In 1970, after four years as a technician in the Air Force, I went to work for GenRad, testing, troubleshooting, repairing, and calibrating new scientific instruments, including portable sound-level meters. One day I got a meter that other techs had been unable to fix.
It worked fine except that randomly, every few seconds, the analog meter would briefly kick up to full scale, making a tiny "bonk" sound when it hit the mechanical stop. The circuit was all transistorized, (no IC's) mostly direct-coupled, but organized into three blocks with capacitive coupling between them, all powered by an 18V NiCd battery pack.
I disconnected the front end block, with no change. I disconnected the middle block, still no change. I probed around with the scope, and found that each time the meter jumped, there was a clean rectangular negative 50V spike, one microsecond wide, throughout the meter driver, and even on the 18V power bus. The meter driver included six transistors, a diode bridge, a fast-slow meter switch, and the meter, all connected by short lengths of RG-178 coax.
I theorized the only thing capable of creating the spike was the inductive meter movement, so I replaced the meter, to no avail. I methodically disconnected and reconnected each end of one coax after another, and found that when the coax going from the switch to the left side of the diode bridge was disconnected from the switch, the pulses went away. But when it was reconnected to the switch and disconnected from the meter, the pulses returned.
I replaced it with a new piece of coax, and the pulses were gone. Just to make sure I wasn't imagining things, I put the old coax back in and the pulses returned. I put the new wire back in, and the pulses were gone again. I see this as a classic example of "Bagg's Law," which states, "An illogical problem requires an illogical solution."
I still can't explain how a three-inch length of RG-178 can generate beautiful precise repeatable negative 50V pulses, but I know what I saw. I brought the coax to the engineering department, but they never really believed my story, and probably tossed it in the trash. It still bothers me that I didn't send it off to a research lab for analysis, or perhaps to the Smithsonian for display.
Do you know what might have gone wrong? Tell us in the comments section below.
Tell us your experience in solving a knotty engineering problem. Send stories to Jennifer Campbell for Sherlock Ohms, and read more cases here.
Chuck Bagg recently retired after 50 years in electronics, mostly in analog circuit design. He holds three patents assigned to major corporations, has had several technical articles published, has presented a scientific paper at an international conference, and passed the Patent Bar exam a few years ago to become a registered patent agent. He is also a member of Mensa, the high-IQ society.
Design & Manufacturing Canada, the largest advanced design and manufacturing trade show serving Canada, will take place in Toronto, June 16-18, 2015. Learn more here.