When I was working for a large military contractor that manufactured high-powered radar systems, I was part of a design team that produced the frequency-generation hardware, which consisted of a 19-inch rack about eight inches high. There were many sub-modules within the rack, and most of the voltages were less that 30V DC. One exception was the use of 120V AC for some secondary circuits.
One day I received a call from the production technician. Apparently, the frequency generator was not working correctly. The plant was located on the property of a municipal airport. It was built during the World War II, when access to the airport was essential. When I arrived at the tech’s test station, he told me that the frequency generator was creating an oscillating burst at the rate of two cycles per second.
That was odd for a frequency generator that normally had internal frequencies in the range of 1,000MHz to 4,000 MHz. When I examined it, I could see that the generator was working normally except for the RF burst that occurred every two seconds. I thought I knew what the problem was, but to make sure, I needed to probe an internal circuit that was next to the 120V AC unit. I asked the tech if the 120V AC had an on/off switch that would allow us to shut it down during the test. He told me that it did. He turned the switch on the test panel and assured me that it was now in the "off" position.
I took a look, just to make sure. I could see that it was indeed "off." I proceeded to perform the necessary probing of the circuit. When my finger touched the terminal, which had the "off" 120V AC, I received a nasty shock. I asked the tech as to why the voltage was still "on." He replied, "Oh, the switch is broken. It doesn't work."
I then insisted I would do no further work on his problem until the switch was replaced. I told him that the cause of the two-second burst of oscillations was due to the airport’s radar sweeping across the test floor and upsetting the frequency generator, which had all of the shielding covers removed for the testing. Moral: never trust a switch for power removal. Test with a reliable tool to make sure the circuit is "dead," and never trust someone else to "assure" you that all is OK!
This entry was submitted by Lee R. Watkins and edited by Rob Spiegel.
Lee R. Watkins is a senior member of IEEE: P.E. He has a BSEE and has worked for Westinghouse, Honeywell, Martin Marietta, Motorola, and ON Semiconductor, and is currently a self-employed consultant. Lee has published several articles, as well as a book on analog filter theory and design.
Tell us your experience in solving a knotty engineering problem. Send stories to Rob Spiegel for Sherlock Ohms.