Microwaves in the Land of Sun and Honey

Gilfillan Brothers

In 1941, Gilfillan Brothers, a Los Angeles manufacturer of radios with an excellent reputation for efficiency, was challenged to produce the Ground Approach Radar developed by MIT.

Figure 1 The AN/MPN-1, the world's first GCA, was Gilfillan's entry into radar.

Gilfillan was contracted in the early days of the program to build ten of these pre-production units, operating under the supervision of Dr. Luis Alvarez of MIT. Alvarez, a pilot, got the idea that if you could use radar to track an airplane to shoot it down, you could track it for landing. The resulting AN/MPN-1 Radar (see Figure 1) with its Airport Surveillance Radar (ASR) at S-band and Precision Approach Radar (PAR) at X-band set the military air traffic control standards.The system was simple, direct, and it worked well, even with previously untrained pilots. It was so successful that the military continued to use it for many years after the war, and it is in use in some countries even today.¹ While a great contributor to the winning of WWII, the GCA became famous for its ability to guide planes to a successful landing during the Russian Blockade of Berlin.

The first working system was called the Mark I. As soon as engineers in Boston got it working and demonstrated, bugs were identified and fixed and the information was passed along to Gilfillan in California. Some Gilfillan people were in Boston for the construction of the Mark I GCA. Among them was Homer Tasker, the chief engineer of Gilfillan, who brought two or three of the regular Gilfillan engineers to work hand-in-glove with the MIT crew.

The pre-production demonstrations held in Washington in February of 1943 were successful. The Army Air Corps then placed an order for 100 production units. Later on the Navy went to Bendix for their production and the Air Corps decided that they needed more than Gilfillan could produce so they also contracted with ITT in New Jersey. In 1962, Gilfillan Brothers Inc.'s name was changed to Gilfillan Corp. On January 30, 1964, all of the assets were transferred to ITT Gilfillan, a wholly owned subsidiary of ITT Corp.² From this auspicious beginning, Gilfillan refined and developed other GCAs and Radar Approach Control (RAPCON) systems over the next 20 years.

During the 1960s, Gilfillan entered the air defense arena by expanding its microwave engineering expertise and its products. The company also developed long-range, three-dimensional radar produced ship-borne and land-based planar array radars (one of the first), both at S-band. Gilfillan also adapted planar arrays for air traffic control (ATC). In the 1980s, Gilfillan further expanded its product line with the development and production of the AN/TLQ-32 Anti-radiation Missile Decoy (ARM-D). Today, ITT-Gilfillan continues to develop microwave components such as low-loss combiners, high power RF amplifier modules and solid-state transmitters for a variety of customers, preserving its legacy as an important Los Angeles company.³

Some of the other early companies like Hoffman Radio (1949-1970) and Packard Bell (1948-1974) were radio companies and were asked to just "up the frequency." Of course, it was natural for these companies to be in the Los Angeles area to serve the broadcast and movie industry, both of which had interests in improving sound and transmission; for example, Gilfillan's Homer Tasker had been Associate Head of Sound at Paramount Pictures.

Hughes Aircraft

Was Howard Hughes (see Figure 2) the catalyst for the microwave industry in Southern California? He gave it one hell of a push, but during WWII he was building the world's largest flying object, the Spruce Goose, which had very little microwave content. It was only after the war that Hughes was awarded a contract to build the M1 fire control system for the F86, which led to Hughes's prominence as a bastion of microwave technology.

Hughes Aircraft was involved in the initial development of a guided air-to-air missile, which began in 1946. The company was awarded a contract for a subsonic missile under the project designation MX-798, which soon gave way to the supersonic MX-904 in 1947. The original purpose of the weapon was as a self-defense weapon for bomber aircraft, but after 1950 it was decided that it should arm fighter aircraft instead, particularly in the intercept role. The missile, known as the Falcon, entered service in 1956 with semi-active radar horning (SARH) and a range of about five miles.

Hughes had a facility in the swamps of Culver City, which had a runway at which the Air Force could test its flight systems and it became Hughes Radar. Eventually, there were a number of organizations with Hughes in their names: Ground Systems in Fullerton, Space and Communication in El Segundo and Electrodynamics in Torrance. At one time, Hughes Aircraft was the largest employer in Los Angeles County. When asked how many people worked at Hughes Aircraft, a known eccentric reportedly replied, "About half of them."
The Hughes Divisions generally worked for individual branches of the military with Radar Handling Air Force work, Ground Systems, Navy jobs and Space and Communications, Comsat and NASA projects.

Hughes, Space & Communications

In October, 1957, Russia successfully launched Sputnik, the first man-made object in Earth orbit. In April 1961, Hughes was awarded its first space contract. Programs with familiar names followed: Surveyor, the lunar soft lander, which provided data for the moon landing; Syncom, the world's first synchronous satellite, which brought live TV from the 1964 Tokyo Olympics; Early Bird, a commercial satellite for the international consortium Intelsat; and ATS-1, which provided a satellite view of Earth's weather, a boon to the TV weatherman. Later programs led to blanketing the Earth's surface with satellite reception. In 2000, GM sold S & C to Boeing.

Bendix

The Bendix radio division was born in 1937 to make radio transmitter/receivers for aircraft and other types of avionics (aviation electronics). During the war, Bendix manufactured about three quarters of all avionics in American aircraft. During and after the war, Bendix made radar equipment of all kinds. Radar developments at Bendix Pacific between 1951 and 1959 included the APS-42A X-band search radar, the APS-55 tail warning radar, the Sparrow II missile, K-band active seeker; The Talos Terminal Guidance Module; and the Position Indicating Beacon Antenna (Ku-band) that was mounted high on the aircraft's vertical tail fin.

By 1955, the major system companies in Southern California included Hughes, Gilfillan, Bendix in North Hollywood, Hoffman in Los Angeles, General Dynamics/Pomona and RCA in West Los Angeles. Rockwell's Autonetics in Anaheim and Ford's Aeronutronic were established next followed in size by Ryan Aeronautical in San Diego (building Ku-band gear) and the General Dynamics Kearny Mesa facility (building the Atlas missile); General Dynamics Convair next to Lindbergh Field was also in the microwave business.

Rockwell Autonetics

On 27 June 1966, Rockwell Autonetics won the contract for the Mark II Avionics system, which consisted of an attack radar, a navigation-attack system, and a lead computing optical sight to be installed on the F-111's A, C and E models. A fixed-price incentive contract with a target price of $145 M was signed with the prime contractor, General Dynamics. Aggressive competition had led bidders to be unduly optimistic. Autonetics delivered the first Mark IIs to General Dynamics on 21 November 1967. Flight testing by an F-111A began on 31 March 1968. Then the "unanticipated unknowns" appeared. During the first full system test, in June, components started interfering with each other. By late 1969, the Mark IIs' snowballing cost reduced the F-111D program from 315 to 96 and component costs swelled as mass production slumped. By June 1972 only 24 F-111Ds were available—two years beyond the time when a 72-plane wing should have been operationally ready. Autonetics' strategy of buying in, accomplished with the complicity of its backers in the Air Force and OSD, resulted in technical problems, remedial cures, and left the F-111D with a complex, highly integrated, one-of-a-kind avionics system.⁶

Ford Aeronutronic

Aeronutronic was a defense and space related division of Ford Motor Co. set up in 1956 in Newport Beach. In 1961 Ford purchased Philco and merged the two companies in 1963. Philco Aeronutronics became NASA's primary communications equipment vendor during the 1960s. Many portions of the Philco side of the company were sold off in the 1970s, until in 1975 all that was left was the original Aeronutronic divisions. These were renamed Ford Aerospace and Communications Corp. in December 1976 and then again to Ford Aerospace Corp. in January 1988. In October 1990 what remained was sold to Loral to become Loral Aeronutronic, before eventually disappearing when Loral was purchased by Lockheed Martin in 1997.

Ryan Aeronautical Co.

The Ryan Aeronautical Co. was founded by T. Claude Ryan in 1934 and became part of Teledyne after 1969. Northrop Grumman purchased Teledyne Ryan in 1999. Ryan built several historically and technically significant aircraft, including two famous V/STOL designs, but its most successful production aircraft would be the Ryan Firebee line of unmanned drones used as targets and unmanned air vehicles.

TRW

The Ramo-Woolridge Corp. was founded by executives who split from Hughes. It was located in nearby Redondo Beach and many Hughes engineers moved to the new enterprise. In 1958, the company merged with Thompson Products, originally called the Cleveland Cap Screw Co., founded in 1901. The new company, Thompson Ramo Wooldridge Inc., shortened its name to TRW Inc. in 1965. TRW Inc. was active in the early development of missile systems and spacecraft, most notably the NASA deep space satellites Pioneer 10 and 11, which sent information back to Earth. TRW was also active in high frequency MMIC devices including GaAs device fabrication. In February 2002 Northrop Grumman launched a $5.9 B hostile bid for TRW. A bidding war between Northrop Grumman, BAE Systems and General Dynamics ended on July 1, 2002, when Northrop's increased bid of $7.8 B (£5.1 B) was accepted.⁵

During the MIMIC Program, TRW ended up being the sole developer of an epitaxial GaAs MMIC, a HBT grown by production quality Molecular Beam Epitaxy (MBE). RF Micro Devices was one of the first to recognize the commercial potential of TRW's GaAs HBTs resulting from the MIMIC program. Early on in the epi game, TRW and RFMD teamed together (RFMD doing the design, test and marketing; TRW doing foundry and fab). In 1993, RFMD figured out not only how to design for manufacture to the TRW process, they also figured out how to sell the component for under $4.⁷

In 1992, the TRW space and electronics MMIC products office merged with Northrup Grumman Space Technology (NGST). In May 2001, a separate operating unit, Velocium, was launched. Today, NGST of Redondo Beach licenses its Velocium MMIC product line and related intellectual property through Hittite Microwave Corp. NGST continues to offer foundry services for a wide array of GaAs and InP technologies, and markets off-the-shelf and custom products above 86 GHz addressing new emerging markets including millimeter-wave imaging, sensor and communication applications.

Rockwell

Rockwell International developed a desktop calculator based on a MOSFET chip for use by its engineers. In 1967 Rockwell set up its own manufacturing plant to produce them, starting what would become Rockwell Semiconductor. Rockwell Semiconductor was then later spun-off from Rockwell in January 1999 and then became known as Conexant (which combined with Alpha Industries to become Skyworks), which spun-off the foundry as Jazz Semiconductor in 2002 (since acquired by Tower Semiconductor). Tower/Jazz is now a pure-play foundry in Newport Beach. The company does not process GaAs chips, offering signal processing ICs for broadband communication systems instead.

The systems companies spawned a number of Southern California component companies, including Micromega, CableAml and GammaF. Other component companies in the area that were established to support the system monsters include Connecting Devices, Daico (established in 1965, producing advanced IF/RF and microwave control products and amplifiers for the defense electronics, aerospace, commercial aircraft and wireless industries), Datcom and Rantec (specializing in waveguide slot arrays and reflector-feed antennas for land, air, sea and space applications, founded in 1957).

Amplica

Founded in 1972 by Chuck Abronson and Jim Cole, with funding from personal savings, family and friends, Amplica grew rapidly and profitably, with initial sales coming from the defense electronics market. The commercial availability of GaAs FETs a few years after the company started enabled an entirely new class of ultra low noise amplifiers (LNA), which very quickly rendered the only competing LNA technologies, tunnel diode and parametric amplifiers, obsolete. During 1974, Abronson modified a GaAs FET LNA that was originally designed for a military radar application to work over the commercial 3.7 to 4.2 GHz TVRO (television receive only) satellite downlink band.

This amplifier outperformed a cooled parametric amplifier in an early TVRO system, which was the first demonstrable commercial GaAs FET LNA for this application. The result was Amplica's entry into the highly profitable, rapidly growing commercial 3.7 to 4.2 GHz TVRO LNA market. Initially the sole producer of such amplifiers, Amplica dominated the LNA market for about the next eight years and went on to become a leading manufacturer of complete TVRO systems for commercial and consumer applications. After a highly successful IPO in 1981, the company merged with COMSAT in 1982.

Test Instruments

Maury Microwave
Mario A. Maury, Sr. founded Maury and Associates in Montclair in 1957. With the help of sons, Mario A. Maury, Jr., and Marc A. Maury, the company focused on the microwave test and calibration industry, developing a comprehensive line of precision instruments, coaxial and waveguide components, and support products.

Simulation Software
According to Jeff Stitt, Hughes engineers in the early ‘60s used slide rules for most calculations. If the program could afford it, they submitted stacks of cards to IBM computers for processing FORTRAN programs that typically ran over night. This was typically reserved for the computation of the performance of various phased array configurations and the design of microwave filters. It was expensive and slow.¹¹

This changed rapidly in the late '60s and early '70s. Hughes signed up for GE time sharing services and started running BASIC programs from punched paper tape to compute the performance of what became the next big thing: Bipolar RF low noise amplifiers. They developed amplifier circuits using home-brewed CAD based on optimization routines, circuit models and transistor data. Hughes's home-brewed CAD eventually gave way to the hot new (at the time) commercial microwave computer-aided-design software from COMPACT and, later on, from EEsof.

EEsof

In 1983, Intel announced a math coprocessor for its IBM-compatible PCs that made fast floating-point calculations practical; this became the first computing platform for EEsof products. EEsof was co-founded by Bill Childs and Chuck Abronson in September 1983 and EEsof introduced its first product, Touchstone, at the 1984 MTT-S in San Francisco. A year later at the MTT-S, EEsof introduced a Touchstone interface to the HP Vector Network Analyzer, which was the first VNA computer interface capable of controlling the instrument to interactively extract and de-embed S-parameters (Touchstone remained the program of choice for such applications for many years).

In the late 1980s Hewlett Packard Co., after first licensing and distributing EEsof's products for sale on their proprietary desktop computers, introduced a product line that directly competed with EEsof. In 1992, Abronson approached HP with a proposal to merge their software operation into EEsof as an independent, standalone company. HP countered with an offer to acquire EEsof and the company was acquired by HP in 1993. Spun out of HP as part of Agilent Technologies Inc. IPO in 1999, EEsof continues today as a successful part of Agilent. In 1997 Abronson started CAP Wireless with Paul Daughenbaugh, the fourth employee to join Amplica. Today, Abronson is Managing General Partner of Accordance Ventures, LP, which holds investments in venture capital partnerships, real estate and a number of private technology companies including Southern California software provider, Applied Wave Research (AWR).

Many noteworthy chapters of microwave history are set in Northern California. Early Bay Area electronics companies are characterized by local venture capitalists; close ties between industry and research universities; a product mix of electronic components, production equipment, advanced communications, instrumentation, and military electronics; inter-firm cooperation; and a tolerance for spin-offs.

The semiconductor and computer industries of Silicon Valley also made this region a hotbed for the development of microwave active components, test equipment, and simulation software. While many associate the beginning of Silicon Valley with the founding of the Shockley Transistor Corp. in 1955, the history of technology in the Bay Area actually goes back to the early days of radio development.

The Federal Telegraph Co.

In 1908, Stanford graduate Cyril Elwell was having trouble getting a spark-based radio telegraph system to work using either a spark transmitter or an alternator. Contacting Dr. Vladimir Poulsen in Copenhagen, Elwell sought the US patent rights for his arc transmitter invention, which could transmit speech messages ten miles, and telegraph signals 180 miles. Elwell then approached Stanford president David Starr Jordan and C.D. Marx, the head of Stanford's civil engineering department, about financing a new company to provide wireless telephone and telegraph services on the Pacific Coast using this technology. Jordan invested $500, establishing a tradition of investing in new technologies by the university.

The company built a small system and invited the public to demonstrations of wireless voice and telegraph communication between Stockton and Sacramento. One year later, with Elwell as the technical leader, the company was renamed the Federal Telegraph Co.

In 1912, Elwell travelled to Washington, D.C. to generate interest in the Poulsen arc at the Navy. The system maintained contact with outbound Navy ships in the Florida Keys long after the transmissions from the competitor's unit faded to nothing. The Navy immediately ordered ten 30 kw arc transmitters for shipboard use.⁸

The Navy's demand for more powerful arc transmitters soon exceeded FTC's technical capabilities. In 1913, a 100 kW unit was requested for the first station in a "high-powered chain" that was to extend southward from Arlington to the Panama Canal Zone and westward to the Philippines. When the United States entered World War I, the entire radio industry was nationalized. FTC received orders for 300 two kW shipboard transmitters, 30 kW battleship transmitters, a 20 kW set for a cruiser and a 5 kW set for a Navy collier. The war work at FTC culminated in the installation of a pair of 1000 kW transmitters at the Lafayette Radio Station, 14 miles southwest of Bordeaux, France. Upon completion in January, 1920, the station was by far the most powerful in the world and had cost $3.5 M to build.⁹

However, the high-powered arcs emitted strong harmonic radio frequencies that interfered with smaller stations. During the war vacuum tubes had been developed that could generate high-power, "short-wave" radio signals that overcame many of the static problems that plagued the long-wave systems. By the war's end, vacuum tubes had been improved to the point where they could be successfully applied to all aspects of radio communications: transmission, reception and signal amplification. It was only a matter of time before they came to dominate the radio industry.

DeForest Develops the Tube Amplifier

In 1910 Lee DeForest came to San Francisco to supervise the installation of wireless telegraph sets on two Army transport ships. The receivers used a vacuum tube, the "audion," invented by DeForest in 1906 and patented in 1907. While DeForest was in San Francisco, he met local radio enthusiasts, including Elwell, who was able to provide him with a laboratory, two assistants and free rein to develop his ideas.⁸ At FTC, DeForest tackled the problem of amplifying incoming telegraph signals to the level at which they could be better received by FTC's "rotary ticker," which sent audible signals to the operator's headset. Within a few months, DeForest and his assistants had invented a three-element vacuum tube that was capable of amplifying the faint electrical signals from long distance telephone and radio transmissions. A few months later, the DeForest team found that the tube could also function as an oscillator.⁹

The vacuum tubes that could be applied to all three stages of wireless radio communication: signal generation (the oscillator), signal reception (the audion) and signal amplification (the amplifier). Because the amplifier could boost weak signals, high-power transmission eventually became less crucial and the cost of long distance wireless systems was radically lowered. That same year, DeForest would hand over the patent rights to AT&T for a mere $50,000.¹⁰

An elaborate patent sharing arrangement between RCA, Westinghouse and GE, along with AT&T and United Fruit, created a virtual patent monopoly that locked out many home-grown electronic companies around the US. The early electronics industry in the Bay Area labored under constant threat of RCA litigation. If the cooperative nature of Bay Area electronics companies during the 1920s, 1930s and 1940s had any one source, it was in opposing the domination of the field by RCA. It also steered Bay Area companies away from consumer electronics and towards specialization in electronic instruments, military electronics, advanced communications technologies, electronic components and production equipment.⁸

The vacuum tube industry in the San Francisco Bay Area, while smaller than the captive operations of the large East Coast firms, was nevertheless an important one. The activities of Elwell, Fuller, DeForest, Farnsworth, Litton, Heintz and Kaufman (H&K), and Eitel and McCullough (Eimac, which was acquired by Varian in 1964), reveal an unbroken lineage in leading-edge electronic component design and production in the San Francisco Bay Area that preceded the region's semiconductor industry by nearly four decades.
Federal's transition from arcs to tubes spanned the years 1927 to 1934. A key figure in that era was engineer Charles V. Litton, a Stanford graduate who became fascinated by tube-blowing as a teenager. In 1931 Federal Telegraph's manufacturing operations were moved to New Jersey after company officials decided Palo Alto was too far away from sources of supplies, skilled labor and key markets. ITT eventually swallowed the company. Among the Federal employees who decided not to go east was Charlie Litton.¹¹

Litton Industries

In 1928 Charles Litton, with degrees in electrical and mechanical engineering from Stanford University, was hired to manage FTC's in-house vacuum tube manufacturing department. Litton built his first ham radio set at the age of 10 and soon made his own vacuum tubes, which he sold to other hams. He even established voice communications with stations as far away as Australia and New Zealand.

When FTC moved to New Jersey in 1932, Litton formed Litton Engineering Laboratories to design and manufacture vacuum tube production equipment. Litton's glass-blowing lathe was able to mass produce glass tube blanks at uniform quality, a huge improvement over the hand-blown blanks of the day. These machines were unique and were used for mass production by virtually all major vacuum tube makers, including GE, Westinghouse and RCA.

In 1940 Litton was the sole source manufacturer of large, high-powered "magnetron" vacuum tubes for ground-based radar systems. Although these tubes, some of which stood four feet high, were highly sought after by the US military, Litton began by fabricating them from scratch in his backyard. During World War II, tube production expanded and, as Litton said, "I woke up one day, and out of the clear blue sky... found myself the sole owner of a million-and-a-half-dollar concern." In 1946, Litton separated the tube business from his research laboratory and machinery business.

In 1953 Litton sold his tube business to former Hughes Aircraft executive Charles "Tex" Thornton and moved his laboratory to Grass Valley, east of the Sierra Nevada Mountains.¹² Thornton's plan was to grow the company, initially dubbed the Electro Dynamics Corp., into a diversified giant through the acquisition of small, innovative electronics companies. The company's first acquisition was Litton Industries, which reached $3 M in sales from Litton's state-of-the-art magnetron tubes in the first year. When he found that the name of Charles Litton carried a great deal of weight with the Navy, Thornton changed the company's name to Litton Industries. By 1959, Litton Industries reached $120 M in sales manufacturing inertial guidance systems for aircraft, duplexers, klystrons and other electronic products. During this period, almost 50 percent of Litton's business was with the US government. By 1961, Litton was the fastest growing company on the New York Stock Exchange. By 1980 Litton Industries had grown to $4.2 B in annual sales.⁸

The Brothers Varian

Russell and Sigurd Varian moved with their family to Halcyon, CA, in 1914. Russell attended Stanford University, switching from the social studies program to physics due to his poor reading skills. Also somewhat weak in math, he took six years to graduate. He returned to Stanford later that year to begin graduate work, receiving his master's degree in 1927. Varian had a succession of jobs, including working on early television systems for Philo T. Farnsworth and Co. In 1935, Stanford University rejected his application for its PhD program because of his poor math and reading skills, and so he began working with his brother instead.

Meanwhile, Sigurd had become a flying enthusiast and barnstorming pilot. The interest spread to Russell and soon both men began to work on ideas to improve the primitive navigation equipment then available. Attempting to develop a radio compass, Russell got the idea for an "electron tube" capable of directing a beam of electrons, which could be used in a number of different applications, including the detection of airplanes by high-frequency radio signal. Approaching Stanford with Russell's concepts, Sigurd was able to obtain use of lab space in the basement of the old physics building, some equipment, and $100 in materials.¹³

William W. Hansen entered Stanford at the age of 16. Receiving his doctorate in 1933, Hansen invented the cavity resonator and began investigating the use of high-frequency waves to accelerate particles to high energy. The Varian brothers, who were in the physics department's basement generating very-high frequency short wavelength signals for radar and direction finding, began working with Hansen. By the summer of 1937 the first klystron tube, as it was called, had been constructed, and it was formally introduced in the Journal of Physics in 1939.¹⁴ The British lost no time in adapting klystron technology for use in radar stations, which helped defeat the Luftwaffe's air raids on London in the summer of 1940.

The klystron represents one of Stanford's best investments: $100 in seed money and use of a small laboratory room were turned into $2.56 M in licensing fees before the patents expired in the 1970s, three major campus buildings and hundreds of thousands of dollars in research funding.

The Stanford Impact

Hansen pioneered much of the development of microwave theory and techniques for testing microwave systems, giving courses on microwave theory at Stanford (and MIT RAD Lab) during World War II to physicists who had been recruited for research on the subject. After the war, Hansen founded Stanford's Microwave Laboratory to develop powerful klystrons and linear accelerators, which has since been named after him. After the war, the close relationship that existed between local industry and the university was extended through the efforts of Frederick Terman (see Figure 3).

Since its inception, Stanford had contributed to Northern California's emergence as a center of technology with high-caliber engineering graduates, research support and financial aid. Stanford's first president David Starr Jordan set this precedence with his support of Elwell and FTC. But it was former FTC employee, Terman who would practically institutionalize this practice.

Along with Hansen, Terman had also been recruited to help with war-related research, directing a staff of more than 850 at the Radio Research Laboratory at Harvard University. This organization was the source of Allied jammers to block enemy radar, tunable receivers to detect radar signals, and aluminum strips ("chaff") to produce spurious reflections on enemy radar receivers. An avid inventor, Terman filed 36 patents between 1930 and 1947 and was elected president of the Institute of Radio Engineers in 1940—the first person west of Pittsburgh to achieve the honor.

From 1925 to 1941 Terman designed a course of study and research in electronics at Stanford that focused on work with vacuum tubes, circuits and instrumentation. During his tenure, Terman greatly expanded the science, statistics and engineering departments in order to win more research grants from the Department of Defense. His students included Oswald Garrison Villard, Jr., William Hewlett and David Packard, whom he encouraged to form their own companies and personally invested in many of them, resulting in firms such as Litton Industries and Hewlett-Packard.

In 1951 he spearheaded the creation of Stanford Industrial Park (now Stanford Research Park), whereby the university leased portions of its land to high-tech firms. First aboard was Varian Associates, which obtained a park lease in 1951. Terman then convinced Hewlett-Packard to head out to Page Mill Road. Soon, a flood of other corporations such as Eastman Kodak, General Electric and Lockheed Corp. moved into what would eventually become known as Silicon Valley.

Hewlett-Packard

Bill Hewlett and Dave Packard first met at Stanford and later decided to form a company in 1938. Even though they were not microwave engineers originally, their story is important because of their contribution to the field of microwave instrumentation. Very few people realize that Hewlett and Packard were not commercially successful with some of their earlier inventions, such as a foul line indicator for bowling alleys, a weight-reducing machine that provided electrical shocks while the user was eating, and a urine flusher system.

Their first successful product, an audio generator using a light bulb to improve oscillator stability, was created from Bill Hewlett's graduate project at Stanford. Terman, who was their electrical engineering professor, recognized the significance of the innovation and suggested that they produce the generator commercially. Bill and Dave's friend, Norman Neely, persuaded a fellow engineer to use the audio generator to produce the special sound effects in Walt Disney's Fantasia. Disney went along with the idea and purchased eight of the generators.

Hewlett and Packard grew rapidly during World War II with about a hundred people on the payroll in 1946. While both founders were skilled engineers, Hewlett concentrated on engineering problems while Packard focused more on the business aspects. After World War II, when Bill Hewlett returned from active duty in the Army Signal Corps, the decision was made to enter into microwave instrumentation. Hewlett and Packard brought in two extremely capable microwave engineers: Bruce Wholey, who had graduated from Stanford with a microwave background from Radio Research Labs at Harvard; and Art Fong, a University of California (Berkeley) graduate with a microwave background from MIT RAD Labs. With the assistance of these two men, the company quickly formed a team that soon rivaled General Radio's. This team developed a line of signal generators from 50 kHz to 21 GHz as well as such components as slotted lines. In the mid-1960s, the first-generation Spectrum Analyzer, followed by the Vector Voltmeter and the Network Analyzer, introduced new measurement capabilities to microwave engineers.

Dalmo-Victor Develops an Airborne Radar Antenna

In 1921, when he was nineteen, Tomlinson Moseley established his own machine shop in San Francisco, the Dalmo Manufacturing Co. In 1934 Moseley hired an immigrant PhD Russian research engineer named Alexander Poniatoff to help with the development work. In 1944, the two began work on a prototype airborne radar antenna for the Navy. According to Poniatoff, Moseley said to him, "Do you know anything about radar?" Poniatoff replied, "I don't, not a damn thing." "Neither do I," Moseley said, "but the contract says the unit must be completed in 100 days, so you can't waste time." In the Dalmo shop in San Carlos, they worked for 100 days without a break, often sleeping in the shop. Surprisingly, they won the contract.

Westinghouse offered to manage the contract, arguing that high-volume production would be too difficult for a small company like Dalmo Manufacturing. Moseley agreed, and the company, now Dalmo-Victor, moved to a larger facility in Belmont to produce the units. By the end of World War II, Dalmo-Victor had emerged as the leading manufacturer of airborne radar antennas. Dalmo-Victor was eventually acquired by the General Instrument Defense Systems Group, which in turn was acquired in 1991 by Litton Industries Applied Technologies Group, headquartered in San Jose.

Defense Funding Creates a Microwave Eco-System

With World War II and the Cold War came technological breakthroughs and military programs, first enabled by the reflex klystron and later by traveling wave tubes and eventually microwave semiconductor devices. This second, "Microwave" era lasted from the late '30s to the mid-'80s and spawned companies such as Varian, Applied Technology, Watkins-Johnson and Avantek.

The number of companies working with the three branches of the armed services, DARPA, NASA and the intelligence agencies would expand throughout the Cold War. Funding included basic research and microwave component development efforts. Perhaps the most well-known R&D program was the filter technology development at Stanford Research Institute (SRI), which resulted in the classic publication by Matthaei, Young and Jones sponsored by the US Army Electronic Research Laboratory.

Competition for funding was intense, particularly for larger programs associated with major military systems. For example, each new aircraft would typically require a specialized avionics suite that resulted in large development and manufacturing programs. Many of the Bay Area companies had core expertise in certain technologies that resulted in niche products that were sold to larger companies who specialized in turn-key systems.¹² This specialization helped support an eco-system of component manufacturers (i.e. Varian Associates, Watkins-Johnson Co. and Avantek), system integrators (i.e. Lockheed, Ford Aerospace, Applied Technology, ESL) and supporting test equipment providers who were diversified by commercial markets (i.e. HP, Wiltron). The emergence of solid-state technology in the Bay Area would contribute to defense funding and, in turn, would benefit from it.

Watkins-Johnson

Dean Watkins was working as a professor of electrical engineering at Stanford University. Richard Johnson was head of Hughes Aircraft Co. microwave laboratory (Southern California) where the two men met while working on TWTs. They both wanted to own their business, but they did not have any money. Once again, Stanford's Terman intervened by arranging financing through a local company that put up the $900,000 capital investment to launch the company in Palo Alto.

The two founders decided in 1957 to develop and manufacture microwave tubes and microwave solid-state devices, then use those products as a foundation from which to diversify into related electronic systems and equipment devices areas. During its first year the company showed a profit, primarily making TWTs, parametric amplifiers and backward wave oscillators. The company's unique contribution was to combine tubes with power supplies, relieving the design engineer of having to address bias details. Watkins-Johnson made many wise acquisitions that allowed extremely rapid growth, purchasing Stewart Engineering for its backward-oscillator business, Communications Electronics for its receiver business and RELCOM for their mixers and the antenna lines of Granger.

TWTs were the multi-octave preamplifier of choice in the '50s thru mid-'70s. An example of a particularly sophisticated application was the SHRIKE anti-radiation missile preamplifier assembly that included three phase and amplitude matched TWTs (manufactured in significant quantity by WJ). While WJ was noted for low noise TWTs, others including Varian and MEC (Teledyne) led in high power TWT technology widely applied in jammer systems. In 2008, WJ was acquired by TriQuint Semiconductor.

Solid-State Microwave Devices

The Fairchild Semiconductor spin-off from Shockley Transistor and the "Fairchildren" that followed are widely believed to be the stimuli that set the Silicon Valley juggernaut in motion. Fairchild started developing Si RF transistors in the late 1950s, delivering "microwave sources" using transistor oscillators and varactor multipliers in 1965. The first GaAs MESFET was demonstrated at Fairchild in late 1960s; the newly formed Microwave and Optoelectronics Division began delivering GaAs devices in 1970, along with MIC components and microwave power transistors. Key employees of this new group, among others, included John Moll of the "Ebers-Moll Transistor Model," and Jim Early, developer of the "Early Effect." Fairchild's management, however, had not fully appreciated the importance of their product line and decided not to fully support the Microwave Division's pioneer work. As a result, most of the key people left to pursue other opportunities.

Hewlett-Packard also started work on GaAs MESFETs in 1969. By 1970, IBM and Fairchild research in this new compound semiconductor had produced reasonably good gain at microwave frequencies, but could not reduce the channel noise in these devices. Around this time Charles Liechti of Hewlett-Packard Laboratories recognized the impact that GaAs transistors might have on microwave systems and in 1971 led his research team to produce very impressive FET performance with low noise and high gain, and was subsequently awarded two Outstanding Paper Awards. Later, he also received the Microwave Prize with his coauthor, Bob Tillman.

A succession of Bay Area spin-offs from Fairchild and Hewlett Packard ended up serving the commercial microwave systems market with products ranging from discrete devices to hybrid microwave integrated circuits (MIC) to monolithic MICs (MMIC). A large number of intermediate-sized local spinoffs and start-ups emerged from both companies during the 1970s and 1980s, but most are no longer in business.

Avantek

In the mid-'60s, Applied Technology INC (ATI), out of Palo Alto, had been working on countermeasures. ATI generally worked on secret "black" programs and built specialized equipment in small lots. James Sterrett, a Senior Engineer at ATI, had developed a proprietary wideband UHF transistor amplifier line. ATI's Director of Marketing, 38 year-old Larry Thielen, quickly noticed that the devices were easy to sell to a wide customer base, simple to manufacture and had a large profit margin. Thielen had wanted to start his own company ever since he was Western Sales Manager at Ampex in Redwood City and now he saw his opportunity. Thielen, along with Sterrett, McVay and Seader (both ATI Engineering supervisors), announced they were leaving to form a startup called Avantek.

The company launched in 1965 with a line of solid-state RF components, quickly expanded into the microwave range, started in-house manufacture of Si transistors and GaAs FETs, and in 1972 introduced a 2 GHz digital radio product line using these devices until 1988 when this product line was sold to Telesciences (later sold to California Microwave in 1993). The company started with five people, used office furniture, their own money and no known orders. After 19 profitable years, they had 2500 people, total capitalization of approximately 380 million (19 million shares x $20/share) and have produced over 25 millionaires. They shipped $160 M in products 1984 and finished 1984 with a backlog of $100 M. By 1991, Avantek's RF devices division was acquired by Hewlett-Packard, which became part of Agilent. As of December 2005, the Avantek division of HP that spun off to Agilent Technologies was sold and morphed into Avago.

California Microwave

ATI also spawned California Microwave, which was founded by David Leeson in 1968. The company started with a product line in competition with microwave sources that originated in 1963 at the Fairchild Microwave Division, but was subsequently acquired by Frequency Sources. The main users were Bell System companies retrofitting its 4, 6 and 11 GHz electron tube systems with solid-state subassemblies, and upgrading to higher capacities. California Microwave later expanded into a variety of microwave system product lines for a variety of customers, mainly through acquisitions. After a change of management in 1997 the company started selling off business entities, changed its name to Adaptive Broadband in 1999, sold its remaining microwave systems product line in 2000 and was acquired by Moseley Associates in 2001.

Farinon

In 1958, Bill Farinon began his firm in a one-room building located in San Carlos. Farinon Electric Co. would become well established in the microwave transmission, digital switching and auxiliary telephone equipment industry.

Beginning in 1970, Farinon introduced Gunn diode local oscillators and 1 W transmitter oscillators and amplifiers into analog and digital radios for the 6, 7 and 11 GHz bands. Farinon started building and using low-noise GaAs FET amplifiers in 1975 in the 2 GHz band, and power amplifiers in 1977. An in-house MIC facility was built in 1972. Farinon spin-off DMC also built an in-house MIC facility. In 1981, Harris Corp. purchased Farinon Electric and created the Harris Farinon Division.

Teledyne

In 1960, Dr. Henry E. Singleton formed a company called Teledyne with Dr. George Kozmetsky, a colleague from Litton Industries. They each invested $225,000, money they earned from their Litton stock options. Loosely translated, Teledyne means Power through Communication. Despite an already crowded market, Singleton believed that producing semiconductors, the "basic building block of electronics," would lead to other high-technology and high-growth inventions. Their backgrounds in high technology and innovative ideas quickly paid off. The company achieved first year sales of $4.5 M and employed nearly 450 people. Second year sales of $10.5 M confirmed their success. The company's products included vacuum devices and integrated solid state microwave subassemblies for electronic warfare, satellite communication and radar applications; microelectronic modules for secure communications; high voltage connectors and cable assemblies; and contract manufacturing of military electronic assemblies.

Measurement Systems

The start of World War II turned a trickle of US government orders for electronic instruments into a stream and then a flood for HP. In 1943, HP entered the microwave field with signal generators developed for the Naval Research Laboratory and a radar-jamming device.

In 1951, HP invented the high-speed frequency counter (HP 524A), greatly reducing the time required to measure high frequencies. Radio stations used it to accurately set frequencies to comply with FCC regulations for frequency stability. Over the years, frequency counters and related products accounted for billions of dollars in revenue. By 1956, HP produced its first oscilloscopes, models 130A/150A.

The S-parameter measuring technique was introduced by Hewlett Packard, and in 1968 the first Automated Network Analyzer (ANA) was brought to market also by HP using "small" minicomputers (see Figure 4). It became an indispensable tool for the design and manufacture of microwave systems by simplified calibration and paved the way to computer-aided measurements. Cost of the system was approximately $250,000 and it required two men to move it around. Thanks to improved technology and circuit integration, a far more capable ANA is now available in a size equivalent to the polar display unit located in the upper left corner of the Figure.

The technology developed for advanced weapons systems, EW and countermeasures would produce advanced instrumentation and require advanced measurement systems. Helped by the invention of the transistor and the analog and digital circuitry that came about with the introduction of minicomputers, Northern California was home to a number of microwave measurement system milestones. Test and measurement equipment also allowed companies to apply their technology to commercial markets any time military spending waned.

Wiltron had started out in the 1960s in Morgan Hill, CA as an electronics outfit operating primarily in the defense sector. With the reduction of defense spending following the collapse of communism in the late 1980s, Wiltron had begun repositioning itself around a core of commercial wireless and wireless test systems. The company was especially strong in the mid-range-frequency sector.
Like the system/component manufacturer eco-system that evolved around the military market, specialized microwave and millimeter-wave test and control instruments operating in the Bay Area include: Giga-tronics, Phase Matrix, Advantest, DCM Industries, Krytar and XL Microwave Inc.

Simulation Software

Computerization of microwave design was relatively slow compared to the other EE design disciplines. Traditionally, microwave circuit design was more of an art than science. Designers often achieved results by tweaking, tuning and shielding circuits on the bench, instead of using a systematic analytical approach.¹⁵ In the 1960s, introductions of Hewlett-Packard's S-parameter test instruments revolutionized microwave component testing and characterization. Soon after, related S-parameter design techniques became available, using measured data directly. By the late 1960s, the newly formed timeshare industry introduced more convenient computer operating systems and programming languages for engineers. Soon after, several software packages were also developed, and became commercially available to microwave designers.

During the early 1970s, COMPACT (Computerized Optimization of Microwave Passive and Active CircuiTs)¹⁶ had gradually become the industry standard due to its speed, nodal circuit description, large transistor S-parameter database and extensive application examples. The program had been available through six international time sharing systems as well as in-house installations. However, it was not until hardware manufacturers introduced powerful minicomputers and companies began to purchase them for dedicated scientific usage in the early 1980s that computer-aided design became widely accepted among microwave circuit designers.

Compact Software merged with COMSAT in 1980 to form Comsat General Integrated Systems in Palo Alto to focus on engineering office automation.¹⁷ In addition to the second-generation version circuit optimization program, SuperCOMPACT, the company also introduced the first automated microwave circuit layout routine, AutoArt.

Wireless Systems

By the late '80s, with dramatic advances in silicon analog and digital devices, and the post-Cold War decline in defense R&D spending, the microwave industry in Northern California began to focus on the "Wireless Revolution." This era reflected an increased focus on commercial microwave applications such as cellular, satellite or wireless local-area communications. The boundaries between microwave and high-speed electronics began to blur, as clock rates of mainstream semiconductors reached the GHz realm and microwave systems began to incorporate more and more digital signal processing. Companies such as ArrayComm, Proxim, Atheros and Airgo are examples of this era. Particularly noteworthy is the high degree of interconnection among the companies in the Bay Area. In a 2006 MTT-S presentation, over 90 percent of 100+ Bay Area companies examined traced their roots to one of the others, and over 50 percent spawned at least one significant spinoff.

Wireless systems suppliers headquartered in the Bay Area or with local engineering and/or manufacturing activities included: Arraycom (1992), Alvarion (1992), Gigabit Wireless (1998, changed name to Iospan in 2000, and acquired by Intel in 2002), Aperto Networks (1999), TeleCIS (2000), Tropos Networks (2000), Airespace (2001, acquired by Cisco in 2005), Aruba Networks (2002), Meru Networks (2002), Trapeze Networks (2002), Beceem Communications (2003), Firetide (2003) and PacketHop (2003).

There were a number of additional manufacturers of point-to-point microwave systems that participated in this market. Among these are (years indicate actual system deliveries): Kebby Microwave (1964-1968), which became Farinon Microwave; Granger Associates (1970-1992) and Granger/Telettra (1986-1991); Culbertson Industries (1971-1974); and Aydin Microwave (1990-1999). While not a manufacturer of complete communications systems, Endwave, founded in 1992, evolved as a supplier of microwave and millimeter-wave subsystems to systems manufacturers in different bands ranging from about 3 to 65 GHz.

Besser Associates

After phasing out of the design software business, I felt that the universities had not adequately prepared engineering students to face practical real-life problems and founded a continuing education group in 1985. The company, Besser Associates (see Figure 5), has trained over 50,000 RF, microwave and wireless engineering professionals and managers worldwide during the past 25 years. The company's instructors are highly experienced and accomplished microwave engineers, most of whom have published extensively in their areas of expertise. Courses are available at various experience levels, ranging from novice to experts, covering circuit and system design as well as RF/microwave measurements.

Conclusion

While most people think of Silicon Valley as the home of the integrated circuit and computer, its history goes back much further to the early days of radio and is inextricably linked to the microwave industry that thrived here. A plaque at 367 Addison Ave. in Palo Alto summarizes the unique combination of entrepreneurs, innovators, venture capitalists, academia and government sponsors that made it all possible: "This garage is the birthplace of the world's first high-technology region, ‘Silicon Valley.' The idea for such a region originated with Dr. Frederick Terman, a Stanford university professor who encouraged his students to start up their own electronics companies in the area instead of joining established firms in the East. The first two students to follow his advice were William R. Hewlett and David Packard, who in 1938 began developing their first product, an audio oscillator, in this garage."

Chuck Swift, who refers to himself as "The Old Peddler", entered sales in 1955 and established his rep/distributor firm, C.W. Swift & Associates, in July, 1958. His product lines were RF and microwave connectors and components, and an occasional pistachio. Chuck attended his first IMS in 1961, and was Chairman of the 1989 Symposium in Long Beach. Awarded the N. Walter Cox Award in 1994, Chuck will be honored with a Luncheon (Chuck Roast) on Tuesday, May 25th. Look for him and his motorcycle in Anaheim.

Les Besser has worked in engineering and management capacities at the microwave divisions of HP and Fairchild, and also at Farinon Electric. He authored COMPACT and founded Compact Software, a pioneer CAD software company. Later, he formed Besser Associates, dedicated to continuing education through instructor-led short courses. He has published numerous technical articles, contributed to and co-authored several textbooks, and is a Life Fellow of the IEEE, receiving numerous awards for his many contributions to the field.

References

1. A. Goldstein, Oral-History: Interview with Kathyn and Charles Fowler, 14 June 1991, Boston, MA, IEEE Spectrum Online, IEEE Org.
2. http://vlex.com/vid/itt-gilfillan-international-telephone-telegraph-38401043.
3. J.M. Milan, "Microwaves at Gilfillan Division, ITT Industries: Past, Present and Future," ITT Industries, Gilfillan Division, Van Nuys, CA.
4. http://en.wikipedia.org/wiki/SCANFAR.
5. http://en.wikipedia.org/wiki/TRW.
6. Jane's Avionics, 1982-1983, p. 28.
7. J.A. McDonald, III-Vs Review, Vol. 10, No. 4, 1997.
8. T.J. Sturgeon, "How Silicon Valley Came to Be," in Understanding Silicon Valley: Anatomy of an Entrepreneurial Region, Stanford University Press, 2000.
9. H. Aitken, The Continuous Wave: Technology and American Radio: 1900-1932, Princeton University Press, Princeton, NJ, 1985.
10. T. Lewis, Empire of the Air: The Men Who Made Radio, HarperCollins, New York, NY, 1991.
11. W.R. MacLaurin, "Invention and Innovation in the Radio Industry," MIT Studies in Innovation, MacMillian, New York, NY, 1949.
12. E.J. Crescenzi, "Bay Area Microwave Technology for Defense: Developments in the Cold War Years," IEEE MTT-S IMS, San Francisco, CA, June 2006.
13. HTTP://hepl.stanford.edu/documents/HEPL_History_opt.pdf.
14. http://books.nap.edu/html/biomems/whansen.pdf.
15. L. Besser, "Combine S-Parameters with Time-sharing," Electronic Design, August 1968.
16. L. Besser and R. Newcomb, "A Scattering Matrix Program for High Frequency Circuit Analysis," IEEE Conference on Systems, Networks and Computers, Mexico, January 1971.
17. L. Besser and C. Abronson, "Computers as Microwave Tools," Microwaves & RF, March 1987.