Re: Mounting the Gecko 3X0's near the servo

--- In CAD_CAM_EDM_DRO@y..., Jon Elson <elson@p...> wrote:

> vrsculptor@h... wrote:
>
> > I've suddenly developed encoder noise problems on the Y axis of my
> > mill. I think I know what the problem is but this got me thinking
> > about relocating the Gecko out by the drive but I wasn't

convinced it

> > would accomplish anything. From an interference perspective

wouldn't

> > noise affect the step signal just as much as it affects the

encoder

> > signal? Any thoughts, recomendations or experiences?
>
> No, the step signals are essentially current loop, and may be quite

a bit

> more noise resistant than the encoder signals. The opto-isolators
> are quite good at cancelling voltage noise.

One could then, in theory, make similar current loop encoder line
drivers. All that is needed is a logic family with good current sink
capability (ie, printer ports were sometimes just octal buffers).
Optoisolators at the receiver end.

Any idea if this would be more or less noise immune than differential
voltage signalling?

And if one is going to optoisolate anyway, why not put a couple feet
of fiber optic capble between the LED and the phototransistor, and
skip the antennas (I mean wires) entirely? I guess one reason might
be that fiber could be more likely to break... and we heard about
what happens then a few days ago.

Chris