PMDG 747 for P3D V4 - Autopilot problem(s)

[CajunRon 8]

I am having numerous problems with the PMDG 747 for P3D V4 autopilot.   First was the problem of the autopilot tripping out continuously.   I solved this by widening the null zone on my CH Products yoke axes beyond the default via the Prepar3d controls calibration (though I've never had this problem with any other other simulator or any other PMDG product but the controller is getting long in the tooth).

My next problem is the autopilot not following the FMC flight plan.   I set up a flight from KJFK to KSFO on the FMC.   (I've been very successful for several years doing this on the 737 NGX and the 777).  I also went through the 747 tutorial flight just to be sure it wasn't a far deviation from what I was use to on the 777.   I've had numerous successful flights with the 747 but then yesterday I installed the "update" issued by PMDG.   This was my first attempt after loading he update.   My inital attempts at getting a stable flight via the autopilot were disastrous.   Though everything seemed stable with the PFD showing SPD, LNAV and VNAV modes active and the autopilot engaged, the plane when into a steep dive after a few minutes of flight while my attention was being given to a PACK 2 error and high temp in the Cargo bay.   Okay, maybe I had left the option for random failures on and maybe this was affecting the flight controls.  So I tried again.  Everything seemed to work fine, no cautions or warnings of any type.   I reached cruise altitude with no problems.   Since this was going to be a long flight I setup for a 4X acceleration through the cruise function in the FMC and went about my business with other things.  About half way through the flight, I checked and found the plane was about 150 miles off course AND heading in a southeast direction (instead of the westerly course established by the flight plan).   The PFD still showed SPD / LNAV / VNAV Path as the active modes.  The FMC still showed the remaining 15 or so waypoints except for the STAR and  approach which I hadn't programmed yet.   The LNAV , VNAV and CMD buttons were all illuminated.  There were no illuminated buttons on the overhead panel.  The only caution messages was a deviation from altitude (about 400 ft or so from plan) and a fuel imbalance message.  Checking the fuel levels in each tank I had 29,900 in each of main tanks 1 and 4 and 12,700 in each of main tanks 2 and 3.  All other tanks were empty.   No fuel imbalance that I could see.   I'm not sure I understand what else would trigger a fuel tank imbalance caution.   I will be the first to admit that I am not anywhere near an expert at all of the and this could possibly be pilot error.   But I do have over 30 years experience in flight siming including years with the PMDG 737 NGX and 777 with not problems like this ever. Certainly I am open to any and all suggestions on what I may have done wrong but at this point in time, my belief is that there is a serious bug in the model...again, I could be wrong about this.  I will exercise reasonable patience as these bugs are worked out.   I am use to that with new products and models and fully understand the need for patience.  But I do have EXTREMELY HIGH expectations of an error free model because of the high price paid.   This high price will cause my patience to wear very thin sometime in the future if the perfection of the model doesn't match the price paid.

FYI, I had the following running as well

Beta version of ActiveSky 2016 for P3D V4

ProATC modified to work with P3D V4

The other change I made was to install FSUIPC 5 and begin using it for joystick axis assignment and calibration.

Any help or comments will be appreciated.  

(I am considering uninstalling the 747 and reinstalling the initial version and also removing FSUIPC to see if I experience the same problem.  I assuming there isn't a limit to the number of installs of the 747 we are allowed as there is with P3D.   Is this correct???)

Thanks

[Captain Kevin 586]

What happens if you get rid of FSUPIC 5. It has been known to cause issues.

[downscc 1,327]

Are you loading a previously saved flight?  Rule out corrupt panel states.  Don't set it on x4 and walk away until you are satisfied all is working as expected.

Rest assured yours is a isolated trouble report so there's not a bug in the product you need to verify, simply figure out what when wrong in your configuration.

FSUIPC is up to version 5.102 now, best to have the latest.

[Brian_A_Neuman 198]

I am fearful that it is the FSUIPC interface with the controller. If it is sending weird signals it could blow the A/P right off. 

 

I am between PANC and KDFW and the A/P is functioning marvelously.

[Budbud 227]

12 hours ago, CajunRon said:

Checking the fuel levels in each tank I had 29,900 in each of main tanks 1 and 4 and 12,700 in each of main tanks 2 and 3.  All other tanks were empty.   No fuel imbalance that I could see.   I'm not sure I understand what else would trigger a fuel tank imbalance caution.

Your fuel imbalance comes from a wrong fuel pump configuration. If the fuel quantity on the main 2 and 3 tanks is higher than on the tanks 1 and 4, then you should have the all the pumps 1, 2, 3, 4 and override/jettison pumps 2 & 3 ON and the 4 xfeed valves open.

When the quantity in mains 2 & 3 decreases to the same quantity as on tanks 1 & 4, then you must switch off the override/jettison pumps 2 & 3 and close the XFEED valves 1 and 4.

Fuel imbalance message comes when you keep the override pumps 2&3 and XFEED valves 1&4 on/open while the quantity in main 2&3 is lower in tanks 1&4.

You should never have less fuel in the tanks 2&3 than in the in the tanks 1&4 in nomal ops.

Check the fuel management in the tutorial and in the FCOM.

Fuel management in the 747 is far less automated than the 777 and more complex than in the 737.

[scandinavian13 5,192]

14 hours ago, CajunRon said:

I'm not sure I understand what else would trigger a fuel tank imbalance caution.

It would be triggered by the clear fuel imbalance between the tanks. It's not referring to a left/right imbalance. It's referring to the fact that you let the plane cook several tons out of the inner tanks such that the balance of fuel between the inner tanks and outer tanks is completely off. Please read the Intro Manual.

[CajunRon 8]

I tried to uninstalled and reinstalled the 747 but I got dark screens and no mouse action in the cockpit.  I then uninstalled removing every trace of the 747 I could find in my directories but still got dark screens and no mouse action.    I had to reinstall the update as well to fix that problem.  I unassigned my CH Product Yoke from FSUIPC and assigned the controls to P3D and widened the dead zone slightly beyond the default.  I also did not run AS2016 nor ProATC.   After stabilizing at cruise altitude and lefting it run a little while to see that it remained stable I then ran the rest of the flight at 4X acceleration until right before the TOD.  Everything went smoothly with no problems.   I switched over to normal time rate and let the FMC begin the decent.  Again everything was fine.   My wife then called me for help with a short chore.   I left the sim running with the FMC still controlling the descent.  The STAR and APPROACH had already been entered into the FMC and all discontinuities removed.    I was gone for about 2 minutes.   When I returned the autopilot had disengaged and the plane was in a hard left bank nose dive.   (I thought that with the autopilot having everything in trim that the plane would continue flying relatively stable in a trimmed state at least for a while if disengaged.)   Anyway for over two hours the autopilot held, even in 4X acceleration, but while away from the computer for 2 minutes (MURPHY'S LAW) with absolutely no movement of the controllers, the autopilot tripped out.  I'm assuming the autopilot in the 747 is programmed to be much more sensitive than that programmed in PMDG's 737 or 777 since since I didn't have to widen the dead zone in those.   I'm suspicious this is part of the problem. 

I did not completely remove the FSUIPC module but should have. I'll try that next.   

Thanks for the education on the fuel tank imbalance.  I was not aware of the need to balance all four tanks.  Just thought it was a left / right imbalance that caused a problem.  I need to review the manual on fuel management again.

[Captain Kevin 586]

2 hours ago, CajunRon said:

I tried to uninstalled and reinstalled the 747

Unless you have exhausted every option and nothing else has worked, this is the last thing you should be doing, as that rarely solves anything. If anything, that usually creates more problems. So far, nobody has told you to do that.

[scandinavian13 5,192]

17 hours ago, CajunRon said:

I tried to uninstalled and reinstalled the 747 but I got dark screens and no mouse action in the cockpit.  I then uninstalled removing every trace of the 747 I could find in my directories but still got dark screens and no mouse action.    I had to reinstall the update as well to fix that problem.

Uninstall/reinstall: don't. As Kevin mentioned, this should be an absolute last resort.

Uninstall/reinstall all traces: even more don't. If done improperly, this will cause even more problems than you're intending to solve.

Installing the update to solve the dark screens / no mouse issue: this is one of the reasons I said the above, and the reason you had to do it is because you have FSUIPC installed, and the update fixes an incompatibility issue with FSUIPC.

17 hours ago, CajunRon said:

I unassigned my CH Product Yoke from FSUIPC and assigned the controls to P3D and widened the dead zone slightly beyond the default.  I also did not run AS2016 nor ProATC.   After stabilizing at cruise altitude and lefting it run a little while to see that it remained stable I then ran the rest of the flight at 4X acceleration until right before the TOD.  Everything went smoothly with no problems.   I switched over to normal time rate and let the FMC begin the decent.  Again everything was fine.   My wife then called me for help with a short chore.   I left the sim running with the FMC still controlling the descent.  The STAR and APPROACH had already been entered into the FMC and all discontinuities removed.    I was gone for about 2 minutes.   When I returned the autopilot had disengaged and the plane was in a hard left bank nose dive.   (I thought that with the autopilot having everything in trim that the plane would continue flying relatively stable in a trimmed state at least for a while if disengaged.)   Anyway for over two hours the autopilot held, even in 4X acceleration, but while away from the computer for 2 minutes (MURPHY'S LAW) with absolutely no movement of the controllers, the autopilot tripped out.  I'm assuming the autopilot in the 747 is programmed to be much more sensitive than that programmed in PMDG's 737 or 777 since since I didn't have to widen the dead zone in those.   I'm suspicious this is part of the problem. 

Still sounds a lot like a hardware issue. Do you have a hardware trim axis assigned? If it's not your yoke, then it might be your motherboard. Do you have an ASUS motherboard?

[CajunRon 8]

Base on what several had said, it looks like I dodge a bullet with the uninstall / reinstall.  Anyway, I completely removed the FSUIPC module from P3D V4,   I had one autopilot disconnect in my KJFK - KSFO test flight  Other than that the only anomaly is that the FMC would not follow the glide slope even though GS was shown to be the active mode.  Likely pilot error.  Will do a little more trial.

I do have an ASUS motherboard and found the reference in the PMDG documentation about the problem with the high precision timer on some of those boards.   I do suspect I have a problem board since I did have system clock problems when I first installed it.   But a BIOS update seemed to have solved that problem (until recently when I had two occasions where the system time was wrong again)...But a new motherboard is not anywhere in the budget right now so I'll have to live with it.   I also disabled the "realistic autopilot" option which is suppose to disable the need to have the plane in perfect trim before the AP is engaged.  This, with the removal of the FSUIPC module, has given me a fairly stable system and the plane is flyable.

Just as a matter of interest I also tried the plane in P3D V3 which has FSUIPC 4 installed.   I had to disable the "realistic autopilot" option on it but was able to leave FSUIPC active and controlling to get a fairly stable autopilot.

Thanks to all for the help.   My only concern now is the idea that the uninstaller is not properly uninstalling the plane.   Having a proper uninstaller and reinstaller is pretty basic to software and the vast majority of software I have installed and uninstalled (1000's in my 50+ years) behave properly.   Since several recommended NOT uninstalling the plane it appears this is a common problem and I would classify it as a bug.

[scandinavian13 5,192]

2 hours ago, CajunRon said:

Thanks to all for the help.   My only concern now is the idea that the uninstaller is not properly uninstalling the plane.   Having a proper uninstaller and reinstaller is pretty basic to software and the vast majority of software I have installed and uninstalled (1000's in my 50+ years) behave properly.   Since several recommended NOT uninstalling the plane it appears this is a common problem and I would classify it as a bug.

Your classification (and assertion) here is incorrect. Just because we do not recommend the action doesn't mean that the installer is not uninstalling correctly. Think, for a moment, that, in order to uninstall a plane, a user might simply just go deleting things instead of using the uninstall function. That could cause issues. Additionally, your extra action of deleting things that the installer did not delete could cause issues, depending on what you were deleting and where.

Don't assume, and don't assert a bug unless you know, categorically, that it is a bug.

[CajunRon 8]

Oh boy...here we go.

The statement "I would classified it as a bug" is an opinion.  It implies that you and others may not classify it as a bug. The statement "this is a bug" is an assertion.  I didn't make that assertion.

As a customer I have certain expectations based on product descriptions and long established standards of performance.  I expect the UNINSTALL feature of Windows 7 to work  and the reinstall to bring the software back to it's original state.   This did not work as it does with the vast majority of other software.   AFTER the conventional method to uninstall/reinstall failed is is when I decided to go one step future and  use the unconventional process of deleting any remaining all references to the PMDG 747 (which the conventional UNINSTALL should have removed in my opinion).  I understood the risk of doing this and take full responsibility for any problems this second unconventional attempt may have caused but the fact is that my first attempt, which was to use the standard Windows 7 uninstall function, did not work.   My expectation is that when I REINSTALL the software, regardless of what may have transpired (dumb things done by the computer owner, hard drive crash, power failure causing corrupt files, etc. etc.)  the software should return to its original, fresh install condition.  This did not occur in the first conventional uninstall/reinstall attempt.   This is a shortcoming of the software and does not meet my expectation as a customer.   For $149 I have the right to hold this expectation.   I guess as a vendor, you have a right to ignore this expectation or to use it as customer critique and feedback to improve....your choice.

[Captain Kevin 586]

27 minutes ago, CajunRon said:

I expect the UNINSTALL feature of Windows 7 to work  and the reinstall to bring the software back to it's original state.   This did not work as it does with the vast majority of other software.   AFTER the conventional method to uninstall/reinstall failed is is when I decided to go one step future and  use the unconventional process of deleting any remaining all references to the PMDG 747 (which the conventional UNINSTALL should have removed in my opinion).  I understood the risk of doing this and take full responsibility for any problems this second unconventional attempt may have caused but the fact is that my first attempt, which was to use the standard Windows 7 uninstall function, did not work.   My expectation is that when I REINSTALL the software, regardless of what may have transpired (dumb things done by the computer owner, hard drive crash, power failure causing corrupt files, etc. etc.)  the software should return to its original, fresh install condition.  This did not occur in the first conventional uninstall/reinstall attempt.   This is a shortcoming of the software and does not meet my expectation as a customer.

I wonder if the stuff that didn't get uninstalled was stuff that was customized and/or otherwise updated later down the line, like navigation data, panel states, or aircraft configurations. I'm not sure about anything else, but I know for a fact that if you already have existing navigation data on your computer, the installer won't touch them. This would result in the plane not quite being in its original, fresh install condition.

[scandinavian13 5,192]

2 hours ago, CajunRon said:

The statement "I would classified it as a bug" is an opinion.  It implies that you and others may not classify it as a bug. The statement "this is a bug" is an assertion.  I didn't make that assertion.

You literally made that exact assertion:

6 hours ago, CajunRon said:

I would classify it as a bug.

Need we go any further on this point?

 

 

2 hours ago, CajunRon said:

As a customer I have certain expectations based on product descriptions and long established standards of performance.  I expect the UNINSTALL feature of Windows 7 to work  and the reinstall to bring the software back to it's original state.   This did not work as it does with the vast majority of other software.   AFTER the conventional method to uninstall/reinstall failed is is when I decided to go one step future and  use the unconventional process of deleting any remaining all references to the PMDG 747 (which the conventional UNINSTALL should have removed in my opinion).  I understood the risk of doing this and take full responsibility for any problems this second unconventional attempt may have caused but the fact is that my first attempt, which was to use the standard Windows 7 uninstall function, did not work.   My expectation is that when I REINSTALL the software, regardless of what may have transpired (dumb things done by the computer owner, hard drive crash, power failure causing corrupt files, etc. etc.)  the software should return to its original, fresh install condition.  This did not occur in the first conventional uninstall/reinstall attempt.   This is a shortcoming of the software and does not meet my expectation as a customer.   For $149 I have the right to hold this expectation.   I guess as a vendor, you have a right to ignore this expectation or to use it as customer critique and feedback to improve....your choice.

If something gets left behind by an installer, then there's a reason it's left behind. You, yourself, note that we are not the only ones who do this with your note of "[...] as it does with the vast majority of other software." If an entire industry is doing it, there's a reason. While you may not understand the reasoning, there's a reason, and it is usually related to inter-dependencies (nav data, as an example). In the exceedingly rare circumstances that an uninstaller leaves files behind that would cause an issue, a simple ticket to any developer would resolve the issue (in fact, we have several bits of text in our own system to describe doing exactly this). Unless you have these directions, you could end up causing more harm than anything else.

The earlier warning wasn't to spark off an entirely pointless debate - it was to tell you that the action was unnecessary, and without guidance, could cause even more issues. In the future, you - with ANY software ($1 or $1000) - may want to keep this in mind. When the product re-installed, it was in exactly the state it was when it was first on your machine. The difference this time, as I noted earlier, was the inclusion of FSUIPC.

It's a cautionary tale. We're here to help you, and if you're reaching out for help, then you likely need the help and are not otherwise knowledgeable enough to solve it on your own. This is normal, and also fine. I get that it all looks a certain way to you, but how things look on the surface isn't always what's going on behind the scenes. That's the piece that I'm here to help solve, and I will occasionally point things out that you should avoid so that both of our lives are easier: you're not wasting time and making things worse, and I (and the rest of us) can help you get things running.

 

Short version: you can do it on your own, or you can allow people to help you. If the people helping you advise you against something, however, it's best to take that advice and continue on. Trying to drag in all kinds of unrelated issues to try and pin the issue back on other people is neither productive nor helpful...as is evidenced by the fact that I'm now finishing this post trying to remember what the actual issue was in the first place, since I've been sidetracked by talking about installers/commonfiles/etc.