Re: [ATM] Which diameter gets worked for a given stroke direction?

[Mike Lockwood <melockwo@uiuc.edu>]

Jay,

You've hit on a topic (astigmatism and the focault test) that's ripe 
for discussion.

Ken's advice below regarding astigmatism removal is sound and will 
definitely help the astigmatism.  However, I question whether you can 
completely eliminate it using this method, due to testing limitations. 
  There might be a significant amount remaining, even when the Focault 
test shows none.

My conservative advice:
Assuming you want the best mirror possible, go back to a sphere. 
That's the only way to make sure (without autocollimation or 
interferometric testing) that there's no astigmatism remaining.  You 
must also identify the cause of the astigmatism.  I would guess that 
it arose during figuring, but you should make sure it wasn't caused by 
uneven work support, not rotating the mirror/tool enough, or other 
factors.  Since you've got some figuring experience, you will likely 
end up with a better figure the second time!

My less conservative advice:
If you really want the mirror to be done, use Ken's method until your 
best possible testing (using multiple readings on multiple axes) shows 
absolutely no astigmatism.  Then continue figuring until the mirror is 
done.  Gentle (and proper) use of the full-size lap for figuring will 
hopefully reduce the astigmatism further.

Everyone,

Background - the test results posted lately (from round-robin testing) 
seem to show that the Focault test can't detect astigmatism down to a 
quantity that is comparable with typical P-V wavefront measurements 
for finished mirrors.  In other words, if you can measure it with the 
Focault test, then it's likely REALLY bad.

This brings up a question - what is the minimum amount of astigmatism 
(cylinder) that the Focault test can dependably detect?  What amount 
causes a detectable tilt in the knife edge shadow as it passes the 
center of the mirror?  I've never seen either of these quantified. 
(I'm hoping to spark a discussion here.  If it has been quantified, I 
apologize for wasting the bandwidth.)

There are commercial opticians that depend exclusively on the Focault 
test for their specs.  (It is only fair to point out that these 
opticians are highly experienced, and use tested and established 
techniques that generally produce excellent figures of revolution.)
However, many amateurs can't claim this.  What's the worst-case 
scenario here?  Someone on the list makes a mirror and it tests to 
0.95 strehl with the Focault test on multiple axes.  Assuming 
reasonable testing accuracy, what is the worst-case amount of 
astigmatism that can be present in the mirror?

	Mike Lockwood


Ken Hunter wrote:
> Just an educated guess here. I haven't actually
> figured out astigmatism with this process.
> 
> With a well pressed full size tool....
> 
> On the long axis. Mirror on top, no pressure but the
> weight of the glass, 30 seconds of "W" strokes
> accenting parallel to the axis.
> 
> Followed by:
> 
> On the short axis. Tool on top, no pressure but the
> weight of the tool, 30 seconds of "W" strokes
> accenting parallel to the axis.
> 
> This should blend both curves together rather than
> just digging a trench parallel to an axis.
> 
> Clean, stabilize and test to ascertain the remaining
> astigmatism (if any) and repeat, adjust if necessary.
> 
> Ken Hunter


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