A lifetime designing PCBs: Job shopper to service bureau
This is the second article in a series about my career experiences in the PCB industry. The first article focused on the early days as I became a designer.
Doing PCB layout and quality assurance (QA) from 1966 to 1978 as a contract designer was a great experience. Working at numerous companies and seeing many different types of designs provided me with a broad awareness of not only the different design technologies, but also how to solve problems in diverse ways.
But it wasn't all fun and games. There was one situation that pushed me to the brink. One night I received a call asking me to help with an emergency. The company had a critical design and the layout designer had failed to complete it. I was asked to review the design, fix the errors, and complete it in less than a week. This sounded like a reasonable challenge…at first.
When I arrived the next morning, the layout and tape-up were on a light table. It didn't look too difficult: there was one very large dual in-line package (DIP), a bunch of smaller DIPs, two connectors, and quite a few discretes. Looking more closely, I saw that the red/blue pencil layout wasn't complete. The designer had started taping the routes without finishing the layout. Warning lights went off in my head. I decided that the first thing I should do was initiate a process. Check and correct the layout against the schematic, then complete the layout, check and correct the tape-up against the layout, finish the tape-up, and finally, check the whole artwork.
I kept getting a feeling that something might be wrong, but at times it seemed that everything was wrong. I was working all day and most of the night, even trying to sleep on the office floor so I could save the commute time.
It was two days before the deadline when a thought popped into my head: "Are the component footprints correct?" I found that the spacing on the large DIP was wrong – a major problem. This added to the stress of the situation, which had already been increasing with each new problem I found.
I started ripping up the routes and moving the pads to the correct locations. It was tedious and I had to fight exhaustion. Then, something very strange happened: The pads and traces on the top and bottom layers were drifting out of alignment with each other. I wondered if I was hallucinating. Even the vias that I had placed earlier appeared to move off the grid. It took a few minutes to fully understand the fatal problem.
The top layer was taped on Mylar, the bottom layer on Acetate. Because Acetate expands as the temperature rises, it should not be used for doing tape-ups. Compared to Acetate, Mylar doesn’t expand at all. When turning on the light table, it heats up and the Acetate begins to stretch. The pads on the Acetate layer would never be aligned with the pads on the Mylar layer. Once the full impact of the problem was realized, I didn't know what to do, and we agreed it was unreasonable to think I could finish the project in time, so I resigned. This project had tested the limits of my capabilities and served as a valuable learning experience.
In 1978, I started a service bureau called Sunshine Design. At first, I was the only employee – working out of my basement. With a drafting table, light table, and boxes of Bishop black-tape supplies, I was confident that it could be a successful endeavor. Since I didn't have a Diazo machine to expose the sepia copies of my artwork and drawings (Figure 1), I mounted each layer of the artwork on top of the sepia paper and secured them on large pieces of cardboard. Then I took the boards into the backyard and exposed them to the sun. Voilà!
Figure 1 Sepia print of PCB artwork
The business was doing well and within a year, Johnnie Travi and Mike Chaney joined the team as we worked out of a rented office in a shopping plaza. The company name was changed to Electrum Design Services. Eventually our careers took each of us in a different direction. However, when I look back at what we accomplished, I am grateful for their dedication, creativity and friendship.
We kept busy because we had a reputation of doing quality work very quickly. The key to keeping customers satisfied was always doing very thorough QA. When we switched to red and blue tape, the PCB design process became a bit easier. Unlike Rubylith, the coloring was on the bottom side of the tape, so there were never any scratches. It also signaled the end of curved traces. With the black tape, an arc could be created by using the X-Acto knife to scrunch together the inside edge of the tape and carefully bend it into an arc. Experienced designers could create concentric arcs. Some designers preferred arcs over angles. But the age of manual design and tape-up had already started to decline.
We were excited when we heard about companies like Racal Redac and SCICARDS that were selling minicomputer eCAD systems for PCB design. Although the cost was still outrageous, especially for a small business, it was significantly less than the systems that required a mainframe computer.
In 1980, everything changed at my service bureau. I bought two used Redac Mini systems, changed the company name to Computer Circuits, hired more designers working multiple shifts, and lost all of my customers.
But that’s a story for the next blog.
— Charles Pfeil is a Senior Product Manager at Altium, working on definition of their products with a primary focus on routing tools.