad for slopes. The auto-baudrate is indeed causing problems in Linux. See: https://github.com/daid/Cura/issues/159 I've done some changes that should help a bit for the next release, but it requires more testing. When the project planner "hangs" on the 20 seconds, just wait. It is doing something, usually copying around some data, and collecting information. That step could use a bit more feedback, but it has not crashed. It is working behind the scenes. The "project planner homing" issue could have something to do with the printer hitting the endstops if you place the object very close to the corner. As with the new start code it will move diagonally from the corner of the machine to the first skirt position. If you use the cura printing interface, or PrintRun, you should be able to see a message in the term window then.

There are no hardware changes that I know off that causes incompatibilities, except for the 1.3 electronics. You could try the firmware from Cura RC4, I am getting some strange reports from the Cura 12.07 firmware from other people... Electronics 1.5.5 was never released. The 1.5.6 schematics are on thingiverse.

By default Cura switches on the Fan on the 2nd layer, not on the first one. This is because a hot first layer sticks better to the tape then a layer which is actively cooled. (And thus improving the success rate of prints) The layer on which the fan is enabled can be adjusted in the expert settings.

But the first few layers are printed at a lower speed, so you could be having issues at higher speeds, which can be anything from lots of slack in the small belts (so they jump over the pulleys), lose pulleys or too little power on the step motors.

If you try to install the wrong Marlin on a Mega (instead of a Mega2560) it will not work, but Cura's updater will fail before that, as it only supports the Mega2560. If you install Marlin on older electronics, it will most likely start, but fail to do anything as the pin settings are wrong. But there are very few 1.3 electronics boards out there. Most run 1.5.4 and new machines get 1.5.6 (which are compatible with each other)