P-61 Gun-Laying Radar

[Koopernic]

That's fascinating
Or drag...
Makes enough sense...

That would be around 6.57 yards at 400 yards?

Blind fire radar was used by the Royal Navy, German Navy and USN. Certainly by the time of Perl Harbour everybody had it. Nobody was in advance of anyone else in a significant way. The Seetakt radar FuMO 26 had an accuracy of 0.25 degrees (about 10000 x tan(0.25) = 43 yards at 10000 yards) and 70m range accuracy. Range accuracy was not absolutely critical because the splash of a 6 inch round was 140 meters high and could easily be seen and compared. British Type 284 radar was unusually accurate at 0.06 degrees because they electronically evaluated the deviation between the lobes and brightened the display beam when strength of the lobes of the actual radar beams was equal and on target. The Germans could have done this but for some reason didn't. Allied radar began to pull ahead in 1942 when for instance type 284 radar jumped from 25kw to 125kw power. The Germans actually matched this but probably never got it to use at sea (only on shore artillery maybe the Tirpitz in 1944) because they were under pressure in the u-boat war and anti bomber war. Also at this time microwave sets started to become common as they moved from the trials stage to deployment. They had significant problems with accuracy and 50cm sets remain the norm till near the end of the war. The US deployed it's outstanding SCR-584 set in 1943 only when the Germans jammed SCR-268 during the Anzio landings. The Germans caught up by the second half of 1944 but by then could produce these sets in quantity. They'd actually had a 13.5cm and 5cm microwave program but gave up much of it when they thought that 50cm could give them all they needed.

 

[Token]

Blind fire radar was used by the Royal Navy, German Navy and USN. Certainly by the time of Perl Harbour everybody had it.

What must be clarified is that blind firing is not necessarily the same as auto tracking. I believe that all Naval blind firing prior to Pearl Harbor was operator-in-the-loop. All of the Navy systems until well after that time relied on operators to align strobes / equalize return size / etc to develop the tracking information. I am pretty sure for the US Navy the first auto tracking gun layers were in early 1943.


So the ability to auto track, vs just blind fire, in field applications came later and started with ground based AAA systems. In fact the British were opposed to auto track, believing a man in the loop was superior.


Gun laying was important, this was Project II for the M.I.T. Radiation Laboratory, with Project I being AI radar. Project III was navigational aids.

The US deployed it's outstanding SCR-584 set in 1943 only when the Germans jammed SCR-268 during the Anzio landings.

The timeline for SCR-584 roll out was established long before that, and stayed pretty much on track, with some minor delays, and the planning for it to replace the SCR-268 goes back to sometime in 1941. Production deliveries of the -584 started in May of 1943 in the US (production orders placed Sep 1942). But training units were still being established, so it was months after that before these first delivery systems were ready to ship out to combat.


Several months before the Anzio landings started in January of 1944 SCR-584's were initially arriving in England, I don't have the exact date but it appears to be about Sep or Nov of 1943. And the first operational kill by an SCR-584 was on February 4, 1944, Battery A, 184th AAA Gun Battalion, Lippets Hill, England. The SCR-584s went ashore on Feb 24, 1944, for Anzio. 24 hours later the -584 got its first kill in Anzio.


Anzio was the first place the SCR-584 became well known, but it was after D-Day that they really started to pay off. 39 sets of SCR-584 and associated guns went ashore on June 6, 1944. But back in England they played a major part in V-1 interceptions, after the V-1 Blitz was initiated in response to the D-Day landings.


T!

 

[Koopernic]

What must be clarified is that blind firing is not necessarily the same as auto tracking. I believe that all Naval blind firing prior to Pearl Harbor was operator-in-the-loop. All of the Navy systems until well after that time relied on operators to align strobes / equalize return size / etc to develop the tracking information. I am pretty sure for the US Navy the first auto tracking gun layers were in early 1943.


So the ability to auto track, vs just blind fire, in field applications came later and started with ground based AAA systems. In fact the British were opposed to auto track, believing a man in the loop was superior.


Gun laying was important, this was Project II for the M.I.T. Radiation Laboratory, with Project I being AI radar. Project III was navigational aids.





The timeline for SCR-584 roll out was established long before that, and stayed pretty much on track, with some minor delays, and the planning for it to replace the SCR-268 goes back to sometime in 1941. Production deliveries of the -584 started in May of 1943 in the US (production orders placed Sep 1942). But training units were still being established, so it was months after that before these first delivery systems were ready to ship out to combat.


Several months before the Anzio landings started in January of 1944 SCR-584's were initially arriving in England, I don't have the exact date but it appears to be about Sep or Nov of 1943. And the first operational kill by an SCR-584 was on February 4, 1944, Battery A, 184th AAA Gun Battalion, Lippets Hill, England. The SCR-584s went ashore on Feb 24, 1944, for Anzio. 24 hours later the -584 got its first kill in Anzio.


Anzio was the first place the SCR-584 became well known, but it was after D-Day that they really started to pay off. 39 sets of SCR-584 and associated guns went ashore on June 6, 1944. But back in England they played a major part in V-1 interceptions, after the V-1 Blitz was initiated in response to the D-Day landings.


T!

Quite in agreeance. Lobe switching is critical for obtains the accuracy required for blind fire, electronic as opposed to visual evaluation improves accuracy further. Auto track of the range combined with electronic evaluation of the deviation then provides an easy path for amplification of the deviation signals and there use in driving servo motors such as the ward Leonard drives used on German radars and the scr-584.

When the Battle od Denmark straights (Bismarck, Hood, PoW etc) happened both sides had radar the German Seetakt FuMO 23 and the British type 284. These radars could range fairly well but were not blind fire capable. The half power beam width of the Seetakt radars was between 4 and 6 degrees depending on antenna size. An operator could locate a target to within a degree by wobbling the antenna of until it was a maximum strength but this was not really enough to shoot guns or torpedos reliably or provide the smooth inputs fire control computers (directors, predictors) needed. The FuMO 26 introduced lobe switching on the receive side called Raddle Peilung (Rattle Pointing) because of the sound the relays made as they switched. This allowed the operator to compare the left versus right returns and get to 0.25 degree accuracy. The British when they upgraded to lobe switching type 284M or 284P went the one step further and electronically evaluated the left versus right return and got 4 times the accuracy.

The Würzburg-D and larger Würzburg-Riese Radar had conical scan and required one operator to track the range manually while the second operator was presented a osciliscope Display of the horizontal and vertical deviation presented as an sine wave amplitude, his job was to make the amplitude zero. Average accuracy was about 0.3 degrees on the 3m diameters Würzburg D and 0.15 on the giant 7.5m Würzburg-Riese. (0.1 degree in the horizontal plane).

Little known is that the Luftwaffe fielded a radar called Mainz (latter Mannheim FuSE 63) also on a 3m dish from 1942 that evaluated the deviation electronically to double the accuracy. Deviation was also presented on dial gauges and the range was also more accurate. They only manufactured about 200.

The late 1943/1944 issue 2 of Mannheim had autotrack for the range gate like SCR-584 and about 140 vacuum tubes like SCR-584. It's not clear to me as to whether they had auto-track only on the critical range or whether they also automatically tracked bearing and elevation. It wouldn't have been hard as the small signal used to drive the indicator dials (already there) could simply have been amplified to drive the ward-Leonard servo motors (already there). The range was accurate to 6m and tracking was so good that even with windows and carpet jamming the circuit could track even though it could be distinguished on the traces by the operators. (They had to find it in the first place)

The Germans had been transferring data directly from radar to FLAK predictor (Kommandogaraet 58) since 1941 with Würzburg-D so Allied Radar wasn't quite as far ahead as often made out, the Germans lagged mainly in producing 9cm microwaves and had to make do with 50cm. Even then 8.4cm sets started to appear in trials by mid 1944 such as Rotterheim (Mannheim using microwaves)

They used 8.4cm so that their passive homing devices could distinguish Allied from German radar.

 

[Token]

The late 1943/1944 issue 2 of Mannheim had autotrack for the range gate like SCR-584 and about 140 vacuum tubes like SCR-584. It's not clear to me as to whether they had auto-track only on the critical range or whether they also automatically tracked bearing and elevation. It wouldn't have been hard as the small signal used to drive the indicator dials (already there) could simply have been amplified to drive the ward-Leonard servo motors (already there). The range was accurate to 6m and tracking was so good that even with windows and carpet jamming the circuit could track even though it could be distinguished on the traces by the operators. (They had to find it in the first place)

If you have the angle error signals, as in the indicator dials / meters, then I would assume the design could have auto angle track if you wanted. You have all the information required.


As for chaff / window and jamming, with a range tracking radar, I mean the expanded or tracking range gate is on the target and AGCing the radar, even a relatively inexperienced operator can see through / track through chaff. And unless the jamming completely removes the return (i.e. negative SNR, or positive J/S) a decent operator can track through that also.


I have seen thousands of chaff events, and to train an operator to see the real target in the chaff is a matter of a few events / days. And as long as the radar allows manual operator inputs to correct the range track a simple twist of the range handwheel puts the gate back on the target, and all is good.


Very tight beamwidths improve the abilities to track through chaff. With narrow beams and aircraft target speeds the chaff quickly moves out of the beam of the radar. Also, chaff tends to take time to bloom, expand, and it might be moved out of the tight beam before the chaff even blossoms into a large target.


Jammers can be even easier to see the target in (unless they completely overpower the return) if you have the target before the jammer comes online. Simple jam signals, cover pulses, RGPO (Range Gate Pull Off), running pulses, etc, tend not to have as much scintilation as a real return. Of course with noise jammers it is a simple matter of J/S (Jam to Signal). If your noise jamming is higher in power than the returned signal the radar, and radar operator, simply is not going to see the return.


Many jam techniques have look through times, times they stop the jammer for a short period to see if the signal they are jamming is still on the air. These short snapshots of target position can really help manual track, especially if the radar has aided rates (manually adjust coast modes).

The Germans had been transferring data directly from radar to FLAK predictor (Kommandogaraet 58) since 1941 with Würzburg-D so Allied Radar wasn't quite as far ahead as often made out, the Germans lagged mainly in producing 9cm microwaves and had to make do with 50cm. Even then 8.4cm sets started to appear in trials by mid 1944 such as Rotterheim (Mannheim using microwaves)


They used 8.4cm so that their passive homing devices could distinguish Allied from German radar.

Although I know much less about German radars than about American(and to a lesser extent British) radar, I am trying to learn.


I am not really sure the Germans were behind much, if any, at all in radar technology. They were behind in a compact microwave source, that is true, but their other techniques were often as good as, or better than, anyone elses. This limited some of what they could do, but at the same time they often took things further than the Allies did in the frequency ranges they had to use.


When the Germans did do 10 cm and shorter wavelengths it seems they tended to have lower power systems than the Allies. They simply could not make 100+ kW at those wavelengths in a device that could easily be in a 2 engine or fighter sized aircraft. If I am not mistaken they also tended to have larger antennas (in ground installations, like the Rotterheim / Marbach) , virtually required because of the lower power, and bringing the benefit of tighter beam widths.


T!

 

[Zipper730]

I know there's a dislike for batch-responses, but considering these messages sort of flow together and I'm listing them in order of response.

British Type 284 radar was unusually accurate at 0.06 degrees because they electronically evaluated the deviation between the lobes and brightened the display beam when strength of the lobes of the actual radar beams was equal and on target.

When you say 'evaluated the deviation', do you mean that it basically measured the signal strength of each lobe to determine how far left/right/up/down the target was? That seems quite advanced for the era...

Allied radar began to pull ahead in 1942 when for instance type 284 radar jumped from 25kw to 125kw power.

That must have been quite impressive...

The US deployed it's outstanding SCR-584 set in 1943 only when the Germans jammed SCR-268 during the Anzio landings.

The SCR-584 was used on the USS Enterprise (CV-6). Because the movement was visible by the gunners on the ship when they saw the radar dish stop going 'round and 'round they realized where A/C were coming from!

What must be clarified is that blind firing is not necessarily the same as auto tracking. I believe that all Naval blind firing prior to Pearl Harbor was operator-in-the-loop. All of the Navy systems until well after that time relied on operators to align strobes / equalize return size / etc to develop the tracking information.

When did the Kriegsmarine and Royal Navy develop auto-track?

In fact the British were opposed to auto track, believing a man in the loop was superior.

Is this the Royal Navy, or both the RN & RAF? Even in aircraft?

Gun laying was important, this was Project II for the M.I.T. Radiation Laboratory

And this is where the AN/APG's came?

Little known is that the Luftwaffe fielded a radar called Mainz (latter Mannheim FuSE 63) also on a 3m dish from 1942 that evaluated the deviation electronically to double the accuracy. Deviation was also presented on dial gauges and the range was also more accurate. They only manufactured about 200.

It seems the Germans were with electronics like they were with aircraft... they produced loads of designs, and many were highly creative, but not produced in sufficient numbers.

Jammers can be even easier to see the target in (unless they completely overpower the return) if you have the target before the jammer comes online.

I'm guessing you'd at least have the initial location of where the target was?

Simple jam signals, cover pulses, RGPO (Range Gate Pull Off), running pulses, etc, tend not to have as much scintilation as a real return.

1. Scintillation is a fluctuation in amplitude on the display, and an apparent movement of the target on the scope?
2. Is amplitude and signal intensity the same thing?
3. What's a Range-Gate Pull Off?
4. What's the difference between a cover pulse and running pulse?

 

[Zipper730]

And yes, bacon all around... I'm actually going to a restaurant in a little while that smokes their own bacon

 

[Token]

Is this the Royal Navy, or both the RN & RAF? Even in aircraft?

Information is scarce, but from what I have read it was a consensus and that it was Dr Bowen and Cpt Faulkner who were the most adamant about it.

And this is where the AN/APG's came?

AI was the initial goal, not necessarily gun laying. So the SCR-520, SCR-720, and Mk IX were the initial successes, the APG's came later but did come out of that Project II work.

I'm guessing you'd at least have the initial location of where the target was?

If you have track before the jammer starts it is easier to see the target in the jammer. Finding the target in a jammer that is already running is much harder if the jammer has enough power. Remember that jammers often have relatively low power compared to radars, hundreds of Watts compared to thousands. But the jammer only has a one way path loss to deal with.

1. Scintillation is a fluctuation in amplitude on the display, and an apparent movement of the target on the scope?

2. Is amplitude and signal intensity the same thing?

3. What's a Range-Gate Pull Off?

4. What's the difference between a cover pulse and running pulse?

1. Fluctuations in the target, amplitude and width.

2. It depends on context. Generally amplitude is target size and signal intensity is a function of the display settings.

3. RGPO is a specific form of jamming. Its function is to break a range auto track the radar may have. It does this by "growing" a cover pulse (a jam pulse that is not delayed or advanced in time) over the target return, and then changing the time of the jam pulse so that the jam pulse walks off of the return pulse. If the jam pulse has grown enough to capture the AGC the range gate will walk off the real target location. After the jam pulse is walked off a distance you kill the jam pulse, and the range gate is left in a location with no target to track. As far as I know RGPO was known and in development during WW II, but may not have been fielded to any operational units.

4. There are two definitions of a cover pulse, note I did not include the term jamming when I said cover pulse. The way I was using the term a cover pulse is a deceptive jamming pulse that is produced with no or minimal time delay between the time the pulse is received on the target and the jam pulse is transmitted, mimicking the return pulse. This means the jammer generated pulse will arrive back at the radar at about the same time as the reflected pulse from the target, potentially covering up the target return. A running pulse is a deceptive pulse moving in time in relationship to the radar pulse timing. That means it does not appear as a fixed pulse to the radar, but instead "runs" in or out on the display.

The other definition of cover pulse jamming involves sending a burst of noise during the time the pulse is on the aircraft. Typically this pulse is wider in time than the transmitted pulse and does not mimic the return but rather obscures it.


T!

 

[Zipper730]

Information is scarce, but from what I have read it was a consensus and that it was Dr Bowen and Cpt Faulkner who were the most adamant about it.

So, this attitude seemed more prevalent in the naval community than the RAF?

AI was the initial goal, not necessarily gun laying. So the SCR-520, SCR-720, and Mk IX were the initial successes

So either Dana Bell was wrong about the aircraft being designed to ultimately have a gun-laying radar, or that was an ultimate-wishlist item?

the APG's came later but did come out of that Project II work.

And they were around in 1943 or 1944?

If you have track before the jammer starts it is easier to see the target in the jammer. Finding the target in a jammer that is already running is much harder if the jammer has enough power.

Oh, I thought it was just because you know where the original path was and at least can extrapolate where it could go from where it first was spotted...

Remember that jammers often have relatively low power compared to radars, hundreds of Watts compared to thousands. But the jammer only has a one way path loss to deal with.

So it's just blasting out signal not blasting out signal then receiving what hasn't been absorbed or dispersed?

1. Fluctuations in the target, amplitude and width.

So the image looks like a jiggly bouncing glob?

2. It depends on context. Generally amplitude is target size and signal intensity is a function of the display settings.

When you say "a function of the display settings" do you mean like how I can make my monitor light up or dim a bit? Or do you mean the amount of power you're deciding to send out?

3. RGPO is a specific form of jamming. Its function is to break a range auto track the radar may have. It does this by "growing" a cover pulse (a jam pulse that is not delayed or advanced in time) over the target return, and then changing the time of the jam pulse so that the jam pulse walks off of the return pulse.

Walks off?

If the jam pulse has grown enough to capture the AGC the range gate will walk off the real target location.

So you make it look like it's somewhere it's not really located?

After the jam pulse is walked off a distance you kill the jam pulse, and the range gate is left in a location with no target to track.

So it's tracking empty space?

4. There are two definitions of a cover pulse, note I did not include the term jamming when I said cover pulse. The way I was using the term a cover pulse is a deceptive jamming pulse that is produced with no or minimal time delay between the time the pulse is received on the target and the jam pulse is transmitted, mimicking the return pulse. This means the jammer generated pulse will arrive back at the radar at about the same time as the reflected pulse from the target, potentially covering up the target return.

So you either mimic the return in such a way as to deceive the scanner or you just brighten it so much that it hides the plane in all the static?

A running pulse is a deceptive pulse moving in time in relationship to the radar pulse timing. That means it does not appear as a fixed pulse to the radar, but instead "runs" in or out on the display.

So it produces errors in timing the location of the aircraft as it's getting different returns at different times?

 

[Dana Bell]

So either Dana Bell was wrong about the aircraft being designed to ultimately have a gun-laying radar, or that was an ultimate-wishlist item?

Hi Zipper,

I can't say when the P-61's AGL was first planned, but it was being planned at least in early 1942 (the first mention I've found). I'm actually writing up the P-61 versus Mosquito story right now, and there's a lot about the turret and radar coming out of the record. Turns out the radar operator (guy in the way back) was being listed as the observer/gunner, and had an aft-facing sight that would have allowed him to take over the turret for defensive fire to the rear.

(While I've found enough to write my article, I still need to dig some more to discover the first mention of the AGL. I've noticed that the early documents don't even mention the radar - though some 1940/41 papers talk about the "special equipment.")

Cheers,



Dana

 

[Token]

So, this attitude seemed more prevalent in the naval community than the RAF?

I have no idea. I also have no idea how long this persisted. All I know is that it is mentioned as being the opinion of those two gentlemen during the initial meetings which established the projects of the M.I.T. Radiation Laboratory.


Obviously, it did not stop the development of such a system.

So either Dana Bell was wrong about the aircraft being designed to ultimately have a gun-laying radar, or that was an ultimate-wishlist item?

When dealing with new weapon systems it is pretty common for multiple systems to be developed in parallel. An ability you desire does not exist when you start the tasking, and so the goal is for multiple new systems to be developed and eventually delivered in a completed end state. Such plans often have phases or milestones of success even short of the originally documented end goal.


I don't know if the aircraft was intended to have an automatic gun laying ability from day one or not, but it would not surprise me if that was at least goal at some point in the life of the system. I.e., something along the lines of "if we can make this happen, great, if not, then we can still move forward with this other capability".


I think it is pretty apparent that if the goal from day one was for the P-61 to have an AGL that this was essentially an unrealized goal. While the APG-1/2 was used on about 6 airframes, for whatever reason the majority of airframes never got these systems and I believe the APGs were eventually removed from the ones that did get them. Requirements change with time, and during the rapid development and changing tactics and techniques of WW II it is very possible that they came to realize the aircraft could do its actual assigned tasking (vs the originally envisioned tasking) just fine without an AGL.


But that is all supposition on my part.

And they were around in 1943 or 1944?

I have not done the research Dana Bell has, but I think it was the P-61B-25-NO aircraft that got the APG-1/2. One could check the records to see when they were being tested at Eglin Field, FL, and at Hammer Field, CA but I think these aircraft were accepted in March and April of 1945.


The data I have indicates the APG-1 itself was a functioning design in early-mid 1944, a result of lessons learned from the SCR-584 and work lightning it to produce the SCR-784.


Obviously the SCR-520 and then 720 was around before the APG-1.

So it's just blasting out signal not blasting out signal then receiving what hasn't been absorbed or dispersed?

Not sure what you mean there.

So the image looks like a jiggly bouncing glob?

Sure, you could say that, a specifically defined jiggling glob.

When you say "a function of the display settings" do you mean like how I can make my monitor light up or dim a bit? Or do you mean the amount of power you're deciding to send out?

Just like your monitor brightness.


Stronger targets appear as taller / bigger targets, not brighter targets. With displays available in WW II any real change in intensity is probably related to the display system, not the size of the target return. Obviously bigger returns can sometimes appear brighter to the user, this is particularly true with PPI type displays, but most often that is a fiction of how your eye/mind integrates light over larger areas, not actual spot brightness.

Walks off?

As you change the delay time of a returned pulse, the jam pulse, it will move in range, either out or in depending on your timing. So the jam pulse will appear to move, or walk, along the displayed trace.

So you make it look like it's somewhere it's not really located?

In time only, and time to a radar amounts to range. So it will move the auto track range gate to a range the real target is not at.

So it's tracking empty space?

So it is tracking nothing at all. You move the jam pulse off of the real target, the radar range goes to the location, range, of the fake target, and then you shut off the fake target, leaving the radar with nothing to track in that location. The radar must then start the entire process over again, find the target, start tracking the target, and then the RGPO jammer can start working on the situation again.

But the case is kind of moot, as I said before, as far as I know there was no RGPO used in WW II, although the technique was known or being worked on.

So you either mimic the return in such a way as to deceive the scanner or you just brighten it so much that it hides the plane in all the static?

Essentially, yes.


One problem with a simple noise jammer is that even if it completely hides the target so you cannot get range, you still have angles to the noise source. Meaning you can point right at the aircraft providing the noise jamming. The noise jammer becomes a beacon showing you where the aircraft is.

So it produces errors in timing the location of the aircraft as it's getting different returns at different times?

Time to a radar is range. Range is a require part of a shooting solution. If I can make you track the wrong range you can shoot at me all day and your shells will not hit...except by bad luck.


But in this case the running pulses are generally meant to confuse the radar or operator. Lots of fake pulses running back and forth to cover the location, in time, of the real signal.


T!

 

[Zipper730]

I can't say when the P-61's AGL was first planned, but it was being planned at least in early 1942 (the first mention I've found).

Thanks! It's nice to tie things down to a specific point in time

I'm actually writing up the P-61 versus Mosquito story right now, and there's a lot about the turret and radar coming out of the record.

I look forward to reading that story... just tell me when you finish the book.

Turns out the radar operator (guy in the way back) was being listed as the observer/gunner, and had an aft-facing sight that would have allowed him to take over the turret for defensive fire to the rear.

Joebaugher mentioned something about the gunner acting as a look-out... this seems to conform to the rear-positioned gunner...

(While I've found enough to write my article, I still need to dig some more to discover the first mention of the AGL. I've noticed that the early documents don't even mention the radar - though some 1940/41 papers talk about the "special equipment.")

Yeah, it was uber secret at the time.

I have no idea. I also have no idea how long this persisted. All I know is that it is mentioned as being the opinion of those two gentlemen during the initial meetings which established the projects of the M.I.T. Radiation Laboratory.

Okay

When dealing with new weapon systems it is pretty common for multiple systems to be developed in parallel.

Makes enough sense

I don't know if the aircraft was intended to have an automatic gun laying ability from day one or not, but it would not surprise me if that was at least goal at some point in the life of the system.

Seems logical enough.

I think it is pretty apparent that if the goal from day one was for the P-61 to have an AGL that this was essentially an unrealized goal. While the APG-1/2 was used on about 6 airframes, for whatever reason the majority of airframes never got these systems and I believe the APGs were eventually removed from the ones that did get them.

Why was the APG-1/-2 removed from the P-61? It seems like a remarkable advantage.

As for the concept of a turret-fighter, it seems like a cool idea, but honestly it seems unnecessary: Particularly if the P-61 was built for a 7.3g x 1.5 load from the outset and had agility figures suitable for a day-fighter. The turret seems like it would be difficult to use if it didn't have gun-laying capability, and it'd save weight not just in the turret, but in a crew member as well.

I'm not sure if the RAF had communicated/hinted at its waning interest in turret-fighters (to either Northrop or the USAAF) after the Battle of Britain, though by 1941 they didn't seem to have much interest, and Northrop had completed the basic design by April 1941...

I'm quite surprised it would have been so hard to have developed two gondolas during the prototype stage: Wuzak said it'd probably delay the program, but I'm not sure how many XP-61's were built, and if parts of the gondola were compatible (the nose, lower-sections, wing-body junction point, with the plexiglass cone replaced with a cut down metal cone, and the upper fuselage redesigned), it seems like it'd be easier than building two structures off the bat (and prototypes are generally not production line).

I figure such designs could be modular and tested in parallel on two prototypes.

The data I have indicates the APG-1 itself was a functioning design in early-mid 1944, a result of lessons learned from the SCR-584 and work lightning it to produce the SCR-784.

The P-61 started use around 1944?

Not sure what you mean there.

You said jammers didn't have to put out as much energy for the same effect. I was curious if this was because the jammer has to send out a pulse one way to interfere with the radar, whereas the radar has to send out a pulse that travels considerable distance, bounces off a target with probably some scatter and absorbtion

Sure, you could say that, a specifically defined jiggling glob.

Okay, I understand

Just like your monitor brightness.

So that doesn't affect the radar beam?

Stronger targets appear as taller / bigger targets, not brighter targets.

I never knew that -- I figured it'd just appear brighter

With displays available in WW II any real change in intensity is probably related to the display system, not the size of the target return. Obviously bigger returns can sometimes appear brighter to the user, this is particularly true with PPI type displays

That's the overhead sweeping view?

In time only, and time to a radar amounts to range. So it will move the auto track range gate to a range the real target is not at.

So it makes the aircraft look progressively closer or further away?

So it is tracking nothing at all. You move the jam pulse off of the real target, the radar range goes to the location, range, of the fake target, and then you shut off the fake target, leaving the radar with nothing to track in that location.

It's kind of an electronic version of a simple ruse...

The radar must then start the entire process over again, find the target, start tracking the target, and then the RGPO jammer can start working on the situation again.

So it makes getting a lock a royal pain in the buttocks, and the hope for the aircraft is that they can deny a lock as long as the plane is in range?

Essentially, yes.

One problem with a simple noise jammer is that even if it completely hides the target so you cannot get range, you still have angles to the noise source.

And that's where track-on-jam comes in: You follow the strobes to their source, and suddenly a protection system becomes a bulls-eye...

 

[Token]

I can't say when the P-61's AGL was first planned, but it was being planned at least in early 1942 (the first mention I've found). I'm actually writing up the P-61 versus Mosquito story right now, and there's a lot about the turret and radar coming out of the record. Turns out the radar operator (guy in the way back) was being listed as the observer/gunner, and had an aft-facing sight that would have allowed him to take over the turret for defensive fire to the rear.


(While I've found enough to write my article, I still need to dig some more to discover the first mention of the AGL. I've noticed that the early documents don't even mention the radar - though some 1940/41 papers talk about the "special equipment.")

This fits the timeline of radar development very well.


The MIT Radiation Laboratory was established to do radar work in Oct 1940. It evolved out of the existing National Defense Research Committee as a result of the Anglo-American combined effort that started in June 1940 and led to, among other things, the Tizard Mission in Sep, 1940. Project I for the Rad Lab was AI radar, Project II was AGL radar, and Project III was Navigation Aids. So from late 1940 they were planning on having AGL, they just did not have the details worked out.


They focused on conical scanning as the best way forward for auto tracking radar from very early in the Project II path. And they had Con scan licked by early 1941, testing the first model in May of that year. Once that problem was beaten they knew AGL was happening. In Feb of 1942 the XT-1 radar, what would become the SCR-584, was being tested. In April 1942 it was standardized for production, this was before the first XP-61 flew I think?


Between the Rad Lab, Bell Labs, and Western Electric they moved pretty quick.


From very early 1941 on auto tracking radar was a done deal, it was just a matter of refinement, so planning for an auto tracking radar in the P-61 in late 1940 / early 1941 was ambitious, but not beyond reason or even particularly risky, although the final form would not yet be known. Planning for one in early 1942 was very safe, the techniques had already been proven and were now ready for production in another system, essentially all that had to be done was trim down the equipment for use in an aircraft. At this point a fairly clear picture of what the equipment on the aircraft would eventually look like had to exist.


Besides which, the AI radar, first the SCR-520 and then the lighter, improved, SCR-720, was an even safer bet, and on an earlier timeline. My understanding is that around Jan of 1942 the decision was made for this radar to go into the P-61?


T!

 

[Token]

Why was the APG-1/-2 removed from the P-61? It seems like a remarkable advantage.

I have no idea, maybe they just found out that the way the aircraft was actually used the AGL was less useful than the AI? Remember that because of the Raster based Palmer scan the AGL was typically less suited for general acquisition than the dedicated AI radar. It may track great, but it searches less well.

I figure such designs could be modular and tested in parallel on two prototypes.

In order to be modular you have to standardize some things first. But this was cutting edge, in a lot of areas they did not know what they did not know yet. It may have been more methodical to use standardization and modular equipments, but it was probably quicker to point a couple teams at the projects and give them their heads.

You said jammers didn't have to put out as much energy for the same effect. I was curious if this was because the jammer has to send out a pulse one way to interfere with the radar, whereas the radar has to send out a pulse that travels considerable distance, bounces off a target with probably some scatter and absorbtion

That is basically it, jammers can get away with lower power levels because the energy only has to travel one path, from the target to the radar, instead of from the radar, to the target, get reflected, and back to the radar.


The jammer has one way path loss to deal with, the radar a two way path loss.

So that doesn't affect the radar beam?

No, no effect on the beam at all.

That's the overhead sweeping view?

Yes, the PPI is the radar display most people think of when they think of radar. An overhead view of a rotating trace on a round scope.

So it makes the aircraft look progressively closer or further away?

Lets say changing in range, this change can be rapid or slow, depending on the desired results. Rapid if you want confusion on the part of the radar operator, slow if you want decpetion.

So it makes getting a lock a royal pain in the buttocks, and the hope for the aircraft is that they can deny a lock as long as the plane is in range?

Either deny lock or keep the radar locked on something else, someplace in time that the aircraft is not. That way even if they do shoot they are shooting in the wrong place.

And that's where track-on-jam comes in: You follow the strobes to their source, and suddenly a protection system becomes a bulls-eye...

Sort of, yes. With TOJ you do not have range. Without range you do not have a shooting solution. However, you can get range form other techniques, such as either coincidence or stereoscopic range finding.


T!

 

[Zipper730]

This fits the timeline of radar development very well.

The MIT Radiation Laboratory was established to do radar work in Oct 1940. It evolved out of the existing National Defense Research Committee as a result of the Anglo-American combined effort that started in June 1940 and led to, among other things, the Tizard Mission in Sep, 1940.

I'm curious about something, and while quite beyond the scope of things here: If the NDRC existed to integrate scientific research with military matters: Why was DARPA ever needed?

While I know that DARPA was created after the USSR beat us into space with Sputnik (something I'm surprised we couldn't have done by just taking a sounding-rocket, as the military wanted, and simply giving it a civilian designation; then having it done by non-military personnel): But they are effectively an advanced scientific research agency for military matters, and the atom bomb definitely was one of the most advanced projects of the day.

Project I for the Rad Lab was AI radar, Project II was AGL radar, and Project III was Navigation Aids. So from late 1940 they were planning on having AGL, they just did not have the details worked out.

The plan for gun-laying was for ships, land-basing, and aircraft, correct?

They focused on conical scanning as the best way forward for auto tracking radar from very early in the Project II path. And they had Con scan licked by early 1941, testing the first model in May of that year.

And that pretty much made it clear that the AGL was feasible...

In Feb of 1942 the XT-1 radar, what would become the SCR-584, was being tested. In April 1942 it was standardized for production, this was before the first XP-61 flew I think?

May 26, 1942 was the first flight of the XP-61...

planning for an auto tracking radar in the P-61 in late 1940 / early 1941 was ambitious, but not beyond reason or even particularly risky

Okay...

all that had to be done was trim down the equipment for use in an aircraft. At this point a fairly clear picture of what the equipment on the aircraft would eventually look like had to exist.

So, the size and shape was established by then...

I have no idea, maybe they just found out that the way the aircraft was actually used the AGL was less useful than the AI?

And this had to do with the slower search ability, despite faster track?

In order to be modular you have to standardize some things first. But this was cutting edge, in a lot of areas they did not know what they did not know yet. It may have been more methodical to use standardization and modular equipments, but it was probably quicker to point a couple teams at the projects and give them their heads.

So it was unfeasible because there were too many unknowns?

Regardless: If the non-turreted version was based on the turreted version, wouldn't all the issues be based on the turreted version (the size of the nose, the antenna and electronic box configurations, gun positions, wing/body junctions)?

That is basically it, jammers can get away with lower power levels because the energy only has to travel one path, from the target to the radar, instead of from the radar, to the target, get reflected, and back to the radar.

Gotcha

No, no effect on the beam at all.

Oh, okay

Yes, the PPI is the radar display most people think of when they think of radar. An overhead view of a rotating trace on a round scope.

Gotcha...

Lets say changing in range, this change can be rapid or slow, depending on the desired results. Rapid if you want confusion on the part of the radar operator, slow if you want decpetion.

So basically deception is to be more convincing?

Either deny lock or keep the radar locked on something else, someplace in time that the aircraft is not. That way even if they do shoot they are shooting in the wrong place.

And that keeps the aircrew breathing

Sort of, yes. With TOJ you do not have range. Without range you do not have a shooting solution. However, you can get range form other techniques, such as either coincidence or stereoscopic range finding.

Stereoscopic range finding means that you use two different sensors which allow the range to be extrapolated?

 

[Token]

I'm curious about something, and while quite beyond the scope of things here: If the NDRC existed to integrate scientific research with military matters: Why was DARPA ever needed?

NDRC was dissolved more than 10 years before DARPA was a thing.


In hind-sight a lot of things done in the draw down after WW II were a tad hasty, but the US wanted to quickly step away from a war footing, and to get back to normal.

The plan for gun-laying was for ships, land-basing, and aircraft, correct?

I do not know for sure, but from what I gather it was to apply to all. Obviously it would probably have been fit to areas of the most benefit the quickest.

And this had to do with the slower search ability, despite faster track?

The solution set available to them at the time was faster search with AI point outs and no auto track, or slower search with AI point outs and auto track. For whatever reason they seem to have gone to the faster search option.


I have no data on why they chose this. Any answer I give will be speculation.


However, with that caveat in place I do have decades of radar experience to lean on, and I can see how they may have arrived at such a decision.


With the technologies of the day they had several options. Using a single radar they could, as I said above, have faster scan/search and no track (SCR-720 option), or slower scan/search and auto track to direct guns (APG-1/2 option). But if you chose auto track the technology supported only a single track at a time, and while tracking you could not also search/scan. You were staring at only the single target you had in track.


In the SCR-720 option you could scan search the volume assigned. You could prosecute an attack on a single aircraft while keeping track of the location of any wingmen / other aircraft in the same general volume of space.


In the APG-1/2 option once you went to auto track you could not see any other targets except for the one you were locked on.


The SCR-720 gave you more situational awareness at the cost of no auto track.


The P-61 was primarily a fighter with guns pointed forward and a single turret (on some models). The AGL may have been a good solution for a bomber with multiple turrets and multiple AGLs, each looking in different directions, but was probably less good in the fighter/pursuit/interceptor type application with targets in the front. Also a larger aircraft could better afford different radars for different tasks, say a radar for search and another radar for track, giving you both capabilities at only a size / weight cost.

So it was unfeasible because there were too many unknowns?


Regardless: If the non-turreted version was based on the turreted version, wouldn't all the issues be based on the turreted version (the size of the nose, the antenna and electronic box configurations, gun positions, wing/body junctions)?

When the P-61 was started, when the specifications were let and the aircraft starting development, did they know the constraints for the AI vs AGL radar would be similar? Since they had no idea what path the AGL would take they could not realistically have known.


It is likely they said something along the lines of "we know the AI will fit in this space, lets try to build an AGL in the same space". But other than size constraints they probably, at least initially, had little idea what the AGL would look like.


You can't standardize if you don't know even the basics of a design. Or rather you can standardize blind, at the potential cost of failure. Your probability of success increases sharply if you give design team their heads and as few constraints as possible. That is, if you can afford to let the design be based only on performance criteria, with no regards to how it looks to your logistics chain.

So basically deception is to be more convincing?

Two different applications. One is used when you don't care if they know you are there or how many of you there are but don't want them to be able to shoot you. The other is there when you want to obscure your numbers or totally prevent tracks of any kind.

Stereoscopic range finding means that you use two different sensors which allow the range to be extrapolated?

Coincident and stereoscopic range finding are visual techniques of range finding. In a jamming situation, particularly noise jamming, you can TOJ the jammer but you have no range information. Fuse your angles information from the radar with range information from a visual range source and you now have a firing solution.


T!

 

[Zipper730]

NDRC was dissolved more than 10 years before DARPA was a thing.

That I didn't know.

I'm guessing RAND took over from 1945-1958?

In hind-sight a lot of things done in the draw down after WW II were a tad hasty, but the US wanted to quickly step away from a war footing, and to get back to normal.

Makes enough sense

I do not know for sure, but from what I gather it was to apply to all. Obviously it would probably have been fit to areas of the most benefit the quickest.

Okay

The solution set available to them at the time was faster search with AI point outs and no auto track, or slower search with AI point outs and auto track. For whatever reason they seem to have gone to the faster search option.

I have no data on why they chose this. Any answer I give will be speculation.

Well, speculation is a good start.

However, with that caveat in place I do have decades of radar experience to lean on, and I can see how they may have arrived at such a decision.

With the technologies of the day they had several options. Using a single radar they could, as I said above, have faster scan/search and no track (SCR-720 option), or slower scan/search and auto track to direct guns (APG-1/2 option). But if you chose auto track the technology supported only a single track at a time, and while tracking you could not also search/scan. You were staring at only the single target you had in track.

In the SCR-720 option you could scan search the volume assigned. You could prosecute an attack on a single aircraft while keeping track of the location of any wingmen / other aircraft in the same general volume of space.

So it had to do with being able to maintain the bigger picture...

In the APG-1/2 option once you went to auto track you could not see any other targets except for the one you were locked on.

Was this known to be a limitation as it was being developed? Because I'm just thinking about the whole turret-fighter, and it seems completely absurd without auto-track (and the limitations make auto-track less than ideally desirable).

When the P-61 was started, when the specifications were let and the aircraft starting development, did they know the constraints for the AI vs AGL radar would be similar?

You wrote in reply #112

"Between the Rad Lab, Bell Labs, and Western Electric they moved pretty quick.

From very early 1941 on auto tracking radar was a done deal, it was just a matter of refinement, so planning for an auto tracking radar in the P-61 in late 1940 / early 1941 was ambitious, but not beyond reason or even particularly risky, although the final form would not yet be known. Planning for one in early 1942 was very safe, the techniques had already been proven and were now ready for production in another system, essentially all that had to be done was trim down the equipment for use in an aircraft. At this point a fairly clear picture of what the equipment on the aircraft would eventually look like had to exist."

You can't standardize if you don't know even the basics of a design. Or rather you can standardize blind, at the potential cost of failure.

I figured, what you said in reply #112, that they basically had a fairly clear picture of what the radar-equipment would look like.

Your probability of success increases sharply if you give design team their heads and as few constraints as possible.

I can't argue with you on that.

I am still puzzled as to a few things

1. According to wuzak (reply #27), the RAF had started to lose interest in the turret-fighter concept (at least as a day fighter) during the battle of Britain: Their interest in a turret-equipped night-fighter clearly persisted through the end of 1940.
2. When did the RAF lose interest in a turret-equipped night-fighter, and did they ever express such a loss of interest to the USAAF or Northrop?
   
3. Being that the SCR-720 was completely useless for the turret-fighter mission: I would assume that the turret was intended for the gun-laying configuration? If not, the turret seemed largely useless otherwise...

Two different applications. One is used when you don't care if they know you are there or how many of you there are but don't want them to be able to shoot you. The other is there when you want to obscure your numbers or totally prevent tracks of any kind.

So confusion is better to obscure numbers, deception is to avoid getting shot once they realize you are present?

Coincident and stereoscopic range finding are visual techniques of range finding. In a jamming situation, particularly noise jamming, you can TOJ the jammer but you have no range information. Fuse your angles information from the radar with range information from a visual range source and you now have a firing solution.

Okay, so you use the optics to triangulate the range coordinate and the X & Y from the radar and pop off the SAM?

 

[Zipper730]

wuzak,

I was wondering about something (this doesn't have to do with radar): There was a conscious decision to put a twin-stage supercharger in the P-61 instead of a turbocharger despite the fact that they projected a considerably higher critical and cruise altitude, and a greatly higher top-speed (this was verified later on in practice with the P-61C).

Why did they forgo this?

 

[Shortround6]

The British may have pitched the turret idea to the US during the initial design stages, as sort of a joint project. The British were buying or getting a lot of different aircraft lend lease and a lot of other equipment went through some sort of common requirement/use filter.

As far as not using the turbo goes, what was the intended target? How many German 1942-43 bombers could outrun or fly higher than a P-61? The Mock-up was inspected in April of 1941, further modifications followed and a letter of intent for 100 aircraft was initiated on Dec 24th 1941, 3 days after the first P-47B (really the 2nd prototype, it takes almost 3 months for the next P-47B to roll out the door) flies.
Considering the problems the USAAC had sorting out the turbo installation in the P-47 during 1942 perhaps it is a good thing they didn't use it.
And please note the P-61C used the same powerplant as the P-47M and N. Not the same engine and turbo used in the P-47B, C and D.
The difference in performance would have been nowhere near as great.

 

[Koopernic]

wuzak,

I was wondering about something (this doesn't have to do with radar): There was a conscious decision to put a twin-stage supercharger in the P-61 instead of a turbocharger despite the fact that they projected a considerably higher critical and cruise altitude, and a greatly higher top-speed (this was verified later on in practice with the P-61C).

Why did they forgo this?

I'm guessing the following. The two stage 4 speed PW R-2800 was a "navy engine" and the navy had already funded its development for the Corsair and Hellcat. This made it cheap and simple to integrate into the Blackwidow because there would be no complicated turbo ducting requiring exotic alloys. The slight improvement in altitude performance may not have been justified given the kind of threats the Germans and Japanese would have fielded.

There would be one exception. The BMW 801TQ engined Ju 488 V401 which was nearing completion in mid July 1944. With the interim turbo supercharged BMW 801TJ engine would have been able to operate at 48,500ft. With the improved BMW 801TQ perhaps 51,500ft. The Ju 388 also might have operated at 44,000ft and the night fighter version had been intended to deal with threats such as the B-29.

Note there were 3 Ju 488 with BMW801TJ, Jumo 222A3/B3 and Jumo 222E3/F3.

The Ju 388 with a Jumo 213E1 was slightly faster but at about 4000ft lower service ceiling.

Out of interest the Germans did the reverse. Most BMW 801 were simple single stage two speed engines (with a 3rd speed possibly to come out) however they did developed the turbo charged BMW 801TJ which saw some limited service on Ju 388 and Ju 88S4. The 801TJ was 'an integral power egg that was unsuitable for fighter use because its large coaxial intercooler diameter and because its exhaust ejected upward into what would be the cockpit wind shield.

The Luftwaffe steadfastly refused to fund development of a 801 version for fighters probably because the work of integrating a turbo charger is so difficult it requires a new aircraft; hence BMW developed the BMW 801R with two stage 4 speed supercharger with intercooler. It would have powered variants of the Ta 152C.

We can surmise that moving a turbo charger around was hard work.

Having said that, the Blackwidow obviously easily could have accommodated a P-47 or P-38 setup given the space in its tail booms

 

[Zipper730]

The British may have pitched the turret idea to the US during the initial design stages, as sort of a joint project. The British were buying or getting a lot of different aircraft lend lease and a lot of other equipment went through some sort of common requirement/use filter.

Had they ever expressed a lack of interest in the turret? Retroactively, I know their interest had faded with time...

As far as not using the turbo goes, what was the intended target? How many German 1942-43 bombers could outrun or fly higher than a P-61?

One could use the same argument for the NF Mosquito... as time goes on, it's generally natural to try and squeeze as much speed and altitude out of a design. This is for numerous reasons, the first being that threats improve with time.

Additionally, I figure a turbocharger would add range

The Mock-up was inspected in April of 1941, further modifications followed and a letter of intent for 100 aircraft was initiated on Dec 24th 1941, 3 days after the first P-47B (really the 2nd prototype, it takes almost 3 months for the next P-47B to roll out the door) flies.
Considering the problems the USAAC had sorting out the turbo installation in the P-47 during 1942 perhaps it is a good thing they didn't use it.

I honestly have little knowledge as to the P-47's issues with its turbocharger.

And please note the P-61C used the same powerplant as the P-47M and N. Not the same engine and turbo used in the P-47B, C and D.

That's a good observation, I missed that: Regardless, the P-47M was available by 1944 right?

The difference in performance would have been nowhere near as great.

To factor in performance, one would compute either TAS/horsepower^3 or IAS/horsepower^3 right?

I'm guessing the following. The two stage 4 speed PW R-2800 was a "navy engine"and the navy had already funded its development for the Corsair and Hellcat.

True, but the P-47 flew May 6, 1941...

This made it cheap and simple to integrate into the Blackwidow because there would be no complicated turbo ducting requiring exotic alloys.

The exotic alloys are the turbine right? As for the ducting, it seemed to have worked on the P-61C...

Having said that, the Blackwidown obviously easily could have accommodated a P-47 or P-38 setup given the space in its tail booms

The booms are an advantage