Re: [CAD_CAM_EDM_DRO] PIC DRO stupid question

Wayne Bengtsson wrote:

> >Really, if someone can come up with accurate grating patterns on film
> or
> >glass, it is fairly
> >easy to make your own linear encoder. Does anyone have any ideas on
> >where to get
> >such grating patterns?
>
> One quick and easy way to make such a pattern would be to draw the
> pattern
> in a good CAD program, and print the pattern on to transparency
> material
> using the best quality injet printer you have access to.

Yes, everybody who hasn't tried this sort of stuff thinks
resolution=accuracy, but
that is not true. Anything that moves paper or transparency film
between rollers has
inherent errors of significant magnitude in it. The ink jet tends to
wobble and
splatter, and probably has accuracy no better than .01" or so, even
though the
resolution is a LOT higher. My standard argument to anyone who proposes
laser
prints, ink jet prints, pen plots, etc. is to make two identical prints
of a regular
grid, turn them 180 degrees with respect to each other, and overlay
them.
The errors will be glaringly obvious.

>
> I am assuming a resolution of 300 DPI would translate into .003"
> accuracy,
> at a minimum. If anyone here has actually tested the output of a
> printer for
> accuracy, I would be very interested to see the numbers.

Well, since I now have a total error of less than .002" on my CNC
machine,
over the entire volume of travel it has, this sounds pretty bad. My
.002" or so
includes not only the encoders, but all the backlash, long-term errors,
etc.
All these errors just keep adding (and sometimes multiplying) and so the

encoder grating needs to be a LOT more accurate than anything else.

Jon