Just one more comment on the issue of light sensitivity.  I had heard of a
method of using exposed 35mm film, and then ran across an entry in Electronic
Design December 2 issue on the method.  Dave Johnson from Colorado writes
that you can expose a roll of 35mm film (preferably Kodacolor 100 ASA) to
fluorescent light for five seconds, and ask the developer to run the roll
without making prints.  The graph of response shown in the article shows high
transmisson up to about 700nM, then a sharp rise in blocking that peaks at
880 and is still high well into 1100.  He adds a caveat (sp?).  The film
filter cannot be used near moisture or where it can be scratched.  I might
add the danger of heat as well, though it would have to be mighy hot.  But
for the lab, you could fabricate quite a nice cocoon for your chip.  If the
film is not conductive, you might even be able to roll the whole chip.  (with
leads protruding through..)

Second comment.  I recently commissioned a new board with the 16C73 on it.
 Had huge troubles uncovering a mistake in board layout, where all of the
analog inputs had a 570mV offset on them.  (IE no reading above 00 until
0.57V).  I measured every pin and thought all of them were connected.  Many
hours later and some words that I cannot repeat, I found that the ground pin
8 was not connected.  In operation, it measured 17mV.  As I often do when
measuring, I assumed that it was grounded.  Arizona confirmed that this pin
is the one of the two grounds that is used heavily by the analog circuit.
 Just remember.  17mV mistake on one end causes 570mV error on the other.

Diddos from Pittsburgh.
Chris Eddy, PE