Chinese Radar Developments - KLJ series and others

The Observer

Junior Member
Registered Member
NRIET skipped the entire PESA cycle entirely, just like they did on the Type 052C. You will have a hard time finding any parallel feed PESA in the entire PLA. This kind of PESA connects the main and central power amplifier and transmitter, with hundreds, or thousands of lines to each element, and so it is actually very difficult and time consuming to produce, not to mention requires a high level of tolerance and quality, with each line must have a very exact length across well over a thousand elements. The kind of PESA you see in service with SAMs, are different animals, ranging from optical feed PESA, where the transmitter feed projects the signal to the back of the array through an optical lens, or the frequency scan planar array, where you have a serial line feed at the side. Compared to the complexity of using a parallel line PESA, an AESA is much simpler. Mind you, there are four different kinds of PESA based on their feed systems, the PLA as a whole only placed into service two of them while the Russians did all four.

People think that you have to develop a PESA first before going to AESA is the ground for this myth, when you can choose to leapfrog it.

Hongdu JL-10 likely uses slotted array. I don't see why a trainer would go through the expense and complexity of using a parallel line feed PESA.

Can you give TL;DR for explanations like this? Because I think we really need those.

Please, Log in or Register to view URLs content!

@Tam did say JL-10/L-15B uses PESA. The article you linked didn't mention what kind of PESA.
 

siegecrossbow

General
Staff member
Super Moderator
Can you give TL;DR for explanations like this? Because I think we really need those.



@Tam did say JL-10/L-15B uses PESA. The article you linked didn't mention what kind of PESA.

China doesn’t mention what kind of AESA is used on the J-10C either. That’s par for the course.
 

Tam

Brigadier
Registered Member
Please, Log in or Register to view URLs content!

Sounds like any Western written article that thinks China is too under developed or low technology for AESA so the array has to be a PESA. When the first Type 052C came out, 'experts' were questioning about it being an AESA, and assumed it was a PESA, because China they claim is too behind for this tech.

I will tell you a good reason why you don't want a PESA on a small fighter, and even less on a plane as small and light as a JL-10.

A PESA is too heavy.

Each element uses a transmit feed that is a heavily shielded co-ax cable, each of the same length, that runs from the element to the transmitter. At the end of the co-ax cable, just below the antenna and the emitter, you got the phase shifter, which looks like a co-ax adapter that is screwed to the end of the cable. At the side of shifter, you got a connector which is how each shifter is controlled. The PESA shifter is made of ferrite-yttrium, and ferrite means its some kind of iron based alloy. So its heavy, and the entire thing is analog.

So you have hundreds of this, with cables sticking out on the back like this that goes to a travelling wave tube which is another bulky thing. The signal going through these cables are analog, and is subject to noise, interference and distortion from all the magnetic fields generated by these components.

download (9).jpegimages (10).jpegimages (11).jpeg

You compare this with an AESA module. A Quad Transmit/Recieve Module is common among aircraft radar types for compactness. Each module has four T/R elements. The phase shifter is a tiny, tiny black IC circuit. You got a D shell connector on the back, the cables are like the ribbon ones you see on a PC, and the communication between this module and the back end computer is digital and binary. Once the signal is converted from analog to digital within this module, the information is saved and there is no distortion and interference. Likewise, on transmit, the signal is digital and converted to analog on this board, then sent to the amps which are tiny black ICs instead of this traveling wave tube thingie, which then goes to the phase shifters, which are also these tiny black ICs, then output at the other end of the board, which is where you put the small sawtooth antennas.

download (10).jpeg

So much of your radar is built like a mini server farm. It consists of ICs and PCBs. In material terms, its plastic, silicon and semiconductor, like Gallium Arsenide or Gallium Nitride. Its light.

The reasons why the US and the Soviet Union went with passive phase array for fighter jets, like that on the MiG-25/31 Zaslon, and AWG-9 on the F-14 Tomcat is because, one word...the Seventies. The digital tech just isn't there yet. The US tried a naval AESA on the USS Long Beach cruiser, and it failed because the concept was too advanced at that time, and the fallback led to the SPY-1 PESA. As for the Soviet Union, they used PESA for a longer period of time because of sanctions that limited their digital technology.

But the J-10B isn't the Seventies. It is by then, the 2000s. China has already implemented its first military applied and accepted AESA at the turn of the decade, the SLC-2, an artillery finder radar. Its first naval AESA, the Type 346, is already being developed for many years, and the design is already finalized before the first Type 052C was laid down. China has no lack of access to digital technology; it is knee deep in it. By the 2000s, its already a major PC manufacturer. You look at other fighter PESA like the first version of RBE2 for the Rafale and that project started in the late eighties. But by far, fighters that started with a slotted array, switched to an AESA on later versions, without going through a PESA stage. The reasons why NRIET never went PESA and went straight to an AESA for the 052C, is probably why you don't see any PESA on the PLAAF, and every airborne phase array in PLAAF service is likely to be an AESA is because NRIET is also the company that does these airborne radars --- including for the J-10, J-11, and J-20. The company started developing AESA early and its built a strong foundation of clear competence on this. Its not likely to develop a PESA already knowing its a dead end technology branch, a deviation from its core competence, would waste valuable development resources and has inherent disadvantages. I don't know if NRIET is behind the JL-10 trainer's radar, but a trainer doesn't need a heavy radar on its nose, and it doesn't need an expensive and powerful radar. Its most likely a slotted array.
 
Last edited:

siegecrossbow

General
Staff member
Super Moderator
Sounds like any Western written article that thinks China is too under developed or low technology for AESA so the array has to be a PESA. When the first Type 052C came out, 'experts' were questioning about it being an AESA, and assumed it was a PESA, because China they claim is too behind for this tech.

I will tell you a good reason why you don't want a PESA on a small fighter, and even less on a plane as small and light as a JL-10.

A PESA is too heavy.

Each element uses a transmit feed that is a heavily shielded co-ax cable, each of the same length, that runs from the element to the transmitter. At the end of the co-ax cable, just below the antenna and the emitter, you got the phase shifter, which looks like a co-ax adapter that is screwed to the end of the cable. At the side of shifter, you got a connector which is how each shifter is controlled. The PESA shifter is made of ferrite-yttrium, and ferrite means its some kind of iron based alloy. So its heavy, and the entire thing is analog.

So you have hundreds of this, with cables sticking out on the back like this that goes to a travelling wave tube which is another bulky thing. The signal going through these cables are analog, and is subject to noise, interference and distortion from all the magnetic fields generated by these components.

View attachment 71356View attachment 71357View attachment 71359

You compare this with an AESA module. A Quad Transmit/Recieve Module is common among aircraft radar types for compactness. Each module has four T/R elements. The phase shifter is a tiny, tiny black IC circuit. You got a D shell connector on the back, the cables are like the ribbon ones you see on a PC, and the communication between this module and the back end computer is digital and binary. Once the signal is converted from analog to digital within this module, the information is saved and there is no distortion and interference. Likewise, on transmit, the signal is digital and converted to analog on this board, then sent to the amps which are tiny black ICs instead of this traveling wave tube thingie, which then goes to the phase shifters, which are also these tiny black ICs, then output at the other end of the board, which is where you put the small sawtooth antennas.

View attachment 71358

So much of your radar is built like a mini server farm. It consists of ICs and PCBs. In material terms, its plastic, silicon and semiconductor, like Gallium Arsenide or Gallium Nitride. Its light.

The reasons why the US and the Soviet Union went with passive phase array for fighter jets, like that on the MiG-25/31 Zaslon, and AWG-9 on the F-14 Tomcat is because, one word...the Seventies. The digital tech just isn't there yet. The US tried a naval AESA on the USS Long Beach cruiser, and it failed because the concept was too advanced at that time, and the fallback led to the SPY-1 PESA. As for the Soviet Union, they used PESA for a longer period of time because of sanctions that limited their digital technology.

But the J-10B isn't the Seventies. It is by then, the 2000s. China has already implemented its first military applied and accepted AESA at the turn of the decade, the SLC-2, an artillery finder radar. Its first naval AESA, the Type 346, is already being developed for many years, and the design is already finalized before the first Type 052C was laid down. China has no lack of access to digital technology; it is knee deep in it. By the 2000s, its already a major PC manufacturer. You look at other fighter PESA like the first version of RBE2 for the Rafale and that project started in the late eighties. But by far, fighters that started with a slotted array, switched to an AESA on later versions, without going through a PESA stage. The reasons why NRIET never went PESA and went straight to an AESA for the 052C, is probably why you don't see any PESA on the PLAAF, and every airborne phase array in PLAAF service is likely to be an AESA is because NRIET is also the company that does these airborne radars --- including for the J-10, J-11, and J-20. The company started developing AESA early and its built a strong foundation of clear competence on this. Its not likely to develop a PESA already knowing its a dead end technology branch, a deviation from its core competence, would waste valuable development resources and has inherent disadvantages. I don't know if NRIET is behind the JL-10 trainer's radar, but a trainer doesn't need a heavy radar on its nose, and it doesn't need an expensive and powerful radar. Its most likely a slotted array.
Please, Log in or Register to view URLs content!

All I can say is that your info is out of date and wrong.
 

Tam

Brigadier
Registered Member
Please, Log in or Register to view URLs content!

All I can say is that your info is out of date and wrong.

With slots like these?

29yqo91 (1).jpg


That's a slotted wave guide phase array. That is a new and design that's very thin and flat. Its a very different kind of PESA you're seeing here as it uses an entire layer behind the slotted guides as a feed. The amplifier is solid state and located right behind the layer feed. This allows the design to be very thin, very flat. That's why its used with flat panel SATCOMs.


1-Figure1-1 (1).pngdownload (17).jpeg'End-feed-single-layer-slotted-waveguide-arrays-a-End-feed-single-layer-slotted.png

You can say this is not like your traditional PESA. Far from it.


It's been long established now that the radar on J-10B (in the picture depicted) is a PESA, not an AESA.

As for JL-10, ultimately we only have rumours to go by, and the rumours have all suggested it is PESA as well.

From what I heard is that it is an AESA, an earlier form. Said radar is the one that is top here. It is referred to as a 1.5 generation, as opposed to the second and third generation of the other two. It describes it as digitalized, with a 2 brick package. Digital and brick, bingo, that's an AESA. Describes it as 12XX units, Gallium (Arsenide?). Guess what 10B stands for.

Second radar there is second generation. Has a new brick package and uses Gallium (Arsenide?). Has 1760 elements. DDS stands for Direct Digital Synthesis. ICs are mounted on both sides of the PCB. Guess what 16 stands for.

Third radar is third generation. Describes it as using Gallium (Nitride?). The T/R description of this differs from the other two. Has 1856 elements. 3D MCM stands for 3D Multi Chip Module with 8 channels and 4 layers. This means the ICs are stacked. Guess what 20 stands for.

This is clearly an infographic describing AESA generation.


Type-1475.jpg
 
Last edited:

SpicySichuan

Senior Member
Registered Member
With slots like these?

View attachment 71361


That's a slotted wave guide phase array. That is a new and design that's very thin and flat. Its a very different kind of PESA you're seeing here as it uses an entire layer behind the slotted guides as a feed. The amplifier is solid state and located right behind the layer feed. This allows the design to be very thin, very flat. That's why its used with flat panel SATCOMs.


View attachment 71362View attachment 71363'View attachment 71364

You can say this is not like your traditional PESA. Far from it.




From what I heard is that it is an AESA, an earlier form. Said radar is the one that is top here. It is referred to as a 1.5 generation, as opposed to the second and third generation of the other two. It describes it as digitalized, with a 2 brick package. Digital and brick, bingo, that's an AESA. Describes it as 12XX units, Gallium (Arsenide?). Guess what 10B stands for.

Second radar there is second generation. Has a new brick package and uses Gallium (Arsenide?). Has 1760 elements. DDS stands for Direct Digital Synthesis. ICs are mounted on both sides of the PCB. Guess what 16 stands for.

Third radar is third generation. Describes it as using Gallium (Nitride?). The T/R description of this differs from the other two. Has 1856 elements. 3D MCM stands for 3D Multi Chip Module with 8 channels and 4 layers. This means the ICs are stacked. Guess what 20 stands for.

This is clearly an infographic describing AESA generation.


View attachment 71366
I remember seeing this a long long time ago. I am still leaning toward the idea that J-10B has a PESA instead of AESA. The image above also says that the supposed J-20 phased-array radar is a third generation prototype dating to 2009. I would not rule out the possibility that the J-16 flankers and J-20 currently in service have something much newer than the 2009 radar prototypes. Keep in mind that it took quite a while for the J-16 to reach IOC, even after China digested the entire flanker design concept and production line in the 1990s.
 

Tam

Brigadier
Registered Member
I remember seeing this a long long time ago. I am still leaning toward the idea that J-10B has a PESA instead of AESA. The image above also says that the supposed J-20 phased-array radar is a third generation prototype dating to 2009. I would not rule out the possibility that the J-16 flankers and J-20 currently in service have something much newer than the 2009 radar prototypes. Keep in mind that it took quite a while for the J-16 to reach IOC, even after China digested the entire flanker design concept and production line in the 1990s.

Based on what? Is there any technical proof that the J-10B array is a PESA? Because of what? Internet momentum?

That diagram is clear as day mentioning a brick architecture for the -10B radar. Modules on AESA are arranged in bricks.

This is what a traditional parallel feed PESA looks like at the edge, using BARS here for the SU-30. Those coils there are the ferrite yttrium phase shifters.

EC9iuYPWwAA3hXj (1).jpg

In contrast the AESA consists of bricks and planks.

HoPk6JF.png

The -10B radar describes a 2 piece brick package on the third line. The -16 radar describes a different brick packaging design. Perhaps for better cooling?

The -10B radar description is also explicit in mentioning T/R. PESA has separate transmit and receive subarrays while in an AESA, an element is both transmit and receive.

It takes some years to develop a military radar, so the timeline around 2009 is about right. Military radars don't advance as fast as commercial and civilian equipment. Phase arrays used in telecom already use a tile based architecture (a) and they make AESA brick designs (b) look like dinosaurs.

11m28f1.jpg


What you see on the diagram with regards to the J-16 and J-20 radars should be more or less same as the diagrams with some up to date refinements. The J-16 radar might be using the 40nm CETIC octacore DSP while the J-20 the 28nm or 40nm CETIC DSP which is not impressive if your idea of modern is 7nm to 14nm. But again, using the stacked chip or Multi Module Chip design can go a long way of compensating for the older fab process, not to mention the larger legacy transistor sizes might be more reliable anyway. The J-10C radar should be more or less along the lines of the J-16 radar but smaller, less elements and less power. Another significant advancement in these years is whether the AESA moved from using Gallium Arsenide for its amplifiers to Gallium Nitride.
 
Last edited:
Top