Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!elroy.jpl.nasa.gov!ncar!virga.rap.ucar.edu!fhage
From: fhage@virga.rap.ucar.edu (Frank Hage)
Newsgroups: comp.robotics
Subject: Re: Sensing a known location ...
Message-ID: <11421@ncar.ucar.edu>
Date: 15 May 91 19:01:41 GMT
Sender: news@ncar.ucar.edu
Reply-To: fhage@virga.rap.ucar.edu (Frank Hage)
Organization: Research Applications Program/NCAR, Boulder, CO
Lines: 27


gcary@SRC.Honeywell.COM (Greg Cary) writes:

>What I need to do is determine when I am passing by a known location... []
>By far the most important thing is that I get repeatable results.

I recently worked on an X-RAY Crystallography goniostat where reproducible
postioning to +/- 0 steps was required (exact positioning). This is probably
overkill for your application.

It used 2 optical sensors(with slots) and associated pins; One to sense
a coarse position and one to sense the motor shaft position. AND the
signals together to detect the reference position. Approach the
reference point from one direction only, a step at a time. My software moved
towards the reference position at speed, looking for the crossing, then
decelerated, and moved back a step at a time.
The sensors were fixed, the pins were attached to moving parts.

Some other Ideas; 1. Have an adjustable physical stop, run the "stage"
against it. As you move away, start counting steps.  2. Mount an
optical sensor on the moving "stage", have a "pin" at each CD
reference position, don't worry about counting steps.

Avoid micro switches if you want repeatable results.

-Frank Hage,  fhage@ncar.ucar.edu
-National Center for Atmospheric Research