The Lithium-Drifted Germanium [Ge(Li)] Detector was a remarkable product in two ways. First, the idea that it would be possible to counter-dope Ge crystals by “drifting” impurity ions centimeters into the bulk under easily achieved conditions was amazing. A process of “drifting” was patented in 1962 by E.M.Pell at General Electric but there is probably earlier work on this technique. Both silicon and germanium were candidates for drifting from the onset. Second, the energy resolution of the Ge(Li) Detector was some 40 times better that of the next best type of detector, the Thallium Doped Sodium Iodide [NaI(Tl)] Scintillation Detector which had ruled in gamma ray detection since World-War 2. It’s no surprise that scientists in Universities and Laboratories jumped at the chance to make Ge(Li) Detectors. This led to a profusion of papers which could have been sub-titled “I Made a Ge(Li) Detector and Look How Good It Is”. Later, when commercial production began, another series of papers could have been sub-titled “I Bought a Ge(Li) Detector and Look How Good It Is”.
ORTEC, Princeton Gamma-Tech, and Nuclear Diodes were well established manufacturers of Ge(Li) Detectors before Canberra entered the business.
In 1968, Dr. Hans Fiedler, who had established Ge(Li) detector manufacturing at Nuclear Diodes, joined Canberra as Technical Director of the Detector Division, which was the first of several divisions started by Canberra over the next few years. Fiedler was a young German Ph.D. who had made detectors as a postgraduate at McMaster University in Canada and then moved to Argonne National Laboratory. The President of Nuclear Diodes, Charlie Walsh, attracted Fiedler from Argonne to Nuclear Diodes which was basically a silicon surface barrier (SSB) detector business at the time. Fiedler was quite successful in establishing production of Ge(Li) detectors there, and Nuclear Diodes quickly joined ORTEC and PGT as major suppliers of these revolutionary products. Fiedler’s contribution to the success of Nuclear Diodes was undeniable and he came to believe that he was not rewarded appropriately. He was particularly incensed when Charlie sold a manufacturing license to Intertechnique in France without giving Fiedler a share of the fee of, reportedly, $200,000.
Charlie was a colorful character and there are many Charlie Walsh stories. One which Fiedler delighted in telling had to do with their attempts to cut a big germanium crystal with a wire saw. The saw wouldn’t go through the crystal in one direction so cuts were made from opposing sides. Unfortunately they did not meet perfectly and there was a big discussion about how to effect the separation. Charlie, ever short on patience happened along at that time and grabbed the crystal by both ends and stretched across his knee as one might try to break a broomstick, In fact it did break, into hundreds of small pieces. This may have been the last time Charlie offered hands-on help to the Germanium Department.
When Fiedler heard about the “Division” concept that Chuck Greer and Emery Olcott has devised as a formula for growing Canberra, he quickly came to a decision to join the company. Of Course, it was now Charlie Walsh’s time to be angry and he did warn Canberra not to use the technique of Lithium Diffusion that Fiedler that devised and that Nuclear Diodes had patented.
As a parting gesture in leaving Nuclear Diodes, Fiedler hired Pete Krantz, who was at Nuclear Diodes for an interview, to join him at Canberra. Pete had worked on Ge(Li) detectors at Harshaw and he spent about a year as Fiedler’s assistant at Canberra.
Sam Knoll, who had joined Canberra earlier on the promise of a future job as a Division Manager, was working as a Sales Engineer in the Southeast. He was selected to be the GM of the Detector Division and planned to join Fiedler in the fall of 1968 after production was underway. Before he came in, however, the new Medical Instruments Division, with Les Hammer as Technical Director, was established and Sam became GM of this division instead. Orren Tench, who had been the Engineering Manager at Canberra for about 18 months, was promoted to GM of the Detector Division, despite the fact that he was not one of the MBA “whiz kids” generally recruited for these positions. Orren was to hold more-or less the same job with Canberra until 2006. `
With the help of Pete Krantz, Fiedler set out to build the 25 or so “drift rigs” and other production equipment. The apparatus went together quickly because it was quite simple. The laboratory went up quickly also in the southwest corner of the basement of 50 Silver Street under the hammer of the indefatigable John Rak. Construction materials were the standard Canberra issue, ¼ inch plywood on two-by-four studs. The drift room was air-conditioned with a big window unit, the first space with the exception of the executive office to be so blessed.
Fiedler claimed no particular expertise in vacuum systems or cryogenics so we hired a mechanical engineer, who had worked in GE’s Vacuum Systems Division, to design cryostats. Jerry Jarvis had a wealth of experience and was perhaps even more headstrong than Fiedler, if this were possible, so the two of them clashed often and loudly in the next few months.
Fiedler had two visions for cryostat design. One involved the use of metal seals and ceramic feedthroughs to improve reliability over conventional O-rings and glass feedthroughs. The second was a cryostat that could operate either vertically or horizontally. The former was realized immediately although the first prototype, more-or-less hewn out of solid stainless steel, looked a lot like a weapon of war. This prototype somehow survived and is pictured below. Production models were somewhat more elegant yet they retained a rather massive appearance until the hard metal seals were replaced with indium seals around 1985.
The Canberra cryostat did turn out to be very reliable with virtually no failures in the first two or three years of production. No doubt the metal seals helped in this regard but we also used a 1.75 inch diameter tailstock on these early models. This dipstick contained enough molecular sieves to pump small leaks for years and must have been a factor in reliability as well. Later the dipstick diameter was reduced to 1.5 inches and then to 1.25 inches and we had to pay more attention to other factors involved in long-term reliability just to stay even.
The second innovation that Fiedler promoted was dubbed the “Calabash” because of its gourd-like shape. An artist’s illustration of this device graced the cover of the Ge(Li) Redbook, an applications/technical reference booklet written by Fiedler and Tench to give credibility to the Detector Division. Jarvis was originally commissioned to design but his design involved the use of 1/8 inch thick stainless steel hemispheres and it would have weighed a ton had it been built. A couple of years later a single prototype was built by Gert Sulfrian of Sulfrian Cryogenics in New Jersey. This unit worked and was sold but it was damaged in shipment to the customer, and after a bitter argument over detector performance, the order was cancelled. This ended the Calabash for all time.
It became clear that Jarvis was not to be around for the long term so an amicable parting was arranged. Jerry stayed on for several months working one or two days per week coaching our young draftsman, Paul Olsen. Meanwhile Jerry bought a commercial fishing boat with the intent of leaving engineering forever. This adventure soon ran aground as well but the parting with Canberra was permanent.
In parallel with tooling for the detector operation a machine shop for cryostat production was established with Dennis Murray as the first employee and foreman. In making way for this facility we endured one final encounter with Jarvis. Most of the basement at 50 Silver Street was now occupied by the Detector Division, the Nuclear Division Engineering Department, and by the embryonic Data Systems Division. The southeast corner was dominated by an ancient 60 horsepower engine-generator that had once provided power to the factory. Jerry convinced his father and brothers, who ran a fish packing house in Rhode Island, that they should remove this monstrosity and install it at their packing house for emergency power for their freezers. Canberra gladly gave then the unit and one nice fall Saturday, the senior Mr. Jarvis and his sons showed up with trucks and tools to effect the move. Jerry, of course, was the only engineer among them and he took it upon himself to bark orders all around, in spite of the fact that nobody had a clear notion of how to move the beast. As the day wore on the temperature rose and tempers grew shorter and shorter. Suddenly everyone realized that the “old man’ was missing. One of the brothers announced “Jerry makes the old man nervous so he is just taking hisself a walk”. Not that day but the generator was eventually raised from the basement and sent off in the direction of Rhode Island. We never saw Jerry or the generator again.
Meanwhile Fiedler was not having much luck with detector production. He had attempted to do lithium diffusions in a way unfamiliar to him because he was afraid of violating the Nuclear Diodes patent. We knew the patent wasn’t worth much for it was taken directly from a guidebook describing electrolytic production of lithium from lithium salts. Nonetheless we did not want to risk a lawsuit from Charlie Walsh. After struggling with the new process without much luck, Fiedler secretly set up the electrolytic process and came at night, diffusing crystals by the method he had developed earlier. He was wary of letting anyone know, including Pete Krantz, so Pete was somewhat ineffective in the job he was hired to do. This situation did not improve and Pete and the company parted ways before a lot longer. The mantle of secrecy, however, still covered the detector operation and it was to remain this way during Fiedler’s entire time at Canberra.
For some time Fiedler continued to make detectors without help. Thanks to one very good ingot of Hoboken germanium we made a number of detectors which were good for that day. This ingot, #1BA-07, made at least nine coaxial detectors measuring 2 to 4% relative efficiency and thus got the division going.
As Canberra detector production came on line the market situation took a turn for the worse. Charlie Walsh hired Ron Wagner, from the University of Maryland, to replace Fiedler and they cranked up production immensely. Charlie was an aggressive salesman, and he was soon selling detectors at what we thought were ridiculously low prices. Fortunately this continued for only a few years when the attention of Nuclear Diodes turned to X-Ray Systems and the company name was changed to EDAX. This brought on explosive growth and attendant problems so after a few high-profile years, EDAX almost went broke, the company was sold to Philips, and Charlie Walsh was forced out.
Canberra detector sales grew slowly. We found that we could sell at higher prices in Europe than in the overheated US market and Fiedler was accepted readily by German scientists who treated him as a favorite son. Tench took and active role in detector sales and sales support domestically and the Detector Division was profitable almost from the time it first began shipping detectors in 1969.
In 1972, when the company was overextended in support of the myriad divisions and prospects for survival looked grim, Fiedler chose to cash in his chips and head back to Germany to set up another detector production effort. It was with trepidation that Orren Tench took on responsibility for detector production. Although Fiedler and Tench had a good personal relationship there was much speculation that Fiedler had departed with the “black magic” necessary to make good detectors.
After Krantz left, another technician, Darrell Smith had come and gone, and Marie Darin, fresh out of college in 1970, was helping Fiedler in the months prior to his leaving. She agreed to stay on, despite Fiedler’s advice, and a rebuilding effort started. It started in a hole however, because the diffusion process wasn’t working properly and much of the in-process inventory had thermal cracks. We devised a method of grinding out the cracks and changed the process to prevent further damage. Fortunately, almost all the inventory was salvaged and, unbeknownst to our customers, we shipped many a detector that looked a bit like Swiss cheese in the next year. We also had a nightmare in the field having changed (for some reason) to an inefficient form of molecular sieves some months earlier. Virtually all the detectors made with this material failed in warranty swamping us with repairs and ruining our hard-won reputation for reliability. With Fiedler and the former Canberra salesman in Germany, Frances Vreven, geared up to produce and sell detectors in Germany, our most attractive market disappeared almost overnight. This was the situation we faced in the early 1970s.
Marie turned out to be a terrific team member and our production efficiency soon reached new highs. Toyo Corporation left ORTEC to become our distributor in Japan and this quickly became a lucrative new market for Canberra. We managed the transition to Zchrochalsky material which allowed us to make larger round detectors and the business grew slowly but surely over the next five years.
It was during this period that we began to hire students from the Wilcox Technical High School in Meriden. Mr. Porter, who taught electronics there, began to send his best and brightest students to Canberra and we brought in many of them to work part-time in their senior year, with the possibility of full-time employment after graduation. Among those who worked in the Detector Division were Jim Colaresi, John Yarmolovich, Mark Cwirka, Dan Smith, Pete Ehmer, and John Costa all of whom made long productive careers in the Detector Division. Joe Cotton started in Detectors and moved to Field Service. Others who worked in other departments include Mike LaForge and Ken Rosengrant.
I don’t know what we would have done without this infusion of talent from Wilcox. These young people formed the nucleus of a talented, energetic, and resourceful workforce. Jim Colaresi started in the Test Department and eventually ran that department as well as the in-house service department. He became Product Manager in 1992 and Engineering Manager around 2006. John Yarmolovich who joined Canberra a year after Jim joined, worked for Marie Darin in Fabrication and took on more and more responsibility in Production through the years.
Not all of the key players came from Wilcox, of course. Karen Crolius joined as a part-time employee to help with Order Administration even after hearing from the Personnel Department that it might not be much of a career. I suppose that theory went out the window after she clocked some forty years in that job.
To exemplify the potential of these students take the case of Chris Souchins, who worked for us during the summer after his senior year at Wilcox. He expected to become a full-time employee but as fall approached, he came to me and said that he had been offered a scholarship to attend the University of Connecticut. We wished him well. About 6 years later I stumbled across Chris in the vicinity of UConn and he revealed that he was finishing up his Ph.D. in Mathematics soon!
The first Ge(Li) detectors were made from material supplied by Sylvania which had a plant in New York state. At about the time that commercial production of detectors began, the Sylvania material suddenly stopped drifting. That is to say, the rate at which the lithium ions would migrate through the bulk crystal in the drifting process, which was slow at best, slowed to a crawl so that it was virtually impossible to make detectors from it. Universities and laboratories all over the world found themselves unable to make detectors and many of them gave up trying. Enter Metallurgie Hoboken-Overpelt (MHO) or “Hoboken” to the trade. This Belgian company, part of Societe General, supplied some of its boat-grown trapezoidal Ge crystals to various laboratories and it drifted as well or better than the good Sylvania material. Thus began a long period in which Hoboken held an absolute monopoly on drift-grade germanium worldwide.
Hoboken germanium was distributed in North America by Nucleonic Products Company (NPC). Bill Wheeler a true gentleman with an unmistakable deep bass voice, was in charge of germanium sales for the duration of the monopoly and it is probably true that NPC made obscene profits on the produce during that time. Perhaps Hoboken did also but we will probably never know. What is known is that there was absolutely no guarantee that the material would make good detectors. Certain electrical and mechanical specifications for the material were in effect, and there was even a drift mobility rating, but there was no guarantee on charge collection or voltage breakdown. The former haunted us always and the latter caused a lot of pain over a period of two years prior to the end of the monopoly. From the beginning until the mid-1970s we were never credited for any bad material supplied by Hoboken.
Bill Wheeler worked hard to keep his customers under control, commiserating with us over the lack of responsiveness at MHO and meeting with us at trade shows or in our factories. But Bill exhibited little in technical understanding of the product. Sometime in the early 1970s, we had a visit from Bill and a factory man, Dr. Walter Schoenmaekers, who was in charge of germanium R&D at MHO, and who was making his first trip to visit customers in the US. Walter figures prominently in the future of Canberra, but he may have been a bit green at that time. We had a strange crystal in process that had a nearly perfect ring defect around the entire circumference. It looked like a very thin cut in the surface as made by a sharp tool. So we put this object under a microscope before the “expert” who, after some time, looked up and pronounced “It’s a crack”. What Walter didn’t know, of course, was that he was talking to the world’s experts on cracked germanium. Orren and Marie Darin has seen more goddamn cracks in germanium than anyone ever dreamed of. It could be that this was the first germanium crystal we ever sent back for credit. They were just too embarrassed to refuse!
Toward the mid-1970s we encountered the voltage breakdown problem in Hoboken germanium. Many crystals simply would not take sufficient reverse bias to fully deplete or to give good charge collection so resolution and peak shape suffered greatly. MHO assured us that the problem was unique to Canberra and we suffered with this delusion for almost two years. Finally, Sandy Wagner, who was the Ge(Li) production manager at ORTEC, called and gingerly inquired about material problems. We learned that our experiences were identical and soon got the same story from Peter Ryge at Princeton Gamma-Tech (PGT). Now, for the first time MHO was forced to acknowledge some responsibility for bad quality, but it would take more to change their attitude in a significant way.
This came about, not surprisingly, when the monopoly was broken. ORTEC started a crystal growing operation in late 1969 under the guidance of John Walter, one of the founders of ORTEC and the man who ran their detector operations with an iron fist for more than 20 years. Rex Trammel, detector maker turned crystal grower, was fairly successful and before the mid-1970s ORTEC was became nearly self-sufficient in drift grade material. Desperate for a solution from the voltage breakdown problem we bought some ORTEC detector elements for re-mount, and they later sold us drifted blanks shipped in dry ice. It was hard to turn a profit this way but we found that the ORTEC detector were generally long, skinny and oversized. We found that we could take an ORTEC true-coax detector of nominal 10% efficiency, re-diffuse it to close both ends, cut it in half, drift it a bit, and make two 8% detectors from it. In this way we eked out a profit selling these odd-sized detectors for a time.
Later, Alden Matsubara, who had joined ORTEC from the EG&G High Energy Instruments business, was charged with selling un-drifted ORTEC crystals and we bought a lot of them. The MHO monopoly was finally broken.
After MHO lost all of ORTECs business and much of ours they finally began to consider their responsibility in pricing and in quality so we resumed using Hoboken crystal right up to the end of the Ge(Li) era which ended soon after July, 1982, when our last Ge(Li) came out of drift.
As a footnote to the drift-grade germanium crystal growing business we should mention a few also rans:
The company Geoscience, which gained control of the TMC MCA design group headed by Harvey Roberts when TMC went bankrupt, attempted to grow crystals in the late 1960s. Their stated goal was to produce a one liter (1000 cc) detector at a time when 50 cubic centimeters loomed large. We believe they had government funding from the military in support of this project. A scientist of Chinese extraction was their crystal grower and they eventually advertised their effort with his picture beside a crystal with the caption “Ah so, that’s my Germanium”. There was no such thing as political correctness at that time. Needless to say the one liter Ge(Li) never materialized. After we bought the Geoscience MCA product line around 1971, they gave us some 20 big clunky drift power supplies, the last evidence of their efforts in this field.
Eagle-PItcher of Quapaw, Oklahoma, was a supplier of germanium metal and infrared lenses, etc. Sometime in the early 1970s, they established a small scale operation to grow drift-grade crystals, but never became a serious supplier.
The behemoth Philips also invested in crystal growing and detector manufacturing in the 1960s and early 1970s. It is not clear to what degree they used their own Ge crystals but their detectors sold well in Europe for a while, probably in large part because Klaus Stock, who later founded Canberra-Positronika was involved.