I'm no expert but I've sat in front of a newer working FACS before. I remember hearing a big part of running them is the sheath buffer, PBS I think it was, gets used once then goes to waste. The lasers for fluorescence can get dim and/or die out altogether. There are likely high-voltage PMTs inside, probably three for each laser color (it may only be a single color machine, but google results show it might come standard with more)... you get back-scatter forward-scattter and side-scatter data, these data channels can be placed on an XY cartesian plane and graphed with false-color representing number of events. So if you get a lot of particles/cells fluorescing the same in side and forward-scatter, there will be a bright spot on the XY plot, and you can draw a conclusion that it is a single type of particle responsible. If you had mixed 2 cell types and you wanted to separate them, you would need to play with your assay (i.e. what dyes you're using) so you'd see two bright spots on the XY plot. Then you basically just draw a circle around the cell group you want to keep, and any events that fall within your circle will get collected, the rest will go to waste.
If it has the FACSworkstation software for running the machine, you could be up and running in a few days. You could probably use distilled water instead of PBS if you're just using calibration plastic microspheres (
bangslabs.com).
Worst case is you get a bunch of cool stuff to hack on! I see at least one or two tektronix oscilloscopes!