quarterwave source Perhaps the most successful spark source we built in our 17 year effort at the University of Wisconsin was the high-frequency, quarter-wave, current injection unit shown here. The design followed from work done by Thomas Bhruns around 1968-69. He introduced us to the fact that radio-frequency power fed into a node of an unterminated transmission line whose length was a quarter of a wavelength of the oscillations driving the line would cause a progressively higher and higher voltage to build at the unterminated anti-node.

Eventually, the breakdown voltage of the dielectric at this point would be exceeded and an electrical discharge would occur, terminating the line and quenching the cavity. This was precisely what we wanted to cause; the electronic initiation of a spark discharge was our primary goal.

If a charged capacitor was placed at the sending node, isolated from the radio-frequency generator with a choke, the capacitor would discharge into the line when the receiving anti-node broke down, injecting a damped pulse of current into the discharge. This gave the spark a "bite" that could be used to erode the grounded, cathodic electrode, thus giving us the spectral information we wanted about the chemical composition of that electrode.

The Bruhns' source was immediately successful. Later, Dr. Tsutomu Araki, a post-doctoral associate from Osaka University, Department of Physics, built the unit shown here with help from mechanicians Robert Schmelzer and Robert Lang of the UW Chemistry Department instrument shops. Dr. Araki was a mechanical wizard, and his skills combined well with those of Schmelzer and Lang to make this marvelous electromechanical device.

Because the radio frequency power responsible for the spark breakdown was so completely defined to be inside the quarter-wave cavity, and because the actual sample receiving the high current density from the injected spark current was a grounded cathode, there was essentially no radio-frequency interference that leaked into the associated electronics nearby. This meant that, for the first time ever, we were able to use modern digital electronics to ignite the spark and to simultaneously position the sampling spot on the electrode. Low current, digital stepping motors could be used to position the electrode beneath the running spark, heretofore an impossibility.

The source was an immense success in providing positionally-stable discharge trains that could be made to attack predictable spots on the surface of a sample electrode by using an a sharpened tungsten needle inside a tapered jet of argon to direct the spark channel while it was being formed by driving the quarter-wave cavity with radio-frequency power.

But, the source was not successful commercially. Instrument manufacturers deemed it "too complicated" for the average foundry operator, largely because of the pulsed radio-frequency power supply we used to drive it. Ironically, this supply turned out later to be much tamer and easier to manage than those used with the commercially accepted ICP. The quarter-wave source now exists only in the picture and the literature (see references 9, 13, and 42 in my bibliography.)

Another spark source that was less sophisticated in both its construction and its spectroscopic applications was the "adjustable waveform" high-voltage spark source that I developed when I first began my work at Wisconsin. This source went through a host of modifications in the hands of my graduate students over the years, and by John Bernier, an electrical engineer at the Jarrell-Ash Company in Waltham, MA. The first source made was derived from the Bardocz circuit I used in my research at the University of Illinois. It is shown here.

This source was based on using a diode and inductor combination placed in parallel with the analytical spark gap G2 to store and shape the current from a parent capacitor discharge so that it did not change direction but did pulsate. The idea was to separate the current cycles into electrode sampling and spectral relaxation periods, and then to integrate the spectral line intensities for elements of analytical interest during the relaxation periods.

The source worked exactly as intended, and my last graduate student, Dr. Steven Barnhart, proved that the concept would indeed translate into routine practice. But, the company that licensed the patent on the source, the Jarrell-Ash Co., buried the source inside a line of its direct reading spectrometers and changed its name to the "electronically controlled waveform source". No synchronization to its current waveform was ever done commercially, and no adjustments to that waveform for optimization purposes were done either.

The source succeeded in providing analysis of higher precision that available in the past, and for all of its technical lifetime that was all it was used for. Most users did not even know what it was or why it produced more precise results.

The adjustable-waveform source now exists only in history and the literature (see references 6, 19 and 22 in my bibliography.) It is shown in its most elegant, industrially-ready form below.

diode spark source