On 2012-02-16, Stanley Schaefer wrote:
On Feb 14, 9:48*pm, "DoN. Nichols" wrote:

[ ... ]

* * * * Most encoders don't have absolute position encoding. *That
requires multiple tracks to generate binary (or more likely, gray code,
which is designed so only one bit changes at a time, so it is harder to
miss something).

* * * * Most rotary encoders on spindles have only two tracks -- one has
only one hole, to show the index point, and the other has as many holes
as needed for the resolution. *If you are counting pulses, reset to zero
every time the index passes. *Yes, I have some encoders which give
binary data, but they are a bit fragile to use on a machine tool, and in
particular, they depend on a single long filament lamp to illuminate all
the tracks and sensors. *Those lamps are made of unobtanium, so if you
design and build a tool using one, and the lamp dies, you have to re-do
the design. *(Yes, you probably could do it using LEDs instead of a
lamp, but still getting that narrow line of illumination will be tricky,
especially inside the small housing provided.

* * * * Better to have the two track encoder (index and count), and set
up a quadrature detector, so you can count the counter up or down
depending on which direction the spindle moves in at the moment.

* * * * Good Luck,
* * * * * * * * DoN.

--
An edge-lit strip of plexiglas would give you that "ribbon of light",
long as you want. Could use incandescent or LED for the light(s).

With LED giving the longer life. The question being how much
LED illumination would be needed to get the intensity the sensors want.
The long filament lamps were run at a lower voltage so their life is
longer. So probably an IR LED would work quite well there.

Enjoy,
DoN.

--
Remove oil spill source from e-mail
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---