Here's an image of the valve I'm trying to recreate. The commercial ones are MAD expensive.
So is precision electronics test gear . . . if you pay retail 
eBay is your friend! I can't imagine why you would want to DIY one of these, there are so many solutions in the life-sciences. Valco valves are the big player. Look for derelict or broken HPLC equipment. I've bought them for under $100.
If you're thinking about amateur HPLC, I'll warn you that the cost of EVERY SINGLE PIECE will make you weep. You won't believe what a single nut, connector or piece of plastic tubing will cost. Then there's the cost (and sourcing) of high purity HPLC grade water and (typically) acetonitrile. You also can't pour the waste down the drain! Hazardous waste disposal fees!
I'm definitely NOT trying to be a wet blanket, just making sure you go in eyes wide open.
Haha I'm well aware that it's not on the cheap end! I'm also not really trying to do HPLC, more like LPLC//flash/FPLC? so HPLC-ish in setup, but at much lower pressures. But you're right, the sample valve might be the one place where finding one used is unbeatable.
There are a lot of truly great scientific papers on analytical standard "DIY" hplc detectors; UV absorbance detectors and fluorescence detectors made from single PEEK cross fittings with column in/out, and optical fibers in the two other ports going to a simple UV LED/photodiode system that I want to try to mess around with.
I'm also hoping to source parts from Ebay, but everything I can avoid buying is a win. For low pressures, I've designed a very cheap fluid connection system consisting of 1/16" PTFE HPLC tubing mashed over lenghts of hypodermic tubing. This gives me both ports to epoxy into things, syringe ports, fittings etc. See attached.
I'm also planning on using non-toxic solvent systems, you can get a long way with water/ethanol or water/acetone mixtures. At this stage I'm not super concerned about solvent purity either.
Thanks for the interest!