Radars of Force Z

[Byron Angel]

Japanese night action successes against radar equipped American ships were mainly in the land return cluttered Slot, not open waters,

It’s hardly that simple. Land return was only one of the problems, and not the major problem often portrayed at that, that caused the deplorable use of radar by the Americans in the Solomons. I think Force Z would have got their rear ends handed to them by the IJN in any night battle, radar or not, open sea or not, at that point in time.

..... My understanding is that land returns in the waters around Guadalcanal were a huge problem for SC used in surface search mode, but posed very much less of a problem for SG.

B

[KevinD]

Kevin, one last thing if you don't mind. Do you know what ships were in Ozawa's and Kurita's forces ? I have :
- Ozawa : Chokai and Sagiri
- Kurita : Sentai 7 (Mogami, Mikuma, Suzuya, Kumano), Kinu (meets at 1907) and Suiraisentai 3 (Sendai + DesDiv 19 : URANAMI, SHIKINAMI, AYANAMI, ISONAMI)
But I am not sure since sources are conflicting, especially for Kurita's screen.
Best,Francis

Francis, I don’t have the above info at hand. A question over on J-Air Naval or TullysPort would get you the answers quick smart though.

Kevin

[Dave Saxton]

..... My understanding is that land returns in the waters around Guadalcanal were a huge problem for SC used in surface search mode, but posed very much less of a problem for SG.

B

That was my understanding as well...until recently. Looking at the accounts more closely there were many instances of the 150cm sets not having problems with land returns, or at least not to the same degree as the other radars at the time. For example, as Kondo's forces began to round Savo it was Washington's CXAM (same as SC but with big antenna) that began to locate the indvidual IJN warships in the shadow of Savo-not the SG. The SG was still blinded by the land returns for several more minutes. At about the same time it was the SC sets on the sacrificial lambs (DDs) that was allowing them to discover it was IJN warships and not shore batteries on Savo that were cutting them to pieces and return fire more effectively-not their FD sets which were foxed by the shadow of Savo. That's not to say that that US 150cm technology was immune to land returns, but it didn't seem to have the problems we have been lead to believe all these years.

Surprizingly many if not most complaints of land return problems involved the SG rather than the SC or the CXAM sets. Lee thought the problem with SG was "side lobes". Indeed "minor lobes" are mentioned as a potential problem when the SG was near land in the technical literature; especially high lands such as cliffs and mountains up to 5,000-6,000 feet above sea level. The SG had a very tiny "cut parabola" antenna (a dish with the tops and bottoms chopped off) with a broad beam on the vertical plane, and with large minor lobes on the vertical plane.

Land returns blinding Blue's SC at Savo don't make sense from a technical stand point. Mikawa's forces passed many miles from Cape Esperance or Savo Is. It was most likely just a simple matter of the IJN warships being out of range at the time. SC's tiny antenna combined with a low to the sea mounting height on a DD was a bad comination in terms of range attainment. Despite plublished figures, the SC from a DD normally had a range of less than 10,000 yards to a surface target. The SC on SLC, mounted high up on the cruiser, detected the IJN forces even before the SGs on Helena and Boise did at Cape Esperance However..

[wadinga]

All,

Firstly, Dave, thanks for posting the very intriguing RAF quote (from Howse?), as you say the more cross-correlated accounts one can accumulate, the closer to the truth we get, avoiding the distortions some are prepared to incorporate. I have not yet found corroboration of this RAF account anywhere, even from sources keen to highlight the unpreparedness of the British for the Japanese assault.

Paddon's account clearly places himself at the centre of radar matters in PoW at the time, and if he is not operating the 281, he is operating the 284 in tracking mode against Bismarck. He describes himself to the Executive officer as the RDF officer. He takes the new Admiral for a tour round the radar installations during Mediterranean service. He says “I” got a new type 273.

He would be remiss indeed in his duties if PoW’s most effective search radar were left unrepaired whilst he was running up a bar bill at the Raffles Hotel. However he says he would have known about it if the Type 273 were unserviceable, and the PRO reference only mentions the failed Type 282s, not that the principal search radar was down. I will certainly keep an eye out for support for one side or the other.

I must correct my earlier post, Type 286 was ASV II (not I) on a ship and far from being ineffectual, Hezlet “Electron and Sea-Power” whilst describing it as “rather primitive” observes it was installed in 97 out of 247 Western Approaches escorts by mid 1941. It detected many U-boats and was directly responsible for terminating the life and career of Joachim Schepke in U-100, amongst others. It is quite likely Electra and Express were equipped with type 286 even if Vampire and Tenedos weren’t.

This business of Type 284 having to be shut off and warmed up again is all very well in principle, but Suffolk would have been incapable of tracking Bismarck for as long as she did in such terrible conditions, if it was not possible to over-ride this practice. Admiral Tom Phillips has received overwhelming criticism over his actions since Force Z’s destruction but to suggest his force was touring the South China Sea at night, when a large and powerfully protected invasion force was nearby, but with all its radars unserviceable or switched off is a put-down too far.

I realise we should not be diverted to US practice in this thread, but an obsession with US radar failings at Savo may tend to colour opinions about subsequent actions too much. At Cape Esperance Helena tracked the approaching Japanese from 27,000 yds down to about 8,000 without reporting to the flagship, who had committed the squadron to a confusing about turn in succession, little knowing the enemy was bearing down at 20 knots. Muddle from the turn, chronically poor TBS discipline and confusion as to where the lost van destroyers were, negated most of the advantage of a pure-luck perfect T crossing occurring before those fantastic night fighting Japanese were even aware US ships were about. At the ensuing battle of Guadalcanal, Helena did it again at a similar range but reported straight away. Further muddled communications and tactics resulted in the US force being even more unprepared and instead of crossing the T ran headlong into the Japanese force. In both cases American radar gave a considerable advantage over Japanese eyeballs but poor TBS discipline resulted in US chaos. Sorry enough off-thread.

All the best

wadinga

[Dave Saxton]

I don't know if the 273 was functional, partly functional, or non-functional, but I would error on the side or partly or non functional for now, because that is the nature of WWII era vacuum tube electronics, especially given the harsh enviroment at sea.

By calling 286 almost useless I'm actually paraphrasing and quoting Sir John Tovey. However, the 286P on Repulse being mounted at 30 meters above the surface of the sea may have been of greater usefulness.

Much of the American issues with radar in 1942 stemmed from inexperience, not neccessarly technical short comings. Radar was brand new in the USN and we lacked the experience gained by the Germans and the British by then. Commanders tended to either want to disregard it as unreliable, or to place too much faith into it, rather than take a more balanced approach that hard experience would eventually teach. Savo Island was in large part a combination of placing undue faith in radar and not understanding the true capabilities and limitations of the new technology. I have to wonder if Phillips with limited experience would have done any better than Scott and Calahgan, or Wright....... or Bey? Would it just be Tassafaronga a year ealier?

Padden highlights Howse's book with some wonderful human insights that give us some insight into the culture of how radar was progressing during middle war years. One storey and I'm only paraphrasing here based on memory:

Once while at Alexandria the Admiral called Padden up to the bridge and asked him to compose a message about the state of the fleet's radars to send to London. After Padden had done so, he gave it to the Admiral to see if that was want he really wanted. The Admiral exclaimed: "Hell No! Tell em the the 281 is useless with out an interrogator and the 284 and 285s fail every time a gun fired. Admiralty Please help! Type 273 is the only radar I have when it's working!" Padden said they never received a reply. LOL

[Byron Angel]

..... My understanding is that land returns in the waters around Guadalcanal were a huge problem for SC used in surface search mode, but posed very much less of a problem for SG.

B

That was my understanding as well...until recently. Looking at the accounts more closely there were many instances of the 150cm sets not having problems with land returns, or at least not to the same degree as the other radars at the time. For example, as Kondo's forces began to round Savo it was Washington's CXAM (same as SC but with big antenna) that began to locate the indvidual IJN warships in the shadow of Savo-not the SG. The SG was still blinded by the land returns for several more minutes. At about the same time it was the SC sets on the sacrificial lambs (DDs) that was allowing them to discover it was IJN warships and not shore batteries on Savo that were cutting them to pieces and return fire more effectively-not their FD sets which were foxed by the shadow of Savo. That's not to say that that US 150cm technology was immune to land returns, but it didn't seem to have the problems we have been lead to believe all these years.

Surprizingly many if not most complaints of land return problems involved the SG rather than the SC or the CXAM sets. Lee thought the problem with SG was "side lobes". Indeed "minor lobes" are mentioned as a potential problem when the SG was near land in the technical literature; especially high lands such as cliffs and mountains up to 5,000-6,000 feet above sea level. The SG had a very tiny "cut parabola" antenna (a dish with the tops and bottoms chopped off) with a broad beam on the vertical plane, and with large minor lobes on the vertical plane.

Land returns blinding Blue's SC at Savo don't make sense from a technical stand point. Mikawa's forces passed many miles from Cape Esperance or Savo Is. It was most likely just a simple matter of the IJN warships being out of range at the time. SC's tiny antenna combined with a low to the sea mounting height on a DD was a bad comination in terms of range attainment. Despite plublished figures, the SC from a DD normally had a range of less than 10,000 yards to a surface target. The SC on SLC, mounted high up on the cruiser, detected the IJN forces even before the SGs on Helena and Boise did at Cape Esperance However..

..... Dave, your above comments appear to contradict not only the wartime SG operational manuals I have read, but a pretty detailed description of the limitations of SC performance in surface search mode that was provided to me by a senior (now retired) US Department of Defense radar specialist and operational analyst.

I did a quick check into USS WASHINGTON's after action report (BB56/A16-3/0155: Action Report, Night of November 14-15, 1942) and turned up the following comments-

Page 21 - Radar Search
"Radar searching was mainly accomplished by the SG radar."
"The CXAM was not particularly effective because innumerable land echoes were present from land over 100 thousand yards away and there were also many confusing side lobes from nearby land and from ships."

Page 29 - Lessons Learned and Recommendations
"The SG radar was invaluable."
"The CXAM was used on stern and quarter bearings but could not be depended upon due to land interference."

Very puzzling.

B

[Dave Saxton]

It's not that puzzling, at least not to me. Those comments are in the rather general, overview type of context, while the examples I pointed out are specific. They are in the action reports. It's true that SG was the primary surace search radar and the 150cm sets did have those problems. But the general rule didn't always translate to specific cases on a consistent basis. The fact is all the radars had problems with minor lobes and land returns, and I would not disagree that they were great with the 150cm designs, but the SG was not immune either. The USN manual I read does point out the problem of minor lobes and high over hanging land masses with SG. Furthermore all the radars including the SG had problems with land returns from distant land masses of 100,000 yards and greater.

because innumerable land echoes were present from land over 100 thousand yards away

The problem of distant land returns is a function of PRF and it doesn't matter the operating frequency, the antenna design, or the instrumentation and so forth. SG had an unambiguous range of 81 miles (150,000 meters) due to its PRF. What happens is that through abnormal propagation an echo returns to the receiver from a distant radar contact after it's associated pulse train has passed, and is now confused with echoes from the present pulse train on the elapsed time base. Therefore, it will show on the display as a contact at say 28,000 yards or 18,000 yards, or whatever distance, when it is really 100,000+ yards away. The Washington records make note of this strange happening with all the radar sets (including SG) when passing by Islands at least 100,000 yards away as they approached the Guadalcanal area. Distant Santa Ysalbel in paticular appears to have been causing problems that night. Washington firing 42 rounds at a imaginary battleship at 18,500 yards was the result. This phenomena was the cause of the Battle of the Pips in northern waters as well.

[Byron Angel]

It's not that puzzling, at least not to me. Those comments are in the rather general, overview type of context, while the examples I pointed out are specific. They are in the action reports. It's true that SG was the primary surace search radar and the 150cm sets did have those problems. But the general rule didn't always translate to specific cases on a consistent basis. The fact is all the radars had problems with minor lobes and land returns, and I would not disagree that they were great with the 150cm designs, but the SG was not immune either. The USN manual I read does point out the problem of minor lobes and high over hanging land masses with SG. Furthermore all the radars including the SG had problems with land returns from distant land masses of 100,000 yards and greater.

because innumerable land echoes were present from land over 100 thousand yards away

The problem of distant land returns is a function of PRF and it doesn't matter the operating frequency, the antenna design, or the instrumentation and so forth. SG had an unambiguous range of 81 miles (150,000 meters) due to its PRF. What happens is that through abnormal propagation an echo returns to the receiver from a distant radar contact after it's associated pulse train has passed, and is now confused with echoes from the present pulse train on the elapsed time base. Therefore, it will show on the display as a contact at say 28,000 yards or 18,000 yards, or whatever distance, when it is really 100,000+ yards away. The Washington records make note of this strange happening with all the radar sets (including SG) when passing by Islands at least 100,000 yards away as they approached the Guadalcanal area. Distant Santa Ysalbel in paticular appears to have been causing problems that night. Washington firing 42 rounds at a imaginary battleship at 18,500 yards was the result. This phenomena was the cause of the Battle of the Pips in northern waters as well.

..... Dave, with all due respect, those comments you consider to be general overview remarks and therefore apparently unimportant were inserted into the radar search section and lessons learned section of the action report for a reason. These comments were penned by the personnel who were physically operating the systems and they make an undeniably clear distinction between the surface search performance of SG (invaluable, despite some issues with land returns) versus CXAM (undependable due to dramatic problems with land returns). The only reasonable inference to be drawn is that SG was the much better performer. I cannot understand how you can argue that SC (in surface search) was somehow equal in overall performance to SG.

The action reports from Tassafaronga make the case even more dramatically. Five US ships lacked SG and were reliant upon their SC/CXAM for surface search. None ofthe five were able to contribute to the action because the land returns from Guadalcanal, which lay behind the Japanese, effectively prevented their SC search radars from acquiring and tracking the enemy. SG could pick up ships against the Guadalcanal background without any real difficulty, and the ships so equipped carried the fight that night.

B

[Dave Saxton]

Byron,

I think your I understand the problem:

I cannot understand how you can argue that SC (in surface search) was somehow equal in overall performance to SG.

That is not what I'm trying to say at all. It's understood that CXAM /SC was inferior to SG for surface search and I stand behind that understanding. What I'm pointing out is that CXAM/SK/SC wasn't as poor as is traditionally understood. And in some cases it performed much better than one would expect.

None ofthe five were able to contribute to the action because the land returns from Guadalcanal, which lay behind the Japanese, effectively prevented their SC search radars from acquiring and tracking the enemy. SG could pick up ships against the Guadalcanal background without any real difficulty, and the ships so equipped carried the fight that night.

The issue here was resolution for range. SG had a resolution for range of 500 yards on the PPI and 300 yards on the its A -scope. SC had a resolution for range of of, IIRC, 1200 yards. The 40cm gunnery sets had a resolution for range of 400 yards. Allow me to disscuss some of scientific principles and practical results of this performance parameter.

Resolution for range determines how far apart for distance two seperate objects need to be from each other to registar as two seperate pips on the indicators. With most conventional pulse radar designs this is governed by the pulse duration. A pulse lasting a duration of millionth of a second covers a distance of 300 meters. However, the leading edge of the pulse rebounds off the target before the trailing edge arrives so the resolution for range will be 1/2 the pulse distance. Hence, SG a conventional pulse radar design, with a resolution for range of 300 yards has a pulse duration of ~ 2 microseconds (the PPI resolution is not as fine because Plan displays are not as accurate or as sharp as A-scopes in their presentation). In practical terms it means that if the IJN warships were farther than 500 yards from the land they could be presented as seperate targets on the presentation, but if they were within 1200 yards of land; SC could not. In the cases of CXAM/ SC determining targets against a land background and SG not, it is a case of the warships not being within 1200 yards of the land, but the CXAM not having the same problems with minor lobes (giving land returns) that SG was having at that moment in time.

[Byron Angel]

The issue here was resolution for range. SG had a resolution for range of 500 yards on the PPI and 300 yards on the its A -scope. SC had a resolution for range of of, IIRC, 1200 yards. The 40cm gunnery sets had a resolution for range of 400 yards. Allow me to disscuss some of scientific principles and practical results of this performance parameter.

Resolution for range determines how far apart for distance two seperate objects need to be from each other to registar as two seperate pips on the indicators. With most conventional pulse radar designs this is governed by the pulse duration. A pulse lasting a duration of millionth of a second covers a distance of 300 meters. However, the leading edge of the pulse rebounds off the target before the trailing edge arrives so the resolution for range will be 1/2 the pulse distance. Hence, SG a conventional pulse radar design, with a resolution for range of 300 yards has a pulse duration of ~ 2 microseconds (the PPI resolution is not as fine because Plan displays are not as accurate or as sharp as A-scopes in their presentation). In practical terms it means that if the IJN warships were farther than 500 yards from the land they could be presented as seperate targets on the presentation, but if they were within 1200 yards of land; SC could not. In the cases of CXAM/ SC determining targets against a land background and SG not, it is a case of the warships not being within 1200 yards of the land, but the CXAM not having the same problems with minor lobes (giving land returns) that SG was having at that moment in time.

..... Dave,

As it was described to me, the biggest problem of SC surface search operation in the Guadalcanal tactical arena was land returns from side lobing - that land masses within 30deg of the beam azimuth would produce land returns on the scope.

One of the other observations made by the same gentleman, and something which can sometimes become lost amid the technological discussion, was that operator competency was a very large component of the overall efficiency of a radar system and there were not a lot of highly skilled radar operators in the USN in 1942.

B

[Dave Saxton]

As it was described to me, the biggest problem of SC surface search operation in the Guadalcanal tactical arena was land returns from side lobing - that land masses within 30deg of the beam azimuth would produce land returns on the scope.

Going back to the original example from the action report that I alluded to, Washington's CXAM was locating IJN warships as they rounded Salvo before the SG could. This doesn't indicate a side lobe problem in this case, but I think the key here is that it was CXAM, not SC. They had two very different antennas. The antenna design and the resulting antenna directional gain determines the side lobe properties as well as the bearing resolution.

In the specific cases of the DD's using their SC's this occurred while the DDs were more or less south of Savo, so they would have no nearby land masses to left or right of the antenna's likely aiming direction. Had that not been the cases, then the SC's side lobe problem probably would have surfaced.

The minor lobes which could give trouble with SG were mainly on the vertical plane rather than the horizontal.

One of the other observations made by the same gentleman, and something which can sometimes become lost amid the technological discussion, was that operator competency was a very large component of the overall efficiency of a radar system and there were not a lot of highly skilled radar operators in the USN in 1942.

Certainly true depending, on degree, on the radar being used. A big potential advantage to SG on this point could be the PPI indication. However, this potential advantage didn't translate in 1942 because of how SG was installed and used in 1942/43. The only PPI provided at that time was in the SG operator's lighted station. The ship's command on the darkened command bridge could not look at the presentation directly, nor could the plot directly in many cases, or the firecontrol centers below decks, or the back up command centers, or the conn. The commands had to depend on the narrative or description provided by the operator or the talker over the phone circuits. It was no different from calling off ranges and bearings from off of an A-scope presentation in practice. The admiral or captain had to depend on the descriptive skills of perhaps a 19 year old kid, while forming and holding a picture of the evolving battle situation in his head. In the case of Calaghan he was getting it third hand over an overloaded TBS as well.

[Dave Saxton]

Perhaps to come back more to topic we could compare Type 273 to SG. This may help to better understand how it could have aided Force Z in the Far East in 1941-42-assuming it was fully functional and remained that way.

Operating frequency: They both operated at 10cm wave length. The primary advantage of using this wave length for surface search was that it could attain good bearing resolution without an extremely large antenna. It was also more likely to provide sufficient antenna gain with out needing to resort to a large antenna as well. Greater antenna gain usually helps provide sufficient range attainment while better suppressing side lobes. Further advantages over metric wave length was that ocean waves were unlikely to attain a height that was 1/2 wave length of the operating wave length causing sea clutter. This is a major problem with meters wave length radars at sea, such as Type 286. Drawbacks of 10cm wave length radars are that they return more noise overall from background clutter compared to decimetric wavelengths, they pickup aircraft poorly, and they are generally unsuitable for phase coherent required operational features.

Antenna design: SG used a single cut parabola for both send and receive. Cut parabola means it is a reflective dish with the tops and bottoms cut off. This reduces the antenna gain on the vertical plane resulting in a large fan shaped beam on the vertical plane (but retaining a narrow horizontal beam). Thus as the ship bobs and rolls on the sea way and the radar beam bobs and rolls with it; a narrow vertical beam could be pointing the beam up into the sky and then down into the sea, but with a wide vertical beam it doesn't matter. It didn’t require a stabilized mounting and can basically be mounted anywhere that was handy. However, in 1942 due to the need to keep the transmission lines as short as possible, it had to be located near the main radar office. This is why the SG antenna was located on the front face of the foretop tower on USN fast battleships in 1942 giving it 60* blind spot aft.

Type 273 differed from Type 271 by using two (uncut) parabolas inside the covering called the “lantern” instead of the “cheese antenna”. One dish was for send and the other for receive. These dishes were each 9 wave lengths in diameter giving a very narrow beam width and superior bearing resolution, and with greater antenna gain. The greater antenna gain improved range attainment to surface targets, and better suppressed side lobes. The vertical beam width was just as narrow as the horizontal in this case. This made it useless for aircraft detection and tracking in most cases. Of course this extremely small pencil beam meant that the antenna mounting had to be fully stabilized even on a large warship, or the beam would be dancing around pointing everywhere and rending the radar useless in a sea way. The resulting antenna mounting and the lantern were relatively large and heavy requiring a robust platform to mount the lantern on. To keep the wave guides short, the lantern platform had to be close by to the radar office containing the transmitter. On a cruiser it was usually right in front of the forward superstructure, or aft, giving the radar a large blind spot. Compared to SG, 273 had superior bearing resolution and was less prone to minor lobe problems. However, Post Receiving Integration was much more difficult, especially prior to fitting 273 with a PPI indicator prior to late 1943.

Indication: The very narrow beam of Type 273 meant that the radar operator had better be paying close attention to his A-scope with the antenna in continuous scan mode or the beam could pass by a target without him noticing it is even there. The range may be very close indeed before he eventually notices the pip. (As far as I know there were no modern automatic detection and labeling circuits in those days).

SG with its PPI (with a momentary afterglow on the scope after the beam had past), and with the wider 5* horizontal beam width, it was less likely that the operator would miss a pip given a moderate scan speed. 5* beam width/bearing resolution is about right for acquiring targets with moderate scan speeds in continuous scan mode.


With extremely narrow beam widths compensation can be made by, very slow scan speeds, longer after glow on the indictor, and/or more pulses per second. The PRF of 273/271 was 3000, giving 273 an unambiguous range of only 52km. (another way to deal with this problem is very high scan speeds such as on FuMO81) I’d imagine that in a restricted battle arena such as the slot or Iron Bottom Sound it would be an absolute nightmare with phantom targets and land returns from more distant islands.

Type 273Q: In the fall of 1943 an improved 273 was made available to RN major units called 273Q. It featured a single, long afterglow, PPI indicator, more power, and other modifications. This was what Duke of York (but as I understand it only DoY or perhaps one other) had at North Cape. It could be operated with two separate pulse durations and range scales. The shorter pulse duration improved resolution for range when going to a fine indication mode.

[lwd]

.... As it was described to me, the biggest problem of SC surface search operation in the Guadalcanal tactical arena was land returns from side lobing - that land masses within 30deg of the beam azimuth would produce land returns on the scope. ....

Neptune's Inferno at least implies that one of the big (if not the biggest) problems with US radar in the Solomons was the lack of knowledge and skill on the part of some of the US command staff. Even when the radars were working quite well the info wasn't always passed to those in charge and when it was they often neglected it.