Re: [CAD_CAM_EDM_DRO] Re: Simple DRO sensors - Cheap !

> > They also cary linear encoder strip, but making protective housings is
> > a real pain.

> Wouldn't these be the best choice for accuracy? It seems to me using data
> from a linear strip would eliminate the problems arising from
> ballscrew/leadscrew tolerances, or rather would allow the software to
> compensate for it in a way that a rotary encoder couldn't.
>
> My concern though is that "CPI" resolution... the highest was 360, does this
> mean it would only read to 1/360 = 0.00277..... inch? As even low end
> ballscrews have runouts around 0.004", it wouldn't seem to be much of an
> improvement, if that is the case.

Most of these rotary and linear encoders operate with two offset
sensors in the read head, which produce two signals 90 degree out of
phase. This is called 'quadrature'. The chip you hook them up to
counts the high and low going transition of each sensor, giving you
four discrete evenly spaced pulses for every line (sometimes called
cycle) of the encoder. Hence 360 lines per inch gives you 1440 counts
per inch, 512 cycles per revolution gives you 2048 counts per revolution.

In short, the good news is that the resolution is four times better
than you thought.

Linear encoders can be more accurate, but the resolution on the
machine is less than an equivelently spaced pattern used as a rotary
encoder on the end of a leadscrew. Don't forget that on a lathe
crossfeed the diameter resolution will be half of the radial
resolution.

I have some of the 360 line per inch stuff, but am eagerly awaiting
the 500 they claim their new system will eventually be capable of.
One can of course get much finer linear resolution encoders with
glass scales for $$$

Chris