* As the Allies took the war back to the Germans, the two sides entered into a race to develop new measures and countermeasures. The result was a technological race for very high stakes.
* By the time the British dropped paratroopers on Bruneval, the Germans had a well-organized air defense system to protect the Reich from Allied air attacks. As mentioned, work on this network began in the summer of 1940, under the direction of Colonel Josef Kammhuber, who would rise through the ranks to major general. When Allied intelligence found out who was the mastermind behind the network, they called it the "Kammhuber Line", and the name stuck.
When Kammhuber began his task, the tools at his disposal included the old visual observer network, sound location gear, searchlights, and two types of radars -- Freya and Wuerzburg. The sound location gear was almost useless and quickly abandoned. Freyas could be used to direct a fighter to the vicinity of a bomber, but since Luftwaffe night fighters didn't have AI at the outset, they were guided to the final attack using searchlights directed by Wuerzburgs. That was obviously a clumsy scheme and the Germans put great effort into developing an AI, as discussed below.
Another problem was that the range of the Wuerzburg was inadequate. The solution was simple: Telefunken simply suggested that they increase the size of the Wuerzburg dish from 3 meters (10 feet) to 7.5 meters (24 feet 7 inches) to increase antenna gain. The rest of the radar was left generally unchanged, though the PRF was cut in half from 3,750 Hz to 1,875 Hz to adjust to the longer range.
The larger dish was tested and proved very effective, and so the "Wuerzburg Riese (Giant Wuerzburg)", mentioned earlier, was promptly put into production. The original Wuerzburg antenna was a simple solid dish, but the Wuerzburg Riese had an unusual lattice framework structure; the dish was built by the Zeppelin company, and its construction reflected techniques used to build airships. About 1,500 Wuerzburg Riese radars were built, with the type going into operational service in 1941. There was also a variant of the Wuerzburg Riese known as the "Gustav", an odd hybrid that had both Wuerzburg and Freya electronics, with a handful built and going into service in 1944.
The original Wuerzburg remained in service with anti-aircraft gun batteries. However, Telefunken felt they could improve on the design, and designed a new set, named "Mannheim", that operated on the same 50 cm (600 MHz) band but had greater accuracy. Some late models even had automatic tracking. Mannheim went into service in mid-1943, and about 400 were built. Since the appearance of Mannheim was similar, though not identical, to Wuerzburg, it was often simply referred to as "Wuerzburg". A handful of "Mannheim Riese" sets were built late in the war to support anti-aircraft missile research, with an oversized antenna like that of the Wuerzburg Riese, but the Mannheim Riese didn't go into production.
It is unclear if the Kriegsmarine ever used either Wuerzburg or Mannheim. The interservice rivalries of the Reich would argue against it, but some sources claim that Wuerzburgs were used at coastal sites, possibly including naval installations. In any case, it does appear that the GEMA developed a gun-laying radar named "Flakleit g" for the Kriegsmarine, which was based on Seetakt and used the same 80 cm (375 MHz) band. Photographs show the Flakleit g to have some similarities to the US SCR-268 set, with a horizontally oriented and a vertically oriented antenna. It is tempting to think it used both vertical and horizontal lobe-switching to accurately track targets, but information on this radar is very hard to find and specifics are unclear.
* From such beginnings, and after some trial and error to find the best procedures and tactics, the Kammhuber Line grew until in maturity it stretched in an arc from northern France, across the Low Countries, and into northern Germany, shielding the Reich from attacks from Britain. The line was divided into defensive cells, or "boxes", each 43 kilometers wide by 34 kilometers deep (27 by 21 miles). Each box contained a Freya and two Wuerzburg-Riese radars. The box was known as a "Himmelbett (Four-Poster Bed)".
The British realized that Freya was an early warning radar, and that Wuerzburg-Riese was used for aircraft tracking. Some British officials believed the second Wuerzburg-Riese was a backup, but R.V. Jones knew that the Germans couldn't afford that level of redundancy. He guessed correctly that one Wuerzburg-Riese was used to track intruding Allied bomber formations, while the other was used to track Luftwaffe interceptors.
The three radars fed fortresslike command posts that were conceptually similar to British GCI stations. However, the German operators lacked PPI displays, and instead worked from staggered rows of seats in front of a huge screen with a map of the battle area. The layout resembled that of a movie theater with bleacher seats, and the command posts were called "Kammhuber-Kinos (Kammhuber's Cinemas)" by German night fighter pilots. Operators in the lower bleacher seats shined lights on the screen to track the movements of aircraft. Blue lights meant friendly interceptors, red lights meant hostile bombers. Other operators standing behind the screen, trained in mirror writing, used marker pens to provide updates on the battle. Fighter controllers in the top seats kept the night-fighter crews up to date over radio. Smaller screens in "balconies" on either side of the "theater" provided updates for what was happening in the neighboring air-defense boxes.
Night fighters would stand by, flying orbits around a light / radio beacon, until a fighter controller got in touch with one and talked it to the vicinity of the target. The night fighter would then turn on its AI, acquire the target, and perform the attack. Although almost comically laborious by modern standards, given what could be done at the time, it was very ingenious. It was highly effective and a great improvement over the disjoint and primitive air-defense network that Kammhuber had at the outset.
* The Kammhuber Line was continuously upgraded throughout its existence. The Wuerzburg-Riese was one improvement, and GEMA had also come up with two derivatives of Freya that had greater range and much better accuracy.
One was named "Mammut (Mammoth)", which essentially consisted of 16 Freyas, linked together in a giant array with 192 dipoles, 30 meters across and 10 meters high (98 by 33 feet). It was mounted on four vertical structural beams, which led British intelligence to call it "Hoarding" (Britlish for the Yank term "Billboard"). About 20 were built, with the first going into service in 1942.
Mammut was a fixed-position radar, but it used electronic steering to scan over a field of view of 100 degrees. In yet another example of the parallel nature of radar discovery, the Germans independently developed phased arrays while the Americans were working on the same technology, and in fact Mammut was the first phased-array radar to go into production. Two Mammuts were often used together to give coverage over two quadrants. Mammut's operating specifications were similar to that of Freya, with the same 2.4 meter (125 MHz) Freya band, 3 microsecond pulse width, and 500 Hz PRF. However, Mammut had a much higher peak power of 200 kW, giving it a range of about 320 kilometers (200 miles). It used horizontal lobe switching to obtain positional accuracy of about half a degree in the horizontal plane.
Mammut had no real ability to determine altitude, being designed as a long-range warning radar to identify groups of intruders. Getting a better fix on the groups once they were detected was the business of the second derivative of Freya, named "Wassermann (Waterman)". There were a number of different versions of Wassermann, but they all essentially amounted to eight or more Freyas mounted vertically on a steerable tower 60 meters (190 feet) high. About 150 were built, with the first going into operation in 1942.
Wassermann used electronic beam steering and lobe switching to achieve a vertical resolution of about 0.75 degree in the middle of its field of view, and a horizontal resolution of about 0.25 degree. Again, signal parameters were the same as Freya's, except that Wassermann produced a peak power of 100 kW and had a range of about 240 kilometers (150 miles). One variant, the "Wassermann S (Schwer / Heavy)", had the array mounted around a tall cylinder, and so the British named the radar "Chimney", applying the same name to the other variants of Wassermann.
Both Mammut and Wassermann were excellent radars and became the backbone of the German early warning network. They proved surprisingly difficult to knock out, though eventually Allied strike fighters found that a barrage of rockets could do the job.
The Germans also operated a passive signals intercept operation, known as the "Y-Dienst (Y-Service)", which used directional antennas and triangulation to locate Allied bomber formations from their radio emissions. The Y-Dienst had the advantage of being impossible to jam, and was an important element of the air-defense system.
* Other German radars saw limited or selective service with the air-defense system. One was known as "Jagdschloss (Hunting Lodge)", which was a wide-area radar produced by Siemens from development work conducted by GEMA. About 80 were built, with the first in service by early 1944.
Jagdschloss was used for tracking bomber formations over long ranges. The radar featured 18 dipoles mounted on a rotating horizontal structural beam 20 meters (66 meters) wide to generate a narrow radar beam in the shape of a vertical fan, with good horizontal resolution but little or no altitude determination capability. Jagdschloss operated at a slightly higher band than Freya, ranging from 2.3 meters to 1.8 meters (129 to 165 MHz); had a pulse length of a microsecond, three times better than Freya; a PRF of 500 Hz; and a peak power of 150 kW.
Jagdschloss was the first German production set with a PPI, transmitting PPI information directly to remote command headquarters over dedicated landlines, or directional 50 cm (600 MHz) radio links.
In late 1943, the Germans also introduced an ingenious passive radar system named "Klein-Heidelberg", probably developed by Telefunken considering the name, that was used only in coastal regions along the North Sea and English Channel. It worked by sensing Chain Home emissions, with one small antenna focused on a CH station to obtain the original radar pulse, and a larger steerable passive antenna, based on the Wassermann-S, that picked up CH reflections on bomber formations.
The time delay between the original and reflected CH signals defined an ellipse on a map, with the CH and Klein-Heidelberg stations at the focal points, that plotted the possible locations of bomber formations. The steerable antenna pinned down the actual location along the ellipse with accuracy adequate for an early-warning system. As with the Y-Dienst, it was effectively impossible to jam. It could only have worked with a floodlight radar like CH, since the Klein-Heidelberg wouldn't pick up the output beam of a steerable or rotating radar set most of the time, or find the signals easy to interpret if it did.
* As a puzzling footnote to German radar development, late in the war the Germans introduced a radar named the "Elefant-Ruessel", a floodlight system with similarities to the British Chain Home longwave stations. It was built by the German Postal Authority with some help from Telefunken. Records on the Elefant-Ruessel are sketchy; it has been seen as a copy of CH, but that seems a simplistic read on the matter. It appears to have been designed primarily as a primitive "over the horizon (OTH)" radar, to pick up air and naval traffic at long ranges over the sea.
Elefant was preceded by "Heidelberg", no close relation to Kleine-Heidelberg, an OTH radar set up north of Amsterdam in 1942, following preliminary experiments on OTH radar using Knickebein technology. Heidelberg obtained range using long HF wavelengths of 13 meters (23 MHz) -- originally 20 meters (15 MHz), but interference was too great -- and high power, with a long PRF to permit distant returns. It could spot sea traffic over the horizon, but it was very vulnerable to sea clutter, with horizontal polarization used in hopes of minimizing the problem.
Heidelberg was followed by two improved experimental OTH radars set up in northern Germany, the first being "Langwellenortung", operating at 36 meters (8.33 MHz) with a PRF of 50 Hz, featuring narrow beam steering of 4 degrees in either direction by phase steering. Using "bounce" of signals from the ionosphere, on occasions it could pick up reflections of lakes deep inside Finland, about 1,500 kilometers (930 miles) away. It was followed by a fully steerable "Rundbluck" system.
The "Elefant" system was set up near the shores of the Netherlands, north of Amsterdam, in 1943. It was, again, similar to CH, operating in a floodlight fashion over a 120 degree arc; it had a peak power of 350 kW, operating in the bands 23:28 and 31.5:37.5 MHz, along with a 10 microsecond pulse and a 25-hertz PRF. Elefant had two transmit towers and a receiver tower like that of Kleine Heidelberg; Elefant could be seen as a follow-on to Kleine Heidelberg.
The See-Elefant system followed in 1944, being set up on the west coast of Denmark, integrated with a Mammut site. It had a single transmit antenna, "Kopfje (Elephant's Head)", consisting of two towers with cables strung between them, "Ruessel (Elephant's Trunk)" was the receiver, which was located about 1.5 kilometers away and used a steerable antenna to determine the direction of the echo. Operational parameters of the See-Elefant were the same as those of Elefant. See-Elefant was apparently designed as a backup system, to give radar ranging in case Allied jamming blinded other radars, the Germans believing that its wideband operation gave it resistance to jamming. It was also used to track V-2 rocket launches, spotting their impacts in Britain.
* The Germans did not set up a comprehensive defense network in the East comparable to the Kammhuber Line. The Soviets were not into strategic bombing, the primary mission of the Red Air Force being battlefield support of the troops -- and the battlefront tended to shift, sometimes drastically, making investments into large fixed-site installations a poor use of resources.
The Germans adapted to the more fluid air-defense environment by being flexible. After the German advance into the USSR bogged down in late 1941, they set up a few Freyas at fixed sites, and then mounted the Himmelbett radar system on trains, shuttling them to where needed. They even mounted such a system on a cargo ship, and operated it in the Baltic beginning in early 1944 and up to the end of the war.BACK_TO_TOP
* As mentioned, early on Luftwaffe night fighters had a desperate need for an AI radar, which was finally supplied by a Telefunken-built radar named "Lichtenstein", codenamed "Emil-Emil". The first Lichtenstein prototype was flying by the summer of 1941, and went into service as the "Lichtenstein B/C" in the spring of 1942. Lichtenstein B/C used an array of four cross-shaped antennas mounted on the nose of a night fighter. The radar operated at 50 cm (600 MHz), with a wide beam that provided a good field of view, a somewhat short maximum range of no more than 4 kilometers (2.5 miles), and an acceptable minimum range of 200 meters (660 feet). It used conical scanning, featuring a rotating element in the antenna electronic systems to spin the beam around. It was Carl Runge's last contribution to the German war effort, since personal feuds eventually forced him out of Telefunken.
Lichtenstein B/C was mounted on the Luftwaffe's primary night-fighter, the Messerschmitt Bf-110 twin-engine fighter, as well as on night-fighter variants of the Junkers Ju-88 bomber. The larger and more powerful Ju-88 was a better night fighter than the Bf-110, but Ju-88s were used for many roles, and there were never enough of them go around. Pilots were initially unhappy with the radar since the antenna arrays cut into the performance of their aircraft, but soon found out that it was far more effective than chasing after bombers illuminated by searchlights.
RAF listening posts heard references between fighters and ground controllers about Emil-Emil, though they didn't know what it was. After an electronic intelligence station in Britain picked up a signal from what seemed to be a German AI, a Vickers Wellington bomber was fitted with a receiver set to the appropriate frequency range and sent over Germany on a ferret mission in early December 1942. The mission worked a little too well, since the bomber was chewed up by a Ju-88 night fighter and almost didn't make it back home.
The Allies actually got their hands on a complete Lichtenstein B/C set on 9 May 1943, when a Ju-88 night fighter landed in Scotland, the crew having decided to defect. However, by this time, it was on the threshold of becoming old news. The Germans were working on an improved version of Lichtenstein, the "SN-2", which provided greater range than Lichtenstein B/C.
The range increase was to be obtained by going to longer wavelengths, which turned out to be in the range of 4.1 to 3.7 meters (73.2 to 81.1 MHz), which were easier to generate at higher power levels. It also produced a wider beam, making it easier to spot intruders. This had been the crippling flaw of British longwave AI radars, since a wide beam led to too much ground clutter at low altitudes to be useful -- but since RAF night bombers generally operated at high altitudes, the wide beam was a benefit to the Luftwaffe.
As an unintended benefit, the new band also successfully camouflaged the new Lichtenstein from the Allies from several months, since the lower frequencies were in the Freya band and Allied ferrets didn't realize that something else was there. In some cases, a Lichtenstein B/C radar was also fitted to a night fighter along with the SN-2 radar to provide close-range targeting, the B/C using a single small antenna array nested inside the SN-2 array.
* Development of a radar-based air-defense network led the Germans to the problem of IFF, just as it had the British, and IFF proved even more troublesome for the Germans. An effective IFF system depends on standardization, and since the Nazi leadership's technology policies were inconsistent and scatterbrained, IFF suffered accordingly.
At first, the Luftwaffe used a modified version of the Y-Geraet blind bombing system known as "Y-Verfahren" for fighter direction. The return signal that gave range though phase shift was transmitted for 20 seconds out of every minute by the aircraft's radio, an idea consciously lifted from the British Pip Squeak scheme. Y-Verfahren basically put the burden of IFF on the ground station, but had the advantage of giving the range to the fighter. It also provided a navigation beam to get the night fighter back home in the dark and foul weather, a task which could be as dangerous as taking on an armed bomber.
However, like Pip Squeak, Y-Verfahren suffered from the fact that it was difficult to integrate with the radar network, and so the Germans worked in parallel on two more sophisticated IFF schemes:
It is less accurate to say that these were competing efforts then they were mutually indifferent. Wolfgang Martini protested the notion of fielding two different IFF systems, each of which only gave half the solution. In fact, Zwilling wasn't even half the solution, since it proved unworkable in practice. Thousands were built, but many of them ended up being cannibalized for parts to build Erstlings.
An improved "Zwilling J1" was built, but the better solution was to make Erstling compatible with Wuerzburg, or as it worked out, make Wuerzburg compatible with Erstling by the straightforward measure of tacking a 2.4 meter (125 MHz) interrogation system onto a Wuerzburg dish. The interrogator was named "Kuckuck (Cuckoo)", with the scheme first introduced in the summer of 1942.
German IFF still didn't get fully on track. The Allies were quick to figure out ways to interfere with it, and German pilots feared, correctly, that the enemy would also develop devices to home in on it. The Germans never even developed an IFF interrogator that could be carried on an aircraft, which hobbled the night-fighter crews. German engineers continued to work on IFF through the rest of the war, but never fielded an effective system.
* Incidentally, the Germans also built an ASV radar, named "Hohentwiel". It was designed by the Lorenz company, which hadn't given up on radar after losing the gun-laying radar contract to Telefunken's Wuerzburg. Hohentwiel operated at a wavelength of 50 cm (600 MHz), and featured various arrangements of cluttered dipole arrays, compared to antlers, mounted on the nose of an ocean patrol aircraft, such as the Focke-Wulf FW-200 Kondor or the Junkers Ju-188.
Hohentwiel was an excellent set. It could spot a merchantman at a range of 80 kilometers (50 miles) and could pick up a submarine periscope at six kilometers (3.75 miles). Like the British ASV.II, it could perform searches by scanning to the sides, and then pinpoint a target in the forward direction using lobe switching. A version of Hohentwiel was built for U-boats as well, but by the time it was available U-boat crews were so jumpy and intimidated that they didn't want to use it, since they feared that the Allies might home in on it.BACK_TO_TOP
* By March 1942, the Kammhuber Line was beginning to seriously bloody the night raiders. R.V. Jones knew that disrupting German radars would blind the Kammhuber Line, and he knew just how to do it. As far back as 1937, he had suggested that a piece of metal foil falling through the air might create radar echoes. In early 1942, a TRE researcher named Joan Curran, the only woman among the boffins, had investigated the idea and come up with a scheme for dumping packets of aluminum strips from aircraft to generate a cloud of false echoes. The strips were to be cut to a half wavelength of the operating frequency of the radar to be jammed, though later quarter-wavelength strips were used as well. Tests against centimetric AI.VII and AI.VIII radar showed it to be highly effective. The scheme was codenamed "Window".
Although R.V. Jones wanted to use Window right away, he was overruled. Lord Cherwell, Watson-Watt, and RAF Fighter Command opposed the use of Window, since they believed that once Window was used, the Germans would immediately learn the trick and use it on raids over the British Isles in turn.
In reality, the trick was much too obvious; like so many other inventions in the Wizard War, the British hadn't been the only ones to think of it. The Germans also performed tests in 1942 on using foil strips to jam radar, calling the scheme "Dueppel". German leadership proved every bit as nervous about the idea as their British counterparts. When Hermann Goering heard about the tests, he ordered them to be stopped immediately and reports on the experiment suppressed. Goering had greater reason for fear than the British: by this time, Allied offensive bombing capabilities clearly outstripped the German ability to retaliate in kind, and in fact the imbalance would only get worse. Dueppel clearly would hurt the Reich far more than it would hurt the British.
* The British used other countermeasures and kept Window in reserve. One of the simplest countermeasures was to simply increase the size of the attacking formations, up to the size of the "thousand bomber raids" conducted by Bomber Command beginning in May 1942. This overwhelmed the defenses, providing far more targets than a Himmelbett cell could tackle. However, it also created a "target rich environment", and Luftwaffe night-fighter pilots eventually adjusted by simply having ground control direct them into the bomber stream, where they hunted targets on their own. The tactic was known as "Zahme Sau (Tame Boar)", and since it placed more load on a night fighter's AI system, it led to the effort to develop the improved Lichtenstein SN-2 AI. Kammhuber protested the use of Zahme Sau, insisting that he could compensate if given more resources, but he was ignored.
Simple numbers of bombers were not a very effective defense against the Luftwaffe, and so in the summer of 1942 the RAF introduced a broadband radar jamming system named "Mandrel". Mandrel was designed by Robert Cockburn's countermeasures group at the TRE, that blinded Freya, Wassermann, and Mammut early-warning radars by throwing out radio noise on the Freya band. Mandrel was originally installed on fighters that escorted bomber formations to their targets. When Mandrel was installed on RAF bombers beginning in December 1942, RAF bomber losses fell substantially.
The Germans adapted by retuning some Freya systems from 2.4 meters (125 MHz) to 2.78 meters (107.9 MHz). The British responded in turn by modifying Mandrel to cover both frequencies, and the race between radar and jammer was on, with the British developing many variants of Mandrel.
The British also developed a jammer named "Shiver" to disrupt Wuerzburg, though Wuerzburg's tight beam made that a much more difficult task -- leading to an improved jammer named "Carpet" that could be tuned to various frequencies in the Wuerzburg band. While jamming had generally been done to that time using sine-wave signals, Cockburn had concluded that using random "white noise" signals was a much better option. It was much harder to compensate for, and it could be misinterpreted by victims as a defect in their own gear or natural interference.
White noise generators were improvised from what was available. Gas-filled tubes that ionized at specific voltages were used in voltage regulators at the time, and it turned out that they produced a nice noisy output if filter capacitors were removed. Photomultiplier tubes, normally used to amplify faint light signals, also produced a noisy "dark current". Both these devices were used as noise generators until purpose-built gas tubes were introduced.
A formation of 20 bombers, each with a Carpet jammer, could blind Wuerzburgs and Mannheims. Unfortunately, a bomber that radiated a continuous jamming signal also announced its presence to enemy fighters. The British came up with a new, more intelligent jammer system named "Carpet II", which was able to scan through the Wuerzburg frequency range, identify a frequency in use, jam it for 30 seconds, and then move on. The jammer did not stay on frequency long enough to allow a fighter to home in on it, and with enough bombers, jamming coverage was effectively continuous. The concept of jumping frequencies, known as "frequency agility", would became more important in the following decades.
Still another jammer, "Airborne Grocer", was developed to jam Lichtenstein 50 cm (600 MHz) transmissions. They also designed a "Ground Grocer" counterpart, but it operated from long range and simply didn't have the power to be effective.
* Jamming radars by dumping noise into them was a crude trick. A more elegant approach was to deceive or "spoof" them, or trick them into seeing things that weren't really there. Like many of the weapons of the Wizard War, spoofing was more or less invented by accident. In July 1941, British technicians were calibrating a Chain Home radar when an aircraft transmitted IFF signals into the radar at relatively high power by accident. To the technicians, the aircraft looked like a large number of fighters approaching, and the British realized that such a technique might be used to confuse German Freya radars.
Cockburn's team developed a device named "Moonshine" to spoof Freya, and tests in March 1942 against Chain Home stations demonstrated its effectiveness. Moonshine was a "pulse repeater". It listened for a pulse from a radar transmitter, and then fired back a spread-out pulse on the same frequency that fooled the radar receiver into "seeing" a reflection that appeared to be from a huge formation of bombers.
Moonshine transmitters were installed in twenty obsolescent Boulton-Paul Defiant two-seat fighters beginning in the spring of 1942. Moonshine went into operation that summer, with the Defiants providing a diversion for the USAAF's first raid on Europe, against Rouen, France, on 17 August. Moonshine proved valuable at first, but it was only used for a few months since the Germans gradually got wise to it and recognized the spoofing for what it was. It was then set aside to be used on special occasions when the Germans weren't expecting it.
* The British also attempted to jam or spoof communications channels between night fighter and their ground controllers. In December 1942, along with Mandrel, the British introduced a radio jammer named "Tinsel", which used microphones placed in the engine nacelles of a bomber to broadcast loud audio noise over Luftwaffe communications links.
Improved radio jamming systems were introduced in 1943. "Ground Cigar" went into operation in July, followed by "Airborne Cigar" in August, and then "Corona" in October. Corona was a full-fledged spoofing operation, with German-speaking British "controllers" breaking in on German ground-controller channels and trying to confuse night fighter pilots. At first, the British "controllers" were given away by obvious accents, but they refined their skills, and there were occasions where Luftwaffe pilots sat through arguments between two controllers, each insisting that he or she was the German and that the other was the Briton.
The RAF also configured aircraft as dedicated electronics warfare aircraft. Boeing B-17G Flying Fortress bombers, known in RAF service as Fortress IIIs, were configured with H2S, jammers, and warning systems and used to protect bomber streams.BACK_TO_TOP
* Over in the US, in December 1941, not long after Pearl Harbor, the push came down from the top to establish a countermeasures group named the "NDRC Division 15" or the "Radiation Research Laboratory (RRL)". The RRL originally existed as an office in the Rad Lab. Its objectives were to develop countermeasures systems; help improve the resistance of US radars and other electronic systems to enemy countermeasures; and develop signals intelligence systems to monitor enemy radars and other electronic systems.
Luis Alvarez was offered the job of running the RRL, but he had other commitments and didn't feel up to a major management job anyway. On his recommendation, the RRL was given to Frederick Terman, previously head of Stanford's Electrical Engineering department.
The RRL was originally set up as an office at the Rad Lab. There was some tension between the RRL and the rest of the Rad Lab; a few of the Rad Lab engineers reacted emotionally when RRL engineers pointed out vulnerabilities in Rad Lab radars, the Rad Lab people feeling that the RRL was making life unnecessarily harder for them. Cooler heads pointed out that the enemy would certainly implement all the countermeasures they could think up and it was important to stay a step ahead; the message was accepted, if grudgingly.
However, the RRL quickly grew large and was soon, as planned, spun off as an independent organization, sited at Harvard University. Terman ran the RRL under an interesting arrangement where the research staff was basically on loan from their normal employers, who were still responsible for pay and benefits. The government reimbursed the companies for their expenses. This scheme allowed the workers to acquire seniority with their normal employers while they were working at RRL.
The RRL also had to spend considerable effort to make sure their most unreplaceable staff weren't snatched up by the draft board and put in the ranks, with the RRL in some cases sending their people overseas on field investigations to get them out of reach. The NRL and Signal Corps didn't have this problem, since if any of their people were targeted by the draft, they could be simply enlisted and come into work in uniforms instead of civilian clothes.
In April 1942, Terman went to the UK for six weeks to visit with Cockburn and his group at the TRE. The two men got along very well, and Terman was impressed by the skills of his British hosts. Cockburn visited the US in October 1942, of course including RRL in his list of places to visit, and was impressed in turn. The cooperation between the TRE and RRL remained good through the conflict. The RRL established an "American-British Laboratory of Division 15", or just "ABL-15", at the TRE site in Malvern to ensure that the two groups worked closely together. ABL-15 eventually grew to be as big as the TRE itself, with the joint lab focusing on immediate operational issues while the TRE conducted more fundamental research.
* Not all of the collaborations worked well. In response to the limitations of the British Ground Grocer Lichtenstein B/C jammer, the RRL developed a monster ground-based jammer named "Tuba", with an oversized horn antenna and a small fleet of support trucks. Tuba could light up fluorescent bulbs a mile away, and flammable things accidentally placed in the output path quickly caught on fire. It was superficially impressive but in practice never very effective -- proving a nuisance to German night-fighters that were just across the Channel, but causing no real bother otherwise.
Some other RRL projects worked much better. The RRL built their own designs of the Mandrel Freya jammer, known as "AN/APT-3" for airborne use and "AN/SPT-3" for shipboard use, and the Carpet Wuerzburg jammer, known as "AN/APT-2" or "AN/SPT-2". As with the British, the RRL built improved versions of Carpet, and also built a number of other jammers, including "Dina", "Rug", and "Broadloom" to cover the spectrum of enemy radars.
The USAAF resisted the introduction of the Carpet for a time, since their initial bombing policy was for daylight, clear-weather strikes where radar jamming of Wuerzburg was of little use, but RRL managed to convince the service to adopt it anyway. The wisdom of this would become obvious once the USAAF began to use H2X to bomb on overcast days.
Late in the war, RRL also introduced a communications jammer, the "AN/ART-3 Jackal", which was used to disrupt German tank radios. It was used in support of Allied counterthrusts against German armor during the Battle of the Bulge in late 1944 and early 1945. RRL also developed microwave radar jammers, but as will be discussed the enemy never got any microwave radars into real operation.
Ironically, even before the RRL had built any jammers, the USAAF had been jamming German radars without trying. The SCR-522 VHF radios carried on USAAF bombers happened to be on the same band as the Freya and could jam the radar when the radios were set to certain channels. The Germans realized this in early 1943, which suggested to them the possibility that they could use their own radios to jam Allied radars, but as it turned out no Allied radars were on the same band. The USAAF was slow to realize that the SCR-522 was jamming the Freya -- but if the SCR-522 jammed the Freya, that meant that Freya was also jamming the SCR-522, and aircrews returning from missions over Europe complained of severe interference on certain radio channels. The problem was eventually traced to Freya, and RRL engineers designed fixes for the problem.
The Germans developed their own longwave jamming gear to blind Allied radars, including the powerful ground-based "Karl" jammer and much less powerful "Kettenhund" jammer, which was carried by bombers. They proved effective in disrupting SCR-268 longwave fire-control radars, leading to hasty efforts by the RRL to modify the SCR-268s to cut through the jamming. However, the German jammers were totally useless against microwave radars.
* Signals intelligence gear was also part of the RRL's charter. The Americans, or at least the US firms Halicrafters and General Radio Company (GRC), seemed to have acquired a particular skill at making radio receivers before the war, and during the conflict these receivers formed the basis for an ever-expanding range of Allied SIGINT gear. The British used the Halicrafters S-27 to pick up Seetakt and Freya emissions in early 1941, and the S-27 would be widely used by the Allies through the rest of the war.
The S-27 was a good piece of gear, but something more purpose-designed was needed. Luis Alvarez, always with a bright idea for somebody else to develop, enlisted an RRL engineer, a Canadian named Dr. Don Sinclair, to work with GRC to work between the RRL and GRC to modify the company's "P-540" receiver into the US Army "SCR-587" (US Navy "ARC-1") SIGINT receiver, built by Philco. The SCR-587 provided a bandwidth from 3 meters to 30 centimeters (100 MHz to 1 GHz) and was a valuable tool through the entire war. Don Sinclair would perform valuable work for the RRL in the lab and the field through the war. After the war, he would go to work for General Radio and eventually become company president.
The NRL also built small numbers of a SIGINT receiver, the "XARD", that operated over the band from 6 meters to 30 centimeters (50 MHz to 1 GHz). It was a crude piece of gear, with tuning performed by adjusting the antenna, a tiresome scheme when the operator was searching back and forth over the band for possible "emitters". It was also not very reliable, but it was one of the first US-built SIGINT receivers to be put into action, being taken on submarine patrols in the fall of 1942.
* The SCR-587 was a much better piece of gear than the XARD, but even the SCR-587 had to be manually dialed over its band, which was tiresome and error-prone. The answer was to automate the process, allowing the receiver to automatically scan through its band.
The requirement was passed on to Peter Goldmark, an RRL engineer who had been borrowed from CBS. After the war, he would develop an alternative color TV system that didn't prove successful but gave TV rival RCA fits, and more successfully developed the 33 1/3 RPM long-playing (LP) phonograph record -- a universal technology, until it was abruptly replaced by the laser compact disk. His scanning receiver, the "AN/APR-2", operated over the same band as the SCR-587. It could be set to scan over its bandwidth at different rates, using a neon indicator to show the frequencies where a signal was detected, and recording up to eight hours of monitoring on an electrochemical tape.
The AN/APR-2 was a complicated piece of gear and development proved troublesome. Goldmark was not amused when one of his colleagues suggested that they drop a prototype on Japan, so that the enemy would struggle in vain for the rest of the war trying to get it to work. However, the RRL did manage to get it into service.BACK_TO_TOP
* At the end of 1941, the Japanese began a wide-ranging offensive that swept through the colonial possessions of the British, Americans, and Dutch in the western Pacific, reaching as far southeast as the north coast of New Guinea to threaten Australia. Among the benefits of this spectacular wave of conquest was the fact that the Japanese obtained a number of British GL-type sets from Singapore, as well as a US SCR-268 set and a damaged US SCR-270 set from Corregidor.
The IJA put a modified version of the GL into production as the "IJA Tachi 3". It operated on a band around 3.75 meters (80 MHz), had a pulse width of one to two microseconds, a peak power of 50 kW, a PRF of 1,000 or 2,000 Hz, and a maximum range of about 40 kilometers (25 miles). About 150 were built by Sumitomo, with the type going into service in early 1944. The Tachi 3 set was the first Japanese set to incorporate Yagi antennas, which was a great irony, since such antennas were the invention of Hidetsugu Yagi, a Japanese electronics researcher of global stature. To add to the irony, Yagi had been involved in the development of the IJA Type A interference detector.
On their part, the IJN recognized the SCR-268 as a good piece of gear and put a derivative of it into production as the "IJN Mark IV Model 1". It operated in a band around 1.5 meters (200 MHz), had a pulse width of 3 microseconds, a peak power of 30 kW, a PRF of 2,000 Hz. and a maximum range of about 48 kilometers (30 miles). It was followed by the improved "IJN Mark IV Model 2", which had basically the same general specifications except that the PRF was reduced to 1,000 Hz. The Japanese built a few hundred of these radars in all.
The IJA also tried to build derivatives of the SCR-268 in the form of the "IJA Tachi 1", "IJA Tachi 2", and "IJA Tachi 4", all operating on the 1.5 meter (200 MHz) band used by the SCR-268, but these radars did not prove satisfactory and were only built in small numbers. Late in the war, the IJA did introduce a much more workable derivative of the Tachi 4, the "IJA Tachi 31", also operating at 1.5 meters (200 MHz), with 70 built.
* In the meantime, both the IJN and IJA fielded derivatives of their earlier fixed-site radars. The IJN Mark I Model 1 was followed in 1942 by about 300 of a lighter transportable 1.5 meter (200 MHz) version, the "IJN Mark I Model 2", and then in 1943 about 1,500 of an even lighter portable version, the 2 meter (150 MHz) "IJN Mark I Model 3".
As if in parallel lockstep, the IJA followed their Tachi 6 in 1943 with about 60 transportable 3 meter (100 MHz) "IJA Tachi 7" sets, and in 1944 followed that with about 400 portable "IJA Tachi 18" sets, operating in the same band.
Other than being lighter, these radars were no great advance over their predecessors, being roughly comparable to the British MRU. However, since the Japanese had developed their own magnetron, in fact well ahead of the Allies, they also developed their own 10 cm (3 GHz) microwave set for naval warfare. The "IJN Mark II Model 2" radar was introduced in 1942, and was well-received by naval crews as a great step ahead of the unsatisfactory longwave Mark II Model 1. About 400 were built and deployed on a range of vessels.
The Mark II Model 2 had a peak power of 2 kW, a pulse width of 2 to 10 microseconds, a PRF of 2,500 Hz, and a range of about 35 kilometers (22 miles) against a large naval surface target. It had separate cone-shaped transmit and receive antennas, giving it the odd appearance of giant toy binoculars. It did not have a PPI, no operational Japanese set ever did, which greatly limited its usefulness for naval operations.
The Japanese also developed a lightweight longwave set, the "IJN Mark II Model 4", operating at 1.5 meters (200 MHz), for use on small vessels and submarines. It is unclear if it saw much service.
* During the first months of the US war against Japan, the Americans were so overwhelmed that worrying about Japanese radar capabilities didn't even make the list. The issue didn't come to the surface until the US Marines landed on Guadalcanal in the Solomon Islands on 7 August 1942. The landings were not heavily opposed -- a situation that gave a completely misleading impression of what to expect in the future -- and the Marines quickly captured an IJN Type I Model 1 radar. The catch came as a surprise, apparently less because anyone thought the Japanese didn't have radar than because few had given the matter much thought. The Japanese radar was dismantled and shipped stateside. NRL researchers found it crude, even in comparison with early American radars such as the SCR-270 and CXAM.
SIGINT receivers were quickly installed on submarines and aircraft to hunt for more Japanese radars. A Consolidated B-24 Liberator ferret that had been fitted with various SIGINT gear, including some lab breadboards, performed probes of the Japanese-held island of the Kiska in the Aleutians in March 1942, and discovered the signatures of two more IJN Type 1 Model 1 radars, which the SIGINT operator reported sounded exactly like the signature of the US SCR-270 longwave radar. Consolidated PBY Catalina flying boats were also configured as ferrets, and more Japanese radars were soon identified.
Submarine ferrets would prove as effective as their flying brethren, possibly more so because the enemy generally didn't know submarines were around and didn't turn off their emitters. However, no other types of Japanese radars were detected through most of 1943, though there were rumors and bogus "sightings" of other types, such as airborne radars that the Japanese simply didn't have at the time.
Better information began to trickle in towards the end of the year, and in February 1944, following the capture of Kwajalein island, the Americans found documents describing a number of Japanese radars, most interestingly the centimetric Mark II Model 2 shipboard radar. Further landings during the spring and summer revealed more data about Japanese radars, including some sets captured intact.BACK_TO_TOP