So I looked up why recommended layer heights are typically “odd,” or not “rounded,” which are both poor terms to describe it. For example why 0.08mm vs. 0.1mm, or .24 instead of .25mm. Because when designing a part, I would typically use more rounded dimension “steps” like 1mm, .25mm, .1mm, etc. (unless of course the requirements call for something finer). Or I’d tend to make something, say, 200mm tall vs. 199.92 or 200.16mm (multiples of 0.24 layer height).
The answers I found seem to indicate that desired layer heights are based on the resolution of the Z axis stepper motors (and geometry of the lead screws, presumably). And that 0.04mm is some kind of “common” step, so layer heights tend to be multiples of this number. Which makes sense (though I’m not sure if that’s the whole story or not).
Is that the correct reason?
And does this magic 0.04mm step/resolution number apply to our H2 printers? I don’t see a spec for that anywhere. I can only assume since all the provided H2 print profiles use those layer heights also. But you know what they say about assuming.
I know I can get the slicer to give me precise total heights by adjusting the final layers, but that’s not the question. Coming from a different design world I just found those layer heights kinda odd. Also from my tests the printer still does an excellent job at “rounded” heights like .25mm, but I’m looking for as much consistency as I can get, and playing within the printer’s limits makes sense.
Modern 3D printers have finer resolution steppers and Z screws, stiffer frames, and better extrusion control that lets them achieve lower layer heights than printers of old. But mostly, slicer defaults are still all using the values imposed by those earlier machine architectures.
I don’t know what the minimum layer heights for our BBL machines are, though I suspect it’s published someplace. But I also suspect that the machines are capable of layer heights that are so low as to not be practical to use.
Thanks, that’s kinda what I was wondering… does it still apply? It’d be simpler to change layer height to rounded values instead of calculating each Z detail of a model in steps of .04mm. (Though with careful parameterization this can be mitigated.)
If Z resolution/accuracy is publishes somewhere, it’s very elusive indeed! Even got the Google AI to admit it didn’t know, though at first it tried to feed me some BS… LOL. It’s quick to remind me that the advertised XY accuracy with vision encoder plate is 0.05mm. I’d think something like Z resolution could be a distinguishing feature between printers, so this is somewhat surprising (I guess they don’t want to commit).
Just to clarify, I’m not looking for the thinnest possible layer, just what the steps might be. I realize the measurement is probably the same, but yes of course at some point it would become too thin to be practical as a single layer (like I’m guessing 0.4mm layer height would be, though I’ve not tried).
Too many variables to say without at least some hard data. Most steppers are 1.8º/step. 200 steps/rev. But modern drive circuits allow “microstepping” that gets that down by 1/2 or 1/4 or even 1/8. Does BBL do microstepping? I’d guess “yes”, but to what degree? No idea. They might even go lower than 1/8. That doesn’t improve stepping accuracy but it does smooth/quiet the stepper in operation. And then, we don’t know the pitch of the lead screw assembly, or if the screws are “geared” (stepper’s drive pulley is larger or smaller than the screw pulleys)… and also, there’s mechanical lash and stiction that’ll mean below some step size, the motor might move but the mechanics might not.
I’ve read that the H2C can get to 0.08mm layer height. But the nozzle diameter is a contributing factor. Best practice (don’t ask me to explain why) dictates a layer height that is no less than 25% of the nozzle diameter, so with a 0.4mm nozzle, 0.1mm is the practical limit.
Most of the reasons are already mentioned here. One thing I wanted to add here.
The steppers in your printer are basically constantly in motion while you are printing. The automatic bed leveling requires the height of the bed to be constantly changing to make sure that you have a consistent layer height across the bed.
Bed tramming helps to minimize that motion, but no bed is really a 100% flat at any given temperature.
If you like to use a 0.1 or 0.25mm layer height, you are free to choose them. It might require some tuning to get it optimized.
Still the rule of thumb is that 50% of the nozzle diameter seems to be optimal.
So maybe you need to get your hands on a 0.5mm nozzle
But that would require a lot of tuning because there is no optimized profile for any material.
I have some fine metric drill bits for my rotary tool… what could go wrong?
Good point about the bed moving anyway for leveling. And I can’t say I’ve noticed any real quality difference with, for example, .25mm height on either a .4 or .6 nozzle, vs. the preset choices. But I do tend to get a slight improvement in Z dimension accuracy, for example with smaller holes on vertical walls (when designing with “round” numbers).
A lot has been explained and reasoned already so I’d only like to add that the stepper motor resolution in combination with the z-axis spindle pitch (I think it is T8-4 but may be wrong) leads to multiples of 0.08mm as the preffered incrememt.
However, 0.08 is already a multiple. I tried to see how low we could go on an X1 last year Minimum Layer Height Limbo: How low can we go?
Nowadays, I frequently run both my X1 and H2C with 0.04mm minimum adaptive layer height.
To be honest, I do not know exactly.
With a step translating to 0.04mm, both 0.2 and 0.24 are multiple integers of that.
0.2 however is an uneven integer, so every 2nd layer would be out of phase while 0.24 would be in-phase.
To me, it does sound plausible although the effect should only really show if there’s something wrong with the lead screws, bushings or heat bed mounting.
Myself, I tend to stick to the defaults or go lower. I will probably replace the 0.2 with a 0.16mm “default” layer height and try to pay closer attention.
I do not think that it is primarily the mechanics. It is more your observation regarding microsteps.
From my limited understanding, the system does not actually know where it is within a given microstep. So I think that it is not so surprising that I got reliably good results down to 0.04mm.
I consider my X1 safe down to 0.03mm, but below that, it is hit and miss. However, the key problem I encountered were the very low flow values.
Let’s see if and when I have a chance to properly throw the problem at my H2C. With a different extruder, maybe the z-axis will show as the limiting factor.
& 
that post on your layer height experiments!