(Warning: Very long boring description of archaic printed circuit board process follows)
My first two jobs while in high school covered most aspects of printed circuit board production, from the initial schematic to the finished boards. The bosses knew each other and had arranged for me to work in both places on alternating days, so I could learn the business.
My original job was as a draftsman, working with electronic circuit drafting. I normally worked on drawings, with pencils and triangles, occasionally getting criticized by my boss for my sloppy lettering.
Once these drawings were finished and approved by the customer, we started preparing for the physical design of the circuit boards.
This very smart fellow in the back of the room would stare at a complex circuit schematic for … a long time, and then he would take out a pair of colored pencils, in red and blue, and begin drawing the layout of the circuit on a fresh sheet of vellum. The red lines represented traces on one side of the circuit board and the blue traces were for the other side. His job was to lay out all of the chips and components in such a way as to minimize the length of traces and the number of “feed through” connections (connections from front to back through a little hole) that were needed.
The original pencil drawing was then passed on to an underling like me. We would put the drawing on a light table and lay a sheet of clear acetate on top. We would then use an x-acto knife and extremely thin pin striping tape, in red and blue, to lay out the traces on the acetate.
The print shop would use red and blue filters to photograph only the top or only the bottom of the board. Those photographs would then go to the other company where I worked, where we would create the actual printed circuit boards.
We would take the artwork from the print shop and create a working positive print from that. Then we would apply a thin layer of photosensitive adhesive film on each side of the copper-clad boards and expose each one to bright light masked by the working positive print.
Wherever the light touched the special plastic, the stuff hardened. This meant that the background stuff (the parts we didn’t want) were hardened, while the plastic covering actual traces was left unexposed. A quick solvent wash removed the plastic covering the traces.
We then ran the boards through a plating line, electroplating a thin layer of solder to the exposed copper, exactly where the traces would be. A quick acid bath then dissolved the previously protected areas, leaving only the copper traces with a thin layer of solder protecting them.
A trip through an oven melted the solder, to allow it to flow into all of the microscopic cracks, forming a shiny silvery surface on all of the traces.
At this point, a skilled worker would take one board and go into a funky darkened booth that had a special magnifier setup that would allow him to precisely center each solder pad over a nifty little drill that would come up from below at the press of a pedal. He would carefully drill each of the many hundreds of holes in this board, positioning each one by hand.
This master board was then given to one of the two burly women who ran the Quad Drill machines. They would put the master board in a special tray and then clamp four stacks of undrilled boards under four pantograph-driven drills. The single handgrip had a tiny pin at the bottom and a button. They would insert the pin in each hole of the master board and hit the button, causing the quad drills to come down, drilling the same hole in thirty or so new boards.
Once drilled, a special router template was created for the new boards, following the machine drawings, and someone like me would first rough-trim the boards with a sheet metal shear, and then use an industrial router table to trim to the final dimensions. That machine spewed huge quantities of fiberglass dust, in spite of the vacuum hose, and consequently killed multiple portable tape players I had taken to work—even though I sealed them in plastic bags.
I would then break the boards’ sharp edges gently by swiping all sides against a big sheet of sandpaper.
The final step of the process involved two silkscreen passes, both using masks created photographically with artwork from the same print shop. The first pass would cover everything that was not a solder pad with translucent green lacquer. The second pass would apply all of the text, such as component names and the company trademark, in white lacquer.
To close the loop on this entire process, my drafting boss had me work on a short production run of assembly (about 50 boards). I assembled and hand-soldered fifty or so six-inch-square boards. That’s where I learned to put the low components in first and how to avoid “cold solder joints”
All of the stages of this process have been supplanted by computers. The circuit design, the drafting, the component layout, print shop work, etcetera.
And much of the mechanical work is now done via CNC machines–computer-driven machinery that does what those tough-looking women did all day long, faster and more accurately.
