The Road To Radar

On June 24, 1930, two employees of the Naval Research Laboratory near Washington, D.C., were trying to find out whether airplane pilots could follow the narrow beams of radio short-waves that a transmitter was sending out from a laboratory building. One of the men, Lawrence Hyland, had set up a receiver a short distance from the laboratory, near the Army airport at Boiling Field. His job was to watch a meter and record the strength of the navigation beams at various locations on the ground. Before long, however, something went wrong. The signal on his meter would go up and down unaccountably for a short time, then return to normal. Hyland inspected his equipment for loose connections and defective parts. Then he realized that the explanation was overhead: airplanes landing and taking off from the field were passing through the beams and reflecting a small part of the energy downward to interfere with the signal coming from the transmitter.

Hyland’s discovery, that distant airplanes could reflect enough radio energy to reveal themselves, was nothing less than revolutionary. It gave hope to admirals pondering the new menace of the torpedo plane and thedive bomber. And it directly stimulated the development of radar in this country, with the result that we had several models capable of spotting distant ships and planes before the British started sharing their radar secrets with us in 1940. Moving from radio theory to radar practice took a great deal of effort over more than a decade. In the United States the radar research that began as the part-time pursuit of a half-dozen scientists ended the war as a sprawling enterprise bigger than the Manhattan Project and six times larger than the whole pre-war radio industry.

Starting with Heinrich Hertz’s experiments using a spark-gap transmitter in the late 1880s, radio scientists had quickly learned that many substances reflect radio waves, and that higher frequencies reflect better than lower frequencies. The electrical inventor Nikola Tesla had proposed locating ships by radio waves as early as 1900, but the era’s transmitters and amplifiers were very feeble, and targets had to be either very close or very large to produce any measurable reflection. When serious reflection research began in the 1920s, scientists spent their time bouncing waves off large objects like ships and mountains—and the ionosphere.

Reflections off the ionosphere—the electrically charged atmospheric layer about sixty miles up that bends long-range radio transmissions around the earth—might not appear closely related to finding airplanes and destroyers, but they pointed the way. The British made the first height measurements of this layer in 1924, using continuous radio waves that caused interference patterns similar to those noticed by Hyland six years later.

A year later two Americans with the Carnegie Institution, Merle Tuve and Gregory Breit, began sending pulsed waves instead. A sudden burst of energy would be sent out, and a remote receiver would pick up both the direct pulse and its echo reflected off the ionosphere. As with radar later, the precise time separation between the pulse and its echo revealed the height of conducting layers in the ionosphere. A stream of pulses would be used, but the brief gaps between them would allow the clear reception of an echo after each. This made unnecessary the interpretation of complex interference patterns as with continuous waves, which were both sent and received at the same time. Tuve and Breit also anticipated radar in making some of their time measurements by using light traces on a cathode-ray tube—an absolute necessity, as it would turn out, for radar. Their articles influenced radio scientists around the world, especially in Britain, and within a few years pulsed waves came to be adopted in most radar research everywhere.

But for all the similarities of the Breit and Tuve system to radar, much more would be required for useful target detection. Starting about 1930, work in this country proceeded along three roughly parallel, largely independent lines—there was no pressing need perceived for the mobilization that would have led to a single unified program. The Naval Research Laboratory pursued a system for shipboard use; the U.S. Army’s Signal Corps Laboratories sought a plane detector and searchlight aimer; and private companies such as RCA experimented with collision-warning systems at microwave frequencies.

The possibility of long-range detection of targets by one means or another had intrigued the War Department for years. During World War I, Maj. Edwin Armstrong of the Signal Corps had worked on a method to find airplanes by homing on the high-frequency radio emissions that emanate from spark gaps in gasoline engines. Armstrong got as far as inventing the superheterodyne radio receiver (which would prove invaluable in radar later), but he couldn’t overcome the fact that it was easy to shield engines so they didn’t emit stray radio noise. Another wartime experiment, pursued through the mid-1930s, employed a parabolic mirror to concentrate infrared radiation from the plane’s hot engine onto a sensor. On a fine day it could pick up planes a few dozen miles away, but fog, smoke, or clouds made it useless.

Through the 1920s the Army sought to locate enemy airplanes at night by use of a sound locator. A bucketlike device amplified the distant drone of engines, and then a searchlight crew would start probing the vicinity; once the plane was pinioned by light, an optical range finder could give firing information to antiaircraft guns. Among the many problems with this system was the fact that since sound travels so slowly, by the time the engine noise reached the locator, a far-off target would have moved miles. Until 1932 the Army’s Signal Corps Laboratories remained sidetracked on infrared detection. Then the laboratories shifted to ranging with microwaves (1,000 megahertz and higher), but a powerful microwave transmitter was a long way off, so the Signal Corps changed its approach again and settled on working with frequencies from 100 to 200 megahertz.

In the meantime the Naval Research Laboratory (NRL) was radio-ranging with continuous waves, following the path opened by Hyland’s 1930 discovery. The Navy encouraged the effort but provided no extra money, and work was strictly part-time. In early 1934 the laboratory demonstrated to the House Subcommittee on Naval Appropriations a system that could warn of planes dozens of miles away but was unable to give precise locations. Worst of all, its transmitter and receiving antennas had to be so far apart there was no hope of fitting the contraption on a ship. The one great achievement of this continuous-wave radar lay in persuading the congressional subcommittee to give the laboratory a special appropriation for further plane-detection research.

At about this time the lid began to come off what had been a very secret project. An article written by three Bell Telephone scientists and published in a radio engineers’ journal of 1933 had described how they were able to detect planes by interference with continuous radio waves. The Navy group responded by filing for a patent, to protect their invention rights. Then, in 1934, a group of scientists from RCA began going around to radio-engineering conferences and demonstrating a collision-warning system using low-power microwaves. Shortly afterward the Signal Corps joined forces with RCA on microwave research.

By the time the Patent Office approved the NRL’s patent claim in late 1934, the outfit had gone beyond continuous waves; it had turned to pulses in search of a more compact and accurate system. (The earlier work would not be in vain; years later inventors would come back to continuous-wave radar, utilizing the Doppler frequency shift to pick out moving targets from a background.) The switch to pulsed-wave research was led by a young physicist named Robert M. Page, who had just started working full-time on the NRL radio-location project.

Page, the seventh of nine children in a Minnesota family, had first trained for the ministry but, on the urging of his college physics professor, transferred into science. He joined the NRL in 1927 and left forty years later with sixty-five patents in the radar field, including postwar innovations such as over-the-horizon radar. Nobody did more than Page to bring forth working radar from the mass of theoretical possibilities.

Page spent the end of 1934 preparing equipment for a December test of pulse ranging. The trial showed that his quick-and-dirty transmitter was adequate for the time being, but his receiver (like the transmitter, a standard model he had modified for the job) was not. It couldn’t shed the energy lingering from the transmitter’s pulse fast enough to leave a blank slate for the faint echo. Page spent the following year designing a very special receiver. Several thousand times a second it shut down during the powerful transmitter pulse and then sprang back to life to take an exceedingly feeble signal a millionth of a billionth the strength of the outgoing pulse and amplify it enough—without adding background noise—to show up on a cathode-ray tube. As Page struggled with this receiver, new high-frequency vacuum tubes came onto the market, and he modified them until they worked in his design.

His next test rig had the precision the earlier one lacked. But it required a transmission antenna 1,000 feet long, strung between two 200-foot towers, so now the effort turned to designing something that could fit on a ship. Page invented a duplexer that allowed transmitter and receiver to share the same antenna by short-circuiting the receiver during the transmission pulse. And the research group raised the radio frequency about six times as high, to 200 megahertz. Higher frequencies permitted smaller antennas and had the important side-effect of making the radar image sharper.

The result was a radar set tested in early 1937 aboard an old four-stacker destroyer, the Leary . Page and his assistants bolted various antennas onto a five-inch gun so they could be swiveled and elevated. “It was an awfully haywire rig,” the inventor recalls. “The transmitter and receiver were right out in the open on the deck over the galley, enclosed in a great big plywood box.”

When the NRL men returned the destroyer after borrowing it for a month, their conclusion was that the Leary ’s set was handicapped by its small antennas; if the transmitter was to see airplanes far enough away to be useful, it would need more power. Shortly thereafter a traveling salesman from a California company provided part of the solution: a new vacuum transmitter tube called the Eimac 100TH, designed to absorb amateur abuse. It would tolerate much higher temperatures and voltages than anything else available. Meanwhile the lab had come up with a design for wiring together multiple transmitter tubes in a ring; they called it a ring oscillator. The combination of ring oscillator and Eimac tube—“the tube is what really made the difference,” says Page—increased the available power a hundredfold.

An improved 200-megahertz set called the XAF went to sea aboard the battleship New York in 1939 and performed excellently, surviving bad weather and the shock of gunfire to spot ships at ten miles and planes at fifty. It even registered cannon shells and birds in flight. It looked a little ragged around the edges (sailors called the flat, wiry antenna the “flying bedspring”), but it was ready for the coming war. The Navy chose RCA as production contractor. RCA had fielded a competing radar set of its own during the same 1939 fleet exercises, but that set had had serious problems. The Navy ordered twenty near-copies of the NRL’s model, renamed CXAM, for battleships, carriers, and cruisers. The CXAM would help win the battles of the Coral Sea, Midway, and Guadalcanal.

By 1940 the Army Signal Corps had made about the same progress as the Naval Reseach Laboratory. Four years of work on pulse ranging at the Signal Corps Laboratories in Fort Monmouth, New Jersey, had resulted in two production sets: the SCR-268, a short-range model for steering searchlights and antiaircraft guns, and the SCR-270 and -271, a long-range early-warning system manufactured in portable and fixed versions. A 270 set at Opana Station, Hawaii, warned operators of the incoming Japanese airplanes on December 7, 1941. Unfortunately, the operators’ superiors mistook the attacking planes for a flight of B-17s due from the mainland. The radar device that would have settled that question, the “Identification, Friend or Foe” transponder, was still in development at the NRL. It came into use shortly thereafter and involved the transmission of a coded radar signal by any friendly plane being tracked.

For all the capability of its radar in 1940, the United States still had a long way to go. Existing sets were satisfactory for spotting ships but inadequate for locating airplanes and surfaced submarines. The antennas were much too bulky to fit in a plane, and the prospect of shrinking them enough was so dim that the laboratories hadn’t pursued airborne radar at all. Antiaircraft radar wasn’t accurate enough to direct flak guns effectively. What was needed was a breakthrough in microwave radar at frequencies of 3,000 megahertz and higher, where much shorter antennas could be used.

Among the host of problems to solve, the biggest was the vacuum tube at the heart of the transmitter. The power unit of a modern household microwave oven, at a thousand watts, puts out dozens of times more radio energy than the best pre-war vacuum tube on the 3,000-megahertz frequency. All microwave experimentation, at the Signal Corps and Naval Research laboratories, at RCA, and later at Bell Labs, ran aground on this power shortage.

Then, in the summer of 1940, Sir Henry Tizard arrived in the United States, at the head of a visiting British technical mission. Britain, aware of its vulnerability to air and sea attack, had pursued an intensive radar-development program since 1935, and Tizard had overseen the effort, producing a multitude of excellent radars in only five years-airborne, shipborne, portable, plus the chain of early-warning radars that enabled the RAF to win the Battle of Britain. But there were many battles yet to be won, and with great effort he had persuaded his superiors to permit the exchange of war research with the United States, in hopes of tapping this country’s production expertise. Tizard arrived by flying boat in August and opened the discussions. His staff followed two weeks later on the steamer Duchess of Richmond , bearing technical reports and diagrams, a formula for a new explosive, and—inside a black metal trunk immune to customs inspection—a small vacuum tube of fabulous value. It was called the cavity magnetron, and it made microwave radar a possibility.

The magnetron was no stranger to this country—a General Electric scientist had invented the first one in 1921 —but our researchers hadn’t been able to get the necessary amount of microwave power out of it. In late 1939 two British scientists at Birmingham University had devised a copper cylinder with six holes slotted to join a central cavity. Electrons coming off a cathode in the center swirled through these cavities on the way to the cylinder wall, which served as an anode, and resonated at microwave frequencies, with plenty of wattage available. The result was to transform an input of simple DC power into a copious output of microwave radio-frequency energy, ready to be broadcast. The new device still had some problems—the worst was a tendency to jump erratically in frequency—and the British hadn’t been able to apply it yet. The Americans whisked it off to the Bell Telephone Laboratories in Whippany, New Jersey, for testing.

The British also offered an improved display screen—the plan-position-indicator, or PPI, the well-known round screen with the streak of light that rotates around the center, painting blips of light for target and terrain. The Americans had developed the PPI simultaneously, but their model’s image faded too quickly. The British solution was a sandwich of two phosphor chemicals, which produced a bright and readable display.

But perhaps the most important item of all that the British brought was the lesson of how they had developed radar so quickly: they had set up a well-funded laboratory mostly staffed and led by young civilian scientists, who were then permitted to solve problems unhindered by strict organizational charts and security compartmentalization. The work was ruled by urgency, and designs often went into the field with minor bugs and loose ends. “Give them the third best to go on with,” said the chief scientist, Robert Watson-Watt: “the second best comes too late, the best never comes.”

Roosevelt’s new National Defense Research Committee was so impressed with the British team’s results that it quickly authorized a similar laboratory, on the campus of the Massachusetts Institute of Technology, in Cambridge, Massachusetts, and ordered it to develop three types of microwave radar as quickly as possible: a portable set for directing antiaircraft guns, one for precision navigation, and an airborne set for night fighters.

The Radiation Laboratory—so called, ironically, to conceal its real purpose and make it seem harmless—would achieve all this and much more. It grew bigger than anyone could have imagined. When it started work in November 1940 with a few dozen employees, all of them could fit into one suite of offices in Building Six, and they used the roof for testing antennas. It ended the war with almost four thousand employees—larger than the entire staff of MIT—and occupied fifteen acres of office space. It had sucked up one out of ten physicists in the country. Working with the British, the military laboratories, and big electronics firms, the “Rad Lab” spawned 150 different-radar systems.

Because radar is a complex, interacting system rather than a couple of breakthrough inventions wired together, there was plenty of work to go around. First the 3,000-megahertz magnetron needed attention, to solve the instability problems and also to figure out why it worked at all. That research led to an equally successful 9,000-megahertz magnetron by mid-1941. Microwaves don’t travel well over wires, so the larger radars required metal tubes called waveguides to pipe the energy between transmitter, antenna, and receiver. There were many new problems of antenna design, precision tracking, amplification, and noise suppression. In the receiver, silicon crystals had to replace certain vacuum tubes which were unusable at frequencies higher than 900 megahertz. A component that solved one problem would frequently cause other problems elsewhere. But the fundamental logjam had been broken. Later in the war the laboratory would expend more effort on production engineering, troubleshooting, and training than on problems of physics and electromagnetism.

By early 1941 the Radiation Laboratory was well under way on a 3,000-megahertz set to control antiaircraft guns. The first test model was so impressive that Army orders for it exceeded in value the whole Boulder Dam project. It became the SCR-584, a trailer-mounted unit that proved amazingly effective—against dive bombers at Anzio Beach, when the Germans were jamming older radars into uselessness; at the Remagen Bridgehead, where it tracked not only airplanes but also commandos swimming downriver; and in the defense of England, during the V-1 campaign, in which 85 percent of the buzz bombs were shot down with cannons controlled by 584 sets. (Part of the credit for this effectiveness against aircraft must go to the radar proximity fuze, a miniaturized radar set that could be installed in a shell and use the Doppler effect to determine the correct distance from a target for detonation. It entered service in 1944.)

Of all the varieties of radar device to come out of the Radiation Laboratory and enter warfare, a huge set called Microwave Early Warning, or MEW, was perhaps the most sophisticated. “The MEW was very, very successful,” says the physicist Albert Hill, who headed the Radiation Laboratory’s section on transmitter components. “It didn’t have too great a role, of course, coming as late as it did, but at times during 1944 it was extraordinarily important.”

MEW used an unconventional “leaky pipe” waveguide—a sort of electronic sprinkler hose—to deliver a whopping million watts of power along the trough of a cylindrical parabolic antenna eight feet high by fifteen feet long. The design work started in the spring of 1942, and problems arose immediately: stray lobes of radiation jetted off from the sides of the antenna, bringing back misleading echoes, and high power tended to arc across connections, melt the magnetron, and burn out the crystal receiver components. But by late November a model mounted on an MIT roof tracked a flying boat 177 miles beyond Nantucket.

In tests in Florida the next year, the set proved so powerful and thorough that the designers had to keep adding displays so more operators could monitor the many thousands of airplanes it was watching. Because time was short, the laboratory started building seven advance models in late 1943, as full-scale production geared up. The first one went into action on the southern coast of England. During D-day it caught a complete view of the air and sea battle, allowing tactical air controllers to truly see what they were doing.

The Radiation Laboratory closed down quickly after the war, and some of its inventions came too late to see battle. One of them, Airborne Early Warning, which used an antenna mounted on a plane to beam down information to a shipboard processing center, is the forefather of the modern Airborne Warning and Control System (AWACS), the familiar military transport with the large, flat radome on top.

Postwar publicity about radar was overshadowed by the growing importance of the atomic bomb. Radar scientists liked to say afterward that although the bomb ended the war, radar won it. Whichever was the victor, it is certainly true that our radar effort made astonishing progress in the fifteen years after Lawrence Hyland first noticed that passing planes were messing up his experiment at Boiling Field.