I made some prototype circuit boards. The scan and digitizer boards were designed
to be the final product, more or less, and the display boards were throwaway samples made on rigid FR-4.
They all worked reasonably well. I had to do a good bit of fiddling around in VHDL to make the
digital signal processing do the right stuff.
I then committed myself to making flex boards. I got some quotes, and selected a small company in Southern California called Circuits Unlimited. I worked closely with the owner, Mario, to ensure that I was designing something that could actually be built in quantity. He had some trial boards made and assembled. They worked, but needed more stiffeners added. That was accomplished, and the final set of boards arrived in mid-June.
The flex cables I chose, made by Parlex, turned out to be more springy than I'd like. The Molex ones are less springy, but I didn't realize at the time how much that would matter. I thought about making a final version of the scan board that would be much smaller and have the flex cables built in. I even got the board design done and quoted on. Unfortunately, it was out of my price range for now. I may well go that route for the next version, as it would make a much sleeker product.
I still had to figure out the batteries. I talked with people at the National Underwater Robotics Challenge, and encountered one fellow who told me that I could get lithium-polymer, or LiPo, batteries at a fraction of the cost I was used to, from a radio-control hobby shop in Washington State called HobbyKing. I checked it out, and got their battery packs. Each one weighs a pound, fits in my pants pocket, and makes enough power to run half of the Video Coat for an hour.
I also bought a video camera at Radio Shack and attached it to a bright LED headlamp, to have a way to let the viewers see themselves on TV.