Re: [ATM] Ain't got no shperometer

[Dominic-Luc Webb <dlwebb@canit.se>]

On Tue, 27 Nov 2007, ArtfulBodger wrote:

> What's the most accurate way to judge sphericity if you don't have a
> spherometer?  Is it the Sharpie test?

First, as others mentioned, the spherometer is not **that** accurate.
I am concerned the Sharpie test may not always work with the very
thin mirrors and non-glass tools we are progressing towards. Also,
others have mentioned that the spherometer won't work for the edge,
but the ring spherometer does in fact measure out rather close to
the edge.

Most of my years of ATMing never involved much money. I have always
managed without a spherometer. In the past year, I bought my first
spherometer from Ken (a 42 mm outer diameter ring spherometer),
which performed as expected, given the precision/accuracy of a
typical guage (0.002 micron or more). I would manage without, but
felt like buying a new toy, and this was my only incentive. This
is very handy during grinding to see if my surface is more or less
spherical (I monitor work on my Cass secondaries), but deviations
well below what this guage can detect can still be large enough to
result in a very bad mirror. That is why we have optical tests,
and why they are not optional replacements for a spherometer.
People here claim this is good tool to accurately measure ROC, but
I still insist optical tests are even better at this as well.

I am not sure why there would be any opposition to using these
optical tests to check the progress of grinding, as earlier stated
in this thread. Wetting the unpolished mirror will make it
reflective enough to check for drastic abberations. Likewise, if
this mirror is to be properly completed, an optical tester of
some kind will ultimately be required anyway. Foucault, Ronchi
and many related tests can all tell you if you are progressing
towards a sphere. Just looking for where a pinpoint lightsource
focuses will allow measurements of ROC. I always reckoned these
optical tests worked very well during any stage of grinding
because they are so sensitive to gross abberations.

For thinner mirrors, I have now learned my lesson. I start with a
good cell. Grinding, polishing, testing and installation into the
scope happens with the mirror and cell together as a unit.

This now connects with an earlier thread about cell design that
I thought to comment on, but didn't. I had a lot of fun with PLOP
and tried out some designs that are well outside our conventional
ATM box. "Trying" is not happening so much in software. I test the
actual mirror and cell in the assembled scope. This way, I know
what my deviations are for whatever center of gravity I choose to
test on the actual instrument. I'd like to think this is a lesson
we all learned from the Hubble telescope fiasco, but reading the
some of the posts on this list, we maybe need to remind ourselves
occasionally. Namely, the most valid test is the one performed on
the completed system. To that end, maybe more could be said
about setting up optical testers on completed scopes on mounts.

Dominic


> I'm sorry, thie is probably a FAQ.
>
> I know there are other ways -- reading the bubbles between tool and
> mirror, judging the feel of the mirror sliding on the tool -- but
> they seem to need a good bit of experience.  Seems like the Sharpie
> test is closest to objective.  Or am I wrong, and is there a more
> precise test that I don't know about?
>
> Gotta make me a spherometer, it's on The List, but my list is as long
> as yours is....
>
> So.  Sharpie, yes?
>
> Pete
>
> --
> Artful Bodger
> http://www.artfulbodger.net
> _______________________________________________
> ATM mailing list http://www.atmlist.net/
>

_______________________________________________
ATM mailing list http://www.atmlist.net/