Frequency modulation or FM radio transformed how we communicate. It was, as Armstrong’s nemesis David Sarnoff said, “a revolution”. There are videos, books, and articles about this history but they seem to skim over the science. In fact, I haven’t seen any videos or articles that simply and accurately describe how Armstrong’s original system actually worked! This video is my attempt to explain how and why Armstrong invented the FM radio and how it was so amazing that it led to his destruction. Ready? Let’s go!
In 1923, Howard Armstrong worked with his friend David Sarnoff (then the VP of RCA) to create the Radiola radio which made Armstrong and Sarnoff millions, and soon broadcast radio took over the culture. There was only one problem, static, and it was due to the nature of AM radio. Let’s start with some basics. Radio waves aren’t sound waves (I know crazy right?) they are actually invisible light waves. Radio waves are used as the medium that is used to transmit or carry sound information. In AM radio, the sound is converted to an electrical signal and that signal is used to change the height (or the amplitude) of the wave. [Thus the name Amplitude Modulation: AM]. However, eveProc. IRE, 24 (5ry time there is a spark it makes a little radio wave which increases the amplitude of the radio wave which creates static in AM radio. Sarnoff had repeatedly said that he wished he had a “little black box” to put in his radio receivers to get rid of static. Howard Armstrong thought he had the solution, why not create a radio wave where the amount the frequency deviated from the original (more squished or stretched) corresponded to the sound wave produced in the microphone instead of changing the height? For this reason, it is called FM radio (for Frequency Modulation).
Actually, many different scientists were interested in using FM radio starting in the late teens and early twenties. However, in February of 1922 John Carson, the chief mathematical theoretician at Bell Laboratories, wrote a paper that it wouldn’t work. Well, to be more accurate he said it would work but it, “inherently distorts without any compensating advantages whatsoever.” Carson eventually concluded that “static, like the poor, will always be with us.” Now Armstrong was even more determined saying, “I could never accept findings based almost exclusively on mathematics.”
By October of 1928, Armstrong focused exclusively on this problem six to seven days a week. He even drove one of his assistants to regularly go to church just to get a break from him. Three years later, he still seemed no closer to a solution when he had a radical notion. When you use FM or frequency modulation you by necessity change the frequency. He knew that with the technology available at the time he needed to keep that fluctuation to a minimum or it distorted the signal (as Carson mathematically proved) but then it was impossible to include the full variations of sound and reduce static. What if, he thought, you started with a low frequency with small changes and then multiplied it to a higher frequency with large changes and made the receiver only receptive to large changes? This is called broadband FM radio. This was not an easy task and it took until the end of 1933 for him to get it right. But when he did, it worked even better than he had hoped. The static was gone. You could transmit the entire range of human hearing instead of a limited range like AM. It could even be made to contain more than one signal at a time to transmit surround sound. It wasn’t an improvement, it was a revelation. And Sarnoff hated it.
You might ask, how can a signal from a microphone be changed into a variation in the frequency of a wave? Armstrong started by creating an AM signal on a very low-frequency radio wave. He then suppressed the original smooth radio signal and reinstalled it 90 degrees out of phase. This creates PM or phase modulation, so, instead of the signal changing the height of the wave, it changed the phase of the wave. He then used a capacitor to “smooth out” the signal (or integrate it) and make it into an FM signal. Finally, he used harmonics to multiply the signal to get to the carrier frequency and frequency variations that he wanted.
How about the receiver? Let’s start with how Armstrong’s AM receiver worked first. With AM Armstrong amplified a signal from an antenna and then mixed it with another signal to create a lower intermediate signal that was easier to deal with. He then filtered it to his desired frequency, amplified that signal, and then used electronics to detect the changes in amplitude. By the way, this “envelope detector” could be as simple as a one-way valve. Finally, he amplified the sound and projected it on a loudspeaker. The FM receiver was only slightly different from his AM receiver. First, with FM you don’t want a narrow frequency filter or you would cut out all your information, so he used a broad frequency filter. Second, he used something he called a “limiter” because it limited the height of the wave. Basically, it saturated a system so that any variation in the height of the waves was cut off since the real signal was in the variation in the frequency and the variation in the height was only due to static anyway. Third, instead of an “envelope detector,” Armstrong used a discriminator that would, as Armstrong put it “translate variations in frequency into variations in amplitude.” So, once again, how? He created two tuned circuits with coils and capacitors, one tuned to above the carrier frequency and one tuned to below the carrier frequency. If a wave was created with no modulation, then it would work equally badly in both circuits. Then, the two coils labeled 48 and 49 would get the same voltage induced in them, which would cause no current to flow between them. However, if the frequency was modulated, then one circuit would resonate better and at higher amplitude and the other would resonate worse with lower amplitude. This would cause one coil to have a higher voltage than the other coil, which would create current between them and correspondingly, current in the speaker or headphones! The bigger the frequency difference, the bigger the amplitude of the current in the headphones. In this way, Armstrong converted changes in frequency to changes in amplitude in a speaker. Whew!
So, if the FM signal was so much better than the AM one, why wasn’t Sarnoff happy? Well, Sarnoff had hoped that Armstrong would invent something to improve the signal, not something that would supplant his entire system. What to do? Sarnoff decided to crush his friend’s new invention, kicked Armstrong out of his laboratory (in the Empire State Building), had his scientists write articles denigrating FM, and forbade RCA from using FM radio. Undeterred, Armstrong sold all of his stock in RCA and started his own FM radio company called “The Yankee Network”. It wasn’t until 5 years later, in 1939, that Sarnoff realized his mistake and tried to get a non-exclusionary license for a million dollars. Armstrong told them to get lost. Armstrong’s lawyer thought his client was crazy saying, “That’s the first time I ever heard of an inventor turning down a million for a non-exclusionary license.” David Sarnoff was incensed. He started a personal war with his former friend and used his influence so that the FCC changed the radio frequencies available for FM, making Armstrong’s equipment and company worthless!
Sarnoff and RCA then started using FM without permission and every other radio and television company followed suit. In July of 1948, Armstrong sued. RCA decided to win by just making the fight as long and as arduous as possible, or as Armstrong wrote to a friend, “they will stall this thing until I am dead or broke”. Armstrong’s lawyer fees went up to $200,000 a year while his income from patents was dwindling from more and more companies simply not paying him for his devices while they waited for the lawsuit to end. And this lawsuit dragged on and on. Armstrong’s deposition alone lasted over a year. By 1952 Armstrong was broke and was getting loans to pay his lawyers. On February 20, 1953, Sarnoff said under oath that, “RCA [has] done more to develop FM than anybody in this country, including Armstrong.” In November Armstrong admitted the state of his finances to his wife, Marion. They had a horrible fight and Armstrong hit her arm with a poker. Marion fled the house, never to see her husband again. On January 31, 1954, exactly forty years after Armstrong stayed up all night with Sarnoff demonstrating his regenerative circuit, a date they celebrated for decades, Armstrong wrote an apology to his wife, removed the air conditioner from his apartment window, and jumped 13 floors to his death. Armstrong was 63 years old.
When David Sarnoff learned about Armstrong’s suicide he blurted, “I did not kill Armstrong”. Despite this protestation, Sarnoff must have known he had a strong hand in his friend’s desperate act and wept openly at his funeral. Marion Armstrong continued all of her husband’s lawsuits. Eventually, she would file twenty-one patent lawsuits and would win all of them. She won over $10 million in damages over the next eleven years.
Let’s go back to 1934. Part of why Sarnoff didn’t want to revolutionize radio with FM is because Sarnoff was cash poor as he had already poured 5 million dollars into controlling a strange item called a cathode ray tube. This is why Sarnoff (and RCA) eventually became the titans of CRT television! But what is a cathode ray tube? How does it work and why (and how) is it connected to television and the oscilloscope, radar, x-ray machines, the discovery of the electron, and the photoelectric effect? That starts next time on the secret history of television.