@LRBS thanks for your clarification. My problem with the word "tweak" comes in its juxtaposition with the statement made by Mr. Randazzo regarding a new Lateral Flight Model having been deployed. When read together it seems to imply that PMDG's founder was being less than forthright in suggesting that a new LNAV system had been created, when I see no evidence having been provided that this was the case. Maybe the new model is just a tweak on the old, maybe it's a totally new model that's just no better than the old, or maybe it's something in between, but unless one has access to the internals, it doesn't seem possible to conclude one way or the other.

That even the most advanced desktop add-ons don't faithfully represent every aspect of the behavior of the real aircraft is really no surprise to anyone, but simulations are just that: simulations. For them to even be in the ballpark of how a real aircraft as complex as an Airbus or Boeing might respond in the real world is endlessly amazing to yours truly.

Thanks to you and those of your colleagues who are kind enough to share your knowledge and experience with the rest of us. It's much appreciated.

@MDFlier that seems an excellent idea, but getting all of the variables under control would seem to be a real challenge. Still it would be interesting to see the outcome.

17 minutes ago, jrw4 said:

Maybe the new model is just a tweak on the old, maybe it's a totally new model that's just no better than the old, or maybe it's something in between, but unless one has access to the internals, it doesn't seem possible to conclude one way or the other.

Not sure why that even matters. The discussion seems to be merely semantic at this point. What matters is the outcome; whether that's achieved through building an entirely new module from scratch or whether it's built upon an existing suite and expanded is irrelevant, imho. It's doing a much better job than before following the magenta line and hopefully they can improve upon that even further.

FWIW, it flew the localizer capture on my flight today just like @Stearmandriver described. Initial hard bank to turn towards the signal followed by leveling off for a moment before banking slightly again to align exactly with the radio beam.

A best practice tip is - when on vectors - to set the inbound course to the approach fix in the lower left of the leg page and engage LNAV once cleared for the approach. This guarantees an appropriate initial bank (as describe above) when intercepting the localizer and should prevent any over- or undershooting.

12 hours ago, Nixoq said:

I threw LNAV and localizer capture in one sentence because since the LNAV update it's capturing the localizer a lot better, so something must have been polished on that front as well and IIRC PMDG said they would bring improvements here in conjunction with the LNAV update.

Fair enough and for what it's worth, I fully agree that they have a good bit of work yet to do with LNAV.  I feel like that last update got them about halfway to where they need to be, in terms of LNAV calculating a full path and then flying it as a smooth continuous path vs recalculating at each waypoint.  This behavior is significantly improved after the update, but with significant improvement still to go to match the airplane behavior.

I'm thinking (and hoping) that at this point, they're waiting until the full changeover to ARINC-424 compatibility with RF leg capability.  Maybe that changeover will itself solve the remaining problems.  I look forward to finding out... Hopefully we see it sometime in the new year!

13 hours ago, Stearmandriver said:

in terms of LNAV calculating a full path and then flying it as a smooth continuous path vs recalculating at each waypoint.

Agree with this. Often when there's a straight line through a waypoint it will bank left and right briefly instead of just keep going level.

No offence to PMDG, but i find Fenix' Airbus is much better in the flying department.

 

And i'm not just talking about MSFS, i found the same behaviour back in P3D.

 

The aircraft struggles to follow the LNAV path, in particular departing airports. It often can't follow the magenta line in the turn, it always overshoots it to the point where i re-input the waypoint its trying to capture to avoid it snaking back trying to intercept the path. And yes, the bank angle selector is fine. 

 

Descents are a problem as well. I find following the VNAV path the aircraft will sometimes lose airspeed below what's set for the VNAV path, up to 20+ knots. I need to turn off Autothrottle and manually bring the speed back up.

 

And i've always believed the aircraft struggles with turbulence. I've had a descent rate of 800fpm, speedbrakes out and the speed really struggles to drop. I've dropped the landing gear to create some more induced drag but it still struggles. I know the 737 is a slippery beast, but that's too much.

 

Under normal circumstances i would re-install, but as i said i have found this behaviour with their previous 737's in past Sim's.

 

With localiser capture as opposed to the Fenix, you don't want the intercept path to be too great in the 737.

 

Just my 2 cents worth.

 

 

 

 

2 minutes ago, pan pan pan said:

No offence to PMDG, but i find Fenix' Airbus is much better in the flying department.

 

And i'm not just talking about MSFS, i found the same behaviour back in P3D.

 

The aircraft struggles to follow the LNAV path, in particular departing airports. It often can't follow the magenta line in the turn, it always overshoots it to the point where i re-input the waypoint its trying to capture to avoid it snaking back trying to intercept the path. And yes, the bank angle selector is fine. 

 

Descents are a problem as well. I find following the VNAV path the aircraft will sometimes lose airspeed below what's set for the VNAV path, up to 20+ knots. I need to turn off Autothrottle and manually bring the speed back up.

 

And i've always believed the aircraft struggles with turbulence. I've had a descent rate of 800fpm, speedbrakes out and the speed really struggles to drop. I've dropped the landing gear to create some more induced drag but it still struggles. I know the 737 is a slippery beast, but that's too much.

 

Under normal circumstances i would re-install, but as i said i have found this behaviour with their previous 737's in past Sim's.

 

With localiser capture as opposed to the Fenix, you don't want the intercept path to be too great in the 737.

 

Just my 2 cents worth.

 

 

 

 

Some of the issues you mentioned, I also found in the P3D version, and that is why I skipped the MSFS version, and instead got the Fenix which I love. I just think the Fenix is just more up to date. 

3 hours ago, pan pan pan said:

Descents are a problem as well. I find following the VNAV path the aircraft will sometimes lose airspeed below what's set for the VNAV path, up to 20+ knots. I need to turn off Autothrottle and manually bring the speed back up.

Keep in mind that VNAV isn't perfect in the real world either. It's normal for it to sometimes be unable to hold speed for example and not uncommon to have to manually help it out. Using the speedbrake is common. I don't know your background but I think simmers usually expect the automation to always be spot on, no matter what, when it sometimes does a really bad job IRL too.

If you're in ARM mode during a VNAV descent in idle the autothrottle is armed but not active so you manage speed manually, that's by design. There are speed undershoot margins that can be chosen by airlines and in the PMDG config menu from 5 kts below to 20 kts below target speed. Within the margin the autothrottle is armed only during an idle descent and FMC SPD will only activate and bring thrust back up to catch up on speed once out of the margin. That's realistic. The PMDG does have a tendency to overshoot speed restrictions in VNAV PATH though from my own experience and I've found myself struggling to slow down a few times even with the gear out, flaps 15 and a slight headwind in a shallow descent on the glideslope. Most of the time it's been performing fine for me though.

As for the Fenix, it does have its moments too. Just recently I've found myself needing to manually control speed via the FCU in a descent because it would lose the target quite a lot following the path. They've refined that a while ago but I see it still likes to sit at the lower end of the speed bracket a lot, even when it should probably have some more energy than that.

19 hours ago, Stearmandriver said:

I'm thinking (and hoping) that at this point, they're waiting until the full changeover to ARINC-424 compatibility with RF leg capability.  Maybe that changeover will itself solve the remaining problems.  I look forward to finding out... Hopefully we see it sometime in the new year!

And how long ago did Rob mention that one? I wouldn’t hold my breath in seeing this anytime soon, much less in 2023. It’s probably on the shelf sitting next to GFO.

Oh, I don't know, considering we've seen work on it already.  I imagine they understand pretty well that this is one of a very few areas where they actually are behind other devs, not because of sim limitations but because of their own workflow.  I'm guessing it has some priority.  😉