The X-Ray “Diversion”

In the early 1970s, the DataNIM  line  continued but the instruments were no longer manufactured independently and, thankfully, were no longer painted in that awful maroon color. Gerry Matthews stayed on as Product Manager for a few years, eventually leaving and forming a company that developed a new product that he described as “an electronic strip chart recorder”.  At that time in history, there was a great deal of excitement around the use of Si(Li) detectors in X-Ray applications. Many of the traditional nuclear instrument companies rushed to get into X-Ray, because it was a growth industry and did not carry the “nuclear taint” that affected (and still does) their perceived value. To distinguish semiconductor-based spectrometers from the traditional crystal (dispersive) spectrometers, the user community dubbed the former “non-dispersive detectors”, despite the negative connotation. The awkward terms “wavelength dispersive” and “energy dispersive” also made their way into the lexicon. The big advantage of non-dispersive systems is that they collect an entire spectrum as once rather than scanning slowly through the energy range of interest. The resultant speed of data collection in many applications made up for the loss of energy resolution which was huge: ~160 eV(FWHM) vs. 1-2 eV(FWHM) for crystal (dispersive) spectrometers.

Among the companies that went X-Ray were Nuclear Diodes, changing its name to EDAX, Princeton Gamma-Tech (PGT), and Tracor Northern (Northern Scientific). KEVEX, once part of TMC, was owned by Dick Frankel, the PT Barnum of X-Ray. Ed Woo was the detector guru at KEVEX for many years and KEVEX was the industry leader. Even ORTEC got into X-Ray in a significant way under the direction of Chuck Greer, who joined ORTEC not very long after leaving Canberra, although he remained a significant owner of Canberra stock.  Interestingly, this diversion by ORTEC into X-Ray gave Canberra a lot of breathing room in the nuclear instruments market. Just saying!

Canberra was not immune to the attraction of X-Ray either. Around 1970 Dr. Gerald Laurer, of the NYU Sterling Forest Research Center, discovered that the 88 keV gamma ray from Cd-109 was just above the K shell absorption edge of lead and thus was very effective in exciting K shell X rays. He used a Canberra planar Ge(Li) detector in a home-made portable XRF Analyzer and got the attention of the NYC Health Department which was fighting a losing battle against lead poisoning among children in the city. Orren incautiously agreed to develop a commercial version of this XRF analyzer and the city gave Canberra an order for 10 units. In record time we designed and built these instruments and dubbed the product “Lead- Ray”.  When Chuck Greer saw the first unit he said to Orren in his typical caustic fashion “There are three misspellings in the word fluorescence on that front panel”.  Orren swears that there were only two! With an LN-2 capacity of about one liter, the holding time was little more than 24 hours. The Health Department hired a bunch of college kids to lug these instruments by subway from one ghetto to another the following summer, and, as could have been predicted, it was a disaster with units failing due to abuse or accidental warm-up. Orren lugged one unit around and demonstrated it in other cities including Chicago, St. Louis, and Cambridge but no other orders were forthcoming. Thus ended Canberra’s first and last attempt to build portable XRF Analyzers.  Pictured below is a Lead-Ray Analyzer that Orren saved from the scrap heap more than 40 years ago. He fired it up again in January 2015 and the Nixie tubes lit up perfectly!

Lead-Ray XRF Analyzer

The attraction of laboratory X-Ray systems continued, however. Si(Li) detector systems were in great demand for use of X-Ray Microprobes and Scanning Electron Microscopes (SEMs). Before Canberra made Si(Li) Detectors or MCAs, for that matter, Bobby Maggi talked Orren into selling a non-dispersive X-ray system to Phil Giles at Bethelem Steel to go on an old MAC-400 Microprobe. We cobbled together a system using a Si(Li) detector kit made by Kevex and an Econ I MCA made by Northern Scientific. Canberra made the special cryostat and supplied the NIM electronics, by far the lesser portion of the order. We built a miniature cryostat that held liquid for about four hours or so and provided a “chicken-feeder” Dewar to which it could be attached when not in use. We’ll forever be indebted to Phil Giles for the remarkable care and feeding he provided to keep this instrument up and running for the next couple of years. Later we replaced the miniature cryostat with a more sensible five liter unit.

In March of 1972, we hired Bob Lothrop, who had been a Si(Li) detector expert for Fred Goulding at Lawrence Berkeley Laboratory. Bob exuded extreme confidence in making Si(Li) detectors and we were  encouraged to flesh out our non-dispersive X-Ray product line. Thus we added dedicated 1700 Series NIMs including the 1713 Amplifier and the 1764 Spectrum Enhancer (really a pile-up rejector and live time corrector). We sent Charlie Reynolds, a clever analog designer, to LBL to learn about pulsed optical reset preamps, and he designed the 1708 preamplifier to complement our new Si(Li) detectors. For neatness, we made a composite cable for the amplifier-preamp connections, using miniature D-15 connectors, burying low voltage, signal and bias (up to 1000Vdc) wiring therein, undoubtedly against all safety regulations, but this never caused a problem in the field.

With this investment and with the Model 8100 MCA, we were ready to tackle the microprobe and SEM application. DataNIM was still alive so our sales force could and did address the X-ray Market with a comprehensive product offering.

One big problem remained, however. Each microprobe and SEM required different, often exotic, cryostat designs for access to the ports that were available for detector entry. In the next couple of years we designed cryostats or interfaces for some 14 different machines before getting our first repeat order. Getting accurate information on the requirements was a challenge so each of these orders was a project unto itself. Finally we sold a second system for a JEOL JSM-U2. With all the information at hand we quickly built and shipped this system, only to get a call, days later, from the user telling us that he wanted the detector on the rear port, not the side port. Back to the drawing board!

The X-Ray effort continued for some three years but it was a struggle to keep up with the companies that were focused on X-ray, many of which had dedicated X-ray analysis software. Canberra had no X-Ray Analysis software in those years but this did not keep some of our clever sales people from selling very expensive “systems”. Klaus Stock, who had the Canberra distributorship in the Benelux countries, was chief among them. Of course, Klaus could have sold ice to Eskimos, he was that good.

Furthermore, Si(Li) detector production was never consistent. Bob Lothrop struggled to make good detectors on a consistent basis, often employing controversial techniques. Eventually he was asked to work on HPGe detector development, but he took little interest in it. He saw the “handwriting”and left Canberra in 1976. We kept the Si(Li) product line alive for a while by buying bare detectors from different manufacturers including Kevex. United Detector Technologies, Tracor, PGT, Jordan Valley (Israell), and Thowe Engineering.  Finally Mike Yocum, in his spare time, established robust production of Si(Li)s which has continued more-or-less routinely to this day.

Orren and Gerry Matthews had proposed a dedicated X-Ray MCA, the 8180, with attractive features including on-screen X-Ray markers and element identification, but management decided that we should focus on nuclear measurements instead and the 8180, introduced in 1975, was re-purposed for gamma ray analysis.  This pretty much ended Canberra’s effort in X-Ray systems.

Since that time, however, X-ray detectors: Si(Li)s , followed by LEGes, Ultra-LEGes, and more recently, PIPS X-ray detectors have been manufactured and sold, with significant success, in OEM and special applications. Danny Meert, in Belgium, secured a license from the University of Antwerp for a “fundamental parameters” analysis software package which he sold in small quantities for years but it did not catch on with Canberra in general. This has gone on largely with direct support from factory experts, and without much support from corporate management, as newly witnessed by the ubiquitous Canberra advertising banner listing the symbols for alpha, beta, gamma and neutron radiation, with no mention of X rays.