A quick poll for everyone: would anyone be upset if a board was fixed at 1/16 microstepping (it’s highest microstepping value) and not offer the other options?
I personally have never used a different value but was curious as to others take on the matter.
Honestly, it depends on the driver used and how it’s hooked up as well as what kind of CPU is controlling them.
I don’t even know what this means…
my opinion…
you are looking at this from wrong angle.
pinching few cents for this option, while competing in price is wrong direction.
providing option to change it on the fly would be more interesting pole…
so would you pay bit extra for programmable option?
reasoning, you could have multiple tools that need/benefit from different settings. that could be implemented in tool change rutine…
i’ve only ever used 16 microsteps for extruder…
It really depends on the driver.
Who’s still using hardwired drivers? I thought we’d all be on SPI interfaced ones by now. Just kidding but we should be. 
@Marko_Novak agreed. I think it makes sense to have more options.
@Neil_Darlow hopefully coming soon;-)
There’s rarely a good reason to use LESS than 1/16… basically just when the controller is hitting a step pulse frequency limit. But going higher is nice sometimes.
Higher microstepping can make steppers run smoother, but between the inaccuracies in stepper mechanical positioning and microstep electrical positioning you’re not gaining anything in accuracy or resolution. Google is your friend, several articles online.
to be clear,
that we are on same page…
higher microsteping for contactless/lightweight applications like laser…
and lower for applications like mill…
i suppose we all agree, that we could skip some options… like anything below 1/8, but not including it… since its commonly used.
what is the upper limit? 256 is kinda… accelerations can become unusable in some cases, but it shouldn’t be top option nowadays…
i recon 512 could be used with lightweight solution and laser tool.
feeding all that can be difficult, since we are kinda stuck with limited chip support…
just a wild idea, why not a board that is driven off x86 like mach compatible boards…
surely this was already discussed somewhere, but i haven’t looked for it. my point being, x86 is getting cheaper, to a point that you could get win10 tablet cheaper than most decently equipped printer boards.
by the way:
does anyone for any “open” plugins for delta mechanics in mach3?
@Marko_Novak MachineKit works on x86 and has Delta support. Needs a realtime Linux kernel to work though.
@Marko_Novak horses for courses - you can buy a nice Porsche and win some races with it - they are getting cheaper these days. but it would be very hard to use it to plough your regular farmer fields ;). point is for most embedded systems you need something that is close to real time execution - the higher you get on the cpu chain the more cpu cycles will be between your “pull that pin up” request and the actual pin changing states. that is one of the main reasons most of the hobby printers are using 8 bit cpu rather than raspberry pi which is i think 6 to 8 times cheaper than mega board.
@Ryan_Carlyle yea, this makes it fall in same category as linuxcnc…
i have a board that works with mach over plugin and Ethernet connection, spare delta frame just no motion planner 
i could make it work in different SW, but why… board has unlocked bootloader so one could make wonders, just not me 
not my cup of tea…
i guess it will sit on the shelf until something with less complicated geometry…
RPi usually has a non-realtime kernel, limiting it’s response to time critical events. Most 8 bit controllers have single purpose firmware and no OS overhead. There is firmware designed for faster processors, such as Smoothieware, that runs significantly faster on faster hardware.