The Age of the Amplifier

William Shockley, John Bardeen, and Walter Brattain, winners of the 1956 Nobel Prize for their work on the “transistor effect.” Via Wikipedia.As we’ve noted more than a few times before, for most of the 20th century AT&T’s Bell Labs was the premier industrial research lab in the US. As part of its ongoing efforts to provide universal telephone service, Bell Labs generated numerous world-changing inventions, and accumulated more Nobel Prizes than any other industrial research lab.1 But the most important of its technical contributions proved to be useful far beyond the confines of the Bell System. Statistical process control, for instance, was invented by AT&T engineer Walter Shewhart to improve the manufacturing of AT&T’s electrical equipment at supplier company Western Electric. Since then, the methods have been successfully applied to all manner of manufacturing, from jet engines to semiconductors to container ships.Interestingly, some of AT&T’s most important technological contributions — namely, the vacuum tube, the negative feedback amplifier, the transistor, and the laser — were (in whole or in part) the product of efforts to make new, better amplifiers for boosting electromagnetic signals. Amplifiers played a crucial role in the Bell System, making it possible to (among other things) connect telephones over long distances, but the value of these four amplifiers extended far beyond telephony. The vacuum tube became a crucial building block for electronics in the first half of the 20th century, used in everything from radio to television to the earliest computers. The negative feedback amplifier helped spawn the discipline of control theory, which is used today in the design of virtually every automated machine. The transistor is the foundation of modern digital computing and everything built on top of it. And the laser is used in everything from fiber-optic communications to industrial cutting machines to barcode scanners to printers.It’s worth looking at why AT&T was so motivated to build better and better amplifiers, and why those efforts produced so many transformative inventions.In 1876 Alexander Graham Bell placed the world’s first telephone call, summoning his assistant Thomas Watson from another room. By 1881, Bell’s company, the Bell Telephone Company (it wouldn’t become American Telephone and Telegraph, or AT&T, until 1899) had 100,000 customers.

By the turn of the century AT&T was operating 1,300 telephone exchanges in the US, connecting over 800,000 customers with 2 million miles of wire.The goal of the Bell System was “universal service” – to connect every telephone user with every other telephone user in the system. But by the early 20th century this quest was bumping up against technological limitations.Telephones converted the sound of someone speaking to electrical signals, which were transmitted along wires until reaching a telephone on the other end, where they were converted back into sound. More specifically, in early telephones the sound from someone speaking would compress and decompress a chamber full of carbon granules, which would alter their electrical resistance, changing how much current flowed through them. At the other end, the electrical current would flow through an electromagnet, which pulled on a thin iron diaphragm; fluctuations in the electrical current would change the motion of the diaphragm, reproducing the speech.But the farther electrical signals travelled, the more they would be attenuated. Resistance from the wire carrying them would convert some of the electrical energy into heat, and electrical current could “leak” between adjacent telephone wires. As the electrical signals got weaker and weaker, the sound would be less and less intelligible when reproduced, until it couldn’t be heard at all. If AT&T wanted to provide universal service, it would need a way to maintain the strength of the electrical signal as it traveled over long distances.AT&T was able to partly resolve this problem using the loading coil, an invention of electrical engineer Michael Pupin. (Lines which had loading coils added to them were sometimes described as being “Pupinized.”) The loading coil added inductance (a tendency to resist changes in current) to telephone lines, which reduced signal attenuation. As a result, the loading coil roughly doubled the effective distance limit of telephone calls, from around 1000-1200 miles to closer to 2000 miles. But the loading coil merely reduced signal attenuation; the signal was still decaying as it traveled along the lines, just more slowly. Without some way of actually amplifying the telephone signals, the maximum distance for a telephone line was enough to connect New York to Denver, but not enough to reach the West Coast from New York and connect the entire country.

AT&T experimented with various mechanical amplifiers, which converted the electrical signals into mechanical movements and then back to electrical signals, but these were found to greatly distort the signal, and were not widely used. What was needed was an electronic amplifier, which could amplify the electrical signals directly, without the lossy and distorting effects of mechanical translation. In 1911, AT&T formed a special research branch to tackle the problem of long-distance transmission, and hired the young physicist Harold Arnold (who would later become the first director of research at Bell Labs) to research possible amplifiers based on the “new physics” of electrons.Electronic amplification: the blue is the voltage of the input signal, which varies over time. The red is the amplified voltage of the output signal. The gain of this amplifier is three: output voltage is 3x input voltage. Similar amplification can be done for electrical current. Via Wikipedia.At first, Arnold had little success. He looked at a variety of possible amplifying technologies, and experimented extensively with mercury discharge tubes (which initially seemed promising), but nothing appeared to fit AT&T’s requirements. But in 1912, Arnold learned of a new, promising amplifier known as the audion, which had been brought to AT&T by American inventor Lee de Forest. De Forest’s audion was, in turn, based on an invention of the British physicist Ambrose Fleming, known as the “Fleming valve.” Fleming was inspired by extensive experimentation with what was known as the “Edison Effect:” the observation that in an incandescent bulb, electric current would flow from the heated filament to a nearby metal plate. Fleming used this effect to create a diode, a device which lets electric current flow in one direction but not another. De Forest modified Fleming’s valve by adding a third element, a metallic grid, between the filament and the plate. By varying the voltage at the metallic grid, De Forest eventually discovered he could control the flow of electrical current from the filament to the plate. This allowed the device to act as an amplifier: a small change in the voltage could create a much larger change in the current flowing from the filament to the plate.A radio receiver built with audions, via Wikipedia.De Forest’s audion had uneven performance — notably, it couldn’t handle the level of energy needed for a telephone line. Moreover, it was clear that De Forest did not quite understand how the device worked.

But Arnold, well-versed in the physics of electrons, recognized its potential, and realized that, with modifications, its various limitations could be overcome. Via “The Continuous Wave,” a history of early radio:Arnold knew exactly what to do about the audion’s limitations. “I suggested that we make the thing larger, increase the size of the plate with the corresponding increases in the size of the grid but particularly at that time I suggested that we were not getting enough electrons from the filament.” What he wanted to do, in fact, was convert the de Forest audion into a different kind of device. He wanted a much higher vacuum in the tube, with residual gas eliminated to the greatest possible extent; and he knew the newly invented Gaede molecular vacuum pump made that possible. He wanted more electron emission from the filament without an increase in filament voltage; and he knew Wehnelt’s new oxide-coated filaments would do that.After paying $50,000 (roughly $1.6 million in 2026 dollars) for the rights to the audion, Arnold and others at AT&T spent the next year turning it into a practical electronic amplifier: the triode vacuum tube. By June 1914, vacuum tube amplifiers were being installed on a transcontinental telephone line connecting New York and San Francisco, and in January of 1915 the transcontinental line was inaugurated at the Panama-Pacific International Exposition with a call between Alexander Graham Bell in New York and Thomas Watson in San Francisco. By the late 1920s, AT&T was using over 100,000 vacuum tubes in its telephone system, and triodes and their descendents (four-element tetrodes, five-element pentodes) would go on to be used in all manner of electronic devices, from radios to TVs to the first digital computers.Via Hong 2001.The vacuum tube, with its ability to amplify electronic signals, represented a sea change in how AT&T engineers thought about telephone service. Prior to the electronic amplifier, a telephone call was essentially a single diminishing stream of electromagnetic energy. The range of that energy could be extended farther and farther from the speaker at the steep cost of its fidelity. The amplifier made it possible to consider a telephone call as a stream of information, as a signal that was distinct from the medium that carried it.

It could be ably renewed, translated, modified in new and exciting ways. As historian David Mindell notes:…a working amplifier could renew the signal at any point, and hence maintain it through complicated manipulations, making possible long strings of filters, modulators, and transmission lines. Electricity in the wires became merely a carrier of messages, not a source of power, and hence opened the door to new ways of thinking about communications…The message was no longer the medium, now it was a signal that could be understood and manipulated on its own terms, independent of its physical embodiment.Thanks to vacuum tube amplifiers, AT&T could finally fulfill its dream of universal telephone service, connecting telephones to each other anywhere in the continental US. But vacuum tubes were far from perfect amplifiers. The ideal amplifier has a linear relationship between the input and the output, effectively multiplying the input current or voltage by some value. If this relationship is non-linear, some inputs will be multiplied more than others, and the signal will become distorted. This distortion can garble speech, and — on a wire carrying multiple telephone calls — can create cross-talk, with speech from one telephone call being heard on another.The vacuum tube was a superior amplifier to anything that preceded it, but it wasn’t a perfectly linear amplifier; its amplification curve formed more of an S-shape, under-amplifying low values and over-amplifying high ones. For a line carrying a single telephone call, the resulting distortion could be mitigated by restricting inputs to the linear portion of the curve, but as AT&T adopted carrier modulation — carrying multiple calls at different frequencies on a single line — distortion became more of a problem.In 1921, Harold Black, a 23 year old electrical engineer, joined AT&T. He soon produced a report analyzing the future potential of a transcontinental telephone line carrying thousands of carrier-modulated conversations. At the time, carrier modulation was being used to carry at most three calls on a single line. Black’s analysis showed that such a line would require an amplifier with far less distortion than existing vacuum tubes,and Black began to work on developing an improved amplifier as a side project.At first, Black simply tried to create vacuum tubes with less distortion, a project that many others at AT&T were also working on. The efforts of Black and others produced higher-quality vacuum tubes, but nothing Black tried reduced the distortion to the degree he was aiming for.