>
>each of the possible phase outputs.  Since you only need relative not
>absolute light measurement, you can calibrate the sensor by holding a point
>light (an LED at fixed current) various distances from it.  Each time the
>distance is doubled, the light intensity is reduced by four.


   This would work best for an LED in a diffuse package.
   Be careful to maintain the axial orientation constant.
   It's also going to tend to fall apart at the near end.
   I wouldn't depend on it any better than a few percent, and
   not for any source-tp-detector distances closer than about
   ten times the diameter of the LED.

   Reg Neale

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.