Yves77

The problem of the fuel overload appears if you change from one PMDG livery to another without exiting and restarting the sim. This is a new bug introduced with the P3Dv4 version of the PMDG planes. I've never had it in FSX, P3Dv2 and v3 (and could not reproduce it now). The PMDG777 is affected the same way. I don't have the other PMDG aircraft yet. The manual states that PMDG planes should not be loaded on top of each other or on top of a 3rd party aircraft. But this was written years before the fuel bug existed. Switching the livery or the PMDG aircraft type has worked anyway in P3Dv3 most of the time, at least the errors (if any) were not directly visible. Now in P3Dv4 the fuel bug forces a restart. This might explain the several threads in this forum about this issue since the P3Dv4 update was released (the first thread here is from July 2017). I did some investigation of what exactly happens with the fuel tanks in the PMDG737-800: Upon reloading the 737NGX the current fuel amount (that was in before reloading) in the center tank is multiplied by 42.99. The right and left tank amount is multiplied by 12.88 each. Where are these strange figures from? They are defined in the aircraft.cfg under [fuel]. The capacity of the center tank is 4299 US Gals, left and right tanks hold up to 1288 US Gals. [fuel] Center1= -1.884, 0.000, -1.800, 4299.000, 0.000 LeftMain= -9.110, -16.500, 1.600, 1288.000, 0.000 RightMain=-9.110, 16.500, 1.600, 1288.000, 0.000 After each consecutive reload or livery switching the fuel amount is again multiplied by these numbers. The gross weight becomes so high that the planes sinks into the ground and starts shaking. The true fuel amount in the tanks is always indicated in the regular payload & fuel menu of P3D. The FMC fuel indication is wrong in this situation, only the gross weight is correct. You can reduce the fuel amount in the FMC fuel page, but it won't help. If you click 1/3 or 2/3 the tanks are correctly refilled, but the FMC shows a much too low fuel level. As a consequence, one gets an engine flameout after a few minutes, because the FMC *thinks* the tanks are empty (the flameout seems to relate on the FMC indication, not on the true fuel amount). I hope PMDG will have a look at this, although it only happens outside of their recommended usage.

Yes, I'm aware of these restrictions stated in the manual. That fuel overload effect is probablly such a side effect of loading "on top of each other" (I'd love to get a more technical explanation). But it has never appeared here until the micro update last week. So far, a livery change was in most cases unproblematic in all these years flying with the NGX. I only ever use the FS Action Fuel Screen to change the fuel and payload in PMDG planes. However, the P3D mass and balance screen shows the true weight used by P3D/FSX and is the reference in case of deviations from the FMC.

That issue with the fuel tanks is really strange, i've never seen it before. The tanks are filled and filled and filled after every livery change. I used the default panel states (NGX short, NGX long), but it had no influence on the issue. I also deactivated fsuipc for the tests. After start-up of P3D v4 everything is fine with the first livery of the NGX. I then set the fuel to "full" (100%) in the FMC fuel page (FS Actions -->Fuel). After that, I select a different livery within the same model type (winglets or no winglets). This does not trigger the warning about loading a default flight first. Once the second livery has initialised, the NGX automatic fuel system gets out of control. The P3D fuel & payload reveals that the center tank was filled from 13 tons to 561 tons (43 x above capacity), tank 1 and 2 each increase from 3.9 to 50.4 tons. The FMC fuel page also shows the correct new gross weight of 717 t and a fuel level of 2290%. But the displayed fuel amount in the three tanks is incorrect (they represent the target 100% fuell.). The plane starts dancing. When I click "Set full" in the FMC fuel page, the physical tanks in P3D are correctly reset to 100% and the bouncing stops. But the FMC shows incorrect values of only 300 kg in each tank. Switching to other NGX liveries sometimes increased my gross weight even more without touching anything else. I stopped at 25'000 tons (330 x max. gross weight). Yves

I've experienced the same or a similar thing today in P3Dv4 HF1 and the latest PMDG update. Once the 20 seconds initialisation of the 738 has finished, there is a huge amount of fuel automatically injected into the tanks, much higher than the tank capacity. My 738 suddenly had a gross weight of almost 800 tons! It then sunk into the ground and bounced all the time. When I switched to a default plane, this one also had massive overweight. In the FMC under FS Actions I then set the fuel level to 2/3 and all was good again. In my case I loaded the scenario screen with a default plane and then selected the PMDG 738 (as recommended). After the initialisation I selected a different livery and only then the problem appeared. I repeated the process several times and it's the same with the 739. I'm not sure yet if it happens with the 777.

Thanks, your hint to the ipcReady.lua was helpful. I added the Multiaxis code in it right below the default Linda entry. This didn't start the script, however when I also keept the Multiaxis code in its own lua file and in the [Auto] section as before, I got it working again. The crash did not appear anymore on aircraft change! I then edited the code in ipcReady.lua many times and did a lot of tests. The result is not clear yet. It seems that just any code in ipcReady.lua could stop the crash, even if there are only comments in it or just the function-end-block. But in some cases the Multiaxis script stopped working randomly on one plane and then restarted when another plane was selected (not related to any specific aircraft). With some code the crash still happened occasionally after maybe 6 to 7 aircraft changes, sometimes earlier or later. I can only guess that there is a timing issue between Simconnect and FSUIPC on aircraft change. Maybe too many tasks or scripts are trying to access simconnect at the same time (my orginal issues was with a fresh installed P3dV4). Even the different log functions of fsuipc could have an impact. I'll do some more tests. I can report my results directly at http://forum.simflight.com/topic/84157-multiple-lua-files-crashing-p3d-on-aircraft-change/ Yves

Thanks for posting the problem on the fsuipc support forum. Unfortunately, Version 5.121a didn't resolve the issue here, neither with my multiaxis.lua nor with the test script flapsdisplay.lua I posted earlier (which just gives a simconnect message when flaps are moved). With ipcReady.lua in the [Auto] section of fsuipc.ini there's still a crash of fsuipc5.dll and p3d on aircraft change exactly like before. The Windows event viewer and fsuipc5.log file also look identical.

Okay, thanks for checking. So let's wait and see what Pete thinks could cause the crash. Yves

Today I re-checked the situation again in P3Dv3 and I got one different result than yesterday. With FSUIPC 4.971 I get the crash also with LINDA 3.0.0b and 2.9.6. I went back to FSUIPC 4.966c (17th April 17) and this one worked fine with LINDA 2.9.6 and 3.0.2c (no crash on aircraft change). FSUIPC 4.971 itself does not cause a crash on aircraft change when multiaxis.lua is running. It's just the combination of ipcready.lua and multiaxis.lua. Let's see what Pete thinks about it. BTW my P3Dv3 log files had the "Debug/Trace Lua Plugins" set in FSUIPC logging page. That's why they got large. Yves

Thanks for the responses. I did some more checks with different versions of LINDA and in P3D v3.4 as well as in v4. Both with the same critical lua script (multiaxis.lua). The results are: P3D v3.4 with FSUIPC 4.71: LINDA 3.0.0b (build 673): No crash upon aircraft change LINDA 3.0.1: Crash when LINDA is started by ipcready.lua. LINDA 3.0.2c: Crash when LINDA is started by ipcready.lua. P3D v4.0 HF1 with FSUIPC 5.113 (and previous versions back to somewhere 5.102 or earlier) All three LINDA versions crash upon aircraft change when LINDA is started by ipcready.lua. Andrew, I sent you these 6 fsuipc.log files per mail.

Hi everyone, I have the same or a similar problem in P3D v4 (HF1) with lua scripts that have an EVENT command in it (e.g. event.param or event.control). The scripts themselves work correctly with the first plane that is loaded. Also LINDA starts up normally ("LINDA ready ..." - all buttons and lua scipts functioning). When I select a different (default) plane in the P3D menu, that plane is loaded, but after 2 seconds I see "LINDA Restarting. PLEASE WAIT! Do not operate buttons". This message remains on-screen and then P3D shows the usual unspecified crash message. This didn't happen in v3. Without LINDA the lua scripts run normally and no crash appears. Workaround: It only happens if LINDA is started by calling ipcReady.lua via fsuipc5.ini. I deactivated ipcReady.lua (renamed .lua extension) and then started P3D. I called LINDA from the system tray and in the GUI I clicked 'Reload LUA Engine' to activate my button assignments in LINDA (no problems here). I could now switch to many different aircraft without a crash and the lua scipts still functioning. The above 'Restarting' message appears for a fraction of a second and is then replaced by "LINDA ready..." (as it should). I have tested this now dozens of times and the results are 100% predictable. Versions used: LINDA 3.0.1 and 3.0.2c FSUIPC 5.113, 5.112, 5.11, 5.10 Two examples for affected Lua scripts below. Only the event commands are critical, ipc comands seems not to cause a crash. flapsdisplay.lua (used for test purposes): function flaps_dec(controlnum, param) ipc.display("Flaps decrement", 2) end event.control(65759, "flaps_dec") Multiaxis.lua (conditional axis based on button state - by Pete Dowson, 2011) : function applyaxis(val) cntrl = ipc.readUD(0x66C0) ipc.control(cntrl,val) end event.param("applyaxis") Win 10 Event viewer: Error 1. .NET Runtime (1026): Application: Prepar3D.exe Error 2. Application Error (1026): Faulty Application: Prepar3D.exe, Faulty Module: FSUIPC5.dll fsuipc5.ini: [Programs] RunIf1=CLOSE,L:\P3D v4\Modules\linda.exe [Auto] 2=lua Multiaxis [LuaFiles] 1=linda 2=PMDG777 Trim 3=logging 4=ipcReady 5=Multiaxis 6=ipcDebug 7=Flapsdisplay FSUIPC5.log (Verbose): The log is very long due to constant queries of the .air-file header, here are just the key lines: ********* FSUIPC5, Version 5.113 (21st August 2017) by Pete Dowson ********* Running inside Prepar3D v4 Module base=7FF8BB720000 Windows 10 Pro 64 Bit reported as Build 15063, Release ID: 1703 (OS 10.0) Prepar3D.exe version = 4.0.28.21686 ... 203 Run: "L:\P3D v4\Modules\linda.exe" 235 LogOptions=70000000 0000001D 235 ### Traffic limited by user parameter to 100 aircraft 235 SimConnect_Open succeeded: waiting to check version okay 235 Opened separate AI Traffic client okay 235 VRI port 1 "com1" failed to open This is the last part where I switch from Beech Baron 58 to Maule M7 and P3D crashes: . 133672 Lua threads being terminated: 133672 1 = "L:\P3D v4\Modules\Multiaxis.lua" 133828 LUA: "L:\P3D v4\Modules\Multiaxis.lua": killed 133828 Aircraft="Maule M7 260C" 133844 LUA.2: LINDA:: [START] Airplane Set event caught! 133844 LUA.2: LINDA:: [START] ************************************************************* 133844 LUA.2: LINDA:: [START] New Aircraft Selected - restarting... 133860 LUA.2: LINDA:: [START] ************************************************************* 133860 LUA.2: LINDA:: [START] Current Aircraft: Maule M7 260C 133860 LUA.2: LINDA:: [START] Air file: L:\P3D v4\SimObjects\Airplanes\Maule_M7_260C\Maule_M7_260C.air 133860 LUA.2: LINDA:: [START] searching available aircraft modules.. 133875 LUA.2: LINDA:: [START] s = FSX DEFAULT 133875 LUA.2: LINDA:: [START] h = FSX Default 133875 LUA.2: LINDA:: [START] ************************************************************* 133875 LUA.2: LINDA:: [START] Aircraft module detected: FSX Default 133891 LUA.2: LINDA:: [START] ************************************************************* 133891 LUA.2: LINDA:: [START] Calling Initialisation... 133891 LUA.2: LINDA:: [START] Path = linda/system/init 134078 Ready Flags: Ready-To-Fly=Y, In Menu=N, In Dlg=N 134094 LUA: "L:\P3D v4\Modules\linda/system/init.lua": killed 134125 LUA.0: beginning "L:\P3D v4\Modules\Multiaxis.lua" 134907 LUA.0: beginning "L:\P3D v4\Modules\linda/system/init.lua" Edited to remove excess lines - scotflieger Regards, Yves