By Don Hall
As a brand new electronics engineer in 1979, I was assigned a rush job to figure out why some circuit boards failed and some didn’t on factory testing. We were making 12″ square boards for telephone office equipment and one particular model of board, a T-1 line interface, was failing dramatically as the factory crew built up the final phone office racks. But not every time, only when a specific technician loaded the racks with cards.
Other technicians never had a failure, while the “unlucky” tech had about a 50 percent failure rate. When the card went bad, a driver transistor or resistor blew off the board with a loud crack when the card was inserted into the powered rack. Live insertion of any card into a functioning phone switch is a design requirement and considered normal.
The “unlucky” tech came to the Engineering Center (3 states away) and repeated his performance by proceeding to destroy multiple lab prototypes. At first I could not get his results, but before long, I too was able to explode TO-92 transistors at will, and got pretty good at repairing them, too.
The sequencing of the various power pins along the length of the 10″ edge connector could cause a momentary “Both On” condition in the discrete push/pull output and the stacked transistors would race to see which would short and blow up first. We called it “exceeding design limits” in the report. I noticed that the “unlucky” tech seated his boards into the card slot by pushing on the bottom front corner of the card, and most everybody else apparently pushed them in a little less askew. So he was the only one who caused the power sequence issue, but it could have happened later in the field at the worst possible time.
The initial fix was to rework the resistor in series with the push/pull chain. For the final production fix, we intentionally redesigned the circuits so that the push/pull output was never able to push and pull at the same time, no matter what order the power arrived onto the board. I learned a lot and had fun showing off my ability to destroy components at will - while get paid for it - until my own designs became capable of mayhem and then it wasn’t quite as much fun anymore.
Don Hall is a 1978 Ohio State grad who has designed telephone office electronics, worked applying energy management systems to buildings, and now is designing LED lighting and Solar power systems. Don has found more ways to ruin small components than even Murphy’s Law would predict. He appreciates others who are also willing to share their adventures and experiences.