Shaft diameter and length is also part of it.

NEMA is just a mechanical aspect and contains zero information about the performance. Equally comparing motors by torque isn’t the full picture as the inductance (and other aspects) can be different. Equally important is the stepper-driver & supply voltage.

Biggest concern for me is creep over time.

Small CNC-mill. Previous 1.85mm NEMA23 wasn’t enough.

Not bad either.

Previously I had a 1.85Nm on this axis which wasn’t enough to reliably pull 1’500 mm^2/s acceleration milling metal. All sort of issues from overheating stepper driver to it losing steps.

I see. Size doesn’t matter to you. It is all about skills.

Let’s me introduce you to the 400W NEMA24 servo.

edit: Sorry. Was only the 400W unit. Somewhere there should be a 750W speciment.

Also be careful with these motor. While a small NEMA17 found on 3D-printer has 0.5 Nm and under normal conditions will stall before serious harm happens this NEMA34 is 9Nm which is enough to break bones. Those 400W servos are equally dangerous. While only around 1Nm they have roughly 4Nm peak and keep this torque at high RPMs.

If you have to mix it use mechanical interlocking parts. Like a PLA cavity that is filled with PETG. The PLA & PETG won’t connect/bond but due to the mechanical design, it will still hold together.

Get a BambuLab P1S without AMS.

I don’t want to be bogged down with having to run proprietary slicers through Wine and things like that. I am not sure how big of an issue that is with e.g. Bambu or Creality (if at all), but I’ve seen enough rug-pulls and enshittification processes that I don’t really want to risk that.

You can use PrusaSlicer, OrcaSlicer, Cura or some other slicer with them. They run perfectly fine in LAN-mode removing the Bamub Lab server stuff/question from the equation.

While not every replacement parts are also made by third party companies things like nozzles, PEI-sheets or fans are available from third parties.

As future-proof as I can possibly hope for.

I would argue money in the bank account is far more future-proof than any 3D printer can be and Bambu Lab costs a fraction of what a Prusa core one costs. So when the future arrives use the saved money to buy a next-gen printer.

Locally produced (I’m EU based).

Bambu Labs are well made so not a safety hazard but I can understand this point.

The Prusa core one looks very promising but at the same time, it isn’t for everybody and the general consumer is likely better of by buying the P1S. For the price of 1 Prusa One you could buy 2 Bambu Lab P1S and 10-20kg of good PETG.

regarding Support: Difficult question. Prusa has excellent support but the last experience I had with their printers before dropping them wasn’t that great and was riddled by issues/bugs. BambuLab on the other have the it just works magic but the support needs to improve. You send them a bunch of log files as requested. They probably only look at the oldest file (that might be months old) and provide a “wrong” reply based on that as a solution without looking at the text you wrote.

or the A1 mini if 18x18x(?)cm^3 is large enough.

For the material: Go with PETG for the coasters.

Samples.

Order each color as 10-20g sample.

SELECT

'Tequila' AS Ingredient0,

'Lime Juice' AS Ingredient1,

'Cointreau' AS Ingredient2,

'Crushed Ice' AS Ingredient 3

FROM

liquor_cabinet

It’s fine.

Not great but won’t cause issues. Might have a look at the configuration to check that the matching thermocouple type is selected/configured.