[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [ATM] PlOp PlOp Fizz Fizz What does PlOp think?
How much does a Plop-designed 9-point cell of the conventional type differ
from the good old equal area design that is reported to work well in real
life? Let's study what PLOP itself says about it!
I have used as example a 317.5 mm mirror, 25.4 mm thick default pyrex, f/5 -
obstruction ignored. With the conventional cell with relative radii 0.4094,
0,8189, PLOP will calculate the RMS error=3.95 nm - this design does not
change focus at all, as it happens (or at least only about half a wavelength
of blue light ;-).
Letting PLOP optimize with no refocus, it comes up with rel radii 0.3403 and
0.8443 and the RMS deformation 3.33 nm - the inner radius a trifle smaller,
the outer a trifle larger, but you would have a hard time to tell the
difference in real life, I'm sure, by any practical method. Even allowing
refocus (by 0.028 mm), Plop says 3.04 nm error - still more than 75% of the
old-fashioned cell.
However, if you allow varying loads (on the inner 3 points), what happens?
Still without allowing refocus, the inner rel radius shrinks to only 0.2172
with the force 0.421 of the outer points' - the outer radius 0.7691 and the
RMS error 2.51 nm. A bit better yet but not dramatically. Another run, this
time allowing refocus, and the RMS error goes down to 1.70 nm (0.035 mm
refocus between horizon and zenith - can you live with that?).
What about a 6-point cell being better than a 9 point one?
Try first a 0.7071 rel radius: at a fixed focus, the error is 7.57 nm - not
so fantastic.
Refocus by 0.008 mm, and the error goes down to 4.73 - still not the best
choice. Optimize for no refocus: radius a tad smaller, 0.6873, error 4.23
nm. But allow refocus: Now error goes down to 2.00 nm, but you need to
refocus by 0.045 mm.
So, the 6-point cell can be an improvement over the 9-point equal load
cell - but only if allowed refocus, and if so, the unequal load 9 point one
is even a little better.
Well, should one believe PLOP?
Nils Olof
_______________________________________________
ATM mailing list http://www.atmlist.net/