ATM Hubble (was NOT "A fairly good mirror, I think")

[Anthony Stillman <atmer@flash.net>]

>>...the Hubble telescope
>>...Perkin-Elmer

>...star tested, only *after* launch.
>...290nm +- 5nm of low order spherical aberration.
>...20nm rms of residual higher order


It should also be noted that in addition to 30th wave zones and a half wave
of spherical aberration, the orbited primary's outer four inches are
painted black.   Yep, that's right, they turned the edge.  The continuation
of this "flat" edge into the un-masked portion of the mirror is the source
of the residual spherical aberration.(1)

Also, technically, this mirror was not manufactured by Perkin-Elmer.
Perkin-Elmer incorporated a subsidiary to do the work.  Immediately
following the shipment of the mirror (literally as it was leaving the
loading dock) the company was de-incorporated.(1)

In many ways it is truly unfortunate that the reflective Offner null test
used to characterize the primary was erroneously re-assembled.  And, that
it was preferentially believed.  However, the silver linings should also be
mentioned.   Namely, both the analysis of residual aberrations in an
optical system based solely on defocused point source images and the
development of post acquisition correction techniques were greatly evolved.
Tinsley Lab's reputation rose by leaps and bounds when it was chosen to
manufacturer the correction optics.  And, professors of engineering and
management gained another excellent example of how not to engineer and
manage.


Anthony


1) Some questions were not meant to be asked.  Besides, the turned edge
puts a human spin on an extraordinarily complex achievement.  And, lets not
forget, building something is much harder than tearing it down.