Just had the idea - has anyone seen anything around someone using igaging beams to serve as positioning aids for a homebrew pseudoservo? I figure if you have enough geeks with arduinos, someone's bound to try to make something like that...
Not even remotely appropriate. They are low resolution, and VERY slow. Besides, what's the point, when real, high-resolution quadrature encoders are available for under $30?
I think the advantage begin that you can use a true position feedback of the table rather than motor position like the quadrature encoders.
People have done it using the Kflop and LinuxCNC systems, but as Ray said, they are extremely slow for any kind of use on a CNC. I'm not sure if anyone has tried using the glass scales though. My guess is that they are still too slow when you rapid.
Glass scales is exactly how it IS done on commercial machines - a rotary encoder on the motor provides velocity feedback, while glass scales provide position feedback, using a dial-loop servo. They are not at all slow, being simply optical linear quadrature encoders.
No, it's not just rapids. Those things are very low resolution - +/-0.0005", which sucks for positioning. They are painfully slow, which means the servo never knows where the axis IS, but only where it WAS some time in the (relatively distant) past. Both dynamic response, at any speed, and final position accuracy, would be very poor indeed. At rapid speeds, it would truly suck, big time.