I'm still interested in knowing what the limits of the printer are ... Now I'm interested in knowing what is the minimum layer height I can print ...

Given that it is a Delta printer, what do I need?

- Engine type -> Nema 17 of 1.8 degrees.
- Movement of the cars in the towers -> By means of belts G.T.2
- Microstepping used -> 16 in Duet

More considerations ....:

- Being a Delta, I suppose that this "resolution" of the movement on the Z axis, will also be valid for X and Y
- If for example, when you find the limit, it is 0.15 mm (Theoretical). What would happen if...:
- I select a lower layer height in the slicer, for example 0.10 mm
- Movements that are not multiples of 0.15 mm, in the previous case, the second layer that would be 0.1 mm x 2 = 0.20 mm, which would also not be multiple of 0.15 mm (Theoretical)
- A third doubt, similar, as would be the printed layers, if when i "slice" the model, I select a layer height of 0.1618 mm? or 0.20 mm

I guess the optimum layer height that I should select in the slicer depends on this. To select a multiplo of the minimum movement that I can do in Z, of the resolution in Z. No?

]]>If you request a layer height/extrusion width combination that requires an amount of filament less than a single microstep of your extruder, you're gonna have underextrusion (or overextrusion if you're using absolute positioning of the extruder).

]]>For Cartesian printers, the calculation is quite simple (in your details you did not mention a pulley number of teeth, though usually 16 or 20, used to drive the belt):

For 1.8degrees per full step, micro-stepped 16x using a GT2 (pitch of 2mm) and a pulley of 20 teeth, we get 80 steps/mm. Inverting this, we get a 0.0125mm/logical step. Therefore, in this case this is the theoretical resolution of the printer.

Now since your printer is a Delta, things get very complicated. In the center of your build plate, you will get extremely high resolution, but the further you go out from the center, the lower the resolution will be. On large Delta printers, with not enough micro-stepping, straight lines close to the edge may appear wavy.

The calculations for a Delta's resolution is simply too complicated to explain here, but luckily there is a tool, which can help you with it: https://configurator.reprapfirmware.org/

Simply start a new custom configuration, and enter the values requested. On the Motors page, use the "Calculate steps per mm" to calculate them - it should be noted that this tool calculates the number of steps needed to move the belt by 1mm, and that does not translate to 1mm of movement of the head.

That said, luckily the Duet can be micro-stepped up to 256x, allowing you to reduce artifacts as mentioned above, if they appear. Also due to the fact that each micro step only contributes a very small amount, the actual resolution remains quite high (many steps per mm of actual head movement), so you will be able to use either 0.1mm, 0.15mm or 0.2mm (or practically any other).

For the delta, yes, the X, Y and Z will have about the same resolutions (though there may be exception positions: thinking about Z which may actually stay the same as long as all towers move the same distance).

If you select a layer height that is not a multiple of the resolution, the printer will typically round this, by going as close to it as it can (may be just over or under). Since resolution is high, the difference should not really be noticeable.

Lets take the 80 steps/mm value, which gave us 12.5microns per step:

- Printing at 0.1mm layer height: it will take 8 steps to raise the nozzle by 0.1mm.
- Printing at 0.15mm layer height: it will take 12 steps
- Printing at 0.2mm layer height: it will take 16 steps
- Printing at 0.1618mm layer height: it will be approximated to 0.1625mm and take 13 steps. It should be noted here that the commanded layer height is slightly higher than requested (0.4326% higher) and therefore, at some point the commanded layer height will be 12 steps, to make up for the higher than requested rounding (though this will be many 100's of layers later).

Typically this rounding does not cause issues - and if you do need the extra resolution, you can increase the micro-stepping.

]]>