Inaccurate Mixture Behavior on Turbocharged Piston Engines

Is this in Sim Update 6 or ongoing?

Looking at the release notes for SU6, I donā€™t see anything related to turbocharger logic. Iā€™m guessing that this bug will not be fixed in SU6.

1 Like

As a follow-up to the data I shared above, I successfully created a workaround solution based on Working Titleā€™s FADEC code for the CJ4. Hereā€™s an overview of what I did:

The data above indicates two primary discoveries:

  1. The fuel/air ratio in the core sim logic is directly proportional to the specific volume (1/density) of the ambient air.
  2. At full-rich mixture, the proportionality constant between fuel/air ratio and ambient specific volume remains constant at all altitudes in the MSFS turbo logic.

Now, for a real-world turbocharged engine, the fuel/air mixture should be related to the intake manifold density, rather than the ambient density. The turbocharger compresses air significantly above ambient density before it enters the engine.

I developed code to calculate the full-power intake manifold density (based on the maximum rated turbo boost pressure, ambient pressure, and ambient temperature). Then, I gathered data from the sim to model the relationship between mixture lever percentage and fuel/air ratio. With these two pieces of information, I could calculate a realistic fuel/air ratio for a userā€™s mixture lever setting.

(As a side note, when I first reported the data I gathered on the relationship between mixture lever percentage and fuel/air ratio in a previous post above, I thought there were 3 or 4 linear functions covering discrete percentage ranges. When I looked closer at the raw data, I found 5 linear functions: 0-20%, 20-60%, 60-75%, 75-90%, and 90-100%. I calculated the slope for these functions based on the raw data I collected from the sim. Using this data, I can mimic the same five-function behavior in my mixture controller.)

Finally, I created a WASM gauge that intercepts the userā€™s hardware mixture inputs, calculates the correct fuel/air ratio for those inputs, and sets the core sim mixture axis at a level that will yield the correct fuel/air ratio. For this portion of the project, I repurposed Working Titleā€™s FADEC controller for the throttle axis to handle the mixture axis instead.

The end result is a WASM module that can be added to the panel of a turbo aircraft to yield more accurate mixture behavior during climb. The mixture lever in the virtual cockpit will show the core sim mixture setting, but your hardware lever will behave the way the real-world airplane should. The code defines a local variable that could be used to set the cockpit mixture lever to mimic the hardware axis, but I have not figured out a way to assign the cockpit mixture lever to a different variable.

I created specific files and instructions for the three main turbo aircraft I am aware of in MSFS right now. I also made the raw source code available. If this code can provide any ideas to Asobo on how to modify the core logic, they are welcome to use it any way they see fit. Anyway, hereā€™s a link to the code I developed:

After a couple of months using and testing this code, I found one other issue that should be addressed in the MSFS turbo logic: turbine inlet temperature (TIT) appears to vary based on ambient atmospheric temperature, rather than intake manifold temperature.

As a turbocharged aircraft climbs, the air in the intake manifold is being compressed more and more compared to the ambient air. Compression results in heating (Boyleā€™s Law). In the troposphere, a good rule of thumb that if you compress air from high altitudes to sea-level, it will be hotter than if you compress air from low altitudes to sea-level. Therefore, as a turbocharged aircraft climbs at constant manifold pressure, the temperature in the intake manifold will generally increase. In turn, the temperature of the exhaust gases after combustion should be similarly higher.

Using my mixture controller code, I observe the opposite. For a constant mixture setting, the turbine inlet temperature decreases with increasing altitude. This would be correct for a naturally-aspirated engine, but not for a turbocharged engine.

12 Likes

Boy am I glad I tracked this topic! I had been avoiding turbocharged aircraft addons due to this very issue but I might have to give them a try with this! Love that you came up with an actual solution to the problem, proving that it can in fact be fixed. If thereā€™s a way to fix the issue you mentioned about the virtual cockpit mixture level positioning, then it would be absolutely perfect.

I hope more people (and Asobo) see this. Voted for sure.

3 Likes

Just tested this out with the Turbo Bonanza mod and wow, it actually works, and it works fantastic. Thank you so much!

1 Like

Youā€™re very welcome! Iā€™ve been using it for a couple of months with the Turbo Arrow and the Seneca V. Itā€™s so much nicer to use than the default mixture logic.

Iā€™d absolutely love for a permanent fix to be developed. This has been one of my pet peeves for generations of MSFS, as itā€™s never been right. :crossed_fingers:

Hi, I just tried for a little bit to get the mod working for the Seneca V but I was having a problem where the aircraft would disappear from the sim and be marked as ā€œNot installedā€ in the content manager after following the directions in your instruction manual. I actually solved the problem on my own and it works for me now, but others may have an issue.

The issue is with the layout.json file. My layout.json file was way different than the one you provided, so instead of replacing mine with yours, I edited the original file to add the .wasm file, by copying that entry from your layout.json and pasting it into mine. I then made sure the size for the panel.cfg was accurate since I had modified it, and it now works.

I donā€™t know if this is just a ā€œmeā€ issue or not but hopefully this fix helps.

Iā€™ve got the Arrow bundle, will the install be the same for this package?

Congrats on the work done, and hopefully the devs will fix this in a later update.

I suspect so, but I donā€™t own the bundle so I am unable to test.

Incredible work! I just bought the Seneca V and have been enjoying it apart from this really obnoxious issue. Canā€™t wait to try your mod, although Asobo really should just fix this properly already.

Thank you for the feedback, @Synoxys787. To everyone: Iā€™ve created a separate thread for discussing the mod so that this thread can stay dedicated to the bug report and sharing data related to it. Here is a link to the mod thread.

Iā€™ve included a suggestion in the initial post for anyone else who runs into the same problem with the layout.json that Synoxys ran into. If anyone is interested in trying the mod, I encourage you to use the new thread to continue the discussion!

3 Likes

Iā€™ve tried this on the Turbo Arrow 4, and Iā€™m a little bit confused.

So, the mod seems to work as described, as I climb the lever moves to lean the engine.
My hardware mixture lever does nothing. I canā€™t lean at cruise, the EGT stays constant depending on throttle setting and rpm. Is this correct and expected behaviour?

Also, is it correct that I have to keep opening the throttle to keep my manifold pressure constant as I climb? Iā€™m a car guy and Iā€™d expect the wastegate to keep closing to keep the manifold pressure constant in a turbo setup. Planes might work differently.

Hmmā€¦you should be able to lean the mixture with your hardware lever at any time. The cockpit lever will move to keep the core-sim mixture set at a level that corresponds to the mixture percentage on your hardware axis, but if you move your hardware lever, the code will react by setting the core sim mixture to match it. Any chance that you have automixture enabled in the sim settings?

The real-life turbo arrow has a fixed wastegate. It requires adjusting the throttle throughout the climb because the wastegate doesnā€™t move. Basically, it was a way for Piper to offer more power at a low price compared to the competition. Most turbocharged planes have a variable wastegate that keeps manifold pressure constant, like you described. The turbo arrow is quirky in that respect.

Iā€™d be happy to continue discussing on the mod thread (linked below) so we can avoid cluttering this bug report thread with mod details.

2 Likes

Iā€™m glad I stumbled across this topic. Iā€™m grabbing the files now. This has irked me ever since I got the Turbo Arrowā€™s and Seneca. Asobo really needs to look into this as is a rather large oversight of a core system on how the engines work.

2 Likes

This topic has definitely been unnoticed by so many. This is a very important issue.

2 Likes

Hello WxMarc,

We investigated in this, but the issue raised is based on a feature that doesnā€™t yet exist on the SIM. Right now, we donā€™t have default Turbocharged aircraft and changing the setting on the Cfg file canā€™t result in activating it.

The Turbocharge line in the Cfg file is an heritage of FSX, this is why this line is displayed.

Instead of a bug, this thread should be turned into a request on the wishlist section to add a Turbocharged feature for specific planes.

Iā€™m letting you read the information above before doing any change on the thread.

6 Likes

Thank you for investigating and providing feedback. Iā€™m a bit confused, though. The turbocharged line in the engines.cfg file definitely has an effect. As such, Iā€™m not sure what you mean by saying that changing the setting ā€œcanā€™t result in activating it.ā€ If turbocharged = 0 in the cfg file, the engine will lose power with altitude like a naturally-aspirated engine should, and no adjustment to the mixture will restore the engine to full sea-level power. On the other hand, if turbocharged = 1, the engine is capable of producing full sea-level power as long as the mixture is adjusted. This is not correct behavior (the mixture should not require adjustment to maintain power in a turbocharged engine), but it is different than the default behavior when turbocharged = 0. To me, this seems like a bug rather than a requested new feature.

If the team would prefer to scrap the old FSX turbocharger logic and start from scratch, I would support making this a wishlist item, but I just want to clarify that the existing logic does have an effect, and the data Iā€™ve reported in this thread documents the bug in the existing logic.

5 Likes

Should we take this statement to mean that turbocharged aircraft are not supported by MSFS? That is, 3rd party devs should write custom code for the engine logic?

Anyway, isnā€™t the PC-6 coming to MSFS as part of the GOTY edition? As this has a turbocharged engine, does that mean that turbochargers are supported from SU7/GOTY onwards? :crossed_fingers:

1 Like