Talk:Microstepping with optical feedback
why are you sampling inputs during a timer interrupt when you could use interrupt-on-change or even the counter subsystem? - Triffid Hunter 03:57, 4 May 2010
Using mouse sensors
Have you considered try this setup with a mouse optical sensor mounted on the carriage measuring the surface of the axis rod? the ADNS2051 sensor has quadrature pins (like most of these days optical mouse sensors) and should act just like the linear encoder. Would be interesting to compare that with the printer-hacked-linear encoder to check how much precision a mouse sensor could get!! ADNS2051 have a maximum of 800DPI, which should give a response of 31.75 microns per measurement. Later revisions can go up to 8000DPI, or 3.175 microns (which is just crazy!!) Hradec 02:55, 27 October 2011 (UTC)
Optical mouse sensors
I have suggested on one of the forums that one could use a more modern optical camera mouse to do two axis feedback using a similar system. Look at the bottom of the build platform and have two random axis geometries that move it in some semi orthogonal way and use the position feedback to establish where you want to go. Using two camera mice one could have incremental signals from one for speed and acceleration processing and then print a pseudo random non repeating pattern on the bottom of the bed and have another camera mouse (in diagnostic mode taking pictures) establish the absolute position of the bed and correct for any missed steps or such. It would possibly allow for the use of two very cheap consumer motors and spectra line drive with capstan pulleys (that may slip but with feedback who cares) and a flexible linkeage system that does not require bearings. I posted a scetch and called it the bamboo printer some time back but cannot find it just now, I will make a page on RepRap Org wiki to document it soon so others can use the ideas. KalleP 06:46, 15 August 2013 (UTC)
Extending to a three phase system
This could be extended to drive three phase brush-less model aeroplane motors like we often find in quad copters and some hobby CNC router spindle motors. Having a feedback for position on a lathe spindle allows for simple thread cutting as the angular position is known. Having a feedback on a small mill allows for thread tapping with synchronised spindle and z-axis moves.
It has 3 wires so could still use the same driver circuitry.
It has many fewer steps (poles) per revolution but this should not cause much difference.
KalleP 06:47, 15 August 2013 (UTC)