Joe,
If money is not your concern I'd suggest two configiurations:
1) Ge monochromator + PSD - no Ka2, extremely good counting statistics,
but somewhat broadened and uncertain peak shapes
2) Geoebel mirror + LiF analyzer crystal - nice peak shapes down to very
low angles, very reliable peak positions but long counting times.
The two set-ups cover almost all powder diffraction cases AND
thin polycrystalline films. Using GSAS you will be able to analyze the
two data types at the same time.
Cheers,
Stan
