You right. This post goes into detail about the Bars hybrid PESA which should be similar:
Naval missile guidance thread - SAM systems
- Thread starter Iron Man
- Start date
Here are example.
Yes, beamforming which is also known as spatial filtering, along with pulse compression and Doppler filtering.
Note how those diagrams match with this physical example which happens to be a module for Phazotron's Zhuk AESA radar.
Note that T/R does not necessarily mean TRM. TRM means T/R per module. DTRM is Dual T/R per Module or two T/R in one module. QTRM is Quad T/R per Module. The above example is a QTRM, and is actually more common with high frequency band like X band. You may not need to put a FPGA behind every T/R, but one in four is good.
This also highlights the problem of getting information from sales literature. They make it sound so special and as if they invented the concept when its been around on others. Like calling a "hybrid" PESA when it should be typical of any PESA or even Planar array.
SPY-1 on AEGIS works the same way, having separate emitters and receivers. Won't be surprised more PESAs are the same. The core terminology of A(ctive) ESA vs. P(assive) ESA lies solely in the transmission side. PESA can have multiple emitters but they are all connected to a single massive signal generator and amplfier that is typically analog. I hate to use the word transmitter since this word is interchangeably used with both emitters and amplifiers. With AESA, each module has its own small amp and signal generator, this being typically solid state and digital.
Last edited:
Brumby
Major
Technically, there is no such thing as a hybrid PESA. It is a marketing spin to give the illusion that somehow the technology adoption is more advanced than its technical architecture NO11 is a MSA radar - period.
Speaking of FPGA, FPGA industry has been trying to sell high end, high scale solutions for cloud computing centers now, and the latest trend is to incorporate AI and deep learning units into the FPGA. Ironically, back in 2016, Xilinx bought a Chinese AI startup called Deephi for this.
Last edited:
Hmm. I just went to check the Wikipedia article on SPY-1. It would appear that it is the true hybrid, having 8 transmitters and 32 amplifiers, unlike the classical PESA with one transmitter. Perhaps that arrangement was necessary to scale up the radar power without introducing too much distortion.
I don't think what is posted in Wiki is correct. I already posted a document within this thread to repeat it again.
SPY-1D has one large transmitter multiplexing on four arrays, while SPY-1B for the Tico has two transmitters, one on each deckhouse front and back, each transmitter is multiplexed for two arrays. The article cites various research papers and gives a summary. Furthermore the way the array is subdivided doesn't support what wiki said. This one large amp or transmitter, as in the Burke, or two, as in the Tico, consists of 32 CFAs or Cross Field Amplifiers.
What do you find wrong with the subdivision? Divide 4096 radiators with 32 subarrays, and you get 128 radiators per subarray. Also, the 32 CFAs align nicely with the 32 radiating subarrays. Not sure how you imagine the "large amp" to work, but I am pretty sure it is not a linear sequence of 32 CFAs. I also don't see what the big deal is with having 8 signal generators, especially seeing to what degree the radar architecture appears to be parallelized.
I checked the source of the info on Wikipedia, and it's a US Naval Institute publication. If they can't get the info right in their 5th edition of the publication, then my only explanation would be that it is so by intention.
I also checked some of the sources from the article you linked. There are some very interesting claims about the agility of the radar that contradict much of what is being discussed in that article.
So this is what the Union of Concerned Scientists article says:
Not 4096 radiators, 4,350 elements. Subarray consists of 2 elements, total of 2175 subarrays. The previous model has 68 subarrays of 64 elements with 4,352 elements.
The description is that a single transmitter consists of 32 CFA multiplexed for all four arrays. Each CFA has 132kw, and 32 of them is about 4.2mw.
The whole ship has 7.5mw of power, 3 generators of 2.5mw. So you can't have more than 32 CFA for the entire ship. You can't have like 32 CFA on a single face, or a total of 4.2mw per face. The power consumption is much greater than what the ship can provide. The Flight III Burke only has 3 x 3.5mw generators.
If the 32 CFA is divided into four arrays, that makes about 8 CFA per array face, or just over 1 mw per face. Compare this with a single 4.2 to 6 MW transmitter that can transmit all of its power into a single face.
The problem with Wiki is that I am not sure if wiki writer is properly reflecting the exact text from Norman Friedman. The article I showed you is also quoting Normal Friedman.
Hmm. The 68 sub arrays were for the receivers, which outnumber the radiators. The radiators have only 32 sub arrays. The original arrangement was 2R arrays for 1T array. The Wikipedia article appears to be describing the original configuration of the radar. In the section on SPY-1B, they state the following:
No mention of the subarray configuration here.
Your numbers are not right. Flight IIA ships come with 3 generators of 3MW. Flight III ships will have 3 generators of 4MW each.
If you fire all the 32 CFAs through a single face, as any sane designer would, the power consumption will not exceed what the ship can provide. The radar is operated in pulse mode, not in CW. Because of that, there is ample time to fire through the other 3 faces while the first face is listening to echoes. I don't understand why you would want to design your radar so that it can use only a 1/4 of the available power.
Now I am intrigued. I will try to get my hands on a copy of this book.