How do you seperate a mixture of sand, sodium chloride, and ammonium chloride?

I would first dissolve out the sodium and ammonium chloride by adding water in excess.  That water needs to be collected and evaporated, which will give back the Na and Amm. chlorides.  Of course you now have the sand isolated.

If the ammonium chloride sublimes at 338° C then heat the chloride mixture until the ammonium chloride sublimes. (The NaCl will be left behind).  

Take that vapor and push it through water, the sublimed material will dissolve in the water, evaporate the water to get the ammonioum chloride.

Are you actually going to attempt the sublimation?  If so here's how I would do it if I only had simple lab equipment. Do separation in the fumehood.  Depending on your amount take a RB flask and fill about a third full with the mixture.  Place flask in iron ring stand or in a sand bath.  Bunsen burner probably OK for heating (Meeker burner if not); glass is good to 350 °C.  Flask has 24/40 joint to which is fitted in a vertical position a straight water-cooled condenser (actually probably won't need to cool it with cold water.  Don't use grease on connecting joint: Teflon sleeve (good to ~200 °C). To the top of the condenser attach an adapter joint and with high pressure tubing to a drying column with Drierite then to a water aspirator .  After it is all set up and aspirator has reduced pressure in flask (several minutes: what's a water aspirator good to? 10 mm of Hg?)
Gently heat evenly with bunsen burner. NH4Cl will sublime at a much lower temp than 338 °C; it will surprise you.   As soon as crystals start appearing on the cold part of the flask back down the heating.  Can gently heat crystals on sides of flask to make sure they condense on the inside of the condenser (think about a slow N2 stream??).  When no more crystals form allow to cool to RT with aspirator still on.  Stop and either dissolve the crystals of NH4Cl in water (wash bottle) or use strong wire to scrap (knock) the NH4Cl onto weighing paper (gives a nicer sample).
Now if we had the equipment in my old research lab we would do the sublimation under high vac (10^-4 atm) to a probe cooled to -78 °C (dry ice) with a short (2cm) path length from sample to probe.  We did hundreds of separations this way.
Good luck!