Page 2 of 4

Re: World War II Japanese radar

Posted: Fri Oct 02, 2009 6:23 pm
This is from "The History of US Electronic Warfare, Volume 1" by Alfred Price


Sources: US official report ‘A Short Survey of Japanese Radar” by the Operations Analysis Section and the Air Technical Intelligence Group of the Far East Air Force, dated November 20, 1945. (Ref. AOC Doc. 32). Declassified February 1962. Royal Air Force Signal Intelligence Report “Japanese Radar Equipment,” November 30, 1944.

Note: Only the more important radars are described in the account which follows.

The development of radar in Japan dates from 1936, when Professor K. Okabe at the University of Osaka began work on an electronic method of detecting aircraft. Working under the famous Dr. Yagi (who gave his name to the Yagi antenna), Okabe concentrated on the development of a bistatic system with a separated transmitter and receiver, to detect the interference to radio signals caused when an aircraft passed between the two. The method of operation was, therefore, the same as that tried at the US Naval Research Laboratory in the 1920s and early 1930s, but which had already been overtaken by pulsed systems. Unlike the work in the US, however, the Japanese were to continue the development of the interference detector in parallel with that of pulsed radars. To differentiate between the two systems, the Japanese referred to the interference detector as the Type A equipment and the pulsed radar as the Type B. Okabe saw the great advantage of the Type A equipment to be that, compared with a pulsed system, for a given transmitted power, far greater detection ranges could be achieved. Since low transmitter power was to dog the Japanese radar program throughout the conflict, this was an important consideration. The great drawback of the interference detector was that it did not give the position of the aircraft along the line between the transmitter and the receiver, and despite much hard work, the Japanese team was never able to resolve this fundamental problem. After a lot of experimentation, the Type A system went into production in 1940, operating in the 40 to 80 MHz band and using transmitters developing between 3 and 400 watts. During 1941, the Type A entered service with the Japanese Army and was deployed in quite large numbers. In all, about a hundred of these equipments were built and they remained in service until the end of the war. The longest Type A line of detection was from Formosa to Shanghai, a distance of more than 400 miles.

Japanese work on the development of pulsed radar began in 1941, with completely separate programs by the Army and the Navy, which resulted in a considerable waste of effort to produce different radars of similar (and, compared with those of the Allies), low performance. Because of their separate lines of development, the Army and Navy programs will be described separately in this account.

The Japanese Army radars followed a fairly logical designation system. Each equipment had a type number preceded by Tachi, meaning land-based (“chi” from tsuchi meaning “earth”), Tase meaning ship-borne (“se” from misui meaning “water”), or Taki meaning air-borne (“ki” from kuki meaning “air”). In each case, the type designation was prefixed “Ta” from “Tama Institute,” the Army’s research institute near Tokyo.

The first Army radar to go into production was the Tachi-6 static early warning set which operated on frequencies in the 68 to 80 MHz band. This unusual radar employed a transmitter with an omnidirectional or wide-angle antenna, and three or four separate receivers each with rotatable directional antennas to search for echoes from targerts this can be compared in principle, to lobe-on-receive only. a technique reintroduced in more recent times as a means of making tracking radars less vulnerable to electronic jamming. The first Tachi-6s were deployed in 1942, and gave ranges up to 185 miles on high flying aircraft. (Production of Tachi-6 is estimated at 350).

Following the Tachi-6, came the only slightly less cumbersome (18 ton) but transportable Tachi-7 for the same purpose, operating on frequencies around 100 MHz and with a maximum range similar to that of the earlier set. About 60 Tachi-7s were built and from October 1943, these were deployed throughout the home islands and occupied territories. For service in the combat zone, a yet lighter set was needed, and early in 1944, the 4-ton Tachi-18 appeared which also operated on frequencies around 100 MHz. (Total production is not known).
None of the early warning radars mentioned above could give indications of altitude, so the Tachi-20, which operated on frequencies in the same band as the Tachi-6, was developed. This set did not go into service until March 1945, and about a dozen were deployed. The Tachi-35, operating on 82 MHz, entered service for the same purpose in May, and only a few were deployed before the war ended.

Japanese Army work on the development of precision radars for searchlight and AA gunlaying received a considerable boost early in 1942, following the capture of the American and British bases at Corregidor and Singapore. At Corregidor, the Japanese captured an intact SCR-268 and a badly damaged early warning radar, probably an SCR-270. On Singapore, they captured badly damaged examples of an early warning set and a gunlaying radar, and also found some useful technical manuals which had not been destroyed.

As a result of these finds, two Japanese searchlight and fire control radars appeared, the Tachi-1 and the Tachi-2, both operating on frequencies around 200 MHz and employing many techniques copied from the SCR-268. The reason for the two different radars was that two companies, Sumitomo (Tachi- 1) and Tokyo Shibaura (Tachi-2) had each received a contract to develop an AA gunlaying radar (a further example of the fragmentation of effort which characterized the Japanese radar program at this time). Neither radar was successful, however, and in total only 65 were built. Sumitomo then switched production to the Tachi-3 (72 to 84 MHz) based on the British Gunlaying Mark II radar. About 150 of these were built, and it became numerically the most important Japanese set in this category.

The Tachi-4 was a mobile set intended to replace the Tachi-2 and operating on the same frequencies, but it was unsuccessful and saw little use. However, a development of the Tachi-4, the Tachi-3 1, proved much more successful. About 70 had been built by the end of the war. It had become the most important precision radar operating in the 200 MHz band.

From the beginning of 1944, the Tama Institute worked to produce a Japanese Army version of the German Wuerzburg radar, one of which had been delivered to Japan by submarine. Before it could go into mass-production, however, it was decided to re-engineer the set to Japanese specifications. As a result of this and the severe disruption in production caused by the B-29 attacks, the Tachi-24, as the set became known, was still in the prototype stage when the war ended.

The only Japanese Army airborne radar to go into mass production was the 150 MHz Taki-1 ship- search equipment, installed on maritime patrol aircraft and torpedo bombers. This radar first saw operational use in the fall of 1944.

As has been said, the development of radar for the Japanese Navy proceeded entirely separately from that for the Army. There was even a security barrier between the two programs to maintain this separation. Thus, the Nihon Musen company, one of the three principal Japanese concerns producing radar during the war, had to divide its main plant at Mataka near Tokyo into two when building equipment for the two services. There was even a ban preventing company engineers working on the different contracts from exchanging information.

The Japanese Navy system of designating radars was quite different from that used by the Army. Sets were categorized by purpose. Thus, the Mark I radars were all ground early warning sets; the Mark us were ship-borne sets, the Mark IVs were ground precision radars for searchlight and/or AA gunnery control and the Mark VIs were airborne ship-search equipments. A notable departure from this system was the Gyoku-3 designation for the Navy airborne interception radar.

The main research and development agency for Navy radars was the Second Naval Technical Institute near Tokyo. The first Navy set to go into production was the Mark I Model 1 early warning radar which operated on frequencies around 100 MHz. The first of these to become operational was installed at Rabaul in the spring of 1942, and altogether about 80 were delivered. The next set in this category to enter production was the Mark I Model 2, a mobile radar operating on frequencies around 200 MHz; about 300 of these were built. This set in its turn was followed into production in 1943, by the lightweight Mark I Model 3 operating on frequencies in the 147 to 165 MHz band. About 1,500 were built and it saw widespread use. The reader will note the many similarities between this family of early warning radars, and the quite separate family developed for the same purpose by the Army.
The first Japanese Navy shipborne radar, the Mark II Model 1 operating on frequencies around 200 MHz, began sea tests on the battleship Ise in March 1942. A few months later, the radically different Mark II Model 2 appeared, operating on frequencies in the 3,000 MHz band and using a 2 kW magnetron as power source (it is interesting to note that at the beginning of 1942, Japanese work on microwave radar was only a few months behind that in the USA, though the gap increased rapidly during the course of the war). Several hundred of these microwave radars were built during the conflict, and versions were installed on submarines and ships of all sizes. The Navy re-engineered its own version of the German Wuerzburg and re-designated it the Mark II Model 3. As in the case of the Army version, however, this radar was only in the prototype stage when the war ended. The Mark II Model 4 was a lightweight set operating on frequencies around 150 MHz and installed on small ships and submarines. In most cases, these ship-borne radars were intended to provide air and surface warning, with antiaircraft and surface gunnery control as an auxiliary function. This requirement for radars to perform conflicting functions was beyond the capability of Japanese technology at the time, and resulted in sets whose performance was mediocre in most respects.

For the control of the searchlights and antiaircraft guns defending its shore bases, the Japanese Navy developed its own version of the SCR-268, the Mark IV Model 1, which also operated on frequencies around 200 MHz. The first of these went into operation at Rabaul in November 1943, and in all 80 were built. Following this set in production, was the Mark IV Model 2, with several improvements to make it easier to build and maintain. The Mark IV Model 3 was a direct copy of the Army’s Tachi- 1 (and a rare example of inter-service cooperation). The Navy found it as unsatisfactory for searchlight control as the Army had for gunnery control, however, and only a few were built.

Towards the end of 1941, the Japanese Navy began work on a lightweight ship-search radar for its patrol aircraft. This resulted in the Mark VI series of equipments operating on frequencies around 150 MHz, which first entered service in 1943. Altogether, some 2,000 examples of this radar were built and the set saw wide scale use.

When the war ended, the Navy had under development an airborne intercept radar for night fighters: the 150 MHz Gyoku-3. Although tests had begun, the radar was too late to see action.

The picture of Japanese radar development throughout and up to the end of World War II is one of piecemeal development, and of sets, which in most cases, brought little improvement in capability over their predecessors. The need to run almost completely separate and parallel radar development programs for the Army and the Navy reduced the effectiveness of the nation’s limited radar research effort, and individual services even placed contracts with different companies to produce radars to do the same job. The outcome was that the small Japanese electronics industry found itself required to turn out relatively small production runs of many different types of radar. It proved quite impossible to keep pace with the quantity and quality of Allied radar development and production. The large scale use of countermeasures against the Japanese radars did not begin until April 1945, and in the four months between then and the end of the war, there was no time to react effectively. The impact of US jamming on the Japanese air defense system is covered in Appendix H.


Note: Because most of these radars ran to numerous models and sub-types, the technical details given below should be regarded as representative for the type only.


Type A Bi-static Doppler Interference Detector.
Quantity built about 100 Power — 3, 10, 100 and 400 watt versions.
First used 1941
Maximum range up to 440 miles
Frequencies 40-80 MHz
Note: although strictly speaking this equipment was not a radar, it has been included in this list for

Tachi-1 Ground Searchlight and AA Fire Control Radar
Quantity built 30 Peak power 5 kW
First used 1943 Pulse Length (approx.) 2 microsecs
Maximum range about 12 miles PRF (approx.) 1,000
Frequencies used: around 200 MHz

Tachi-2 Ground Searchlight and AA Fire Control Radar
Quantity built 35 Peak power 10 kW
First used 1943 Pulse Length 2 microsecs
Maximum range about 25 miles PRF 1,000
Frequencies used: around 200 MHz

Tachi-3 Ground Searchlight and AA Fire Control Radar
Quantity built about 150 Peak power 50 kW
First used 1944 Pulse Length 1, 2 microsecs
Maximum range about 25 miles PRF 1, 2 thousand
Frequencies used: 72 to 84 MHz

Tachi-6 Static Early Warning Radar
Employed omni-directional or wide-angle transmitter antenna, and up to four separate directional and movable receiver antennas.
Quantity built 350 Peak power 10-50 kW
First used 1942 Pulse Length 25-35 microsecs
Maximum range 185 miles PRF 500 or 1,000
Frequencies used: 68 to 80 MHz

Tachi-7 Transportable Early Warning Radar
Quantity built about 60 Peak power 50 kW
First used 1943
Maximum range 185 miles
Frequencies used: around 100 MHz

Tachi-18 Mobile Early Warning Radar
Quantity built 400 Peak power 50 kW
First used 1944
Maximum range 185 miles
Frequencies used: 94 to 106 MHz

Tachi-31 Ground Searchlight and AA Fire Control Radar
Quantity built 70 Peak power 50 kW
First used 1945
Maximum range 25 miles
Frequencies used: 187 to 214 MHz

Taki-1 Airborne Ship-search Radar
Quantity built about 1,000 Peak power 10 kW
First used 1944
Maximum range about 60 miles against a large ship
Frequencies used: around 150 MHz.


Mark I Model 1 Static Ground Early Warning Radar
Quantity built about 80 Peak power 5 kW
First used 1942 Pulse Length 10-30 microsecs
Maximum range about 90 miles PRF 530-1,250
Frequencies used: 92 to 108 MHz

Mark I Model 2 Transportable Ground Early Warning Radar
Quantity built about 300 Peak power 5 kW
First used 1942 Pulse Length 3-20 microsecs
Maximum range about 90 miles PRF 750-1,500
Frequencies used: 187 to 214 MHz

Mark I Model 3 Portable Ground Early Warning Radar
Quantity built about 1,500 Peak power 10 kW
First used 1943 Pulse Length 3-12 microsecs
Maximum range about 90 miles PRF 400-600
Frequencies used: 146 to 165 MHz

Mark II Model 1 Shipborne Air and Surface Search Radar
First used 1942 Peak power 5 kW
Maximum range 90 miles for aircraft,18 miles against large ship Pulse Length 3-20 microsecs
PRF 500-1,100
Frequencies used: 185 to 210 MHz

Mark II Model 2 Shipborne Surface Search and Fire Control Radar
Quantity Built 400 Peak power 2 kW
First used 1942 Pulse Length 2-10 microsecs
Maximum range 22 miles against large ship PRF 2,500
Frequencies used: 2,857 to 3.125 MHz

Mark IV Model 1 Ground Searchlight and AA Fire Control Radar
Quantity built 80 Peak power 30 kW
First used 1943 Pulse Length 3 microsecs
Maximum range 30 miles PRF 2,000
Frequencies used: around 200 MHz

Mark IV Model 2 Ground Searchlight and AA Fire Control Radar
Improved version of the Mark IV Model 1
First used 1944 Peak power 30 kW
Maximum range 30 miles Pulse Length 3 microsecs
Frequencies used: around 200 MHz PRF 1,000

Mark VI Airborne Ship-search Radar
Quantity built more than 2,000 Peak power 3-6 kW
First used 1943 Pulse Length 3-10 microsecs
Maximum range 43 miles against a PRF 700-1,200
large ship
Frequencies used: 140 to 160 MHz

Re: World War II Japanese radar

Posted: Fri Oct 02, 2009 6:31 pm
Also from "The History of US Electronic Warfare, Volume 1" by Alfred Price - the Allied reaction.


Note: This Appendix is extracted from the post-war report “American Radar Countermeasures VS. Japanese Flak & EW Radar,” prepared by the Air Technical Intelligence Group and dated December 10, 1945. (Ref. AOC Doc. 32). Declassified 1962.

AA Procedures — Army

Japan was divided into four main Army areas for protection by AA Divisions; the Tokyo area, Nagoya area, Osaka area and the northern part of the island of Kyushu. Each Area Hq seems to have handled all air warning information that affected its own area of control. There was very little liaison between the ground and air forces in such matters as GCI or SLC (ground controlled interception or searchlight control radars). The only way they could keep track of their own fighters was by constant plane to ground radio chatter, and the various Flak batteries were notified from time to time of the movements of friendly planes by telephone.

The coastal areas around the main cities and military target areas of Japan were protected both by EW (Early Warning) stations and by a “radio fence” using the Doppler system. . . Early warning radars were never used in the actual “setting on” of Flak radars,but in the last months of the war some thought seems to have been given to this as additional equipment were dreamed up (Tachi-20 and Tachi-35) to measure elevation angle as well as range and azimuth. The only real use they got out of the additional equipment, however, was to use it as an aid to vectoring fighters for intercept purposes.

According to weather conditions, cloud cover, time of day etc., tracking was done by radar or a combination of radar and optical means. When possible, radar was used for obtaining range, and optical means for determining azimuth and elevation (sonic devices seem to have been used in isolated cases, without very much success). Several officers who were interrogated, admitted that on different occasions the operators had difficulty in determining if both the radar and optical devices were tracking the same target.
An effort was made to determine what in percentage of the air raids radar and/or optical firing data was used, but no statistics were available—same old story “all records destroyed,” and anyway the Japanese made little effort to keep track of such information. Records of this sort would have been valuable to us in determining the over-all effectiveness of our jamming program.

AA Procedures — Navy

In the early part of the war the Navy was primarily responsible for the defense of all islands in Japan’s extensive “outer circle,” but as the war progressed they also took over from the Army part of the responsibility for defending from air raids certain sections in the “home” islands and a good part of Formosa. These naval Flak positions were mainly around naval bases and the larger seaports of Kyushu, Honshu and Formosa.

When the Flak radars and early warning radars were first set up, little consideration was given to any other problem than convenience of installation. Thus, in most cases, it happened that the EW sites might be miles from the Flak radars and all data on an approaching raid had to be telephoned to them. This resulted in considerable delay and confusion and so an overall plan was prepared to set up the EW and GL radars side by side. This plan, however, they never got around to carrying out.

To a great extent the Flak radars depended almost entirely on the EW radars for specific data to “get on” the target. Their range was comparatively short and their searching ability very poor so it was necessary to get warning of an approaching raid as early and as accurately as possible so that the batteries themselves could be prepared. . . Thus, when the EW radars were partially jammed on top of this, the meager information obtained had to be telephoned to the Flak radar. It left the whole system in a state of confusion and unpreparedness. And now, add to that the severe jamming of the Flak radars themselves.

Effectiveness of RCM

Japanese A.J Equipments — Army

This subject can be summed up in one word, “none.”

Japanese AJ Techniques — Army

The Japanese Army had received from Germany the general principals of the “Wuerzlaus” method and had done considerable work experimenting with it. However, the Army operators decided they could read through our “Rope” well enough and so this method was never used in the field.

Against electronic jamming they had practically no means of counteracting its effect. In some cases azimuth was obtained by a crude sort of D/Fing on the source of the jamming signal by using the point of maximum disturbance on the receiver-indicator, and the range was obtained, if possible, by some radar not too badly jammed. Various Japanese personnel questioned stated that it seemed the 200 Mc frequency was more often jammed in the Tokyo area, while the 78 Mc GL frequency would be jammed west of the Tokyo area. No reason can be found for this opinion. Since they had sets using both frequencies scattered through the various Flak defenses, usually some radar would be partially free ofjamming — not that it ever did them much good. (They had no plan or order in the selecting of frequencies for any particular Flak battery. They simply used whatever set was available at the time the battery was placed. Some estimates state that only one out of every eight AA positions was radar controlled.)

When our electronic jamming became serious, they attempted to hasten the research and the solving of production problems they were encountering on the Japanese version of the German Wuerzburg. They felt confident that since its directivity was good, any jamming effectiveness would be greatly minimized except from an aircraft directly within the beam-width of the antenna. The net result was, one Wuerzburg set up for final tests. Also they planned to modify their existing 1.5 meter (200 Mc) equipment so that when jammed they could switch bands either to 1.7 m (176 Mc) or to 1.3 m (230 Mc). This seems to be rather extreme band shifting and no information is available as to how this was to be accomplished. By the end of the war, the Army had only gotten as far with the plan as to send an order through to the labs and manufacturers to start research on the problem. However, very high priority was assigned to the project.

Indications of Effectiveness of Jamming from Japanese Sources

Many Army officers, both technical and operational, were questioned, as well as numerous civilian technicians and radar engineers, concerning the effectiveness of our jamming efforts. In almost every case, the person questioned, in effect, threw up his hands and said, “our antiaircraft firing radars were useless under the ‘wave disturbing’ from your aircraft.” They admitted an almost 100% reduction in effectiveness of their Flak radars. And unless means for obtaining optical data were good, the batteries just did not fire when the radars were jammed. As for Rope and/or Window jamming, the story was different. Most personnel questioned claimed that experienced operators could track aircraft through the clutter but “with difficulty.”

The following statements are direct quotes from statements made by several responsible officers connected with technical research and with the tactical use of radar: Col. K. Sataki, Tama Research Institute (radar research and production), “Highest priority given to anti-jamming studies and to efforts to modify existing radars so that band shifting would be possible.” Lt. Gen. Tada, Chief of Tama Institute, “About 10% of the high frequency research effort in all fields, or about 80% of the research effort on radar projects alone, was diverted to the development of anti-jamming equipments.” Lt. Col. Hiroshi Tominaga, Sig. Sect., Imperial Gen. Staff (was in charge of planning and setting up radar sites in China and some on Honshu), “American jamming was so complete that maximum research effort was expended to develop our radar to combat it. Most ‘search’ experiments (i.e. with intercept receivers), were dropped” (considerable research had been in progress to develop search receivers by which Allied radars might be Ferreted).

Lt. Gen Tada also expressed an opinion that, through air raids,Japan’s capacity for radar production was knocked down to about one third of its former level and its capacity for research to about one half. Still, he believed that six months would have been sufficient to greatly improve Japanese radar in its ability to combat our jamming. The above statements and opinions can be accepted as the sum and substance of the answers received from almost every person questioned.

There seem to be no available figures on the number of rounds of AA ammunition per aircraft kill, using radar controlled batteries either with or without jamming, but operational officers state that jamming reduced the detection or tracking range of B-29s from approximately 18 to 25 miles down to 1 to 3 miles.

Japanese AJ Equipments — Navy

The Navy, like the Army, had had no time to develop anything to combat our jamming.

Japanese AJ Techniques — Navy

They claim that no foreign information was used in any way in their AJ research. And like the Army, they state that Window and/or Rope jamming caused negligible difficulty. Operators were given “on the spot” instruction by officers from the naval radar schools, who travelled around to the various sites and conducted tests on Chaff suspended from balloons. Operators were taught to recognize the difference in “beat” between the Window return and the target return. Also they attempted to discriminate between the wave shape of a Window return and that of a plane. Window jamming began very early compared with electronic jamming and was first carried out by B-29s (more likely B-24s — author) attacking from China bases. The Japanese claim that this gave them lots of practice tracking through Window echoes, especially in Formosa, and that they had considerable success with their Flak radars there.

But with electronic jamming it was a much different story. Like the Army, the Navy very freely admits that its Flak radars were practically useless at any time electronic jamming was done by American aircraft. They claim to have first had trouble in the Kyushu district in “April or May 1945,” and later, in “June 1945,” in the Kanto or Tokyo district.

Their plans to minimize the effect of electronic jamming consisted of two parts:

1. Research was started to modify their existing GL radars so that a frequency shift of as much as 10 Mc either side of the original frequency might be attained (this however was never completed “due to design difficulties,” though some experiments were carried out with EW radar so modified and they claimed these were quite successful.).

2. An attempt was made to modify the Mark II Model 2, 10 cm shipborne to measure elevation angle. Experiments against aircraft indicated a range up to 20km (12 mi) for this set. However, work on this was not completed either, before the end of the war.

The Mark II Model 2 (10 cm) was also experimented with as an SLC radar and one such set was found at Tsukishima, a branch of the 2nd Naval Tech. Institute in Tokyo.

The S-8A (also known as the Mark II Model 3), a 58 cm Flak radar, was felt to be another answer to the jamming problem, and it was given high priority, but again they failed because of little time and loss of production facilities due to air raid damage.

There was no special training given to radar operators because, as in the words of one Japanese officer, “the concrete plan to anti-jamming did not exist.” However, some efforts along that line were made but with doubtful success. Operators were advised to use the “gain” control in attempting to bring the echo out of the noise, and an attempt was made to do “lobe comparison by noise amplitude.” What they did was use the receiver indicator as a rough D/Fing unit, using the point of greatest noise amplitude as a means of determining azimuth and elevation; then to get the range they would use the Mark I Model 3 150 Mc EW set. They admitted this method was very inefficient and inaccurate — “The method to use the enemy’s disturbance waves in reverse as our direction finder, afforded us only to discern the approximate direction of the enemy owing to the bad condition for minuteness (they found that pip matching on the maximum strength of our electronic noise was almost impossible due to the wide side-bands of ourjamming frequency) and shooting was also difficult.”That classic of understatement just quoted was made by Vice Admiral Nawa, chief of the Megura Park branch of the 2nd Naval Technical Institute.

Many of the individuals questions stated flatly that at night or in bad weather “When your aircraft ‘disturbed our locators’ we could not and did not shoot.” This seems to have been especially true in Kyushu and the Tokyo areas. Around the Sasebo Naval base in western Kyushu, and in Formosa, they claimed to have had better success — more practice, as they put it.

Very high priority was given to anti-jamming research but nothing was accomplished in the short time they had from April 1945 until the end of the war. Captain Ideura of the Naval Tech Dept., stated that their only hope was the modification of their shipborne 10 cm job, and if this was susceptible to jamming “then we could not see the way for defense.” And he and other officers and engineers of his department were quite frank in admitting that modification of any radar in the last months of the war was next to impossible. Laboratories and manufacturing plants were being further dispersed; small factories supplying badly needed components were bombed out; and lastly, the airplanes used to conduct tests against their experimental radars “had to take refuge occasionally” from our planes — and to crown it all, they quite often did not have enough gas to fly the planes to conduct the tests to see whether or not they had a radar that would do the work planned for it.

Vice Admiral Nawa, chief of the Mogura Park branch of the 2nd Naval Technical Institute, complained that “to do effective research we must get rid of the B-29s, but to get rid of the B-29s we must have radars not susceptible to jamming” — he left off there, unable to come to a definite conclusion. As Lt. Cdr. Masaki so ably put it, “In brief, we were troubled by American jamming from beginning to end; our Flak radars did not perfectly get out its difficulties.”

After discussing it with numerous officers and engineers of the 2nd Naval Tech Institute, the consensus of opinion was that, by the end of the war, production facilities had been reduced by air raids to about 10% of normal, and electronic research to about 15 to 20% of normal. Since the Navy was not primarily interested in land-based GL radar, only about 10% of its total electronics research effort was turned to anti-jamming. But this 10% represented practically all the “brains” involved with the design and production of this type of radar.

They explain this extreme loss of production and research ability not so much by the actual damage from the air raids, but because of:

1. Shortage of parts (mechanical and electrical).

2. Faulty production (experts and technicians drafted, mental distress due to air raids).

3. Shortage of transportation.

4. Loss of production volume by dispersal of factories.

5. Frequent changes of plans.

Of course, 1, 3 and 5 above are plainly due to air raid damage, and behind their inability to maintain any sort of adequate defense against our air blows, was their almost total lack of countermeasures against our radar and our jamming.

Re: World War II Japanese radar

Posted: Mon Oct 05, 2009 7:52 am
by RF
This is a huge mass of information, which will take quite a while to digest.

Thanks for all the posts, USS ALASKA.

Re: World War II Japanese radar

Posted: Wed Oct 07, 2009 1:14 pm
RF wrote:This is a huge mass of information, which will take quite a while to digest.

Thanks for all the posts, USS ALASKA.
Not a problem sir - hope this was the info you were looking for.

Re: World War II Japanese radar

Posted: Mon Oct 31, 2011 2:30 pm
by marjan
Type 32

Became Operational: September 1944
War Status: under preparation for practical use
Installed: ground shore batteries, heavy cruisers and battleships
Purpose: anti-surface ship gunfire control
Wavelength: 10 cm
Peak Output: 2 kw
Transmitter: magnetron
Receiver: crystal
Detector: n/a
Detected: large surface ships 30 km
Weight: 5000 kg
Number built: 60
Antennae: square horn, send and receive separate us

Re: World War II Japanese radar

Posted: Mon Oct 31, 2011 6:07 pm
by RF
Was this Type 32 used for internal AA defences - given that Japan was under heavy bombing assault in this period?

Re: World War II Japanese radar

Posted: Tue Nov 01, 2011 7:11 pm
by Dave Saxton
The Mark3model2 was an adaption of the Mark2Model2 or Type22 which had been in use since 1942. It incorporated lobe switching for accurate bearing determination were as the Mark2Mod2 did not. The Mark3Mod2 or Type 32 was strictly a surface gunnery control radar and was not suitable for AA work. Brown wrote this concerning the Type3Mod2:
IJN surface fire direction radar, mark3model2, This set used lobe switching by a receiver having two horns below the single transmitter horn on top. By the time it was developed the Japanese Navy was no longer capable of fighting surface actions. The set shown here was mounted for coast artillery but was never used for that either....
The first number in the IJN designation system indicates the mission of the radar:

1 = air warning
2 = surface search
3= surface gunnery
4 = AA gunnery
6 =airborne
Was this Type 32 used for internal AA defences
The Japanese Navy had several different radar sets that could be used for flak direction -on land- but not on warships.

There was the Mark4Mod2 which was a direct copy of the US Army's excellent SCR-268 blind fire AA radar. The Mark4Mod3 (an improved Type 42) and the Mark4Mod4 which was captured on Peleliu. American records indicate that these were all very effective flak radars capable of fully blind flak direction.

The IJA also had the Tachi -4 which was a direct copy of the British SLC (search light control) captured at Singapore and the Toshiba developed Tachi-1 and Tachi-2 search light control sets all of which did not impress the Americans as being very effective.

The Japanese did try to bring into operation copies of German Wuerzburg flak control radar. This was the Tachi-24 but Japanese industry could not produce electron valves of the required quality to make it work. Only one Tachi-24 prototype was fully completed before VJ day. There was also the Tachi-31 which was basically a Tachi-24 but used Mark4Mod2 electronics operating 1.5 meters wave length, pressed into srevice late war in the Tachi -24's stead.

Re: World War II Japanese radar

Posted: Wed Nov 02, 2011 9:28 am
by RF
My impression from the last post is that all the technologies in radar development were straight copies of acquired and captured equipment, copying imported technology was something the Japanese had a renowned reputation for doing both before and particulary after WW2.

Was any of the Japanese radar technology indigenously developed as opposed to being copied?

Re: World War II Japanese radar

Posted: Wed Nov 02, 2011 2:34 pm
by Dave Saxton
Yes, there were several indigenous Japanese radar designs. The Mark2Mod2 (Type22) and it's gun laying derivative Mark3Mod2 (Type32) were wholey original. These radar sets were centimetric and powered by an independantly invented and developed cavity magnetron. The IJN deployed operationally the Type22 10cm sea search radar within months of the American operational deployment of the 10cm SG sea search radar.

However, it would have been quicker to copy such existing flak directing radars as the proven SCR-268 than it would to develop an original equalvilant. Wuerzburg was the best flak directing radar and AA concept in the world during WWII. It made sense to either attempt to directly copy Wuerzburg or to develop an equalvilant employing the concept.

Re: World War II Japanese radar

Posted: Wed Nov 02, 2011 6:26 pm
by RF

Looking at the record of the US firebombing of the Japanese cities, it seems that there is little evidence of this radar technology achieving concrete results for the defence of the home islands. I gather that this bombing by the B29's was almost completely unopposed, as the Japanese had neither the heavy AA guns or planes capable of reaching them, and thus why most of the world had remained unaware of these radar developments.

Thanks for your posts Dave.

Re: World War II Japanese radar

Posted: Wed Nov 02, 2011 7:19 pm
by Dave Saxton
The IJN effort appears to have been more productive than the IJA effort, not that it was productive compared to the Allied or German efforts. However, how responsible was the IJN for the defense of city's and industrial areas? Did they have enough Type 42s to make a difference even for harbors and naval air stations? Naval radar technology like the Type 22 and Type 32 were fairly useless for command and control of night fighters or directing flak against air attack.

The IJA only got one Tachi-24 completed by the war's end, and the ersatz Tachi-31 while based on the Tachi-24 woud not have had the precision of real Wuerzburg.

Even if the Tachi-31's were deployed in sufficient numbers we don't find the kind systemic support to make them effective, that we find in the British, American, and German air defense systems.

In the German system there were long range air warning radars (Mammut, Jagdschloss, Wasserman..) whose data was then coordinated through combat information centers to develop an overall situational awarenss and plan of action. The Himmilbett system divided up the airspace into many relatively small cells. Each cell had at least one Freya which provided an overview of the tactical situation locally, supported by at least two Giant Wuerzburgs. One Giant Wuerzburg tracked the enemy while the other Giant Wuerzburg could track and direct a friendly night fighter. Finally, once the enemy fell within range of the substantial flak batteries, the small Wuerzburgs could take over Flak direction.

The Germans and the Allies also had IFF systems and significant ECM assets at their disposal which the Japanese never did.

Re: World War II Japanese radar

Posted: Sun Nov 20, 2011 2:09 pm
by Siegfried
The 10cm Mark 2 Model 2 Modification 2 was the type with fire control capabillity. It had 3 horn antenas: ostensibly a central one as a transmitter with one either side, presumably splayed out a degree or two that could be used to compare left right signal returns. Allied intelligence assesed the sets as being installed in IJN destroyers.

I have a pdf with a sketch, can't seem to upload on this site but it can be obtained from the technical->radar section of the forum. You have to sign in to upload.

The basic Mark 2 Model 2 (Often called the Type 22) had a seperate transmit and receive horn. Later versions had a single horn achieved by a sort of quater wave transformer effect in the waveguide and the Mod 2 had the 3 waveguides for fire control. Several hundred were built though one immagines the fire control version was in limited supply.

The Japanese actually beat Randall and Boot to the high performance magnetron and even developed stappred version but the moves slower in converting it to radar. The Germans also had multicavity magnetrons of quite reasonable power (I specifically say multicavity magnetron not just split anode) however they lacked the optimal cavities of the Randall and Boot version. One version by the Sanitas company of 1939 could produce 100W continious at about 18cm (probably good for 3kW impulse). The Germans were not rushing to develop microwave radar and the program they had both at Lorentz and Telefunken for 25cm sets used disk triodes (these could be used at about 12kW at 9cm), they liked the flexibillity, while they also obsessed about tunable magnetrons.

However had the Japanese shared their magnetrons with the Germans in by early 1942 when they were fielding their first type 22 radars the Germans might have benefitted enormously. As it was type 22 wasn't practical till the sophistikcated German Rhebok calibration circuit made this radar practical from small destroyers so the japanese were helped more by the Germans even in this.

The interesting think about the early Japanese magnetrons was they were hollow molydenum steel in a glass envelop, water cooled. The sheet could be distorted and this was used to retune the magnetron.

Re: World War II Japanese radar

Posted: Tue Nov 22, 2011 4:51 pm
by Dave Saxton
Siegfried wrote: The Germans also had multicavity magnetrons of quite reasonable power (I specifically say multicavity magnetron not just split anode) however they lacked the optimal cavities of the Randall and Boot version. One version by the Sanitas company of 1939 could produce 100W continious at about 18cm (probably good for 3kW impulse). The Germans were not rushing to develop microwave radar and the program they had both at Lorentz and Telefunken for 25cm sets used disk triodes (these could be used at about 12kW at 9cm), they liked the flexibillity, while they also obsessed about tunable magnetrons.

However had the Japanese shared their magnetrons with the Germans in by early 1942 when they were fielding their first type 22 radars the Germans might have benefitted enormously.
I wrote this on another forum to explain why the Germans didn't consider the the cavity magnetron a high priority and why the cavity magnetron isn't a magical device:
A magnetron is a type of vacuum tube which uses a magnetic field to cause the electrons traveling from anode to cathode to resonate (hopefully) at a given frequency. (A tube called a magnetron was invented in America in the early 1920s but this is actually a triode in function and uses the magnetic field to replace the control grid) The magnetron as for radar was invented in Japan by Kinjiro Okabe in 1927. These were called split anode magnetrons. Then in 1935 Hans Hollman applied for a secret German patent for the first cavity magnetron. This was at the time when Hollman was employed by GEMA as a consultant and when GEMA was preparing the prototype Seetakt powered by a four cavity magnetron for demonstration to the Kriegsmarine.

However, as explained by von Kroge GEMA dropped the magnetron from the design:
The magnetron oscillator generated no great love for itself at GEMA…….an electrical field that results from a potential being applied between anode and filament (cathode) is at right angles to the magnetic field and the values of the two fields are critical in determining if and how oscillations occur. A variation in the strength of one or the other of these two fields causes an undesired change in the frequency, a deficiency that frequently provoked von Willisen and had disrupted a number of experiments…..
The problem here was that the GEMA radar design required that the timing and phasing of the pulses be “coherently” coordinated. Seetakt was not only the first operational radar but it was also the first coherent radar design. Modern American radar engineer Russell Rzemien of John Hopkins University explains why frequency instability is detrimental to a coherent radar functions:

…(frequency) stability in both the transmitter and receiver is the primary prerequisite for effective coherent signal processing…a coherent radar compares the phase or frequency of a target echo with a stable oscillator or reference source….range resolution is obtained (in coherent design) by transmitting a long time duration pulse that is either phase or frequency modulated, and then processing the pulse on reception to produce a narrow range response. This has the advantage of increasing the total amount of energy reaching a target, hence increasing sensitivity, without sacrificing range resolution. It also permits transmitters at lower peak power which is desirable for number of reasons beyond the scope of this introduction…
Commenting on WWII era magnetron powered microwave radar vs triode powered decimetric or UHF radar, Rzemien wrote:

During WWII the advantages and problems associated with high powered microwave radars became apparent. The primary advantages was the ability to build reasonably sized antennas that provided the desired gain, hence increased target detection range, and narrow beam (improved bearing resolution)…(however) the increase in operating frequency and corresponding decrease in wave length increased backscatter from natural objects in the environment…..
Although the magnetron allowed microwave wave lengths, as needed by the Allies, in small relative antenna size applications, the instabilities of magnetrons made it generally unsuitable for some the German radar designs, such as Seetakt and Freya, and anti- chaff modules, which used phase coherent dependant features. Dr Weil of Raytheon mentions that it’s better to not use magnetrons if the following are required:

1) If precise frequency control is needed…
2) If precise frequency agility is required….
3) If the best possible stability is required…..
4) If coherence is required….
5) If coded or shaped pulses are required……
6) If the lowest possible spurious power levels are required…..

Overall, magnetron powered radar wasn’t necessarily better than triode powered radar and it did offer additional draw backs as pointed out by Dr. Weil:

1) Sparking, especially when the magnetron is first started and extended warm up and tuning times..
2) Moding (random frequency jumping)
3) Noise rings (interference with short range target echoes on a PPI display)
4) Spurrious Radio Frequency Output……
5) Radio Frequency Leakage out of the Cathode Stem….
6) Frequency Drift….
7) Frequency pushing…
8) Frequency Pulling…….
9) Short life span…
10) Short Tuner life span…
Intersting that you note that the Japanese discovered strapping. I did not know that.
As it was type 22 wasn't practical till the sophistikcated German Rhebok calibration circuit made this radar practical from small destroyers so the japanese were helped more by the Germans even in this.
This is most interesting. Was this the GEMA messkette or the Telefunken markentiel or both? Brown mentions that a GEMA engineer was sent to Japan, but the Japanese refused to allow him top secret clearance, so he spent his time designing and building test modules. However, the GEMA test modules were closely associated to their messkette system. Coherent signal proccessing technology would certainly have drastically helped the Japanese 10cm radars with their low power and enourmous pulse widths.

Re: World War II Japanese radar

Posted: Sat Nov 26, 2011 8:02 pm
by Siegfried
The note that the Japanese had strapping is made in Louis Brown's "A radar history of WW2 technical and military imperatives".
A good history of the japanese M3 multicavity magentron and type 22 radar is given in the Russel Burns edited Radar Development to 1945: Nakajima, S., "The history of Japanese radar development to 1945," Ch. 18 (pp. 243–258). The inventor was Yoji Ito but his brother, Shigeru Nakajima, also a researcher in the field has written much on this early work. Unfortunatly I am travelling internationally and can't provide much more detail than that. In this article Nakajima makes the point that the Rehbok circuit was important in making it possible to transfer type 22 to smaller ships such as destroyers.

Yoji Ito's initial multicavity magnetrons were called 'mandarin magnetrons' or 'chrysanthenum magnetrons' after the shape of the cavities. Development of a complete theoretical understanding of these lead to a full theoretical understanding and then optimal cavity shape of a hole and a narrow slit.

In Germany this kind of magnetron was called a 'radmagnetron' or 'wheel magnetron'. A look at the Lorentz 8cm Herringer Magnetron (refered to below) shows that the Germans seem to have developed optimal cavity shape as well! It was however a glass tube without water cooling.

"Microwave tube development in Germany from 1920-1945 H. Döring" contains some interesting snippets:

"The most important tube, the multi-cavity magnetron, was already in existence
in Germany in 1937. For example, Telefunken built an eight-cavity magnetron
operating at 1.5cm and capable of delivering 50mW ouput CW. At Sanitas a
water-cooled magnetron was developed to deliver IOOW CW at 25cm. At this
time, such tubes were called 'Radmagnetron' ('Rad' meaning 'wheel' in German). A
tube developed in 1'938 at C. Lorenz AG by F. Herriger for a wavelength of 8cm
is shown in Fig. 18. However the advantages of this particular design for pulse
operation were not fully appreciated in Germany at the time."

"At the Flugfunk-Forschungsinstitut FFO in Grafelfing L. Mayer, an ingenious
tube designer, developed several magnetrons and klystrons. In his work he was
skillfully supported by a technological group under W. Knecht-jointly they
developed, for example, a single-anode magnetron mounted inside a coaxial
transmission line, tunable aver the range 2.6-12cm and capable of delivering an
output power of 50mW-1 W (see Fig. 15). Using a somewhat similar water cooled
tube they obtained an output of 10W at 6cm. In pulse operation a tunable
continuous anode magnetron generated 600W at a wavelength of 3.8cm"

NB Pulse ratings are generally 30-50 times continious ratings.

As you say the Germans were not fond of some of the problems of magnetrons: the Randall and Boot magnetron
did not solve all of these, for instance frequency drift was dealt with by long warm up times and automatically
continiously tunning the receiver of the radar to the wandering transmitter.

They instead focused on microwave triodes which promised to provide them with the flexibillity
sophisticated signal processing they were after.

In these the limitations of conventional tubes are overcome in serveral ways.
1 The transit time of electrons issue was solved by developm grid spacings
of only a few thousands of an inch.
2 The effect of indirect heating of the grid causing secondary emissions was dealth with by
special coatings (gold, silver or nickel)
3 The whole tube was built as a coaxial transmission line.

You mention that Seetakt was the first coherant radar. One example of coherant pulse doppler
radar was the Wurzlaus circuit add on the Wurzburg FLAK radar to overcome windows (foil drops)

In this a stable crystal oscilator was multiplied up to the 560MHz opperating frequency of the
radar and the main output tube locked in using a chain of frequency doublers.

The method was to generate a sinusoidal oscilation and then distort that waveform to generate a second harmoinic. An oscilator tuned to the higher frequency would then be forced to oscilate in phase and frequency with the harmonic. It was a powerfull technique.

This property would incidently have allowed the Germans to increase the power of their radars by using phased arrays of transmitters eg a seperate amplifier for each of the rods of the dielectric rod aerials used on Berlin radar.

One key problem for the Germans was that the half power beam widths of their 50-80cm radars was so broad that the beams intersected to much jamming both in the form of noise and windows.

Their next generation of radars, most likely opperating at 25cm and able to produce several times the power would likely have overcome this.
Tow tubes were the:
The LD12 could produced 100kW pulses at 20cm.
The LD7 could produce 15kW pulses at 9cm and 25kW at 25cm. Variants could even oscilate at 10GHz. (3cm)

These tubes continued to be produced in East Germany and the Soviet Union and formed the basis of many soviet radars into the vietnam war.

Re: World War II Japanese radar

Posted: Sun Nov 27, 2011 9:32 pm
by Dave Saxton
Thank's for the insights.

Regarding 1/2 power beam width. The 1/2 power beam width is product of this equation:

wl(k)/A= HPBW.

Essentially it's the wave length by the aperture (or effective antenna size in wave lengths). The smaller the wave length the easier it is to attain narrow focused beams. However, if too small of antenna is used then advantages of microwaves can be lost. For example, Mark3 Model 2 (and Type22) had a very wide beam width, despite the 10cm wavelength, of 15*, simply because of the relatively tiny antenna horns.

In the Technical Mission to Japan, the Mk3model 2 data is given as:

Wavelength: 10cm
Radiated power: 2kw
Pulse width: 10us
Range BBto BB: 35km
Range accuracy: +/-100 meters
Bearing accuracy with lobe switching:0.5*

No spec is given for resolution for range. Bearing resolution is 15*.

The Germans valued the advantages of coherent function on resolution for range and other performance parameters over using smaller wave lengths in some cases/designs. In many cases the disadvantages of using longer wave lengths could be off set by using a larger, or as large as practical, antenna.

On edit: there were other advantages of slightly longer wave lengths besides avoiding the instabilities of magnetrons when using 1940's technology. One was that the noise factor of the receiving pentodes increases drastically from 400mhz through 1000mhz operating frequency. Thus, for example, by increasing the wavelength of Seetakt from 60cm (500mhz) to 80cm (IIRC 378hz) this noise problem was eliminated and superior signal to noise ratios attained. Additionally, the reliable power ouput of triodes usually increases with increased wavelength.

GEC used grid grounded triodes instead of receiving pentodes to minimize this problem at 50cm (600mhz). Cyrstal receiving componants typically had the same noise factor as pentodes at 50cm regardless of the wavelength.