Jump to content

Searching for proof of underperforming AN/APG-73 radar


GumidekCZ

Recommended Posts

18 minutes ago, Harker said:

Yep, that's the main issue at this point.

 

I mean its one of them. RCS isn't 1 number, it varies with target aspect, it varies with radar absobent coatings, it varies when you add stores IRL. Yeah you see some "average" number presented but its 100% wrong in detail. 

 

The funny thing is, you can use open source matlab software to calculate RCS from whatever aspect you want, and it was doable on a decade + old machine. So the question really is, why doesn't ED do that offline, then make some aspect/stores table for each A/C. I mean yeah, you need some compute time for each model, but do the flyable planes first and you are 1000% better in terms of realism if you do. It won't neccisarily be "right" but it will be "relatively right" between all DCS jets and thats all we are looking for.

 


Edited by Harlikwin
  • Like 4

New hotness: I7 9700k 4.8ghz, 32gb ddr4, 2080ti, :joystick: TM Warthog. TrackIR, HP Reverb (formermly CV1)

Old-N-busted: i7 4720HQ ~3.5GHZ, +32GB DDR3 + Nvidia GTX980m (4GB VRAM) :joystick: TM Warthog. TrackIR, Rift CV1 (yes really).

Link to comment
Share on other sites

 
I mean its one of them. RCS isn't 1 number, it varies with target aspect, it varies with radar absobent coatings, it varies when you add stores IRL. Yeah you see some "average" number presented but its 100% wrong in detail. 
 
The funny thing is, you can use open source matlab software to calculate RCS from whatever aspect you want, and it was doable on a decade + old machine. So the question really is, why doesn't ED do that offline, then make some aspect/stores table for each A/C. I mean yeah, you need some compute time for each model, but do the flyable planes first and you are 1000% better in terms of realism if you do. It won't neccisarily be "right" but it will be "relatively right" between all DCS jets and thats all we are looking for.
 
Yeah, the RCS tables in general need to be addressed. They could easily be approximated with a 2 variable function for say, 5 degree (or lower, so we don't see jumps) increments in aspect and a simple delta function coefficient for stores. Even something small like that, which would be computationally inexpensive and easy to plug into the existing radar API, would greatly increase fidelity.
  • Like 2

The vCVW-17 is looking for Hornet and Tomcat pilots and RIOs. Join the vCVW-17 Discord.

CVW-17_Profile_Background_VFA-34.png

F/A-18C, F-15E, AV-8B, F-16C, JF-17, A-10C/CII, M-2000C, F-14, AH-64D, BS2, UH-1H, P-51D, Sptifire, FC3
-
i9-13900K, 64GB @6400MHz RAM, 4090 Strix OC, Samsung 990 Pro

Link to comment
Share on other sites

3 hours ago, GumidekCZ said:

Could this performance for APG-73 be from RW, or close to real?

https://www.key.aero/forum/modern-military-aviation/15964-rafale-vs-gripen?page=7

APG-73 (F/A-18E/F, Block1) - (from Block2 was eqiped with APG-79 AESA)

For RCS 0.0001 m2 class target: 5~6 km+ ???
For RCS 0.001 m2 class target: 10~11 km+ (5.5-6.1nm)->in DCS 5,5nm RWS / 3.0nm FLANK / 3.0nm BEAM (10nm ACM all)
For RCS 0.1 m2 class target: 32~36 km+  (18-20nm)->in DCS 17.5nm HOT /9.5nm FLANK /9.5nm BEAM
For RCS 1.0 m2 class target: 56~64 km+ (30-35,5nm)->in DCS 31nm /31nm FLANK /31nm BEAM
For RCS 5.0 m2 class target: 84~96 km+ (46.6-53,3nm)->in DCS 47nm HOT / 47 FLANK / 47nm BEAM
For RCS 10.0 m2 class target: 100~114 km+ (55.5-63nm)->in DCS 56nm HOT / 56 FLANK / 52nm BEAM

 

DCS values checked by me, editing RCS values and than testing at 36000 ft, targets 2000 ft higher alt than me in Hornet. Scan zone AZ20° 1BAR. INT and HI PRF.

Acording to values given from link page, DCS APG-73 is in average some 6% bellow average detection distance... which is far better than I was expecting.

 

Big question now is the RCS values of each aircraft type in DCS.

 

 

I mean those numbers at a guess are from CMNAO or similar game, they get those numbers from refs like Janes most likely or other similar sources.

 

The problem is that "max" range on a radar isn't a 100% thing, ever. You have a ton of factors that determine detection range IRL, including weather, your attitude, bandit altitude. Specific radar mode and so forth. So those ranges are a guess. They look "ok" ish to me and the nice thing seems to be that they are using some consistent data set. Which more or less tells us, the hornet isn't underperforming, other modules (cough viper) are overperforming "maybe" in certain modes, "maybe". And yet other planes are underperforming (cough F15)...

 

Bottom line. ED needs to do a better job with radars, and then it needs some sort of consistency Czar, so that when they make changes to 1 module they consider in the broader context of ALL modules...

 

  • Like 2

New hotness: I7 9700k 4.8ghz, 32gb ddr4, 2080ti, :joystick: TM Warthog. TrackIR, HP Reverb (formermly CV1)

Old-N-busted: i7 4720HQ ~3.5GHZ, +32GB DDR3 + Nvidia GTX980m (4GB VRAM) :joystick: TM Warthog. TrackIR, Rift CV1 (yes really).

Link to comment
Share on other sites

When testing ranges, I also come to one weird thing, and thats upper hemisphere detection with AN/APG-73 is NOT automaticly striped of Doppler speed limits. Which are now 100kph or 27,78m/s.

With such "modern" radar which also served shortly on Super Hornet, I would expect some automatic function like MLC AUTO in F-14 tomcat had many years before APG-73.

 

Edit:      For others, reading this topic, I must point out, that detect ranges were found with smallest radar scan settings possible. Any larger scan area will prolong time to detect the target after it pass by limit range for such RCS and thus the actual detection range can be much shorter (also depends on closure speed).


Edited by GumidekCZ
Link to comment
Share on other sites

There is likely a small underperformance indeed, compared to usually quoted ranges of detection of the real one and hearsay. We will see when ED adjust the APG-68 how both compare (i.e. if they get the same detection performance then ED are definitely doing something wrong).

 

A related and probably more important problem are the RCS inconsistencies in the database across aircraft, and the external stores not affecting RCS.

Link to comment
Share on other sites



 
Bottom line. ED needs to do a better job with radars, and then it needs some sort of consistency Czar, so that when they make changes to 1 module they consider in the broader context of ALL modules...
 

I'd argue that it'd be even better if ED worked on developing a basic, physics-based radar API that takes basic inputs like diameter, power, PRF, mode etc and allows for certain coefficients and automatically determines the probability of detection at range x like that.

Then, module developers need only to include the specifications of the radar, add the correct modes and use the coefficients to achieve published results in certain modes (I'm including coefficients to deal with things like the APG-68(V)5's MPRF, which is not the same as the APG-73's MPRF, for example). That would eliminate much of the speculation and any changes made to the radar physics API would affect DCS as a whole. Everything would be on the same level.
  • Like 6

The vCVW-17 is looking for Hornet and Tomcat pilots and RIOs. Join the vCVW-17 Discord.

CVW-17_Profile_Background_VFA-34.png

F/A-18C, F-15E, AV-8B, F-16C, JF-17, A-10C/CII, M-2000C, F-14, AH-64D, BS2, UH-1H, P-51D, Sptifire, FC3
-
i9-13900K, 64GB @6400MHz RAM, 4090 Strix OC, Samsung 990 Pro

Link to comment
Share on other sites

5 hours ago, Harker said:


 


I'd argue that it'd be even better if ED worked on developing a basic, physics-based radar API that takes basic inputs like diameter, power, PRF, mode etc and allows for certain coefficients and automatically determines the probability of detection at range x like that.

Then, module developers need only to include the specifications of the radar, add the correct modes and use the coefficients to achieve published results in certain modes (I'm including coefficients to deal with things like the APG-68(V)5's MPRF, which is not the same as the APG-73's MPRF, for example). That would eliminate much of the speculation and any changes made to the radar physics API would affect DCS as a whole. Everything would be on the same level.

 

I partly agree. The issue is finding enough info. And really to get "good" numbers out of some generalized equation you have to make assumptions. There is a 50% detection probability for the radar equation. And while you could probably find things like antenna size, or maybe even figure out average power out for a given mode (not trivial), I really doubt you are gonna find any info on the noise floor of the receiver, or how post processing is improving that. So you are back to some of those tables gumidek posted, and are those ranges for a 100% detection, 50% detection, what?

 

The other thing, is that factors like weather or even ownship altitude can effect those ranges fairly significantly depending on the radar as well as RCS and aspect, there are lots of stories from "earlier" generations of radar seeing a track at long range only to have it disappear as the aspect/rcs changed, of course more modern radars will try to keep looking for that guy and still show you their best "guess" of where he is. So I'm not really concerned too much with matching performance within 6% of some table on the internet. 

New hotness: I7 9700k 4.8ghz, 32gb ddr4, 2080ti, :joystick: TM Warthog. TrackIR, HP Reverb (formermly CV1)

Old-N-busted: i7 4720HQ ~3.5GHZ, +32GB DDR3 + Nvidia GTX980m (4GB VRAM) :joystick: TM Warthog. TrackIR, Rift CV1 (yes really).

Link to comment
Share on other sites

On 6/7/2021 at 6:10 AM, Gierasimov said:

Its detecting and tracking A2A missiles now.

 

Wait, huh?? It seriously can do that??

Intel 13900k @ 5.8ghz | 64gb GSkill Trident Z | MSI z790 Meg ACE| Zotac RTX4090 | Asus 1000w psu | Slaw RX Viper 2 pedals | VKB Gunfighter Mk3 MCE Ultimate + STECS/ Virpil MongoosT50+ MongoosT50CM | K-51 Collective + custom AH64D TEDAC | HP Reverb G2 | Windows 11 Pro | |Samsung Odyssey G9 | Next Level Racing Flight Seat Pro

The Boeing MQ-25A Sting Ray = Dirt Devil with wings
 My wallpaper and skins

Link to comment
Share on other sites

Just now, Hammer1-1 said:

Wait, huh?? It seriously can do that??

Apparently this is a feature as Chizh said. 

However in a future patch we should see some tweaks to the missile interception success rate as currently I am getting 100% success when shooting down R-77 with either 120C or 9X

Intel i7-13700KF :: ROG STRIX Z790-A GAMING WIFI D4 :: Corsair Vengeance LPX 64GB ::  MSI RTX 4080  Gaming X Trio  :: VKB Gunfighter MK.III MCG Ultimate :: VPC MongoosT-50 CM3 :: non-VR :: single player :: open beta

Link to comment
Share on other sites

I didnt even know that was a thing; is the real AN/APG73 capable of doing so? Cruise missiles I understand, but cant imagine anything smaller than that as being filtered out by the search radar.


Edited by Hammer1-1
  • Like 1

Intel 13900k @ 5.8ghz | 64gb GSkill Trident Z | MSI z790 Meg ACE| Zotac RTX4090 | Asus 1000w psu | Slaw RX Viper 2 pedals | VKB Gunfighter Mk3 MCE Ultimate + STECS/ Virpil MongoosT50+ MongoosT50CM | K-51 Collective + custom AH64D TEDAC | HP Reverb G2 | Windows 11 Pro | |Samsung Odyssey G9 | Next Level Racing Flight Seat Pro

The Boeing MQ-25A Sting Ray = Dirt Devil with wings
 My wallpaper and skins

Link to comment
Share on other sites

16 minutes ago, Hammer1-1 said:

I didnt even know that was a thing; is the real AN/APG73 capable of doing so? Cruise missiles I understand, but cant imagine anything smaller than that as being filtered out by the search radar.

 

 

Yeah its a point of contention. Apparently they have some docs that say a SU-27 tactic was to shoot down incoming AAM's, so surely the APG-73 which is more capable could have the same capability. The issue frankly is a complicated one, I'm sure the radar can technically "see" a missile coming at you (high doppler shift) at some range (RCS varying on missile type of course). So "seeing it pop up" is probably something that "can" happen, the other question is does it happen, and can the radar get a firing solution on the missile to shoot it down, and then does the missile have good enough fusing to actually hit/damage the incoming missile. If you think about it, even late 70's era SAM's could see and shoot missiles, but they were pretty specifically designed with that capability in mind. And in that case its simpler problem since the launcher is static. 

 

22 minutes ago, Gierasimov said:

Apparently this is a feature as Chizh said. 

However in a future patch we should see some tweaks to the missile interception success rate as currently I am getting 100% success when shooting down R-77 with either 120C or 9X

 

You are airborne IRON DOME. 

  • Like 1

New hotness: I7 9700k 4.8ghz, 32gb ddr4, 2080ti, :joystick: TM Warthog. TrackIR, HP Reverb (formermly CV1)

Old-N-busted: i7 4720HQ ~3.5GHZ, +32GB DDR3 + Nvidia GTX980m (4GB VRAM) :joystick: TM Warthog. TrackIR, Rift CV1 (yes really).

Link to comment
Share on other sites

41 minutes ago, Harlikwin said:

 

Yeah its a point of contention. Apparently they have some docs that say a SU-27 tactic was to shoot down incoming AAM's, so surely the APG-73 which is more capable could have the same capability. The issue frankly is a complicated one, I'm sure the radar can technically "see" a missile coming at you (high doppler shift) at some range (RCS varying on missile type of course). So "seeing it pop up" is probably something that "can" happen, the other question is does it happen, and can the radar get a firing solution on the missile to shoot it down, and then does the missile have good enough fusing to actually hit/damage the incoming missile. If you think about it, even late 70's era SAM's could see and shoot missiles, but they were pretty specifically designed with that capability in mind. And in that case its simpler problem since the launcher is static. 

 

 

You are airborne IRON DOME. 

Capability that's not documented. But if it if ito speculate maybe they set vs to some velocity to see them see them and even lock them ( mode thats not planned)


Edited by IkarusC42B Pilot
Link to comment
Share on other sites

21 minutes ago, IkarusC42B Pilot said:

Capability that's not documented. But if it if ito speculate maybe they set vs to some velocity to see them see them and even lock them ( mode thats not planned)

 

 

Yeah the whole thing is a bit too "unknown" for my taste. Plus at that point you are playing robotech IMO.

New hotness: I7 9700k 4.8ghz, 32gb ddr4, 2080ti, :joystick: TM Warthog. TrackIR, HP Reverb (formermly CV1)

Old-N-busted: i7 4720HQ ~3.5GHZ, +32GB DDR3 + Nvidia GTX980m (4GB VRAM) :joystick: TM Warthog. TrackIR, Rift CV1 (yes really).

Link to comment
Share on other sites

So take this with a grain of salt, but I asked one of my squadron leaders about this. He flew backseats in Growlers once upon a time (and we have more than a few IRL warthog pilots in our ranks as well), but apparently AT LEAST in the Growler you can track incoming missiles. Thats not to say its true, thats not to say its indicative of an AN/APG-73, and thats not to put at ease the question whether or not its accurate. Merely pointing out that maybe theres a bit of truth in that statement. I wont say for certain simply because I personally dont know, to me its just hearsay. So other than the fact that it could most likely be classified information in HOW it does it, but in all likelihood it can.

 

I never even thought to ask...

Intel 13900k @ 5.8ghz | 64gb GSkill Trident Z | MSI z790 Meg ACE| Zotac RTX4090 | Asus 1000w psu | Slaw RX Viper 2 pedals | VKB Gunfighter Mk3 MCE Ultimate + STECS/ Virpil MongoosT50+ MongoosT50CM | K-51 Collective + custom AH64D TEDAC | HP Reverb G2 | Windows 11 Pro | |Samsung Odyssey G9 | Next Level Racing Flight Seat Pro

The Boeing MQ-25A Sting Ray = Dirt Devil with wings
 My wallpaper and skins

Link to comment
Share on other sites

54 minutes ago, Hammer1-1 said:

So take this with a grain of salt, but I asked one of my squadron leaders about this. He flew backseats in Growlers once upon a time (and we have more than a few IRL warthog pilots in our ranks as well), but apparently AT LEAST in the Growler you can track incoming missiles.

May be you can ask him also about this:

 

11 hours ago, GumidekCZ said:

When testing ranges, I also come to one weird thing, and thats upper hemisphere detection with AN/APG-73 is NOT automaticly striped of Doppler speed limits. Which are now 100kph or 27,78m/s.

With such "modern" radar which also served shortly on Super Hornet, I would expect some automatic function like MLC AUTO in F-14 tomcat had many years before APG-73.

 

Link to comment
Share on other sites

Ill show this thread to him and see if he wants to chime in. Cant guarantee anything as a lot of these guys shy away from here because this place is kind of toxic...and I dont really blame them.

  • Like 1

Intel 13900k @ 5.8ghz | 64gb GSkill Trident Z | MSI z790 Meg ACE| Zotac RTX4090 | Asus 1000w psu | Slaw RX Viper 2 pedals | VKB Gunfighter Mk3 MCE Ultimate + STECS/ Virpil MongoosT50+ MongoosT50CM | K-51 Collective + custom AH64D TEDAC | HP Reverb G2 | Windows 11 Pro | |Samsung Odyssey G9 | Next Level Racing Flight Seat Pro

The Boeing MQ-25A Sting Ray = Dirt Devil with wings
 My wallpaper and skins

Link to comment
Share on other sites

I don't think it's surprising at all that APG-73 could detect and track an A2A missile, especially at look-up and at high closure, but keep in mind that tactically actually trying to intercept enemy A2A missiles with your own missiles really only makes sense in the last 5-10 years with the proliferation of longer ranged and more advanced ARH weapons. I would almost say its out of the scope of DCS and our circa 2005 Hornet sim except that the scope has never been really defined.

Link to comment
Share on other sites

23 hours ago, GumidekCZ said:

Big question now is the RCS values of each aircraft type in DCS.

In DCS they are static. There is no RCS change based on aspect, for example Flanker IIRC has 5m2 RCS head on and will havesame rcsin 3/4 or 1/2. Or from below.

There are datamined RCSvalues on this forum. 

Link to comment
Share on other sites

Some of these RCS values are another proof, that ED didn't made any deeper search for more less "correct" values. And RCS is now roughly based on size.

One example for all: All sources across the internet will tell you that F-16A version has twice or more RCS than its C version block 50and newer. In DCS... Both variants have same value.

But even bigger question raises, if we start debate, if this all matters, if ED is not counting with any loadout on pylons. Problem is, that ED would than need defined RCS value for every pylon and every weapon. But this should be discussed in another separate topic.


Edited by GumidekCZ
  • Like 1
Link to comment
Share on other sites

Anyway something is just clearly wrong with the latest patch: Against AI Mig-23's at 20K altitude I have relatively useful detection range when I narrow my sweep, but I loose lock continuously (in whatever mode). The AI Migs can engage me first 100% of the time. They are actually not really manoeuvring as they are AI. 

It is only when I boresight the radar when the bandits are within 10 NM that I can lock and engage. 

  • Like 1

[sIGPIC][/sIGPIC]

Link to comment
Share on other sites

Anyway something is just clearly wrong with the latest patch: Against AI Mig-23's at 20K altitude I have relatively useful detection range when I narrow my sweep, but I loose lock continuously (in whatever mode). The AI Migs can engage me first 100% of the time. They are actually not really manoeuvring as they are AI. 
It is only when I boresight the radar when the bandits are within 10 NM that I can lock and engage. 
Increase your ageout setting, in the DATA sublevel and press SET in the RWS mode page. The setting is implemented wrongly in DCS and ties the radar memory to the bricks ageout setting. For the AMRAAM, it defaults to 4, so until the bug is fixed, increase it.

The vCVW-17 is looking for Hornet and Tomcat pilots and RIOs. Join the vCVW-17 Discord.

CVW-17_Profile_Background_VFA-34.png

F/A-18C, F-15E, AV-8B, F-16C, JF-17, A-10C/CII, M-2000C, F-14, AH-64D, BS2, UH-1H, P-51D, Sptifire, FC3
-
i9-13900K, 64GB @6400MHz RAM, 4090 Strix OC, Samsung 990 Pro

Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...