Unidentified subject!

[dib@lb6.jpl.nasa.gov (David I Brown)]

I ask your patience with my questions as I'm new to the TASS
program.

Some observations:

1) A >= 100 mm dia optic will be needed to get to 16 mag with 

a CCD.

2) With a 2kx2k CCD with 15 micron pixels, you'll get 4+ 

degrees on a side with an F/4 system (now I understand the 

400 mm fl).

3) A 500 nm Airy disc will be greatly undersampled at this 

F/#. There is a conflict between field of view size and 

astrometry, and fov wins. 

Are these simple observations consistent with your thoughts? 

 
Questions:

1) What is your ideal goal for the system? 

a) 3 sets of optics looking at adjacent regions of the sky, 

each with it's own bvri filter wheel or

b) 4 co-aligned optics each with its own filter ( b or v or 

..)looking at the same region of sky simultaneously?

Is there another possibility I haven't touched on?


2) How many of the mark iv's do you want in operation, total?


Comments:

1) If you can point the optics around the sky automatically,
then you can trade the optically complex first possibility
for the simpler second to make up for total system fov. This
takes more time to cover a region of sky, of course.

2) In my view, it's best to keep it simple mechanically and 

operationally - one filter, one optical design for it. No 

moving parts of a filter wheel, no potential focus problems. 

You've got mount problems to handle, but that's no mystery.

3)  The more mark iv's you make, the more you can justify a 

different design tailored to each filter (which can be much 

cheaper in glass).

Practical comments:

1) I've had great luck with imaging quality of off-the-shelf 

Hoya filter glass. The peak-to-valley  wavefront aberration 

across a 1.5" dia of several 2mm thick filter glass samples 

was 1/20 wave (HeNe). I'm not joking. I had it measured 

interferometrically and confirmed in practice. There are Hoya 

and Schott glass equivalents.

2) If you go with an independent optical design for each 

filter bandpass,
it may be possible to tailor some parts of the overall design 

to some off-the-shelf elements saving a lot.  Or it may be 

possible to design to standard test plates used by mfgrs in 

making optics. This might help. Of course, performance/cost 

has to be the final criterion. I'm sure Elliot is well aware 
of this.