Re: *** Powerbook From Hell

On Tue, 30 May 2006 20:54:43 +0000 (UTC), "Mike Andrews"
<mikea@xxxxxxxxxxxx> wrote:
Jasper Janssen <jasper@xxxxxxxxxxxx> wrote:

Was it really calibration? The story I've heard is that the optical test
device was assembled with one of the lens elements backwards.

The figure tests (Foucault knife-edge or some such) were done with a

Yeah, I'd heard Foucault testing as well. Which kind of gave me pause --
they figured the Hubble with the same method (if a slight bit more
precise) as amateur mirror grinders use for their small (by comaprison, at
least) mirrors, I mean.. just wow.

gauge rod inserted backwards. Apparently the rod had a thick bit in
the middle, and both ends were turned to a smaller diameter, but not
for the same distance: one end was supposed to be inserted in the
gauge facility, and the other was supposed to be a precise distance
from the turned shoulder on the first end. Thus, schematically:


A --------------====---------- B

But ends A and B were the _SAME_ diameter, and it was but the work of
a moment for Murphy to turn the rod around -- especially since it was
unmarked and you had to _know_ which end went into the fixture.

Aha. Thanks, that makes more sense than assembling a lens element wrongly
-- the latter sounds like something that'd be noticed (although a subtle
correction plate.. who knows).

The real fsckup was that the tools necessary for measuring the fsckup
in the mirror figure were in the same building, but in use by a NRO
project and therefore required clearances to go in and use them then.
The bureaucratic hoo-ha in getting the clearances, or in covering up
the classified NRO hardware, was judged too much when someone asked to
do it, and so the Hubble mirror flew as-is, unmeasured, and fscked-up.

Yay for no backups. You'd think at the very least an enterprising soul
would have trundled it outside and set up a diagonal for a star test. Who
could resist *that* temptation...

The root of the whole thing was that ends A and B were the same dia.,
and both could be inserted into the gauge fixture. That's at least as
bad as LockHead handing JPL stuff in English units when JPL expected
(and thought) they were metric.

I'd say the root was that they were both the same *and* the rod was still
asymmetric -- if they'd just made a rod of uniform diameter (easier to do,
too!) there wouldn't have been a problem. I'm beginning to suspect now
that that was originally the intention, and then the fuckup fairy visited
when they found they couldn't get the required diameter rod in the
required material and they ended up having to turn down an existing piece
of metal -- and they took a shortcut on that.

Jasper
.

Loading