Aug 19, 2014

When Radiation Meant Radio Communications, and then Harris

Most of the Analog Footsteps seemed to have trodden over neighborhoods around MIT and Stanford. So I always know I’ll find an interesting story for companies based elsewhere. For this post, I’ve chosen Harris Semiconductor in Melbourne, Florida.

According to its website, Alfred and Charles Harris founded the Harris Automatic Press Company in December, 1895. From their day job in a jewelry store in Niles, Ohio, they invented an automatic sheet feeder and a new printing press to better accommodate it. By the middle of the last century, it had grown into one of the world's largest and most successful manufacturers of printing equipment under a new name – Harris-Seybold. In 1957, Harris-Seybold merged with Intertype Corporation, a world leader in typesetting equipment. The 1957 acquisition of Gates Radio brought Harris in to the radio equipment industry, specifically broadcasting transmitters and microwave equipment as an extension of the media industry. Even with the acquisition of Gates Radio, the focus of the company remained on the printing press. Now called Harris-Intertype, they foreshadowed that the printing press would move away from a strictly mechanical product and depend more on electronics. Therefore, in 1967, they acquired Radiation Inc., a manufacturer of space and military technology located in Melbourne, Florida, just south of the Kennedy Space Center (Cape Canaveral to people of my age). In 1974, the company changed its name to Harris Corporation, and moved its headquarters from Cleveland, Ohio, to Melbourne. Many other acquisitions followed. It later sold its printing business to focus exclusively on electronics.

So, who was Radiation, Inc. and why would Harris abandon print and broadcast media for what Radiation did?
Homer Denius and George Shaw
Let’s go back to World War II. Henry Stimson was the Secretary of War and therefore was in close coordination with Vannevar Bush, particularly on the Atomic Bomb. One of Stimson’s assistants was Thomas Meloy. After the war, the U.S. Navy suggested that he and Joseph Parks move from New York City to Washington, DC, to get government contracts. Putting their names together, they formed Melpar, Inc. Homer Denius and George Shaw worked at Melpar in Virginia. They presented plans to the Melpar management for expansion into instrumentation, specifically telemetry. Melpar rejected this proposal, so Homer and George decided to try it on their own.

In 1950, they founded Radiation, Inc. The context of “radiation” actually referred to radio frequency radiation, (communication by radio) and not nuclear fallout or electronics suitable for satellite applications. They started in Melbourne, Florida, in cheap buildings available on the former World War II Naval Air Station in close proximity to Cape Canaveral and its missile range activities. Actually, the good boating in the area was probably a major factor to the founders, since they were both enthusiastic yachtsmen. The company was funded from their own savings and likely some money borrowed from friends. As with Melpar, their focus was on government contracts. The company soon developed expertise in several areas, including: PCM telemetry systems (airborne and ground); digital data acquisition and processing systems and large tracking antenna. In many cases they were working decades ahead of the practical state of the art. John Hartley joined the firm in 1956 after serving on the faculty of Auburn University. That was the year that Radiation stock went public. Hartley later went on to become CEO of the company.

Radiation's involvement with the aerospace program included equipment for the Telstar and Courier communication satellites and the Nimbus and Tiros weather satellites. Military systems that relied upon Radiation equipment included the Atlas, Polaris, and Minuteman missiles.

Source: Radwiki (see notes)
The earliest Radiation hardware known to still exist is the "Force Amplifier"; this one pictured was purchased on E-Bay by a Radiation memorabilia collector at Harris. Yes, apparently there is at least one Radiation memorabilia collector! It dates back to about 1955. It was designed to measure and indicate the forces applied by the pilot to the various control surfaces of an aircraft in flight. The amplifier contains three vacuum tubes: 6U8 triode-pentode; 12AX7 dual triode; and 6AQ5 power pentode.

Source: Radwiki (see notes)
Radiation also had several facilities in Orlando. There was an Instrumentation Division located on the McCoy Airport property; a Research Division near the Herndon Airport. The Space Communications Division was established in Mountain View, CA, near Moffett Field. In 1960 Radiation acquired Levinthal Electronics Inc, in Palo Alto, CA (founded by Elliott Levinthal from Stanford University). Levinthal built high power transmitters, including a 430MHz radar transmitter at the Arecibo Observatory in Puerto Rico producing 2.5MW of peak power. In 1959, Radiation started a new facility in Palm Bay, Florida, and by the end of the year were occupying the first three buildings of what eventually became a large complex there.

Another person who left academia to join the staff at Radiation was Dr. Joseph Boyd. In the late 1950s and early 1960s Boyd taught electrical engineering at the University of Michigan. At the same time he was also director of the Willow Laboratories, a prestigious science and technology research institute with a staff of more than 1,000 scientists and engineers. Boyd joined Radiation in 1962 and within a year was made president. His first significant action as president was to set up a microelectronics plant to develop and produce integrated circuits. For PCM telemetry products, Radiation had contracts with Fairchild and Motorola for integrated circuits. But Boyd wasn’t quite satisfied, so he said “I’m going to start my own.” It’s difficult to verify, but it seems that a strong proponent for a captive semiconductor group was Peter Petroff.

Who was Peter Petroff? Petroff was born in the Bulgarian village of Brestovitsa, in 1919, to an Eastern Orthodox priest. After attending a religious seminary, Petroff enlisted in the French Foreign legion in October 1939. He was captured by the Germans while defending the French Maginot Line in 1940, and sent to a German Prisoner Of War camp in Poland. There are two post-war stories that are in conflict (in my mind, at least), one is that during the war he was conscripted into military service as a Nazi soldier. At war’s end, the Bulgarian government offered a short period of amnesty, but after that anyone who had served in the German army would be declared a war criminal and sentenced to death. The other story is that he returned to Bulgaria in March 1941 and became an officer in the Bulgarian Army. His duties included being a palace guard to King Boris III of Bulgaria and participating in the Honour Guard for the funeral of Turkey's President, Kemal Ataturk, the founder of modern Turkey. If I can resolve the conflict, I will edit this post.

In 1944, he moved to Germany to study engineering. He graduated from Darmstadt and Stuttgart universities with a master's degree in electrical, mechanical, and civil engineering. While in Germany he also studied his life-long passion, naval architecture, and designed and built the first of over 60 boats in 1947.

The Canadian government offered a visa and he immigrated there. Petroff arrived in Toronto in 1951. He worked on arctic engineering and construction projects for the US Air Force at Goose Bay, Labrador, and Thule, Greenland. He joined an overseas construction company and went to Indochina in 1956 for assignments in bridge and power plant construction. While in Vietnam, he built a 65-foot catamaran of his own design. He had the catamaran shipped to Canada, and sailed south on the intercoastal waterway to Florida with his wife and two sons.
After berthing in the Melbourne city harbor, his unique catamaran was noticed by George Shaw, the co-founder of Radiation. George always had yachts and after a tour of the boat with Peter he pulled strings to get him a position at Radiation. This chance meeting gave Peter access to executive row at the company and he never ceased to use it.

The family version of the story is slightly different, but not necessarily in conflict. Ralph Petroff, his son, said he might still be living in Canada if it hadn't been for Sputnik. Peter Petroff, tried to emigrate from Bulgaria to the United States after World War II, but he got only as far as Canada. Petroff said his father was told by U.S. immigration authorities that, with the waiting list at the time, it would be 2007 before he and his family could get into the country. That changed after Sputnik. His father was an engineer, and the United States was scrambling to find technically trained people. His father got his green card about 18 months after Sputnik launched, Petroff said.

I found an unauthorized obituary written by former colleague, Don Sorchych, who concluded: “…as far as the semiconductor division, which I later managed, it never would have existed without Peter. Peter was viewed as an expert in microminiaturization. And he correctly argued integrated circuits, then in their infancy, would be the ultimate solution in future systems.”
“With his connections in the executive suite, he badgered and bullied them, virtually wearing out the carpet in visits followed by lunches in the company cafeteria. He was the ultimate salesman, bright, persuasive and persistent.”

Petroff’s story continues long after his brief employment at Radiation, Inc., but I will save that for some other blogger. I’ll only offer a few highlights. His most unique distinction was to be officially declared an Enemy of the People by the communist regime in Bulgaria, for which he received a death sentence in absentia. The sentence was later lifted. He is also credited for renovating the Gemini II, a boat that served as the base of operation for Lee Taylor's successful assault on the world water speed record on Lake Guntersville in 1967.

In its obituary for Peter Petroff, The New York Times quoted Ralph Petroff, who said it was ironic that his father had died a peaceful death. "He always laughed at danger and he laughed at death. He should have never made it to his 83rd birthday, let alone his 20th," Ralph Petroff said. "I guess if you were to combine Indiana Jones with Thomas Edison, the result would be Peter Petroff." (The New York Times also credits Petroff with inventing the first digital LED wrist watch – which isn’t true, although he was at Electro/Data, Inc., headed by R&D manager John Bergey, where the first Pulsars were developed. I’ll avoid that whole story, thank you.)

The idea of developing semiconductors was radical. Executive row rocked with dissent, but Shaw’s booming voice could be heard arguing the strategic necessity, according to Sorchych. “It was a huge risk for a small company, which at that time was a 100 percent U.S. government contractor.”

Finally the decision was made to proceed, solely due to Peter’s persistence according to Sorchych. So, “although he never ‘organized’ the division as stated in his obit, it wouldn’t have happened without him.”

It is unclear to me what role Petroff played after that. To start the group, Dr. Boyd hired Uryon Davidsohn in 1962 to be the director of the new group, called the Physical Electronics Department, and the charter was to introduce new techniques for thin-film, silicon and hybrid forms of integrated circuits. Urie was the first person hired. He came from General Electric, where he was a consultant, and claimed to have a PhD – but didn’t. According to Don Sorchych, Urie was a short, humpbacked, bug-eyed, homely guy, exceedingly brilliant, weird, but with a certain charm. Urie hired Clyde Combs from Texas Instruments, who in turn brought over many more engineers from T.I. Don worked on the “systems” side of the company and was brought over because of his extensive knowledge of circuits. Don says, “I was offered the Director of Engineering position, I refused. Pressure was applied and I reluctantly agreed to take the position.” Don said the group was consuming lots of money but not producing much at first. He added, “Urie’s staff meetings were a riot, …these engineers were out of control, undermining each other and trying to figure out why the limited output of circuits suffered from a failure mode no one understood.”

He said, “a technician named Ed Guerra, also from Texas Instruments, found the answer in a technical paper published by IBM, but what did a technician know? Eventually, Guerra’s findings solved the problem.”

Don says that Urie was an extremely anal person and a clean freak. Apparently, Urie invited Hartley to his home (Jack Hartley, the division VP at the time). “It had been raining and he told Hartley to clean his shoes on an outside carpet. When Hartley came inside, Urie pulled a four barrel .22 caliber Derringer from his pocket and shoved it in his face. That resulted in Urie being involuntarily committed. When he was released, he would call me at all hours of the night and threaten suicide unless I came and had a brandy with him.”

Don tells another story about Urie and his ex-wife that ends with Urie asking her to bring him and Don a glass of brandy.

“She handed each of us a drink and sat down. He looked at his glass, looked at mine and yelled, ‘Bring me Don’s glass.’ She did. He held the glasses up side by side and screamed, ‘You f***king bitch you gave him more than me!’”

Eventually, they found that Urie didn’t have a Ph.D. as claimed, among other things. Urie left Radiation and joined his friend Dr. Arnie Lesk at Motorola in Phoenix. “One day Urie told one of his sons to clean the pool. His son resisted. Urie insisted, apparently strongly. The boy went in the bedroom, retrieved a shotgun, came out and blew Urie away.”

But let’s not let Peter Petroff or Urie Davidsohn affect the story. Let’s just say that Dr. Boyd decided to start a semiconductor group and the first director hired Clyde Combs away from Texas Instruments.

Clyde Combs was the son of sharecropper in North Carolina. He worked the oilfields of Texas and then joined the Air Force. He was stationed out of West Palm Beach, Florida, during the Korean War. After the service, he went to college at Southern Methodist University and worked as a co-op student at Texas Instruments on Lemmon Avenue, its first manufacturing site. This was 1955. For context, remember that in early 1952 Texas Instruments purchased a patent license to produce germanium transistors from Western Electric Co. for $25,000. Then in April 1954, Gordon Teal (formerly of Bell Labs) created the first commercial silicon transistor at TI. In 1954, TI designed and manufactured the first transistor radio.

Clyde worked on a little bit of everything, including helping Jack Kilby. The engineers took courses in semiconductor processing, taught by Walt Runyan. Walt was a former T.I. engineer who moved to academia and for many years Walt’s text was the only one on the subject. Each engineer was assigned a chapter, Clyde took diffusion. At the time, they had grown junction and mesa transistors. Fairchild developed the planar process but TI stayed with mesa for awhile. Clyde worked with Al Stein on “special projects” and that’s how they got TI in the planar technology. Al was also an expert in EPI technology through his research at Arizona State University.

Clyde heard about the group at Radiation in Florida through his “engineering network.” Having grown up in North Carolina and served in the Air Force in West Palm Beach, he was interested in moving. He joined Radiation in 1963 before there was anything more than office space for the group. He set up the whole process, all the fabrication equipment, diffusion ovens, etc. Other early members of the team were Phil Garner for assembly and test, Dane MacNeil for thin-film, Ernie Poplan for administration. These were the first four guys. Later, Clyde brought many of his colleagues over from T.I., including John Short and Ed Guerra. A bit later on, Don Sorchych was brought in as the first circuit guy.

I found an old issue of Microelectronic News by Don Hoefler with a spin-off tree. It went back further than Fairchild, and included Bell Labs, T.I., Hughes Aircraft and Motorola among others. It showed Radiation, Inc. and listed Sorchych, Combs and Garner and implied that they left T.I. to start Radiation in 1963. Well, obviously, Radiation started before 1963 but that’s when the semiconductor group started. Clyde Combs was from T.I. but not the others.

Clyde had quite a few stories including when his boss, Urie, would pull out his Derringer and put it on the table during meetings. Urie had lots of ideas but didn’t fully understand the technology. But he was the boss, so they tried to implement his ideas. Clyde wanted to build mixed signal circuits but worried about keeping the tubs isolated. So he had some logic wafers thinned, then cut out a single dice. He did the same with an analog circuit. Then he deposited carbon dioxide on a new wafer, planarized it and attached the two die. Ultimately, this led to the development of what became Radiation’s (and therefore Harris’) dielectrically isolated process.  Urie, the boss, put his name on the patent but Clyde did much of the work along with John Short and Ed Guerra. Years later, Dick Casey from General Electric suggested that the process should work well for radiation-hardened circuits for space applications.

To reinforce how small the world is, Clyde also told me about the time he interviewed with Charlie Sporck and Bob Widlar at National Semiconductor. Widlar had his ax in the office. He told Clyde that when he got frustrated with a problem, he’d take the ax up into the woods and chop logs. Earlier, he had interviewed with Andy Grove and Gordon Moore at Fairchild. Clyde did not join Fairchild or National.
Clyde's notes on the EPI reactor, note the actual wafers!

The D-I process helped Radiation, and later Harris, win many government projects. The Navy was a big customer with Posiden, Trident, Peacekeeper and Minuteman missile programs. They did 34 circuits for the W71 program (familiar if you’ve seen “The Falcon and the Snowman”; W71 was history's first completely operational Early Warning and Surveillance Satellite, at the time it was so classified they didn't even give it a name. It is now called the Defense Support Program and is the program that Reagan used to scare the crap out of Gorbachev).

In 1964 Jack Hartley was chosen to run the physical electronics department, later elevated to be called the Microelectronics Division. "Radiation was by far the smallest company – by a factor of 10 – to enter into the integrated circuit business," he says. The technology was difficult and demanding. Initially the company wanted the circuits to support its digital telemetry products. On Jack's recommendation, the company expanded operations to become a merchant supplier, providing integrated circuits to other companies. Radiation developed early programmable devices and the dielectrically-isolated process generated a lot of revenue for radiation-hardened devices for space and defense applications.

Radiation invested heavily in semiconductors, but it evolved into a stand-alone business and not an integral part of the “systems” businesses. In the early years of the Microelectronics Division, “systems” projects were pushed to design-in circuits that were being developed by internally. It was not an easy sell for management, because the “systems” guys had to choose the optimum device for their project. Only rarely, did they have to resort to internally-developed custom circuits to accomplish their goals.

Now let’s return to 1967, when Radiation merged with Harris-Intertype. (Notice that I said “merged” because we’re now looking at it from the Radiation perspective, compared to “acquisition” from the Harris perspective.) Radiation wanted the merger because it was felt that major investments in integrated circuit manufacturing capabilities were needed to remain competitive, but that the investment would stress the company. Harris-Intertype wanted the merger because they wanted to branch out into communications. The name Radiation persisted as a division of Harris-Intertype through 1974. The merger was eventually reasonably successful, although the organizations had quite different cultures.

The systems side of Harris had too many mergers and acquisitions to mention. On the semiconductors side, more acquisitions occurred. In 1988 Harris acquired General Electric’s semiconductor operations, which included GE’s prior acquisition of the RCA solid-state business and its 1981 acquisition of Jean Hoerni’s and Jack Gifford’s Intersil. 1988 happened to be the year I entered the semiconductor industry at National Semiconductor. I distinctly remember having one of the Harris/RCA/GE/Intersil data books and thinking it looked like a crazy consolidation of a company. I’m told that the RCA and GE-Solid State acquisitions were useful for their patents but the products were dogs. The Intersil acquisition was very good and a nice complement to the military-focused Harris products.

In 1999 the entire semiconductor unit was spun off under the name Intersil.

As expected, I found a very interesting story with no direct path to Silicon Valley or MIT. There were some mentions of Jack Kilby, Bell Labs and Vannevar Bush, but they were not crucial to the story. Radiation, Incorporated was mostly a home-grown enterprise. And I found a strong sense of loyalty to the old Radiation group within Harris. It has very much a family feel, as you might expect. If you are one of the loyal Radiation people or a Radiation memorabilia collector, go visit and contact Frank Perkins.  And if I've gotten anything wrong or left something out, please let me know.


Radiation, Inc. (wow!)

My View, Don Sorchych

My View, Don Sorchych

Wikipedia Times-Picayune, “Newhouse News Story on Sputnik: Beeping ball lights fire that launches U.S. to moon” Sputnik's ascent pushed Americans 50 years ago Wednesday, October 03, 2007 By Kent Faulk


  1. From a recent email on another topic, Clyde Combs gave me this description of the development of the dielectric isolation process at Radiation Inc.:

    A starting wafer of digital starting material, used for digital transistors, was lapped and polished to a thickness of 2 mils. The same was done for an analog slice. Next, we diced up the respective wafers into squares. The squares were then placed on a handle wafer which was oxide coated and
    put into an epi reactor and over coated with CO2 silicon dioxide which held the squares in place. As the surface was rough, we then again lapped and polished the slice leaving analog and digital islands which were oxide isolated. Note this was originally done to be able to build analog and digital transistors side by side such that the diffusion of dopants for further processing could be contained
    within the respective island.

    This isolation technology was originally fabricated as a diffusion barrier, however we soon found that this would also have important advantages in radiation environments. After the dielectric isolation operation, we used standard wafer/photoresist processing techniques to successfully build analog and digital transistors side by side. In a five year time frame Radiation, Incorporated became the #1 supplier of radiation hardened integrated circuits for strategic market. During this time Radiation, Incorporated was competing against the major semiconductor companies of the time: Texas Instruments,
    Fairchild, Motorola, and RCA. Starting from this concept stage, there were many days of hard engineering work to make this a cost-effective process for the manufacturing of integrated circuits for military and commercial applications.

    After many discussions with other engineers both outside and within our department, I realized the potential of this isolation concept and decided to file a patent on this idea. After consulting with Radiation's lawyers, George Lane and Dub Hudson, I filed a patent request with myself, Ed Guerra, and John Short as inventors. (I added Ed and John to the patent as they were key
    in helping me to develop the first model.) I then took this application to my boss, Ury Davidon, and he demanded that if I wanted to keep my job at Radiation, he was going to be first on the patent. At that time I decided to let the patent go and not argue the point.

  2. This comment has been removed by a blog administrator.

  3. Interesting

    Paul Reinecke
    Radiation - Harris SC - Intersil: 1969 - 2009

    MED Materials Dept Bldg 4, 6
    Fab 54 Yield Enhancement
    Military Space World-Wide Planning