First, the Red and Blue linear dispersion equations (caveat emptor, observers will still need to obtain their own Argon dispersion equations - the ones below are just for planning observations):
"row" = the row number on the detector, as seen by DS9 (reduce by 1 for IDL)
"slit" = the row number of the center of the slit, as seen in a direct (no grism) image (also for DS9).
Red_Lambda = 628.20 nm + 0.1414 nm/row * (row - slit)
Blue_Lambda = 435.92 nm + 0.1057 nm/row * (row - slit)
Thus, a MOS slit positioned at row 860 will have wavelength limits of
506.8nm at row 1 and 798.69 nm at row 2064.
Further, I recommend keeping the MOS row 'zone' to row 860 +/- 180 rows.
This biases things a bit away from the optical axis center (at row
2032), but puts more of the optical spectra onto the detector.
So, I've been designing for slits within a 2.3 arcmin tall zone, centered
at row 860, of width 10 arcmin.
The slits I've used so far (and that seem OK) are 9 arcsec wide (i.e., EW)
by 2 arcsec tall (NS - the dispersion direction). I've also tried to put in
at least 4 arcsec of separation between slits along EW, resulting in a
slit 'pitch' (center to center separation) of no closer than 13 arcsec.
(PRISM plate scale is about 0.3923 "/pix)
I've used the digital POSS in DS9 to locate star center coordinates (with
Mimir overlays to verify) for the slits and to pick 4 alignment stars for
each MOS plate. These alignment stars are near the ends of the 10 arcmin
wide MOS swath and have holes with 10 arcsec diameters. Be sure to
pick fairly bright alignment stars in the optical (not the NIR, as I mistakenly
did the first time!).
An example from DS9 is attached. Note that it is not shown centered on
row 860 or turned this into an AUTOCAD drawing for the shop just yet.
