The Idea Factory: Bell Labs and the Great Age of American Innovation
Page 24
When Kelly said something was dead, it was dead. Except to Pierce. He and Kompfner prepared a long memo in early October 1958 that made their case to Kelly yet again. Pierce showed the memo to his boss, Baker, but it was not sent on to Kelly, almost certainly on Baker’s wise suggestion.15 The strategy was to wait. “Kelly was a great hero of mine,” Pierce later reflected, “and a great leader of Bell Laboratories. Even great men can be wrong.” When Kelly retired at the end of 1958, Fisk took charge of the Labs. Pierce and Kompfner then circulated another proposal on January 6, 1959, suggesting that the cost for the satellite would be $1 million during 1959 “and more in subsequent years.” It seemed possible, they noted, “that satellites could provide very broad-band transoceanic channels more cheaply than cable or other means.” The two men also suggested that if Bell Labs didn’t act, someone else would. “It is important for the Bell Laboratories to assume a leading role in research on satellite communications. This would enable us to evaluate satellite communication technically and to exploit it expeditiously if it proves advantageous to us.”16 This time, Pierce’s proposal was authorized by AT&T as Case 20564. It was also given an official name: Project Echo. NASA, in turn, scheduled a tentative launch for the fall of 1959.17
Pierce and Kompfner asked Bill Jakes, a young electrical engineer, to take charge of the project. Jakes was wiry and athletic and quietly commanding; after growing up in Evanston, Illinois, he had spent most of World War II in charge of the radio room on an aircraft carrier in the South Pacific. Jakes’s team on Echo quickly swelled to several dozen men, and to Pierce’s relief Jakes soon took almost complete control. As Pierce would later say, Jakes’s word was law. At times the younger man would remark to Pierce, empathetically, that it must have bothered him that he didn’t get to run Echo’s day-to-day operations. “I don’t know what I said, but it certainly wasn’t the full truth,” Pierce later wrote. “The Echo ground terminal would never have got built or worked had I been the project engineer. Through a mixture of ineptitude or boredom, I would have flubbed.” For Pierce, it was plenty of work just being an idea man. Jakes was the kind of guy who would actually make ideas happen.
AT THE TIME Echo was authorized, Pierce held a high position in Bell Labs’ research department. His office was located at the Labs’ Murray Hill location, the large campus of brick buildings twenty-five miles west of New York City where Shockley and Shannon had also worked before they made their respective moves to California and Massachusetts. Every day at noon at Murray Hill, he stretched out on a table in his office and took a nap. “I tried to call him one day,” Bill Jakes recalls, “and his secretary said, ‘Oh, you can’t talk to him now, he’s sleeping.’” At Murray Hill, Pierce, with several hundred employees reporting to him, held the title of executive director of communications research, which in organizational terms put him one step below Bill Baker, the Labs’ vice president for research. (Baker, in turn, was one of several vice presidents who reported to Labs president Jim Fisk.) The satellite work, however, was centered not on Pierce and Baker and Fisk in Murray Hill but on Jakes in Holmdel, thirty miles south. Holmdel was where Pierce’s friend Harald Friis had worked and where he had invented the horn antenna. And it was where Pierce’s other good friend Rudi Kompfner, the traveling wave tube inventor, had now taken charge upon Friis’s retirement in 1958.
Holmdel was a world unto itself. The laboratory and its environs comprised hundreds of acres of open land, assembled from three farms purchased by Bell Labs in 1929, where several dozen men did research on microwave transmissions. They worked just a few minutes from Sandy Hook beach and the Atlantic Ocean. “You drove off the main road,” Bill Jakes recalls of his commute to work, “and it would be about a half mile until you came to the laboratory, with nothing but beautiful mown fields on both sides.” In the 1940s, the vast and isolated property had been used for some of the early research on silicon and germanium. But mostly Holmdel was the Labs’ central facility for the experimental installation of large antennas and to test emerging technologies on wireless microwave transmissions. The families of the men who worked at Holmdel would congregate there on the weekends for group picnics and to throw boomerangs in the empty fields; their children could play with an ancient water pump and explore an abandoned Dutch farmhouse on the grounds dating from the early 1700s. Years later some men, John Pierce included, would be moved nearly to tears as they recalled this vanished outpost. Its main laboratory building was a modest one-story wooden clapboard structure with three wings radiating from a peaked central core. The wings had so many windows that the building resembled an exceedingly long glassed-in porch. Some of the men affectionately called the building the Turkey Shed, after a traveling poultry-feed salesman, seeing the old farmhouse and the clapboard laboratory, mistook it for a working farm during a sales call. To work in Holmdel was to work in a place that was breezy and bucolic and hushed in a country way. “Lord, it was just beautiful,” Jakes’s wife, Mary, says—so beautiful that you couldn’t possibly think it was on the vanguard of radio research, though in fact it was. Holmdel’s appearance simply didn’t jibe with any conventional notions of laboratory science.18
Rudi Kompfner’s deputy at Holmdel was named C. Chapin Cutler. Technically speaking, Bill Jakes reported to both men, but on matters of Echo they formed a triumvirate with Jakes in the lead. Around the time Echo was authorized, Kompfner, Cutler, and Jakes went about buying the equipment for building their ground receiving station. About a mile north of the Holmdel lab was a flat-topped rise known as Crawford Hill, about five hundred feet above sea level, with an expansive view to the ocean just a few miles away, that the Holmdel men used for research on antennas.19 In the summer of 1959 plows arrived and construction began atop the hill. Several small prefabricated buildings were set up to house men and equipment. More important, two circles were cleared for the installation of two large antennas.
One of the novel aspects of Echo was the intention to use sophisticated computer systems to track the satellite as it moved across the sky and then automatically align the antennas in accordance with the balloon’s trajectory. For about $230,000, Jakes’s Echo crew purchased a huge dish antenna, sixty feet in diameter, that would transmit signals to the orbiting satellite—signals that the balloon would reflect to a ground site being built in tandem by the engineers at the Jet Propulsion Laboratory in Goldstone, California, just south of Death Valley. The Echo team, meanwhile, had also designed and requisitioned an immense horn antenna (the cost was about $128,000) resembling a kind of huge empty tobacco pipe, which would receive signals sent from the Jet Propulsion station in California as they bounced off the orbiting satellite.20 It was, in essence, a simple two-way conversation, but one made possible through an immensely complex electronic infrastructure on each edge of the continent. The horn antenna on Crawford Hill would be “steerable.” That is to say, it would be mounted on a circular track so that it could swivel according to the predicted path of the satellite. Near the base of the horn antenna, the Echo team would install the supercooled maser that would amplify the faint signals.
It was true, to a large extent, that the satellite project was about technology. But it was also about a vast logistical coordination—worked out on numerous trips to Washington and California—between Bell Labs, the Jet Propulsion Lab, NASA, and several other scientific affiliates (such as Douglas Aircraft and MIT’s Lincoln Labs) that were providing technological support. In mid-June 1959, the teams convened at NASA headquarters in Washington to offer progress reports. A series of tests were agreed upon in preparation. NASA would try out a new Delta-Thor rocket it intended to use for the satellite launch and would begin overseeing “shotput” experiments—abbreviated launches to heights of a few hundred miles—to investigate whether the huge reflective balloon, folded within a round canister only two feet in diameter, could be deployed and inflated correctly.21 The teams met at NASA headquarters again in late October. The date for the fall launch had apparently been scrapped, with hopes r
esting on a balloon launch during the spring of 1960. The team at the Jet Propulsion Lab and Jakes’s team in Holmdel meanwhile planned a series of prelaunch communication tests. Each would direct a radio signal toward the moon while the station on the opposite coast would try to receive the reflection. The first dates were set for late November 1959.
Moonbounces, they were called.
THE ECHO TEAM on Crawford Hill now ceased working normal hours.
“Poor Mary,” Bill Jakes says of his wife. “I’d go out there at ten o’clock at night. And come home at three o’clock. And we’d do this night after night after night after night to see how things were going.”
There were about a dozen moonbounces, some more successful than others, during the fall of 1959 and the winter of 1960. By the spring the equipment of all the various teams around the country appeared to be working satisfactorily. NASA’s hopes for a balloon launch from Florida’s Cape Canaveral in March were eventually pushed back to the morning of May 13, 1960. The space agency issued a lengthy press release that morning explaining the equipment and the launch procedure. The rocket’s payload—the thin Echo balloon, “about half the thickness of the cellophane on a cigarette package”—would be filled with about thirty pounds of “sublimating powders” (a combination of benzoic acid and anthraquinone) and folded in an accordion fashion into a twenty-six-and-a-half-inch magnesium container, which was put into the Delta rocket’s third stage. After separating from the rocket, the magnesium container would be split open by an explosive charge and the balloon would release. The powders inside, warmed by the sun, would turn to gas and help inflate the balloon. Moving in a southeasterly direction, Echo would then circle the earth at a height of about 1,000 miles once every two hours at an estimated speed of 16,000 miles per hour. It would be about as bright in the night sky as the star Vega. NASA predicted that Holmdel and Goldstone would have their first chance to connect during the satellite’s first pass, 150 minutes later. At that point, both teams would have the silvery balloon in their direct line of sight.22
There were four small buildings on Crawford Hill. One was a command center where Jakes and his team monitored the transmission and receiving equipment. Jakes would park himself in front of a laboratory bench loaded with equipment that showed whether the satellite dishes and horn antenna were tracking properly. He would watch the dials constantly. On May 12, Jakes and his colleagues spent the night there following the countdown from Cape Canaveral, which was sent to them by Teletype. Their families gathered nearby to wait. Pierce, Kompfner, and other Bell Labs executives convened in a small support building, perhaps fifteen by twenty feet, located near the command center.
“Here we all were,” Mary Jakes recalls. “All of us waiting, sitting on a grassy hill. And here was this building with the brains inside and the women and children outside. All of this stuff had gone into this launch. And it went up and it was a failure. Everyone else felt that a member of the family had died.”23 A bulletin from Bell Labs that afternoon explained that the third stage of the Delta rocket did not achieve orbit. The balloon payload burned up in the atmosphere.
What followed for the distraught members of the Echo team were several more months of tests—more moonbounces, more computer tracking experiments, more waiting. NASA scheduled a second launch for early August, but that, too, faced delays due to mechanical complications and weather. Jakes and his crew and Pierce showed up at the command shack around midnight on August 9–10 only to have their hopes dashed. And then, forecasting a night of clear weather on August 11, NASA gave the go-ahead for an early-morning August 12 launch.24 To allow his crew to concentrate on the task at hand, Jakes ordered everybody in the control shack—Pierce and Kompfner included—to another building. “I just wanted them the hell out of there,” he says, somewhat surprised and amused that his bosses actually followed his order.25 The countdown came from Florida over the Teletype machine. And at 5:39 a.m., the Delta rocket with the balloon payload went up.
The men on Crawford Hill waited anxiously. Around the world, several tracking stations had been enlisted to follow Echo’s progress, both by telescope and by monitoring its small solar-powered radio beacon, as it sped forth at 16,000 miles per hour. A station in Trinidad got it first. Then Johannesburg confirmed a signal. At 7:05 word came to Jakes from Woomera, a station in Australia—“Woomera has acquired it!” he announced over the loudspeaker—which virtually assured the men that the balloon was functioning and on the correct path. Pierce, drinking coffee and eating a doughnut in the support building, began jumping up and down; his glasses bounced on his nose. It was a startling sight to colleagues who had never before seen him visibly happy. “It’s in orbit,” he yelled. “Echo is in orbit.” He telephoned his wife to tell her.26
Within about thirty minutes the balloon was in the sights of both Goldstone and Holmdel and their antennas began their automated tracking movements. Goldstone intended to broadcast to Holmdel a recorded message from President Eisenhower prepared especially for the occasion. At 7:41 Jakes called California and asked them to start playing the message: A microwave signal was beamed from Goldstone’s antennas to the orbiting balloon, reflected down, and then sucked in by the giant horn in New Jersey, where it was amplified four thousand times by the maser in the horn’s base. Eisenhower’s voice came through the loudspeakers on Crawford Hill so clearly—the president lauded Echo as “one more significant step in the United States’ program of space research and exploration”—that Pierce and other visitors didn’t even realize at first it was coming from the other coast. When they did, a cheer went up.
Pierce quickly regained his composure. Soon afterward, he talked with reporters and answered questions about the experiment before departing for his house in Berkeley Heights, New Jersey. He’d been up all night. At home, he took a short rest, then went outside to paint his garage.27
It was possible that Pierce, the most restless thinker, was already on to the next idea. For Jakes, meanwhile, the work on Echo was actually just starting. During the following days and weeks he and his counterparts in California did a series of successful experiments, mostly at night—the first-ever two-way phone conversations over a communications satellite, for instance, and the first-ever satellite fax transmissions. Jakes and Pierce, now minor celebrities—Echo was on the front page of virtually every newspaper in the country—also began receiving piles of fan mail and crank calls. (“Your balloon made me pregnant,” one woman told Jakes.) In the course of its first year, Echo went around the earth 4,481 times; the ten-story balloon meanwhile began to wrinkle slightly as it was pelted with micrometeors and its interior gases leaked slowly out, giving it a twinkling effect to the naked eye.28 By then it had become an international curiosity, though, a bright dot streaking regularly across the night sky. Soon after, the Bell System set up a phone service in New York City to tell callers at which times Echo would be visible. A schedule was published in various newspapers as well. At Jones Beach, not far from New York City, the bandleader Guy Lombardo would sometimes look at his watch and momentarily stop his orchestra’s summer evening concert. Then he would suggest that everyone in the audience pause to look up and see Echo, at one point just a curious idea, and now something more, pass overhead.29
NOT LONG AFTER the Echo experiment, Pierce was walking down the hall in Bell Labs’ Murray Hill office when he bumped into a big, intimidating executive named Frederick Kappel. Kappel, a plainspoken Minnesotan, had begun his career at AT&T in the 1920s digging holes for telephone poles at $25 a week; he had risen through the ranks over a long career to become the company’s president. He and Pierce liked each other. They’d had an amicable encounter several years before, when an executive at AT&T had complained that the psychological research being done at Bell Labs under Pierce was too distant from communications science to deserve funding. Kappel headed a committee that looked into the matter, and after a review he let Pierce’s group proceed. Now, on the day when Kappel saw Pierce at Murray Hill, the executive was in
the midst of a series of speeches extolling the potential of a global satellite communications system, run in large part by AT&T. “Look what you’ve got me into, John,” Kappel joked.30
His point was that satellites, only a year after Echo, had progressed from an intriguing experiment to a cutthroat business. A number of multinational corporations—RCA, General Electric, ITT, among others—had already made clear their intentions, as Fortune magazine put it in July 1961, to install “the great cable in space.” Satellites offered an alternative to the increasingly burdened underseas cables; more important, they could carry live television, which the current underseas cables could not. Within a decade, some economists predicted, orbiting relays could be a billion-dollar-a-year business.31 “We think the best practical system would use about 50 satellites in space,” Kappel explained to U.S. News & World Report. Would AT&T spend $25 million to do it? “I’m talking, really, about a system costing more than that,” Kappel remarked. He put the cost at $200 million. Even at such an expense, he said, he believed that it could eventually make transatlantic calling cheaper.32
Pierce would later observe that Project Echo proceeded quickly and smoothly in part because it was considered eccentric: Few people in the business community perceived its practical importance, and as a result Pierce and his crew on Crawford Hill were largely left alone. In truth, apart from Mervin Kelly, there had been little skepticism about satellites at Bell Labs. Even as Echo and its ground stations were being planned in late 1959 and early 1960, Pierce was attending meetings about building a more complex active satellite, rather than a passive one like Echo.33 In simplest terms, the active satellite—it was soon approved and given the name Telstar—would be able to receive a radio signal beamed from the ground, amplify it ten billion times, and then retransmit it, on another frequency, back to earth. It could carry hundreds of phone calls or several television programs. (Television required far more bandwidth than phone calls.) Kappel, in other words, didn’t want to spend $200 million of AT&T’s money on big silver balloons. He wanted to spend it on an array of active satellites.