Su-27 / Su-33: No Radar when inverted below 1500m

Yes you could say it doesn’t matter where the radar is pointed, but it matters where the compensation horn is pointed and how far from the ground

As long as the target is flying towards the radar faster than the ground the compensation antenna can point wherever it wants, the frequency range of target and ground return is completely different. Even the F-15C manual says the compensation antenna issue happens with fleeing targets.

Imagine locking an approaching target when flying level. Now the radar knows the frequency of the closing target and sets its filters to only pass frequencies in this region. If you now go inverted, it still only looks for frequencies in that region, anything else will be ignored.

Yes, when the target is flying away the lock will be less stable and can be broken by rolling. But not when you are engaging a head on bandit.

And in DCS the radar 100% shuts off independend from the target aspect when going inverted <1500m.

If Doppler worked that perfect why would there be a compensation horn?

Because when the target is not flying towards the radar then the spectra from ground reflections and the target itself will overlap. There you want to minimize any return that is not coming directly from the direction of the target.

Yes, and it’s still there even if there is a high closure rate and you are in HPRF

What is still here? Sidelobe returns? Yes they are, but as I wrote multiple times and as you can even see in the MiG-29B document, the ground return that comes in through the sidelobes is in a different spectrum in this case. This is what you have a bank of bandpass filters for, or FFT in the digital computer of the plane. Frequencies that you don't want get removed, either via analog or digital filtering.

But does the compensation horn care about frequency or spectra? Just because it’s made to block side lobe and ground returns doesn’t necessarily mean that’s all it blocks. Signals going into that lower half of that radar have to go through the horn before being touched by the filter. What can FFT do if the signal to noise ratio goes down from unblocked ground returns and a dampened signature because the compensation horn is front of the target?

What do you mean with the compensation horn in front of the target?

The compensation antenna is not blocking radio waves from entering the main radar. It is a seperate antenna with a very wide beamwidth, but with a very low gain in the area of the main antennas beamwidth.

The radar then compares the signals received from the main channel and from the compensation channel. When a certain return is strong just in the main channel but weak in the compensation channel, then this is a valid signal from the main lobe. When a signal stronger in the compensation channel, then it is not a valid return as it is coming from a sidelobe. The radar processing will then ignore this combination of time delay and frequency.

It is using a seperate receiver. It is not just like sunglasses blocking the radar returns of a certain area.

So the radar in the MiG-29/Su-27 has to be aligned with the horizon on roll otherwise it cannot track?

 

The radar is locked on roll into its place without any stabilisation in any mode except the BVR mode. This would mean that a target locked in any other mode would be lost as soon as you bank more than 0° (because then the radar has the same roll angle as it has when you tilt more than 120° in BVR mode).

 

 

In general it would be nice to have an image of the compensation antenna, or at least the location. I would really like to know where it is exactly.

So, basically rolling only would affect BVR mode, that you wouldn't use when below 1500m, and otherwise there ain't problems as you wouldn't be flying inverted below 1500m...

It all depends on where the compensation antenna is exactly. When it is fixed to the fuselage then all modes will suffer the same way from rolling and will have the same amount of sidelobe compensation degradation at certain degrees of roll angle (This does of course not mean that the radar will lose tracking, especially when the target is in front aspect).

When the compensation antenna is somewhere on the roll gimbal, then it will not suffer from worse sidelobe compensation in BVR mode at roll angles below 120°, but it will suffer in close combat modes on any roll angle above 0°.

The amount of degradation of course depends on the roll angle and also on the beamwidth of the compensation antenna. Maybe it still covers the ground when inverted, but with less gain?

I don't see why the radar would be roll stabilized at all when the compensation antenna is not mounted somewhere on the moving radar. But maybe it wants to keep the the polarisation of the radiowaves optimal (either horizontal or vertical) for long range engagements for an improved range and detection.

How important it really required to be in a with visual range (10 km or less) considering that you can have a IRST assisting the targeting?

Like, you are chasing someone, so your speeds match periodically and then not, meaning your doppler radar will lose a lock by filtering. Same thing when releasing chaff that spreads wide between target and radar, effectively blocking radar seeing through.

Is the IRST capable keep radar locked on target that is pointing where IRST is locked by heat, regardless that radar can't get clear return? Like a AIM-7 and guiding just by beaming radar in FLOOD move, and let missile fly to radar return, and by guidance of the datalink that IRST is providing as target heading changes, even if range ain't calculated anymore?

As if the IRST can just keep radar locked, even when radar is incapable at the moment to lock at target because interference, does it matter is the radar stabilized or not?

If it is already more effective to use a IR R-27T/ET missiles in chase, and R-60/R-73 in turning fight, and leave the R-27R/ER radar guided missiles to BVR engagement or targets heading toward where in situations you anyways would be flying above 1500m altitude and not rolling, does it really then matter do you have gimbal limits or not? Like how many would be engaging targets at 200km range, while flying below 1500m or having target flying below 1500m?

If Su-27 flies at 1500m altitude, it is horizon limited at 138km distance. To see a target at 200km range on ground, it would need to fly at 3135m altitude (why as well 3000m altitude is important one as you are maximizing range at lower altitudes.).

So again I don't see a problem, as reason to fly inverted by passing 120° gimbal limit at those ranges past 10-20km.

I am just questioning that when does such a limit really occur?

So again I don't see a problem, as reason to fly inverted by passing 120° gimbal limit at those ranges past 10-20km.

 

I am just questioning that when does such a limit really occur?

In DCS at the end of every engagement that startet with BVR. The high altitude R-27ER shots will always get chaffed, and you end up in low altitudes closer to the target, where the best thing you can do after firing a missile is going inverted and pulling towards the ground into the thicker air.

With the current radar behaviour it means broken radar lock + trashed R-27ER, because due to the Fox 1 bug it will go for a chaff as soon as the radar stops emitting for a splitsecond.

I am not sure how loss of radar tracking while keeping IRST tracking works when guiding SARH missiles with the real Flanker. The pulses for the missiles and the pulses that the radar emitts for its own tracking purposes are different ones and multiplexed in time. There is a good chance that the radar can still emit the missile pulses while being slaved to the IRST, and at the same time still trying to regain the lock. A lock that will not be broken by just rolling inverted aganist a head on target in the first place.

I finally found where the compensation antenna is located on the MiG-29:

The long yellow part in the top under the cover is the waveguide, and on the circular end of the cover (where the fixed reflector of the main antenna sits) you can see the yellow end of the waveguide peeking out forming a horn antenna.

Here is an image without the cover, where the waveguide + horn antenna are removed. You can still see the yellow part of the waveguide where it all attaches to, with a red rectangular cover to protect it:

Now, that was the MiG-29. But what about the Su-27?

Here is a side view, the whole radar is slightly rolled towards the camera, so we can see a little from the top of the radar cover. This means that a horn antenna on the front of the cover like on the MiG-29 radar would be seen in this image. But there is none. The horn antenna could of course sit behind the radio transparent cover.

Another image:

Here the cover is removed, and the radar is tilted in such a way that we can see through the slow in the moveable reflector (where the waveguide for the main antenna sits). If the radar was build the same way as the N019, then we would see the waveguide for the guard channel throug the whole, and extending beyond the reflector. But it is not there:

Maybe the waveguide was temporarily removed in this case like on the N019 in the one image above?

But there would need to be a slot in the moveable reflector for the waveguide to pass through. The reflectors diameter is as much as the cover, so there is no space for the waveguide to fit between reflector and cover without a cut in the reflector. For example like in this MiG-radar, where you can see a rectangular cut in the top of the reflector (and the open waveguide in yellow where the compensation antenna would attaches to).

And last but not least, even the DCS model of the Su-27 does not have a seperate horn antenna for the compensation channel above the moveable reflector:

With the MiG-23 for examples some versions had the seperate horn antenna as on the MiG-29, but some variants had the compensation antenna pointed on the movable refector, according to some forum. Maybe this is the case with the N001 of the Flanker.

With either of the two locations for the compensation antenna, it is clear that the guard horn follows the main antenna on roll. So in BVR mode, the compensation antenna will START to tilt away from the ground only when the roll angle is beyond 120°. As the antenna will have a very wide beam, the sidelobe compensation will not instantly be ruined, as it is right now in DCS. The maximum angle that the compensation antenna will be turned away from the ground is 60° anyway. So it is still mostly pointing on the ground. Also on the MiG-29 the compensation antenna does not follow the main antenna elevation or azimuth, from which we can expect that the beamwidth is propably wide enough to cover the whole azimuth and the negative elevation range of the main antenna.

As the radar will be locked on roll in close combat modes (vertical scan on the MiG for example locks any target in a range of up to 10 km), it is very likely that the compensation antenna beamwidth is wide enough to work during maneuvering (with a locked roll axis, the comensation antenna is never perfectly aligned unless the bank angle is exactly 0°).

The current behaviour in DCS of losing any radar function when rolling just 1° beyond 120° in BVR STT is just wrong. Both antennas will be rotated by 1° in this case, so there is almost no difference. In DCS there is a big difference though: instant loss of the complete radar function.

And even ith a max tilt of 60° when flying inverted the guard channel will still point on the ground, but a little more to one side than to the other. Still the sidelobe compensation will not be fully broken, just to a certain extend.

And with a target that is flying towards the fighter the compensation antenna position should not play any role, as ground returns do not really matter in this case.

While I think ground returns still play a role in this, amazing finds and analysis of the pictures, great job! Much easier to visualize now:) I get what you are saying now, it would have to be in the perfect spot to lose lock after 121 degrees of role, it should lose lock any time after 120 degrees of roll depending on where the target is. Is that right?

In BVR mode for roll angles smaller than 120° the main antenna and the sidelobe suppression antenna will both be fully aligned with the horizon, there should be no degradation at all, so here DCS is correct (but there it ends).

For roll angles above 120° the compensation antenna will rotate with the plane, so it will no longer point at the center below the main radar beam but in the direction that the aircraft rolls. The amount of left/right shift depends on how much the plane rolls beyond 120°. For smaller angles above 120° there will not be a big change, it will still cover almost all of the area. But in DCS the radar is unable to do anything at any roll angle beyond 120°.

Even for larger angles (largest angle is 60° when the plane is exactly inverted) the compensation antenna is still covering much of the ground.

If we had a better image of that horn antenna we could even calculate the beamwidth. It is very likely that it has a rather high one, as it looks like the diameter of the antenna is not really larger than the actual waveguide itself, which results in a large beamwidth. The higher the beamwidth is, the less the roll angle will cause issues.

To me it seems likely that the compensation channel is only really usefull when scanning, so less false targets will appear on the radar scope. Once a proper target has been locked it is unlikely that it gets lost because the sidelobe return would have to match the target in azimuth, elevation, speed, flight direction and signal strength. Pretty unlikely.

In all the close combat modes where the targets are not shown on any scope but selected automatically (vertical scan) or by pointing the beam directly at it (helmet mode, optical mode) the sidelobe returns are less of an issue as the radar is looking at the target directly.

On the real MiG-29 the compensation channel is only used when the pilot enables it via a switch. It will then use some of the radar resources (filter etc.) that would otherwise be used for the main channel, thus resulting in a slight degradation of radar performance.

And for the SU-27 it is not even clear where the compensation antenna is. Very likely it is different than the MiG, because it cannot be seen on any image. Maybe it is incorporated into the main antenna? The IFF channel also has a sidelobe compensation channel in the Su-27 for more accurate IFF direction finding. Question is whether it is using the same compensation antenna as for the radar operation or a different one.

Still the Su-27 suffers from the same problem, without any explanation by ED.

Edit:

The IFF sidelobe suppression works quite interesting on the Su-27:

The sidelobe suppression antenna (wherever it is) and then the main antenna are emmiting IFF signals seperated in time, and the target is then comparing their signal strengths. Only when the pulse from the main antenna is stronger than the one from the sidelobe suppression antenna it will respond. So here the sidelobe detection is actually done by the target. Clever and simple!

The IFF sidelobe suppression antenna has to cover the whole azimuth and elevation range to work properly (it is intended to increase azimuth accuracy of IFF during the scanning mode. Elevation accuracy is less important, as targets will not be seperated by altitude on the radar topdown view anyway, but even in elevation it performs the sidelobe suppression by exactly the same prinicples as in azimuth). This all means that the sidelobe compensation antenna used for IFF has to have a wide beamwidth OR its beam is moving with the main beam, otherwise it could not be used in BVR scan mode.

And if the IFF is using the same antenna for sidelobe compensation as the radar is, then the radar's sidelobe comensation will not really be affected by rolling at all due to the high beamwidth OR because it is moving with the main antenna!

Let's take a look at the AN/APG-63 from the F-15C:

There are two horn antennas. The orange one at the top, and the grey one at the bottom.

Note how the bottom one is pointing downwards. If the F-15 is using this downward pointing antenna for sidelobe compensation, then it will face EXACTLY THE SAME situation as our MiG-29 with DCS logic.

Is the F-15C radar roll stabilized? What are the gimbal limits?

The top horn antenna seems to be pointing straight ahead. Height is about 3.27 cm, width is about 4.1 cm. Enter these values on the following page:

https://www.rfwireless-world.com/calculators/Horn-Antenna-Calculator.html

The radar is X-Band, so with a frequency of about 10 GHz we end up with a horizontal 3dB- beamwidth of about 50° (+/-25°) and a vertical 3dB-beamwidth of about 48° (+/-24°). The lower horn antenna is not fully visible in the image, but its dimensions appear to be the same.

So, the guard antenna(s) of the F-15 radar already have a very wide beam. Now looking at the MiG-29 radar the guard antenna is significantly smaller. The dimensions of the opening are barely any wider than the waveguide itself. With horn antennas a smaller width/height equals to a wider beam. So the MiG-29's compensation antenna covers significantly more in azimuth and elevation than that of the F-15C. This makes alot of sense, as the compensation antenna of the MiG does not follow the main beam as the compensation antenna on the APG-63 does.

The downtilt of the compensation antenna on the MiG-29 is very small, it looks like it is in the order of about or less than 10°. With a vertical beamwidth of significantly more than 48° there is almost no difference in coverage, whether the MiG is flying level or inverted.

It is time for ED to remove that unrealistic restriction for russian/soviet aircraft from the game.

Or to come up with an explanation that makes sense.

Bonus:

Downward pointing fixed compensation antennas on F-15 AESA without roll stabilisation. Inverted radar loss confirmed by Eagle Dynamics Logic!

Its about time ED make Russian missiles hit, 14% in MP no matter parameters, we fly MP not SP.

Aren't that things are on the showed image the CW transmitters for the SARH guided missiles like for the F-15A/C and you can see even on E model the "funnel" antenna?

Does the R-27R/ER need CW illumination like the AIM-7F or it can be guided without CW like the AIM-7M?

No, these russian radars don't have a dedicated CW antenna, they time-multiplex the radar operation and the target illumination.

It is roughly 20 ms of radar operation followed by roughly 30 ms of target illumination, and then the cycle repeats.

This is why it is called a discrete continuous CW.

The yellow antenna on the russian radars really is for sidelobe compensation.

The antenna in your picture also has one or two of those horn antennas for sidelobe compensation.

14 hours ago, molnibalage said:

Aren't that things are on the showed image the CW transmitters for the SARH guided missiles like for the F-15A/C and you can see even on E model the "funnel" antenna?

Does the R-27R/ER need CW illumination like the AIM-7F or it can be guided without CW like the AIM-7M?

There probably never was a CW illuminator on the F-15.   There was provision to include it should the USAF wish to use AIM-7E-2 and later variants, but AFAIK the radar never gained those components.  The FLOOD horn repeats the HSTT signal as far as I can tell.

6 hours ago, BlackPixxel said:

No, these russian radars don't have a dedicated CW antenna, they time-multiplex the radar operation and the target illumination.

Is there any note on the roll limitations in the manuals?   There is none for the F-15 but that doesn't necessarily mean anything.

14 hours ago, molnibalage said:

Aren't that things are on the showed image the CW transmitters for the SARH guided missiles like for the F-15A/C and you can see even on E model the "funnel" antenna?

Does the R-27R/ER need CW illumination like the AIM-7F or it can be guided without CW like the AIM-7M?

 

APG-63 doesnt have CW, it uses HPRF STT + Injected signal for PDI illumination. And the ER is a PD, not CW. And FWIW, the 7F was compatible with both CW and PD. only the 7D/E was CW only

4 minutes ago, dundun92 said:

APG-63 doesnt have CW, it uses HPRF STT + Injected signal for PDI illumination. And the ER is a PD, not CW. And FWIW, the 7F was compatible with both CW and PD. only the 7D/E was CW only

R-27 is discrete continuous CW, at least the R-27R1. It still uses the doppler for velocity gating.

1 minute ago, BlackPixxel said:

 

R-27 is discrete continuous CW, at least the R-27R1. It still uses the doppler for velocity gating.

The Su-27 doesnt have a CW illuminator though, and the missiles doesnt guide off of a true CW signal. While it is called "discrete continuous CW", its really just a very high PRF STT ultimately.

19 minutes ago, dundun92 said:

The Su-27 doesnt have a CW illuminator though, and the missiles doesnt guide off of a true CW signal. While it is called "discrete continuous CW", its really just a very high PRF STT ultimately.

No, during the illumination periods it is 30 ms of CW with some frequency modulation for the radio correction signal + additional commands. Not pulsed at all.