Edwin Howard Armstrong

2.1. Regenerative Circuit

Armstrong began working on his first major invention while still an undergraduate at Columbia. In late 1906, Lee de Forest had invented the three-element (triode) "grid Audion" vacuum tube. At the time, the precise operation of vacuum tubes was not fully understood. De Forest's initial Audions did not have a high vacuum and would emit a blue glow at moderate plate voltages. By 1912, the principles of vacuum tube operation were better comprehended, and the potential of regenerative circuits using high-vacuum tubes became apparent.

Armstrong, having experimented with the early, temperamental Audions during his youth, was spurred by these later discoveries to gain a detailed scientific understanding of how vacuum tubes functioned. In collaboration with Professor Morecroft, he conducted comprehensive studies using an oscillograph. His breakthrough discovery was realizing that employing positive feedback (also known as "regeneration") could produce amplification hundreds of times greater than previously achieved. This amplified signals to a strength sufficient for receivers to use loudspeakers instead of just headphones. Further investigation revealed that when the feedback was increased beyond a certain level, a vacuum tube would go into oscillation, meaning it could also function as a continuous-wave radio transmitter.

Beginning in 1913, Armstrong prepared a series of comprehensive demonstrations and papers meticulously documenting his research. In late 1913, he applied for patent protection for the regenerative circuit. On October 6, 1914, US Patent 1,113,149 was issued for his discovery. Although Lee de Forest initially dismissed Armstrong's findings, beginning in 1915, de Forest filed a series of competing patent applications that largely replicated Armstrong's claims, asserting that he had discovered regeneration first. This claim was based on a notebook entry made on August 6, 1912, while working for the Federal Telegraph Company, predating Armstrong's recognized date of January 31, 1913. This led to an interference hearing at the patent office to determine priority. De Forest was not the only other inventor involved; the four competing claimants included Armstrong, de Forest, General Electric's Irving Langmuir, and Alexander Meissner, a German national whose application was seized by the Office of Alien Property Custodian during World War I.

Following the end of World War I, Armstrong retained legal representation from the law firm of Pennie, Davis, Martin and Edmonds. To finance his legal expenses, he began issuing non-transferable licenses for the use of his regenerative patents to a select group of small radio equipment firms. By November 1920, 17 companies had been licensed. These licensees paid 5% royalties on their sales, which were restricted to "amateurs and experimenters." Meanwhile, Armstrong explored options for selling the commercial rights to his work. Although the RCA was the obvious candidate, on October 5, 1920, the Westinghouse Electric & Manufacturing Company secured an option for 335.00 K USD for the commercial rights to both the regenerative and superheterodyne patents, with an additional 200.00 K USD to be paid if Armstrong prevailed in the regenerative patent dispute. Westinghouse exercised this option on November 4, 1920.

Legal proceedings related to the regeneration patent were divided into two groups of court cases. An initial court action began in 1919 when Armstrong sued de Forest's company in district court, alleging infringement of patent 1,113,149. This court ruled in Armstrong's favor on May 17, 1921. A second line of court cases, stemming from the patent office interference hearing, had a different outcome. The interference board had also sided with Armstrong, but he was unwilling to settle with de Forest for less than what he considered full compensation. Under this pressure, de Forest continued his legal defense and appealed the interference board decision to the District of Columbia district court. On May 8, 1924, that court ruled that de Forest should be considered the inventor of regeneration. Armstrong, along with much of the engineering community, was shocked by this decision, and his side appealed. Although the legal proceeding twice went before the US Supreme Court, in 1928 and 1934, he was unsuccessful in overturning the decision. This 12-year legal battle was the longest patent dispute of its time, and the Supreme Court's ruling, which upheld de Forest's patent, is now widely considered to have been based on a technical misunderstanding.

In response to the second Supreme Court decision upholding de Forest as the inventor of regeneration, Armstrong attempted to return his 1917 IRE Medal of Honor, which had been awarded "in recognition of his work and publications dealing with the action of the oscillating and non-oscillating audion." The organization's board refused to allow him, issuing a statement that it "strongly affirms the original award."

2.2. Superheterodyne Receiver

The United States entered World War I in April 1917. Later that year, Armstrong was commissioned as a captain in the U.S. Army Signal Corps and assigned to a laboratory in Paris, France, to assist in developing radio communication for the Allied war effort. He returned to the US in the autumn of 1919, having been promoted to the rank of Major. During both world wars, Armstrong granted the US military free use of his patents.

During this period, Armstrong's most significant accomplishment was the development of a "supersonic heterodyne" - soon shortened to "superheterodyne" - radio receiver circuit. This circuit made radio receivers far more sensitive and selective, and it remains extensively used today. The key feature of the superheterodyne approach is the mixing of the incoming radio signal with a locally generated signal of a different frequency within the radio set. This mixing occurs in a component called the mixer. The result is a fixed, unchanging intermediate frequency (IF) signal, which is then easily amplified and detected by subsequent circuit stages. This technology represented a major advancement over previous methods, which suffered from significant interference.

In 1919, Armstrong filed an application for a US patent for the superheterodyne circuit, which was issued the following year. This patent was subsequently sold to Westinghouse on May 22, 1920, for 335.00 K USD. The patent was challenged, triggering another patent office interference hearing. Armstrong ultimately lost this patent battle, although the outcome was less controversial than the regeneration proceedings. He reportedly respected Lucien Lévy, the French challenger, in a way he could not respect de Forest, possibly recognizing that Lévy had some legitimate claim to the invention.

The challenger was Lucien Lévy of France, who had also worked on developing Allied radio communication during World War I. He had been awarded French patents in 1917 and 1918 that covered some of the same basic ideas used in Armstrong's superheterodyne receiver. AT&T, interested in radio development primarily for point-to-point extensions of its wired telephone exchanges, purchased the US rights to Lévy's patent and contested Armstrong's grant. The subsequent court reviews continued until 1928, when the District of Columbia Court of Appeals disallowed all nine claims of Armstrong's patent, assigning priority for seven of the claims to Lévy, and one each to Ernst Alexanderson of General Electric and Burton W. Kendall of Bell Laboratories.

Although most early radio receivers used regeneration, Armstrong approached RCA's David Sarnoff, whom he had known since demonstrating his regeneration receiver in 1913, about the corporation offering superheterodynes as a superior option to the general public. The ongoing patent dispute was not a hindrance because extensive cross-licensing agreements signed in 1920 and 1921 between RCA, Westinghouse, and AT&T meant that Armstrong could freely use the Lévy patent. Superheterodyne sets were initially considered prohibitively complicated and expensive, as the initial designs required multiple tuning knobs and used nine vacuum tubes. In conjunction with RCA engineers, Armstrong developed a simpler, less costly design. RCA introduced its superheterodyne Radiola sets to the US market in early 1924, and they were an immediate success, dramatically increasing the corporation's profits. These sets were considered so valuable that RCA would not license the superheterodyne to other US companies until 1930.

2.3. Super-regeneration Circuit

The legal battle over the regeneration patent had one serendipitous outcome for Armstrong. While he was preparing apparatus to counteract a claim made by a patent attorney, he "accidentally ran into the phenomenon of super-regeneration." This involved rapidly "quenching" the vacuum-tube oscillations, which allowed him to achieve even greater levels of amplification.

A year later, in 1922, Armstrong sold his super-regeneration patent to RCA for 200.00 K USD plus 60,000 shares of corporation stock. This was later increased to 80,000 shares in payment for consulting services. This transaction made Armstrong RCA's largest shareholder, and he noted that "The sale of that invention was to net me more than the sale of the regenerative circuit and the superheterodyne combined." RCA initially envisioned selling a line of super-regenerative receivers until superheterodyne sets could be perfected for general sales, but it turned out the circuit was not selective enough to be practical for broadcast receivers.

3.1. Background: The Challenge of Radio Static

"Static" interference - extraneous noises caused by sources such as thunderstorms and electrical equipment - plagued early radio communication using amplitude modulation (AM). This problem perplexed numerous inventors who attempted, with little success, to eliminate it. Many ideas for static elimination were investigated, but none proved effective. In the mid-1920s, Armstrong began researching a solution. He initially, and unsuccessfully, attempted to resolve the problem by modifying the characteristics of AM transmissions.

One approach that had been considered was using frequency modulation (FM) transmissions. Instead of varying the strength of the carrier wave as with AM, the frequency of the carrier was changed to represent the audio signal. In 1922, John Renshaw Carson of AT&T, known for inventing Single-sideband modulation (SSB), had published a detailed mathematical analysis which concluded that FM transmissions did not provide any improvement over AM. While the Carson bandwidth rule for FM is important today, Carson's review turned out to be incomplete, as it analyzed only what is now known as "narrow-band" FM.

3.2. Principles and Advantages of FM

In early 1928, Armstrong began researching the capabilities of FM. Although others were involved in FM research at this time, he was aware of an RCA project to determine if FM shortwave transmissions were less susceptible to fading than AM. In 1931, RCA engineers constructed a successful FM shortwave link that transmitted the Schmeling-Stribling fight broadcast from California to Hawaii, and they noted at the time that the signals seemed to be less affected by static. However, the project made little further progress.

Working in secret in the basement laboratory of Columbia's Philosophy Hall, Armstrong developed "wide-band" FM, discovering significant advantages over the earlier "narrow-band" FM transmissions. In a "wide-band" FM system, the deviations of the carrier frequency are made to be much larger than the frequency of the audio signal, which can be shown to provide superior noise rejection. This meant that FM produced clearer sound with significantly less noise compared to AM. He was granted five US patents covering the basic features of the new system on December 26, 1933. Initially, the primary claim was that his FM system was effective at filtering out the noise produced in receivers by vacuum tubes.

Armstrong challenged Carson's earlier findings, and in 1936, he published a landmark paper in the Proceedings of the IRE that meticulously documented the superior capabilities of using wide-band FM. This paper would be reprinted in the August 1984 issue of Proceedings of the IEEE. A year later, a paper by Murray G. Crosby (inventor of Crosby system for FM Stereo) in the same journal provided further analysis of the wide-band FM characteristics and introduced the concept of "threshold," demonstrating that there is a superior signal-to-noise ratio when the signal is stronger than a certain level (known as the FM threshold characteristic).

3.3. Patenting and Commercialization Efforts

Armstrong had a standing agreement to give RCA the right of first refusal to his patents. In 1934, he presented his new system to RCA president David Sarnoff. Sarnoff was somewhat taken aback by its complexity, as he had hoped it would be possible to eliminate static merely by adding a simple device to existing receivers. From May 1934 until October 1935, Armstrong conducted field tests of his FM technology from an RCA laboratory located on the 85th floor of the Empire State Building in New York City. An antenna attached to the building's spire transmitted signals for distances up to 80 mile. These tests helped demonstrate FM's static-reduction and high-fidelity capabilities. However, RCA, which was heavily invested in perfecting TV broadcasting, chose not to invest in FM and instructed Armstrong to remove his equipment.

Denied the marketing and financial clout of RCA, Armstrong decided to finance his own development and form ties with smaller members of the radio industry, including Zenith and General Electric, to promote his invention. Armstrong believed that FM had the potential to replace AM stations within five years, which he promoted as a boost for the radio manufacturing industry, then suffering from the effects of the Great Depression. Making existing AM radio transmitters and receivers obsolete would necessitate that stations buy replacement transmitters and listeners purchase FM-capable receivers.

In June 1936, Armstrong gave a formal presentation of his new system at the US Federal Communications Commission (FCC) headquarters. For comparison, he played a jazz record using a conventional AM radio, then switched to an FM transmission. A United Press correspondent was present and recounted in a wire service report that: "if the audience of 500 engineers had shut their eyes they would have believed the jazz band was in the same room. There were no extraneous sounds." Moreover, "Several engineers said after the demonstration that they consider Dr. Armstrong's invention one of the most important radio developments since the first earphone crystal sets were introduced." Armstrong was quoted as saying he could "visualize a time not far distant when the use of ultra-high frequency wave bands will play the leading role in all broadcasting," although the article noted that "A switchover to the ultra-high frequency system would mean the junking of present broadcasting equipment and present receivers in homes, eventually causing the expenditure of billions of dollars."

3.4. FCC Proceedings and the FM Band Shift

In the late 1930s, as technical advances made it possible to transmit on higher frequencies, the FCC investigated options for increasing the number of broadcasting stations, in addition to ideas for better audio quality, known as "high-fidelity." In 1937, it introduced what became known as the Apex band, consisting of 75 broadcasting frequencies from 41.02 to 43.98 MHz. As on the standard broadcast band, these were AM stations but with higher quality audio - in one example, a frequency response from 20 Hz to 17,000 Hz +/- 1 dB - because station separations were 40 kHz instead of the 10 kHz spacings used on the original AM band. Armstrong worked to convince the FCC that a band of FM broadcasting stations would be a superior approach. That year he financed the construction of the first FM radio station, W2XMN (later KE2XCC) at Alpine, New Jersey. FCC engineers had believed that transmissions using high frequencies would travel little farther than line-of-sight distances, limited by the horizon. However, when operating with 40 kilowatts on 42.8 MHz, the station could be clearly heard 100 mile away, matching the daytime coverage of a full power 50-kilowatt AM station.

FCC studies comparing the Apex station transmissions with Armstrong's FM system concluded that his approach was superior. In early 1940, the FCC held hearings on whether to establish a commercial FM service. Following this review, the FCC announced the establishment of an FM band effective January 1, 1941, consisting of forty 200 kHz-wide channels on a band from 42 to 50 MHz, with the first five channels reserved for educational stations. Existing Apex stations were notified that they would not be allowed to operate after January 1, 1941, unless they converted to FM.

Although there was interest in the new FM band by station owners, construction restrictions that went into place during World War II limited the growth of the new service. Following the end of WWII, the FCC moved to standardize its frequency allocations. One area of concern was the effects of tropospheric and Sporadic E propagation, which at times reflected station signals over great distances, causing mutual interference. A particularly controversial proposal, spearheaded by RCA, was that the FM band needed to be shifted to higher frequencies to avoid this problem. This reassignment was fiercely opposed as unneeded by Armstrong, but he lost. The FCC made its decision final on June 27, 1945. It allocated 100 FM channels from 88 to 108 MHz and assigned the former FM band to 'non government fixed and mobile' (42-44 MHz) and television channel 1 (44-50 MHz), now sidestepping the interference concerns. A period allowing existing FM stations to broadcast on both low and high bands ended at midnight on January 8, 1949, at which time any low band transmitters were shut down, making obsolete 395,000 receivers that had already been purchased by the public for the original band. Although converters allowing low band FM sets to receive high band were manufactured, they ultimately proved to be complicated to install and often as (or more) expensive than buying a new high band set outright.

Armstrong felt the FM band reassignment had been inspired primarily by a desire to cause a disruption that would limit FM's ability to challenge the existing radio industry, including RCA's AM radio properties that included the NBC radio network, plus the other major networks including CBS, ABC, and Mutual. The change was thought to have been favored by AT&T, as the elimination of FM relaying stations would require radio stations to lease wired links from that company. Particularly galling was the FCC assignment of TV channel 1 to the 44-50 MHz segment of the old FM band. Channel 1 was later deleted, since periodic radio propagation would make local TV signals unviewable. While technically, the 88-108 MHz band is more suitable for FM radio due to issues like sunspot activity affecting the 50 MHz range, the shift was a significant blow to early FM radio stations, with some being forced to close.

Although the FM band shift was an economic setback, there was reason for optimism. A book published in 1946 by Charles A. Siepmann heralded FM stations as "Radio's Second Chance." In late 1945, Armstrong contracted with John Orr Young, a founding member of the public relations firm Young & Rubicam, to conduct a national campaign promoting FM broadcasting, especially by educational institutions. Articles promoting both Armstrong personally and FM were placed with general circulation publications including The Nation, Fortune, The New York Times, Atlantic Monthly, and The Saturday Evening Post.

In 1940, RCA offered Armstrong 1.00 M USD for a non-exclusive, royalty-free license to use his FM patents. He refused this offer because he felt this would be unfair to the other licensed companies, which had to pay 2% royalties on their sales. Over time, this impasse with RCA dominated Armstrong's life. RCA countered by conducting its own FM research, eventually developing what it claimed was a non-infringing FM system. The corporation encouraged other companies to stop paying royalties to Armstrong. Outraged by this, in 1948 Armstrong filed suit against RCA and the National Broadcasting Company, accusing them of patent infringement and that they had "deliberately set out to oppose and impair the value" of his invention, for which he requested treble damages. Although he was confident that this suit would be successful and result in a major monetary award, the protracted legal maneuvering that followed eventually began to impair his finances, especially after his primary patents expired in late 1950.