Re: ATM Wave Length Foucault Data! (long)

[mjc5 <mjc5@psu.edu>]

Thomas A Simmons wrote:
> 
> Hi Peter,
> 
>   I agree with 99% of your message, but not about leaving it a sphere.
> I'm bad with remembering names, but didn't we talk about this some
> time ago? The bottom line was that leaving the mirror a sphere doesn't
> teach (me) the patience and skills that will be necessary to create
> larger and faster mirrors.

	Patience is pretty hard to teach! But you'll get it by the time you're
done making a mirror. ;^)
	
>   What perplexes me is how the slope can come out almost perfect,
> and yet the inner portion was still rough. I spent a long time
> inspecting this mirror last night, and (1) the inner 1/2 IS irregular,
> but not noticeable with the mask. (2) when nulling on the inner portion,
> the waviness (ripple?)is small, but noticeable. (3) the brighter LED
> appears to be the factor that revealed it (hmmm, what about that other
> mirror?). (4) When I substituded the green LED on the tester, shadows
> were difficult to discern. This is the one I wrote earlier just wasn't
> bright enough. (5) My tester has too much slop!! (6) digital camera to
> record, communicate errors?

	What I think we have here is a combination of problems.

	First is that the mirror isn't really completely polished. If it's wavy
it needs more work. Granted, you could have well polished waves, but
still the mirror isn't done.

	You need to get a good sphere and be sure the surface is "perfect"
before proceeding to put a mask on the mirror. It really is possible to
make a mirror that tests well, but has problems. The mask can hide a lot
of sins.

	You need to settle on one type of lighting situation. LED's are kind of
cool, but if you use a lot of different types, you can just add to the
confusion. Also, since many people sand the top off their LED's, I
wonder how consistent the resulting images they get are. It isn't chic,
but there is something to be said for the old bulb illuminating through
a nicely cut hole and diffuser.

>   So the bottom line is that there remains work to be done on this mirror.
> Smoothing the inner portion of the mirror shouldn't be any different than
> a thoughtful application of a figuring session. If I overshoot and make a
> hole, I guess I'd be "chasing the plane" as Mel posted some time ago.
> I also want to experiment further with brighter LEDs of other colors.

	What you might want to do, and this is just a suggestion, is to take
the mirror back to a sphere instead of trying to fix the problem
directly from where you are now. The sphere will show every defect
glaringly, especially things like ripple and roughness. After you have
the proper sphere, then you can parabolize, and unless you have a gross
figuring problem, the mirror will remain smooth when you reach the parabola.

	- Mike -