Aerosoft's F-16, Eagle Dynamics aircraft, NGX, MD-11, 747-400, CS 777.  Pretty much anything with LARGE hi-bypass fans.  You'll rarely hear this in RL or on the sim in smaller aircraft like a CRJ or MD-80/90, unless youre sitting in the last seats of the AC in real life.

 

It does change very slightly in different conditions, but not to the point that you would notice without a full-blown engine test cell.

Aerosoft's F-16, Eagle Dynamics aircraft, NGX, MD-11, 747-400, CS 777.  Pretty much anything with LARGE hi-bypass fans.  You'll rarely hear this in RL or on the sim in smaller aircraft like a CRJ or MD-80/90, unless youre sitting in the last seats of the AC in real life.

 

It does change very slightly in different conditions, but not to the point that you would notice without a full-blown engine test cell

I find it to be easily noticeable in real life aircraft actually, but only if you're sitting in front of the aircraft. I've heard it in the 747, 777 and the 737NG. Not yet sure about Airbus, I've only been on an A319, wasn't sitting in front of the engines.

 

By the way, apart from this overshoot effect that is present during initial spool-up, I have also noticed the engines in general are much more responsive. In case of the 737NG that is only at higher RPMs by the way, at lower RPMs they are very irresponsive. And in the 747 and 777 the engines seem to be extremely responsive, even to very small thrust changes.

OK, as promised I have worked on mimicking the engine spool-up + loop gain (overshoot) from the real-world 737NG. I did it as best as I could.

Note what kind of erratic movements I needed to make with my thrust levers, to simulate this, while the engines should be spooling-up this way by default if I set and hold my thrust levers at 45% N1.

http://www.youtube.com/watch?v=3RdyfJSBRVM&feature=youtu.be

Thank you. In which add-on(s) exactly have you noticed this realistic overshoot effect? Or do you mean you have never noticed it yet, just like in the NGX?

 

It's an interesting post you made. I explains the overshoot effect and also proves my findings even more. It looks like it much be something that is present at all times, and doesn't magically arise when atmospheric conditions change.

 

I will also try to mimic the real-world engine spool-up in the NGX, I will record it and post it on YouTube, and you'll see what kind of a antics I have to make with my thrust levers to mimic it. 

Interesting, do you think the effect I described, and as shown by the red line on my graph, can be simulated in the NGX? Or are there limiations in the .air files that don't allow for this? Just curious.

 

Limitations of the air file, and FSX's modeling of Turbine engines.

 

There is no way to modify the air files, to achieve the spool up dynamics you have so accurately measured in the real plane.

 

There are so many limitations in FSX's modeling of a Turbine engine,  that this is really one of the more minor deviations from real world.

 

The fact that even a Real World 737 Captain did not notice the difference, would tend to imply that there are more important things to correct, than the N1 spool up characteristics,

 

In the Sim, I would just pretend that my engine needed some minor adjustment at its next service, and maybe even an ECM software update.

 

Could be worse  -- your batteries could be melting down and catching on fire !!

 

.

 

By the way, apart from this overshoot effect that is present during initial spool-up, I have also noticed the engines in general are much more responsive. In case of the 737NG that is only at higher RPMs by the way, at lower RPMs they are very irresponsive. And in the 747 and 777 the engines seem to be extremely responsive, even to very small thrust changes.

The overshoot during initial spool-up is not correct and is due to an incorrectly set-up table 1505.

In real life if engines surged like that during start-up it would be a hot start and could damage the engines severely.

 

There is no way to modify the air files, to achieve the spool up dynamics you have so accurately measured in the real plane.

 

Of course one can modify the air files! You just need to know what you're doing. You can set-up the tables to mirror your engine performance  within 1% across the entire performance spectrum. That's a lot of work but it can and has been done.

Your starting point is the engine limitations schedule. Also find out how long it takes during spool-up from ignition to idle. Set those numbers in table 1505 adjust fuel flow etc. There's a lot to do to be accurate. It takes time and perseverance to get it exact but it can be done.

 

 

The overshoot during initial spool-up is not correct and is due to an incorrectly set-up table 1505.
In real life if engines surged like that during start-up it would be a hot start and could damage the engines severely.

You're confusing me a little here.

 

1. You say the NGX has an overshoot simulated because of an incorrect table 1505. The NGX does not have an overshoot simulated, the real-world plane has it, and I want it to be simulated in the NGX, that's the whole point of the thread here.

2. I'm not referring to the start-up of the engines, but to the initial spool-up before TO/GA (to 45% N1). If you say the engines would get damaged because of an overshoot, then why do all 737NG's do this? It was easily noticeable on all of my 737 flights, AND it is noticeable on lots of YouTube videos as well, as you can see in in the video in my OP. 

Note that the overshoot is very slight. So let's say the pilot sets thrust to 45%, the engines would overshoot their RPM to around 47%, then go back. As far as I can understand from FSMP's posts, this is called the "loop gain".

The thing that I showed in my own video a few posts above this one, was simply a simulation of the real-world engine spool-up like shown in my graph and in the video from the OP. I matched it as closely as I could. If you look at my thrust inputs in that simulation video, you can see how erratic they are. Of course in reality one is not supposed to do that, but, it was the only way to simulate the real-world engine spool-up and looking at the erratic inputs I had to make, it shows much it's off in the NGX. The engines in the NGX should be spooling-up this way by simply setting and holding the thrust levers at 45%.

I thought this was all discussed soon after the release and PMDG recognised there were some differences and either said they wouldn´t fix it or  couldn´t fix it as it was a FSX limitation.

I do remember PMDG stating that the engine behavior is currently limited by FSX. The discussion at that time was about the slower overall spool-up when setting 45% N1 in comparison to the real engine (which is observable in the graphic of the initial post). The NGX is within certified limits, albeit barely so (posted by Mr. Randazzo). Nevertheless, this thread proved to be very educational.

I do remember PMDG stating that the engine behavior is currently limited by FSX. The discussion at that time was about the slower overall spool-up when setting 45% N1 in comparison to the real engine (which is observable in the graphic of the initial post). The NGX is within certified limits, albeit barely so (posted by Mr. Randazzo). Nevertheless, this thread proved to be very educational.

The thread from back then was also created by me. And because of my bad explaining skills of back then, people interpretated my post as "slower engine spool-up", while I actually meant the exact same as I do now.

 

What is surprising me though this time, is that there hasn't been a single response PMDG in this thread so far. I am a 100% sure Ryan or perhaps some other PMDG staff has read it. Perhaps that's positive sign and are re-evaluating the possibilities.

Arjen,

 

The overshoot isn't called loop gain. Adjusting the loop gain in a control system will adjust the overshoot. Note you can also adjust other things in the control loop like damping but loop gain was what was mentioned.

The thread from back then was also created by me. And because of my bad explaining skills of back then, people interpretated my post as "slower engine spool-up", while I actually meant the exact same as I do now.

 

Oh, I see ;-)

 

 

 

What is surprising me though this time, is that there hasn't been a single response PMDG in this thread so far. I am a 100% sure Ryan or perhaps some other PMDG staff has read it. Perhaps that's positive sign and are re-evaluating the possibilities.

 

I hope PMDG is very busy with the 777 beta ;-)

You know, I recall Mr. Randazzo talk about the future of flight simulation at one conference or other and stating that he was very much looking forward into the future and what's in store. My hope really is that at some point in time we will have a platform that enables developers to spend their time on incorporating subtleties such as this engine behavior and not on battling with a limited and fault-riddled software that FSX is.

You're confusing me a little here.

 

1. You say the NGX has an overshoot simulated because of an incorrect table 1505. The NGX does not have an overshoot simulated, the real-world plane has it, and I want it to be simulated in the NGX, that's the whole point of the thread here.

2. I'm not referring to the start-up of the engines, but to the initial spool-up before TO/GA (to 45% N1). If you say the engines would get damaged because of an overshoot, then why do all 737NG's do this? It was easily noticeable on all of my 737 flights, AND it is noticeable on lots of YouTube videos as well, as you can see in in the video in my OP. 

 

Note that the overshoot is very slight. So let's say the pilot sets thrust to 45%, the engines would overshoot their RPM to around 47%, then go back. As far as I can understand from FSMP's posts, this is called the "loop gain".

 

The thing that I showed in my own video a few posts above this one, was simply a simulation of the real-world engine spool-up like shown in my graph and in the video from the OP. I matched it as closely as I could. If you look at my thrust inputs in that simulation video, you can see how erratic they are. Of course in reality one is not supposed to do that, but, it was the only way to simulate the real-world engine spool-up and looking at the erratic inputs I had to make, it shows much it's off in the NGX. The engines in the NGX should be spooling-up this way by simply setting and holding the thrust levers at 45%.

 

 +1

 

It would help if those responding to something posted, actually read and understand what others have said before, before they launch into disputing what has been previously said.

 

Could cut down on a lot of disagreements !!

 

An NO,  no matter want you do with the air files,  you cannot change the basic code algorithms that FSX is using, when it uses the air files as parameters for its engine modeling calculations.

 

The bottom line, is that FSX's modeling of Turbine engines does not provide the flexibility to simulate this sort of detail.

If you want the N2 gauge to follow a profile, that the air file cannot provide, then the N2 gauge  would need to be  "programmed"  to "appear"  to be indicating the characteristic that are required.

 

Same goes for EGT,  to accurately "simulate" Hot Starts, and Compressor Stalls.

If you have a strong interest in the technical side of control system, in Flight Simulators,  there is a good article about the development of the Autopilot  in FlightGear.

 

While this is not the Engine Control System of a 737NGX,   it will give you some idea of the potential complexities  of such systems, and how complex they can be to model.

 

http://www.geoffair.com/fg/site/Docs/XMLAutopilot/             :p0504:

I am not completely sure but I do not believe that PMDG modeled a FADEC. This could be the variations you are noticing. 

 

Thanks,

Ron

 

Not sure about you guys but it is still the best plane in my stable. I still have a lot to learn about the systems due to time constraints but I love to fly her and a lot by hand on the numbers. I would have never noticed this if it had not been brought up and probably still won't.

Do you think the T7 will have the same problem? If so will you buy it?

Sometimes Perfection can be a matter of opinion.

 

Thanks,

Ron