Hi Jon,

My program consists of spectral observations of B- and A-stars using ARCES. Since these are bright stars, short observations work best so as to avoid the detector going into non-linear counts (>35000); they are not all the same apparent magnitude, though, so your students will need to figure out how long to observe each star for. I typically aim for 25-30 thousand peak counts. For reference, on my last observing run, I found that a 5-minute observation on a 6th magnitude star produced a peak of around 28,000 counts. Considering that some of my stars are much brighter than this (Many stars on my target list have Arabic names), this needs to be scaled to match, and you may end up with some extremely short integration times (e.g. ~3 seconds for a 1st magnitude star). As a result I typically find that the detector readout time is a large portion of my schedule.

I've attached the full target list (~900 stars). Python tools such as astroquery.simbad can be used to pull apparent magnitude for the stars, as well as RA/Dec and proper motion, to build a TUI-compatible target list. I also like to put 30-second ThArs in roughly 1/hr.

Also, I follow a pretty non-descriptive naming pattern, mostly to avoid 3-AM confusion: all science frames are simply named "science.xxxx". The target information is stored in the header. Calibration and ThAr frames have more descriptive names ("ThAr.xxxx", "bias.xxxx", etc). I would like the students to follow this pattern as well, otherwise my analysis code will get confused.

Thanks - 
Adam J. R. W. Smith