Re: [CAD_CAM_EDM_DRO] PIC DRO _now_ i've made it a stupid question

Where is our optics/grating expert when we need him?

In the meantime, let me add a couple of comments.

There is no need to get a very small light source or detector! A rather large one is fine.

The goal is to line up two identical gratings, one stationary, one moving close to each other. The lines need to be accurately
parallel to each other. With light shining form one side and being detected on the other side. Ideally, when they are offset
one line with, no light gets through and when they line up, 50% light gets through. The detector looks at the total average
light that passes through.

Try it with two identical pocket combs and you see what I mean.

That is why changing encoder wheel lines needs to match the detector grating and why it is not a good idea to take a detector
from a rotary encoder and try to use it on a linear encoder.

For the quadrature decoders, each should be adjusted for best alignment (largest analog signal swing on a scope) and then the 90
degree phase adjusted while watching that the "lines" alignment is not lost.
Bertho Boman
Vinland Corporation
====================================
Jon Elson wrote:

>
> Yes, precisely! You could either take the sensors and logic out of an HP encoder, and
> just use the finer grating, or make an entire assembly and PC board. You put the
> 'analyzer' grating on a slant with respect to the long one, and adjust it until you get the
> right quadrature relationship on the two signals. Without exotic collimating optics, the
> higher the frequency of the grating, the closer the gratings have to be to each other.
>
> Jon