Jump to content

[REPORTED]Is the GBU-38 JDAM still Borked ?


Recommended Posts

Hi, I have just got back into DCS after some time doing other things.

 

I always use the Easy-East Georgia- Spring mission for refresher training have been doing this for years. On waypoint 3 Patrol as Rockets seem to be useless I use one of my GBU-38's to take out a truck and all of the soldiers, then come back and take out the other truck with guns - always previously worked a treat. Now with the latest 'Stable' version GBU-38's miss every time. On waypoint 5 I use my wingman to take out the tanks with guided bombs - he dropped two GBU-12's and killed two tanks with direct hits then he used a GBU-38 JDAM and the bomb went into the other field ??

 

I found a thread going back to April 2019 with people having the same problem but couldn't find any discussion on this subject since. The GBU-38 was always my weapon of choice for static targets so I know how to use them (even went back to the manual to make sure), but something has gone very wrong with the accuracy of these bombs.

 

Have DCS made any acknowledgment of this problem, as no fix seems to have come out after several updates.

GAJ52

 

Intel i7-7700K @ 4.60GHz | 32 GB Ram | Win 10 Pro 64 | GTX 1080i 11.00MB | Saitek X52 Pro

Link to post
Share on other sites
I found a thread going back to April 2019 with people having the same problem but couldn't find any discussion on this subject since. The GBU-38 was always my weapon of choice for static targets so I know how to use them (even went back to the manual to make sure), but something has gone very wrong with the accuracy of these bombs.

No problems here when dropping 2/3rd to 1/3rd between the min/max range index.

 

It's possible the flight profile of the GBU-38 has changed a little with the weapon updates for the F/A-18C, etc. so there might be an issue when dropping from lower altitudes and at max range, but it's rare for me to miss provided I give a couple of seconds after the "Man Release" cue appears before releasing the bomb.

 

If you continue to have issues please post a track and I'll take a look.

i9 9900K @4.7GHz, 32GB DDR4, RTX2070 Super 8GB, 1TB NVMe, 6+4TB HD, 1+1TB SSD, MSFFB2 joystick, X52 Pro Throttle, TIR5, Win 10 Pro x64, 1920X1080

Link to post
Share on other sites

We do have an issue with GBU-38, it has been reported to the team

 

 

thanks

smallCATPILOT.PNG.04bbece1b27ff1b2c193b174ec410fc0.PNG

Forum rules - DCS Crashing? Try this first - Cleanup and Repair - Discord BIGNEWY#8703 - Youtube - Patch Status

Windows 10 Pro x64, NVIDIA MSI RTX 2080Ti VENTUS GP, Intel® i9-10900K 3.70GHz, 5.30GHz Turbo, Corsair Hydro Series H150i Pro, 32GB DDR @3000, ASUS ROG Strix Z490-F Gaming, HP Reverb G2

Link to post
Share on other sites
We do have an issue with GBU-38, it has been reported to the team

 

 

thanks

Interesting, I haven't noticed any issues. Can you provide any details about the GBU-38 issue? Just curious.

Intel i7-4790K @ 4x4GHz + 16 GB DDR3 RAM + Nvidia Geforce RTX 2080 (8 GB VRAM) + M.2 SSD + Windows 10 64Bit

 

DCS Panavia Tornado (IDS) really needs to be a thing!

 

Tornado3 small.jpg

Link to post
Share on other sites

Many thanks for your feedback on the GBU-38.

 

Its strange how some people have no problems using this weapon. I use the JDAM as per pages 563/564 of the A-10C user manual. The graphic of the HUD shows the aircraft at 6580 ft with the min and max carets far apart, this is how it was for me in the past but not now, at this height the two carets are barely apart.

 

Thanks BIGNEWY, I hope the team look at this as a matter of urgency as its definitely floored compared to previous versions and I've been flying the Warthog from day one.

GAJ52

 

Intel i7-7700K @ 4.60GHz | 32 GB Ram | Win 10 Pro 64 | GTX 1080i 11.00MB | Saitek X52 Pro

Link to post
Share on other sites

We have seen a repeat of the original bug, sometimes the GBU-38 will not hit target even if the a target has been designated correctly.

smallCATPILOT.PNG.04bbece1b27ff1b2c193b174ec410fc0.PNG

Forum rules - DCS Crashing? Try this first - Cleanup and Repair - Discord BIGNEWY#8703 - Youtube - Patch Status

Windows 10 Pro x64, NVIDIA MSI RTX 2080Ti VENTUS GP, Intel® i9-10900K 3.70GHz, 5.30GHz Turbo, Corsair Hydro Series H150i Pro, 32GB DDR @3000, ASUS ROG Strix Z490-F Gaming, HP Reverb G2

Link to post
Share on other sites

No issues here.

Mainboard: ASUS Maximus X Hero Intel Z 370

CPU: Intel Core i7-8086K @ 4.0 GHz

Memory: 32GB Corsair Dominator Platinum DDR4-3000

Graphics Card: ASUS NVIDIA GeForce RTX 3080 10GB

Monitor ASUS PA 329 32" @ 4K

1 SSD Samsung 860 PRO 256 GB

1 SSD Samsung 860 PRO 4 TB

Windows 10 - 64 V. 2004

CH Pro combatstick, throttle and pedals

Link to post
Share on other sites

Before I begin, I would like to note I performed these tests before I saw this thread was marked [Reported], but regardless I thought they might be of interest.

 

I have been part of a number of “GBU38 broken?” type threads and what has always frustrated me is there has never been a defined standard of performance we seem to be going for as a community. What I mean is that someone will pose an often anecdotal type question about GBU38 performance, some people will second it, and others will have some variety of “working fine here,” and a discussion over proper switchology will ensue. What never seems to be addressed is a quantifiable measure of what is acceptable and what isn’t.

 

I wanted to address this issue with at least an attempt to create a repeatable, measurable test that would remove as many variables as possible so we could look simply at the 38’s performance. The standard of performance I decided on was a K-Kill on a static tank. I’m sure some of the scripting wizards here would use MOOSE to determine CEP, or something to that effect; I would find that fascinating but I don’t have the time to sort that out, and to be honest I think complete destruction of a static tank could be the ultimate “gold standard” of GBU38 performance for our needs.

 

CONDITIONS

That said, I placed a static T72 at 37T GG 15858 10365. This point is the beginning of the first runway stripe on the arrival end of Rwy 13 at Batumi. The tank was aligned longitudinally with the centerline.

 

Meteorological conditions were ISA, winds calm.

 

I then placed an A10 at 8000 feet MSL 5 miles northwest of the tank, aligned down runway 13. It was armed with GBU38s on stations 3 and 9, 4000 lb fuel, full CM gun. Using the mission editor’s “setup cockpit” function, I created a waypoint in the A10 at 37T GG 15858 10365, 33 feet MSL, named it “TANK”, and saved the setup. (Interestingly enough, every time I launched the mission this waypoint would show up as GG 15857 10364. I tried numerous times to place a new aircraft and “setup cockpit” again, double checking my entry, but for some reason when I would launch the mission the waypoint was always 1 meter west and 1 meter south of where it was placed. This means the steerpoint was 1.4 meters from the center of the tank, still under its footprint, but I thought it worth mentioning in the spirit of complete disclosure).

 

Next, I started this mission 20 times. I performed the following steps each time…

-Select FUNC and MSN on the UFC to give me access to the TANK waypoint

-DMS FWD (short) to select the TANK waypoint (STRP is already SPI)

-DMS RIGHT (short) to select GBU38 CCRP

-Fly to the following target parameters and release a single GBU38: 275 KIAS, 8000 MSL, 2.1 NM to steerpoint (~1/2 through the GBU38 CCRP engagement envelope), 1.0 G.

-Let the bomb land, save the track, and repeat.

 

After 20 iterations I went through the tracks and documented my actual flight parameters at release and the result against the tank. I would also point out I took no “alibi” passes or any such thing. Once I resigned myself to not being able to resolve the 1.4 meter waypoint discrepancy discussed above I recorded tracks of 20 consecutive passes regardless of result.

 

RESULTS

Drop parameters were held as consistent as possible. Across the 20 drops, standard deviation from the target airspeed of 275 KIAS was 1.47, with an extreme maximum of 279 and minimum of 274. Standard deviation from 8000 feet MSL was 6.54 feet, with all drops occurring between 8010 and 1790 MSL. Distance from the 2.1 NM release target had a standard deviation of .03 NM and a range of 2.2 to 2.0. G was kept at .9 on 5 of the runs, with the remaining 15 at 1.0 G.

 

In every drop the tank was damaged to some extent. (Hopefully someday M-Kills and F-Kills will be modeled in DCS). 75% of the drops resulted in a catastrophic kill of the tank.

 

CONCLUSION

This trial presented the A10C and GBU38 with conditions for weapons release as close to ideal as possible; no enemy threat, no wind, clear skies, and most importantly a SPI set perfectly at the target elevation and within 1.4 meters of its center. All that said, 25% of the time the target was not destroyed in one shot. Said another way, an A10C carrying 4 JDAMs can reasonably expect one of them to miss under near perfect conditions, to say nothing of the errors that can creep in when trying to designate targets with the targeting pod, let alone considerations for enemy fire or adverse weather. Under the current state of the game, this tells me that when a tank positively, absolutely has to go down in one pass, I will be leaning on the Maverick and LGBs.

 

I will post an example track of my trials, #16 (a miss) and a PDF of my recorded data from the entire series. I welcome any questions or critique of my method.

GBU38 Trial Results.pdf

GBU 38 Take 16.trk


Edited by tom_19d
Typo in results paragraph corrected

Multiplayer as Variable

 

Asus Z97-A - I7 4790K - 32 GB HyperX - EVGA GTX 1080 Ti - Corsair 750i PSU

 

TM Warthog HOTAS - TM Cougar MFDs - CH Pedals - TrackIR 5 - Samsung RU8000 55”

Link to post
Share on other sites

Did you try creating a mission point via the CDU using that 10-digit grid as if passed by a JTAC or FAC(A)?

 

I wonder if there’s some kind of “rounding” going on in the ME and I wonder if that happens in the CDU as well?

 

I’m curious what MGRS coordinates the TGP would yield in Point Track on that tank?

[sIGPIC][/sIGPIC]

http://www.476vfightergroup.com/content.php

High Quality Aviation Photography For Personal Enjoyment And Editorial Use.

www.crosswindimages.com

Link to post
Share on other sites
Did you try creating a mission point via the CDU using that 10-digit grid as if passed by a JTAC or FAC(A)?

 

I’m not sure what you are asking here, as I only created the target waypoint via the CDU. Setup cockpit in the ME puts you in engine so you can start the mission with the avionics programmed, I used zero ME generated waypoints in the trials.

 

 

I’m curious what MGRS coordinates the TGP would yield in Point Track on that tank?

 

TGP 3.9 miles with a point track, Laser on, GG 15857 10366, it is within a meter both directions.

Multiplayer as Variable

 

Asus Z97-A - I7 4790K - 32 GB HyperX - EVGA GTX 1080 Ti - Corsair 750i PSU

 

TM Warthog HOTAS - TM Cougar MFDs - CH Pedals - TrackIR 5 - Samsung RU8000 55”

Link to post
Share on other sites
I’m not sure what you are asking here, as I only created the target waypoint via the CDU. Setup cockpit in the ME puts you in engine so you can start the mission with the avionics programmed, I used zero ME generated waypoints in the trials.

 

 

 

TGP 3.9 miles with a point track, Laser on, GG 15857 10366, it is within a meter both directions.

 

 

I misunderstood your post. I took it to mean you assigned the waypoint IN the ME. My apologies...

[sIGPIC][/sIGPIC]

http://www.476vfightergroup.com/content.php

High Quality Aviation Photography For Personal Enjoyment And Editorial Use.

www.crosswindimages.com

Link to post
Share on other sites
I misunderstood your post. I took it to mean you assigned the waypoint IN the ME. My apologies...

 

No worries at all. I was just trying to do everything I could (direct entry of a perfect grid and elevation into the CDU, no TGP, etc) to take away every variable of employment and putting all the performance on the bomb itself...

Multiplayer as Variable

 

Asus Z97-A - I7 4790K - 32 GB HyperX - EVGA GTX 1080 Ti - Corsair 750i PSU

 

TM Warthog HOTAS - TM Cougar MFDs - CH Pedals - TrackIR 5 - Samsung RU8000 55”

Link to post
Share on other sites
RESULTS

In every drop the tank was damaged to some extent. (Hopefully someday M-Kills and F-Kills will be modeled in DCS). 75% of the drops resulted in a catastrophic kill of the tank.

 

Interesting.

 

For info.that's in agreement a Weapons Matrix for the AV-8B that gives a tank PK of 50-75% for a GBU-38 with the best impact angle and fusing (top preference with a PK of 75-100% was LMav or GBU-32).


Edited by Ramsay
typo

i9 9900K @4.7GHz, 32GB DDR4, RTX2070 Super 8GB, 1TB NVMe, 6+4TB HD, 1+1TB SSD, MSFFB2 joystick, X52 Pro Throttle, TIR5, Win 10 Pro x64, 1920X1080

Link to post
Share on other sites
Interesting.

 

For info.that's in agreement a Weapons Matrix for the AV-8B that gives a tank PK of 50-75% for a GBU-38 with the best impact angle and fusing (top preference with a PK of 75-100% was LMav or GBU-32).

 

Thanks, that is interesting too, and it is just the kind of data driven discussion I think is needed! Do you have any insight on the difference there between the GBU 38 and 32? I would speculate accuracy isn’t really at issue, but that the PK is higher with the 32 simply due to increased mass.

 

I’m just fine with the idea that our DCS GBU38 should miss on occasion even under the best conditions, but when the discussion is always binary (aka “it is broke” or “it works fine here”), rather than percentage driven, it is hard to judge what is actually going on in the game.

Multiplayer as Variable

 

Asus Z97-A - I7 4790K - 32 GB HyperX - EVGA GTX 1080 Ti - Corsair 750i PSU

 

TM Warthog HOTAS - TM Cougar MFDs - CH Pedals - TrackIR 5 - Samsung RU8000 55”

Link to post
Share on other sites
  • 6 months later...

@BIGNEWY

 

any updates on the status of this?

 

Since the OB 2.5.6.47224 I've experienced some inconsistencies in the precision of the GBU38s

 

Thanks

ASUS ROG STRIX Z490 F-GAMING | i7-10700K | RTX3090 TUF OC | 32GB DDR4 3200Mhz | Windows 10 64bit

Acer Predator X34P | TrackIR 5 | TM Warthog | TM T.Flight Rudder Pedals

HP Reverb

 

A-10C | A-10C II | F/A-18C | F-16C | FC3 | PG | Syria | SC

Link to post
Share on other sites

I've done some testing this morning using the last OB 2.5.6.47404 and Stable 2.5.5.41371.

 

 

Attached are the following tracks:

 

- A10C GBU38

- A10C GBU38-2

- A10C GBU38-3

 

these are done with OB.

 

I've tried various method to assign the SPI and then releasing the bomb but the result are inconsistent. It's either a perfect hit or a miss by many meters.

 

- slew TGP over the boat

- TMS fwd short to point track

- TMS fwd long to assign SPI

- pickle

 

next step was using the markpoint

- slew TGP over the boat

- TMS fwd short to point track

- TMS right to assign markpoint

- CDU mark point

- TMS aft long SPI to markpoint

-pickle

 

Also tried lasing before assigning the SPI

- slew TGP over the boat

- TMS fwd short to point track

- firing laser and hold

- TMS fwd long to assign SPI

- stop lasing

- pickle

 

Releasing is within parameters at various altitudes but in general above 8/9000ft, decent speed, wings level and either level flight or a very shallow descend before releasing.

 

I'm playing the sturmovik campaign where you have to target the 2 missile boats in the harbor of Batumi but I've seen the same hit/miss in any other missions/campaigns

 

Sometimes if you go to the F6 view the you can see the bomb flight path is not steady but it's oscillating quite a bit. Not sure if this is intended or else :dunno:

 

 

 

I am not sure if I'm doing something wrong or there's still a bug related to the GBU38s. (I am aware that the precision of GBU38s is being worked on since 2.5 but I would like to give some more evidence to the devs.)

 

 

 

The last track (A10C GBU38-4 stable) is done with the stable version but the results are the same (1 miss by a lot and the second perfect hit)

 

If anyone can have a look at the tracks and help me it would be very much appreciated. :helpsmilie:

 

Thanks!

A10C GBU38.trk

A10C GBU38-2.trk

A10C GBU38-3.trk

A10C GBU38-4 stable.trk

ASUS ROG STRIX Z490 F-GAMING | i7-10700K | RTX3090 TUF OC | 32GB DDR4 3200Mhz | Windows 10 64bit

Acer Predator X34P | TrackIR 5 | TM Warthog | TM T.Flight Rudder Pedals

HP Reverb

 

A-10C | A-10C II | F/A-18C | F-16C | FC3 | PG | Syria | SC

Link to post
Share on other sites
I've done some testing this morning using the last OB 2.5.6.47404 and Stable 2.5.5.41371.

 

Excellent tests, thanks a lot! :thumbup:

 

I'll group the tracks with the respective description to make sure I'm understanding you correctly.

 

- A10C GBU38

[...]

- slew TGP over the boat

- TMS fwd short to point track

- TMS fwd long to assign SPI

- pickle

 

Looked good. I observed one perfect hit for the first GBU-38. Can't say whether or not the second one was a good hit because of TGP masking.

 

Good execution, both weapons should have hit.

 

With stationary targets, it's not necessary to go to Point Track. Area will do just fine. However, if the boats were to start moving, Point Track will keep the TGP on target, so that's a plus. Then again, it won't help the JDAMs, so all things considered... Point Track doesn't hurt in this example, but it's not necessary either.

 

- A10C GBU38-2

[...]

next step was using the markpoint

- slew TGP over the boat

- TMS fwd short to point track

- TMS right to assign markpoint

- CDU mark point

- TMS aft long SPI to markpoint

-pickle

 

In this track, I couldn't see you attacking a markpoint. SPI Generator was set to TGP for both attacks, just like in the first track.

 

Anyway, TGP was spot-on and both weapons should have hit. Instead, I observed one close and one far miss, if memory serves (should have taken notes :huh:).

 

Note: never ever push the stick forward when releasing bombs. Doesn't matter that you were in a dive, but the dive was actually increasing for one of the attacks. Yes, I learned this the hard way - my jet was once detonated by the bomb it had just released. :music_whistling:

 

- A10C GBU38-3

[...]

Also tried lasing before assigning the SPI

- slew TGP over the boat

- TMS fwd short to point track

- firing laser and hold

- TMS fwd long to assign SPI

- stop lasing

- pickle

 

Again, I observed two misses, but with the TGP spot-on, both weapons should have hit.

 

There's an error in the methodology, though. The SPI is continuously generated. As long as you lase, it'll be more accurate (very precise distance measurement). As soon as you stop lasing, it'll be back to a little less precise math (your aircraft's position in space, with TGP line of sight intersecting the ground at the elevation given by the digital elevation database - and not intersecting the ground at the exact position you're looking at).

 

In this track, it doesn't matter much; the angle between your aircraft and the target at weapon release was so high, the LOS ground intersection would be almost identical with and without lasing, and would not account for the 20+ meters offset that the weapon missed in my replay.

 

It's important to understand, though: with TGP as SPI Generator, it's perfectly useless to lase the target, unless you do something with that precise measurement, like dropping a weapon or creating a markpoint. Otherwise, the measurement will have been more precise while lasing, and again less precise just as soon as the lasing stops, for no good reason.

 

I am not sure if I'm doing something wrong or there's still a bug related to the GBU38s. (I am aware that the precision of GBU38s is being worked on since 2.5 but I would like to give some more evidence to the devs.)

 

GBU-38s used to be pretty spot-on in earlier versions.

 

From what I've read, they now tend to miss by maybe a little too much, and too often, compared to their expected RL performance, and it seems to be fairly random. Whether or not that's a bug in DCS, I don't know. But I'm fairly sure that in your tracks, you didn't have any switchology errors and each and every GBU-38 had a 100% chance to hit the intended target.

 

The last track (A10C GBU38-4 stable) is done with the stable version but the results are the same (1 miss by a lot and the second perfect hit)

 

Didn't look at this track, but I think the failure rate of GBU-38 far pre-dates 2.5.5, so it's no surprise that current stable and current OB show the same hit/miss behavior.

Link to post
Share on other sites

Thanks a lot Yurgon for your reply! Lot's of very nice points!

 

 

 

 

In this track, I couldn't see you attacking a markpoint. SPI Generator was set to TGP for both attacks, just like in the first track.

 

Anyway, TGP was spot-on and both weapons should have hit. Instead, I observed one close and one far miss, if memory serves (should have taken notes :huh:).

 

Note: never ever push the stick forward when releasing bombs. Doesn't matter that you were in a dive, but the dive was actually increasing for one of the attacks. Yes, I learned this the hard way - my jet was once detonated by the bomb it had just released. :music_whistling:.

 

 

For track no.2 I might have saved the wrong track regarding the markpoint... don't really know sorry :P

 

Duly noted about the pitch down when releasing bombs :thumbup: not really sure why I did that :doh:

 

 

Again, I observed two misses, but with the TGP spot-on, both weapons should have hit.

 

There's an error in the methodology, though. The SPI is continuously generated. As long as you lase, it'll be more accurate (very precise distance measurement). As soon as you stop lasing, it'll be back to a little less precise math (your aircraft's position in space, with TGP line of sight intersecting the ground at the elevation given by the digital elevation database - and not intersecting the ground at the exact position you're looking at).

 

In this track, it doesn't matter much; the angle between your aircraft and the target at weapon release was so high, the LOS ground intersection would be almost identical with and without lasing, and would not account for the 20+ meters offset that the weapon missed in my replay.

 

It's important to understand, though: with TGP as SPI Generator, it's perfectly useless to lase the target, unless you do something with that precise measurement, like dropping a weapon or creating a markpoint. Otherwise, the measurement will have been more precise while lasing, and again less precise just as soon as the lasing stops, for no good reason.

 

 

Regarding lasing: I have read a lot of controversy about this method and to be honest I have never used it before this testing, however I wanted to give it a try to see if it would make any difference (which it didn't)

 

Your explanation about lasing makes total sense and I agree with you =)

 

GBU-38s used to be pretty spot-on in earlier versions.

 

From what I've read, they now tend to miss by maybe a little too much, and too often, compared to their expected RL performance, and it seems to be fairly random. Whether or not that's a bug in DCS, I don't know. But I'm fairly sure that in your tracks, you didn't have any switchology errors and each and every GBU-38 had a 100% chance to hit the intended target.

 

I remember that as well, used to be almost a guaranteed hit every time you were dropping a JDAM. Hopefully they'll be fixed soon, happy to hear that at least my method is somewhat correct :D

 

Thanks again!

ASUS ROG STRIX Z490 F-GAMING | i7-10700K | RTX3090 TUF OC | 32GB DDR4 3200Mhz | Windows 10 64bit

Acer Predator X34P | TrackIR 5 | TM Warthog | TM T.Flight Rudder Pedals

HP Reverb

 

A-10C | A-10C II | F/A-18C | F-16C | FC3 | PG | Syria | SC

Link to post
Share on other sites
  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...