When I was a mechanical engineer in the late 1980s, I worked at a disk drive company in Longmont, Colorado, owned by Fujitsu. It was my job to design the motor that spun the platter of disks. To save space, the motor was integrated with the drive's baseplate, which was made of aluminum. After we had some prototypes of the design made, we discovered that there was a terrible resonance problem. While the motor was spinning, the baseplate vibrated like crazy, making it hard for the read-write heads to track the data. I wrote a Fast Fourier Transform (FFT) program using Borland Turbo Basic to isolate the offending frequency. I attached a small accelerometer to the baseplate with a bit of beeswax and spun up the motor. The program took the signal from the accelerometer and isolated the different frequencies. It turned out that the vibration was either 3.1kHz or 3.6kHz (I can't remember). I tried to dampen the vibration by adding different structural elements to the baseplate, but it took a long time to get prototypes made. I ended up quitting and moving to California with Carla to work on bOING bOING (the zine) full time before I solved the problem.
This video by the excellent 3Blue1Brown about the Fourier Transform was a trip down memory lane.
