VNAV descent p3d

[Pilot53 544]

Im having a great time flying the 777 in p3d (too bad though now I cant ever bear to use fsx again), but vnav descents have been strange now on just about every flight.  The airplane cannot seem to calculate a decent path without having to go into speed mode.  At the tod the aircraft will start to descend and go to idle thrust mode but then will level off and speed will bleed off to the point that spd mode must activate.  Then it starts to descent again but at a rate too shallow for idle thrust until it reaches lower altitude.  This is while using a descent forcast and asn wx.

[Guest]

To be honest that's the same problem clearly visible also into the FSX version.

 

I've almost never seen a good & smooth VNAV path, I assume that's because I am under the impression that the T7 has been extensively tested with the most important US an UK airports and therefore their SID and STARS.

 

Reading what forumers says ( that's also my opinion and personal experience ) a good VNAV mostly happens when dealing with bigger and well known airports and therefore well tested SID and STARS , that's not the same when the FMC has to deal with minor airports.

 

Same speech when programming the route against f.i. Heathrow...everything is perfect but when programming a route against f.i. a small italian airport the FMC goes crazy drawing magenta lines like it was owned by a schizophrenic evil. Most of the times I have to correct the FMC routes cutting waypoints, adding waypoints in order to fix the issues of the T7 FMC when dealing with not tested airports SIDs and STARS. Of course vectors are not part on my opinion, they are not involved.

[scandinavian13 5,192]

Not sure I agree with this. When in testing, I consistently flew the 777 on old A-Nav STARs, new RNAV STARs, and new RNAV OPD STARS. Without knowing where we actually flew it, I do find it somewhat irritating that one would insinuate that we simply stuck to the major airports and didn't fly anywhere else. That would be false. As part of testing, I frequently flew the 777 into airports that had no business taking a 777. None of the variations in airport or STAR types resulted in problems, provided you did a few things:

1. You absolutely need to know how to fly the plane. I don't mean to insult anyone, but remember that this is a study sim. It isn't meant to be flown causally. It can be, but if flying causally, expect casual results (casual meaning "not precise" - remember the "trash in, trash out" concept).
2. The FMC is only as good as the data you give it. With a lot of the old A-Nav STARs, the FMC gets no data fed to it from the database. As an example, the COATT STAR into IAD has no crossing restriction at FALKO. The reason is that it's an "expected" crossing and not a hard assignment (like the later altitudes on the newer BARIN STAR). This can throw a wrench in your plans because the end of the COATT is a long vector leg that continues away from the airport. As a result, the FMC - without the additional data - will leave you high. This isn't an FMC problem; this is a data entry problem.
3. The descent needs to be properly forecast. The FMC is only as good as the data you give it (sound familiar?). In this case, the FMC can't predict things unless you give it data to work with. With the datalink function, this is made easier, but don't forget to update it closer to the actual T/D (instead of when it first imports, likely before you even take off).
4. Ensure you close your (hardware) throttles prior to the T/D. Some of the level-offs (and even climbs) are related to people not closing their hardware throttles. When the throttles send a spike, the sim is reminded of the hardware position, and since VNAV DES is a HOLD A/T mode, it will use the new throttle settings and shallow out the descent.
5. Remember that this is automation. Automation is good for most cases, but it cannot handle all of them. Sometimes you need to take over. This is both realistic and correct.

If either of you could give a specific route or STAR, I'd be glad to check it out.

[Guest]

No Kyle, I didn't want to insinuate that the T7 has not properly tested on minor airports, it would be impossible to test it over all FSX airports, that would be an insane and useless effort.

 

I told that in my own experience I have seen the T7 FMC performing like a charm into biggest and well known airports and the same person ( aka as myself ) have seen the very same plane FMC bahave in a weird way into certtain airports, therefore the VNAV has been affected accordingly.

 

I am sure it happened to everyone at least on time to view weird magenta lines connecting waypoints of a given SID or STAR present into the FMC latest Airac. I mean, the experience to spot weird magenta lines arcs or worst to be forced to manually delete a certain waypoint because the FMC simply can't deal with it even if it a perfect legitimate SID or STAR used in the reality.

 

I think to remember ( I am at work now ) Malpensa in Italy --> LIMC from Sim Giants 2010 version. Each time I take off or land into this major italian airport the FMC after inserting a landing or departing runway and after choosing a certain SID and STAR it draws comic magenta lines looking like a ball of wool after a kitty has fiddled with it.

 

So, if the very same person flies the very same plane but he experience very different behaviour of the FMC and therefore of the VNAV in your opinion is it really always a matter of data input or it maybe the FMC performing differently depending on the situation/airports/SID and STARS ?

 

Told that, I really appreciate a lot your effort here and the support you give to everyone, that's without any doubt.

[Pilot53 544]

Thanks for the support kyle, If you are interested in replicating this behavior it was enroute from SBGR to KIAH, Flying the Gill1 star ILS 26L transition.  I always update the descent forecast about 100nm prior to tod and have seen this happen with just about every vnav descent flying into phnl, klax, egll, eddf, and a few others.  It is always flown with the lastest published stars, asn wx,and vnav descent forecasts.

[scandinavian13 5,192]

 

 

KIAH, Flying the Gill1 star ILS 26L transition.

 

I'll have a look.

[rjack1282 316]

 

 

Ensure you close your (hardware) throttles prior to the T/D. Some of the level-offs (and even climbs) are related to people not closing their hardware throttles. When the throttles send a spike, the sim is reminded of the hardware position, and since VNAV DES is a HOLD A/T mode, it will use the new throttle settings and shallow out the descent.

 

This.  I was experiencing the same thing with a massive bleed off of speed.  The plane would then say, "wait a second I should be going faster" and then accelerate.  I noticed it seemed to be occurring in high headwinds and chalked it up to that even though I had just updated the decent forecast prior to TOD.  The first flight I did was the tutorial where I learned to pull back my saitek throttles during the hold phase; however, at some point I forgot to do that on every flight.  This was the cause at least for me.  Thanks.

[dave_paige 11]

I don't own the 777 but just for fun I'll try it with the NGX. Maybe not all the way from SBGR but I'll fly the same arrival, which transition were you using?. I get very similar complaints about the NGX's FMC as well and I find it's always either, the user doesn't understand how to use the systems and/or fly the plane, or, very occasionally, missing or faulty navdata, which can be always be fixed.

 

Dave

[Pilot53 544]

This.  I was experiencing the same thing with a massive bleed off of speed.  The plane would then say, "wait a second I should be going faster" and then accelerate.  I noticed it seemed to be occurring in high headwinds and chalked it up to that even though I had just updated the decent forecast prior to TOD.  The first flight I did was the tutorial where I learned to pull back my saitek throttles during the hold phase; however, at some point I forgot to do that on every flight.  This was the cause at least for me.  Thanks.

What you are describing is the same behavior that I am experiencing, however I dont think that is what kyle is talking about.  He is referring to what happens when you dont pull your controllers back to idle and the fms mode goes into "hold" the throttles will spike to their commended position.  The issue we are experiencing occurs when our throttles are pulled all the way back but the calculated vnav path is too shallow to maintain speed at idle thrust.

I don't own the 777 but just for fun I'll try it with the NGX. Maybe not all the way from SBGR but I'll fly the same arrival, which transition were you using?. I get very similar complaints about the NGX's FMC as well and I find it's always either, the user doesn't understand how to use the systems and/or fly the plane, or, very occasionally, missing or faulty navdata, which can be always be fixed.

 

Dave

It was KELPP.GILL1 26L transition.

[rjack1282 316]

 

 

What you are describing is the same behavior that I am experiencing, however I dont think that is what kyle is talking about.  He is referring to what happens when you dont pull your controllers back to idle and the fms mode goes into "hold" the throttles will spike to their commended position.  The issue we are experiencing occurs when our throttles are pulled all the way back but the calculated vnav path is too shallow to maintain speed at idle thrust.

 

Perhaps it was coincidence and the specific vnav path I was flying didn't trigger the issue.

[albertz 0]

Its not because of the geometric path descent or the way FSX transitions the winds aloft is it??

I know that the geometric path descent causes issues with the VNAV going way below profile in VNAV Speed with thrust added. That can be altered by entering a new cruise altitude at intermediate level offs and calculating a new TOD. Most of the arrivals into IAH that I've seen seem to have pretty well optimised paths though.

Cheers

Matthew Knight