[3.0] Microwave Radar & The MIT Rad Lab

v2.0.6 / chapter 3 of 12 / 01 jan 15 / greg goebel

* In 1940, a group of British researchers all but stumbled onto a new electronic component, the "cavity magnetron", that permitted the development of effective microwave radars. The British understood the importance of the cavity magnetron and kept it secret. However, they also understood the necessity of obtaining American aid in the war. In another far-sighted gesture, in the summer of 1940 the British government sent a mission to the US to share the cavity magnetron and other British electronic technologies with the Americans.

By this time, the Americans were realizing that they would be drawn into the conflict sooner or later, and were beginning to mobilize for war. The Americans quickly set up a major research organization, the "Radiation Laboratory", to develop the technologies handed to them. This was far-sighted as well. The British were perfecting their longwave radars, making them more reliable and capable, but microwave technology promised to be a major step forward, and it had to be developed as quickly as possible.

cavity magnetron



* While the Americans were working on radar at a relatively leisurely pace, the British were pushing the technology as hard as they could. Whatever progress that was being made on AI and ASV still wasn't fast enough. Fortunately, while the work struggled forward, a pair of researchers at the University of Birmingham, John Randall and Henry Boot, had come up with a new invention that would help make radar more effective all up and down the line, and put the Allies well ahead of the Axis in radar for the rest of the war.

Shorter wavelengths provided a number of advantages for radar technology, including finer resolution, a tighter beam, and greater immunity to noise. However, there was simply no technology available in 1940 to generate radio waves of sufficient energy at short wavelengths. The British Admiralty, in collaboration with other interested parties, had set up a special committee to investigate radar that would operate at ten centimeter wavelengths, in the microwave band. Clarendon Laboratory at Oxford was assigned to work on a microwave receiver, while a team from the physics department at the University of Birmingham was to work on a microwave transmitter.

The University of Birmingham effort was led by an Australian named Marcus Oliphant. Randall and Boot were not at the heart of the transmitter development project. In the fall of 1939, they were simply trying to develop microwave detector circuits. To test their designs, they had to generate microwaves for their circuits to detect.

Every very rare now and then people who are newcomers to a field make a great discovery, simply because they don't know what works and what doesn't. Randall and Boot didn't know much about generating microwaves, so they set about learning how. There were two devices available at the time for the task. The first was the "magnetron", which was basically a classic vacuum diode with a magnetic field placed across it. The interaction between the external magnetic field and the electron flow through the tube produced microwaves. The other was the "klystron", much more recently invented by the brothers Sigurd and Russell Varian at Stanford University in California, and based on a "resonant cavity" through which streams of electrons flowed. Oliphant's team believed the klystron was the solution for short-wavelength radar.

Randall and Boot didn't want to spend a lot of time and effort generating microwaves for test purposes. They focused on the less sophisticated magnetron simply because it seemed simpler to work with. As they learned about the magnetron, however, they realized that they could combine features of the magnetron and the klystron and come up with something new. Working on a shoestring budget, the two men pieced together their new "cavity magnetron", as they called it. The core of the cavity magnetron was a thick copper cylinder, with a large central tunnel bored through it. Six smaller tunnels, or "resonant cavities", were bored around the central tunnel, and connected to the central tunnel through slots running down their length.

The copper cylinder was positively charged, forming the "anode" of the tube. A metal conduit was inserted down the central tunnel. The conduit was negatively charged, forming the "cathode" of the tube. The cylinder assembly was sealed at the ends, and a magnetic field placed across it. Under the combined influence of the electrical potential between anode and cathode and the magnetic field, electrons circulated in the central tunnel, producing electromagnetic radiation in the resonant cavities. The electromagnetic radiation from the cavities coupled together in the central tunnel, interacting with the electron flow to efficiently extract energy from it with high efficiency.

Physicists working with the device would later describe it as a kind of "whistle", where the flow of electrons generated electromagnetic waves of a specific wavelength, just as the flow of air through a whistle generates sound waves of a specific wavelength. The frequency of a whistle is dependent on its size, with a big whistle generating a low sound and a small whistle generating a shrill one. Similarly, the frequency output of the cavity magnetron was dependent on the size of the cavities. The cavities had a diameter of 1.2 centimeters (a little under a half inch), confining the electromagnetic radiation to produce "standing waves" at 9.1 cm (3,300 MHz or 3.3 gigahertz / GHz). A tap was bored through the side of the cylinder to provide an outlet for the microwave energy generated inside.

Randall and Boot performed the first microwave transmission using their cavity magnetron system on 21 February 1940. Within a few days, they were lighting up fluorescent tubes from some distance away, which indicated power output on the order of 500 watts. They found this unbelievable, and rechecked their figures and experimental setup. Nothing was wrong. The cavity magnetron was an entirely unexpected leap forward in microwave technology.

The cavity magnetron was so promising that Oliphant's group abandoned their work on the klystron to work with the new device. Through the spring and early summer, they improved their crude microwave transmitter into something resembling an operational system, with a maximum output power of 15 kilowatts, three orders of magnitude greater than the output power available with any other device.

The cavity magnetron opened doors to new technological possibilities. The TRE received its first cavity magnetrons on 19 July 1940. They were supplied to a centimeter-technology group under Philip I. Dee, of the Cavendish Laboratory at Cambridge University. Dee's team quickly put together a microwave radar system operating at 9.1 centimeters, though the technology was referred to as "10 centimeter" for convenience, and tracked an aircraft with it on 12 August 1940. The next day, the radar tracked a technician riding a bicycle carrying a tin sheet. Ground "clutter" would have simply blinded any long-wavelength radar under such circumstances. Microwave radar had arrived.

* It should be no great surprise, given the fact that radar itself arose simultaneously in several countries, that several other nations discovered the cavity magnetron at about the same time. Two Soviet engineers, N.F. Alekseev and D.D. Malairov, developed the technology in the late 1930s and actually published a description of it in a public technical journal in 1940.

The fact that it wasn't kept a secret indicates the importance, or lack of it, assigned to it, all the more so because the Soviets took secrecy to an extreme. Soviet radar design was badly hampered by bureaucratic indifference and incompetence, plus the fact that a number of first-class engineers were arbitrarily purged and sent off to forced labor camps, from which many never returned. The USSR didn't exploit the cavity magnetron until after the Western Allies had put it into extensive use, and Soviet radar designs lagged badly. They received radar gear from Britain and the US, building copies when it seemed like a good idea, as they did with GL Mark II. Postwar Soviet radar work was based on technology provided by the Western Allies.

Magnetron technology was also invented roughly in parallel in Switzerland, France, and Japan. In a further irony, the Japanese design was based on a device built in the mid-1930s by an American engineer, Arthur L. Samuel of Bell Labs, that he never got to work very well. Samuel did make a major contribution to early Allied longwave radar work by designing a high-frequency triode vacuum tube known as a "doorknob" for its appearance.

It appears that Randall and Boot, whose naivete in electronics was obvious, knew little or nothing of any efforts similar to their own. Ironically, both the Americans and the Germans had worked hard before the war to come up with a device to generate high power, short wavelength radio signals and missed the cavity magnetron, while the British basically stumbled onto it by accident and ran with it.



* The first operational British 10 cm (3 GHz) or "S-band" set was the shipboard "Type 271", which was rushed into production within months, with sea trials performed on a production set in March 1941. The Type 271 was a crude system, with manual direction and separate transmit and receive antennas, each in the form of a wide short open-faced box with a parabolic back and stacked on top of each other. The antenna configuration was nicknamed "Cheese", apparently because the antennas looked like they had been cut from the side of a round of cheese.

The Type 271 led to a long series of naval and ground-based radars, worth listing here though it is getting ahead of the story. The next step was the "Type 273" for major warships, with sea trials conducted in August 1941. The Type 273 used side-by-side parabolic dishes, each 90 centimeters (a yard) in diameter, and it was the first radar to be mounted on a gyrostabilized platform, like a naval gun, to keep it on target. Later versions had a PPI display.

A "CD Mark IV" version of the Type 271, sometimes just called the "Type 271 CD", was designed for coastal defense, featuring antennas with an aperture of 2 meters (6 feet 7 inches). Since the narrow microwave beam was effective at very low angles, the CD Mark IV, and its "Mark V" and "Mark VI" successors, were sometimes referred to as "Chain Home Extra Low (CHEL)". The Army CD Mark VI was also used by the RAF, being designated the "AMES Type 52", going through various refinements that were designated "AMES Type 53" through "AMES Type 56".

The Royal Navy continued to improve on the Type 271/273 design to come up with the "Type 277" 10 cm (3 GHz) radar, which was also used by the RAF in a mobile installation as the "AMES Type 14". The initial marks of the AMES Type 14 used dual cheese-style antennas, mounted horizontally, though late marks had a lighter and more effective antenna.

The Royal Navy Type 277 was used as the basis of an RAF "centimetric heightfinder (CMH)" radar, the "AMES Type 13", which preceded the AMES Type 14 into service. The CMH used Type 277 electronics, coupled to a stacked dual cheese-style antenna that was mounted vertically, creating a horizontally flattened beam that was ideal for determining heights of intruders. The AMES Type 13's antenna rocked back and forth or "nodded" vertically, from one degree down to 20 degrees up, to scan for targets. Later marks of the Type 13 discarded the heavy cheese antenna for a lighter and more effective mesh-style antenna.



* The naval Type 271 was the first operational centimetric radar only because naval radar was the easiest to do. In 1940, the priority was S-band AI for nightfighters.

It was clear that developing "centimetric" radar systems would take time, money, and engineering resources; Britain was desperately short of all three. Churchill knew that Britain needed American scientific, engineering, and manufacturing resources to win the war. He decided to unconditionally share the cavity magnetron and other British technical secrets with the US to ensure rapid development and deployment of the new technology.

Sir Henry Tizard had been promoting a meeting between British and American scientists, his contacts across the Atlantic having told him that there were many American scientists who really wanted to help defeat Hitler. At the beginning of August 1940, Churchill authorized Tizard to form a team to take Britain's most promising new technologies and demonstrate them to the Americans. Tizard formed up a party of seven, including himself. He recruited a prominent physicist, John Cockcroft, as his deputy. The team needed a radar expert. Eddie Bowen had been transferred from Saint Athan to more comfortable surroundings at Swanage to work on microwave AI. He was not doing anything of great significance, one of the advantages of having been sidelined by the management, and so he was recruited, rescuing him from bureaucratic oblivion.

Tizard and one of the team's military officers flew across the Atlantic, arriving in Washington on 22 August 1940, while the rest of the team left England on the Canadian ocean liner DUCHESS OF RICHMOND on 30 August. The liner spent so much time zig-zagging to avoid being torpedoed that the group called the vessel the "Drunken Duchess".

Bowen had been assigned to collect papers and hardware for the mission, including a cavity magnetron, and hand them over to military handlers for the voyage. He threw everything into a black metal box that had been designed to carry deeds and other commercial papers. Though Bowen was a relaxed sort, carrying around so many of Britain's vital secrets in one small box was nerve-wracking, and he was happy to pass the box off to its military escort.

The DUCHESS OF RICHMOND arrived at Halifax, Nova Scotia, on 5 September. The black box was shuttled off to the British embassy in Washington, arriving, after some worrying moments, on 9 September. The Tizard mission spoke with Canadian authorities but made no real effort to engage in discussions on radar technology, since the British thought the Canadians had no interest in the subject. In reality, the Canadians had built up their own radar lab after the outbreak of war in Europe, and developed their own prototype longwave sets, with very little knowledge of what the British were doing. The Tizard mission didn't find out about this work until they had gone south and were well into discussions with the Americans. The decision would then be made to inform the Canadians and get them involved.

* Tizard had made arrangements for the mission with his American opposite number, Dr. Vannevar Bush. In late June, Bush had managed to push through the creation of the US government "National Defense Research Committee (NDRC)", with Bush as chairman. The NDRC's mandate was to coordinate technology development between civilian scientists and military officials, and conduct preliminary technical research studies using such funds as were available. The materials provided by the Tizard mission fell precisely within the NDRC's charter.

The electronics technology brought over by the British was inspected by Alfred Loomis, a successful businessman and dedicated amateur scientist. Loomis had made a fortune as an investment banker and then gone into research, almost as a rich man's hobby that turned into a professional passion. He had extensive contacts in government and industry, and was in charge of a special "Microwave Committee" set up by the NDRC.

On 19 September, at a party hosted by Loomis, the British revealed the cavity magnetron. The Americans had roughly kept pace with the British in longwave radar development, and in fact members of the Tizard mission were surprised and impressed at what the Yanks had done, but the US had nothing like the cavity magnetron. Loomis immediately recognized the potential of the device, believing it would save US radar development two years of their own work.

Loomis, Bush, and other NDRC officials realized that a civilian research lab had to be set up outside of military control, using NDRC funding, to ensure that cavity magnetron technology was developed and deployed as quickly as possible. Bush had his roots in the Massachusetts Institute of Technology (MIT), and Loomis had strong ties to MIT as well, so MIT was chosen to host the new laboratory. This decision was made without any discussion with the president of MIT, Karl Compton, but Compton quickly agreed to the arrangement. The MIT radar research laboratory was originally named the "Microwave Laboratory", but within a few weeks it became the blandly named "Radiation Laboratory", or simply "Rad Lab".

At the end of October 1940, a conference on nuclear physics was in progress in Boston, and the officials organizing the emerging Rad Lab used it as an opportunity to make contacts and recruit researchers. Loomis and Compton set up a meeting at Boston's exclusive Algonquin Club, where the charismatic Bowen charmed the physicists and excited them with the new British technology.

Among the recruits was Isidor Isaac "I.I." Rabi, one of the most highly respected physicists in America, who brought along some of his best students. Another highly regarded physicist, Ernest O. Lawrence of the University of California at Berkeley, who had close ties to Loomis, enlisted several of his students in the effort, most prominently the brilliant and pragmatic Luis Alvarez. Alvarez was a superb "ideas man", able to come up with imaginative concepts, validate them, and then pass them on to others who would follow through with the exhausting "grunt work" of turning them into operational systems.

By mid-November 1940, the Rad Lab was a going operation, with physicists beginning research, while the lab's founders scrambled to get the operating details in place. By mid-December, the laboratory included thirty physicists, as well as the hired help needed to keep the place running. Alfred Loomis had executive oversight through the NDRC Microwave Committee, but day-to-day Rad Lab management was in the hands of Lee DuBridge, one of Ernest Lawrence's proteges.



* The Rad Lab of course wasn't the only American organization involved with radar. Bell Labs was still ramping up their radar work, but would become a major player in time. The service labs, the US Army Signal Corps and the NRL, had been in the game well before the creation of the Rad Lab, and had put their own longwave radars into production. American longwave radars were generally competitive with British longwave radars; the American military researchers thought the claims of the British were overblown, and found them "snooty, crusty, scornful, and antagonizing."

The antagonism would be aggravated when Watson-Watt came across the Atlantic at the end of 1941, just after America entered the war, to inspect American radar defenses. He ended up writing a report for US Secretary of War Henry L. Stimson that gave the Rad Lab high marks, but had nothing but contempt for Signal Corps work in the field. Partly Watson-Watt had been put off by the extreme confusion which had seized the US Army at the time, which was entirely to be expected given the abrupt outbreak of war and which was not news to the people caught up in it. Partly he was simply being parochial, sneering at the perfectly functional SCR-270 and pushing the US government to order a hundred of the unimpressive MRUs as an interim step towards obtaining Chain Home stations. After he went back to Britain, Signal Corps officers went through a mad fire drill, hastily cranking out reports to counter Watson-Watt's misinformation, and got the order for the MRUs canceled.

On top of this irritation, the Army Air Forces (USAAF), which was involved with radar development as a customer, decided that British radar was way ahead of the Americans in all respects, which even the Tizard mission didn't believe. The excessive awe in which the USAAF held the British had the unfortunate effect that the service pushed for US firms to simply build copies of British gear, much to the annoyance of American workers in the field.

Despite such difficulties, the Signal Corps was still trying to get along. They in fact sent a contingent of officers to the UK in the fall of 1941 to work with the British on radar. These US Army officers took orders from British superiors, and since America was officially a neutral at the time, the presence of these Yanks in British forces was a secret. The Signal Corps even lost track of the location of some of their people in Britain for a time. Two were killed before America entered the war.

The NRL was a bigger headache to the Rad Lab, since Admiral Harold Bowen was openly hostile, and things didn't improve until he was replaced by Rear Admiral Van Keuren.

For the moment, the Signal Corps and the NRL continued to work on longwave radars, while the Rad Lab focused on microwaves. Longwave radar had its uses, still does, and there was plenty of room for better longwave systems. The Rad Lab could work out the bugs in microwave technology and then pass it on to Army and Navy researchers, who could then perform the unglamorous but definitely nontrivial work of turning it into something that could be used by military forces in the field.

Everyone finally managed to get more or less on the same page, though there would always be rivalries and frictions, and it might not have always seemed that chummy to the players involved; bureaucracies are like that. Allied radar development benefited from cooperation, and as described later, Axis radar development suffered severely from the lack of it.



* The Tizard mission team suggested to the Americans that Rad Lab research be focused on three items, in order of priority:

The Rad Lab had the best and the brightest America could offer, as well as access to major financial, technical, and industrial resources. The problem was that they lacked experience. Eddie Bowen decided to stay and help the Rad Lab, likely motivated in part to put an ocean between himself and A.P. Rowe. Bowen, with his substantial practical experience in the field, became a demigod to the researchers. The fact that he was bright, hard-working, and easy to like didn't hurt, either. Luis Alvarez felt Bowen was the Rad Lab's most important member.

Bowen became known as "Taffy", an old slang term for a Welshman, derived from the River Taff in Wales. He kept the nickname for the rest of his life. Bowen acted as a conduit for transfer of information and hardware across the Atlantic, and spent a good deal of time traveling around America, visiting and evaluating military and industrial research facilities. He send detailed reports back to England as he surveyed the state of radar development in the US. The Americans clearly had a great deal of expertise that the British desperately needed. However, American radar work was unfocused and moving slowly, though Bowen hoped the Rad Lab would change that.

Rad Lab researchers worked hard to meet a 6 January 1941 deadline for demonstration of an experimental centimetric AI prototype. The sprawling prototype was set up on top of MIT Building 6, and was ready on 4 January. Tradition states that the first echoes were off the dome of the Christian Science Mother Church in Boston. It was not a practical system, featuring separate transmit and receive antennas, but it was a start.

The next deadline was to get a prototype flying in an aircraft by 1 February 1941. That meant obtaining a duplexer element to allow the same antenna to be used to transmit and receive. Within a month, they had a duplexer, what they called a "transmit-receive (TR)" element, that worked. They fitted a klystron tube as a preamplifier to the receiver crystal detectors, and the klystron's output saturated when the transmit pulse was sent out. The klystron would then recover before the reflected pulse came back.

That was a clumsy fix. The ultimate solution to the problem, obtained from the TRE in the spring, was a low-pressure gas tube named the "Sutton tube", developed by a TRE researcher named A.H. Cooke. The transmit pulse ionized the gas, making the tube conductive. The tube was in a circuit that turned off the receiver input when the conductivity of the tube dropped. The ions in the tube recombined immediately after the passage of the pulse, making the tube nonconductive and switching the receiver input on again. To ensure that the tube switched on quickly enough to protect the receiver, an electric filament was inserted to maintain a low state of ionization at all times, which encouraged a rapid "cascade" to full ionization when the transmit pulse arrived.

That was in the future. For the moment, problems were accumulating, and the group was diverging from a solution. Like the British crew in the early days of Bawdsey Manor, they were physicists and didn't always have the best grasp of practical electronic design. The fact that some of them were amateur radio enthusiasts who were used to tinkering with tubes and resistors helped, but the 1 February deadline came and went without a workable radar.

The researchers grew frustrated and redoubled their efforts, and within a week they began to converge to a solution again. On 10 March 1941, a Douglas B-18 Bolo bomber fitted with a prototype centimetric AI radar took to the air for preliminary tests. The aircraft-mounted system underwent a more serious test on 27 March, with Bowen, Alvarez, and two other Rad Lab members on hand. The radar successfully picked up surface vessels and a few US Navy submarines cruising on the surface. British radar development was slightly ahead of the Americans at the time, the British having already demonstrated airborne centimetric radar and in fact had performed an air-to-air detection on 10 March, but the Americans were making serious progress, and their detection of surfaced submarines was a first.

The detection of submarines led the Rad Lab to the formation of a new team to work on ASV and shipboard radar derived from the AI set. The team began work in May 1941 on fitting a centimetric radar to an old US Navy World War I "four-piper" destroyer, the USS SEMMES. Sea trials began in June, and the SEMMES became something of a floating extension of the Rad Lab for a year while the engineers tweaked the design, with continuous inputs from the Navy men on the destroyer.

In the meantime, the AI set was handed out of the Rad Lab and turned into an operational system. Western Electric began work on an operational version that would emerge as the "SCR-520", the first US AI radar. It would not be an operational success in itself, but it would be a basis for a number of successful operational systems.

* The Rad Lab settled down to a more structured existence, with Lee DuBridge in overall command. Routine management was conducted by F. Wheeler Loomis of the University of Illinois, while technical direction was provided by I.I. Rabi. Rad Lab staffers said that DuBridge and Loomis had a division of labor, with DuBridge always saying "yes" and Loomis always saying "no". It turned out to be a highly effective arrangement. Rabi, a Polish-born Jew who hated Nazis, always responded to technical demonstrations with the question: "How many Germans will it kill?"

By early spring 1941, the Rad Lab's progress had been remarkable, with Bowen writing Cockcroft that the gear was "ready for manufacture". The Rad Lab was still expanding its activities, setting up an investigation into 3 cm (10 GHz) or "X-band" radar headed by Norman Ramsay, one of Rabi's students. Unfortunately, the poor communications between the Rad Lab and the US military's labs remained in effect for the moment, and to Eddie Bowen's disgust, the British military was doing nothing with their American counterparts to improve matters, either. To make things worse, communications between the Rad Lab and the TRE were also not very good, partly because of the bad blood between Eddie Bowen and A.P. Rowe.

That began to change later in the spring. Few top American officials doubted that the US would be fighting Hitler soon, and had been taking increasingly strong measures to prepare the country for war and quietly assist the British, such as the "Lend-Lease" military assistance bill that US President Franklin D. Roosevelt passed into law in March 1941.

Vannevar Bush believed that war preparations should include the creation of a powerful government organization to coordinate wide-ranging scientific research outside of parochial military interests, something like the NDRC, but with much greater authority and much less haphazard funding arrangements. Bush's lobbying towards this end resulted in the creation of the "Office of Scientific Research & Development (OSRD)" in June 1941, with himself as director. The NDRC would become an element in the OSRD.

That same month, the British ordered ten AI sets from the Rad Lab for preliminary evaluation, with a possible order of 200 to follow if the results of the evaluation were good. The US Navy was impressed by the radar experiments on board the USS SEMMES, and placed an order for shipboard radars.

The Rad Lab began building production prototypes of X-band radars, and tested an X-band radar on the roof of Building 6. Cooperation with the TRE began to improve. A team from the Rad Lab, including Taffy Bowen, went to Britain in July 1941 for a competitive evaluation of American and British S-band AI radar gear. At first, the visit didn't seem to go well. Bowen was annoyed to find that while his American colleagues were almost in awe over British radar work, Dee seemed to believe that the Yanks had nothing to offer. Bowen found the attitude hard to understand, since early British longwave radar work was heavily indebted to American vacuum tubes.

Dee thawed out quickly, however. The prototype Rad Lab centimetric AI set seemed about as effective as the TRE set. Oddly, however, the Rad Lab transmitter was much better than the TRE transmitter, while the TRE receiver was much better than the Rad Lab receiver. The researchers mated the American transmitter to the British receiver and tripled the range of the radar.

As it turned out, the detector circuitry in the Rad Lab receiver used a triode vacuum tube, while the TRE receiver used silicon crystal diodes. The Americans had actually started out with silicon crystal diodes, but had poor results. The Rad Lab workers found out later that they had got hold of defective diodes. The little fiasco led to a strong American research effort on solid-state detectors that would have major implications for solid-state electronics technology after the war. In any case, the trip to Britain had paid off well for both sides, and cooperation improved significantly.



* By the time the Rad Lab team was performing tests in England with the TRE in July 1941, the British had decided that AI was no longer the top priority in radar development. What was really needed was improved ASV.

Partly the reason AI had dropped in priority was that even with longwave radars, RAF night fighter forces had become vastly more effective, due to a number of measures. The 1.5 meter (200 MHz) AI Mark IV sets had become much more reliable, and their operators far more competent thanks to better training. The AI Mark IVA included a feature that allowed the radar operator to send the location of the target being tracked to a display on the pilot's control panel, improving coordination between the two crewmen.

The RAF had also phased out the pathetic Blenheim night fighter, replacing it with the Bristol Beaufighter, a stubby brute of an aircraft that was produced, though not originally designed, specifically for the task. The Beaufighter was faster than the Blenheim and much more heavily armed, with six 7.7 millimeter (0.303 caliber) machine guns in the wings and four Hispano 20 millimeter cannon in the belly. The cannon had a nasty tendency to jam that took some time to work out, but when they worked, they had devastating killing power. Later on, the faster De Havilland Mosquito light bomber was pressed into service as an even more effective night fighter.

A particularly important innovation was a new layer in the radar network, known as "Ground Controlled Intercept (CGI)" stations, to help deal with the limited range of the AI sets in the night fighters. As mentioned, during the Battle of Britain Chain Home sets had been used to vector interceptors against intruders, but CH was inaccurate. That was not too much of a problem if pilots could look around the sky for the intruders, but that wasn't an option in dark night or murky weather.

Experiments had been conducted with CHL sets in early 1940 to investigate a GCI scheme. The lack of a rotating antenna and PPI made the scheme clumsy since it implied using one radar to track the intruder, another radar to track the interceptor, and then coming up with a workable scheme to link the two tracks together and get the interceptor on target. A PPI display made the task much easier, because it gave a ground controller a "picture" of both the interceptor and target.

The beginning of the night Blitz in the fall of 1940 put GCI on top of the priority queue. The TRE modified a CHL, adding an early PPI display, to operate as a GCI radar. Trials showed the new scheme to be very effective. The first GCI-controlled intercept took place on the night of 18 October 1940, with the controller guiding a Beaufighter with AI Mark IV to the target.

The TRE scratch-built six more GCI sets, with the first of them going into service on New Year's Day 1941. These early GCI sets of course operated at 1.5 meters (200 MHz) like their parent, the CHL, and were mobile systems. Following the six operational prototype GCIs, the TRE developed a full production GCI, the "AMES Type 7", which was in production by the end of 1941. It also operated at 1.5 meters (200 MHz), using a rotating dipole array antenna. Beam width was about 15 degrees, which was acceptable for its role though not all that accurate. but the AMES Type 7 permitted several ground controllers to use the same radar, and also provided longer range and more precise height determination. The set was built in both fixed-site and transportable versions, with the ground-controller tracking and communications kit part of the system.

Given that many British longwave radars operated at 1.5 meters (200 MHz), the British began to fear mutual interference and jamming. Microwave radars were the proper answer over the long run, but in the interim, in January 1942 six 50 cm (600 MHz) "AMES Type 11" CHL/GCI sets were ordered, and were delivered by the end of the year.

* With the technical improvements, the number of RAF night fighter kills had gone from a miserable two in December 1940, to 102 in May 1941. And then, abruptly, the targets went away. Although the Germans staged a raid on London on the night of 10 May 1941 that killed three thousand people and flattened the House of Commons, it was the last stab of the Blitz. The Luftwaffe was needed to support Hitler's invasion of Russia, planned for late June, and large-scale bomb raids on England had to be called off. The Luftwaffe's air assault on Britain was over.

The Germans still conducted small-scale raids, generally with fast FW-190 fighter-bombers, on selective targets; in early 1944, they would also attempt another night-bombing campaign, but the "Baby Blitz" would only result in heavy losses of aircraft and little damage to Britain, to be called off after a few months. From the summer of 1944, as discussed later, the Germans would try a different approach to bombarding Britain that would prove more effective.

A US Army Air Corps team visited Britain in mid-1941 and was so impressed by GCI and AI that they put the wheels in motion for the US to produce copies of them. The 1.5 meter (200 MHz) GCI became the "SCR-527" and AI Mark IV became the "SCR-540". However, American production of these radars was delayed. The SCR-527 wasn't delivered until the spring of 1943 and didn't prove all that useful when it was, since it had to operate in terrain much more varied than the gently rolling English countryside and wasn't up to the task. The SCR-540 never went into large-scale production, since it was completely obsolete by the time the Americans were ready to build it in quantity.

* The Luftwaffe's bombers ceased to be a deadly threat to Britain in the spring of 1941; the Kriegsmarine's (German Navy's) submarines, or "U-boats", took their place. The U-boats had been troublesome since the first days of the war, but the Royal Navy hadn't considered them particularly dangerous at first. The British had developed echo-locating systems, which they called "ASDIC" and the US called "sonar", after the First World War. Convoy systems had proven effective against U-boats in WWI, and with ASDIC and the convoy system, the Admiralty felt that the U-boat threat was manageable.

They were wrong, but the realization took much too long to sink in. The U-boats, under the command of Admiral Karl Doenitz, were using new tactics, particularly "wolf pack" attacks, in which a number of U-boats cooperated in nighttime surface assaults on convoys, overwhelming thinly spread escorts. After the fall of France in the spring of 1940, U-boats were able to operate from French Atlantic ports, increasing access of the submarines to the open sea. By the spring of 1941, U-boats were sinking at least a hundred ships a month and threatening to reduce Britain to starvation. The Germans called their rampage "the First Happy Time". The British finally realized the extent of the threat and took every measure to fight back. They built more escort vessels, obtained bases in Iceland for air and sea patrols, and after many difficulties were able to crack and read German naval ciphers.

Radar was potentially a strategic weapon in the "Battle of the Atlantic", as Winston Churchill called the war for the sea lanes. A World War II submarine was really a surface vessel that could hide underwater for relatively short periods of time while running at low speed on battery power. Radar allowed a submarine to be detected on the surface from long range; it could then be attacked, which at the very least would force it to stay under the surface, restricting its mobility.

Consolidated Liberator with ASV & cannon tray

The British had finally put longwave ASV Mark II in service, fitting it to over a hundred aircraft. While the antenna schemes varied between different aircraft or sometimes different installations on the same type of aircraft, a Short Sunderland flying boat or Consolidated Liberator patrol bomber generally featured a row of transmitting dipole aerials down its back, with two rows of shorter aerials on each side, and a receiving Yagi antenna near the tip of each wing for lobe switching.

Aircraft fitted with ASV Mark II had contributed to the sinking of the German battleship BISMARCK at the end of May, but longwave ASV was still not effective enough against U-boats. Submarines were much smaller targets than typical surface warships, and in the clutter of rough sea conditions they could be very difficult to detect. ASV.II had a very long minimum range of about 1,500 meters (5,000 feet), and pressing an attack in the dark on a moving target very troublesome.

Centimetric ASV was therefore a priority. The request went back across the Atlantic to the Rad Lab, and they reacted quickly. A TRE engineer named Denis Robinson was sent across the Atlantic to help work on centimetric ASV, where he was impressed with the expertise of the physicists at the Rad Lab, and astounded at the resources at their disposal. By December 1941, Rad Lab development of centimetric ASV was in high gear.