Back in 1978, I was working at a company that manufactured magnetic memory products. We were starting on our second disk drive product that had one fixed and one removable 8 inch disk, plus a linear head servo. We found that a lot of disk heads were ringing at about 6MHz upon shutoff of write current.
The head consisted of a ferrite bar on which the coil was wound in contact with a roughly U-shaped ferrite piece, which was gapped at the tip. The whole thing was encased in a ceramic block attached to a cantilever spring.
LC resonance was ruled out immediately, leaving mechanical resonances under scrutiny. The folks at our subsidiary were convinced it was the coil. They tried several variations on epoxying the winding to damp it. Some of these tests resulted in increased damping. I was skeptical the coil could possibly have a mechanical resonance at that high a frequency.
I questioned why the epoxy didn't cause a huge change in the frequency when it failed to completely damp the oscillation. I noted there wasn't much variety in resonant frequencies among heads exhibiting the problem. It seemed it would take deliberate effort to make coils wound on a core with such consistent mechanical resonance frequencies and not a chorus of different frequencies from each one.
The experiment that finally identified the culprit was a simple one. I brought a magnet within a few inches of the head. The ringing frequency varied smoothly with changes in distance. There was no dc in the winding during the ringing; therefore, the magnet could not have strained the coil to cause this smooth shift in force-free ringing frequency. It was magnetostriction in the core. The epoxy experiments had generally resulted in bonding of the core along with the coil, and so we were partially successful at damping it.
Dick Neubert has a long and diverse history in electronics and (mostly real-time) programming. While a co-op student at Northeastern University, he contributed bug fixes and general improvements to a satellite antenna positioning system for Queen Elizabeth II. His design work ranges from high-performance disk head servo systems to computerized automation systems for sawmills. He hold an MS in Engineering Sciences.
This entry was submitted by Dick Neubert and edited by Rob Spiegel.
Tell us your experience in solving a knotty engineering problem. Send examples to Rob Spiegel for Sherlock Ohms.