The Chip: How Two Americans Invented the Microchip and Launched a Revolution

Home > Other > The Chip: How Two Americans Invented the Microchip and Launched a Revolution > Page 6
The Chip: How Two Americans Invented the Microchip and Launched a Revolution Page 6

by T. R. Reid


  All this THINKING about THINKING thrust Shockley into a furious racial and political controversy that he initially provoked in the late 1960s and that continues, at lower intensity, to this day. Perhaps it was the self-confidence that came from the knowledge that he had hit on one of the most important scientific ideas of the century; perhaps it was some political inclination. For whatever reason, though, the great physicist decided that he was also an expert in the field of human intelligence. In a letter to the National Academy of Sciences, and then in a series of lectures and interviews, Shockley urged detailed study of a problem he named “dysgenics” and defined as “retrogressive evolution through the disproportionate reproduction of the genetically disabled.” In plain English, Shockley was claiming that, on the average, blacks are dumber than whites. Thus high birthrates among blacks could lead to “decline of our nation’s human quality.” “My research leads me inescapably to the opinion,” Shockley said, “that the major cause of American Negroes’ intellectual and social deficits is hereditary and is racially genetic in origin.” Shockley went on to propose a social policy to deal with this situation: government should work to reduce birthrates among low-IQ elements of the population, through programs such as tax breaks for voluntary sterilization.

  This theory, proposed by a physicist with no training in genetics and set forth during a fairly tumultuous period in American history, made Shockley one of the most despised and vilified men in the United States. He was denounced as a pseudoscientist, a fanatic, a fascist. He was burned in effigy on both coasts and denied the right to speak at some of the nation’s most prestigious colleges, including Harvard, Dartmouth, and Yale. He was permitted to appear at Princeton, speaking in a small hall (chosen for its tight security) while hundreds of enraged demonstrators screamed protests outside. Back at Stanford, groups of students equipped with bullhorns, the portable and immensely powerful loudspeakers made possible by the transistor, would gather under his office window to chant “Off Pig Shockley”—affording Shockley the experience, probably unique in engineering history, of watching his own invention used to provide hundredfold amplification of demands for his death.

  The instigator seemed to relish the stir he had created. A photograph of one of Shockley’s classes that has been disrupted by demonstrators in Ku Klux Klan robes and pointed hats shows the professor standing calmly aside, arms folded, taking in the scene with an appearance of benign unconcern. At one of the countless “Off Shockley” rallies at Stanford, the microphone went on the blink; the inventor, who was present, stepped forward and repaired the transistorized device so that the speakers could continue their denunciation of his ideas. No matter how bitter the attacks on him, Shockley never stopped seeking new forums to convey his message; in 1982, to the dismay of party leaders, he briefly pursued the Republican nomination for the U.S. Senate. His platform called for a public inquiry into “dysgenics.” Perhaps the saddest aspect of the whole thing was that the argument about intelligence became Shockley’s central concern; the sheer pleasure he had always taken in pondering the mysteries of the physical world was overwhelmed by his immersion in politics. When I pushed and pulled at him to get him to remember the thought processes that led to his greatest invention, he kept steering the conversation back to intelligence and race. He was still fixated on that issue when he died in 1989.

  One of the ironies of this extended controversy was that Shockley, for all his intensity and determination, seemed much more a pleasant grandfatherly type than a public ogre. Those who worked with him over the years invariably described him as “charming”; at scientific meetings he was known for telling jokes and even performing magic tricks at the podium while delivering his papers. He also had a reputation for getting the most out of the people who worked with him. This he certainly did at Bell Labs when, just after World War II, he was put in charge of a team investigating new semiconductor applications.

  The senior member of the group, Walter Houser Brattain, was forty-five when the transistor was invented. Brattain grew up on the family ranch in Washington State and went to Whitman College. He did graduate work at Oregon and Minnesota, and upon receiving his Ph.D. in 1929 went to work for the newly organized research institution Bell Labs. It was a time, as he recalled later, when “the vacuum tube and thermionics were just shedding their baby teeth,” and he was first put to work on tubes. Later he moved into semiconductor research under the great scholar C. J. Davisson and was present on the day in 1937 when word arrived that Davisson had won the Nobel Prize. A horde of reporters swarmed onto the premises, and the lab was quickly engulfed in microphones, newsreel cameras, and banks of klieg lights. Davisson, noticing the astonishment on his young assistant’s face, stepped over to Brattain and whispered, “Don’t worry, Walter—you’ll win one some day.”

  The most unassuming of men, Brattain seemed almost embarrassed about winning awards. At the ceremony during which King Gustav VI of Sweden presented him the Nobel Prize, Brattain observed that “one indeed needs to be humble about accepting such an award when he thinks how fortunate he was to be in the right environment at the right time.” He was equally diffident about the scientist’s role in society. His job, Brattain said, was merely to enhance our understanding of the physical world; “I feel strongly, however, that the scientist has no right to dictate how his understanding is used.” Nonetheless, when he took a look back on the twenty-fifth anniversary of the transistor, Brattain did voice one complaint: “The thing I deplore most is the use of solid state electronics by rock and roll musicians to raise the level of sound to where it is both painful and injurious.”

  Brattain was the experimentalist of the transistor trio; he had an intuitive feeling for the way semiconductors ought to work, and it was he who turned his colleagues’ theories into working apparatus. The first transistor was an ungainly construct of germanium and wire that could be made to work only by Walter Brattain and only, he wrote later, “if I wiggled it just right.”

  The third member of the transistor team was John Bardeen. Born in 1908, the son of the dean of the medical school at the University of Wisconsin, he grew up in Madison and took his engineering degree at Wisconsin. He went to work for Gulf Oil in Pittsburgh but decided after three years that he preferred pure to applied science. He enrolled at Princeton, studied semiconductor physics under the future Nobel laureate Eugene Wigner, and took a Ph.D. in 1936. He taught at Minnesota, worked in the Naval Ordnance Laboratory during the war, and then agreed to join his friend Shockley at Bell Labs in 1946. In 1951, with the basic transistor work completed, he suggested that Bell undertake work on superconductors—that is, materials that show almost no electrical resistance at extremely low temperatures. Bell rejected this proposal, foolishly, on the grounds that superconductivity was too pie-in-the-sky to work. Bardeen left the lab and took a position at the University of Illinois (where a young engineer named Jack Kilby heard him lecture). While there, Bardeen joined with two other physicists, Leon Cooper and John Schrieffer, to develop the theoretical basis of superconductivity. Their scheme, known as the BCS theory, has turned out to work and has been the basis for almost all superconductor development. For this theory, Bardeen won a second Nobel Prize.

  Bardeen was a perfect complement to Brattain. His forte is theory. On the transistor project, Brattain would work his experiments and then Bardeen would sift through the results and explain what they meant. He was not the type of person who could wiggle a piece of machinery just right and make it work. On the day in 1972 when his second Nobel was announced, the public relations people at Illinois, proud as punch, arranged a press conference and summoned every science reporter within traveling distance. The reporters showed up, but not Bardeen. He apologized later, explaining that he had been stuck at home because he couldn’t get his (transistorized) garage door opener to work.

  Brattain the experimentalist, Bardeen the theorist, and Shockley, who did some of both and was clearly the leader of the group, began in 1946 to study the different attrib
utes of the two types of semiconductor: P-type, the material that had been doped with excess positive charges (holes), and N-type, which had excess negative charges (electrons). They worked with “bipolar” germanium—that is, a strip of germanium that had been doped so as to be N-type on one end and P-type at the other. The men focused intensively on the point in the center where N-type and P-type meet. This point, the most important meeting place in modern physics, is called the P-N junction. The P-N junction makes wonderful things happen.

  The P-N junction works like the turnstile you pass through when you enter the subway or a stadium: you can go through easily in one direction but not the other. The P-N junction is a one-way door for electrons; they can pass through it going one direction, but not the other. When the semiconductor strip is hooked up to a source of current—a battery, for example—electrons can flow easily from the N-type material, across the P-N junction, to the P end. But they can’t cross the junction in the other direction.

  A device that lets current pass in only one direction—that’s a rectifier, just like John A. Fleming’s vacuum tube rectifier, which made reliable radios possible. Since the two ends of the bipolar strip act like the two electrodes of Fleming’s tube, the semiconductor rectifier, like the tube, is called a diode. But instead of the large, hot, fragile, power-hungry vacuum tube that Fleming used, this semiconductor diode, with its minute P-N junction, is tiny, low-power, unheated, and unbreakable. The semiconductor diode needs no vacuum, either, and thus no glass bulb. The electronic action takes place within the solid; hence the term “solid state.”

  By 1946, when the Bell Labs team began its work, the operation of the semiconductor diode was fairly well understood. Semiconductor diodes were not at all important in electronics, however—except for highly specialized applications, like radar— because they did not perform the essential task of amplification. As long as electronic equipment still needed vacuum tubes—with all their attendant problems of high power, heat, and size—as amplifiers, it was just as convenient to use vacuum tube diode rectifiers as well. The revolution in electronics came when Shockley, Bardeen, and Brattain, after two years of intense and frequently exasperating work, devised a semiconductor triode—a tiny, low-power, unheated, solid-state amplifier. This was the very device Shockley had thought of back in December 1939.

  The world’s first solid-state amplifier was a jury-rigged affair put together by Bardeen and Brattain. It involved two fine wires placed extremely close to each other precisely at the P-N junction on a small piece of germanium. If the wires were set at exactly the right spot, and if Brattain wiggled them just right, a small current flowing into one wire could be amplified to a current one hundred times as great.

  The inventors demonstrated this device to the Bell Labs brass on December 23, 1947. At first they measured the amplification with standard electronic meters so that the observers could see the dial on the meter swing far to the right, showing a huge amplification of current. But the test that really brought home what had been achieved came when they hooked up a microphone to one end of their invention and a loudspeaker to the other. One by one, the men picked up the microphone and whispered “Hello”; the loudspeaker at the other end of the circuit shouted “HELLO!” Shockley, with an acute sense of history, realized the nice symmetry of the moment: another major breakthrough in communications had occurred in Bell’s lab. “Hearing speech amplified by the transistor,” he wrote later, “was in the tradition of Alexander Graham Bell’s famous ‘Mr. Watson, come here, I want you.’ ”

  This first device, known as a point-contact transistor, was a cumbersome, imprecise instrument that of itself would probably not have made a significant impact. The true importance of that first transistor was that it inspired Shockley to perfect an improved solid-state amplifier, the device that would revolutionize electronics: the junction transistor. Shockley had only a small role in the development of the point-contact invention. “My elation with the group’s success was tempered . . . ,” he recalled later. “I experienced some frustration that my personal efforts . . . had not resulted in a significant inventive contribution of my own.” This frustration, coupled with the knowledge that solid-state amplification was indeed possible, gave Shockley “the will to think.” For the next five weeks, he dedicated himself to “the effort that tedious and precise thinking demands.” Working alone on New Year’s Eve, he filled nineteen pages in his lab notebook with ideas and sketches. On January 23, 1948, he set down in his notebook the concept of the junction transistor.

  Shockley’s idea—the basis of all transistors today—was a semiconductor sandwich, a strip of germanium with three different regions—N-type at one end, P-type in the middle, N-type at the other end. This created two P-N junctions, back-to-back. Three wires were hooked up—one to each N-type region, and one to the P-type in the middle. Now, if a charge was sent to the center region, it would vigorously suck electrons from one N-region, across the center, and out the other end. The stronger the charge on the center region, the more electrons it drove across the junction. That is, the charge in the center section could amplify the flow of current down the strip.

  In one common type of transistor, the first N-type region is called the source; the center region is the gate; the second N-type region the drain. Current flows out of the source and down the drain. How much current flows is controlled by the gate in the middle. A small change in the current hooked to the gate causes a big change in the current from source to drain. More important, variations in the gate current are exactly mimicked in the current flowing to the drain.

  The three sections of this solid-state sandwich are analogous to the three electrodes of Lee De Forest’s vacuum triode. The semiconductor “gate” acts like the vacuum tube “grid.” A radio signal sent to the gate can shape a much stronger current flowing from

  source to drain. Just as in the vacuum tube, the effective result is an amplified reproduction of the weak signal. In addition to its application as an amplifier, the solid-state triode can serve as an extremely high-speed switch. If the device is adjusted properly, a signal to the gate will cut off the drain current completely; another signal will open the gate and turn the drain current back on. The important thing is that, in a transistor, this on-off switching takes place in billionths of a second. This is faster than any vacuum tube could be made to switch. This capability for ultra-high-speed switching made possible the modern computer and all its myriad offspring and networks.

  Smaller, lighter, faster, more sensitive, more reliable, and far more power-efficient than the vacuum tube, the transistor could not have been anything other than a stupendous success. Two important decisions by the Bell System accelerated its progress. Mindful of its founder’s lifelong interest in helping the deaf, Bell waived all patent royalties on the first important transistor product, the miniature hearing aid. For all other applications, Bell Labs, moved by a sense of public service (and, perhaps, by a pending antitrust action against AT&T, its parent company), established a bargain-basement license fee of $25,000 and ran training programs for all firms interested in producing transistors.

  By the mid-fifties, semiconductor sales were in the billion-dollar range and the vacuum triode was becoming a museum piece. Each new year brought hundreds of intriguing new inventions based on the transistor. The popular press treated the new technology as a full-fledged miracle. “To all industrial needs, and most human physical needs,” Time magazine noted in a typically breathless report in 1957, “the electronics magicians are sure they have the key.”

  Or were they? By 1957, the electronic magicians were sure only that they had a serious problem—the problem posed by the tyranny of numbers. Unless the interconnections dilemma could be resolved, the enormous promise of a transistorized future might never be realized. The central importance of the problem—and the profits to be gained from its solution—instilled in governments and research labs and manufacturing concerns around the globe “the will to think” about a solution, and t
he will to spend large sums in pursuit of it.

  One of the firms in the forefront of this pursuit was Texas Instruments, which mounted a large-scale research project to deal with the numbers problem early in 1958 and began recruiting semiconductor experts from throughout the world for the task. Among the men TI hired was a lanky thirty-four-year-old engineer from Milwaukee named Jack Kilby.

  3

  A NONOBVIOUS SOLUTION

  Jack St. Clair Kilby landed his first job in electronics the year the transistor was born and has been working ever since at the front lines of high technology. He was among a small group of pioneers in the 1950s who developed the transistor from a laboratory specimen to a mainstay of industry. He conceived and built the world’s first integrated circuit, or semiconductor chip, and then proceeded to invent what is probably the chip’s most famous offspring, the handheld calculator. He spent a half dozen years after that working on the photovoltaic effect, an as yet imperfectly understood phenomenon of semiconductor physics, in an effort to construct an electric generating station fueled by the sun. He has won countless scholarly awards, including the National Medal of Science, the technologist’s equivalent of the Congressional Medal of Honor; the Kyoto Prize, Japan’s equivalent of the Nobel Prize; and the Nobel Prize itself. His picture hangs in a hall of fame in Washington between those of Henry Ford and Ernest Lawrence, inventor of the cyclotron.

 

‹ Prev