A-6 Intruder, Any Weather, Any Time....

[XBe02Drvr]

I thought they were tracking him but didn't fire because they didn't want a KAL007 all over again...

Ground radars were getting occasional paints on him, enough to plot his track, but jets in the air vectored to his location failed to see him, and their look-down radars weren't good enough to dredge him out of the clutter. True, they didn't want another KAL, but they had to verify it was truly some innocent civilian, and not a sophisticated stealthy attack of some kind, and they were having a hard time doing it.
That Cuban AN-2 cropduster that dropped in on us could have been lugging a pretty sizable nuke instead of the 28 refugees that were packed into its pesticide hopper.
Cheers,
Wes

 

[XBe02Drvr]

Moscow would have suffered a crater, some serious blast-forces and a decent firestorm that would have burned up a sizeable portion of the city.

I don't think a 172 could carry that potent of a nuke, but if he detonated a battlefield nuke on the steps of the Kremlin, he might have made a good stab at decapitating the USSR.
Cheers,
Wes

 

[Zipper730]

Ground radars were getting occasional paints on him, enough to plot his track, but jets in the air vectored to his location failed to see him

Okay

their look-down radars weren't good enough to dredge him out of the clutter.

I thought pulse-doppler radars work by measuring relative velocities and using a signal filter to remove anything which isn't moving above a certain velocity from the scope... can't you just turn off the filter?

That Cuban AN-2 cropduster that dropped in on us could have been lugging a pretty sizable nuke instead of the 28 refugees that were packed into its pesticide hopper.

So they were trying to flee?

I don't think a 172 could carry that potent of a nuke

Well, if you look at the size of the W80 and W84 nuclear warhead that could be fitted into the older BGM-109's, it could almost certainly fit in a C-172: I'm not sure if the weight is compatible with the plane, but that is beyond my pay-grade.

Regardless the W84 has a yield up to 150 kt, and the W80 up to 200 kt, and a 14-22 kt yield could do the damage I described to a city in the manner I described.

if he detonated a battlefield nuke on the steps of the Kremlin, he might have made a good stab at decapitating the USSR.

Yep...

 

[XBe02Drvr]

I thought pulse-doppler radars work by measuring relative velocities and using a signal filter to remove anything which isn't moving above a certain velocity from the scope... can't you just turn off the filter?

That filter is a Moving Target Indicator or MTI, and on western radars it's generally adjustable as to the threshold velocity for target depiction. The Soviets back in cold war days tended to be more rigid about stuff like this. Like the rest of their people, pilots generally had less autonomy, and had systems with less flexibility and fewer adjustable parameters.
I used to hang out in NAS Approach Control on slow nights playing cards with the controllers, and one night I turned the MTI speed gate down to 10 MPH, and suddenly the scope was full of targets! Every vehicle on the Overseas Highway, every fishing boat out on the water, all traffic in downtown Key West was suddenly a visible target.
I have no trouble believing those interceptors couldn't find a low flying slow moving Cessna.
Cheers,
Wes

 

[Zipper730]

That filter is a Moving Target Indicator or MTI, and on western radars it's generally adjustable as to the threshold velocity for target depiction.

Okay

The Soviets back in cold war days tended to be more rigid about stuff like this. Like the rest of their people, pilots generally had less autonomy, and had systems with less flexibility and fewer adjustable parameters.

Didn't work out too good for them...

I used to hang out in NAS Approach Control on slow nights playing cards with the controllers, and one night I turned the MTI speed gate down to 10 MPH, and suddenly the scope was full of targets! Every vehicle on the Overseas Highway, every fishing boat out on the water, all traffic in downtown Key West was suddenly a visible target.

So I guess the radar's range of view is partially downward and partially upward because if it didn't aim downward you wouldn't see cars and fishing boats...

I have no trouble believing those interceptors couldn't find a low flying slow moving Cessna.

They either saw nothing or everything

 

[XBe02Drvr]

So I guess the radar's range of view is partially downward and partially upward because if it didn't aim downward you wouldn't see cars and fishing boats...

Think about it! It's an air traffic control radar. It has to show all targets from surface to almost straight overhead, and do it without having to be dialed up and down like an air intercept radar. It has a vertically oriented fan shaped beam, unlike the pencil beam that airborne radars have. It's all about power. An airplane can't lug a hydroelectric power plant around with it.
Cheers,
Wes

 

[Token]

Jets vectored to intercept him failed to see him. SAM sites couldn't get a launchable target return off him. He was too slow for their Moving Target Indicators, just like the Cuban AN-2 that gave us a surprise visit one morning at the Naval Air Station.

For a couple of reasons I am not going to buy he was too slow for all of their MTIs, regardless of what has been published. Lets go with the easiest to confirm, most of the systems ringing Moscow have (and had at the time, 1987) advertised abilities to engage helicopters. The second is that not every radar has MTI, that gets harder to confirm for specific radars as they tend to not talk about those details all that often.


I could easily buy that some of the acquisition radars would have trouble pointing him out to the shooters, possibly because of their MTI's, but the shooters should have had no problem with him. And if the Acqs could not see him, after they knew something was there (and they would have to know something was out there to vector aircraft), it was due to poor operators. MTI is generally adjustable, and can also be turned off totally. A good op knows to flip it off and on when looking for a known present but undetected aircraft.


As a radar technician and then engineer I have a few decades of military radar experience, and have ALWAYS been able to track a Cessna if I was actually looking for it. Birds, party balloons, cars on the road, swarms of bugs, whatever. If I know something is probably there I can get hits on it, particularly with a shooter.

I thought pulse-doppler radars work by measuring relative velocities and using a signal filter to remove anything which isn't moving above a certain velocity from the scope... can't you just turn off the filter?

The description really depends on if you are talking about ground radar (such as ACQs and SAM sites) or airborne radars, they work slightly differently


Ground based Pulse Doppler radars do work roughly as you describe, but don't think of it as relative velocity, rather it is radial velocity, velocity inbound or outbound. A 450 knot target crossing on a tangent will have zero radial velocity at that tangential point, and so will have no Doppler, regardless of its velocity.


Further, ground based PD radars tend to have a filter around zero Doppler. Their returned signal from ground clutter will be around zero Doppler, the Earths surface not having any significant radial velocity to other parts of the Earths surface. So at zero Doppler the returns will be huge. It is relatively simple to design a sharp, narrow, bandstop (notch) filter that blocks your transmitted frequency and any small Doppler shifts to that frequency. So now you have killed ground clutter and very slow moving targets. Typical velocities for easily designed filters at X band are on the order of 25 to 40 meters / second. So ground clutter is rejected, having no Doppler shift, and targets below + and - 25 meters per second radial velocity are also killed.


With such a system and a 25 m/s notch a target under 50 knots would not be seen, or at least not be seen well.


For airborne PD radars it is a similar setup, but now instead of a fixed filter at the radar transmitted frequency it is an adjustable filter. And now it is relative velocity.


This notch filter is not necessarily an MTI although it ends up working as an affective one, and it can be exploited by aircraft. I think every driver knows that if you are being painted by a PD ground radar going to the notch (zero radial velocity to the radar, such as turning to put him on your wingtip) may cause him issues. Of course, modern radars are a bit better, today it is common for a PD radar to change techniques when looking for a target at or near zero Doppler.


Not all PD radars have a notch filter. Whether they have a notch filter or not the end result is the same. If they do notch out ground return then they cannot see relatively small targets at or near zero Doppler. If they do not notch it out then they can't see the targets anyway as they are lost in ground clutter. Filter or not, the scopes can be made clear by only displaying Doppler shifts above a certain value.


Not all radars are PD though. Those radars must have some way to kill ground clutter. And that is an MTI.

That filter is a Moving Target Indicator or MTI, and on western radars it's generally adjustable as to the threshold velocity for target depiction.

I have worked on radars from a lot of countries, and every radar I have seen with MTI includes an ability to adjust it.

T!

 

[Token]

Think about it! It's an air traffic control radar. It has to show all targets from surface to almost straight overhead, and do it without having to be dialed up and down like an air intercept radar. It has a vertically oriented fan shaped beam, unlike the pencil beam that airborne radars have. It's all about power. An airplane can't lug a hydroelectric power plant around with it.

But even ATC radars don't normally look downwards, that would just increase the ground clutter the MTI would have to reduce. However, there are always sidelobes. So even when they shape the beam to be a fan looking from horizon upwards some energy goes down. And since those down targets tend to be much closer they get significant energy reflected off them.


Most of these ground tracks, trucks and cars and such, are actually from sidelobes rather than the main lobes.

T!

 

[XBe02Drvr]

Hey Token, thanks for the info. It's great to have an engineer's perspective. My understanding and experience of radar is largely from a user's point of view. I operated, maintained and instructed on a radar interception weapons system trainer in the Navy back in the early seventies (F-4B). Since then it's just been airborne weather radar thunderbumper dodging. (as the dodger, not the tech) I drank a lot of beers with the radar techs in the fighter squadron, the air test and evaluation squadron, ATC, and the Hawk missile sites. They all loved to talk shop.
I don't know what the timeframe of your experience is, but back in the day precise notch filters were not so easy as in the modern era. Phase lock loops were big, clunky,complicated, and expensive. Putting four of them in an IC chip was still in the future. I built one with TUBES and discrete components in Avionics A School.
Did you get any hands-on experience with Soviet radars? I was told they were unsophisticated and clunky compared to the west. Apparently the Wild Weasels and Iron Hands were fairly adept at zero-dopplering the Fansong? (I think it was) radar of the SA-2 SAM sites.
It's pretty amazing what a good operator can do with a good radar, and here in the west we have plenty of good operators. But I suspect such was not so much the case in the USSR, with their two year conscript military system. And the Air Defense System around Moscow in 1987 was quite likely like Oahu on December 6, 1941; peacetime mentality, not much recent practice, no "bunker mentality", low espirit d'corps morale, and probably a fair bit of corruption. Under these circumstances virtuoso radar operator performance may not be a sure thing. It's one thing to have a specified anti-helicopter capability, and quite another to reliably perform it on any given day.
Appreciate your take on it.
Cheers,
Wes

 

[Zipper730]

Think about it! It's an air traffic control radar. It has to show all targets from surface to almost straight overhead, and do it without having to be dialed up and down like an air intercept radar. It has a vertically oriented fan shaped beam, unlike the pencil beam that airborne radars have. It's all about power. An airplane can't lug a hydroelectric power plant around with it.

So the greater power-output makes it better able to see through clutter?

For a couple of reasons I am not going to buy he was too slow for all of their MTIs, regardless of what has been published. Lets go with the easiest to confirm, most of the systems ringing Moscow have (and had at the time, 1987) advertised abilities to engage helicopters.

Good point.

I could easily buy that some of the acquisition radars would have trouble pointing him out to the shooters, possibly because of their MTI's, but the shooters should have had no problem with him. And if the Acqs could not see him, after they knew something was there (and they would have to know something was out there to vector aircraft), it was due to poor operators. MTI is generally adjustable, and can also be turned off totally. A good op knows to flip it off and on when looking for a known present but undetected aircraft.

So a lot of them were not properly proficient?

Ground based Pulse Doppler radars do work roughly as you describe, but don't think of it as relative velocity, rather it is radial velocity, velocity inbound or outbound. A 450 knot target crossing on a tangent will have zero radial velocity at that tangential point, and so will have no Doppler, regardless of its velocity.

Wait, so if you were to put the plane into a huge wide ass turn, it'll disappear off their radar?

t is relatively simple to design a sharp, narrow, bandstop (notch) filter that blocks your transmitted frequency and any small Doppler shifts to that frequency.

I'm not sure how that even works in concept, provided it's not classified.

For airborne PD radars it is a similar setup, but now instead of a fixed filter at the radar transmitted frequency it is an adjustable filter. And now it is relative velocity.

Okay, so that's the linear velocity of your plane and theirs?

Not all radars are PD though. Those radars must have some way to kill ground clutter. And that is an MTI.

So it's not called an MTI on a PD radar?

But even ATC radars don't normally look downwards, that would just increase the ground clutter the MTI would have to reduce. However, there are always sidelobes.

So there are some reflections that manage to make it through the clutter and into the beam?

 

[XBe02Drvr]

So the greater power-output makes it better able to see through clutter?

No, it's not about seeing through the clutter, that's what the MTI is for. It takes a lot more power to drive that fan-shaped beam of the ATC radar than the pencil beam an air intercept radar has. When a broad beam radar such as Airborne Early Warning or Surface Search is carried in an aircraft, it's generally something with four engines and four generators such as an EC-121, C-130, P-3 or the like. Our Navy Flying Club mechanic was a retired chief who had worked on E-1s back in the pre-MacNamara days when they were called WF-2s ("Willy Fudd" or "stoof-with-a-roof"). He said that big dish on top of the plane consumed every ampere the two overtaxed, overheated generators could crank out. Lose one generator and the "eye in the sky" is suddenly blind.
Cheers,
Wes

 

[Token]

Hey Token, thanks for the info. It's great to have an engineer's perspective. My understanding and experience of radar is largely from a user's point of view. I operated, maintained and instructed on a radar interception weapons system trainer in the Navy back in the early seventies (F-4B).

While Navy active duty I did SIGINT/ELINT from the 70's to the mid 80's. Mid 80's I left the Navy and shifted over to the noisy side of the house, as a radar operator and tech. And a few years later I shifted to engineering on the same side of the house for T&E facilities. From one end or the other I have done EW since the 70's.

Since then it's just been airborne weather radar thunderbumper dodging. (as the dodger, not the tech) I drank a lot of beers with the radar techs in the fighter squadron, the air test and evaluation squadron, ATC, and the Hawk missile sites. They all loved to talk shop.

I don't know what the timeframe of your experience is, but back in the day precise notch filters were not so easy as in the modern era. Phase lock loops were big, clunky,complicated, and expensive. Putting four of them in an IC chip was still in the future. I built one with TUBES and discrete components in Avionics A School.

My professional electronics experience is from the mid 70's to today. My hobby electronics experience goes back to 1967 when I was first licensed as a ham.


Tracking notch filters (as used by airborne radars) used to be difficult, but tight fixed notch filters (as used by ground based radars) have been pretty easy for a long time. Today it is all done in DSP anyway.

Did you get any hands-on experience with Soviet radars? I was told they were unsophisticated and clunky compared to the west. Apparently the Wild Weasels and Iron Hands were fairly adept at zero-dopplering the Fansong? (I think it was) radar of the SA-2 SAM sites.

If I did have such experience I probably would not discuss it in a forum like this one.


The Fan Song is the radar for the SA-2, but it is not a Doppler tracker.

It's pretty amazing what a good operator can do with a good radar, and here in the west we have plenty of good operators. But I suspect such was not so much the case in the USSR, with their two year conscript military system. And the Air Defense System around Moscow in 1987 was quite likely like Oahu on December 6, 1941; peacetime mentality, not much recent practice, no "bunker mentality", low espirit d'corps morale, and probably a fair bit of corruption. Under these circumstances virtuoso radar operator performance may not be a sure thing. It's one thing to have a specified anti-helicopter capability, and quite another to reliably perform it on any given day.

Appreciate your take on it.

I would not be so quick to discredit Russian operators, even in the USSR period of time. While you are correct on their enlisted folks, because of their mandatory service and essentially no professional NCO corps, they tended to use officers in positions we would have a senior NCO. These were trained, well trained, individuals with at least Associates equivalent levels education and training on their systems.


T!

 

[Token]

So the greater power-output makes it better able to see through clutter?

Higher power levels increase ground clutter return, requiring better MTI performance. Sometimes lower power results in a higher probability of detection, partially because you reduce the false alarm rates.


Radar design is not a simple case of "more power better", it is a balancing act of power, receiver sensitivity, clutter rejection (whatever technique is used), and a dozen + other factors. And then you must match radar performance to the anticipated target set. A radar that excels at detecting high speed, high altitude, targets may be crap for low speed, low altitude targets.

So a lot of them were not properly proficient?

Who said that? While this is opinion I believe there was a large factor of they had not been trained for this specific scenario. A single, low speed, low altitude, small'ish RCS, target not presenting any of the classic attack profiles. They may have been very adequately trained to kill US bombers and fighters on low-low-low attack profiles, but that will not be the same as a Cessna hedge hopping.

Wait, so if you were to put the plane into a huge wide ass turn, it'll disappear off their radar?

If the radar is a pulse Doppler radar with no pulse compression techniques or alternate techniques to track zero Doppler targets then yes, it is that simple. Many of the early PD radars fit this description. The target will either disappear into the notch filter or it will disappear into the non-Doppler shifted energy transmitted by the radar. Alternately the radar can transition out of a PD mode of operation, into a more traditional simple return mode, while the target is near zero Doppler, but then it has lost the benefits of being a PD system during this period.


But it is very important to remember that not all radars are PD, in fact most are not. This is particularly true of long range acquisition radars, they tend to be more simple in nature, or at least they did until relatively recently.

I'm not sure how that even works in concept, provided it's not classified.

Not classified at all, the basic physics are well documented in many texts. The specifics of a certain system might be classified, but that is true of many things regarding radar design / performance.

Okay, so that's the linear velocity of your plane and theirs?

It basically breaks down to relative radial velocity for airborne radars and absolute radial velocity for ground based radars.

So it's not called an MTI on a PD radar?

A PD radar can have MTI, but the basic zero Doppler rejection of a PD radar is not generally referred to as MTI. In other words, just because it kills zero Doppler targets does not make it an MTI.

So there are some reflections that manage to make it through the clutter and into the beam?

The beam of an antenna is not perfect. I mean if the radar makes a one degree pencil beam pointed at a specific target all of the transmitted energy does not just go into that beam. Some of it goes into side lobes and back lobes. A side lobe may be 30+ dB down in power from the main lobe, but that still equates to real power in the sidelobe.

Lets take a notional antenna with a one degree beamwidth (@3dB points) looking 1 degree above the horizon. This is not a real radar, just throwing some numbers out to demonstrate the issues. At 1 degree Elevation and a 1 degree beamwidth (+/- 0.5 dB either side of center) in a perfect world the radar would not see the ground, as all the ground is outside the beam.


But looking at the above picture you can see a couple of things.


First is that there is energy in the main beam outside the 3 dB points. Second is that there are other lobes of power besides the main beam. These other lobes are called side lobes. For our notional example lets call the first side lobe 40 dB down in power. If that main lobe is pointed slightly above the horizon then that first side lobe on the lower side is pointed right at the ground out in front of the radar.


If the transmitter has 200 kW of power and the antenna gain is 40 dBi (typical of a 1 degree beamwidth) then the ERP (Effective Radiated Power, or the transmitter power plus the gain of the antenna) in the main beam will be about 2,000,000 kW, or 10000 times the transmitter power. Crap ton of power, yes?


But the first side lobe, at 40 dB less gain, still has some very real power. 0 dB gain still makes it unity gain. So that 200 kW transmitter power becomes an ERP of about 200 kW pointed at the ground in front of the radar.


And now the inverse square law comes into play. The main beam is looking at and for smaller targets many miles away, while the side lobe, because of the geometry, is looking at targets not so far away. Due to the inverse square law the targets not so far away will be much stronger than the targets far away. There are many techniques radars can use to reduce the impact of this near clutter, but the fact remains it is there and you must do something about it. So it is not unusual to be able to see targets resultant from these side lobes or other unintended reflections.


T!

 

[Zipper730]

It takes a lot more power to drive that fan-shaped beam of the ATC radar than the pencil beam an air intercept radar has.

Because you have a wider beam...

When a broad beam radar such as Airborne Early Warning or Surface Search is carried in an aircraft, it's generally something with four engines and four generators such as an EC-121, C-130, P-3 or the like.

Okay, I understand.

 

[Token]

Because you have a wider beam...

It is a physics problem.


Think of it this way, the total energy you can transmit, or receive (it works both ways), is defined as power density on the surface of a sphere. With no gain at all, no bias towards a certain direction, the energy on the surface of the sphere is equal all over the place, a uniform power density across the whole surface of the sphere. When you have an antenna that does this it is called an isotropic radiator, and for the most part these exist in the theoretical world.


If I want some directionality, I want the energy to look in a specific direction, I have to pull energy from some parts of the sphere and direct it towards other parts of the sphere. The total power across the entire surface of the sphere is the same, but now there is less in some areas and more in other areas, directing the energy in desired directions. This is antenna gain.


If I take all of the energy across the sphere and only point it to one half of the sphere, so that only a hemisphere is now powered (wrong term, but I am painting a verbal picture here) then I have taken the power from the unwanted hemisphere and added it to the power in the desired hemisphere. In theory the power density in the desired hemisphere is now twice as strong, and there is nothing in the unwanted hemisphere. This is a 3 dB increase in antenna gain over the isotropic radiator. All of the power is now concentrated in one half the area of the surface of the sphere. To increase 3 more dB I would have to half the area again, so that all of the power is now concentrated in one quarter the surface area of the sphere, this would be 6 dB of antenna gain over the isotropic.


A shooter, a fire control radar, wants to maximize energy on a given target, minimize power on unwanted targets, and also maximize angular resolution. So these radars tend to have very small beam widths and correspondingly higher gain. Think of it as looking at the World through a straw. You really only see the one thing you are concentrating on, and little else. So typically a shooter does not function alone, it typically does a poor job of finding unknown targets, but a great job of seeing targets initially detected by some other source (yes, one radar can do all of this, but typically by having different modes of operation, some modes optimized towards search / acq and others optimized towards engaging the target).


But a search radar wants to look over a wider instantaneous area, to increase its probability of finding an unknown target. Angular resolution is less important because you are not directing a missile / projectile to a point in space, requiring high accuracy, you are only determining that something is in an area. So the beams are wider in some aspect. They may be tall thin beams, wide short beams, etc, but they tend to spread the energy out over a larger part of the sphere. This means they have, by definition, less possible gain. So all other things being equal they have to make up for this lower gain by having more power.


There are a whole lot of variables not even vaguely touched on here, like the fact that shooters are typically higher in frequency because of how frequency plays in gain and antenna size, or the fact that space loss decreases as frequency goes down so search / acq radars tend to be lower in frequency, etc. But you have to start someplace.


Sorry man, we have drifted hard from the A-6 thing.

T!

 

[XBe02Drvr]

Yes, but it's all relevant. The A-6 was nothing if not an airborne radar station. This is good stuff.
Cheers,
Wes

 

[Zipper730]

Tracking notch filters (as used by airborne radars) used to be difficult, but tight fixed notch filters (as used by ground based radars) have been pretty easy for a long time. Today it is all done in DSP anyway.

DSP?

I would not be so quick to discredit Russian operators, even in the USSR period of time. While you are correct on their enlisted folks, because of their mandatory service and essentially no professional NCO corps, they tended to use officers in positions we would have a senior NCO. These were trained, well trained, individuals with at least Associates equivalent levels education and training on their systems.

Understood

Sometimes lower power results in a higher probability of detection, partially because you reduce the false alarm rates.

Fascinating...

A radar that excels at detecting high speed, high altitude, targets may be crap for low speed, low altitude targets.

It might not be able to filter out ground-clutter?

If the radar is a pulse Doppler radar with no pulse compression techniques or alternate techniques to track zero Doppler targets then yes, it is that simple.

Wow...

Alternately the radar can transition out of a PD mode of operation, into a more traditional simple return mode, while the target is near zero Doppler, but then it has lost the benefits of being a PD system during this period.

So it would pick up stationary targets at altitude but miss things in the weeds?

Not classified at all, the basic physics are well documented in many texts.

Okay

It basically breaks down to relative radial velocity for airborne radars and absolute radial velocity for ground based radars.

Okay

A PD radar can have MTI, but the basic zero Doppler rejection of a PD radar is not generally referred to as MTI.

So it's just called zero Doppler rejection?

The beam of an antenna is not perfect. I mean if the radar makes a one degree pencil beam pointed at a specific target all of the transmitted energy does not just go into that beam. Some of it goes into side lobes and back lobes. A side lobe may be 30+ dB down in power from the main lobe, but that still equates to real power in the sidelobe.

Lets take a notional antenna with a one degree beamwidth (@3dB points) looking 1 degree above the horizon. This is not a real radar, just throwing some numbers out to demonstrate the issues. At 1 degree Elevation and a 1 degree beamwidth (+/- 0.5 dB either side of center) in a perfect world the radar would not see the ground, as all the ground is outside the beam.

Makes sense

If the transmitter has 200 kW of power and the antenna gain is 40 dBi (typical of a 1 degree beamwidth) then the ERP (Effective Radiated Power, or the transmitter power plus the gain of the antenna) in the main beam will be about 2,000,000 kW, or 10000 times the transmitter power.

Is that like energy density? The energy per area or volume?

Due to the inverse square law the targets not so far away will be much stronger than the targets far away. There are many techniques radars can use to reduce the impact of this near clutter, but the fact remains it is there and you must do something about it.

Which would basically somehow reduce the exponential effect to something that's more linear?

It is a physics problem.

Think of it this way, the total energy you can transmit, or receive (it works both ways), is defined as power density on the surface of a sphere. With no gain at all, no bias towards a certain direction, the energy on the surface of the sphere is equal all over the place, a uniform power density across the whole surface of the sphere.

Which means the antenna is transmitting equally in all directions?

When you have an antenna that does this it is called an isotropic radiator, and for the most part these exist in the theoretical world.

If I want some directionality, I want the energy to look in a specific direction, I have to pull energy from some parts of the sphere and direct it towards other parts of the sphere. The total power across the entire surface of the sphere is the same, but now there is less in some areas and more in other areas, directing the energy in desired directions. This is antenna gain.

I vaguely follow you in concept...

In theory the power density in the desired hemisphere is now twice as strong, and there is nothing in the unwanted hemisphere. This is a 3 dB increase in antenna gain over the isotropic radiator.

I follow...

All of the power is now concentrated in one half the area of the surface of the sphere.

Followed...

To increase 3 more dB I would have to half the area again, so that all of the power is now concentrated in one quarter the surface area of the sphere, this would be 6 dB of antenna gain over the isotropic.

And that has to do with a smaller area with the same energy?

A shooter, a fire control radar, wants to maximize energy on a given target, minimize power on unwanted targets, and also maximize angular resolution. So these radars tend to have very small beam widths and correspondingly higher gain.

Ok

Think of it as looking at the World through a straw.

All at once anyway...

But a search radar wants to look over a wider instantaneous area, to increase its probability of finding an unknown target. Angular resolution is less important because you are not directing a missile / projectile to a point in space, requiring high accuracy, you are only determining that something is in an area.

So wide-angle is good for searching, narrow angle for tracking?

So the beams are wider in some aspect. They may be tall thin beams, wide short beams, etc, but they tend to spread the energy out over a larger part of the sphere. This means they have, by definition, less possible gain. So all other things being equal they have to make up for this lower gain by having more power.

Ok

Sorry man, we have drifted hard from the A-6 thing.

No biggie

Yes, but it's all relevant. The A-6 was nothing if not an airborne radar station. This is good stuff.

Good point

 

[Token]

DSP?

DSP is Digital Signal Processing, or Processor. Just a way of saying it is done digitally. Filters, delay lines, and phase detectors that would have been hardware in the past are now digital. You digitize the IF (Intermediate Frequency) of the radar and process it all in software

It might not be able to filter out ground-clutter?

That may be one aspect, but don't fixate on ground clutter, that is only one factor of radar performance.


Examples of what I am talking about might be centered on the PRF (Pulse Repetition Frequency) of a radar (among many other factors). The PRF is how many pulses per second the radar emits.


Lets step back to the very basics of radar. At the most basic level a radar emits a pulse of RF, that pulse propagates (travels) outwards to a reflective target, and then the pulse reflects and propagates back to the radar. By measuring the time all of this takes, dividing that time by 2 (since it is a two way trip), and then multiplying the result by the speed of light, the radar can calculate distance from the radar to the target. And then the next pulse is made, starting the process all over.


This time between pulses can be a limiting factor.


A lower PRF radar, fewer pulses per second, can potentially see longer ranges by allowing longer time for the pulse to go out to the target and get back before the next pulse is sent. When the next pulse is sent it resets the timer that keeps track of all of this, so your limit on range is the distance light can travel in half the time from the first pulse to the second (assuming a simple unencoded pulse train).


By this description a low PRF radar can unambiguously see further targets in range than a high PRF radar. Why not just make them all low PRF then? There are many reasons not to, but the most basic is because higher PRFs mean more reflected energy from any given target, increasing the radars probability of detection.


Now when you combine this whole PRF thing with an MTI you end up with radar blind speeds, simple radial velocities that are NOT zero Doppler that the radar cannot see because of the specific PRF used. A fixed PRF radar, one PRF only being used, has multiple blind speeds. But a radar that uses staggers, multiple PRFs used alternately, can insure that while one PRF may be blind at a given speed the other(s) will not be.


And then there are Doppler ambiguities caused by PRF, so that a low PRF radar has high Doppler ambiguity, while a high PRF radar has low Doppler ambiguity, at the cost of range ambiguities, as I demonstrated above.


Anyway, there are lots of factors that determine the performance of a specific radar configuration. You design a radar with a specific target set in mind and you also must consider the platform the radar is on, you can't just say "I will build one radar that does everything, anywhere". With the target set considered you optimize features of the radar for that set and from that source.

So it would pick up stationary targets at altitude but miss things in the weeds?

Basically yes. Unless some additional technique is used to reduce clutter (such as a more traditional MTI system) the weeds all just started looking like targets, so you need to find your desired target in a stack of targets.

So it's just called zero Doppler rejection?

It has a couple of different terms, depending on the audience, but mainlobe clutter rejection or mainlobe clutter filtering is one of them.


Here is a decent basic source on MTI and PD radar. http://tentzeris.ece.gatech.edu/FebMarLectures_2013.pdf


A really good text on radar would be The Radar Handbook, edited by Merrill Skolnik. I think it is currently in its 3rd edition. You can probably find a PDF of it online.

Is that like energy density? The energy per area or volume?

Yes, just like that.

Which would basically somehow reduce the exponential effect to something that's more linear?

No.

Which means the antenna is transmitting equally in all directions?

In the case of the theoretical isotropic radiator, yes, that would be the case.

And that has to do with a smaller area with the same energy?

Exactly, TANSTAAFL (There Ain't No Such Thing As A Free Lunch). You get gain by cramming the same amount of energy, or power, into a smaller volume of the sphere.

So wide-angle is good for searching, narrow angle for tracking?

That is pretty much it, yes. Of course, you can get to more complex techniques, such as rapidly scanned narrow angle beams, and then this description goes out the window.

T!

 

[Zipper730]

DSP is Digital Signal Processing, or Processor.

Ok.

That may be one aspect, but don't fixate on ground clutter, that is only one factor of radar performance.

Understood

Examples of what I am talking about might be centered on the PRF (Pulse Repetition Frequency) of a radar (among many other factors). The PRF is how many pulses per second the radar emits.

Lets step back to the very basics of radar. At the most basic level a radar emits a pulse of RF, that pulse propagates (travels) outwards to a reflective target, and then the pulse reflects and propagates back to the radar.

Is the pulse the same as the wavelength, or is a pulse a timed emission of a given wavelegnth?

A lower PRF radar, fewer pulses per second, can potentially see longer ranges by allowing longer time for the pulse to go out to the target and get back before the next pulse is sent. When the next pulse is sent it resets the timer that keeps track of all of this, so your limit on range is the distance light can travel in half the time from the first pulse to the second (assuming a simple unencoded pulse train).

That's why low pulse-repetition rates are better for long range?

higher PRFs mean more reflected energy from any given target, increasing the radars probability of detection.

Because they send more pulses and ping the target more?

Now when you combine this whole PRF thing with an MTI you end up with radar blind speeds, simple radial velocities that are NOT zero Doppler

How does that occur?

But a radar that uses staggers, multiple PRFs used alternately, can insure that while one PRF may be blind at a given speed the other(s) will not be.

A staggered system would be like slow, fast-fast-fast, slow, fast-fast-fast?

And then there are Doppler ambiguities caused by PRF, so that a low PRF radar has high Doppler ambiguity, while a high PRF radar has low Doppler ambiguity

What's doppler ambiguity?

Anyway, there are lots of factors that determine the performance of a specific radar configuration. You design a radar with a specific target set in mind and you also must consider the platform the radar is on, you can't just say "I will build one radar that does everything, anywhere".

Because it would be overly complicated and cost an absurd amount of money to build?

Basically yes.

So if there were slow speed targets you'd want to be switching routinely back and forth?

It has a couple of different terms, depending on the audience, but mainlobe clutter rejection or mainlobe clutter filtering is one of them.

Okay

Here is a decent basic source on MTI and PD radar. http://tentzeris.ece.gatech.edu/FebMarLectures_2013.pdf

I downloaded it.

A really good text on radar would be The Radar Handbook, edited by Merrill Skolnik. I think it is currently in its 3rd edition. You can probably find a PDF of it online.

Sounds good

Yes, just like that.

Okay

No.

So the system would have to calculate signal intensity to pulse repetition rate and timing?

In the case of the theoretical isotropic radiator, yes, that would be the case.

Ok

Exactly, TANSTAAFL (There Ain't No Such Thing As A Free Lunch).

All too true...

That is pretty much it, yes.

Okay

Of course, you can get to more complex techniques, such as rapidly scanned narrow angle beams

Which means the beam goes back & forth real fast and allows you to cover some space, particularly if the computer can store some data?

 

[fubar57]

.49 seconds............Radar Basics - Doppler Dilemma (Doppler ambiguity)