Dim3nsioneer

Sure there is. Use a brim made from PVA.

I had good results with 10mm for the join distance. Default value for the horizontal expansion is in most cases not so bad but sometimes it does itself create islands on the outside of a print. Then I set it to 0.

Have you checked the model in the x-ray view? I would not be surprised if you would find some model flaws indicated in red there...

Why do the supports fail? Are they too small and break off? In this case play with the settings "support horizontal expansion" and "support join distance" until you get a solid block of support.

Maybe someone from @Ultimaker can shed some light on the glass question?

Ja, das ist der Grund. Am besten mit dem Fingernagel oder einer Messingbürste reinigen wenn es zuviel Kruste dran hat.

The obvious one that PFA is transparent.

Ohne es selbst getestet zu haben würde ich sagen nein. Dafür gibt es zwei Gründe: einerseits ist ja anders als beim PVA mit der Luftfeuchtigkeit kein Lösungsmittel in der Luft vorhanden; andererseits ist die Zersetzungstemperatur mit 350-400°C doch deutlich weiter weg als beim PVA wo sie nicht allzuviel oberhalb der Drucktemperatur liegt.

Depends where the filament is broken. If it's broken inside the Bowden tube, simply remove the Bowden tube on the print head. If it is broken inside the feeder, open the feeder (check https://ultimaker.com/en/resources/50859-clean-the-feeders for this). If it is broken inside the print head or even inside the print core the only way is to push it through. For this, move the print head to a corner, remove the Bowden tube on the print head, heat the core up and push from the top with some filament. If it is completely stuck, take a thin needle and push from the bottom (be careful not to burn your fingers as liquid plastic could suddenly come out of the nozzle. If this doesn't work as well, I recommend to replace the core. The origin of the clog could be a lot of things including a small piece of dust picked up with the filament.

Wenn ich das richtig verstanden habe, ist die Länge der kürzeren E3D Düsen das Problem. Abhilfe bietet ein anderer Fanduct, z.B. dieser hier: https://www.youmagine.com/designs/umo-fanduct-for-3dsolex-and-e3d-nozzles

Wenn man in den Cura Einstellungen die Option des automatischen Herunterfallens auf die Druckplatte deaktiviert kann man mit dem Verschiebe-Werkzeug das Objekt in die Luft setzen. Ist dann das Stützmaterial aktiviert wird der Bauraum unter dem Objekt mit Stützstruktur gefüllt.

While I understand your frustration about not being able to restart a failed print at a specific height there are good arguments against making those routines available as changeable gcode. First, it's not all gcode. On both UM3 and S5 those routines are written in Python. Most commands do actually send a line of gcode to the motion controller. But in between them there are commands which use different routines and functionalities not on the motion controller but on the Olimex board. You also have to keep in mind that those two models are not designed for makers but for professionals. Their main request is to have an automated start-print-procedure which cannot be accidentally changed. If you want to do extrusion tests, I recommend to build a test station maybe based on a printcore with some simple electronics or even a standard printer control board. You may get some ideas from a project to use a printcore on an Ultimaker 2+ by @ultiarjan: As for the restart of a failed print: This does not happen on a S5 due to the filament sensor. Running out and printing air is UM3 specific. But please be aware that such a runout does not happen at once but starts with an underextrusion which is getting worse. Depending on the size of the print this can go over multiple layers. Restarting such a failed print might result in a very bad print quality around the layer it started to run out. Finding the right layer might also be a bit too tricky for most users. However, a restart after a short power cut would be a great feature (hint, hint) but I guess it's not what you meant. We know that it is quite unlikely that it will be possible to upgrade an UM3 with the filament sensor from the S5. But there is a community project driven by @foehnsturm going on which puts a filament sensor to an UM2(+): As you can see, bringing it to the UM3 is an option but requires that Ultimaker releases the source code of the UM3.

Mit dem BB core gehst Du sicher bezüglich Verstopfungen ein kleineres Risiko ein als mit dem AA core. Es könnte aber zu Tropfen während der Standbyphase kommen.

Die frühen Hardcores waren nichts anderes als umgebaute UM-Printcores und sehen/sahen damit auf den ersten Blick nahezu identisch aus. Der Feeder des S5 hat eine Spezialbeschichtung auf der Feeder-Welle, welche tatsächlich verhindert, dass abrasives Material die Welle übermässig abnutzt. Mit einem Hardcore und einer Everlast-Düse zusammen hat man da also ein Abrasiv-Paket. Das geht übrigens auch für den UM3; dort benötigst Du einfach noch einen Bondtech DDG-Extruder-Upgrade-Kit dazu; dann kannst Du dort auch CF, GF & Co. verarbeiten.

The standard tube is PFA. PTFE is milky white. If that does not bother you, you can use also a 1/4"-1/8" PTFE tubing which is usually about half the price of PFA.

Maybe you want to check a Bondtech...

Dann verrate ich hier doch mal ein Geheimnis: es sind die gleichen Printcores... ?

until then...

Most of the usually not displayed parts are not used very often. So the effort for making them presentable on the webshop might be a bit too big, especially for tiny parts in the low Euro or even Cent regime.

Das scheint alles seine Richtigkeit zu haben. Tough PLA wird vermutlich mit dem AA 0.25 nicht so schön und Breakaway kann gemäss https://ultimaker.com/en/resources/52121-how-to-print-with-ultimaker-breakaway nur mit dem 0.4mm AA Core gedruckt werden.

This picture show a possible solution. You can see the narrow column which represents the object to be printed. The larger box on top is set as a mesh which modifies settings. Those settings are specified by per object settings on the left. Assuming you start printing with a 0.4mm nozzle and 20% infill (settings on the right) you can change to a line width corresponding to a 0.25mm nozzle and 50% infill setting inside this modification mesh. The only thing left to do is to stop the printer before it starts the part inside the modification mesh and this can be done with the PauseAtHeight postproc script (box), stopping the printer at the height in question. When that happened you can use "change material" on your printer and reload the same filament but exchange the nozzle during the procedure. I hope this helps.

That one is now called ChangeAtZ and is a postprocessing script.

@foehnsturm: First test shows: - sensor works - flow is around 95% during print - flow variation is pretty large (90-100%) - some correlation between flow rate and infill quality can be seen - retractions influence the flow rate for a very short time (almost not visible by eye)

There might be a small difference in the hotend cooling fan logic depending on how old the Ultimaker 2 is. On older UM2 the fan runs continuously while at some point Ultimaker updated the design of the electronics and included a logic circuit which makes that fan run only when the hotend is hot. Every UM2+ had this right from the start.

No problem, I meant no offense. It's probably the material choice which decides in this case whether or not the design has to be adjusted.