PMDG 737 models updated

[lwt1971 7,693]

13 minutes ago, Zangoose said:

I had a quick look at the 737-800 flight_model.cfg file and I don't see anything with those lines. 

Ah ok, thanks for checking. As per https://forum.pmdg.com/forum/main-forum/general-discussion-news-and-announcements/212095-new-ground-friction-variables-in-su10 , some did try adding these parameters themselves and got some good results, hopefully PMDG is still working on implementing these with the right values for their aircraft FMs.
 

[penta_a 434]

with latest update under SU11 BETA, 737-800 is like riding a camel's back, up-down all the time.

[jarmstro 2,406]

AutoThrottle doesn't seem right to me now. It took an age for NI to reduce once I'd reached cruise at 37000ft and it went over speed.

[tgsweat 146]

41 minutes ago, jarmstro said:

AutoThrottle doesn't seem right to me now. It took an age for NI to reduce once I'd reached cruise at 37000ft and it went over speed.

I noticed something like this as well, and its been doing it before this update, but when I do speed intervene and set my speed lower than my current speed, the plane will actually increase throttle for a little while causing the plane to speed up first, then slow to my new set speed.  

[Jeeeno 314]

To me in general the aircraft worsened a lot after this update especially if you are flying with the AP. VNAV is much more aggressive, ATHR is much more aggressive and it still overshoots the selected speed sometimes. I don't like it at all. 

[MDFlier 2,031]

I'm on my 1st flight in the 800 since the update.

I was climbing through FL220 and I noticed that the yaw damper was off. Shouldn't I have seen a FLT CONT indication on the captain's 6 pack?

And while not an autothrottle issue, I did have a problem with the engines only going down to 39% when the throttle was in idle. Went into the Windows joystick properties, which launched the Thrustmaster utility. I moved the throttle and made sure the green bar followed it properly, and then exited. I did not calibrate. Went back into MSFS and throttle was fine. Maybe there's something odd happening with the throttle in general?

Edited November 4, 2022 by MDFlier

[tweekz 2,165]

Quote

The MSFS environment modeling is excellent in very many regards- but it also provides some levels of mathematical instability that would be truly terrifying if you hit them in real life. We have adapted the 737 to manage these moments so that they are not disruptive to the flying experience and still yield a realistic result. This required a comprehensive re-study of the control logic used in the actual airplane, with adaptation for extremes. It has been a fantastic exercise and we think you will notice the results right away as improved stability in roll, pitch and now also in the thrust channel.

This answers one of my questions: Does PMDG critizise the MSFS environment as "overexaggerated" or do they just have to adapt.

Seems like both is the case. And it also sheds some light on the accuracy of the physics from an experts point of view, that produces high quality add ons for decades - so this should be of value.

It seems like the physics are very good, but there are edge cases (mathematical instability) where it does not behave correct. This pretty much confirms my theory, that the MSFS flight model / physics are quite good with lots of potential, but (stil)l need some tuning.

Edited November 4, 2022 by tweekz

[Twenty6 694]

21 hours ago, penta_a said:

with latest update under SU11 BETA, 737-800 is like riding a camel's back, up-down all the time.

Don't think that's limited to the 737s. I've noticed this more pronounced oscillation in the 310R, Kodiak and Hjet (which are all I've been flying.) Add turbulence when flying GA over mountains and this new beta SU11 becomes stupid.

[Alpine Scenery 1,609]

IMHO, the issue with the modeling is they left too much up to the developers to smooth out. In theory, this might be ok, but unless you have a clearly established API that doesn't change much, then it means they will keep having to patch it as Asobo keeps changing stuff. That is my understanding, most of the sub-$20 GA planes are just plugging in basic numbers though and using a lot of defaults I think, as a lot of them aren't flying that well. It seems like Asobo expected the devs to add smoothing algorithms to stuff.

Instead, I believe it's basically just a bunch of queryable data you can take in, opt to use some of their default functions to interpret the movement, or just make your own. The problem is, who has time to write something like that when you are just building a single aircraft, sure PMDG maybe, but most others do not.

Edited November 4, 2022 by Alpine Scenery

[lwt1971 7,693]

48 minutes ago, tweekz said:

It seems like the physics are very good, but there are edge cases (mathematical instability) where it does not behave correct. This pretty much confirms my theory, that the MSFS flight model / physics are quite good with lots of potential, but (stil)l need some tuning.

As is almost invariably the case usually, tuning has to be done in the individual aircraft's flight models, which is why the best sim aircrafts are those developed by expert and established devs who know the ins and outs of their birds. There is no one "global" flight model in a sim per se, there is just the aerodynamics/physics engine and framework the sim provides, and then it's all down to how the flight models *per* aircraft are implemented/tuned on top of this. Both work in tandem and the core MSFS aerodynamics/physics engine has improved a lot and gained more capabilities SU afer SU, but of course there's room for more growth and fixing. But at the end of the day, it's the individual aircraft's flight model that makes or breaks a bird.. i.e. a well implemented aircraft FM based on legacy table-lookup methods can fly better and more realistically than a poorly implemented aircraft FM based on the latest CFD and other tech.

 

[Alpine Scenery 1,609]

Right, there is no one specific flight model, but there is a flight model in the sense that you can fully develop a plane using only what Asobo provides just by plugging in the numbers, but to really give the plane any character you have to use a lot of modifiers (from what I have read).

I'm not a plane developer, though I may get back into that at some point, but from my somewhat limited understanding, most developers are just using the modifiers to kind of force a plane to meet the numbers it needs to match things like climb rate. However, it's not actually constraining to very specific data like drag co-efficients at X altitude at X pitch at X speed at X weight. Most developers aren't doing this last thing I guess, some are though. So it leaves a lot of planes flying kind of generic. It does however generate a stream of data for winds and vector whatever, but it's up to devs I guess to smooth that data as others have noted. 

It's just a system that is too heavily biased to "only experts" instead of ease-of-use. 
 

Edited November 4, 2022 by Alpine Scenery

[tweekz 2,165]

You can change a lot on the individual flight model, but there are some things that are just overdone. Like the intensity of wind changes. I regularly fly over the alps and I've never witnessed turbulences like there are in MSFS.

[Jeeeno 314]

PMDG apparently in the latest update tried to address the behaviour of the plane in response to the irrealistic atmosphere simulation in MSFS: however the AP now behaves oddly. 

[Alpine Scenery 1,609]

48 minutes ago, tweekz said:

You can change a lot on the individual flight model, but there are some things that are just overdone. Like the intensity of wind changes. I regularly fly over the alps and I've never witnessed turbulences like there are in MSFS.

Too much cold weather turbulence is probably the issue, though I'm surprised there isn't a lot of turbulence in the Alps.
There is a lot in the Western US, but it's much warmer.

I believe parts of Nevada and Utah are some of the most turbulent areas in the US (excluding storms), guessing partly because of the severe temperature change from day-night creates a huge differential between the ground and air temperature at certain times of day, which ramps up the gradient effect in the mountains.
 

Edited November 4, 2022 by Alpine Scenery

[tweekz 2,165]

15 minutes ago, Alpine Scenery said:

Too much cold weather turbulence is probably the issue, though I'm surprised there isn't a lot of turbulence in the Alps.

There certainly is a lot of turbulence, but it's close to the mountains and in smaller planes. With a 737 or A320 you rarely notice anything special other than the usual stuff.