The rectangular frame-like object seemingly fastened above the pilothouse of the USS New York (BB-34) in 1938 wasn’t an oversized mattress spring or an early-model solar panel. It was the antenna for the XAF, the first radar set installed on board a major U.S. warship.
Successful tests of the new device—including three months of 20-hour-a-day operation for aircraft detection, navigation, and gunnery practice—convinced the Navy that radar would be a godsend. The awkward-looking, 17-foot-square antenna could reliably detect aircraft as far as 100 nautical miles out and spot surface ships 15 miles away. And it could track projectiles and falling shot while they were in flight.
Suddenly it was practicable to defend an aircraft carrier using fighters. Without radar, the carrier never would have had sufficient warning of an attack, and never would have been able to keep enough fighters aloft continuously to fend off or destroy attackers. Radar solved the problem by making surprise attacks far more difficult. It took several more years for the Navy to understand how to use radar for defense, but when it finally did—in 1943–44, by creating the combat information center—the new technology proved to be the key to victory for America’s fast-carrier task forces. Radar’s greatest triumph in World War II was the “Marianas Turkey Shoot” during the June 1944 Battle of the Philippine Sea, which destroyed Japan’s carrier airpower.
The United States and almost all the other major combatants had independently discovered the radar concept. The Naval Research Laboratory (NRL) began developing pulse radar in 1934 and started testing it in 1935. It was the NRL that hit upon the idea of turning the radar receiver off briefly during the transmission phase so operators could use a single antenna both to transmit and receive radar pulses. (The British, working at about the same time, required two separate antennas for their early radars.) By the end of 1936, the NRL’s invention was routinely detecting aircraft 60 miles away. And in December 1938, the NRL-developed XAF was installed on board the New York.
The Radio Corporation of America (RCA) produced a pilot version, labeled CXAM, using NRL’s design. The Navy installed six of these in July and August 1940 on board the carrier Yorktown (CV-5), the battleship California (BB-44), and the heavy cruisers Chester (CA-27), Chicago (CA-29), Northampton (CA-26), and Pensacola (CA-24). The antennas could all be tipped back to elevate the pulse-beam—presumably in hopes of measuring the elevation of a target—but the beam proved far too broad for that. A hastily assembled second model, the CXAM-1, eliminated the beam-elevation feature. Installation of the new model began late in 1941, on board the carriers Lexington (CV-2), Saratoga (CV-3), Ranger (CV-4), Enterprise (CV-6), and Wasp (CV-7); the new battleships North Carolina (BB-55) and Washington (CV-56); the older battleships Texas (BB-35), Pennsylvania (BB-38), and West Virginia (BB-48); the light cruiser Cincinnati (CL-6); and the large seaplane tenders Curtiss (AV-4) and Albemarle (AV-5). The follow-on full-production version—the large, flat SK-1—was visible on many U.S. warships during the remainder of the war.
The choice of ships on which the early CXAM models were installed says a lot about the various roles that the Navy hoped the new technology would play. The California was the Pacific Fleet flagship; the carriers operated away from the main battle line, in part to make it more difficult for an enemy to spot them; the cruisers were assigned to scout ahead of the fleet. All needed their own early warning of an air attack.
The CXAM-1 installation list began with the rest of the carriers, naturally, and it included both the California and the Texas, the flagship of the newly formed Atlantic Fleet. The two big seaplane tenders were considered near-equivalents of carriers because the Navy depended heavily on long-range seaplanes for reconnaissance and bombing. For both kinds of vessels, having search-radar capability not only helped detect enemy planes, it also enabled them to track and control their own aircraft.
At this early stage, the only display a CXAM could provide was an A-scan, which showed signal amplitude at various ranges. A large blip indicated an echo, and the operator could pick out its range. The antenna could scan (at a standard speed of five revolutions per minute), and operators could plot range against bearing, but there was no chart-like graphic presentation as there is with today’s plan position indicator (PPI).
CXAM operated at a relatively long wavelength (1.5 meters, equivalent to 200 megahertz), and its 14-degree-wide beam offered a resolution of three degrees. It could distinguish two objects 400 yards apart, and its range was accurate within 300 yards. It could easily detect a fighter flying at 10,000 feet at 50 nautical miles.
As it turned out, that wasn’t good enough. By 1945, with jet-propelled aircraft in the offing, the Navy was looking for a way to track planes more precisely—and that meant using short-wavelength transmitters that could send out narrow radar beams. On came a string of short-wavelength successors to the CXAM—the SR, SPS-6, and SPS-12. It seemed as though the 1938-era system was on its way out.
But when these new models were tested, they produced surprising results. The new, more streamlined jet fighters reflected these short-wavelength pulses away. So while a jet carrying external bombs or fuel tanks was relatively easy to detect and track, one that wasn’t carrying such bulky appurtenances was more difficult to see.
The radar experts at NRL understood what was happening. The longer the radar wavelength, the less the signal is affected by details of target shape. As a result, the Navy developed new long-wavelength radars—the SPS-17, SPS-29, SPS-37, and SPS-43—whose antennas were reminiscent of the “big mattresses” of the 1930s and 1940s—with an extra-wide antenna version for carriers and guided-missile cruisers. These later were traded for the short-wavelength SPS-49, which is still in service. It made up for any deficiency in detecting streamlined jets by adopting much better signal processing.
The underlying physics has not changed: It is still true that the longer the wavelength, the less the radar is affected by the shape of its target. Shape also is the most important key to stealth. The less a radar system is affected by shape, the more likely it can detect stealthy aircraft. That may well be why Chinese warships often still sport really long wavelength radars, operating at frequencies far below those used by CXAM and its successors. Their beams are far too broad for good definition, but theoretically they can detect stealthy aircraft and missiles. There is a vast difference between a radar operator who is searching for targets that may not be present, and one who is seeking to refine position data on a target that the operator knows is there, within an area defined by a broader-beam set. With the increasing use of stealth technology, we may once again think of CXAM not merely as a historical artifact, but as a forerunner of today’s innovations.