Page 242

Chapter 9: Signal Equipment: Radar (June–October 1942)

Airborne Radars on the Increase During 1942 airborne radars multiplied, and multiplied again, as newer and better sets emerged from the laboratories, tending always toward shorter wavelengths. The seemingly endless variety of their types and applications, the overwhelming quantities in which the Air Forces demanded them, amazed and sometimes stupefied the initiate hardly less than the occasional outsiders who were favored with glimpses of these highly secret developments.1

IFF—Identification: Friend or Foe Radar

One of the airborne radar types to which BOLERO,, the American build-up in England, gave an immediate and mighty boost in 1942 was IFF. American aircraft flying over Britain had to be identified in the oscilloscopes of the ground radars which kept vigil over all the approaches. The agreed -upon IFF for BOLERO was the British II, which, pending the development of the more advanced IFF Mark III, the Signal Corps copied as the SCR-535. Colonel Marriner in the Air Forces Communications Directorate had assured the British on 8 May that “the American SCR-535 be installed in all aircraft assigned to BOLERO in lieu of the British Mark II (R. 3003).”2

The SCR-535 could pick up radar beams on the SCR-270’s frequency band and also on the bands of both the SCR-268 and the numerous British radar types. It threw back an intense echo, which would flash prominently on the oscilloscopes of ground search radars and thereby serve to identify the aircraft as friendly. Despite production troubles and shortages, Philco produced thousands of SCR-535’s during the year. By 30 July 1942 the Air Forces had on order 18,000 of these radars.3 Although IFF

Page 243

Mark II and SCR-535 were stopgap sets, serving in place of the more universally usable IFF Mark III, which the Allies had agreed to adopt ultimately, the 535’s met all immediate requirements and their production kept up with the new aircraft, so that by the year’s end the Signal Corps could assert that no airplanes had been delivered to combat areas without them.4 Yet by then the SCR-535 was yielding to the Mark III SCR-595 and 695. As early as October, after purchasing some 38,000 SCR-535 components and after getting 18,000 SCR-595’s through the Navy, the Signal Corps had also procured 62,000 SCR-695’s.5 The British Mark III would become the universal Allied IFF of World War II, replacing both Mark II and the so-called Mark IV, the last being IFF of American design—the American radio recognition sets SCR’s-515, 532, and 533, which were held in readiness in case Mark III sets became compromised through capture.

Signal Corps Altimeters; Secretary Patterson’s Objections

High-level bombing had always held a top place in Air Forces thinking. Therefore another airborne radar application that began to assume great importance as American aircraft moved outward to carry the offensive to the enemy was the radio altimeter, actually a radar whose downward rays and upward reflections could give exact clearance above the ground or water below. The year 1942 saw rapid developments touching the American pulse altimeter SCR-518.6

The Signal Corps had ordered 11,579 of these heavy 100-pound altimeters which were effective to at least 20,000 feet. It had ordered them well before Pearl Harbor on a letter contract with RCA dated 3 September 1941. Meanwhile, Signal Corps engineers in the Aircraft Radio Laboratory developed an improved version, the SCR-618, efficient to 40,000 feet and weighing only about 60 pounds. On the Air Forces’ request, the Signal Corps now reduced the SCR-518 order to 6,114 and in February issued a letter of intent to RCA and a subcontractor, Stewart-Warner, for 15,000 SCR-618’s. In mid-1942 the AAF placed its entire 1943 altimeter requirements at 24,527 SCR-618’s and only 7,989 SCR-518’s.7

By this time research successes again further complicated procurement problems by developing a still better and lighter weight set, the SCR-718, which weighed only some 30 pounds. The fighting men must have the best, and once more the AAF altered its altimeter orders. It now asked for several thousand 718’s and cut its SCR-518 to 2,000 sets. This the AAF did on 20 September. Three weeks later, on 14 October, it asked the Signal Corps to stop

Page 244

The SCR-718, radio 
altimeter

The SCR-718, radio altimeter

production altogether on the 15,000 SCR-618’s which had been ordered in February.8

Thereupon, procurement people in the Signal Corps and in the Office of the Under Secretary of War protested that established programs, factory dispositions, and money spent on tooling and materials were lost or largely wasted. But such was the inevitable toll exacted by progressive technology. In this particular example of lost motion and money, General Somervell, head of the Services of Supply, getting his figures no doubt from the Signal Corps, estimated that $960,000 in critical materials, now abandoned in a partly fabricated condition, had been wasted; likewise, $210,000 in tools, jigs, and dies—to say nothing of a loss of 70,000 man hours spent in fabricating the wasted material. He estimated, too, that it would cost $472,500 to change the factories over to other types of manufacture. On top of all this he piled “cancellation cost and tool costs to the United States in excess of $1,600,000, with nothing to show for it.”9 This case, and several others like it, all bore out a prophecy which Col. Tom Rives, head of the Radar Division in the Office of the Chief Signal Officer, had made in April 1942, when he said, “The Air Corps’ plans for use of signal equipment are changing so rapidly that purchases based on past

Page 245

requirements are likely to cause appreciable trouble.”10

For exact determination of altitude at low levels, the pulsed high-level altimeter, such as the SCR-518 and SCR-718, would not do. At levels measured in the tens and hundreds of feet of altitude, a pilot had to have another kind of radar altimeter, the continuous-wave, frequency-modulated (FM) type. In this radar variety, the Air Forces had previously taken slight interest. But now in mid-1942, submarine warfare all of a sudden menaced American shores, and the airmen found themselves called upon to help combat the new peril. They needed the low-level altimeter for pressing close-in attacks upon subs. “Our operational people have indicated,” Col. Robert G. Breene, director of the AAF Technical Services, informed the Joint New Weapons Committee on 29 June, “that an accurate absolute altimeter of this type is almost as important as the ASV equipment.” Navy men had had such an altimeter developed and in production, their AYB, extremely accurate from zero to several hundred feet and especially useful to them as a landing aid for carrier-borne aircraft. By mid-July Colonel Marriner had asked General Olmstead for 1,000 sets to meet the new needs. On 21 September the device received approval for standardization from the Signal Corps Technical Committee and became the RC-24, subsequently the AN/APN-1.11

Thus, still another radar joined the large numbers of electronic gadgets crowding the Army’s airplanes. Airborne radar and radio were multiplying almost by the day, to the pleasure of some, to the dismay of others. Electronic diversification into myriads of new military applications merely paralleled in the air what was happening on the ground—the whole constituting a tide which rose as inexorably against the command of military lords in World War II as the ocean had once defied King Canute.

A year earlier, shortly before Pearl Harbor, the Chief of Staff, General Marshall, had given Olmstead an inexecutable order—to simplify and reduce Army’s radio types.12 Obediently, Olmstead had sought to comply, but his efforts were in vain. Now, in September 1942, Robert P. Patterson, Under Secretary of War, repeated Marshall’s wish, with respect to airborne equipment especially. He directed his comments not to Olmstead directly, as Marshall had done, but to the Chief Signal Officer’s new superior, General Somervell, who of course merely passed them down to the Signal Corps for reply. Specifically, Patterson wrote that he constantly questioned “the need for so much and so many types of radio, especially for airplanes.” He expressed the hope that David Sarnoff, president of the Radio Corporation of America, who had recently joined the War Department for a 30-day tour of duty as a

Page 246

Signal Corps colonel, might help to review air electronic military requirements, in the interests both of decreasing airborne weight and of easing the shortages which plagued factory production. Acquiescing in the demand, General Olmstead’s deputy, General Code, suggested possible simplifications. Code listed the twelve items which the Signal Corps was currently installing in bombers, the eight items that went into transport planes, and so on, down to the single interphone set installed in primary trainers. He then proposed various reductions, especially in the nine types of radio compasses, which he thought might be simplified into two. As for altimeters, he suggested eliminating the high altitude SCR-518 from airplanes equipped with ASV radar and using the low level RC-24 only. On the whole, Code felt that the “situation is well in hand,” for already the number of aircraft radio types had been reduced from forty-one to sixteen. Presumably he was thinking only of radio communication and navigation sets, not of radars, which were now irresistibly on the increase.13

The AAF responded at once, and vehemently. It rejected any notion that airborne electronic devices be simplified or reduced. It wanted more, not fewer, of them. Brig. Gen. Harold M. McClelland, in Air Forces Headquarters, after reading a copy of Signal Corps’ proposed simplifications, wrote emphatically to the Chief of the Air Forces. General Code, McClelland commented, was giving Somervell and Patterson a false impression. A reduction in the number of radio compass types, for example, would not reduce the total quantity needed and would not therefore alleviate the load on industrial production, which was Patterson’s chief concern. Moreover, Maj. Gen. Carl Spaatz and other air commanders in the field wanted more and more equipment, of all types. “Send everything,” was the tenor of their reports, “we will decide what to leave out”—a demand which, McClelland dryly observed, offered no relief to production problems. He agreed with Spaatz that the Air Forces had to have more radio and radar, and he passed on to General Arnold his belief that those responsible for producing this equipment had “held their sights too low.” The Air Forces, he believed, far from following the Signal Corps’ effort to comply with Patterson’s wishes, should demand greater production. McClelland was emphatic:

Insist that the SOS (Chief Signal Officer) take steps (similar to those taken by you when you saw the need for more aircraft plants) ; which will result in the provision of production facilities to meet the increasing demand for radio equipment. It is apparent to me that those responsible for procurement of radio have held their sights too low and have not anticipated demand. The development of radar is finding increasing application to airborne use. Such use is very materially enhancing the potency of the airplane e. g., sea search, blind bombing, air gun laying, etc. We are certainly not going to go without these important devices and we can expect no compensating reduction in the normal communication sets. Therefore, we should face the fact that we need increased production facilities, make the best estimate we can, and start getting new plants built.14

Page 247

AI—Airborne Interception Radar

Not until the late spring of 1942 had Signal Corps’ copies of the puny British AI IV (long wave) begun at long last to come off Western Electric’s production lines as SCR-540. Only 580 sets had been ordered; no more were wanted, so the AAF had asserted in May.15 Late in June the AAF issued a priority schedule for delivery of the first SCR-540’s: to Navy, 3 sets; to the ARL, 1 set; to the Douglas Aircraft Company, 1 set; to the Hawaiian Department, 3 sets; to the Signal Corps School, 6 sets; to the AAF Technical School No. 2, 6 sets; to the Western Electric Training School, Kearny, New Jersey, 5 sets; then 6 more sets to the AAF Technical School No. 2. Finally, after all the above had been delivered, 59 sets would go to the Douglas Aircraft Company.16

Airmen installed their 540’s in oversize fighter planes such as P-70’s, converted from the twin-engined A-20 attack bomber. By the end of August, 25 sets had been marked for Hawaii, consigned to the signal property officer at the air depot there. Ten days later, Colonel Marriner asked the Signal Corps to allocate another 25 sets to Hawaii for spares. Simultaneously, he asked also for 7 spare sets to go to Panama. On 12 September the Navy Bureau of Aeronautics asked for 12 sets for the Marine night fighter squadrons.17

No sooner did the airmen begin to receive this highly complicated airborne equipment than dire difficulties respecting maintenance arose, difficulties touching all airborne radar alike, IFF and ASV as well as AI. Two Signal Corps lieutenants who had served in England as members of the Electronic Training Group and who were now contributing their experience to the ARL recommended to Colonel Bayer, the chief of the Radar Division at the laboratory, that each Air Corps squadron be assigned maintenance men who could make minor repairs on the spot. In their opinion, each squadron should adopt the British practice and set up a maintenance section of a score or so men equipped with the “megger” (which was a special device to test insulation), together with test set I-48, the multi-range volt-ohmmeter TS-189, and suitable signal generators and oscilloscopes. Colonel Gardner, director of the laboratory, agreed, excepting only the signal generators and oscilloscopes which he believed should be reserved for maintenance work at depots.18

Page 248

Test and maintenance equipment constituted but one trial. Another, especially severe in airborne radar, was training equipment. Airborne radar required that operators be well trained, or else the costly equipment was useless. Yet to train men in actual flight was out of the question. There were not enough airplanes, and what radar existed was desperately needed for patrol and combat operations. The answer lay in sets which could simulate airborne operation in a classroom. The British had devised such radar trainers. The ARL, also, had long before asked the Link Aviation Devices Company, manufacturers of the well-known Link trainer for pilots, to develop a radar crew trainer, RC-110, to train men in AI (SCR-540) operation. But before production of the trainer had begun, the 540 itself was becoming obsolete for air intercept, being replaced by the much larger microwave AI radar, the SCR-520, which called for a quite different trainer set. Because this new radar was already well developed, Colonel Marriner was able to cancel further work on the development of the 540 trainer.19

By October 1942, therefore, America’s first air intercept radar, the SCR-540, was obsolescent. Further, it suffered from many defects. Maj. James W. McRae, charged with the Airborne Radar Section in the Radar Branch of the Office of the Chief Signal Officer, reported concerning SCR-540’s which had been installed in P-70’s at the Orlando air base, Florida, that though “the external appearance and general impression created on first examining the SCR-540A equipment is one which augurs well for the future ... unfortunately it is not borne out in practice.” He attributed many breakdowns to inferior materials used in its construction, especially in the cables interconnecting the components of the radar. He noted defects in the cathode ray tubes which were “seriously astigmatic,” so that accurate focusing of target echoes was impossible. Some breakdowns he blamed upon bad installation (as in the case of the VHF command radio, SCR-522) and others attributed to poor workmanship or even sheer carelessness on the part of crews.20

Fortunately, night fighter aircraft no longer had to depend upon the long-wave SCR-540 to hunt down enemy planes. Since early in 1942 the Radiation Laboratory, where scientists worked on radar for Division 14 of the National Defense Research Committee, had been building numbers of its prize AI-10. The Army’s version of this big microwave American radar, the SCR-520, was so superior to the older British long wave AI IV that the British themselves pressed eagerly to get quantities of it in a special form suited to their planes, designated SCR-520-UK.21 When factory

Page 249

production began in late spring, the Air Forces set up priorities for the first deliveries: the first two sets to the Northrup Aircraft Corporation and thereafter one to Western Electric (the manufacturer), one to the Douglas Aircraft Company, thirteen to the Air Forces Technical School No. 2, six to the Signal Corps School, and four to the Signal Corps Maintenance School at Kearny, New Jersey.22 Most of these SCR-520’s were destined to be delivered, not as AI’s, air intercept radars, but as air-to-surface-vessel sets. They would be converted into SCR-517’s, to serve in the capacity which the immediate needs of the war most urgently demanded—aircraft search for enemy submarines.

ASV—Air-to-Surface-Vessel Microwave Radar

This most important airborne radar development during 1942 was directly stimulated by the necessity to clear the seas. ASV, air-to-surface-vessel radar, was the electronic instrument which contributed more than anything else to the defeat of German submarines. ASV first helped in the attack on raiders along the Atlantic seaboard; then, in British hands, it aided the passage of the Allied invasion fleets through the Bay of Biscay and on toward North Africa. Now, in mid-year, production of the original longwave ASV, British ASV-II, copied by the Signal Corps as SCR-521, was well under way. But the microwave ASV was a much better set, and by September the AAF was planning to replace ASV II (SCR-521) with the new SCR-517.23 Combat needs could not await factory production of the SCR-517. The British received the Laboratory’s prototype microwave ASV, DMS-1000—mounted in a B-24—dubbed Dumbo I, followed by several more crash-built sets, that is, hastily handmade by the laboratory’s own shop, the Research Construction Corporation. Dumbo I, and its dozen or so successors, called Dumbo II’s, went to work in the hands of the English against subs in the Atlantic and in the Bay of Biscay late in 1942, winning British acclaim as “the first substantial radar contribution made directly to Britain’s war effort by the United States.”24

Secretary of War Stimson and his technical advisers had sensed early in 1942 the great importance of ASV radar. So had the AAF which, burdened with the general responsibility for shore-based air operations, was very much interested indeed. When the Radiation Laboratory fitted its first experimental microwave ASV into a B-24 Liberator bomber and flew it in the spring of 1942, Stimson himself rode on a demonstration flight. Henry Guerlac, historian of the Radiation Laboratory, writes that when the scientists took their ten-centimeter ASV to Washington in April 1942:

... not much interest was shown in the higher reaches of the Pentagon until the Secretary of War was induced by his radar adviser to fly with the equipment. This he did, before his generals had responded to the invitation, in a plane loaded with depth charges in case an enemy submarine should actually be sighted. Without difficulty the radar located a ship and the plane was able to home on it so that the Secretary could look out of the window and see the results of the pursuit.

Page 250

He was convinced. “That’s good enough for me. Let’s go home.” The next day General Marshall and General Arnold found identical notes on their desks from Mr. Stimson saying, in effect: “I’ve seen the new radar equipment. Why haven’t you?”25

The Navy did not welcome the use of Army patrol bombers to sweep the Atlantic shipping lanes. Admiral Ernest J. King stated to General Marshall that in escort ships alone lay the promise of success against U-boats and raiders. The Navy doubted the efficacy of aircraft as submarine killers and wished to restrict Army aircraft to the control of the several Sea Frontier commands, to be employed by the Navy to cover convoys.26 But ASV-equipped Army aircraft emphatically disproved the Navy’s doubts, beginning with the first ten preproduction ASV-10’s, which had been hastily converted from AI-10’s and mounted in B-18 bombers during the early months of 1942.

By May Western Electric, the manufacturer of the first microwave sets, was struggling to convert most of its production of AI-10 (SCR-520) into ASV-10 (SCR-517-A). Production sets 3 through 102 were thus to be converted. “This action,” wrote 1st Lt. Wilbur H. Vance, Jr., in the Radar Division of the Office of the Chief Signal Officer, “was directed by the Secretary of War approximately April 7, 1942, because only 2 SCR-520 (AI-10) equipments are needed during the months of April and May, whereas 100 ASV-10 equipments can be used during the same months.” Actually it was to be well past mid-summer before production totaled 100. The conversion, Vance explained, was a matter “of changing the short range AI equipment long range surface vessel detection equipment. The modifications necessary consist of changing the speed of rotation of the Spinner and the angle of tilt of the Reflector. The production of the SCR-517-A equipment,” Vance added, “has been made greater than the original delivery schedule planned for the SCR-520 equipment because of the higher priority rating assigned to the converted equipment.”27

The work progressed slowly, both because severe engineering problems had to be solved and because changes were introduced into the equipment itself. Large-scale production would come, but with agonizing slowness, only as submarine outrages multiplied. Production engineers and Army officers who blandished each other and the War Department heads with fair promises either did not realize how complex ten-centimeter ASV radar was or else were indulging in wishful thinking.

For example, on 30 January 1942, Robert A. Lovett, Assistant Secretary of War for Air, had written to Colonel Marriner saying he understood the performance of the American microwave ASV was “very excellent.” “Will you please advise me,” he asked, “how many of these we have on order and how soon we are to receive the first items. I am told,” he added, “that the British are putting in for a large number and I want to make sure that production is

Page 251

accelerated to meet our urgent present needs, even if we have to use hand-built sets at the outset and until the production line comes into being.” Marriner replied, giving the delivery schedule which Western Electric had set up for the 50 sets on order: 10 to be delivered in February and 20 in each of the next two months. Then he added brightly, “Believe this schedule will be met with no delays.” Two months later there was still no production of ten-centimeter ASV’s. Thereupon, at the end of March, Brig. Gen. Bennett E. Meyers, executive officer of the AAF’s Materiel Command, addressed a frantic note to Marriner. Lovett was on the warpath for ASV. “Mr. Lovett,” Meyers wrote, “is very concerned about ASV equipment ... will you please,” he besought Marriner, “give me a squib that I can hand on to Mr. Lovett. I promised this information today but I had been unable to get together with you. Please take a couple of minutes and write up a brief for me.” Meyers explained Lovett’s impatience touching the ten-centimeter ASV: Lovett now was expecting that the first production set would be ready by 15 May, that 15 sets would be built by 15 June, 300 sets every month thereafter!28

Radar engineers and production experts at Western Electric had their hands full with Signal Corps’ two microwave AI and ASV radars, the 520 and the 517 respectively. By the end of May Western Electric had given the Signal Corps to understand that it could provide the Navy during June with 60 SCR-517’s, designated ASC radar, followed by 40 in July; the Army Air Forces with 28 SCR-520’s in June, 35 in August; the British with 60 SCR-520-UK’s in July, 100 in August, and 40 in September. All of these figures were thoroughly reshuffled several times during the summer of 1942, as demands shifted from AI (SCR-520) to ASV (SCR-517) and as production goals proved totally impossible. By 30 June, for example, “... the delivery dates on the SCR-520-UK [had] been changed a number of times because of their interference with delivery of the SCR-517-A. ... The Office of the Secretary of War and the Army Air Forces [had] indicated that still more SCR-517-A’s [might] be ordered, in which case, the schedule for delivery of the British SCR-520-UK equipment would again be changed.”29

Meanwhile a new customer had appeared among those who desired the SCR-517 airborne sea search radar. It was the Navy. Yet the Navy did not want SCR-517’s for airborne use at all. Instead, it intended to mount them on small subchasers. This usage would have sharply curbed the potentiality of ASV. Obviously an airplane flying a mile high could scan a far greater area than could any surface craft from masthead height. Nonetheless, the Vice Chief of Naval Operations, Rear Adm. Frederick J. Home, wrote to General Arnold in May asking that a number of SCR-517’s be diverted for use in naval submarine chasers.30

Page 252

Accordingly, Colonels Rives and Metcalf of Signal Corps Radar Division conferred with representatives from the Air Forces, from Western Electric, and from Navy’s Bureau of Ships. The Navy group, headed by Rear Admiral C. A. Jones, “explained that there was an extreme need for Radar Equipment on small surface craft to be used in searching for submarines on our East coast and the Gulf of Mexico.” Naval officers had found that ASV-10, in the form which Western Electric was building as SCR-517-A, could “be used with considerable effectiveness on shipboard for the detection of submarines.” The Navy had quantities of long-wave shipboard radar on order as SW2C from War Supplies Limited but admitted that “the Army SCR-517-A is about twice as effective as these are.” Further, the naval officers explained that “the delivery of the SW2C is far behind the production rate of the ships ... acquiring SCR-517-A equipments would help the Navy solve the present shortage in the most rapid way possible; and also will give them an instrument which is more effective against submarines.” When they asked Rives if the Signal Corps would give them some sets, one hundred in fact, he naturally replied that that was a matter for the Air Forces to decide.

This request for sets the AAF did agree to, a considerable favor in view of the fact that airmen were already sending out antisubmarine bombers equipped with ASV-10 under Col. William Dolan’s command. It would seem a waste to install any of the valuable airborne search sets on surface vessels. The ranking AAF representative present at this May conference, Col. J. K. DeArmond from the Directorate of Communications, granted “that the Air Forces would release these 100 SCR-520’s [to be converted to 517’s] provided that first, the delivery schedule of the 100 SCR-517’s ordered by the Army was not altered; secondly, that a quantity of 25 or 30 SCR-520 equipments were delivered to the Army for service tests before delivery began on the quantity of 100 SCR-517-A’s to be delivered to the Navy.”31

The Navy was not blind to the idea of submarine search from aircraft, however. Admiral Jones and his subordinates explained at this conference that they too had been getting some microwave ASV equipment, which Western Electric was building as Navy Type ASC, an ASV-10 set almost identical with the Army SCR-517-A. Five sets were expected in June, 5 in July, none in August, 25 in September, and 55 in October.32 They would be getting it, at this rate, much more slowly than the Signal Corps and the Army. But following Admiral Home’s request and Colonel De Armond’s acquiescence, the AAF agreed that 100 “of the 300 SCR-517-A equipments so obtained ... be turned over to the Navy. ... “ By now, in mid-1942, the AAF total requirement for improved types of the SCR-517 (B and C versions) stood at 2,450.33

Meanwhile, nobody was getting many microwave ASV’s or AFs either, principally because one essential component was in short supply. This was the spinner, the rapidly rotating and highly accurate antenna mechanism which enables the narrow beam of a microwave radar to scan a wide area. On 17 June 1942, at another conference with

Page 253

Navy, Air Forces, and Western Electric representatives, Signal Corps spokesmen “explained that the schedule for the delivery of SCR-517-A’s to both the Army and the Navy was being held up because of the lack of Spinners produced by the General Electric Company. The Spinner Production,” they added, “has been far behind the promised schedule because of the lack of certain vital tools and machinery and it has been necessary to divert these machines from other users by obtaining higher priorities for them. Philco has been set up as a secondary source of Spinners.”34

According to Comdr. A. M. Granum from Navy’s Bureau of Ships, Secretary of Navy Frank Knox, and his assistant, James V. Forrestal, together with Admiral Jones, had given notice to General Electric that the production of the SCR-517-A’s, for use in PT boats, “was a national emergency.” The Navy wanted to divert a number of the 517’s from the AAF that very month of June, but Colonel DeArmond asserted that “there were 11 Army airplanes completely equipped with all SCR-517-A parts except the Spinners,” and he added that “this deficiency will have to be supplied before any Spinners can be diverted anywhere else.” He did make some concessions, so that in the end the Navy might get some of the earlier types of microwave ASV, the early, 750-pound version, SCR-517-A, while the Army would receive the SCR-517-B and C, the latter of which would be much lighter in weight.35

By August Fred Lack, vice president of Western Electric, wrote to Col. William M. Mack, in the Wright Field Signal Corps Procurement District, saying that his company had delivered, as of the first of the month, 199 sets of SCR-517-A and 4 sets of SCR-520-A. General Electric’s production of spinners and antennas had improved during July, Lack said; but to meet the Signal Corps requirements of 1,142 sets in 1942, 939 more would have to be built. And as usual, obstacles to this rate of production had already risen; for example, the Army had asked that beacon and IFF features be added to the SCR-517-A at once, constituting a new version, SCR-517-B. This was in addition to a “small package set,” the SCR-517-C, delivery of which was to begin in October. Naturally, these changes did not simplify production problems. Lack asked for more help from Signal Corps expediters in order to increase his company’s supply of mica and of such components as switches, condensers, and potentiometers.

By 3 October 200 517-A’s had been procured, inspected, and shipped, 100 each to the Air Forces and the Navy; 48 of the new 517-C’s had also gone to the Air Forces. Colonel Marriner informed the Chief Signal Officer on 11 September that 91 517-A’s

Page 254

had come to the Air Corps, 62 of which had been installed in B-18A bombers; the remainder had gone to schools or had been used as spares because—Marriner added the universal complaint whenever new equipment came in—there were no spare parts. The Air Corps had to cannibalize some of its new ASV’s in order to provide spare parts to keep other sets in operation.36

Thus all through the summer of 1942 ASV production problems sputtered and smoked. There was too much promising, too little production. Air Forces officers prepared a case history for the Services of Supply at the end of the year, checking off against the fair promises the dismal quantities actually delivered. The report commented:

... only properly evaluated production schedules should be set up. Production schedules were established which actual deliveries of equipment have never paralleled. Through failure to receive the equipment as scheduled, it has been impossible to vigorously prosecute that campaign of anti-submarine warfare which is dictated by the sinkings of surface vessels along our sea frontiers and in other areas in which we are vitally interested. Failure to receive equipment on scheduled dates results in loss of use of aircraft or ineffective use. When installations are contemplated, it is necessary to order airplanes to depots in advance of receipt of equipment. Failure to receive the equipment as scheduled requires the aircraft to remain at the depots for unnecessarily long periods during which time the Air Forces lose both the services of the crew and the plane.37

ASV production during the summer of 1942 had probably been about as good as the state of the radar art and the experience, or rather inexperience, of both the manufacturers and the military users allowed. The demand was terrific, especially on the part of those who did not comprehend all the difficulties. Meanwhile AAF users, under the spur of Secretary Stimson and his adviser, Dr. Edward Bowles, worked out the tactics whereby Army bombers might best thwart the submarine menace. The bombers would fly over the sea to seek out submarines, not defensively merely, accompanying convoys to protect them, as the Navy wished, but offensively, sweeping the seas far and wide. Once a submarine was discovered, if not immediately bombed to destruction it could be hounded. It would have to come to the surface eventually to recharge the storage batteries upon which it depended for underwater propulsion. If continually forced under, it would be rendered helpless.38 A scant ten B-18’s had been working out such operational tactics under Col. William C. Dolan, using the first preproduction ASV-10’s, that is, SCR-517-A’s converted from the ten-centimeter AI radar SCR-520. General Arnold had told Secretary Stimson that Dolan’s pilots flying four ASV-10’s from Langley Field and two from Jacksonville, Florida, were “enthusiastic over the possibilities of this equipment.” A month later Dr. Bowles urged the Secretary to set

Page 255

up a “special bombardment group for submarine destruction.”39

Bowles, backed by Dolan’s ten B-18’s and their repeated successes in sighting and attacking submarines, got action. By the summer of 1942, the Army set up the Sea Search Attack Development Unit (SADU), soon renamed the First Sea Search Attack Group, “at Langley Field to be an operational testing ground for new types of antisubmarine weapons. It is planned,” Capt. A. B. Martin wrote on 25 July, “that it be of an experimental nature. However, at the present time due to the concentration of ASV equipment and trained personnel at Langley Field a certain amount of training of ASV-10 operators and mechanics is being carried on. In addition SADU is called upon from time to time by the First Bomber Command to undertake tactical antisubmarine missions.”40

By the end of 1942 Secretary Stimson’s firm stand on behalf of ASV radar for patrol airplanes, and not just for seacraft, seemed vindicated. “Experience in combatting submarine activities along the North Atlantic and Caribbean sea frontiers,” an AAF spokesman wrote on the first anniversary of Pearl Harbor, “has demonstrated that the most effective means of curbing this type of enemy activity is through the use of radar equipped patrol aircraft. Such activities have been materially reduced in those zones in which the Army Air Forces have been in a position to provide airplanes equipped with sea-search radar sets for patrol work.”41

Although ASV radar would soon fall from its high place in Air Forces and Signal Corps interests, going over to the Navy, it nonetheless set off a long train of developments. In particular, it led to blind bombing of land targets by air, BTO (bombing-through-overcast) radar. Even the earlier, long-wave ASV-II (SCR-521) had proved valuable for “seeing” the ground. So General Stoner, Acting Chief Signal Officer, informed General Arnold on 22 August 1942, adding that he now had on hand some 250 complete sets ready to ship wherever the airmen might want them.42 In the Aleutians the SCR-521 had proved not so much for vessel search as for locating islands and coast lines, and so aiding navigation amid cloud and fog. “All airmen in the Aleutians,” commented one AAF officer, “are enthusiastic about ASV radar, with which they can ‘see’ mountains and prominences and thereby locate themselves. Their interest in ASV for search purposes,” he added, “appears secondary.”43

Page 256

At about the same time General Olmstead asked that the National Defense Research Committee set up a project at its Radiation Laboratory in order to “determine the most desirable procedure for the use of existing ASV equipment over land.” The radar maps which the Plan Position Indicator, PPI-type oscilloscope, made possible, “would obviously be useful,” he wrote, “guiding ASV equipped bombers to particular cities or targets.” Three-centimeter gear, he suggested, would be even better than the existing ten-centimeter radar.44

Thus the last months of the year 1942 found ASV microwave radar, far from being an end product, rather the beginning of more and more microwave developments in airborne applications. Ten-centimeter equipment would yield to sets operating on less than one-third as long a wavelength—on three centimeters or about one inch—whereas the wavelength of Signal Corps’ first radars was measured in feet. Numerous new microwave radar types were on their way, such as ETO, bombing-through-overcast; LAB, low-altitude-bombsight; ARO, airborne-range-only; AGL, airborne-gunlaying. All these constituted luxuriant proliferations of the flourishing radar growth which would confound the critics of electronic expansion in the air and bemuse even the advocates.

Ground Radar Potentialities Multiplied by Microwave Techniques

American ground radars for air defense—first, searchlight directors (SCR-268); next, long-range aircraft detectors (SCR-270 and 271); then, following British experience, radars suitable for GCI (SCR-516, 527, and 588)—all formed but one family of radars, all for use on the ground but all in some way involving aircraft.45

SCR-296, Seacoast Artillery Fire Control Radar

There was another application for ground radar, unrelated to aircraft, in which the Army, specifically the Coast Artillery Corps, had had an interest since 1937. This was a radar which could assist the Coast Artillery Corps’ mission of coastal defense and which could detect approaching vessels, in the first instance giving advance warning and in the second providing range and azimuth data to shore batteries. In 1937 the Coast Artillery Corps had asked for such a set. The Signal Corps had typed it as SCR-296.46 But then, deferring to the greater need for air defense, specifically for the need to develop long-range detectors of aircraft, the Coast Artillery Corps had permitted the Signal Corps

Page 257

to concentrate on the 270 and did not reopen the matter of the 296 until late 1940.

Development got under way in 1941 when the Signal Corps Laboratories obtained a 296 from Western Electric and added lobe-switching mechanism so that it could track a target. After putting it to service tests, the Coast Artillery approved it, had it standardized, and placed an order for 20 sets. All this occurred before Pearl Harbor. Immediately after that attack, although fear of air onslaughts transcended all other fears, there also arose apprehension for the safety of harbors and anchorages, into which the enemy might send small forays of deadly torpedo boats or submarines. Therefore, on 18 December 1941, Admiral Harold L. Stark had asked General Marshall to consider the use of detector systems to stand guard at harbors and anchorages. The chief of Coast Artillery, to whom the inquiry was referred, replied that while he did have searchlights which could illuminate a motor torpedo boat up to 7,000 yards, he did not have radio detection equipment able to discern such surface craft. The SCR-268, which he was now receiving in quantities, was “not designed to operate at zero elevation,” he explained. But he did expect that the SCR-296 would prove to be an effective detector of small surface vessels. Delivery of the 296, he believed, would commence in the spring of 1942.47

Production began slowly, the first set being delivered in April 1942, the second in July. By then the Coast Artillery order stood at 176 sets, to be delivered before mid-1943. This radar served for fire control functions through most of World War II, although it had several defects, arising especially from the rather long wavelengths it employed at a frequency of about 700 megacycles. Before the end of 1942, the Signal Corps embarked upon the development of a better tracking set for coastal defense, the SCR-598. This radar would lead, before the war’s end, to a superb three-centimeter microwave Coastal Artillery radar, the AN/MPG-1, operating at the extremely high frequency of 10,000 megacycles. These refined sets did not see service in the war. But another microwave radar did come to the aid of Coast Artillery in 1942. This was the ten-centimeter SCR-582, designed by the Radiation Laboratory in Cambridge, Massachusetts. It was one of the first ground applications of microwave radar using the new PPI, or Plan Position Indicator.48

SCR-582, Harbor Surveillance Radar

The Coast Artillery men first saw the 582, or rather its laboratory prototype XT-3, late in December 1941. A few days earlier a Signal Corps officer, Colonel Corput, had ordered 50 sets from the Radiation Laboratory’s own shops, the Research Construction Corporation. Between

Page 258

27 and 30 December Lt. Col. James E. McGraw, of the Coast Artillery Corps Board, had tested the XT-3 on Deer Island (Fort Dawes) in Boston Harbor. He favored it, rather conservatively, with the opinion that the “SCR-582 is desirable equipment for use by the Coast Artillery in harbor defense observation stations.” The eye of its constantly rotating four-foot dish antenna flashed upon the PPI a map-like view of the harbor area together with all the vessels therein. This it did with the detail which only microwaves make possible. Whereas the SCR-296 could track but one target at a time, meanwhile leaving the remainder of the harbor unsurveyed, the 582 could watch over all targets within its range. As a surveillance set it valuably supplemented the fire control and tracking radar SCR-296.49

In May 1942 the first of the 50 SCR-582’s crash-built by the Research Construction Corporation underwent preliminary tests. The Coast Artillery men were deeply impressed. They expected the set “to be of exceedingly great value.” Further service tests in July at Fort Story, Virginia, led the Coast Artillery Board to urge increased production, beyond the 55 sets already on order, these being the 50 from the Research Construction Corporation and 5 from the Crosley Radio Corporation. The board recommended that the 582 be restudied, in the light of its capabilities, “with a view to the widest possible justifiable application of this set.”50

The 582 successes soon came to the ears of the Air Forces, whereupon in September General Arnold asked the Signal Corps to test the new microwave ground detector against aircraft. Though the first sets could not scan the sky since no provisions had been built into them for tilting the antenna dish upwards, nonetheless Coast Artillery men had noticed how well this radar was able to track low-flying craft, 500 to a 1,000 feet above the sea, out to ranges of 40,000 yards.51

Thus by the middle of 1942 the new microwave radars, all radiating ten-centimeter waves at about 3,000 megacycles, all American sets developed by the Radiation Laboratory out of descendants from the British cavity magnetron, were receiving tremendous favor. Added to the success of the ground set SCR-582 were the of the ASV-10, which, in the form of the airborne SCR-517, was actually helping to sink submarines off American shores. Added to these was the scintillating promise of the gun layer SCR-584, or rather of its laboratory prototype XT—1. Microwaves were coming in for all radar applications, and long-wave radar was on the way out, though the long-wave types would continue to do duty all through the war because they

Page 259

Coast artillery fire 
control radars

Coast artillery fire control radars. The SCR-582 (upper left) and SCR-296 (upper right) at Charleston, S. C., and the SCR-296 and fire control towers at San Juan, Puerto Rico (below)

Page 260

had been developed first and were now in full production.

The Army put the new microwave ground radars to test in Panama. In mid-1942 Dr. Bowles, radar specialist on Secretary Stimson’s staff, visited Panama, and was followed some weeks later by another radar scientist and director of the Radiation Laboratory, Dr. Lee A. DuBridge. There the army commander, General Andrews, who was still faced with the possibility of carrier-based air attack, asked that the Panama Canal Zone become the testing ground for the newest and best in radar. The Radiation Laboratory scientists therefore kept in mind, as they developed microwave applications, the problems of Panama defense, especially the problems presented by reflections from mountains. One of the scientists, Dr. Ralph Bown, wrote to DuBridge on 6 August: “We thought that the sharp beams characteristic of microwave equipment, such as Dr. Bainbridge’s HPG, might be one way of solving the problem.” Another way, which Colonel Corput suggested to Dr. Bown, was continuous wave doppler radar, of the “fence” type.52

Microwave radar would be tried first and would indeed prove much better able to detect aircraft over high land masses than long-wave radar. Panama, then, was the ideal place to try out the SCR-582 after it had been modified with a tilting antenna for aircraft detection. Accordingly, in September DuBridge recommended two 582’s, so modified, and one SCR-615, the latter being still another microwave detector which the Radiation Laboratory had under design patterned after Navy’s HPG, a radar intended for shipboard use both as a medium-range aircraft warning set and as a GCI.53 Representatives of the Signal Corps, the Air Forces, the General Staff, and the Radiation Laboratory thereupon agreed that toward the year’s end the 708th Aircraft Warning Company in Panama would receive one SCR-615 and two SCR-582’s, modified for vertical search. While the National Defense Research Committee would supply the radars, together with the installation crews and air-conditioned buildings, the Chief Signal Officer would supply the power equipment.54 Subsequently, the modified SCR-582, made mobile, installed in three 2½-ton trucks, and modified with a tilting antenna dish for aircraft detection, became the SCR-682, a Coast Artillery long-range early-warning radar employed in World War II against both surface vessels and aircraft.55

SCR-615, Microwave Radar for GCI, Ground-Controlled Interception

The SCR-615 was another Radiation Laboratory ten-centimeter radar, whose production, distribution, and maintenance worries the Signal Corps now inherited. The Radiation Laboratory had begun developing this SCR early in the year at the behest of Colonel Saville, Director of Air

Page 261

Defense, who had frantically prodded the Signal Corps and the British for a GCI set. He had prodded all through 1941. Then early in 1942, realizing that the Canadian-built CHL/GCI (the SCR-588) would not be available for a while and that the SCR-527 (Signal Corps’ copy of British GCI) would not come out of General Electric factories for quite some months, Saville turned to the National Defense Research Committee in a third effort to get some sort of GCI soon. He had hoped that the Radiation Laboratory scientists might quickly modify for GCI use the Navy’s HPG.56

The scientists made the modification, but not so quickly as the impatient airmen wished—not until after the 588’s became relatively abundant late in 1942 (but before the 527 appeared). And anyway, SCR-615 was far better, for it was a step in the right direction, into microwaves and away from the long waves which both the 588 and the 527 employed for long-range search. It was the first ten-centimeter GCI, a large fixed set having a six-foot antenna dish. Of the first two models, crash-built by the Research Construction Corporation, one underwent service tests at the Army Air Forces School of Applied Tactics (AAFSAT), Orlando, Florida, toward the end of 1942; the other went to Taboga Island, at the Pacific entrance of the Panama Canal, where it replaced an SCR-271 which had failed to work well there because of reflections from the rugged terrain round about.

This first microwave GCI opened some eyes, especially among the proponents of long-wave radars, though it had some drawbacks, too, and failed to win complete approval at first (unfortunately true of the initial installation in Florida). Its range was not inconsiderable, up to 90 miles, refuting Watson-Watt’s insistence that only long-wave radar could fetch distance. Moreover, its coverage at low levels, always a problem with long-wave radar, was a revelation. Even the set in Florida, which had been poorly sited, tracked large numbers of flights below 5,000 feet, flights which other radars in the area, all radiating long waves, failed to detect at all. Similarly, in Panama, Lt. Gen. George H. Brett, successor to Andrews, got a surprise when he thought he would show up the radar detection system, as so often had been done before, by flying his plane in close to the water. On landing, he expected that his flight had been entirely “unseen” by the SCR radars in the zone. But he received a pleasant surprise when he was handed a plot of his flight, for which the SCR-615 crewmen and their microwave helpmate provided the data, and found that his plane had been kept in “view” all the way to its landing.57

SCR-602, Lightweight Warning Thinking in terms of defense had set the tone of American radar efforts early in 1942. Fear for the west coast and for the Panama Canal had led the military to demand long-range detectors and GCI radars, even the fixed long-wave gear of the early British CH and CHL/GCI sets. But as 1942 wore on, concern over passive air defense of the United States coast and possessions gave way to thinking in terms of Allied

Page 262

offensives. A fresh reappraisal found long-range detectors necessary at only a few places, most notably, in Panama. In general the SCR-270 or 271 continued to prove good enough. Even Saville had acknowledged that the 270 was good in isolated uses, as on an island, for aircraft detection.

But Watson-Watt’s belief that such long-wave radars alone could attain great ranges was soon disproved by the new microwave ground sets. And the Air Forces’ original stress upon mobility, which had led to the first American aircraft detector, the mobile SCR-270, received renewed emphasis in 1942 as military men planned to invade hostile shores. Contemplating attack upon distant beaches where air raids could sow vast harm among congested landing areas, the airmen gave thought to a set which might be taken onto a beachhead in pieces, carried by a few men who could assemble the radar in an hour or two. Its power and range, of course, would be limited but would suffice. The AAF, with its predilection for British long-wave sets, turned to them once again, to the British LW, or lightweight warning radar which the Signal Corps copied as the SCR-602, in time for use in North Africa.

Early in 1942 Saville had set forth the airmen’s requirement for an “equipment similar to the British light mobile or portable early warning set.” But this LW radar prototype was not yet in production. Maj. Gen. James E. Chancy had cabled from England that the British would send two sets to the United States as soon as they became available.58 In February 1942 General Olmstead asked for information about this British “light mobile and portable aircraft warning RDF set.” Two months later he had gotten nothing, whereupon he had asked the Chief of Staff to make another petition. But even as he asked, the blueprints were at last on their way, followed late in June by one of the first British prototypes. On 25 June General Eisenhower in London notified Washington that Lt. W. W. Debenham, a Signal Corps officer who had been working with British radars, was that night leaving Prestwick, Scotland, the English terminus of the transatlantic air ferry. With him he had 1,000 pounds of lightweight warning radar.59 While the Signal Corps Radar Laboratory at Camp Evans, Belmar, New Jersey, looked over this LW, the AAF asked in mid-July for 100 service test sets and 100 spares. The AAF doubled the figures before the end of the month, asking for 200 sets from American sources as well as for 200 sets from the British. These last the Signal Corps contracted for with the Canadian firm, Research Enterprises Limited, at $15,000 each, delivery to begin in the following January.60 Thus the SCR-602 Type 1 became another British copy, the last copy of a British radar and the

Page 263

last copy of a long-range set (wavelength about one and a half meters, on 212 megacycles) which the Signal Corps would have to produce for the AAF.

By mid-1942 the pressure for LW radar was becoming tremendous and, as so often happens with new developments, the demand built up suddenly with little advance warning to the laboratory and procurement officers. At the outbreak of World War II the American Army had no really portable early warning radar. It had only the heaviest kind of detectors. Even the mobile 268 and 270 could go only where heavy trucks could transport them. To ship them over water and to land them was no light matter. Obviously their weight and bulk would be handicaps in the island hopping, in the diversified landing and other highly mobile operations which would figure so prominently in World War II. What the Army needed was a detector which could be packaged for hand-carrying by a small number of men, landed with the first waves of an assault, and assembled quickly for operation whether in the early stages of an assault or in moving through jungles. LW radar was wanted now, wanted desperately, and the pressure for it led to an amazing variety. Contemplating invasions, the General Staff and the Air Forces wanted at once all the sets they could get (their plans already called for upwards of 1,000 sets). The Air Forces gave them priority over all other sets except IFF radar. They wanted them lightweight, and still lighter, and they wanted sets operating at different frequencies in order to minimize the likelihood of jamming, an art in which the enemy was becoming proficient, especially in the 200-megacycle bandwidth of the original British LW.

Before the end of the year the LW types numbered no less than ten, ten varieties of SCR-602, either in production or under development by Research Enterprises, by the Navy, by the International Telephone and Radio, by the Bell Laboratories, by Radio Corporation of America, by General Electric, and by the Signal Corps Radar Laboratory. The ten types varied in frequency from the 212 megacycles of the long-wave British prototype to 1,000 megacycles. They varied in weight, too, from the 1,200 pounds of Type 1, which was a copy of the British LW, to a 250-pound parachute set (SCR-602 Type 7) which General Electric was developing.61

Thus, though last of the long-wave radars, the SCR-602 was not least in but size. Indeed, in its several forms, it came to be one of the most important and numerous of aircraft detectors. Only the first 25 off the production lines of Research Enterprises Limited were exact copies of the British LW prototype. They were designated SCR-602-T1. The next 250 productions differed in design and in designation, which was SCR-602-T6. Thereafter, still further modifications of this Canadian-built British copy resulted in yet another change of designation, SCR-602-A. Of all these types of the SCR-602, was Type 8, the unique Signal Corps development, that proved to be the best.62

This Signal Corps creation, SCR-602-T8, was destined to become the most efficient set of its type, used late in World War II as the AN/TPS-3 (the subsequent Army-Navy terminology: T for transportable, P for radar, S for search). For some time prior to 1942 the Signal Corps

Page 264

SCR-602 Type 8 lightweight 
warning radar

SCR-602 Type 8 lightweight warning radar

Laboratories had been working on 600-megacycle radar, radiating waves 50 centimeters long, intermediate between long-wave and microwave radar. Dr. Harold A. Zahl, a Signal Corps radio engineer, had developed vacuum tube VT-158 in Signal Corps’ own thermionic laboratory (organized in 1940 in order to turn out radar tubes of types so complex that industry hesitated to attempt them). The VT-158 was capable of generating 50-centimeter waves with remarkable power output in the order of hundreds of kilowatts, remarkable power for a triodetype tube at that date. The Laboratories tried the tube in the various sets, such as in the SCR-268. But not until the AAF set up requirements for an LW did the VT-158 really find its place, in SCR-602-T8. Type 8, triumphing, would win even AAF as well as British acclamations, a triumph indeed for the Signal Corps. Asking for 200 sets, Air Vice Marshal of the Royal Air Force, R. B. Mansell, told General Colton that “this development is one of the most important in Ground Radar technique in recent years and that the designers are to be congratulated in producing a receiver, display and high power transmitter in a single unit measuring only 42 inches by 20 inches by 20 inches.” SCR-602-T8, or the AN/TPS-3, was Signal Corps’ most substantial contribution to radar after the ancestral Army sets SCR-268, SCR-270, and SCR-271.63

Page 265

The remainder of ground radar and all airborne radar in World War II is essentially, as far as the initial stages of research and development are concerned, a story of other institutions such as industrial laboratories and especially the Radiation Laboratory, under Division 14 of the National Defense Research Committee, within the Office of Scientific Research and Development. The Signal Corps kept in touch with the work and took over the final details of development, inheriting of course all the multitudinous adjustments necessary to meet military requirements and to fit the laboratory design for mass production. So it was, as already shown, with SCR-582 and 615. They were the Radiation Laboratory’s offspring, although the Signal Corps made many tests and improvements upon them thereafter to suit military needs and saw to all the multifarious details of procurement and distribution. So it was also with two most significant ground radars which the Radiation Laboratory developed for use in World War II: the SCR-584 and the MEW, short for microwave early warning, or in AN terminology, the AN/CPS-1.

SCR-584, Microwave Tracking or GL, Gun-Laying Radar

The SCR-584 was a gun layer, the most successful single application of the microwave ten-centimeter technique to ground fighting in World War II. It could automatically track an unseen target at night or in cloud or fog, supplying range, azimuth, and elevation data to a gun director, which aimed the guns of a battery. It doubled the usefulness of big guns. From the time the 584 first appeared on the Anzio beachhead—where it enabled gunners to play havoc with the air attacks of the enemy, who had been successfully jamming the SCR-268’s there—it was the indispensable aid to Allied antiaircraft gun batteries. The 584, in cooperation with the proximity fuze, which was actually a tiny radar built into a projectile, was to nip the buzz bomb menace and find new applications—detecting land targets, tanks, and vehicular convoys at night.64 More 584’s were to be built in World War II than any other American radar, except the first, the patriarchal SCR-268.

Subsequent to first tests of the SCR-584 at the Signal Corps Laboratories in December 1941, the XT-1, the experimental prototype of the 584, was returned to the Radiation Laboratory. There the scientists refitted it with a larger six-foot antenna dish and stepped up its range.65 Next, the set went to Fort Monroe, Virginia, for further tests by the Antiaircraft Artillery Board, then under the Coast Artillery Corps. This in March 1942. Any skepticism which had arisen from the first tests at Fort Monmouth in December 1941, when the range of the set had proved disappointing, evaporated during these second tests at the mouth of the Chesapeake Bay. The range was now far greater than the 15,000-yard minimum the service users required. The accuracy was uncanny. And the set was stalwart, even in its laboratory-built prototype form. While it stood upon the sea wall at Fort Monroe, overlooking Hampton Roads, a violent storm came up; waves beat against the wall and salt spray doused the XT-1. But the morning after, as the set dried out, it went into action as usual, none the worse for its exposure. The Antiaircraft Artillery

Page 266

Board concluded: “The Radio Set XT-1 is superior to any radio direction finding equipment yet tested for the purpose of furnishing present position data to an antiaircraft director.” The Antiaircraft Artillery Board urged the Signal Corps to standardize the set and procure it in sufficient quantities to supply one to each Antiaircraft Artillery gun battery.66

Two weeks later General Olmstead recommended that XT-1 be standardized as gun layer SCR-584, simultaneously asking a change in the fifth supplemental national defense appropriation for the fiscal year 1942 whereby 622 sets might be purchased with funds that had been already allocated to another gun-laying radar, the SCR-545. This he had urged after a conference in his office a few days before with representatives of General Electric, Westinghouse, the National Defense Research Committee, and the Antiaircraft Artillery Command. General Electric and Westinghouse had been so optimistic as to estimate completion of the 622 sets by July 1943, whereas the bitter fact would be that by then actual delivery would scarcely have begun.67

There were cogent reasons for the delay of the SCR-584, reasons arising out of both preproduction engineering changes and priority problems during production itself. Some of these the Signal Corps inherited from the Radiation Laboratory; others resulted from the military characteristics which the users desired the Signal Corps to satisfy.

In the spring of 1942 the original XT-1, despite its potencies, was still a long way from a usable practical military set. When the Antiaircraft Artillery Board members wrote their report after testing the set in March 1942, they commented that the XT-1, being a laboratory model, contained many makeshift parts which rendered it unfit as it stood for field use. They recommended some sixteen changes needed to bring the set up to the military characteristics they desired in the production gun layer. For example, the Antiaircraft Artillery men wanted tracking to be as automatic in range as it was in azimuth and elevation (the laboratory model required hand-aided tracking in range). They wanted the antenna mount, its driving mechanism and pedestal, completely redesigned, a little matter that was to prove a severe engineering and production problem.68 They wanted this set, intended for short-range gun laying, to include provisions for early warning, illuminating upon a PPI-type oscilloscope the reflection of planes many miles away, long before they came within gunshot.69

In asking this double function, early warning and gun laying combined in one radar, the artillerymen were not just

Page 267

conjuring up engineering and production troubles to harass the Signal Corps. They were following a precedent. The SCR-545, which the Signal Corps and Western Electric already had under development for the Antiaircraft Artillery Command, incorporated this double function, as did the British GL-3. Both the British set and the American counterpart employed two different frequencies: (1) a low frequency emanating long waves for long-range early warning, and (2) a higher frequency radiating ten-centimeter microwaves for short range precise gunfire data.70

This use of two different frequencies for two quite different functions was a decided complication, which the designers of the 584 discarded. The Radiation Laboratory scientists found that they could extend the range of the microwave XT-1 up to 90,000 yards, quite enough for early warning purposes. Then, as targets flew within about 32,000 yards, the set could begin to serve as a gun layer; the antenna would lock onto the target’s reflection, automatically follow it and feed azimuth and elevation data to the gun directors.

All this could be done without recourse to long waves. The ten-centimeter waves of the 584 could do everything longer waves could do, and do it much better. The SCR-584 was freed from the handicaps which inevitably accompanied long waves, siting troubles especially. There were no nulls or blind spots. There were no limitations upon low coverage, a boon when German buzz bombs began to fly in 1944, darting in level courses only a few hundred feet above the ground. The 584 tracked them, whereas long-wave radar would have failed utterly.

During the summer of 1942 the Signal Corps Radar Laboratory labored with the production design of the many components of the SCR-584. The laboratory report for August listed as complete the designs for:71

power unit

automatic tracking mount

receiver power

azimuth and elevation tracking mount

amplidyne mount

remote video amplifier

position indicator

Nearly completed were the designs for:

modulator

pedestal and antenna

receiver

magnetron

positioning control unit

Still in the development stage were the designs for:

connection diagram

equipment assembly

wiring

interconnection cables

Consider for a moment the complexity of such a device. Basically any radio or radar consists of three or four elements: a power supply, an antenna, a transmitter, and a receiver. As contemplated in the summer of 1942, the SCR-584 required about 140 electronic tubes, of about a score of types. The basic transmitter required only one tube, the very powerful cavity magnetron which made centimeter radar possible. The basic receiver did not require an excessive number, a total of 14. But around the receiver-transmitter core with its total of 15 tubes thronged a dizzy array of special circuits, each doing a vital job and each

Page 268

requiring its own gamut of tubes. For example:72

Tubes Azimuth and Elevation Tracking 16
Automatic Tracking Unit 5
Altitude Converter (Control Unit) 6
Altitude Converter (Power Unit) 8
Modulator 22
Range Power 10
PPI Power 7
PPI Unit 18
PPI Scope 1

All this mass, together with operational and maintenance accessories, was packed into a van which housed the operators. The van, its load, which included IFF equipment (RC-184), together with a trailer containing an electric generator driven by its own gasoline engine, constituted the SCR-584, ten tons in all. Small wonder that one half the total cost of a large 90-millimeter antiaircraft gun battery (four guns) went to the radar equipment, valued at $100,000.

Despite its superiority, the SCR-584 did not go into production until mid-1943. This was not because of any failure on Signal Corps’ part to push matters. It pushed hard from the very first. For example, the Signal Corps succeeded during April and May 1942 in getting the Army to stop production of a searchlight director radar, the ten-centimeter SCR-541, and to divert this effort into the 584. The Coast Artillery Corps had long pressed for a new searchlight director, lighter in weight and handier than the original SCR-268. But as the SCR-541 approached production in 1942, the Army found that in some ways the new microwave set would be no improvement over the 268; for example, it would be nearly as heavy. Furthermore, the 541 tied up components, production facilities, and raw materials which could be put to better account in the fabrication of the much more valuable SCR-584. As early as April, therefore, Rives, Metcalf, and others in the Signal Corps Radar Division in Washington had debated with representatives of the Services of Supply, Antiaircraft Artillery Command, Ordnance, Westinghouse, and War Production Board, with a view to recommending cancellation of a contract which the Signal Corps had made with Westinghouse for some four to six hundred SCR-541’s. It was Signal Corps’ contention that the radio parts accumulated for the 541, together with raw materials, factory production facilities such as tools, space, workers, and so on, should be channeled into the production of the new gun layer SCR-584.73 This recommendation, as just noted, was carried out.

But another suggestion, the one General Olmstead had made a few days earlier, that 584’s be purchased with SCR-545 funds, was not adopted. The order for

Page 269

Radar sets SCR-545 
(above) and SCR-584 (below)

Radar sets SCR-545 (above) and SCR-584 (below). Note IFF antenna in left foreground of each photograph

Page 270

273 SCR-545’s remained, and they would begin coming off production lines in 1943 a few weeks before the 584’s. No more were ordered, however. Meanwhile, the 584 went on order, to the number of nearly 3,000, with high hopes from Maj. Gen. Joseph A. Green, Antiaircraft Artillery commander. “By use of this equipment,” Green commented, “antiaircraft gun batteries will, for the first time, be capable of directing accurate fire against unseen targets.” One set he considered a must for each gun battery.74

Neither the 3,000 or so sets which General Green hoped to get in 1943 nor even General Olmstead’s more sober figure of 1,175 would attain completion by 1943’s end. As already pointed out, priority quirks during 1942 were one cause. Priorities which were supposed to speed the things most needed in this case caused delay. Four test models had been put on order with General Electric in April 1942 for early delivery. They were needed for preliminary tests, which would inevitably reveal desired changes and improvements. These changes would then be incorporated in the subsequent production. The four test models therefore should have enjoyed a priority at least as high, if not higher, than the entire production. But not so. Their priority was much lower. Summer came and went, and yet the four pilot test models failed to appear. General Olmstead thereupon complained to the Army and Navy Munitions Board priority makers that the reason was clear—production of the four sets suffered from a lower priority, AA-3, than the rest of the 584 program, which enjoyed nearly a top priority of AA-1. Illogical test models, upon which the entire production awaited, should languish unrecognized by the very priority makers who evaluated the rest of the program at AA-1. general of the Army Ground Forces was aroused. He particularly wanted the first four sets in order to initiate training. Otherwise, the first procurement sets now expected in April 1943 would have no one who knew how to operate them, however desperate the need of combat Antiaircraft Artillery units. He asked General Somervell to grant the manufacturer, General Electric, top AAA priority.75

One reason for the low priority was secrecy. Priority makers in the Army-Navy Electronics Production Agency and the War Production Board could not get the facts. The Signal Corps’ contracts with manufacturers stipulated, for example, that delivery schedules and other data touching the 584 could be made only to Signal Corps members. Such restrictions arose again and again, whenever secret equipment was procured. Representatives of Chrysler, General Electric, Westinghouse, and Fruehauf, among others, all complained of the trouble they encountered when seeking assistance from the Army-Navy Electronics Production Agency and the War Production Board toward getting the 584’s produced. Why? Because they were forbidden to provide these agencies with the delivery schedules of highly classified Signal Corps equipment. Yet how could the priority judges make

Page 271

equitable judgments if they could not be adequately informed?76

The Signal Corps’ complaint about the lower priority of the first four 584’s touched off a minor explosion over the GL radar program. The explosion uncovered some interesting things, as, for example, the fact that the allocation of machine tools to electronic production was nearly nil. The men who made allocations were vastly impressed by the need for aircraft, and they allotted 60 percent of all machine-tool facilities to aircraft plant expansion in 1941-1942, but none to plant expansion for the manufacture of the electronics equipment which the aircraft needed. Yet this equipment represented a large percent of a bomber’s cost (it would approach 33 percent by the war’s end). Likewise, ordnance got large allocations, but not GL radar, though, as pointed out earlier, it would soon constitute one half the cost of a 90-millimeter gun battery. Not till late 1942 did Signal Corps production needs begin to win recognition among machine-tool allocators, when the Corps was granted a parsimonious 4 percent. No wonder manufacture of the huge radars and of other complex electronic devices lagged, while the Army screamed for them and glowered at the Signal Corps as the most laggard of suppliers. Ignorance and secrecy doubtless helped to prolong the delay. Everyone knew airplanes needed aluminum and guns needed steel, together with myriads of machine tools to fabricate the metals. But not everyone knew about the electronic devices that controlled and guided the massive metal mechanisms coming off aircraft and ordnance production lines. It was easy to see the growing body of the war juggernaut and add to it, but it was not so easy to perceive that it had eyes and a network of electronic nerves of radar. Besides, many of the electronic components, and all of the radar, were secret. Most men, even those in uniform, were not permitted to see it, or only dimly.

Airmen concerned over air defense had for many months been making clear their need for electronic equipment. Colonel Saville, for example, had been especially insistent, with the strong support of the Secretary of War. Yet it was not till late 1942 that General McClelland in AAF headquarters asked General Arnold to demand more money, more factories, more tools for electronics, just as the AAF chief had done months before for aircraft. By now, late in 1942, the artillerymen began to clamor too, having just begun to learn the possibilities which GL radar offered. At the same time the Air Forces, backed by Secretary Stimson himself, had gotten the best priorities obtainable for much of their airborne electronic needs. But the priority makers had not yet given much thought to ground radar for artillery and very little thought at all to the necessary preliminaries to the allocation of raw materials and machine tools. At the level of the Joint Chiefs of Staff, Army Ground Forces needs in electronics

Page 272

were consistently overridden during 1942 by the Navy and the Air Forces.77

During the autumn of that year, correspondence flew back and forth between the heads of the Antiaircraft Artillery and the Army Ground Forces recounting the significance of the new gun-laying and other ground radars in a manner which reveals how new it all was to the ground Army. The high-ranking correspondents took pains to explain the sets, what they were, and what they did. The sets, they wrote, were essential for protection of harbors and of invasion efforts generally.

Mighty important, then, was this new application of radar if overseas invasions must depend on it. General Eisenhower would soon be writing back from North Africa saying that he must have something better than the SCR-268 for gun laying, a refinement for which the 268 was never intended, but for which it proved better than no radar at all.78 Thus the land Army was slow finding out how badly it needed gunlaying radar, just as the Air Forces had not appreciated airborne radar until the turn of 1940-1941. And until the Army found out its needs and brought pressure to fill them, the funds and the support which the Signal Corps had to have in order to develop and produce the equipment simply were not forthcoming. When the Army belatedly found out what it wanted and how badly, then it wanted the equipment immediately, if not sooner. Naturally and lamentably, it had to wait, and the Signal Corps had to endure its fulminations, a state of affairs to which military supply agencies must, in the nature of things it seems, become resigned.

On 8 October General Green, in the headquarters of the Antiaircraft Artillery Command, complained to the Commanding General, Army Ground Forces, that on information from the Chief Signal Officer he had discovered he would not get any GL’s, either SCR-545 or SCR-584, until some six or seven months after the delivery date originally promised. If the sets arrived too late, it would be too bad, to state it mildly. As Green put it:

If radar equipment is to be used to advantage in this war, the development and procurement of newer types of radar equipment must be pressed to the utmost, and the importance and urgency of this work should be impressed on all concerned. It is strongly recommended that every effort be made to expedite the antiaircraft radar program, both in development and procurement, and that particular stress be placed on the procurement of gun control sets and the development and procurement of a new searchlight control set.

General Moore, writing for the Commanding General, Army Ground Forces, agreed. Adding details about other Coast Artillery radars, the 296 and the 582 (whether for his own enlightenment or for that of General Somervell to whom he addressed a first indorsement on Green’s letter), he attributed the delay to the action of the priority makers.

This procurement program [of the 296, 545, 582, and 584] was disrupted by the action of the Precedence Committee of the

Page 273

Combined Communications Board which set up a precedence list for the manufacture of all radar equipment and relegated Coast Artillery radar procurement to Categories III to IX inclusive, thereby delaying all Coast Artillery radar manufacture for many months. By this action the procurement of Radio Set SCR-296 was stopped and certain critical parts were diverted for the manufacture of radar equipment in categories I and II [doubtless aircraft radar] in spite of the fact that a contract had been let, manufacture was in progress and many of the SCR-296’s were nearing completion. Similar action has delayed the initiation of procurement of Radio Sets SCR-584 and 545 several months.

General Moore urged Somervell to take steps “to correct the existing situation” and to jack up the precedence of these GL radars “to a point which will assure uninterrupted procurement in the immediate future.” He regarded such sets as vital safeguards in the invasions which the military chiefs were contemplating. “Our harbor defenses,” he wrote, “must have the necessary antiaircraft fire control apparatus to assure a safe haven for allied naval forces. Our antiaircraft units must be supplied with gun laying radars before they can enter combat with any reasonable prospect of success.”

Somervell, no authority on electronics, sought information from the electronic section of his empire, the Signal Corps—information “upon which a reply may be drafted.” Colton drafted the reply. So far, he showed, the delays at which Green and Moore chafed were not all the consequence of priority maladjustments at the production level. They were rather a consequence of a priority oversight at a lower, and earlier, stage in the development that resulted in insufficient machines to tool up production plants. The changes which the Antiaircraft Artillery Board had wanted made in the original XT-1, when it had tested the sets in March 1942, called for $10,000,000 in tools alone, before the raw materials were assigned and before production could even begin. Yet the Signal Corps had been favored with no such tool allotment. What it had gotten for its manufacturers it had obtained by diversion from other services. That had been true till very recently when the Signal Corps had been allowed a 4-percent tool allotment, still ridiculously small when one considered the vastly larger percentage electronics shared, on a cost basis, with the rest of the war production effort.79

The little matter of low priority for the pilot models was only one obstacle. There were many others: shortage of raw materials and critical components, such as tubes, for example, which consumed such rarities as tantalum, molybdenum, and tungsten and whose manufacture was most delicate and exacting. Then there were selsyns, tricky little electric motors and coordinating mechanisms which enabled the different parts of a large robot to keep in perfect step, its motions perfectly synchronized. The order for 584’s was going to require some 60,000 selsyns immediately. Still another tricky component, rejoicing in the name amplidyne, consisted of electrical motors and gears which enabled heavy platforms bearing radar antennas and mounts to rotate, smoothly, precisely, whether fast or slow. The obstacles to production scheduling which all these complexities presented

Page 274

Metcalf summed up for General Somervell late in September 1942:80

Delivery of these sets [2,750 SCR-584’s and 273 SCR-545’s, by General Electric and Western Electric respectively] was previously scheduled to be completed by the end of 1943. However, the production of Radio Set SCR-584, which comprises the bulk of this procurement, is contingent upon a great many uncertain factors, for example, the extension of plant facilities and provision of assembly of these sets and for production of the special semi-trailer and precision antenna elevating and tracking mechanism, the allocation of various critical raw materials and procurement of such critical component parts as tubes, selsyns, amplidynes, etc. Accordingly, on the basis of the best information now available, the Chief Signal Officer estimates that no more than about eleven hundred seventy-five (1175) Radio Sets SCR-584 can be safely counted upon for delivery by the end of 1943.

Thus, production of Signal Corps equipment such as this giant microwave gunlaying radar was big business indeed, in which many an agency participated. In one of the early procurement debates respecting the SCR-584 there sat representatives of the Signal Corps Radar Laboratory, the Navy, the War Production Board, the National Defense Research Committee, Westinghouse, General Electric, and Chrysler. Rives and Elder presided, while Dr. Bowles, Secretary Stimson’s radar consultant, attended.81 Eventually, the difficult mount design, with all the exacting requirements that daunted several manufacturers, was accepted by the Chrysler Corporation. General Electric and Westinghouse, each with a contract for 1,375 sets, began production, and by May 1943 the first production SCR-584 was to arrive at the Signal Corps Radar Laboratory.82

MEW, Microwave Early Warning Radar

As the gun layer SCR-584 was one brilliant application of microwave radar to ground military use, so was another, MEW, also a Radiation Laboratory ten-centimeter development. MEW never received an SCR number because it was typed after the Army-Navy nomenclature replaced the SCR designations, and so its official military title came to be AN/CPS-1 (C for air-transportable, P for radar, S for search). Its development began in 1942 as an outgrowth of microwave research at ten centimeters applied to detection of aircraft at very great distances. Named microwave early warning, it became abbreviated to MEW. Its range from the start excelled that of all other radars and vanquished forever any claim that long range might be the exclusive prerogative of long-wave sets. Moreover, older long-wave, long-range radars were notoriously affected by their site. They were not very accurate either, and they could be jammed easily by enemy transmitters operating on their relatively low frequencies. All these defects the MEW swept aside.

MEW introduced a new type of antenna and reflector. The majority of previous microwave sets had employed a dish-like

Page 275

antenna reflector called a parabola. MEW employed a reflector which resembled a cylinder or pipe cut in half lengthwise. Instead of a dipole antenna which radiated from the center or focus of the dish-type reflector, MEW employed a long pipe or wave guide running the length of the half-cylinder reflector. The wave guide conducted radiations from the transmitter’s magnetron and sprayed them, through a number of slots in the guide pipe, onto the concave surface of the half-cylinder reflector, whence flashed forth a flattened beam. Some of the transmitter and receiver components were attached directly to the back of the large antenna shell, so that the set literally perched on its own antenna array. MEW had a tremendously high power output of 500 kilowatts at impulse peaks.83

It was the members of the so-called High Power Group (HPG) at the Radiation Laboratory who developed the MEW, and just before it the SCR-615. Late in November 1942, using a makeshift antenna and reflector atop the main building of the Massachusetts Institute of Technology, the HPG tested their prodigious MEW against a target (a PBY in some accounts, a B-18 in others) flying southeast from Cambridge, beyond the shore and out over the ocean, even beyond Nantucket Island. An excited group of scientists watched the oscilloscope and stuck pins on a wall map as the target’s reflection came back from ever farther away. The pins marched on across the map, off its edge, and onto the wall as MEW continued to track the plane out to 177 miles.84

The Signal Corps had a representative at the Radiation Laboratory on that day, Wendell L. Rehm of the Radar Laboratory’s general engineering section. On the next day, back at his laboratory desk at Camp Evans, Rehm addressed a memo to Maj. John J. Slattery. He called the results of the MEW test “quite amazing.” The test airplane, which he called a B-18, was finally lost at 177 miles at an altitude of 16,500 feet. He added that radar scientists were already wondering how they could present all the targets which so powerful a radar could pick up—whether they might not have to use a whole battery of oscilloscopes since a single scope could not begin to display everything such a set could “see” in an area of land and sky approximately 400 miles in diameter. One of the scientists, Dr. Lawson, had cautioned him not to be “too optimistic” until further tests. Obviously, though, the Radiation Laboratory workers were agog and could well be optimistic.

It is worth recording, however, that not all Signal Corps officers and civilians were as enthusiastic. A cautious engineer from the Office of the Chief Signal Officer, Norman A. Abbott, attended a conference a few days later at the Radiation Laboratory and viewed the MEW itself with lusterless eye, principally because the set employed so huge an antenna and reflector and because it had no provision for determining target altitude, points sure to irk the Air Forces. Abbott reported that Army liaison men at the Radiation Laboratory thought it “just another long-range detector of which there were

Page 276

several already in existence.”85 This cautious military estimate would soon yield to the enthusiasm of the scientists who knew how revolutionary this new departure in microwave radar was.

Thus, during 1942, United States Army microwave radar, developed from the pooled military and civilian efforts of the British and the Americans, began to prove its worth. In aircraft applications the proof came quickly. The ground application came more slowly, but the promise of SCR-582 and 584 and of the MEW was already clear by the year’s end. These radars would establish the excellence of microwave sets on the ground, although not until 1944 would the 584 and the MEW participate in combat operations. Then their contributions would be scarcely less impressive than those of the airborne BTO radar in blind bombing or of the ASV in driving German submarines from the seas.