747-400 auto pilot problem

[Randalicous 0]

Hi.I have been using this product without issue for a while now. But in my last 2 flights I have had issues with the autopilot.I programme it as per normal, set my cruise alt in it and activate the flight plan. Climb out using lnav and vnav, however, when I get to 30000 (my cruise alt) the plane keeps on climbing at like 5000 fpm. I press the V/S button and attempt to manually bring the plane down, but it ignores what im doing. The light on the v/s activates and i set it to -1800fpm but it does not react with it. Can anyone tell me why this would happen and what causes it.ThanksSteve.

[sheepfu 3]

Have you enabled one of the three CMD buttons?

[Randalicous 0]

Tried it on all three im afraid.. no good. I did a flight today and it behaved. (headscratch)

[Martin Connor 0]

My first flight on VATSIM and I had a similar problem, but with bells on.  Aircraft failed to follow the hdg, and as with yours kept climbing without command.  However, as I got to lower altitude, it started to work normally, captured the localizer, and the Glide Slope, then decided to fail whilst on approach, which I was able to catch early enough to complete the landing safely on manual. 

 

Only clue I saw during the flight was that the OAT was registering 62C at 29,000.  I know there is a problem with the weather data if loaded from VATSIM, but I had it set to disabled, but perhaps I was getting it anyway.  Use a remote PC for FSINN so maybe I have to tweak things.  Is this a clue to your problem?

[kevinh 409]

I find the 747 AP logic occasionally gets confused and VNAV stops functioning, though the indications are that it's working. I usually switch to FLCH and then once it's stable at the correct altitude I can reselect VNAV and it carries on as if nothing had happened. I'm surprised V/S didn't help in the same way, but perhaps you could try FLCH next time it happens?

[Martin Connor 0]

Thanks, the plot is thickening somewhat.  I have two partitions, with O.S. on each.  The idea was to have one very clean one just for FSX.  I upgraded the Nav Data for the first time on this new clearn partition, and it is doing odd things. i.e. The FMS is not calculating the altitude between waypoints correctly, so my VNAV path would be like a roller coaster.  Tried manually programming a nearly similar course into the old version, and all went well.  I am wondering if the navdata has triggered something.  Other noticable issues:

 

• FMS not deleting completed legs from PFD display. 
• Aircraft appears to be attempting to refly previous legs. 
• Heading selector is doing nothing, but the Flight Director appears to be showing the correct information.

 

I believe I have the PMDG updates correct to 2.10.04, but not sure how to check this.  The usual About PMDG does give a number, but it doesn't correlate with those we are used to seeing.

 

Any help with this would be much appreciated.

[kevinh 409]

Puzzling indeed. I don't see how a navdata update could trigger that kind of thing. For the most part it's just position information, radio aids and waypoint lat/longs.FMC VNAV calculations shouldn't be affected by such updates.

 

The first two things are related, if the aircraft didn't get close enough to a waypoint it will try and return to it and the leg doesn't get removed. If a basic mode like heading isn't working then things are bad. It sounds more like an installation problem, but if it worked before the nav update then it must have got corrupted somehow. Have you tried reinstalling the 747? Apart from anything else you would be sure what version you had installed.

[Martin Connor 0]

This evening I have found the answer.  Not sure I understand it, but I had the co-pilot's flight director on, which was preventing me from either engaging VNAV or cancelling V/S mode.  I suspect that somehow, the FMS on the CP side must have had conflicting information.  Seems more likely in the real world than this one, but there it is.  Thanks for your thoughts Kevin.

[kevinh 409]

This evening I have found the answer.  Not sure I understand it, but I had the co-pilot's flight director on, which was preventing me from either engaging VNAV or cancelling V/S mode.  I suspect that somehow, the FMS on the CP side must have had conflicting information.  Seems more likely in the real world than this one, but there it is.  Thanks for your thoughts Kevin.

That's odd, it shouldn't do that, but glad you found the cause.  Worth remembering in case it ever happens to me.

[Martin Connor 0]

Ok, so much for finding the problem.  My old install works fine, my new doesn't.  Example: Runway 26L Gatwick EGKK Biggin 7m departure.  Things go wrong right from the start.  Flight director suddenly goes from around 10 degrees to zero causing speed to climb rapidly.  It then starts the right turn - a tight one - but only banks to around 10 degrees.  Switch is set to Auto.It then gentle comes around to the required heading, but has missed the waypoints by miles.  It then decides to turn right completely away from the track, and keeps going, or tries to fly back to an earlier waypoint.

 

Airnav data is identical, FSX SP2 and PMDG 2.10.040 are installed.  Old version has Traffic X but I deinstalled it from new version as part of fault diagnosis.  New version has FSInn installed and old doesn't, but I disabled the fsinn dll on new version and it still doesn't make any difference.

 

People sometimes mention a Flight 1 registry fix, but the symptoms don't seem to match.  The hardware is the same for both installs, as are video drivers.  I am at a total loss right now of what to try next.  I wanted this separate install to maximise memory and cpu when flying on vatsim.  Not looking hopeful right now.

[Martin Connor 0]

SUCCESS:-  Looking for differences between my two installations, it suddenly occurred to me that my new install has the default Virtual Cockpit where the F/O screens are alive.  My old setup had them turned off.

 

Turned off F/O screens, and the autopilot is fine and dandy. :rolleyes:

[Martin Connor 0]

Nope! Got it wrong again, but I do think I now have the solution, and feel a total wally.  The biggin 7m departure from runway 26L requires a 180 degree turn soon after leaving the ground.  As I was only flying to Manchester, I loaded the plane with passengers and as much cargo as it would take.  It was therefore very heavy, and must have an effect on the maximum turn radius.  With the reduced radius, it not only missed the first waypoint, but the second too.  I allowed the plane to fly itself, and it gently flew in a circle, flew through the waypoints it had missed on the second lap, and then began to track according to the flight plan in the FMS.

 

I call this a learning experience.  It seemed unlikely that such a significant but could exist with the F/O screen.  I must have had a different load on the other install and not realised it.

[YMMB 234]

Nope! Got it wrong again, but I do think I now have the solution, and feel a total wally.  The biggin 7m departure from runway 26L requires a 180 degree turn soon after leaving the ground.  As I was only flying to Manchester, I loaded the plane with passengers and as much cargo as it would take.  It was therefore very heavy, and must have an effect on the maximum turn radius.  With the reduced radius, it not only missed the first waypoint, but the second too.  I allowed the plane to fly itself, and it gently flew in a circle, flew through the waypoints it had missed on the second lap, and then began to track according to the flight plan in the FMS.

 

I call this a learning experience.  It seemed unlikely that such a significant but could exist with the F/O screen.  I must have had a different load on the other install and not realised it.

I need to check the SID, however if you keep your speed low but safe enough to avoid a stall, say 160-180 knots with flaps still extended, either by manually setting the speed bug or setting the speeds into the FMC could it make the turns as required and then once past this continue your climb and clean up? As a matter of interest what was your fuel load as well?

[Martin Connor 0]

A new discovery may shed light on what is going on.  I tried a lower speed, and using the rudder to assist the turn,and it managed the turn and tracked left to the waypoint, then steadily tracked off to the right.  Then I noticed that I could gently bring it back on track just turning the yoke, without the A/P disconnecting.  Odd since it had always done this in the past.  I checked AFDS settings for the A/P and found them set to true by default, which I thought provided that disconnect function.  A little reading disclosed this to not be the case.  True = the yoke controlling the aerilons even when in A/P. 

 

I then looked at the calibration of the yoke, and discovered it has taken on a right turn bias.  So, now my problem is how to cure that.  The Saitek Site gives instructions for deleting old registry settings prior to recalibrating, but my registry doesn't contain the values.  More later perhaps.

 

Looking at the SID chart and the FMS programming also raises issues.  It's not right, close but doesn't provide for an accurate intercept of the Biggin 111 Radial.  Something else to poke around with.