This is another story with a nebulous date. I am pretty sure it happened in early 1987, but not sure when.
The sonar imaging application that Scott wrote was amazing…it was way ahead of anything else that was available. But there are some processes that can suck up a lot of computer processing time when done in software, but could be really fast if done in hardware. Examples of this were the analog to digital conversion of the sonar signal, and the slant-range-correction that had to be applied to the returning sonar pings.
Scott understood this and one day he walked in with an enormous stack of manuals…I think they were from Intel and Texas Instruments. He spent a couple days looking for information about the specific chips that would be useful in these processes, and a few others, and then he set down and designed the circuit to perform this magic.
You can get a good overview of the history of integrated circuits here.
When going from a circuit design to a printed circuit board the first step is to create a prototype. Scott did this on on a board similar to this one. This allows you to test the board without committing a lot of money to designing and fabricating the PC board.
Once he was sure the circuit would do what he wanted, he hired someone to design the circuit board. I forget this persons name, but will try to track him down and get a story from him.
Within a year the computer Scott was using for surveys looked like a standard PC clone from the outside, but the motherboard on the inside had about eight long sockets for optional plug-in boards, and Scott had designed and built what were basically full-blown computers to go in each one. This really jacked up the processing power of the unit, and the number of input signals that it could handle.