B747-400 - Erratic LNAV Oscillations w/ FSUIPC Turbulence

[Guest nd55b]

Hi Folks,I wasn't sure if this was already covered by another topic; my apologies if I am being redundant. I've been having a dialogue with Pete Dowson about a phenomenon I observed while flying the 747 through FSUIPC-modelled turbulence. Specifically, in the presence of even mild turbulence, the autopilot loses complete lateral control while LNAV is commanded. The deviations can be pretty severe and in some instances have resulted in the aircraft rolling over on its head.The problem is immediately quelled by suppressing turbulence (in fact control is regained almost immediately). The behavior does appear to be unique to the 747. Pete and a few others have been able to reproduce this. Although Pete did mention that he developed the FSX turbulence code w/ someone from PMDG and he did not experience this behavior in an earlier release of the product. Since upgrading to 2.10.00.40, however, he was able to reproduce this condition. I am fairly new to the '47, and have only known 2.10.00.40, so I have nothing else with which to compare it to. I do know other aircraft are seemingly unaffected by FSUIPC's modelling of turbulence.Pete made some changes in his most recent beta, 4.294, which smooth out the frequency with which he is altering the wind data, and this seems to keep the aircraft from entering "the death spiral", but the aircraft still gyrates +/- 25 degrees with respect to the desired flight path even in minimal turbulence (+/- 2 or 3 kts/degrees).Anyway, here is the thread. It's rather verbose and probably best experienced from the bottom-up:http://forums.simflight.com/viewtopic.php?...=441649#p441649I wanted to know if you were aware of the issue or had additional insight into what might be causing this and/or if any workarounds exist (apart from disabling turbulence altogether).Thanks for your time.Best Regards,Jason BarlowND55B

[mccoy007 0]

Hi there used to be an discussion herehttp://forums.avsim.net/dcboard.php?az=sho...id=149015&page=I also had this problem with my 744X, which is the same version as yours. I am using the FSUIPC 4.28 with ASX SP3.I disable the ASX's own wind smooth function and using FSUPIC wind smooth. So right now I won't have sudden wind direction change in 180 degrees, but the turbulence still cause my queen unstable, just like the behavior your said. Sometime I will lost my VNAV. The only solution now is disable the all cloud turbulence.But I think for serious simulator, this seemed not acceptable, may be this is the reason why we use weather radar. The only thing I think I can prevent my plane into such a extreme condition is to using weather radar, satellite imagine for pre-flight planning, also if I really encounter such a thing, I will try to fly the plane manually without a/p.I am not sure about if you flying other aircraft such as 767-300ER from LD, I don't have time to test their behavior in Turbulence, if they have the same behavior may be it is the PMDG's problem, if same we had to look again to Peter and ASX, to see if the new version of FSUPIC or SP will work.Yufeng

[Guest nd55b]

Thanks for your comments (and the link). With regard to your question, the Level-D 67 handles the turbulence just fine for me. I did have some additional feedback with regard to Mr. Randazzo's comments in the thread you referenced (hopefully he is available for further dialogue :^)>> If you are seeing such extreme behavior from the airplane, >> then this would indicate to me that you have your turbulence >> settings unrealistically high, and/or you have extraordinary wind >> shifts taking place. In my test cases, turbulence was light. Wind variance was +/- 5 degrees and wind speed was +/- 2 or 3 knots. Worth noting, however, is that while the magnitude of the change may be very small, the frequency with which the wind speed was updated by FSUIPC's turbulence emulation seemed quite high (observing it empirically).>> I haven't used FSUIPC's new wind smoothing, so I cannot comment>> on it- but I do know that Dr. Vaos provided quite a bit of>> insight to Pete Dowson as he was working on this capability a few >> months ago.Pete never mentioned any names, but did say he worked closely with PMDG in the development of this feature to create a much more realistic turbulence model (which I believe most feel he has achieved). During development, Pete was testing with version 2.00.00.31 of the product, and this phenomenon was not reported. In response to this issue, Pete installed the latest update, and was immediately able to generate significant lateral oscillations with minimal turbulence.>> At any rate- if you are seeing 60-90 degree bank angle changes >> happening suddenly- I would bet the house that if you watch the >> wind vector on the nav display- you'll see significant and >> massive changes. The changes are actually very small. As an additional test, I disabled wind variance. This causes FSUIPC to model its turbulence only by adjusting wind speed; no directional changes in wind were observed. Small, but high frequency updates to the wind speed alone really seem to bugger the autopilot system in the 744. Disabling turbulence immediately allows the autopilot to recover.Anyway, I realize this problem is "fixable" by suppressing turbulence, but it does take away an element of realism (especially now that Pete has got the model working so well). I hope someone at PMDG could get back in touch with Pete to see what might have changed. Two great products deserve each other :)J