P-61 Gun-Laying Radar

[wuzak]

Thanks for the history Token.

I wasn't suggesting that the British invented the cavity magnetron, but rather that it advanced the development of airborne radar in the US when it was brought over in 1940.

Are you aware of development of gun-laying radar for aircraft in the US during the war?

EDIT: There were gun-laying radars developed for AA guns, at least.

 

[Token]

I wasn't suggesting that the British invented the cavity magnetron, but rather that it advanced the development of airborne radar in the US when it was brought over in 1940.


Are you aware of development of gun-laying radar for aircraft in the US during the war?

Yeah, no doubt, what the British provided to the US gave the Allies a huge step up in radar in general, not just airborne. Combining the British development with the US's ability to further develop without fear of enemy bombing and over a larger pool of researchers was just the ticket.


If the British had kept these things to themselves they would not have been able to get them into their own wide use as fast, and the US would have been at least a year, if not two, behind in the tech curve. It was a win-win.


Even with the radar this thread is discussing, the SCR-720, the British adopted it themselves, slightly modified, as the AI Mk X.


WW II airborne gun layers is not really an area I have much depth in, but I think there were a few airborne gun layers in WW II for the US, all right at wars end.


Weren't most of the AN/APG designated series during WW II gun layers of one type or another? Some as aids to manual tracking, some as automatic systems, and I know a few were also bomb / torpedo radars. The AN/APG-1 (SCR-580) and the AN/APG-2 (SCR-702) come to mind (or maybe switch those two SCR numbers, see below). I remember both of those as auto tracking systems, similar in function to the ground based SCR-584 (a system I do have firsthand experience with, and no, I am not that old) and tried on several aircraft, including, I think, the P-61. Weren't the AN/APG-3 and APG-15 both 10 GHz auto tracking gun layers developed for the B-29 or B-32? Others, like the AN/APG-5 (SCR-726), were range only radars.


A US system similar to the British turret shown earlier in this thread, and intended for heavy bomber, was designated AN/AGL and was under development. I seem to remember reading that it was completed before V-J day but never was installed in any combat aircraft before the end of hostilities.


By the way, when I went to refresh my memory on some of these I found some real confusion in online sources, and these make me doubt some of my own information. For example, some sources calling the SCR-702/580 on the P-61 the same radar as the SCR-720 (like maybe 702 is 720 typoed), others confusing the correlation between APG-1 and APG-2 (some sources calling APG-1 the SCR-702, others calling it the SCR-580, I don't know which is correct and my statements a couple paragraphs above are to the best of my memory). Since the APG-1 and APG-2 are specifically described in the MIT Rad Lab series, Volume 26, pg 237, as having nutating feeds these cannot be the SCR-520 and -720 mistyped, as the -520 and -720 had fixed feeds (the -720 is also touched on in Volume 26). It appears as if the APG-1 and APG-2 might actually be the same model radar, built by two different vendors, and that might be leading to a lot of the confusion on which is which. Maybe.

(edit) Later, found some good gouge here Summary Technical Report of Division 14, NDRC. Volume 2. Military Airborne Radar Systems (MARS)


T!

 

[Token]

A little further. Always keeping in mind I have zero P-61 specific firsthand knowledge, so what I think I know is from reading and from 50-70 year old stories by people who did have firsthand knowledge. When I started working in the field a lot of the "old timers" who had done this kind of work were still around, and a lot of the old hardware was still squirreled away and being used periodically for other projects.


My understanding is the SCR-520 and -720 were both used in the P-61 as intercept radars, with no auto track or gun laying ability. The -720 became the standard intercept radar for the P-61. It turns out I have actually been hands on with the -720, at least the antenna and HV / transmitter decks as used in a modified truck based radar built in the early 60's and still around when I started work at a Navy facility. I did not realize the source of those parts until I started researching as a result of this thread. Looking at pictures of the -720 I realized I had seen that antenna and transmitter section before. By the same token I think I have worked with parts of an APG-2 in similar applications, but I am less sure about that as I have found few detailed images of the APG-2.


The APG-1 and APG-2 radars appear to have been both gun laying, and auto tracking (I know the APG-2 was auto tracking), radars used on several platforms, including the P-61. Specifically how these were used in the P-61 (I mean displays installed) I have no idea, but they must have replaced the SCR-720 in the P-61, as I am pretty sure they are too large for both the -720 and APG-2 to have been on the aircraft without some modification. The APG radar fed a computer, which controlled the turret.


From what I have found it appears as if maybe 6 or so aircraft were used with first the APG-1 and some of them later APG-2. These seem to have been P-61 B aircraft, Block 25, serial numbers 43-8231 to 43-8236. Pictures I have found of 2 of these aircraft do seem to show a differently painted radome than on other P-61's. The painting of the radome is potentially important as this can be a way to control unwanted reflected radar energy.


So far I have not been able to find any pictures of P-61s with the radome off and anything other than an SCR-720 antenna in place, you can easily tell the difference between the SCR-720 and the APG-2 (and I assume SCR-720 vs APG-1) by the feed assembly of the antenna.


T!

 

[wuzak]

I can't find any information about the APG-1 or APG-2.

Do you have nay links that would help?

 

[wuzak]

The Black Widow night fighter used its on-board radar only to plot intercept courses when pursuing enemy aircraft. Once having used closed with his target, the pilot of the Black Widow sighted his prey by eye and used a conventional optical gunsight to fire his guns at the enemy. Operationally-practical radar-directed airborne fire control was still many years in the future. Nevertheless, there were some experiments with the Black Widow in which automatic airborne fire control was tried out. The P-61B-25-NO was a block of seven experimental aircraft which were fitted with a Western Electric APG-1 gun-laying radar which was coupled with a General Electric remote-controlled turret system. The radar fed data into an analogue computer, which in turn directed the turret guns onto the target. One P-61B-15-NO was also modified in this fashion, and the first six P-61B-20-NO aircraft were also modified to this configuration. All of these aircraft were tested by the Air Proving Command at Elgin Field, Florida and at the night fighter training establishment at Hammer Field in California. However, I don't think that this innovation ever made it to the field.

Northrop P-61B Black Widow

 

[wuzak]

I would have thought that the gun laying radar would be shorter range than the search radar, so you would need both?

 

[Token]

I would have thought that the gun laying radar would be shorter range than the search radar, so you would need both?

Today we would be tempted to think that way, but remember these were a couple of the first air intercept and gun laying designs used, the specialization was just beginning. As systems were optimized for each job they would have diverged further and further, but initially they could have been very similar, drawing on parts and processes on hand.


The first question is why do you think the gun layer would be shorter ranged? Sure, when talking about actual .50 caliber guns we don't need 10 mile range, the affective range of the guns is going to be something less than 2000 yards, so that is one driver to short range. However why can't the gun layer also serve as the AI radar? It could guide you to the target and, when close enough, direct the guns also.


The SCR-720 and the APG-2 antennas are similar in size, and they work in the same frequency ranges at similar transmitter power levels. The size determines the gain for a given frequency. Gain and transmitter power level, when combined with receiver MDS (primarily driven by technology) are the main contributors to maximum detection range. So it would be reasonable to think that the two radars had similar maximum detection ranges. There is the issue of maximum unambiguous range as defined by pulse repetition interval, but I don't think either of these radars were working at particularly high PRI's (I can't find anything definitive, but the magnetron duty cycle was probably the limiter).


Further, some documentation calls the APG-1 and APG-2 both air intercept and gun laying. Why not combine the functions?


The SCR-720 used a simple helical scan, just its main beam swept out in space in a helical fashion. Viewed from above the RC-94 antenna swept around and around continuously clockwise with 2 different scan speeds, 360 RPM and 100 RPM (100 RPM being use din long range and beacon modes). The transmitter was inhibited whenever it was pointed back towards the aircraft, so this resulted in an arc in front of the aircraft. The beam was about 10 degrees wide in azimuth and elevation. So a single sweep was an arc 10 degrees tall. The end points of the arc, the angle off the nose in azimuth, was controllable, and might be as wide as 180 degrees, +/- 90 off the nose. Each revolution of the antenna it could be automatically (or manually) stepped up in 5 degree increments (half the 10 degree beamwidth). The antenna hardware could mechanically see up and down 60 degrees, or twelve 5 degree steps up and down. However there were issues with low altitude scans (not only ground clutter issues but also platform issues) so the usable limitations were more like -30 to +60 degrees elevation. The SCR-720 controller allowed the selection of 4 preset elevation ranges, a sub set of the total hardware limits. Typically this was a single level scan at 0 deg elevation, covering -5 to +5 deg with the beam, and 3 other multiple elevation level scans.


The APG-1/2 used a Palmer scan, combining a raster function with its conical scan. The conical scan was required for the angles aspect (azimuth and elevation) of the automated track capability, if they had not used conical they would have had to use some other node switching scan, and the conical scan had been perfected for application to the SCR-584. The raster function allowed the conical scan to be rapidly swept across large areas of the sky.


Because the APG-1/2 used this Palmer / raster scan instead of a helical scan it was likely a bit less good at finding targets, taking longer for scan cycles to complete. Instead of just slinging the antenna around and around the antenna had to travel to one end of the sweep, stop, step up or down in elevation, reverse direction, sweep back to the other end, stop, repeat. However once it found a target more precise information on the target was available. So you trade speed for precision.


The minimum range for each system would be tied to pulse width and receiver recovery time. I have no data for the APG-1/2, but the data I can find for the SCR-720 indicates a minimum range of either 100 or 150 yards. The maximum radar range of the SCR-720 against an airborne target was 20 miles or less. These -720 numbers would not have been unreasonable for a combined AI/AGL radar of similar technology. Yes, the -720 also had a 100 mile radar range scale, but this appears to have been more of a beacon mode than a skin track mode, despite the range being selectable in radar mode.


T!

 

[Token]

I can't find any information about the APG-1 or APG-2.


Do you have nay links that would help?

I ran across a couple of online sources that tied the APG-1/2 to the P-61, and that allowed me to find more references in print. The best was that DTIC document I quoted earlier, Military Airborne Radar Systems, but several other DTIC docs helped also.


Most of the technical information I found on the APG-1 / 2 was in the MIT Radiation Laboratory Series of books. I have an original 1947 first edition of this 28 volume set. And you really had to dig to find it, there is not, for example, a section dedicated to them. Rather there are sections on different scan types, antenna designs, display types, etc, and if you dig deep in each of those they reference specifics of the SCR-720 and the APG-1/2, as well as other similar radars. Looking them up in the indexes or searching them by key word will yield a few pages, reading the text will show more related info.


You can find the Rad Lab series online in PDF from several sources, such as here: https://www.jlab.org/ir/MITSeries.html But as I said, you have to dig in those to find the info. Volume 26 has the easiest to find information, but not the most detail. From there you have to work without the designations and more on the pictures and descriptions since those books dwell less on system specifics, by name, and more on techniques.


T!

 

[wuzak]

The first question is why do you think the gun layer would be shorter ranged? Sure, when talking about actual .50 caliber guns we don't need 10 mile range, the effective range of the guns is going to be something less than 2000 yards, so that is one driver to short range. However why can't the gun layer also serve as the AI radar? It could guide you to the target and, when close enough, direct the guns also.

I though the gun-laying radar would be shorter range as that's all it needs to be.

You make a good point about the radar being able to do both search and track and gun-laying. It should be possible.

And it may be possible to put a pipper on the pilot's gun sight for the fixed guns. Help him line up the shot. It may have even been the case with the Mosquito and the AI MK.X (SCR 720).

But the forward facing radar is little usefulness for gun-laying for the turret, as the majority of the turret's field of fire is not covered by the radar. Sure, it will give some direction in offensive operations when the turret is being used in the forward position, but not for all of its elevation and certainly not for defensive fire to the rear.

 

[XBe02Drvr]

But the forward facing radar is little usefulness for gun-laying for the turret, as the majority of the turret's field of fire is not covered by the radar. Sure, it will give some direction in offensive operations when the turret is being used in the forward position, but not for all of its elevation and certainly not for defensive fire to the rear.

Forget defensive fire. The only value of the turret in night/blind fighting is offensive in the forward quadrant. Technology of the time didn't allow for 360 degree spherical radar coverage. Defensive fire was a strictly visual business.
The turret could either be locked forward and aligned with the ventral cannons, or aimed at a target in the radar's cone of coverage. Or it could be elevated "schrage musik" style and fired at an overhead target.
Cheers,
Wes

 

[Token]

I though the gun-laying radar would be shorter range as that's all it needs to be.

Today that might be the case in some applications, and from the APG-3 on that appears to have generally been the case. But that is more typically talking about a turret on a bomber for defensive fire. The turret on the P-61 was more of an offensive tool than defensive.

You make a good point about the radar being able to do both search and track and gun-laying. It should be possible.

Since the APG-1/2 and the SCR-720 were approximately the same size, I doubt there was any other way to do it. There simply was not enough room for two systems that size in the P-61, so one system was going to have to do both tasks if you wanted them done.


By the way, it appears the APG-1 might have been originally conceived of for the XA-26A night fighter, 41-19505.

And it may be possible to put a pipper on the pilot's gun sight for the fixed guns. Help him line up the shot. It may have even been the case with the Mosquito and the AI MK.X (SCR 720).

I don't think there is any way the SCR-720 could put a target in the gun sight. I am very sure that was not ever possible in the P-61 installation. The -720 simply has no way to isolate one target into the sight.


The installation of the -720 in the P-61, and pictures of the Mossy look similar, was a C scope mounted just under the gun sight and centered in front of the pilot. The Radar Operator had both a B and C scope, but the pilot had only a C scope. Information on how a C Scope displays here Radar display - Wikipedia That page presents a round C Scope, but in the P-61 it was rectangular, still the same kind of scope.


So the pilot had only a position relative to the nose on his display, with NO indication of range. It was, in essence, an electronic sight with no lead computation. And he also had a 5.8 power set of night binoculars with range estimation pips in them. The probable use was for the pilot to use the C Scope to get lined up with the target, then switch to the glasses for fine tuning the shot.


Interesting you bring up the AI Mk X. While this is essentially an SCR-720 it appears to have retained the SCR-520 ability to be set to a much wider elevation coverage area. On the SCR-520, and apparently on the AAI Mk X, you set the scan elevation window with two switches. One set the low level and one set the high level. So for example if you set the low level to -20 degrees and the high level to +30 degrees the radar would search between -20 and +30 degrees. The US -720 just had 4 preset selections, one 4 position switch, for total window height.

But the forward facing radar is little usefulness for gun-laying for the turret, as the majority of the turret's field of fire is not covered by the radar. Sure, it will give some direction in offensive operations when the turret is being used in the forward position, but not for all of its elevation and certainly not for defensive fire to the rear.

I really don't think the turret on the P-61 was there for defensive applications. Just looking at the general layout it really looks to me like it was there to improve the fighters chances during attacks.


Turrets on bombers with gun laying radars, such as the APG-3 and APG-15, often put the radar right on the turret, so it always pointed in the general direction the turret was physically pointed. And those radars also often operated at much lower power levels, as little as 1/100, and less, the transmitted power of the APG-1/2.


In the P-61 the use of a gun laying radar would have allowed the pilot to not have to sight. That means he could fly the aircraft, intentionally keep a slight elevation bias on the target, target slightly above his nose, and concentrate on not running into the target, while the gunner or RO locked the target up and triggered the guns.


T!

 

[Zipper730]

The radar scanned.

Okay

In the original versions the radar operator would direct the gunner as to where to aim.

Like he'd call out elevation, azimuth, and range?

I don't believe that the operator directed the radar as such, but monitored it and adjusted the operating parameters.

How were the operating parameters adjusted? Basically was this like what a lock-on does but manually adjusting everything?

 

[wuzak]

Like he'd call out elevation, azimuth, and range?

Yes, that would be logical.

How were the operating parameters adjusted? Basically was this like what a lock-on does but manually adjusting everything?

Wouldn't have a clue.

 

[Zipper730]

By the end of the war the Japanese, completely independent of any knowledge of British or US efforts, had a working short wavelength cavity magnetron. The Russians also had an indigenously developed version.

I never knew the Japanese and Russians developed them, fascinating.

Weren't most of the AN/APG designated series during WW II gun layers of one type or another?

Quite a few according to this site AN/APG to AN/APH - Equipment Listing

Some as aids to manual tracking, some as automatic systems, and I know a few were also bomb / torpedo radars. The AN/APG-1 (SCR-580) and the AN/APG-2 (SCR-702) come to mind (or maybe switch those two SCR numbers, see below).

I'm not sure as for their SCR designators, but from what it would appear that the following radars did the following things...

AN/APG-1: S-Band interception radar: Was used on the P-61B
AN/APG-2: S-Band interception/gun-laying radar: Was used on the P-61
AN/APG-3: Gun-laying radar, designed for tail-gun use: Was used on the B-29 and B-36B
AN/APG-4: Low-altitude torpedo-release radar: Was used on the TBM
AN/APG-5: S-Band gun-laying/range-finding radar: Was used on the B-17, B-24, F-86A-5
AN/APG-6: Low-altitude bomb-release radar: Was a modified AN/APG-4
AN/APG-7: Glide-bomb control radar
AN/APG-8: S-Band turret fire-control radar: Was used on the B-29B
AN/APG-9: Low-altitude bomb-release radar: Was an improved AN/APG-6
AN/APG-10: Non specified weapon-system radar
AN/APG-11: L-Band toss-bombing radar
AN/APG-12: L-Band low-altitude bomb-release radar: Was yet another improved AN/APG-6
AN/APG-13: S-Band gun-laying radar: Used with the B-25H's 75mm nose-cannon
AN/APG-14: S-Band gunsight-radar: Used on the B-29
AN/APG-15: S-Band tailgun-radar: Used on the B-29B, and PB4Y
AN/APG-16: X-Band gun-laying radar: Was a odified AN/APG-2; Used on the B-32 and XB-48
AN/APG-17: S-Band low-altitude bomb-release radar: Was yet another improved AN/APG-4
AN/APG-18: X-Band gun-laying turret-control radar: Was a modified AN/APG-5, and used with the S-4 gunsight
AN/APG-19: X-Band fire-control radar: Improved AN/APG-8 and -18
AN/APG-20: S-Band low-altitude bomb-release radar: Was yet another improved AN/APG-6
AN/APG-21: Ground-ranging radar
AN/APG-22: X-band gunsight-radar: Was used with the Mk.18/23 lead-computing gunsight
AN/APG-23: Weapons-system radar: Was used on the B-36A
AN/APG-24: Weapons-system radar: Was used on the B-36B​

I remember both of those as auto tracking systems, similar in function to the ground based SCR-584 (a system I do have firsthand experience with, and no, I am not that old) and tried on several aircraft, including, I think, the P-61.

Auto-track means blind-shooting or lock-on right?

Weren't the AN/APG-3 and APG-15 both 10 GHz auto tracking gun layers developed for the B-29 or B-32? Others, like the AN/APG-5 (SCR-726), were range only radars.

The APG-3 was used on the B-29 and B-36B; the APG-15 was used on the B-29B, and PB4Y.

From what I have found it appears as if maybe 6 or so aircraft were used with first the APG-1 and some of them later APG-2. These seem to have been P-61 B aircraft, Block 25, serial numbers 43-8231 to 43-8236. Pictures I have found of 2 of these aircraft do seem to show a differently painted radome than on other P-61's.

Can you put those images up?

Looking at these numbers the AN/APG-1 entered service around 1942-1943?

The SCR-720 used a simple helical scan, just its main beam swept out in space in a helical fashion. Viewed from above the RC-94 antenna swept around and around continuously clockwise with 2 different scan speeds, 360 RPM and 100 RPM (100 RPM being use din long range and beacon modes). The transmitter was inhibited whenever it was pointed back towards the aircraft, so this resulted in an arc in front of the aircraft. The beam was about 10 degrees wide in azimuth and elevation. So a single sweep was an arc 10 degrees tall. The end points of the arc, the angle off the nose in azimuth, was controllable, and might be as wide as 180 degrees, +/- 90 off the nose. Each revolution of the antenna it could be automatically (or manually) stepped up in 5 degree increments (half the 10 degree beamwidth). The antenna hardware could mechanically see up and down 60 degrees, or twelve 5 degree steps up and down. However there were issues with low altitude scans (not only ground clutter issues but also platform issues) so the usable limitations were more like -30 to +60 degrees elevation.

What kind of platform issues?

The APG-1/2 used a Palmer scan, combining a raster function with its conical scan. The conical scan was required for the angles aspect (azimuth and elevation) of the automated track capability, if they had not used conical they would have had to use some other node switching scan, and the conical scan had been perfected for application to the SCR-584. The raster function allowed the conical scan to be rapidly swept across large areas of the sky.

So the conical scan provides some kind of triangulation function (from what I vaguely remember as the beam spins round and round, it produces two lobes as it does) and more precision while the raster function basically swings left and right?

Because the APG-1/2 used this Palmer / raster scan instead of a helical scan it was likely a bit less good at finding targets, taking longer for scan cycles to complete.

How long would it take in comparison. To do a 360-degree scan on the SCR-720 it 6 revolutions every second, or 2160 degrees per second.

I'm surprised you couldn't just use a raster-scan for search and the combination of conic/raster for target-tracking.

The minimum range for each system would be tied to pulse width and receiver recovery time. I have no data for the APG-1/2, but the data I can find for the SCR-720 indicates a minimum range of either 100 or 150 yards.

Gunnery range is around 400 yards right?

I ran across a couple of online sources that tied the APG-1/2 to the P-61, and that allowed me to find more references in print. The best was that DTIC document I quoted earlier, Military Airborne Radar Systems, but several other DTIC docs helped also.

Most of the technical information I found on the APG-1 / 2 was in the MIT Radiation Laboratory Series of books. I have an original 1947 first edition of this 28 volume set. And you really had to dig to find it, there is not, for example, a section dedicated to them. Rather there are sections on different scan types, antenna designs, display types, etc, and if you dig deep in each of those they reference specifics of the SCR-720 and the APG-1/2, as well as other similar radars. Looking them up in the indexes or searching them by key word will yield a few pages, reading the text will show more related info.

You can find the Rad Lab series online in PDF from several sources, such as here: https://www.jlab.org/ir/MITSeries.html But as I said, you have to dig in those to find the info.

I'll take a look at that...

 

[XBe02Drvr]

I'm surprised you couldn't just use a raster-scan for search and the combination of conic/raster for target-tracking.

The design of an antenna optimized for helical scan and the machinery to move it was quite different from raster scan systems. Besides, you can generally fit a larger diameter raster dish in the same size radome, giving it an advantage in gain, and hence detection range. An antenna designed to do both probably wouldn't do either very well.
Later radars (as in jet interceptors) used a nodding (raster scan) antenna with a feedhorn that could rotate (nutate), giving the beam a conical effect when in lock-on mode that made for accurate tracking and a secure lock. That nutating beam was what generated the distinctive "spinscan" sound in the threat warning receiver heard just before an AAM blew your tail off. In addition to helping the radar keep lock, the reflections from that nutating beam guided the missile to the target. If the target could maneuver sharply enough to break lock (usually by exceeding the gimbal limits of the opponents antenna), he could defeat the missile.
Cheers
Wes

 

[Token]

Auto-track means blind-shooting or lock-on right?

Yes, in Auto Track you are locked on the target. The radar is automatically generating Azimuth, Elevation, and Range errors and applying those errors to the auto track to keep the target position known.


Range only radars may just keep the range gate locked on the aircraft, but other radars, such as the APG-1/2, would track in all axis. The angular track accuracy of the APG-2 was about +/- 0.15 degrees.

Can you put those images up?

I'll have to find them again, I did not keep them.

What kind of platform issues?

When looking too far to the side or too far down (assuming your antenna is bottom mounted) while radiating you start to get a lot of RF energy (often from side lobes of the antenna) reflected back into the radar from the structure of the aircraft. By limiting the off axis look angle in those planes you can reduce this, and you can further control it by strategically placed sheets of reflective and absorptive material. And yes, some kinds of paint can be used for this purpose.

So the conical scan provides some kind of triangulation function (from what I vaguely remember as the beam spins round and round, it produces two lobes as it does) and more precision while the raster function basically swings left and right?

A tracking radar will produce the elevation and azimuth errors by either amplitude changes or phase changes, depending on the system design. I am pretty sure the APG-1/2 used amplitude changes.


Think of it this way, the feed of the antenna is intentionally mechanically skewed, so the beam points off center slightly. Now the feed is spun on its central axis, so this skewed beam "wobbles" around the center of the antennas pointing axis. If you compare the signal return amplitude when the scan is at the top and bottom of the wobble you can determine if the target is up or down in relationship to the scan axis. The same technique is applied to azimuth.


Now you know what directions to drive the antenna to center the target.


When the target is centered in the antennas beam the target amplitude at the top and bottom, as well as left and right, positions of the conical scan (feed rotation) will be the same. This is the desired state for the tracking circuits, equal amplitude for the target in all feed positions, this means the antenna is looking right at the target.


Of course all of the above only works if you can get a Range automatic track or can manually keep a range gate on the target. This Range track gate establishes the sample gate times for Elevation and Azimuth errors. Enter Electronic Counter Measures, if you can break the Range lock you have gone a long way to killing the auto track.


The Raster scan simply scans right and left and up and down. So that it might start in the lower left corner of the "box" (scanned volume), sweep to the lower right corner, step up by a pre-determined elevation, scan to the left, hit the left edge, step up, etc. The example I have described here would be a Bidirectional Raster Scan.

How long would it take in comparison. To do a 360-degree scan on the SCR-720 it 6 revolutions every second, or 2160 degrees per second.

The SCR-720 had two scan rates, 360 RPM and 100 RPM.

I have no idea the exact volume scan rates of the APG-1/2, the texts I have seen only indicate it is slower than the SCR-720 scan rate. And since it uses a raster, instead of a helical, I can understand that, to do a raster you have to stop and change directions at each edge instead of just continuing the scan around.

I'm surprised you couldn't just use a raster-scan for search and the combination of conic/raster for target-tracking.

The problem is that skewed feed and the fact it is mechanically derived. To use a pure raster search you would have to stop the feed rotation. Now you have to index the feed to a known location in its rotation, so that you know the direction of the skew and can compensate for the skew. Then you can do the raster. But to transition to track you now must start the feed rotating, zero the offset you put in to compensate for skew, and initiate track.


Simpler just to leave the conical scan running while you search.


And this combination of a Conical scan with a Raster scan is called a Palmer scan.

T!

 

[Token]

Later radars (as in jet interceptors) used a nodding (raster scan) antenna with a feedhorn that could rotate (nutate), giving the beam a conical effect when in lock-on mode that made for accurate tracking and a secure lock.

The APG-1/2 used exactly this setup. Except, like most systems, the conical scan is on all the time, even in search modes. Systems that stop the con scan are few and far between, far easier to just leave it scanning and don't bother using the amplitude / phase data. Or have separate feeds / antennas and do a COSRO (Conical Scan On Receive Only) kind of thing.


By the way, if the feed rotates it is not nutating. While both rotation and nutation can result in a conical scan, nutation does not rotate the feed. A rotating feed results in a changing polarization, while a nutating feed does not. Also, some designs nutate by moving the reflector with the feed staying stationary. These later are the most likely to stop the con scan for search, as it is relatively simple to index the reflector to on axis with the feed.

That nutating beam was what generated the distinctive "spinscan" sound in the threat warning receiver heard just before an AAM blew your tail off.

But you really know you are having a bad day when the sound goes to steady, this means the conical scan is bore sighted on you. As long as you hear the warble / varying tones / amplitudes of the scan they are probably not dead on you. A steady tone in the RHAW / RWR / ESM gear is always bad, as it means whatever you are hearing has either stopped scanning and is staring at you, or the scan variations have become so small you can't detect them by ear.

T!

 

[Zipper730]

Yes, in Auto Track you are locked on the target.

Understood

I'll have to find them again, I did not keep them.

Oh, I thought you had the pictures with you...

When looking too far to the side or too far down (assuming your antenna is bottom mounted) while radiating you start to get a lot of RF energy (often from side lobes of the antenna) reflected back into the radar from the structure of the aircraft.

So the limitations are the aircraft's structure reflecting back into the radar, and the terrain providing the aircraft is low to the ground and within the resolution cell?

The conical scan will help produce the elevation and azimuth errors by either amplitude changes or phase changes, depending on the system design. I am pretty sure the APG-1/2 used amplitude changes.

I thought you wanted to reduce the errors?

Think of it this way, the feed of the antenna is intentionally mechanically skewed, so the beam points off center slightly. Now the feed is spun on its central axis, so this skewed beam "wobbles" around the center of the antennas pointing axis. If you compare the signal return amplitude when the scan is at the top and bottom of the wobble you can determine if the target is up or down in relationship to the scan axis. The same technique is applied to azimuth.

Now that makes sense to me

The Raster scan simply scans right and left and up and down.

That I get, I said what I said because the palmer-scan was better for holding a lock, while the raster scan was faster for going left-right/up-down.

I have no idea the exact volume scan rates of the APG-1/2, the texts I have seen only indicate it is slower than the SCR-720 scan rate. And since it uses a raster, instead of a helical, I can understand that, to do a raster you have to stop and change directions at each edge instead of just continuing the scan around.

I would not have thought the antenna would have carried that much inertia... learn something new everyday

The problem is that skewed feed and the fact it is mechanically derived.

Meaning it's based on the device that swivels the antenna and rotates it? And I'm guessing the feed-horn is not designed to not rotate round and round, so it wouldn't be designed to simply center itself so it would be able to simply point in whatever direction the swivel is aimed so it just goes left, right, left, right, left, right, or left, right, up, left-right, up, left, right, up, left right, down... and so on?

 

[XBe02Drvr]

Meaning it's based on the device that swivels the antenna and rotates it? And I'm guessing the feed-horn is not designed to not rotate round and round, so it wouldn't be designed to simply center itself so it would be able to simply point in whatever direction the swivel is aimed so it just goes left, right, left, right, left, right, or left, right, up, left-right, up, left, right, up, left right, down... and so on?

Imagine a plate sitting on a table. That is your antenna dish. Now imagine there's a small depression in the exact center of the plate in which the point of a tall slender top is spinning. Gyroscopic rigidity in space keeps the top perpendicular to the plate. That is your rotating feedhorn. Now imagine you can suspend the laws of physics enough to pick up this plate/spinning top assembly and bolt it into an articulated gimbal structure on the front end of your aircraft. This gimbal has powerful little motors that drive the dish left-right-up-a-notch, left-right-up-a-notch, and so forth until it reaches the upper gimbal limit, then repeats the process down to the lower limit and back to level. Since a feedhorn isn't as symmetrical as a top, the beam it projects from the dish at any point in its rotation is not symmetrical around the central axis of the dish. Rotation of the feedhorn causes this asymmetrical beam to form a very slender cone of signal centered on the axis of the dish, strongest on axis and weaker towards the edges. This oprovides the signal strength comparisons that generate the "pointing error" signals used to drive the antenna to put its axis "on target". Once the antenna is "on target", the azimuth and elevation are known and return time provides range.
Cheers,
Wes

 

[Token]

So the limitations are the aircraft's structure reflecting back into the radar, and the terrain providing the aircraft is low to the ground and within the resolution cell?

Those are some of the limitations, to be sure. And your look down angle is also limited by your altitude. Without MTI (and these early radars did not have MTI) looking down will show you the ground out at the end of the range trace. The more you look down, for a given aircraft altitude, the shorter you maximum range to track or detect another aircraft will be.


I think you are using resolution cell incorrectly though. The resolution cell is the time occupied by the radar pulse. It is a moving space defined in depth by the duration of the pulse and in width and height by the beamwidth of the radar. The resolution cell is the area where a radar can detect a target is present but cannot resolve the number of targets present in the cell. And it has nothing to do with maximum range.

I thought you wanted to reduce the errors?

Errors are how the radar knows where the target is. Yes, you want to keep the errors as small as possible, but to do this you must first find the errors, so you can counter or null them. Every track has errors, no track is perfect, so you want to make the errors as small as possible.


The Conical scan is how the radar, in this case, derives the errors to know what direction to drive the antenna, so as to null the errors. There are other ways to develop these angle errors, Con scan is just one of the oldest techniques.


Kind of like this, only substitute target track for missile:


View: https://www.youtube.com/watch?v=F4Dvc1NrZJI

That I get, I said what I said because the palmer-scan was better for holding a lock, while the raster scan was faster for going left-right/up-down.

The Palmer scan is not better at anything, it is simply the combination of two different scan types at the same time. In this specific application the Conical scan is required to track, the Raster is required to search. When the radar has both types of scans active at the same time it is called a Palmer scan. The radar cannot both scan and track at the same time, it is one or the other.


So while searching, or scanning a large volume of space, it is using a Raster scan, it just so happens the beam also has a Conical scan at the same time. While tracking the radar stops the Raster scan, stares at the target, and uses its Conical scan to derive tracking errors.

I would not have thought the antenna would have carried that much inertia... learn something new everyday

For a bidirectional Raster scan, which by definition changes direction for every bar of the raster, you must overcome that inertia, no matter how small or large it may be. For a Helical scan there is no need to overcome anything except friction, the antenna just goes around and around in the same direction at the same rate, stepping up at some point in each revolution. A unidirectional Raster may do the same thing as a Helical, and have no inertia to overcome, or it might stop and retrace, having to overcome inertia each stop.

Meaning it's based on the device that swivels the antenna and rotates it? And I'm guessing the feed-horn is not designed to not rotate round and round, so it wouldn't be designed to simply center itself so it would be able to simply point in whatever direction the swivel is aimed so it just goes left, right, left, right, left, right, or left, right, up, left-right, up, left, right, up, left right, down... and so on?

No, the portion of the antenna mechanism that rotates the antenna and grossly moves the beam around in space in elevation and azimuth is unrelated to the skewed feed that provides the Conical scan.


A fixed feed antenna, like the SCR-720, just stares (with its beam) straight ahead of the dish. The device the antenna is mounted on (often called a pedestal) then moves the beam around in space, in the case of the SCR-720 in a Helical fashion (360 degrees of rotation, step up, 360 degrees of rotation, step up, etc), in the case of the APG-1/2 in a Raster.


Completely separate from that is the Conical scan. It rotates the beam of the antenna in small circles out in front of the antenna, independent of how the pedestal moves the dish. To do this it is intentionally skewed, mechanically, so that the beam does not go out directly ahead of, and on axis with, the dish. It goes slightly off center of where the dish is looking.


This explains the basic Conical portion of the operation pretty well Conical scanning - Wikipedia


The key is above, when I said to do this the antenna feed is skewed, mechanically, to make this happen. This is a simple design. To allow the feed to index from this mechanically skewed position to on boresight would complicate the feed assembly. It is done on some systems, but not many at all.


More often when you want to allow both Conical scan (a spinning off boresight axis beam) and on boresight axis fixed beam operation you don't use a mechanically skewed feed assembly. Instead you wobble the dish itself to create the circular beam motion about the boresight axis. This would result in a nutating beam.


T!