At Reuter Organ Co. we built real pipe organs. My position was director of production and senior engineer. I had a problem once that I had to solve from 600 miles away. As a musician, I can use my musical skills to analyze various mechanical sounds scientifically. As a licensed amateur radio operator, my electronic skills help to make some problems relatively easy to solve. This particular problem was not so easy.
I received a call from our area representative who told me that the large pipe organ console we had just built and installed was blowing fuses in the three large DC power supplies. The fuses would blow during only about 5 percent of the power-up cycles and all three supplies would blow at the same time.
These supplies delivered power to three independently separate circuits. The representative revealed that it was all or none and whenever it happened, there was a short mild "burp" sound. I asked him to check for any shorts or any other abnormal circuit conditions that would cause over current conditions -- there were none. Everything was normal except for the few times that all three fuses would blow randomly, all at the same time.
Since this console was located almost 600 miles away, I decided to set up the exact circuit conditions in my lab with the one 50 amp and two 35 amp 12V DC supplies delivering power to three similar loads. I connected the supplies to a similar console. I starting latching relays to that in the console that activates the console lights, power supplies, and main air blower controller located in the utility room.
Everything operated correctly with no blown fuses. Since the problem occurred randomly about 5 percent of the time, I continued to cycle the system on and off many times. Finally, I heard a mild burp and all three supplies blew their fuses. I had to repeat this experiment many times to locate the burp sound again. It came from the latching relay.
Now, I could listen for the musical pitch, a low B, from the relay to identify the frequency and duration of the burp, which was 60 Hz for about 1/4 second. But why did it buzz only about 5 percent of the time? Putting a variac on the system, I could replicate the buzz more often with lower voltages. I determined that the spring and armature of the relay mechanically resonated at about 60 Hz and that the inrush current from the largest supply was sufficient to drop the supply voltage just enough to cause the relay armature to drop out for half a cycle.
Once that happened only during the short voltage peak of the AC cycle, the armature would oscillate at 60 Hz, thus operating as a mechanical rectifier causing pulsating 120V DC to continue to the three supplies. While power transformer primaries appear as relatively high impedance devices under little load to AC, they are a virtual dead short at DC. That is why all three supplies always blew fuses at the same time during those 5 percent random times. We fixed the problem by changing out the relay to another brand with an armature/spring resonance at a different frequency than 60 Hz. We also now use switching supplies.
This entry was submitted by Robert J. Vaughan and edited by Rob Spiegel.
Robert J. Vaughan received a BMus in Pipe Organ Performance in 1965 from Bethany College. He built a large three-manual pipe organ just out of college and rebuilt several other pipe organs until he was hired by Reuter Organ Co. in 1969 as a draftsman. Having a keen interest in music, structural engineering, and amateur radio electronics, Robert integrated all three disciplines into his final position with the company as director of production from the late 80s until retirement in 2008. He still does contract design work for the company and likes to restore old wooden tube radios from the 1930s.
Tell us your experience in solving a knotty engineering problem. Send stories to Rob Spiegel for Sherlock Ohms.