Re: Work List

[<aah@nofs.navy.mil>]

Regarding work list:
  1) you need to work on the "slop" on the declination
     drive as well.  Just leaving the motor on does not
     help if the axis has a degree or more of dead zone.
     
  3) overscan.  My message from July 22, 1998 spelled
     out my wishes -- you didn't keep it around? :-)
     Here is the ideal setup for each serial line:
        16 prescan pixels
         8 covered pixels
      2032 active, exposed to light, pixels
         8 covered pixels
        32 overscan pixels
     for a total of 2096 pixels/line.  In real life,
     this is overkill, especially for overscan if your
     RC constant drops off reasonably quickly.  I thought the
     covered pixels would be more valuable than they have been;
     I'm still suspicious they are not being clocked correctly.
     So lets use the following scheme:
        read and drop 16 pixels (prescan)
        read 2040 pixels and save (covered + active region)
        read and drop 16 pixels (covered plus some overscan)
        read 8 overscan pixels and save
     for a total of 2048 pixels saved out of 2080 converted.

  6) Our machinist made up a cadkey drawing of the part
     he machined for the lens shades.  Basically he took a
     piece of 1" aluminum and cut a ring out with inner diameter
     equal to the lens assembly, and outer diameter equal to
     a piece of 6" galvanized vent pipe.  He then went and milled
     the aluminum ring to reduce the weight.  Looks nice, but
     overkill.  I'll try to get the drawing into a reasonable
     file format when I get back from Europe and post a picture
     as well.

 17) If the Stamp has a reset line, you might bring it back to the
     host computer so that you can reset the Stamp without having
     to run outside and cycle the power.  Often you can cheat and
     use one of the serial cable handshake lines for this feature.
     
You've covered everything pretty thoroughly for now.  There
will be more after I've used the camera a few more weeks!
Arne