Daid

I'm no longer responsible for the firmware at Ultimaker. As I left Ultimaker 2 months ago. Didn't make a huge fuzz about it. But I left to focus more on my family. The travel time to Ultimaker was no longer acceptable for me anymore. Also, I suggested to develop a feature to import/export network settings to a text file on USB to handle all "odd" configuration options. But it never got priority. And: connman enable ethernet Won't do anything, because the network managment service that is part of the firmware will disable it again, which is linked to the printer settings network configuration. So enabling cabled connection in the firmware menu, will enable the ethernet in connman.

That shouldn't happen, a id+key should remain valid unless the machine is factory reset. So you only need it once. (On HTTP Digest level things do expire, but I would expect PHP to handle this for you seamlessly)

The 5.X firmware has about 900MB for the root partition, up from the 500MB, so that should help a lot.

You are right that F2FS does not support online resizing. This is also made more difficult by the fact that we allocated the rest of the space for Cura Connect. Updating to a 5.X release will also resize your partitions to the same size as the S5 firmware. And this update is available at the moment, so that might give you enough space. Another way would be to boot from the internal SD card slot and do the same method as with a live-linux-system. Finally, you could also opt for the dirty method, where you place for example /usr/share on the 2nd partition, and symlink /usr/share to that partition. As always, if you mess up, it's your own risk.

Als cappaciteit je grootste probleem is, dan zou ik eerder meer UM3's dan minder UM S5s aanschaffen. Software wise, zijn deze identiek voor netwerk functies enzo. Dus met de een of de andere ga je niet meer of minder ICT op je nek krijgen. Je "electrical conductive material" opmerking, nee. Dit is er nog niet. Alles wat er is kun je hoogstens voor ESD en cap-touch gebruiken. Niet om electrische schemas mee te printen, daarvoor geleid het gewoon niet goed genoeg.

Vergeet de flowsensor niet. Minder failed prints wint zichzelf ook terug afhankelijk van je toepassing.

Silly question, why are your motor cables white? The brownish we see on all board that have a long runtime. Nothing to worry about there. The burn around the C60 is strange, could have been a bad solder joint. Both C60 and C64 aren't super critical components. And your board will most likely function fine without them. But might not pass EMC emission tests.

Actually, started on this design less then a year ago, sketching it out. I have drawings for like 3 other machines as well, that I never build. Actually started ordering parts 2 months ago. I always had an interest in making my own CNC router. And I did help out with the large one in FabLab Amersfoort: http://fablabamersfoort.nl/nl/boek/portaalfrees so maybe you are confused with that one? I don't think the frame is accurate and fast enough for good 3D printing. And the Z movement is only like ~40mm. So not that good for 3D printing. Well, when we bought the house, we agreed that I have control over the garage. Which means, I have a corner in the garage that I have control over :-) still space is precious, which is where the design came from. And now, i'm discovering, shocker, CAM (aka slicing for CNC routing) software is shitty.

Hey, it has 5 3D printed parts, so I can show this off here right? ? It's my home build CNC router. With a 120x120cm cut area (if I put an extra desk in front of it) It's called LimpyCNC due the fact that it has 1 guide rail on the left side, while the right side only has a belt and rides on top of the table. So it's a-symmetrical. It uses 2 belts on both sides for accurate positioning. As well as a belt trough the X axis. All GT2. Because, as you would expect from me, I'm using quite a few Ultimaker parts. Discarded parts for various reasons. The X/Y motors are Ultimaker 2 feeder motors. With our normal pulleys on them, the Z motor is a UM2 Z motor, but shorter. The electronics are UM2 electronics (with broken heater output) The spindle mount, belt idles on the right side, and the Z nut mount are 3D printed. The plywood parts are lasercut. With all the "free" parts, I've only spend around 500 euros on this machine. I'm still tinkering with it, the Z stage has some flex in the Y direction, causing backlash effects. Just installed a fix for that, but haven't done a new cut with it yet. Firmware is 100% custom, based on Marlin's motion planner, and the UM2 display code. But nothing else, doing a lot from scratch here.

The S5 sensor is a A1335. The S5 and 3 are running on the same code base, so the older run-out code has been replaced with the newer S5 code (which works a lot better then what we had at the time of the 3)

It's on the list to have a good static-ip-configuration option, there is just no priority on it, so other things get pushed above it. I don't like it that you need to use the connman commandline way, trust me. I'm not seeing the current ssh-to-connman "solution" is actually a solution to the static IP issue. It's an ugly workaround at best.

Update on this one. We've found the problem related to abort and too much or too little priming in the next print. We're still working on the proper fix, as it's a complex interaction of motion being aborted while the rest of the system is continuing with movements to abort the print. All this causes filament positions to be out of sync and thus the large amount of priming later on.

Actually, something the config files could be causing this problem. So it would be nice to have those as well.

Nope. Linear Advance isn't a new idea. There where attempts at it back in 2014 or so. But getting it to work really properly in all situation is hard, the bowden style UM printers make it even harder. Currently, we are actually hitting performance limits in the motion controller already (something we are investigating and improving), making it more complex will not help on this area. And will most likely make high detailed prints look worse instead of better.

If there is no material present in the printer, you can just swap a printcore without using the menu. (As long as it is cold) bit of a workaround, but saves a lot of time ?

3/4 of a nozzle size is pretty much the upper limit for a layer height. So 0.3 is possible with a 0.4 nozzle. But with thicker layers, you want a larger nozzle size. 0.2mm is our highest layer height for our provided profiles for a 0.4mm nozzle.

Note that the active leveling system is very sensitive. Most likely, the system is introducing some kind of "softness" in the system that the active leveling cannot handle. Which looks for the sharp transition of not touching the bed to touching the bed. It has not been engineered with this upgrade in mind. And it was already complex enough to get working without this. (As far as I know, these upgrades are not officially supported, sorry)

Amount of filament left will be be an indication. Note that we already track this with NFC spools, but only by the amount that we think we use, and thus is off due to material slip. Also, with the end-of-filament detection of the flow sensor, why not start a print even if there isn't enough material on a spool? Use up ALL your material. Finally. Apparently, the S5 feeders current design almost fits the UM3, mechanically. There is the point where the motor needs to turn the other way around. And some part will grind against the back panel. And if you have an UM3 with flow-sensor cables, those cables are a bit on the short side for the right feeder. With a bit of work, it can fit according to our mechanical engineers. But there are currently no plans to put this out as an upgrade kit, as I said before. The general consensus seems to be against upgrade kits.

Ah, you got that wrong. ColorFabb buys their raw materials from these companies. Or in a car analogy, ColorFabb is your local car shop+repair, these companies are the car manufacturers. (Does not go up entirely, but shows the difference in operation scale) These material companies are the type where you ask for a sample of material, and they ask how many metric tons of pellets you want, not a few kg, tons. They make the plastics themselves, not the filament, the knowledge they have is huge. (Most likely the production chain is something like, Oil->Shell->DSM->ColorFabb->Ultimaker3->3D-Print, but I could be slightly wrong here)

Actually, you have that history the wrong way around. Marketing wanted BuildCore and SupportCore. R&D corrected them, and explained that the cores are not specific to build/support materials but to specific material properties. It wouldn't be strange that some exotic build material needs to be printed in a BB core (I look at you "woodfill")

Which wouldn't be called AA then, but a different letter. After all, the letters indicate the type of PrintCore.

Sounds to me like a problem area actually. PLA or PVA above 45C is problematic to catastrophic. So you wouldn't want that in a heated build environment...

We found and fixed the cause of that (the cable isn't properly clamped at the head, causing it to wiggle in the connector and lose connection), I think you can get a replacement top of the printhead, or put some thick tape in the clamping to solve that problem.

I always jam the Hex2mm screw driver into the UM3 feeder lever to push it at the office (At home I have an UMO, that always had a quick release) No guarantees on that, but it would have been silly not to have the option in the future after we identified this issue. It's a bit like the new extension port, better be prepared.

The sensor is an encoder wheel touching the filament, so mechanical. Optical simply doesn't work. If you want it to work for all filaments you need 2 different wavelengths of light, and it will get dirty. (If I remember right, black nylon almost completely absorbs IR light for example) As far as I know, the feeder wheel is almost the same, except for it no longer wearing down on carbon filled materials. Those materials will still ruin the normal PrintCores. Oh, but there is a quick-release lever on the feeder now. Instead of the tiny handle that you couldn't hold down for longer then 10 seconds without hurting your fingers.