Its not about putting TR modules alone.
you need to accommodate antennae+every subsystem in that size ,
means if you are putting 1500 module on 650 mm antennae ( size of MMR) you require to put that many wave-shifters and cooling mechanism within that cranked space.
When you have RF signal of 1500 module , you need that much more sub-array portioning and choreography . In effect that much more powerful COTS processor .
Yes but AESA radars are developed with modular design concepts. That mean you have specific space for every component. If the number of individual T/R modules and radiating elements does not increase gradually that means beyond 2000 you don't need larger beamformer, exciter, receiver or cooling mechanism. Building blocks are same, technologies are different. Power, cooler and calibration are some complexities but they don't cause problem until certain limits.
About radar computer RC (processor), its not usually COTS but developed by DRDO itself and DRDO is the best in this area. We have our own RC for MKI's Bars radar as well as LCA's radar. DRDO already developed beam steering computer for other projects.
For exampple look at typical AESA radar configuration for figher aircrafts... note the cooling package, power supply and LRUs....
It mainly depends on technologies. Look at the nose cone length of F-22 and F-18E/F, F-18E has longer nose cone than F-22 that means it can have larger number of T/R modules but in reality APG-79 has around ~1100 while APG-77 has nearly ~2000 T/R modules, still F-22 can accommodate APG-77.
We all know how much amplifiers and esp GA-AS chips generate heat , biggest drawback of current AESA is only this.
Now SAAB overcame this by putting less modules but moving the antennae mechanically beyond 60 degrees in Azimuth to cover more area.
Thales corp is trying with GA-Nitride chips in place of regular GA- Arsenide to have less heating issues.
Heat is not the only reason for putting less number of but size of MMIC components (mainly built with GaAs, SiGe, SiC, GaN) and costs. More number of T/R modules you can put on your array or radiating elements keeping the radiating power in mind its better for you. But everyone cannot just do that. Heating is a issues usually taken care with controlled radiation, low-peak power.
That is why even Fighters of size of
F18 (APG79 have 1100 modules)
Rafale has close to 1000
Mig 35 earlier ZHUK-AE model had 700 ( Zhuk- MFAE to get 1100 in future)
Eurofighter ( VIXEN 1000 will have 1000 module)
Our fav Gripen has (Vixen 600 module)
Only Raptor has more than 1800-2000
Pak-FA is suppose to have 1400-1500 only.
Same reason why IBRIS and BARS even though not Actively beam steering radar , still have range and performance better than all AESA radar at this moment .( except APG-77 which has close to 2000)
F-22 is not huge wrt F-18 or Eurofighter than how it's APG-77 managed to get nearly 2000 T/R modules, much more than others? Because it depends on technology.
AESA doesn't mean increased range. It depends on number of T/R modules, power per module and transmit duty. IBRIS and BARS as powerful and large PESA radars have medium to high peak power resulting high duty for a passive radar. For example 1000 elements PESA needs 10 kW of power while 1000 elements AESA radar needs 0.5 kW of average power resulting higher radiation of power from PESA. So IBRIS or BARS need much more power to be good but still don't have features like LPI, high transmit duty and reliability.
Here is combined packages of LCA's PD radar and CAESAR AESA radar including all components... cooler, wave shifter, BSC etc....
http://2.bp.blogspot.com/_DOqSTiSBP3U/S56a8A3dtlI/AAAAAAAAAfk/Mesf5TQPC9U/s400/7.jpg
http://www.bharat-rakshak.com/media/35321-1/IMG_1074.JPG
They barely different in over all size.
