Daid

That looks like a proper alternative. I do not know the build quality, but it says 3mm, which is the same size as our thermocouple. Could be a big money save on shipping to get those. EDIT: He seems to have some more "Ultimaker knockoff" products. Small note of warning for his teflon coupler, I know the dimensions of this one are very critical, a difference of 0.1mm can cause plugs instead of resolving plugs.

70mm/s shouldn't be a problem, but extrusion issues could also be caused by too little or to much tension on the spring at the feeder. It should be about 11mm long when filament is inserted.

All of them :-) I just print random stuff to test my printer, settings, and software tweaks (then again, I develop the software, so doing lots of different things is important for me) Actually, as a 3D modeler, what would be important to learn is that "manifold meshes" are important, as well as "no intersecting faces". As for the kit, the first 80-90% of the print quality comes from properly tweaking the mechanics and proper filament, the last 20-30% comes from tweaking the settings. (On some complex this can be different, but for most prints this is the case) First thing you will most likely need to adjust is the short-belt tension. It usually goes wrong there for kit builds quality wise. As the printer works fine with lose belts, but won't achieve super quality.

It's a customized part, and it shames me a bit to tell you that you can only buy it from Ultimaker, who buys it in large bulk. Send an email to support, they can help you out.

Your example isn't that bad, I've seen much worse. It comes from the Skeinforge code which is still behind Cura. Which does some strange things and as with all odd code, things sometimes go wrong. I'm working on replacing Skeinforge, which is quite a tough job, but can be done.

Not sure why the defaults are low, but it's important to never skip a step on it, and the platform full with PLA, can be quite heavy. And mass is slow.

You did nothing wrong, the server is breaking down. (quite literal)

The linux has a Cura.py script instead of the cura.sh script

The only "true" way to check if the pulleys are properly secured is by putting a mark on them and on the axes, and see if they do not move independently. The pulleys can feel secured but could be a tiny bit lose, as you only have a tiny bit of slanting.

Usually this is a backlash problem caused by slack in the short belts. Which can be solved by unscrewing the X/Y motors a bit, pushing them down, and screwing them down tightly. EDIT: Also, I'm thinking about adding a disclaimer, "Every time someone suggests the infill not touching the walls problem is caused by software I will kill a puppy." Because if you print with normal PLA (flex is different) and your machine doesn't have backlash, then the walls will touch the infill fine. Increasing the infill-overlap is NOT the correct way to counter this. It's a mechanical problem with a mechanical fix.

I'm pretty much in charge of the LCD code right now. But because of internal Ultimaker priorities I'm delaying work on it. (I also have some internal test versions on my laptop that I need to merge and it's becoming a bit of a mess) Cura has my priority right now. So all LCD related pull requests will be dealt with, but not right now. Also, a few requests are hanging on "waiting for response from submitter".

The leak doesn't seem that serious. You could extrude some ABS (if you have it) this usually plugs it. Else use some tools to slightly tight it while the hotend is hot, this usually also seals the leak. (Do while HOT, else you'll break it!)

Nice job! Looks like you are feeding the PLA from the wrong side of the roll

In hindsight, the 2 cables are not the best solution. A single ribbon cable would have been easier (and most certainly just as available, as we used tons of 20 wire cables on my previous job)

You cannot wreck the extruder wheel unless you use a hammer. The motor is not strong enough. And the firmware refuses to extrude if the head temperature is below 170C.

This happens if the 2 band cables are connected the wrong way around.

Actually, most people forgot this already. It's quite normal for a crisp clean hotend to smoke a tiny bit the first time. I think it's some kind of oil or coating that's burning up. It's only a tiny amount of smoke, but it even catches me off-guard when it happens and I see it. The "burned up PEEK" which correctly Joergen warns you for, is quite nasty. And one of the more common failures. But if you got trough the initial checkup wizard, this is much less likely to happen. If you want to be sure, you could run that wizard again to check. There are quite a few things in place trying to prevent the PEEK burnout, but it isn't 100% fool proof. (a disconnection of the sensor board while the machine is hot is the biggest issue, as the machine will keep reading the "old" temperature from before the disconnection) A simple tip, insert PLA into your machine. I see your tube is still empty. The reason for having PLA in your machine is simple, PLA will start to smell quite nasty before you reach "PEEK destroying temperatures". You can react on this nasty smell and save your hotend. (It's really a last resort, and I hope you never have to. But it has saved a few hotends in the past)

Seriously? You are basing an "often" claim on a sample size of 2 with 1 being bad? I know your customer experience was bad. But generalizing your problem as if it happens a lot is just bull. I've never seen a bad bolt in real life, and I've only seen 1 on photos. And I've seen a lot of printers, with all kinds of problems. Joergen helped out at the 3DEA store, so he knows what breaks down when you have 12 printers running none-stop. So our sample size for knowing what usually goes wrong is a lot bigger. The thermocouple comment from Joergen is right, you better add one of those to be sure, they are hard to source as they are not sold as spare part trough the normal shop and damage quite easy. So see if customer support can add one of those to your order. (The horse-clips have been replaced by better quality ones, also solving the problem with chewing up the bowden tube)

PEEK and the brass tube are most common to break (during assembly or a temperature sensor connection failure) I've never seen the alu-plate being damaged. Electronics can be damaged, but they are quite expensive to have "on stock". Spare stepper driver might be nice to have, as sometimes they break (but not very often)

Small note, the extruder upgrade should fit fine, if you use the correct number of washers. Which is most likely why it didn't fit for you. I've installed the upgrade on a Rev1 machine (older then yours) and there it fits fine.

Sorry? The swapped endstops are not a bug in Cura, and are actually the reason I added the test in the first place. To avoid head slamming into the sides of the machine.

Oh, I could have broken something then...

Everything except the UltiController is included in the DIY kit. The assembled machines can be bought from the shop (they are in their own category). Note that the assembled have a longer lead-time (due to high demand) we are working on reducing this. The assembled machine includes the Ulticontroller and all upgrades. Properly calibrated and tested, including "proof of proper printing" left on the machine. Only thing you would need to do is calibrate the bed level again (as this gets moved due to shipping) Dual-extrusion might be released as a highly experimental (no warranty) kit. We're not sure yet. I've build 3 dual-extrusion machines so far, and 2 of those have major problems (one of those so major that I removed the 2nd extruder). And I never got the printing results that are in the "good to great" category (I do not think any other producer gets great quality multi-color prints, as I see little to none multi-color photos online)

It uses the profile from the normal mode in the main window. Yes, this isn't clear at all, and bound to change in the May release.

The slowness comes from the 3D GCode view. It's doing a crazy amount of slow math. Switching the view (top right button) to normal object view will stop the GCode render and thus makes it fast again (older Cura versions used to continue the GCode render in the background, no matter what) The "mirror turns object black" is something I also noticed, it's some combination between mirroring and resetting which causes this. It doesn't do any harm, and I'm working on fixing the cases that it goes wrong. Moving the objects in the project planner requires "manual placement" to be enabled in the project planner. This can be done in the project planner settings. In the normal interface moving is not possible. (I'm working on merging the project planner and the normal interface for the release in May, which will be a big release and a lot of work)