P3D v3 crash but no Event Viewer entry

[Ray Proudfoot 8,954]

I flew a Bonanza from the WAL VOR on a heading of 120° replicating yesterday’s flight in Concorde. No crash over the problematic area. I’ll try it again with Concorde to see if it’s consistently reproducible. If it is then that would be unique.

[SteveW 1,792]

On 11/23/2019 at 5:01 PM, Ray Proudfoot said:

Departing Manchester and flying westward there wasn't a problem but after reaching WAL VOR

I turned around and flew back along the same heading that previously caused a crash.

And it crashed for a third time.

Something to do with the heading perhaps. But that would be bizarre. I am truly baffled by this behaviour.

I've seen this before with navigation data embedded in the scenery for a VOR or some other navigation type, cause an issue like yours, but not with all aircraft. Could be the navigation system is choking on something unrecognisable to it.

[SteveW 1,792]

...another idea could be data from an AIRAC file set if in use somewhere.

[Ray Proudfoot 8,954]

8 minutes ago, SteveW said:

...another idea could be data from an AIRAC file set if in use somewhere.

FSL Concorde has no database of navigation data unlike the PMDG for example. I need to repeat this test and see how consistent it is. And if I can get Process Monitor running that may give a clue.

[Ray Proudfoot 8,954]

I flew my test route with Concorde and with Process Monitor running. Somewhat unusually it failed to crash P3D around the area where it has before.

The log was 4.2Gb so given it contains no useful info I have binned it. I shall try again in a day or so probably with an actual flight from Lisbon to Manchester and starting PM as I approach the danger area.

Should it crash then the hard work of decyphering the log starts!

[Ray Proudfoot 8,954]

After a couple of months flying a B737 in P3Dv4 I have returned to this crash problem in P3D v3.4.

I'm flying FS Labs Concorde from EGCC to LPPT using a standard route. The sim is crashing around the same area probably 8 times out of 10. I then reload a saved flight (FSUIPC autosave) and the rest of the flight is fine.

I was running ProcMon which captures a huge amount of data even though it's filtered for P3D only. I have compressed the file but it's still 43Mb. I've looked through it but with no experience of ProcMon I don't really know what I'm looking for. If anyone can offer some help it would be appreciated.

What I find bizarre is reloading after a crash doesn't result in a crash at the same location. I've checked the timing of AutoSave in FSUIPC but that isn't responsible as it's not the first one. I've been airborne for around 10 mins when it crashes. For those if you familiar with NW England it's an EKLAD departure out of EGCC then overflying the WAL VOR and then turning west and flying north of Anglesey then turning to the south to start the supersonic climb. The crash occurs when directly north of Anglesey.

I can make the file available via DropBox if necessary.

[Jim Young 2,356]

You are running a 32-bit app.  When I was running 32 bit apps, I kept getting crashes in the same area.  Ran ProcMon.  ProcMon indicated it was loading MyTrafficX when the crash occurred.  I disabled MyTrafficX.  Crashes stopped.  Later on I heard MyTrafficX had compatibility issues with the textures belonging to Air Force aircraft and you had to disable the AF bgl.  I did.  Problem fixed.

The above is why I keep telling everyone they have to investigate their crashes by following the instructions in the AVSIM CTD Guide, page 10, How to Fix Almost any Crash in P3Dv3.4.  You rename the P3D,cfg, scenery,cfg, and the dll.xml.  Run the scenario and see if the crashes continue.  If a crash occurs, the issue is with one of those files.  The scenery.cfg and dll.xml load possible issues.  The p3d.dll may have tweaks not compatible or not entered correctly.  I suspect the dll.xml. 

If disabling those three files does not fix the issue, then it could be a commercial aircraft installed that is incompatible.

Best regards,

Jim

[Ray Proudfoot 8,954]

Hi Jim,

I appreciate the guide can help in many situations but I'm flying a 3rd party aircraft that has 3 entries in DLL.XML so cannot revert to a default one.

It happens with no Ai traffic package running. I switched from MTX to UTL a while back.

However, I reviewed my scenery.cfg and noticed 3 entries for ORBX which I feel are unnecessary for the type of flights in Concorde. I've disabled those layers and successfully passed the point where it crashed earlier.

I'm very cautiously optimistic this may have fixed it. Enquiries on the FS Labs forum reveal others do not have this problem so it's something specific to my system.

I'll continue to fly this test route and see if it remains stable. ORBX do some unusual things with their scenery at times and it might just be a glitch in a file around the crash area.

Edited February 2, 2020 by Ray Proudfoot

[Ray Proudfoot 8,954]

Flight completed with no CTD. Further test flights will tell me one way or another if the problem has been solved.

Funny enough it only ever happens at two positions. Either north of Anglesey or close to the decel point west of Porto. Never anywhere else which is why I suspect bad BGLs.

[Jim Young 2,356]

8 hours ago, Ray Proudfoot said:

I appreciate the guide can help in many situations but I'm flying a 3rd party aircraft that has 3 entries in DLL.XML so cannot revert to a default one.

You can disable all of the entries except the 3 entries for the third party.

It looks like you might have solved your mystery or at least I hope so. 

[Ray Proudfoot 8,954]

2 minutes ago, Jim Young said:

You can disable all of the entries except the 3 entries for the third party.

It looks like you might have solved your mystery or at least I hope so. 

If I thought the DLL entries were responsible I would certainly look at editing that file. But there are only three more and they are “trusted” entries from GoFlight and Active Sky.

The fact the problem appears to be location-based and kills P3D stone dead with no warnings or entry in Event Viewer tells me it’s something out of the ordinary. Hard to recall now but there’s a possibility I copied the P3D v4 scenery.cfg and edited it for v3. The ORBX entries may relate to v4 scenery and that could be responsible for the unusual CTD.

I’ll have a better idea after more flights starting tomorrow. Appreciate your help with this Jim and that of Steve too.

[Ray Proudfoot 8,954]

I’ve given up with this problem. Performed an uninstall of P3D v3.4 and after a reboot installed a new copy. Only addons are GSX, GoFlight drivers and Active Sky 16. Nvidia drivers are 382.05. Those work with W7 and a 900 Nvidia card.

One hour into a Concorde flight from Lisbon to Manchester P3D just disappears off the display. Exactly the same place as it did yesterday on the previous install. Lefteris at FSL says Nvidia drivers no newer than 376.33 are supported. Trouble is they do not run on the 1080Ti or newer cards. Only the 900 series.

Other Concorde pilots have gone to extreme lengths to keep flying her. No Windows updates and you have to be running a 900 series card. So any 1000 or 2000 card is useless.

It’s a legacy product no longer sold and no decision yet which platform a new version will be built for.

Sad state of affairs for such an iconic aircraft. Let’s hope the future is brighter. It couldn’t get any worse really.

I’m grateful for the advice given but it seems there is no solution to this problem. 🙄 I shall now revert to subsonic mode. 😪

[Ray Proudfoot 8,954]

I’m like a dog with a bone. I can’t leave this problem unresolved. Yesterday on a virgin P3D v3 install with only GoFlight drivers, GSX and Active Sky running I flew EGLL-KJFK. For 3 hours the flight was fine but then 60nm from JFK the sim just CTD.

Looking back in the Concorde forum on FSL I was posting about my flights with the previous version - 1.40 and I never mentioned this CTD problem. 1.41 was provided in early October to resolve a bug with user settings being overwritten.

In early November I started having these CTD problems but put it down to drivers. It now seems likely these problems are now caused by 1.41 and not anything in my setup. I’ll be raising a ticket with FSL and I believe I have enough evidence to prove the problem lies in the Concorde code.

[Jim Young 2,356]

Only way to fix is to investigate.  I don't have time to check but are you monitoring your VAS.  99.9% of the time, if there is a CTD near your destination and during descent/approach, it was caused by the sim running out of VAS.  When I use 32-bit applications I always monitor VAS and then look at the FSUIPC log to see where the VAS went up and down.  I have done several FSX flights from Chicago to Wash Dulles in the PMDG 777 and run out of VAS on the approach phase 100% of the time.  VAS dives to 400MB's.  There is always a nasty storm outside too using AS16.  When there are nasty storms, I have found the CMeteoXml.dll will give me an NTDLL error.  I think it must be disabled when flying through nasty weather.  It is the weather radar gauge for Carenado. 

[Ray Proudfoot 8,954]

Jim, my post wasn't a cry for help. More a case of updating this topic with info. I can assure you it is not a VAS issue. 100%. I monitor it via a LUA plugin and over 600Mb was free.

Having heard from another user running the same version of Concorde who has no issues but has locked his OS from updates I suspect that is the reason for my problem as I allow them. I just need to find the troublesome one and uninstall it. Despite raising a ticket at FSL there has been no response so far.