| I spent the first half of my professional career doing basic research in optical emission spectroscopy. This occurred in graduate school at the University of Illinois, Urbana, under the direction of Howard V. Malmstadt, between 1960 and 1965, and at the University of Wisconsin, Madison, between 1965 and 1982, under my own direction. I worked at Wisconsin for 17 years, graduated 17 Ph.D. students, and 17 MS students. My work occurred during the time when the students at Wisconsin were in full riot, protesting the Viet Nam war and developing the student power movement.
My interest in this area came from observing work at the General Motors Technical Center in Warren, MI during the late 50's and early 60's. This was directed toward improving spark plug electrode design by studying material erosoin using time-resolved spectroscopy. I visited this lab in mid-August of 1961, and learned how to build a Bardocz type electronic spark source from work going on there.
Professor Arpad Bardocz (Hungary) himself visited my Illinois lab in 1962, and, seeing my home-made recreation of his original source, became quite excited. He showed me how to place a lens between the spark and a vignetting slit and how to use a rotating mirror to pick up the image of that slit as a function of time. With that simple addition to my apparatus, I was able to get time-resolved spectra that accompanied the movement of ionized material away from the spark gap cathode. The next 20 years of my career were altered by that one observation.
| My spark sources, those of some of my students, and the results we obtained on how the spark works to erode electrodes and produce useful spectra, have all been published. The Wisconsin lab was closed in 1982 when I came to St. Olaf College to spend the second half of my career teaching. It was a wild ride.
Sparks come in various sizes and shapes. The ones that I was interested in occur at atmospheric pressure, in either air, nitrogen, helium, or argon. They are called "condensed" because the current rises in them sufficiently fast that they are radially pinched into a tight, highly conducting core of plasma by the toroidal magnetic field that forms just after the spark gap first breaks down and accepts current. Typically, current densities of 10^6 to 10^8 amps/cm^2 are needed to form such condensed discharges.
When lesser current densities are used (usually due to external restrictions of current rates of rise), non-condensed discharges with quite different physical properties result.
In my research, most of my spark sources used capacitive current waveshaping, with sets of series and parallel connected inductors (open coils) forming the resonant circuits that controlled the peak current amplitudes and rate of rise. I also developed a way to keep the current in the spark unipolar while still pulsating (follow the link to publications or patents above) by using a diode in parallel with the spark gap.
The photo was taken by Dave Coleman. The third floor lighted office was mine.
The scene is not too far fetched.