It has been widely discussed, to my knowledge without any written intervention from PMDG.Now i did a "scietific" test.I used this routeDEP: LIRF 25ARR: LIRA ILS15 TRANS URBECRZ FL 120Waypoints:medaltaqcmp /6000Aurb 200/3000Avarious points inserted by the STARThis has been the same for various flights with the 700, in many weather and weight conditions.The FMC once i have inserted the URB hard constraints, automatically creates a DECEL point a few miles before URB.There has been not a SINGLE TIME that the aircraft has been able to maintain the VNAV path.Every single time the Vnav disconnected saying "VNAV Path Unachievable".Now this should prove that there is at least a small problem: the FMC designate the position of the DECEL point too late, and it isn't able to respect the altitude constraint at the next point in the route.It is not up to me to find why and how the fmc does its calcs wrong, but i hope that PMDG will have a look at it, as it is affecting a wide variety of situations.This is just one exasperated situation (by the repetition, as it is plainly a natural operation to put a deceleration once nearing the IP), but many of us know how many times the VNAV disconnected saying the same thing, and that we've been told countless times it was our fault.I think that many times it can be pilot's fault, but not this: here is the FMC that calculates the decel points, not me! So the error is its.And i even tryed going down since FL90 with spoilers (after the FMC called for Drag Required), even at the decel point.No joy.So please tell me why this revision 10.4 Unit can't be able to do such a stupid thing as decelerate in time.Thanks

Well,I think this has something to do with the route and what you're trying to make the aircraft do.If you have a speed restriction at one point B and an altitude restriction as well. And an altitude restriction on the point A before that. You can end up in trouble.THe FMC is calculating a geometrical path between the points. If this path exceeds a certain angle you will have a disconnect and "VNAV Path unachieavable". If you to this adds a decel segment, which baiscally means that the FMC will command pitch to a descent of 500-1000 fpm, you have even less distance for the total descent from A to B making it even more likely for a disconnect.What's the distance between CMP and URB?http://forums.avsim.net/user_files/91604.jpgHope it helps,

cmp /6000A Well this appears to be one reason. This FMC cannot handle a restriction that is an altitude without the speed also restricted on the same wpt I believe. There is the ability to do this in real life of course. I have not tried that route but all one has to do is look at the ANGLE between each wpt to see if it is within VNAV limits. And do not expect VNAV PATH to do very well with speed. Have you tried this same route in VNAV SPD? It will honor restrictions too but keep the target all the way down. Best Wishes,[h4]Randy J. Smith[/h4]http://www.rawbw.com/~bdoolin/shinault/Animation1.gifCaution! Not a real pilot, but do play one on TV ;-)AMD 64 3200+ | ASUS KV8 DELUXE | GFORCE 5700 ULTRA @535/1000 | Maxtor 6Y080M0 SATA 80 GIG | 512 DDR 400 | Windows Xp Pro | Windows Xp Pro 64 |

>This FMC cannot handle a restriction that is an altitude without the >speed also restricted on the same wpt I believe.I think this shouldn't be a problem for the FMC Randy. Or are you saying that in this respect the FMC differs from the real one?If the former, I believe you are refering to the speed without alt restriction problem on earlier versions of the FMC software? Cheers,

6000A at CMP is obliged, as it is a minimum radar of the ATC.I do not know how the reals do this approach, probably they are vectored as it often happens in rome TMA.Nonetheless i think it should be able to do this.6000ft at CMP is a radar minimum while 3000ft at URB is the IP of Ciampino's ILS.That said, planes will usually overfly CMP high above 6000ft, and still make it for URB at 3000ft and a velocity slow enough to intercept (i doubt they will overfly the IP at 240kt...).The experience i had, is that the FMC does not calculate the right descent angle BEFORE the deceleration point: it ends too high.And the DECEL point is near to URB, rather then halfway the distance or nearer to CMP.So i think it's a fault in how the FMC calculates the position of deceleration points.From CMP to URB it's a little less then 12nm.Edit.I just this some math:12nm = 22.2km3000ft = 1kmangle of the slope is 2.58

> Well this appears to be one reason. This FMC cannot handle a>restriction that is an altitude without the speed also>restricted on the same wpt I believe.This is nowhere to be seen in the manual Randy.Where did you take this from?> There is the ability to do this in real life of course. I>have not tried that route but all one has to do is look at the>ANGLE between each wpt to see if it is within VNAV limits.As i wrote in the above message, the pure angle of this segment is 2.58

Claudio,I've flown your approach a couple of times and I get some kind of anomaly when doing this. I remembered a thread dealing with the LOGAN arrival as well and we found out that something was not working correctly. Can you check and see if this applies to your situation?http://forums.avsim.net/dcboard.php?az=sho...57114&mode=fullI also noticed that after passing CMP at 6000 I got the V/S bug visible. Once at vertical speed 0 and once at about 500 fpm. See pitcure. This of course got you way high of the path. Recycling VNAV would help in this case. http://forums.avsim.net/user_files/91693.jpgHope it helps,

As shown in the pic, the AC does not descend but should (bug I guess, which probably can be worked around by recycling VNAV). When the V/B to wpt exceed 7.xxx degrees, VNAV will disconnect.I will check it out hopefully tonight in the debugger and let you know if something other than recycling VNAV can be done.anthonyAnthony MertonPrecision Manuals Developmenthttp://www.precisionmanuals.com

>6000A at CMP is obliged, as it is a minimum radar of the>ATC.>I do not know how the reals do this approach, probably they>are vectored as it often happens in rome TMA.>Nonetheless i think it should be able to do this.>6000ft at CMP is a radar minimum while 3000ft at URB is the IP>of Ciampino's ILS.>That said, planes will usually overfly CMP high above 6000ft,>and still make it for URB at 3000ft and a velocity slow enough>to intercept (i doubt they will overfly the IP at 240kt...).>The experience i had, is that the FMC does not calculate the>right descent angle BEFORE the deceleration point: it ends too>high.>And the DECEL point is near to URB, rather then halfway the>distance or nearer to CMP.>So i think it's a fault in how the FMC calculates the position>of deceleration points.>From CMP to URB it's a little less then 12nm.>>Edit.>I just this some math:>12nm = 22.2km>3000ft = 1km>angle of the slope is 2.58

I'll see if the plane gets fixed on a shallow dive, will let you know.But could you please elaborate on the V/S bug?I just see your V/S to be zero, where is the problem?

Ok i redid the route, and noticed the bug: autopilot stops commanding pitch.It gets stuck when you need to decelerate the speed or the descent, and doesn't move from there anymore.It happened both nearing FL100 when i engaged spoliers, and again near CMP.I had to recicle VNAV to make it work again.Then that's definetevely a bug, isn't it?

Definitely a bug, yes.I flew it with the debugger and what happens is that the AFDS does not command the AP to resume the descent because of some bad synchronization. The AP captures the 6000 alt and after commanded alt is set to 3000, it enters default VS mode internally. The AFDS should take this in account but for some not yet very clear reason does not...I'm afraid that re-cycling VNAV as Mats suggested is the only thing I can think of in the current state of affairs.anthonyAnthony MertonPrecision Manuals Developmenthttp://www.precisionmanuals.com

>I'm afraid that re-cycling VNAV as Mats suggested is the only>thing I can think of in the current state of affairs.I think that it's a good idea to pin a bug list in the forum, or at least on your site.You should update the "Maintenance Status Board".Anyway, i guess you're deep into the 747 development cycle, is there an hope for a fix once you get over the 747?

> Maintenance status board yes, I'll request it is done.> is there an hope for a fix once you get over the 747?yesanthonyAnthony MertonPrecision Manuals Developmenthttp://www.precisionmanuals.com

This is something that mystifies me as well and it always happens on whatever route/descent I'm making. I always get a 'Descent Path Unacheivable' message or 'Drag required'. Like someone else has asked, if the FMC calculates the top of descent point and the decel point, then surely you shouldn't get these messages - why doesn't it start the descent earlier - although I have tried this manually but still get these messages. I've tried increasing the rate of descent to get back on path, but then the speed increases even with the spoilers out. I'm sure I'm using far more air brakes than i've seen on a real 737 on approach, although on most occasions I've been able to 'rescue' the descent and re-engage VNAV quite successfully.My main critism of 737NG is that my wife is likely to leave me if I spend anymore time on it - has anyone any suggestions for this problem please ?PS I liked the message from the guy who said his woman brings him a drink whenever he rings the cabin chime - don't think I'd even suggest it !Sorry forgot name !Dave